Best Practice & Research Clinical Gastroenterology 28 (2014) 771e781
Contents lists available at ScienceDirect
Best Practice & Research Clinical Gastroenterology
2
Role of antiviral treatment for HCC prevention Massimo Colombo, Prof. *, Massimo Iavarone, MD, PhD “A. M. and A. Migliavacca” Center for Liver Disease, Division of Gastroenterology and Hepatology, degli Studi di Milano, Milan, Italy Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Universita
a b s t r a c t Keywords: Hepatocellular carcinoma HBV HCV Interferon Nucleoside analogues
Chemoprevention of hepatocellular carcinoma (HCC) is listed as a yet highly debated long-term benefit of a successful treatment of patients with chronic viral hepatitis. In the hepatitis B virus (HBV) arena, the retrospective scrutiny of both interferon and nucleos(t) ide analogues (NUC) studies failed to provide robust evidence for HCC chemoprevention, due to a number of confoundings in the studies that were originally designed to assess the antiviral activity of interferon therapy. However, the reanalysis of outcomes following patients stratification for risk factors of HCC, provided a clue to find an association between NUC therapy and a reduced risk of liver cancer in non cirrhotic patients, only. In the hepatitis C scenario, a meta analysis of 30 observational studies of patients treated with interferon demonstrated a more than 70% reduction of HCC risk occurring independently of severity of underlying liver fibrosis which was less pronounced in aged patients and those with more advanced liver fibrosis. While the reasons for the residual risk of HCC in virological responders remain largely unexplained, international societies recommend surveillance for HCC of both HBV and HCV responders to antiviral therapy. © 2014 Elsevier Ltd. All rights reserved.
Introduction Prevention of population exposure to the etiologic agents of liver disease stands as the most effective, practical approach to reduce liver-related mortality from hepatocellular carcinoma (HCC), worldwide [1]. This is the clear message conveyed by the program of mass vaccination of newborns
* Corresponding author. Division of Gastroenterology and Hepatology, Fondazione IRCCS Ca' Granda Ospedale Maggiore degli Studi di Milano, Via F. Sforza 35, 20122 Milan, Italy. Tel.: þ39 0255035432; fax: þ39 0250320410. Policlinico, Universita E-mail address: [email protected] (M. Colombo).
http://dx.doi.org/10.1016/j.bpg.2014.07.017 1521-6918/© 2014 Elsevier Ltd. All rights reserved.
772
M. Colombo, M. Iavarone / Best Practice & Research Clinical Gastroenterology 28 (2014) 771e781
against hepatitis B virus (HBV), the dominant risk factor for HCC on a global scale, that was started in early eighties in Taiwan and now is in place in more than 180 WHO nations [2]. Owing to the significant contribution of HBV and hepatitis C virus (HCV) to HCC and the availability of effective antiviral regimens to permanently suppress or remove etiologic agents, antiviral treatments of patients chronically infected by hepatitis viruses stand as a complementary to primary prophylaxis strategy in the prevention of HCC, and this is even more so in the HCV scenario where a sterilizing vaccine is far from being available [1]. However, the evidence that patients with chronic hepatitis B and hepatitis C can effectively be protected against HCC risk by treatment with interferon and/or oral anti-HBV analogues is rather controversial, due to the lack of randomized controlled trials that are ideally needed to establish efficacy, but are logistically and ethically challenging. Reanalysis of studies with antivirals suggest that virus-induced HCC was more likely to be prevented in younger patients with mild liver inflammation rather than in older patients with advanced liver fibrosis or cirrhosis, who in fact are those at higher risk of developing liver cancer [3]. While this calls for a reassessment of current strategies of patient prioritization to antiviral therapies, that are mostly dictated by cost-utility criteria and therefore target the most in need patients with advanced liver disease, we became progressively aware that uncertainty regarding rates and pattern of HCC chemoprevention by antiviral regimens is mainly the consequence of methodological flaws generated by the retrospective scrutiny of the literature. As these studies were originally designed to evaluate the antiviral potency of interferon and oral analogues, treatment outcomes could not properly be weighed for important predictors of liver cancer like hepatitis severity and comorbidities, both common causes of interferon ineligibility, and life style factors like alcohol abuse and tobacco smoking [4]. This was particularly true for patients with chronic hepatitis C, who until a few months ago were bound to receive treatment with interferon-based regimens only, which may suffice to cause permanent virus sterilization in susceptible patients, depending on the infecting genotype, severity of associated liver disease, presence of co-morbidities and genetically determined sensitivity to interferon, but remains of limited applicability and efficacy in patients with cirrhosis and comorbidities who are closest to develop an HCC [4]. As a matter of fact, treatment ineligibility of patients with severe comorbidities and those with advanced or decompensated HCV might have caused two competing risk factors of shortened survival, like HCC and all cause mortality to appear reduced in the selected cohorts of HCV patients who could achieve a sustained virological response (SVR) to interferon therapy [5,6]. This could no longer be the case once safe, universally applicable and highly effective oral regimens against HCV become available to treat patients not only with decompensated hepatitis but also those at risk of death for extrahepatic complications, provided however that antiviral therapies will reach all the infected population.
