Sharon Winters

A multidisciplinary approach to honest broker services for tissue banks and clinical data: a pragmatic and practical model.

Honest broker services are essential for tissue‐ and data‐based research. The honest broker provides a firewall between clinical and research activities. Clinical information is stripped of Health Insurance Portability and Accountability Act‐denoted personal health identifiers. Research material may have linkage codes, precluding the identification of patients to researchers. The honest broker provides data derived from clinical and research sources. These data are for research use only, and there are rules in place that prohibit reidentification. Very rarely, the institutional review board (IRB) may allow recontact and develop a recontact plan with the honest broker. Certain databases are structured to serve a clinical and research function and incorporate ‘real‐time’ updating of information. This complex process needs resolution of a variety of issues regarding the precise role of the HB and their interaction with data. There also is an obvious need for software solutions to make the task of deidentification easier.

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

BackgroundRecent advances in genomics, proteomics, and the increasing demands for biomarker validation studies have catalyzed changes in the landscape of cancer research, fueling the development of tissue banks for translational research. A result of this transformation is the need for sufficient quantities of clinically annotated and well-characterized biospecimens to support the growing needs of the cancer research community. Clinical annotation allows samples to be better matched to the research question at hand and ensures that experimental results are better understood and can be verified. To facilitate and standardize such annotation in bio-repositories, we have combined three accepted and complementary sets of data standards: the College of American Pathologists (CAP) Cancer Checklists, the protocols recommended by the Association of Directors of Anatomic and Surgical Pathology (ADASP) for pathology data, and the North American Association of Central Cancer Registry (NAACCR) elements for epidemiology, therapy and follow-up data. Combining these approaches creates a set of International Standards Organization (ISO) – compliant Common Data Elements (CDEs) for the mesothelioma tissue banking initiative supported by the National Institute for Occupational Safety and Health (NIOSH) of the Center for Disease Control and Prevention (CDC).MethodsThe purpose of the project is to develop a core set of data elements for annotating mesothelioma specimens, following standards established by the CAP checklist, ADASP cancer protocols, and the NAACCR elements. We have associated these elements with modeling architecture to enhance both syntactic and semantic interoperability. The system has a Java-based multi-tiered architecture based on Unified Modeling Language (UML).ResultsCommon Data Elements were developed using controlled vocabulary, ontology and semantic modeling methodology. The CDEs for each case are of different types: demographic, epidemiologic data, clinical history, pathology data including block level annotation, and follow-up data including treatment, recurrence and vital status. The end result of such an effort would eventually provide an increased sample set to the researchers, and makes the system interoperable between institutions.ConclusionThe CAP, ADASP and the NAACCR elements represent widely established data elements that are utilized in many cancer centers. Herein, we have shown these representations can be combined and formalized to create a core set of annotations for banked mesothelioma specimens. Because these data elements are collected as part of the normal workflow of a medical center, data sets developed on the basis of these elements can be easily implemented and maintained.

A decade of experience in the development and implementation of tissue banking informatics tools for intra and inter-institutional translational research

