Is female sex a significant favorable prognostic factor in hepatocellular carcinoma?

OBJECTIVE: As sex favorably modulates the natural history of chronic liver diseases and the risk for neoplastic evolution, our study aimed to ascertain whether female hepatocellular carcinoma (HCC) patients are also characterized by better prognosis. METHODS: The ITA.LI.CA (Italian Liver Cancer) database was used, including 1834 HCC patients (482 females, 1352 males) that were consecutively diagnosed. The following variables were considered: age, etiology, modality of diagnosis, earlier interferon treatment, bilirubin, alpha-fetoprotein levels, constitutional syndrome, portal thrombosis, metastasis, number and size of nodules, grading, Child-Pugh class, tumor-nodes-metastases and Cancer of the Liver Italian Program staging, and treatment. RESULTS: Female HCC patients were characterized by older age (P=0.0001), higher prevalence of HCV infection (P=0.0001), diagnosis more frequently by surveillance (P=0.003), higher alpha-fetoprotein levels (P=0.0055), lower prevalence of constitutional syndrome (P=0.03), portal thrombosis (P=0.04), and metastasis (P=0.0001). HCC in females was more frequently unifocal (P=0.0001), smaller (P=0.001), well differentiated (P=0.001), and of lower Cancer of the Liver Italian Program and tumor-nodes-metastases stage (P=0.0001 and 0.0001). However, females underwent curative treatments (transplantation, resection, percutaneous ablation) in the same percentage of cases as males. Finally, females had a significantly longer survival (median 29 95% confidence interval (CI): 24-33 vs. 24 (22-25) months, P=0.0001). The difference was sharper median 36 (CI: 31-41 vs. 17 (CI: 15-19)] when females undergoing surveillance were compared with males diagnosed incidentally or for symptoms. The Cox model also identified sex as an independent predictor of survival. When only patients undergoing surveillance were considered, no significant difference was observed. CONCLUSION: HCC in females has better prognosis, but this is possibly more because of higher compliance with surveillance than to real biological differences.