Interim analysis of immunomodulatory low-dose radiotherapy plus chemoimmunotherapy as conversion therapy for locally advanced potentially/borderline resectable esophageal squamous cell carcinoma (ROICE trial): A phase II single-arm study.

412 Background: The optimal treatment strategy for locally advanced potential/borderline resectable esophageal squamous cell carcinoma (LAPBR-ESCC) remains controversial. In this phase II study, we aim to evaluate the safety and efficacy of immunomodulatory low-dose radiotherapy (iRT) combined with chemoimmunotherapy (iCT) as conversion therapy for LAPBR-ESCC. Methods: This prospective study plans to enroll 43 patients (pts) with LAPBR-ESCC. Treatment begins with iRT (2 Gy × 3 fractions or 3 Gy × 2 fractions) combined with chemotherapy (nab-paclitaxel plus carboplatin) and tislelizumab. If the tumor becomes resectable, surgery is performed; if still unresectable, definitive chemoradiotherapy (dCRT) is administered. The primary endpoint is R0 resection rate. Secondary endpoints include safety, conversion rate, objective response rate (ORR), pathological complete response (pCR), major pathologic response (MPR), 2-year event-free survival (EFS), and overall survival (OS). Exploratory endpoints involve efficacy-related biomarkers and predictive models. Results: From April 2024 to May 2025, 25 pts were enrolled. The cohort consisted of 21 males and 4 females, with a median age of 66 years (range: 53–77). Among them, 20 pts (80.0%) were initially diagnosed with stage IVA disease, and 3 pts (12.0%) had stage IVB disease (limited to supraclavicular lymph node metastasis). As of May 9, 2025, all 25 pts were evaluable for efficacy: 5 achieved complete response (CR), 16 achieved partial response (PR), and 4 had stable disease (SD), resulting in an ORR of 84.0% (21/25). Surgery was performed on 23 pts, with R0 resection achieved in 21 pts. The R0 resection rate was 84.0% in the intention-to-treat (ITT) population and 91.3% in the surgical cohort (21/25 and 21/23, respectively), exceeding the prespecified threshold of 65%. Among the 23 resected cases, pCR was observed in 5 pts (21.7%), and MPR was achieved in 10 pts (43.5%). Primary lesion pCR and MPR were observed in 6 (26.1%) and 15 pts (65.2%), respectively. All pts (25/25) experienced at least one treatment-related adverse event (TRAE), and 17 pts (68.0%) had at least one grade 3/4 TRAE. The most common TRAEs were leukopenia (92.0%), lymphopenia (88.0%), neutropenia (88.0%), alopecia (88.0%), and anemia (84.0%). The most common grade ≥3 TRAEs were leukopenia and neutropenia, each occurring in 40.0% of pts (10/25). No grade 5 events or perioperative deaths were observed. Conclusions: Interim results from this phase II trial demonstrate that iRT combined with iCT followed by surgery is feasible and safe for patients with LAPBR-ESCC. Clinical trial information: ChiCTR2400083999 .