Goals of Treatment in the Patient With Chronic HBV Infection CME
The ultimate goals of treatment for chronic hepatitis B virus (HBV) infection are indisputable — prevention of the life-limiting outcomes of liver failure and liver cancer. These goals are not, however, practical treatment goals, because liver failure and liver cancer are the results of decades of chronic infection. Rather than rely on these clinical endpoints, HBV treatment trials have relied on surrogate endpoints that reflect viral replication and liver disease activity. Importantly, appropriate surrogate endpoints should be selected based on evidence that they do indeed reflect improvement in clinical outcomes for patients.
Improvement in serum liver enzymes (alanine aminotransferase [ALT] and aspartate aminotransferase [AST]) and liver histopathological findings, while used in some clinical trials as surrogate endpoints for successful HBV treatment, are not valid endpoints to be used in the clinical setting. The factors that prevent the use of these 2 potential endpoints in the clinic are discussed below. Appropriate surrogate endpoints for hepatitis B treatment include:
- Suppression of HBV DNA;
- Hepatitis B e antigen (HBeAg) seroconversion; and
- Hepatitis B surface antigen (HBsAg) loss.
Data supporting the use of these endpoints and practical aspects of applying these endpoints to the clinical care of patients follow.
What’s Not Useful? Liver Enzymes and Liver Histopathology
One of the first improvements noted by the clinician after the initiation of antiviral therapy is a reduction in serum ALT and AST. After 1 year of pegylated interferon (PEG-IFN)-based or nucleos(t)ide analog (NUC) therapy, ALT normalization occurs in 34%-51% and 41%-78% of patients, respectively.[1-10] After 5 years, 66%-80% of patients on NUC therapy demonstrate normalization of serum ALT.[11-14] Regardless of the end-of-treatment reduction in ALT level, ALT improvements were not durable. Follow-up studies demonstrate that ALT remained normal in only 32%-60% and 32%-49% of patients 6-12 months after discontinuation of PEG-IFN or NUC therapy, respectively.[1,2,14-16] Reduction in serum ALT is usually among the first laboratory improvements noticed in patients receiving antiviral therapy, preceding complete viral load suppression or serologic conversions. Therefore, stopping therapy based on ALT normalization alone, without consideration of other factors, will usually result in virologic and biochemical rebound.
Another factor abrogating the use of serum ALT as a surrogate endpoint is controversy regarding the definition of normal ALT level. Hepatology experts generally agree that the true upper limit of normal (ULN) ALT for men should be 30 IU/L for men and 19 IU/L for women. Many clinical laboratories will, however, use a ULN level as high as 79 IU/L for men and 49 IU/L for women. A study of 3233 patients with chronic HBV infection demonstrated that patients with ALT levels 1 or 2 times ULN and even 0.5-1 times ULN experienced more liver complications than those with an ALT level less than 0.5 times ULN. Uncertainty regarding the correct ULN of this biochemical marker renders serum ALT as an endpoint for treatment response not easily applied in the clinic setting.
Improvement in liver histopathology is an attractive surrogate endpoint because it directly measures liver disease activity, and thus should be linked to liver disease outcomes, particularly cirrhosis. Antiviral therapy has been shown not only to prevent progression of liver disease but also to reverse fibrosis. No study has examined the predictive value of this reduction in fibrosis in terms of other virologic and serologic outcomes, or in terms of long-term clinical outcomes. In addition, no study has examined the durability of fibrosis reduction after treatment is discontinued. Given the lack of data and the risks involved with liver biopsy, on-treatment liver biopsy is not a practical means of measuring treatment response.
Suppression of HBV DNA
The association between serum HBV DNA levels and clinical outcomes of chronic HBV infection was emphasized by data derived from the REVEAL study, a large-scale, long-term prospective study of a population-based cohort of Taiwanese adults with chronic HBV infection. These data demonstrated that both risk for progression to cirrhosis and risk for development of hepatocellular carcinoma (HCC) were associated with elevated serum HBV DNA levels.[20,21]
The effect of treatment-induced reduction in HBV DNA levels on improving clinical outcomes has been examined in patients with advanced liver fibrosis and cirrhosis. In a large randomized trial comparing the efficacy of lamivudine to that of placebo, patients assigned to lamivudine had better clinical outcomes, including reduced rates of progression in Child-Turcotte-Pugh score, variceal hemorrhage, spontaneous bacterial peritonitis, or HCC. Another study demonstrated lower rates of HCC and mortality in patients with cirrhosis who maintained viral suppression on lamivudine and maintained viral suppression, compared with those who had virologic breakthrough.
Clinical endpoints are difficult to measure in the absence of cirrhosis, but virologic response has been shown to be associated with histologic outcome. An analysis of 26 HBV treatment trials found that a 1-log reduction in median serum HBV DNA level corresponded with a 2-point decrease in median histological grading, suggesting that HBV DNA reduction during antiviral therapy is a good surrogate for histological improvement.
