Download Incidental Extra-cardiac Findings on Clinical CMR

Incidental Extra-cardiac Findings on Clinical CMR; A Comparison of 3 HASTE
Techniques.
Bruce Irwin1*, Tom Newton2* Charles Peebles3, Alexander Borg1, David Clark4, Chris
Miller1, Nick Abidin1, Melanie Greaves2, Matthias Schmitt1,5
1
Department of Cardiology, University Hospital of South Manchester (UHSM)
Department of Radiology, UHSM
3
Department of Radiology, Southampton General Hospital
4
Alliance Medical, Wythenshawe CMR unit
5
Biomedical Imaging Institute, The University of Manchester
2
BI and TN contributed equally to this work
Word count;
Abstract: 393
Text (including Tables/Figures): 4163
References: 416
Address for Correspondence
Dr Matthias Schmitt
Department of Cardiology
Office 4, North West Heart Centre
University Hospital of South Manchester
Southmoor Road
Manchester M23 9LT
Tel./Fax 01612914940
e-mail; [email protected]
ABSTRACT
Objectives: First, we sought to assess the frequency of incidental extra-cardiac
findings (IEF) found in a consecutive series of clinical Cardiac MR scans. Second, we
compared the 3 clinically used HASTE acquisition protocols in this context. Third, we
determined the impact of the three different HASTE protocols on acquisition time and
image quality.
Methods: Three consecutive groups of 238 patients (total 714), referred for clinically
indicated CMR, were scanned with either breath-hold HASTE (BH, Group 1), free
breathing HASTE (FB, Group 2) or diaphragmatic navigated HASTE (NAV, Group
3) in addition to multi-slice, single shot steady state sequences in 3 orthogonal planes.
All 714 clinical reports were reviewed regarding the presence of incidental extracardiac findings and the recommendations on the need for further investigation,
follow up, and/or clinical correlation. Finally, to determine the impact of each
HASTE protocol on acquisition time and image quality an additional cohort of 15
patients underwent all 3 protocols back to back in a random fashion. The length of
each acquisition was timed and image quality was reviewed and scored externally.
Results: A total of 180 IEF were found in 162 (22.7%) out of 714 patients. There was
no significant difference in frequency of IEF between the 3 HASTE groups. Out of
180 IEF 88 were considered minor and 92 major findings. Of the latter, 8 (1.1%) were
considered highly significant including one bronchoalveolar carcinoma stage 1B
requiring lobectomy, 2 cases of florid sarcoidosis in patients presenting with VT and
“structurally normal hearts” on Echo, one case of pulmonary aspergillosis, 2 cases of
advanced pulmonary fibrosis, one ascending thoracic aortic aneurysm and a case of
iatrogenic liver haemorrhage following placement of a pericardial drain. FB HASTE
acquisition (69±2.5s) was significantly faster than BH (105±3.8s) and NAV
(121±2.7s), p<0.001 but also produced the lowest image quality on a 5 point scale; 3.5
(FB) versus 3.9 (BH) versus 3.8 (NAV), p=0.08.
Conclusion: Overall, IEF are common and lead to follow on investigations in a
substantial minority of cases. However, the overall incidence of highly significant
findings in the current study was low (~1%). There was no difference in the frequency
of incidental extra-cardiac findings between the 3 HASTE protocols. Whilst the free
breathing HASTE technique is statistically significantly faster than breath hold and
navigated HASTE the absolute time saving is small and probably out-weight by the
resulting lesser image quality.
INTRODUCTION:
Cardiac magnetic resonance (CMR) is now an established imaging modality with well
described clinical indications1 and appropriateness criteria2.
In addition to the heart, a typical CMR examination will also image adjacent thoracic
and abdominal structures. Indeed most, but not all, centres initially perform gradient
echo “scout”-imaging of several slices in coronal, saggital and axial views followed
by axial imaging of the entire chest, conventionally using half fourier turbo spin echo
imaging (e.g. HASTE). Consequently, findings incidental to the cardiac examination
may be encountered some of which may be clinically relevant.
Although well described in the computed tomography (CT) literature3-7 there is
paucity of data regarding the frequency and clinical impact of incidental extra-cardiac
findings (IEF) discovered during routine CMR.
