Department of Medicine and Molecular Science, Division of Frontier Medical Science, Hiroshima University, Hiroshima, Japan,
Hiroshi Aikata, Department of Medicine and Molecular Science, Division of Frontier Medical Science, Programs for Biomedical Research, Graduate School of Biomedical Sciences, Hiroshima University, 1-2-3 Kasumi, Minami-ku, Hiroshima 734-8551, Japan. Email: firstname.lastname@example.org
Background and Aims: We evaluated the prognosis and associated factors in patients with small hepatocellular carcinoma (HCC; up to 3 nodules, each up to 3cm in diameter) treated with percutaneous radiofrequency ablation (RFA) as first-line treatment.
Methods: Eighty-eight consecutive patients who underwent percutaneous RFA as first-line treatment were enrolled, among whom 70 who had hypervascular HCC nodules which were treated by a combination of transcatheter arterial chemoembolization and RFA. RFA was repeated until an ablative margin was obtained.
Results: The rate of local tumor progression at 1 and 3 years was 4.8% and 4.8%, respectively. The rate of overall survival at 3 and 5 years was 83.0% and 70.0%, and the rate of disease-free survival at 3 and 5 years was 34.0% and 24.0%, respectively. On multivariate analysis, age (< 70 years; hazard ratio [HR] = 2.341, 95% confidence interval [CI] = 1.101–4.977, P = 0.027) and indocyanine green retention rate at 15 min (< 15%; HR = 3.621, 95% CI = 1.086–12.079, P = 0.036) were statistically significant determinants of overall survival, while tumor number (solitary, HR = 2.465, 95% CI = 1.170–5.191, P = 0.018) was identified for disease-free survival. Overall survival of patients with early recurrence after RFA was significantly worse than that of patients with late recurrence. Tumor size was the only independent risk factor of early recurrence after RFA of HCC (tumor size > 2 cm; risk ratio [RR] = 4.629, 95% CI = 1.241–17.241, P = 0.023).
Conclusion: Percutaneous RFA under the protocol reported here has the potential to provide local tumor control for small HCC. In addition to host factors, time interval from RFA to recurrence was an important determinant of prognosis.
Hepatocellular carcinoma (HCC) is one of the most common malignancies in the world today. Currently, owing to periodic surveillance of patients with chronic hepatitis and liver cirrhosis, small HCC are occasionally found during imaging examinations.1,2 Although surgical resection is the standard treatment modality for HCC, its use is usually limited because the majority of patients, even with small HCC, have associated severe liver dysfunction.3,4 Various percutaneous ablation therapies are now in clinical use as alternative treatment options for small HCC, including percutaneous ethanol injection (PEI), microwave coagulation therapy (MCT), cryoablation, laser and radiofrequency ablation (RFA).5–10 Although originally restricted to HCC that were unsuitable for resective surgery, percutaneous ablation therapies are now proving valuable in small operable cases as well.11,12
Radiofrequency ablation is currently considered the most effective percutaneous ablation therapy, and some centers now use it as a first-line treatment option, even in patients suitable for surgery.13 Among the various clinical studies of RFA, however, only a few have focused on percutaneous RFA for small HCC as a first-line treatment option.14,15 Moreover, few studies have evaluated the long-term outcome and prognostic factors of percutaneous RFA for small HCC (up to 3 nodules, each up to 3cm in diameter).
Radiofrequency ablation procedures and treatment evaluation after RFA differ according to facility. In our institution, a combination transcatheter arterial chemoembolization (TACE) and RFA was performed in patients with hypervascular HCC nodules. Efficacy was evaluated by dynamic computed tomography (CT) 2–3 days after each treatment session, and RFA sessions were repeated until an ablative margin was obtained.
Using this protocol, we performed percutaneous RFA in 88 consecutive patients with small HCC from February 2001 to September 2007. Here, we retrospectively evaluated the long-term survival and complications with RFA as first-line treatment in patients with small HCC up to 3 nodules, each up to 3cm in diameter, and assessed prognostic factors that affected therapeutic outcomes. In addition, we analyzed the relationship between the time interval from RFA of HCC to recurrence and prognosis.
