Study population, technical success rate, and safety
As described in Fig. 1, 126 patients with HCC in high-risk locations were enrolled in this study, including 70 patients in the TACE-RFA group and 56 patients in the TACE-RFA-125I group, and 46 pairs of patients were matched after the PSM analysis. The baseline characteristics of patients between the two groups were balanced after the PSM analysis, which were shown in Table 1. All of the TACE, RFA, or 125I seed implantation procedures of the 46 pairs of patients were successfully performed, and there was no occurrence of major complications or procedure-related deaths. The median follow-up period was 43.5 months (range, 31–68 months). One representative case of TACE-RFA-125I therapy for HCC in high-risk locations is shown in Fig. 2.
Table 1
The patients’ baseline characteristics between the TACE-RFA and TACE-RFA-125I groups before and after PSM analysis
Variable | Before PSM (Median, IQR; No., %) | p value | After PSM (Median, IQR; No., %) | p value |
| TACE-RFA (n = 70) | TACE-RFA-125I (n = 56) | | TACE-RFA (n = 46) | TACE-RFA-125I (n = 46) | |
Age (years) | 61 (54–69) | 59 (56–65) | 0.443 | 61 (54–72) | 59 (53–68) | 0.356 |
Gender | | 0.200 | | | 0.765 |
Male | 63 (90.0%) | 46 (82.1%) | | 39 (84.8%) | 40 (87.0%) | |
Female | 7 (10.0%) | 10 (17.9%) | | 7 (15.2%) | 6 (13.0%) | |
Child-Pugh class | | 0.026* | | | 0.216 |
A | 45 (64.3%) | 46 (82.1%) | | 42 (91.3%) | 38 (82.6%) | |
B | 25 (35.7%) | 10 (17.9%) | | 4 (8.7%) | 8 (17.4%) | |
BCLC stage | | | 0.936 | | | 0.676 |
A | 32 (45.7%) | 26 (46.4%) | | 23 (50.0%) | 21 (45.7%) | |
B | 38 (54.3%) | 30 (53.6%) | | 23 (50.0%) | 25 (54.3%) | |
Tumor size (cm) | | 0.498 | | | 0.677 |
≤ 3 | 38 (54.3%) | 27 (48.2%) | | 24 (52.2%) | 22 (47.8%) | |
3–5 | 32 (45.7%) | 29 (51.8%) | | 22 (47.8%) | 24 (52.2%) | |
Tumor number | | 0.846 | | | 0.901 |
1 | 50 (71.4%) | 38 (67.9%) | | 35 (76.1%) | 35 (76.1%) | |
2 | 11 (15.7%) | 11 (19.6%) | | 8 (17.4%) | 7 (15.2%) | |
3 | 9 (12.9%) | 7 (12.5%) | | 3 (6.5%) | 4 (8.7%) | |
Ascites | | 0.014* | | | 0.503 |
Absent | 50 (71.4%) | 50 (89.3%) | | 40 (87.0%) | 42 (91.3%) | |
Mild | 20 (28.6%) | 6 (10.7%) | | 6 (13.0%) | 4 (8.7%) | |
HBV infection | | 0.302 | | | > 0.999 |
No | 4 (5.7%) | 6 (10.7%) | | 4 (8.7%) | 4 (8.7%) | |
Yes | 66 (94.3%) | 50 (89.3%) | | 42 (91.3%) | 42 (91.3%) | |
AFP (µg/L) | | 0.151 | | | 0.832 |
≦ 400 | 31 (44.3%) | 32 (57.1%) | | 27 (58.7%) | 28 (60.9%) | |
> 400 | 39 (55.7%) | 24 (42.9%) | | 19 (41.3%) | 18 (39.1%) | |
Abbreviations: TACE, transarterial chemoembolization; RFA, radiofrequency ablation; PSM, propensity score matching; IQR, interquartile range; No., Number; BCLC, Barcelona Clinic Liver Cancer; HBV, hepatitis B virus; AFP, α-fetoprotein |
*p value < 0.05 was considered to indicate statistical significance |
The tumor recurrence rates between the two groups after PSM
The 1-, 2-, 3-, 4-, and 5-years of local recurrence rates in the TACE-RFA-125I group and TACE-RFA group were 0%, 11.1%, 16.9%, 16.9%, 20.1%, and 6.5%, 15.8%, 33.2%, 46.1%, 46.1%, respectively. The overall recurrence rates (including local, intrahepatic distant, and extrahepatic recurrences) at 1-, 2-, 3-, 4-, and 5-years were 2.1%, 24.4%, 38.5%, 53.8%, and 61.5% in the TACE-RFA-125I group and 17.5%, 40%, 66.1%, 81.1%, and 90.9% in the TACE-RFA group, respectively. Both the local and overall recurrence rates in the TACE-RFA-125I group were significantly lower than those in the TACE-RFA group (p = 0.037, p = 0.004).
