Fax: (011) 886-3-3272236
Prospective randomized controlled study of interferon-alpha in preventing hepatocellular carcinoma recurrence after medical ablation therapy for primary tumors
Article first published online: 12 DEC 2003
Copyright © 2003 American Cancer Society
Volume 100, Issue 2, pages 376–382, 15 January 2004
How to Cite
Lin, S.-M., Lin, C.-J., Hsu, C.-W., Tai, D.-I., Sheen, I.-S., Lin, D.-Y. and Liaw, Y.-F. (2004), Prospective randomized controlled study of interferon-alpha in preventing hepatocellular carcinoma recurrence after medical ablation therapy for primary tumors. Cancer, 100: 376–382. doi: 10.1002/cncr.20004
- Issue published online: 5 JAN 2004
- Article first published online: 12 DEC 2003
- Manuscript Accepted: 13 OCT 2003
- Manuscript Revised: 3 OCT 2003
- Manuscript Received: 9 JUN 2003
- interferon (IFN-α);
- hepatocellular carcinoma (HCC);
- medical ablation therapy
Hepatocellular carcinoma (HCC) recurrence after ablation therapy for primary tumors is common.
To evaluate the effectiveness of interferon-alpha (IFN-α) in preventing HCC recurrence, 30 eligible patients were randomized into three groups: 11 patients treated with three mega units (MU) of IFN-α three times weekly for 24 months (IFN-α-continuous group), 9 patients treated with 3 MU of IFN-α daily for 10 days every month for 6 months followed by 3 MU of IFN-α daily for 10 days every 3 months for a further 18 months (IFN-α-intermittent group), and 10 patients who received no IFN-α therapy (control group). The three groups were comparable in terms of etiology, demographics, and laboratory data at entry and HCC characteristics.
After a median follow-up of 27 months (range 4–53 months), 9 patients (90%) in the control group and 9 patients (45%) in 2 treatment groups (6 patients in the IFN-α-continuous group and 3 patients in the IFN-α-intermittent group) developed an HCC recurrence (P = 0.021). Cumulative HCC recurrence rates in the IFN-α-intermittent, IFN-α-continuous, and control groups were 22.2%, 27.3%, and 40% at the end of 1 year and 33.3%, 54.6%, and 90% at the end of 4 years (P = 0.0375), respectively (control vs. IFN-α-intermittent group, P = 0.0123; vs. IFN-α-continuous group, P = 0.0822). If both IFN-α groups were combined, the cumulative HCC recurrence rate of the patients treated with IFN-α and the control group was 25% and 40% at the end of 1 year and 47% and 90% at the end of 4 years, respectively (P = 0.0135).
The data suggested that IFN-α therapy may reduce HCC recurrence after medical ablation therapy for primary tumors. Cancer 2004;100:376–82. © 2003 American Cancer Society.
Hepatocellular carcinoma (HCC) is a common malignancy in Asia.1, 2 Although various locoregional therapies including percutaneous ethanol injection (PEI), percutaneous acetic acid injection (PAI), or radiofrequency ablation (RFA) can achieve complete necrosis of small HCC,3–5 HCC recurrence after these therapies is still common.3–6 Previously we have successfully treated HCC using medical modalities such as transcatheter arterial chemoembolization (TACE), PEI, or PAI.7–9 HCC recurrence after these treatment modalities is also common. Although candidates for medical ablation usually exhibit compensated hepatic functional status, the frequent recurrence of HCC after successful ablation contributes to short survival. Therefore, attempts to prevent HCC recurrence are essential to prolong patient survival.
Interferon-alpha (IFN-α) has immunomodulatory, antiviral, antiproliferative, and antitumoral effects.10–12 It is effective in treating hepatitis B virus (HBV) and hepatitis C virus (HCV) infection, improving the prognosis of patients with chronic liver disease, reducing cirrhosis or HCC development, and prolonging survival.13–16 Several investigators have demonstrated that IFN-α is effective in preventing HCV-related HCC recurrence after resection17–19 and one investigation has shown the effectiveness of IFN-α in preventing HBV-related HCC recurrence.20 These favorable results prompted us to conduct a randomized controlled study to evaluate the effectiveness of IFN-α therapy in preventing HCC recurrence after successful medical ablation therapy for primary tumors.