While the benefits of suppressing HBV DNA are well supported by these data, the application of this endpoint in clinical practice has not been well defined. PEG-IFN and NUC agents have been shown effective in suppressing HBV replication, but only a small percentage of patients maintain virologic suppression after treatment is discontinued. Emerging data suggest that the rapidity of viral load suppression once antiviral therapy is initiated may be the most important virologic parameter to guide decisions about treatment endpoints. The GLOBE study comparing telbivudine with lamivudine found that an undetectable HBV DNA level (< 300 copies/mL) at week 24 was associated with an HBeAg seroconversion rate of 41% at 1 year, compared with 4% in those with HBV DNA levels greater than 10,000 copies/mL at 24 weeks. Another study evaluating the predictive value of HBV DNA levels at various points during lamivudine therapy found that HBV DNA levels at week 4 (< 2000 IU/mL) and week 16 (< 800 IU/mL) were the best predictors of a combined endpoint of HBeAg seroconversion, ALT normalization, and absence of lamivudine-resistant mutations after 5 years of treatment. Viral kinetics also proved important during treatment with PEG-IFN. In a study of 66 patients receiving PEG-IFN and lamivudine, HBV DNA suppression at week 8 was a predictor of sustained virologic response, defined as loss of HBeAg, anti-HBe seroconversion, and HBV DNA < 10,000 copies/mL at 1 year. In contrast, patients without early virologic response were highly unlikely to experience a sustained virologic response, with a negative predictive value of 92%.
These data support the use of HBV DNA levels to make clinical decisions for patients with chronic HBV infection on antiviral therapy. For patients with cirrhosis or advanced fibrosis, antiviral treatment is lifelong, as the benefit of continued virologic suppression outweighs the risk for virologic rebound and biochemical flare if treatment is withdrawn. For these patients, the treatment endpoint is complete virologic suppression (generally defined as < 60 IU/mL), but treatment is never discontinued unless viral resistance develops or the patient is unable to tolerate medication.
For patients with chronic HBV infection without advanced fibrosis or cirrhosis, HBV DNA is not used as the single parameter to define treatment success but is considered in the context of serologic parameters described below.
HBeAg seroconversion remains a valid clinical endpoint in patients who are positive for HBeAg. In an early study of IFN-alpha, survival until liver transplantation or death and lack of clinical complications was significantly better in treated patients who cleared HBeAg than in patients who did not (P = .004 for survival and P = .018 for absence of clinical complications). This finding has been confirmed by other studies. It is, however, important to emphasize to patients that seroconversion can take years to occur, and it does not occur in all patients. Seroconversion occurs at a rate of approximately 20% of patients per year of antiviral therapy and increases over time.
The validity of HBeAg seroconversion as a treatment endpoint has been questioned, because low levels of HBV DNA persist even after HBeAg seroconversion. Furthermore, HBeAg reversion may occur. In one study, 24% of patients developed HBeAg-negative hepatitis with detectable HBV DNA, and 4.1% developed HBeAg reversion after first experiencing HBeAg seroconversion. Conversely, 92% of study patients who achieved HBeAg seroconversion during lamivudine treatment maintained seroconversion after 5 years. Patients had completed at least 12 months of lamivudine after seroconversion.[33,34] When tSimilar results have been reported with entecavir, telbivudine, and PEG-IFN.[35-37]
HBeAg remains a worthy therapeutic endpoint because of its association with improved clinical endpoints. Most experts recommend treatment of at least 6 months and preferably 12 months after seroconversion with an NUC. After treatment discontinuation, reappearance of low-level HBV DNA almost always occurs. The threshold at which this low-level HBV DNA becomes concerning is not known, although most providers would consider reinstituting antiviral therapy if the HBV DNA level rises above 2000 IU/mL, or if HBeAg reversion occurs.
HBsAg Loss and Decrease in HBsAg Levels
HBsAg loss has long been the desired endpoint of treatment for chronic HBV infection. Prospective studies of patients with chronic HBV infection have reported lower rates of all adverse liver outcomes in patients who lost HBsAg spontaneously, compared with those who did not.[38,39] A study of 218 patients who were followed for a mean of 63 months after spontaneous HBsAg clearance found that none of the patients who were cirrhosis free and HBV monoinfected at the time of HBsAg loss developed cirrhosis or HCC during follow-up. In contrast, the incidence of cirrhosis and HCC was 3.4% and 0.7%, respectively, in those who remained HBsAg positive.
Unfortunately, the rate of HBsAg loss with available HBV treatments is low. For patients undergoing NUC therapy, HBsAg loss has been reported to occur in 0%-2% of HBeAg-positive patients and in less than 1% of patients who are HBeAg negative at the end of 1 year. After 3-5 years of treatment, 2%-8% of HBeAg-positive patients have been reported to have developed HBsAg loss, but the rates in HBeAg-negative patients continue to be discouragingly low (0%-5%).[1-13]
The low rate of HBsAg loss with currently available therapies renders this an impractical endpoint. However, as data emerge and use of new assays increases, decline of quantitative HBsAg levels may become more widely used as a means to guide therapy.
Because the mechanism of action of NUC therapy is to inhibit the reverse transcription of the pregenomic HBV RNA to HBV DNA, these drugs have no effect on the covalently closed circular DNA reservoir within infected hepatocytes. Therefore, the endpoint of viral eradication is not feasible with these drugs.
- For hepatitis B patients with cirrhosis, the endpoint of therapy is prevention of clinical complications, including liver cancer and hepatic decompensation; for these patients, treatment is generally considered lifelong. For patients without cirrhosis, these clinical sequelae take decades to occur, so that the use of surrogate endpoints is necessary to guide treatment for chronic HBV infection.
- For patients who have chronic HBV infection and who are HBeAg positive, HBeAg seroconversion remains a valid endpoint, and the standard of care is to discontinue treatment 6-12 months after hepatitis B seroconversion is achieved. Such patients must still be monitored for the development of HBeAg-negative chronic hepatitis B, or HBeAg reversion.
- For patients with HBeAg-negative chronic hepatitis B, the only clear endpoint is HBsAg seroconversion. Because this endpoint occurs so rarely, many consider the treatment of HBeAg-negative chronic hepatitis B to be lifelong. As further data emerge, the use of viral kinetics or quantitative HBsAg titers to guide duration of therapy may become possible in this challenging group of patients.