Anatomical and Histologic Response to Neoadjuvant Therapy: Recurrence Patterns Among Poor Responders. A Multi-institutional Observational Cohort Study.

Abstract

To characterize the prognostic implications after hepatectomy for colorectal liver metastases (CRLM) by integrating anatomical response to neoadjuvant therapy (NAT), quantified using the Tumor Burden Score ratio (TBSr), with histologic response assessed by Tumor Regression Grade (TRG). Accurate assessment of response to NAT is central to the management of CRLM. While both macroscopic and histologic tools exist, their combined ability to characterize recurrence patterns remains poorly defined. Patients undergoing curative-intent hepatectomy for CRLM after NAT (2004-2022) were identified from an international, multi-institutional database. Macroscopic response was classified using TBSr with data-driven and prespecified cutoffs, while TRG categorized histologic regression as good (1-3) or poor (4-5). Six combined TRG-TBSr patterns were collapsed into 3 risk groups using maximally selected log-rank statistics. Primary outcomes were recurrence-free survival (RFS) and timing of recurrence (early <12 mo; late ≥12 mo); liver-limited recurrence was secondary. Among 483 patients, the median RFS was 15.87 months. Both TRG (aHR: 1.22, 95% CI: 1.12-1.33) and TBSr (aHR: 1.30, 95% CI: 1.12-1.52) were independently associated with worse RFS (both P<0.001). Early recurrence occurred in 36.9% of patients and was independently associated with TRG (OR: 1.34, 95% CI: 1.15-1.57, P<0.001), TBSr (OR: 1.35, 95% CI: 1.03-1.80, P=0.0036), and synchronous disease (OR: 1.85, 95% CI: 1.21-2.87, P=0.005). Intrahepatic recurrence occurred in 25.1% of patients and was predicted by TRG (hazard ratio: 1.36, 95% CI: 1.15-1.61) and TBSr (hazard ratio: 1.59, 95% CI: 1.31-1.93) (both <0.001). The integrated TRG-TBSr risk categories were strongly associated with RFS, early recurrence, and intrahepatic recurrence (all P<0.001). Combined anatomical and histologic response provided a simple, interpretable, and biologically meaningful framework to stratify recurrence risk after hepatectomy for CRLM.