Download MINER VA ANESTESIOLOGIC A . V OL. 82 . N o. 11 . PA

CURE PALLIATIVE E DI
SUPPORTO IN ONCOLOGIA
INDICE
Volume di 336 pagine con 16 figure in nero e a colori e 100 tabelle
ISBN 978-88-7711-864-6
INTRODUZIONE
Cure palliative e di supporto
Bioetica
Comunicazione
Fattori prognostici
Qualità di vita
Spiritualità
DOLORE
Neurofisiologia
Caratteristiche del dolore
Valutazione del dolore
Trattamento farmacologico
Oppioidi
Aspetti specifici del
trattamento con oppioidi
Procedure invasive
SINTOMI
Sintomi gastrointestinali
Sintomi respiratori
Sintomi psichiatrici
Sindromi sistemiche e
situazioni specifiche
Emergenze
Cure di fine vita
€ 45,00
Le cure palliative sono state progressivamente adottate negli ospedali e nelle
istituzioni accademiche di tutto il mondo,
ma tale processo è stato lento e difficile.
Le risorse umane ed economiche sono
limitate rispetto ad altre discipline che si
sono sviluppate nella corrente principale
della medicina clinica e accademica. In
questo volume l’autore introduce il lettore
lungo un percorso in cui vengono trattati
tutti gli aspetti delle cure palliative, dalla
valutazione al trattamento del paziente
oncologico. Questo testo sarà pertanto
fonte di ispirazione per tutti i professionisti del settore nella loro pratica quotidiana e permetterà di migliorare le cure dei
pazienti sofferenti.
3HUOHYRVWUHRUGLQD]LRQL
COME ORDINARE LIBRI
x ULWDJOLDWHHVSHGLWHODFHGRODD/LEUHULD0LQHUYD(GL]LRQL0LQHUYD0HGLFD&RUVR%UDPDQWH7RULQR
x SHUXQFRQWDWWRGLUHWWR7HO)D[
x FROOHJDWHYLDOVLWR,QWHUQHWZZZPLQHUYDPHGLFDLW,QGLUL]]RHPDLOERRNGHSW#PLQHUYDPHGLFDLW
ACQUISTA ON-LINE
*OL(QWL3XEEOLFLVRQRSUHJDWLGLHIIHWWXDUH,HRUGLQD]LRQLXWLOL]]DQGRLOEXRQRG·RUGLQHGHOO·(QWHVWHVVR
/·DFTXLUHQWHKDGLULWWRGLUHVWLWXLUHODPHUFHHQWURJLRUQLGDOODGDWDGLULFHYLPHQWRGHOSDFFRSRVWDOH/DUHVWLWX]LRQHVDUjDFDULFRGHOO·DFTXLUHQWHVWHVVRHODPHUFHQRQGRYUjHVVHUHLQ
DOFXQPRGRGHWHULRUDWD'HFUHWROHJLVODWLYR1,GDWLWUDVPHVVLFLHFKH/HLDLVHQVLGHOOD/HJJHFLDXWRUL]]DDWUDWWDUHVDUDQQRXWLOL]]DWLDLVROLILQLFRPPHUFLDOLH
SURPR]LRQDOLGHOODQRVWUDDWWLYLWj$JJLRUQDPHQWLRFDQFHOOD]LRQLGHLGDWLGRYUDQQRHVVHUHULFKLHVWLD(GL]LRQL0LQHUYD0HGLFD&RUVR%UDPDQWH7RULQR
Accedi al sito www.minervamedica.it, clicca su LIBRI e procedi con l’ordine on-line.
Metodi di pagamento: carta di credito e, solo per spedizioni in Italia, contrassegno.
MODALITÀ
E SPESE DI SPEDIZIONE
6HQG1RFRSLHVRI
Spedizione tramite corriere espresso.
Le spese di spedizione saranno calcolate in base al luogo di consegna, peso dei volumi e, solo per le spedizioni in
Italia, metodo di pagamento.
Contrassegno al ricevimento del volume
Pagamento anticipato a mezzo assegno o su c /c postale n. 00279109
TEMPI DI SPEDIZIONE
Carta diincredito
Spedizioni
Italia: 2-4 giorni lavorativi.
Spedizioni fuori Italia: 3-7 giorni lavorativi a seconda della destinazione.
Diners Club
International
Carta n.
®
scadenza
ALTRE
MODALITÀ
ACQUISTO
Cognome
e nome DI
............................................................................................................................................................................................................................................................
Per altre modalità di acquisto e pagamento contattaci:
Indirizzo
....................................................................................................................................................................................................................................................................................
- email:
[email protected]
- tel. 011 678282 int. 328.
InCAP
caso........................
di istituti,
enti
o
ditte
che
effettuano
le ordinazioni con buoni d’ordine contattaci:
Città ............................................................................................................................................................................................................
Prov. ...............................
- email: [email protected] - tel. 011 678282 int. 328.
Partita IVA o Codice Fiscale
......................................................................................................................................................................................................................................
Ai sensi del D.Lgs 196/2003 Minerva Medica garantisce la riservatezza dei dati e la possibilità di modificarli o cancellarli con richiesta scritta al
Data ........................
Email .............................................................................................
...............................................................................................................................
responsabile
dei dati personali.
Per ulteriori informazioni rivolgersi a: EdizioniFirma
Minerva
Medica - Privacy policy - Corso Bramante 83-85 10126 Torino
(',=,21,0,1(59$0(',&$
MINERVA ANESTESIOLOGIC A . VOL. 82 . No. 11 . PAGES 1129 - 1240 . NOVEMBER 2016
S. MERCADANTE
V O L U M E 8 2 · N o. 1 1 · N O V E M B E R 2 0 1 6
MINERVA ANESTESIOLOGICA
ITALIAN JOURNAL OF ANESTHESIOLOGY AND ANALGESIA
MONTHLY JOURNAL FOUNDED IN 1935 BY A. M. DOGLIOTTI
OFFICIAL JOURNAL OF ITALIAN SOCIETY OF ANESTHESIOLOGY, ANALGESIA,
RESUSCITATION AND INTENSIVE CARE (S.I.A.A.R.T.I.)
EDITORIAL BOARD
EDITOR IN CHIEF
F. Cavaliere
Roma, Italy
CRITICAL CARE
ANESTHESIA
General Critical Care
Associate Editor
G. M. Albaiceta (Oviedo, Spain)
Section Editor
G. Biancofiore (Pisa, Italy)
E. De Robertis (Napoli, Italy)
General Anesthesia
Associate Editor
M. Rossi (Roma, Italy)
Section Editor
E. Cohen (New York, USA)
P. Di Marco (Roma, Italy)
J. T. Knape (Utrecht, The Netherlands)
O. Langeron (Paris, France)
P. M. Spieth (Dresden, Germany)
Circulation Critical Care
Section Editor
S. Scolletta (Siena, Italy)
E. Bignami (Milano, Italy),
Pediatric Anesthesia
Section Editor
M. Piastra (Roma, Italy)
Respiration Critical Care
Section Editor
S. Grasso (Bari, Italy)
P. Terragni (Sassari, Italy),
Obstetric Anesthesia
Section Editor
E. Calderini (Milano, Italy)
Neurocritical Care
Section Editor
F. S. Taccone (Brussels, Belgium)
Regional Anesthesia
Section Editor
A. Apan (Giresun, Turkey)
M. Carassiti (Roma, Italy)
ETHICS
PAIN
Section Editor
A. Giannini (Milano, Italy)
Section Editor
M. Allegri (Parma, Italy)
F. Coluzzi (Roma, Italy)
MEDICAL STATISTIC
Section Editor
B. M. Cesana (Brescia, Italy)
MANAGING EDITOR
A. Oliaro
Torino, Italy
Vol. 82 - No. 11
MINERVA ANESTESIOLOGICA
III
GENERAL INFORMATION
This journal is PEER REVIEWED and is quoted in:
Current Contents, SciSearch, PubMed/MEDLINE, EMBASE, Scopus
Impact factor: 2.036
Published by
Edizioni Minerva Medica - Corso Bramante 83-85 - I-10126 Turin (Italy)
Tel. +39 (011) 678282 - Fax +39 (011) 674502
Website: www.minervamedica.it
Editorial office: [email protected] - Subscriptions: [email protected]
Advertising: [email protected]
Presidential Office: Antonio Corcione, UOC Anestesia e TIPO, Azienda Ospedaliera dei Colli, V. Monaldi, Via Leonardo
Bianchi 1, 80131 Naples, Italy - Tel. 081 7065214. E-mail: [email protected]
Technical Secretariat: S.I.A.A.R.T.I., Viale dell’Università 11, 00185 Rome, Italy - Tel. 06 4452816
E-mail: [email protected] - Website: www.siaarti.it
Administrative Secretariat S.I.A.A.R.T.I. - Edizioni Minerva Medica - Corso Bramante 83-85 - 10126 Turin, Italy - Tel. +39
(011) 678282 - Fax +39 (011) 674502 - E-mail: [email protected]
ANNUAL SUBSCRIPTIONS
ITALY
EUROPEAN UNION
OUTSIDE THE
EUROPEAN UNION
Individual: Print € 130,00
Print+Online
€ 135,00
Online
€ 140,00
Individual: Print € 225,00
Print+Online
€ 235,00
Online
€ 245,00
Individual: Print € 250,00
Print+Online
€ 260,00
Online
€ 270,00
Institutional:
Print € 180,00
Institutional:
Print € 300,00
Institutional:
Print € 335,00
Online
Small Medium Large € 350,00
€ 407,00
€ 429,00
Online
Small Medium Large € 351,00
€ 409,00
€ 429,00
Online
Small Medium Large € 383,00
€ 446,00
€ 462,00
Print+Online
Small Medium Large € 367,00
€ 420,00
€ 446,00
Print+Online
Small Medium Large € 378,00
€ 430,00
€ 456,00
Print+Online
Small Medium Large € 404,00
€ 467,00
€ 483,00
Subscribers: Payment to be made in Italy: a) by check; b) by bank transfer to: Edizioni Minerva Medica, INTESA SANPAOLO Branch
no. 18 Torino. IBAN: IT45 K030 6909 2191 0000 0002 917 c) through postal account no. 00279109 in the name of Edizioni Minerva
Medica, Corso Bramante 83-85, 10126 Torino; d) by credit card Diners Club International, Master Card, VISA, American Express.
Foreign countries: a) by check; b) by bank transfer to: Edizioni Minerva Medica, INTESA SANPAOLO Branch no. 18 Torino. IBAN:
IT45 K030 6909 2191 0000 0002 917; BIC: BCITITMM c) by credit card Diners Club International, Master Card, VISA, American
Express. Members: for payment please contact the Administrative Secretariat SIAARTI.
Notification of changes to mailing addresses, e-mail addresses or any other subscription information must be received in good time.
Notification can be made by sending the new and old information by mail, fax or e-mail or directly through the website www.minervamedica.it at the section “Your subscriptions - Contact subscriptions department”.
Complaints regarding missing issues must be made within six months of the issue’s publication date.
Prices for back issues and years are available upon request.
© Edizioni Minerva Medica - Torino 2016. All rights reserved. No part of this publication may be reproduced, stored or transmitted in
any form or by any means, without the prior permission of the copyright owner.
Monthly publication. Authorized by Turin Court no. 281 of 24-1-1953.
This Journal is associated with
IV
This Journal complies with the Code of Self-Discipline of Medical/Scientific Publishers associated with FARMAMEDIA and may accept advertising
MINERVA ANESTESIOLOGICA
November 2016
IN­S TRUC­T IONS TO AU­T HORS
Minerva Anestesiologica is the journal of the Italian National
Society of Anaesthesia, Analgesia, Resuscitation, and Intensive
Care (SIAARTI - http://www.siaarti.it). Minerva Anestesiologica
publishes scientific papers on Anesthesiology, Intensive care,
Analgesia, Perioperative Medicine and related fields.
Manuscripts are expected to comply with the instructions
to authors which conform to the Uniform Requirements for
Manuscripts Submitted to Biomedical Editors by the International
Committee of Medical Journal Editors (http://www.icmje.org).
Articles not conforming to international standards will not be
considered for acceptance.
Submission of manuscripts
Papers should be submitted directly to the online Editorial Office
at the Edizioni Minerva Medica website:
http://www.minervamedicaonlinesubmission.it
Duplicate or multiple publication
Submission of the manuscript means that the paper is original
and has not yet been totally or partially published, is not currently
under evaluation elsewhere, and, if accepted, will not be published
elsewhere either wholly or in part.
Splitting the data concerning one study in more than one publication could be acceptable if authors justify the choice with
good reasons both in the cover letter and in the manuscript.
Authors should state the new scientific contribution of their
manuscript as compared to any previously published article
derived from the same study. Relevant previously published
articles should be included in the cover letter of the currently
submitted article.
Permissions to reproduce previously published material
Material (such as illustrations) taken from other publications must
be accompanied by the publisher’s permission.
Copyright
The Authors agree to transfer the ownership of copyright to
Minerva Anestesiologica in the event the manuscript is published.
Ethics committee approval
All articles dealing with original human or animal data must
include a statement on ethics approval at the beginning of the
methods section, clearly indicating that the study has been
approved by the ethics committee. This paragraph must contain
the following information: the identification details of the ethics
committee; the protocol number that was attributed by the ethics committee and the date of approval by the ethics committee.
The journal adheres to the principles set forth in the Helsinki
Declaration (http://www.wma.net/en/30publications/10policies/
b3/index.html) and states that all reported research concerning
human beings should be conducted in accordance with such
principles. The journal also adheres to the International Guiding
Principles for Biomedical Research Involving Animals (http://
www.cioms.ch/index.php/12-newsflash/227-cioms-and-iclasrelease-the-new-international-guiding-principles-for-biomedical-research-involving-animals) recommended by the WHO and
requires that all research on animals be conducted in accordance
with these principles.
Patient consent
Authors should include at the beginning of the methods section
of their manuscript a statement clearly indicating that patients
have given their informed consent for participation in the research
study.
Every precaution must be taken to protect the privacy of patients.
Conflicts of interest
Authors must disclose possible conflicts of interest including
financial agreements or consultant relationships with organizations involved in the research. All conflicts of interest must be
declared both in the authors’ statement form and in the manuscript
file. If there is no conflict of interest, this should also be explicitly
stated as none declared. All sources of funding should be acknowledged in the manuscript.
Authorship and contributorship
All persons and organizations that have participated to the study
must be listed in the byline of the article (authors) or in the notes
(contributors). The manuscript should be approved by all coauthors, if any, as well as, tacitly or explicitly, by the responsible authorities of the institution where the work was carried out.
Authors and contributors must meet the criteria for authorship
and contributorship established by the Uniform Requirements for
Manuscripts Submitted to Biomedical Editors by the International
Committee of Medical Journal Editors (http://www.icmje.org).
Authors’ statement
Papers must be accompanied by the authors’ statement (http://
www.minervamedica.it/en/journals/minerva-anestesiologica/
index.php) relative to copyright, originality, authorship, ethics
and conflicts of interest, signed by all authors.
Disclaimer
The Publisher, Editors, and Editorial Board cannot be held responsible for the opinions and contents of publications contained in
this journal.
The authors implicitly agree to their paper being peer-reviewed.
All manuscripts will be reviewed by Editorial Board members
who reserve the right to reject the manuscript without entering
the review process in the case that the topic, the format or ethical
aspects are inappropriate. In addition, the Editorial Board admits
for submission manuscripts previously submitted for publication
to another journal but rejected with reviewers’ comments. In such
cases, authors are requested to submit the scientific comments of
the other journal and their point-by-point reply to the reviewers’
comments in the manuscript file. This will help to speed up the
peer review process.
Once accepted, all manuscripts are subjected to copy editing. If
modifications to the manuscript are requested, the corresponding
author should send to the online Editorial Office the revised manuscript under two separate files, one file containing the revised
clean version and another containing both a letter with pointby-point responses to the reviewers’ comments and the revised
version with corrections highlighted.
Correction of proofs should be limited to typographical errors.
Substantial changes in content (changes of title and authorship,
new results and corrected values, changes in figures and tables)
are subject to editorial review. Changes that do not conform to
the journal’s style are not accepted. Corrected proofs must be
sent back within 3 working days to the online Editorial Office
of Minerva Anestesiologica. In case of delay, the editorial staff
of the journal may correct the proofs on the basis of the original
manuscript.
Publication of manuscripts is free of charge. Colour figures and
excessive alterations to proofs will be charged to the authors.
Authors will receive instructions on how to order reprints and a
copy of the manuscript in PDF.
For further information about publication terms please contact the
Editorial Office of Minerva Anestesiologica, Edizioni Minerva
Medica, Corso Bramante 83-85, 10126 Torino, Italy - Phone +39011-678282 - Fax +39-011-674502
E-mail: [email protected].
ARTICLE TYPES
Original papers are reports of original researches on the topics
of interest to the journal and do not exceed 3000 words, 6 figures or tables, and 40 references. Papers reporting randomized
controlled trials should be in accordance with the CONSORT
statement (http://www.consort-statement.org).
Review articles are comprehensive papers that outline current
knowledge, usually submitted after prior consultation with the
Editors and do not exceed 4000 words, 6 figures or tables, 70
references. Reviews should be in accordance with the PRISMA
statement (http://www.prisma-statement.org).
Experts’ opinions are papers reviewing the literature, usually
invited by the Editor in Chief and do not exceed 3000 and 2000
words with 50 and 40 references respectively.
Editorials are brief comments, commissioned by the Editor-inChief and express author’s opinion about published papers or
topical issues and do not exceed 1000 words and 20 references.
Letters to the Editor (Correspondence) are brief comments on
articles published in the journal or with other topics of interest
and do not exceed 500 words, 5 references and 1 figure or table.
PREPARATION OF MANUSCRIPTS
Text file
The formats accepted are Word (.DOC) and RFT. The text file
must contain title, authors’ details, abstract, key words, text, references, notes and titles of tables and figures. Tables and figures
should be submitted as separate files. The file should not contain
active hyperlinks.
Title and authors’ details
Short title, with no abbreviations. First name in full, middle
name’s initial, surname of the authors. Collective name, if any,
as last author. Corresponding author marked with an asterisk.
Affiliation (section, department and institution) of each author.
Name, address, e-mail of the corresponding author.
Abstract and key words
Articles should include an abstract not exceeding 250 words. For
original papers and technical notes, the abstract should be structured as follows: background (status of knowledge and aim of the
study), methods (experimental design, patients and interventions),
results (what was found), conclusion (meaning of the study). Key
words should refer to the terms from Medical Subject Headings
(MeSH) of the Index Medicus. No abstracts are required for editorials or letters to the Editor.
Text
Units of measurement, symbols and abbreviations must conform
to international standards. Measurements of length, height, weight
and volume should be given in metric units (meter, kilogram, liter)
or their decimal multiples. Temperatures must be expressed in
degrees Celsius. Blood pressure must be expressed in millimeters
of mercury. All hematological and clinicochemical measurements
should be expressed in metric units using the International System
of Units (IS). The use of unusual symbols or abbreviations is
strongly discour-aged. The first time an abbreviation appears in
the text, it should be preceded by the words for which it stands.
For original papers the text must be subdivided into the following
sections: introduction, materials and methods, results, discussion,
conclusions. The introduction should describe the present state of
knowledge on the subject and clearly state the aim of the research.
The Materials and methods section should provide enough information to allow other researchers to reproduce results and should
include a statistics section. Detailed statistical methodology must
be described in order to enable the reader to evaluate and verify
the results. Please report power analysis, randomized procedure,
tests of normality, descriptive parameters, statistical tests and the
probability level chosen for statistical significance. Drugs should
be identified by their international non-proprietary name. If special equipment is used, then the manufacturer’s details (including
town and country) should be given in parentheses. The Results
section should describe the outcome of the study. Data presented
in tables and graphs should not duplicate those presented in the
text. Very large tables should be avoided. The Discussion should
include the interpretation of the results obtained, an analysis of
the limits of the study, and references to work by other authors.
Conclusions should highlight the significance of the results for
clinical practice and experimental research.
References
The list of references should only include works that are cited in
the text and that have been published or accepted for publication
(a copy of the acceptation letter can be requested) in journals
cited in peer-reviewed Index Medicus journals. Websites can also
be cited providing the date of access. Bibliographical entries in
the text should be quoted using superscripted Arabic numerals.
References must be set out in the standard format approved by
the International Committee of Medical Journal Editors (http://
www.icmje.org).
Footnotes and endnotes must not be used in the preparation of
references.
Journals
Each entry must specify the author’s surname and initials (list all
authors when there are six or fewer; when there are seven or more,
list only the first six and then “et al.”), the article’s original title,
the name of the Journal (according to the abbreviations used by
Index Medicus), the year of publication, the volume number and
the number of the first and last pages. When citing references,
please follow the rules for international standard punctuation
carefully.
Examples
– Standard article
Sutherland DE, Simmons RL, Howard RJ. Intracapsular technique
of transplant nephrectomy. Surg Gynecol Obstet 1978;146:951-2.
– Organization as author
International Committee of Medical Journal Editors. Uniform
requirements for manuscripts submitted to biomedical journals.
Ann Int Med 1988;108:258-65.
– Issue with supplement
Payne DK, Sullivan MD, Massie MJ. Women’s psychological
reactions to breast cancer. Semin Oncol 1996;23(1 Suppl 2):
89-97.
Books and monographs
For occasional publications the names of the Authors, the title,
edition, place, publisher and year of publication must be given.
Examples
– Books by one or more Authors
Rossi G. Manual of Otorhinolaryngology. Turin: Edizioni Minerva
Medica; 1987.
– Chapter from book
De Meester TR. Gastroesophageal reflux disease. In: Moody FG,
Carey LC, Scott Jones R, Ketly KA, Nahrwold DL, Skinner DB,
editors. Surgical treatment of digestive diseases. Chicago: Year
Book Medical Publishers; 1986. p. 132-58.
– Congress proceedings
Kimura J, Shibasaki H, editors. Recent advances in clinical neurophysiology. Proceedings of the 10th International Congress of
EMG and Clinical Neurophysiology; 1995 Oct 15-19; Kyoto,
Japan. Amsterdam: Elsevier; 1996.
Notes
Authors’ contribution statement; list of the members of the collective name (author’s name in full, middle name’s initial in capital
letters and surname, with relevant affiliation); contributors’ names;
mention of any funding, research contracts; conflicts of interest;
dates of any congress where the paper has already been presented;
acknowledgements.
Titles of tables and figures
Titles of tables and legends of figures should be included both in
the text file and in the file of tables and figures.
File of tables
Each table should be submitted as a separate file. Formats accepted
are Word (.DOC) and RTF. Each table should be created with
the Table menu of the word processing software of the operating
system employed, by selecting the number of rows and columns
needed. Tabulations are not allowed. Each table must be typed correctly and prepared graphically in keeping with the page layout of
the journal, numbered in Roman numerals and accompanied by the
relevant title. Notes should be inserted at the foot of the table and
not in the title. Tables should be referenced in the text sequentially.
File of figures
Each figure should be submitted as a separate file. Formats
accepted: JPEG set at 300 dpi resolution preferred; other formats
accepted are TIFF and PDF (high quality). Figures should be numbered in Arabic numerals and accompanied by the relevant legend.
Figures should be referenced in the text sequentially.
Histological photographs should always be accompanied by the
magnification ratio and the staining method.
If figures are in color, it should always be specified whether color
or black and white reproduction is required.
MINERVA ANESTESIOLOGICA
ITALIAN JOURNAL OF ANESTHESIOLOGY AND ANALGESIA
MONTHLY JOURNAL FOUNDED IN 1935 BY A. M. DOGLIOTTI
OFFICIAL JOURNAL OF ITALIAN SOCIETY OF ANESTHESIOLOGY, ANALGESIA,
RESUSCITATION AND INTENSIVE CARE (S.I.A.A.R.T.I.)
Vol. 82
November 2016
No. 11
OFFICIAL JOURNAL OF ITALIAN SOCIETY OF ANESTHESIOLOGY, ANALGESIA,
RESUSCITATION AND INTENSIVE CARE (SIAARTI)
CONTENTS
1129
1149
EDITORIALS
Propofol-remifentanil and spontaneous breathing, a
magnificent pair
Sbaraglia F., Sammartino M.
Diaphragmatic ultrasonography as an adjunct predictor tool of weaning success in patients with difficult
and prolonged weaning
Flevari A., Lignos M., Konstantonis D., Armaganidis A.
1158
1132
Ultrasonography during weaning: a support to a comprehensive skilful strategy
De Robertis E., Romano G. M., Piazza O.
Non-invasive hemodynamic optimization in major
abdominal surgery: a feasibility study
Broch O., Carstens A., Gruenewald M., Nischelsky E.,
Vellmer L., Bein B., Aselmann H., Steinfath M., Renner J.
1170
1135
Perioperative hemodynamic therapy: goal-directed or
meta-directed?
Propofol versus midazolam for premedication: a placebo‑controlled, randomized double‑blinded study
Uhlig C., Spieth P. M.
Elvir Lazo O. L., White P. F., Tang J., Yumul R., Cao X.,
Yumul F., Hausman J., Hernandez Conte A., Anand K. K.,
Hemaya E. G., Zhang X., Wender R. H.
1138
1180
ORIGINAL ARTICLES
Propofol-remifentanil anesthesia for upper airway
endoscopy in spontaneous breathing patients: the
ENDOTANIL Randomized Trial
Besch G., Chopard-Guillemin A., Monnet E., Causeret A.,
Jurine A., Baudry G., Lasry B., Tavernier L., Samain E.,
Pili-Floury S.
Vol. 82 - No. 11
The optimal time between clinical brain death diagnosis and confirmation using CT angiography: a retrospective study
Kerhuel L., Srairi M., Georget G., Bonneville F., Mrozek S.,
Mayeur N., Lonjaret L., Sacrista S., Hermant N., Marhar
F., Gaussiat F., Abaziou T., Osinski D., Le Gaillard B.,
Menut R., Larcher C., Fourcade O., Geeraerts T.
MINERVA ANESTESIOLOGICA
I
CONTENTS
1189
Serum S100β as a prognostic marker in patients with
non-traumatic intracranial hemorrhage
Junttila E. K., Koskenkari J., Ohtonen P. P., Karttunen A.,
Ala-Kokko T. I.
1199
REVIEWS
Perioperative hemodynamic goal-directed therapy
and mortality: a systematic review and meta-analysis
with meta-regression
Giglio M., Manca F., Dalfino L., Brienza N.
1214
Targeting blood products transfusion in trauma: what
is the role of thromboelastography?
Figueiredo S., Tantot A., Duranteau J.
1235
LETTERS TO THE EDITOR
High-sensitivity troponin and extubation failure after
successful spontaneous breathing trial
Mottard N., Renaudin P., Wallet F., Thiollière F.,
Bohe J., Friggeri A.
1236
Posterior reversible encephalopathy syndrome in
acute pancreatitis
Compagnone C., Bellantonio D., Pavan F., Tagliaferri F.,
Barbagallo M., Fanelli G.
1238
Successful treatment of life-threatening hemorrhaging due to amniotic fluid embolism
Aurini L., Rainaldi M. P., White P. F., Borghi B.
1230
JOURNAL CLUB CRITIQUE
Failure of statins in ARDS: the quest for the Holy Grail
continues
Grimaldi D., Durand A., Gleeson J., Taccone F. S.
1240
TOP 50 MINERVA ANESTESIOLOGICA
REVIEWERS
V O L U M E 8 2 · N o. 1 1 · N O V E M B E R 2 0 1 6
About the cover: The cover shows the Serge’s sphygmo-oscillograph produced by Viola Gaetano company, Milan, around 1930. From the SIAARTI collection of medical devices at Viale
dell’Università, 11; Rome, Italy.
II
MINERVA ANESTESIOLOGICA
November 2016
© 2016 EDIZIONI MINERVA MEDICA
Online version at http://www.minervamedica.it
Minerva Anestesiologica 2016 November;82(11):1129-31
EDITORIAL
Propofol-remifentanil and spontaneous
breathing, a magnificent pair
Fabio SBARAGLIA*, Maria SAMMARTINO
Institute of Anesthesia and Intensive Care, Sacro Cuore Catholic University, Rome, Italy
*Corresponding author: Fabio Sbaraglia, Institute of Anesthesia and Intensive Care, Sacro Cuore Catholic University, Largo F Vito
1, 00135 Rome, Italy. E-mail: [email protected]
T
he maintenance of spontaneous breathing
during surgical/endoscopic procedures is
a real challenge for the anesthesiologist. An
increasing number of pathologies are treated
with mini-invasive therapies; clinical status of
our patients (poor conditions, neuromuscular
diseases, high risk for weaning process, etc.)
and new settings will force us to elaborate new
strategies for the breathing control.
For many years, the use of halogenated anesthetics has been mandatory but, the development of rapid off-set intravenous drugs seems
to be a key source for the anesthesia community. Since the 1990s indeed, both propofol and
remifentanil have gradually revolutionized our
mindset, leading to experiment their potentiality in several fields.
In this issue of Minerva Anestesiologica,
Besch et al.1 have gone a step forward. The
ENDOTANIL randomized trial he designed,
assesses the efficacy of both propofol and
remifentanil administered in target controlled
infusion (TCI) in tubeless anesthesia with
spontaneous breathing in patients undergoing
pan-endoscopy. This procedure (a sequence
of direct laryngoscopy, rigid bronchoscopy
and esophagoscopy) is a high-risk challenge 2
and requires a close control over the respiratory trigger. The choice of this combination is
Comment on p. 1138.
Vol. 82 - No. 11
a pragmatic approach based on the synergistic
interaction between propofol and remifentanil3
in a large number of patients, reporting good
results and a safe profile.
In the past, several papers looked into this
matter with positive results. Anesthesiologists
utilized propofol-remifentanil technique to
achieve sedation while maintaining spontaneous ventilation in a large number of settings:
awake intubation,4 bronchoscopy,5 complex
digestive endoscopy,6 surgical procedures,7, 8
dentistry 9 and Monitored Anesthesia Care in
adults 10 and children.11
The development of a more accurate administration by TCI models, further promoted
the use of the intravenous technique in clinical practice. Despite the reluctance of the
Food and Drug Administration in the United
States 12 and the lack of a strong evidence of
better outcomes,13 the number of papers published about this topic increased.14
TCI technique has been compared with
traditional total intravenous anesthesia 15 and
with manual controlled anesthesia too,16 but it
does not provide yet sufficient evidence to allow firm recommendations about the TCI use.
The approach of ENDOTANIL in the titration
of propofol TCI is the paradigm of the success of this technique in the clinical practice,
despite strong evidences are lacking. In that
study infusion rate was increased step by step
Minerva Anestesiologica
1129
SBARAGLIA
Propofol-remifentanil and spontaneous breathing
until the loss of consciousness (interindividual
variability resulted in a range between 4.0 and
6.5 mg/mL), and, before starting with procedure, increased to reach a mild overdosage
without affecting spontaneous breathing. Propofol administration is therefore tailored to the
single patient, drawing attention from reached
concentration to end results. We can apply the
same remarks for remifentanil that, however,
was maintained at fixed concentration in order
to investigate the outcome of this study.
The trend of clinical practice pushes toward
an opening up of intravenous drugs, reaching
the extreme in the patient controlled sedation
(PCS). PCS seems to be the new frontier of
sedation: patient, if well-trained, could dose
the level of sedation thought the procedure.
In this way, the risk of over-dosage, should
be reduced, and the satisfaction of patient
improved. First experiences in delivery room
with remifentanil,17 and with propofol in endoscopy suite,18 reported positive results and
the combination of these drugs has been satisfactory utilized in the early 2000s.19
Independently of who is the provider performing sedation, most papers emphasize the
paramount importance of respiratory adverse
events control by an adequate monitoring. Apnea, hypopnea and hypoventilation frequently
occur during spontaneous breathing anesthesia
and the use of SaO2 control is too late in detecting respiratory side effects. By contrast, end-expiratory carbon dioxide control (EtCO2), allows
analyzing in real time the adequacy of alveolar
ventilation and of respiratory rate. The immediate adverse events detection can help the provider in reducing the seriousness of them. After
many authors confirmed these findings, the evidence of a need of an early EtCO2 monitoring
has become so strong that even the American
Society of Anesthesiologists modified its standards for basic anesthetic monitoring, including
it as mandatory (July 2011).20
In this field, the clinical report by Besch
finds its weakness. The number of respiratory
adverse events is distorted by the absence of
a reliable monitoring and the global amount
of apneic events could not be accurately measured. The difference between the two arms
1130
of randomization (remifentanil vs. placebo)
did match for this point, but only with the focus on the severe hypoxia. It is possible that
a closer monitoring, increasing the number of
light/moderate disventilation detected, could
enhance the differences between these groups.
Waiting for new trials about continuous
transcutaneous measurement of pCO2 to guide
sedation,21 EtCO2 monitoring must remain
mandatory during spontaneous breathing sedation and anesthesia.
Despite this bias, ENDOTANIL is a trial
rich of positive suggestions. The combination
of propofol and remifentanil in spontaneous
breathing is a current reality, even if their potential is far from being understood.
References
  1. Besch G, Chopard-Guillemin A, Monnet E, Causeret A,
Jurine A, Baudry G, et al. Propofol-remifentanil anesthesia for upper airway endoscopy in spontaneous breathing
patients: the ENDOTANIL randomized trial. Minerva
Anestesiol 2016;82:1138-48.
  2. Passot S, Servin F, Allary R, Pascal J, Prades JM, Auboyer C, Pascal J, Prades JM, Auboyer C, et al. Target-controlled versus manually-controlled infusion of propofol
for direct laryngoscopy and bronchoscopy. Anesth Analg
2002;94:1212-6.
  3. Mertens MJ, Olofsen E, Engbers FH, Burm AG, Bovill
JG, Vuyk J. Propofol reduces perioperative remifentanil
requirements in a synergistic manner: response surface
modeling of perioperative remifentanil-propofol interactions. Anesthesiology 2003;99:347-59.
  4. Cafiero T, Esposito F, Fraioli G, Gargiulo G, Frangiosa A,
Cavallo LM, et al. Remifentanil-TCI and propofol-TCI
for conscious sedation during fibreoptic intubation in the
acromegalic patient. Eur J Anaesthesiol 2008;25:670-4.
  5.Chen L, Yu L, Fan Y, Manyande A. A comparison between total intravenous anaesthesia using propofol plus
remifentanil and volatile induction/ maintenance of anaesthesia using sevoflurane in children undergoing flexible fibreoptic bronchoscopy. Anaesth Intensive Care
2013;41:742-9.
 6.Borrat X, Valencia JF, Magrans R, Gimenez-Mila M,
Mellado R, Sendino O, et al. Sedation-analgesia with
propofol and remifentanil: concentrations required to
avoid gag reflex in upper gastrointestinal endoscopy.
Anesth Analg 2015;121:90-6.
  7. Berkenstadt H, Perel A, Hadani M, Unofrievich I, Ram
Z. Monitored anesthesia care using remifentanil and propofol for awake craniotomy. J Neurosurg Anesthesiol
2001;13:246-9.
  8.Rosati M, Bramante S, Conti F, Rizzi M, Frattari A, Spina T. Laparoscopic salpingo-oophorectomy in conscious
sedation. JSLS 2015;19.
  9. Nagels AJ, Bridgman JB, Bell SE, Chrisp DJ. Propofolremifentanil TCI sedation for oral surgery. N Z Dent J
2014;110:85-9.
10.Höhener D, Blumenthal S, Borgeat A. Sedation and regional anaesthesia in the adult patient. Br J Anaesth
2008;100:8-16.
Minerva Anestesiologica
November 2016
Propofol-remifentanil and spontaneous breathingSBARAGLIA
11. Malherbe S, Whyte S, Singh P, Amari E, King A, Ansermino JM. Total intravenous anesthesia and spontaneous
respiration for airway endoscopy in children--a prospective evaluation. Paediatr Anaesth 2010;20:434-8.
12.Egan TD, Shafer SL. Target-controlled infusions for intravenous anesthetics: surfing USA not! Anesthesiology
2003;99:1039-41.
13.Wilhelm W. Target-controlled infusion (TCI): clinical
tool or scientific toy? Anaesthesist 2008;57:221-2.
14.Absalom AR, Glen JI, Zwart GJ, Schnider TW, Struys
MM. Target-controlled infusion: a mature technology.
Anesth Analg 2016;122:70-8.
15. Weninger B, Czerner S, Steude U, Weninger E. Comparison between TCI-TIVA, manual TIVA and balanced anaesthesia for stereotactic biopsy of the brain. Anasthesiol
Intensivmed Notfallmed Schmerzther 2004;39:212-9.
16. Moerman AT, Herregods LL, De Vos MM, Mortier EP,
Struys MM. Manual versus target-controlled infusion
remifentanil administration in spontaneously breathing
patients. Anesth Analg 2009;108:828-34.
17.Schnabel A, Hahn N, Broscheit J, Muellenbach RM,
Rieger L, Roewer N, et al. Remifentanil for labour analgesia: a meta-analysis of randomised controlled trials.
Eur J Anaesthesiol 2012;29:177-85.
18. Mazanikov M, Udd M, Kylänpää L, Mustonen H, Lindström O, Färkkilä M, et al. A randomized comparison of
target-controlled propofol infusion and patient-controlled
sedation during ERCP. Endoscopy 2013;45:915-9.
19. Joo HS, Perks WJ, Kataoka MT, Errett L, Pace K, Honey
RJ. A comparison of patient-controlled sedation using either remifentanil or remifentanil-propofol for shock wave
lithotripsy. Anesth Analg 2001;93:1227-32.
20. Weaver J. The latest ASA mandate: CO(2) monitoring
for moderate and deep sedation. Anesth Prog 2011;
58:111-2.
21.Hannam JA, Borrat X, Trocóniz IF, Valencia JF, Jensen
EW, Pedroso A, et al. Modeling respiratory depression
induced by remifentanil and propofol during sedation and
analgesia using a continuous noninvasive measurement
of PCO2. J Pharmacol Exp Ther 2016;356:563-73
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material
discussed in the manuscript.
Article first published online: August 31, 2016. - Manuscript accepted: August 24, 2016. - Manuscript received: July 14, 2016.
(Cite this article as: Sbaraglia F, Sammartino M. Propofol-remifentanil and spontaneous breathing, a magnificent pair. Minerva
Anestesiol 2016;82:1129-31)
Vol. 82 - No. 11
Minerva Anestesiologica
1131
© 2016 EDIZIONI MINERVA MEDICA
Online version at http://www.minervamedica.it
Minerva Anestesiologica 2016 November;82(11):1132-4
EDITORIAL
Ultrasonography during weaning:
a support to a comprehensive skilful strategy
Edoardo DE ROBERTIS 1*, Giovanni M. ROMANO 1, Ornella PIAZZA 2
1Department
of Neurosciences, Reproductive and Odontostomatologic Sciences, University of Naples Federico II,
Naples, Italy; 2Department of Medicine and Surgery, University of Salerno, Salerno, Italy
*Corresponding author: Edoardo De Robertis, Department of Neurosciences, Reproductive and Odontostomatologic Sciences, University of Naples Federico II, via S. Pansini 5, 80131 Naples, Italy. E-mail: [email protected]
M
ost of critical ill patients can be extubated without complications, but 26-42%
of those with mechanical ventilation (MV) fail
the first attempt of weaning.1 Failure of extubation with subsequent reintubation is an independent risk factor for increased mortality.2
Therefore, deciding when a patient is ready for
discontinuation of MV continues to be a challenge.
Causes of weaning failure are:
—— respiratory failure (increased ventilatory
work; reduced compliance from cardiogenic or
non-cardiogenic edema, etc.);
—— cardiac failure (pre-existing cardiac dysfunction, increased metabolic demand, weaning-induced cardiac dysfunction);
—— neuromuscular causes (depression of the
respiratory center, muscular failure, including
the diaphragm),3 ICU-acquired weakness;4
—— metabolic causes: malnutrition, hypoglicemia, electrolytes abnormalities, anemia;
—— neuropsychological causes: anxiety, delirium.
Early recognition and treatment of these
adverse conditions help to reduce the risk of
failure.
Diaphragm is remarkably prone to dysfunction in critically ill patients receiving mechanical ventilation.5 Controlled mechanical ventiComment on p. 1149.
1132
lation induces an oxidative stress which leads
to proteolysis and diaphragm atrophy.6, 7
Ultrasound is an excellent tool for anatomical and functional assessment of the diaphragm, with a valuable cost/ benefit ratio,
since it is a non-invasive procedure. Diaphragmatic performance may be evaluated measuring the excursion of the muscle during inspiration with a 3- to 5-MHz probe (B- or M-mode),
or by measuring the thickness with a 10-MHz
probe (B- or M-mode).8 Both these methods
have shown to correlate adequately with transdiaphragmatic pressure values in spontaneously breathing cardiac surgery patients.9
The study conducted by Flevari et al. in this
issue of Minerva Anestesiologica adds new interesting insight on the potential value of ultrasonography evaluation of diaphragmatic function during weaning in critically ill patients.10
These authors examined 27 patients with difficult and/or prolonged weaning in a prospective
cohort study. They found that during a spontaneous breathing trial, a left hemidiaphragm excursion at a cut-off 10 mm, assessed by M-mode
ultrasonography, was the best index to predict
weaning success (sensitivity 86%, specificity 85%, negative predictive value 94%). The
cut-off values for predicting weaning failure as
determined by the area under the receiver operating characteristic curve, were ≤10 mm for
right hemidiaphragm excursion and ≤7 mm
Minerva AnestesiologicaNovember 2016
Ultrasonography during weaning: a support to a comprehensive skilful strategy
for left hemidiaphragm excursion. The authors
concluded that ultrasonographic measurement
of diaphragmatic excursion may be used as an
adjunct predictor for the weaning process in patients with difficult and/or prolonged weaning.
Indeed, focusing on the diaphragm excursion
during the weaning process may be very intriguing. As a matter of fact, ultrasound method is
by far easier to perform compared with more
complex methods such as transdiaphragmatic
pressure or maximal inspiratory pressure measurements. However, major drawbacks hide beyond the easiness and rapidity of the ultrasound
technique. One is the operator-dependence and
reproducibility of the results. In their study, Flevari et al., did not test interobserver and intraobserver variability, even though diaphragmatic
excursion studied by M-mode ultrasonography
seems to be a highly reproducible method.11
Moreover, the ultrasonographic evaluation of
diaphragmatic function needs to take into account different variables:
—— the precise identification of the phase of
the respiratory cycle and its duration;12
—— the type and quality of breathing (quiet,
deep, superficial);
—— the patient’s position;
—— the presence of PEEP (the increase in
end-expiratory lung volume lowers the diaphragmatic dome and may decrease the diaphragmatic excursion);13
—— the presence of anatomic and/or pathologic alterations which can hinder the ultrasound scan.
The measurement of diaphragmatic excursion requires a thorough understanding and
identification of the various ultrasound diaphragmatic patterns in different illnesses and
the corresponding changes in breathing mechanics. One strength of the study by Flevari
et al., is the fact that the authors did measurements of both hemidiaphragmatic excursions,
which may give a hint in the global diaphragmatic function evaluation and help understanding how the hemidiaphragms are working
together. However, extrapolating the global
respiratory function from the evaluation of
the diaphragmatic excursion may be oversimplified. The interplay between thoracic cage,
Vol. 82 - No. 11
DE ROBERTIS
intercostal muscle and diaphragm contraction determine the force which generates flow
through the airways, and the measurement of
the diaphragmatic excursion is only one link
of this complex biological network.14, 15 Nevertheless, we know that the causes of failure in
retirement of MV are multifactorial and they
cannot be related to the diaphragm dysfunction only. To remove mechanical ventilation
the earliest is possible, it is necessary to reconsider the need for mechanical ventilation daily.
As the authors suggest, ultrasonographic
evaluation of the diaphragm is undoubtedly
a powerful adjunct tool to predict weaning
failure. This non-invasive technology offers
the advantage of daily monitoring of the diaphragmatic function, assessing the progressive
improvement over time of the patients and the
readiness to wean from mechanical ventilation. However, ultrasonographic evaluation
during the weaning needs to be integrated into
defined protocols, which consider specific cutoff for predicting weaning success in patients
with different illnesses, and in a comprehensive skillful strategy that envisions all possible
causes of a difficult weaning.
References
 1. Boles JM, Bion J, Connors A, Herridge M, Marsh B,
Melot C, et al. Weaning from mechanical ventilation. Eur
Respir J 2007;29:1033-56.
 2.Frutos-Vivar F, Esteban A, Apezteguia C, González
M, Arabi Y, Restrepo MI, et al. Outcome of reintubated patients after scheduled extubation. J Crit Care
2011;26:502-9.
  3. McCool FD, Tzelepis GE. Dysfunction of the diaphragm.
N Engl J Med 2012;366:932-42.
  4. Jung B, Moury PH, Mahul M, de Jong A, Galia F, Prades
A, et al. Diaphragmatic dysfunction in patients with ICUacquired weakness and its impact on extubation failure.
Intensive Care Med 2016;42:853-61.
  5.Sieck GC, Mantilla CB. Effect of mechanical ventilation on
the diaphragm. N Engl J Med 2008;358:1392-4.
  6.Levine S, Nguyen T, Taylor N, Friscia ME, Budak MT,
Rothenberg P, et al. Rapid disuse atrophy of diaphragm
fibers in mechanically ventilated humans. N Engl J Med
2008;358:1327-35.
  7. Dres M, Dubé B-P, Mayaux J, Delemazure J, Reuter D,
Brochard L, et al. Coexistence and impact of limb muscle
and diaphragm weakness at time of liberation from mechanical ventilation in medical icu patients. Am J Respir
Crit Care Med 2016 [Epub ahead of print].
  8. Mayo P, Volpicelli G, Lerolle N, Schreiber A, Doelken
P, Vieillard-Baron A. Ultrasonography evaluation during
the weaning process: the heart, the diaphragm, the pleura
and the lung. Intensive Care Med 2016;42:1107-17.
Minerva Anestesiologica
1133
DE ROBERTIS
Ultrasonography during weaning: a support to a comprehensive skilful strategy
  9.Lerolle N, Guérot E, Dimassi S, Zegdi R, Faisy C, Fagon
J-Y, et al. Ultrasonographic diagnostic criterion for severe diaphragmatic dysfunction after cardiac surgery.
Chest 2009;135:401-7.
10. Flevari A, Lignos M, Konstantonis D, Armaganidis A.
Diaphragmatic ultrasonography as an adjunct predictor
tool of weaning success in patients with difficult and prolonged weaning. Minerva Anestesiol 2016;82:1149-57.
11. Boussuges A, Gole Y, Blanc P. Diaphragmatic motion
studied by m-mode ultrasonography: methods, reproducibility, and normal values. Chest 2009;135:391400.
12. De Robertis E, Uttman L, Jonson B. Re-inspiration of
CO2 from ventilator circuit: effects of circuit flushing and
aspiration of dead space up to high respiratory rate. Crit
Care 2010;14:R73.
13. Uttman L, Bitzén U, De Robertis E, Enoksson J, Johansson L, Jonson B. Protective ventilation in experimental
acute respiratory distress syndrome after ventilatorinduced lung injury: a randomized controlled trial. Br J
Anaesth 2012;109:584-94.
14. De Robertis E, Zito Marinosci G. The luxury of breathing
oxygen. Minerva Anestesiol 2013;79:1324-5.
15.Iannuzzi M, De Sio A, De Robertis E, Piazza O, Servillo G, Tufano R. Different patterns of lung recruitment
maneuvers in primary acute respiratory distress syndrome: effects on oxygenation and central hemodynamics. Minerva Anestesiol 2010;76:692-8.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material
discussed in the manuscript.
Article first published online: September 16, 2016. - Manuscript accepted: September 15, 2016. - Manuscript received: August 17,
2016.
(Cite this article as: De Robertis E, Romano GM, Piazza O. Ultrasonography during weaning: a support to a comprehensive skilful
strategy. Minerva Anestesiol 2016;82:1132-4)
1134
Minerva AnestesiologicaNovember 2016
© 2016 EDIZIONI MINERVA MEDICA
Online version at http://www.minervamedica.it
Minerva Anestesiologica 2016 November;82(11):1135-7
EDITORIAL
Perioperative hemodynamic therapy:
goal-directed or meta-directed?
Christopher UHLIG, Peter M. SPIETH*
Department of Anesthesiology and Critical care Medicine, University Hospital Dresden, Technische Universität
Dresden, Dresden, Germany
*Corresponding author: Peter Markus Spieth, Department of Anesthesiology and Critical Care Medicine, University Hospital Dresden, Technische Universität Dresden, Fetscherstraße 74, 01307 Dresden, Germany. E-mail: [email protected]
P
erioperative hemodynamic stability is one
of the primary objectives during anesthesia. To achieve this goal, a blood pressure
within a predefined range, fluid homeostasis as
well as sufficient oxygen delivery is warranted
to avoid impairment of organ function. To improve patient outcome different protocols have
been developed focusing mainly on the use of
fluids, red blood cells and vasoactive drugs to
reach normal or in some cases supra-normal
values of cardiac output and/or oxygen delivery.
In the author’s opinion, many factors contribute to hemodynamic instability or impairment of postoperative organ function. The following factors are of paramount importance
considering perioperative patient outcome:
—— the patient. Individual preoperative assessment of the patients’ medical history, ASA
status and preoperative risk factors are essential for a good outcome after surgery;
—— the surgeon. Maybe the most important
factor. The type of surgery and the resulting
surgical trauma as well as the experience of
the surgeon are key factors for patient survival. The type of surgery may force the anesthetists to be restrictive with fluids, since this
influence postoperative outcome especially in
surgical procedures where anastomoses of the
Comment on p. 1199.
Vol. 82 - No. 11
gastrointestinal tract are performed. The influence of the surgeon is often underestimated or
underrepresented by evidence based medicine.
Different types of surgery were often combined for trials performed by anesthetists and/
or the surgical procedures were performed by
the same, usually experienced team. This may
not reflect the reality especially in teaching
hospitals. Since the surgeon is the main reason
for intraoperative blood loss, duration of anesthesia and postoperative complications such
as sepsis caused by impaired wound healing
or insufficiency of anastomoses, these complications might by underrepresented in randomized controlled trials, or patients withdraw
consent postoperatively due to the experience
of a complication;
—— the anesthesiologist. The experience of
the anesthesiologist and his reaction to fast
changing intraoperative conditions are also
contributing to patient outcome. Like the surgeon, experience and patient centered decision
making are the keys to a successful treatment
of severe intraoperative hemorrhage or other
complications like myocardial infarction, lung
embolism or anaphylactic reaction;
—— the team and resources. The experience
of the whole team treating the patient intraoperatively, their interaction, their awareness
and prediction of critical situations are crucial. Technical resources, such advanced he-
Minerva Anestesiologica
1135
UHLIG
PERIOPERATIVE HEMODYNAMIC THERAPY
modynamic monitoring, devices for warming
of infusion and blood products and access to
adequate amount of blood products in case of
severe hemorrhage;
—— the goal. A standard established protocol
for goal-directed therapy may help to guide the
anesthesiologist during the procedure. Different protocols have been developed and investigated in randomized controlled trials over
the last two decades. However, the definition
of a goal for fluid therapy remains tricky since
hemodynamic parameters like central venous
pressure, pulse pressure variation or pulse contour derived cardiac output analysis are not reliable as single value or only valid in case of
sinus rhythm in absence of cardiac valve pathology, respectively.
In this issue of Minerva Anestesiologica,
Giglio et al. performed a meta-analysis regarding the effects of goal-directed therapy
protocols in a mixed surgical population consisting mainly of abdominal surgical patients
on mortality.1 This meta-analysis was state of
the art and is reported in accordance with the
PRISMA statement.2 Giglio et al. analyzed 58
articles including the data of 8171 patients.
Giglio et al. found that goal-directed protocols
reduce perioperative mortality. In addition,
a meta-regression model was used to determine the cut-off and the influence of risk of
bias, mortality rate and publication year. Metaregression is a method which provides an extension of standard meta-analysis techniques,
where the influence of statistical heterogeneity
among results of multiple studies can be related to one or more characteristic of the trial.3
A subgroup analysis dividing the trials in two
groups according to risk of bias and mortality
in the control group, which were significant in
the meta-regression, revealed that this effect is
only consisting in the case of high risk of bias
and a mortality in the control group greater than
10%. In trials with low risk of bias no statistic
significant difference was detected. However,
the analyses have several limitations. First of
all, the risk of bias assessment was not correct
according to the Cochrane collaboration, since
the item “other risk of bias” was not assessed
and the overall risk was not rated according to
1136
the GRADE approach.4, 5 Since most included
trials had unclear risk of bias and only five trials fulfil the criteria for low risk of bias in all
domains. The results of subgroup analysis related to risk of bias may be invalid. Second,
mortality was not defined as a fixed time point,
however the authors state that in-hospital mortality was considered in case of more than one
type of mortality data. Furthermore, mortality
might not be the favorable outcome parameter
since pulmonary or extrapulmonary complications e.g. insufficiency of bowel anastomoses
or cardiac events are more likely to be affected by goal directed therapy than mortality. Third, the analyses protocol was not registered before. Fourth, since death was a rare
event in most trials the Peto odds ratio might
be more appropriate for the analyses of binary
outcomes.6, 7 Fifth, the goal parameters differ
among trials as well as the type and amount
of fluids and vasoactive drugs, which were no
further analyzed. Sixth, no subgroup analysis
among the different surgical procedures (e.g.
cardiac surgery vs. non cardiac surgery). Since
most trials had unclear or high risk of bias the
results of the work performed by Giglio et al.
should be interpreted with caution.
