Tertiary Prevention of Hepatitis B-related Hepatocellular Carcinoma: Do We Have a Verdict?
Vincent Wai–Sun Wong,Francis Ka–Leung Chan
DOI: https://doi.org/10.1053/j.gastro.2013.03.038
IF: 29.4
2013-01-01
Gastroenterology
Abstract:Wu CY, Chen YJ, Ho HJ, et al. Association between nucleoside analogues and risk of hepatitis B virus-related hepatocellular carcinoma recurrence following liver resection. JAMA 2012;308:1906–1913.In Asia, hepatitis B virus (HBV) infection is the leading cause of cirrhosis and hepatocellular carcinoma (HCC; Ann Hepatol 2012;11:284–293). HCC is a deadly cancer, so much so that the number of people dying from this cancer is almost the same as the number of newly diagnosed HCCs each year in many countries. This dismal outcome stems from 2 main causes. First, many patients present with advanced HCCs or suffer from decompensated liver disease at the time of diagnosis. In either case, curative treatments such as liver resection and local ablative therapy cannot be offered. This is particularly true in countries where a reliable HCC screening program is not in place. In that case, typically less than half of the patients with HCC can undergo curative treatments (Liver Int 2008;28:79–87). Second, even the lucky few who can receive curative treatments have high risk of recurrence. In 5 years, ≤75% of patients would develop intrahepatic recurrence of HCC (Gastroenterology 1993;105:488–494; J Hepatol 2003;38:200–207; J Hepatol 2007;47:684–690).To prevent recurrence after liver resection, we need to understand HCC recurrence. Traditionally, HCC recurrence is classified into early recurrence and late recurrence using an arbitrary cutoff of 2 years since the time of liver resection. Early recurrence is closely associated with tumor characteristics, including resection margin, number of tumors, vascular invasion, and alpha fetoprotein level (J Hepatol 2003;38:200–207; J Hepatol 2009;51:890–897). In contrast, factors associated with late recurrence more resemble risk factors for the first occurrence of HCC. These include histologic necroinflammation, high HBV DNA, and number of tumors. Because both hepatic necroinflammation and HBV DNA are amenable to treatment, it is of interest to test whether antiviral therapy may prevent HCC recurrence in patients who have undergone liver resection.In a recent article in JAMA, Wu et al examined HCC recurrence in patients with HBV-related HCC using the Taiwan National Health Insurance Research Database (JAMA 2012;308:1906–1913). Among 100,938 incident HCC cases, they included 4569 who had HBV infection, received liver resection between 2003 and 2010, and had follow-up of >3 months. The untreated cohort comprised 4051 patients who did not receive antiviral therapy, and the treated cohort included 518 patients who received oral nucleoside analogs, most being entecavir or lamivudine.The mean follow-up duration was 2.2 ± 1.8 years in the untreated cohort and 2.6 ± 1.7 years in the treated cohort. Overall, 1765 (43.6%) patients in the untreated cohort and 106 (20.5%) patients in the treated cohort had HCC recurrence (hazard ratio; 0.67; 95% confidence interval [CI], 0.55–0.81, after adjusting for competing mortality). Deaths owing to HCC or HCC treatment occurred in 314 (7.8%) patients in the untreated cohort and 20 (3.9%) patients in the treated cohort. By Kaplan-Meier analysis, the 6-year cumulative incidence of HCC recurrence was 54.6% in the untreated cohort and 45.6% in the treated cohort (P < .001). The number needed to treat to prevent 1 HCC recurrence within 6 years was 12 (95% CI, 7.4–22.6). The authors concluded that nucleoside analog use was associated with lower risk of HCC recurrence among patients with HBV-related HCC after liver resection.CommentIn theory, the efficacy of antiviral therapy in preventing HCC recurrence is best tested by a randomized, placebo-controlled trial using overall mortality and HCC recurrence as outcomes. In reality, however, such a trial is almost impossible to conduct. First, oral nucleoside analogs have already been shown to be effective in preventing disease progression and HCC in chronic hepatitis B patients with cirrhosis or advanced liver fibrosis (N Engl J Med 2004;351:1521–1531; Aliment Pharmacol Ther 2008;28:1067–1077), whereas the majority of patients with HCC have underlying cirrhosis. In other words, if one is to design a trial testing the use of antiviral therapy to prevent HCC recurrence, it is only ethical to recruit the small proportion of patients without