Background: Preoperative CRT with a FU/platinum regimen has been used for years for esophageal or junctional cancer before the CP regimen became a standard of care following the results of the CROSS study (van Hagen 2012). We aimed at evaluating the complete resection (R0) rate and severe postoperative morbidity rate associated with these 2 neoadjuvant regimens, each being combined with the radiation (RT) regime used in the CROSS trial. Methods: PROTECT is a multicenter, randomized, non-comparative, phase 2 trial (NCT02359968) in patients (pts) with resectable esophageal or Siewert type I-II junctional cancer, stage II (T1-2N1 or T3N0) or stage III (T3N1 or T4anyN) tumors (UICC-7 classification), and ECOG PS ≤2. Following randomization (balanced by ECOG PS 0 vs 1-2, stage II vs III, squamous-cell (SCC) vs adenocarcinoma (ADK), center), pts received CP (AUC2 CBP plus PCL 50mg/m² / week x 5 weeks), or Fx (FU 400 mg/m² bolus Day 1, then FU 1600 mg/m² continuous infusion over 2 days, plus OX 85 mg/m², and Folinic acid 200 mg/m², 2-h infusion, Day 1; 3 cycles every 2 weeks). RT technique was similar in both arms: 3D-conformal as published in the CROSS trial or IMRT (n = 35); total dose of 41.4Gy, 5 fractions of 1.8Gy / week, starting at Day 1 of chemotherapy. Surgery was performed 4 to 8 weeks after completion of CRT through a transthoracic or mini-invasive approach with a two field extended lymphadenectomy. Co-primary endpoints were R0 (failure: R1 or disease progression under CRT), and severe postoperative morbidity rate ≤30 days after surgery (Clavien-Dindo grade ≥ III). Based on a Bryant and Day 2-stage design (p0 = 75% and p1 = 90% for resection; p0 = 45% and p1 = 25% for morbidity; α = 10% and β = 15%), 48 evaluable pts were required by arm. Results: 100/104 pts recruited from 02/2015 to 08/2020, started the study treatment: 50 CP & 50 Fx. Overall, median age = 64 (range, 33-79); 82/100 males; 62 ADK and 38 SCC; 66 esophageal and 34 junctional site; 31 stage II; 68 stage III, 1 Nx. R0 resection was obtained in 46/50 CP pts (92.0%, 95% CI: 80.8-97.8%), and in 42/48 Fx pts (87.5%, 74.8-95.3%); 2 non evaluable pts because of event unrelated to disease progression. Severe postoperative adverse events (AEs) occurred in 34/91 pts who underwent surgery: 21/48 CP (43.8%, 29.5-58.8%) and 13/43 Fx (30.2%, 17.2-46.1%). Severe AEs were respiratory disorders (CP 26%; Fx 26%), esophageal fistula (CP 18%; Fx 6%), infection (CP 5%; Fx 3%), haemorrhage (CP 5%; Fx 0%) and gastric tube necrosis (CP 6%; Fx 3%). 5 pts died from AEs (3 CP, 2 Fx). A TRG1-2 was observed in 29/48 (60.4%, 95% CI: 44.3-74.2%) CP pts, and in 19/43 (44.2%, 29.1-60.1%) Fx pts. Conclusions: When combined to preoperative radiation therapy at 41.4Gy, both regimens (CP and Fx) provided short-term benefit on R0 resection; however, CP is associated with a severe postoperative morbidity rate higher than expected. Clinical trial information: NCT02359968.