Effectiveness and durability of benefit of mTOR inhibitors (mTORi) in a real-world cohort of metastatic prostate cancer (mPCA) patients with PI3K pathway alterations.

Abstract

e17034 Background: Alterations in the PTEN/PI3K/AKT/mTOR pathway are prevalent in prostate cancer, as frequent as 41% by prospective comprehensive genomic profiling (CGP) (PMID: 31218271). The use of mTORi in unselected patients (pts) with metastatic castrate-resistant prostate cancer (mCRPC) has limited efficacy in inducing a PSA response or clinical response. In this retrospective study, we evaluated the clinical effectiveness of mTORi specifically in mPCA pts with tissue assessed PI3K pathway alterations. Methods: This study used the nationwide (US-based) de-identified Flatiron Health-Foundation Medicine mPCA clinico-genomic database (FH-FMI CGDB), originating from approximately 280 US cancer clinics (̃800 sites of care). We evaluated treatment data for pts with mPCA from October 2014 to February 2020 and selected all pts who received mTORi therapy. Unique patients who received mTORi therapy were included if they had PI3K pathway alteration by tissue CGP. Demographic and clinical data were extracted and analyzed for PSA response and time to next treatment (TTNT) on mTORi therapy. Results: In a cohort of 2301 mPCA pts with 7208 evaluable treatment lines in the database, there were 20 pts who underwent mTORi treatment. After excluding 1 pt who had liquid CGP and 2 pts who did not have PI3K pathway alterations, we included 17 mPCA pts with tissue assessed PI3K pathway alterations who received mTORi treatment. Pts had a median age of 72y (IQR 68.0, 76.0) and were heavily pretreated (11 pts, 64.7% on 4th- or later-line therapy). Two pts (11.8%) were treated in first-line metastatic hormone-sensitive prostate cancer (mHSPC) setting and the remaining 15 pts (88.2%) had mCRPC. All 15 mCRPC pts received prior novel hormone therapy (NHT), and 10 pts (58.8%) were previously treated with a taxane. The PI3K pathway alterations included PTENdel (10 pts, 58.8%), AKT1mut (4 pts, 23.5%), PTENmut (2 pts, 11.8%), and dual PTENmut and PIK3CAmut (1 pt, 5.9%). Most (15 pts, 88.2%) were treated with everolimus (EVE) monotherapy; 1 pt received EVE + abiraterone (EVE/ABI); 1 pt had EVE + docetaxel (EVE/DOC). The median pre-treatment PSA (14 pts available) was 148.9 ng/dL (range 1.16, 3430). Among 10 pts with on treatment PSA available, 3 pts had PSA decrease at ̃ week 12 (range -0.1%, -44.0%) and 5 patients had PSA decrease overall. Only 1 pt (with PTEN homozygous deletion) had a PSA response with nadir -95.1% and TTNT of 7.5 months (mo). The median TTNT was 3.62 mo (range 0, 8.52). Interestingly, the ABI/EVE pt had PSA nadir -17.2% and TTNT 2.0 mo; the EVE/DOC pt had nadir of -21.9% and TTNT 7.5 mo. Conclusions: The clinical use of mTORi in patients with prostate cancer is low overall. In this small cohort of mPCA pts with tissue assessed PI3K pathway alterations, mTORi therapy was not effective with few PSA responses and short duration of therapy.