The few studies5, 8, 9 and preliminary reports10, 11 available to date demonstrate
considerable variability, with IEF rates ranging from 7.6% to 81%, and indicate
unsurprisingly, that the prevalence of IEF is dependent on multiple factors, not least
the population studied, the image sequences applied and the definitions utilised to
classify and categorise IEF. Furthermore, it is unclear if the various methods of
performing axial HASTE sequences (i.e breath-hold (BH), free breathing (FB), or
diaphragmatic navigated (NAV)) impact on the frequency of detecting IEF. In the
current study we sought to assess the frequency of IEF in consecutive patients
referred for clinically indicated CMR, in an NHS tertiary care setting. We also aimed
to determine the impact of the 3 most commonly applied HASTE acquisition
protocols on the frequency of IEF, as well as on image acquisition time and image
quality.
METHODS:
The study was planned in May 2008 and carried out from 12/08/2008 to 14/09/2009
in a dedicated CMR unit run collaboratively as a Managed Care Service by Alliance
Medical at the University Hospital of South Manchester (UHSM), in Wythenshawe.
Patients: Prospectively, three subsequent groups of 238 adult patients each (i.e 714
patients in total), all referred for clinically indicated CMR, were scanned with either
BH HASTE (Group 1), FB HASTE (Group 2), or NAV HASTE (Group 3) sequences.
All 714 clinical studies were acquired by 4 CMR trained radiographers, employed by
Alliance Medical, and reported by one of 3 Physicians (2 Cardiologists (MS, NA) + 1
Radiologist (MG)) employed by UHSM Foundation Trust. All reports were
subsequently reviewed and audited by either a CMR-Level II trained Cardiology (BI)
or CMR-Level II trained Radiology Fellow (TN) regarding the presence of IEF and
the recommendations on the need for further investigation, follow up and/or clinical
correlation. Both Fellows closely collaborated and assured that they applied identical
criteria for classification/categorisation of IEF in line with the following definition.
Definition: We prospectively decided to classify IEF in minor and major findings
whereby minor findings were considered benign or of no clinical importance not
requiring clinical correlation or follow up. Major findings were those IEF potentially
or definitely considered to be of clinical significance and/or requiring clinical
correlation or further follow and/or work-up.
Whilst there was variability in case load amongst the reporting physicians, their
relative reporting contributions did not change between the HASTE groups.
Furthermore, pleural effusions and/or ascites co-existing with significant ventricular
impairment were in contrast to some previous studies not classified as IEF.
CMR: All clinical CMR studies were vetted and coded by a single Consultant (MS)
and for auditing purpose grouped in to one of 8 clinical indication groups (UK
national MRI codes MCORV = Cardiac Ventriculogram and MCVVS = LV volume
study were grouped together). Also, as we did not perform coronary artery imaging in
isolation, we decided for audit purposes to classify MACOA (= MRA coronary
arteries) as either a ventricular function study if no contrast was administered, or as a
MCVIA (viability study) if prior contrast administration took place (in which case
“late enhancement” imaging was always performed additionally).
All scans were performed according to locally agreed scanning protocols, tailored
where necessary to the individual patient. These were largely based on standardised
protocols recommended by the Society of Cardiac Magnetic Resonance (SCMR). As
a minimum all studies included 3 localising single shot steady state sequences in 3
orthogonal planes followed by one of three HASTE protocols (see below), assessment
of left and right ventricular function by means of breath hold cine steady state free
precession (SSFP) sequences, predominantly retrospectively gated but triggered in the
presence off significant arrhythmia. Additional sequences, were performed as
clinically indicated. Eighty nine percent of patients received 0.1 to 0.2mmol/kg of
intravenous gadolinium either as dimeglumine gadopenetetate (Magnevist®, Bayer
Schering Pharma), Gadobutrol (Gadovist®, Bayer Schering Pharma), or
Gadobenate® dimeglumine (MultiHance, Bracco).
Group 1; Axial breath hold HASTE (BH)
Axial multi-slice HASTE sequence was acquired from the top of the aortic arch to the
diaphragm. The field of view (FOV) chosen was dependant on both patient size and
shape and typically ranged from 340mm x 233mm to 390mm x 344mm. Base
resolution and phase resolution were 256 and 59% respectively. Slice thickness and
slice gap were 8mm and 2mm respectively.
This resulted in a spatial resolution typically ranging from 2.3mm x 1.3mm x 8mm to
2.5mm x 1.5mm x 8mm.
Repetition time (TR) varied with patient heart rate. Interleaved slices were therefore
acquired in 3, 4 or 5 concatenations (and therefore 3, 4 or 5 breath holds) in order to
keep scan time to approximately 10 to 15 seconds per breath hold acquisition.
Group 2; Axial free breathing HASTE (FB)
Axial slices were acquired in a single concatenation and with the patient breathing
shallowly.The total acquisition time was heart rate dependent and typically varied
between 60-80 seconds.