Primary prevention Primary prevention of HCC not only means prevention of population exposure to the hepatitis B virus (HBV) and hepatitis C virus (HCV) infection, but it also includes all approaches aiming to interrupt or prevent virus-induced liver damage. There are, however, major differences in the primary prophylaxis of HCC in relation with aetiology, beginning with the lack of a sterilizing vaccine against HCV to finish with differences in the outcome of antiviral therapy which leads to universal virus eradication in hepatitis C responders to interferon and to permanent suppression, but not eradication of the virus, in patients with chronic hepatitis B who successfully respond to either interferon or nucleos(t)ide analogs (NUC) therapy [3,4]. In the latter patients, in the face of a clinically silent persistence of virus replication leading to tapering down of hepatic inflammation along with reversal of HBV induced hepatic fibrosis, the same virological response does not suffice to prevent cancer, since anti-HBV therapy fails to remove the potentially carcinogenetic sequences of HBVDNA integrated into liver cell chromosomes [7]. Conversely, the virus sterilizing effects of anti-HCV therapy in sustained responders is the consequence of virus inability to integrate into human DNA sequences whereas suppression of intracytoplasmic virus replication allows for the restoration of cell innate immunity resulting in permanent clearance of infection [8]. As a consequence, a proportion of responders experience a substantial reduction of all cause and liver specific mortality, possibly due to interruption of virus induced liver fibrogenesis, with evidence of cirrhosis reversal in some patients [5,6]. These
M. Colombo, M. Iavarone / Best Practice & Research Clinical Gastroenterology 28 (2014) 771e781
773
clinical benefits may not reach those patients with chronic hepatitis C who also suffer co morbidities linked to diabetes, overweight and alcohol abuse. Hepatitis B HCC may occur at any clinical stage of HBV infection, yet tumor incidence prevails in patients with higher viremia and those with advanced liver fibrosis compared to low viremic patients and those with an early infection [9,10]. This notwithstanding, cancer risk appears to be modulated by several other virus and non virus-related factors like genotype C of HBV, mutations of pre-S, enhancer-H and core promoter, increased age, male sex, alcohol abuse and tobacco smoking, which in part account for the epidemiological and clinical heterogeneity of the tumor worldwide [11e13]. There is, however, mounting evidence that clinical heterogeneity of HCC may in part be accounted for by genetic diversity, too. While the pathogenetic mechanisms of HBV-related tumors are mainly linked to inflammatory reactions causing liver cells to transform into a mitogenic and mutagenic environment, yet virus integration within the liver cell DNA results in a number of molecular events eventually causing the activation of cell oncogenes and inactivation of tumor suppressor genes, all working to boost neoplastic transformation of the infected liver cell [7]. All in all, while these pathogenetic mechanisms of HBVrelated HCC may in part explain how genetic diversity influences the onset of HCC, they also sustain the hypothesis that liver carcinogenesis may be prevented or attenuated by pharmacological inactivation of HBV replication. Yet, systematic analysis of studies of interferon therapy of patients with chronic hepatitis B have provided conflicting evidenced for HBV-related HCC chemoprevention by interferon therapy going forward. Indeed, the rates of the tumor in treated patients was found to be reduced in three meta analyses whereas it appeared to be unchanged in other three, a chasm that can be