The local progression-free survival between the two groups after PSM analysis
The 1-, 2-, 3-, 4-, and 5-years LPFS rates in the TACE-RFA-125I group were 100%, 82.4%, 74.8%, 63.5%, and 54%, respectively, which were significantly higher than those of 91.3%, 69.4%, 50.7%, 29.4%, and 26.7% in the TACE-RFA group, respectively (p = 0.004; Fig. 3A). The uni- and multi-variate analyses demonstrated that the TACE-RFA-125I treatment, tumor size≤ 3 cm, a solitary HCC, and BCLC stage A were the protective factors for patients’ LPFS (Table 2).
Table 2
The uni- and multi-variate analyses of LPFS between the TACE-RFA and TACE-RFA-125I groups after PSM analysis
Variables | Univariate analysis | Multivariate analysis |
HR (95% CI) | p value | HR (95% CI) | p value |
Age (years) | 0.99 (0.97–1.02) | 0.710 | | |
Treatment method | | | | |
TACE-RFA-125I | Ref | | Ref | |
TACE-RFA | 2.32 (1.30–4.16) | 0.005* | 3.35 (1.79–6.26) | < .001** |
Gender | | | | |
Female | Ref | | | |
Male | 2.17 (0.78–6.03) | 0.138 | | |
HBV infection | | | | |
No | Ref | | | |
Yes | 0.87 (0.37–2.05) | 0.752 | | |
Tumor size (cm) | | | | |
≤ 3 | Ref | | Ref | |
3–5 | 6.27 (3.24–12.14) | < .001* | 2.56 (1.14–5.73) | 0.022** |
Tumor number | | | | |
1 | Ref | | Ref | |
2–3 | 2.66 (1.48–4.80) | 0.001* | 2.84 (1.49–5.43) | 0.002** |
BCLC stage | | | | |
A | Ref | | Ref | |
B | 6.84 (3.50–13.36) | < .001* | 3.80 (1.69–8.53) | 0.001** |
AFP (µg/L) | | | | |
≤ 400 | Ref | | | |
> 400 | 0.80 (0.45–1.43) | 0.445 | | |
Ascites | | | | |
Absent | Ref | | | |
Mild | 1.50 (0.64–3.54) | 0.349 | | |
Child-Pugh class | | | | |
A | Ref | | Ref | |
B | 3.83 (1.77–8.33) | < .001* | 2.85 (1.15–7.07) | 0.024** |
Abbreviations: LPFS, local progression-free survival; TACE, transarterial chemoembolization; RFA, radiofrequency ablation; PSM, propensity score matching; HR, hazard ratio; CI, confidence interval; Ref, reference; HBV, hepatitis B virus; BCLC, Barcelona Clinic Liver Cancer; AFP, α-fetoprotein |
*p value≤0.1 in uni-variate analysis were included in multi-variate analysis, **p value < 0.05 was considered to indicate statistical significance in multi-variate analysis |
The progression-free survival between the two groups after PSM analysis
The 1-, 2-, 3-, 4-, and 5-years PFS rates in the TACE-RFA-125I group were 97.8%, 69.3%, 53.7%, 38.2%, and 31.8%, respectively, which were significantly higher than those of 84.8%, 54.3%, 29.3%, 16.3%, and 7.2% in the TACE-RFA group, respectively, and the median PFS in the TACE-RFA-125I group was also significantly longer than that of the TACE-RFA group (42.0 months vs. 29.0 months, p = 0.002; Fig. 3B). The uni- and multi-variate analyses indicated that the TACE-RFA-125I treatment, tumor size≤ 3 cm, a solitary HCC, BCLC stage A, and absence of ascites were the protective factors for patients’ PFS (Table 3).