In summary, we do not know whether goaldirected therapy improves patient survival
since most randomized controlled trials are
not powered for mortality as primary outcome.
Interpreting the data of this meta-analysis is
more likely that there is no effect on mortality. As pointed out by the authors there is a
need for a large randomized controlled trial
powered for mortality. Although goal-directed
therapy may be a useful tool in high-risk patients, the experience and skills of surgeons
and anesthetists are more likely to contribute
to patient outcome.
References
  1. Giglio M, Manca F, Dalfino L, Brienza N. Perioperative
hemodynamic goal-directed therapy and mortality: systematic review and meta-analysis with meta-regression.
Minerva Anestesiol 2016;82:1199-213.
  2. Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA
Group. Preferred reporting items for systematic reviews
and meta-analyses. Br Med J 2009;339:b2535.
Minerva Anestesiologica
November 2016
PERIOPERATIVE HEMODYNAMIC THERAPYUHLIG
  3. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Statistical heterogeneity in systematic reviews of clinical trials:
A critical appraisal of guidelines and practice. J Health
Serv Res Policy 2002;7:51-61.
  4. Higgins JP, Altman DG, Gotsche PC, Jüni P, Moher D,
Oxman AD, et al; Cochrane Bias Methods Group; Cochrane Statistical Methods Group. The Cochrane Collaboration’s tool for assessing risk of bias in randomized
trials. Br Med J 2011;343:d5928.
  5. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter
Y, Alonso-Coello P, et al; GRADE Working Group.
GRADE: An emerging consensus on rating quality of
evidence and strength of recommendations. Br Med J
2008;336:924-6.
  6. Yusuf S, Peto R, Lewis J, Collins R, Sleight P. β-Blockade
during and after myocardial infarction: An overview of the
randomized trials. Prog Cardiovasc Dis 1985;27:335-71.
  7. Brockhaus AC, Bender R, Skipka G. The Peto odds
ratio viewed as a new effect measure. Stat Med J
2014;33:4861-74.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material
discussed in the manuscript.
Article first published online: September 16, 2016. - Manuscript accepted: September 15, 2016. - Manuscript revised: September 6,
2016. - Manuscript received: June 1, 2016.
(Cite this article as: Uhlig C, Spieth PM. Perioperative hemodynamic therapy: goal-directed or meta-directed? Minerva Anestesiol
2016;82:1135-7)
Vol. 82 - No. 11
Minerva Anestesiologica
1137
© 2016 EDIZIONI MINERVA MEDICA
Online version at http://www.minervamedica.it
Minerva Anestesiologica 2016 November;82(11):1138-48
ORIGINAL ARTICLE
Propofol-remifentanil anesthesia for upper
airway endoscopy in spontaneous breathing
patients: the ENDOTANIL Randomized Trial
Guillaume BESCH 1, 2, Angeline CHOPARD-GUILLEMIN 1, Elisabeth MONNET 3, 4,
Arnaud CAUSERET 1, Amelie JURINE 1, Gerald BAUDRY 1, Benjamin LASRY 1,
Laurent TAVERNIER 5, 6, Emmanuel SAMAIN 1, 2, Sebastien PILI-FLOURY 1, 2 *
1Department
of Anesthesiology and Intensive Care Medicine, University Hospital of Besancon, University of
Franche-Comte, Besancon, France; 2EA 3920, INSERM SFR 133, Besancon, France; 3Clinical Investigation Center,
University Hospital of Besancon, Besancon, France; 4SFR-FED 4234, University of Franche-Comte, Besancon,
France; 5Department of ENT Surgery, University Hospital of Besancon, Besancon, France; 6EA 481 and SFR-FED
4234, University of Franche-Comte, Besancon, France
*Corresponding author: Sebastien Pili-Floury, Department of Anesthesiology and Intensive Care Medicine, University Hospital of
Besancon, 3 bvd Alexander Fleming, F-25000 Besancon, France. E-mail: [email protected]
A B STRACT
BACKGROUND: The ENDOTANIL Trial aimed at comparing an association of target-controlled infusion (TCI) of
remifentanil and propofol to TCI of propofol alone on the clinical conditions during pan endoscopy for assessment of the
upper airway (pan endoscopy) performed under tubeless general anesthesia.
METHODS: This double-blind, single center, parallel, randomized, placebo-controlled trial was conducted in a French
tertiary level of care, from June 2009 to February 2013. Patients scheduled for elective pan endoscopy were anesthetized
using propofol TCI combined to either remifentanil TCI (effect-site concentration=1.5 ng.mL-1; remifentanil group) or
placebo (control group). The main outcome measure was the percentage of clinically acceptable conditions for pan endoscopy, using a 5-criteria score (ease of laryngoscopy, position and movements of the vocal cords, cough and movements of
the limbs to stimulation). The secondary outcomes were hemodynamic and respiratory safety.
RESULTS: In this study 218 patients (mean±SD age 60 [10] yrs) were included. Clinically acceptable conditions were
observed in 68% and 64% of the patients included in Remifentanil and Control group, respectively (P=0.39). None of
the 5 parameters of the pan endoscopy score was significantly different between the 2 groups. Hemodynamic alterations
were significantly lower in the Remifentanil as compared to the control group. Incidence of hypoxemia or need for rescue
mechanical ventilation did not significantly differ between the 2 groups.
CONCLUSIONS: The adjunction of remifentanil to propofol TCI, at a dose that maintain spontaneous breathing, did not
improve the conditions for pan endoscopy, but attenuates the hemodynamic response induced by upper airway stimulation.
(Cite this article as: Besch G, Chopard-Guillemin A, Monnet E, Causeret A, Jurine A, Baudry G, et al. Propofol-remifentanil
anesthesia for upper airway endoscopy in spontaneous breathing patients: the ENDOTANIL Randomized Trial. Minerva
Anestesiol 2016;82:1138-48)
Key words: Anesthesia, General - Anesthesia, Intravenous - Endoscopy - Bronchoscopy - Laryngoscopy - Propofol.
P
an endoscopy for assessment of the upper
aero digestive (pan endoscopy) tract consists in a three steps endoscopy: a direct larynComment in p. 1129.
1138
goscopy, a rigid bronchoscopy and an esophagoscopy, performed under general anesthesia.
As endotracheal intubation is considered inadequate by many surgeons for both laryngoscopy and bronchoscopy, either high frequency
Minerva AnestesiologicaNovember 2016
PROPOFOL-REMIFENTANIL ANESTHESIA FOR UPPER AIRWAY ENDOSCOPYBESCH
jet ventilation or spontaneous ventilation without endotracheal intubation (tubeless anesthesia) has been proposed during these first 2
steps of the procedure.1-3 Tubeless anesthesia
requires cautious titration of short-acting anesthetic drugs to achieve a highly controlled
anesthetic depth which warrants both optimal
conditions for pan endoscopy and minimal risk
of drug-induced apnea. The main challenge for
anesthesiologists is to enable both optimal exposure of patient’s oropharyngeal tract and adequate oxygen supply. Total intravenous general anesthesia using target-controlled infusion
(TCI) of propofol has been shown to be safe
and efficient and is commonly used for the anesthetic management for pan endoscopy.4
However, the noxious stimulations observed during pan endoscopy may be similar
to those described during endotracheal intubation performed by direct laryngoscopy and
can be responsible of cough, movement of the
patient, peri-procedure awareness, and/or haemodynamic alterations.5, 6 Coadministration of
opioids could be considered to reduce the occurrence of these events,7, 8 and improve the
conditions of pan endoscopy.9-12 Remifentanil
is a powerful short-acting esterase-metabolized opioid whose pharmacokinetic profile
could be well suited for the anesthetic management of pan endoscopy.13-15 Coadministration
of remifentanil and propofol were reported to
allow for titration of the anesthetic depth that
enables adequate oxygen supply during spontaneous breathing.16, 17
The aim of the randomized, placebo-controlled ENDOTANIL Trial was to test the hypothesis that the addition of remifentanil to
propofol TCI anesthesia improved the clinical
conditions of the two first steps, laryngoscopy
and bronchoscopy, in spontaneously breathing
anaesthetized patients undergoing pan endoscopy.
Materials and methods
Design of the study
The ENDOTANIL Trial was a randomized, placebo-controlled, double-blind, single-
Vol. 82 - No. 11
center clinical study, conducted according to
the French bioethics law (Art. L. 1121-1 of
the law no. 2004-806, August 9th, 2004), at a
University Hospital in France. The study was
approved by the Institutional Review Board
Est-II, University Hospital of Besancon, Besancon, France (no. 08/501, Chairperson Prof
E. Haffen, MD, PhD) on April 9th, 2009. It was
registered in the European Union Drug Regulating Authorities Clinical Trials (EudraCT no.
2008-007758-36) on November 18th, 2008. All
patients gave written informed consent to participate to the study. Propofol (Diprivan®, AstraZeneca, London, UK) and remifentanil (Ultiva®, GlaxoSmithKline UK Ltd, Brentford,
UK) used for anesthesia are drugs approved
by the French Drug Regulatory Agency for
human administration. All patients above 18
years, scheduled for pan endoscopy under general anesthesia between June 2009 and February 2013 were eligible. Non-inclusion criteria
were: age <18 or >80 years, inability or refusal
to give informed consent, pregnancy and/or
breast feeding, long-term opioid treatment,
known drug abuse or drug dependency, allergy
to one of the study drugs, and chronic obstructive lung disease requiring oxygen therapy for
more than 12 hours a day.
Anesthetic management
Upon arrival at operating room 1 hour after oral premedication using hydroxyzine 1
mg.kg-1, a standard monitoring was set up
(General Electric Healthcare, Fairfield, CT,
USA). Heart rate (HR), mean arterial pressure
(MAP), respiratory rate (RR) and pulse oximetry (SpO2) values were collected before the
induction of anesthesia and then every 3 min
during the study period by an independent
observer. After IRB approval on April 27th
2010, a BIS™ monitoring was added (Aspect BIS™ XP™ or BIS™ VISTA™, Aspect
Medical Systems Inc., Newton, MA, USA)
and BIS value measured before induction of
anesthesia and just before laryngoscopy. The
anesthesiologist in charge of the patient was
blinded to BIS values throughout the study
period.
Minerva Anestesiologica
1139
BESCH
PROPOFOL-REMIFENTANIL ANESTHESIA FOR UPPER AIRWAY ENDOSCOPY
A 20-gauge catheter was placed in a forearm vein and connected to a 3-way needle-free
system (Smartsite®, Alaris Medical Systems,
San Diego, CA, USA) allowing the infusion of
both saline at a constant flow of 3 mL.kg-1.h-1
for hydration, and anesthetic drugs, delivered
using a Base Primea™ infusion system (Fresenius Vial SAS, Brezins, France).
After preoxygenation with 100% oxygen
via a face mask (oxygen expiratory fraction
>90%), anesthesia was induced with propofol
TCI to achieve an initial propofol effect-site
concentration (propofol-Ce) = 3 µg.mL-1, using Schnider pharmacokinetic model.18, 19 Infusion rate was then increased by step of 0.5
µg.mL-1 every 2 min, and the propofol-Ce at
loss of consciousness (propofol-Ce LOC) was
noted. Facial mask ventilation was tested, and
the patient excluded in case of difficult or impossible facial mask ventilation. When stable
spontaneous breathing was obtained, oxygen
3 L.min-1 was delivered via a nasal cannula,
and maintained throughout the pan endoscopy. Then, the propofol-Ce was increased at a
value equal to 1.3 * propofol-Ce LOC; and the
infusion of the study drug (placebo or remifentanil according to randomization) was
started. Study period ended when both direct
laryngoscopy and rigid bronchoscopy were
completed, and infusion of the study drug was
stopped. Esophagoscopy was then performed
after endotracheal intubation, using an anesthetic protocol managed by the attending anesthesiologist.
Randomization and study drug administration
Patients included in the study were randomly assigned the day before pan endoscopy into either Control or Remifentanil group.
Randomization was performed in-block (ratio
1:1) using a computer-generated randomnumber table (nQuery Advisor® software
v6.0, nQuery Advisor®, Los Angeles, CA,
USA), prepared by our biostatistician (EM).
Patients, investigators, practitioners in charge
of patients, evaluators and data analysts were
blinded to both block sizes of randomization and the study treatment administered.
1140
The study drug allocated by the randomization was prepared and delivered the day of
the pan endoscopy by the Hospital pharmacist to the anesthesiologist in charge of the
patient. The appearance and the volume of
the syringes of placebo (40 mL of saline) and
of remifentanil (2 mg of remifentanil diluted
in 40 mL of saline, to obtain a remifentanil
concentration of 50 µg.mL-1) were identical.
Infusion rate was adjusted to reach a study
drug effect-site concentration (study drug-Ce)
= 1.5 ng.mL-1, calculated using Minto pharmacokinetic model.18 This study drug-Ce value was chosen as the median value allowing
spontaneous breathing reported in the study
by Murdoch et al., and was maintained steady
throughout the study period.16 After steadystate effect-site concentrations were obtained
for both drugs, the surgeon started the pan
endoscopy.
If a moderate hypoxemia and/or apnea lasting more than 60 sec occurred, both pan endoscopy and study drug infusion were stopped,
propofol-Ce was decreased and oxygen delivery was increased to 5 L.min-1. If both a
spontaneous breathing at a RR >8 c.min-1 and
SpO2 value >95% were obtained in the following next minute, the study drug infusion
was started again to reach a study drug-Ce =
0.75 ng.mL-1, i.e. half the former study drugCe. If the respiratory status worsened to severe
hypoxemia and/or RR remained <8 c.min-1,
emergency manual ventilation via a facial
mask with 100% oxygen was performed. The
decision for tracheal intubation or surgical tracheostomy was left to the discretion of the attending anesthesiologist.
Data collected and endpoint measures
Demographic data, past medical history,
ASA physical status, Revised Cardiac Risk Index described by Lee et al.,20 were collected
at inclusion. Tumor localization, extent and
histology of the tumor were obtained after biopsies analysis.
The condition for the pan endoscopy was
assessed using a 5-criteria score (Table I),
adapted from the clinical research practice
Minerva AnestesiologicaNovember 2016
PROPOFOL-REMIFENTANIL ANESTHESIA FOR UPPER AIRWAY ENDOSCOPYBESCH
Table I.—Score to assess pan endoscopy* conditions during general anesthesia.
Pan endoscopy conditions
Variables
Reactivity during pan endoscopy*
Coughing during pan endoscopy*
Movement of limbs during pan endoscopy*
Initial vocal cords position at laryngoscopy
Vocal cords movements during laryngoscopy
Laryngoscopy
Excellent
Good
Poor
None
None
Abducted
None
Easy
Diaphragm
Slight
Intermediate
Moving
Fair
Sustained (>10 s)
Vigorous
Closed
Closing
Difficult
Laryngoscopy: Easy = jaw relaxed, no resistance to blade during laryngoscopy; Fair = jaw not fully relaxed, slight resistance to blade; Difficult
= poor jaw relaxation, active resistance by the patient.
Condition for laryngoscopy and vocal cords position and movements were assessed by the surgeon, blinded to the study drug used.
*Pan endoscopy consists in a three steps endoscopy for assessment of the upper aero digestive tract: a direct laryngoscopy, a rigid bronchoscopy and an esophagoscopy. Conditions were assessed during the first 2 steps of the pan endoscopy: the direct laryngoscopy and the rigid
bronchoscopy.
guidelines described by Viby-Mogensen et
al.21 Condition of pan endoscopy was graded
“excellent”, if all criteria were rated excellent;
“good” if all criteria were at least good; or
“poor” if at least one criteria was graded poor.
The primary endpoint measure was the percentage of clinically acceptable conditions for
pan endoscopy, defined as conditions graded
as either excellent or good.
The secondary endpoint measures were: 1)
percentage of direct laryngoscopy or bronchoscopy interruption for inadequate anesthesia; 2) hemodynamic alterations observed
during pan endoscopy, and 3) occurrence of
respiratory adverse events, 4) complications
related to pan endoscopy.
Hemodynamic alterations were defined
by the maximal and minimal variation (expressed in percent of preinduction value)
of HR (∆-HRmax, and ∆-HRmin) and MAP
(∆-MAPmax, and ∆-MAPmin) observed during
the first 2 steps of pan endoscopy.
Respiratory adverse events were defined by
the number of episodes of moderate (SpO2 value <94%) or severe (SpO2 <90%) hypoxemia,
and the need for rescue endotracheal intubation and mechanical ventilation.
Pan endoscopy-related complications were
defined as follows: significant hemorrhage
and dental and/or mucosal injuries related to
the endoscopic procedure or to the biopsy. A
double data entry was done a posteriori by an
independent data manager on a randomized
sample of 22 patients to assess the reliability
of the data collected.
Vol. 82 - No. 11
Statistical analysis
The Shapiro-Wilk Test was used to check
whether each quantitative variable was normally distributed or not. Normally distributed
and non-normally distributed quantitative variables, expressed respectively as mean (SD)
and as median [interquartile range], have been
compared between the Control and Remifentanil groups by using respectively the Student’s
t-test and the Mann-Whitney U-Test. The intergroup comparisons for qualitative data, expressed as number of patients (percentage),
were done by using the χ2 Test. All analyses
were performed on an intention-to-treat basis.
The reliability of the data collected by the
two independent observers in the randomized sample of 22 patients was assessed using
the non-parametric tests of agreement of the
Cohen’s kappa coefficient for the qualitative
variables and of the Kendall rank correlation
coefficient for the quantitative variables (data
not shown).
A sample size of 222 patients was necessary
to identify an increase in the percentage of
clinically acceptable conditions for the pan endoscopy from 50% to 70% in the Remifentanil
as compared to Control group, with a two-sided type I error of 0.05 and a power of 90%. We
planned to include 256 patients in case of patients drop out. Two interim analysis were performed by an independent biostatistician (EM)
after the inclusion of 80 and 160 patients, with
stopping boundaries allowing early termination of the study, as described by O’Brien and
Minerva Anestesiologica
1141
BESCH
PROPOFOL-REMIFENTANIL ANESTHESIA FOR UPPER AIRWAY ENDOSCOPY
Fleming (P=0.0006 and P=0.016 for the 1st and
the 2nd interim analysis, respectively).22
Statistical analyses were performed using
SAS 9.3 software (SAS Institute Inc., Cary,
NC, USA). All P-value are two-sided and
P<0.05 was considered statistically significant.
Results
Four hundred and fifty eight patients were
scheduled for pan endoscopy during the study
period. As no interim analysis reached the statistical significance, the study was conducted
until 256 patients, were randomized in either
Control or Remifentanil groups. Two hundred
and eighteen patients were finally analyzed
(Figure 1). Demographic data, co-morbidities, and type and location of tumors of the 2
groups are given in Table II. Direct laryngoscopy and bronchoscopy could be completed
in all patients while maintaining spontaneous
breathing without endotracheal intubation, and
a biopsy was done in 216 cases. The results
of intraoperative measures are given in Table
III. The values of propofol-Ce LOC, total dose
Table II.—Demographic data of patients in the CONTROL and in the REMIFENTANIL groups.
CONTROL
group
(N.=106
patients)
REMIFENTANIL group
(N.=112
patients)
Age (yr)
60 (10)
60 (9)
BMI¥† (kg.m-²)
23 [21-26]
24 [21-27]
Male gender*
89 (84)
91 (81)
Obesity* (defined as BMI† ≥
11 (10)
9 (8)
30 kg.m-²)
History of chronic alcohol
60 (57)
64 (57)
abuse*
ASA§ physical status
28 (26)
25 (22)
III-IV*
Revised cardiac risk index‡ 100 (94)/6 (6) 110 (98)/2 (2)
value 0-1/≥ 2*
Type of tumour
Squamous cell carcinoma*
90 (86)
85 (75)
Oral cavity or larynx*
58 (55)
43 (38)
Naso-, oro- or
44 (42)
39 (35)
hypo-pharynx*
Esophagus*
1 (1)
1 (1)
Multiple tumour sites*
6 (6)
5 (4)
Other type of tumour*
13 (12)
24 (21)
Negative biopsy*
5 (5)
7 (6)
Values are mean (SD). ¥Values are median [interquartile range].
*Values are number (proportion).
†BMI: Body Mass Index = weight/height².
†‡Revised cardiac risk index is calculated according to Lee et al.19
§ASA: American Society of Anesthesiologists.
Patients scheduled for pan endoscopy: N.=458
Patients randomized: N.=256
Patients allocated to control group: N.=128
Patients excluded: N.=22
— Consents withdrawal: N.=3
—  Pan endoscopy cancelled: N.=9
—  Preoperative death*: N.=0
— Incomplete data: N.=8
—  Difficult face mask ventilation: N.=2
Patients analyzed: N.=106
Patients allocated to Remifentanil group: N.=128
Patients excluded: N.=16
— Consents withdrawal: N.=2
—  Pan endoscopy cancelled: N.=5
—  Preoperative death*: N.=1
— Incomplete data: N.=6
—  Difficult face mask ventilation: N.=2
Patients analyzed: N.=112
Figure 1.—Flow chart of patient inclusions in the study. *preoperative pulmonary embolism. Pan endoscopy consists in a
three steps endoscopy for assessment of the upper aero digestive tract consists: a direct laryngoscopy, a rigid bronchoscopy
and an esophagoscopy.
1142
Minerva AnestesiologicaNovember 2016
PROPOFOL-REMIFENTANIL ANESTHESIA FOR UPPER AIRWAY ENDOSCOPYBESCH
Table III.—Intraoperative parameters of patients in the CONTROL and in the REMIFENTANIL groups.
CONTROL group (N.=106)
Anesthesia parameters
Infusion of the study drug
Total volume infused (mL)
Total dose infused (µg.kg-1.min-1)
Total dose infused (µg)
Duration of the infusion (min)
Infusion of propofol
Total dose infused (mg.kg-1.min-1)
Total dose infused (mg)
Duration of the infusion (min)
Propofol-Ce LOC ‡(µg.ml-1)
BIS values**†
BISt0†
BISlaryngo†
Safety parameters
Hemodynamic safety§
HRmax (bpm)
HRmin (bpm)
HRt0 (bpm)
Δ-HRmax (%)
Δ-HRmin (%)
MAPmax (mmHg)
MAPmin (mmHg)
MAPt0 (mmHg)
Δ-MAPmax (%)
Δ-MAPmin (%)
Respiratory safety¶
Moderate hypoxemia*
Severe hypoxemia*
Surgical complications*
REMIFENTANIL group (N.=112)
P value
1.17 [0.95-1.43]
0
0
10 [8-14]
1.18 [0.96-1.44]
0.085 [0.074-0.102]
59.2 [48.0-72.0]
10 [7-13]
0.94
0.56 [0.45-0.71]
525 [393-700]
14 [11-17]
5.0 [4.0-6.5]
0.53 [0.42-0.67]
500 [379-674]
13 [10-17]
5.5 [4.0-6.3]
0.37
0.80
0.54
0.72
97 [96-98]
29 [23-40]
97 [94-98]
27 [24-37]
91 [80-100]
80 [67-87]
68 [62-81]
31 [13-44]
10 [0-24]
105 [95-122]
89 [77-100]
103 [93-113]
5 [(-6)-20]
13 [5-24]
90 [79-100]
78 [66-85]
74 [65-85]
17 [8-33]
1 [(-5)-11]
101 [90-115]
81 [71-94]
101 [92-115]
0 [(-13)-12]
21 [12 -29]
19 (18)
11 (11)
4 (4)
25 (22)
11 (10)
1 (1)
0.53
0.13
0.78
0.71
0.31
0.01
0.0016
0.0001
0.06
0.0022
0.94
0.02
0.002
0.44
0.87
0.20
Values are median [interquartile range]. *Values are number (proportion).
†BIS: bispectral index. BISt0 and BISlaryngo are BIS values before induction of anesthesia and just before the laryngoscopy, respectively.
‡ propofol-Ce LOC: effect-site concentration of propofol predicted by the Schnider model at loss of consciousness during induction of anesthesia.
§t0, max and min values for heart rate (heart rate) and MAP (mean arterial pressure) are values observed before induction of anesthesia, highest and lowest values during the procedure, respectively. Δ-HRmax, Δ-HRmin, Δ-MAPmax and Δ-MAPmin are the maximal and minimal variation
(expressed in percent of preinduction value) of HR (∆-HRmax, and ∆-HRmin) and MAP (∆-MAPmax, and ∆-MAPmin) observed during the first 2
steps of the pan endoscopy: the direct laryngoscopy and the rigid bronchoscopy.
¶Moderate hypoxemia and severe hypoxemia are defined by a decrease in SpO2 <94% and <90%, respectively.
**Data obtained in 87 and 95 patients in the CONTROL and REMIFENTANIL groups, respectively.
Pan endoscopy consists in a three steps endoscopy for assessment of the upper aero digestive tract: a direct laryngoscopy, a rigid bronchoscopy
and an esophagoscopy.
of propofol infused, average dose of propofol,
total volume of study-drug infused, were not
different between the 2 groups (Table III). BIS
value just before laryngoscopy, obtained in
182 patients (respectively 87 and 95 patients
in the Control and in the Remifentanil groups)
was not significantly different in the 2 groups
(Table III).
Clinically acceptable conditions for pan endoscopy (primary endpoint) were observed in
77 (68%) of the Remifentanil patients and 67
(64%) of Control patients, a difference not significant (P=0.39). None of the 5 parameters of
Vol. 82 - No. 11
the score (cough, movement of limbs, position
of the vocal cords, movements of the vocal
cords, laryngoscopy) were significantly different between the groups (Figure 2).
The hemodynamic response to endoscopyrelated noxious stimulation was significantly
lower in the Remifentanil group as compared
to the Control group (Table III). A significantly
larger decrease in arterial blood pressure was
observed in the Remifentanil group in comparison to the Control group (Table III). The
incidence of respiratory adverse events or surgical complications was not different between
Minerva Anestesiologica
1143
BESCH
PROPOFOL-REMIFENTANIL ANESTHESIA FOR UPPER AIRWAY ENDOSCOPY
P=0.09
P=0.33
P=0.36
P=0.32
P=0.32
Figure 2.—Intergroup comparison of the 5 parameters of the score used to assess pan endoscopy conditions during general
anesthesia. No significant difference between groups. Pan endoscopy consists in a three steps endoscopy for assessment of
the upper aero digestive tract: a direct laryngoscopy, a rigid bronchoscopy and an esophagoscopy. Conditions were assessed
during the first 2 steps of the pan endoscopy: the direct laryngoscopy and the rigid bronchoscopy.
the 2 groups. No patient required emergency
endotracheal intubation or surgical tracheostomy during the study period.
Discussion
The result of this study showed that adjunction of remifentanil to a TCI of propofol, at a
dose that allow spontaneous breathing did not
improve the clinical conditions for the pan
endoscopy compared to the TCI of propofol
alone. Hemodynamic response to the noxious
stimulation of pan endoscopy was attenuated
by remifentanil and the incidence of moderate
or severe hypoxemia was not increased.
ENDOTANIL trial aimed to assess the combination of both TCI propofol and remifentanil in tubeless anesthesia with spontaneous
breathing in patients undergoing pan endoscopy. In our study, the incidence of clinically
acceptable conditions for pan endoscopy was
not increased by the addition of remifentanil.
A beneficial effect was expected as a dose-
1144
dependent improvement in the clinical conditions of procedures such as laryngoscopy or
endotracheal intubation has been described in
several papers.8, 23 Several hypotheses could
be raised to explain the lack of effect we observed in our study. First, the dose of remifentanil administrated to the patient may be
too low to add a beneficial effect to propofol
anesthesia. As a matter of fact, 1) the Ce of
remifentanil reported in the literature to produce excellent conditions for endotracheal
intubation, were higher than the Ce we used
in the ENDOTANIL trial,9-12, 24 and 2) the
continuous infusion of remifentanil was frequently preceded by an initial bolus dose.25-27
Other authors have reported that spontaneous
breathing could be maintained during fiberoptic intubation or flexible bronchoscopy by
using remifentanil alone at a mean effect-site
concentration as high as 2.4 ng.mL-1 and 2.5
ng.mL-1 respectively.28, 29 However, we think
that these higher doses of remifentanil could
not be used safely for tubeless management of
Minerva AnestesiologicaNovember 2016
PROPOFOL-REMIFENTANIL ANESTHESIA FOR UPPER AIRWAY ENDOSCOPYBESCH
general anesthesia for the pan endoscopy that
requires spontaneous breathing.30 Remifentanil has depressive respiratory effect that is
dose-dependent and the association of propofol and remifentanil increases the risk of respiratory depression in a synergistic manner.17, 30
Thus, we chose a remifentanil-Ce equal to 1.5
ng.mL-1, i.e. the median value reported by
Murdoch et al. to perform spontaneous breathing sedation for flexible fiber optic bronchoscopy.16 In our study, this Ce of remifentanil
allowed to obtain the same low rate of druginduced apnea and/or of facial mask ventilation requirement. We assume that half of the
patients included in the study of Murdoch et
al. maintained a spontaneous breathing with
higher doses of remifentanil and that using
higher dose of remifentanil could provide different results on both hemodynamics and quality of pan endoscopy.
Second, the relative role of the analgesic
and the hypnotic components of anesthesia
on the response to several noxious stimuli has
been studied graphically for several noxious
procedures, using a three-dimension response
surface modeling of the response/no response
to a standardized stimulation (1st dimension)
versus the corresponding remifentanil-Ce (2nd
dimension: analgesia axis) versus propofolCe (3rd dimension: hypnotic axis).11, 12, 31 Interestingly, Kern et al. reported that the morphologic feature of the response surface for
laryngoscopy was skewed toward the hypnotic axis and the difference in the probability of no response to laryngoscopy at a
concentration of propofol equal to 5 µg.mL-1
(mean value in our study) appeared very low
between patient receiving or not an infusion
of remifentanil.11
Third, we used the pharmacokinetic models
of Schnider and of Minto for the TCI of propofol and of remifentanil, respectively.18, 19 These
pharmacokinetic models have been commonly
used in clinical studies to assess the association
of propofol and remifentanil in many clinical
situations, including laryngoscopy and endotracheal intubation.31, 32 However, the models
have been validated in healthy volunteers,
and the environment of the operating room
Vol. 82 - No. 11
may have generated a stress-induced increase
in cardiac output that potentially may modify
drug distribution.33, 34
The beneficial effect of either low or high
doses of remifentanil on the cardiovascular responses observed during laryngoscopy, rigid
bronchoscopy or endotracheal intubation has
been previously described in elective surgical patients.8, 23, 35-39 In our study, the infusion
of remifentanil lowered the increase of the
heart rate during pan endoscopy without any
clinically relevant hypotension. Thus, the use
of opioids during pan endoscopy could have
a potential interest to prevent perioperative
myocardial ischemia in this population of
smokers at risk of undiagnosed coronary artery
disease.40, 41
As a continuous monitoring of capnography
was not performed during the ENDOTANIL
trial, the number of apneic events could not
be reliably measured. We have chosen to focus on the clinically relevant apneic episodes
that were responsible for hypoxemia. As none
of the patients included in the study were hypoxic at baseline, we assume that all episodes
of moderate or severe hypoxemia could potentially be related to drug-induced respiratory depressive effect. Considering the lack of
difference between the 2 groups, we suppose
that the moderate and severe hypoxic episodes
were mainly related to the propofol infusion
and to the presence of the endoscope in the
upper airway during rigid bronchoscopy that
could have been responsible for significant airway obstruction.
Paralyzing patients during rigid bronchoscopy is matter of debate. The anesthetic
protocol in the ENDOTANIL trial aimed to
maintain spontaneous breathing during direct
laryngoscopy and rigid bronchoscopy to provide adequate oxygen supply during the procedure. Therefore, neuromuscular blocking
agents were not used to avoid apnea and its
consequences.
Limitations of the study
In the absence of specific score developed
to assess the conditions for the pan endos-
Minerva Anestesiologica
1145
BESCH
PROPOFOL-REMIFENTANIL ANESTHESIA FOR UPPER AIRWAY ENDOSCOPY
copy, the pan endoscopy-score was adapted
from previously published tools that are based
on the good clinical research practice guidelines.21 We hypothesized that the stimulations
of the upper airway observed during the direct laryngoscopy or the rigid bronchoscopy
of the pan endoscopy were close to those induced by laryngoscopy and endotracheal intubation. However, we cannot exclude that
the absence of effect of remifentanil on the
conditions for the pan endoscopy can only be
explained by an inadequate assessment tool.
In this regard, pan endoscopy-score may be
considered too sensitive, as, although only
2/3 of the patients had clinically acceptable
condition as assessed by the pan endoscopyscore, the endoscopic exploration of the upper
airway and the biopsies during the procedure
could be completed in all patients of the study.
Nonetheless, movements of the limbs and/or
vocal cords could occur in patients with clinically acceptable conditions. The results of
our study could not support that therapeutic
intervention, such as laser treatment or injection into the vocal cords, could be safely performed by using the same anesthetic protocol.
Finally, the intra and inter observer variability
of the pan endoscopy-score has not been assessed and could have impact the results of
the study.
The same Ce of remifentanil was used in all
patients included in the Remifentanil group,
for the reasons explained above. A dose adapted to each patient, using recently developed
pain monitoring may have produced better
results.41
The increase in the effect-site concentration of propofol just before the beginning of
the pan endoscopy has been chosen arbitrarily
and could have contributed to an overdosing.
Considering that the same anesthetic protocol
has been applied in the 2 groups, we assume
that it has probably no impact on the results of
the study.
Our study was not designed to assess a reduction in intraoperative awareness by adding remifentanil to TCI of propofol. Even if
a higher rate of this event has been reported
in ENT surgery, it remains infrequent. Ran-
1146
domized controlled trials aimed to test this
hypothesis should include a very large number of patients to reach a sufficient statistical
power.
Conclusions
In this study, addition of remifentanil
to propofol TCI anesthesia, at a dose that
maintains spontaneous breathing did not improve clinical condition for laryngoscopy
and bronchoscopy during pan endoscopy,
although it blunted partially hemodynamic
response during the procedure. Further studies are required to assess whether titration of
remifentanil could improve condition for pan
endoscopy.
Key messages
—— The adjunction of low doses of remifentanil to target-controlled infusion of
propofol failed to significantly improve the
conditions of direct laryngoscopy and rigid
bronchoscopy during pan endoscopic assessment of the upper aero digestive tract.
—— The infusion of low doses of remifentanil during pan endoscopy did not increase
the risk for respiratory adverse events.
—— Remifentanil could lower the hemodynamic response to direct laryngoscopy
and rigid bronchoscopy.
References
 1.Bourgain JL, Desruennes E, Fischler M, Ravussin P.
Transtracheal high frequency jet ventilation for endoscopic airway surgery: A multicentre study. Br J Anaesth
2001;87:870-5.
  2. Jaquet Y, Monnier P, Van Melle G, Ravussin P, Spahn
DR, Chollet-Rivier M. Complications of different ventilation strategies in endoscopic laryngeal surgery: A 10year review. Anesthesiology 2006;104:52-9.
 3.Perrin G, Colt HG, Martin C, Mak MA, Dumon JF,
Gouin F. Safety of interventional rigid bronchoscopy using intravenous anesthesia and spontaneous assisted ventilation. A prospective study. Chest 1992;102:1526-30.
  4. Passot S, Servin F, Allary R, Pascal J, Prades JM, Auboyer C, et al. Target-controlled versus manually-controlled
infusion of propofol for direct laryngoscopy and bronchoscopy. Anesth Analg 2002;94:1212-6.
  5. Billard V, Moulla F, Bourgain JL, Megnigbeto A, Stan-
Minerva AnestesiologicaNovember 2016
PROPOFOL-REMIFENTANIL ANESTHESIA FOR UPPER AIRWAY ENDOSCOPYBESCH
ski DR. Hemodynamic response to induction and intubation. Propofol/fentanyl interaction. Anesthesiology
1994;81:1384-93.
  6. Kelly JS, Roy RC. Intraoperative awareness with propofol-oxygen total intravenous anesthesia for microlaryngeal surgery. Anesthesiology 1992;77:207-9.
  7. Kazama T, Ikeda K, Morita K. Reduction by fentanyl of
the cp50 values of propofol and hemodynamic responses
to various noxious stimuli. Anesthesiology 1997;87:21327.
  8.Thompson JP, Hall AP, Russell J, Cagney B, Rowbotham
DJ. Effect of remifentanil on the haemodynamic response
to orotracheal intubation. Br J Anaesth 1998;80:467-9.
  9. Bouillon TW, Bruhn J, Radulescu L, Andresen C, Shafer
TJ, Cohane C, et al. Pharmacodynamic interaction between propofol and remifentanil regarding hypnosis,
tolerance of laryngoscopy, bispectral index, and electroencephalographic approximate entropy. Anesthesiology
2004;100:1353-72.
10.Grant S, Noble S, Woods A, Murdoch J, Davidson A. Assessment of intubating conditions in adults after induction with propofol and varying doses of remifentanil. Br
J Anaesth 1998;81:540-3.
11. Kern SE, Xie G, White JL, Egan TD. A response surface
analysis of propofol-remifentanil pharmacodynamic interaction in volunteers. Anesthesiology 2004;100:137381.
12. Mertens MJ, Olofsen E, Engbers FH, Burm AG, Bovill
JG, Vuyk J. Propofol reduces perioperative remifentanil
requirements in a synergistic manner: Response surface
modeling of perioperative remifentanil-propofol interactions. Anesthesiology 2003;99:347-59.
13.Egan TD, Lemmens HJ, Fiset P, Hermann DJ, Muir
KT, Stanski DR, et al. The pharmacokinetics of the new
short-acting opioid remifentanil (gi87084b) in healthy
adult male volunteers. Anesthesiology 1993;79:881-92.
14. Westmoreland CL, Hoke JF, Sebel PS, Hug CC, Jr., Muir
KT. Pharmacokinetics of remifentanil (gi87084b) and its
major metabolite (gi90291) in patients undergoing elective inpatient surgery. Anesthesiology 1993;79:893-903.
15. Mastronardi P, Dellacasa P. Observational study on the
use of remifentanil in general anesthesia. Drug utilisation
research. Minerva Anestesiol 2004;70:605-16.
16. Murdoch JA, Hyde RA, Kenny GN. Target-controlled
remifentanil in combination with propofol for spontaneously breathing day-case patients. Anaesthesia
1999;54:1028-31.
17. Peacock JE, Luntley JB, O’Connor B, Reilly CS, Ogg
TW, Watson BJ, et al. Remifentanil in combination with
propofol for spontaneous ventilation anaesthesia. Br J
Anaesth 1998;80:509-11.
18. Minto CF, Schnider TW, Shafer SL. Pharmacokinetics
and pharmacodynamics of remifentanil. Ii. Model application. Anesthesiology 1997;86:24-33.
19.Schnider TW, Minto CF, Gambus PL, Andresen C,
Goodale DB, Shafer SL, et al. The influence of method
of administration and covariates on the pharmacokinetics of propofol in adult volunteers. Anesthesiology
1998;88:1170-82.
20.Lee TH, Marcantonio ER, Mangione CM, Thomas EJ,
Polanczyk CA, Cook EF, et al. Derivation and prospective validation of a simple index for prediction of
cardiac risk of major noncardiac surgery. Circulation
1999;100:1043-9.
21.Viby-Mogensen J, Engbaek J, Eriksson LI, Gramstad L,
Jensen E, Jensen FS, et al. Good clinical research practice
(gcrp) in pharmacodynamic studies of neuromuscular
blocking agents. Acta Anaesthesiol Scand 1996;40:5974.
22. Fleming TR. One-sample multiple testing procedure for
phase ii clinical trials. Biometrics 1982;38:143-51.
Vol. 82 - No. 11
23.Hall AP, Thompson JP, Leslie NA, Fox AJ, Kumar
N, Rowbotham DJ. Comparison of different doses of
remifentanil on the cardiovascular response to laryngoscopy and tracheal intubation. Br J Anaesth 2000;84:1002.
24. Bouvet L, Stoian A, Rimmele T, Allaouchiche B, Chassard D, Boselli E. Optimal remifentanil dosage for providing excellent intubating conditions when co-administered with a single standard dose of propofol. Anaesthesia
2009;64:719-26.
25.Stevens JB, Wheatley L. Tracheal intubation in ambulatory surgery patients: Using remifentanil and propofol
without muscle relaxants. Anesth Analg 1998;86:45-9.
26.Troy AM, Huthinson RC, Easy WR, Kenney GN. Tracheal intubating conditions using propofol and remifentanil
target-controlled infusions. Anaesthesia 2002;57:1204-7.
27. Klemola UM, Mennander S, Saarnivaara L. Tracheal intubation without the use of muscle relaxants: Remifentanil or alfentanil in combination with propofol. Acta
Anaesthesiol Scand 2000;44:465-9.
28.Lallo A, Billard V, Bourgain JL. A comparison of propofol and remifentanil target-controlled infusions to facilitate fiberoptic nasotracheal intubation. Anesth Analg
2009;108:852-7.
29.Chalumeau-Lemoine L, Stoclin A, Billard V, Laplanche
A, Raynard B, Blot F. Flexible fiberoptic bronchoscopy
and remifentanil target-controlled infusion in icu: A preliminary study. Intensive Care Med 2013;39:53-8.
30.Nieuwenhuijs DJ, Olofsen E, Romberg RR, Sarton
E, Ward D, Engbers F, et al. Response surface modeling of remifentanil-propofol interaction on cardiorespiratory control and bispectral index. Anesthesiology
2003;98:312-22.
31. Johnson KB, Syroid ND, Gupta DK, Manyam SC, Egan
TD, Huntington J, et al. An evaluation of remifentanil
propofol response surfaces for loss of responsiveness,
loss of response to surrogates of painful stimuli and laryngoscopy in patients undergoing elective surgery. Anesth Analg 2008;106:471-9.
32.Trabold F, Casetta M, Duranteau J, Albaladejo P, Mazoit
JX, Samii K, et al. Propofol and remifentanil for intubation without muscle relaxant: The effect of the order of
injection. Acta Anaesthesiol Scand 2004;48:35-9.
33.Avram MJ, Krejcie TC, Henthorn TK, Niemann CU.
Beta-adrenergic blockade affects initial drug distribution
due to decreased cardiac output and altered blood flow
distribution. J Pharmacol Exp Ther 2004;311:617-24.
34. Kurita T, Uraoka M, Jiang Q, Suzuki M, Morishima Y,
Morita K, et al. Influence of cardiac output on the pseudo-steady state remifentanil and propofol concentrations
in swine. Acta Anaesthesiol Scand 2013;57:754-60.
35.Guignard B, Menigaux C, Dupont X, Fletcher D, Chauvin M. The effect of remifentanil on the bispectral index
change and hemodynamic responses after orotracheal intubation. Anesth Analg 2000;90:161-7.
36.Prakash N, McLeod T, Gao Smith F. The effects of
remifentanil on haemodynamic stability during rigid
bronchoscopy. Anaesthesia 2001;56:576-80.
37.Albertin A, Casati A, Deni F, Danelli G, Comotti L, Grifoni F, et al. Clinical comparison of either small doses
of fentanyl or remifentanil for blunting cardiovascular
changes induced by tracheal intubation. Minerva Anestesiol 2000;66:691-6.
38.Iannuzzi E, Iannuzzi M, Cirillo V, Viola G, Parisi R, Cerulli A, et al. Peri-intubation cardiovascular response during low dose remifentanil or sufentanil administration in
association with propofol tci. A double blind comparison.
Minerva Anestesiol 2004;70:109-15.
39.Iannuzzi E, Iannuzzi M, Cirillo V, Viola G, Parisi R,
Chiefari M. Small doses of remifentanil and alfetanil in
continuous total intravenous anesthesia in major abdomi-
Minerva Anestesiologica
1147
BESCH
PROPOFOL-REMIFENTANIL ANESTHESIA FOR UPPER AIRWAY ENDOSCOPY
nal surgery. A double blind comparison. Minerva Anestesiol 2003;69:127-36.
40. Warltier DC, Pagel PS, Kersten JR. Approaches to the
prevention of perioperative myocardial ischemia. Anesthesiology 2000;92:253-9.
41. Boselli E, Bouvet L, Begou G, Torkmani S, Allaouchiche
B. Prediction of hemodynamic reactivity during total intravenous anesthesia for suspension laryngoscopy using
analgesia/nociception index (ani): A prospective observational study. Minerva Anestesiol 2015;81:288-97.
Authors’ contributions.—Guillaume Besch had substantial contributions to 1) design of the study, and acquisition and interpretation
of data; and 2) drafting the article manuscript. Angeline Chopard-Guillemin had substantial contributions to 1) conception and design
of the study, and acquisition and interpretation of data; and 2) drafting the article manuscript. �������������������������������������
Elisabeth Monnet���������������������
had substantial contributions to 1) methodological design, and analysis and interpretation of data; and 2) revising the manuscript critically for important
intellectual content. Arnaud Causeret had substantial contributions to 1) design of the study, acquisition and interpretation of data; and
2) revising the manuscript critically for important intellectual content. Amelie Jurine had substantial contributions to 1) acquisition
and interpretation of data; and 2) revising the manuscript critically for important intellectual content. Gerald Baudry had substantial
contributions to 1) acquisition and interpretation of data; and 2) revising the manuscript critically for important intellectual content.
Benjamin Lasry had substantial contributions to 1) acquisition and interpretation of data; and 2) revising the manuscript critically for
important intellectual content. Laurent Tavernier had substantial contributions to 1) conception and design of the study and interpretation of data; and 2) drafting the article. Emmanuel Samain had substantial contributions to 1) conception and design of the study and
interpretation of data; and 2) drafting the article. Sebastien Pili-Floury had substantial contributions to 1) conception of the study, and
analysis and interpretation of data; and 2) drafting the manuscript.
Funding.—The work was not funded by: National Institutes of Health (NIH), Howard Hughes Medical Institute (HHMI), Medical
Research Council (MRC), and/or Wellcome Trust. The work was not supported by official funding or a grant. The financial cost of
the study was supported by the Department of Anesthesiology and Intensive Care Medicine, 3 Bvd Alexander Fleming, University
Hospital of Besancon, F-25000 Besancon, France.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material
discussed in the manuscript.
Congresses.—Data have been presented at the French Society of Anesthesia and Intensive Care Medicine Annual Meeting in Paris,
France, on September 2014, and at the American Society of Anesthesiologists Annual Meeting in New Orleans, Louisiana, USA, on
October 2014.
Acknowledgements.—We would like to thank Dr Eviane Farah for her assistance in preparing the manuscript.
Article first published online: June 17, 2016. - Manuscript accepted: June 15, 2016. - Manuscript revised: May 23, 2016. - Manuscript
received: March 3, 2016.
1148
Minerva AnestesiologicaNovember 2016
© 2016 EDIZIONI MINERVA MEDICA
Online version at http://www.minervamedica.it
Minerva Anestesiologica 2016 November;82(11):1149-57
ORIGINAL ARTICLE
Diaphragmatic ultrasonography as an adjunct
predictor tool of weaning success in patients
with difficult and prolonged weaning
Aikaterini FLEVARI *, Michael LIGNOS, Dimitrios KONSTANTONIS,
Apostolos ARMAGANIDIS
Second University Critical Care Clinic, Attikon University Hospital, Athens, Greece
*Corresponding author: Aikaterini Flevari, 1 Rimini str., Chaidari, Athens, 12462, Greece. E-mail: [email protected]
A B STRACT
BACKGROUND: Diaphragmatic ultrasonography has been recently considered as a new weaning predictor method.
Previous studies checked diaphragmatic excursion and thickness during quiet or deep breathing in unselected populations
of critically ill patients. Our study aimed to investigate diaphragmatic excursion during quiet and unassisted breathing, in
comparison to standard predictor tools, such as Rapid Shallow Breathing Index (RSBI) and Maximal Inspiratory Pressure
(Pimax), in patients with difficult and/or prolonged weaning.
METHODS: Patients with difficult and/or prolonged weaning, who met the criteria for spontaneous breathing trial, were
assessed. The excursion of each hemidiaphragm (DEx) was evaluated by B-mode and M-mode ultrasonography while
patient was on quiet breathing and at supine position. RSBI and Pimax were simultaneously recorded and weaning outcome was recorded.
RESULTS: Thirteen male and fourteen female patients were included. DEx [median and interquartile range, mm] was 14
(8.5-22) for the right hemidiaphragm (RDEx) and 12 (7-23) for the left (LDEx). We found no difference in DEx between
sexes. Among the four weaning predictor tools, LDEx at a cut-off 10 mm was the best index to predict weaning success
(sensitivity 86%, specificity 85%, Negative Predictive Value 94%). The optimal cut-off values, as determined by the area
under the receiver operating characteristic curve, were 10 mm for RDEx, 7 mm for LDEx, 57 breaths/min/L for RSBI
and -20 cmH2O for Pimax.
CONCLUSIONS: Our results suggest that DEx threshold of 10 mm and 7 mm for right and left hemidiaphragms respectively could be used as adjunct tool in the predictive algorithm of weaning in difficult to wean patients.
(Cite this article as: Flevari A, Lignos M, Konstantonis D, Armaganidis A. Diaphragmatic ultrasonography as an adjunct
predictor tool of weaning success in patients with difficult and prolonged weaning. Minerva Anestesiol 2016;82:1149-57)
Key words: Ventilator weaning - Ultrasonography - Diaphragm.
A
lthough the process of discontinuing
(weaning) mechanical ventilation (MV)
remains one of the most challenging steps
throughout the period of MV, weaning predictor tools often fail to attain their role. Weaning
covers the entire process of liberating patients
from MV until the final extubation. In 2005
a Task Force of experts came to a consensus
Comment in p. 1132.
Vol. 82 - No. 11
regarding guidelines for “Weaning from Mechanical Ventilation”. Tobin therein defined
weaning as a six step process: treatment of
the underlying condition (stage 1), suspicion
that the patient is ready to wean (stage 2), assessment of the readiness to wean (stage 3),
spontaneous breathing trial (SBT) (stage 4),
extubation (stage 5) and assessment of probable reintubation (stage 6).1 In stage 3, apart
from clinical stability which implies resolution
Minerva Anestesiologica
1149
FLEVARI
DIAPHRAGMATIC ULTRASONOGRAPHY IN PATIENTS WITH WEANING
of the acute phase disease and cardiovascular,
metabolic and neuropsychiatric stability, a series of objective measurements are continuously monitored in order to assess weaning capability. Commonly used parameters include
Rapid Shallow Breathing Index, defined as the
ratio of respiratory frequency to tidal volume
(RSBI=f/VT), Maximum Inspiratory Pressure
(Pimax), pulmonary gas exchange (e.g. PaO2/
FiO2, PaCO2), Vital Capacity, Minute Ventilation and static Compliance.2-4 Among them,
RSBI has been acknowledged as the most important predictor of weaning failure with best
cut-off value that of 105 breaths/min/L.2, 5
On the other hand, weaning process is not
always straightforward and therefore Brochard
in the same consensus statement introduced a
classification of critically-ill patients based on
the difficulty and length of weaning process.1
Difficult weaning may in fact be due to different or mixed etiologies, the diagnosis of which
requires meticulous monitoring of all the aforementioned parameters. The commonly missed
diagnosis of diaphragmatic dysfunction, from
weakness to total paralysis, required until now
strenuous methods, such as electrical or magnetic phrenic nerve stimulation, or thoracic
fluoroscopy. During the last decade, portable
ultrasonography has become an invaluable
tool, since it is a simple, non-invasive and reproducible method of imaging, introducing a
new and less studied weaning predictor tool:
the diaphragmatic ultrasonography.5, 6 Several
researches have studied diaphragmatic motion,7-10 but none of them has been performed
in the intriguing sub-population of critically ill
patients with difficult and prolonged weaning.
Aim of this study was to test if ultrasonographic M-mode diaphragmatic excursion (DEx),
during regular, quiet and effortless breathing,
as compared to standard predictors, (RSBI and
Pimax), could be a successful extubation tool
in difficult to wean critically ill patients.
Materials and methods
Setting and ethical standards
This is a prospective cohort clinical trial,
conducted in a general university hospital
1150
adult Intensive Care Unit (ICU) from June to
December 2014, in accordance with the principals enunciated in the World Medical Association Declaration of Helsinki. All data were
part of routine treatment and their collection
involved no risk for the subjects, whose anonymity has been preserved. The study was approved by the hospital’s Ethics Committee (No
9-13-16/May/14), before patients’ recruitment.
For the aforementioned reasons the Committee
waived the need for obtaining informed consent.
Study population
Demographic characteristics were recorded:
sex, age, Body Mass Index (BMI), and Acute
Physiology and Chronic Health Evaluation
(APACHE II) Score on admission to ICU.
Figure 1 shows flow diagram of patient recruitment. As per study protocol, among all patients who underwent SBT, only patients with
difficult and prolonged weaning were eligible
for inclusion. According to scientific data and
definitions provided,1 70% of intubated critically ill patients can easily be extubated. Of
the remaining 30%, 25% experience “Difficult
Weaning” (DW-defined as failure of the first
SBT, but success within one week of the first
attempt), whereas 5% experience “Prolonged
Weaning” (PW-defined as repeated failed attempts within more than a week from the first
trial). In patients with weaning failure, mortality raises up to 25%.11 We therefore investigated patients which have had either more
than one episode of failed SBT but successful
extubation within a week (DW) or repeated
failed episodes of SBT in more than a week
time period (PW).1, 5 Patients were included in
the study if they fulfilled all of the following
criteria: age≥18 years; Glascow Coma Scale
≥14; hemodynamic stability; use of oxygen
fraction ≤0.5; pressure support <12 cmH2O
and positive end expiratory pressure (PEEP)
<6 cmH2O; respiratory rate (RR) <35/min,
tidal volume Vt >5 mL/kg Ideal Body Weight
(IBW). The use of sedatives and/or analgesics
was not contraindicated as long as the patient
remained calm, cooperative and easily arous-
Minerva Anestesiologica
November 2016
DIAPHRAGMATIC ULTRASONOGRAPHY IN PATIENTS WITH WEANINGFLEVARI
Figure 1.—Flow chart of patients for the study population.
able (Ramsey score 2-3). Patients were excluded from the study if they had at least one of the
following: lack of cooperation, incapability to
initiate an inspiratory effort, intubation due to
elective surgery or upper airway obstruction,
ventilation via tracheostomy tube, hemodynamic instability.