explained by a difficulty of extrapolating HCC chemoprevention through the retrospective scrutiny of the studies that were originally designed to assess antiviral efficacy of interferon therapy. Studies outcomes, in fact, were biased by the lack of pretreatment patient stratification for relevant cancer predictors and exclusion of patients at higher risk of developing HCC like those unfit to receive interferon owing to advanced hepatitis. To some extent, the re-analysis of these studies was also biased by the lack of a separate analysis of treatment outcomes between sustained responders and non responders, who represent a majority of all patients with chronic hepatitis B receiving interferon (Table 1). Working against the constraints due to poor applicability and tolerability of interferon based regimens, was the advent of user-friendly oral nucleos(t)ide analogs (NUC) that allowed for a broader and safe access of patients to effective anti-HBV therapy. A randomized controlled study in patients with advanced fibrosis (Ishak stage 4e6) from Hong Kong and the Pacific area, provided evidence that response to the first generation nucleoside analogue lamivudine delays progression of HBV-related liver disease as it was cumulatively assessed in terms of increased Child Pugh score, liver failure or development of HCC [15]. While the clinical benefits were maintained in patients with a persistent virological response, in patients who experienced lamivudine resistance the incidence rate of clinical endpoints including HCC was higher (11% vs 5%) than in the placebo treated patients at month 32 Table 1 Meta-analysis studies of risk of hepatocellular carcinoma (HCC) in patients with chronic hepatitis B who received interferon therapy [14]. Authors
No. Studies
No. treated vs controls
HCC relative risk/risk difference* (95% CI)
p Value
Comments
Sung et al, 2008
12
1292 vs 1458
0.66 (0.48e0.89)
0.006
Yang et al, 2009 Miyake et al, 2009
11 8
1006 vs 1076 553 vs 750
0.59 (0.43e0.81) 5.0%* (9.4e0.5)
0.001 0.028
Camma et al, 2001 Zhang et al, 2011 Jin et al, 2011
7 2 9
853 vs 652 176 vs 171 1,291 vs 1,048
4.8%* (0.11e0.015) 0.23 (0.05e1.04) 0.274 (0.059e1.031)
NS NS NS
No effect in non-cirrhotic patients Normal ALT excluded Effect not shown in Europeans All cirrhotic patients
*¼means “risk difference” used in ref Miyake and Camma, while the other references used HCC relative risk.
774
M. Colombo, M. Iavarone / Best Practice & Research Clinical Gastroenterology 28 (2014) 771e781
(p < 0.031). Despite some caveats related to pooling clinical outcomes with a tumor end point that would require a separate assessment to guide anticipated interruption of a randomized study when results had crossed the predefined boundary for showing efficacy, the Asia Pacific study on lamivudine is universally taken as evidence of antiviral agents being able to prevent deterioration of chronic hepatitis B, a fact that for ethical reasons discouraged further controlled studies in the field. This notwithstanding, a most intriguing finding of the study in Hong Kong, were the patients who, despite achieving a virological response to lamivudine, ultimately went on developing a HCC, which was in fact the only complication arising in virological responders (Table 2). Other findings added to the debate of HCC arising in patients responding long term to NUCs. In a systematic review of studies of NUCs treatment of patients with HBV it was clearly defined that HCC was prevented in patients with chronic hepatitis but not in those with cirrhosis, and in general in patients that could not achieve complete virological suppression [16]. A recent cohort study from Greece confirmed that cirrhotic patients longterm responding to lamivudine remained at risk of developing liver cancer [17]. However, all these studies enrolled patients treated with lamivudine or rescued with adefovir, i.e. regimens characterized by limited potency and low to moderate genetic barrier, which are not recommended any more by International guidelines to treat patients with chronic hepatitis B in general, and especially in patients with compensated cirrhosis. Based on the assumption that the use of more potent NUC might result in a stronger prevention of HCC, interest has focused on HCC incidence in patients treated long term with entecavir or tenofovir through the reanalysis of observational studies either with a single arm or with a historical comparison group. In Japan, the cumulative five-year HCC incidence in a population treated with entecavir was significantly lower than in patients unexposed to treatment (adjusted HR ¼ 0.37; p ¼ 0.03) [20], this time the reduction being more than four times greater in cirrhotic patients than in non-cirrhotics, a difference that faded in non cirrhotic patients. Such positive outcomes in cirrhotics were confirmed by an observational study in Hong Kong, this time in a mixed population of NUC-naïve and NUC-experienced (30%) patients (HR 14% vs 26%) [21], which however found no significant differences in HCC incidence between entecavir-treated and untreated patients in the overall population. Another study in Japan [22] found no treatment-related differences in HCC incidence in unselected patients, but once more HCC incidence was reduced by NUC therapy in cirrhotics (3% vs 11%, p < 0.01) It should however be annotated that all these studies were underpowered to assess incidence of HCC in patients without cirrhosis, thereby making it necessary to conduct larger studies with longer follow-up to detect potential benefits by treatment in non-cirrhotic patients. However, the hope that more potent NUC like entecavir and tenofovir, might finally prevent liver cancer in responders with cirrhosis, rapidly faded away when a multicenter study in Italy in patients with compensated cirrhosis who had persistently undetectable serum HBV-DNA during four years of entecavir monotherapy, showed substantial annual rates of neoplastic transformation of the liver on the same order of magnitude as in untreated historical controls [19,22]. To circumvent a biased interpretation of tumor chemoprevention by NUC in underpowered studies, HCC incidence observed in treated patients was compared to that predicted by a risk calculator. By combining virus and liver disease variables three scoring systems for HCC risk have been developed that might be used for pretreatment stratification of HBV patients [23e25], yet these models (Table 3) need to be validated in patients of other ethnicities and geographical origin. The REACH-B score was constructed from a database of the Risk Evaluation of Viral Load Elevation and Associated Liver Disease (REVEAL) cohort in Taiwan and was subsequently validated in other independent sets of Asian patients [25], whereas the scoring systems CU-HCC and GAG-
Table 2 Rates of hepatocellular carcinoma in naïve patients with advanced liver fibrosis or cirrhosis long-term treated with NUC [15e19]. Author, year
Study
Continent
NUC
Patients
HCC/year
Liaw, 2004 Papatheodoridis, 2010 Papatheodoridis, 2011 Kurokawa, 2012 Lampertico, 2013
RCT Review R R CP
Asia Asia/Europe Europe Japan Europe
Lamivudine Lamivudine Lamivudine Lamivudine Entecavir
211 w/o resistance 81 responders 62 responders 42 responders 164
1.5% 2.4% 2.5% 2.8% 2.5%
RCT ¼ randomized controlled trial; R ¼ retrospective study; CP ¼ cohort prospective study.
M. Colombo, M. Iavarone / Best Practice & Research Clinical Gastroenterology 28 (2014) 771e781
775
Table 3 Comparison of risk scoring systems for hepatocellular carcinoma in chronic hepatitis B.