MATERIALS AND METHODS
Patient Characteristics and Treatment of HCC
Excluding patients with evidence of decompensated liver disease (Child–Pugh class C, ascites, bleeding esophageal or gastric varices, serum albumin level < 3.0 g/dL, serum bilirubin level > 3 mg/dL, or hepatic encephalopathy) or contraindication for IFN therapy (such as platelet count < 50,000/mm3), 30 patients with 35 nodular HCCs were successfully treated with nonsurgical methods, including PAI for 10 HCC tumors and TACE with subsequent PAI for 25 HCC tumors between July 27, 1998 and September 21, 1999. Successful treatment of HCC was defined as an absence of enhancement on dynamic computed tomography scan (CT) plus necrosis on multiple fine-needle aspiration cytology smears at least 4 months after treatment. Table 1 lists the clinical characteristics of all patients. Nineteen HCCs were diagnosed by histology and 16 HCCs were diagnosed by cytology. HCCs were quantified by dynamic CT scan, arterial angiography, and CT scan during arterial portography. None of the 30 patients underwent surgery, because of older age (> 70 years) in 1 patient, cirrhosis with Child-Pugh Grade B in 2 patients, concurrent uremia in 1 patient, refusal of reresection for postoperative HCC recurrence in 3 patients, and refusal of resection in 23 patients.
|No. of patients||10||11||9|
|Median age (range)||59 (49–72)||60 (26–70)||63 (31–65)||NS|
|Child–Pugh class (A/B/no cirrhosis)||9/1/0||9/0/2||8/1/0||NS|
|HCC characteristics before trial|
|No. of HCC (single vs. two)||8/2||9/2||8/1||NS|
|Median size of maximal tumor (mm) (range)||25 (20–36)||20 (10–69)||23 (12–37)||NS|
|Edmondson's Grade 1/2/3||2/2/8||2/6/5||2/4/4||NS|
|HCC therapy before trial (PAI/PAI+TACE)||2/10||4/9||4/6||NS|
|Median laboratory values (range)|
|Albumin (g/dL)||3.8 (3.5–4.8)||4.1 (3.5–5.0)||4.3 (3.1–4.7)||NS|
|AST(IU/L)||47 (35–200)||33 (15–161)||46 (30–78)||NS|
|ALT(IU/L)||61 (29–304)||46 (17–239)||54 (21–76)||NS|
|Platelet (× 1000/mm3)||72 (58–101)||131 (51–189)||84 (51–176)||NS|
|α-fetoprotein (ng/ml)||20 (9 to > 10,000)||16 (3–7898)||18 (4–420)||NS|
PAI was performed using 25% or 50% acetic acid under sonographic guidance (Aloka SSD-650, Tokyo, Japan). PAI was performed twice weekly until the tumor displayed a nearly completely hyperechoic image on sonography, for a maximum of three sessions for HCC < 3 cm, or a maximum of six sessions for HCC measuring 3.1–6 cm. Chemoembolization was performed by lipiodol (Lipiodol Ultra fluide, Mitsui, Tokyo, Japan) or doxorubicin injection followed by a mixture of gelatin sponge articles (Gelfoam; Upjohn Pharmacia, Tokyo, Japan) insertion into the tumor feeding artery. Booster PAI was performed for tumors identified as viable using helical CT scan, and displayed enhancement of the partially embolized tumor 2–4 weeks after chemoembolization. Booster PAI was performed similarly as the monotherapy PAI.
Study Design and IFN-α Therapy
After achieving complete tumor necrosis, all 30 patients were randomized to the following 3 groups: 11 patients received an intramuscular injection of 3 mega units (MU) of IFN-α-2b 3 times a week for ≥ 24 months (IFN-α-continuous group), 9 patients received 3 MU of IFN-α2b daily for 10 days every month for 6 months followed by 3 MU IFN-α daily for 10 days every 3 months for an additional 18 months (IFN-α-intermittent group), and the remaining 10 patients received only nonspecific hepatotropic tablets without injection (control group). Two patients in the IFN-α-intermittent group developed severe soreness and malaise after the first two and three IFN-α doses, respectively, and refused further therapy. However, these two patients who dropped out of IFN-α-intermittent therapy were also included in the intention-to-treat analysis of HCC recurrence. During IFN-α therapy, complete blood cell count was evaluated at 2–4-week intervals. The IFN-α dose was reduced to 1.5 MU if the neutrophil count was < 2500/mm3 or the platelet was < 50,000/mm3. If the neutrophil count was < 500/mm3 or the platelet count was < 20,000/mm3, IFN-α therapy was ceased for 1 week to allow the cell counts to return to suitable levels.