Between February 2001 and September 2007, 263 consecutive Japanese patients with small HCC were referred to our hospital for treatment. None had been previously treated for HCC, up to 3 nodules, each up to 3cm in diameter, absence of portal venous thrombosis and known extrahepatic metastases, and Child–Pugh class A or B liver cirrhosis. RFA treatment was selected in consideration of the site, size and number of tumors, functional reserve of the liver, tumor marker level, and the patient's general status. Of the total number of patients, 89 (34%) were treated with TACE, 69 (26%) with surgical resection, nine (3%) with PEI, six (2%) with liver transplantation (LT) and two (1%) with intraoperative RFA. The remaining 88 (34%) patients with 116 small HCC underwent percutaneous RFA as a first-line treatment option and were included in the study population. Patient characteristics are given in Table 1. The 54 men and 34 women (age range, 45–80 years; median, 68 years) were followed for 14–90 months (median, 36 months). All patients underwent ultrasound-guided percutaneous RFA, at which time 73 had Child–Pugh class A and 15 had Child–Pugh class B cirrhosis. Most patients (76%, 67/88) had a single tumor, 14 (16%) had two tumors and seven (8%) had three tumors. Maximum tumor diameter ranged 1–3 cm (median, 1.8 cm). Of the 88 patients, the HCC was 2 cm or less in 61 and more than 2 cm in the remaining 27. The RFA procedure was explained fully to all patients, and informed consent was obtained. Despite the feasibility and availability of surgery, all patients voluntarily preferred ablation under informed consent. This study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki.
Table 1. Characteristics of patients treated by radiofrequency ablation
Values are median (range).
Values are number of patients unless stated otherwise.
AFP, α-fetoprotein; DCP, des-γ-carboxyprothrombin; HBV, hepatitis B virus; HCV, hepatitis C virus; PT, prothrombin time; RFA, radiofrequency ablation.
Pretreatment imaging studies included abdominal ultrasonography (US), contrast-enhanced dynamic CT, and angiography combined with CT during arterial portography and hepatic arteriography.
In 91 nodules, HCC was diagnosed based on the following classic imaging manifestations: (i) early enhancement at the arterial phase and hypoattenuation at the portal venous phase or at the equilibrium phase on contrast-enhanced dynamic CT; and (ii) hyperattenuation on CT during hepatic arteriography and hypoattenuation on CT during arterial portography.16–18 The remaining 25 nodules in 18 patients were diagnosed as HCC by pathological methods.
All 116 nodules were percutaneously treated under US guidance, with all patients under conscious sedation with pentazocine (5–10 mg, Pentagin; Sankyo, Tokyo, Japan) and midazolam (1–4 mg, Dormicum; Astellas, Tokyo, Japan) administrated i.v. Vital signs were continuously monitored during the procedure. A 200-W, 480-kHz monopolar radiofrequency generator (Cool-tip RF system; Radionics, Burlington, MA, USA) was used as the energy source. Radiofrequency energy was delivered with a 17-G, Cool-Tip electrode with a 2-cm (nodules ≤ 1.5 cm in diameter) or 3-cm (nodules 1.5–2.5 cm) exposed metallic tip (Radionics) introduced into the center of the nodule. For nodules greater than 2.5 cm, multiple needle punctures were performed with a 3-cm exposed metallic tip. In the 70 (80%) of 88 patients with hypervascular HCC nodules confirmed on CT during hepatic arteriography, TACE was performed an average 3 days before RFA. TACE was performed through the femoral artery using the technique of Seldinger under local anesthesia. An angiographic catheter was inserted selectively into the hepatic feeding artery of a segment or subsegments containing the target tumor. We used cisplatinum (Randa; Nippon Kayaku, Tokyo, Japan) as an anticancer drug mixed with iodized oil (Lipiodol; Nihon Schering, Tokyo, Japan) at a concentration of 10 mg/mL and injected at a dose of 10–40 mg/body. The selected dose was based on tumor size. Injection was discontinued upon full accumulation of iodized oil in the tumor vessels. No gelatin sponge or coil embolization was used after TACE in the present study.
Efficacy of RFA and follow up
To evaluate the efficacy of RFA, dynamic CT was performed at 2–3 days after each treatment session using the protocol described for pretreatment imaging studies. CT findings were confirmed by consensus between the two radiologists. On dynamic CT images, the non-enhancing area was measured as the ablated area. When the diameter of the non-enhancing area was greater than that of the ablated nodule, RFA was considered to have produced a complete effect and the treatment was terminated. When the diameter of the necrotic area was closely similar to that of the tumor without any ablated margin or when the only partial enhancement of a portion of the tumor was seen, RFA was considered to have produced an incomplete effect, and an additional session of ablation was accordingly performed 3–5 days later, and in principle repeated until a complete effect was confirmed. The outcome of RFA was evaluated in CT images taken 3–4 weeks after the final RFA session.
