Liver resection versus transarterial chemoembolization for the treatment of intermediate-stage hepatocellular carcinoma

Abstract

Background: The role of transarterial chemoembolization (TACE) as the standard treatment for intermediate-stage hepatocellular carcinoma (HCC) is being challenged by increasing studies supporting liver resection (LR); but evidence of survival benefits of LR is lacking. We aimed to compare the overall survival (OS) of LR with that of TACE for the treatment of intermediate-stage HCC in cirrhotic patients.

Methods: A Markov model, comparing LR with TACE over 15 years, was developed based on the data from 31 literatures. Additionally, external validation of the model was performed using a data set (n = 1735; LR: 701; TACE: 1034) from a tertiary center with propensity score matching method. We conducted one-way and two-way sensitivity analyses, in addition to a Monte Carlo analysis with 10 000 patients allocated into each arm.

Results: The mean expected survival times and survival rates at 5 years were 77.8 months and 47.1% in LR group, and 48.6 months and 25.7% in TACE group, respectively. Sensitivity analyses found that initial LR was the most favorable treatment. The 95% CI for the difference in OS was 2.42-2.46 years between the two groups (P < 0.001). In the validation set, the 5-year survival rates after LR were significantly better than those after TACE before (40.2% vs. 25.9%, P < 0.001) and after matching (43.2% vs 30.9%, P < 0.001), which was comparable to the model results.

Conclusions: For cirrhotic patients with resectable intermediate-stage HCC, LR may provide survival benefit over TACE, but large-scale studies are required to further stratify patients at this stage for different optimal treatments.

Keywords: Markov Model; intermediate-stage hepatocellular carcinoma; liver resection; propensity score matching; transarterial chemoembolization.

Figure 1

Flow diagram of the Markov cohort model. Each pane represents a state of health. Straight lines with arrows indicate transition from one state to another one while circular arrows mean that some patients may stay at the same state for more than one cycle. Two therapy strategies were designed in this model with initial treatments of LR and TACE. Patients with progressive HCC in both groups were assumed to receive no further treatments. For TACE group, patients with PR, SD, or recurrence after CR will be considered candidates for repeated TACE except those with progressive disease. For LR group, patients with recurrent HCC after initial LR were assumed to receive repeated TACE treatment. Patients with positive resection margin will be assumed to have progressive HCC. LR, liver resection; TACE, transarterial chemoembolization; HCC, hepatocellular carcinoma; PR, partial response; SD, stable disease; CR, complete response

Figure 2

Overall survival curves for LR and TACE groups in the treatment of compensated cirrhotic patients with intermediate‐stage HCC. The survival curves of the LR group were better than that of the TACE group. LR, liver resection; TACE, transarterial chemoembolization; HCC, hepatocellular carcinoma

Figure 3

Cumulative survival curves of LR group and TACE group for the validation cohort patients with intermediate‐stage HCC before (A) and after (B) matching. Note the significant differences of cumulative survival rate, median survival time and survival proportion at 5‐year between LR group and TACE group. LR, liver resection; TACE, transarterial chemoembolization; HCC, hepatocellular carcinoma