Incidence and predictors of abdominal pain after transarterial chemoembolization of hepatocellular carcinoma: a single-center retrospective study.

PURPOSE: To assess the incidence and predictive factors of abdominal pain following transarterial hepatic chemoembolization (TACE) in patients with hepatocellular carcinoma.

METHODS

In this single-center retrospective cohort study, abdominal pain was defined as a score of 4 or more within 72 h after TACE and requiring additional drug intervention. Patient, tumor characteristics, and technical factors associated with severe pain were identified using the decision tree and binary logistic regression model.

RESULTS

Of 220 patients who were included in the study, 126 (57.3%) had abdominal pain after 206 of 420 TACE procedures (49.0%). A predictive model built based on the logistic regression identified the drug-eluting bead TACE (DEB-TACE) (odds ratio [OR] = 3.340; 95% confidence interval [CI] 2.169-5.141), the number of tumors (OR = 2.235; 95% CI 1.060-4.713), embolization of both hepatic lobes (OR = 2.310; 95% CI 1.109-4.813), and concomitant extrahepatic artery embolism (OR = 2.654; 95% CI 1.227-5.739) as the independent predictors of severe abdominal pain. Similarly, the decision tree confirmed the DEB-TACE as the strongest predictor of subsequent performance, followed by the history of hepatectomy and the embolization in the right or both lobes. The area under the receiver operating characteristic curve (AUC) of the classification prediction effect of the two models was 0.706 for the logistic regression and 0.676 for the decision tree. Internal validation results show that the accuracy of logistic regression model prediction was 71.4%.

CONCLUSION

The model suggests that DEB-TACE and multiple treatment sites are predictors of abdominal pain after TACE in patients with hepatocellular carcinoma. It may help improve nursing management practices.