Download National Mesothelioma Virtual Bank (NMVB)

AUTOMATED BIOSPECIMENS ANNOTATION
MODEL FOR SCALABLE MESOTHELIOMA
BIOBANKING
(HTTP://WWW.MESOTISSUE.ORG)
SUPPORTED BY CDC/NIOSH GRANT 2U24OH009077-08
Waqas Amin, MD
Senior Research Scientist
Department of Biomedical Informatics
University of Pittsburgh
Disclosure
"I disclose that neither I nor my partner have relevant
financial relationships with commercial interests."
Overview:
 The National Mesothelioma Virtual Bank (NMVB) provides de-identified




annotated mesothelioma biospecimens to the mesothelioma research
community.
We collect retrospective and prospective mesothelioma cases at New
York University (NYU), University of Pennsylvania (U Penn), University
of Pittsburgh (U Pitt) and Rowell Park Cancer Institute (new collection
site).
Key to this resource is continued annotation of mesothelioma
biospecimens including demographic, epidemiologic, clinicopathologic,
follow-up and recurrence data for all biospecimens collected.
The NMVB database allows a user to query the biospecimen resource to
identify cohorts for translational research.
We then provide well-annotated high quality mesothelioma
biospecimens and data to approved researchers via a Letter of Intent –
see http://www.mesotissue.org/apply.cfm
Accomplishments 2006-2014:
 Established biobanking infrastructure at four collaborative sites.
 Collected more than 1150 mesothelioma pleural and peritoneal
biospecimens (biopsies and surgical resections).
 Established a centralized NMVB database for specimen annotation and
query.
 Fulfilled over 35 specimen and data requests across the country and
overseas. Represent a total of 531 patients’ material shared with the
mesothelioma research community.
 Developed three distinct Tissue Microarray slides to examine the
distribution of marker molecules in hundreds of different tissues displayed
on a single slide.
NMVB-CAE Model:
Built upon the caTISSUE Clinical Annotation Engine (CAE)
developed by DBMI as part of the NCI-caBIG project.
Takes a UML Domain Model as input.
 Generates core application components:
 Database
 Metadata
 Query capability
 Data viewers
 Data entry forms
NMVB-CAE Model:
Achieved semantic and syntectic interoperability by describing the
common data elements in the form of metadata or data descriptors
and by using a controlled vocabulary.
Emphasis has been placed to provide user access at three levels.
NMVB statistical data for public view
 Approved investigator database query that allows seeing
individual patient de-identified clinical data
Data manger access to query and edit the stored data. Patient
privacy is of utmost importance at all levels of user's access.

