Validation d'un score pré-opératoire prédisant la survie à long terme du cholangiocarcinome intrahépatique réséqué: étude de cohorte multicentrique internationale

PURPOSE
To predict outcome of patients with intrahepatic cholangiocarcinoma (ICC) treated by liver surgery (LS), a preoperative risk score (PRS) could be clinically relevant.
External validation on independent datasets is crucial for evaluating accuracy and generalizability of these models.
The objective of this study was to externally validate the PRS developed by Sasaki et al. on 250 patients for prediction of long-term outcomes after LS for ICC, and based on preoperative albumin, neutrophil-on-lymphocites-ratio, CA19-9 and tumor size.
METHODS
Patients treated by LS for ICC from 10 international high-volume HPB centers from 2001 to 2018 were included in the external validation cohort. Harrell’s c-index and a calibration plot were used to test discrimination and calibration performance of PRS. Kaplan–Meier curve for risk groups as described in the original study were displayed.
RESULTS
A total of 355 patients with 174 deaths during the follow-up period (median=41.7 months, IQR 32.8-50.6) were included in this external-validation cohort. The median PRS value was 14.7 (IQR 10.7 – 20.6), with normal distribution across the cohort. A Cox regression on PRS covariates found coefficients similar to those of the derivation cohort, except for tumor size. Measures of discrimination estimated by Harrell’s c-index was 0.61(95%CI:0.56-0.67). The Kaplan-Meyer estimation showed reasonable discrimination across risk groups, with 5y survival rate ranging from 20.1% to 0%.
CONCLUSION
In this external validation cohort, the PRS has mild discrimination and poor calibration performance, similarly to the original publication. Nevertheless, given its clinical usefulness, a weak model is better than no model at all.