Tislelizumab (TIS; BGB-A317) plus chemotherapy (CT)/chemoradiotherapy (CRT) as positron emission tomography (PET)–guided neoadjuvant (n) treatment (tx) for resectable esophageal squamous cell carcinoma (R-ESCC): RATIONALE-213 final analysis.

4027 Background: Studies have shown overall survival improvements with nCRT + surgery vs surgery alone in locally advanced ESCC. However, preoperative CRT may have additional safety concerns, leading to some patients (pts) receiving nCT rather than nCRT. PET-computed tomography maximum standardized uptake value (SUV max ) change after induction CT (IC) has been shown to have reliable predictive value of pathological complete response (pCR) for R-ESCC in pts with nCRT and may optimize neoadjuvant tx selection. TIS (anti-PD-1) has improved survival in pts with ESCC. We report the final analysis of RATIONALE-213, a phase 2, open-label, multicenter study in China evaluating PET-guided nTIS + CT/CRT in R-ESCC (NCT04974047). Methods: Eligible adult pts had histologically confirmed R-ESCC (cT1-2N + M0 or cT3N any M0), ECOG performance status 0/1, adequate organ function, no fistula risk, and had received no prior tx. Pts had a baseline (BL) PET scan, 1 cycle of IC (cisplatin-paclitaxel [Cis-Pac]), and a PET scan 15-21 days later. Pts were grouped into 2 cohorts by response to IC based on the percentage decrease in 2nd PET SUV max in the primary tumor: responders (R, ≥35%) or nonresponders (NR, < 35%). Both cohorts received 3 cycles of TIS 200 mg IV Q3W, the first 2 with CT (2 cycles Cis-Pac) for R, or with CRT (2 cycles investigator [Inv]-chosen CT [Cis-Pac, or 5-FU + Cis] + RT [40 Gy/20 fractions]) for NR, then surgery. Primary endpoint was pCR per local pathologist. Secondary endpoints were 1-year disease-free survival (DFS), 1-year event-free survival (EFS), objective response rate (ORR) before surgery, R0 resection rate by Inv, and safety. Results: Of 70 pts enrolled, 15 (21.4%), 48 (68.6%), and 7 (10.0%) had stage II, III, and IVA disease at BL, respectively. As of 25 Oct 2024 (median follow-up 25.5 mo), 30 pts were R and 40 NR. Of R, 20 (66.7%) had surgery. Of NR, 32 (80.0%) had surgery. Efficacy endpoints are shown in the table. Median DFS and EFS were not reached for R and NR. Grade ≥3 treatment-related adverse events (TRAEs) in R (15 [50.0%]) and NR (33 [82.5%]) were consistent with known CT or CRT toxicity; serious TRAEs occurred in 5 R (16.7%) and 7 NR (17.5%). No TRAEs led to surgery cancellation or death. Conclusions: APET-guided approach may help optimize neoadjuvant tx of R-ESCC. nTIS + CT/CRT showed promising efficacy and a tolerable safety profile in both responders and nonresponders. Clinical trial information: NCT04974047 . R NR pCR, a n (%)(95% CI) 6 (30.0) (11.9, 54.3) 11 (34.4)(18.6, 53.2) 1-year DFS, b %(95% CI) 79.0 (47.9, 92.7) 74.2(53.3, 86.8) 1-year EFS, c %(95% CI) 87.1 (64.3, 95.8) 67.8 (48.3, 81.2) R0 resection, a n (%) 19 (95.0) 29 (90.6) ORR, d n (%) 15 (71.4) 14 (42.4) a Efficacy analysis set (EAS) R=20; NR=32. b EAS with R0 resection R=19; NR=29. c Safety analysis set (SAS) R=30; NR=40. d SAS with measurable disease at BL R=21; NR=33.