Definition of early recurrence of hepatocellular carcinoma based on the concordance index and optimal treatment strategy

Abstract

Background

Hepatocellular carcinoma (HCC) has a high rate of recurrence and requires multiple treatment options. Early recurrence has been associated with poor prognosis; however, the timing of early recurrence varies in the literature. Therefore, in this study, we aimed to determine the timing of early recurrence with the best prognostic value and clarify the appropriate treatment strategy.

Methods

This retrospective study conducted to determine prognosis and treatments included 274 patients with HCC who underwent liver resection at the Saitama Cancer Center. The prognostic value of early recurrence at each surveillance period was calculated using Harrell's concordance index (C-index).

Results

The C-index at 9 months postoperatively was 0.656, which was the highest cut-off value for early recurrence. Notably, 122 patients had no recurrence, 58 had early recurrence (<9 months), and 94 had late recurrence (≥9 months). The 5-year overall survival (OS) of patients with early recurrence was worse (16.5 %, p<0.001), whereas that of those with late and no recurrences were similar (79.2 % vs. 70.9 %, p=0.86). The 5-year OS rate of patients with late intrahepatic recurrence in the surgery group was 100 %.

Conclusion

The optimal cutoff value for the timing of early recurrence was 9 months postoperatively. The early recurrence group had a high rate of extrahepatic recurrence and poor OS, whereas late recurrences were mostly intrahepatic, with no difference in prognosis between the late and no recurrence groups. Patients with intrahepatic recurrence after >9 months are considered for rehepatectomy because of its good outcomes.

Synopsis

The C-index-based optimal cutoff for the timing of early HCC recurrence was 9 months postoperatively. The early recurrence group had a high rate of extrahepatic recurrence and poor OS, whereas late recurrences were mostly intrahepatic with a good prognosis.