Table 3
The uni- and multi-variate analyses of PFS between the TACE-RFA and TACE-RFA-125I groups after PSM analysis
Variables | Univariate analysis | Multivariate analysis |
HR (95% CI) | p value | HR (95% CI) | p value |
Age (years) | 1.01 (0.98–1.03) | 0.656 | | |
Treatment method | | | | |
TACE-RFA-125I | Ref | | Ref | |
TACE-RFA | 2.07 (1.28–3.34) | 0.003* | 2.96 (1.75–5.01) | < .001** |
Gender | | | | |
Female | Ref | | Ref | |
Male | 2.07 (0.98–4.36) | 0.057* | 1.25 (0.56–2.77) | 0.583 |
HBV infection | | | | |
No | Ref | | | |
Yes | 1.70 (0.72–4.03) | 0.230 | | |
Tumor size (cm) | | | | |
≤ 3 | Ref | | Ref | |
3–5 | 5.34 (2.89–9.89) | < .001* | 2.67 (1.25–5.68) | 0.011** |
Tumor number | | | | |
1 | Ref | | Ref | |
2–3 | 2.39 (1.38–4.16) | 0.002* | 3.98 (2.06–7.71) | < .001** |
BCLC stage | | | | |
A | Ref | | Ref | |
B | 7.44 (3.83–14.43) | < .001* | 5.90 (2.57–13.55) | < .001** |
AFP (µg/L) | | | | |
≤ 400 | Ref | | | |
> 400 | 0.74 (0.46–1.20) | 0.228 | | |
Ascites | | | | |
Absent | Ref | | Ref | |
Mild | 3.72 (1.79–7.74) | < .001* | 2.93 (1.34–6.42) | 0.007** |
Child-Pugh class | | | | |
A | Ref | | Ref | |
B | 1.96 (0.95–4.02) | 0.067* | 1.08 (0.46–2.53) | 0.862 |
Abbreviations: PFS, progression-free survival; TACE, transarterial chemoembolization; RFA, radiofrequency ablation; PSM, propensity score matching; HR, hazard ratio; CI, confidence interval; Ref, reference; HBV, hepatitis B virus; BCLC, Barcelona Clinic Liver Cancer; AFP, α-fetoprotein |
*p value≤0.1 in uni-variate analysis were included in multi-variate analysis, **p value < 0.05 was considered to indicate statistical significance in multi-variate analysis |
The overall survival between the two groups after PSM analysis
At the end of follow-up, 28.3% (13/46) patients in the TACE-RFA-125I group and 56.5% (26/46) patients in the TACE-RFA group died. The 1-, 2-, 3-, 4-, and 5-years OS rates were 100%, 93.4%, 80.7%, 74.9%, and 64.7% in the TACE-RFA-125I group, respectively, which were significantly higher than those of 97.8%, 78%, 68.6%, 51.1%, and 45.3% in the TACE-RFA group, respectively (p = 0.011; Fig. 3C). The uni- and multi-variate analyses showed that the TACE-RFA-125I treatment, tumor size≤ 3 cm, a solitary HCC, BCLC stage A, and Child-Pugh class A were the protective factors for patients’ OS (Table 4).