Follow-up US and dynamic CT were performed at 3- to 4-month intervals using protocols similar to those applied in the pretreatment studies. Serum HCC-specific tumor markers, including α-fetoprotein (AFP), its lectin fraction 3, and des-γ-carboxy prothrombin (DCP), were measured every 1–2 months. Local tumor progression was determined when a subsequent follow-up CT demonstrated any tumor growth or enhancement in the ablation zone, where complete primary effectiveness (i.e. no evidence of residual tumor) was previously obtained. Secondary effectiveness was evaluated in patients who underwent follow-up CT 1 year or more after RFA,19 and included tumors that underwent successful repeat RFA after the identification of local tumor progression. However, locally progressive tumors treated with intraoperative RFA, TACE, PEI or surgical resection were excluded from consideration of complete secondary effectiveness. Early recurrence was defined as that occurring within 12 months and late recurrence as that after more than 12 months.
Survival analysis was performed on a patient-by-patient basis. Disease-free survival was considered to be survival time from the first RFA to the last follow up, local tumor progression, occurrence of new HCC in the liver, distant metastasis or death, whichever occurred first.
Complications were assigned to major and minor categories.19 Major complications were defined as those which required treatment or additional hospitalization, or which resulted in permanent adverse sequelae. All other complications were considered to be minor.
Common major complications that occurred after percutaneous RFA were hemorrhages requiring transfusion, liver abscesses requiring percutaneous drainage, bile duct injuries requiring biliary drainage, pleural effusions or homotraces requiring thoracentesis, bowel perforations, cancer seeding, hepatic failure and death. Complications were assessed on the basis of the number of treatments and sessions.
Cumulative rates of local tumor progression were assessed using the Kaplan–Meier method. Univariate analysis was performed to identify parameters predicting overall survival, and to identify parameters predicting disease-free survival. Rates of overall survival and disease-free survival were assessed using the Kaplan–Meier method and compared with the log–rank test. In addition, a univariate Cox proportional hazards model was fitted to each variable, and all variables of P < 0.10 were subjected to multivariate analysis to assess their value as independent predictors of overall and disease-free survival. Moreover, we compared the differences in clinical features between the early recurrence and late recurrence groups: continuous data were expressed as median (range) and compared using the Mann–Whitney U-test, while categorical variables were compared using the χ2-test. Multivariate analysis of risk factors for early recurrence was performed by the stepwise logistic regression model.
P < 0.05 was considered to be a significant difference. Data processing and analysis were performed with commercially available software (SPSS ver. 9.0 for Windows; SPSS, Chicago, IL, USA).
Of a total of 263 patients with small HCC, 88 patients were treated with percutaneous RFA, 70 of whom were treated with a combination of TACE with RFA. The remaining 18 patients were treated by RFA alone. Fifty-eight patients obtained complete ablation in one session, 21 in two sessions and nine in three sessions, giving 87 of 88 patients with complete ablation. Complete ablation was not obtained in the remaining patient. Of the 87 patients with complete ablation, three patients developed local tumor progression, as did the one patient without complete ablation (Fig. 1).
Local tumor control
Of 116 tumors, 115 (99%) showed complete primary effectiveness on 1-month follow-up CT, while 107 (97.2%) of 110 showed complete secondary effectiveness with follow-up CT of 1 year or more.
Of 115 tumors with complete primary effectiveness, local tumor progression was identified in four (3.4%) at follow-up CT from 4–11 months (mean, 8.2 months) after RFA. Cumulative rates of local tumor progression, estimated at 1 and 3 years, were 4.8% and 4.8%, respectively. No further local tumor progression was detected after month 11. One (25%) of four locally progressive tumors was treated again with percutaneous RFA. Of the remaining three, one was treated with PEI, one with surgical resection and one with hepatic arterial infusion therapy. All four of these patients developed local tumor progression within 12 months.
Cumulative disease-free survival rates estimated at 3 and 5 years were 34.0% and 24.0%, respectively (Fig. 2). Univariate analysis identified tumor size, tumor number, AFP, serum albumin level, platelet count, indocyanine green 15-min retention rate (ICG-R15) and hepatitis B virus (HBV) infection as significant determinants of disease-free survival. Multivariate analysis identified tumor number as the only statistically significant determinant of disease-free survival (solitary, hazards ratio [HR] = 2.465, 95% confidence interval [CI] = 1.170–5.191, P = 0.018).
Of a total of 88 patients, 17 (19.3%) died due to HCC (n = 5), hepatic failure or complications of cirrhosis (n = 6) and other causes (n = 6).
Cumulative overall survival rates estimated at 3 and 5 years were 83.0% and 70.0%, respectively (Fig. 3). Univariate analysis identified sex, age, serum bilirubin level, serum albumin level, and (ICG-R15; %) as significant determinants of overall survival. Multivariate analysis identified age and ICG-R15 as the statistically significant determinants of overall survival (aged < 70 years; HR = 2.341, 95% CI = 1.101–4.977, P = 0.027; and ICG-R15 < 15%; HR = 3.621, 95% CI = 1.086–12.079, P = 0.036).