The study protocol was reviewed by the hospital ethics committee and informed consent was obtained from the patients. Patient enrollment ended in September 1999.
Laboratory Tests and Detection of HCC Recurrence
Complete blood cell count and levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), and albumin were monitored every 2–4 weeks in the first 2 years during IFN-α therapy and levels of ALT, AST, and albumin were monitored every 1–3 months afterward. Additional tests also were included, namely, levels of viral markers and α-fetoprotein (AFP). Serum levels of HBV DNA or HCV RNA were checked if necessary. Recurrent HCC was monitored by AFP measurement and sonography every 2 months and by dynamic CT scan every 6 months, or more frequently if required. All patients were followed up at monthly intervals during therapy and at 3-month intervals during follow-up. The end point of the study was tumor recurrence. More frequent visits or evaluations were performed given suspicions of HCC recurrence or hepatitis flare ups. If sonography or CT scan raised suspicions of HCC recurrence, arterial angiography, core needle biopsy, or aspiration cytology was used for verification (Table 2).
|Characteristics||IFN-α group||Control group|
|No. of patients||4||5||9||4||5||9|
|Computed tomography scan||5||5||10||5||7||12|
|Median α-Fetoprotein level (ng/mL) (range)||1700 (5–1953)||88 (3–476)||84 (3–1953)||363 (10–461)||421 (229–798)||363 (10–798)|
|Characteristics of HCC|
|Median size (mm) (range)||15 (14–20)||16 (12–22)||16 (12–22)||16 (10–20)||16 (12–20)||16 (10–20)|
|No. of tumors (one/two/three)||3/1/0||3/1/1||5/2/1||3/1/0||3/2/0||6/3/0|
|Edmondson's Grade 1 or 2/3||5/0||5/3||10/2||5/0||5/2||10/2|
|Median age (range)||60 (41–66)||64 (48–68)||60 (41–68)||62 (59–65)||51 (49–69)||59 (49–69)|
|Therapy of HCC before trial|
|PAI and TACE||2||3||5||3||5||8|
Biochemical tests and complete blood cell count were performed using routine automated techniques. The viral markers and AFP were assayed using commercially available radioimmunoassay kits (Ausria II, HBeAg-RIA; Abbott Laboratories, North Chicago, IL). Serum anti-HCV was assayed using a second or third-generation enzyme immunoassay kit (UBI-HCV-EIA II or III, United Biochemical Inc., Lake Success, NY). Finally, serum levels of HBV DNA were measured by polymerase chain reaction (PCR) as described in an earlier study.21 HCV RNA was assayed using a sensitive reverse transcription-PCR method (AMPLICOR HCV test; Roche Diagnostic, NJ).
Data were expressed as mean ± standard deviation. The Student t test, chi-square test, or Fisher exact test were used to compare various variables among the groups. Multiple comparisons were performed for three or more groups or subgroups. HCC recurrence analysis and multivariate analysis with age, gender, history of previous HCC resection, Child–Pugh grade, tumor number, tumor size, histocytologic grade, baseline albumin level, ALT level, gamma-glutamyl transpeptidase (GT) level, platelet count, AFP level, IFN-α therapy, and regimen of IFN-α therapy were performed using the Cox regression model with proportional hazards.22 Data analysis was conducted using SAS software version 6.12 (SAS Institute Inc., Cary, NC).23 Significance was set at a P < 0.05.
