| |
Optimizing Management Strategies in Special Patient Populations, part 5
|
| |
| |
The American Journal of Gastroenterology
Volume 101 Page S26 - January 2006
Morris Sherman, M.B. B.Ch., Ph.D., F.R.C.P.(C)
University of Toronto and University Health Network, Toronto, Ontario, Canada
Chronic hepatitis B infection presents a number of challenges to clinicians. There are additional considerations when defining management strategies for individuals with advanced liver disease, or coinfection, or those at high risk of developing hepatocellular carcinoma (HCC). Treatment of decompensated cirrhosis is particularly important. Evidence suggests that suppression of viral replication through nucleos(t)ide analog therapy leads to longer time to transplantation, improved liver function, and improved survival times. The use of interferon in patients with decompensated hepatitis B is associated with serious complications and is currently contraindicated for these patients by the AASLD Practice Guidelines. Hepatitis B coinfection is often associated with more extensive disease. In patients with HBV/HCV coinfection, one disease is usually dominant and consequently should be the focus of therapy. HIV/HBV coinfection increases the risk of progressive liver disease. Therapeutic agents active against both viruses should be utilized at the correct dose to limit the development of resistance. Agents specific for HBV, e.g., entecavir, enable hepatitis to be treated while avoiding the risk of HIV resistance developing. Dual infection with HBV and HDV is particularly challenging. Nucleos(t)ide analogs are ineffective in treating HDV infection, and there is limited data concerning the efficacy of interferon in this setting. The association between chronic hepatitis B infection and hepatocellular carcinoma (HCC) is well established. In patients at high risk of HCC, screening regimes may be effective. Furthermore, there is an increasing body of evidence indicating that effective suppression of viral replication may be associated with a reduced risk of HCC.
INTRODUCTION
The treatment of chronic hepatitis B in the average patient has been dealt with in other articles in this supplement. In this article, I will deal with special situations where additional considerations have to be taken into account in selecting therapy for chronic hepatitis B. These include patients with decompensated liver disease, and patients coinfected with hepatitis C, hepatitis D, or HIV. The discussion will also consider the role of antiviral therapy in prevention of hepatocellular carcinoma (HCC), and will end with a short discussion of screening for HCC. Note that the term "hepatitis B carrier" is used to denote chronic hepatitis B infection regardless of the presence or absence of liver disease.
Treatment of Patients with Hepatic Decompensation
Patients with liver failure resulting from chronic hepatitis B are obviously the hepatitis patients with the greatest need for therapy. There is evidence that effective therapy can either delay the need for transplantation, or even allow patients to be removed from the transplant list altogether. This has been accomplished with interferon and with lamivudine, establishing the principle that suppression of viral replication allows recovery to occur. It is likely that adequate viral suppression with any other drug will be equally effective.
The studies investigating treatment of decompensated hepatitis B all suffer from the same limitations; the sample sizes are small and there is no control group. There are four published series of patients treated with interferon. Hoofnagle et al. studied 18 HBeAg-positive patients who were treated with interferon (1). In six patients (33%) there was biochemical improvement, but five patients (28%) suffered a significant bacterial infection, and in nine there was an exacerbation of disease. Perrillo et al. reported a series of 26 patients treated with a regimen of low-dose interferon, titrated upward as the patient's condition allowed (2). Eleven patients (42%) achieved a sustained loss of HBV DNA (as measured by insensitive hybridization assays) and a reduction in ALT activity. Three patients (12%) developed serious bacterial infections. Survival was improved in the responders, compared to the nonresponders, but since 50% of the responders had Child's A cirrhosis at baseline and none of the nonresponders had Child's A, the contribution of interferon to this improved survival remains uncertain. Furthermore, given the absence of control groups in these two studies, it cannot be determined whether the incidence of infections and other complications were any more frequent than could be expected in any group of patients with liver disease of this severity. Marcellin et al. treated HBV-related decompensated cirrhosis with low-dose interferon for up to 48 months (3). Sustained loss of HBV DNA with normalization of transaminase levels and clinical improvement was achieved in 10 of 15 patients. Seven were long-term survivors. Complications were less frequent, but were severe (one bacterial infection and one variceal bleed). The AASLD Practice Guidelines currently contraindicate the use of interferon in patients with decompensated cirrhosis (4).