Weaning protocol
In our clinic, a weaning protocol is being
implemented: Patients if sedated receive assistcontrol ventilation. After gradual withdrawal
of sedatives/ and or analgesics and if patients
are clinically stable, (RR <25 bpm, PaO2 >80
mmHg on FiO2 <60%, PEEP <7 cmH2O, minute ventilation VE<10 L/min), we implement
pressure-support ventilation (PSV). PSV is
initially set at 15-20 cmH2O (above PEEP) and
attempts to reduce this level of support by 2 to
4 cmH2O are made at least twice daily according to patient’s adaptation and recovery. When
pressure-support and PEEP level are less than
10 and 6 cmH2O respectively and provided
Vol. 82 - No. 11
that the patient is alert, calm, co-operative and
has adequate cough, we withdraw the ventilator and allow him to breathe spontaneously
through a T-tube circuit for up to two hours.
In rare cases, patients may require small doses
of analgosedation in order to be calm and cooperative (Ramsey score 2-3, absence of delirium), preferably propofol or remifentanyl.
Per study protocol, we used remifentanil (at a
dosage of ≤0.02 μgr/kg/min) which thanks to
its favorable pharmacokinetics, allows rapid
elimination from blood plasma. Remifentanil
infusion was interrupted 10 minutes before any
test performance, so as not to interfere with our
measurements. Criteria for SBT discontinuation were any of the following: change in mental status, discomfort or diaphoresis, RR>35/
min, hemodynamic instability (systolic blood
pressure >180 mmHg, or <90 mmHg, heart
rate >130 or increase or 20% increase). The
decision to discontinue SBT or to extubate was
made by ICU-doctors, according to their clinical and parametrical judgement and so was the
decision to implement NIV or re-intubate. All
Minerva Anestesiologica
1151
FLEVARI
DIAPHRAGMATIC ULTRASONOGRAPHY IN PATIENTS WITH WEANING
SBTs were classified into a success group (SG)
or a failure group (FG). Weaning failure was
the trial after which the patient required mechanical ventilation (invasive or non-invasive)
within forty-eight hours.
Measurement of standard weaning parameters, namely RSBI and Pimax
Pimax has long been used in ICU setting to
assess inspiratory muscle strength. The maneuver consists of a maximum inspiratory effort
via a unidirectional valve, which permits exhalation while inhalation is temporarily blocked,
allowing patients after 3-4 breathing efforts to
perform maximal effort at a lung volume approaching residual. Its measurement is noninvasive and simple to perform, but requires
considerable degree of cooperation.12 Pimax
was measured while the patient was on SBT
on a T-piece trial. On the other hand, RSBI
was measured with the patient on a continuous
positive airway pressure of 5 cmH2O, an alternative to SBT method.13 This was provided
by setting the ventilator at PSV mode and adjusting pressure support level at 0 cmH2O and
PEEP at 5 cmH2O, enabling us to directly view
tidal volume and respiratory frequency.
Ultrasonography
Diaphragmatic evaluation was performed
with a multifrequency 5-MHz vector transducer placed immediately below costal margins.
We used right midclavicular or anterior axillary line to visualize right DEx and left middle-posterior axillary line to visualize left DEx
with liver and spleen serving as acoustic windows respectively. We started our exploration
by B-mode sonography, a two dimensional
image, essential to select the exploration line.
Movement was recorded by M-mode sonography and diaphragmatic inspiratory amplitude
(excursion) was measured along the selected
line. The first endpoint was placed at the foot
of the inspiration slope on the diaphragmatic
echoic line and the second endpoint at the apex
of the slope. A detailed description of the sonographic technique is provided by Matamis
1152
et al.6 In our study all patients were lying in
0-10° recumbent position and all measurements were performed during quiet breathing
by two operators (among the three first writers). When patients were accustomed to our
technique, the DEx of five consecutive normal
respiratory cycles were recorded and the mean
value was calculated. The mean value of the
recordings made by the two operators was finally recorded. Apart from DEx, ultrasonography enabled us to visualize other pathologic
entities, such as atelectasis, pleural effusion or
diaphragmatic palsy, that would probably interpret weaning difficulty. Physicians responsible for weaning decisions were blinded to
sonographic measurements.
Statistical analysis
All groups of data were tested for normality
by Kolomogorov-Smirnov Test. If data were
normally distributed, Student’s t-test tested the
difference between the means; if not, non-parametric Mann-Whitney U Test was performed.
Therefore descriptive statistics are presented as
mean (±standard deviation) and median values
(25-75th interquartile ranges) respectively. Sensitivity, specificity, positive and negative predictive value (PPV-NPV), positive and negative
likelihood ratio (PLR-NLR) were calculated for
all predictive variables, namely RDEx, LDEx,
RSBI and Pimax. Sensitivity and specificity are
dependent on the cut-off value used, above or
below which a test is positive. For RDEx, we decided to test two cut-off points at 10 mm and 15
mm, whereas for LDEx we tested only 10 mm
excursion during quiet breathing. This threshold
is based on previous studies performed in normal subjects, which found that the lower normal
limit of DE for both hemidiaphragms was 9 mm
in women and 10mm in men.7 This cut-off limit
has been adopted by other researchers as well.8
For RSBI and Pimax we adopted standard cutoff values, namely <100 breaths/min/L and
≤-30 cmH2O respectively.2, 14 Receiver Operative Characteristic (ROC) curves were plotted
in order to examine the diagnostic accuracy of
each variable and choose the optimal cut-off
value for our study, based on our measurements.
Minerva Anestesiologica
November 2016
DIAPHRAGMATIC ULTRASONOGRAPHY IN PATIENTS WITH WEANINGFLEVARI
Table I.—Characteristics and data of study population.
No of
patients
% of
patients
No of patients
Male/Female
Etiology of ICU submission
Cardiovasuclar surgery
Non-surgical cardiac disease
Trauma
Non-traumatic post-surgical disease
Sepsis - ARDS - MOF
Miscellaneous
27
13/14
100
48/52
8
3
7
2
5
2
30
11
26
7
19
7
Data
Median
value
95% CI
65
7
54-73
5-10
3
3-4
Characteristics
Age (years)
Duration of MV before study
inclusion (days)
No of failed trials before study
inclusion
ICU: Intensive Care Unit; CABG: coronary artery bypass graft;
COPD: chronic obstructive pulmonary disease; MOF: multiple organ failure; MV: mechanical ventilation.
For all analyses, statistical significance was set
at 5% (a=0.05) and testing was 2-tail. Statistical
analysis was carried out by MedCalc Statistical Software version 16.4.3 (MedCalc Software
bvba, Ostend, Belgium; 2016).
Results
In Table I, we present study population and
primary reason for ICU admission. Among patients, eleven were admitted after elective or
emergent surgical procedure and sixteen were
admitted due to medical conditions. We also
provide the duration of mechanical ventilation
and the number of failed trials before study inclusion. Among the 27 patients enrolled, only
two required remifentanil infusion, which as
per study protocol was discontinued ten minutes before study performance. Trials were
for both patients successful. Table II tabulates
anthropometric characteristics, APACHE II
score and results referring to the study population (A), as well as to male and female patients (B1-B2). Apart from results appearing
in Tables, we state that we found no difference
between RDEx and LDEx (P=0.6) in the overall study group, as well as between genders. In
SG we found no difference between right and
left DEx (P=0.86) and the same was found for
FG as well (P=0.11). Findings from SG and
FG are presented in Table II and Figure 2. As
we can see, all four weaning predictors had
statistically significant difference. Table IV
tabulates sensitivity, specificity, PLR, NLR,
PPV and NPV of all weaning predictors, standard (f/Vt and Pimax) and sonographic. We
present data for two different RDEx cut-offs,
namely 10 mm and 15 mm. The cut-off value
of DEx for predicting successful extubation
was found 10mm for the right hemidiaphragm
and 7 mm for the left by receiver operating
characteristic (ROC) curve analysis (Table V).
For the other two traditional predictors, f/Vt
and Pimax, the cut-off values were found at
57 br/min/L and -20 cmH2O respectively. At
these cut-off limits, all four tests seem to have
Table II.— Anthropometric characteristics (Age/BMI), Illness Severity Score (APACHE 2) and predictors of weaning
success in A. Overall study group, B. Male and Female subpopulations (in median and 25-75th percentile values).
Variables
Age (years)
BMI (kg/m2)
APACHE 2
Days ventilated on inclusion
Weaning predictors
f/Vt (RSBI) (breaths/min/L)
Pimax (cmH2O)
RDEx (mm)
LDEx (mm)
A.
All patients (100%)
B1.
Male patients (48%)
B2.
Female patients (52%)
P value
Median
25th- 75th
Median
25th- 75th
Median
25-75th
65
22
20
7
53-75
20-28
15-25
5-10
54
22
18
7
46-67
20-24
15-20
5-11
69
25
23
6
61-75
20-30
19-25
4-10
0.05*
0.13
0.08
0.36
70
25
14
12
53-100
20-34
8.5-22
7-23
70
25
15
18
46-80
19-31
12-30
10-25
74
28
13
10
57-119
21-35
7-20
5-19
0.19
0.9
0.11
0.24
BMI: Body Mass Index; APACHE II Score: Acute Physiology and Chronic Health Evaluation; f/Vt (RSBI): ratio of respiratory frequency to
tidal volume, representing Rapid Shallow Breathing Index; Pimax: Maximal Inspiratory Pressure; RDEx: Right Diaphragmatic Excursion;
LDEx: Left Diaphragmatic Excursion. Numbers with an asterisk (*) indicate difference of a statistical significance.
Vol. 82 - No. 11
Minerva Anestesiologica
1153
FLEVARI
DIAPHRAGMATIC ULTRASONOGRAPHY IN PATIENTS WITH WEANING
A
B
Figure 2.—Right (A) and Left (B) Diaphragmatic Excursions (RDEx, LDEx) in Success and Failure groups.
Table III.—Comparison of the characteristics (Age / BMI), APACHE II Score and predictors of weaning between
Success and Failure groups (in median and 25-75th percentile values).
SG
Variables
Median
Age (years)
BMI (kg/m2)
APACHE II
Weaning predictors
f/Vt (RSBI) (breaths/min/L)
Pimax (cmH2O)
RDEx (mm)
LDEx (mm)
FG
25th-
75th
Median
25-75th
P value
Median
60
22
21
48-74
20-28
16-25
70
28
20
67-77
20-32
13-25
0.06
0.45
0.64
64
30
19
19
45-85
25-35
13-27
11-25
116
15
7
5
68-120
11-24
5-10
2-6
0.03*
0.01*
0.004*
0.0004*
SG: success group; FG: failure group; BMI: Body Mass Index; APACHE II Score: acute physiology and chronic health evaluation; f/Vt
(RSBI): ratio of respiratory frequency to tidal volume, representing Rapid Shallow Breathing Index; Pimax: maximal inspiratory pressure;
RDEx: right diaphragmatic excursion; LDEx: left diaphragmatic excursion. Numbers with an asterisk (*) indicate difference of a statistical
significance.
Table IV.—Sensitivity, specificity, AUC, positive/negative likelihood ratio, positive/negative predictive value of
standard predictors of weaning failure, and of right and left diaphragmatic excursion during quiet breathing.
RDEx was tested against two critical values.
Prognostic variables
RDEx <10 mm
RDEx <15 mm
LDEx <10 mm
F/Vt >100 breaths/min/L
Pimax <│-30 cmH2O│
Sensitivity (%)
Specificity (%)
AUC
Positive Likelihood Ratio
Negative
Likelihood
Ratio
Positive
Predictive
Value (%)
71
100
86
71
85
85
65
85
85
65
0.78
0.83
0.85
0.78
0.75
4.8
2.85
5.7
4.8
2.4
0.3
0
0.16
0.33
0.2
63
50
67
63
46
Negative
Predictive
Value (%)
89
100
94
89
93
RDEx: right diaphragmatic excursion; LDEx: left diaphragmatic excursion; f/Vt: ratio of respiratory frequency to tidal volume; Pimax: maximal inspiratory pressure; AUC: area under the curve.
comparable diagnostic accuracy as this is expressed by the area under the curve (AUC) of
the ROC curve analysis. Pairwise Comparison
of ROC curves showed no statistical difference between AUC.
1154
Discussion
To our knowledge, this is the first study to
investigate sonographic diaphragmatic excursion as a weaning tool in patients with
Minerva Anestesiologica
November 2016
DIAPHRAGMATIC ULTRASONOGRAPHY IN PATIENTS WITH WEANINGFLEVARI
Table V.—Receiver operating characteristic (ROC) curves for all weaning predictor variables and optimal cut-off
points for excluding weaning failure.
Prognostic variables
RDEx (mm)
LDEx (mm)
F/Vt (breaths/min/L)
PImax (cmH2O)
Optimal
cut-off point
AUC
≤10
<7
≥57
<│-20│
0.87
0.95
0.77
0.81
95% CI
Lower
Upper
0.69
0.79
0.57
0.61
0.97
0.99
0.91
0.93
P value
0.001*
0.0001*
0.006*
0.01*
Sensitivity (%) Specificity (%)
86
86
100
71
85
95
60
85
AUC: area under the curve; CI: Confidence Interval; RDEx: right diaphragmatic excursion; LDEx: Left diaphragmatic excursion; F/Vt: ratio
of respiratory frequency to tidal volume; PImax: maximal inspiratory pressure. Numbers with an asterisk (*) indicate difference of a statistical
significance.
prolonged and difficult weaning, in whom
diaphragmatic contractile dysfunction is well
described.15 We found that the excursion of
both hemidiaphragms (RDEx-LDEx) performs well in terms of diagnostic accuracy. We
also found that a cut-off limit of 10mm for the
right hemidiaphragm and of 7 mm for the left
can predict weaning success. The much lower
left hemidiaphragmatic threshold (7 mm) was
rather surprising. This could be probably attributed either to differences in probe positioning, or to the study population per se. As far
as probe positioning is concerned, we handled
the probe on the middle-posterior left axillary
line, so as the angle of ultrasound beam to
reach perpendicularly left diaphragm, whereas
we placed the probe on the midclavicular or
the anterior right axillary line in order to visualize right diaphragm. That is to say, we probably faced LDEx closer than RDEx. As far
the study population is concerned, in patients
with difficult weaning, who may have diminished DEx, this could be counterbalanced by
an extra effort exerted by accessory inspiratory
muscles, finally making weaning feasible. We
found no gender difference in diaphragmatic
excursion unlike the findings by previous research.7
Ultrasonographic indices as predictors of successful extubation are rather new in the scientific
literature.5, 6, 8-10, 16 Our findings suggest that if
we want to find all failed trials, RDEx <15 mm
and LDEx<10 mm should be included (high
sensitivity). On the other hand, if we want to correctly diagnose success (high NPV), we should
rely on RDEx>15 mm, LDEx>10 mm and MIP
above │-30 cmH2O│. Moreover based on optimal cut-off points, if LDEx<7 mm, RDEx≤ 10
Vol. 82 - No. 11
mm, and MIP<│-20 cmH2O│ weaning is highly probable to fail (high specificity).
Among the researchers investigating sonographic indices as weaning predictor
rules,8, 9, 16, 17 DiNino et al. 10 as well as Ferrari
et al. 16 measured the alteration of diaphragmatic thickness on deep breathing, whereas Kim et
al.,8 as well as Jiang et al.9 measured diaphragmatic excursion. All of them concluded that
sonographic indices of the diaphragm could be
used as weaning predictors, since they perform
well, if not better, in comparison to standard
predictors. Both studies investigating DEx included all patients eligible for a weaning trial
and not only those with prolonged or/and difficult weaning. Jiang et al. investigated patients,
already scheduled to be extubated, along the
right anterior axillary and left posterior axillary line, to find the same cut-off threshold (11
mm) for both hemidiaphragms. As per study
protocol, Jiang et al. measured real time liver
and spleen displacements and not DEx, and
as they stated, this may integrate the contraction of all inspiratory muscles both diaphragm
and accessory inspiratory muscle. This study
has a different concept of measuring, which
along with the difference in probe positioning
could explain the discrepancy between their
LDEx threshold and ours. Kim et al. also included all ICU patients along the right anterior
axillary and the left midaxillary line. Patients
were classified according to a DEx cut-off of
10mm in two groups, with and without diaphragmatic dysfunction (DD), to find that 29%
of the patients had indeed DD with an excursion of a median value of 3.0 mm for the right
diaphragm and 2.6 mm for the left. Their cutoff limits for predicting primary (within 48 h)
Minerva Anestesiologica
1155
FLEVARI
DIAPHRAGMATIC ULTRASONOGRAPHY IN PATIENTS WITH WEANING
weaning failure are higher than ours (14mm
for RDEx and 12mm for LDEx), probably reflecting differences in methodology in terms
of study population (all patients, instead of
patients with difficult/prolonged weaning) and
probe positioning.
We therefore believe that patients with difficult and/or prolonged weaning, whose diaphragm has some degree of weakness due to
prolonged mechanical ventilation,18 recruit all
inspiratory muscles, making final extubation
feasible. The concept that in this particular
sub-population of patients, a lower diaphragmatic threshold would be acceptable, warrants
further validation in future studies.
As far as standard weaning predictors are
concerned, we examined the performance
of RSBI and Pimax. RSBI (with a threshold
value of 105 breaths/min/L) was first introduced in 1991 by Yang and Tobin 2 as the best
among many weaning predictors. Since then it
has been re-examined in many studies and its
value, though doubted,5, 14 remains high.19-21
These studies demonstrate a wide range of
threshold values (60-105) and positive likelihood ratios (0.84-4.67),14 indicating that the
index has great heterogeneity in posttest probability, probably due to differences in methodology and variations in pretest probability
(Bayes’ theorem).4, 14, 21 In our study, RSBI
was performed at CPAP 5 cmH2O, in order to
be able to directly visualize tidal volume and
respiratory frequency. Adding this minimal
support is probably needed to compensate for
the resistance and the dead space of the circuitry of the ventilator.22-24 Besides this method is
considered as an alternative of SBT by many
studies.13, 25 Nevertheless we are aware of the
fact that differences in methodology would
alter the cut-off threshold of the index, since
even minimal support would affect work of
breathing.25, 26 In our study, the RSBI cut-off
value was 57 breaths/min/L, which lies at the
lowest point of the aforementioned range. The
fact that RSBI was recorded while the patient
was on CPAP 5 cmH2O could probably explain
the lower cut-off value, as has previously been
stated.27 In terms of diagnostic accuracy (AUC,
sensitivity and specificity (0.77, 100% and
1156
60% respectively), its low specificity diminishes its value. Nevertheless our results were
comparable to those of other researches.2, 9
Finally, the standard index that determines
inspiratory muscle strength is Pimax.12 Pimax
is simple to perform but difficult to interpret. First of all, the lower cut-off limit (not
uniformly accepted) is largely dependent on
patient’s voluntary effort, which is very hard
to obtain in critically ill patients. Moreover
Pimax reflects the integrated pressure exerted
by all inspiratory muscles and not by the diaphragm itself. We found the threshold of │-20
cmH2O│, as a good predictor of extubation
failure, which lies in the range of │20-30 cmH2O│ described in literature.1
Limitations of the study
There are some limitations for this study.
First, interobserver and intraobserver variabilty were not tested. We accepted the fact
that past studies have given good results in
this field.7, 28 Second, we did not measure diaphragmatic thickness, since this requires both
patient’s cooperation and an ultrasound probe
of a higher frequency (>7.5 MHz). Finally the
cut-off value of LDEx warrants validation in a
new and larger cohort of patients.
Conclusions
Our results suggest that diaphragmatic excursion, as assessed by M-mode ultrasonography, can reliably estimate weaning failure in
patients with difficult and prolonged weaning.
In fact the sonographic thresholds of ≤10 mm
for the right hemidiaphragm and <7 mm for the
left could be used as adjunct tools in predicting
failed extubation in the subpopulation of critically ill patients with difficult and prolonged
weaning.
Key messages
—— Difficult and prolonged weaning is
associated with increased mortality.
—— Standard weaning parameters do not
Minerva Anestesiologica
November 2016
DIAPHRAGMATIC ULTRASONOGRAPHY IN PATIENTS WITH WEANINGFLEVARI
suffice to predict weaning failure in patients
with difficult and prolonged weaning.
—— We suggest that if LDEx<7 mm and
RDEx≤10 mm weaning trial is highly probable to fail.
References
 1. Boles JM, Bion J, Connors A, Herridge M, Marsh B,
Melot C, et al. Weaning from mechanical ventilation. Eur
Resp J 2007;29:1033-56.
  2. Yang K, Tobin M. A prospective study of indexes predicting the outcome of trials of weaning from mechanical
ventilation. N Engl J Med 1991;324:1445-50.
  3. Nemer SN, Barbas CS, Caldeira JB, Cárias TC, Santos
RG, Almeida LC, et al. A new integrative weaning index
of discontinuation from mechanical ventilation. Critical
Care 2009;13:R152.
  4.Conti G, Montini L, Pennisi MA, Cavaliere F, Arcangeli
A, Bocci MG, et al. A prospective, blinded evaluation of
indexes proposed to predict weaning from mechanical
ventilation. Intens Care Med 2004;30:830-6.
 5.Sellarés J, Ferrer M, Torres A. Predictors of weaning after acute respiratory failure. Minerva Anestesiol
2012;78:1046-53.
  6. Matamis D, Soilemezi E, Tsagourias M, Akoumianaki E,
Dimassi S, Boroli F, et al. Sonographic evaluation of the
diaphragm in critically ill patients. Technique and clinical
applications. Intens Care Med 2013;39:801-10.
 7. Boussuges A, Gole Y, Blanc P. Diaphragmatic motion
studied by m-mode ultrasonography: methods, reproducibility, and normal values. Chest 2009;135:391-400.
  8. Kim WY, Suh HJ, Hong SB, Koh Y, Lim CM. Diaphragm
dysfunction assessed by ultrasonography: influence on
weaning from mechanical ventilation. Critical Care Med
2011;39:2627-30.
  9. Jiang JR, Tsai TH, Jerng JS, Yu CJ, Wu HD, Yang PC.
Ultrasonographic evaluation of liver/spleen movements
and extubation outcome. Chest 2004;126:179-85.
10. DiNino E, Gartman EJ, Sethi JM, McCool D. Diaphragm
ultrasound as a predictor of succesful extubation from
mechanical ventilation. Thorax 2014;69:423-7.
11.Esteban A, Alía I, Tobin MJ, Gil A, Gordo F, Vallverdú I,
et al. Effect of Spontaneous Breathing Trial Duration on
Outcome of Attempts to Discontinue Mechanical Ventilation. Am J Respir Crit Care Med 1999;159:512-8.
12. ATS/ERS Statement on respiratory muscle testing. Am J
Respir Crit Care Med 2002;166:518-624.
13.Ely WE, Baker AM, Dunagan DP, Burke HL, Smith AC,
Kelly PT, et al. Effect of the duration of mechanical ventilation of identifying patients capable of breathing spontaneously. N Engl J Med 1996;335:1864-9.
14. MacIntyre NR, Cook DJ, Ely EW, Epstein SK, Fink JB,
Heffner JE, et al. Evidence-Based Guidelines for Weaning and Discontinuing Ventilatory Support. A Collective
Task Force Facilitated by the American College of Chest
Physicians; the American Association for Respiratory
Care ; and the American College. Chest 2001;120:37595.
15. Powers SK, Kavazis AN, Levine S. Prolonged mechanical ventilation alters diaphragmatic structure and function. Crit Care Med 2009;37(10 Suppl):S347-353.
16. Ferrari G, Filippi G, De Elia F, Panero F, Volpicelli G,
Aprà F. Diaphragm ultrasound as a new index of discontinuation from mechanical ventilation. Crit Ultrasound J
2014;6:8.
17.Umbrello, M, Formenti, P, Longhi, D, Galimberti, A,
Piva, I, Pezzi, A, et al. Diaphragm ultrasound as indicator
of respiratory effort in critically ill patients undergoing
assisted mechanical ventilation : a pilot clinical study.
Crit Care 2015;19:161.
18.Sassoon, C.S.H. Ventilator-associated Diaphragmatic
Dysfunction. Am J Respir Crit Care Med 2002;166:10178.
19. de Souza LC, Lugon JR. The rapid shallow breathing index as a predictor of successful mechanical ventilation
weaning : clinical utility when calculated from ventilator
data. J Bras Pneumol 2015 41:530-5.
20. Epstein SK. Etiology of Extubation Failure and the Predictive Value of the Rapid Shallow Breathing Index. Am
J Respir Crit Care Med 1995;152:545-9.
21.Tobin MJ, Jubran A. Variable performance of weaningpredictor tests : role of Bayes’ theorem and spectrum and
test-referral bias. Intens Care Med 2006;32:2002-12.
22.Straus C, Louis B, Isabey D, Lemaire F, Harf A, Brochard
L. Contribution of the Endotracheal Tube and the Upper
Airway to Breathing Workload. Am J Respir Crit Care
Med 1998;157:23-30.
23. Nathan SD, Ishaaya AM, Koerner SK, Belman MJ. Prediction of minimal Pressure Support during weaning
from mechanical ventilation. Chest 1993;103:1215-9.
24. Pelosi P, Solca M, Ravagnan I, Tubiolo D, Ferrario L,
Gattinoni L. Effects of heat and moisture exchangers on
minute ventilation, ventilatory drive, and work of breathing during pressure-support ventilation in acute respiratory failure. Crit Care Med 1996;24:1184-8.
25. Desai NR, Myers L, Simeone F. Comparison of 3 different methods used to measure the rapid shallow breathing
index. J Crit Care 2012;27:418.e1-418.e6.
26.Tobin MJ. Extubation and the myth of “minimal ventilator settings”. Am J Respir Crit Care Med 2012;185:34950.
27. El-Khatib MF, Zeineldine SM, Jamaleddine GW. Effect
of pressure support ventilation and positive end expiratory pressure on the rapid shallow breathing index in intensive care unit patients. Intens Care Med 2008;34:50510.
28.Houston JG, Angus RM, Cowan MD, McMillan NC,
Thomson NC. Ultrasound assessment of normal hemidiaphragmatic movement: relation to inspiratory volume.
Thorax 1994;49:500-3.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material
discussed in the manuscript.
Acknowledgements.—We would like to express our sincere thanks to all patients who patiently suffered our services. We would also
like to gratefully acknowledge the critical reviews of anonymous reviewers made on an earlier, as well as on the present version of
the manuscript.
Article first published online: July 12, 2016. - Manuscript accepted: July 7, 2016. - Manuscript revised: June 16, 2016. - Manuscript
received: March 5, 2016.
Vol. 82 - No. 11
Minerva Anestesiologica
1157
© 2016 EDIZIONI MINERVA MEDICA
Online version at http://www.minervamedica.it
Minerva Anestesiologica 2016 November;82(11):1158-69
ORIGINAL ARTICLE
Non-invasive hemodynamic optimization in
major abdominal surgery: a feasibility study
Ole BROCH 1 *, Arne CARSTENS 2, Matthias GRUENEWALD 1,
Edith NISCHELSKY 3, Lukas VELLMER 3, Berthold BEIN 4, Heiko ASELMANN 5,
Markus STEINFATH 1, Jochen RENNER 1
1Department
of Anesthesiology and Intensive Care Medicine, University Hospital Schleswig-Holstein, Campus
Kiel, Germany; 2Department of Anesthesiology and Intensive Care Medicine, Imland Hospital, Rendsburg,
Germany; 3Christian-Albrechts-University Kiel, Schleswig-Holstein, Germany; 4Department of Anesthesiology and
Intensive Care Medicine, Asklepios Hospital St. Georg, Hamburg, Germany; 5Department of General and Thoracic
Surgery, University Hospital Schleswig-Holstein, Campus Kiel, Germany
*Corresponding author: Ole Broch, University Hospital Schleswig-Holstein, Campus Kiel, Department of Anesthesiology and Intensive Care Medicine, Schwanenweg 21, D-24105 Kiel, Germany. E-mail: [email protected]
ABSTRACT
BACKGROUND: Today, most of the pre-emptive hemodynamic optimization algorithms are based on variables associated with invasive techniques like arterial cannulation. The non-invasive Nexfin™ technology is able to estimate continuous
Cardiac Index (CI) and pulse pressure variation (PPV). However, the efficiency of an early goal directed therapy (EGDT)
algorithm based on non-invasive variables has to be proven. The aim of our study was to investigate the feasibility of a
non-invasive driven EGDT protocol and its impact on patient’s outcome.
METHODS: Seventy-nine patients (ASA II-III) undergoing elective major abdominal surgery were randomized to either
study group (SG, N.=39) or control group (CG, N.=40). The SG was treated according to an algorithm based on noninvasive CI and PPV, whereas the CG received standard of care. Postoperative complications up to 28 days and length of
hospital stay (LOS) in both groups were recorded.
RESULTS: There was no significant difference between the groups regarding demographics, hemodynamic variables,
preoperative risk scores and duration of surgery. The total amount of complications was higher in the CG (SG 94 vs. CG
132 complications, P=0.22) without reaching statistical significance. LOS revealed no difference between both groups
(SG, 9 [7-15] vs. CG, 9 [7-15.25] days, P=0.82). We have seen no impact of the non-invasive optimization protocol with
respect to postoperative mortality.
CONCLUSIONS: In this patient collective, we could demonstrate the feasibility of a non-invasive approach for hemodynamic optimization. However, EGDT based on non-invasive variables was not able to significantly improve outcome.
(Cite this article as: Broch O, Carstens A, Gruenewald M, Nischelsky E, Vellmer L, Bein B, et al. Non-invasive hemodynamic
optimization in major abdominal surgery: a feasibility study. Minerva Anestesiol 2016;82:1158-69)
Key words: Anesthesia - Surgical procedures - Feasibility studies.
M
ajor surgical procedures include the
threat of hypoperfusion and possible
mismatch in oxygen delivery and demand.
Several studies have demonstrated that a preemptive hemodynamic optimization by early
goal directed fluid and catecholamine therapy
(EDGT) was associated with beneficial impact
on both morbidity and mortality.1 However,
1158
these beneficial effects are more pronounced
in critically ill patients.2 It must be noted, that
“basic” perioperative monitoring, i.e. electrocardiogram, oxygen saturation and non-invasive or invasive blood pressure measurements,
is not able to sufficiently detect hypovolemia
and ongoing organ hypoperfusion.3 Flow related variables like cardiac index (CI), Stroke
Minerva AnestesiologicaNovember 2016
NON-INVASIVE HEMODYNAMIC OPTIMIZATION IN MAJOR ABDOMINAL SURGERY
Volume Index (SVI) and their surrogates,
stroke volume variation (SVV) and pulse pressure variation (PPV) are claimed to be advantageous for guiding fluid and catecholamine
therapy.4 Therefore, less invasive monitoring
systems based on transpulmonary thermodilution or uncalibrated pulse contour analysis
have gained increasing interest by gradually
minimizing the reluctance of physicians to use
advanced hemodynamic monitoring.5 Over the
last years several studies focused on EGDT
protocols, yielding a strong proof of feasibility of different algorithms and, moreover providing evidence that EDGT improves patient’s
outcome in various clinical scenarios.1 The
Nexfin™ monitoring system (previous manufacturer: BMEYE, Amsterdam, The Netherlands, current manufacturer: Clearsight™,
Edwards Lifesciences LLC, Irvine, CA, USA)
offers a completely non-invasive approach
providing continuous beat-to-beat CI, SVI,
SVV, PPV and arterial pressure by using an
inflatable finger cuff. Several clinical investigations have demonstrated sufficient trending
ability of the Nexfin™ technology.6-8 At present
there is no evidence, whether an EGDT protocol based on a non-invasive monitoring technology in general is feasible in a wide range of
patients on the one hand, and if it is able to reduce postoperative complications on the other
hand. Therefore, the aim of our study was in
first line to verify the feasibility of a non-invasive EDGT protocol in patients undergoing
major open abdominal surgery and in second
line to compare their postoperative outcome
with a standard of care approach.
Materials and methods
The study was conducted as a single center, prospective randomized trial from March
2014 to October 2015 at the University Hospital Schleswig-Holstein, Campus Kiel, in
accordance with the Helsinki declaration.
After approval from institutional ethics committee (Ethikkomission UKSH Kiel, AZ
B260/11, Christian Albrecht University Kiel,
Schwanenweg 20, D 24105 Kiel), written informed consent for participation in the study
Vol. 82 - No. 11
BROCH
was obtained preoperatively from all patients.
The trial was registered on ClinicalTrials.gov
(NCT02559141).
Eighty-six ASA II and III patients undergoing elective major abdominal surgery were
included in the study. Seven patients have to
be excluded due to various reasons. Patients
undergoing major abdominal procedures with
an estimated duration ≥120 minutes and a high
transfusion probability due to an anticipated
blood loss ≥1000 mL were screened for elegibility. Flow chart with enrolment, intervention allocation, follow-up and data analysis is
represented in Figure 1. The Physiological and
Operative Severity Score for the enumeration
of Mortality and Morbidity (POSSUM) was
used for risk evaluation. Exclusion criteria
were patients less than 18 years old, ASA I or
IV classification, heart rhythm disorders, advanced peripheral artery occlusive disease, arteriovenous shunts concerning upper extremities and laparoscopic abdominal procedures.
Patients received midazolam 0.1 mg/kg
orally 30 minutes before induction of anesthesia. A five lead electrocardiogram (ECG;
S/5, GE Healthcare, Helsinki, Finland) and
peripheral oxygen saturation (SpO2) were established as routine monitoring. As suggested
by recent literature, non-invasive blood pressure was obtained on both upper arms to exclude a difference ≥10 mmHg.9 In case of a
non-invasive blood pressure difference ≥10
mmHg, the Nexfin™ monitoring system was
positioned on the ipsilateral side to the arterial
line. Prior to induction of general anesthesia
patients received a peripheral venous access
and a radial arterial line (Arrow International,
Inc. Reading, PA, USA) under local anesthesia. After installation of the pneumatic finger
cuff, the Nexfin™ monitoring system was feeded with patient specific data as advised by the
manufacturer. The underlying physiological
background and functionality of this monitoring system is described elsewhere.10-13 Adjustment of the transducer was followed by zeroing at the height of the midaxillary position.
The non-invasive device was established at the
side contralateral to the indwelling radial catheter, followed by attachment of a temperature
Minerva Anestesiologica
1159
BROCH
NON-INVASIVE HEMODYNAMIC OPTIMIZATION IN MAJOR ABDOMINAL SURGERY
(N.=115)
(N.=9)
(N.=13)
(N.=7)
(N.=86)
(N.=3)
(N.=3)
(N.=1)
(N.=79)
(N.=40)
(N.=39)
(N.=40)
(N.=39)
(N.=1)
(N.=40)
(N.=38)
Analyzed (N.=38)
Analyzed (N.=40)
(N.=0)
(N.=0)
Figure 1.—Enrolment, allocation, follow-up and analysis of participants through the trail.
probe (Siemens AG, S1138, Erlangen, Germany) for measuring skin temperature at the
fingertip. Temperature was also measured with
a nasopharyngeal probe. All variables were automatically indexed to body surface area. After
induction of general anesthesia with 0.5 µg/
kg sufentanil and 1.5 mg/kg propofol, orotracheal intubation was facilitated with 0.6 mg/
1160
kg rocuronium. Anesthesia was maintained in
accordance with our standard operating procedure using sevoflurane and sufentanil. Thereafter, a central venous catheter was introduced
in the right internal jugular vein. Patients were
ventilated in a pressure controlled modus with
an oxygen/air mixture using a tidal volume of
8 mL/kg ideal body weight and a positive end-
Minerva AnestesiologicaNovember 2016
NON-INVASIVE HEMODYNAMIC OPTIMIZATION IN MAJOR ABDOMINAL SURGERY
expiratory pressure of 5-10 cmH2O. Respiratory rate was adjusted to achieve normocapnia
(pCO2 35-40 mmHg). Ventilation parameters
were documented every 60 minutes after intubation. During surgery, the patients received a
continuous infusion of crystalloid solution (6
mL/kg/ h).
Data collection
Patients were randomized either into study
group (SG, N.=39) or control group (CG,
N.=40). Each patient received a central venous catheter and an arterial line. The Nexfin™
monitoring system was also established in
each group but was strictly covered and inaccessible for the attending anesthetist in the CG.
Another investigator who was not involved in
management of anesthesia and hemodynamic
therapy collected non-invasive hemodynamic
variables estimated by the Nexfin™ monitoring
system in the same manner in the CG as well
in the SG. In accordance to the study algorithm, hemodynamic variables were recorded
at baseline, after induction of anesthesia and
subsequently every 15 minutes till the end of
surgery. Central venous and arterial blood gas
samples were drawn hourly after baseline. Patients in the CG were treated with a basic algorithm, targeting an invasive estimated mean
arterial pressure (MAP) of ≥65 mmHg, a CVP
≤12 mmHg and a consistent hemoglobin level
≥8 g/dL. To preserve a MAP ≥65 mmHg, vasoactive agents like bolus injection of theodrenaline/cafedrine or continuous infusion of
norepinephrine (0.01-0.08 µg/kg/min) were
administered. Due to the absence of advanced
hemodynamic monitoring and no possibility to
control the therapeutic effect, dobutamine was
not used in the CG.
The SG was treated on the basis of an EDGT
algorithm which among others, intended to
maintain preload (Figure 2). In presence of a
PPV >10%, volume substitution was initiated
using 500 mL of crystalloids and/or colloids
as long as CI increased ≥2.5 L/min/m2. Maintenance of CI ≥2.5 L/min/m2 and MAP ≥65
mmHg was achieved by using norepinephrine
(0.01-0.08 µg/kg/min), theodrenaline/cafe-
Vol. 82 - No. 11
BROCH
drine and dobutamine (5-10 µg/kg/min). Catecholamine administration, blood loss, urine
output, fluid input was recorded every 15 minutes up to the end of surgery in the SG as well
as in the CG. Time between end of surgery and
extubation was also recorded.
Hemodynamic monitoring was continued
in all patients in the post-anesthesia care unit
(PACU) or intensive care unit (ICU) for at
least 180 minutes. Again, data collection was
performed every 15 minutes in all patients
but there was no defined algorithm postoperatively. Central venous and arterial blood gas
samples were drawn hourly after arrival in the
PACU or ICU. Data on catecholamine use, estimated blood loss, urine output and fluid intake were obtained 24 hours postoperatively.
Postoperative complications were divided into
different categories: general (bleeding, reoperation), infection (wound, pneumonia, urinary
tract, abdominal), cardiovascular (pulmonary
edema, myocardial infarct, stroke, new onset
atrial fibrillation), respiratory (respiratory support >24 h, pleural effusion) and abdominal
(ascites, prolonged paralytic ileus, need of parenteral feeding, ongoing nausea and vomiting)
complications. Length of hospital stay (LOS)
and complications were followed up 28 days
after surgery and documented by daily visits and aligning data from the patient charts.
Documentation of LOS was carried out by the
electronic chart.
Statistical analysis
Sample size calculation was based on our
own hospital registry, which revealed an incidence of postoperative complications of 45%
for a similar historical patient collective (unpublished data). For a decrease in morbidity
from 45 to 30%, we calculated a sample size of
35 patients for each group with a 0.05 difference
(two-sided) and a power of 80%. Considering a
drop out of patients of 10-20% due to different
reasons, we decided to include 43 patients in
each group. Statistical analysis was performed
by using commercially available statistics software (GraphPad Prism 5, GraphPad Software
Inc., San Diego, CA, USA; SigmaPlot 13.0
Minerva Anestesiologica
1161
BROCH
NON-INVASIVE HEMODYNAMIC OPTIMIZATION IN MAJOR ABDOMINAL SURGERY
Figure 2.—Fluid and catecholamine management in the study group based on hemodynamic variables.
CI: Cardiac Index; PPV: pulse pressure variation; MAP: mean arterial pressure.
for Windows version 7, Systat Software, Inc.,
San Jose, CA). Parametric data were analyzed
with the student`s two-tailed t-test, whereas the
Mann-Whitney U-test was used for analysis of
non-parametric data. Categorical data were
compared using χ2 and Fisher’s Exact Test and
P<0.05 was considered statistical significant.
Data are listed in mean, standard deviation
(SD) or median, interquartile range (IQR) as
appropriate. Comparison of data was restricted
to the mean operation procedure time at 225
minutes chosen as the average.
1162
Results
A total of 86 patients were randomized. Seven patients had to be excluded from the study
for various reasons. Finally, data of 79 patients, 58 males and 21 females, were included
and allocated; 39 patients in the SG and 40
patients in the CG. One patient in the SG has
to be excluded from further analysis due to ongoing heart rhythm disorders. No patient has
to be excluded due to persistent deficiencies
regarding the non-invasive monitoring tool.
Minerva AnestesiologicaNovember 2016
NON-INVASIVE HEMODYNAMIC OPTIMIZATION IN MAJOR ABDOMINAL SURGERY
We observed no side effects by the long lasting continuous non-invasive monitoring nor
signal loss or any monitoring related complication. There were no significant differences
between both groups regarding demographics,
initial hemodynamic and laboratory variables
and preoperative risk scores. Preoperative demographic, hemodynamic and laboratory data
are represented in Table I.
MAP did not differ significantly at any time
from baseline up to 225 min. CI was comparable between both groups from baseline until 75
min, differences in CI between groups reached
significance from 90 min until the end of data
collection. At this time, CI was significant
higher in the SG with the exception of CI at 105
min achieving no significance (Figure 1). PPV
revealed significant differences at various time
points from baseline to end of surgery (Figure
2). For lactate concentration analysis showed
no significant difference between groups at any
time from baseline. At baseline, lactate concen-
BROCH
tration was 0.88 (0.30) in the SG vs. 0.85 (0.23)
mmol/L in the CG (P=0.51). After 60 min, lactate concentration was 0.78 (0.34) in the SG
vs. 0.72 (0.18) mmol/L in the CG (P=0.46), at
120 min 0.84 (0.28) in the SG vs. 0.82 (0.26)
mmol/L in the CG (P=0.42), at the end of operation 1.21 (1.39) in the SG vs. 0.97 (0.48)
mmol/L in the CG (P=0.31) and 180 min after
operation 1.11 (0.39) in the SG vs. 0.99 (0.43)
mmol/L in the CG (P=0.33). During operation,
temperature of the fingertip was 35.5 (0.7) °C
in the study group and 35.7 (0.4) °C (P=0.39)
in the CG. Values were given as mean (SD).
Regarding the amount of administered fluid,
packed red blood cells, urine output or blood
loss intra- and postoperatively, no significant
differences have been revealed. Fresh frozen
plasma and noradrenaline have been administered more frequently in the CG. Dobutamine
was preserved intraoperatively only for the SG.
None of the patients in both groups received
inotropes after the end of surgery (Table II).
Table I.—Characteristics, demographic data, preoperative haemodynamic and laboratory parameters.
Age (yr)
Sex m/w
Height (cm)
ABW (kg)
ASA II / III
POSSUM (physiological score)
POSSUM (operative score)
NIBP SP (mmHg)
NIBP DP (mmHg)
NIBP MAP (mmHg)
CI (L/min/m2)
HR (min-1)
SV (mL)
O2 Sat (%)
PPV (%)
SVV (%)
Eadyn
WBC (103/µL)
Lactate (mmol/L)
Hb (g/dL)
Epidural analgesia
PONV history
Study group
N.=39
Control group
N.=40
P value
67 (9)
28 / 11
174 (9.6)
79.7 (19.3)
28 / 11
18.0 (3.7)
17.4 (4.1)
135 (27)
72 (12)
95 (17)
2.8 (0.7)
71 (11)
79 (23)
95 (2)
10 (6)
8 (5)
1.25 (0.13)
7.78 (0.48)
0.88 (0.30)
12.9 (2.0)
6
1
65 (13)
30 / 10
176 (8.1)
80.8 (14.3)
28 / 12
17.5 (3.1)
18.0 (4.5)
138 (23)
75 (9)
98 (13)
2.9 (0.9)
72 (13)
79 (21)
96 (5)
11 (6)
9 (5)
1.22 (0.17)
7.42 (0.44)
0.85 (0.23)
13.0 (1.9)
6
2
=0.40
=0.49
=0.33
=0.78
n.a.
=0.52
=0.53
=0.58
=0.17
=0.40
=0.65
=0.56
=0.95
=0.52
=0.43
=0.66
=0.59
=0.58
=0.56
=0.73
n.a.
n.a.
Data comparison performed by Student´s t-test, Mann Whitney U-Test, χ2-test or Fisher´s Exact Test for detection of preoperative group differences. Values are given as mean (SD). ABW: actual body weight; ASA: physical status classification by the American Society of Anesthesiologists; POSSUM: physiological and operative severity score for the enumeration of mortality and morbidity; NIBP: non-invasive blood
pressure; SP: systolic pressure; DP: diastolic pressure; MAP: mean arterial pressure; CI: cardiac index; HR: heart rate; SV: stroke volume;
O2Sat: oxygene saturation; PPV: pulse pressure variation; SVV: stroke volume variation; Eadyn: arterial elastance; WBC: white blood cell
count; Hb: hemoglobin; PONV: postoperative nausea and vomiting; n.a.: not assessed.
Vol. 82 - No. 11
Minerva Anestesiologica
1163
BROCH
NON-INVASIVE HEMODYNAMIC OPTIMIZATION IN MAJOR ABDOMINAL SURGERY
Table II.—Intraoperative fluids: vasopressors and inotropes.
Crystalloids (mL)
Colloids (mL)
PRBC (mL)
FFP (mL)
Blood loss (mL)
Urine output (mL)
Fluid Balance (mL)
Noradrenaline (μg)
Theo/Caf (mL)
Dobutamine (µg)
Study group
N.=39
Control group
N.=40
P value
3000 (2000-4000)
500 (250-1000)
0 (0-0)
(N.=14)
0 (0-0)
(N.=0)
800 (300-1500)
300 (200-700)
2300 (1500-3300)
0 (0-186)
(N.=15)
0.5 (0-1.5)
(N.=25)
11 (0-50)
(N.=22)
3000 (2250-3875)
250 (0-1000)
0 (0-0)
(N.=8)
0 (0-0)
(N.=2)
675 (400-1150)
300 (150-637)
2350 (1500-3400)
60.6 (0-232)
(N.=21)
0.8 (0-2)
(N.=24)
0 (0-0)
(N.=0)
=0.99
=0.06
=0.22
=0.59
=0.44
=0.58
=0.95
=0.18
=0.40
=0.57
Data comparison by Student´s t-test (and nonparametric tests) to detect intraoperative group differences. PRBC: packed red blood cells; FFP:
fresh frozen plasma; Theo/Caf: theodrenaline/cafedrine; *P<0.05 (vs. control group); Data are median (interquartile range [IQR]).
Table III.—Length of stay in hospital and postoperative complications.
Hospital stay (days) total
ASA II
Hospital stay (days) total
Hospital stay (days)
ASA II no complication
Hospital stay (days)
ASA II ≥1 complication
ASA III
Hospital stay (days) total
Hospital stay (days)
ASA III no complication
Hospital stay (days)
ASA III ≥1 complication
Total amount of complications
ASA II
Complications
ASA III
Complications
Study group
N.=39
Control group
N.=40
P value
9 (7-15)
9 (7-15.25)
=0.82
8 (7-12.5)
8 (7-11.75)
=0.71
7 (6-9.5)
7 (6-8)
=0.50
9 (7-21)
10 (8-13)
=0.65
13 (7-18)
14.5 (7.5-33.5)
6.5 (3.75-8.5)
16 (13-46)
94
6.5 (6-7)
17.5 (9.75-39.75)
132
=0.64
=0.81
=0.92
=0.22
61 (2.1 per patient)
71 (2.5 per patient)
=0.43
33 (3.0 per patient)
61 (5.1 per patient)
=0.17
Data analysis by Mann Whitney U-Test and Fisher´s Exact Test for detection of group differences. Data are median (interquartile range [IQR]).
Duration of operation was comparable in
both groups (SG: 227 (93) vs. CG: 230 (80)
min, P=0.92) and time period from end of surgery until extubation differed not significantly
(SG: 15 (7) vs. CG: 17 (8), P=0.15). Data were
mean (SD). The total amount of complications
was higher in the control group (SG: 94 vs.
CG: 132, P=0.22). This difference in complication rate was also obtained for ASA II (SG:
61 vs. CG: 71, P=0.43) and ASA III patients
1164
(SG: 33 vs. CG: 61, P=0.17) without reaching statistical significance (Table III). Detailed
analysis of complication rate in relation to the
ASA classification is represented in Figure 3.
In the SG, less patients developed ≥1 complication (SG, N.=29 vs. CG, N.=34, P=0.27).
The number of complications for ASA III patients (SG, N.=7 vs. CG, N.=10, P=0.45) differed not significantly. Patients without any
complication were higher in the SG (SG,
Minerva AnestesiologicaNovember 2016
NON-INVASIVE HEMODYNAMIC OPTIMIZATION IN MAJOR ABDOMINAL SURGERY
BROCH
Figure 3.—Intraoperative variance of CI. Hemodynamic variables for study group (grey dots) and control group (balck
dots). *P<0.05 (vs. control group).
Figure 4.—Intraoperative chronological sequence of pulse pressure variation (PPV). Hemodynamic variables for study group
(grey dots) and control group (black dots). *P<0.05 (vs. control group).
N.=10 versus CG, N.=6, P=0.27). Again, no
significant differences for ASA II and ASA III
patients were detected (Figure 3).
LOS in both groups was comparable (SG,
9 (7-15) vs. CG, 9 (7-15.25) days, P=0.82).
Looking at ≥1 complications and LOS, ASA
III subgroup (SG, 16 (13-46) vs. CG, 17.5
(9.75–39.75) days, P=0.92) and ASA II subgroup (SG, 9 (7-21) vs. CG, 10 (8-13) days,
P=0.48) showed no significant difference.
Data were presented as median (IQR). With
Vol. 82 - No. 11
respect to the patient collective without any
complication and LOS, ASA III subgroup and
ASA II subgroup again revealed no significant
differences between SG and CG (Table III).
Discussion
The present single center study demonstrated that EGDT based on the non-invasive Nexfin™ technology was feasible in ASA II and
III patients undergoing major open abdominal
Minerva Anestesiologica
1165
BROCH
NON-INVASIVE HEMODYNAMIC OPTIMIZATION IN MAJOR ABDOMINAL SURGERY
Figure 5.—Number of postoperative complications in the study group and the control group listed according to the ASA classification. *P<0.05 (vs. control group).
surgery. Furthermore, preemptive hemodynamic optimization using an algorithm based
on continuous non-invasive CI and PPV calculations was able to reduce the rate of postoperative complications, however without reaching
statistical significance. Patients at higher risk
(ASA III) benefited more from EGDT than
patients at lower risk (ASA II). There was no
difference in hospital LOS between the SG and
the CG.
To our best knowledge, this is the first study
investigating the feasibility of a completely
non-invasive approach based on the Nexfin™
technology for EGDT in a larger patient collective. The concept of early, preemptive algorithm-based titration of fluids and inotropes has
been shown to reduce postoperative morbidity
and mortality in different clinical scenarios.4
However, it is recognized that early implementation of GDT has a higher beneficial impact
in patients exhibiting more severe comorbidities.2 Therefore, definition of high risk surgical procedures and identification of appropriate patients are indispensable prerequisites for
performing efficient EGDT. Several years ago,
Shoemaker and colleagues defined criteria for
selecting surgery and patients associated with
1166
higher risk.14 These criteria have been slightly
modified over the years and firmly established
in daily clinical routine.4, 15 In this context, we
obtained a reduced total amount of complications in ASA III classified patients in the SG
compared to the ASA III patients of the CG,
however without reaching statistical significance. Moreover, we observed a higher rate of
patients in the SG without any complication.
Again, these observations were statistically
insignificant. A possible explanation of these
results could be the composition of patients
with less comorbidities in the present study.
In this context, we observed considerable
more ASA II patients in the SG as well in the
CG. This is of crucial clinical importance as
a recent investigation could demonstrate that
EGDT performed in low risk patients could
be associated with adverse effects on primary
outcome.16 It must be noted, however, that we
did not observe detrimental effects regarding
our EGDT protocol in both patient collectives. On the contrary, by applying an EGDT
algorithm based on non-invasive monitoring,
we recognized a slight trend towards less or
zero complications in ASA III patients of the
SG compared to ASA III patients of the CG.
Minerva AnestesiologicaNovember 2016
NON-INVASIVE HEMODYNAMIC OPTIMIZATION IN MAJOR ABDOMINAL SURGERY
Furthermore, by avoiding complications, ASA
III patients revealed a similar hospital LOS
compared to ASA II patients exhibiting less
comorbidities. This potential benefit for the
ASA III subgroup was even more pronounced
considering that occurrence of ≥1 complication increased the hospital LOS more than
three times in this patient population.
