Study population and patient characteristics
From September 2009 to September 2018, a total of 2447 patients received TACE-RFA or TACE alone, and 2048 patients were excluded because they did not meet the study requirements, as shown in Figure 1. Finally, a total of 399 patients were included in this study, 128 were treated with combination therapy and 271 were treated with TACE alone. The detailed clinical characteristics of the 399 patients are summarized in Table 1.
The median follow-up period was 38.1 months (range, 5.7-110.5 months) in the TACE-RFA group and 27.8 months (range,14.4-103.9 months) in the TACE group. In the TACE group, 171 patients died during the observation period, whereas in the TACE-RFA group, only 42 patients died.
The objective tumor regression rate of patients in the TACE-RFA group was 85.9%, and that in the TACE group was 44.7%, which was statistically significant between the two groups. In addition, the disease control rate in the TACE-RFA group was 91.4%, and that in the TACE group alone was 72.0%. Hence, compared with the TACE group, the TACE-RFA group had better tumor response.
In the TACE group, three patients had severe complications, with an incidence of 1.1%. One patient died 3 days after TACE due to acute liver and kidney failure, and two patients developed biloma after TACE. Two cases of severe complications in TACE-RFA group, with an incidence of 1.6%. Subcapsular hematoma of the liver was found on CT scan in 1 patient after multi-hook probe puncture, one patient had bile duct injury during RFA. There was no significant difference in the incidence of major complications between the two groups (P=0.66).
Median OS was 59 months in the TACE-RFA group and 16 months in the TACE group(P<0.001) (Fig 2A). In the TACE-RFA group, the 1-, 3-, 5- and 8-year survival rates were 90.6%, 76.6%, 68.0%, 68.0%. In the TACE group, the 1-, 3-, 5- and 8-year survival rates were 64.5%, 15.1%, 10.8%, 10.8%. Univariable analyses showed that mean tumor size, total bilirubin, AFP>400 ng/mL, BCLC B and therapy method (TACE-RFA) were related to OS (Table 2). Then, through multivariable analysis (Table 3), we found that AFP>400 ng/mL was an independent risk factor for OS and TACE-RFA combination therapy was significantly in connection with better OS.
Median PFS was 45 months in the TACE-RFA group and 4 months in the TACE group (Fig 2B). The cumulative PFS rates of 1, 3, 5, and 8 years in the TACE-RFA group were significantly higher than that in the TACE group. Univariable analyses indicated that mean tumor size, AFP>400 ng/mL, BCLC B and TACE-RFA was related to PFS (Table 2). Multivariable analysis revealed that albumin, hepatitis B and TACE-RFA were associated with PFS (Table 3).
Subgroup Analysis by Tumor Size
In the subgroup analysis, for HCC patients with tumor diameter less than 3cm, the cumulative OS rates (Fig 3A) and cumulative PFS rates (Fig 4A) at 1, 3, 5,8 years were better in patients treated with TACE-RFA than those treated with TACE alone. For HCC patients with tumor diameter of 3-5cm, there were no difference in 1-year cumulative OS rates between the two groups (Fig 3B). The cumulative PFS rates of 1,3,5,8 years were different between the two groups (Fig 4B).
Similarly, the 1,3,5,8 years cumulative OS rates (Fig 3C) and PFS rates (Fig 4C) of HCC patients in the TACE-RFA group with tumor diameter of 5-10cm were significantly better than those in the TACE group. And it was the same with patients with tumor diameter greater than 10cm (OS: Fig 3D; PFS: Fig 4D).
Death risk curve
As the tumor size increases, the death risk curve of the TACE group increased significantly faster than the TACE-RFA group, and the difference in risk score between them also increased (Fig 5). In other words, the larger the tumor size, the greater the reduction in death risk in the TACE-RFA group.