Performance validation of the ALPPS risk model.

Area of Science:

• Hepatobiliary Surgery
• Surgical Oncology
• Clinical Risk Stratification

Background:

• Two risk models for predicting 90-day mortality in Associating Liver Partition and Portal Vein Ligation for Staged Hepatectomy (ALPPS) were previously proposed.
• Validation of these predictive models was pending.

Purpose of the Study:

• To validate two proposed risk models for predicting 90-day mortality in patients undergoing ALPPS.
• To assess the performance of these models in an independent validation cohort.

Main Methods:

• A validation cohort (VC) comprising 258 patients from the ALPPS registry and external centers was assembled.
• The VC included 70 patients from centers outside the original ALPPS registry.
• The predictive performance of the pre-stage 1 and pre-stage 2 ALPPS risk models was evaluated using c-statistics and Brier scores, comparing VC results to the development cohort (DC).

Main Results:

• The VC confirmed the predictive capability of both models, with acceptable performance for pre-stage 1 (c-statistic 0.64) and good performance for pre-stage 2 (c-statistic 0.77).
• Model performance in the VC was comparable to the DC, with similar Brier scores for both pre-stage 1 (0.089 vs. 0.081) and pre-stage 2 (0.079 vs. 0.087) models.
• The validation cohort included 32 early mortalities (12%).

Conclusions:

• The ALPPS risk score models are now fully validated for predicting individual patient risk.
• These validated models can assist in clinical decision-making to mitigate procedure-related early mortality following ALPPS.