Group 3; Axial diaphragmatic navigated HASTE (NAV)
As with the breath hold HASTE sequence, concatenations were set to maintain a
breath hold time of 10 to 15 seconds.
A single navigator echo was placed over the right hemi-diaphragm in order to monitor
its position at each breath hold. If the position of the diaphragm varied then the slice
position for each concatenation was altered in order to maintain contiguous slices.
In the current study therefore, spatial resolution, was predominantly affected by the
patientssize and shape rather than a result of the HASTE modality chosen.
Acquisition time (AT) and Image quality (IQ); To determine the impact of each
HASTE protocol on acquisition time and image quality an additional cohort of 15
patients underwent all 3 protocols back to back in a random fashion. The length of
each acquisition was timed with a stopwatch and IQ was reviewed and scored
separately in a blinded fashion by an external investigator (CP) using a 5 point scoring
system whereby 1 marked unsatisfactory IQ, 2 = satisfactory IQ, 3=good IQ, 4= very
good IQ, 5=excellent IQ.
Statistics: The prevalence of IEF is presented in absolute and per-terms. Data with
respect to image quality and acquisition speed are presented as mean ± SEM.
Comparisons for image acquisition speed and image quality were performed by
repeated measures ANOVA with post-hoc Bonferoni for pair-wise testing where
appropriate. Proportionate comparison between HASTE protocols were analysed by
Pearson Chi Square tests. A two-sided p value of <0.05 was considered significant.
RESULTS
A)
A total of 714 studies were included in the analysis, split evenly between each cohort.
Demographic characteristics are shown in table 1. Group 3 (NAV) had a somewhat
greater proportion of male subjects, although age distribution and BSA remained
consistent.
Table 1. Demographic data
Mean Age
Group 1
Group 2
Group 3
p
53.9 (17-85)
54.8 (16-85)
54.8 (15-85)
ns
1.93±0.016
1.90±0.014
1.92±0.015
ns
1.42
1.69
2.2
0.07
(Range)
Body Surface
Area (BSA)
Male: Female
ratio
An analysis of the clinical indications for the study cohort is displayed in figure 1 and
tables 2. This represents the typical workload of our centre, with little variation seen
in the 2 most prevalent clinical indications, i.e. stress perfusion and viability imaging,
over the time period of the study. However, overall there was a statistical difference
between clinical indications between the 3 groups (p=0.009) which was
predominantly driven by the proportionally larger number of Stress function (i.e.
Dobutamine stress) and “pure” LV volume studies (i.e non-contrast studies) in Group
1. There was no correlation of IEF with the CMR indication.
Figure 1. Indications for CMR scanning across all groups
Because of the small number of first pass rest perfusion this indication was
pooled with stress perfusion for the purpose of statistical analysis.
Table 2. Indication for CMR study by group
Indication
Group 1
Group 2
Group 3
No. Patients (%)
No. Patients (%)
No. Patients (%)
Myocardial stress perfusion
92 (38.7)
89 (37.4)
105 (44.1)
Stress function study*
10 (4.2)
3 (1.3)
3 (1.3)
Myocardial viability study†
74 (31.1)
92 (38.6)
88 (37.0)
Congenital anomaly study
20 (8.4)
20 (8.4)
11 (4.6)
MRA thoracic aorta
15 (6.3)
15 (6.3)
16 (6.7)
LV volume study ‡
23 (9.7)
10 (4.2)
11 (4.6)
Valvular function study
1 (0.4)
8 (3.4)
3 (1.3)
Rest perfusion study
3 (1.3)
1 (0.4)
1 (0.4)
*
Dobutamine stress with wall motion analysis
Includes investigation of arrhythmogenic right ventricular cardiomyopathy
‡ No gadolinium administered.
†
From the 714 patients scanned, a total of 180 IEFs were discovered in 162 patients.
This gives an overall prevalence of IEFs in the study of 25.2%, affecting 22.7% of
patients scanned.
As outlined above, these findings were further characterised into those perceived as
clinically significant or ‘major’, and those considered insignificant or ‘minor’. Eightyeight extra-cardiac findings of minor and ninety-two of major significance were
reported (see tables 3 + 4). Seven patients (1.0%) were diagnosed with more than one
IEF.
Highly significant Findings:
Amongst the major IEF 8 (1.1% ) findings were considered highly significant
including one bronchio-alveolar carcinoma stage 1B, 2 cases of florid pulmonary
sarcoidosis in patients presenting with VT and “structurally normal hearts” on Echo,
one case of pulmonary aspergillosis, 2 cases of advanced pulmonary fibrosis, one
ascending thoracic aortic aneurysm (6.6cm) requiring surgery, in a patients with
atypical chest pains, and a case of iatrogenic liver haemorrhage following placement
of a pericardial drain.