cirrhosis. Second, HCC recurrence is a serious outcome, whereas oral nucleoside analogs carry few side effects. Because the patent for entecavir has expired, oral nucleoside analogs will also become more affordable. Therefore, few patients with HCC would likely agree to participate in such a trial. As such, observational studies like the report by Wu et al are probably the best evidence we will ever have.Among patients with chronic hepatitis B, high HBV DNA level is strongly associated with future risk of HCC development (JAMA 2006;295:65–73, J Clin Oncol 2010;28:1660–1665). Numerous studies have also confirmed the association between HBV DNA and HCC recurrence in patients receiving liver resection or local ablative therapy (J Hepatol 2003;38:200-207; J Hepatol 2009;51:890–897). High viral load is also associated with increased mortality in HCC patients receiving systemic chemotherapy (Hepatology 2007;45:1382–1389). A number of small, observational studies showed that patients receiving oral nucleoside analogs after liver resection or local ablative therapy have lower risk of HCC recurrence (J Gastroenterol Hepatol 2007;22:1929–1935; Am J Gastroenterol 2008;103:1663-1673; J Gastroenterol 2009;44:991–999). In a meta-analysis of 9 cohort studies, antiviral therapy is associated with a 41% relative risk reduction in HCC recurrence (Aliment Pharmacol Ther 2011;33:1104–1112). This further translates into lower overall and liver-related mortality. However, previous studies were limited by small sample size. Besides, patients who received antiviral therapy likely had significant difference from those who did not.The current study by Wu et al is remarkable for its large sample size. However, for all studies based on patient registries, the method of analysis and limitations should be examined carefully (JAMA 2012;308:1906–1913). First, unlike prospective cohort studies, many important clinical parameters are often missing from the registry databases. In the current study, for example, liver function parameters, HBV DNA level and tumor characteristics were unavailable. Based on diagnostic coding, only 40%–50% of the study cohort was labeled as having cirrhosis. Compared with other Asian cohorts, this proportion is likely the result of underreporting. The reasons for starting and withhold antiviral therapy are also uncertain. Although one may argue that the treated cohort should be more likely to have cirrhosis and abnormal liver function tests and this is thus a negative bias against the survival benefit in the antiviral group, it is also possible that some patients did not receive treatment because their condition was too poor. Furthermore, the authors did not ascertain the causes of death with case records or death registries. Instead, the primary diagnosis of hospitalization in the 3 months preceding death was assumed to be the cause of death. Nevertheless, because this was a group of sick patients, numerous complications might have developed since the last hospitalization.In addition, survival analysis of the treated and untreated cohorts was performed differently. In the untreated cohort, the baseline was defined as the date of hospitalization for liver resection. In contrast, patients who had received oral nucleoside analogs for <90 days were excluded from the treated cohort. For those who had received oral nucleoside analogs for >90 days, the baseline was defined as the date of starting antiviral therapy. Overall, there were 518 patients in the treated cohort, and 281 were excluded because they received antiviral therapy for <90 days. Because many cases of interruption of antiviral therapy were likely owing to clinical deterioration and/or deaths, the exclusion of these patients might have dramatically affected the results and conclusions.Given the merits and limitations of the study by Wu et al, where should we go from here? Oral nucleoside analogs are well-tolerated and have benefits beyond potential prevention of HCC recurrence. Treated patients have lower risk of hepatitis flares and disease progression. Therefore, the risk-to-benefit ratio strongly favors the prescription of oral nucleoside analogs to chronic hepatitis B patients after curative treatment for HCC. Meanwhile, similar studies with longer duration of follow-up and more detailed clinical data are encouraged. Wu CY, Chen YJ, Ho HJ, et al. Association between nucleoside analogues and risk of hepatitis B virus-related hepatocellular carcinoma recurrence following liver resection. JAMA 