Another explanation of our results could be
related to the experience of the attending anesthesiologist. Anesthesia for both patient collectives was performed by a consultant with
long-standing experience. Although patients
in the CG received a hemodynamic therapy
simply based on invasive variables like MAP
and CVP we observed similar intraoperative
lactate levels in both groups which revealed
no statistic significant differences. As we used
dynamic variables like PPV, some confounders should be noted. Several studies identified heart rhythm disturbances, respiratory
rate17, spontaneous breathing efforts18, presence of intraabdominal hypertension19 and/
or right ventricular failure20 as clinical entities with a potentially relevant impact on the
predictive accuracy of dynamic variables. To
evaluate whether differences in arterial tone
between both groups have influenced fluid responsiveness, dynamic arterial elastance was
determined as suggested by recent literature.21
However, we observed no significant differences in arterial elastance between the study
and the control group. Since none of these
confounders has been identified in our study,
one main requirement regarding the application of these variables was fulfilled. Moreover,
several studies in recent times have demonstrated that Nexfin-based non-invasive assessment of PPV allows an acceptable prediction
of fluid responsiveness.22 In contrast, another
study reported that SVV and PPV given by the
non-invasive Nexfin device were not able to
predict fluid responsiveness.23 Both studies
were performed in patients after cardiac surgery. However, two recently published studies could demonstrate that dynamic variables
calculated by non-invasive finger cuff were
able to adequately predict fluid responsiveness.24, 25 Consequently, and this was our as-
Vol. 82 - No. 11
BROCH
sumption, in combination with the literature
proved sufficient CI-trending ability of the
Nexfin technology,26 an EGDT protocol based
on these non-invasive variables per se should
be able to primarily avoid occult hypovolemia
and secondarily maintain stable hemodynamics. Although our results are not significant
enough to claim that non-invasive driven
EDGT protocols per se have the potential to
reduce postoperative outcome in this patient
population, we have proven the feasibility.
Obviously, the next reasonable step, based on
our presented results, would be an appropriate
powered multicenter study. We therefore performed a power analysis with a two sided 0.05
difference and a power of 80% to achieve a
reduction in postoperative morbidity of 10%,
yielding a sample size of 193 patients for each
group.
Limitations of the study
Some limitations of our study should be
noted. We investigated patients without heart
rhythm disturbances and advanced peripheral artery occlusive disease or arteriovenous
shunts. Keeping in mind, that pulse contour
analysis could be impaired in presence of
these conditions 27-29 we cannot extrapolate
our results to patients exhibiting different comorbidities. With respect to our EGDT protocol, recent literature suggested that a threshold
value of PPV ≥13% should be better used for
volume therapy control.30 Further research is
needed to more clearly define the indications
and limitations of EGDT protocols based on
non-invasive haemodynamic variables, especially in patients exhibiting more severe comorbidities.
Conclusions
In conclusion, our results suggest that an
EGDT protocol based on non-invasive haemodynamic variables in patients undergoing open
abdominal surgery is feasible. Although postoperative complications could be reduced, the
EGDT protocol was not able to reach statistical significance and had no impact on LOS.
Minerva Anestesiologica
1167
BROCH
NON-INVASIVE HEMODYNAMIC OPTIMIZATION IN MAJOR ABDOMINAL SURGERY
Key messages
—— Implementation of an EGDT protocol based on non-invasive haemodynamic
variables in patients undergoing open abdominal surgery was feasible.
—— The non-invasive driven EGDT algorithm induced a slight trend towards less
or zero complications in ASA III patients.
—— By avoidance of complications, ASA
III patients revealed a similar hospital LOS
compared to ASA II patients.
—— Although postoperative complications could be reduced, the EGDT protocol
was not able to reach statistical significance
and had no impact on LOS.
References
1. Hamilton MA, Cecconi M, Rhodes A. A systematic review and meta-analysis on the use of preemptive hemodynamic intervention to improve postoperative outcomes
in moderate and high-risk surgical patients. Anesth Analg
2011;112:1392-402.
  2. Kern JW, Shoemaker WC. Meta-analysis of hemodynamic optimization in high-risk patients. Crit Care Med
2002;30:1686-92.
  3. Mythen MG, Webb AR. Intra-operative gut mucosal hypoperfusion is associated with increased post-operative
complications and cost. Intensive Care Med 1994;20:99104.
  4. Gurgel ST, do Nascimento P, Jr. Maintaining tissue
perfusion in high-risk surgical patients: a systematic review of randomized clinical trials. Anesth Analg
2011;112:1384-91.
  5. McGuinness S, Parke R. Using cardiac output monitoring
to guide perioperative haemodynamic therapy. Curr Opin
Crit Care 2015;21:364-8.
  6. Bogert LW, van Lieshout JJ. Non-invasive pulsatile arterial pressure and stroke volume changes from the human
finger. Exp Physiol 2005;90:437-46.
  7. Stover JF, Stocker R, Lenherr R, Neff TA, Cottini SR,
Zoller B, et al. Noninvasive cardiac output and blood
pressure monitoring cannot replace an invasive monitoring system in critically ill patients. BMC Anesthesiol
2009;9:6.
  8. Broch O, Renner J, Gruenewald M, Meybohm P, Schottler J, Caliebe A, et al. A comparison of the Nexfin(R)
and transcardiopulmonary thermodilution to estimate
cardiac output during coronary artery surgery. Anesthesia
2012;67:377-83.
  9. Clark CE, Taylor RS, Shore AC, Ukoumunne OC, Campbell JL. Association of a difference in systolic blood
pressure between arms with vascular disease and mortality: a systematic review and meta-analysis. Lancet
2012;379:905-14.
10. Eeftinck Schattenkerk DW, van Lieshout JJ, van den
Meiracker AH, van den Meiracker AH, Wesseling KR,
Blanc S, et al. Nexfin noninvasive continuous blood pressure validated against Riva-Rocci/Korotkoff. Am J Hypertens 2009;22:378-83.
1168
11. Penaz J, Voigt A, Teichmann W. Contribution to the continuous indirect blood pressure measurement. Z Gesamte
Inn Med 1976;31:1030-3.
12. Wesseling KH, Jansen JR, Settels JJ, Schreuder JJ.
Computation of aortic flow from pressure in humans
using a nonlinear, three-element model. J Appl Physiol
1993;74:2566-73.
13. Bogert LW, Wesseling KH, Schraa O, Van Lieshout EJ,
de Mol BA, van Goudoever J, et al. Pulse contour cardiac
output derived from non-invasive arterial pressure in cardiovascular disease. Anaesthesia 2010;65:1119-25.
14. Shoemaker WC, Wo CC, Thangathurai D, Velmahos G,
Belzberg H, Asensio JA, et al. Hemodynamic patterns of
survivors and nonsurvivors during high risk elective surgical operations. World J Surg 1999;23:1264-70.
15. Lees N, Hamilton M, Rhodes A. Clinical review: Goaldirected therapy in high risk surgical patients. Crit Care
2009;13:231.
16. Challand C, Struthers R, Sneyd JR, Erasmus PD, Mellor N, Hosie KB, et al. Randomized controlled trial of
intraoperative goal-directed fluid therapy in aerobically
fit and unfit patients having major colorectal surgery. Br J
Anaesth 2012;108:53-62.
17. De Backer D, Taccone FS, Holsten R, Ibrahimi F, Vincent
JL. Influence of respiratory rate on stroke volume variation in mechanically ventilated patients. Anesthesiology
2009;110:1092-7.
18. Heenen S, De Backer D, Vincent JL. How can the response to volume expansion in patients with spontaneous respiratory movements be predicted? Crit Care
2006;10:R102.
19. Renner J, Gruenewald M, Quaden R, Hanss R, Meybohm
P, Steinfath M, et al. Influence of increased intra-abdominal pressure on fluid responsiveness predicted by pulse
pressure variation and stroke volume variation in a porcine model. Crit Care Med 2009;37:650-8.
20. Mahjoub Y, Pila C, Friggeri A, Zogheib E, Lobjoie E,
Tinturier F, et al. Assessing fluid responsiveness in critically ill patients: False-positive pulse pressure variation
is detected by Doppler echocardiographic evaluation of
the right ventricle. Crit Care Med 2009;37:2570-5.
21. Monge Garcia MI, Gil Cano A, Gracia Romero M. Dynamic arterial elastance to predict arterial pressure response to volume loading in preload-dependent patients.
Crit Care 2011;15:R15.
22. Lansdorp B, Ouweneel D, de Keijzer A, van der Hoeven
JG, Lemson J, Pickkers P. Non-invasive measurement of
pulse pressure variation and systolic pressure variation
using a finger cuff corresponds with intra-arterial measurement. Br J Anaesth 2011;107:540-5.
23. Fischer MO, Coucoravas J, Truong J, Zhu L, Gérard JL,
Hanouz JL, et al. Assessment of changes in cardiac index
and fluid responsiveness: a comparison of Nexfin and
transpulmonary thermodilution. Acta Anaesthesiol Scand
2013;57:704-12.
24. Stens J, Oeben J, Van Dusseldorp AA, Boer C. Noninvasive measurements of pulse pressure variation and
stroke volume variation in anesthetized patients using
the Nexfin blood pressure monitor. J Clin Monit Comput
2016;30:587-94.
25. Vos JJ, Poterman M, Salm PP, Van Amsterdam K, Struys
MM, Scheeren TW, et al. Noninvasive pulse pressure
variation and stroke volume variation to predict fluid responsiveness at multiple thresholds: a prospective observational study. Can J Anesth 2015;62:1153-60.
26. Ameloot K, Palmers PJ, Malbrain ML. The accuracy of
noninvasive cardiac output and pressure measurements
with finger cuff: a concise review. Curr Opin Crit Care
2015;21:232-9.
27. Maj G, Monaco F, Landoni G, Barile L, Nicolotti D, Pieri
M, et al. Cardiac index assessment by the pressure re-
Minerva AnestesiologicaNovember 2016
NON-INVASIVE HEMODYNAMIC OPTIMIZATION IN MAJOR ABDOMINAL SURGERY
cording analytic method in unstable patients with atrial
fibrillation. J Cardiothorac Vasc Anesth 2011;25:476-80.
28. Barile L, Landoni G, Pieri M, Ruggeri L, Maj G, Nigro
Neto C, et al. Cardiac index assessment by the pressure
recording analytic method in critically ill unstable patients after cardiac surgery. J Cardiothorac Vasc Anesth
2013;27:1108-13.
29. Talts J, Raamat R, Jagomagi K. Influence of pulse pres-
BROCH
sure variation on the results of local arterial compliance
measurement: A computer simulation study. Comput Biol
Med 2009;39:707-12.
30. Cannesson M, Le Manach Y, Hofer CK, Goarin JP, Lehot
JJ, Vallet B, et al. Assessing the diagnostic accuracy of
pulse pressure variations for the prediction of fluid responsiveness: a “gray zone” approach. Anesthesiology
2011;115:231-41.
Authors’ contributions.—Broch O and Carstens A equally contributed to this manuscript.
Funding.—The present study was supported by departmental funding only.
Conflicts of interest.—Jochen Renner has received honoraria from Edwards Lifesciences (Irvine, CA, USA) for giving lectures.
Berthold Bein is a member of the medical advisory board of Pulsion Medical Systems (Munich, Germany) and has received honoraria for consulting and giving lectures. Matthias Gruenewald has received honoraria from GE Healthcare (Munich, Germany) and
Pulsion Medical Systems (Munich, Germany) for giving lectures. No other author has a conflict of interest with regard to any device
employed in this study.
Acknowledgements.—The authors wish to thank the anesthesiological nursing stuff for their help and assistance in conducting this
study.
Article first published online: June 28, 2016. - Manuscript accepted: June 23, 2016. - Manuscript revised: June 21, 2016. - Manuscript
received: February 12, 2016.
Vol. 82 - No. 11
Minerva Anestesiologica
1169
© 2016 EDIZIONI MINERVA MEDICA
Online version at http://www.minervamedica.it
Minerva Anestesiologica 2016 November;82(11):1170-9
ORIGINAL ARTICLE
Propofol versus midazolam for
premedication: a placebo‑controlled,
randomized double‑blinded study
Ofelia L. ELVIR LAZO 1, Paul F. WHITE 1, 2, 3*, Jun TANG 1, Roya YUMUL 1, 4,
Xuezhao CAO 1, 5, Firuz YUMUL 1, Jonathan HAUSMAN 1, Antonio HERNANDEZ CONTE 1,
Kapil K. ANAND 1, Emad G. HEMAYA 1, Xiao ZHANG 1, Ronald H. WENDER 1
1Department
of Anesthesiology, Cedars‑Sinai Medical Center, Los Angeles, CA, USA; 2White Mountain Institute,
The Sea Ranch, CA, USA; 3Istituto Ortopedico Rizzoli, University of Bologna, Italy; 4David Geffen School of
Medicine‑UCLA, Los Angeles, CA, USA; 5First Affiliated Hospital of China Medical University, Shenyang, China
*Corresponding author: Paul F. White, The White Mountain Institute, 41299 Tallgrass, The Sea Ranch, CA 95497, USA.
E‑mail: [email protected]
A B STRACT
BACKGROUND: It has been previously reported that subhypnotic doses of propofol could offer an advantage over
midazolam for premedication. This study was designed to test the hypothesis that a 20 mg IV dose of propofol would be
more effective than a standard 2 mg IV dose of midazolam for reducing acute anxiety prior to induction of anesthesia.
METHODS: One hundred twenty outpatients scheduled to undergo orthopedic surgery were randomly assigned to one
of three study groups: control (saline); propofol (20 mg); or midazolam (2 mg). Immediately before administering the
study medication, each patient evaluated their level of acute anxiety and sedation on 11‑point verbal rating scales (VRSs)
0=none- 10=highest, and they were also shown a picture. Upon arrival in the OR ~5 min after administering the study
medication, anxiety and sedation levels were reassessed and a second picture was shown. At discharge from the recovery
area, anxiety and sedation levels and their ability to recall the two pictures were reassessed.
RESULTS: Compared to the saline group, both propofol and midazolam produced significant increases in the patient’s
level of sedation upon entering the OR (+2.5±2.4 vs. +4.6±2.5 and +5.2±2.3, respectively [p<0.001]). Propofol was ef‑
fective as midazolam compared to saline in reducing the patient’s level of preinduction anxiety (from 3.2±2.2 to1.8±1.8
vs. 3.1±2.2 to 2.3±2.1 and 2.7±1.8 to 2.8±2.1, respectively). Propofol produced more pain on injection and midazolam
significantly reduced recall of the second picture.
CONCLUSIONS: When administered ~5 min prior to entering the OR, propofol, 20mg IV, was as effective as midazolam
2mg IV in reducing anxiety.
(Cite this article as: Elvir Lazo OL, White PF, Tang J, Yumul R, Cao X, Yumul F, et al. Propofol versus midazolam for pre‑
medication: a placebo‑controlled, randomized double‑blinded study. Minerva Anestesiol 2016;82:1170-9)
Key words: Preoperative care - Anxiety - Amnesia - Propofol - Midazolam.
E
arlier studies 1‑3 have reported that the
vast majority of outpatients would chose
to receive a sedative‑anxiolytic drug prior to
entering the OR. Administration of a small IV
dose of midazolam for premedication prior to
the patient entering the operating room (OR)
is a common clinical practice in the ambula‑
tory setting. However, even small doses of IV
1170
midazolam can contribute to residual sedation
and amnesia in the postanesthesia care unit
(PACU).1‑2, 4
Propofol has become the most popular IV
sedative‑hypnotic medication for both anes‑
thesia and sedation in the ambulatory setting
because of its rapid onset and excellent recov‑
ery profile.5, 6 It has been previously reported
Minerva AnestesiologicaNovember 2016
PROPOFOL VERSUS MIDAZOLAM FOR PREMEDICATIONELVIR LAZO
that subhypnotic doses of propofol could of‑
fer an economic advantage over midazolam
for premedication.7 However, there is only one
published study evaluating the use of propo‑
fol for IV premedication.7 In this earlier study,
Quario and Thompson reported that propofol
(0.4 mg/kg IV) had anxiolytic effects compa‑
rable to midazolam (40 µg/kg IV) with less
memory impairment, respiratory depression
and dizziness.7 The dose of midazolam used
in the earlier study was almost twice the stan‑
dard IV premedication dosage and these inves‑
tigators failed to assess the effect of propofol
and midazolam premedication on preoperative
recall and the early postoperative recovery
profile. Therefore, we designed a prospective,
randomized, double‑blind, placebo‑controlled
study to test the hypothesis that a 20 mg IV
bolus dose of propofol would be more effec‑
tive than 2 mg IV bolus dose of midazolam in
reducing acute anxiety prior to induction of
anesthesia. The secondary objectives were to
compare the effects of a small premedication
dose of propofol or midazolam on the level of
sedation and recall immediately prior to induc‑
tion of anesthesia.
Materials and methods
This clinical study was approved by Ce‑
dars‑Sinai Medical Center’s Institutional Re‑
view Board (IRB) with # Pro00025204 on
08/30/2013 (which was registered in ClinTri‑
als.gov, Reg. # NCT01976845). It also com‑
plied with all 25 items on the Consort 2010
checklist. After obtaining IRB approval and
written informed consent, 120 ASA physical
status I‑III adult patients (range: 18‑68 years)
scheduled to undergo elective orthopedic sur‑
gery procedures were enrolled in this study and
randomly assigned to one of three treatment
groups (N.=40/group): control (Saline), pro‑
pofol (20 mg), or midazolam (2 mg). Random‑
ization assignment was generated with a 1:1:1
allocation ratio using a computer software pro‑
gram. The randomization number specifying
the study medication (namely, saline, propofol
or midazolam) was placed in sealed envelope
and given to a non‑participating anesthesi‑
Vol. 82 - No. 11
ologist on the day of the scheduled procedure
after completing the patient’s preoperative as‑
sessment. The study was conducted between
November 2013 and December 2014.
Prior to the day of surgery, potential study
patients received a packet of materials from
their surgeon describing the study. The pack‑
et contained an initial patient contact letter,
a HIPAA information sheet, and the IRB‑ap‑
proved informed consent form.
Patients who were interested in participating
in the study were asked to bring the materi‑
als to the hospital on the day of surgery. After
reviewing all the details of the study with the
patient and verifying that they met all the in‑
clusionary criteria, the study coordinator col‑
lected the signed consent forms. The patients
were informed that it would be important to
complete all aspects of the study, including
answering questions in the early postoperative
period.
Inclusion criteria included the following:
Patients to undergo orthopedics procedures,
Willingness and ability to sign an informed
consent document, no allergies to midazolam
or propofol, 18‑70 years of age, American
Association of Anesthesiologists Class I‑III
adults.
Exclusion criteria included the following:
Patients with known allergy, hypersensitivity
or contraindications to midazolam, propofol,
standardized anesthetic and analgesic medi‑
cations administered during the study period,
unstable preexisting medical conditions in‑
volving the central nervous system (CNS),
pregnant or lactating women, history of alco‑
hol or drug abuse within the past 3 months,
chronic use of sedative‑hypnotic or anxiolytic
medications prior to the operation, morbidity
obesity (Body Mass Index [BMI] >40 kg/m2),
unstable angina, diseases requiring oxygen
therapy, obstructive sleep apnea (OSA) and
sleep disordered breathing.
Withdrawal criteria included the follow‑
ing: Allergies or hypersensitivity to propofol
or midazolam, patients present any clinical‑
ly‑significant medical conditions involving the
brain, heart, kidney, liver or endocrine diseas‑
es, patients agitated or confused with a level
Minerva Anestesiologica
1171
ELVIR LAZOPROPOFOL VERSUS MIDAZOLAM FOR PREMEDICATION
of anxiety greater than 6 prior to receiving the
study drug, patient who received additional
premedication upon entering into the OR.
In the preoperative holding area, consent‑
ing patients were asked to provide a detailed
medical history including demographic infor‑
mation (e.g., age, weight, height, BMI), smok‑
ing history, chronic medications, recreational
drug usage, and history of motion sickness or
previous postoperative nausea and vomiting
(PONV). Each patient was connected to moni‑
toring devices for recording their respiratory
rate, oxygen saturation, heart rate, and blood
pressure values as part of the standard of care.
In the preoperative holding area immedi‑
ately before administering the study medi‑
cation each patient evaluated their level of
acute anxiety and sleepiness (sedation) using
11‑point verbal rating scales (VRSs) (0=none
to 10=extremely high). The VRSs have been
used previously to assess sedation and anxiety
levels.8‑11 In addition, each patient was evalu‑
ated using the Observer’s Assessment of Alert‑
ness/Sedation (OAA/S) Scale 12 (Appendix I)
and the Ramsay Scale (Appendix II).13 After
performing these assessments, the patients
were shown a picture of an apple. The patients
were then administered 2 mL IV of the study
medication in an opaque syringe containing
saline (2 mL), propofol (20 mg), or midazolam
(2 mg) 3‑6 min prior to being transported to
the OR by an anesthesiologist who was not in‑
volved in caring for the patient or performing
the perioperative assessments. The syringes
were covered with an opaque tape to blind the
medication from the patient and the individual
responsible for administering the study medi‑
cation.
Each patient was asked whether or not they
experienced any discomfort during the injec‑
tion of the study medication and a binary “yes
or no” was recorded. The investigators in‑
(N.=161)
(N.=0)
(N.=0)
(N.=22)
(N.=0)
(N.=139)
(N.=48)
(N.=48)
(N.=48)
(N.=48)
(N.=43)
(N.=43)
(N.=0)
(N.=0)
(N.=0)
(N.=0)
(N.=8)
(Analyzed N.=40)
(N.=0)
(N.=0)
(N.=0)
(N.=8)
Analyzed (N.=40)
(N.=0)
(N.=3)
Analyzed (N.=40)
(N.=0)
Figure 1.—Consort 2010 flow diagram.
1172
Minerva AnestesiologicaNovember 2016
PROPOFOL VERSUS MIDAZOLAM FOR PREMEDICATIONELVIR LAZO
volved in performing the perioperative assess‑
ments were blinded as to the patient’s response
the question regarding pain on injection of the
study medication. Upon arrival in the OR ~5
min after administering the study medication,
anxiety and sedation levels were reassessed
using the same 11‑point VRSs and a second
picture (a dollar bill) was shown to the pa‑
tients. The patient’s level of consciousness was
also reassessed using the OAA/S and Ramsay
scales prior to induction of anesthesia. After
applying the standard OR monitoring devices,
general anesthesia was induced with a combi‑
nation of lidocaine 30‑50 mg IV and propofol
2‑2.5 mg/kg IV followed by desflurane 3‑6%
or sevoflurane 1‑2% titrated to maintain he‑
modynamic stability. A laryngeal mask airway
(LMA) device was used in all cases except in
3 cases where the anesthesiologists decided to
perform tracheal intubation with an endotra‑
cheal tube.
A local anesthetic infiltration was adminis‑
tered to all patients before performing the sur‑
gical incision and a peripheral nerve block at
the end of the surgical procedural for periop‑
erative analgesia. After the operation, patients
were transported to the PACU. Upon discharge
from the recovery area, anxiety and sedation
levels were again assessed using the VRSs and
patients were queried regarding their ability to
recall the two pictures.
The standardized perioperative assessments
included: 1) patient demographic data; 2) lev‑
el of anxiety (i.e., nervousness) and sedation
(i.e., sleepiness) at three time points (In the
preoperative area prior to being transported
to the OR, before induction of anesthesia, and
prior to discharge from the recovery area); 3)
dosages of all anesthetic drugs, local anesthet‑
ics, and IV fluid therapy during the operation;
4) duration of surgery (from skin incision until
closure) and anesthesia (from IV induction un‑
til discontinuation of the volatile anesthetic);
5) duration of the PACU stay; 6) requirements
for “rescue” analgesic and antiemetic medica‑
tion in the PACU, 7) Side effects in the PACU
prior to discharge home; 8) time from baseline
assessment to study drug administration; 9)
time from study drug administration until the
Vol. 82 - No. 11
pre‑induction assessments were performed;
and 10) ability to recall the two pictures shown
before the start of surgery.
None of the study patients received any type
of sedative‑anxiolytic premedication prior to
enrollment in this study. All consenting pa‑
tients received their premedication in the pre‑
operative holding area ~5 minutes prior to en‑
tering the operating room (OR). If the patient
received additional premedication upon enter‑
ing into the OR, their participation in the study
was terminated and their data was excluded
from the final statistical analysis.
Statistical analysis
A sample size of 34 patients in each group
was determined using a power analysis based
on the presumption than both propofol and
midazolam will decrease anxiety score by
25% or more compared to the Control (sa‑
line) group (VRS=3, 3, and 4, respectively),
SD=1.5, with a power of 80% and significance
level of 0.05 (using 2‑sided t‑test). However,
to reduce the risk of a type II statistical er‑
ror, we decided before starting the study to
increase the sample size to 40 patients per
group.7 The analysis was performed using
SAS 9.3 for Windows (SAS Institute, Cary,
NC, USA) and R. 3.0.1. Our dataset contained
both categorical and continuous measure‑
ments. For categorical measures, we presented
total numbers (N.) with the percentages (%)
and used χ2 test (or Fisher’s Exact Test) to
conduct the group comparisons. For continu‑
ous measures, we presented mean values with
their standard deviations and used the Kol‑
mogorov‑Smirnov test to check the normality.
To conduct the multiple comparisons among
those 3 groups, we performed the one‑way
ANOVA if the measure is approximately nor‑
mal and the Kruskal Wallis test if the measure
fails to pass the normality test. A Bonferroni
correction was applied when multiple com‑
parisons were performed over time. All tests
were two‑sided; and P values ≤0.05 were con‑
sidered statistically‑significant. Data are pre‑
sented as mean values±SD, numbers (N.), and
percentages (%).
Minerva Anestesiologica
1173
ELVIR LAZOPROPOFOL VERSUS MIDAZOLAM FOR PREMEDICATION
Results
The three groups were comparable with re‑
spect to demographic characteristics, duration
of anesthesia, surgery and recovery stay, type
of surgery, vital signs, oxygen saturation and
baseline sedation and anxiety scales among the
three treatment groups (Table I, Figure 2).
Compared to the saline group, both propofol
and midazolam produced significant increases
in the patient’s level of sedation (or sleepiness)
upon entering the OR (+2.5±2.4 vs. +4.6±2.5
and +5.2±2.3, respectively [P<0.001] and
compared to their respective baseline values).
Propofol and midazolam were equivalent in
effectiveness in reducing the patient’s level of
preinduction anxiety compared to saline (from
3.2±2.2 to 1.8±1.8 vs. 3.1±2.2 to 2.3±2.1 and
2.7±1.8 to 2.8±2.1, respectively) (Figure 3).
Although propofol produced more pain on
injection than midazolam and saline (40% vs.
23% and 10%, respectively P≤ 0.007), mid‑
azolam significantly reduced recall of the sec‑
ond picture compared to propofol and saline
(30% vs. 75% and 95%, respectively, P≤0.001)
(Table I).
Table II summarizes the intraoperative drug
usage in the three premedication treatment
Table I.—Demographic characteristics, duration of anesthesia, surgery and recovery stay, surgery type, OAAS and
Ramsay sedation scale, picture recall and pain on administration of the study drug for the three premedication
treatment groups.
Age (yr)
BMI (kg/m2)
Gender (N.) F/M
ASA (N.) I/II/III
Race (N.) Asian/Black/Caucasian/Hispanic
Smoker (N.)
Alcohol user (N.)
History of PONV (N.) yes/no
Took sleeping or anti‑anxiety drugs the night before (N.)
General anesthesia
Sedation
Peripheral nerve block at the end of surgery
Anesthesia time (min)
Surgery time (min)
Time to discharge (min) to the postsurgical ward for arthroplasty cases
Time to discharge (min) to home for the non‑arthroplasty cases
Time from baseline assessment to study drug (min)
Time from administration of study drug until the pre‑induction assess‑
ments were performed (min)
Type of orthopedic surgery (N.):
Arthroplasty
Arthroscopy
Open reduction internal fixation (ORIF)
Tendon/ligament repair/biopsy
OAAS§ (N.) (4 /5) baseline level
(4/5) before induction level
Ramsay (N.) baseline level
Before induction level
Memory [picture] recall:
Pre‑picture # 1 (N.) yes/no
Post‑picture # 2 (N.) yes/no
Pain on administration (N.)
Saline
(N.=40)
Propofol
(N.=40)
Midazolam
(N.=40)
P
49±14
28±5
15/25
16/19/5
1/4/35/0
2
21
7
12
13
27
18
107±45
79±41
172±92
116±70
28±28
3.8±2
52±13
28 ± 5
24/16
6/28/6
3/5/31/1
5
17
4
13
14
26
17
116±48
84±44
199±105
102±50
32±42
3.9±2
51 ±15
28 ± 6
17/23
12/24/4
3/6/31/0
5
21
9
11
16
24
23
130±61
102±54
184±100
109±30
24±26
4.6±2
0.7
0.9
0.1
0.2
0.8
0.4
0.6
0.3
0.9
0.8
0.8
0.6
0.2
0.07
0.8
0.7
0.5
0.1
21
7
6
5
0/40
2/37
40
40
25
7
4
4
0/40
6/34
40
40
23
6
6
5
0/40
7/33
40
40
0.9
0.9
0.8
0.9
NA
0.2
NA
NA
40/0
38/2
4
39/1
30/10*†
16*
40/0
12/28*†‡
9
NA
<0.001
0.007
Values are mean±SD or numbers (N.)
*Significantly different from Saline group, P<0.05.
†Significantly different from the Baseline value, P<0.05. ‡Significantly different from the Propofol group, P<0.05.
§OAAS: Observer Assessment of Alertness and Sedation Scale.
1174
Minerva AnestesiologicaNovember 2016
PROPOFOL VERSUS MIDAZOLAM FOR PREMEDICATIONELVIR LAZO
P value=0.6
P value=0.6
P value=0.8
P value=0.8
P value=0.7
P value=0.8
Figure 2.—Comparative effects of midazolam, propofol and saline on mean arterial pressure (MAP), the heart rate (HR), and
oxygen saturation during the perioperative period for the three premedication treatment groups.
Values are mean±SD or numbers (n). * Significantly different from Saline group, P<0.05. Baseline: before the administration
of the premedication. Before Induction: approximately 5 minutes after the administration of the premedication.
groups. There were no significant differences
in intraoperative drug usage or emergence
times from anesthesia among the three treat‑
ment groups. In addition to the length of the
PACU stay; the incidence of side effects and
requirements for rescue analgesic and anti‑
emetic medications were also similar among
the three treatment groups (Table II).
All patients were conscious and were re‑
sponsive to questions from the investigator at
all times prior to induction of anesthesia. The
patient’s level of consciousness was unchanged
from their baseline values when assessed using
Vol. 82 - No. 11
the OAA/S 12 and Ramsay 13 scales (Table I).
Compared to saline, the administered dosag‑
es of propofol and midazolam failed to delay
time to discharge to the postsurgical ward for
arthroplasty cases or to discharge to home for
the non‑arthroplasty cases.
Discussion
Premedication is considered to be a valuable
part of the perioperative surgery experience by
patients and anesthesia providers.14‑16 In the
current study, premedication with propofol 20
Minerva Anestesiologica
1175
ELVIR LAZOPROPOFOL VERSUS MIDAZOLAM FOR PREMEDICATION
Anxiety Level
Saline
Propofol
Baseline Level
Sedation Level
Midazolam
Saline
Before Induction Level
Propofol
Midazolam
Postoperative Level
Figure 3.—Comparative effects of propofol and midazolam (vs. saline) on the level of sedation and anxiety for the three
premedication treatment groups.
Values are mean±SD or numbers (n). * Significantly different from Saline group, P<0.05. † Significantly different from the
Baseline value, P<0.05. Anxiety level (0=none to 10= extremely nervous) and Sedation level (0=none to 10= extremely se‑
dated/sleepy). Baseline: before the administration of the premedication. Before Induction: approximately 5 minutes after the
administration of the premedication.
mg IV, was found to produce significant anx‑
iolytic, sedative and amnestic effects without
cardiorespiratory depression. The only disad‑
vantage of using propofol was more frequent
mild discomfort upon IV injection. The use of
a relatively small‑dose of propofol 20 mg IV
avoided excessive CNS depression and avoid‑
ed delaying the early recovery processes after
these minor (non‑arthroplasty) ambulatory
orthopedic procedures. Both midazolam and
propofol produced greater amnesia than saline;
however, propofol predication was associated
with less impairment of recall than midazolam
(P<0.001). These findings are consistent with
earlier studies documenting the significant an‑
terograde amnestic effects of midazolam pre‑
medication.1, 7 However, this is the first study
documenting the amnestic effect of a subhyp‑
notic dose of propofol.
1176
The benzodiazepine drug class has been al‑
leged to produce dose‑dependent anxiolytic,
sedative and amnestic effects when adminis‑
tered for premedication.10 However, the dos‑
age which is commonly administered for pre‑
medication, namely 2 mg IV (20‑25 µg/kg
IV), failed to produce a statistically‑significant
decrease in the patient’s level of preoperative
anxiety compared to saline. Given its more fa‑
vorable anxiolytic effects (and lesser degrees
of amnesia), low‑dose propofol may offer ad‑
vantages over midazolam for IV premedica‑
tion of anxious patients in the ambulatory set‑
ting. A larger scale follow‑up study is clearly
needed to evaluate the impact of propofol pre‑
medication in the preoperative holding area
and its impact on other clinical outcomes after
ambulatory surgery.
Propofol has the advantage of a more rapid
Minerva AnestesiologicaNovember 2016
PROPOFOL VERSUS MIDAZOLAM FOR PREMEDICATIONELVIR LAZO
Table II.—Intraoperative Drugs Administered, Incidence of Side Effects, Need for Rescue Analgesic and Antiemetic
Drugs Administered to the three premedication treatment groups.
Intraoperative Drugs
Lidocaine 50‑100 mg (n)
Fentanyl µg (mcg)
Propofol (mg)
Glycopyrrolate 0.2‑1mg (n)
Bupivacaine (mg)
Ondansetron 4 mg (n)
Ephedrine 10‑20 mg (n)
Famotidine 20 mg (n)
Hydromorphone (n)
Morphine mg (n)
Total Fluids (mL)
Side Effects
Nausea (n)
Itching (n)
Rescue analgesic requirement:
Hydromorphone 0.2‑3 mg (n)
Fentanyl 50‑100 mcg (n)
Ketorolac 50 mg (n)
Norco Tab (n)
Lortab (n)
Oxycodone (mg)
Tramadol 50‑100 mg (n)
Acetaminophen 650 mg‑1000 (n)
Celebrex 200mg (n)
Ondansetron (n)
Saline (N.=40)
Propofol (N.=40)
Midazolam (N.=40)
P value
6
88±40
500±357
4
31±23
20
1
12
0
0
1471±818
10
113±65
577±360
1
31±15
13
4
7
3
3
1646±708
3
125±82
681±505
5
40±19
18
2
9
3
0
1595±651
0.1
0.35
0.15
0.24
0.30
0.26
0.3
0.4
0.2
NA
0.54
5
0
6
1
4
0
0.8
NA
16
3
0
8
2
4
2
6
2
5
16
2
1
6
1
6
1
4
1
7
14
1
0
5
0
4
1
5
1
5
0.8
0.6
NA
0.6
0.4
0.7
0.8
0.8
0.8
0.2
Values are mean±SD or numbers (n). * Significantly difference from Saline and midazolam group, P<0.05
onset of action (30 seconds vs. 120 seconds
for midazolam) and a shorter duration of ef‑
fect (biologic half‑life of 4 minutes vs. 30
minutes for midazolam), and can also produce
dose‑dependent sedation.17 Consistent with the
earlier study,7 we did not find any difference in
the level of consciousness (OAA/S and Ram‑
say scales [Table I]) produced by propofol 20
mg and midazolam 2 mg prior to induction of
anesthesia. Importantly, all patients were fully
conscious, cooperative and responsive to ques‑
tions from the anesthesiologists immediately
prior to induction of anesthesia with the doses
of propofol and midazolam administered in the
preoperative holding area prior to transferring
the patient to the OR.
One of the major concerns regarding the
routine use of propofol for premedication re‑
lates to the possible need for additional moni‑
toring and/or supplemental oxygen in the pre‑
operative holding area. Although induction
doses of propofol can produce dose‑dependent
Vol. 82 - No. 11
decreases in blood pressure and heart rate,
as well as apnea and respiratory depression,
doses of propofol ranging from 40 to 150 mg
(total dose) were not found to be associated
with significant cardiovascular or respiratory
changes.18 The findings of a meta‑analysis by
Wang et al.19 are also consistent with our find‑
ings that low‑dose propofol sedation does not
increase the risk of bradycardia, hypotension,
or hypoxemia compared with midazolam seda‑
tion.
Compared to saline, neither the anesthesia
time nor the duration of the PACU stay were
prolonged by use of propofol (20 mg e.v.) or
midazolam (2 mg IV) for premedication. Fur‑
thermore, the postoperative pain scores, inci‑
dence of PONV, and need for rescue analge‑
sics and antiemetic medications did not differ
among the three treatment groups, demonstrat‑
ing that these premedicant dosages of midazol‑
am or propofol did not increase postoperative
side effects compared to saline.
Minerva Anestesiologica
1177
ELVIR LAZOPROPOFOL VERSUS MIDAZOLAM FOR PREMEDICATION
Although the earlier study by Quario and
Thompson 7 examining the use of propofol as
an alternative to midazolam for premedication
reported similar findings, these authors admin‑
istered larger doses of both propofol (0.4 mg/kg
vs. 0.26 mg/kg) and midazolam (40 µg/kg vs.
24 µg/kg). The standard premedication dosage
of midazolam is 20‑25 µg/kg e.v.). With larger
dosages of these CNS depressant drugs, elderly
and higher risk patients’ populations might ex‑
perience both cardiovascular and respiratory
depression. In the earlier study,7 midazolam
premedication in the absence of supplemental
oxygen was associated with a decrease in SpO2
values <98% in 45% of the patients (vs. 10%
in the propofol and saline groups, respectively).
Despite breathing room air in the current study,
SpO2 values did not show any significant differ‑
ences among the three treatment groups. In the
earlier study, both the larger dosages of propo‑
fol and midazolam also produced small but sta‑
tistically significant decreases in MAP values.
Although propofol was highly effective when
administered for IV premedication, 16 patients
(40%) in our study reported mild discomfort on
injection of the low‑dose of propofol. The addi‑
tion of a small amount of lidocaine to the propo‑
fol formulation would have further reduced (or
even eliminated) the discomfort on injection of
the predication.20
Another potential criticism of this study
is that fixed dosages of the two study drugs
were administered to all patients irrespective
of their body weight, they were chosen arbi‑
trarily based on standard clinical practices in
North America to administer fixed doses of the
medications which are commonly used for pre‑
medication. However, in the future, a proper
dose‑response study should be performed to
determine the optimal doses of each drug for
endovenous premedication. Another limita‑
tion was the exclusion of patients with anxi‑
ety >6/10; this could biased the patient selec‑
tion, but this was a requirement from IRB in
order to perform the study. Fortunately, none
of the prospective patients were excluded from
participating in the study due to high levels of
anxiety. Finally, the absence of any compara‑
tive studies that analyze other types of memory
1178
(non‑declarative) testing, and the lack of ad‑
ditional memory evaluations in our study, are
noteworthy study limitations.
Conclusions
When administered ~5 minutes prior to en‑
tering the OR, propofol, 20mg IV, appears to
be an effective anxiolytic as midazolam, 2 mg
IV. Compared to midazolam, propofol pro‑
duced a similar level of sedation but caused
less impairment of recall prior to induction of
anesthesia.
Key messages
—— Premedicant doses of propofol (20
mg) and midazolam (2 mg) which pro‑
duced comparable levels of sedation were
associated with similar reductions in pre‑
operative anxiety.
—— At equivalent levels of sedation,
midazolam produced greater impairment
of recall at induction of anesthesia com‑
pared to propofol.
—— Premedicant dose of propofol (20
mg) or midazolam (2 mg) did not produce
significant cardiorespiratory changes.
—— When propofol (20 mg) or midazol‑
am (2 mg) were administered for IV pre‑
medication immediately prior to entering
the operating room, these sedative-hypnot‑
ics failed to delay discharge home (outpa‑
tients) or to the surgical ward (inpatients).
References
  1.Shafer A, White PF, Urquhart ML, Doze VA. Outpatient
premedication: use of midazolam and opioid analgesics.
Anesthesiology 1989;71:495‑501.
  2.Van Vlymen JM, Sá Rêgo MM, White PF. Benzodiaz‑
epine premedication: can it improve outcome in patients
undergoing breast biopsy procedures? Anesthesiology.
1999;90:740‑7.
 3.Raeder JC, Breivik H. Premedication with midazolam
in out‑patient general anaesthesia: a comparison with
morphine‑scopolamine and placebo. Acta Anaesthesiol
Scand 1987;31:509‑14.
  4. Kain ZN, Mayes LC, Bell C, Weisman S, Hofstadter MB,
Rimar S. Premedication in the United States: a status re‑
port. Anesth Analg 1997;84:427‑32.
Minerva AnestesiologicaNovember 2016
PROPOFOL VERSUS MIDAZOLAM FOR PREMEDICATIONELVIR LAZO
 5.White PF. Ambulatory anesthesia and surgery: past,
present and future. In: White PF, editors. Ambulatory an‑
esthesia & surgery. London: WB Saunders Co; 1997. p.
1‑34.
  6. Doze VA, Westphal LM, White PF. Comparison of pro‑
pofol with methohexital for outpatient anesthesia. Anesth
Analg 1986;65:1189‑95.
  7. Quario Rondo L, Thompson C. Efficacy of propofol
compared to midazolam as an intravenous premedication
agent. Minerva Anestesiol 2008;74:173‑9.
  8. White PF, Coe V, Shafer A, Sung ML. Comparison of al‑
fentanil with fentanyl for outpatient anesthesia. Anesthe‑
siology 1986;64:99‑106.
  9.Shafer A, White PF, Urquhart ML, Doze VA. Outpatient
premedication: use of midazolam and opioid analgesics.
Anesthesiology 1989;71:495‑501.
10. White PF, Tufanogullari B, Taylor J, Klein K. The effect of
pregabalin on preoperative anxiety and sedation levels:a
dose‑ranging study. Anesth Analg. 2009;108:1140‑
5.
11.Tanasale B, Kits J, Kluin PM, Trip A, Kluin‑Nelemans
HC. Pain and anxiety during bone marrow biopsy. Pain
Manag Nurs 2013;14:310‑7.
12.Chernik DA, Gillings D, Laine H, Hendler J, Silver JM,
Davidson AB, et al. Validity and reliability of the Ob‑
server’s Assessment of Alertness/ Sedation Scale: study
with intravenous midazolam. J Clin Psychopharmacol
1990;10:244‑51
13.Ramsay MAE, Savege TM, Simpson BRJ, Goodwin R.
Controlled sedation with alpaxalone‑alphadolone. Br
Med J 1974;2:656‑9.
14.Cressey DM, Claydon P, Bhaskaran NC, Reilly CS. Ef‑
fect of midazolam pretreatment on induction dose re‑
quirements of propofol in combination with fentanyl in
younger and older adults. Anaesthesia 2001;56:108‑13.
15. Bergendahl H, Lonnqvist PA, Eksborg S. Clonidine: an
alternative to benzodiazepines for premedication in chil‑
dren. Current Opin Anaesthesiol 2005;18:608‑13.
16. Bergendahl H, Lonnqvist PA, Eksborg S. Clonidine in
paediatric anaesthesia: review of the literature and com‑
parison with benzodiazepines for premedication. Acta
Anaesthesiol Scand 2006;50:135‑43.
17. Triantafillidis JK, Merikas E, Nikolakis D, Papalois AE.
Sedation in gastrointestinal endoscopy: current issues.
World J Gastroenterol 2013;19:463‑81.
18.Paspatis GA, Manolaraki M, Xirouchakis G, Papan‑
ikolaou N, Chlouverakis G, Gritzali A. Synergistic seda‑
tion with midazolam and propofol versus midazolam and
pethidine in colonoscopies:a prospective, randomized
study. Am J Gastroenterol 2002;97:1963‑7.
19. Wang D, Wang S, Chen J, Xu Y, Chen C, Long A, et al.
Propofol combined with traditional sedative agents versus
propofol- alone sedation for gastrointestinal endoscopy:a
meta‑analysis. Scand J Gastroenterol 2013;48:101‑10.
20. Jalota L1, Kalira V, George E, Shi YY, Hornuss C, Radke
O, et al. Perioperative Clinical Research Core. Preven‑
tion of pain on injection of propofol: systematic review
and meta‑analysis. BMJ. 2011 15;342:d1110.
Funding.—This study was funded by Department of Anesthesiology, Cedars‑Sinai Medical Center, Los Angeles, CA, USA.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material
discussed in the manuscript.
Congresses.—This report was previously presented in part at the ASA and WARC meetings.
Article first published online: September 9, 2016. - Manuscript accepted: September 7, 2016. - Manuscript revised: July 28, 2016. Manuscript received: February 4, 2016.
Appendix I
The observer’s assessment of alertness/sedation (OAAS) rating scale
OAAS score 5—awake and responds readily to name spoken in normal tone.
OAAS score 4—lethargic responses to name in normal tone.
OAAS score 3—responds only after name is called loudly and/or repeatedly.
OAAS score 2—responds only after name called loudly and mild shaking.
OAAS score 1—does not respond when name is called loudly and mild shaking or prodding.
OAAS score 0—does not respond to noxious stimulation.
Appendix II
Ramsey Sedation Scale
Ramsey 1 Anxious, agitated, restless.
Ramsey 2 Cooperative, oriented, tranquil.
Ramsey 3 Responsive to commands only.
Ramsey 4 Brisk response to light glabellar tap or loud auditory stimulus.
Ramsey 5 Sluggish response to light glabellar tap or loud auditory stimulus.
Ramsey 6 No response to light glabellar tap or loud auditory stimulus.
Vol. 82 - No. 11
Minerva Anestesiologica
1179
© 2016 EDIZIONI MINERVA MEDICA
Online version at http://www.minervamedica.it
Minerva Anestesiologica 2016 November;82(11):1180-8
ORIGINAL ARTICLE
The optimal time between clinical brain
death diagnosis and confirmation using
CT angiography: a retrospective study
Lionel KERHUEL 1, Mohamed SRAIRI 1 *, Gilles GEORGET 1,
Fabrice BONNEVILLE 2, Ségolène MROZEK 1, Nicolas MAYEUR 1, Laurent LONJARET 1,
Sandrine SACRISTA 1, Nathalie HERMANT 1, Fouad MARHAR 1, François GAUSSIAT 1,
Timothée ABAZIOU 1, Diane OSINSKI 1, Benjamin LE GAILLARD 1, Rémi MENUT 1,
Claire LARCHER 1, Olivier FOURCADE 1, Thomas GEERAERTS 1
1Department
of Anaesthesia and Intensive Care, University Hospital of Toulouse, University Toulouse 3 Paul
Sabatier, Toulouse, France; 2 Department of Neuroradiology, University Hospital of Toulouse, University Toulouse
3 Paul Sabatier, Toulouse, France
*Corresponding
author: Mohamed Srairi, Department of Anaesthesia and Intensive Care, University Hospital of Toulouse, 1 Place
Baylac TSA 40031, 31059 Toulouse cedex 09, France. E‑mail: srairi.m@chu‑toulouse.fr
A B STRACT
BACKGROUND: In several countries, a computed tomography angiography (CTA) is used to confirm brain death (BD).
A six‑hour interval is recommended between clinical diagnosis and CTA acquisition despite the lack of strong evidence
to support this interval. The aim of this study was to determine the optimal timing for CTA in the confirmation of BD.
METHODS: This retrospective observational study enrolled all adult patients admitted between January 2009 and De‑
cember 2013 to the intensive care units of a French university hospital with clinically diagnosed BD and at least one
CTA performed as a confirmatory test. The CTAs were identified as conclusive (e.g. yielding confirmation of BD) or
inconclusive (e.g. showing persistent brain circulation).
RESULTS: One hundred and four patients (sex ratio M/F 1.8; age 55 years [41‑64]) underwent 117 CTAs. CTAs con‑
firmed cerebral circulatory arrest in 94 cases yielding a sensitivity of 80%. Inconclusive CTAs were performed earlier
than conclusive ones (2 hours [1‑3] vs. 4 hours [2‑9], P=0.03) and were associated with decompressive craniectomy (5
cases [23%] vs. 6 cases [7%], P=0.05) and the failure to complete full neurological examination (5 cases [23%] vs. 4 cases
[5%], P=0.02). Six hours after BD clinical diagnosis, the proportion of conclusive CTA was only 51%, with progressive
increase overtime with more than 80% of conclusive CTA after 12 hours.
CONCLUSIONS: A 12‑hour interval might be appropriate in order to limit the risk of inconclusive CTAs.
(Cite this article as: Kerhuel L, Srairi M, Georget G, Bonneville F, Mrozek S, Mayeur N, et al. The optimal time between clin‑
ical brain death diagnosis and confirmation using CT angiography: a retrospective study. Minerva Anestesiol 2016;82:1180-8)
Key words: Brain death - Tomography, X‑Ray computed - Angiography.
I
n 50% of European countries, the confirma‑
tion of a clinically diagnosed brain death
(BD) by ancillary tests is mandatory before or‑
gan donation.1, 2 Brain computed tomography
angiography (CTA) is the preferred confirma‑
tory test in approximately 40% of European
countries,1 but also outside Europe.3 In coun‑
1180
tries where confirmatory tests are not manda‑
tory like Canada, United States of America or
United Kingdom, CTA remains recommended
when uncertainty exists, and particularly when
neurological examination cannot be fully com‑
pleted.4 BD can be affirmed using CTA when
absence of contrast enhancement is found on a
Minerva AnestesiologicaNovember 2016
OPTIMAL TIMING FOR CTA IN THE CONFIRMATION OF BD
selected number of brain vessels.5 In France,
a four‑vessel CTA interpretation model is cur‑
rently used, according to recent guidelines.6
The use of the four‑vessel score is supported
by the study of Frampas et al.7
The French Society of Neuroradiology
recommends a minimum six‑hour interval
between clinically diagnosed BD and the ac‑
quisition of CTA, in order to allow the intra‑
cranial pressure (ICP) to exceed the mean arte‑
rial pressure (MAP) both in the infratentorial
and the supratentorial compartments, leading
to a cerebral circulatory arrest (CCA) 6 despite
the fact that no strong evidence supports this
six‑hour time interval. Interestingly, the rec‑
ommended time lag is the same regardless of
BD etiology especially anoxic brain damage
that is not managed differently from the other
conditions. Determining the optimal timing
for CTA may prevent unnecessary transport
to the radiology facility and may improve the
diagnostic performance of CTA for BD con‑
firmation. Moreover, several limits of CTA
regarding its insufficient sensitivity (e.g. pre‑
served intracranial vascular filling in clinically
brain‑dead patients) have been described, rais‑
ing serious concerns about its use as a compul‑
sory test in particular settings, such as posterior
fossa primary lesion, skull fracture, decom‑
pressive craniectomy, ventricular drainage,
and anoxic encephalopathy.8, 9 Most relevant
studies assessing CTA sensitivity used clinical
examination as a gold standard, have shown a
sensitivity between 65% and 100%.7, 10‑14 This
variability may preclude routine CTA use in
common practice.
The main objective of this study was to de‑
termine the optimal CTA timing for BD confir‑
mation. We also aimed to study the risk factors
for inconclusive CTA (e.g. showing persistent
brain circulation).
Materials and methods
Ethical considerations
According to French law, informed consent
from each patient and/or next of kin was not
necessary because of the retrospective design
Vol. 82 - No. 11
KERHUEL
nor was the local ethical committee agreement.
Data were anonymized before statistical analy‑
sis.
Study population
All adult patients admitted to the General
Intensive Care Unit (ICU) or the Neurosur‑
gical ICU of a French university hospital be‑
tween 1 January 2009 and 31 December 2013,
and subsequently developing clinical symp‑
toms of BD, were considered. Final inclusion
in the study required at least one CTA acquisi‑
tion for BD confirmation during the ICU stay.
Exclusion criteria were: a BD diagnosis prior
admission to the ICU and substantial miss‑
ing data restricting statistical analysis. As the
hospital is a referral center for organ procure‑
ment, some patients were referred from level
II centers after BD was confirmed. ICU admis‑
sion policies in both units remained unchanged
throughout the study period.