Number of patients Place of development Age (years) HBeAg-negative (%) Cirrhosis (%) Follow-up (years) Antiviral therapy (%) HCC (n,%) Scoring system
CU-HCC (Wong et al) [23]
GAG-HCC (Yuen et al) [24]
REACH-B (Yang et al) [25]
1005 Hong Kong 48a Not reported 38.1 9.94b 15.1 105, 10.4 Variable Age >50 years Albumin1.1 mg/dL
820 Hong Kong 40.6b 56.6 15.1 5.62b 0 40, 4.9 Variable Age Per year Male BCP Mutation
3584 Taiwan 45.7a 84.8 0 12.0b 0 131, 3.7 Variable Age Per 5 years over 30 Male ALT U/L 15/44 45 HBeAg-positive HBV-DNA
Contents lists available at ScienceDirect
Best Practice & Research Clinical Gastroenterology
2
Role of antiviral treatment for HCC prevention Massimo Colombo, Prof. *, Massimo Iavarone, MD, PhD “A. M. and A. Migliavacca” Center for Liver Disease, Division of Gastroenterology and Hepatology, degli Studi di Milano, Milan, Italy Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Universita
a b s t r a c t Keywords: Hepatocellular carcinoma HBV HCV Interferon Nucleoside analogues
Chemoprevention of hepatocellular carcinoma (HCC) is listed as a yet highly debated long-term benefit of a successful treatment of patients with chronic viral hepatitis. In the hepatitis B virus (HBV) arena, the retrospective scrutiny of both interferon and nucleos(t) ide analogues (NUC) studies failed to provide robust evidence for HCC chemoprevention, due to a number of confoundings in the studies that were originally designed to assess the antiviral activity of interferon therapy. However, the reanalysis of outcomes following patients stratification for risk factors of HCC, provided a clue to find an association between NUC therapy and a reduced risk of liver cancer in non cirrhotic patients, only. In the hepatitis C scenario, a meta analysis of 30 observational studies of patients treated with interferon demonstrated a more than 70% reduction of HCC risk occurring independently of severity of underlying liver fibrosis which was less pronounced in aged patients and those with more advanced liver fibrosis. While the reasons for the residual risk of HCC in virological responders remain largely unexplained, international societies recommend surveillance for HCC of both HBV and HCV responders to antiviral therapy. © 2014 Elsevier Ltd. All rights reserved.
Introduction Prevention of population exposure to the etiologic agents of liver disease stands as the most effective, practical approach to reduce liver-related mortality from hepatocellular carcinoma (HCC), worldwide [1]. This is the clear message conveyed by the program of mass vaccination of newborns
* Corresponding author. Division of Gastroenterology and Hepatology, Fondazione IRCCS Ca' Granda Ospedale Maggiore degli Studi di Milano, Via F. Sforza 35, 20122 Milan, Italy. Tel.: þ39 0255035432; fax: þ39 0250320410. Policlinico, Universita E-mail address: [email protected] (M. Colombo).
http://dx.doi.org/10.1016/j.bpg.2014.07.017 1521-6918/© 2014 Elsevier Ltd. All rights reserved.
772
M. Colombo, M. Iavarone / Best Practice & Research Clinical Gastroenterology 28 (2014) 771e781
against hepatitis B virus (HBV), the dominant risk factor for HCC on a global scale, that was started in early eighties in Taiwan and now is in place in more than 180 WHO nations [2]. Owing to the significant contribution of HBV and hepatitis C virus (HCV) to HCC and the availability of effective antiviral regimens to permanently suppress or remove etiologic agents, antiviral treatments of patients chronically infected by hepatitis viruses stand as a complementary to primary prophylaxis strategy in the prevention of HCC, and this is even more so in the HCV scenario where a sterilizing vaccine is far from being available [1]. However, the evidence that patients with chronic hepatitis B and hepatitis C can effectively be protected against HCC risk by treatment with interferon and/or oral anti-HBV analogues is rather controversial, due to the lack of randomized controlled trials that are ideally needed to establish efficacy, but are logistically and ethically challenging. Reanalysis of studies with antivirals suggest that virus-induced HCC was more likely to be prevented in younger patients with mild liver inflammation rather than in older patients with advanced liver fibrosis or cirrhosis, who in fact are those at higher risk of developing liver cancer [3]. While this calls for a reassessment of current strategies of patient prioritization to antiviral therapies, that are mostly dictated by cost-utility criteria and therefore target the most in need patients with advanced liver disease, we became progressively aware that uncertainty regarding rates and pattern of HCC chemoprevention by antiviral regimens is mainly the consequence of methodological flaws generated by the retrospective scrutiny of the literature. As these studies were originally designed to evaluate the antiviral potency of interferon and oral analogues, treatment outcomes could not properly be weighed for important predictors of liver cancer like hepatitis severity and comorbidities, both common causes of interferon ineligibility, and life style factors like alcohol abuse and tobacco smoking [4]. This was particularly true for patients with chronic hepatitis C, who until a few months ago were bound to receive treatment with interferon-based regimens only, which may suffice to cause permanent virus sterilization in susceptible patients, depending on the infecting genotype, severity of associated liver disease, presence of co-morbidities and genetically determined sensitivity to interferon, but remains of limited applicability and efficacy in patients with cirrhosis and comorbidities who are closest to develop an HCC [4]. As a matter of fact, treatment ineligibility of patients with severe comorbidities and those with advanced or decompensated HCV might have caused two competing risk factors of shortened survival, like HCC and all cause mortality to appear reduced in the selected cohorts of HCV patients who could achieve a sustained virological response (SVR) to interferon therapy [5,6]. This could no longer be the case once safe, universally applicable and highly effective oral regimens against HCV become available to treat patients not only with decompensated hepatitis but also those at risk of death for extrahepatic complications, provided however that antiviral therapies will reach all the infected population.