No significant difference was found between each of the groups when analysed for
the total extra-cardiac findings identified, both minor and major.
Table 3. Findings of minor significance
Abnormality
Group 1
Group 2
Group 3
Total all groups
Atelectasis
3
0
3
6
Large Axillary lymph node
0
1
0
1
Azygos lobe
2
0
0
2
Bronchogenic cyst
1
0
0
1
Diaphragmatic Lipoma
0
1
0
1
Hiatus Hernia
3
5
2
10
High take-off of Bronchus
1
0
0
1
Hepatic cyst (solitary)
8
4
3
15
Hepatic haemangioma
0
1
0
1
Kyphoscoliosis
0
0
1
1
Lipomatosis (epicardial/
2
3
1
6
5
1
10
16
1
0
0
1
Small pleural effusions
1
0
2
3
Simple Renal cysts ≤Bosniak II
0
2
0
2
Thyroid goitre
0
3
2
5
Vascular anomaly
3
2
7
12
Vertebral body haemangioma
0
4
0
4
Total minor abnormalities
30
27
31
88
(solitary)
mediastinal)
Mediastinal lymph nodes (<1cm in
short axis)
Pleural thickening (postpleurodesis)
P value for inter-group difference
Table 4. Findings of major clinical significance
Abnormality
Group 1
Group 2
Group 3
Total all groups
Dilated aorta*
2
3
2
7
Dilated bile ducts
0
1
0
1
Dilated oesophagus (achalasia)
1
1
0
2
Dilated main pulmonary artery†
7
5
4
16
Scarring of/around defect breast
1
0
0
1
Empyema
0
1
0
1
Hepatic cysts (multiple)
6
3
4
13
Hepatic mass awaiting diagnosis
0
0
1
1
(Iatrogenic) liver haemorrhage
1
0
0
1
Lung nodule
0
0
1
1
Lung mass
3
2
3
8
Thyroid mass
0
0
1
1
Parenchymal lung abnormality
6
11
6
23
Pleural effusion, cause unknown
0
0
2
2
Pleural thickening
2
6
1
9
Renal cysts (multiple/septated)
1
2
1
4
0
1
0
1
30
36
26
92
implant
≥Bosniak IIF
Vertebral body lesion, aetiology
unknown
Total major abnormalities
P value for inter-group difference
* Defined as > 4.0cm at pulmonary artery bifurcation level
†Defined as > 3.0cm or larger then ascending aorta
Lung nodule defined as <or equal 3cm, Lung mass >3cm
B) Image Acquisition Time and Quality
Data for the 15 patients who underwent all three methods of HASTE acquisition is
summarised in table 5. FB HASTE took the shortest time, with average scan duration
of 69 seconds. BH and NAV methods took significantly longer, on average 105 and
121 seconds, respectively. The individual image quality scores were averaged for
each of the three methods. FB resulted in the lowest image quality, with an average
quality score of 3.5, where 5 is the theoretical maximum. BH and NAV methods
displayed superior quality, with average scores of 3.9 and 3.8 respectively. The
overall difference in image quality between the acquisitions showed a trend towards
significance.
Table 5. Acquisition Time (in seconds) and Image Quality (5 point scoring system)
Patient
No.
Group
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Average
SD
SEM
Total sequence time
(s)
BH
FB
NAV
G1
G2
G3
107
81
130
112
79
135
134
60
134
123
70
115
95
85
111
102
69
125
98
62
103
105
80
128
112
61
131
75
48
102
89
65
127
92
67
116
103
62
116
103
74
121
123
68
115
105
69
121
14.8
9.8
10.5
3.8
2.5
2.7
p<0.001
Image quality score
(1-5, 5 = best)
BH
FB
NAV
G1
G2
G3
5
4
4
4
4
4
3
3
2
4
3
4
4
4
5
5
5
5
4
4
4
4
3
4
4
3
3
3
3
3
4
3
4
4
3
4
4
4
4
3
4
4
4
3
3
3.9
3.5
3.8
0.59
0.64
0.78
0.15
0.17
0.2
BH= Breath hold; FB= Free breathing; NAV= diaphragmatic navigation
Image
Comparison
(best 1st)
BH>NAV>FB
BH>NAV>FB
FB>BH>NAV
BH>NAV>FB
NAV>BH>FB
FB>BH>NAV
BH>NAV>FB
NAV>BH>FB
BH>NAV>FB
FB>BH>NAV
BH>NAV>FB
NAV>BH>FB
BH>NAV>FB
FB>NAV>BH
BH>NAV>FB
p=0.08
DISCUSSION:
The literature describing the prevalence of IEF during cross-sectional cardiac imaging
is predominated by publications based on cardiac CT. In contrast there is a paucity of
data characterising the prevalence and clinical significance of IEF on CMR, especially
in a pure clinical setting.