BD diagnosis
A nurse performed sequential clinical exam‑
inations at least every two hours (composed of
the Glasgow Coma Scale, assessment of pupil
diameter and their reactivity to light, urine out‑
put and urine density measurement) as part of
our ICU clinical monitoring protocol. When a
clinical modification was noticed, the intensive
care physician was immediately called to per‑
form a full neurological examination to assess
clinical BD. After the exclusion of confounding
factors (hypothermia, mean arterial pressure
less than 65 mmHg, residual sedation, electro‑
lyte or acid‑base disturbances), a clinical diag‑
nosis of BD was confirmed in patients with an
identified critical brain injury compatible with
a diagnosis of BD, a Glasgow Coma Scale of 3,
and a cessation of all brainstem reflexes includ‑
ing spontaneous ventilation assessed by the
Apnea Test (AT) according to practical guid‑
ance of the American Academy of Neurology
for determination of BD.4 AT was performed as
follows: a duration of 10‑15 minutes, preoxy‑
genation with 100% oxygen, initial equilibra‑
tion of arterial carbon dioxide, the absence of
Minerva Anestesiologica
1181
KERHUEL
OPTIMAL TIMING FOR CTA IN THE CONFIRMATION OF BD
spontaneous breathing was confirmed by both
a clinical examination and a partial pressure of
carbon dioxide ≥ 60 mmHg.15
According to French legislation,2 either one
CTA or two electroencephalograms (EEG)
separated by a four‑hour interval are mandatory
for BD confirmation before organ donation. Re‑
sidual opacified vessels on a CTA or residual
electrical activity on an EEG precluded BD
confirmation. In the case of an inconclusive an‑
cillary test, the French procedure requires the
repetition of the EEG or CTA until BD confir‑
mation is reached. All CTAs were performed lo‑
cally, following the protocol recommended by
the French Society of Neuroradiology.6 A pre‑
liminary unenhanced cerebral acquisition was
used as a reference. Then, a contrast medium
(Iomeprol; concentration: 400 mg/mL; volume:
90‑120 mL) was injected intravenously at a rate
of 4 mL/s. A second acquisition was performed
60 seconds after the injection and was used to
evaluate the contrast enhancement of the ves‑
sels. Local attending neuroradiologists inter‑
preted CTA images. Opacification of the super‑
ficial temporal arteries was assessed to confirm
the correct injection of the contrast medium. Re‑
sidual vessel opacification was determined by
a visual comparison between unenhanced and
enhanced CT images. From 2009 to 2010, BD
was confirmed on CTA when an absence of ves‑
sel contrast enhancement was noticed accord‑
ing to the seven‑point score outlined by Dupas
et al.16 Since 2011, the analysis has been limited
to the four‑point CTA score defined by Fram‑
pas et al.,7 as recommended by the most recent
guidelines.6 CTAs showing a CCA according
to the seven‑score or the four‑score scale were
defined as “conclusive CTA”. CTAs showing a
residual cerebral perfusion in at least one vessel
were classified as “inconclusive CTA”. Patients
were allocated into two subgroups depending
on whether one CTA only was necessary to con‑
firm BD or at least one CTA was inconclusive
during the BD confirmation procedure.
Data collection
Cases were identified through the Organs
and Tissues Harvesting Department database.
1182
All data were obtained from medical records
and patient charts. Baseline patient character‑
istics were collected, including demographics,
causes for ICU admission, cause of death, and
ICU length of stay (LoS). Variables recorded
regarding BD diagnosis were: completeness of
brainstem reflex assessment, AT achievement,
and ancillary tests (e.g. EEG and/or CTA). We
reviewed the time interval between the seda‑
tion withdrawal and the clinical BD diagnosis,
the time interval between the clinical BD di‑
agnosis and the BD confirmation with an an‑
cillary test, and the time interval between the
discontinuation of sedation and the final BD
declaration.
Statistical analysis
The main outcome was the rate of con‑
clusive CTA over time (in other words, the
probability to confirm BD with CTA). Un‑
fortunately, this probability has never been
reported before, nor was the proportion of
CTA to the total of ancillary tests. Therefore
the needed sample size was impossible to de‑
termine. We assumed that a minimum of 100
confirmed BD cases using CTA would allow
performing parametric statistical tests consid‑
ering the missing data expected to be around
5%. Data were analyzed using XLSTAT
2014.3.03 software and R software (R De‑
velopment Core Team 2008, Vienna, Austria.
http://www.R‑project.org). Categorical vari‑
ables were reported as proportions and con‑
tinuous variables were reported as medians
and interquartile ranges (IQR, 25‑75th percen‑
tile). Statistical analysis used the chi‑squared
test or Fisher’s exact test for categorical vari‑
ables, Mann‑Whitney test for comparisons
between two groups and Kruskal‑Wallis Rank
Sum Test for comparison among three groups.
A p‑value of 0.05 or less was considered sig‑
nificant. The Kaplan‑Meier estimator includ‑
ed all the performed CTAs in order measure
the proportion of conclusive CTAs (Y‑axis)
over time (X‑axis). The time of clinical BD
diagnosis was considered as the origin of the
X‑axis.
Minerva AnestesiologicaNovember 2016
OPTIMAL TIMING FOR CTA IN THE CONFIRMATION OF BD
Results
KERHUEL
BD diagnosis
Study population
During the study period, 104 out of 190 clin‑
ically brain‑dead patients were included (Fig‑
ure 1). The sex ratio M/F was 1.8 and the me‑
dian age was 55 years (41‑64). Demographic
data are reported in Table I. The most common
causes for ICU admission were brain trauma
(36 cases, 35%), anoxic encephalopathy (22
cases, 21%) and aneurysmal subarachnoid
hemorrhage (20 cases, 19%). The median ICU
length of stay was 3 days (2‑6).
The median time interval from sedation
withdrawal to clinical BD diagnosis was 9
hours (6‑13). A complete examination of brain‑
stem reflexes was not possible in nine cases
(9%). AT was contra‑indicated or prematurely
stopped in 46 cases (44%) owing to thoracic
trauma, ventilator acquired pneumonia and/or
pre‑existing medical condition such as chronic
obstructive lung disease or respiratory failure.
CTA or EEG confirmed BD in 102 patients.
CTA accounted for 90% of final BD diagnosis
(94 patients). When a first CTA did not confirm
Figure 1.—The diagram shows different brain death confirmation strategies. BD: Brain death; CTA: computed tomography
angiography; EEG: electroencephalogram; ICU: Intensive Care Unit.
Vol. 82 - No. 11
Minerva Anestesiologica
1183
KERHUEL
OPTIMAL TIMING FOR CTA IN THE CONFIRMATION OF BD
Table I.—Demographic data.
Table II.—Characteristics of brain death diagnosis.
Variables
All patients
(N.=104)
Age (years)
Sex ratio (male/female)
Reasons for ICU admission
–– Stroke
–– Cardiac arrest
–– Household accident
–– Road accident
–– Suicide attempt
–– Other
Causes of death
–– Traumatic brain injury
–– Anoxic encephalopathy
–– Aneurysmal subarachnoid haemorrhage
–– Haemorrhagic stroke
–– Ischaemic stroke
–– Brain infection
Length of stay in the ICU (days)
55 [41‑64]
1.8 (67/37)
42 (40)
19 (18)
16 (15)
15 (14)
9 (9)
3 (3)
36 (35)
22 (21)
20 (19)
13 (12)
11 (11)
2 (2)
3 [2‑6]
Data are reported as No (%) or median [IQR]. No: number; IQR:
interquartile range; BD: brain death; ICU: Intensive Care Unit.
BD, a second CTA was performed in 2 cases
out of 3, otherwise EEG was chosen (Figure 1
and Table II). A withdrawal of life support was
decided in 2 patients before an ancillary test
could confirm BD. CTA sensitivity was not
significantly different before and after 2011
(80% and 81% respectively; P=0.96).
Inconclusive CTA
One hundred and seventeen CTAs were
conducted on the 104 clinically brain‑dead
patients. Ninety‑four CTAs (80%) were con‑
clusive for BD, 21 CTAs (18%) showed a per‑
sistent cerebral blood flow in at least one ves‑
sel (e.g. inconclusive CTA), and 2 CTAs (2%)
were not interpretable because of artefacts.
Two inconclusive CTAs were reported for the
same patient. The main causes of inconclusive
CTAs were residual opacification of at least
one M4‑segment of the middle cerebral artery
(47% of inconclusive CTAs) or of the internal
cerebral vein (34% of inconclusive CTAs) or
both (Figure 2).
In 12 patients out of the 94 patients who un‑
derwent CTA for the final BD confirmation, it
was not possible to found the exact hour of BD
clinical diagnosis. So, only 82 patients were
included for the analysis of CTA performances
1184
All patients
(N.=104)
Variables
Clinical BD determination
Time from sedation withdrawal to clinical
BD diagnosis (hours)
Complete clinical examination not being
possible
Apnoea test failure
Final ancillary test for BD confirmation
Time from clinical BD diagnosis to confir‑
mation with an ancillary test (hours)
Ancillary tests
–– CTA confirmed BD
–– EEG confirmed BD
–– Inconclusive
Time from sedation withdrawal to final BD
declaration (hours)
9 [6‑13]
9 (9)
46 (44)
6 [2‑11]
94 (90)
8 (8)
2 (2)
21 [12‑35]
Data are reported as No (%) or median [IQR]. No: number; IQR:
interquartile range; BD: brain death; CTA: computed tomography
angiography; EEG: electroencephalography.
Artifact,
N.=2
6%
Not available,
N.=4
13%
Internal cerebral
vein,
N.=11
34%
Terminal segment of
the middle cerebral artery,
N.=15
47%
Great cerebral vein,
N.=0
0%
Figure 2.—Frequency of opacification for each vessel in case
of inconclusive computed tomography angiography (CTA).
According to guidelines 7 a four‑vessel CTA interpretation
model is used in France to confirm cerebral circulatory arrest.
It relies on the lack of opacification of the M4 segment of both
middle cerebral arteries that is to say the terminal segment
for the cortex (M4‑MCA) and also the lack of opacification
of the internal cerebral vein (ICV) and the great cerebral vein
(GCV). This figure represents the reason why CTAs were con‑
sidered as inconclusive in 22 patients. Two inconclusive CTAs
were reported for the same patient so the results of 23 CTAs
are showed. Residual opacification of both the M4‑MCA and
the ICV was found in 9 CTAs. Technical artifact precluded the
diagnosis of brain death in 2 cases. NA: not available.
to confirm BD according to the delay from
clinical BD (Figures 3, 4). The time needed to
achieve BD confirmation in 50% of the popula‑
Minerva AnestesiologicaNovember 2016
OPTIMAL TIMING FOR CTA IN THE CONFIRMATION OF BD
Figure 3.—Survival curve describing the probability of con‑
clusive CTA over time. The survival function describes the
probability for each case to reach a final state (e.g. cerebral
circulatory arrest confirmed by CTA) from an initial state (e.g.
clinical brain death). The X‑axis represents the time between
clinical brain death diagnosis (whether an apneoa testing was
performed or not) and conclusive CTA. Dotted lines represent
the lower and upper limits of the 95 % confidence interval of
the survival function. Full line represents the study population.
Abbreviation: CTA: Computed Tomography Angiography.
KERHUEL
tion using CTA was 5.5 hours (4‑7.5). Six hours
after clinical BD, the proportion of conclusive
CTA was 51%. Moreover, 12 hours after BD
clinical diagnosis, CTA were conclusive in 65
cases (79%) and this proportion reached 85%
and 89% respectively after 20 and 23.5 hours.
CTA sensitivity was significantly associated
with the time lapse from clinical diagnosis to
CTA examination (P=0.007) when dividing
our cohort into three subgroups based on this
time lapse (Table III).
Univariate analysis showed that inconclu‑
sive CTAs were acquired significantly ear‑
lier than conclusive CTAs (2 hours (1‑3) vs. 4
hours (2‑9), P=0.008). The failure to complete
full neurological examination (P=0.02) and
decompressive craniectomy (P=0.05) were
associated with the occurrence of at least one
inconclusive CTA (Table IV). The time from
sedation withdrawal to final BD confirmation
was significantly longer in the subgroup with
at least one inconclusive CTA than in the con‑
clusive CTA group (48 hours (34‑84) vs. 17
Figure 4.—Computed tomography angiography sensitivity over time for brain death confirmation. This figure is derived from
the survival analysis and represents the cumulated sensitivity of computed tomography angiography (CTA) over time. The
X‑axis represents the time (expressed in hours) between clinical brain death diagnosis and the first CTA (if more than one
were performed) showing cerebral circulatory arrest. The Y‑axis represents CTA sensitivity (expressed in percents) mean‑
ing the proportion of conclusive CTAs to the total number of CTAs performed at each time points. The statistical analysis
included data from the 82 patients with a precise time for brain death clinical diagnosis and who underwent CTA for the final
brain death confirmation.
Vol. 82 - No. 11
Minerva Anestesiologica
1185
KERHUEL
OPTIMAL TIMING FOR CTA IN THE CONFIRMATION OF BD
Table III.—Effect of the delay from brain death on computed tomography angiography performance to confirm
brain death.
Time lapse (h)
Total number of CTA performed
–– Conclusive CTA
–– Inconclusive CTA
Sensitivity (%)
Group 1
0‑6 hrs
Group 2
6‑12 hrs
Group 3
After 12 hrs
49
42
7
86
25
23
2
92
19
17
2
89
The statistical analysis included data from the 82 patients with a precise time for brain death clinical diagnosis and who underwent CTA for the
final brain death confirmation. The three groups are significantly different in terms of proportion of conclusive and inconclusive CTAs using
Kruskal Wallis Sum Test (group 1 vs. group 2 vs. group 3, P=0.007). h: hour; CTA: computed tomography angiography.
Table IV.—Bivariate analysis of the association of inconclusive CTA occurrence with demographics and brain death
diagnosis characteristics.
No
inconclusive CTA
during BD assessment
(N.=82)
Variables
Demographics
Male gender
Age (years)
Cause of death
–– Brain trauma
–– Brain anoxia
–– Ruptured brain aneurysm
–– Hemorrhagic stroke
–– Ischemic stroke
–– Brain infection
BD diagnosis
Not able to complete clinical examination
Apnoea testing failure
Decompressive Craniectomy
Time from sedation withdrawal to clinical BD diagnosis (hours)
Time from clinical BD diagnosis to first CTA acquisition (hours)
54 (66)
55 [42‑63]
1 or more inconclusive
CTA during BD
assessment
(N.=22)
13 (57)
54 [41‑63]
P value
0.62
0.83
31 (38)
17 (21)
14 (17)
10 (12)
9 (11)
1 (1)
5 (23)
5 (23)
6 (27)
3 (14)
2 (9)
1 (4)
0.56
4 (5)
38 (46)
6 (7)
9 [5‑14]
4 [2‑9]
5 (23)
8 (36)
5 (23)
9 [3‑22]
2 [1‑3]
0.02
0.47
0.05
0.89
0.008
Data are reported as No (%) or median [IQR]. No: number. IQR: interquartile range. BD: brain death. CTA: computed tomography angiog‑
raphy.
hours (9‑28), respectively; P<0.001). The con‑
sent for organ donation of the patients’ rela‑
tives decreased from 55% to 45% if an incon‑
clusive CTA occurred during the diagnostic
procedure without reaching statistical signifi‑
cance (P=0.43).
Discussion
The main result of this study is that the rec‑
ommended time lapse of 6 hours from clini‑
cal BD seems to be insufficient since the pro‑
portion of conclusive CTA is only 51%, with
progressive increase overtime with more than
80% of conclusive CTA after 12 hours. Incon‑
clusive CTAs (e.g. persistent cerebral flow) are
associated with a shorter time from clinical BD
1186
diagnosis. The duration of the whole BD diag‑
nostic procedure was almost three times longer
when an inconclusive CTA occurred. These re‑
sults highlight the fact that performing a CTA
very early after BD significantly increase the
likelihood of having a inconclusive CTA and
increase the total time needed to confirm BD,
possibly affecting relatives compliance for or‑
gan donation and grafts quality.17
Previous observational studies have fo‑
cused on time interval between clinical ex‑
amination and EEG: the median duration of
the BD procedure varied from 4 hours to 27
hours.18‑20 Part of this variability was linked to
differences between the diagnostic procedures
across countries. For example, in a recent
Spanish cohort study of 289 patients, Fernan‑
Minerva AnestesiologicaNovember 2016
OPTIMAL TIMING FOR CTA IN THE CONFIRMATION OF BD
dez et al. reported a 3.5% rate of inconclusive
EEG, and a time to BD diagnosis that varied
strongly among patients.19 Notably, according
to the Spanish legislation, an ancillary test is
optional and may be used to shorten the rec‑
ommended six‑hour period of clinical observa‑
tion. In a recent US multicenter retrospective
study of 1311 patients, Lustbader et al. found
that the median duration time for BD diagnosis
was 19 hours.18 In the present study, the me‑
dian time from clinical BD diagnosis to CTA
is relatively short (6 hours). BD confirmation
was obtained for 47 patients (40%) in the first
six hours. Five CTAs that were performed be‑
yond the six‑hour time were inconclusive, and
only 2 CTAs were still inconclusive beyond a
12‑hour time point. In the case of a 12‑hour in‑
terval, the rate of inconclusive CTA decreased
to 2% (2 CTAs out of 104 patients).
Clinical experience shows that an intensiv‑
ist faced with an inconclusive CTA is used to
postponing the additional ancillary test to the
next day, in order not to accumulate inconclu‑
sive results; this strategy may partly explain
the time extension. Interestingly, Orban et al. 21
have published a prospective controlled study
showing that the transcranial Doppler (TCD)
was useful for identifying the precise time of
brain circulation cessation, thus decreasing the
time to CTA acquisition with a low risk of in‑
conclusive results.
Univariate analysis confirmed that a decom‑
pressive craniectomy, complete clinical exami‑
nation not being possible, and an early CTA
acquisition were significantly associated with
the occurrence of an inconclusive CTA. These
factors had already been described in expert
opinions,9 but to our knowledge this is the first
quantitative analysis of risk factors for incon‑
clusive CTA ever reported. In our study, AT was
contra‑indicated or prematurely stopped in 44%
of cases, a proportion twice as high as that re‑
ported in the literature.22 Our recruitment of se‑
verely injured patients with a high prevalence of
chest trauma or ventilator‑acquired pneumonia
may partly explain this result. Cause of death
was not a significant risk factor for inconclusive
CTA, although a primary anoxic lesion associ‑
ated with brain perfusion has been previously
Vol. 82 - No. 11
KERHUEL
reported.23 However, this may be in relation
with lack of power given the small number of
patients in this subgroup of the present study.
Limitations of the study
The present study has several limitations.
Firstly, the recruitment of the patients through
the Organs and Tissues Harvesting Department
database may have led to selection bias. It is
possible that several patients presenting with
clinical BD were not reported to the department
when the intensivist considered the patient not
eligible for organ donation. Secondly, 51 pa‑
tients whose BD was confirmed by EEG only
were excluded from the study. It is possible that
some of these patients had a risk factor for an
inconclusive CTA, guiding the choice of the
intensivist to EEG. This exclusion could have
influenced demographic data and have under‑
estimated the proportion of inconclusive CTAs.
Finally the replacement of the seven‑point score
with a four‑point CTA score in 2011 according
to the French guidelines 6 may have influenced
the rate of inconclusive CTAs. However, the
reported sensitivity was not significantly dif‑
ferent before and after 2011. The impact of the
variability in interpretation of CTA by radiolo‑
gists is probably low, since a study published
in 2009 by Sawicki et al. showed an 89% and
95% agreement in the inter‑radiological inter‑
pretation regarding the seven‑point score and
the four‑point score, respectively.24
Conclusions
The recommended six‑hour interval between
clinical BD diagnosis and CTA is probably in‑
sufficient, as a 12‑hour interval might be more
appropriate in order to limit the risk of incon‑
clusive CTA. This study reveals sources of in‑
conclusive CTA (decompressive craniectomy,
incomplete neurological examination, CTA ac‑
quisition before 6 hours after clinical BD). The
choice of an optimal timing should nevertheless
also take into account the medical complications
and the ethical issues that may occur in the case
of an extended time interval. These results have
to be confirmed in a larger prospective study af‑
Minerva Anestesiologica
1187
KERHUEL
OPTIMAL TIMING FOR CTA IN THE CONFIRMATION OF BD
ter exclusion of the cases for which brainstem
reflexes cannot be completely assessed.
Key messages
—— The recommended time lapse of 6
hours between clinical brain death and
computed tomography angiography is
probably insufficient.
—— The optimal time for CT angiog‑
raphy from clinical signs of brain death
might be 12 hours.
—— Decompressive craniectomy, incom‑
plete clinical brain stem assessment and the
timing of computed tomography angiogra‑
phy are predictors of inconclusive CTAs
(e.g. false negative CTAs).
References
  1. Citerio G, Crippa IA, Bronco A, Vargiolu A, Smith M.
Variability in brain death determination in europe: look‑
ing for a solution. Neurocrit Care 2014;21:376‑82.
  2. [No authors listed]. Décret relatif au constat de la mort
préalable a prélèvement d’organes, de tissus et de cellules
à des fins thérapeutiques ou scientifiques et modifiant le
code de la santé publique (deuxième partie: Décrets en
Conseil d’Etat). JORF n°282 du 4 décembre 1996; 1996.
  3. Gardiner D, Shemie S, Manara A, Opdam H. Interna‑
tional perspective on the diagnosis of death. Br J Anaesth
2012;108 Suppl 1:i14‑28.
  4. Wijdicks EF, Varelas PN, Gronseth GS, Greer DM,
American Academy Of N. Evidence‑based guideline
update: determining brain death in adults: report of the
Quality Standards Subcommittee of the American Acad‑
emy of Neurology. Neurology 2010;74:1911‑8.
  5. Wijdicks EF. The diagnosis of brain death. N Engl J Med
2001;344:1215‑21.
  6. Societe Francaise De N, Societe Francaise De R, Agence
De La B. [Recommendations on diagnostic criteria of
brain death by the technique of CT angiography]. J Neu‑
roradiol 2011;38:36‑9.
  7. Frampas E, Videcoq M, De Kerviler E, Ricolfi F, Kuoch
V, Mourey F, et al. CT angiography for brain death diag‑
nosis. AJNR Am J Neuroradiol 2009;30:1566‑70.
  8. Sawicki M, Bohatyrewicz R, Walecka A, Solek‑Pastusz‑
ka J, Rowinski O, Walecki J. CT Angiography in the Di‑
agnosis of Brain Death. Pol J Radiol 2014;79:417‑21.
  9. Flowers WM, Jr., Patel BR. Persistence of cerebral blood
flow after brain death. South Med J 2000;93:364‑70.
10. Rieke A, Regli B, Mattle HP, Brekenfeld C, Gralla J,
Schroth G, et al. Computed tomography angiography
(CTA) to prove circulatory arrest for the diagnosis of
brain death in the context of organ transplantation. Swiss
Med Wkly 2011;141:w13261.
11. Bohatyrewicz R, Sawicki M, Walecka A, Walecki J,
Rowinski O, Bohatyrewicz A, et al. Computed tomo‑
graphic angiography and perfusion in the diagnosis of
brain death. Transplant Proc 2010;42:3941‑6.
12. Escudero D, Otero J, Marques L, Parra D, Gonzalo JA,
Albaiceta GM, et al. Diagnosing brain death by CT per‑
fusion and multislice CT angiography. Neurocrit Care
2009;11:261‑71.
13. Leclerc X, Taschner CA, Vidal A, Strecker G, Savage J,
Gauvrit JY, et al. The role of spiral CT for the assessment
of the intracranial circulation in suspected brain‑death. J
Neuroradiol 2006;33:90‑5.
14. Welschehold S, Kerz T, Boor S, Reuland K, Thomke F,
Reuland A, et al. Detection of intracranial circulatory ar‑
rest in brain death using cranial CT‑angiography. Eur J
Neurol 2013;20:173‑9.
15. Société Française D’anesthésie Réanimation ADLB, So‑
ciété Française De Réanimation De Langue Française,
Conférence D’experts. Prise en charge des sujets en état
de mort encéphalique dans l’optique d’un prélèvement
d’organes. 2003.
16. Dupas B, Gayet‑Delacroix M, Villers D, Antonioli D,
Veccherini MF, Soulillou JP. Diagnosis of brain death
using two‑phase spiral CT. AJNR Am J Neuroradiol
1998;19:641‑7.
17. Inaba K, Branco BC, Lam L, Salim A, Talving P, Plurad
D, et al. Organ donation and time to procurement: late is
not too late. J Trauma 2010;68:1362‑6.
18. Lustbader D, O’hara D, Wijdicks EF, Maclean L, Tajik W,
Ying A, et al. Second brain death examination may nega‑
tively affect organ donation. Neurology 2011;76:119‑24.
19. Fernandez‑Torre JL, Hernandez‑Hernandez MA, Mu‑
noz‑Esteban C. Non confirmatory electroencephalogra‑
phy in patients meeting clinical criteria for brain death:
scenario and impact on organ donation. Clin Neurophysi‑
ol 2013;124:2362‑7.
20. Vicenzini E, Pro S, Pulitano P, Rocco M, Spadetta G,
Zarabla A, et al. Current practice of brain death deter‑
mination and use of confirmatory tests in an Italian Uni‑
versity hospital: a report of 66 cases. Minerva Anestesiol
2013;79:485‑91.
21. Orban JC, El‑Mahjoub A, Rami L, Jambou P, Ichai C.
Transcranial Doppler shortens the time between clinical
brain death and angiographic confirmation: a randomized
trial. Transplantation 2012;94:585‑8.
22. Goudreau JL, Wijdicks EF, Emery SF. Complications
during apnea testing in the determination of brain death:
predisposing factors. Neurology 2000;55:1045‑8.
23. Ala TA, Kuhn MJ, Johnson AJ. A case meeting clini‑
cal brain death criteria with residual cerebral perfusion.
AJNR Am J Neuroradiol 2006;27:1805‑6.
24. Sawicki M, Bohatyrewicz R, Safranow K, Walecka A,
Walecki J, Rowinski O, et al. Computed tomographic an‑
giography criteria in the diagnosis of brain death‑compar‑
ison of sensitivity and interobserver reliability of differ‑
ent evaluation scales. Neuroradiology 2014;56:609‑20.
Authors’ contributions.—Lionel Kerhuel, Mohamed Srairi, and Thomas Geeraerts designed the study. Lionel Kerhuel collected
the data and performed the statistical analysis. All others participated in collecting the data. Lionel Kerhuel, Mohamed Srairi, and
Thomas Geeraerts wrote the paper. All authors read and approved the final manuscript.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material
discussed in the manuscript.
Article first published online: September 13, 2016. - Manuscript accepted: September 9, 2016. - Manuscript revised: September 6,
2016. - Manuscript received: March 20, 2016.
1188
Minerva AnestesiologicaNovember 2016
© 2016 EDIZIONI MINERVA MEDICA
Online version at http://www.minervamedica.it
Minerva Anestesiologica 2016 November;82(11):1189-98
ORIGINAL ARTICLE
Serum S100β as a prognostic marker in patients
with non-traumatic intracranial hemorrhage
Eija K. JUNTTILA* 1, 2, 3, Juha KOSKENKARI 2, 3, Pasi P. OHTONEN 4,
Ari KARTTUNEN 5, Tero I. ALA-KOKKO 2, 3
1Department
of Anesthesiology, Tampere University Hospital, Tampere, Finland; 2Division of Intensive Care,
Department of Anesthesiology, Oulu University Hospital, Medical Research Center Oulu, Oulu, Finland; 3Research
Unit of Surgery, Anesthesiology and Intensive Care, Oulu University, Oulu, Finland; 4Department of Anesthesiology
and Surgery, Oulu University Hospital, Oulu, Finland; 5Department of Radiology, Oulu University Hospital, Oulu,
Finland
*Corresponding author: Eija Junttila, Department of Anesthesiology, Tampere University Hospital, PO BOX 2000, 33521 Tampere,
Finland. E-mail: [email protected]
ABSTRACT
BACKGROUND: The serum concentration of S100β protein reportedly predicts outcomes after brain injury. We examined the prognostic accuracy of S100β in patients with non-traumatic intracranial hemorrhage.
METHODS: This was a prospective, observational study of patients with non-traumatic intracranial hemorrhage treated
in the intensive care unit at our university hospital. Computed tomography imaging findings and the level of consciousness on admission were recorded. Serum S100β concentration was measured serially during the first six days of admission. Patients with subarachnoid hemorrhage (SAH group) or intracerebral hemorrhage (ICH group) were analyzed separately. The 3-month and 1-year functional outcomes were assessed using the Glasgow Outcome Scale (GOS).
RESULTS: Of 108 patients enrolled, 66 were included in the SAH group and 42 in the ICH group. High initial S100β
concentration was associated with Glasgow Coma Score 3-6 on admission (SAH group 0.61 μg/L versus 0.15 μg/L,
P=0.001 and ICH group 1.00 μg/L versus 0.42 μg/L, P=0.005). Initial S100β concentration correlated with ICH volume
(rho=0.50, P<0.001) and IVH Sum Score (rho=0.30, P=0.013). The thresholds for the initial S100β concentration with
100% specificity for poor outcome (GOS 1-3) were 1.40 μg/L for SAH and 1.76 μg/L for ICH group. ORs varied between
3.1 and 6.1 for S100β on poor outcome in the SAH group. Increasing S100β level during study period was associated
with poor outcome in the SAH group.
CONCLUSIONS: Serum S100β concentration corresponds with the severity of neurological insult and predicts poor
outcome in patients with non-traumatic intracranial hemorrhage.
(Cite this article as: Junttila EK, Koskenkari J, Ohtonen PP, Karttunen A, Ala-Kokko TI. Serum S100β as a prognostic marker
in patients with non-traumatic intracranial hemorrhage. Minerva Amestesiol 2016;82:1189-98)
Key words: Glasgow Outcome Scale - Intracranial hemorrhages - S100 Calcium Binding Protein beta Subunit.
N
on-traumatic intracranial hemorrhage has
high mortality and morbidity.1 The level
of consciousness and the computed tomography (CT) imaging findings in the acute phase
are strong predictors of outcome.2-5 Predicting
outcome during intensive care unit (ICU) stay
remains challenging.
One prognostic marker of interest has been
S100β protein, with the highest concentrations
in astrocytes, oligodendrocytes and Schwann
Vol. 82 - No. 11
cells.6 Necrosis of these brain parenchymal
cells releases S100β into the extracellular fluid,
which can pass through the disrupted bloodbrain barrier to the systemic circulation.6 Although the association between elevated serum
S100β concentration and worse outcome has
been reported in patients with different types
of intracranial hemorrhage,7-9 few have considered S100β as a long-term predictor of outcome. In patients with SAH there have been
Minerva Anestesiologica
1189
JUNTTILA SERUM S100β AS A PROGNOSTIC MARKER IN PATIENTS WITH NON-TRAUMATIC INTRACRANIAL HEMORRHAGE
only few studies that have examined 1-year
functional outcome as a primary endpoint 8, 10
and in patients with ICH there has been none.
The aims of this study were to examine the
association between serum S100β concentration and the level of consciousness and CT imaging findings on admission, and to evaluate
the predictive value of serum S100β concentration for outcome 3 months and 1 year after in patients with non-traumatic intracranial
hemorrhage.
Materials and methods
Patients and clinical management
This was a subsidiary study using data collected from a prospectively enrolled cohort of
consecutive patients admitted with non-traumatic intracranial hemorrhage to the tertiary
level ICU of our institution between December
2007 and December 2009, the details of which
are described in more detail elsewhere.11, 12
This study was approved by the Ethics Committee of the Oulu University Hospital. Informed consent was obtained from the patient
or a legal surrogate in all cases.
Patient characteristics including age, sex
and medical history were recorded. The level
of consciousness on hospital admission (or
before sedation, if earlier) was assessed using
the Glasgow Coma Scale (GCS) 13 and categorized into one of four groups: GCS 15; 13-14;
7-12 and 3-6. The data during the ICU stay
was prospectively collected into an electronic
database (Critical Care Information Management System, CCIMS, Clinisoft, GE, Helsinki,
Finland).
The diagnosis of intracranial hemorrhage
was established by CT imaging of the head
on admission. Images were retrospectively reviewed by a neuroradiologist (AK); SAH and
IVH were judged against the Hijdra SAH CT
score (SAH Sum Score 0-30 and IVH Sum
Score 0-12) 14 and ICH volume (mL) was calculated ([d1×d2×d3]/2, [three perpendicular
diameters, d]). The presence or absence of diffuse cerebral edema, acute hydrocephalus and
midline shift were also noted. The etiology
1190
of the hemorrhage was recorded based on the
data from head CT scans, CT-angiographies,
digital subtraction angiographies, operation
and autopsy reports. Based on radiological
findings patients were divided into two groups:
the SAH group included patients with aneurysmatic SAH and/or IVH/ICH or perimesencephalic SAH, while the ICH group was comprised of patients with primary and secondary
ICH and/or IVH, which may have enlarged to
the subarachnoid space.
Criteria for ICU admission were the need
for neurosurgical or endovascular intervention
and/or derangement of physiologic functions
in patients judged by the treating physician
to have a credible chance of survival. During
the ICU stay, all patients were treated according to our normal clinical practice, consistent
with latest guidelines 15-17 and as previously
described.11, 12
Blood samples for determination of serum
S100β concentration
Each patient’s participation in the study
commenced on the day of ICU admission (day
0) and lasted for a further 5 days (days 1-5),
unless the patient died or was transferred to another hospital. Study period was divided into
three 2-day sections: days 0-1, days 2-3 and
days 4-5. The first arterial blood sample for
serum protein S100β analysis was taken after
a consent was obtained shortly after recruitment on day 0 or 1, and thereafter once in every 2-day period. Serum protein S100β specimen collection, preparation and analysis were
carried out according to the manufacturer’s
instructions. Blood for serum specimen was
collected using standard serum sampling tubes
and preserved one hour in room temperature
before 10 minutes centrifuging to separate
blood cells and serum. S100β was measured
using an immunochemiluminometric method
(Elecsys 2010, Roche Diagnostics, Roche
Diagnostics GmbH, Mannheim, Germany).
The initial (on day 0 or 1), and the maximum
(between days 0 and 5) S100β concentrations
were used for analysis. In patients with two to
three S100β values trends in blood concentra-
Minerva Anestesiologica
November 2016
SERUM S100Β AS A PROGNOSTIC MARKER IN PATIENTS WITH NON-TRAUMATIC INTRACRANIAL HEMORRHAGE JUNTTILA
tions were categorized as 1) increasing or 2)
decreasing or unchanged.
Outcomes
The 3-month and 1-year outcome was assessed using the Glasgow Outcome Scale
(GOS). (18) Data were collected by means of a
telephone interview by one of the authors (EJ)
or a trained research assistant (MV) with either
the patient or carer. Outcome was categorized
as good (GOS 4–5: no disability to moderate
disability) or poor (GOS 1–3: severe disability,
vegetative state, death).
Statistical analysis
SPSS (IBM Corp. Released 2012. IBM
SPSS Statistics for Windows, Version 21.0.
Armonk, NY: IBM Corp.) software was used
for statistical analysis. Summary measurements are expressed as a mean with standard
deviation (SD) or as a median with 25th–75th
percentile, unless otherwise stated. Continuous variables were analyzed using Student’s
t-test or the Mann-Whitney U-test, the latter
for non-normally distributed data. The χ² test
or Fisher’s exact test were used for categorical
variables. Spearman’s correlation coefficient
(rho) was used to assess correlations between
continuous variables. The area under the receiver operating characteristic (ROC) curve
(AUC) was used as a measure of the diagnostic
accuracy of S100β in predicting poor outcome.
The sensitivity and specificity were calculated
with exact 95% binominal confidence intervals. The lowest S100β value giving 100%
specificity was chosen as a threshold value
with the idea that with a highly specific test a
positive test result one could be almost certain
of the presence of the outcome, ie. GOS 1-3 in
our study.
A multivariable logistic regression model
was built for SAH group using GOS 1-3 at
3-month as outcome variable. A maximum
of two adjusting covariates was used due to
small data (SAH group patients with GOS 1-3,
N.=31) to assess the impact of S100β value on
GOS, thus the variable S100β was forced into
Vol. 82 - No. 11
each model. The adjusting covariates used in
logistic regression model were age, sex, aneurysmatic bleeding, comorbidity, diffuse brain
edema, ����������������������������������
ventriculostomy, neurogenic pulmonary edema, ischemic electrocardiography abnormalities, craniotomy, SAH sum score, IVH
sum score and ICH volume. Since the linearity
assumption of S100β did not hold, the S100β
was categorized into two classes according
to median value (≤0.225 μg/L versus >0.225
μg/L). Of the adjusting factors age was used as
continuous variable and all other variables as
dichotomous variables. Adjusting factors were
chosen according to clinical interest or having
P value <0.2 in univariable model. The results
of these models are presented as minimum and
maximum ORs. Multivariable logistic regression model was no created for ICH group patients due the small number of GOS 4-5 in this
group (N.=7).
Two-tailed P values were reported and
P<0.05 was considered statistically significant.
Results
During the 2-year study period, 191 patients
with non-traumatic intracranial hemorrhage
were treated in our ICU, 108 of whom were included in our analysis. The detailed flowchart
of this study is presented in Figure 1. The main
reasons for exclusion were: unavailability of
study personnel in 19 cases, treatment withdrawal in 17 cases, admission delay >48 hours
in 14 cases and other reasons in 24 cases. Sixty-six patients were included to the SAH group
and 42 to the ICH group.
Patient characteristics, primary head CT
findings and the neurosurgical interventions
undertaken are presented in Table I, and in our
previous study.12 Sixty-three patients had aneurysmal hemorrhage of which 38 (60%) underwent coiling. With the exception of more
impaired level of consciousness on admission
in ICH group patients, there were no significant differences in the clinical characteristics
between the groups. Intraventricular hemorrhage and hydrocephalus were common in
both groups. Diffuse cerebral edema was more
common in SAH group and midline shift in
Minerva Anestesiologica
1191
JUNTTILA SERUM S100β AS A PROGNOSTIC MARKER IN PATIENTS WITH NON-TRAUMATIC INTRACRANIAL HEMORRHAGE
ICU admission for non-traumatic
intracranial hemorrhage
N.=191
Excluded, N.=81
age <18 y, N.=2
admission delay >48 h, N.=14
AVM bleeding, N.=2
decision to withdraw from the active
treatment before recruitment, N.=17
decision to transfer to the ward before
recruitment, N.=8
consent was not obtained, N.=7
study personnel absent, N.=19
other reason, N.=12
Eligible for the study
N.=110
Excluded after review
N.=2
AVM bleeding, N.=1
tumor bleeding, N.=1
Analyzed
N.=108
SAH group
N.=66
Follow-up interview
N.=54
ICH group
N.=42
Died
N.=12
Follow-up interview
N.=30
Died
N.=12
Figure 1.—Flowchart, until 1-year follow-up.
ICH group. Intracranial pressure (ICP) monitoring and ventriculostomy were often required in both patient groups, but the need for
craniotomy was more frequent in ICH group
patients.
1192
The 3-month and 1-year GOS are presented
in Table II. Mortality rates were 15% and 18%
in the SAH group and 26% and 29% in the ICH
group, respectively. Functional outcome data
were available for all 84 patients alive 1 year
Minerva Anestesiologica
November 2016
SERUM S100Β AS A PROGNOSTIC MARKER IN PATIENTS WITH NON-TRAUMATIC INTRACRANIAL HEMORRHAGE JUNTTILA
Table I.—Clinical characteristics, initial imaging findings and neurosurgical interventions undertaken according
to the study group.
Clinical characteristics
Age (Y), mean (range)
Sex, male, N. (%)
Co-morbidities, N. (%)
The level of consciousness on admission, N. (%)
GCS 15
GCS 13-14
GCS 7-12
GCS 3-6
APACHE II, mean (SD)
Primary head CT findings
Presence of SAH, N. (%)
SAH Sum Score, med (25th-75th)
Presence of IVH, N. (%)
IVH Sum Score, med (25th-75th)
Presence of ICH, N. (%)
ICH volume, med (25th-75th)
Hydrocephalus, N. (%)
Diffuse brain edema, N. (%)
Midline shift, N. (%)
Performed neurosurgical interventions
ICP monitoring, N. (%)
Ventriculostomy, N. (%)
Craniotomy*, N. (%)
SAH group
N.=66
ICH group
N.=42
57 (25-80)
34 (52)
40 (61)
57 (28-75)
22 (52)
26 (62)
19 (29)
21 (32)
8 (12)
18 (27)
19.5 (6.6)
62 (94)
17 (10-22)
41 (62)
5 (2-6.5)
22 (33)
29 (5-46.25)
34 (52)
31 (47)
18 (27)
46 (70)
42 (64)
24 (36)
2 (5)
10 (24)
13 (31)
17 (41)
20.9 (6.2)
6 (14)
4.5 (1.75-11.75)
27 (64)
10 (2-10)
40 (95)
46.5 (17.75-82.5)
21 (50)
8 (19)
26 (62)
31 (74)
25 (60)
27 (64)
SAH: subarachnoid hemorrhage; IVH: intraventricular hemorrhage; ICH: intracerebral hemorrhage; GCS: Glasgow Coma Scale; APACHE:
Acute Physiology And Chronic Health Evaluation; ICP: intracranial pressure.
*Craniotomy for aneurysm clipping and/or hematoma evacuation.
Table II.—Functional outcome at 3 months and 1 year
assessed by the Glasgow Outcome Scale according to
the study group.
3-month GOS
Good, N. (%)
5 (good recovery)
4 (moderate disability)
Poor, N. (%)
3 (severe disability)
2 (vegetative state)
1 (dead)
1-year GOS, N. (%)
Good, N. (%)
5 (good recovery)
4 (moderate disability)
Poor, N. (%)
3 (severe disability)
2 (vegetative state)
1 (dead)
SAH group
N.=66
ICH group
N.=42
35 (53)
22
13
31 (47)
20
1
10
7 (17)
2
5
35 (83)
23
1
11
39 (59)
27
12
27 (41)
14
1
12
13 (31)
6
7
29 (69)
17
0
12
GOS: Glasgow Outcome Scale; SAH: subarachnoid hemorrhage;
IVH: intraventricular hemorrhage; ICH: intracerebral hemorrhage.
Vol. 82 - No. 11
after the insult: outcomes were less favorable
in the ICH group at both 3 months and 1 year.
Between 3 months and 1 year, four patients in
the SAH group and six in the ICH group recovered sufficiently for their GOS category to
improve from poor to good.
High initial S100β concentration was associated with GCS 3-6 on admission (0.61
[0.18-1.55] μg/L versus 0.15 [0.08-0.25] μg/L,
P=0.001 in SAH group patients with and 1.00
[0.47-2.34] μg/L versus 0.42 [0.20-0.87] μg/L,
P=0.005 in ICH group patients). There was a
significant correlation between initial S100β
concentration and ICH volume (rho=0.50,
P<0.001), as well as IVH Sum Score (rho =
0.30, P=0.013). In contrast, there was no correlation between initial S100β concentration
and SAH Sum Score (rho=0.17, P=0.17). Initial S100β concentration was higher in patients
with midline shift (0.78 [0.29-1.68] μg/L ver-
Minerva Anestesiologica
1193
JUNTTILA SERUM S100β AS A PROGNOSTIC MARKER IN PATIENTS WITH NON-TRAUMATIC INTRACRANIAL HEMORRHAGE
sus 0.17 [0.09-0.44] μg/L, P<0.001), but not
in patients with diffuse brain edema or hydrocephalus (0.31 [0.15-1.11] μg/L versus 0.27
[0.12-0.82] μg/L, P=0.60 and 0.38 [0.16-1.01]
μg/L versus 0.23 [0.11-0.66] μg/L, P=0.061,
respectively). The median delay from the onset of symptoms to the ICU admission was 7
(5-11) hours. It did not correlate with initial
S100β values (Spearman´s correlation coefficient rho=0.017).
Maximum S100β concentration was higher
in patients with poor 1 year functional outcome
compared with patients with good outcome in
both SAH group (0.68 [0.20-2.22] versus 0.15
SAH group, 3-month
[0.09-0.28], P<0.001) and ICH group (0.95
[0.69-1.98] versus 0.49 [0.24-0.81], P=0.002).
The statistical powers of these results were
84.4% and 89.1%, respectively. The data with
two to three S100β measurements was available in 57 SAH group and in 38 ICH group
patients. Increasing trend in S100β concentration levels were more frequent in patients with
poor outcome (SAH group, 16/24 versus 6/33,
P<0.001 and ICH group 10/11 versus 16/27,
P=0.059). Receiver operator characteristic
curves for initial and maximum S100β concentration are shown in Figure 2. Maximum
S100β concentration produced the best AUC
ICH group, 3-month
ROC Curve
ROC Curve
S100B max d 0-1
S100B max d 0-5
Sensitivity
Sensitivity
S100B max d 0-1
S100B max d 0-5
1-Specificity
1-Specificity
SAH group, 1-year
ICH group, 1-year
ROC Curve
ROC Curve
S100B max d 0-1
S100B max d 0-5
Sensitivity
Sensitivity
S100B max d 0-1
S100B max d 0-5
1-Specificity
1-Specificity
Figure 2.—Receiver operating characteristic curves of serum S100β concentration on initial (days 0 to 1) and maximum
serum S100β concentration between days 0 and 5 predictive of poor outcome at 3-month and 1-year according to the type of
intracranial hemorrhage.
1194
Minerva Anestesiologica
November 2016
SERUM S100Β AS A PROGNOSTIC MARKER IN PATIENTS WITH NON-TRAUMATIC INTRACRANIAL HEMORRHAGE JUNTTILA
Table III.—Prognostic accuracy according to the area under the curve of receiver operator characteristic curves in
Figure 2, and thresholds for serum S100β concentration with 100% specificity and the corresponding sensitivity.
Cox-Snell R2 (R2) values are given to assess the goodness of fit of the models.
SAH group
N.=66
3-month*
S100β
days 0-1, μg/L*
days 0-5, max, μg/L
1-year
S100β
days 0-1, μg/L*
days 0-5, max, μg/L
ICH group
N.=42
AUC
95% CI
Threshold
with 100%
specificity
Sensitivity
R2
AUC
95% CI
Threshold
with 100%
specificity
Sensitivity
R2
0.80
0.81
0.69-0.91
0.71-0.92
1.40
1.47
0.21
0.29
0.28
0.28
0.75
0.74
0.53-0.98
0.50-0.97
1.76
1.76
0.21
0.23
0.07
0.09
0.74
0.77
0.62-0.83
0.66-0.89
1.40
1.47
0.24
0.33
0.17
0.26
0.71
0.80
0.54-0.88
0.64-0.95
1.76
1.76
0.24
0.28
0.14
0.22
SAH: subarachnoid hemorrhage; IVH: intraventricular hemorrhage; ICH: intracerebral hemorrhage; AUC: area under the curve; CI: confidence interval.
*N.=62 in SAH group.
in SAH group patients at 3 months and for ICH
group patients at 1 year (Table III). The threshold values for initial and maximum S100β
concentrations with 100% specificity for unfavorable outcome at 3 months and 1 year were
identical (Table III).
According to adjusted logistic regression
models the OR for the impact of S100β on
poor outcome in SAH group at three months
varied between 3.1 and 6.1 reaching statistical significance at 5% level in all tested models, except the one in which ICH volume and
ventriculostomy were adjusting factors. In
that model the OR for S100β was 3.1 (95% CI
0.87-11.1, P=0.082).
Discussion
We found that serum protein S100β concentration predicted poor outcome in both
SAH and ICH groups. Cut-off values with
100% specificity for poor outcome at both 3
months and 1 year were 1.40 μg/L for SAH
group and 1.76 μg/L for ICH group. Patients
with poor outcome had more frequently increasing S100β values during the study period.
Higher initial serum S100β concentration was
associated with diminished conscious level on
admission in both groups, and although serum
S100β concentration correlated with ICH volume and IVH Sum Score, there was no correlation with the SAH Sum Score.
Vol. 82 - No. 11
Higher serum S100β concentration was associated with poor outcome in both groups of
patients at both follow-up time points, a finding that is consistent with studies on SAH patients,8, 9, 19 although in two previous studies
outcome was only assessed after 6 months.9, 19
Delgado et al.7 reported a clear association between elevated S100β concentration and poor
outcome 3 months after ICH, but their study
population was substantially less neurologically impaired on admission than our cohort,
and those who subsequently underwent surgical intervention were excluded from their
analysis. Notably, 20% (6 out of 30) patients in
our study with high-grade ICH recovered sufficiently well to improve from the poor to good
outcome category between 3 months and 1
year, suggesting that longer-term assessments
of recovery of function may better reflect outcome in these patient populations.
In this study patients with increasing trend
in S100β levels had more frequently poor outcome. In SAH group the result was clear and
consistent with previous reports 8 despite the
short sample collection time in our study. In
ICH group it did not reach statistical significance, which might be explained just by small
study group.
Our study is the first to report thresholds for
initial and maximum S100β concentrations
giving 100% specificity for poor outcome
at both 3 months and 1 year in these patient
Minerva Anestesiologica
1195
JUNTTILA SERUM S100β AS A PROGNOSTIC MARKER IN PATIENTS WITH NON-TRAUMATIC INTRACRANIAL HEMORRHAGE
populations. It is striking that these thresholds
were identical for outcomes at both evaluation
times, supporting the validity of our results
and the usefulness of S100β as a long-term
prognostic indicator. In patients with traumatic brain injury, threshold values for serum
S100β have been shown to vary widely and are
markedly higher for unfavorable neurological
prognosis (GOS 1-3) than in this study.20 Potential explanations include more extensive
and widespread central nervous cell injury or
extracerebral release of S100β in patients with
multiple injuries. However, our AUC analyses
yielded plausible values for our patient population, and are in concordance with previous
studies.8, 21
The level of consciousness and the computed tomography (CT) imaging findings in the
acute phase are strong predictors of outcome
in these patient populations.2-5 In agreement
with previous reports,9, 21, 22 we also found that
the initial low GCS scores were associated
with elevated S100β concentration in both
SAH and ICH groups. Previous investigators
have used the Fisher Scale to report the radiological abnormalities in SAH,8, 9, 19, 22, 23 two
of which identified an association between elevated protein S100β concentration and high
Fisher grade.9, 22 We elected to use the Hijdra
SAH CT score instead of the Fisher grading
scale as it more precisely defines the site and
amount of blood. Interestingly we did not find
a correlation between S100β concentration
and the amount of subarachnoid blood defined
by the SAH Sum Score. This contrast between
our findings and those of other investigators
can likely be explained by the presence of IVH
in patients with Fisher grade 4 hemorrhage.
This conclusion is supported by reports that
the presence of IVH increases S100β concentration in patients with ICH 7 and that IVH is
associated with worse outcome in both SAH
and ICH;4, 24 in our study there was a significant correlation between S100β concentration
and IVH Sum Score. The correlation between
S100β concentration and ICH volume that we
identified concurs with previous studies and
can likely be explained by the larger parenchymal injury.7, 21
1196
Limitations of the study
Our study had some limitations. First, there
is wide variation in delays between the onset
of neurological symptoms and blood sampling
for measurement of initial serum S100β concentration; in addition the delays caused by
the patients and health care system and difficult to control, we were not able to collect
research blood samples without first gaining
the consent of the patient or the assent of a
relative. However, the delay to sampling
did not correlate with S100β levels. Second,
S100β was not measured daily. All these may
affect to the results and peak S100β values
may be missed. Third, only critically ill patients with severe ICH requiring neurosurgical
interventions were included in the study, and
care must be exercised when extrapolating our
findings to populations with less severe ICH.
Instead, nearly all patients with aneurysmatic
SAH arrive to the ICU prior the aneurysm securing. Fourth, the frequency of craniotomies
in ICH group patients was almost double that
of those in SAH group patients in this study,
which must take into account when interpreting our findings. The difference is explainable
by the ICU admission criteria described above
and the priority of endovascular intervention
for aneurysm securing in our hospital. Entering ICH volume and ventriculostomy into the
multivariate model the OR for S100β loses
statistical significance. This is explained by
the small samples size as can been seen from
the wide confidence interval (0.87-11.1). Finally, the period over which serum S100β
concentration was measured may have been
too short to account for the influence of vasospasm on outcome in patients with aneurysmal bleeding.
Conclusions
We found that serum protein S100β concentration predicted poor long-term functional outcome in both SAH and ICH group patients, and
corresponded with the severity of neurological
insult assessed by the level of consciousness
and imaging findings on admission. These re-
Minerva Anestesiologica
November 2016
SERUM S100Β AS A PROGNOSTIC MARKER IN PATIENTS WITH NON-TRAUMATIC INTRACRANIAL HEMORRHAGE JUNTTILA
sults suggest that prospective trials looking
more precisely at the relationship between protein S100β and long-term neurologic outcome
in hemorrhagic stroke would be useful.
  8.
  9.
Key messages
—— Outcome prediction during intensive
care unit stay remains challenging in patients with non-traumatic intracranial hemorrhage and consequently there has been a
great interest in finding a laboratory test for
it.
—— Studies evaluating the usability of
blood S100β protein concentration as a
prognostic marker have promising results,
but the data considered it as a long-term
predictor of outcome with these patients is
scarce.
—— In this study S100β concentration
predicted poor long-term functional outcome and corresponded with the severity
of neurological insult in this patient population.
10.
11.
12.
13.
14.
15.
16.
References
  1. Fernandes HM, Mendelow AD. Non-traumatic intracranial hemorrhage. In: Webb A, Shapiro MJ, Singer M,
Suter PM, editors. Oxford textbook of critical care. New
York: Oxford University Press; 1999. p. 464-73.
  2. Christensen MC, Broderick J, Vincent C, Morris S, Steiner T. Global differences in patient characteristics, case
management and outcomes in intracerebral hemorrhage:
the Factor Seven for Acute Hemorrhagic Stroke (FAST)
trial. Cerebrovasc Dis 2009;28:55-64.
  3. Navarrete-Navarro P, Rivera-Fernandez R, Lopez-Mutuberria MT, Galindo I, Murillo F, Dominguez JM, et al.
Outcome prediction in terms of functional disability and
mortality at 1 year among ICU-admitted severe stroke patients: a prospective epidemiological study in the south of
the European Union (Evascan Project, Andalusia, Spain).
Intensive Care Med 2003;29:1237-44.
  4. Rosengart AJ, Schultheiss KE, Tolentino J, Macdonald RL.
Prognostic factors for outcome in patients with aneurysmal
subarachnoid hemorrhage. Stroke 2007;38:2315-21.