NMVB-CAE Model: Limitations
Existing informatics solution is expensive, laborious and quickly
exhausts the resource available to sustain the operation of tumor
bank.
Discouraging to share datasets to multi-institutional studies.
Self contained model that doesn’t offer modularity and self scaling.
Data elements are not recorded or normalized to standardized
terminologies (SNOMED-CT, ICD-9 etc…).
Sustaining the Expansion of the National Mesothelioma Virtual
Bank (NMVB)
 Ongoing Aim: To continue to serve the needs of the mesothelioma cancer
research community by collecting tissue, blood and clinical data and
providing efficient access to these federated resources.
 Expansion Aim: To expand the NMVB to the Roswell Park Cancer Institute
(RPCI) as a collection site.
 Sustainability Aim: To automate biospecimen annotation through
electronic extraction of clinical and pathology data from electronic health
records (EHR) and cancer registry system. Implementing a more
sustainable informatics federation model by deploying i2b2 (Informatics
for Integrating Biology and the Bedside) and SHRINE (Shared Health
Research Information Network) to maximize the effectiveness of the data
mining process across NMVB sites.
Design Principles (Sustaining Aim):
 To automate the biospecimen annotation through electronic extraction
of clinical and pathology data from electronic health records (EHR), TIES
and cancer registry system.
 Adopting the Text Information Extraction System (TIES), a clinical
document search engine, to identify the cases and reports.
 Adopting a new informatics federated model by deploying i2b2
(Informatics for Integrating Biology and the Bedside) and SHRINE
(Shared Health Research Information Network) to maximize the
effectiveness of the data mining process across collaborations.
 New Informatics model will leverage the existing funding from other
funding sources (CTSA-ACTS and PCORI).
Design Principles (Sustaining Aim):
 Provide data contributors with full ownership of and access to their own data.
 Minimize barriers for data owners to collaboratively contribute their data to
new or existing datasets.
 Support a tiered sharing model which provides a granular, permissioned,
and audit-capable data sharing framework.
 Enable near real-time access to data, supporting a virtuous cycle in which
immediate data access promotes further data contribution and collaboration.
 Encourage ongoing incorporation of outside datasets from multiple sources.
Common Data Elements (CDEs):
 With leadership of the MVB Coordinating Committee, we
established a CDE subcommittee to develop CDEs pertinent to
mesothelioma: demographic, epidemiologic, clinical, pathologic
specimen and block annotation, follow up and outcome.
 Major standards used to build CDEs:
 North American Association of Central Cancer Registry (NAACCR)
core elements.
 College of American Pathology (CAP) checklists.
 American Joint Commission on Cancer (AJCC) staging.
 Association of Directors of Anatomic and Surgical Pathology
(ADASP) guidelines.
TIES NLP process for Surgical Pathology Reports
Specimen Collection:
Specimen Types:
 Tissue Microarray with Clinical Data Annotation
 Fresh Frozen Tissues
 Blood Products: Serum, RBC, Plasma, Whole blood, Buffy Coat,
etc…
 Paraffin embedded blocks
NMVB Accrual-Blood Products:
Type of Blood Products
Number of Cases
Serum
400
Whole Blood
305
Buffy Coat
225
Plasma
401
Red Blood Cells
245
NMVB Accrual-Surgical Specimens:
Specimen
Type
Surgical Procedure Type
Biopsy Specimen
Resected Specimen
Paraffin
166
362
Fresh Frozen
42
360
Bulk Frozen
12
108
http://www.mesotissue.org/
NMVB Related and Supported Publications:
1. Amin W, Parwani AV, Melamed J, Raja F, Pennathur A, Valdevieso F, Whelan NB, Landreneau R, Luketich J, Feldman M,
Pass HI, Becich MJ. National Mesothelioma Virtual Bank: A Platform for Collaborative Research and Mesothelioma
Biobanking Resource to Support Translational Research, Lung Cancer International, vol. 2013, Article ID 765748, 9
pages, 2013. doi:10.1155/2013/765748.
2. Amin W, Srinivasan M, Song S, Parwani AV, Becich MJ. Use of Automated Image Analysis in Evaluation of
Mesothelioma Tissue Microarray (TMA) from National Mesothelioma Virtual Bank. Pathology Research and Practice.
2014 February, 210(2) 79–82 PMID: 24359722.
3. Amin W, Kang HP, Egloff AM, Singh H, Trent K, Ridge-Hetrick J, Seethala RR, Grandis J, Parwani AV. An informatics
supported web-based data annotation and query tool to expedite translational research for head and neck
malignancies. BMC Cancer. 2009 Nov 13;9:396. PMID: 19912644 PMCID: PMC2780457.
4. Mohanty SK, Mistry AT, Amin W, Parwani AV, Pople AK, Schmandt L, Winters SB, Milliken E, Kim P, Whelan NB, Farhat
G, Melamed J, Taioli E, Dhir R, Pass HI, and Becich MJ. The development and deployment of Common Data Elements
for tissue banks for translational research in cancer - an emerging standard based approach for the Mesothelioma
Virtual Tissue Bank. BMC Cancer. 2008 Apr 8;8:91. PMID: 18397527 PMCID: PMC2329649.
5. Pass HI, Lott D, Lonardo F et al. Asbestos exposure, pleural mesothelioma, and serum osteopontin levels. N Engl J
Med. 2005 Oct 13;353(15):1564-73.
6. Zhang X, Shen W, Dong X, Fan J, Liu L, Gao X, Kernstine KH, Zhong L. Identification of novel autoantibodies for
detection of malignant mesothelioma. PLoS One. 2013 Aug 19;8(8):e72458. doi: 10.1371/journal.pone.0072458.
PMID: 23977302.
7. Newick K, Cunniff B, Preston K, Held P, Arbiser J, Pass H, Mossman B, Shukla A, Heintz N. Peroxiredoxin 3 is a redoxdependent target of thiostrepton in malignant mesothelioma cells. PLoS One. 2012;7(6):e39404. doi:
10.1371/journal.pone.0039404. Epub 2012 Jun 25.
Acknowledgment:
Collaborators:
 Center for Disease Control and Prevention (CDC)
 National Institute of Occupational Safety & Health (NIOSH)
 Mesothelioma Foundation (Meso Fndn)
 Mount Sinai School of Medicine (MSSM)
 New York University (NYU), New York City, NY
 University of Pennsylvania (U Penn), Philadelphia, PA
 University of Pittsburgh (U Pitt), Pittsburgh, PA
 Roswell Park Cancer Institutes (RPCI), Buffalo, NY
Future Partners: NCI Meso SPORE Core to NMVB in 2015 with the University of Hawaii
(UHCC) via Meso SPORE to also include Mayo Clinic.
Leadership:
Michael J. Becich MD, PhD (U Pitt)
Michael Feldman MD (U Penn)
Harvey I. Pass MD (NYU)
Jonathan Melamed MD (NYU)
Rebecca Jacobson MD, MS (U Pitt)
Anil V. Parwani MD, PhD (U Pitt)
Steven Abelda, MD (U Penn)
David Bartlett, MD (U Pitt)
Carl Morrison, MD (RPCI)
Carmelo Gaudioso (RPCI)
Angela R. Omilian (RPCI)
James Luketich M.D (U Pitt)
Mary Hesdorfer (Meso Fndn)
James Pingpank, MD (U Pitt)
Raja Flores, MD (MSSM)
Thank you