HCC Recurrence and Treatment
After a median of 27 months (range, 4–53 months) after entry, HCC recurrence was detected in 9 patients (90%) in the control group, in 6 (55%) patients in the IFN-α-continuous group, and in 3 patients (33%) in the IFN-α-intermittent group. If the 2 IFN-α groups were combined, the recurrence of HCC was lower than in the control group (45% vs. 90%; P = 0.021). Of the HBV patients, HCC recurred in 4 of 4 (100%) untreated patients and in 4 of 12 (33.3%) IFN-α-treated patients (P = 0.0384, Fisher exact test). Of the HCV patients, HCC recurred in 5 of 6 (83.3%) untreated patients and in 5 of 8 (62.5%) IFN-α-treated patients (P = 0.335, Fisher exact test; Table 2). The 1-year, 2-year, 3-year, and 4-year cumulative HCC recurrence rates were, respectively, 40%, 70%, 90%, and 90% in the control group, 22.2%, 22.2%, 33.3%, and 33.3%, respectively, in the IFN-α-intermittent group, and 27.3%, 36.4%, 45.5%, and 54.6%, respectively, in the IFN-α-continuous group. The difference in cumulative HCC recurrence rates across the IFN-α-intermittent, IFN-α-continuous, and control groups was significant (P = 0.0375). The cumulative HCC recurrence rate was higher in the control group than in the IFN-α-intermittent (P = 0.0123) and IFN-α-continuous groups (P = 0.0822), whereas no difference (P = 0.4389) was found between the IFN-α-continuous and IFN-α-intermittent groups. If the two IFN-α groups were combined, the 1-year, 2-year, 3-year, and 4-year cumulative HCC recurrence rates were 25%, 30%, 40.4%, and 47%, respectively, compared with 40%, 70%, 90%, and 90%, respectively, in the control group (control vs. IFN-α, P = 0.2317 at the end of 1 year; P = 0.045 at the end of 2 years; P = 0.0111 at the end of 3 years; and P = 0.0208 by the Fisher exact test or P = 0.0135 by Mantel–Cox analysis at the end of 4 years; (Fig. 1).
Among the 30 patients in the current study, 3 patients (1 each in the 3 groups) received previous surgical resection for HCC before study entry. These three patients developed HCC recurrence during the follow-up period of this trial. Excluding these three patients, the 1-year, 2-year, 3-year, and 4-year cumulative HCC recurrence rates were, respectively, 44.4%, 66.7%, 88.9%, and 88.9% in the control group, 25%, 25%, 25%, and 25%, respectively, in the IFN-α-intermittent group, and 30%, 40%, 40%, and 50%, respectively, in the IFN-α-continuous group. The difference in cumulative HCC recurrence rates across the IFN-intermittent, IFN-α-continuous, and control groups was still significant (P = 0.0419). The cumulative HCC recurrence rate was higher in the control group than in the IFN-α-intermittent group (P = 0.021), whereas no difference was present between the IFN-α-continuous and IFN-α-intermittent groups (P = 0.3495) and between the control and IFN-α-continuous groups (P = 0.1253). If the 2 IFN groups were combined, the 1-year, 2-year, 3-year, and 4-year cumulative HCC recurrence rates were 27.8%, 33.3%, 33.3%, and 40.7%, respectively, compared with 44.4%, 66.7%, 88.9%, and 88.9%, respectively, in the control group (control vs. IFN-α, P = 0.235 at the end of 1 year; P = 0.091 at the end of 2 years; P = 0.0125 at the end of 3 years; and P = 0.0174 by the Fisher exact test or P = 0.0264 by Mantel–Cox analysis at the end of 4 years).
Treatment of recurrent HCC included resection in one patient, PEI or PAI only in four patients, combined TACE and PEI in four patients, PEI and RFA in six patients, and liver transplantation after PEI and TACE in one patient. RFA was performed as described elsewhere.24 In the control group, one patient died of hepatic failure and four patients died of hepatic failure and HCC progression. In the two treatment groups, three patients died of HCC progression and hepatic failure and one patient who dropped out of intermittent IFN therapy died of hepatic failure.