What the current study adds in the context of previous publications on IEF:
The present study is to our knowledge the second largest consecutive series reporting
the prevalence of (previously unknown) IEF in clinically indicated CMR studies. In
keeping with the cardiac CT literature (~21% prevalence) our study found that the
overall detection rate of IEF on clinically indicated CMR is relatively high (22.7%).
Interestingly, our pick up rate was almost identical to a preliminary report of identical
size also performed within the UK-NHS setting 11. The group from the London Chest
Hospital found, in a slightly older population (61.3 ± 15years, range 16-86 years, 67%
male), IEF in 218 (24%) out of 714 patients. These pathologies included renal cysts
(9%), aortic pathology (3%), lung collapse/consolidation (2%), liver cysts (2%), lung
mass (1%). They also found 2 cases of mediastinal masses, 1 mediastinal
lymphadenopathy, 6 increased signal from the hila, a single renal mass, and one large
hiatus hernia. In contrast with these 2 UK based CMR studies, both reporting
comparable detection rates to that of cardiac CT literature, the so far largest published
US based study by Chan et al 8 concluded that non-cardiac pathology is uncommonly
reported. In their cohort (1534 patients, 62%male, age 50±15years) 116 studies
(7.6%) had at least one non-cardiac finding of which 48 (3.1%) reports were deemed
to demonstrate major and 70 (4.6%) minor findings.
Major findings were; lymphadenopathy (n=22 / 40% of major findings), Lung
abnormalities (encompassing nodules, masses and infiltrates; n=19 / 35%),
Mediastinal masses (n=6 / 11%), Breast lesion (n=4 / 7%), Ascites (n=3 / 5%), soft
tissue mass (n=1 / 2%). Minor findings were: Pleural effusion (n=30 /40%), benign
liver cyst (n=15 / 20%), renal cysts (n=14 / 19%) ), Hiatus Hernia (n=7 / 9%),
diaphragmatic abnormality (n=2 / 3%), Splenic abnormality (n=2 / 3%), paraspinal
lipoma (n=2 /3%), anomalous vasculature non-cardiac (n=2 / 3%).
Importantly only 8 findings in 6 reports (0.4%) of 1534 reports were ultimately
deemed to be new and clinically significant. Of note, not unexpected, the age of those
with clinically significant non-cardiac pathology was greater (54±16).
Why these (apparent) discrepancies ?
The few studies5, 8, 9 and preliminary reports10 available demonstrate considerable
variability in IEF diagnosis with rates ranging from 7.6% (0.4% for major findings) to
81% (17%) and indicate that, not surprisingly, the prevalence of IEF is dependent on
multiple factors including the cohort studied, the “clinical setting”, the sequences
applied, and the reading session format as well as, possibly most importantly the
definitions utilised to categorise a IEF. With respect to minor IEF it is possibly not
unreasonable to assume, especially given that most (if not all) of the published
literature is based on retrospective review of reports rather than images, that the bulk
of variation in the literature reflects different attitudes and weighting of the most
common abnormalities (such as simple renal and hepatic cysts and vascular
abnormalities) by the reporting clinicians. This assumption is possibly supported by
the significantly smaller variation in “major” IEF in the same cohorts. What
constitutes a “major” IEF, is unfortunately not standardised and confused in the
literature with definitions “ranging from one that requires follow-up or clinical
correlation to one that needs immediate evaluation or treatment.” 12
With respect to referral bias it is apparent that the population in the study by Chan
clearly differs, not only in age but equally important in scan indication, significantly
from that in both UK based studies. Whilst the assessment of IHD (i.e. MCORP and
MCVIA) was the predominant clinical indication in both UK studies the 2 largest
referral cohorts in the US based study were assessment of ventricular function (34%)
and interestingly assessment of pulmonary vein anatomy (25%). The difference, in
particular with respect to the latter indication may be more representative of local
practise variation and the academic nature of the US institution rather than being
representative of differing technology application between healthcare systems.
