A snapshot of potentially personalized care: Molecular diagnostics in gynecologic cancer.

Abstract

5029 Background: Genetic abnormalities underlie the development of cancer. It has been proposed that tumors be recategorized by gene mutation such as BRAF in LG serous, TP53 in HG serous, and PIK3CA in clear cell and endometrioid tumors. These targets potentially represent an opportunity for personalizing cancer therapy. Methods: Gynecologic Oncology patients at the MGH Cancer Center can have their tumor genotyped for a panel of mutations by SNaPshot, a validated, CLIA approved assay developed by MGH that uses DNA from FFPE tissue to interrogate 160 site-specific mutations across 15 genes (AKT1, APC, BRAF, CTNNB1, EGFR, ERBB2, IDH1, KIT, KRAS, MAP2K1, NOTCH1, NRAS, PIK3CA, PTEN, TP53). At present SNaPshot has no validated endpoints in GYN Cancers but may help identify a useful clinical trial. Results: Between 5/17/10 and 10/17/11, 125 patients consented to SNaPshot genotyping. Patients had a median age of 59 (24-78) yrs. Tumors were ovarian 70(56%), uterine clear, UPSC, or MMMT 16(13%), uterine endometrioid 10(8%), fallopian tube 8(6%), PPC 7(6%), cervical 6(5%), uterine sarcomas (3), ACUP (2), vulvovaginal (2), metastatic (1). A mutation was identified in 41(33%), with 9 of these (23%) having 2 or 3 (n=2) mutations. In the 85 ovarian, FT, and PPC cancers 33% were +ve, but 50% were in TP53. The low mutation rate for TP53 is likely explained by copy number abnormalities (Amplification). 50% of the 10 uterine tumors were +ve, with 3 of those 5 having multiple mutations in the PIK3CA pathway, while 69% of the non-endometrioid uterine tumors had mutations. Only 20% of the vulvo-vaginal and Cx tumors had mutations, both PIK3CA. 19% of the purely serous tumors (n=58) had TP53 mutations, and 37% of the purely clear/endometrioid tumors (n=19) had mutations in PIK3CA, PTEN or AKT. Certain rare tumors did not have identifiable mutations: granulosa cell tumors (2), ovarian small cell (2). 5 pts with a PIK3CA mutation were enrolled on a clinical trial (2 phase II, 3 phase I, 3 uterine, 1 ovary, 1 cervix). Conclusions: SNaPshot can identify potentially important therapeutic targets. However, the incidence of "drugable" targets in ovarian cancer is low, and <5% subjects eventually were treated on a relevant clinical trial.