# Residual disease subtyping predicts survival and guides adjuvant immunotherapy in esophageal squamous cell carcinoma after neoadjuvant chemoimmunotherapy

## TL;DR

This study introduces a new way to classify esophageal cancer patients who don't fully respond to initial treatment, helping to predict survival and decide on further immunotherapy.

## Contribution

A novel pathological subtyping system for non-pathological complete response ESCC patients to guide adjuvant immunotherapy.

## Key findings

- The T+N+ subtype had the worst survival outcomes, while T0N+ had similar outcomes to those with complete response.
- Adjuvant immunotherapy significantly improved survival only in the T+N0 subtype.
- Prognostic nomograms with high accuracy were developed for individualized risk assessment.

## Abstract

The majority of locally advanced esophageal squamous cell carcinoma (ESCC) patients do not achieve a pathological complete response (NPCR) after neoadjuvant chemoimmunotherapy (NCIT), and their prognosis exhibits significant heterogeneity. This study aimed to establish a pathological subtyping system for NPCR patients to guide precision adjuvant therapy.

We conducted a retrospective analysis of 243 patients with locally advanced ESCC who underwent NCIT followed by esophagectomy. NPCR patients were categorized into three pathological subtypes based on the anatomical sites of residual disease: T+N+ (residual tumor in both primary site and lymph nodes), T+N0 (residual tumor confined to primary site only), and T0N+ (residual disease in lymph nodes only). Survival outcomes were compared, and the efficacy of adjuvant immunotherapy was evaluated within each subtype. Prognostic nomograms for disease-free survival (DFS) and overall survival (OS) were constructed and validated.

Among 176 NPCR patients, the novel subtyping system achieved significant prognostic stratification. The T+N+ subtype demonstrated the poorest survival outcomes, while the T0N+ subtype did not significantly differ from the PCR cohort. Notably, adjuvant immunotherapy provided significant survival benefits exclusively in the T+N0 subtype (DFS: HR = 3.45, 95% CI: 1.17-10.17, P = 0.025; OS: HR = 4.17, 95% CI: 1.07-16.23, P = 0.039), with no significant benefits observed in either the T+N+ or T0N+ subtypes. Based on these findings, we developed and internally validated prognostic nomograms integrating pathological subtype, ypTNM stage, and other key clinicopathological variables, which demonstrated good predictive accuracy (C-index >0.75) for individualized risk assessment.

We propose a novel and practical pathological subtyping framework for NPCR ESCC patients that effectively resolves prognostic heterogeneity and identifies, the T+N0 subtype as the primary beneficiary of adjuvant immunotherapy. The developed nomograms provide a user-friendly tool to facilitate personalized postoperative management and adjuvant therapy decisions.