Univariate and Multivariate analysis in Relation to Background Variables and Recurrence of HCC
Using univariate analysis of background factors in relation to HCC recurrence revealed no significant differences in patient age (> 60 years vs. ≤ 60 years; P = 0.247) or gender (male vs. female; P = 0.110), maximum tumor dimension (> 20 mm vs. ≤ 20 mm, P = 0.15), tumor number (single vs. double, P = 0.327), TNM stage (Stage I vs. Stage II, P = 0.068), Edmondson grade (Grade 1 and 2 vs. 3, P = 0.362), cirrhosis at entry (cirrhosis vs. no cirrhosis, P = 0.438), or laboratory parameters (Table 3). However, significant differences were observed for IFN-α therapy versus no therapy (P = 0.021) and for IFN-α therapy regimens (comparisons among IFN-α-continuous, IFN-α-intermittent, and control groups, P < 0.05, Fisher exact test) (Table 3). Multivariate analysis showed that IFN-α therapy (P = 0.0168) and regimen of IFN-α therapy (P = 0.036) were the only two significant factors related to HCC recurrence. However, history of previous HCC resection (P = 0.146), age (P = 0.745), gender (P = 0.329), hepatitis B surface antigen (HBsAg) seropositivity (P = 0.101), anti-HCV seropositivity (P = 0.247), cirrhosis at entry (P = 0.775), Edmondson grade (P = 0.0649), PAI therapy (P = 0.334), chemoembolization therapy (P = 0.465), baseline albumin level (P = 0.548), baseline AST level (P = 0.812), baseline ALT level (P = 0.872), baseline platelet count (P = 0.999), and baseline AFP level (P = 0.434) were not significant factors. Intermittent IFN administration was the most effective regimen in the prevention of HCC recurrence.
|Factors||No. of recurrences (%)||P|
|≤ 60 (n = 15)||10 (66.7)||0.247|
|> 60 (n = 15)||8 (53.3)|
|Male (n = 23)||12 (52.2)||0.110|
|Female (n = 7)||6 (85.7)|
|Tumor dimension (mm)|
|≤ 20 (n = 16)||8 (50)||0.150|
|> 20 (n = 14)||10 (71.4)|
|1, 2 (n = 13)||7 (53.8)||0.362|
|3 (n = 17)||11 (64.7)|
|I (n = 14)||6 (42.9)||0.068|
|II (n = 16)||12 (75)|
|HBV (n = 16)||8 (50)||0.152|
|HCV (n = 14)||10 (71.4)|
|≤ 4 (n = 15)||11 (73.3)||0.104|
|> 4 (n = 15)||7 (46.7)|
|≤ 50 (n = 17)||10 (58.8)||0.291|
|> 50 (n = 13)||8 (61.5)|
|≤ 50 (n = 15)||9 (60)||0.293|
|> 50 (n = 15)||9 (60)|
|≤ 80,000/mm3 (n = 15)||9 (60)||0.293|
|> 80,000/mm3 (n = 15)||9 (60)|
|≤ 15 (n = 13)||6 (46.2)||0.124|
|> 15 (n = 17)||12 (70.6)|
|Yes (n = 20)||9 (45)||0.021|
|No (n = 10)||9 (90)|
|HBV patients (n = 12)||4 (33.3)||0.192|
|HCV patients (n = 8)||5 (62.5)|
|HBV patients (n = 4)||4 (100)||0.600|
|HCV patients (n = 6)||5 (83.3)|
|Patients with positive HBsAg|
|IFN-α therapy (n = 12)||4 (33.3)||0.038|
|Placebo control (n = 4)||4 (100)|
|Patients with positive anti-HCV|
|IFN-α therapy (n = 8)||5 (62.5)||0.335|
|Placebo control (n = 6)||5 (83.3)|
During the study period, acute exacerbation with ALT elevation up to 394 IU/L was encountered in one patient with chronic HCV infection in the control group. In addition, acute exacerbation of ALT levels up to 452 IU/L occurred in one patient with chronic HBV infection during IFN-α therapy withdrawal because of HCC recurrence. No flare up of ALT was observed in the remaining patients. HBeAg seroconversion occurred in one patient 4 months after IFN-α therapy (IFN-α-continuous group). AST and ALT levels normalized in patients with HBV infection but not in patients with HCV infection during IFN-α therapy.
Side Effects of Interferon-Alpha
Common side effects after initial IFN injection included fever, chills, myalgia, and headache. Thrombocytopenia and neutropenia requiring IFN-α dose reduction during the course of IFN-α therapy occurred in all patients in the IFN-α-continuous group except for a 28-year-old male patient. No significant adverse effects of IFN-α therapy were observed. However, for three patients in the IFN-α-continuous group, IFN-α therapy was discontinued because of severe malaise, arthralgia, and depression, respectively.