  5. Junttila EK, Koskenkari J, Romppainen N, Ohtonen PP,
Karttunen A, Ala-Kokko TI. ����������������������������
Risk factors for 1-year mortality in patients with nontraumatic intracranial hemorrhage requiring intensive care. Acta Anaesthesiol Scand
2011;55:1052-60.
  6. Brunkhorst R, Pfeilschifter W, Foerch C. Astroglial proteins as diagnostic markers of acute intracerebral hemorrhage-pathophysiological background and clinical findings. Transl Stroke Res 2010;1:246-51.
  7. Delgado P, Alvarez Sabin J, Santamarina E, Molina CA,
Vol. 82 - No. 11
17.
18.
19.
20.
21.
22.
23.
Quintana M, Rosell A, et al. Plasma S100B level after
acute spontaneous intracerebral hemorrhage. Stroke
2006;37:2837-9.
Sanchez-Pena P, Pereira AR, Sourour NA, Biondi A,
Lejean L, Colonne C, et al. S100B
����������������������������
as an additional prognostic marker in subarachnoid aneurysmal hemorrhage.
Crit Care Med 2008;36:2267-73.
Weiss N, Sanchez-Pena P, Roche S, Beaudeux JL, Colonne C, Coriat P, et al. Prognosis value of plasma S100B
protein levels after subarachnoid aneurysmal hemorrhage. Anesthesiology 2006;104:658-66.
Lai PM, Du R. Association between S100B Levels and
Long-Term Outcome after Aneurysmal Subarachnoid
Hemorrhage: Systematic Review and Pooled Analysis.
PLoS One 2016;11:e0151853.
Junttila E, Ala-Kokko T, Ohtonen P, Vaarala A, Karttunen
A, Vuolteenaho O, et al. Neurogenic pulmonary edema
in patients with nontraumatic intracerebral hemorrhage:
predictors and association with outcome. Anesth Analg
2013;116:855-61.
Junttila E, Vaara M, Koskenkari J, Ohtonen P, Karttunen
A, Raatikainen P, et al. Repolarization abnormalities in
patients with subarachnoid and intracerebral hemorrhage:
predisposing factors and association with outcome. Anesth Analg 2013;116:190-7.
Teasdale G, Jennett B. Assessment of coma and impaired
consciousness. A practical scale. Lancet 1974;2:81-4.
Hijdra A, Brouwers PJ, Vermeulen M, van Gijn J. Grading the amount of blood on computed tomograms after
subarachnoid hemorrhage. Stroke 1990;21:1156-61.
Connolly ES,Jr, Rabinstein AA, Carhuapoma JR,
Derdeyn CP, Dion J, Higashida RT, et al. Guidelines for
the management of aneurysmal subarachnoid hemorrhage: a guideline for healthcare professionals from the
American Heart Association/american Stroke Association. Stroke 2012;43:1711-37.
Diringer MN, Bleck TP, Claude Hemphill J,3rd, Menon
D, Shutter L, Vespa P, et al. Critical care management
of patients following aneurysmal subarachnoid hemorrhage: recommendations from the Neurocritical Care
Society’s Multidisciplinary Consensus Conference. Neurocrit Care 2011;15:211-40.
Morgenstern LB, Hemphill JC,3rd, Anderson C, Becker
K, Broderick JP, Connolly ES,Jr, et al. Guidelines for the
management of spontaneous intracerebral hemorrhage: a
guideline for healthcare professionals from the American
Heart Association/American Stroke Association. Stroke
2010;41:2108-29.
Jennett B, Bond M. Assessment of outcome after severe
brain damage. Lancet 1975;1:480-4.
Wiesmann M, Missler U, Hagenstrom H, Gottmann
D. S-100 protein plasma levels after aneurysmal subarachnoid haemorrhage. Acta Neurochir (Wien)
1997;139:1155-60.
Mercier E, Boutin A, Lauzier F, Fergusson DA, Simard
JF, Zarychanski R, et al. Predictive value of S-100beta
protein for prognosis in patients with moderate and severe traumatic brain injury: systematic review and metaanalysis. BMJ 2013;346:f1757.
Hu YY, Dong XQ, Yu WH, Zhang ZY. Change in plasma
S100B level after acute spontaneous basal ganglia hemorrhage. Shock 2010;33:134-40.
Vos PE, van Gils M, Beems T, Zimmerman C, Verbeek
MM. Increased GFAP and S100beta but not NSE serum
levels after subarachnoid haemorrhage are associated
with clinical severity. Eur J Neurol 2006;13:632-8.
Moritz S, Warnat J, Bele S, Graf BM, Woertgen C. The
prognostic value of NSE and S100B from serum and cerebrospinal fluid in patients with spontaneous subarachnoid hemorrhage. J Neurosurg Anesthesiol 2010;22:2131.
Minerva Anestesiologica
1197
JUNTTILA SERUM S100β AS A PROGNOSTIC MARKER IN PATIENTS WITH NON-TRAUMATIC INTRACRANIAL HEMORRHAGE
24. Hallevi H, Albright KC, Aronowski J, Barreto AD, Martin-Schild S, Khaja AM, et al. �����������������������
Intraventricular hemor-
rhage: Anatomic relationships and clinical implications.
Neurology 2008;70:848-52.
Authors’ contributions.—Eija K. Junttila, Juha Koskenkari, Tero I. Ala-Kokko designed the study, analyzed and interpreted the data,
drafted and critically revised the manuscript. Ari Karttunen analyzed the CT findings, contributed to the interpretation of the data,
provided critical revision of the article. Pasi
��������������������������������������������������������������������������������������������������
P. Ohtonen�����������������������������������������������������������������������������������
contributed substantially to the acquisition of the data, analysis and interpretation of the data, provided critical revision of the article.
Funding.—This study was financially supported by a small project grant from the Oulu University Hospital EVO grant for Dr Junttila.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material
discussed in the manuscript.
Acknowledgments.—The expert help of research assistant Maarika Vaara, MD, Elina Paloniemi, MD, and study nurse Sinikka Sälkiö
in the collection of data, and Heikki Koskinen, MD in the figure editing was much appreciated.
Article first published online: September 15, 2016. - Manuscript accepted: September 14, 2016. - Manuscript revised: September 13,
2016. - Manuscript received: February 1, 2016.
1198
Minerva Anestesiologica
November 2016
© 2016 EDIZIONI MINERVA MEDICA
Online version at http://www.minervamedica.it
Minerva Anestesiologica 2016 November;82(11):1199-213
REVIEW
Perioperative hemodynamic goal-directed
therapy and mortality: a systematic review
and meta-analysis with meta-regression
Mariateresa GIGLIO 1*, Fabio MANCA 2, Lidia DALFINO 1, Nicola BRIENZA 1
1Anaesthesia
and Intensive Care Unit, Department of Emergency and Organ Transplantation, University of Bari,
Policlinico, Bari, Italy; 2Department of Education, Psychology, Comunication studies, University of Bari, Bari, Italy
*Corresponding author: Mariateresa Giglio, Anaesthesia and Intensive Care Unit, Department of Emergency and Organ Transplantation, University of Bari, Policlinico, Piazza G. Cesare 11, 70124 Bari, Italy. Email: [email protected]
A B STRACT
INTRODUCTION: Recent data found that perioperative goal directed therapy (GDT) was effective only in higher control
mortality rates (>20%) with a relatively high heterogeneity that limited the strength of evidence. The aim of the present
meta-analysis was to clearly understand which high risk patients may benefit of GDT.
EVIDENCE ACQUISITION: Systematic review of randomized controlled trials with meta-analyses, including a metaregression technique. MEDLINE, EMBASE, and The Cochrane Library databases were searched (1980-January 2015).
Trials enrolling adult surgical patients and comparing the effects of GDT versus standard hemodynamic therapy were
considered. The primary outcome measure was mortality. Data synthesis was obtained by using Odds Ratio (OR) with
95% confidence interval (CI) by random-effects model.
EVIDENCE SYNTHESIS: Fifty eight studies met the inclusion criteria (8171 participants). Pooled OR for mortality was
0.70 (95% CI 0.56-0.88, P=0.002, no statistical heterogeneity). GDT significantly reduced mortality when it is >10% in
control group (OR 0.43, 95% CI 0.30-0.61, P<0.00001). The meta-regression model showed that the cut off of 10% of
mortality rate in control group significantly differentiates 43 studies from the other 15, with a regression coefficient b of
-0.033 and a P value of 0.0001. The significant effect of GDT was driven by high risk of bias studies (OR 0.48, 95% CI
0.34-0.67, P<0.0001).
CONCLUSIONS: The present meta-analysis, adopting the meta-regression technique, suggests that GDT significantly
reduces mortality even when the event control rate is >10%.
(Cite this article as: Giglio M, Manca F, Dalfino L, Brienza N. Perioperative hemodynamic goal-directed therapy and mortality: a systematic review and meta-analysis with meta-regression. Minerva Anestesiol 2016;82:1199-213)
Key words: Perioperative care - Mortality - Therapy.
T
Introduction
he strategy of hemodynamic goal-directed
therapy (GDT) refers to monitoring and
manipulation of physiological hemodynamic
parameters by means of therapeutic interventions,1 based mainly on fluids, red blood cells
and inotropic drugs. This regimen was origiComment in p. 1135.
Vol. 82 - No. 11
nally applied in surgical patients with the aim to
reach normal or supranormal values of cardiac
output and oxygen delivery 2 and later applied
to critically ill patients. A first meta-analysis 3
including surgical and critically ill patients did
not demonstrate any significant overall benefit.
Some years later, a lower mortality was observed
only in very severe surgical, trauma and medical
patients in whom optimization treatment was
performed before organ failure occurrence.4 A
Minerva Anestesiologica
1199
GIGLIO
GOAL DIRECTED THERAPY AND MORTALITY
subsequent meta-analytic study,5 differentiating
surgical patients from patients with sepsis and
organ failure, showed that in septic patients no
benefit was observed, although the matter is still
under debate,6 while mortality was improved in
perioperative subgroup, although the presence
of statistical heterogeneity and inconsistency
reduced the strength of evidence. The benefit
was recently confirmed only in very high risk
patients (control group mortality >20%) on the
basis of a priori subgroup division.7, 8 In these
papers, however, no clear explanation was provided about how this cut-off was obtained and a
relatively high heterogeneity was still observed
in all the analyses,7, 8 thus reducing the strength
of evidence.
We therefore conducted a meta-analysis
with meta-regression to clearly evaluate which
high risk patient can benefit from perioperative
GDT.
Information sources
Evidence acquisition
Eligibility criteria
RCTs were selected according to the following inclusion criteria:9
—— types of participants. Adult patients
(ages 18 years and older) undergoing major
surgery were considered. Studies involving
mixed populations of critically ill, nonsurgical
patients, or postoperative patients with sepsis
or organ failure were excluded;
—— types of interventions. GDT was defined as monitoring and manipulation of hemodynamic parameters to reach normal or
supranormal values by fluid infusion alone or
in combination with inotropic therapy in the
perioperative period within eight hours after
surgery. Studies including late hemodynamic
optimization treatment were excluded;
—— types of comparisons. Trials comparing
the beneficial and harmful effects of GDT versus standard hemodynamic therapy were considered. RCTs with no description or no difference in optimization strategies between groups,
as well as RCTs in which therapy was titrated
to the same goal in both groups or was not titrated to predefined end points were excluded;
1200
—— types of outcome measures. The primary outcome measure was mortality. For
those RCTs providing more data on mortality (i.e. in-hospital, 30-day, 90-day), the inhospital mortality was considered. Sensitivity
analysis was planned including only low risk
of bias trials (see below). Moreover, another
sub-group analysis was planned on the basis
of the result of the meta-regression model
(see below). A third subgroup analysis was
planned combining the results of the previous
two analyses (i.e. high mortality/high risk of
bias, high mortality/low risk of bias, low mortality/high risk of bias, low mortality/low risk
of bias);
—— types of studies. RCTs on perioperative
GDT in surgical patients were included. No
language, publication date, or publication status restrictions were imposed.
Different search strategies (last update January 2015) were performed to retrieve relevant
randomized controlled trials (RCTs) by using MEDLINE, The Cochrane Library and
EMBASE databases. No date restriction was
applied for MEDLINE and The Cochrane Library databases, while the search was limited
to 2008-2014 for EMBASE database.10 Additional RCTs were searched in The Cochrane
Library and the Database of Abstracts of Reviews of Effects (DARE) databases and in the
reference lists of previously published reviews
and retrieved articles. Other data sources
were hand-searched in the annual proceedings
(2008-2014) of the Society of Critical Care
Medicine, the European Society of Intensive
Care Medicine, the Society of Cardiovascular
Anesthesiologists, the Royal College of Anesthetists, the American Society of Anesthesiologists. In order to reduce publication bias, abstracts were searched.11 Publication language
was not a search criterion.
Search terms
Trials selection was performed by using
the following search terms: randomized con-
Minerva Anestesiologica
November 2016
GOAL DIRECTED THERAPY AND MORTALITYGIGLIO
trolled trial, controlled clinical trial, surgery,
goal directed, goal oriented, goal target, cardiac output, cardiac index, DO2, oxygen consumption, cardiac volume, stroke volume, fluid therapy, fluid loading, fluid administration,
optimization, supranormal. The search strategies used for the MEDLINE, The Cochrane
Library and EMBASE databases are reported
in the Supplementary Appendix I, online content only.
Study selection
Two investigators (FM, NB) examined at
first each title and abstract to exclude clearly
irrelevant studies and to identify potentially
relevant articles. Other two investigators (LD,
MG) independently determined eligibility of
full-text articles retrieved. The names of the
author, institution, journal of publication and
results were unknown to the two investigators
at this time.
Data abstraction and study characteristics
Data were independently collected by two
investigators (MG, NB), with any discrepancy
resolved by re-inspection of the original article. To avoid transcription errors, the data were
input into statistical software and rechecked by
different investigators (LD, FM).
RCT data gathered
Data abstraction included surgical risk (defined by the authors on the basis of POSSUM
score,12 ASA physical status classification, age
>60 years, pre-operative morbidity, as previously adopted),13 mortality of control group,
type of surgery (i.e., elective or emergent,
abdominal, thoracic, vascular, etc.), anesthesiological management, hemodynamic goaldirected therapy (end-points, therapeutic intervention and monitoring tools).
Risk of bias in individual studies
A domain-based evaluation, as proposed by
the Cochrane Collaboration, was used to eval-
Vol. 82 - No. 11
uate the methodological quality of RCTs.14
This is a two-part tool, addressing seven specific domains (namely, sequence generation,
allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective
outcome reporting) that are strongly associated
with bias reduction.15, 16 Each domain in the
tool includes one or more specific entries in a
“Risk of bias” table. Within each entry, the first
part of the tool describes what was reported to
have happened in the study, in sufficient detail
to support a judgement about the risk of bias.
The second part of the tool assigns a judgement relating to the risk of bias for that entry.
This is achieved by assigning a judgement of
“Low risk”, “High risk”, or “Unclear risk”
of bias. After each domain was completed,
a “Risk of bias summary” figure presenting
all of the judgements in a cross-tabulation of
study by entry are generated. The green plus
indicates low risk of bias, the red minus indicates high risk of bias, the white color indicates unclear risk of bias. For each study the
number of green plus obtained for every domain was calculated: RCTs with 5 or 6 green
plus were considered as having an overall low
risk of bias.
Summary measures and planned method of
analysis
Meta-analytic techniques (analysis software RevMan, version 5.3.5, Cochrane Collaboration, Oxford, England, UK) were used
to combine studies using odds ratios (ORs)
and 95% confidence intervals (CIs). A statistical difference between groups was considered to occur if the pooled 95% CI did
not include 1 for the OR. An OR less than 1
favored GDT when compared with control
group. Two-sided P values were calculated.
A random-effects model was chosen for all
analyses. Statistical heterogeneity and inconsistency were assessed by using the Q
and I2 tests, respectively.17, 18 When the P
value of the Q-Test was <0.10 17 and/or the
I2 was >25%, heterogeneity and inconsistency were considered significant.19
Minerva Anestesiologica
1201
GIGLIO
GOAL DIRECTED THERAPY AND MORTALITY
Meta-regression. Assessing the impact of the
slope
In the present meta-analysis the chosen
covariates were the mortality rate in control
group, the risk of bias evaluation (considering
how many green plus the study obtained) and
the year of publication. The meta-regression
model 20-22 was applied to all the included studies. The software used was Comprehensive
Meta Analysis Version 3.0
Evidence synthesis
Study selection
Identification
The search strategies identified 3244 (MEDLINE), 9948 (Cochrane Library) and 3054 (EMBASE) articles. Thirteen articles were identified
through other sources (congress abstracts, ref-
16246 records identified
through database searching
erence lists). After initial screening and subsequent selection, a pool of 98 potentially relevant
RCTs was identified. The subsequent eligibility process (Figure 1) excluded 40 articles and,
therefore, 58 articles 23-80 with a total sample of
8171 patients, were considered for the analysis.
Study characteristics
All included articles evaluated the effects of
hemodynamic optimization on mortality as primary or secondary outcome and had a population
sample of adult surgical patients, undergoing
both elective or emergent procedures (Supplementary Table I, online content only).23-80 The
studies were performed in Australia, United
States, Europe, Canada, Brazil, China, Israel and
India from 1991 to 2014 (Supplementary Table
I) and were all published in English.
13 additional records identified
through other sources
Screening
3516 records after duplicates removed
116 records excluded
Eligibility
98 full-text articles assessed for eligibility
Included
214 records screened
58 studies included in
qualitative synthesis
40 full-text articles excluded:
— 26: hemodynamic optimization
titrated to the same end-point
or not titrated to predefined end
points, or no difference between
groups in the optimization protocol;
— 13: mixed population of critically ill, not surgical patients, with
already established sepsis or organ failure and undergoing late
optimization;
— 1: only protocol
58 studies included in
qualitative synthesis
Figure 1.—Flow chart summarizing the studies selection procedure for the meta-analysis.
1202
Minerva Anestesiologica
November 2016
GOAL DIRECTED THERAPY AND MORTALITYGIGLIO
Data concerning RCTs morbidity/mortality
risk definition, population and type of surgery
are presented in Supplementary Table I. The
risk of bias assessment for each trial is showed
(in Supplementary Table II online content
only).
Figure 2.—Rates of mortality for each of the studies with Odds Ratios (ORs) and 95% Confidence Intervals (CI). The pooled
OR and 95% CI are shown as the total. The size of the box at the point estimate of the OR gives a visual representation of
the “weighting” of the study. The diamond represents the point estimate of the pooled OR and the length of the diamond is
proportional to the CI.
Vol. 82 - No. 11
Minerva Anestesiologica
1203
GIGLIO
GOAL DIRECTED THERAPY AND MORTALITY
Figure 3.—Rates of mortality with Odds Ratios (ORs) and 95% Confidence Intervals (CI) in subgroup according to risk of
bias. Studies were divided in high risk of bias and low risk of bias according to the Cochrane domain-based evaluation (see
text for details). The pooled OR and 95% CI are shown as the total. The size of the box at the point estimate of the OR gives
a visual representation of the “weighting” of the study. The diamond represents the point estimate of the pooled OR and the
length of the diamond is proportional to the CI.
1204
Minerva Anestesiologica
November 2016
GOAL DIRECTED THERAPY AND MORTALITYGIGLIO
Quantitative data synthesis
In 58 RCTs, 482 patients died: 205 out of
4137 (5%) were randomized to perioperative
goal-directed therapy, and 277 out of 4034
(7%) were randomized to control. Pooled OR
for mortality was 0.70 and 95% CI was 0.560.88. No statistical heterogeneity and inconsistency were detected (Figure 2). Excluding the
largest study,61 the result was confirmed: OR
was 0.62 with 95% CI 0.48-0.80 (P=0.0002,
6177 pts), and no significant statistical heterogeneity (Q statistic P=0.56; I2=0%) was observed.
The sensitivity analysis showed that the significant effect of GDT on mortality was driven by high risk of bias RCTs (OR 0.48, 95%
CI 0.34-0.67, P<0.0001, Q statistic P=0.78;
I2=0%, 32 RCTs), while no effect was demonstrated in low risk of bias trials (OR 0.94,
95% CI 0.73- 1.20, P=0.61, Q statistic P=0.57;
I2=%, 26 RCTs) (Figure 3).
Meta-regression
Figure 4 showed the plot of log odds ratio
on control group mortality: the meta-regression model identified, by inspection to the
plot, a point (cut-off) on the upper confidence
interval of the regression line that separates
positive log odds ratios to negative log odds
ratios: this cut-off coincides with the mortality
rate in control group of 10%. In other words,
the meta-regression model, applied to all 58
RCTs, showed that the cut off of 10% of mortality rate in control group significantly differentiated 43 studies from the other 15, with a
regression coefficient b of -0.033 and a P value
of 0.0001 (see Figure 4). Supplementary Table
III (online content only), showed the results
for meta-regression using control group mortality to predict the log odds ratio.
The subgroup analysis including only studies in which the mortality rate in the control
group was lower than 10% showed no significant results (OR 0.99, 95% CI 0.78-1.27,
P=0.95, Q statistic P=0.97 I2=0%, 43 RCTs),
while a statistical significant effect was observed in those RCTs with a mortality rate in
Vol. 82 - No. 11
Figure 4.—Regression model applied to all 58 studies, basing on mortality rate in control group. The cut off of 10% of
mortality rate in control group significantly differentiates 43
studies on the right side of the graph from 15 on the left side.
The regression coefficient b=-0.033 with a P-value of 0.0001
confirms the analysis (see text for details).
control group >10% (OR 0.43, 95% CI 0.300.61, P<0.00001, Q statistic P=0.41; I2=3%,
15 RCTs, (Figure 5).
Figure 6 showed the plot of log odds ratio
on risk of bias evaluation. The meta-regression
model was applied to all 58 RCTs. The inspection to the plot showed that only studies with
high risk of bias (<5 green plus obtained with
the Cochrane domain-based evaluation for risk
of bias) had a significant reduction in mortality
rate, with a regression coefficient b of 0.225
and a P value<0.00001 (Figure 6), while no
significant reduction was observed in low risk
of bias RCTs. Supplementary Table IV (online
content only) showed the results for meta-regression using risk of bias evaluation to predict the log odds ratio. Figure 7 showed the
results of the combined 4 subgroups (i.e. mortality <10%/high risk of bias, mortality <10%/
low risk of bias, mortality >10%/high risk of
bias, mortality >10%/low risk of bias): only
the group with mortality rate >10% and high
risk of bias reached statistical significance (OR
0.38, 95% CI 0.25-0.57, P<0.00001, Q statistic P=0.66; I2=0%, 9 RCTs). The group with
mortality rate >10% and low risk of bias did
not show any statistical significance (OR 0.53,
95% CI 0.25-1.13, P=0.10, Q statistic P=0.19;
I2=33%, 6 RCTs).
In order to look to the effect of time as a
covariate, another meta-regression was per-
Minerva Anestesiologica
1205
GIGLIO
GOAL DIRECTED THERAPY AND MORTALITY
Figure 5.—Rates of mortality with Odds Ratios (ORs) and 95% Confidence Intervals (CI) in subgroups defined according
to the mortality rate in control group. RCTs were divided in studies with a control mortality rate < of 10% or > of 10%. The
pooled OR and 95% CI are shown as the total. The size of the box at the point estimate of the OR gives a visual representation
of the “weighting” of the study. The diamonds represent the point estimate of the pooled ORs and the length of the diamonds
is proportional to the CI.
1206
Minerva Anestesiologica
November 2016
GOAL DIRECTED THERAPY AND MORTALITYGIGLIO
Figure 6.—Regression model applied to all 58 studies, basing on risk of bias evaluation (number of green plus obtained
by each RCT according to the domain-based risk of bias
evaluation, as proposed by the Cochrane Collaboration).
The regression coefficient b=0.225 with a P-value of 0.0001
confirms the result of the sensitive analysis: the global effect
of GDT on mortality was driven by high risk of bias studies, while the higher quality studies did not demonstrate any
benefit in mortality reduction. (see text for details).
formed, adopting the publication year as a
covariate: this analysis (Figure 8) showed that
there is no statistical dependence between the
effect on mortality and the year of publication
(regression coefficent 0.045).
Discussion
This systematic review and meta-analysis
confirmed that perioperative GDT reduced
mortality after surgery. This significant effect
was maintained when the mortality rate in control group was >10% and when high risk of
bias RCTs were considered. No significant effect was observed in low risk of bias trials.
GDT has been originally applied in surgical patients in order to face the perioperative
increase in oxygen demand and to prevent organ failure. When performed in patients with
already established organ failure, no outcome
improvement was found.5 These different results may rely on the basis that only in the
early stage of systemic inflammatory response
syndrome (i.e. in the early preoperative period) it is possible to prevent the deleterious effects of hypoperfusion and decreased oxygen
delivery, while, when oxygen debt is no longer
reversible, increasing oxygen transport is no
more effective.
Vol. 82 - No. 11
Expected mortality appears to be a very
striking factor affecting effectiveness of GDT.
Shoemaker et al.2 found that GDT was effective only when optimization treatment
was performed in high risk medical, surgical
and trauma patients (control group mortality
>20%) before organ failure occurrence. Recent data 7, 8 have confirmed GDT benefits only
in surgical patients, but have again limited its
effectiveness only in patients with very high
control mortality (i.e., >20%). This reported
cut-off, however, may result from a priori classification and may not reflect the “real life” of
every day practice.81, 82 The present meta-analysis, adopting the meta-regression technique,
demonstrated that preoperative hemodynamic
optimization significantly reduced mortality
even when the event control rate is >10%. It
should be underscored, however, that, in this
meta-regression model, the relationship between effect estimates and the control group
risk is complicated by a technical phenomenon
known as regression to the mean. This arises
because the control group risk forms an integral part of the effect estimate. A high risk in
a control group, observed entirely by chance,
will on average give rise to a higher than expected effect estimate, and vice versa. This
phenomenon results in a false correlation between effect estimates and control group risks.
This is the reason why we decided to perform
also a subgroup analysis that is another way
to investigate heterogeneous results or to answer specific questions about particular patient
groups. The subgroup adopting the cut-off of
mortality rate in control group >10% reinforced the meta-regression result. Moreover,
despite the existence of methodological heterogeneity among studies dealing with GDT,
such as timing, monitoring and protocols, a
strong statistical homogeneity and consistency
(even using conservative cut-off values) was
still observed in the main as well as in the
sensitive analyses, whereas moderate to high
heterogeneity and inconsistency has limited
precedent results.7 A reason for this discrepancy is that the present meta-analysis did not
include two 83, 84 of the 32 studies of previous
papers, because in these two papers dopex-
Minerva Anestesiologica
1207
GIGLIO
GOAL DIRECTED THERAPY AND MORTALITY
Figure 7.—Rates of mortality with Odds Ratios (ORs) and 95% Confidence Intervals (CI) in subgroup according to the combination of mortality > or <10% and risk of bias. Four subgroup were obtained: mortality <10%/high risk of bias, mortality
<10%/low risk of bias, mortality >10%/high risk of bias, mortality >10%/low risk of bias (see text for details). The pooled
OR and 95% CI are shown as the total. The size of the box at the point estimate of the OR gives a visual representation of
the “weighting” of the study. The diamond represents the point estimate of the pooled OR and the length of the diamond is
proportional to the CI.
1208
Minerva Anestesiologica
November 2016
GOAL DIRECTED THERAPY AND MORTALITYGIGLIO
Figure 8.—Regression model applied to all 58 studies,
basing on the publication year as a covariate: this analysis
shows that there is no statistical dependence between the
effect on mortality and the year of publication (regression
coefficient 0.045).
amine was used without predefined hemodynamic end points. Another paper 8 reviewed
the use of GDT to reduce mortality, and again
arbitrarily adopted a 20% cut-off of mortality rate in control group. Once again, heterogeneity reduced the strength of the evidence.
One included paper 85 was not included in the
present meta-analysis for methodological reason (both the control and the GDT group were
treated with the same protocol). Moreover, the
present meta-analysis was updated with new
19 studies published from 2010 to 2014. All
these reasons may have reduced the variability
in the outcome observed.
The ineffectiveness of GDT in reducing
mortality in the low control mortality subgroups could be due to a low statistical power
needing much larger numbers of patients to
show statistical significance. However, this
subgroup analysis including 6900 patients was
enough powered to exclude the latter hypothesis. An alternative hypothesis is that patients
that are not very ill may not respond as clearly
to increased hemodynamic.
The quality analysis showed that the global
effect of GDT on mortality was driven by high
risk of bias studies, while higher quality studies did not demonstrate any benefit in mortality reduction. The meta-regression analysis
further confirmed this figure. The risk of bias
for each study was evaluated and studies were
Vol. 82 - No. 11
classified as low and high risk of bias according to the domain-based evaluation, as that
proposed by the Cochrane Collaboration. Out
of 58 studies, 26 reached a low risk of bias
evaluation. Many studies presented some important limitations since were conducted in
single centers with limited patient samples,
and only few RCTs were adequately randomized and double-blinded. The overall low quality of individual studies on GDT has been previously called into question,3 although the trial
quality seemed to influence the outcome in the
studies including perioperative patients less
than in the subset of patients with established
sepsis and multiple organ failure.5, 7 However,
it is well established that studies with high risk
of bias often overestimate the true effect, reducing the clinical significance of any result,13
and this could explain the results of the present study. Interestingly, most of the studies
with low risk of bias are also low mortality
studies: we therefore tried to combine the two
subgroup analyses making high mortality/low
bias, high mortality/high bias, low mortality/
low bias and low mortality/high bias groups.
This analysis confirmed that no effect was seen
in mortality rate <10%, both in low and in high
risk of bias, while the benefit on mortality was
driven by high risk of bias trials (9 RCTs),
while the subgroup including mortality >10%
and low risk of bias (6 RCTs) did not reach
statistical significance.
It has been proposed that the year of publication could affect the risk of bias, since older
paper are more prone to high risk of bias, while
newer ones are less affected by risk of bias.
The meta-regression adopting the publication
year as a covariate showed that there is no
statistical dependence between the effect on
mortality and the year of publication, suggesting that also older RCTs, if well planned and
conducted, could be considered as low risk of
bias, while recent ones, if not adequately designed, could be affected by high risk of bias.
Therefore, another possible interpretation of
the present meta-analysis could be that, when
dealing with the effect of perioperative hemodynamic optimization on mortality, one should
consider not only the “risk of bias” per se, but
Minerva Anestesiologica
1209
GIGLIO
GOAL DIRECTED THERAPY AND MORTALITY
also the design of the study, and, maybe more
important, the type of population enrolled, including co-morbidities, ASA class and mortality risk.
This study had a number of limitations.
No attempt was made to correct for the type
or quantity of fluids or inotropes given, because they are inconsistently reported in the
literature and have a demonstrable wide variability in their dosing across studies. Moreover, the studies included varied in terms of
hemodynamic monitoring, the goals proposed
and achieved, the timing of intervention: this
could have introduced a relatively high clinical
heterogeneity, although the results remained
consistent across a number of subgroups and
sensitivity analyses. However, the high heterogeneity among the tools and goals used to
define GDT is still a major clinical problem. It
is hard to believe that GDT by means of a Masimo pulse oxymeter can in anyway be equal
to GDT conducted by a pulmonary artery catheter which is the goal standard to measure cardiac output.
Additional well-designed randomized controlled studies are necessary to clarify these
discrepancies and to determine whether mortality can be reduced through the maintenance
of perioperative tissue perfusion in high-risk
surgical patient. Moreover, several issues need
to be clarified, such as timing, monitoring tools
and protocols adopted, as well as the targets
adopted, as recently underscored.86
Conclusions
This meta-analysis, within the limitations
of existing data, the high heterogeneity among
adopted protocols, and the analytic approaches
used, suggested that preoperative GDT significantly reduced mortality when the event control rate is >10%. In well conducted non-biased studies no mortality benefit was observed
but the effect may be limited due to inclusion
of small number of high mortality trials. Additional well-designed randomized controlled
studies are still necessary to clarify several discrepancies among monitoring tools, goals and
timing.
1210
Key messages
—— The present meta-analysis, adopting
the meta-regression technique, suggested
that preoperative hemodynamic optimization significantly reduced mortality even
when the event control rate is >10%.
—— The global effect of GDT on mortality was driven by high risk of bias studies,
while higher quality studies did not demonstrate any benefit in mortality reduction.
—— When dealing with the effect of perioperative hemodynamic optimization on
mortality, one should consider not only the
“risk of bias” per se, but also the design of
the study, and, maybe more important, the
type of population enrolled, including comorbidities, ASA class and mortality risk.
References
  1. Harten J, Kinsella J. Perioperative optimisation. Scott
Med J 2003;49:6-9.
  2. Shoemaker WC, Appel PL, Kram HB. Hemodynamic
and oxygen transport responses in survivors and nonsurvivors of high-risk surgery. Crit Care Med 1993;21:97790.
  3. Heyland DK, Cook DJ, King D, Kernerman P, BrunBuisson C. Maximizing oxygen delivery in critically ill
patients: a methodologic appraisal of the evidence. Crit
Care Med 1996;24:517-24.
  4. Kern JW, Shoemaker WC. Meta-analysis of hemodynamic optimization in high-risk patient. Crit Care Med
2002;30:1686-92.
  5. Poeze M, Greve JWM, Ramsay G. Meta-analysis of hemodynamic optimization: relationship to methodological
quality. Critical Care 2005;9:R 771-R779.
 6.Dell’Anna AM, Taccone FS. Early-goal directed therapy for septic shock: is it the end? Minerva Anestesiol
2015;81:1138-43.
  7. Hamilton, MA, Cecconi M, Rhodes A. A Systematic
Review and Meta-Analysis on the Use of Preemptive
Hemodynamic Intervention to Improve Postoperative
Outcomes in Moderate and High-Risk Surgical Patients.
Anesth Analg 2011,112:1392-40.
 8.Gurgel ST, do Nascimento P. Maintaining Tissue Perfusion in High-Risk Surgical Patients: A Systematic
Review of Randomized Clinical Trials. Anesth Analg
2011;112:1384-91.
  9. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred
Reporting Items for Systematic Reviews and Meta-Analyses: the PRISMA statement. BMJ 2009;339:b2535.
10. Lefebvre C, Manheimer E, Glanville J. Searching for
studies. In Cochrane Handbook for Systematic Reviews
of Interventions Version 5.0.1. Volume Chapter 6. Edited
by: Higgins JPT, Green S. The Cochrane Collaboration
(updated September 2008); [http://www.cochrane-handbook.org/].
11. McAuley L, Pham B, Tugwell P, Moher D. Does the
inclusion of grey literature influence estimates of inter-
Minerva Anestesiologica
November 2016
GOAL DIRECTED THERAPY AND MORTALITYGIGLIO
vention effectiveness reported in meta-analyses? Lancet
2000;356:1228-31.
12. Copeland GP, Jones D, Walters M. POSSUM: a scoring
system for surgical audit. Br J Surg 1991;78:355-60.
13. Brienza N, Giglio MT, Marucci M, Fiore T. Does perioperative hemodynamic optimization protect renal function
in surgical patients? A meta-analytic study. Crit Care Med
2009;37:2079-90.
14. Higgins JPT, Altman DG, Sterne JAC. Chapter 8: Assessing risk of bias in included studies. In: Higgins JPT,
Green S, editors. Cochrane Handbook for Systematic
Reviews of Interventions Version 5.1.0 (updated March
2011). The Cochrane Collaboration, 2011. Available
from www.cochrane-handbook.org.
15. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds
DJ, Gavaghan DJ, et al. Assessing the quality of randomized clinical trials: is blinding necessary? Control Clin
Trials 1996;17:1-12.
16. Juni P, Witschi A, Bloch R, Egger M. The hazards of
scoring he quality of clinical trials for meta-analysis.
JAMA 1999;282:1054-60.
17. Higgins JPT, Thompson SG, Deeks JJ, Altman DG.
Measuring inconsistency in meta-analyses. BMJ
2003,327:557-60.
18. Higgins JPT, Thompson SG. Quantifying heterogeneity
in a meta-analysis. Stat Med 2002;21:1539-58.
19. Deeks JJ, Higgins JPT, Altman DG. Chapter 9: Analysing data and undertaking meta-analyses. In: Higgins JPT,
Green S, editors. Cochrane Handbook for Systematic
Reviews of Interventions Version 5.0.1 (updated September 2008). The Cochrane Collaboration, 2008. Available
from www.cochrane-handbook.org
20. Thompson S, Higgins J. How should meta-regression
analyses by undertaken and interpreted? Stat Med
2002;21:1559-74.
21. Van Houwelingen HC, Arends LR, Stijnen T. Advanced
methods in meta-analysis: multivariate approach and
meta-regression. Stat Med 2002;21:589-624.
22.White IR. Multivariate random-effects meta-analysis.
Stata J 2009;9:40-56.
23. Bartha E, Arfwedson C, Imnell A, Fernlund ME, Andersson LE, Kalman S. Randomized controlled trial of goaldirected hemodynamic treatment in patients with proximal femoral fracture. Br J Anaesth 2013;110:545-53.
24. Bender JS, Smith-Meek MA, Jones CE. Routine pulmonary artery catheterization does not reduce morbidity and
mortality of elective vascular surgery: results of a prospective, randomized trial. Ann Surg 1997;226:229-36.
25. Benes J, Chytra I, Altmann P, Hluchy M, Kasal E, Svitak
R, et al. Intraoperative fluid optimization using stroke
volume variation in high risk surgical patients: results of
prospective randomized study. Crit Care 2010;14:R118.
26. Berlauk JF, Abrams JH, Gilmour IJ, O’Connor SR,
Knighton DR, Cerra FB. Preoperative optimization of
cardiovascular hemodynamics improves outcome in
peripheral vascular surgery: a prospective, randomized
clinical trial. Ann Surg 1991;214:289-299.
27. Bisgaard J, Gilsaa T, Rønholm E, Toft P. Hemodynamic
optimisation in lower limb arterial surgery: room for improvement? Acta Anaesthesiol Scand 2013;57:189-98.
28. Bisgaard J, Gilsaa T, Rønholm E, Toft P. Optimising
stroke volume and oxygen delivery in abdominal aortic
surgery: a randomised controlled trial. Acta Anaesthesiol
Scand 2013;57:178-88.
29. Bishop MH, Shoemaker WC, Appel P, Meade P, Ordog
GJ, Wasserberger J, et al. Prospective, randomized trial
of survivor values of cardiac index, oxygen delivery, and
oxygen consumption as resuscitation endpoints in severe
trauma. Arch Surg 1992,127:1175-9.
30. Bonazzi M, Gentile F, Biasi GM, Migliavacca S, Esposti
D, Cipolla M, et al. Impact of perioperative hemodynam-
Vol. 82 - No. 11
ic monitoring on cardiac morbidity after major vascular
surgery in low risk patients. A randomised pilot trial. Eur
J Vasc Endovasc Surg 2002;23:445-51.
31. Boyd O, Grounds M, Bennett D. A randomized clinical
trial of the effect of deliberate perioperative increase of
oxygen delivery on mortality in high-risk surgical patients. JAMA 1993;270:2699-2708.
32. Brandstrup B, Svendsen PE, Rasmussen M, Belhage B,
Rodt SÅ, Hansen B, et al. Which goal for fluid therapy
during colorectal surgery is followed by the best outcome: near-maximal stroke volume or zero fluid balance?
Br J Anaesth 2012;109:191-9.
33. Buettner M, Schummer W, Huettemann E, Schenke S,
van Hout N, Sakka SG. Influence of systolic-pressurevariation-guided intraoperative fluid management on
organ function and oxygen transport. Br J Anaesth
2008;101:194-9.
34.Cecconi M, Fasano N, Langiano N, Divella M, Costa
MG, Rhodes A, et al. Goal-directed hemodynamic therapy during elective total hip arthroplasty under regional
anaesthesia. Crit Care 2011;15:R132.
35. Challand C, Struthers R, Sneyd JR, Erasmus PD, Mellor N, Hosie KB, et al. Randomized controlled trial of
intraoperative goal-directed fluid therapy in aerobically
fit and unfit patients having major colorectal surgery. Br J
Anaesth 2012;108:53-62.
36.Chytra I, Pradl R, Bosman R, Pelnár P, Kasal E, Zidková
A. Esophageal Doppler-guided fluid management decreases blood lactate levels in multiple-trauma patients:
a randomized controlled trial. Crit Care 2007;11:R24.
37. Conway DH, Mayall R, Abdul-Latif MS, Gilligan S,
Tackaberry C. Randomised controlled trial investigating
the influence ofintravenous fluid titration using Esophageal Doppler monitoring during bowel surgery. Anaesthesia 2002;57:845-9.
38. Donati A, Loggi S, Preiser JC, Orsetti G, Münch C, Gabbanelli V, et al. Goal-directed intraoperative therapy reduces morbidity and length of hospital stay in high-risk
surgical patients. Chest 2007;132:1817-24.
39. Fleming A, Bishop M, Shoemaker W, Appel P, Sufficool
W, Kuvhenguwha A, et al. Prospective trial of supranormal values as goals of resuscitation in severe trauma.
Arch Surg 1992;127:1175-9.
40. Forget P, Lois F, de Kock M. Goal-directed fluid management based on the pulse oximeterderived pleth variability
index reduces lactate levels and improves fluid management. Anesth Analg 2010;111:910-4.
41. Gan TJ, Soppitt A, Maroof M, el-Moalem H, Robertson
KM, Moretti E, et al. Goal-directed intraoperative fluid
administration reduces length of hospital stay after major
surgery. Anesthesiology 2002;97:820-6.
42. Goepfert MS, Richter HP, Eulenburg CZ, Gruetzmacher
J, Rafflenbeul E, Roeher K, et al. Individually Optimized Hemodynamic Therapy Reduces Complications
and Length of Stay in the Intensive Care Unit: A Prospective, Randomized Controlled Trial. Anesthesiology
2013;119:824-36.
43. Harten J, Crozier JEM, McCreath B, Hay A, McMillan
DC, McArdle CS, et al. Effect of intraoperative fluid
optimisation on renal function in patients undergoing
emergency abdominal surgery: a randomised controlled
pilot study (ISRCTN 11799696). Int J Surg 2008;6:197204.
44. Jammer I, Ulvik A, Erichsen C, Lødemel O, Ostgaard
G. Does central venous oxygen saturation-directed fluid
therapy affect postoperative morbidity after colorectal
surgery? A randomized assessor-blinded controlled trial.
Anesthesiology 2010;113:1072-80.
45. Jhanii S, Vivian-Smith A, Lucena-Amaro S, Watson D,
Hinds CJ, Pearse RM. Hemodynamic optimisation improves tissue microvascular flow and oxygenation after
Minerva Anestesiologica
1211
GIGLIO
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
GOAL DIRECTED THERAPY AND MORTALITY
major surgery: a randomised controlled trial. Crit Care
2010;14:R151.
Jones C, Kelliher L, Dickinson M, Riga A, Worthington
T, Scott MJ, et al. Randomized clinical trial on enhanced
recovery versus standard care following open liver resection. Br J Surg 2013;100:1015-24.
Kapoor M, Kakani M, Chowdhury U, Choudhury M,
Lakshmy, Kiran U, et al. Early goal-directed therapy in
moderate to high-risk cardiac surgery patients. Ann Card
Anaesth 2008;11:27-34.
Lobo SM, Salgado PF, Castillo VG, Borim AA, Polachini
CA, Palchetti JC, et al. Effects of maximizing oxygen delivery on morbidity and mortality in high-risk surgical
patients. Crit Care Med 2000;28:3396-404.
Lopes MR, Oliveira MA, Pereira V, Lemos I, Auler J,
Michard F. Goal-directed fluid management based on
pulse pressure variation monitoring during high-risk
surgery: a pilot randomized controlled trial. Crit Care
2007;11:R100.
Mayer J, Boldt J, Mengistu A, Rohm K, Suttner S. Goaldirected intraoperative therapy based on autocalibrated
arterial pressure waveform analysis reduces hospital stay
in high-risk surgical patients: a randomized, controlled
trial. Crit Care 2010;14:R18.
McKendry M, McGloin H, Saberi D, Caudwell L, Brady
AR, Singer M, et al. Randomised controlled trial assessing the impact of a nurse delivered, flow monitored protocol for optimisation of circulatory status after cardiac
surgery. BMJ 2004;329:258.
Mckenny M, Conroy P, Wong A, Farren M, Gleeson N,
Walsh C, et al. A randomised prospective trial of intraoperative Esophageal Doppler-guided fluid administration in major gynaecological surgery. Anaesthesia.
2013;68:1224-31.
Moppett IK, Rowlands M, Mannings A, Moran CG,
Wiles MD. LiDCO-based fluid management in patients
undergoing hip fracture surgery under spinal anaesthesia:
a randomized trial and systematic review. Br J Anaesth
2015;114:444-59.
Mythen MG, Webb AR. Perioperative plasma volume
expansionreduces the incidence of gut mucosal hypoperfusionduring cardiac surgery. Arch Surg 1995;130:4239.
Noblett SE, Snowden CP, Shenton BK, Horgan AF. Randomized clinical trial assessing the effect of Doppleroptimized fluid management on outcome after elective
colorectal resection. Br J Surg 2006;93:1069-76.
Pearse R, Dawson D, Fawcett J, Rhodes A, Grounds
RM, Bennett ED. Early goal-directed therapy after
major surgery reduces complications and duration of
hospital stay. A randomised, controlled trial Crit Care
2005;9:687-93.
Pearse R, Harrison DA, MacDonald N, Gillies MA,
Blunt M, Ackland G, et al. Effect of a perioperative,
cardiac output-guided hemodynamic therapy algorithm
on outcomes following major gastrointestinal surgery: a
randomized clinical trial and systematic review. JAMA
2014,311:2181-90.
Peng K, Li J, Cheng H, Ji F. Goal-Directed Fluid Therapy Based on Stroke Volume Variations Improves Fluid
Management and Gastrointestinal Perfusion in Patients
Undergoing Major Orthopedic Surgery. Med Princ Pract
2014,23:413-20.
Pestana D, Espinoza E, Eden A, Nájera D, Collar L, Aldecoa C, et al. Perioperative goal-directed hemodynamic
optimization using noninvasive cardiac output monitoring in major abdominal surgery: a prospective, randomized, multicenter, pragmatic trial: POEMAS Study
(PeriOperative goal-directed thErapy in Major Abdominal Surgery). Anesth Analg 2014;119:579-87.
Polonen P, Ruokonen E, Hippelainen M, Pöyhönen M,
1212
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
Takala J. A prospective, randomized study of goal-oriented hemodynamic therapy in cardiac surgical patients.
Anesth Analg 2000;90:1052-9.
Poso T, Winso O, Aroch R, Kesek D. Perioperative Fluid
Guidance with Transthoracic Echocardiography and
Pulse-Contour Device in Morbidly Obese Patients. Obes
Surg 2014;24:2117-25.
Sandham JD, Hull RD, Brant RF, Knox L, Pineo GF,
Doig CJ, et al. A Randomized, Controlled Trial of the
Use of Pulmonary-Artery Catheters in High-Risk Surgical Patients. N Engl J Med 2003;348:5-14.
Schereen TWL, Wiesenack C, Gerlach H, Marx G. Goaldirected intraoperative fluid therapy guided by stroke
volume and its variation in high-risk surgical patients: a
prospective randomized multicentre study. J Clin Monit
Comput 2013;27:225-33.
Senagore AJ, Emery T, Luchtefeld M, Kim D, Dujovny N, Hoedema R. Fluid management for laparoscopic
colectomy: a prospective, randomized assessment of
goal-directed administration of balanced salt solution or
hetastarch coupled with an enhanced recovery program.
Dis Colon Rectum 2009;52:1935-40.
Shoemaker WC, Appel PL, Kram HB, Waxman K, Lee
TS. Prospective trial of supranormal values of survivors
as therapeutic goals in high-risk surgical patients. Chest
1988;94:1176-86.
Sinclair S, James S, Singer M. Intraoperative intravascular volume optimisation and length of hospital stay after
repair of proximal femoral fracture: randomised controlled trial. BMJ 1997;315:909-12.
Smetkin AA, Kirov MY, Kuzkov VV, Lenkin AI, Eremeev AV, Slastilin VY, et al. Single transpulmonary thermodilution and continuous monitoring of central venous
oxygen saturation during off-pump coronary surgery.
Acta Anaesthesiol Scand 2009;53:505-14.
Szakmany T, Toth I, Kovacs Z, Leiner T, Mikor A, Koszegi T, et al. Effects of volumetric vs. pressure-guided
fluid therapy on postoperative inflammatory response:
a prospective, randomized clinical trial. Intensive Care
Med 2005;31:656-63.
Ueno S, Tanabe G, Yamada H, Kusano C, Yoshidome S,
Nuruki K, et al. Response of patients with cirrhosis who
have undergone partial hepatectomy to treatment aimed
at achieving supranormal oxygen delivery and consumption. Surgery 1998;123:278-86.
Valentine RJ, Duke ML, Inman MH, Grayburn PA, Hagino RT, Kakish HB, et al. Effectiveness of pulmonary artery catheters in aortic surgery: a randomized trial. J Vasc
Surg 1998;27:203-11.
Van der Linden PJ, Dierick A, Wilmin S, Bellens B, De
Hert SG. A randomized controlled trial comparing an intraoperative goal-directed strategy with routine clinical
practice in patients undergoing peripheral arterial surgery. Eur J Anaesthesiol 2010;27:788-93.
Velmahos GC, Demetriades D, Shoemaker WC, Chan LS,
Tatevossian R, Wo CC, et al. Endpoints of resuscitation of
critically injured patients: normal or supranormal? A prospective randomized trial. Ann Surg 2000;232:409-18.
Venn R, Steele A, Richardson P, Poloniecki J, Grounds
M, Newman P. Randomized controlled trial to investigate
influence of the fluid challenge on duration of hospital
stay and perioperative morbidity in patients with hip fractures. Br J Anaesth 2002;88:65-71
Wakeling HG, McFall MR, Jenkins CS, Woods WG,
Miles WF, Barclay GR, et al. Intraoperative Esophageal
Doppler guided fluid management shortens postoperative hospital stay after major bowel surgery. Br J Anaesth
2005;95:634-42.
Wenkui Y, Ning L, Jianfeng G, Weiqin L, Shaoqiu T, Zhihui T, et al. Restricted peri-operative fluid administration
adjusted by serum lactate level improved outcome after
Minerva Anestesiologica
November 2016
GOAL DIRECTED THERAPY AND MORTALITYGIGLIO
76.
77.
78.
79.
80.
major elective surgery for gastrointestinal malignancy.
Surgery 2010;147:542-52.
Wilson J, Woods I, Fawcett J, Whall R, Dibb W, Morris C, et al. Reducing the risk of major elective surgery:
Randomised controlled trial of preoperative optimisation
of oxygen delivery. BMJ 1999,318:1099-103.
Zakhaleva J, Tam J, Denoya PI, Bishawi M, Bergamaschi
R. The impact of intravenous fluid administration on
complication rates in bowel surgery within an enhanced
recovery protocol: a randomized controlled trial. Colorectal Dis 2013;15:892-9.
Zhang J, Chen CQ, Lei XZ, Feng ZY, Zhu SM. Goaldirected fluid optimization based on stroke volume variation and cardiac index during one-lung ventilation in
patients undergoing thoracoscopy lobectomy operations:
a pilot study. Clinics 2013;68:1065-70.
Zheng H, Guo H, Ye JR, Chen L, Ma HP. Goal-Directed
Fluid Therapy in Gastrointestinal Surgery in Older Coronary Heart Disease Patients: Randomized Trial. World J
Surg 2013;37:2820-9.
Ziegler DW, Wright JG, Coban PS, Flancbaum L. A
prospective randomized trial of preoperative “optimization” of cardiac function in patients undergoing elective peripheral vascular surgery. Surgery 1997;122:
584-92.
81. Ghaferi AA, Birkmeyer JD, Dimick JB. Variation in hospital mortality associated with inpatient surgery. N Engl J
Med 2009;361:1368-75.
82. Pearse RM, Moreno RP, Bauer P, Pelosi P, Metnitz P,
Spies C, et al. Mortality after surgery in Europe: a 7 day
cohort study. Lancet 2012;380:1059-65.
83. Stone MD, Wilson RJ, Cross J, Williams BT. Effect of
adding dopexamine to intraoperative volume expansion
in patients undergoing major elective abdominal surgery.
Br J Anaesth 2003;91:619-24.
84. Takala J, Meier-Hellmann A, Eddleston J, Hulstaert P,
Sramek V. Effect of dopexamine on outcome after major abdominal surgery: a prospective, randomized, controlled multicenter study. European Multicenter Study
Group on Dopexamine in Major Abdominal Surgery. Crit
Care Med 2000;28:3417-23.
85. Lobo SM, Lobo FR, Polachini CA, Patini DS, Yamamoto AE, de Oliveira NE, et al. Prospective, randomized
trial comparing fluids and dobutamine optimization of
oxygen delivery in high-risk surgical patients. Crit Care
2006;10:R72.
86. Berger MM, Gradwohl-Matis I, Brunauer A, Ulmer H,
Dünser MW. Targets of perioperative fluid therapy and their
effects on postoperative outcome: a systematic review and
meta-analysis. Minerva Anestesiol 2015;81:794-808.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material
discussed in the manuscript.
Article first published online: April 14, 2016. - Manuscript accepted: April 12, 2016. - Manuscript revised: April 8, 2016. - Manuscript received: November 12, 2015.
For supplementary materials, please see the online version of this article.
Vol. 82 - No. 11
Minerva Anestesiologica
1213
© 2016 EDIZIONI MINERVA MEDICA
Online version at http://www.minervamedica.it
Minerva Anestesiologica 2016 November;82(11):1214-29
REVIEW
Targeting blood products transfusion in trauma:
what is the role of thromboelastography?