2012;308:1906–1913. In Asia, hepatitis B virus (HBV) infection is the leading cause of cirrhosis and hepatocellular carcinoma (HCC; Ann Hepatol 2012;11:284–293). HCC is a deadly cancer, so much so that the number of people dying from this cancer is almost the same as the number of newly diagnosed HCCs each year in many countries. This dismal outcome stems from 2 main causes. First, many patients present with advanced HCCs or suffer from decompensated liver disease at the time of diagnosis. In either case, curative treatments such as liver resection and local ablative therapy cannot be offered. This is particularly true in countries where a reliable HCC screening program is not in place. In that case, typically less than half of the patients with HCC can undergo curative treatments (Liver Int 2008;28:79–87). Second, even the lucky few who can receive curative treatments have high risk of recurrence. In 5 years, ≤75% of patients would develop intrahepatic recurrence of HCC (Gastroenterology 1993;105:488–494; J Hepatol 2003;38:200–207; J Hepatol 2007;47:684–690). To prevent recurrence after liver resection, we need to understand HCC recurrence. Traditionally, HCC recurrence is classified into early recurrence and late recurrence using an arbitrary cutoff of 2 years since the time of liver resection. Early recurrence is closely associated with tumor characteristics, including resection margin, number of tumors, vascular invasion, and alpha fetoprotein level (J Hepatol 2003;38:200–207; J Hepatol 2009;51:890–897). In contrast, factors associated with late recurrence more resemble risk factors for the first occurrence of HCC. These include histologic necroinflammation, high HBV DNA, and number of tumors. Because both hepatic necroinflammation and HBV DNA are amenable to treatment, it is of interest to test whether antiviral therapy may prevent HCC recurrence in patients who have undergone liver resection. In a recent article in JAMA, Wu et al examined HCC recurrence in patients with HBV-related HCC using the Taiwan National Health Insurance Research Database (JAMA 2012;308:1906–1913). Among 100,938 incident HCC cases, they included 4569 who had HBV infection, received liver resection between 2003 and 2010, and had follow-up of >3 months. The untreated cohort comprised 4051 patients who did not receive antiviral therapy, and the treated cohort included 518 patients who received oral nucleoside analogs, most being entecavir or lamivudine. The mean follow-up duration was 2.2 ± 1.8 years in the untreated cohort and 2.6 ± 1.7 years in the treated cohort. Overall, 1765 (43.6%) patients in the untreated cohort and 106 (20.5%) patients in the treated cohort had HCC recurrence (hazard ratio; 0.67; 95% confidence interval [CI], 0.55–0.81, after adjusting for competing mortality). Deaths owing to HCC or HCC treatment occurred in 314 (7.8%) patients in the untreated cohort and 20 (3.9%) patients in the treated cohort. By Kaplan-Meier analysis, the 6-year cumulative incidence of HCC recurrence was 54.6% in the untreated cohort and 45.6% in the treated cohort (P < .001). The number needed to treat to prevent 1 HCC recurrence within 6 years was 12 (95% CI, 7.4–22.6). The authors concluded that nucleoside analog use was associated with lower risk of HCC recurrence among patients with HBV-related HCC after liver resection. CommentIn theory, the efficacy of antiviral therapy in preventing HCC recurrence is best tested by a randomized, placebo-controlled trial using overall mortality and HCC recurrence as outcomes. In reality, however, such a trial is almost impossible to conduct. First, oral nucleoside analogs have already been shown to be effective in preventing disease progression and HCC in chronic hepatitis B patients with cirrhosis or advanced liver fibrosis (N Engl J Med 2004;351:1521–1531; Aliment Pharmacol Ther 2008;28:1067–1077), whereas the majority of patients with HCC have underlying cirrhosis. In other words, if one is to design a trial testing the use of antiviral therapy to prevent HCC recurrence, it is only ethical to recruit the small proportion of patients without cirrhosis. Second, HCC recurrence is a serious outcome, whereas oral nucleoside analogs carry few side effects. Because the patent for entecavir has expired, oral nucleoside analogs will also become more affordable. Therefore, few patients with HCC would likely agree to participate in such a trial. As such, observational studies like the report by Wu et al are probably the best evidence we will ever