Checkpoint inhibitor benefit in perioperative gastro-esophageal cancer: A meta-analysis of phase III trials.

BACKGROUND

The integration of immunotherapy (IO) with perioperative chemotherapy represents an advance in locally advanced, resectable gastroesophageal cancers. However, randomized controlled trials (RCTs) have yielded discordant findings with respect to event-free survival (EFS) and overall survival (OS), particularly when differing chemotherapy backbones and IO agents are employed. Understanding the sources and implications of these discrepancies is essential for optimizing treatment strategies. Here, we sought to compare outcomes between perioperative FLOT-based and cisplatin/fluoropyrimidine based regimens when combined with IO, and to evaluate the consistency of FLOT-only arms across major RCTs in locally advanced gastroesophageal cancers.

METHODS

RCTs investigating the role of perioperative combination chemotherapy with IO in patients with locally advanced gastro-esophageal cancer were included. Kaplan-Meier curves were digitally reconstructed to obtain individual patient data. Survival analyses incorporated testing for the proportional hazards assumption and were supplemented with piecewise and pooled random-effects analyses to address time-dependent effects and between-study heterogeneity.

RESULTS

No significant difference in EFS was observed between the FLOT-durvalumab and FLOT-pembrolizumab arms (HR = 0.907, 95 %-CI: 0.637-1.290, p = 0.586). FLOT-IO regimens showed superior EFS compared to Cis/fluoropyrimidine-IO (HR = 0.790, 95 %-CI: 0.647-0.966, p = 0.021) as well as FLOT-only (HR = 0.732, 95 %-CI: 0.610-0.878, p < 0.001). While EFS curves for FLOT-only arms converged on long-term follow-up, OS curves diverged, with increased heterogeneity across FLOT-only arms apparent beyond 24 months. Notwithstanding, these analyses should be interpreted with caution due to the lack of patient-level covariate adjustment across trials.

CONCLUSION

As we await the mature OS data from MATTERHORN, the addition of IO to perioperative FLOT should be considered the preferred standard-of-care in resectable, locally-advanced gastro-esophageal adenocarcinoma. Our comparative analyses suggest that FLOT remains a favored chemotherapy backbone for perioperative IO, but confirmation from future randomized trials with mature survival data is needed.