The results of the current study show that the HCC recurrence rate among untreated patients was 40% at 1 year, 70% at 2 years, and 90% at 4 years and was higher than the rate among patients treated with IFN-α (25% at 1 year, 30% at 2 years, and 47% at 4 years). The benefit of preventing HCC recurrence after IFN-α therapy becomes evident in the patients observed for > 2 years (IFN-α-treated patients vs. IFN-α-untreated patients, P > 0.05 at 1 year and P = 0.045 at 2 years; P < 0.05 at > 2 years). A similar finding was also observed in the study by Kubo et al.17 HCC recurrence within the first 2 years was assumed to be the result of original HCC whereas HCC recurrence > 2 years after ablation therapy may be the result of new hepatocarcinogenesis.17 In addition to demonstrating that chemoprevention of HCC could also be achieved using IFN-α therapy in patients with HCC after medical ablation therapy, this small study also demonstrated that intermittent IFN-α administration was the most effective regimen. However, further stratified analyses comparing the two treatment protocols (intermittent IFN-α administration and continuous IFN-α administration) may not be appropriate in such a small group of patients as in the current study.
Oon20 reported successful prevention of HCC recurrence using intermittent IFN-α therapy for 3 years in HBV-related HCC after resection. The effective prevention of HCC recurrence using IFN-α therapy was also achieved in HCV-related HCC after resection.17, 18 The results of the current small study showed that the prevention of HCC recurrence using IFN-α therapy is effective in HBV-related HCC, but not in HCV-related HCC (Table 3). The effect of IFN-α therapy in preventing HCC recurrence was believed to be caused by the inhibition of inflammation in relation to HCV infection, because the HCV viremia is frequently present even after long-term therapy of IFN-α.17–19 Persistent abnormal ALT levels throughout the course of HCV patients may be responsible for this difference. The ongoing inflammation may contribute to hepatocarcinogensis in patients with chronic HCV or HBV infection.25
The sample size in the current study was small. This is similar to the controlled studies conducted by Ikeda et al.18 and Kubo et al.17 in which there were 20 patients and 30 patients, respectively. The likely reason for the small sample size is that most patients with HCC receiving medical ablation were in the advanced stage of chronic liver disease. Old age, thrombocytopenia, and cirrhosis with mild decompensation limit the use of IFN-α therapy. Another reason is the therapeutic consideration. Both the current study and the study of of Kubo et al.17 demonstrate that significant differences between the IFN-α therapy and control groups become evident at the end of 2 years of observation. Despite the small sample size, the differences observed were significant. Nonetheless, a larger sample size would increase the power of the study.
Severe adverse effects were uncommon in the current trial, although neutropenia requiring IFN-α dose reduction was common. The IFN-α dose used in the current study approximates that used in Oon20 and was tolerated in most patients in these two trials. The dose used in our study (6.3 MU of IFN-α weekly) was smaller than that used in the studies by Kubo et al. (12.7 MU of IFN-α weekly)17 and Ikeda et al.18 (12 MU of natural IFN-β weekly). Dose reduction was required for all but one patient in the current study. Therefore, the optimal dose for IFN-α therapy remains unclear.
Although the sample size in the current study might have been too small to reach a firm conclusion, the results derived from this randomized controlled study suggest that IFN-α therapy may reduce HCC recurrence after medical ablation for primary tumors and that intermittent IFN-α therapy may be the preferred method of administration.
The authors thank Miss Ivy Chen for her assistance.
- 2Survival of relatives of patients with hepatocellular carcinoma. Eight-year nationwide survival analysis in relatives of patients with hepatocellular carcinoma: role of viral infection. J Gastroenterol Hepatol. 2002; 17: 682–689., , , et al.
- 18Interferon beta prevents recurrence of hepatocellular carcinoma after complete resection or ablation of the primary tumor—a prospective randomized study of hepatitis C virus-related liver cancer. Hepatology. 2000; 32: 228–232., , , et al.
- 22Regression model and life tables. J R Stat Soc [B]. 1972; 34: 187–220..
- 23SAS Institute Inc. SAS/STAT user's guide, release 6.11 edition. Cary, NC: SAS, 1996.