Association of Adjuvant Immunotherapy with Improved Survival for Stage II-III Esophageal Cancer: Three-Year National Perspective

S. Sakowitz¹, S. Bakhtiyar², S. Mallick³, T. Coaston⁴, Y. Sanaiha⁵, P. Benharash⁶, J. Yanagawa^7
1UCLA David Geffen School of Medicine, Los Angeles, California ²University of Colorado, Aurora, Colorado ³UCLA, Los Angeles, California ⁴David Geffen School of Medicine at University of California-Los Angeles, Los Angeles, California ⁵University of California, Los Angeles, Los Angeles, California ⁶UCLA Division of Cardiac Surgery, Los Angeles, California ⁷UCLA, David Geffen School of Medicine, Los Angeles, California

Purpose: Among esophageal cancer patients who are at high risk for recurrence after neoadjuvant chemoradiotherapy and esophagectomy, recent trials have demonstrated adjuvant immunotherapy to confer survival benefit at 1-year. Evaluating a national cohort, we sought to evaluate trends in utilization of adjuvant immunotherapy, and elucidate survival outcomes beyond the 1-year mark.

Methods: This was a retrospective analysis of all patients diagnosed with histologically-proven adenocarcinoma or squamous-cell carcinoma esophageal cancer within the 2018-2021 National Cancer Database. Only patients with clinical stage II-III disease, who received neoadjuvant chemoradiotherapy followed by esophagectomy with a complete (R0) resection, were considered. We then limited the cohort to those who were subsequently found to have residual pathologic disease, with a Tumor and Node classification of at least ypT1 or ypN1. Among the final cohort, patients were stratified by receipt of adjuvant immunotherapy (Immunotherapy) versus no immunotherapy (No-Immunotherapy).

The significance of temporal trends was considered using Cuzick’s test. We evaluated survival at 1- and 3-years using Kaplan-Meier time-to-event analyses and Cox proportional hazard models, adjusting for patient, disease, and hospital factors. A sensitivity analysis applied entropy balancing to further adjust for potential confounding variables and ensure covariate balance. As a subgroup analysis, we repeated our analysis stratifying by stage.

Results: Of 4,331 patients meeting criteria, 801 (18%) received adjuvant immunotherapy. The proportion of patients treated with adjuvant immunotherapy increased from 3% in 2018 to 47% in 2021 (P < 0.001; Figure1A).

On average, and relative to No-Immunotherapy, the Immunotherapy group was of comparable age, sex, race, and comorbidity burden. Further, the two groups similarly often were treated at academic centers. While Immunotherapy and Non-Immunotherapy were of comparable stage, Immunotherapy faced larger tumor sizes (4.3[3.0-6.0] vs 4.0cm[2.5-5.5], P< 0.001; Table).

The median follow-up was 19 months for Immunotherapy (Interquartile Range[IQR] 13-24 months), and 22 months for No-Immunotherapy (IQR 11-33 months). Following comprehensive risk-adjustment, receipt of adjuvant immunotherapy was linked with significantly reduced mortality at 1-year (Hazard Ratio[HR] 0.53, 95%Confidence Interval[CI] 0.38-0.74). Adjuvant immunotherapy continued to confer a survival advantage at 3-years (HR 0.72, CI 0.57-0.90; Figure1B).

Upon sensitivity analysis, receipt of adjuvant immunotherapy remained associated with reduced mortality at 1- (HR 0.53, CI 0.39-0.75) and 3-years (HR 0.72, CI 0.58-0.91). Stratifying by stage, receipt of adjuvant immunotherapy was linked with improved survival for Stage III patients (1-Year HR 0.54, CI 0.39-0.75; 3-Year HR 0.67, CI 0.52-0.86), but comparable outcomes for Stage II (1-Year HR 0.85, CI 0.42-1.69; 3-Year HR 1.02, CI 0.61-1.71).

Conclusion: Among patients who underwent esophagectomy following neoadjuvant chemoradiotherapy at high risk for recurrent esophageal disease, receipt of adjuvant immunotherapy was significantly linked with improved survival at 1- and 3-years. Our findings underscore the importance of ensuring adjuvant immunotherapy is accessible to all appropriate candidates, particularly those with Stage III disease.

Identify the source of the funding for this research project: None