Comparison with clinical features between the early and late recurrence groups
Table 2 summarizes the characteristics of the 43 (48.8%) of 88 patients with recurrence during the follow-up period after RFA, classified according to time to recurrence into early (n = 18) and late recurrence (n = 25) subgroups. Correlations between time to recurrence from RFA and prognosis were analyzed. Kaplan–Meier curves for overall survival after RFA according to time to recurrence are shown in Figure 4. Overall survival of patients with early recurrence was significantly worse than that of patients with late recurrence (P = 0.014). On subgroup analysis, tumor size showed a significant association with early recurrence (P = 0.031). Multivariate analysis identified tumor size of more than 2 cm (risk ratio [RR] = 4.629, 95% CI = 1.241–17.241, P = 0.023) as the only independent risk factor for early recurrence of HCC after RFA.
Table 2. Characteristics of 43 patients with recurrence after RFA (according to time to recurrence)
Early recurrence group (n = 18)
Late recurrence group (n = 25)
Values are median (range).
Values are number of patients (%) unless stated otherwise.
AFP, α-fetoprotein; DCP, des-γ-carboxyprothrombin; HBV, hepatitis B virus; HCV, hepatitis C virus; PT, prothrombin time; RFA, radiofrequency ablation.
Of all patients with recurrence, four developed local tumor progression, all of whom were in the early recurrence group. Two of these four patients had perivascular tumors. Of the remaining two, treatment evaluation at CT scan was insufficient in one patient, while in the second, introduction of the radiofrequency electrode was difficult because of the insufficient imaging provided by US.
Of the 43 patients with recurrence, 20 (46%) were initially treated with TACE, 13 (30%) with RFA, five (12%) with surgical resection, two (5%) with PEI and one (2%) with hepatic arterial infusion therapy. The remaining two patients (5%) received no specific treatment prior to death. None of the 88 patients developed extrahepatic metastases during the follow-up period, nor was neoplastic seeding identified.
A total of 88 RFA treatments in 127 sessions were performed as first-line treatment for 116 HCC in 88 patients (mean, 1.47 sessions/treatment). A total of five complications (5.7% per treatment, 3.9% per session) were observed during the follow-up period.
Among complications, pleural effusion was observed in three patients, but drainage was not required. Two patients with hepatic infarction showed an increase in serum aspartate aminotransferase levels (range, 207–447 IU/mL; mean, 270.8 IU/mL). Fever greater than 38°C was observed in seven patients after RFA, all of whom showed complete recovery within 5 days without special treatment. No major complications were encountered in any patient, and no procedure-related death occurred.
In the present study, combination TACE and RFA was performed in patients with hypervascular HCC nodules. On the other hand, patients with hypovascular HCC nodules were treated by RFA alone. Efficacy was evaluated by dynamic CT 2–3 days after each treatment session, and RFA sessions were repeated until an ablative margin was obtained.
Using this protocol, we performed percutaneous RFA in 88 consecutive patients with small HCC (up to 3 nodules, each up to 3cm in diameter) and assessed prognostic factors that affected therapeutic outcomes.
Results from recent retrospective studies of long-term survival with RFA treatment have been promising.15,20–22 In their trial of 664 patients with HCC treated with percutaneous RFA, the largest to date, Tateishi et al.15 reported cumulative survival rates at 1, 3 and 5 years of 94.7%, 77.7% and 54.3% for primary HCC and 91.8%, 62.4% and 38.2% for recurrent HCC, respectively. They performed TAE with Lipiodol to tumors of more than 2 cm to delineate the border of the tumors at CT scan for treatment evaluation after RFA. Our present long-term (5-year) overall survival rate of 70% is better than those in these previous studies. Results showed no significant difference in overall survival between RFA with and without TACE. In an Italian study in 187 patients with Child–Pugh class A or B cirrhosis and early-stage HCC who were excluded from surgery, overall survival rates at 1, 2, 3, 4 and 5 years were 97%, 89%, 71%, 57% and 48%, respectively.20 The only significant prognostic factor seen in both these two studies was Child–Pugh class. Similarly, we also identified host factors (age and ICG-R15) as prognostic factors in our present study. In contrast with reports of hepatic resection of HCC, the present and previous studies of RFA did not identify tumor factors as prognostic. Taken together, these results indicate the strong potential of percutaneous RFA as a treatment modality for small HCC.