Context: Tissue banking informatics deals with standardized annotation, collection and storage of biospecimens that can further be shared by researchers. Over the last decade, the Department of Biomedical Informatics (DBMI) at the University of Pittsburgh has developed various tissue banking informatics tools to expedite translational medicine research. In this review, we describe the technical approach and capabilities of these models. Design: Clinical annotation of biospecimens requires data retrieval from various clinical information systems and the de-identification of the data by an honest broker. Based upon these requirements, DBMI, with its collaborators, has developed both Oracle-based organ-specific data marts and a more generic, model-driven architecture for biorepositories. The organ-specific models are developed utilizing Oracle 9.2.0.1 server tools and software applications and the model-driven architecture is implemented in a J2EE framework. Result: The organ-specific biorepositories implemented by DBMI include the Cooperative Prostate Cancer Tissue Resource ( http://www.cpctr.info/ ), Pennsylvania Cancer Alliance Bioinformatics Consortium ( http://pcabc.upmc.edu/main.cfm ), EDRN Colorectal and Pancreatic Neoplasm Database ( http://edrn.nci.nih.gov/ ) and Specialized Programs of Research Excellence (SPORE) Head and Neck Neoplasm Database ( http://spores.nci.nih.gov/current/hn/index.htm ). The model-based architecture is represented by the National Mesothelioma Virtual Bank ( http://mesotissue.org/ ). These biorepositories provide thousands of well annotated biospecimens for the researchers that are searchable through query interfaces available via the Internet. Conclusion: These systems, developed and supported by our institute, serve to form a common platform for cancer research to accelerate progress in clinical and translational research. In addition, they provide a tangible infrastructure and resource for exposing research resources and biospecimen services in collaboration with the clinical anatomic pathology laboratory information system (APLIS) and the cancer registry information systems.

The importance of pathology informatics in translational research.

Pathology informatics involves management and analysis of large complex data sets derived from various tests performed in clinical and anatomic pathology laboratories, annotated biorepositories, image analysis, telepathology, and large scale experiments, including gene expression analysis, proteomics, and tissue array studies. It facilitates intelligent use of computing technologies to improve patient care and understand the natural history of disease. Herein, we describe the various bioinformatics tools used to support translational research at the University of Pittsburgh Medical Center.

FOLFIRINOX and gemcitabine/nab-paclitaxel efficacy in the treatment of locally advanced unresectable pancreatic adenocarcinoma.

399 Background: Locally advanced (LA) unresectable pancreatic adenocarcinoma (PDA) historically portends a poor prognosis with a median OS of 9-11 months. Recently, two multi-drug regimens – FOLFIRINOX and gemcitabine/nab-paclitaxel – have proven effective in the metastatic setting. We hypothesized that use of these regimens in the LA setting may improve survival. Methods: A retrospective review of a single institution’s cancer registry of all consecutive LA (unresectable) PDA patients between 2010 and 2014 was performed. LA status was verified by review of the triphasic, pancreas protocol CT scan at diagnosis using the 2015 NCCN criteria for resectability. Patients were divided into 4 groups: Group 1 = no therapy, Group 2 = “old” gemcitabine or 5-FU-based chemotherapy (CTX), Group 3 = “new” CTX (FOLFIRINOX and/or Gem/nab-paclitaxel), and Group 4 = resection after downstaging. Demographic, tumor related variables, and treatment outcomes were analyzed. Results: LA disease was verified in 107 consecutive pa...

Survival differences in colorectal cancer (CRC) on the basis of race: University of Pittsburgh Medical Center (UPMC) experience

4041 Background: CRC has higher 5-yr survival in whites (W) (65%) as compared to African Americans (AA) (55%). Our study was aimed to detect factors that might affect survival in these two groups at UPMC. Methods: Of 2,494 patients (pts) diagnosed with CRC at UPMC from 1994–2005, complete data (age, sex, disease status, tumor characteristics, treatment & overall survival) were available for 1,800 pts (181 AA & 1619 W). Results: Average age was 68 yrs in AA vs 69 in W. Pts with stage I, II, III, & IV in AA vs W were 22% vs 18%, 27% vs 34%, 32% vs 29% & 19% vs19%, respectively. Tumor grade differentiation in AA vs W was well (WD) in 2% vs 3%, moderate (MD) in 83% vs 76 %, poor (PD) in 14%vs 20% and undifferentiated (UD) in 1% vs 1% of pts, respectively. Median (M) time from diagnosis to starting treatment was 8 days in AA vs 11 in W. M number (no.) of lymph nodes (LN) examined were 12 in AA vs 14 in W. Overall M survival time (MST) was 49 months (m) in AA vs 67 m in W (p=0.09). MST for stage II+III pts was ...