Primary prevention Primary prevention of HCC not only means prevention of population exposure to the hepatitis B virus (HBV) and hepatitis C virus (HCV) infection, but it also includes all approaches aiming to interrupt or prevent virus-induced liver damage. There are, however, major differences in the primary prophylaxis of HCC in relation with aetiology, beginning with the lack of a sterilizing vaccine against HCV to finish with differences in the outcome of antiviral therapy which leads to universal virus eradication in hepatitis C responders to interferon and to permanent suppression, but not eradication of the virus, in patients with chronic hepatitis B who successfully respond to either interferon or nucleos(t)ide analogs (NUC) therapy [3,4]. In the latter patients, in the face of a clinically silent persistence of virus replication leading to tapering down of hepatic inflammation along with reversal of HBV induced hepatic fibrosis, the same virological response does not suffice to prevent cancer, since anti-HBV therapy fails to remove the potentially carcinogenetic sequences of HBVDNA integrated into liver cell chromosomes [7]. Conversely, the virus sterilizing effects of anti-HCV therapy in sustained responders is the consequence of virus inability to integrate into human DNA sequences whereas suppression of intracytoplasmic virus replication allows for the restoration of cell innate immunity resulting in permanent clearance of infection [8]. As a consequence, a proportion of responders experience a substantial reduction of all cause and liver specific mortality, possibly due to interruption of virus induced liver fibrogenesis, with evidence of cirrhosis reversal in some patients [5,6]. These
M. Colombo, M. Iavarone / Best Practice & Research Clinical Gastroenterology 28 (2014) 771e781
773
clinical benefits may not reach those patients with chronic hepatitis C who also suffer co morbidities linked to diabetes, overweight and alcohol abuse. Hepatitis B HCC may occur at any clinical stage of HBV infection, yet tumor incidence prevails in patients with higher viremia and those with advanced liver fibrosis compared to low viremic patients and those with an early infection [9,10]. This notwithstanding, cancer risk appears to be modulated by several other virus and non virus-related factors like genotype C of HBV, mutations of pre-S, enhancer-H and core promoter, increased age, male sex, alcohol abuse and tobacco smoking, which in part account for the epidemiological and clinical heterogeneity of the tumor worldwide [11e13]. There is, however, mounting evidence that clinical heterogeneity of HCC may in part be accounted for by genetic diversity, too. While the pathogenetic mechanisms of HBV-related tumors are mainly linked to inflammatory reactions causing liver cells to transform into a mitogenic and mutagenic environment, yet virus integration within the liver cell DNA results in a number of molecular events eventually causing the activation of cell oncogenes and inactivation of tumor suppressor genes, all working to boost neoplastic transformation of the infected liver cell [7]. All in all, while these pathogenetic mechanisms of HBVrelated HCC may in part explain how genetic diversity influences the onset of HCC, they also sustain the hypothesis that liver carcinogenesis may be prevented or attenuated by pharmacological inactivation of HBV replication. Yet, systematic analysis of studies of interferon therapy of patients with chronic hepatitis B have provided conflicting evidenced for HBV-related HCC chemoprevention by interferon therapy going forward. Indeed, the rates of the tumor in treated patients was found to be reduced in three meta analyses whereas it appeared to be unchanged in other three, a chasm that can be explained by a difficulty of extrapolating HCC chemoprevention through the retrospective scrutiny of the studies that were originally designed to assess antiviral efficacy of