Pathological Diagnoses in Cases of Indeterminate or Unknown Primary Submitted for Molecular Tumor Profiling
Brock E Schroeder, Marianne Laouri, Er Chen, Mark G Erlander, Catherine A Schnabel. bioTheranostics, Inc., San Diego, CA; Deloitte, San Francisco, CA; Quorum Consulting, Inc., San Francisco, CA
Background: Conclusive determination of tissue-of-origin has significant therapeutic implications and presents a major diagnostic challenge. This case series examines the pathologic characterization and role of diagnostic immunohistochemistry (IHC) in malignant tumor samples submitted for molecular tumor profiling using the 92-gene cancer classifier, CancerTYPE ID (CTID).
Design: De-identified cases (N=815) submitted for CTID testing were retrospectively analyzed. Sample characteristics including morphology, tumor grade, IHCs, suspected primary site, and time to completion (TTC; based on dates of sample collection and pathologist sign-out) were abstracted from pathology reports. Pathologic diagnoses were categorized as unknown, single suspected primary, differential (≥2 suspected), or no primary site reported. CTID data included probabilities, % tests with reportable results and TTC.
Results: 754 cases had evaluable pathology reports (age 62 y; 49% male; 82% carcinoma/ adenocarcinoma; 46% excision, 40% core needle, 14% FNA/cell block/other). Conventional pathologic work-up included a mean of 7 IHCs (range: 0-35) and a mean TTC of 7.7 d (range: 0-127). CK7 (75%), CK20 (70%), TTF1 (63%) were the most common IHCs. 55% underwent all 3. Pathology reports indicated that upon submission for molecular testing 20% of cases had a single suspected primary, 53% a differential diagnosis, 8% were unknown, and 20% had no specified primary site. Notably, although cases with >7 IHCs vs ≤7 IHCs had longer TTC (9.3 vs 6.4d, p=.0017), they were not associated with an increase in single suspected primaries or a reduction in unknown diagnoses. CTID predicted a primary site in 91% of cases (median probability: 78%, range: 23-96%; mean TTC: 5.1d), and yielded similar results across biopsy types (excision: 92%, core needle: 92%, FNA/cell block/other: 89%). CTID provided a prediction in 93% (n=59) of cases with a pathologic diagnosis of unknown. In differential cases (n=399), CTID resolved the pathologic diagnosis in 55% and provided a new primary in 36%.
Conclusions: This large retrospective analysis demonstrated current IHC practices are nonstandardized in protocol, interpretation and TTC and there is a significant need for additional analytical approaches to aid in resolving indeterminate primary and differential diagnoses. CTID is an objective, standardized test with a high analytical success rate that may facilitate diagnosis by providing adjunctive molecular data in tumors with indeterminate or unknown origin.
Wednesday, March 21, 2012 1:00 PM
Poster Session VI # 80, Wednesday Afternoon