[Dynamically monitoring circulating tumor DNA as a biomarker for immunotherapy in advanced esophageal squamous cell carcinoma].

Objective: To investigate the clinical value of dynamic monitoring of plasma circulating tumor DNA (ctDNA) in evaluating the efficacy of immunotherapy for esophageal squamous cell carcinoma (ESCC). Methods: Plasma samples were collected at baseline and after every 2-3 treatment cycles from 94 patients with advanced ESCC receiving second-line sintilimab monotherapy in the ORIENT-2 study. Targeted sequencing was performed to analyze somatic variants in ctDNA, and the molecular tumor burden index (mTBI) was calculated to assess dynamic changes in ctDNA. Imaging examinations were conducted synchronously with plasma sample collection, and treatment response was evaluated according to the Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST 1.1). Results: A total of 614 somatic mutations were detected in 93 eligible baseline plasma samples, with a median of 6 mutations per sample. Missense mutations were the most frequent mutation type. The baseline mutational profile revealed frequently mutated genes including TP53 (82%), CDKN2B (23%), and NOTCH1 (22%). In the 68 patients with both baseline and post-2-cycle plasma samples, no significant changes were observed in the variant allele frequencies (VAFs) of core driver genes before and after treatment (all P＞0.05), and no newly emerged core driver gene mutations were identified. CCND1 copy number variation was associated with shorter progression-free survival (PFS) (HR=1.88, 95% CI: 1.08-3.27), while mutations in other genes, including TP53 and NOTCH1, showed no association with PFS (all P＞0.05). None of the frequently mutated genes were associated with overall survival (OS) (all P＞0.05). Among the 68 patients, 11 (15.9%) achieved ctDNA clearance, none of whom showed tumor progression on concurrent imaging evaluation. The remaining 57 patients had ctDNA that became positive or remained persistently positive, of whom 30 (52.6%) showed tumor progression on imaging, with a statistically significant difference between groups (P=0.002). Compared with patients whose ctDNA became or remained positive, those with ctDNA clearance exhibited significantly delayed tumor progression (HR=2.05, 95% CI: 1.06-3.96), but ctDNA status after 2 cycles did not significantly affect OS (HR=1.38, 95% CI: 0.62-3.09). After 2 cycles of treatment, patients in the low mTBI group had a higher disease control rate (DCR) than those in the high mTBI group [73.5% (25/34) vs. 38.2% (13/34), P=0.007], as well as superior PFS (HR=2.80, 95% CI: 1.65-4.75) and OS (HR=3.54, 95% CI: 1.95-6.42). Dynamic changes in mTBI were highly consistent with concurrent imaging response assessments. The molecular response group had a significantly higher DCR than the non-response group [87.5% (21/24) vs. 42.2% (19/45), P＜0.001], as well as superior PFS (HR=2.39, 95% CI: 1.41-4.06) and OS (HR=2.77, 95% CI: 1.46-5.22). Among 32 patients with stable disease (SD) at the first imaging evaluation after 2 cycles, those in the molecular response group had comparable PFS with the non-response group (HR=1.03, 95% CI: 0.51-2.08, P=0.942), but significantly longer OS (HR=3.12, 95% CI: 1.20-8.06). Conclusion: Dynamic monitoring of the ctDNA-based molecular tumor burden index (mTBI) provides real-time and sensitive molecular information for evaluating the efficacy of immunotherapy in advanced ESCC, demonstrating definite clinical value for personalized treatment management.