In our study, the estimated 3- and 5-year disease-free survival rates were 34% and 24%, respectively. In their study of 570 patients with early-stage HCC treated with percutaneous RFA, Choi et al.23 reported cumulative disease-free survival rates at 3 and 5 years of 26.5% and 21.0%, respectively, which were consistent with our present results. In our analysis, only tumor factor (no. of tumors: multiple) was significantly associated with disease-free survival. Latent tumors might already have existed at the time of RFA.
In our study, local tumor progression rate during a median of 36 months of follow up was 4.8%, a markedly low rate compared with those reported previously. In accordance with our institutional protocol for small HCC, combination of TACE and RFA was performed in patients with hypervascular HCC nodules. Vascular occlusion by TACE permits the formation of larger thermal lesions by reducing heat loss.13 In addition, accumulation of Lipiodol might be useful for obtaining the border of the tumors at CT scan after RFA. To ensure complete ablation, cases evaluated as incompletely ablated following the first session of RFA were subject to a second treatment session 3–5 days later. Our RFA protocol might have contributed to our results of local tumor control. Nevertheless, four patients with local tumor progression after RFA were seen. For perivascular tumors in particular, the possible heat-sink effect of intrahepatic blood flow means that the possibility of incomplete ablation is high. This hypothesis is supported by a study conducted by Lu et al.,24 in which perivascular tumor location was an independent and dominant predictor of treatment outcome of RFA in terms of both the completeness of ablation and local tumor progression. On this basis, RFA combined with PEI might be useful in preventing local tumor progression of perivascular HCC. For those cases in which poor conspicuity of the tumor at US hampered introduction of the radiofrequency electrode, we should have used contrast-enhanced US.25 In our study, review of CT images in a patient who developed local tumor progression showed that the initial evaluation of therapeutic response was insufficient. Although the therapeutic response of HCC to RFA is often evaluated by comparing pre- and post-RFA CT, it is sometimes difficult to determine whether an ablative margin has been achieved. One solution to this problem may be fusion of pre- and post-RFA CT images,26 but any achievement of a local tumor progression rate of 0% might be difficult as long as the evaluation of response to RFA is restricted to imaging examination only. This restriction should be noted in a treatment of RFA.
Moreover, comparison of clinical features between the early and late recurrence groups showed that overall survival was significantly worse in patients with early recurrence after RFA than in those with late recurrence. Recent studies have shown that the time interval from resection of HCC to recurrence is an independent prognostic factor of survival after recurrence,27,28 suggesting that early recurrence arises primarily from intrahepatic metastases, whereas most late recurrences are likely of multicentric origin. Our present results might accord with results of these studies. Further, all patients with local tumor progression were in the early recurrence group, among whom only one was treated with percutaneous RFA, and other treatments were selected in the remaining three patients. In the present study, patients with local tumor progression had poor prognosis. Our RFA protocol might have the potential to provide local tumor control for small HCC. Moreover, this RFA protocol might decrease the number of patients with early recurrence of HCC, and contribute to the improvement of the prognosis.
In addition, these findings also suggest the need for different therapeutic approaches to the prevention of early and late recurrence after RFA for HCC. In our analysis, an association with early recurrence was limited to a single tumor factor (tumor size > 2 cm) only. For patients with this risk factor, treatment modalities with potential of more curative intent, such as RFA combined with TACE13 or hepatic resection might have to be selected if possible. A randomized controlled trial might be necessary to solve this issue.
To prevent late recurrence, therapeutic approaches effective at suppressing multicentric occurrence such as polyprenoic acid and interferon (IFN) therapy may be indicated in patients with cirrhotic liver.29–31 Of 88 patients who underwent RFA, 79 were hepatitis C virus-positive, 21 of whom received IFN therapy. Of these, a sustained virological response was achieved in five. Because the number of cases is small, the effect of IFN therapy could not be analyzed. Our policy is to evaluate for the complete ablation after RFA and to implement rigorous CT and US surveillance. On this basis, effective treatment modalities (hepatic resection, repeated RFA, or TACE) can be considered as early as possible before recurrent tumor progression.
A total of five complications (5.7% per treatment, 3.9% per session) were observed during the follow-up period, but none of these was major or required the cessation of therapy.
In conclusion, under our RFA protocol percutaneous RFA is considered a reliable treatment for small HCC in terms of therapeutic efficacy and safety. Although the present study has some limitations, such as the small number of patients and retrospective design, our results demonstrate that percutaneous RFA can be used successfully as first-line treatment for small HCC. In addition, we also demonstrated that early and late intrahepatic recurrence after RFA of HCC were associated with prognosis. These findings may suggest a need for different strategies in the prevention and management of early and late intrahepatic recurrence.