Molecular gene expression profiling to predict the tissue of origin and direct site-specific therapy in patients (pts) with carcinoma of unknown primary site (CUP): Results of a prospective Sarah Cannon Research Institute (SCRI) trial.

Abstract

10530 Background: Tumor profiling is an emergent technique to determine tissue of origin in CUP patients. However, the value of these predictions in improving treatment efficacy is unknown. In this prospective trial, we used tumor profiling results to direct site-specific therapy for CUP pts. Methods: A 92-gene RT-PCR assay (CancerTYPE ID; bioTheranostics, Inc.) was performed on tumor biopsies from previously untreated CUP pts who consented. When a tissue of origin was predicted, pts who were treatment candidates were assigned standard site-specific first-line therapy. Results: Between 10/08 and 12/11, 289 pts were enrolled, 252 had successful assays performed, and 247 (98%) had a tissue of origin predicted. 224 pts were eligible for treatment; 197 pts received assay-directed treatment. 120 of 224 treated pts (54%) had assay diagnoses of tumor types known to derive substantial benefit from standard site-specific treatment (bladder 27, colorectal 26, NSCLC 24, breast 10, ovary 10, kidney 9, prostate 4, germ cell 4, others 6 (3 sites), while 104 pts (46%) had assay diagnoses of relatively resistant tumors (biliary tract 45, pancreas 12, gastroesophageal 10, liver 7, sarcoma 5, cervix 5, others 20 (8 sites). Median OS for all treated pts was 10.8 months (mos); OS for 197 pts with assay-directed treatment was 12.2 mos (versus 6.0 mos for 27 pts receiving empiric therapy). Median OS was better in the 120 pts with assay diagnoses of more responsive tumor types (12.8 vs 7.4 mos; p = .027). Median OS (mos) in specific subgroups: pancreas 9, kidney 12, colon 12, NSCLC 16, ovary 30. Conclusions: This is the first prospective trial in which molecular profiling has directed site-specific therapy in CUP pts. Assay-directed therapy in 197 pts produced a median OS (12.2 mos) that compares favorably with previous empiric CUP therapy. CUP pts predicted to have more responsive tumor types had longer survival compared to less responsive types, suggesting accurate identification by the assay. These results strengthen the rationale for molecular profiling in CUP management.

Blinded Comparator Study of Immunohistochemical Analysis versus a 92-Gene Cancer Classifier in the Diagnosis of the Primary Site in Metastatic Tumors

“…In this prospective clinical trial of patients diagnosed as having cancer of unknown primary site, many of whom had poorly differentiated or undifferentiated tumors, those who were treated with site-specific therapy based on the 92-gene assay molecular diagnosis had overall survival of 12.5 months, a result that compared favorably with patients from previous prospective trials of patients with cancer of unknown primary site who were treated with empirical regimens (9.1 months) and in a subset of patients in this trial treated with empirical regimens (4.9 months). 27,28 There are several limitations to this study. First, case selection was not representative of daily practice in that case selection specifically identified difficult-to-diagnose tumors; thus, the study is not reflective ofdand likely underestimatesdthe overall accuracy of both methods.…”