Lamividine has also been used in the management of patients with end-stage liver disease. Again, the series are small, with the largest published study including 35 patients. Not all of these studies differentiated between HBeAg-positive and anti-HBe-positive patients, so that HBe antigen seroconversion was not always used as an endpoint. Villenueve et al. showed that lamivudine effectively suppressed viral load (5). This was associated with a slow improvement in liver function in most patients, as measured by a decrease in the Child-Pugh score. Similar findings have been reported in several other studies (6-8). Yao et al. showed that the rate of transplantation, time to transplantation, and survival were all improved in treated patients compared with a matched control cohort (6). Manolakopoulos et al. further showed that emergence of lamivudine resistance was associated with progression of disease and a high risk of death (8). However, not all studies showed improved survival compared with nontreatment. For example, Fontana et al. found no difference in survival in a retrospective review of data from 162 treated and 147 untreated patients (9). However, in this study the inclusion of patients who were HBV DNA-negative probably decreased the potential size of the treatment effect.
Neither adefovir nor tenofovir has been systematically studied in patients with decompensated liver disease. Presumably, however, as long as adequate viral suppression is achieved biochemical and clinical improvement will follow. That said, adefovir is less potent than tenofovir, and less reliably achieves adequate viral suppression (defined as <105 copies/mL in this comparative study) (10). Furthermore, the potential of nephrotoxicity with adefovir may be increased in patients with advanced liver disease, who are already at increased risk of impaired renal function.
Keeffe et al. have suggested that patients with decompensated liver disease should be treated with the objective of reducing HBV DNA to undetectable levels (11). The rationale is that since there is no known "safe" level of HBV DNA, the better the suppression the lower the likelihood of relapse due to the resistant phenotype emerging. Although both lamivudine and adefovir may reduce HBV DNA to undetectable levels, relatively few patients achieve this endpoint. Interferon therapy seldom achieves undetectable viral levels. Therefore nucleoside analogs should be used as first-line therapy for patients with hepatic decompensation because they have a higher rate of normalization of liver function than interferon, with fewer complications. In patients treated with lamivudine, the risk of developing resistance is high and the consequences grave. This would suggest that these patients should be treated with two agents, the current best choices being lamivudine and tenofovir. There are no reports of the efficacy of this combination in patients with decompensated disease, and it is not clear that the combination will produce more effective viral suppression than monotherapy, but given the risk of progressive liver disease and the serious outcomes when lamivudine resistance develops (and, in all probability, if resistance to adefovir/tenofovir were to develop), combination therapy may be more effective than monotherapy in preventing the emergence of resistance.
As more potent nucleoside analogs (entecavir, telbivudine, etc.) become available they may be preferable to less potent agents because they are likely to have a lower rate of resistance. Interferon can also be used, however, given the risks of complications, it may not be the best first-line therapy but can be considered in patients who have failed both lamivudine and adefovir/tenofovir. The regimen should either be low dose and titrated upward as tolerated, or as long-term low-dose therapy.
Hepatitis B HIV Coinfection
Patients coinfected with hepatitis B and HIV are at high risk for progressive liver disease and HCC (12, 13). Furthermore, the HCC seems to progress more rapidly, particularly if the CD4 count is low (14). Thus, as with hepatitis C, there is possibly a greater imperative to treat these patients. However, the fact that many of these patients are also receiving HAART (highly active antiretroviral therapy), or may do so in the future, complicates the choice of therapy.
Interferon is less effective in inducing seroconversion in patients with HIV infection (12, 15, 16). Many of the early interferon studies included patients with HIV and the presence of HIV was a predictor of nonresponse. In one study the response rate to interferon was 12/31 (39%) in the HIV-negative group, compared with 0/9 in the HIV-positive group (12). However, a poor initial, but sustained response rate was seen in patients with low CD4 counts. Both HIV-infected individuals with well-compensated disease and mono-infected patients responded.
The first studies of the effect of nucleos(t)ide analogs on hepatitis B in HIV-positive patients came from studies of HAART in which it was noticed that regimens that included lamivudine (3TC) resulted in lower ALT levels and reduced liver-related mortality (17). Three available nucleos(t)ide analogs have anti-HIV activity, which paradoxically limits their use in HIV/HBV coinfected individuals. The use of lamivudine, adefovir, or tenofovir monotherapy for treatment of hepatitis B in HIV-infected individuals is associated with the development of HIV resistance to these agents. Thus, they should be used as a component of HAART. Furthermore, although adefovir has anti-HIV activity, at the doses used for hepatitis B treatment it has minimal activity against HIV and is not suitable for use in these patients. It should also be noted that exposure to adefovir will likely lead to resistance to tenofovir, since adefovir and tenofovir exhibit cross resistance. The weak anti-HBV activity of adefovir also means that it may be ineffective in patients with very high viral loads, such as those seen in coinfected patients (18). There are several reports that tenofovir is effective against hepatitis B in HIV-positive patients, both in lamivudine-naive patients and in patients who have developed lamivudine resistance (19, 20). (Note from Jules Levin: the utility of tenofovir is addressed in more detail in other parts of this series. In HBv/HIV coinfection Truvada, a fixed-dose combination of tenofovir+FTC once daily, is a popular therapy as it treats both HIV & HBV. Tenofovir studies in HBv shows it is potent & ongoing studies are being conducted for the purpose of establishing tenofovirfs database for HBV & in coinfection. Tenofovir also appears to be an effective choice for patients with lamivudine (3TC) resistance.)
