Treatment Allocation and Survival Outcomes in Potentially Resectable Colorectal Liver Metastases: A Real-World Cohort Study from Thailand

Background/Objectives: Liver resection is a potentially curative treatment for colorectal liver metastases (CRLM), but some potentially resectable patients may not reach specialist hepatobiliary assessment. This study evaluated treatment allocation, survival outcomes, and the potential gap between resectability and actual liver resection in a real-world Thai cohort. Methods: We retrospectively reviewed 265 patients with CRLM treated at a tertiary-care hospital in Thailand between 2012 and 2022. Patients were classified into liver resection (LR; n = 57) and no liver resection (NLR; n = 208) groups. NLR patients without extrahepatic metastases were retrospectively reassessed for potential resectability by two hepatobiliary surgeons using predefined anatomical, oncological, and liver functional criteria. Baseline characteristics and overall survival (OS) were compared between LR patients and potentially resectable NLR patients. Cox regression analysis was also performed in the combined resectable cohort. Results: Median OS was 59 months in the LR group and 23 months in the NLR group. In the NLR group, age < 65 years (adjusted HR 0.663, 95% CI 0.458–0.960; p = 0.030) and absence of extrahepatic metastases (adjusted HR 0.637, 95% CI 0.417–0.973; p = 0.037) were independently associated with longer OS. Among 67 NLR patients without extrahepatic metastases, 27 (40.3%; 13.0% of all NLR patients) were considered potentially resectable. All 27 had undergone primary colorectal tumor resection but had not been referred for hepatobiliary surgical assessment. No significant differences were observed in the measured baseline characteristics between these 27 patients and the LR group. Median OS was 29 versus 59 months, respectively (log-rank p = 0.018). In the combined resectable cohort (n = 84), LR remained associated with longer OS after adjustment for timing of liver metastases, number of liver lesions, and bilobar distribution (adjusted HR 0.40, 95% CI 0.20–0.81; p = 0.011). Conclusions: A clinically relevant subgroup of non-resected patients met predefined criteria for potential resectability on retrospective specialist reassessment. These findings should be interpreted as real-world treatment-allocation observations rather than evidence of a causal survival benefit of LR and support early hepatobiliary referral and repeated multidisciplinary assessment of resectability.