interferon therapy. Studies outcomes, in fact, were biased by the lack of pretreatment patient stratification for relevant cancer predictors and exclusion of patients at higher risk of developing HCC like those unfit to receive interferon owing to advanced hepatitis. To some extent, the re-analysis of these studies was also biased by the lack of a separate analysis of treatment outcomes between sustained responders and non responders, who represent a majority of all patients with chronic hepatitis B receiving interferon (Table 1). Working against the constraints due to poor applicability and tolerability of interferon based regimens, was the advent of user-friendly oral nucleos(t)ide analogs (NUC) that allowed for a broader and safe access of patients to effective anti-HBV therapy. A randomized controlled study in patients with advanced fibrosis (Ishak stage 4e6) from Hong Kong and the Pacific area, provided evidence that response to the first generation nucleoside analogue lamivudine delays progression of HBV-related liver disease as it was cumulatively assessed in terms of increased Child Pugh score, liver failure or development of HCC [15]. While the clinical benefits were maintained in patients with a persistent virological response, in patients who experienced lamivudine resistance the incidence rate of clinical endpoints including HCC was higher (11% vs 5%) than in the placebo treated patients at month 32 Table 1 Meta-analysis studies of risk of hepatocellular carcinoma (HCC) in patients with chronic hepatitis B who received interferon therapy [14]. Authors
No. Studies
No. treated vs controls
HCC relative risk/risk difference* (95% CI)
p Value
Comments
Sung et al, 2008
12
1292 vs 1458
0.66 (0.48e0.89)
0.006
Yang et al, 2009 Miyake et al, 2009
11 8
1006 vs 1076 553 vs 750
0.59 (0.43e0.81) 5.0%* (9.4e0.5)
0.001 0.028
Camma et al, 2001 Zhang et al, 2011 Jin et al, 2011
7 2 9
853 vs 652 176 vs 171 1,291 vs 1,048
4.8%* (0.11e0.015) 0.23 (0.05e1.04) 0.274 (0.059e1.031)
NS NS NS
No effect in non-cirrhotic patients Normal ALT excluded Effect not shown in Europeans All cirrhotic patients
*¼means “risk difference” used in ref Miyake and Camma, while the other references used HCC relative risk.
774
M. Colombo, M. Iavarone / Best Practice & Research Clinical Gastroenterology 28 (2014) 771e781
(p < 0.031). Despite some caveats related to pooling clinical outcomes with a tumor end point that would require a separate assessment to guide anticipated interruption of a randomized study when results had crossed the predefined boundary for showing efficacy, the Asia Pacific study on lamivudine is universally taken as evidence of antiviral agents being able to prevent deterioration of chronic hepatitis B, a fact that for ethical reasons discouraged further controlled studies in the field. This notwithstanding, a most intriguing finding of the study in Hong Kong, were the patients who, despite achieving a virological response to lamivudine, ultimately went on developing a HCC, which was in fact the only complication arising in virological responders (Table 2). Other findings added to the debate of HCC arising in patients responding long term to NUCs. In a systematic review of studies of NUCs treatment of patients with HBV it was clearly defined that HCC was prevented in patients with chronic hepatitis but not in those with cirrhosis, and in general in patients that could not achieve complete virological suppression [16]. A recent cohort study from Greece confirmed that cirrhotic patients longterm responding to lamivudine remained at risk of developing liver cancer [17]. However, all these studies enrolled patients treated with lamivudine or rescued with adefovir, i.e. regimens characterized by limited potency and low to moderate genetic barrier, which are not recommended any more by International guidelines to treat patients with chronic hepatitis B in general, and especially in patients with compensated cirrhosis. Based