Trimodality therapy with carboplatin/paclitaxel (CP) or FOLFOX (FFX) for esophageal/esogastric junctional cancer (EC/EGJ): Expanded safety and efficacy data from PROTECT.

Abstract

370 Background: When combined to preoperative radiation therapy (RT), CP and FFX regimen provide both high complete resection (R0) rate for EC/EGJ cancer (Adenis, ASCO 2022). However, it appeared that neoadjuvant chemoradiation (nCRT) with CP is associated with a severe postoperative morbidity rate higher than expected. We present here the expanded safety and efficacy analyses from the PROTECT trial. Methods: PROTECT is a randomized, phase 2 trial which included stage II/III and ECOG PS ≤2 EC or Siewert I-II EGJ cancers. Patients (pts) received FFX or CP with concurrent RT (41.4Gy, 1.8Gy, 23 fractions), followed by surgery 4-8 weeks after completion of nCRT (Messager, BMC Cancer 2016). Co-primary endpoints were proportion of R0 rate and proportion of Clavien-Dindo severe postoperative morbidity. Main secondary endpoints were nCRT and postoperative safety (NCI CTCAE v.4), DFS and OS. Results: 41/50 (82%) and 39/50 (78 %) pts received the planned chemo cycles and concurrent RT in FFX and CP arms, respectively. Grade (gr.) 3–4 AEs related to nCRT (FFX 14/50, 28%; CP 14/50, 28%) occurring in ≥5% of pts included lymphopenia (n=3, 6%; n=4, 8%), neutropenia (n=1, 2%; n=3, 6%), fatigue (n=2, 4%; n=0) and esophagitis-related to RT (n=1, 2%; n=1, 2%). No death was reported during nCRT. Surgery (FFX and CP; mini invasive: 15 and 15, hybrid: 22 and 21, open: 7 and 12) was performed in 44 and 48 pts, in FFX and CP groups, respectively. The main gr. III-V surgical complications (Clavien-Dindo scale) occurring in ≥5% of pts included esophageal fistula (n=2/43 evaluable pts, 6%; 8/48, 17%), conduit necrosis (n=2, 5%; n=1, 2%), ARDS (n=3, 7%; n=3, 6%), pleural effusion (n=3, 7%; n=4, 8%), and haemorrhage (n=0; n=3, 6%). There were no postoperative deaths. With a median follow-up of 54m, median DFS were 12.3m and 20m (HR=0.84; 95%CI: 0.52-1.35; p=0.48) and median OS were 31.7m and 45.8m (HR=0.79; 95%CI: 0.47-1.32; p=0.36) in FFX and CP arms, respectively. Prognostic factors significantly associated to DFS in univariate analysis were: R0 resection, TRG1-2 and ypT0N0 status. Conclusions: A higher than expected number of severe esophageal fistula was observed in the CP arm. We could not demonstrate a significant benefit of CP compared to FFX in terms of survival outcomes, but the study was not primarily designed to specifically address this issue. Clinical trial information: NCT02359968 .