Active hepatitis, as indicated by HBV DNA concentrations of >104 to 105 copies/mL is associated with a high likelihood of progression of disease, even if the ALT elevation is not impressive. This is particularly likely if there is significant fibrosis. These patients will likely benefit from treatment of the hepatitis B infection. An indication for treatment of chronic hepatitis B in coinfected individuals is also an indication for HAART, even if the CD4 count and HIV viral load would not otherwise warrant therapy.
Entecavir can be used as hepatitis B monotherapy, because it has no activity against HIV. A recent study has reported the efficacy of entecavir in coinfected individuals on existing lamivudine therapy (21).
Hepatitis B and Hepatitis C Coinfection
For reasons that are not clear, when HBV/HCV coinfection occurs, usually only one virus actively replicates. Patients coinfected with HBV and HCV tend to have more advanced fibrosis (22), and have a higher incidence of HCC (23). If the patient is HBeAg-positive it is usually the hepatitis B that is active, but if the patient is anti-HBe-positive either the hepatitis B or the hepatitis C may be active. If HBV DNA is detectable and HCV RNA is present in low titer or is undetectable, particularly if the HBV DNA level is >104 to 105 copies/mL, then hepatitis B is the active disease. When the patient is positive for HCV RNA and negative for HBV DNA, hepatitis C is the active disease. Replication of both viruses can occasionally be detected (22, 24).
In these coinfected patients the dominant disease should be treated. Patients with active hepatitis B should be treated in the standard manner. The choice between interferon or a nucleoside analog should be made on the usual grounds, with no concession made to the presence of hepatitis C. The absence of HCV RNA has the same significance in this situation as in patients who are anti-HCV-positive only. The hepatitis C has, for practical purposes, been eradicated, and will not reactivate once the hepatitis B has been treated. Patients with active hepatitis C should be treated in the standard manner with pegylated interferon and ribavirin. What happens to the hepatitis B infection once the hepatitis C has been cleared is not known. However, if it flares, it can be controlled with nucleoside analog therapy.
Treatment of a true HBV/HCV coinfection with high serum concentrations of both viruses is a little more complicated. First, it is not possible to determine whether one disease or the other, or both, are contributing to the hepatic inflammation. Therefore both diseases have to be treated. The simplest strategy is to start therapy with pegylated interferon and ribavirin, and add lamivudine or adefovir, either during therapy if the HBV DNA response is sub-optimal, or after therapy if virologic relapse occurs posttherapy. If the patient is HBeAg-positive, treatment should continue until seroconversion, and if the patient is anti-HBV-positive nucleoside analog treatment should continue indefinitely. In anti-HBe-positive patients, relapse after interferon therapy is likely. Once again, nucleos(t)ide analogs can be instituted on relapse following interferon therapy and continued indefinitely.
Hepatitis B and Hepatitis D Coinfection
The outcome of interferon therapy for hepatitis D is uncertain. Farci et al. reported that 10/14 patients (71%) treated with 9 MIU interferon achieved biochemical improvement and seven patients (50%) cleared HDV RNA (25). The response was sustained in five patients (36%). A subsequent report indicated that long-term survival was improved in patients treated with this regimen, compared with those treated with a lower dose regimen and with untreated patients (26). Others have described similar on-treatment responses, but with virtually 100% posttreatment relapses (after 48 weeks of therapy) (27). Addition of lamivudine to high-dose interferon did not improve the sustained remission rate (28). Nucleoside analogs are ineffective in treating hepatitis D (29, 30). This is probably because the persistence of HBsAg allows the HDV to enter and be released from cells, while therapy does not reduce HDV replication.
Prevention of HCC
Approximately 25% of Asian males with chronic hepatitis B will have a hepatitis B-related cause of death, mainly HCC and, to a lesser extent, cirrhosis and its complications. Therefore, one of the long-term objectives for the treatment of HBV is to prevent HCC. There is now clear evidence that active viral replication and active liver disease correlate with increased cancer risk after years of follow-up (31, 32). Although most patients who develop HCC are anti-HBe-positive, the presence of HBeAg years prior to the development of HCC imparts a greater risk of HCC than that in HBeAg-negative patients (32). Thus, it would seem logical that if hepatic inflammation could be suppressed, cirrhosis prevented, and viral load reduced, the risks for cirrhosis would also decrease, and the incidence of HCC in this population would fall. Thus, it would be expected that effective treatment of chronic hepatitis B infection might reduce the risk of HCC.