Samy FIGUEIREDO *, Audrey TANTOT, Jacques DURANTEAU
1Département
d’Anesthésie et de Réanimation, Hôpitaux Universitaires Paris-Sud, Université Paris-Sud, Hôpital de
Bicêtre, Assistance Publique – Hôpitaux de Paris, Le Kremlin-Bicêtre, France
*Corresponding author: Samy Figueiredo, Département d’Anesthésie et de Réanimation, Hôpitaux Universitaires Paris-Sud, Université Paris-Sud, Hôpital de Bicêtre, Assistance Publique – Hôpitaux de Paris, 78 rue du Général Leclerc 94275 Le Kremlin-Bicêtre
CEDEX, France. E-mail: [email protected].
A B STRACT
Viscoelastic hemostatic assays (VHAs), mainly thromboelastography (TEG) and the rotational thromboelastometry (ROTEM), provide global information on clot formation and dissolution at patient bedside, allowing fast identification of coagulation disorders. In trauma patients, VHAs are able to predict massive transfusion and mortality. These devices might
also be used for applying targeted administration of procoagulant factors (e.g. fibrinogen concentrate) as an alternative
to or in addition to using predefined fixed ratios of red blood cells: platelets: fresh frozen plasma/cryoprecipitate. These
goal-directed, individualized treatment algorithms seem to reduce blood product transfusion without deleterious effects
on patient outcome. Nevertheless, a clear outcome benefit of using VHAs remains to be demonstrated in trauma patients.
(Cite this article as: Figueiredo S, Tantot A, Duranteau J. Targeting blood products transfusion in trauma: what is the role of
thromboelastography? Minerva Anestesiol 2016;82:1214-29)
Key words: Blood coagulation disorders - Hemostatics - Thromboelastography - Blood transfusion - Blood coagulation.
U
ncontrolled hemorrhage is the leading
cause of death in trauma patients.1 Hemorrhage may lead to coagulopathy and coagulopathy is able to exacerbate bleeding, thus creating
a lethal vicious circle. Coagulation disorders
are present in about 25-35% of trauma patients
at their admission and impact significantly on
morbidity and mortality.2-4 Current knowledge
considers tissue injury and hypoperfusion as
the main drivers of acute traumatic coagulopathy, leading to systemic anticoagulation, hypocoagulability and hyperfibrinolysis through
the activation of the protein C pathway.4 Acidosis, hypothermia and hemodilution secondary to inadequate resuscitation will exacerbate
coagulopathy. Fast and accurate diagnostic of
coagulopathy is a key element for achieving
1214
adequate hemostatic resuscitation in bleeding
trauma patients. Standard laboratory coagulation tests (SLCTs) including prothrombin time
(PT), activated partial thromboplastin time
(aPTT), fibrinogen concentration (Clauss
method), D-dimer and platelets count are usually used to evaluate patients’ coagulation
status. With the descriptions of the cell-based
model of hemostasis 5 and the acute traumatic
coagulopathy,4 the limitations of SLCTs have
been increasingly recognized. First, SLCTs
only evaluate plasmatic activation without taking into account the important interactions between clotting factors, platelets and other cellular components of whole blood involved in
thrombin generation. Secondly, SCLTs do not
evaluate the overall clot strength or the balance
Minerva AnestesiologicaNovember 2016
TARGETING BLOOD PRODUCTS TRANSFUSION IN TRAUMAFIGUEIREDO
between clot formation and fibrinolysis. Thirdly, the variable processing time, from blood
sampling to obtaining results (median>80 minutes) 6 may impair the usefulness of SCLTS in
emergency settings such as trauma.
In recent years, there has been an increasing interest for viscoelastic hemostatic assays
(VHAs), mainly thromboelastography (TEG;
Hemoscope Corporation, Niles, OH, USA) and
the rotational thromboelastometry (ROTEM;
Tem International GmbH, Munich, Germany),
to provide global and functional assessment of
coagulation. These devices could enable clinicians to rapidly identify which component(s)
of the coagulation process should be targeted
during hemostatic resuscitation. This manuscript reviews recently published data on the
use of TEG/ROTEM in trauma patients and
discusses their contribution to coagulation
management and their impact on morbidity
and mortality.
Materials and methods
We searched Medline, Embase and Cochrane Library databases to identify studies in
English on thromboelastography and/or rotational thromboelastometry in trauma. We used
the MeSH terms and keywords “thromboelastography” AND “trauma”, “rotational thromboelastometry” AND “trauma”, “TEG” AND
“trauma”, “ROTEM” AND “trauma”. Abstracts were screened for eligibility. The reference lists of all articles selected for detailed
review and all relevant published reviews
were searched to find any other studies potentially eligible for inclusion. The date of the
last search was 31st January 2016. All original
studies and major review articles concerning
the use of TEG and/or ROTEM in trauma were
included. ��������������������������������������
The characteristics of the studies included in the present review are presented in
Table I.
Principles and theoretical advantages
of thromboelastography
VHAs such as TEG and ROTEM can
be performed at patient bedside and give a
graphic presentation of clot formation and
subsequent lysis. Briefly, the collected wholeblood sample is placed in a special designed
Mechanical-electrical transducer
Torsion wire
4° 75
Pin
Blood sample
Cup
4°45
A
B
TEG
ROTEM
Figure 1.—Principles of thromboelastography (A) and rotational thromboelastometry (B). Citrated or native whole-blood
is inserted in a specific cup with an activator of coagulation. A pin suspended by a torsion wire is immerged in the cup and
connected to a detector system. Cup and pin are oscillated relative to each other with movement initiated from either the cup
(TEG) or the pin (ROTEM). As fibrin forms between the cup and pin, the transmitted rotation from the cup to pin (TEG) or
the impedance of the rotation of the pin (ROTEM) are detected and a tracing is generated.
Vol. 82 - No. 11
Minerva Anestesiologica
1215
FIGUEIREDOTARGETING BLOOD PRODUCTS TRANSFUSION IN TRAUMA
cup with an activator of coagulation. Inside
the cup is suspended a pin connected to a detector system, and the cup and pin are oscillated relative to each other with movement
initiated from either the cup (TEG) or the pin
(ROTEM). As fibrin forms between the cup
and pin, the transmitted rotation from the cup
to pin (TEG) or the impedance of the rotation
of the pin (ROTEM) are detected and a trace
is generated as seen in Figure 1. This trace is
divided into several parts, each one reflecting a different stage of the hemostatic process (Figure 1). Although TEG and ROTEM
nomenclatures are different, both provide
information on the speed of clot initiation
(coagulation factors), kinetics of clot growth
(thrombin generation and cleavage of fibrinogen), clot strength, and fibrinolysis (Table I).
The contribution of fibrinogen to clot strength
can be evaluated by adding platelet inhibition
agents to the (Functional Fibrinogen assay
for TEG and FIBTEM assay for ROTEM).
Analysis of the tracings morphology and parameters will determine which component(s)
of the hemostatic process needs to be targeted
(Table II).
COAGULATION
Clot
Clot
initiation kinetics
Thromboelastography-guided
hemostatic resuscitation
Given that hemorrhage-related mortality in
trauma patients occurs within the 2 or 3 first
hours after injury,7, 8 it is crucial to start the management of coagulopathy as soon as hemorrhage is clinically suspected. This management
is part of an overall approach of damage-control resuscitation: correction of hypothermia
and acidosis, restrictive fluid administration,
permissive hypotension in selected populations
and restoration of blood volume.9 Unbalanced
transfusion strategies systematically lead to depletion and dilution of coagulation factors, increasing exsanguination. Consequently, hemostatic resuscita­tion combines repletion of red
blood cells (RBC), platelets (PLT) and coagulation factors by transfusion (RBC, PLT, fresh
frozen plasma (FFP)/cryoprecipitate) and/or
administration of pharmaceutical hemostatic
agents such as fibrino­gen concentrate (FC),
prothrombin complex concentrate (PCC), antifibrinolytic agents (mainly tranexamic acid)
and less frequently desmopressin, recombinant
factor VIIa and factor XIII.
FIBRINOLYSIS
Clot
strength
K
TEG
R
MA
LY30
MCF
Li30
α angle
α angle
CT
ROTEM
CFT
10
20
50
Time (min)
Figure 2.—Classical representation of TEG and ROTEM tracings. TEG: thrombelastography; ROTEM: rotational thromboelastometry; R-time: reaction time; CT: clotting time; K-time: coagulation time; CFT: clot formation time; MA: maximum
amplitude; MCF: maximum clot firmness; LY 30: lysis at 30 min; Li30: clot lysis at 30 min.
1216
Minerva AnestesiologicaNovember 2016
TARGETING BLOOD PRODUCTS TRANSFUSION IN TRAUMAFIGUEIREDO
Targeting red blood cell transfusion
In bleeding trauma patients, RBC transfusion aims not only at improving perfusion and
oxygen delivery but also at providing sufficient cellular phospholipid surfaces to enhance
thrombin generation.10 RBCs also increase
platelets activation and promote clotting at the
bleeding site by pushing platelets to the vascular wall.11
In addition, logistic issues make it crucial to
anticipate massive transfusion (MT) as soon as
possible. VHAs are able to predict MT, usually defined as transfusion of ≥10 RBC units
within 24 hours of trauma. Most of the studies
showed that TEG/ROTEM had similar ability
to predict MT than standard coagulation tests
performed at the laboratory, with shorter turnaround times. The following TEG values were
found to be predictive of MT: ACT>128 (odds
ratio=5; 95% CI: 1-19; P=0.01),12 α-angle
<56,13 G value 14 and hyperfibrinolysis defined by LY30>15% (OR=19) 15 or even >3%
(with specificity and PPV > 90% but very low
31% sensibility).16 TEM values predictive of
MT were: FIBTEM MCF and FIBTEM A10
(AUC-ROC curves >0.83),17 clot amplitude
at 5 minutes [CA5] ≤35 mm, which had better sensitivity [71% vs. 42%, P<0.001] but
worse specificity [85% vs. 94%] compared to
a PTr>1.2,18 abnormal MCF (OR=8 and AUCROC curve=0.8 [95% CI: 0.7-0.9]; P<0.001)
even though Hb<10g/dL was a stronger predictor of MT (OR=18). Besides MT prediction, rTEG ACT,12 TEG MA 19 and ROTEM
MCF FIBTEM 20 were found to be correlated
with the amount of blood products transfused
whereas rTEG α-angle <74.7 was predictive of
any transfusion.21 One study with a relatively
small population of transfused patients (22 of
161) found that common TEG parameters (R,
K, angle, MA, LY60) were not statistically different between transfused and not transfused
patients; only MA-ADP from Platelet-Mapping® was statistically different (P=0.004).22
Studies focusing on the performances of the
different VHAs to predict MT are summarized
in Table III.
Vol. 82 - No. 11
Targeting repletion of coagulation factors
(PLT, FFP, cryoprecipitate or PCC): VHAs
or massive transfusion protocols?
Rapid and efficient blood product administration may be achieved either by massive
transfusion protocols (MTPs) using predefined
fixed ratios of RBC:PLT:FFP/cryoprecipitate or
by applying goal-directed administration of coagulation factor concentrates guided by VHAs.
Benefits reported from MTPs using fixed
ratios are related to facilitation of communication in emergency settings between the blood
transfusion service and the clinical team and
to standardization of patient care between all
medical providers in one center. However, optimal ratio is not so easy to be defined: if it is too
low, treatment of coagulopathy is delayed and
bleeding continues; if it is too high, indiscriminate administration of blood products may lead
to transfusion-related complications (immunomodulation, acute lung injury, cardiac overload)
or waste. Recently, the PROPPR Study 8 showed
that among 680 patients predicted to receive
MT and randomly assigned to a 1:1:1 or 1:1:2
FFP:PLT:RBC transfusion ratio, no significant
difference in overall mortality at 24 hours or
30 days was detected. However, more patients
achieved hemostasis and fewer patients died of
exsanguination in the 1:1:1 group, without any
increase in morbidity outcomes. It is noteworthy that the authors used a platelet first strategy
for administering the 1:1:1 ratio whereas more
than 80% of the patients had a platelet count
≥150.000/mm3. The recently recognized phenomenon of platelet dysfunction after trauma
patients might be a rationale for this strategy.23
The “theragnostic approach” using point-ofcare VHAs allows an individualized and goaldirected hemostatic therapy with the use of coagulation factor concentrates such as FC, PCC
and less frequently rFVIIa and Factor XIII.
This strategy is primarily used in Austria and
Switzerland. A first cohort study on 131 trauma
patients described the ROTEM-guided administration of FC and PCC in addition to FFP and
PLT, and found a favourable survival rate compared with a predicted survival rate by the Trauma Injury Severity Score (TRISS).24 Given that
Minerva Anestesiologica
1217
FIGUEIREDOTARGETING BLOOD PRODUCTS TRANSFUSION IN TRAUMA
Table I.—Characteristics of all the studies included in the present review.
Reference
Year
Design
Cotton 12
2011
Prospective,
observational, single
center
Holcomb 13
2012
Pezold 14
Patients (n)
Population
Technique
Trauma patients;
Median ISS=14
rTEG
citrated blood
Retrospective,
1974
cohort study, single center
Trauma patients;
Median ISS=17
rTEG
citrated blood
2012
Retrospective
80
Trauma patients with ISS>16
42 non-MT vs.
38 MT receiving more than 10U of RBC in
the first 24 hours
rTEG
fresh whole blood
Ives 15
2012
Prospective,
single center
118
Trauma patients with an ISS>16
TEG
citrated blood
Chapman 16
2013
Prospective,
single center
289
Trauma patients;
TEG
Median ISS=30
citrated blood
MTP criteria at admission (N.=73) vs.
all non-MTP trauma patients (N.=216) in the
same period
Schöchl 17
2011
Retrospective,
observational
323
Trauma patients with an ISS >16
MT group (78) receiving more than 10U of
RBC in 24h versus non MT group (245)
ROTEM
citrated blood
Davenport 18 2011
Prospective,
observational,
cohort study
300
Trauma patients
Median ISS=12
ROTEM
citrated blood
Nystrup 19
2011
Retrospective
89
Trauma patients
Median ISS=21
TEG
citrated blood
Tauber 20
2011
Prospective,
observational,
single center
334
Trauma patients
Median ISS=34
Polytrauma (274) and isolated head injury
(60)
ROTEM
citrated blood
Jeger 21
2012
Prospective,
observational
single center
76
Trauma patients
Median ISS=29
rTEG
citrated blood
Carroll 22
2009
Prospective,
observational,
single center
161
Trauma patients
Median ISS=20
TEG and
PLT Mapping
citrated blood
Schöchl 24
2010
Retrospective
131
Trauma patients receiving more than 5U
RBC/24h
Median ISS=38
ROTEM
citrated blood
Schöchl 25
2011
Retrospective
681
Trauma patients
ROTEM
citrated blood
1218
272
Minerva AnestesiologicaNovember 2016
TARGETING BLOOD PRODUCTS TRANSFUSION IN TRAUMAFIGUEIREDO
Objectives
The ability of r-TEG to predict
early blood transfusion (<2h)
Main Results
Linear regression demonstrated that ACT predicted RBC, plasma and PLT transfusions within
the first 2 h of arrival.
ACT>128 predicted MT (>10U of RBC) in the first 6 hours (OR 5; 95% CI, 1-19; P=0.01).
The ability of r-TEG to reliably ACT-predicted RBC transfusion, and the α-angle predicted massive RBC transfusion better
predict blood component trans- than PT, aPTT or INR (P<0.001).
fusion compared to SLCTs
The α-angle was superior to fibrinogen for predicting FFP transfusion (P<0.001).
MA was superior to PLT count for predicting PLT transfusion (P<0.001).
rTEG for prediction of MT and The predictive power for MT did not differ among INR (adjusted AUC ROC=0.92), aPTT
mortality
(AUC ROC=0.90, NS), or G (AUC ROC=0.89, NS).
For MT-death, G had the greatest adjusted AUC ROC (0.93) compared with the AUC ROC
for BD (0.87, P=0.05), INR (0.88, NS), and PTT (0.89; P=NS).
MT and mortality associated
with HF
HF was present in 13 patients (11%).
HF patients had a greater need for MT (77% vs. 9%; adjusted OR=19; 95% CI, 4-101;
P<0.001) and had a greater early mortality (69% vs. 2%; adjusted OR=56; 95% CI, 7-432;
P<0.001) and in-hospital mortality (92% vs. 9%; adjusted OR=56; 95% CI, 5-650; P=0.001).
LY30 threshold associated with HF was present in 19 patients (26%) and 11 patients (15%) had LY30≥3%.
MT and mortality
LY30≥3%: performance for prediction of MT (in the first 6 hours): PPV=91%, Sens=31%,
Spe=98%, NPV=65%.
The ability of ROTEM for early The best predictive values for MT were provided by Hb and Quick value (AUC ROC=0.87
prediction of MT.
for both parameters).
Similarly high predictive values were observed for FIBTEM MCF (0.84) and FIBTEM A10
(0.83).
Detection of ATC (defined
by PTr> 1.2) with ROTEM
compared with SLCTs, POC
PT, and BD
CA5 threshold ≤35mm had a detection rate of 77% for ATC with a false positive rate of 13%.
Patients with CA5 ≤ 35 mm were more likely to receive RBC (46% vs. 17%, P<0.001) and
FFP (37% vs. 11%, P<0.001)
Better detection rate of MT with CA5=71% (vs. 43% for PTr>1.2, P<0.001).
Transfusion requirement and
mortality associated with low
clot strength (MA)
MA correlated with the amount of RBC (P=0.01), FFP (P=0.04) and PLT (P=0.03) transfused
in the first 24 h.
Patients with reduced MA demonstrated increased 30-day mortality (47% vs. 10%, P<0.001).
Low MA is independently associated with mortality after adjusting for age and ISS.
ROTEM threshold associated
with transfusion requirement
and mortality
Significant differences in mortality were detected for defined ROTEM-thresholds: FIBTEM 7
mm (21% vs. 9%, P<0.01), EXTEM MCF 45 mm (25% vs. 9%, P<0.001).
MCF FIBTEM was correlated with RBC transfusion (OR 0.9, 95% CI 0.9-0.98).
HF increased fatality rates and occurred as frequently in isolated TBI as in polytrauma.
Correlation between SLCTs and The r-TEG α-angle was the parameter with the greatest sensitivity (84%) and validity (77%)
rTEG parameters in the predic- at a cut-off of 75 degrees.
tion of any transfusion
Physicians blinded to TEG:
transfusion guided clinically
and with SLCTs results
TEG and Platelet mapping for
prediction of transfusion and
mortality
No TEG parameter was significantly different for the 22 patients who required transfusion,
except MA-ADP (P<0.01)
R and MA were correlated with mortality (both P<0.001)
ROTEM (FIBTEM, EXTEM)guided FC and PCC therapy
Difference in mortality was 14% observed vs. 28% predicted by TRISS (P<0.01) and 24 %
predicted by RISC (P=0.014).
Standard FFP transfusion com- RBC transfusion avoided in 29% of patients in the FC-PCC group compared with only 3% in
pared to PCC and FC guided by the FFP group (P<0.001).
ROTEM
PLT transfusion avoided in 91% of patients in the FC-PCC group, compared with 56% in the
FFP group (P<0.001).
Mortality was comparable between groups: 7.5% in the FC-PCC group and 10% in the FFP
group (NS).
(To be continued)
Vol. 82 - No. 11
Minerva Anestesiologica
1219
FIGUEIREDOTARGETING BLOOD PRODUCTS TRANSFUSION IN TRAUMA
Table I.—Characteristics of all the studies included in the present review (Continues).
Reference
Patients (n)
Population
Technique
Tapia 26
2013
Year
Retrospective,
before/after study
Design
289
TEG
fresh whole blood
Gonzalez 28
2015
Prospective,
randomized control trial,
single center
111
Moore 30
2014
Prospective,
observational,
single center
180
Rourke 38
2012
Prospective,
cohort study,
2 centers
517
Patients receiving≥6U of RBC in the first
24h before (N.=165) and after (N.=124)
MTP initiation
Median ISS=25
PreMTP group (165) and standardized MTP
group (N.=124)
Trauma patients meeting criteria for MTP
activation (SBP<70 mmHg and/or HR
>108/min)
Median ISS=30
TEG group (N.=56) vs. SLCT group (N.=55)
Trauma patients
Median ISS=29
3 groups according to fibrinolysis:
HF (LY30≥3%); physiologic (LY30: 0.0812.9%), and shutdown (LY30: 0-0.08%)
Trauma patients
Meyer 40
2015
Prospective,
observational,
cohort study,
single center
182
Trauma patients
Median ISS=17
TEG and
FIBTEM
citrated blood
Kornblith 41
2014
Prospective,
observational,
single center
251
Trauma patients
Median ISS=9
TEG
citrated blood
Agren 44
2014
Prospective,
observational,
single center
101
TEG
citrated blood
Schlimp 46
2013
Retrospective,
single center
157
Chapman 55
2015
Retrospective,
cohort study
572
63 patients with trauma or surgery with
ongoing bleeding and
38 healthy blood
donors
Trauma patients receiving more than 1 g
Fibrinogen within 24h
3 groups: FC group, FC-PCC group and FCPCC-FFP group
Trauma patients with≥1U of RBC
Raza 56
2013
Prospective,
cohort study,
single center
288
Trauma patients
Median ISS=10
ROTEM
citrated blood
Kashuk 68
2012
Retrospective,
before (no rTEG)/after
(with rTEG) study
68
Patients receiving more than 6 RBCs/6 hours rTEG
62% ISS >36
fresh whole blood
rTEG
fresh whole blood
TEG
citrated blood
ROTEM
citrated blood
ROTEM
citrated blood
rTEG
fresh whole blood
ATC: acute traumatic coagulopathy; BD:
����������������������������������������������������������������������������������������������������
base deficit; FC:
����������������������������������������������������������������������������������
fibrinogen concentrate; FFP: fresh frozen plasma; Hb: hemoglobin; ������������
HF: Hyperfibrinolysis; MT: massive transfusion; MTP: massive transfusion protocol; NPV: negative predictive value; NS: non significant; OR: odds ratio;
PAP: plasmin–antiplasmin complex; PCC: prothrombin complex concentrate; PLT: platelet; PPV: positive predictive value; PTr: PT ratio;
RBC: red blood cell; Sens: sensibility; SLCT: Standard laboratory coagulation test; Spe: specificity.
1220
Minerva AnestesiologicaNovember 2016
TARGETING BLOOD PRODUCTS TRANSFUSION IN TRAUMAFIGUEIREDO
Objectives
Main Results
TEG-guided resuscitation
vs. standardized MTP
TEG-guided resuscitation is equivalent to standardized MTP for patients receiving≥6U of
RBC and for blunt trauma receiving≥10U of RBC.
TEG-directed resuscitation is superior to MTP in patients with penetrating trauma
receiving≥10U of RBC.
MTP therapy worsened mortality in penetrating trauma patients receiving≥10U of RBC.
rTEG-guided vs. SLCT-guided
MTP
28-days survival in the rTEG-guided MTP group was significantly higher than the SLCT
group (P=0.032).
SLCT group required similar number of RBC units as the rTEG group (SLCT: 5 [2-11], TEG:
4.5 [2-8] (NS), but more plasma units and more PLT units in the first 2 hours of resuscitation).
Fibrinolysis threshold associated with mortality rates
Mortality rates were lower for the physiologic group (3%) compared with the HF (44%) and
shutdown (17%) groups (P=0.001).
U-shaped distribution of death related to the fibrinolysis system in response to major trauma,
with a nadir in mortality, with level of fibrinolysis after 30 minutes between 0.81% and 2.9%
Response and outcome associated with Fibrinogen replacement therapy, correlation to
ROTEM parameters.
Pre-fixed MTP including administration of RBC, FFP, PLT,
CRYO and FC and ex-vivo FC.
To compare TEG FF and
FIBTEM with each other and
with Clauss method
Patients who received fibrinogen supplementation maintained their admission fibrinogen level
(1.6 g.L-1).
Fibrinogen level was an independent predictor of mortality at 24 h and 28 days (P<0.001).
Patients who received cryoprecipitate maintained their fibrinogen levels, and had lower mortality rates than those who did not.
FF MA and FIBTEM MCF had identical correlation coefficients (both ρ=0.64,
P<0.001) and identical explained variances (both R2=0.41, P<0.001)
Performance of FF MA and FIBTEM MCF versus Clauss fibrinogen levels of 1.5, 2.0, and 2.5
g/L by ROC analysis. For all three cut-offs, FF MA and FIBTEM MCF had AUROC curves
above 0.8 with P<0.001.
Patients requiring plasma transfusion had a significantly lower admission %MA(FF) (27% vs..
31%, P<0.05).
Higher admission %MA(FF) was predictive of reduced mortality (hazard ratio, 0.815,
P<0.001).
Level of FF and contribution
of fibrinogen to clot strength
associated with coagulopathy,
transfusion requirements, and
outcomes
TEG-FF compared with Clauss TEG-FF was on average 1.0 g/L higher than the plasma fibrinogen concentration.
method for fibrinogen concentration measurement
Role of FC for rapidly increasing fibrinogen plasma levels.
FFP given only to the most
severely injured patients
To determine which rTEG tracing pattern is exclusively found
within the nonsurvivors
Comparison of Fibrinolytic
activation (FA) detected by
ROTEM vs.
plasmin–antiplasmin (PAP)
complex and D-dimer levels.
rTEG-guided resuscitation
compared with patients admitted prior to the rTEG® period
implementation.
Vol. 82 - No. 11
FIBTEM-CF at 10 min (CA10) was maintained, with a small increase in the FC-PCC group.
Fibrinogen concentration and FIBTEM CA10 were within the normal range in all groups at
24 hours.
A “diamond-shaped” tracing, with short time to MA (≤ 14 min) and complete lysis before
LY30 point, had 100% PPV for mortality.
TEM detected clot lysis only when PAP complex levels were increased to 30 times normal
(P<0.001) and antiplasmin levels were <75% of normal.
Patients with FA had increased 28-day mortality as compared with those with no FA (12% vs.
1%, P<0.001).
Patients with a MRTG > 9.2 received significantly less RBCs, FFP, and cryoprecipitate (all P
values <0.05).
G value from r-TEG was associated with survival (P=0.03).
Minerva Anestesiologica
1221
FIGUEIREDOTARGETING BLOOD PRODUCTS TRANSFUSION IN TRAUMA
Table II.—Main TEG and ROTEM parameters: definitions, representative coagulation process and therapeutic
options.
TEG parameter
R-time (min)
ROTEM parameter
CT (min)
ACT (sec)
K-time (min)
Coagulation parameter
reflected
Definition
Time from the start of the test until 2
mm amplitude
Enzymatic clotting
factors
Surrogate of R-time in the rapid-TEG Enzymatic clotting
assay, which uses tissue factor to
factors
obtain a quicker tracing
CFT (min)
angle α (degrees) angle α (degrees)
MA (mm)
MCF (mm)
G (dynes/cm2)
MCE (dynes/cm2)
LY30 (%)
LI 30 (%)
MA-FF (mm)
FIBTEM (mm)
Time from 2mm to 20mm amplitude
Angle of a tangent line between the
initial split point of the tracing and
the growing curve
Point at which clot strength reaches
its maximum measure
Overall total clot strength, calculated
from MA
Percentage of clot strength loss 30
min after reaching MA
Contribution of fibrinogen to clot
strength
Enzymatic clotting
factors
Thrombin generation
Therapeutic options
to consider
FFP or PCC if
prolonged R-time
(protamine if heparin
present)
FFP or PCC if
prolonged ACT
(protamine if heparin
present)
Fibrinogen, platelets,
factor XIII
All coagulation
interactions
Fibrinolysis
FC and/or PLT if MA
decreased
Fibrinogen
FC or cryoprecipitate
if MA decreased
TXA
TEG: thrombelastography; ROTEM: rotational thromboelastometry; R-time: reaction time; ACT: activated clotting time; CT: clotting time;
K-time: coagulation time; CFT: clot formation time; MA: maximum amplitude; MCF: maximum clot firmness; MCE: maximum clot elasticity; G: total clot strength; LY-30: lysis at 30 min; Li30: clot lysis at 30 min; MA-FF: maximum amplitude using Functional Fibrinogen assay;
FFP: fresh frozen plasma; PCC: prothrombin complex concentrate; FC: fibrinogen concentrate; TXA: tranexamic acid; PLT: platelets.
there was no control group, a second study compared these findings with the German Trauma
Registry selected by transfusion of two units or
more of FFP and then compared on transfusion
therapy. RBC transfusion was avoided in 29%
of patients in the FC-PCC group compared with
only 3% in the FFP group (P<0.001). Transfusion of PLT was avoided in 91% of patients
in the FC-PCC group, compared with 56% in
the FFP group (P<0.001). Mortality was comparable between groups: 7.5% in the FC-PCC
group and 10% in the FFP group (P=0.69).25 In
a retrospective before/after study on 289 trauma
patients without traumatic brain injury (TBI),
Tapia et al. compared TEG-directed (=before)
with MTP-directed (=after) resuscitation. The
authors reported a survival benefit with TEG-directed resuscitation compared to a fixed 3RBC:
2PFC MTP in a subgroup of penetrating trauma
patients requiring >10 units of RBC.26 All these
before/after studies should be interpreted with
caution since the results observed might be due
to the use of TEG/ROTEM and/or only to the
implementation of a protocol. Few randomized
1222
control trials (RCTs) have been conducted in
order to determine the clinical impacts of VHAguided coagulation management on mortality and morbidity in trauma patients. Weber et
al.27 included 100 cardiac surgical patients with
coagulopathy in an RCT aiming at evaluating
ROTEM-guided hemostatic therapy. The group
treated according to a ROTEM algorithm had
reduced chest tube drainage, reduced transfusion requirements and, importantly, reduced 30day mortality compared with the group treated
according to SLCTs. Gonzalez et al. included
111 trauma patients in a very recently published
RCT aiming at comparing MTP goal-directed
by TEG and MTP guided by SLCTs. Patients
resuscitated with TEG-guided MTP had significantly improved 6-hour and 28-day survivals
and received less plasmas and platelets in the
first 2 hours than patients receiving SLCTsguided MTP28. Some important points should
be noted before generalizing the findings of
this valuable study: 1) the first units of RBC
and plasma were administered according to the
clinician’s practice regardless of randomization
Minerva AnestesiologicaNovember 2016
TARGETING BLOOD PRODUCTS TRANSFUSION IN TRAUMAFIGUEIREDO
Table III.—Performances of the viscoelastic haemostatic assays to predict massive transfusion.
Reference
Year
Cotton 12
2011 Prospective,
observational,
single center
2012 Retrospective,
cohort study
single center
2012 Retrospective
Holcomb 13
Pezold 14
Design
Patients (N.)
Population
Technique
272
Trauma patients;
median ISS=14
rTEG
citrated blood
1974
Trauma patients;
median ISS=17
rTEG
citrated blood
80
Trauma patients with
rTEG
ISS>16
fresh whole
42 non-MT vs.
blood
38 MT receiving more
than 10U of RBC in the
first 24 hours
Ives 15
2012 Prospective,
single center
118
Trauma patients with
an ISS>16
TEG
citrated blood
Chapman 16
2013 Prospective,
single center
289
TEG
citrated blood
Schöchl 17
2011 Retrospective,
observational
323
Davenport 18
2011 Prospective,
observational,
cohort study
300
Trauma patients;
Median ISS=30
MTP criteria at admission (N.=73) vs.
all non-MTP trauma
patients (N.=216) in
the same period
Trauma patients with
an ISS>16
MT group (78) receiving more than 10U of
RBC in 24h vs. non
MT group (245)
Trauma patients
Median ISS=12
Main Results
ACT>128 predicted MT (>10U
of RBC) in the first 6 hours
(OR 5; 95%CI, 1-19; P=0.01)
α-angle predicted massive RBC
transfusion better than PT,
aPTT or INR (P<0.001)
The predictive power for MT
did not differ among INR (adjusted AUC ROC=0.92), aPTT
(AUC ROC=0.90, NS), or G
(AUC ROC=0.89, NS)
For MT-death, G had the greatest adjusted AUC ROC (0.93)
compared with the AUC ROC
for BD (0.87, P=0.05), INR
(0.88, NS), and PTT (0.89;
P=NS)
HF was present in 13 patients
(11%).
HF patients had a greater
need for MT (77% vs. 9%;
adjusted OR=19; 95%CI,
4-101; P<0.001) and had a
greater early mortality (69%
vs. 2%; adjusted OR=56;
95% CI, 7-432; P<0.001) and
in-hospital mortality (92% vs.
9%; adjusted OR=56; 95% CI,
5-650; P=0.001).
LY30≥3%: performance for
prediction of MT (in the first 6
hours): PPV=91%, Sens=31%,
Spe=98%, NPV=65%.
ROTEM
citrated blood
The best predictive values for
MT were provided by Hb and
Quick value (AUC ROC=0.87)
ROTEM
citrated blood
Better detection rate of MT
with CA5=71% (vs. 43% for
PTr>1.2, P<0.001).
ATC: acute traumatic coagulopathy; AUC: area under the ROC curve; BD: base deficit; FC: fibrinogen concentrate; FFP: fresh frozen plasma;
Hb: hemoglobin; HF: Hyperfibrinolysis; MT: massive transfusion; MTP: massive transfusion protocol; NPV: negative predictive value; NS:
non-significant; OR: odds ratio; PCC: prothrombin complex concentrate; PLT: platelet; PPV: positive predictive value; PTr: PT ratio; RBC:
red blood cell; ROC: Receiver Operating characteristic; Sens: sensibility; SLCT: Standard laboratory coagulation test; Spe: specificity.
group; 2) tranexamic acid was not widely used
(12% of the whole cohort); 3) there are no data
on the time and the methods used to achieve hemostasis. Additional studies, including RCTs,
are warranted to confirm these beneficial effects of VHA-guided coagulation management
on mortality and morbidity in trauma patients.
Vol. 82 - No. 11
Noteworthy, using TEG/ROTEM devices
does not necessarily imply using only FC and
PCC. Thrombin generation does not represent a
major issue in trauma patients. Since the administration of PCC will increase thrombin generation for several days PCC could increase the risk
of thrombosis and should therefore be used with
Minerva Anestesiologica
1223
FIGUEIREDOTARGETING BLOOD PRODUCTS TRANSFUSION IN TRAUMA
caution.29 Factor replacement with FC and PCC
is not considered equivalent to plasma or cryoprecipitate transfusion, as they do not contain important proteins present in these blood products,
particularly Factors V, VIII and XIII. In addition,
plasma transfusion may confer more benefits
than FC and PCC beyond factor replacement.30
Plasma is an iso-osmolar resuscitation fluid with
higher oncotic pressure than 0.9% saline and
contains amount of proteins such as albumin,
immunoglobulins, apolipoproteins and protease
inhibitors. Interestingly, resuscitation with FFP
compared to lactated Ringer’s or normal saline
has been shown to attenuate endothelial glycocalyx disruption following hemorrhagic shock 31-33.
Further studies are needed to better characterize
potential benefits of plasma transfusion beyond
restoration of coagulation.
In summary, individualized hemostatic resuscitation guided by VHAs and use of MTPs
with predefined FFP:PLT:RBC ratios have both
strengths and weaknesses. VHAs allow goaldirected management of coagulopathy but it
takes time to run the samples and to obtain the
results. Activation of the MTP by the clinician
leads to rapid and timely delivery of all predefined blood products but the optimal ratio is
still being debated. Interestingly, it is not mandatory to choose a single strategy and to reject
the other one: both strategies may be used at
different times the resuscitation. Indeed, some
authors advocate to initiate balanced transfusion therapy with transfusion packages using
a 1:1:1 ratio in the early phase of massive
bleeding, and then to use goal-directed administration of hemostatic blood components and
coagulation factor concentrates during further
resuscitation when information become available from laboratory or point-of-care tests
(Copenhagen concept).34 Although this third
strategy has recently been recommended,35 additional research is required in this field to precisely determine the respective role of VHAs
and MTPs in coagulation resuscitation.
Targeting fibrinogen supplementation
Fibrinogen concentration decreases early after injury in coagulopathic trauma patients and
1224
low levels have been associated with worse
outcome.36-38 Thus, fibrinogen supplementation is recommended in the management of
trauma-induced coagulopathy.39 The optimal
concentration of fibrinogen in this context
has yet to be determined, but current consensus guidelines suggest a target level of 1.5-2.0
g/L.39 Clauss method is considered as a gold
standard for measuring fibrinogen concentration in plasma. However, this method has slow
turnaround times and do not reflect the qualitative (or functional) aspect of fibrinogen. Specific VHAs exist for estimating fibrinogen levels, including functional fibrinogen assay (FF)
for TEG and FIBTEM for ROTEM. A recent
study showed that both FF and FIBTEM correlated with fibrinogen level measured by Clauss
method (both ρ=0.64, P<0.001), as well as with
each other, 40 confirming previous results.41
Despite these correlations, absolute values
obtained by using FF and FIBTEM are not
equivalent. FF MA values have been reported
to be higher than FIBTEM MCF values.42-44
Technological differences between the two
devices along with differential effects of the
platelets inhibitors used in the two tests (cytochalasin D for FIBTEM and abciximab for
FF assay) may contribute to this phenomenon.
Target values and thresholds for hemostatic
therapy could consequently depend on the device used and clinicians should be aware of
this issue. Fibrinogen supplementation may be
provided by FFP, cryoprecipitate and/or FC.
Plasma contains only low levels of fibrinogen 45 and some authors have shown that using
plasma as the unique source of fibrinogen supplementation is not sufficient enough to correct fibrinogen depletion.38, 46 Cryoprecipitate
contains variable amounts of fibrinogen (along
with von Willebrand factor, factor VIII and
factor XIII), is not virus inactivated and has to
be pooled and thawed before administration. In
contrast, FC contains a well-defined concentration of fibrinogen and is immediately available for use.47 In Europe, FC administration
is authorized both in congenital and acquired
coagulopathies, whereas its use is restricted to
congenital coagulopathies in the United States.
In summary, fibrinogen supplementation is
Minerva AnestesiologicaNovember 2016
TARGETING BLOOD PRODUCTS TRANSFUSION IN TRAUMAFIGUEIREDO
required in bleeding trauma patients and has
probably to be done by administration of FC or
cryoprecipitate guided by VHAs (initially 3-4
g or 50 mg/kg, ����������������������������������
respectively����������������������
) ��������������������
in addition to plas39
ma transfusion. Further studies are warranted
to standardize TEG/ROTEM specific fibrinogen assays and to determine their clinical usefulness for guiding fibrinogen administration.
Targeting antifibrinolytic therapy
The CRASH-2 trial (Clinical Randomisation
of Antifibrinolytic therapy in Significant Hemorrhage) 48 assessed the effects of early administration of tranexamic acid (TXA) on 28-day
hospital mortality, vascular occlusive events,
and transfusions in adult trauma patients. This
trial randomized 20,211 patients with or at risk
of significant bleeding to either TXA (loading
dose 1 g over 10 minutes followed by infusion
of 1 g over 8 hours) or placebo within 8 h of
trauma. All-cause mortality was significantly
reduced with TXA (1463 [14.5%] TXA vs.
1613 [16.0%] placebo; relative risk 0.91, 95%
CI 0.85-0.97; P=0.0035), and the risk of death
due to bleeding was also significantly reduced
(489 [4.9%] vs. 574 [5.7%]; relative risk 0.85,
95% CI 0.76-0.96; P=0.0077). All cause mortality reduction was 1.5%, giving a number
needed to treat (NNT) of 67 to save one life over
28 days. On the basis of the CRASH-2 trial results, it has been estimated that TXA use could
save between 70,000 and 100,000 lives per year
worldwide.49 Greatest impact of TXA on mortality was in patients with severe shock (systolic
blood pressure [SBP]<75 mmHg).50 There was
no difference in the rate of venous thrombotic
events or stroke between groups, whereas there
was a statistically significant reduced risk of
myocardial infarction in the TXA group. Significant limitations of the CRASH-2 Trial have already been described: only 50% of study cohort
received blood transfusion and TXA did not
reduce RBC transfusion requirements; no data
available regarding additional blood products
(plasma, platelets) administered and whether
damage-control resuscitation and/or MTP were
used; fibrinolysis and coagulation assessments
were not part of the study design; moreover,
Vol. 82 - No. 11
TXA treatment given after 3 h after injury was
associated with an increased risk of death caused
by bleeding (4.4% vs. 3.1%; RR, 1.44; 95% CI,
1.12-1.84; p=0.004). For all these reasons, the
mechanism by which TXA reduced mortality in
the CRASH-2 trial remains unclear. Recently,
studies evaluating the benefits of TXA administration in mature trauma systems with MTP
protocols and early hemodynamic resuscitation
have shown conflicting results.51-52 This calls
for caution of indiscriminate use of antifibrinolytic drugs and raises the question of whether
coagulation assessments should be performed
prior to their administration. The traditional
laboratory test for diagnosing pathologic fibrinolysis is the euglobulin lysis time (ELT) but it
requires 4 hours to be prepared.50 Increased values of ROTEM LI30 or TEG LY30 can reflect
hyperfibrinolysis. Thus, some authors have recommended using TEG or ROTEM to determine
which patient to treat with TXA. However, definitions and thresholds values for pathologic fibrinolysis are not clearly defined. For example,
TEG manufacturer reports the normal range of
clot lysis at 30 min to be 0% to 7.5% whereas
some authors use the cut-point of > 3% lysis
since this value was associated with an increase
in mortality.16,53
Interestingly, Moore et al. identified a Ushape distribution of death related to fibrinolysis in response to trauma, with an increased
risk of death with level of fibrinolysis after 30
min above 2.9% and below 0.8%.54 The same
team identified a particular “diamond-shaped”
TEG tracing pattern (short time to MA and
complete lysis before LY30 point) with 100%
positive predictive value for mortality.55 First
derivative velocity curve values generated by
rTEG measuring lysis such as maximum rate
of lysis (MRL) and total lysis (TL) were also
described as a potential tool to define fibrinolysis.53 Another issue regarding the use of ROTEM or TEG to detect fibrinolysis is a possible lack of sensitivity to measure endogenous
fibrinolytic activity. Indeed, Raza et al. showed
that 5% of patients had TEM hyperfibrinolysis (defined as maximum clot lysis (ML) exceeding 15% of the maximum clot firmness)
whereas 57% of these patients had evidence
Minerva Anestesiologica
1225
FIGUEIREDOTARGETING BLOOD PRODUCTS TRANSFUSION IN TRAUMA
of fibrinolysis according to plasmin-antiplasmin (PAP) complex and D-dimer levels.56 In
summary, additional research is required to
improve the ability of TEG/ROTEM to detect fibrinolysis, to standardize HF measurement and to determine whether these devices
are useful to guide antifibrinolytic therapy in
trauma patients. Pending further studies, TXA
should be administered within 3 h after injury
to the trauma patient who is bleeding or at risk
of significant hemorrhage.39
Targeting whole blood transfusion?
Whole blood (WB) was the primary resuscitation fluid in military settings through the
beginning of the Vietnam War. After crucial
advances in whole blood fractionation in the
1960s and the 1970s, blood donor centers began supplying hospitals with individual components (RBCs, plasma, PLT) and removed WB as
a readily available product. Damage control resuscitation has recently emerged and promotes
earlier and more aggressive use of plasma and
PLT, in ratios approximating that of WB. This
has renewed interest in the potential benefits of
WB for some authors. A recent single-center
randomized trial found that PLT-modified WB
(MWB) was not superior to blood component
therapy in reducing transfusion volumes or
decreasing 24-hour and 30-day mortality in
severely injured civilian patients predicted to
receive massive transfusion. A benefit was seen
only in a sensitivity analysis where patients
with brain injury were excluded, where the use
of MWB resulted in significantly reduced transfusion volumes compared to blood component
therapy.57 However, many issues remain to be
addressed: use of fresh warm or cooled WB,
PLT-supplementation of WB requirement or
not, maximal duration of WB storage?
Targeting other hemostatic drugs: factor XIII,
factor VIIa?
Factor XIII (FXIII), which cross-links fibrin
monomers, plays a key role in the stabilisation of the fibrin clot. Recent in vitro ROTEM
studies using different models of hemodilution
1226
have shown that FXIII administration can reverse dilutional coagulopathy, although activated platelets and fibrinogen supplementation
are required to observe this effect.58, 59 FXIII
seems to exhibit some in vitro antifibrinolytic
properties.60 Further studies with in vivo FXIII
administration are needed to assess the clinical impact of FXIII supplementation in trauma
settings. Thromboelastography can be used to
monitor rFVIIa administration although there
is no rFVIIa-specific parameter and moreover,
recent European guidelines 39 suggest that the
use of recombinant factor recombinant activated coagulation factor VII (rFVIIa) should
be considered only if first-line treatment with
a combination of surgical approaches, bestpractice use of blood products and antifibrinolytics, correction of acidosis, hypothermia and
hypocalcemia fail to control bleeding.
Limits of thromboelastography
Although viscoelastic hemostatic assays
(VHA) provide global information on clot formation and dissolution, these devices do not
assess the endothelial contribution to hemostasis since an activator is directly added to initiate coagulation during the test.61 Consequently,
diagnosis of conditions such as von Willebrand
disease or endothelial dysfunction is not possible with VHAs. Additionally, platelet inhibition or dysfunction may not be identified with
the standard assays, unless thrombin is inhibited and platelet ago­nists are utilized (TEGPlatelet Mapping assay for example). A recent
study also showed that TEG results might be
normal in patients receiving warfarin 62 and
capacity of TEG to detect new oral anticoagulants remains to be determined in vivo.63
Some conditions have to be met for proper
use of TEG/ROTEM as a POC device: the machines must be calibrated daily and, moreover,
training and qualification must be put in place
for anyone performing the test since running
samples requires technical skills. Despite the
material costs of all tests and devices, the use
of VHAs has been proven to be cost-effective
in cardiac surgery and in trauma settings.64, 65
Another limitation is the lack of standardiza-
Minerva AnestesiologicaNovember 2016
TARGETING BLOOD PRODUCTS TRANSFUSION IN TRAUMAFIGUEIREDO
tion of VHAs due to the considerable heterogeneity in methods, reagents and parameters
measured. Indeed, TEG and ROTEM may not
be fully equivalent, as the results provided by
these devices do not seem to be interchangeable.66, 67 Different TEG/ROTEM-based transfusion algorithms for trauma patients have recently been proposed.14, 26, 28, 68 However, the
data used to create these algorithms are often
missing and no prospective validations or comparisons to other algorithms have been reported. Moreover, there is no consensus either on
the threshold values that justify transfusion or
on which hemostatic product should be administered for a given abnormal value (for example
a low MA value on TEG may lead to PLT transfusion or desmopressin administration depending on which algorithm one is referring to).
Finally, the lack of large and well-conducted
RCTs assessing the real benefit of VHAs on
patient outcomes represents the main limitation of these devices in trauma patients.69
Conclusions
In trauma patients, TEG and ROTEM allow
prediction of massive transfusion requirement
and mortality, and some parameters show good
correlation with the amount of blood products
transfused. VHAs can identify at patient bedside which component(s) of the hemostatic process is (are) impaired and should be targeted:
coagulation factors, fibrinogen, platelets and/or
fibrinolysis. Lack of standardization among the
different available devices/assays and the various therapeutic options (e.g. FFP or coagulation
factors concentrates) limits the widespread use
of VHAs-based coagulation algorithms. Goaldirected, individualized treatment algorithms
may improve patient outcome, as it has been
shown for cardiac surgery in prospective randomized trials. This outcome benefit remains
to be clearly demonstrated in trauma settings.
Key messages
—— Viscoelastic hemostatic assays
(VHAs), thromboelastography (TEG) and
Vol. 82 - No. 11
rotational thromboelastometry (ROTEM),
provide global and functional assessment
of coagulation at patient bedside.
—— In trauma patients, VHAs have been
shown to predict massive transfusion and
mortality.
—— These devices allow goal-directed
administration of hemostatic blood components (mainly coagulation factor concentrates such as fibrinogen concentrate)
as an alternative to or in conjunction with
transfusion protocols using fixed ratios of
red blood cells, platelets and fresh frozen
plasma.
—— Standardization of the different TEG
and ROTEM assays as well as high quality
evidence of improvement in patient outcomes are mandatory to enable the widespread use of VHAs.
References
 1.Evans JA, van Wessem KJP, McDougall D, Lee KA,
Lyons T, Balogh ZJ. Epidemiology of traumatic deaths:
comprehensive population-based assessment. World J
Surg 2010;34:158-63.
 2.MacLeod JBA, Lynn M, McKenney MG, Cohn SM,
Murtha M. Early coagulopathy predicts mortality in trauma. J Trauma 2003;55:39-44.
 3.Maegele M, Lefering R, Yucel N, Tjardes T, Rixen D,
Paffrath T, et al. Early coagulopathy in multiple injury:
an analysis from the German Trauma Registry on 8724
patients. Injury 2007;38:298-304.
 4.Brohi K, Singh J, Heron M, Coats T. Acute traumatic coagulopathy: The Journal of Trauma: injury, Infection, and
Critical Care 2003;54:1127-30.
 5.Roberts HR, Hoffman M, Monroe DM. A cell-based
model of thrombin generation. Semin Thromb Hemost
2006;32(Suppl 1):32-8.
 6.Toulon P, Ozier Y, Ankri A, Fléron M-H, Leroux G, Samama CM. Point-of-care versus central laboratory coagulation testing during haemorrhagic surgery. A multicenter study. Thromb Haemost 2009;101:394-401.
  7. Holcomb JB, del Junco DJ, Fox EE, Wade CE, Cohen MJ,
Schreiber MA, et al. The prospective, observational, multicenter, major trauma transfusion (PROMMTT) study:
comparative effectiveness of a time-varying treatment
with competing risks. JAMA Surg 2013;148:127-36.
  8. Holcomb JB, Tilley BC, Baraniuk S, Fox EE, Wade CE,
Podbielski JM, et al. Transfusion of plasma, platelets, and
red blood cells in a 1:1:1 vs a 1:1:2 ratio and mortality
in patients with severe trauma: the PROPPR randomized
clinical trial. JAMA 2015;313:471-82.
  9. Kaafarani HMA, Velmahos GC. Damage Control Resuscitation In Trauma. Scand J Surg 2014;103:81-8.
10. Whelihan MF, Mann KG. The role of the red cell membrane in thrombin generation. Thromb Res 2013;131:37782.
11. Du VX, Huskens D, Maas C, Al Dieri R, de Groot PG, de
Minerva Anestesiologica
1227
FIGUEIREDOTARGETING BLOOD PRODUCTS TRANSFUSION IN TRAUMA
Laat B. New insights into the role of erythrocytes in thrombus formation. Semin Thromb Hemost 2014;40:72-80.
12.Cotton BA, Faz G, Hatch QM, Radwan ZA, Podbielski J,
Wade C, et al. Rapid Thrombelastography Delivers RealTime Results That Predict Transfusion Within 1 Hour of
Admission: The Journal of Trauma: Injury, Infection, and
Critical Care 2011;71:407-17.
13. Holcomb JB, Minei KM, Scerbo ML, Radwan ZA, Wade
CE, Kozar RA, et al. ������������������������������
Admission rapid thrombelastography can replace conventional coagulation tests in the
emergency department: experience with 1974 consecutive trauma patients. Ann Surg 2012;256:476-86.
14. Pezold M, Moore EE, Wohlauer M, Sauaia A, Gonzalez
E, Banerjee A, et al. Viscoelastic clot strength predicts
coagulation-related mortality within 15 minutes. Surgery
2012;151:48-54.
15.Ives C, Inaba K, Branco BC, Okoye O, Schochl H, Talving P, et al. Hyperfibrinolysis elicited via thromboelastography predicts mortality in trauma. J Am Coll Surg
2012;215:496-502.
16.Chapman MP, Moore EE, Ramos CR, Ghasabyan A, Harr
JN, Chin TL, et al. Fibrinolysis greater than 3% is the
critical value for initiation of antifibrinolytic therapy. J
Trauma Acute Care Surg 2013;75:961-7;discussion 967.
17.Schöchl H, Cotton B, Inaba K, Nienaber U, Fischer H,
Voelckel W, et al. FIBTEM provides early prediction of
massive transfusion in trauma. Crit Care 2011;15:R265.
18. Davenport R, Manson J, De’Ath H, Platton S, Coates
A, Allard S, et al. Functional definition and characterization of acute traumatic coagulopathy. Crit Care Med
2011;39:2652-8.
19.Nystrup, Kristin B, Nis A Windeløv, Annemarie B Thomsen, and Pär I Johansson, et al. Reduced clot strength
upon admission, evaluated by thrombelastography
(TEG), in trauma patients is independently associated
with increased 30-day mortality. Scand J Trauma Resusc
Emerg Med 2011;19:52.
20.Tauber H, Innerhofer P, Breitkopf R, Westermann I, Beer
R, El Attal R, et al. Prevalence and impact of abnormal
ROTEM(R) assays in severe blunt trauma: results of the
“Diagnosis and Treatment of Trauma-Induced Coagulopathy (DIA-TRE-TIC) study.” Br J Anaesth 2011;107:37887.
21. Jeger V, Willi S, Liu T, Yeh DD, De Moya M, Zimmermann H, et al. The Rapid TEG α-Angle may be a sensitive
predictor of transfusion in moderately injured blunt trauma patients. ScientificWorldJournal 2012;2012:821794.