have.Among patients with chronic hepatitis B, high HBV DNA level is strongly associated with future risk of HCC development (JAMA 2006;295:65–73, J Clin Oncol 2010;28:1660–1665). Numerous studies have also confirmed the association between HBV DNA and HCC recurrence in patients receiving liver resection or local ablative therapy (J Hepatol 2003;38:200-207; J Hepatol 2009;51:890–897). High viral load is also associated with increased mortality in HCC patients receiving systemic chemotherapy (Hepatology 2007;45:1382–1389). A number of small, observational studies showed that patients receiving oral nucleoside analogs after liver resection or local ablative therapy have lower risk of HCC recurrence (J Gastroenterol Hepatol 2007;22:1929–1935; Am J Gastroenterol 2008;103:1663-1673; J Gastroenterol 2009;44:991–999). In a meta-analysis of 9 cohort studies, antiviral therapy is associated with a 41% relative risk reduction in HCC recurrence (Aliment Pharmacol Ther 2011;33:1104–1112). This further translates into lower overall and liver-related mortality. However, previous studies were limited by small sample size. Besides, patients who received antiviral therapy likely had significant difference from those who did not.The current study by Wu et al is remarkable for its large sample size. However, for all studies based on patient registries, the method of analysis and limitations should be examined carefully (JAMA 2012;308:1906–1913). First, unlike prospective cohort studies, many important clinical parameters are often missing from the registry databases. In the current study, for example, liver function parameters, HBV DNA level and tumor characteristics were unavailable. Based on diagnostic coding, only 40%–50% of the study cohort was labeled as having cirrhosis. Compared with other Asian cohorts, this proportion is likely the result of underreporting. The reasons for starting and withhold antiviral therapy are also uncertain. Although one may argue that the treated cohort should be more likely to have cirrhosis and abnormal liver function tests and this is thus a negative bias against the survival benefit in the antiviral group, it is also possible that some patients did not receive treatment because their condition was too poor. Furthermore, the authors did not ascertain the causes of death with case records or death registries. Instead, the primary diagnosis of hospitalization in the 3 months preceding death was assumed to be the cause of death. Nevertheless, because this was a group of sick patients, numerous complications might have developed since the last hospitalization.In addition, survival analysis of the treated and untreated cohorts was performed differently. In the untreated cohort, the baseline was defined as the date of hospitalization for liver resection. In contrast, patients who had received oral nucleoside analogs for <90 days were excluded from the treated cohort. For those who had received oral nucleoside analogs for >90 days, the baseline was defined as the date of starting antiviral therapy. Overall, there were 518 patients in the treated cohort, and 281 were excluded because they received antiviral therapy for <90 days. Because many cases of interruption of antiviral therapy were likely owing to clinical deterioration and/or deaths, the exclusion of these patients might have dramatically affected the results and conclusions.Given the merits and limitations of the study by Wu et al, where should we go from here? Oral nucleoside analogs are well-tolerated and have benefits beyond potential prevention of HCC recurrence. Treated patients have lower risk of hepatitis flares and disease progression. Therefore, the risk-to-benefit ratio strongly favors the prescription of oral nucleoside analogs to chronic hepatitis B patients after curative treatment for HCC. Meanwhile, similar studies with longer duration of follow-up and more detailed clinical data are encouraged. In theory, the efficacy of antiviral therapy in preventing HCC recurrence is best tested by a randomized, placebo-controlled trial using overall mortality and HCC recurrence as outcomes. In reality, however, such a trial is almost impossible to conduct. First, oral nucleoside analogs have already been shown to be effective in preventing disease progression and HCC in chronic hepatitis B patients with cirrhosis or advanced liver fibrosis (N Engl J Med 2004;351:1521–1531; Aliment Pharmacol Ther 2008;28:1067–1077), whereas the majority of patients with HCC have underlying cirrhosis. In