Pathologic complete response among patients with esophageal cancer receiving neoadjuvant chemotherapy or chemoradiation: A meta-analysis.

Abstract

305 Background: Although neoadjuvant therapy improves overall survival in esophageal cancers, the benefit is largely confined to those with the greatest pathologic response. Current estimates of pathologic complete response (pCR) rates come from individual studies and lack precision. With this study, we sought to estimate summary pCR rates in patients with non-metastatic esophageal cancer who received either neoadjuvant chemotherapy or chemoradiation by conducting a systematic review and meta-analysis of clinical trials. Methods: Studies were identified from the Medline, EMBASE, and CENTRAL database searches. Eligible studies were published in English from 1992-2022 and consisted of clinical trials that focused on incident non-metastatic esophageal cancer or cancer of the gastroesophageal junction. Trials were required to contain at least one arm of patients receiving neoadjuvant chemotherapy (including perioperative) or neoadjuvant chemoradiation (including treatment with induction chemotherapy). Eligible studies had to report pCR in at least 20 patients who underwent neoadjuvant therapy and surgery. Pooled pCR prevalence was determined using the Freeman-Tukey double arcsine transformation and a random effects model. Results: Among 6,575 records first identified, we included 75 studies with 5,988 patients in the meta-analysis. Of the included studies, 38 (51%) were single arm trials, 35 (47%) were randomized trials, and 2 (3%) were non-randomized trials of multiple treatments. Geographically, 28 (37%), 24 (32%), 21 (28%), and 2 (3%) of the included trials were performed in Asia, Europe, North America, and Australia, respectively. Across 91 trial arms that delivered either neoadjuvant chemotherapy or chemoradiation, platinum and fluorouracil-based regimens were the most common (48.4%), followed by platinum and taxane-based regimens (26.4%), regimens that contained platinum-based agents, fluorouracil, and a taxane (13.2%), and other regimens (12.1%). The pooled prevalence of pCR after neoadjuvant chemotherapy in squamous cell carcinomas was 9% (95% CI: 6% – 14%) across 16 studies, ranging from 0%-32%. The pooled prevalence of pCR after neoadjuvant chemoradiation in squamous cell carcinomas was 32% (95% CI: 26% – 39%) across 21 studies, ranging from 8% to 66%. For adenocarcinomas, the pooled prevalence of pCR was 2% (95% CI: 0% – 7%) after neoadjuvant chemotherapy, across 3 studies, and 22% (17% - 27%) after neoadjuvant chemoradiation across 11 studies. Conclusions: In this meta-analysis, we found that under one-third of patients with non-metastatic esophageal cancer who receive neoadjuvant chemo(radiation) experience pCR. As pCR represents an increasingly utilized as an endpoint in neoadjuvant trials, these estimates of pooled pCR rates may serve as an important benchmark for future trial design.