22.Carroll RC, Craft RM, Langdon RJ, Clanton CR, Snider
CC, Wellons DD, et al. Early evaluation of acute traumatic coagulopathy by thrombelastography. Transl Res
2009;154:34-9.
23. Wohlauer MV, Moore EE, Thomas S, Sauaia A, Evans E,
Harr J, et al. Early platelet dysfunction: an unrecognized
role in the acute coagulopathy of trauma. J Am Coll Surg
2012;214:739-46.
24.Schöchl H, Nienaber U, Hofer G, Voelckel W, Jambor
C, Scharbert G, et al. Goal-directed coagulation management of major trauma patients using thromboelastometry
(ROTEM)-guided administration of fibrinogen concentrate and prothrombin complex concentrate. Crit Care
2010;14:R55.
25.Schöchl H, Nienaber U, Maegele M, Hochleitner G, Primavesi F, Steitz B, et al. Transfusion in trauma: thromboelastometry-guided coagulation factor concentratebased therapy versus standard fresh frozen plasma-based
therapy. Crit Care 2011;15:R83.
26.Tapia NM, Chang A, Norman M, Welsh F, Scott B,
Wall MJ, et al. TEG-guided resuscitation is superior to
standardized MTP resuscitation in massively transfused
penetrating trauma patients. J Trauma Acute Care Surg
2013;74:378-85;discussion 385-6.
1228
27. Weber CF, Görlinger K, Meininger D, Herrmann E, Bingold T, Moritz A, et al. Point-of-care testing: a prospective, randomized clinical trial of efficacy in coagulopathic
cardiac surgery patients. Anesthesiology 2012;117:53147.
28.Gonzalez E, Moore EE, Moore HB, Chapman MP,
Chin TL, Ghasabyan A, et al. Goal-directed hemostatic
resuscitation of trauma-induced coagulopathy: a pragmatic randomized clinical trial comparing a viscoelastic assay to conventional coagulation assays. Ann Surg
2016;263:1051-9.
29.Schöchl H, Voelckel W, Maegele M, Kirchmair L,
Schlimp CJ. Endogenous thrombin potential following
hemostatic therapy with 4-factor prothrombin complex
concentrate: a 7-day observational study of trauma patients. Crit Care 2014;18:R147.
30. Moore EE, Chin TL, Chapman MC, Gonzalez E, Moore
HB, Silliman CC, et al. Plasma first in the field for postinjury hemorrhagic shock. Shock 2014;41(Suppl 1):35-8.
31. Peng Z, Pati S, Potter D, Brown R, Holcomb JB, Grill R,
et al. Fresh frozen plasma lessens pulmonary endothelial
inflammation and hyperpermeability after hemorrhagic
shock and is associated with loss of syndecan 1. Shock
2013;40:195-202.
32. Dekker SE, Sillesen M, Bambakidis T, Jin G, Liu B, Boer
C, et al. Normal saline influences coagulation and endothelial function after traumatic brain injury and hemorrhagic shock in pigs. Surgery 2014;156:556-63.
33.Torres LN, Sondeen JL, Ji L, Dubick MA, Torres Filho
I. Evaluation of resuscitation fluids on endothelial glycocalyx, venular blood flow, and coagulation function after
hemorrhagic shock in rats. J Trauma Acute Care Surg
2013;75:759-66.
34.Stensballe J, Ostrowski SR, Johansson PI. Viscoelastic guidance of resuscitation. Curr Opin Anaesthesiol
2014;27:212-8.
35.Rossaint R, Bouillon B, Cerny V, Coats TJ, Duranteau J,
Fernández-Mondéjar E, et al. The European guideline on
management of major bleeding and coagulopathy following trauma: fourth edition. Crit Care 2016;20:100.
36.Inaba K, Karamanos E, Lustenberger T, Schöchl H, Shulman I, Nelson J, et al. Impact of fibrinogen levels on outcomes after acute injury in patients requiring a massive
transfusion. J Am Coll Surg 2013;216:290-7.
37. Hagemo JS, Stanworth S, Juffermans NP, Brohi K, Cohen M, Johansson PI, et al. Prevalence, predictors and
outcome of hypofibrinogenaemia in trauma: a multicentre observational study. Crit Care 2014;18:R52.
38.Rourke C, Curry N, Khan S, Taylor R, Raza I, Davenport
R, et al. Fibrinogen levels during trauma hemorrhage,
response to replacement therapy, and association with
patient outcomes. J Thromb Haemost 2012;10:1342-51
39.Spahn DR, Bouillon B, Cerny V, Coats TJ, Duranteau J,
Fernández-Mondéjar E, et al. Management of bleeding
and coagulopathy following major trauma: an updated
European guideline. Crit Care 2013;17:R76.
40. Meyer MAS, Ostrowski SR, Sørensen AM, Meyer ASP,
Holcomb JB, Wade CE, et al. Fibrinogen in trauma,
an evaluation of thrombelastography and rotational
thromboelastometry fibrinogen assays. J Surg Res
2015;194:581-90.
41. Kornblith LZ, Kutcher ME, Redick BJ, Calfee CS, Vilardi RF, Cohen MJ. Fibrinogen and platelet contributions to clot formation: implications for trauma resuscitation and thromboprophylaxis. J Trauma Acute Care Surg
2014;76:255-6;discussion 262-3.
42.Solomon C, Sørensen B, Hochleitner G, Kashuk J, Ranucci M, Schöchl H. Comparison of whole blood fibrinbased clot tests in thrombelastography and thromboelastometry. Anesth Analg 2012;114:721-30.
43.Solomon C, Baryshnikova E, Schlimp CJ, Schöchl H,
Minerva AnestesiologicaNovember 2016
TARGETING BLOOD PRODUCTS TRANSFUSION IN TRAUMAFIGUEIREDO
Asmis LM, Ranucci M. FIBTEM PLUS provides an
improved thromboelastometry test for measurement of
fibrin-based clot quality in cardiac surgery patients. Anesth Analg 2013;117:1054-62.
44.Agren A, Wikman AT, Ostlund A, Edgren G. TEG® functional fibrinogen analysis may overestimate fibrinogen
levels. Anesth Analg 2014;118:933-5.
45. Theusinger OM, Baulig W, Seifert B, Emmert MY, Spahn
DR, Asmis LM. Relative concentrations of haemostatic
factors and cytokines in solvent/detergent-treated and
fresh-frozen plasma. Br J Anaesth 2011;106:505-11.
46.Schlimp CJ, Voelckel W, Inaba K, Maegele M, Schöchl
H. Impact of fibrinogen concentrate alone or with prothrombin complex concentrate (+/- fresh frozen plasma)
on plasma fibrinogen level and fibrin-based clot strength
(FIBTEM) in major trauma: a retrospective study. Scand
J Trauma Resusc Emerg Med 2013;21:74.
47. Maung AA, Kaplan LJ. Role of fibrinogen in massive injury. Minerva Anestesiol 2014;80:89-95.
48.CRASH-2 trial collaborators, Shakur H, Roberts I,
Bautista R, Caballero J, Coats T, et al. Effects of tranexamic acid on death, vascular occlusive events, and
blood transfusion in trauma patients with significant
haemorrhage (CRASH-2): a randomised, placebo-controlled trial. Lancet 2010;376:23-32.
49. Ker K, Kiriya J, Perel P, Edwards P, Shakur H, Roberts
I. Avoidable mortality from giving tranexamic acid to
bleeding trauma patients: an estimation based on WHO
mortality data, a systematic literature review and data
from the CRASH-2 trial. BMC Emerg Med 2012;12:3.
50.Napolitano LM, Cohen MJ, Cotton BA, Schreiber MA,
Moore EE. Tranexamic acid in trauma: how should we
use it? J Trauma Acute Care Surg 2013;74:1575-86.
51.Cole E, Davenport R, Willett K, Brohi K. Tranexamic
acid use in severely injured civilian patients and the effects on outcomes: a prospective cohort study. Ann Surg
2015;261:390-4.
52. Harvin JA, Peirce CA, Mims MM, Hudson JA, Podbielski JM, Wade CE, et al. The impact of tranexamic acid
on mortality in injured patients with hyperfibrinolysis. J
Trauma Acute Care Surg 2015;78:905-9;discussion 90911.
53.Cotton BA, Harvin JA, Kostousouv V, Minei KM, Radwan ZA, Schöchl H, et al. Hyperfibrinolysis at admission
is an uncommon but highly lethal event associated with
shock and prehospital fluid administration. J Trauma
Acute Care Surg 2012;73:365-70;discussion 370.
54. Moore HB, Moore EE, Gonzalez E, Chapman MP, Chin
TL, Silliman CC, et al. Hyperfibrinolysis, physiologic
fibrinolysis, and fibrinolysis shutdown: the spectrum of
postinjury fibrinolysis and relevance to antifibrinolytic
therapy. J Trauma Acute Care Surg 2014;77:811-7;discussion 817.
55.Chapman MP, Moore EE, Moore HB, Gonzalez E, Morton AP, Chandler J, et al. The “Death Diamond”: Rapid
thrombelastography identifies lethal hyperfibrinolysis. J
Trauma Acute Care Surg 2015;79:925-9.
56.Raza I, Davenport R, Rourke C, Platton S, Manson J,
Spoors C, et al. The incidence and magnitude of fibrinolytic activation in trauma patients. J Thromb Haemost
2013;11:307-14.
57.Cotton BA, Podbielski J, Camp E, Welch T, del Junco D,
Bai Y, et al. A randomized controlled pilot trial of modified whole blood versus component therapy in severely
injured patients requiring large volume transfusions. Ann
Surg 2013;258:527-32;discussion 532-3.
58.Schlimp CJ, Cadamuro J, Solomon C, Redl H, Schöchl
H. The effect of fibrinogen concentrate and factor XIII on
thromboelastometry in 33% diluted blood with albumin,
gelatine, hydroxyethyl starch or saline in vitro. Blood
Transfus 2013;11:510-7.
59. Kind SL, Spahn-Nett GH, Emmert MY, Eismon J, Seifert
B, Spahn DR, et al. Is dilutional coagulopathy induced
by different colloids reversible by replacement of fibrinogen and factor XIII concentrates? Anesth Analg
2013;117:1063-71..
60. Dirkmann D, Görlinger K, Gisbertz C, Dusse F, Peters J.
Factor XIII and tranexamic acid but not recombinant factor VIIa attenuate tissue plasminogen activator-induced
hyperfibrinolysis in human whole blood. Anesth Analg
2012;114:1182-8.
61. Haas T, Görlinger K, Grassetto A, Agostini V, Simioni P,
Nardi G, et al. Thromboelastometry for guiding bleeding
management of the critically ill patient: a systematic review of the literature. Minerva Anestesiol 2014;80:132035.
62. Dunham CM, Rabel C, Hileman BM, Schiraldi J, Chance
EA, Shima MT, et al. TEG® and RapidTEG® are unreliable for detecting warfarin-coagulopathy: a prospective
cohort study. Thromb J 2014;12:4.
63. Dias JD, Norem K, Doorneweerd DD, Thurer RL, Popovsky MA, Omert LA. Use of Thromboelastography
(TEG) for Detection of New Oral Anticoagulants. Arch
Pathol Lab Med 2015;139:665-73.
64.Whiting P, Al M, Westwood M, Ramos IC, Ryder S,
Armstrong N, Misso K, Ross J, Severens J, Kleijnen J.
Viscoelastic point-of-care testing to assist with the diagnosis, management and monitoring of haemostasis: a
systematic review and cost-effectiveness analysis. Health
Technol Assess 2015;19:1-228.
65.Nardi G, Agostini V, Rondinelli B, Russo E, Bastianini
B, Bini G, et al. Trauma-induced coagulopathy: impact of
the early coagulation support protocol on blood product
consumption, mortality and costs. Crit Care 2015;19, 83.
66.Venema LF, Post WJ, Hendriks HGD, Huet RCG, de
Wolf JTW, de Vries AJ. An assessment of clinical interchangeability of TEG and RoTEM thromboelastographic
variables in cardiac surgical patients. Anesth Analg
2010;111:339-44.
67.Sankarankutty A, Nascimento B, Teodoro da Luz L, Rizoli S. TEG® and ROTEM® in trauma: similar test but different results? World J Emerg Surg 2012;22;7(Suppl 1):S3.
68. Kashuk JL, Moore EE, Wohlauer M, Johnson JL, Pezold
M, Lawrence J, et al. Initial experiences with pointof-care rapid thrombelastography for management of
life-threatening postinjury coagulopathy. Transfusion
2012;52:23-33.
69. Hunt H, Stanworth S, Curry N, Woolley T, Cooper C,
Ukoumunne O, et al. Thromboelastography (TEG) and
rotational thromboelastometry (ROTEM) for trauma induced coagulopathy in adult trauma patients with bleeding. Cochrane Database Syst Rev 2015;CD010438.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material
discussed in the manuscript.
Article first published online: September 8, 2016. - Manuscript accepted: September 7, 2016. - Manuscript revised: August 30, 2016. Manuscript received: January 30, 2016.
Vol. 82 - No. 11
Minerva Anestesiologica
1229
© 2016 EDIZIONI MINERVA MEDICA
Online version at http://www.minervamedica.it
Minerva Anestesiologica 2016 November;82(11):1230-4
JOURNAL CLUB CRITIQUE
Failure of statins in ARDS:
the quest for the Holy Grail continues
David GRIMALDI, Arthur DURAND, James GLEESON, Fabio S. TACCONE *
Department of Intensive Care, Hôpital Erasme, Université Libre de Bruxelles, Brussels, Belgium
*Corresponding author: Fabio S. Taccone, Department of Intensive Care, Hôpital Erasme, Université Libre de Bruxelles (ULB),
Route de Lennik 808, 1070 Brussels, Belgium. E‑mail: [email protected]
A B STRACT
Experimental and clinical observational studies have shown potential benefits of statin administration in the acute respiratory distress syndrome (ARDS) by modulating inflammation and preventing worsening respiratory function. More
recently, two randomized clinical trials failed to demonstrate an improved survival of ARDS patients treated with statins.
In the first study, conducted by the ARDS Network, 745 patients with sepsis‑associated ARDS were randomized within
48-hours of onset to receive either rosuvastatin or placebo. There was no significant difference between the rosuvastatin
and placebo groups for hospital mortality (primary outcome, 29% vs. 25%, P=0.21) or ventilator‑free days (15±11 vs.
15±11, respectively; P=0.96). In rosuvastatin‑treated patients, renal and hepatic failure free‑days were significantly lower
than in the placebo group, raising serious safety concerns. In the second study (HARP-2 trial), 540 patients with ARDS
were randomized within 48-hours of onset to receive either simvastatin (80 mg/day) or placebo. There was no significant
difference between the study groups for number of ventilator‑free days (primary outcome, 13±10 in the simvastatin vs.
12±10 in the placebo group, P=0.21) or 28-day mortality (22% vs. 27%, respectively; P=0.23). No significant difference
in serious adverse events was reported between groups. Herein, we discuss the main reasons for these negative findings
and consider where there could be a role for statins in ARDS patients.
(Cite this article as: Grimaldi D, Durand A, Gleeson J, Taccone FS. Failure of statins in ARDS: the quest for the Holy Grail
continues. Minerva Anestesiol 2016;82:1230-4)
Key words: Adult respiratory distress syndrome - Hydroxymethylglutaryl‑CoA reductase inhibitors - Patient outcome
assessment - Randomized controlled trial.
A
cute respiratory distress syndrome
(ARDS) is a major cause of morbidity
and mortality in the intensive care unit.1 ARDS
is characterized by protein‑rich pulmonary
edema due to alveolar‑capillary barrier injury
caused by overwhelming inflammation.2 In
particular, sepsis syndrome, either due to pneumonia or from an extra‑pulmonary source, is
the leading cause of ARDS over all other inflammatory conditions. Protective ventilatory
management, using low tidal volume and high
positive end‑expiratory pressure ventilation to
lower driving pressure, together with restrictive fluid management after hemodynamic sta-
1230
bilization, are the most effective therapeutic
interventions to reduce length of mechanical
ventilation and mortality in ARDS patients.3-5
To date, no targeted pharmacological therapy
has shown substantial beneficial effects in this
disease.
Given the pathophysiology of ARDS, the
modulation of the inflammatory response is
an attractive and logical strategy. In particular, statins, which inhibit 3-hydroxy-3-methylglutaryl coenzyme A (HMG‑CoA) reductase
and are commonly used to treat dyslipidemia,
have pleiotropic effects, including antioxidant,
anti‑inflammatory and endothelial‑protective
Minerva AnestesiologicaNovember 2016
STATINS AND ARDSGRIMALDI
actions.6 These properties led to the hypothesis that statins could improve lung function
and survival in ARDS patients. Several animal
studies showed that statins were able to prevent experimentally‑induced ARDS.7, 8 For example, in mice receiving an intra‑tracheal administration of endotoxin, pre‑treatment with
simvastatin decreased the severity of histological lung injury and alveolar‑capillary leakage.7
These results were later confirmed in a small
cohort of healthy volunteers, who received
oral simvastatin for 4 days before endotoxin
inhalation.9 Moreover, observational data
showed an association between previous statin
treatment and a better outcome in ARDS and
septic patients.10, 11 In a randomized phase‑II
trial, Craig et al. observed that treatment with
simvastatin (N.=30) decreased biomarkers of
inflammation measured in the broncho‑alveolar lavage fluid of patients with ARDS without
significant adverse events.12 Taken together,
these promising data prompted the design of
two large randomized multicenter studies (the
SAIL study from the ARDS Network using rosuvastatin and the HARP-2 study using simvastatin) aiming to demonstrate a clinical benefit of statin treatment in ARDS patients.13, 14
Discussion
Both trials failed to demonstrate a clinical
benefit of statin administration during ARDS
on mortality or ventilator free‑days; these negative results were in line with those from another recent study on simvastatin as an adjunctive
therapy for ventilator‑associated pneumonia
(VAP).15 A summary of the studies evaluating statin‑therapy in critically ill patients with
severe respiratory failure is shown in Table I.
Importantly, some issues deserve further discussion when interpreting these results.
The biological rationale for using statins in
ARDS is based on the presence of an acute
systemic inflammatory response, which is enhanced by an increased pulmonary production
of TNF-α and release of matrix metalloproteinases (MMPs), by the alteration of endothelium and leucocyte translocation.9 Indeed,
damaged pulmonary vascular endothelium and
Vol. 82 - No. 11
alveolar accumulation of neutrophils are histological hallmarks of ARDS.16 In pre‑clinical
studies where statins have been shown to be
protective, they were administered before the
inflammatory insult, which could have given
the drug a chance to “prepare” the endothelium. Furthermore, experimental lung‑injury
models, such as those using an endotoxin challenge, do not mimic the full effects of a real
infection neither reproduce the same inflammatory status.
The choice of a hydrophilic statin (e.g. rosuvastatin) by the SAIL investigators, claiming a
higher free drug‑fraction and less adverse effects, compromised the study as there is a lack
of significant pre‑clinical data on this molecule
in the ARDS setting. Also, it still remains unclear whether all statins have the same immune‑modulatory effects. Most pre‑clinical
trials working on the anti‑inflammatory effects
of these drugs in sepsis or in ARDS used lipophilic statins, such as atorvostatin or simvastatin.7, 9 Lipophilic molecules are known
to have greater tissue penetration; moreover
lipophilic statins demonstrate a stronger antibacterial effect.17 When the authors measured
rosuvastatin concentrations in a subgroup
of 404 patients, mean peak concentrations
reached 7.3 ng/mL, which were largely below target drug‑levels (10-70 ng/mL) able to
provide significant immune‑modulatory effects. These results could be explained by the
use of a moderate daily dose (20 mg), similar
to chronic therapy given in healthy subjects;
however this therapeutic strategy did not take
into account the altered drug pharmacokinetics
recognised in septic patients, which may lead
to inadequate concentrations in the early phase
of therapy if standard regimens are used.18 In
the HARP-2 study,14 the authors evaluated
simvastatin at a higher dose than in chronic
conditions, which had been previously shown
to provide adequate concentrations in critically‑ill patients.12 Furthermore, no significant
difference in ventilator‑free days and mortality was shown between groups, although drug
levels were not specifically assessed.
Neither trial considered the severity of
ARDS in the inclusion criteria and mean
Minerva Anestesiologica
1231
GRIMALDISTATINS AND ARDS
Table I.—Clinical studies on statins and ARDS.
Study
Targeted
pathology
Methods
Inclusion Criteria
SAIL 13
Prospective
Randomized
Double‑blinded
Placebo‑controlled
Multicenter
100% sepsis
Rosuvastatin
71% pneumonia (40 mg loading
PaO2/
dose, 20 mg
FiO2=170±70
daily)
on inclusion*
HARP-2 14
Prospective
Randomized
Double‑blinded
Placebo‑controlled
Multicenter
ARDS<48 hours
PaO2/FiO2<300
Sepsis (SIRS
criteria)
No treatment
with statin
in the last 48
hours
ARDS<48 hours
PaO2/FiO2<300
No treatment
with statin
in the last 2
weeks
Papazian et al.15
Prospective
Randomized
Double‑blinded
Placebo‑controlled
Multicenter
100%
pneumonia
PaO2/FiO2=189
[140-240] on
inclusion
Simvastatin
(60 mg daily)
HARP 12
Prospective
Randomized
Double‑blinded
Placebo‑controlled
First VAP
episode
No PaO2/FiO2
criteria
No treatment
with statin
since
intubation
ARDS
PaO2/FiO2<300
PaO2/
FiO2=170±55
on inclusion*
Simvastatin
(80 mg daily)
Primary
outcome
Intervention
42% sepsis
Simvastatin
58% pneumonia (80 mg daily)
PaO2/
FiO2=130±50
on inclusion*
Results
Hospital
mortality
N.=745
No effects on:
–– Mortality
–– MV‑free days
–– CRP levels
Less days free of
hepatic or renal
failure
MV‑free days N.=540
(28 days)
No effects on:
–– Mortality
–– MV‑free days
–– CRP levels
–– Non‑pulmonary
organ failure
28-day
N.=284
mortality
No effects on:
–– Mortality
–– MV‑free days
Physiological N.=60
and
–– Decreased
biological
pulmonary
outcomes
inflammation
–– Decrease CRP
–– No increase in
adverse events
MV: mechanical ventilation; CRP: C‑reactive protein; VAP: ventilator‑associated pneumonia.
*PaO2/FiO2 data were approximated by considering the two study groups (statins vs. placebo) altogether.
PaO2/FiO2 ratios vary considerably across
patients included in these studies. It would,
however, make sense that the benefit of
statins, if any, would be in patients with the
greatest inflammation e.g. those with the
most severe hypoxemia, as suggested by
a recent observational study.19 Moreover,
ARDS is not a well‑characterized disease
but is rather a syndrome caused by numerous conditions; the HARP studies enrolled
patients with ARDS whatever the cause,
and the SAIL study enrolled sepsis‑related
ARDS patients corresponding to a wide variety of conditions. It is reasonable to think
that a given pharmacological intervention
may be beneficial only in specific sub‑groups
of ARDS patients while being detrimental
in others, resulting in high statistical‑noise
in large trials. As an example, patients de-
1232
veloping ARDS secondary to a VAP should
be potentially excluded as in this particular condition, simvastatin has already been
shown to be ineffective.15 Another important limitation is the possibility that some
of the included patients were suffering from
acute cardiogenic pulmonary edema rather
than ARDS, as sometimes the diagnosis can
be clinically challenging. Although in both
studies the investigators cited elevated left
atrial pressures as an exclusion criterion, it
was not clearly indicated if or how this was
assessed. Thus, the inclusion of non‑ARDS
patients may have further diluted any effect
of statins in this setting.
Once again, disappointment arises from
large clinical studies despite promising results in animal data, illustrating the difficulty
to move from experimental studies to clinical
Minerva AnestesiologicaNovember 2016
STATINS AND ARDSGRIMALDI
outcomes.20 Overenthusiasm about suggestive
results from observational studies needs to be
tempered by the risk of overlooking unmeasured confounders. As an example, the trend
towards statin protection in a prospective observational study reported by the Irish Critical Care Trials group did not reach statistical
significance, even after multivariate analysis.12 Moreover, one potential confounder that
has not been measured in recent observational
statin/ARDS studies is the socio‑economic status, e.g. patients from lower socio‑economic
groups are less likely to receive appropriate
primary health care and are less likely to be
taking statins than others.21 Finally, the concomitant prescription of macrolide antibiotics in these patients, which can inhibit statin
metabolism and increase blood concentrations,
went also unaccounted in these observational
studies.
The last point to underline is the fact that
patients pre‑treated with statins were excluded
from the trials (Table I). This is in contrast to
the statin dosing approach taken in pre‑clinical
and clinical observational studies. In the SAIL
study, a pre‑specified small subgroup analysis
of patients who previously took statins, but
not during the 48 hours prior to enrolment,
showed that continuation of therapy was not
associated with a clinical benefit. However, as
conflicting results have been reported in septic patients,22 it seems reasonable to pursue a
statin treatment unless there is renal or hepatic
failure (exclusion criteria in all trials), concern
about rhabdomyolysis or the development of
another recognized side‑effect. When considering the potential advantage of continuing a
drug in critically ill patients it is also important
to consider the potential negative impact of
withdrawing the drug. Indeed, stopping statins
in at‑risk hospitalized patients increases their
risk of a cardiac event.23
The tolerance of statins in critically ill patients is uncertain, and few data exist. In the
HARP and HARP-2 studies, simvastatin therapy appears to be well tolerated; however, in the
SAIL study, rosuvastatin was associated with
an increase duration of renal and hepatic dysfunction. Clinical relevance of these adverse ef-
Vol. 82 - No. 11
fects needs to be clarified, but these safety concerns plead against the use of higher doses of
rosuvastatin. Furthermore, statins may worsen
sepsis‑induced immune dysfunction 24 and then
increase the risk of ICU‑acquired infection.
Conclusions
In two prospective randomized trials, statin
administration in the early inflammatory phase
of ARDS was not associated with improved
60-day mortality. Statins did not show any
benefit in ventilator‑free days or in reducing
inflammatory biological markers. To date,
statin treatment cannot be recommended in
ARDS patients.
Key messages
—— Early statin therapy in patients with
ARDS was not associated with any benefit
on mortality or ventilator‑free days.
—— Continuation of statin therapy in patients with early ARDS may be beneficial.
—— Statins use in critically ill patients
may induce renal and hepatic failure.
—— The use of statin in other cause of
sepsis remains to be studied by randomized
trials.
References
  1. Schell‑Chaple HM, Puntillo KA, Matthay MA, Liu KD;
National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome Network. Body temperature and
mortality in patients with acute respiratory distress syndrome. Am J Crit Care 2015;24:15-23.
 2. Fanelli V, Ranieri VM. Mechanisms and clinical consequences of acute lung injury. Ann Am Thorac Soc
2015;12(Suppl 1):S3-8.
 3.The Acute Respiratory Distress Syndrome Network.
Ventilation with lower tidal volumes as compared with
traditional tidal volumes for acute lung injury and the
acute respiratory distress syndrome. N Engl J Med
2000;342:1301-8.
  4.National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome (ARDS) Clinical Trials Network:
Wiedemann HP, Wheeler AP, Bernard GR, Thompson
BT, Hayden D, deBoisblanc B, et al. Comparison of two
fluid‑management strategies in acute lung injury. N Engl
J Med 2006;354:2564-75.
  5.Amato MB, Meade MO, Slutsky AS, Brochard L, Costa
EL, Schoenfeld DA, et al. Driving pressure and survival
in the acute respiratory distress syndrome. N Engl J Med
2015;372:747-55.
Minerva Anestesiologica
1233
GRIMALDISTATINS AND ARDS
  6. Ito MK, Talbert RL, Tsimikas S. Statin‑associated pleiotropy: possible beneficial effects beyond cholesterol reduction. Pharmacotherapy 2006;26:85S-97S.
  7. Jacobson JR, Barnard JW, Grigoryev DN, Ma SF, Tuder
RM, Garcia JG. Simvastatin attenuates vascular leak and
inflammation in murine inflammatory lung injury. Am J
Physiol Lung Cell Mol Physiol 2005;288:L1026-32.
  8. Yao HW, Mao LG, Zhu JP. Protective effects of pravastatin in murine lipopolysaccharide‑induced acute lung
injury. Clin Exp Pharmacol Physiol 2006;33:793-7.
  9. Shyamsundar M, McKeown ST, O’Kane CM, Craig
TR, Brown V, Thickett DR, et al. Simvastatin decreases lipopolysaccharide‑induced pulmonary inflammation in healthy volunteers. Am J Respir Crit Care Med
2009;179:1107-14.
10.Irish Critical Care Trials Group. Acute lung injury and
the acute respiratory distress syndrome in Ireland: a prospective audit of epidemiology and management. Crit
Care 2008;12:R30.
11. Mansur A, Steinau M, Popov AF, Ghadimi M, Beissbarth
T, Bauer M, et al. Impact of statin therapy on mortality in patients with sepsis‑associated acute respiratory
distress syndrome (ARDS) depends on ARDS severity:
a prospective observational cohort study. BMC Med
2015;13:128.
12.Craig TR, Duffy MJ, Shyamsundar M, McDowell C,
O’Kane CM, Elborn JS, et al. A randomized clinical trial
of hydroxymethylglutaryl- coenzyme a reductase inhibition for acute lung injury (The HARP Study). Am J
Respir Crit Care Med 2011;183:620-6.
13.National Heart, Lung, and Blood Institute ARDS Clinical Trials Network, Truwit JD, Bernard GR, Steingrub J,
Matthay MA, Liu KD, Albertson TE, et al. Rosuvastatin
for sepsis‑associated acute respiratory distress syndrome.
N Engl J Med 2014;370:2191-200.
14. McAuley DF, Laffey JG, O’Kane CM, Perkins GD, Mullan B, Trinder TJ, et al.; HARP-2 Investigators; Irish
Critical Care Trials Group. Simvastatin in the acute respiratory distress syndrome. N Engl J Med 2014;371:1695703.
15. Papazian L, Roch A, Charles PE, Penot‑Ragon C, Perrin
G, Roulier P, et al.; STATIN‑VAP Study Group. Effect of
statin therapy on mortality in patients with ventilator‑associated pneumonia: a randomized clinical trial. JAMA
2013;310:1692-700.
16. Preira P, Forel JM, Robert P, Nègre P, Biarnes‑Pelicot
M, Xeridat F, et al. The leukocyte‑stiffening property
of plasma in early acute respiratory distress syndrome
(ARDS) revealed by a microfluidic single‑cell study:
the role of cytokines and protection with antibodies. Crit
Care 2016;20:8.
17. Masadeh M, Mhaidat N, Alzoubi K, Al‑Azzam S, Alnasser Z. Antibacterial activity of statins: a comparative study
of atorvastatin, simvastatin, and rosuvastatin. Ann Clin
Microbiol Antimicrob 2012;11:13.
18. Kruger PS, Freir NM, Venkatesh B, Robertson TA, Roberts MS, Jones M. A preliminary study of atorvastatin
plasma concentrations in critically ill patients with sepsis.
Intensive Care Med 2009;35:717-21.
19. Mansur A, Steinau M, Popov AF, Ghadimi M, Beissbarth
T, Bauer M, Hinz J. Impact of statin therapy on mortality in patients with sepsis‑associated acute respiratory
distress syndrome (ARDS) depends on ARDS severity:
a prospective observational cohort study. BMC Med
2015;13:128.
20. Perrin S. Preclinical research: Make mouse studies work.
Nature 2014;507:423-5.
21. Wallach‑Kildemoes H, Andersen M, Diderichsen F,
Lange T. Adherence to preventive statin therapy according to socioeconomic position. Eur J Clin Pharmacol
2013;69:1553-63.
22. Thomas G, Hraiech S, Loundou A, Truwit J, Kruger P,
Mcauley DF, et al. Statin therapy in critically‑ill patients
with severe sepsis: a review and meta‑analysis of randomized clinical trials. Minerva Anestesiol 2015;81:92130.
23. Heeschen C, Hamm CW, Laufs U, Snapinn S, Böhm M,
White HD; Platelet Receptor Inhibition in Ischemic Syndrome Management (PRISM) Investigators. Withdrawal
of statins increases event rates in patients with acute coronary syndromes. Circulation 2002;105:1446-52.
24. Monneret G, Venet F. Statins and sepsis: do we really
need to further decrease monocyte HLA‑DR expression
to treat septic patients? Lancet Infect Dis 2007;7:697-9.
Conflicts of interest.—The authors certify that there is no conflict of interest with any financial organization regarding the material
discussed in the manuscript.
Article first published online: May 27, 2016. - Manuscript accepted: May 25, 2016. - Manuscript revised: May 3, 2016. - Manuscript
received: January 29, 2016.
1234
Minerva AnestesiologicaNovember 2016
LETTERS TO THE EDITOR
© 2016 EDIZIONI MINERVA MEDICA
Online version at http://www.minervamedica.it
Minerva Anestesiologica 2016 November;82(11):1235-6
High-sensitivity troponin and
extubation failure after successful
spontaneous breathing trial
Dear Editor,
Myocardial dysfunction is one of the causes of failure
to wean from mechanical ventilator support in intensive
care. A multitude of clinical, biological and echocardiographic risk factors are associated with weaning failure
linked to myocardial dysfunction.
However, weaning failure must be differentiated
from extubation failure, as the causes of failure are
different.1 The incidence of extubation failure varies
from 10-20% and is defined as the need for reintubation
within 24-72 hours after properly conducted weaning.1
This failure is associated with increased morbidity and
mortality. Numerous clinical risk factors are associated
with extubation failure.2 The value of considering troponin levels in this context is a question that has already
been asked.3
Weaning from mechanical ventilator support increases myocardial effort, respiratory effort, the myocardial oxygen demand and adrenergic stress.3 These
conditions contribute to an imbalance between myocardial oxygen supply and demand. So, silent myocardial
ischemia could occur during weaning.3 The dosage of
high-sensitivity troponin (hs-troponin assays) allows
hitherto undetectable troponin levels to be detected as
well as faster variations. Our hypothesis is that an increase in hs-troponin during weaning is a sign of subclinical myocardial damage and could be related to
early reintubation.
We conducted a single-center study approved by the
research ethics committee (Comité de Protection des
Personnes SUD-EST IV) between February 2014 and
September 2015. We prospectively included patients
who were on ventilation for more than 48 hours and
were extubated after successful respiratory weaning,
Spontaneous breathing trial lasts actually one hour. The
extubation was considered a failure if the patient was
reintubated within 72 hours. Hs-troponin was measured
within two hours before weaning (T1) and one hour after weaning (T2). The absolute value of this biomarker
as well as its evolution between T1 and T2 were recorded (progression = [T2–T1]/T1).
Seventy patients were included in the study. Twelve
patients (17%) were reintubated within 72 hours. Both
groups were not statistically different in terms of SAPS
II score at admission, age, sex, number of days on ventilation before extubation and cardiac disease (Table
I). In univariate analysis, troponin level at T1, at T2 or
its variation between T1 and T2 were not significantly
different between the group of patients with or without
extubation failure at 72 hours.
BNP was the only biomarker associated with a risk
of reintubation after properly conducted weaning.4 In
the context of myocardial ischemia, a significant increase in hs-troponin can be observed during the first
hour.5 Spontaneous breathing trial could be considered
as a cardio-pulmonary stress test. Despite a pathophysiological rationale for hs-troponin measure in the context of weaning, this study suggests that hs-troponin
may not predict extubation failure. These results must
Table I.—Clinical and biological characteristics.
Age (years)
SAPS II at admission
Ventilation (days)
Male sex
Anterior cardiac disease
Hs-Troponin (ng/L)
Before weaning trial
After weaning trial
Progression (%)
Entire cohort
(N.=70)
Success of weaning
(N.=58)
Failure of weaning
(N.=12)
68±13
63±17
11.5±8.0
64%
32 (46%)
68±13
61±16
10.5±7.0
62%
25 (43%)
69±13
72±17
15.5±9.5
75%
8 (67%)
0.69*
0.053**
0.08**
0.39
0.14
196±602
184±593
-2.4±17.2
214±658
201±647
-1.8±18.2
110±170
102±159
-5.4±11.6
0.46**
0.49**
0.48**
P value
Values are expressed as mean±SD or as N. (%).
Progression = (T2–T1)/T1
SAPS: Simplified Acute Physiologic Score.
*t-test;**Mann-Whitney test.
Vol. 82 - No. 11
Minerva Anestesiologica
1235
LETTERS TO THE EDITOR
be interpreted with caution owing to the limited number
of patients.
Nicolas MOTTARD 1 *, Pauline RENAUDIN 1,
Florent WALLET 1, Fabrice THIOLLIÈRE 1,
Julien BOHE 1, 2, Arnaud FRIGGERI 1
1Service
of Reanimation, Hospices Civils de
Lyon, Centre Hospitalier Lyon-Sud,
Pierre Bénite, France;
2Université Claude-Bernard, Lyon, France
*Corresponding author: Nicolas Mottard, Service de Réanimation, Centre Hospitalier Lyon-Sud, 165 chemin du grand Revoyet, 69495 Pierre Bénite, France.
E-mail: [email protected]
References
 1.Epstein SK. Decision to extubate. Intensive Care Med
2002;28:535-46.
 2.Sellarés J, Ferrer M, Torres A. Predictors of weaning after acute respiratory failure. Minerva Anestesiol
2012;78:1046-53.
  3.Teboul J-L. Weaning-induced cardiac dysfunction: where
are we today? Intensive Care Med 2014;40:1069-79.
 4.Chien J-Y, Lin M-S, Huang Y-CT, Chien Y-F, Yu C-J,
Yang P-C. Changes in B-type natriuretic peptide improve
weaning outcome predicted by spontaneous breathing
trial. Crit Care Med 2008;36:1421-6.
  5.Reichlin T, Irfan A, Twerenbold R, Reiter M, Hochholzer
W, Burkhalter H, et al. Utility of absolute and relative
changes in cardiac troponin concentrations in the early
diagnosis of acute myocardial infarction. Circulation
2011 Jul;124:136-45.
Conflicts of interest.—The authors certify that there is no conflict
of interest with any financial organization regarding the material
discussed in the manuscript.
Article first published online: May 27, 2016. - Manuscript accepted: May 25, 2016. - Manuscript revised: May 10, 2016. Manuscript received: January 5, 2016.
(Cite this article as: Mottard N, Renaudin P, Wallet F, Thiollière
F, Bohe J, Friggeri A. High-sensitivity troponin and extubation
failure after successful spontaneous breathing trial. Minerva Anestesiol 2016;82:1235-6)
© 2016 EDIZIONI MINERVA MEDICA
Online version at http://www.minervamedica.it
 ;(): ;():;():():
):
Minerva
Anestesiologica
2016 November;82(11):1236-7
Posterior reversible encephalopathy
syndrome in acute pancreatitis
Dear Editor,
We report a case of posterior reversible encephalopathy syndrome (PRES) during the hospitalization of
acute pancreatitis.
1236
PRES is clinical and radiologic syndrome characterized by a constellation of clinical signs and symptoms including headache, altered mental status, visual change, seizures and occasionally focal neurologic
signs. Neuroimaging is essential to the diagnosis:
typical findings are symmetrical white matter edema
in the posterior cerebral hemispheres, particularly
the parieto-occipital regions, most often visualized
on magnetic resonance imaging.1 Many conditions
could be associated with PRES development such as
hypertension, preeclampsia, renal disease, sepsis or
autoimmune diseases.
A 73-year-old man with a medical history of hypertension, chronic ischemic heart disease and diabetes
mellitus type 2 was first admitted with acute pancreatitis caused by gallstones. Endoscopic retrograde cholangiopancreatography with papillosphincterotomy were
performed providing him with an overall wellness. Few
days later, the patient showed a rapidly decreasing of
consciousness and generalized seizures. After resolution of the seizures, a cerebral computed tomography
(CT) scan was performed and the patient was admitted
to Intensive Care Unit (ICU).
Upon his arrival at the ICU, the patient was unconscious and able to localize painful stimuli without neurological deficits. He had uncontrolled hypertension,
with systolic blood pressure more than 200 mmHg. The
cerebral CT scan showed absence of ischemic or hemorrhagic lesions.
Routine laboratory and cerebrospinal fluid parameters were normal. Cerebral magnetic resonance imaging was finally performed, showing areas of altered
signal at the level of midbrain, cerebellum and posterior regions; no basilar artery occlusion was observed
(Figure 1).
With the achievement of blood pressure control, the
neurologic function improved in few days, then the patient was discharged from the ICU in absence of neurological deficits. Follow-up MR imaging was performed
two weeks later showing normal findings. The clinical
features and the MR imaging findings of the reported
case are strongly suggestive of PRES. Imaging is essential for the differential diagnosis of PRES, which
includes strokes, cerebral venous thrombosis and encephalitis.2 The pathophysiologic mechanism of PRES
is unknown, although it would result from a dysfunction
of the cerebrovascular auto-regulatory mechanism due
to hypertension. Severe acute pancreatitis is often associated with capillary leak syndrome and an increasing
in endothelial permeability. These changes in vascular
permeability could be the link between acute pancreatitis and PRES. Altered cerebral microcirculation is also
the main mechanism underneath sepsis-related brain
dysfunction which frequently shows clinical and radiological signs of PRES.³
The preferential involvement of the parietal and occipital lobes is thought to be related to the relatively
poor sympathetic innervation of the vertebrobasilar circulation. In patients with PRES, the myogenic component of autoregulatory mechanism is blunted by passive
Minerva Anestesiologica
November 2016
LETTERS TO THE EDITOR
Figure 1.—A T2-weighted MRI image of the patient shows bilateral, slightly asymmetric, multifocal T2 hyperintensities in
the posterior temporo-occipital subcortical white matter.
overdistension of the vessel due to elevations in blood
pressure. Thus, autoregulation is strictly dependent on
the neurogenic response: the more poorly innervated
areas in the posterior circulation are most vulnerable.
Prompt lowering of blood pressure, treatment of associated seizures and removal of causative agent are recommended in the management of PRES. The prognosis is
usually benign with complete reversal of clinical symptoms within several days, when adequate treatment is
immediately initiated.
In the literature only few cases of association between
acute pancreatitis and PRES with documentation of cerebral vasculopathy has been reported. In 2008 Shen et
al.4 described a case of PRES after acute pancreatitis
in a patient affected by acute intermittent porphyria.
Later, Yamada et al.5 reported a case of recurrent PRES
in a pediatric patient with pancreatitis subsequent to nephrotic syndrome. In all these cases posterior reversible
encephalopathy was completely recovered without any
neurological sequelae because of the early recognition
of the symptoms and timely treatment.
Christian COMPAGNONE,
Daniele BELLANTONIO *,
Federica PAVAN, Fernanda TAGLIAFERRI,
Maria BARBAGALLO, Guido FANELLI
Department of Anesthesia, Intensive
Care and Pain Therapy,
University Hospital of Parma, Parma, Italy
Vol. 82 - No. 11
*Corresponding author: Daniele Bellantonio, Department of
Anesthesia, Intensive Care and Pain Therapy, University Hospital of Parma, via Gramsci 14, Parma, Italy.
E-mail: [email protected].
References
  1. Fugate JE, Claassen DO, Cloft HJ, Kallmes DF, Kozak
OS, Rabinstein AA. Posterior reversible encephalopathy
syndrome: associated clinical and radiologic findings.
Mayo Clin Proc 2010;85:427-32.
  2. Bartynski WS. Posterior reversible encephalopathy syndrome, part 1: fundamental imaging and clinical features.
AJNR Am J Neuroradiol 2008;29:1036-42.
  3.Oddo M, Taccone FS. How to monitor the brain in septic
patients? Minerva Anestesiol 2015;81:776-88.
 4.Shen FC, Hsieh CH, Huang CR, Lui CC, Tai WC,
Chuang YC. Acute intermittent porphyria presenting as
acute pancreatitis and posterior reversible encephalopathy syndrome. Acta Neurol Taiwan 2008;17:177-83.
  5. Yamada A, Atsumi M, Tashiro A, Hiraiwa T, Ueda N. Recurrent posterior reversible encephalopathy syndrome in
nephrotic syndrome: case report and review of the literature. Clin Nephrology 2012;78:406-11.
Conflicts of interest.—The authors certify that there is no conflict
of interest with any financial organization regarding the material
discussed in the manuscript.
Article first published online: June 7, 2016. - Manuscript accepted: June 6, 2016. - Manuscript revised: May 31, 2016. Manuscript received: March 20, 2016.
(Cite this article as: Compagnone C, Bellantonio D, Pavan F,
Tagliaferri F, Barbagallo M, Fanelli G. Posterior reversible encephalopathy syndrome in acute pancreatitis. Minerva�������
��������������
Anestesiol 2016;82:1236-7)
Minerva Anestesiologica
1237
LETTERS TO THE EDITOR
© 2016 EDIZIONI MINERVA MEDICA
Online version at http://www.minervamedica.it
 ;(): ;():;():():
):
Minerva
Anestesiologica
2016 November;82(11):1238-9
Successful treatment of
life-threatening hemorrhaging
due to amniotic fluid embolism
Dear Editor,
The current report of a successful treatment of a case
of amniotic fluid embolism (AFE) is of importance to
the readership of this Journal because AFE is a rare but
catastrophic event with a high mortality for both the
mother and the fetus.1 Although the pathophysiology
remains uncertain, an anaphylactic-type reaction is the
most likely initiating factor. Unfortunately, AFE cannot
be prevented and the diagnosis is based on the patient’s
clinical presentation (and is essentially one of exclusion).
During the second stage of labor, a healthy
39-year-old woman developed acute severe hypotension (systolic blood pressure [SBP]<30 mmHg),
cyanosis (SpO2<70%), altered mental status (Glasgow Coma Scale [GCS] Score of 6 -opened her
eyes in response to painful stimuli, made no verbal
sounds, and displayed abnormal reflex responses to
painful stimuli) and severe fetal bradycardia (HR
of 40 bpm). A presumptive diagnosis of AFE was
made and the patient was immediately transferred
to the operating theatre for an emergency caesarean
delivery. Following the extraction of the placenta,
the patient developed uterine atony accompanied
by severe haemorrhage which failed to respond to
the infusion of uterotonic drugs (namely, oxytocin
20 and 10 mg IV bolus injections and sulprostone 8.3 ug/min IV) necessitating an emergency
hysterectomy. The patient remained hypotensive
with a SBP=70 mmHg, MAP=43 mmHg, DBP=30
mmHg despite aggressive fluid replacement (2 L of
warmed crystalloid), intraoperative blood salvage
reinfusion (washed red cells, 485 mL),2 packed red
blood cells (5 units) and of fresh frozen plasma (4
units). Vasoactive therapy was initiated with a norepinephrine infusion at 0.09 ug/kg/min. The patient
displayed signs of disseminated intravascular coagulation (DIC) and the routine clotting tests revealed
a severe coagulation disorder (Table I). The coagulopathy was immediately treated using fibrinogen 2
g IV and recombinant activated factor VII (rFVIIa)
90 ug/kg IV,3 resulting in a prompt reversal of the
life-���������������������������������������������
threatening bleeding. The
�����������������������
mother and the newborn recovered promptly and completely without
residual cardiac or neurological deficits.
This case illustrates the importance of the early
diagnosis of AFE followed by an aggressive treatment plan in order to achieve a good clinical outcome for both the mother and the child������������
. The availability of testing such as thromboelastometry and
thrombelastography as an adjunct to conventional
coagulation studies in the operating room may have
facilitated the more precise treatment of the coagulation disorder (with hemostatic therapeutic agents)
during post-partum hemorrhage.
The use of rFVIIa remains controversial. However, current treatment guidelines for AFE suggest
that rFVIIa only be used for hemostatic therapy in
cases of ongoing obstetrical bleeding. Some experts
have recommended that rFVIIa only be used after
massive coagulation factors replacement has proven to be insufficient. 4 A large retrospective series
presented by the Australian and New Zealand Haemostasis Registry suggested that earlier administration of rFVIIa might improve uterine preservation
rates in this population. Despite an increased risk
of thromboembolism in these patients, the complication has rarely been reported following rFVIIa
use. This
����������������������������������������������
is a unique report involving the successful treatment of a parturient with a suspected AFE
and life-threatening coagulopathy by early use of
fibrinogen followed by rFVIIa. This approach may
obviate the need for other more invasive and more
costly therapeutic interventions. Both the parturient and the neonate survived without any residual
morbidity.
Table I.—Temporal changes in the patient’s perioperative laboratory test results.
Hemoglobin (g/dL)
Hematocrit (%)
Platelets (n)
Prothrombotin activity ratio (%)
INR (%)
Activated partial thromboplastin time (sec)
Fibrinogen (mg/dL)
1238
Baseline
Before
procoagulation
therapy
After
procoagulation
therapy
At 12 hours
At 36 hours
11.6
35.0
359,000
1.1
1.0
1.0
469
04.5
13.1
247,000
4.00
3.84
Not coagulated
Undetectable
08.7
25.5
138,000
0.98
1.00
1.61
101
07.8
23.9
114,000
0.98
1.00
1.13
140
08.4
24.7
133,000
Minerva Anestesiologica
1.01
0.96
168
November 2016
LETTERS TO THE EDITOR
Lucia AURINI 1*, Maria Pia RAINALDI 2,
Paul F. WHITE 3, Battista BORGHI 4
1Unit of Anesthesia, Department of Medical and
Surgical Sciences, Sant’Orsola Hospital, University of
Bologna, Bologna, Italy; 2Department of Maternal and
Child Health, Unit of Anesthesia, Sant’Orsola Hospital,
Bologna, Italy; 3Department of Anesthesiology at
Cedars Sinai Medical Center, Los Angeles, CA, USA;
White Mountain Institute, Tallgrass, The Sea Ranch,
CA, USA; 4Department of Biomedical and Neuromotor
Sciences, Unit of Anesthesia, Rizzoli Orthopedic
Institute, University of Bologna, Bologna, Italy
*Corresponding
author: Lucia
�������������������������������������
Aurini, Unit
�����������������������
of Anesthesia, Department of Medical and Surgical Sciences, �����������������
University of Bologna, Sant’Orsola Hospital, via Massarenti 9, 40138 Bologna,
Italy. E-mail: [email protected]
References
 1.Gist RS, Stafford IP, Leibowitz AB, Beilin Y. Amniotic
fluid embolism. Anesth Analg 2009;108:1599-602.
Vol. 82 - No. 11
  2.Rainaldi MP, Tazzari PL, Scagliarini G, Borghi B, Conte
R. Blood salvage during caesarean section. Br J Anaesth
1998;80:195-8.
  3.Abdul-Kadir R, McLintock C, Ducloy AS, El-Refaey H,
England A, Federici AB, et al. Evaluation and management of postpartum hemorrhage: consensus from an international expert panel. Transfusion 2014;54:1756-68.
 4.Leighton BL, Wall MH, Lockhart EM, Phillips LE, Zatta
AJ. Use of recombinant factor VIIa in patients with amniotic fluid embolism: a systematic review of case reports.
Anesthesiology. 2011 115:1201-8.
Authors’ contribution.—Lucia Aurini wrote the main part of the
text (with the assistance of Dr. PF White); Maria Pia Rainaldi
treated the patient, collected the data and revised the manuscript;
Battista Borghi assisted with the revision of the manuscript
Conflicts of interest.—The authors certify that there is no conflict
of interest with any financial organization regarding the material
discussed in the manuscript.
Article first published online: July 22, 2016. - Manuscript accepted: July 6, 2016. - Manuscript revised: June 21, 2016. Manuscript received: May 3, 2016.
(Cite this article as: Aurini L, Rainaldi MP, White PF, Borghi
B. Successful treatment of life-threatening hemorrhaging due to
amniotic fluid embolism. Minerva Anestesiol 2016;82:1238-9)
Minerva Anestesiologica
1239

T O P 5 0 M I N E RVA A N E S T E S I O L O G I C A R E V I E W E R S
Top 50 reviewers April 2016-September 2016
1240
Surname
Name
Savoia
Bertini
Agrò
Ezzeldin
Camporesi
Biasucci
Faraoni
El Tahan
Brazzi
Aceto
Benes
Cattano
Alston
Brogly
Ting
Abad Gurumeta
Caruselli
Kinoshita
Gottumukkala
Montini
Gómez-Ríos
Vaida
Deflandre
Arnal
Donadello
Onutu
Cavallone
Riem
Pabelick
Donatelli
Kayhan
Borghi
Turnbull
De Blasi
Tritapepe
Gentili
Max
Dalfino
Kim
Bruder
De Pascale
Clendenen
Lautrette
Kakodkar
Eden
Beilin
Sbaraglia
Cinnella
Lirk
Guarracino
Gennaro
Pietro
Eugenio Felice
Ibrahim Saleh
Enrico Mario
Daniele Guerino
David
Mohamed
Mohamed
Paola
Jan
Davide
Theodore
Nicolas
Chien-Kun
Alfredo
Marco
Hiroyuki
Vijaya Narasimha Raju
Luca
Manuel Ángel
Sonia
Eric P.
Jean-Michel
Katia
Adela Hilda
Laura Francesca
Nicole
Christina Maria
Francesco
Zeynep
Battista
David
Roberto Alberto
Luigi
Andrea
Martin
Lidia
Tae W.
Nicolas
Gennaro
Steven R
Alexandre
Prashant Shivaji
Arieh
Yaakov
Fabio
Gilda
Philipp
Fabio
MINERVA ANESTESIOLOGICA
Number
of revisions
10
9
9
9
8
8
8
7
7
7
7
7
6
6
6
6
6
5
5
5
5
5
5
5
5
4
4
4
4
4
4
4
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
November 2016