other words, if one is to design a trial testing the use of antiviral therapy to prevent HCC recurrence, it is only ethical to recruit the small proportion of patients without cirrhosis. Second, HCC recurrence is a serious outcome, whereas oral nucleoside analogs carry few side effects. Because the patent for entecavir has expired, oral nucleoside analogs will also become more affordable. Therefore, few patients with HCC would likely agree to participate in such a trial. As such, observational studies like the report by Wu et al are probably the best evidence we will ever have. Among patients with chronic hepatitis B, high HBV DNA level is strongly associated with future risk of HCC development (JAMA 2006;295:65–73, J Clin Oncol 2010;28:1660–1665). Numerous studies have also confirmed the association between HBV DNA and HCC recurrence in patients receiving liver resection or local ablative therapy (J Hepatol 2003;38:200-207; J Hepatol 2009;51:890–897). High viral load is also associated with increased mortality in HCC patients receiving systemic chemotherapy (Hepatology 2007;45:1382–1389). A number of small, observational studies showed that patients receiving oral nucleoside analogs after liver resection or local ablative therapy have lower risk of HCC recurrence (J Gastroenterol Hepatol 2007;22:1929–1935; Am J Gastroenterol 2008;103:1663-1673; J Gastroenterol 2009;44:991–999). In a meta-analysis of 9 cohort studies, antiviral therapy is associated with a 41% relative risk reduction in HCC recurrence (Aliment Pharmacol Ther 2011;33:1104–1112). This further translates into lower overall and liver-related mortality. However, previous studies were limited by small sample size. Besides, patients who received antiviral therapy likely had significant difference from those who did not. The current study by Wu et al is remarkable for its large sample size. However, for all studies based on patient registries, the method of analysis and limitations should be examined carefully (JAMA 2012;308:1906–1913). First, unlike prospective cohort studies, many important clinical parameters are often missing from the registry databases. In the current study, for example, liver function parameters, HBV DNA level and tumor characteristics were unavailable. Based on diagnostic coding, only 40%–50% of the study cohort was labeled as having cirrhosis. Compared with other Asian cohorts, this proportion is likely the result of underreporting. The reasons for starting and withhold antiviral therapy are also uncertain. Although one may argue that the treated cohort should be more likely to have cirrhosis and abnormal liver function tests and this is thus a negative bias against the survival benefit in the antiviral group, it is also possible that some patients did not receive treatment because their condition was too poor. Furthermore, the authors did not ascertain the causes of death with case records or death registries. Instead, the primary diagnosis of hospitalization in the 3 months preceding death was assumed to be the cause of death. Nevertheless, because this was a group of sick patients, numerous complications might have developed since the last hospitalization. In addition, survival analysis of the treated and untreated cohorts was performed differently. In the untreated cohort, the baseline was defined as the date of hospitalization for liver resection. In contrast, patients who had received oral nucleoside analogs for <90 days were excluded from the treated cohort. For those who had received oral nucleoside analogs for >90 days, the baseline was defined as the date of starting antiviral therapy. Overall, there were 518 patients in the treated cohort, and 281 were excluded because they received antiviral therapy for <90 days. Because many cases of interruption of antiviral therapy were likely owing to clinical deterioration and/or deaths, the exclusion of these patients might have dramatically affected the results and conclusions. Given the merits and limitations of the study by Wu et al, where should we go from here? Oral nucleoside analogs are well-tolerated and have benefits beyond potential prevention of HCC recurrence. Treated patients have lower risk of hepatitis flares and disease progression. Therefore, the risk-to-benefit ratio strongly favors the prescription of oral nucleoside analogs to chronic hepatitis B patients after curative treatment for HCC. Meanwhile, similar studies with longer duration of follow-up and more detailed clinical data are encouraged.