on the assumption that the use of more potent NUC might result in a stronger prevention of HCC, interest has focused on HCC incidence in patients treated long term with entecavir or tenofovir through the reanalysis of observational studies either with a single arm or with a historical comparison group. In Japan, the cumulative five-year HCC incidence in a population treated with entecavir was significantly lower than in patients unexposed to treatment (adjusted HR ¼ 0.37; p ¼ 0.03) [20], this time the reduction being more than four times greater in cirrhotic patients than in non-cirrhotics, a difference that faded in non cirrhotic patients. Such positive outcomes in cirrhotics were confirmed by an observational study in Hong Kong, this time in a mixed population of NUC-naïve and NUC-experienced (30%) patients (HR 14% vs 26%) [21], which however found no significant differences in HCC incidence between entecavir-treated and untreated patients in the overall population. Another study in Japan [22] found no treatment-related differences in HCC incidence in unselected patients, but once more HCC incidence was reduced by NUC therapy in cirrhotics (3% vs 11%, p < 0.01) It should however be annotated that all these studies were underpowered to assess incidence of HCC in patients without cirrhosis, thereby making it necessary to conduct larger studies with longer follow-up to detect potential benefits by treatment in non-cirrhotic patients. However, the hope that more potent NUC like entecavir and tenofovir, might finally prevent liver cancer in responders with cirrhosis, rapidly faded away when a multicenter study in Italy in patients with compensated cirrhosis who had persistently undetectable serum HBV-DNA during four years of entecavir monotherapy, showed substantial annual rates of neoplastic transformation of the liver on the same order of magnitude as in untreated historical controls [19,22]. To circumvent a biased interpretation of tumor chemoprevention by NUC in underpowered studies, HCC incidence observed in treated patients was compared to that predicted by a risk calculator. By combining virus and liver disease variables three scoring systems for HCC risk have been developed that might be used for pretreatment stratification of HBV patients [23e25], yet these models (Table 3) need to be validated in patients of other ethnicities and geographical origin. The REACH-B score was constructed from a database of the Risk Evaluation of Viral Load Elevation and Associated Liver Disease (REVEAL) cohort in Taiwan and was subsequently validated in other independent sets of Asian patients [25], whereas the scoring systems CU-HCC and GAG-
Table 2 Rates of hepatocellular carcinoma in naïve patients with advanced liver fibrosis or cirrhosis long-term treated with NUC [15e19]. Author, year
Study
Continent
NUC
Patients
HCC/year
Liaw, 2004 Papatheodoridis, 2010 Papatheodoridis, 2011 Kurokawa, 2012 Lampertico, 2013
RCT Review R R CP
Asia Asia/Europe Europe Japan Europe
Lamivudine Lamivudine Lamivudine Lamivudine Entecavir
211 w/o resistance 81 responders 62 responders 42 responders 164
1.5% 2.4% 2.5% 2.8% 2.5%
RCT ¼ randomized controlled trial; R ¼ retrospective study; CP ¼ cohort prospective study.
M. Colombo, M. Iavarone / Best Practice & Research Clinical Gastroenterology 28 (2014) 771e781
775
Table 3 Comparison of risk scoring systems for hepatocellular carcinoma in chronic hepatitis B.
Number of patients Place of development Age (years) HBeAg-negative (%) Cirrhosis (%) Follow-up (years) Antiviral therapy (%) HCC (n,%) Scoring system
CU-HCC (Wong et al) [23]
GAG-HCC (Yuen et al) [24]
REACH-B (Yang et al) [25]
1005 Hong Kong 48a Not reported 38.1 9.94b 15.1 105, 10.4 Variable Age >50 years Albumin1.1 mg/dL
820 Hong Kong 40.6b 56.6 15.1 5.62b 0 40, 4.9 Variable Age Per year Male BCP Mutation
3584 Taiwan 45.7a 84.8 0 12.0b 0 131, 3.7 Variable Age Per 5 years over 30 Male ALT U/L 15/44 45 HBeAg-positive HBV-DNA
