Development and validation of a prognostic nomogram for predicting the progression risk of HCC after lenvatinib therapy.

Although lenvatinib presents promising results in advanced hepatocellular carcinoma (HCC), the treatment responses have distinct individual variability. To address this, we aimed to construct a prognostic model to forecast the risk of progression in HCC patients who underwent lenvatinib therapy. Accordingly, two hundred twenty-three HCC patients who received lenvatinib treatment at the First Affiliated Hospital of Fujian Medical University were enrolled. Statistically significant parameters were identified by univariate analysis and multivariate Cox regression analysis. Subsequently, receiver operating curves (ROC) and calibration curves were plotted to estimate the predictive accuracy and discriminative ability of the model. Decision curve analysis (DCA) was carried out to assess the clinical utility of the nomogram by quantifying the net benefits under all threshold probabilities, and the bootstrap re-sampling method was chosen for the internal validation. Finally, tumor number, tumor size, metastasis, alpha-fetoprotein (AFP), protein induced by vitamin K absence or antagonist II (PIVKA-II) and Child-Pugh grade were finally included in the nomogram to predict the 6-, 12- and 18- months progression-free survival (PFS) rates of lenvatinib-treated HCC patients. An unadjusted C-index of 0.725 and a bootstrap-corrected C-index of 0.689 indicated good prediction accuracy of the model. The AUC for 6-, 12- and 18-month PFS rates were 0.663, 0.677 and 0.810 repetitively. The calibration curve showed that there was a good agreement between predicted progression probability and actual observation one, and DCA indicated a favorable clinical benefit of the nomogram. In summary, Aa practical and well-calibrated model was developed, which could objectively predict the prognosis of lenvatinib treatment in HCC patients.