Sex-based differences in survival after liver transplantation for colorectal cancer liver metastases: A multivariable analysis.

BACKGROUND & AIMS: Liver transplantation (LT) for colorectal liver metastases (CRLMs) is attracting increasing interest, especially after publication of the TransMet trial. However, multivariable survival analyses are lacking. Here, we performed such an analysis in a multicentre cohort.

METHODS

We conducted a retrospective multicentre study of 82 patients with CRLMs undergoing LT (from 2006 to 2020) across seven US and European centres, using multivariable Cox, competing-risk models, and extensive sensitivity analyses.

RESULTS

Overall survival rates after 1, 3, and 5 years were 93.7%, 73.4%, and 54.9%, respectively. The findings align with an association between higher risk and the female sex (estimated hazard ratio (HR) 4.1, 95% CI: 1.8-9.2), and the following variables: carcinoembryonic antigen >80 μg/L, right-located colorectal cancer (CRC), largest diameter >5.5 cm, KRAS mutation, and absence of previous liver therapy. Other possible associations with higher uncertainty were pN2-positive CRC and the number of nodules (>10). Variables such as progressive disease after pretransplant chemotherapy and time from primary CRC surgery to LT of ≤24 months, exhibited weaker, less consistent associations.

CONCLUSIONS

This first multivariable survival analysis of LT for CRLM suggests that female sex is associated with worse outcomes, whereas the prognostic strength of the model currently used in clinical practice is not confirmed. Our findings challenge current selection criteria, highlighting the need for improved prognostic models with better discrimination and calibration.

IMPACT AND IMPLICATIONS

This multicentre retrospective study analysed survival outcomes in 82 patients undergoing liver transplantation for colorectal liver metastases across seven US and European centres. Several factors, including female sex, high carcinoembryonic antigen levels, right-sided colorectal cancer, larger tumours, KRAS mutation, pN2-positive CRC, number of nodules, and no prior liver therapy, were linked to poorer outcomes. The study questions current prognostic models and selection criteria, emphasizing the need for more accurate tools to guide decision-making in patients with colorectal liver metastases.