Treatment outcomes with first-line (1L) nab-Paclitaxel + gemcitabine (AG) and FOLFIRINOX (FFX) in metastatic pancreatic adenocarcinoma (mPAC).

e18147 Background: AG and FFX have demonstrated improved overall survival (OS) compared to gemcitabine (GEM) as 1L treatment (tx) for mPAC. Limited comparative data exists on AG vs FFX. This study compared time to treatment failure (TTF), OS, and patient (pt) demographics in mPAC pts receiving 1L AG vs FFX, and evaluated outcomes by performance status (PS) in a real-world setting. Methods: This was a retrospective observational cohort study of adult pts diagnosed with stage IV mPAC treated with 1L AG, FFX, or GEM monotherapy between 04/2013 and 10/2015 with at least 3 mo follow-up, to 01/2016. We report results for AG vs FFX; AG vs GEM analysis is ongoing. Data were programmatically extracted from the iKnowMed electronic health record database of pts in The US Oncology Network. Eligibility included ≥3 doses of A (for AG) or ≥2 doses of fluorouracil (for FFX), and excluded pts on clinical trial or with other cancer diagnoses during the study period. Categorical variables were compared with chi-squared tests. TTF and OS were assessed using the Kaplan-Meier method. Results: 2,901 pts with mPAC receiving chemotherapy during the study period were initially identified. 486 total pts met all eligibility criteria, with 255 AG and 159 FFX pts. Median age was 68 and 61 yrs for AG and FFX respectively (p < .0001). ECOG PS was 0-1 in 77% of AG pts and 91% of FFX pts (p = 0.0004). Median TTF and mOS in all pts was 3.5 [95% CI 3.0-3.9] and 9.8 mo [95% CI 8.2-11.5] respectively. There was no statistical difference in mTTF or mOS for the AG vs FFX cohorts (mTTF 3.7 vs 4.3 mo, log-rank p = 0.25; and mOS 9.8 vs 11.4 mo, log-rank p = 0.38). Among pts with ECOG PS 0-1, TTF for AG vs FFX was 4.2 vs 4.3 mo respectively (log-rank p = 0.47); and OS for AG vs FFX was 12.1 vs 11.4 mo (log-rank p = 0.68). Conclusions: Although pts in the AG cohort were older and had worse PS, mTTF and mOS were not statistically different from pts in FFX. Results were similar after stratifying by PS. Limitations include the retrospective and non-randomized nature of the study. However, this represents real world outcomes where randomized data do not exist. Analysis of toxicities and resource utilization is ongoing and will aid in optimizing tx selection in these pts.