However, whilst partly coming to different conclusions with respect to the frequency
of IEF, the overriding commonality between the current study and both above detailed
studies, and in fact most of the published literature, is that highly significant findings
(i.e. a finding that in line with the 2010 Expert Consensus requires immediate
evaluation or treatment) occur with a prevalence varying between 0.4%-2%.
Nevertheless, even taken together this literature remains insufficient to answer the
important question if the incidental finding of an extra-cardiac pathology during
cross-sectional cardiac imaging has a significant impact on patient health and long
term outcome. Equally, little is known with respect to the cost-effectiveness and
potential increased anxiety of follow on investigations triggered by IEF on crosssectional cardiac imaging. On the other hand in the increasingly litigious healthcare
environment of western societies it appears unwise not to have policies in place that
report and deal with potentially significant IEF.
What is the impact of the sequences used ?
A recent meta-analysis investigating the prevalence of incidental findings on magnetic
resonance imaging (MRI) of the brain found unsurprisingly that sequences applied
and sequence resolution were important determinant of the frequency with which IEF
where identified.13 Also Chan et al pointed out that: “a better understanding of how
various sequences impact on the ability to detect non-cardiac pathology would help
the development of guidelines in CMR training”.8 The current study is the first to
compare three different clinically utilised HASTE protocols with respect to the
frequency of IEF, image quality and acquisition time and found no significant
differences with respect to the IEF detection rate. Whilst the FB HASTE (Group 2)
was acquired significantly faster than the two other acquisition protocols the overall
time saving (of 36 or 51 seconds respectively) was small (on the background of an
average study length of 43 minutes) and occurred against a significant trade off in
image quality.
Should one alter protocols during scanning to address IEF ?
This important question off course only arises if the IEF is actually identified during
the acquisition of the study which will be affected by the local set-up with respect to
physician supervised or unsupervised study acquisition. However, even if detected
during the acquisition one must consider to what degree additional sequences
performed to evaluate the IEF further would impact on the workflow within a unit.
Equally important, one must not ignore the possibility that the deviation (in particular
if lengthy) from the original protocol may detrimentally effect the likelihood of
answering the original clinical question posed. In this light whilst minor changes to
cover the IEF with further slices/sequences may be appropriate more significant
changes to the protocol may not be desirable as dedicated specialised sequences, and
or contrast and expertise may be required, and /or alternative imaging modality may
be more appropriate and / or cost effective.14
The true significance of IEF?
Beyond the medico-legal aspects, the true clinical relevance probably more likely
relates to the fact that a number of IEF may actually account for patient symptoms
and thereby at least partly negate the originally entertained differential diagnosis.
Common examples in support of this are lung pathology such as atelectasis,
Emphysema, pleural effusions, pleural and interstitial lung diseases causing shortness
of breath. Equally important IEF are not infrequently a clue to either a symptom
relevant co-morbidity or again intimately related to the patients presenting problems.
Examples in the current study are the patients presenting with VT in the setting of
pulmonary sarcoidosis. Furthermore, IEF can provide indirect explanation for cardiac
pathology such as abnormal pulmonary venous drainage contributing to SOB or coarctation, causing increased afterload and driving hypertension, leading to LVH,
diastolic dysfunction and breathlessness.
Also, a range of IEF, whilst unrelated to a subjects symptoms and at the time of
diagnosis of no immediate clinical significance may become highly relevant during
the patients future diagnostic workup and potential subsequent treatment and therefore
should be documented and communicated in a way that the benefit of this information
is not subsequently lost. Examples of such cases are the finding of vascular
abnormalities such as interrupted IVC or retro-oesophageal course of a right
subclavian artery, both highly relevant for potential future cardiac catheterisation.
Finally, it is intuitive to assume that in individual circumstances action taken based
on the detection of an IEF, e.g. such as the lung malignancies found in the current
study, will improve outcome.
Study limitations:
Overall study and subgroup size was only moderate. However, this work is still the
second largest published series and even our sub-groups are larger than most currently
published work in this field. Despite prospective study design, report analysis for IEF
reports (not images!) and review of follow on investigations and therapy was
retrospective and limited to medical and electronic records of in-house patients.
Furthermore, patients were not randomised to one of the HASTE groups but for
organisational reasons scanned in subsequent cohorts with the inherent risk of
temporal bias. With respect to this care was given that avoid protocol changes and
unchanged relative reporting contribution amongst the 3 Consultant during the study
period. Most importantly, like with all previous work in this field the definition of
minor and major criteria, also pre-specified and in line with the subsequently
published 210 Expert Consensus on Coronary computed Tomographic angiography 12,
remains a subjective one and at least minor variations in weighting of IEF, e.g. if
simple hepatic, even if multiple, and / or renal cysts are mentioned in a report can not
be excluded.