Table 4
The uni- and multi-variate analyses of OS between the TACE-RFA and TACE-RFA-125I groups after PSM analysis
Variables | Univariate analysis | Multivariate analysis |
HR (95% CI) | p value | HR (95% CI) | p value |
Age (years) | 0.98 (0.95–1.02) | 0.315 | | |
Treatment method | | | | |
TACE-RFA-125I | Ref | | Ref | |
TACE-RFA | 2.34 (1.20–4.56) | 0.013* | 4.60 (2.11–10.00) | < .001** |
Gender | | | | |
Female | Ref | | | |
Male | 2.47 (0.76–8.06) | 0.133 | | |
HBV infection | | | | |
No | Ref | | | |
Yes | 1.21 (0.43–3.43) | 0.713 | | |
Tumor size (cm) | | | | |
≤ 3 | Ref | | Ref | |
3–5 | 11.76 (5.07–27.26) | < .001* | 7.46 (2.68–20.74) | < .001** |
Tumor number | | | | |
1 | Ref | | Ref | |
2–3 | 2.66 (1.38–5.14) | 0.004* | 3.44 (1.66–7.11) | < .001** |
BCLC stage | | | | |
A | Ref | | Ref | |
B | 8.38 (3.71–18.90) | < .001* | 3.40 (1.30–8.89) | 0.013** |
AFP (µg/L) | | | | |
≤ 400 | Ref | | | |
> 400 | 1.23 (0.65–2.33) | 0.532 | | |
Ascites | | | | |
Absent | Ref | | | |
Mild | 1.13 (0.40–3.20) | 0.819 | | |
Child-Pugh class | | | | |
A | Ref | | Ref | |
B | 6.79 (2.83–16.30) | < .001* | 5.84 (2.00–17.00) | 0.001** |
Abbreviations: OS, overall survival; TACE, transarterial chemoembolization; RFA, radiofrequency ablation; PSM, propensity score matching; HR, hazard ratio; CI, confidence interval; Ref, reference; HBV, hepatitis B virus; BCLC, Barcelona Clinic Liver Cancer; AFP, α-fetoprotein |
*p value≤0.1 in uni-variate analysis were included in multi-variate analysis, **p value < 0.05 was considered to indicate statistical significance in multi-variate analysis |
The subgroup analyses by tumor size after PSM
In the subgroup analysis of patients with tumor size≤ 3 cm, the 1-, 2-, 3-, 4-, and 5-years LPFS, PFS, and OS rates in the TACE-RFA-125I group (LPFS rates: 100%, 92.9%, 89.1%, 85.3%, and 72.5%; PFS rates: 100%, 85.7%, 74.7%, 53.1%, and 44.2%; OS rates: 100%, 100%, 96.3%, 92.4%, and 79.8%, respectively) were all significantly better than those in the TACE-RFA group (LPFS rates: 100%, 80%, 69.4%, 44.9%, and 44.9%; PFS rates: 93.3%, 70%, 45.2%, 25.1%, and 11.2%; OS rates: 96.7%, 93.3%, 89.9%, 75%, and 66.5%, respectively) (p = 0.005, p = 0.005, p = 0.040, Fig. 4A-C). Meanwhile, the median PFS in the TACE-RFA-125I group was significantly longer than that in the TACE-RFA group (50.0 months vs. 34.0 months, p = 0.005, Fig. 4B).
As described in Fig. 4D-F, in the subgroup analysis of patients with tumor size > 3 cm and ≤ 5 cm, the 1-, 2-, and 3-years LPFS and OS rates in the TACE-RFA-125I group (LPFS rates: 100%, 82.6%, and 49.1%; OS rates: 100%, 82.6%, and 48.7%, respectively) were significantly higher than those in the TACE-RFA group (LPFS rates: 75%, 48.6%, and 13.9%; OS rates: 93.8%, 48.1%, and 24.8%, respectively) (p = 0.034, p = 0.011, respectively). The 1-, 2-years PFS rates in the TACE-RFA-125I group (PFS rates: 94.4%, 42.4%, respectively) were also significantly higher than those in the TACE-RFA group (PFS rates: 56.3%, 25%, respectively) (p = 0.011). In addition, the median LPFS, PFS, and OS in the TACE-RFA-125I group were significantly longer than that of in the TACE-RFA group (median LPFS: 36.0 months vs. 23.0 months, p = 0.034; median PFS: 23.0 months vs. 17.0 months, p = 0.011; median OS: 36.0 months vs. 23.0 months, p = 0.011).