Whether interferon therapy for HBV infection reduces the risk of HCC is controversial. Some of the differences in studies from different parts of the world might be explained by the differences in the natural history of infection with different genotypes. Infection with HBV genotype A, for example, is generally associated with milder disease, and with a higher seroconversion rate following interferon therapy. Studies in Europe, where genotype A is more common, have suggested that there is a reduced incidence of HCC following interferon therapy (33). Studies from Taiwan and Hong Kong, however, have produced conflicting results (34, 35). Lin et al. found that interferon therapy reduced the proportion of study subjects who developed HCC from about 15% to 1.5-4%, depending on the type of interferon used (34). However, Yuen et al., who followed a larger cohort over a longer period, found no difference in the HCC rate in patients who were treated with interferon compared with those who did not receive interferon (35). This study was not randomized, but the cohorts were matched.
Liaw et al. were the first to show that cirrhotic hepatitis B carriers taking lamivudine have a reduced incidence of HCC over a median follow-up period of 32 months compared with a placebo-treated group (36). However, those who developed lamivudine resistance were once again at higher risk for HCC. This emphasizes the need for effective therapy in cirrhotic patients.
These studies are proof of principle that if it is possible to reduce hepatic inflammation it should be possible to reduce the risk of HCC.
Screening for Hepatocellular Carcinoma
The risk of developing HCC in hepatitis B carriers is well known. As a result, hepatitis B carriers have been entered into screening programs. However, not all hepatitis B carriers are at similar risk. For example, the risk of HCC in a 20-year-old female is much lower than the risk of HCC in a 60-year-old man. This raises the question of whether all hepatitis B carriers should be screened, or only those at higher risk, and if so, what level of risk warrants screening. Unfortunately, there are no experimental data to indicate at what point it becomes worthwhile to screen a population. Disease models and decision analyses have been used to investigate the value of HCC screening in patients with cirrhosis, or with hepatitis C (37, 38), but there are no published analyses of the same question in hepatitis B. Our analysis (M. Sherman, unpublished data) suggests that if the incidence of HCC exceeds 0.2% per year, screening becomes effective and cost-effective in reducing mortality. This suggests that all hepatitis B carriers in whom the incidence might be expected to be higher than 0.2% per year should undergo screening. These groups include all hepatitis B carriers with cirrhosis and, among noncirrhotic patients, males over 40 years, and women over 50 years. Patients with anti-HBe-positive disease that has been inactive over many years, and who do not have cirrhosis, have a reduced risk of HCC compared to those with active disease, and screening in this population may not be effective. Patients who have responded to therapy for hepatitis B, or those who, although cirrhotic have lost HBsAg, should also undergo screening, because, while the incidence in these patients is reduced, they remain at risk for HCC.
There is a single randomized controlled trial of screening in hepatitis B carriers. Subjects were screened with 6-monthly ultrasound and alpha-fetoprotein (AFP). There was a significant reduction in mortality from HCC in the screened group compared to the control group (39).
Screening should be done by ultrasound at 6-month intervals. Although ultrasonography in the cirrhotic liver is difficult, it should be possible to identify nodules larger than 1 cm, which can then be subject to either biopsy or enhanced follow-up. Smaller nodules do not require any intervention. CT scanning is possibly more sensitive than ultrasound, but is associated with significant amounts of radiation, particularly if patients are to undergo 6-monthly scanning for many years. CT scanning has not been properly investigated as a screening tool, and is not recommended for screening.
AFP screening is too insensitive and insufficiently specific for use as a general screening test, despite the results from the randomized controlled trial described above (40). The only population in which AFP screening has more impressive sensitivity and specificity is in the Alaskan native population (41). Other biochemical tests have not been adequately validated as screening tools at this stage.
There is as yet no good data to suggest that a shorter screening interval results in better outcomes than the longer interval. Indeed, the available evidence suggests that the outcomes are similar, whether patients are screened every 6 or every 12 months (42, 43). However, it can be argued that the likelihood of complete eradication of an HCC by local ablation decreases as the tumor size increases, even in tumors smaller than 2-3 cm (44). Thus, the earlier a tumor is found, the more likely that complete eradication can be achieved, and this may be more likely with 6-monthly screening. This may not be true if the likely treatment is resection or liver transplantation.
It should be noted that the screening interval does not have to be shortened for patients deemed to be at higher risk, since the screening interval depends on the tumor growth rate, not the degree of risk.
Also, not all patients who are at risk for HCC require screening. For example, patients in whom no possible treatment would be offered, e.g., the very elderly or those with other serious medical problems, should not be screened. However, with the ease and low complication rate of local ablative techniques, such as alcohol injection or radiofrequency ablation, fewer and fewer patients have contraindications to some form of therapy.
|
|
| |
| |
|
|
|
