A prediction model of major complications after radiofrequency ablation for recurrent hepatocellular carcinoma patients

Background: Postoperative major complications are potentially fatal to recurrent hepatocellular carcinoma (RHCC) patients. We aimed to construct a prediction model of major complications after RFA for RHCC patients.

Methods: We retrospectively reviewed the medical records of 407 RHCC patients who underwent RFA as second treatment. Patients were divided into two groups according to the date of RFA: training cohort (277 patients treated in 2010-2016) and validation cohort (130 patients treated in 2017-2019). 23 clinicopathological variables were recorded and analyzed. The logistic regression model was used to build a prediction model.

Result: Major complications developed in 3.6 % of RHCC patients after RFA. In the multivariate analysis, tumor adjacent vessels (p = 0.004) and hepatitis C (p = 0.022) were associated with postoperative complications. The prediction model was described as follow: Risk score (major complication) = 5.180 + 3.391*tumor location+3.389*hepatic etiology, the Youden index was 0.642, the best cut-off value of the model was 8.57 (sensitivity, 78.57 %, specificity, 84.03 %). The area under the receiver operating characteristic curve of the predictive model was 0.85 (95 % CI, 0.82 to 0.88). The validation of the model demonstrated acceptable results, the sensitivity was 80.00 %, specificity was 98.40 %.

Conclusions: This study developed a simple and reliable prediction model of postoperative major complications after RFA for RHCC patients.