Conclusions
Overall, IEF are common and require follow or additional investigations in a
substantial minority of cases. However, the overall incidence of highly significant
findings in the current study was low (~1%) and very similar to that found in the CT
literature. Furthermore, there appears to be no substantial difference in the frequency
of incidental extra-cardiac findings between the 3 clinically applied HASTE
protocols. Whilst the FB HASTE technique is statistically significantly faster than BH
and NAV HASTE the absolute time saving is small and probably out-weight by the
resulting lesser image quality.
Figure 2:
A)
B)
C)
D)
Figure 3:
A)
C)
B)
D)
Figure 4:
A)
B)
C)
Figure 5:
A)
C)
D)
B)
E)
Figure 6:
A)
B)
C)
Figure Legend:
Figure 1: Clinical indication for CMR. Studies could have more than clinical
indication / question but for coding and auditing purposes were allocated to a single
indication as displayed in the pie chart. Data are presented as percentages.
Figure 2: Fifty-six year old male patient, referred for stress perfusion imaging to
evaluate cardiac sounding chest pain, is found to have 2 IEF on CMR.
Images A-D; A: Depicts a well demarcated lung lesion in the left lower zone (yellow
arrow) with low central and higher peripheral signal on coronal scout image.
Additionally, (red arrow) a well rounded liver lesion with lower than surrounding
liver signal was seen. The liver lesion was further investigated with Ultrasound and
judged to reflect a haemangioma. No change in size has occurred over 24 months. The
presence of the lesion was confirmed by lateral chest x-ray (B) and PA (C) chest
radiographs and further evaluated with contrast enhanced CT (D). The nodule
demonstrated a benign pattern of dense calcification and as such was followed with
plain chest radiographs. It has remained unchanged in size for 2 years.
Figure 3: 43 year old female non-smoker referred for evaluation of atypical chest
pains. Coronal Localiser (A) and axial HASTE imaging (B) depict an ill defined
region of increased signal in the apical segment of the left lower lobe.
Axial CT “lung windows” (C) confirms the presence of a 2.7cm ill defined soft tissue
nodule. Image D) illustrates the tumour (red arrow) axial CT “bone window” guided
needle (yellow arrow) biopsy.
The patient underwent curative resection of a broncho-alveolar carcinoma T2 M0 N0
stage 1B. Follow up CT 18/12 post surgery demonstrates freedom from local
recurrence.
Figure 4: Images of 3 Patients with Sarcoidosis.
Image A: HASTE images in a 35 year old man being investigated for sustained
ventricular tachycardia. The three axial cuts from the upper thorax demonstrate right
supraclavicular and bilateral upper paratracheal lymphadenopathy (yellow arrows).
The diagnosis of sarcoidosis was subsequently confirmed with transbronchial biopsy.
There was no evidence of localised myocardial oedema on T2 weighted imaging nor
localised hyperenhancement on myocardial delayed gadolinium imaging.
Image B: Axial HASTE image at the level of the upper arch demonstrating bilateral
pulmonary parenchymal high signal within the posterior upper lobes (yellow arrows)
in a patient presenting with non-sustained VT and subsequently diagnosed with
sarcoidosis. Again the patient had no evidence of localised hyperenhancement on
myocardial delayed gadolinium imaging.
Image C: Phase sensitive inversion recovery (PSIR) delayed Gadolinium image of a
patient investigated for heart failure symptoms. The image shows multiple areas of
localised hyperenhancement (yellow arrows) affecting both ventricles. The patient
also had mild mediastinal lymphadenopathy and subsequent mediastinal lymph node
biopsy confirmed the diagnosis of sarcoidosis.
Figure 5: Examples of additional incidental findings:
Image A; Coronal Scout image demonstrating a large hiatus hernia (arrow).
Image B: Axial HASTE image at the level of the pulmonary bifurcation in an 80 year
old patient with known extrinsic allergic alveolitis demonstrating a rather patulous air
filled oesophagus.
Image C: Lower thoracic HASTE demonstrating isolated rounded high signal
vertebral body lesion judged to reflect a vertebral body haemangioma.
Image D: Axial HASTE image though the upper thorax reveals an azygos fissure
which is a normal variant of no clinical significance.
Image E: Axial HASTE through the lung apex demonstrates generalised enlargement
of the thyroid with a 25mmx25mm nodule, with mildly increased signal as compared
to surrounding thyroid tissue, within the left lobe. Further investigation with
ultrasound demonstrated the nodule to be hypoechogenic with a hyperechogenic
(calcific) rim. Further smaller nodules of the same echogenicity but no
lymphadenopathy was seen. Features were judged to be in keeping with a
multinodular goitre.
