Five-year outcomes of perioperative or only adjuvant gemcitabine plus nab-paclitaxel for resectable pancreatic cancer (the NEONAX trial): A randomized phase II trial of the AIO pancreatic cancer group.

4193 Background: Perioperative chemotherapy (CTX) in resectable pancreatic ductal adenocarcinoma (rPDAC) is still not considered standard of care and data are limited. NEONAX examined gemcitabine (Gem)/nab-paclitaxel (nab-P), in the perioperative or adjuvant therapy of resectable PDAC (NCCN criteria). Methods: NEONAX is a prospective, randomized phase II trial with two independent arms (127 patients, 22 German centers) and randomization 1:1 to perioperative (2 pre- and 4 postoperative cycles, po-arm A) or adjuvant (6 cycles, ad-arm B) of gem (1000mg/m2 BSA, d1, 8, 15) and nab-P (125mg/m2 BSA, d1, 8, 15), q4w. Results: As we previously reported were R0- and N0-resection rates high (po-R0 88%, ad-R0 67%, po-N0 33%, ad-N0 29%) in the ITT-population (all randomized pts.). The primary endpoint DFS rate of 55% @ 18 months in the mITT population (defined as R0/R1 resected pts. that either started perioperative (A) or adjuvant (B) CTX), was not reached in both arms (arm A: 32%, arm B 41%). Whereas 91.5% of pts. in po-arm A started and 84.7% completed neoadjuvant CTX, only 42.4% of pts. in ad-arm B started and 25% completed adj. CTX, so the CTX dose intensity was higher in the po-arm A. (Seufferlein et al., Ann Onc 2023) Here we report long-time 5-year outcomes according to PFS and OS and present a preplanned subgroup analysis of potential prognostic factors. The mOS in the ITT population (all randomized pts.) was 25.5 mo. in the po-arm and 16.8 mo. in the ad- arm. This corresponds to an mDFS of 11.4 mo. in the po-arm and 5.1 mo. in the ad-arm, respectively. The mOS in the mITT population (all randomized pts. that started po-ctx (po-arm) or started ad-ctx (ad-arm)) was 27.9 mo. in the po-arm and 26.8 mo. in the ad-arm. This corresponds to an mDFS of 14.1 mo. in the po-arm and 16.1 mo. in the ad-arm, respectively. In the preplanned analysis of subgroup factors impacting outcome, the benefit of po-treatment was independent of tumor size, N-status and baseline Ca19-9 level. This benefit was not visible in the mITT population where only patients in the ad-arm were included who had started adjuvant treatment postoperatively. Conclusions: These 5 year data confirm the outcome of patients receiving gem/nab in the po-setting in the ITT population. DFS and OS effect were numerically better compared to the ad-arm although the trial was not powered for direct comparison of the arms. This difference disappeared in the mITT population when patients received more CTX in the ad arm. We conclude that the difference between ITT and mITT is likely because more CTX could administered in the po-arm when the ITT population was considered and may constitute one of the major effects of neoadjuvant chemotherapy in resectable PDAC, particularly when a high rate of patients as in our multicenter trial could get neoadjuvant, but not adjuvant treatment. Clinical trial information: NCT02047513 .