Hepatocellular carcinoma (HCC) is the second leading cause of cancer-related deaths worldwide [1]. The Barcelona Clinic Liver Cancer (BCLC) staging system has been widely accepted as a guideline for HCC treatment [2]. According to this system, liver resection, liver transplantation (LT), and local ablation are curative treatment modalities that improve survival outcomes in selected patients with good reserve liver function. Among these modalities, LT results in the best survival outcome. However, liver resection remains the first-line treatment option in many centers because of the severe organ shortage and increased waiting time. Over the past decade, the survival outcome of patients after liver resection has improved significantly because of advances in surgical skills and devices as well as postoperative management. However, long-term survival outcomes are unsatisfactory owing to the high recurrence rate after treatment [34567]. Most studies show a high incidence of intrahepatic HCC recurrence [89101112]. Moreover, to date, well-designed comparative studies on the treatment of intrahepatic HCC recurrence are sparse, and it is still unclear which treatment modality will guarantee better survival outcomes in patients with intrahepatic HCC recurrence.

Therefore, this study aimed to investigate the survival outcomes according to the treatment modality in selected patients with small (<2 cm), solitary, intrahepatic recurrent HCCs after primary liver resection that were classified as recurred Barcelona Clinic Liver Cancer stage O (reBCLC-O) and to determine the risk factors associated with survival outcomes in these patients.

This study included all the patients with HCC recurrence who underwent primary hepatic resection at the Seoul National University Hospital and Samsung Medical Center in Korea, between 2005 and 2011. To evaluate survival outcome according to the treatment modality, patients with intrahepatic reBCLC-O were divided into 2 groups; the curative treatment group (treated with re-resection, salvage LT, or radiofrequency ablation [RFA]) and the transarterial chemoembolization (TACE) group.

The reBCLC-O was defined as a small (<2 cm), solitary HCC with a performance status score of 0–1 and Child-Pugh score A or B, regardless of the primary HCC stage and time interval to recurrence.

After primary hepatic resection, all patients were followed up every 3–4 months to check for recurrence by monitoring α-FP levels and using dynamic CT or MRI. Hepatic recurrence was defined as new lesions observed with at least one imaging modality according to the guidelines provided by the European Association for Study of the Liver and Korean Association for Study of the Liver [1314].

Statistical analysis was performed using IBM SPSS Statistics ver. 20 (IBM Corp., Armonk, NY, USA). Categorical data were compared using the Pearson chi-square test or Fisher exact test, as appropriate. Continuous data were compared using the Wilcoxon rank-sum test. A P-value of <0.05 was considered to indicate statistical significance. The overall survival (OS) rate was defined as the interval between the time of recurrence and patient's death. Survival curves were generated using the Kaplan-Meier method and compared using the log-rank test. To determine the risk factors for OS in the entire cohort, a stratified Cox proportional hazard regression model was used.

The propensity score matching (PSM) method was performed for 1:1 matching of the primarily selected 353 patients. The 2 matched groups (i.e., the curative resection group and the TACE group) were compared to examine co-variable balance and determine any significant differences in baseline co-variables. Propensity score values were generated for characteristics at baseline (age, sex, tumor size, and albumin level) and the time of recurrence (time interval to recurrence and α-FP level).

This study was performed in accordance with the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the Institutional Review Boards of Samsung Medical Center (2013-05-013) and Seoul National University Hospital (H1303-061-474). This study does not require patient consent.

Among 917 patients with HCC recurrence after primary hepatic resection, 394 patients with reBCLC-O were selected (Fig. 1). Untreated patients (n = 9), patients who received percutaneous ethanol injection (n = 14), and patients who had positive pathologic margins (n = 18) were excluded. No patient underwent systemic therapy for reBCLC-O. Among the 353 included patients, 150 (42.5%) were in the curative treatment group and 203 (57.5%) in the TACE group. Several characteristics of the patients were significantly different between the 2 groups (Table 1). In the curative treatment group, we observed smaller tumors, lower baseline albumin levels, patients with a time interval of ≤12 months until tumor recurrence, and lower α-FP levels at the time of recurrence (P < 0.05 for all).

