pCR shows poor surrogacy for OS in neoadjuvant GEA trials overall, but strong for esophageal adenocarcinoma

This systematic review and meta-analysis, following PRISMA and ReSEEM guidelines, evaluated whether pathologic complete response (pCR) is a valid surrogate endpoint for overall survival (OS) in randomized controlled trials of neoadjuvant therapy for resectable distal esophageal, gastroesophageal junction (GEJ), or gastric adenocarcinoma. The analysis included 26 RCTs comprising 7452 patients. The study assessed both individual-level surrogacy (correlating pCR rates with median OS across trial arms) and trial-level surrogacy (correlating treatment effects on pCR, measured by odds ratios, with treatment effects on OS, measured by hazard ratios) using weighted linear regression and calculating the coefficient of determination (R²). Results showed that, overall, pCR demonstrated no meaningful correlation with OS at the individual level (R² = 0.006; p = 0.210). At the trial level, treatment effects on pCR explained very little of the variability in treatment effects on OS (R² = 0.060; p = 0.206). Most trials evaluated cytotoxic chemotherapy or chemoradiotherapy. The analysis revealed significant heterogeneity in surrogacy across subgroups. Surrogacy was strong for trials in esophageal adenocarcinoma (R² = 0.886; p < 0.001), moderate for gastric cancer trials (R² = 0.386; p = 0.041), and absent for trials focusing on GEJ cancer. Furthermore, trials that incorporated radiotherapy demonstrated strong surrogacy, whereas chemotherapy-only trials did not. The authors conclude that interpreting pCR results in gastroesophageal adenocarcinoma trials requires caution and that the validity of pCR as a surrogate might need to be reassessed in the context of upcoming immunotherapy trials.

For people with stomach or esophageal cancer, a major goal of treatment before surgery is to shrink or even eliminate the tumor—what doctors call a 'pathologic complete response' or pCR. It's a hopeful sign, and for years, clinical trials have used it as a shortcut to predict whether a new treatment will help patients live longer. But a new, large review of 26 randomized trials involving over 7,400 patients suggests we should be very careful with that hope. The analysis found that, overall, whether a patient achieved a pCR had almost no meaningful connection to how long they ultimately lived. At the trial level, when a treatment improved pCR rates, that improvement explained very little about whether it also improved overall survival. The story gets more complicated when you look at specific cancers. For cancers starting in the esophagus, a complete response was a very strong predictor of longer life. For stomach cancers, the link was moderate. But for cancers at the junction of the stomach and esophagus, there was no link at all. The type of treatment also mattered: trials that included radiation showed a strong link, while chemotherapy-only trials did not. The bottom line: a tumor disappearing before surgery is a positive event, but it's not a reliable guarantee of survival for most patients with these cancers, and doctors need to interpret trial results with caution.