Figure 6: 55 year old woman referred for bilateral neck swelling and vague history of
an abnormal aorta.
Image A: Still frame of a coronal SSFP cine demonstrating an elongated right sided
cervical aortic arch (yellow arrow) as well as a large left sided septated cystic thyroid
mass (red arrow).
Images B, C: Volume rendered contrast enhanced Aortogram (anterior view - image
B, posterior view - image C) highlight the abnormally arising arch vessels whereby
the first arch vessel is the left common carotid, followed by the right vertebral artery
(normally arising from the right subclavian), the right common carotid, the right
subclavian (arising from the posterior aspect of the arch), and the left subclavian
arising from an aortic diverticulum (Kommeral).
Reference List
(1) Pennell DJ, Sechtem UP, Higgins CB et al. Clinical indications for
cardiovascular magnetic resonance (CMR): Consensus Panel report. Eur
Heart J 2004 November;25(21):1940-65.
(2) Hendel RC, Patel MR, Kramer CM et al.
ACCF/ACR/SCCT/SCMR/ASNC/NASCI/SCAI/SIR 2006 appropriateness
criteria for cardiac computed tomography and cardiac magnetic resonance
imaging: a report of the American College of Cardiology Foundation Quality
Strategic Directions Committee Appropriateness Criteria Working Group,
American College of Radiology, Society of Cardiovascular Computed
Tomography, Society for Cardiovascular Magnetic Resonance, American
Society of Nuclear Cardiology, North American Society for Cardiac Imaging,
Society for Cardiovascular Angiography and Interventions, and Society of
Interventional Radiology. J Am Coll Cardiol 2006 October 3;48(7):1475-97.
(3) Haller S, Kaiser C, Buser P, Bongartz G, Bremerich J. Coronary artery
imaging with contrast-enhanced MDCT: extracardiac findings. AJR Am J
Roentgenol 2006 July;187(1):105-10.
(4) Onuma Y, Tanabe K, Nakazawa G et al. Noncardiac findings in cardiac
imaging with multidetector computed tomography. J Am Coll Cardiol 2006
July 18;48(2):402-6.
(5) Dewey M, Schnapauff D, Teige F, Hamm B. Non-cardiac findings on
coronary computed tomography and magnetic resonance imaging. Eur Radiol
2007 August;17(8):2038-43.
(6) Hunold P, Schmermund A, Seibel RM, Gronemeyer DH, Erbel R. Prevalence
and clinical significance of accidental findings in electron-beam tomographic
scans for coronary artery calcification. Eur Heart J 2001
September;22(18):1748-58.
(7) Horton KM, Post WS, Blumenthal RS, Fishman EK. Prevalence of significant
noncardiac findings on electron-beam computed tomography coronary artery
calcium screening examinations. Circulation 2002 July 30;106(5):532-4.
(8) Chan PG, Smith MP, Hauser TH et al. Noncardiac pathology on clinical
cardiac magnetic resonance imaging. JACC Cardiovasc Imaging 2009
August;2(8):980-6.
(9) McKenna DA, Laxpati M, Colletti PM. The prevalence of incidental findings
at cardiac MRI. Open Cardiovasc Med J 2008;2:20-5.
(10) Romney BP, Khosa F, Costa DN, et al. Non-cardiac findings on
cardiovascular magnetic resonance imaging are common: impact of imaging
sequence and reading session format. Circulation 2008;18.S784.
(11) Burchel Thomas LD, Mathur Anthony, Davies Ceric, Westwood Mark.
Prevalence of non-cardiac incidental findings during routine clinical CMR
assessment. Journal of Cardiovascular Magnetic Resonance 11 (Suppl 1):P48.
2009.
(12) Mark DB, Berman DS, Budoff MJ et al.
ACCF/ACR/AHA/NASCI/SAIP/SCAI/SCCT 2010 expert consensus
document on coronary computed tomographic angiography:areport of the
American College of Cardiology Foundation Task Force on Expert Consensus
Documents. Circulation 121[22], 2509-2543. 8-6-2010.
(13) Morris Z, Whiteley WN, Longstreth WT, Jr. et al. Incidental findings on brain
magnetic resonance imaging: systematic review and meta-analysis. BMJ
2009;339:b3016.
(14) Charles Peebles. My approach to extra-cardiac findings on CMR. 2010.
Ref Type: Online Source