Accuracy of detecting residual disease after neoadjuvant chemoradiotherapy for esophageal squamous cell carcinoma (preSINO trial): A prospective multicenter diagnostic cohort study in Asia.

196 Background: The benefit of standard esophagectomy in patients with clinical complete response (cCR) after neoadjuvant chemoradiotherapy (nCRT) is disputable and the alternative of active surveillance should be studied. PreSANO study demonstrated that clinical response evaluations (CREs) are accurate to detect residual tumor after nCRT in patients with mainly adenocarcinoma, however, its applicability to esophageal squamous cell carcinoma (ESCC) is unknown. The preSINO trial (NCT03937362) aimed to assess the accuracy of CREs based on bite-on-bite biopsy, endoscopic ultrasonography with fine needle aspiration (EUS-FNA) of suspected lymph nodes, PET-CT in patients with ESCC, and to provide a basis for organ-sparing strategy of ESCC in Asia. Methods: Patients were eligible when they were planned for nCRT (CROSS regimen) followed by standard surgery. Patients received the first CRE (bite-on-bite biopsies) 4-6 weeks after completion of nCRT. Patients with locoregional residual tumor and without distant metastases underwent immediate surgery. In patients with a cCR at CRE-1, a second CRE (CRE-2) was done 10-12 weeks after completion of nCRT using PET-CT, bite-on-bite biopsies and EUS-FNA of suspicious lymph nodes. All patients underwent surgery irrespective of the outcome of CRE-2, in the absence of distant metastases. Circulating tumor DNA (ctDNA) analyses based on tumor-informed assay were performed at baseline and CREs for exploratory analysis. Primary endpoint was the accuracy of CREs for detecting TRG3-4 or TRG1-2 with ypN+ residual tumor. A false-negative rate (FNR) of 19.5% was considered acceptable according to the study protocol. Results: From Aug 20, 2019, to Jan 15, 2023, 309 patients were included of whom 250 patients underwent nCRT plus surgery. Eighteen of 133 patients with TRG3-4 or TRG1-2 with ypN+ residual tumor were not detected by bite-on-bite biopsies and EUS-FNA (FNR: 13.5%). Sensitivity, specificity, negative predictive value and positive predictive value of detecting any residual tumor were 82%, 93%, 69% and 97%, respectively. PET-CT detected interval distant metastases in 13 (5%) of 268 patients prior to surgery. Postoperative distant recurrence rate among patients with ctDNA-positive and ctDNA-negative during CREs were 15.1% (11/73) and 3.3% (2/59), respectively. Conclusions: Bite-on-bite biopsies and EUS-FNA for lymph nodes were accurate for detecting locoregional residual disease after nCRT in patients with ESCC. Post-nCRT ctDNA-positive during CREs may indicate an increased risk of long-term distant metastasis, potentially serving as a diagnostic tool to identify patients who would benefit from postponement of surgery and additional systemic therapy. The long-term follow-up results of this trial will further answer this question. Clinical trial information: NCT03937362 .