After PSM, both groups were well matched (89 patients in each group); the patient characteristics are summarized in Table 1. Of the 89 patients in the curative treatment group, 63 (70.8%) underwent RFA, 19 (21.3%) underwent re-resection, and 7 (7.9%) underwent salvage LT.

In the present study, we demonstrated that curative treatment is the preferred option for patients with small (<2 cm), solitary, recurring intrahepatic HCC carcinoma that was defined as reBCLC-O. There were 2 major findings. One is that curative treatment did result in significantly better survival outcomes than TACE after patient selection using PSM. The second finding was that curative treatment was an independent prognostic factor in multivariate analysis.

Several curative treatments have been proposed for early-stage intrahepatic recurrent HCC over the past decades. Curative treatments consist of 3 treatment modalities: repeated hepatectomy, RFA, and salvage LT. Repeated hepatectomy has provided favorable outcomes in treating recurrent HCC; however, only a minority of patients are eligible for this because most patients have liver cirrhosis [151617]. Therefore, a strategy of primary resection and salvage LT for intrahepatic recurrent HCC has been suggested [1819]. This strategy showed the best long-term survival outcome among the curative treatments; the outcome was comparable to that of primary LT. In several studies, RFA has yielded a survival outcome comparable to that of repeated hepatectomy. This is considered more effective and safer than repeated resection in the aspects of lower severity; when compared to repeated hepatectomy, RFA shows a lower incidence of complication [202122].

The results of the studies mentioned above support our results. One of the limitations seen in these studies was that the indications for each treatment modality were not consistent. To overcome this issue, this study limited the size and number of recurrent tumors, which enabled us to define specific treatment targets. In addition, this study is unique in that the results were derived using PSM to overcome the limitation of the retrospective study design.

Our study demonstrated that a single tumor at an initial stage, longer time interval to recurrence (more than 12 months), normal INR value, and lower serum α-FP level (<200 ng/mL) were correlated with better patient survival.

The causes of tumor recurrence after hepatic resection may vary depending on the timing of the recurrence after initial treatment. In many studies, early recurrence means recurrence within 1 year after surgery caused by micrometastasis around the tumor at the time of surgery, and late recurrence means recurrence of de novo tumor in background cirrhosis [232425].

Among these, the number of tumors and recurrence time interval have been considered as the tumor characteristics that are strongly correlated with the survival outcome [192627]. It is noteworthy that the initial tumor characteristics have a significant impact on patient survival after recurrence. Several studies have reported that α-FP and INR levels are also closely related to the prognosis in the setting of primary resection. The α-FP has value both as a prognostic marker and predicting the response to therapy [28]. INR level is well known and widely used as a component of the model for end-stage liver disease score, which predicts survival expectancy in end-stage liver disease [29]. However, our study demonstrates that the α-FP and INR levels even at the time of recurrence have an impact on the survival outcome.

Therefore, initial tumor characteristics as well as the liver function and α-FP level at the time of recurrence should be considered as impact factors that influence the outcome of treatment in the reBCLC-O patient group.

This study has some limitations. A selection bias was present as the study had a retrospective design. The choice of treatment modality might have been influenced by patient characteristics and preferences owing to the lack of guidelines for treatment of HCC recurrence. Thus, the patients in the 2 groups exhibited some differences in clinicopathological characteristics, which may have acted as confounders that affected the oncologic outcome. Nevertheless, the results of our study are significant because several confounding factors were corrected for using PSM. Therefore, the results of this study could help in deciding a suitable treatment modality for patients with recurrent HCC.

In conclusion, the OS of patients in the curative treatments group was better than that of patients in the TACE treatment group after PSM. Curative treatments for recurred HCC, the nature of the original HCC (number and time interval to recurrence), and the INR and α-FP levels at the time of recurrence were found to be important prognostic factors for patients with reBCLC-O. Therefore, based on our results, considering these prognostic factors, curative treatment is strongly recommended in the patients with reBCLC-O recurrence for better survival. Nevertheless, further prospective randomized studies are warranted to confirm these results.