Early Progression (progr) in Patients (pts) with Metastatic Pancreatic Cancer (mpac) and a Germline BRCA Mutation (gbrcam): Phase III POLO Trial of Olaparib (O) Versus Placebo (P).

750 Background: In POLO (NCT02184195), maintenance O was associated with significant progr-free survival benefit vs P in pts with a gBRCAm and mPaC (Golan NEJM 2019). Early progr or death (within 4 months [m]) occurs in ~35−45% of pts on standard-of-care first-line (1L) chemotherapy for mPaC (Conroy NEJM 2011; von Hoff NEJM 2013); however, predictive factors are currently unknown and early progr has not been addressed in the maintenance setting. We examined factors potentially associated with early progr in POLO. Methods: Following ≥16 weeks of 1L platinum-based chemotherapy (PBC) without progr, pts were randomized to maintenance O (tablets; 300 mg bd) or P until progr or unacceptable toxicity. Early progr was defined as progr (by blinded independent central review) or death within 4 m of randomization. A stepwise logistic regression model included baseline (BL) factors age, albumin, lactate dehydrogenase (LDH), global health status (GHS) and physical functioning (PhysF) as continuous variables, and discrete variables listed in the Table. Results: 62/154 randomized pts (40%) were defined as early progressors (EP; Table). Due to missing BL data, the multivariate analysis included 127 pts (56 EPs [44%]). Lower BL PhysF score (continuous) was significantly associated with early progr ( P= 0.02); no difference for partial/complete response (PR/CR) vs stable disease (SD). Conclusions: While small sample size limited analysis power, PhysF score was the only BL factor significantly associated with early progr in pts with a gBRCAm and mPaC in the POLO trial of maintenance O vs P. Clinical trial information: NCT02184195 . [Table: see text]