Recalibrating survival prediction among patients receiving trans‐arterial chemoembolization for hepatocellular carcinoma

Background & Aims The Pre-TACE-Predict model was devised to assess prognosis of patients treated with trans-arterial chemoembolization (TACE) for hepatocellular carcinoma (HCC). However, before entering clinical practice, a model should demonstrate that it performs a useful role. Methods We performed an independent external validation of the Pre-TACE model in a cohort that differs in setting and time period from the one that generated the original model. Data from 826 patients treated with TACE for naïve HCC (2008-2018) were used to assess calibration and discrimination of the Pre-TACE-Predict model. Results The four risk-categories identified by the Pre-TACE-Predict model had gradient monotonicity, with median survivals of 52.0, 36.2, 29.9, and 14.1 months respectively. However, predicted survivals systematically underestimated observed survivals (R2: 0.667). A recalibration was adopted maintaining fixed the prognostic index and modifying the baseline survival function. This resulted in an almost perfect calibration (R2: 0.995) in all the four risk categories. Cox regressions showed that aetiology and macrovascular invasion, included in the Pre-TACE-Predict model, had no prognostic impact in the present study population, and that coefficients for tumour size and multiplicity were overestimated. The c-index was similar to that of the m-HAP-III, but higher than those of HAP, m-HAP-II and the six-and-twelve models. Conclusions The recalibration of Pre-TACE-Predict model improved the estimation of survival probabilities of HCC patients treated with TACE. The highest discriminatory ability of the Pre-TACE-model in comparison to other available models, together with risk stratification and recalibration, makes it the best prognostic tool currently available for these patients.