Overwhelming Array of Variables Confounds Interpretation of Co-Infection Studies
Answers Not Forthcoming
Estimates of HCV prevalence among people with HIV in the United States range from 30% to 50%, with even higher rates among people infected through injection drug use. In the U.S., end stage liver disease has become a leading cause of death in people with HIV. Yet national guidelines for the management of hepatitis C infection have conspiculously lagged the current state of the art. A recent government sponsored panel seeks to rectify the situation. Once again, Daniel Raymond attended on behalf of TAG and returned with all the data and dish.
The release on June 12th of a new NIH draft consensus statement, “Management of Hepatitis C: 2002” represents an important milestone in treatment, care, and research in hepatitis C virus (HCV) infection. The consensus statement was drafted by an expert panel chaired by Dr. James Boyer following two days of presentations by researchers with a question and answer period open to the public. The NIH characterizes the consensus statement as an independent report, rather than a policy paper. When the previous consensus statement was released in 1997, alpha interferon monotherapy was the standard of care, as ribavirin had not yet been approved for use in combination therapy. The new statement endorses combination therapy in all patients with biopsy results indicating portal or bridging fibrosis and inflammation and necrosis.
The panel recommends first line therapy using ribavirin in combination with pegylated interferon, a newer formulation requiring only once-weekly dosing and providing an improved sustained virological response (undetectable HCV viral load six months after treatment) approaching 50% in mono-infected patients with genotype 1 and about 80% in patients with genotypes 2 and 3. These response rates are roughly equivalent in studies using ribavirin in combination with Peg-Intron (pegylated interferon alfa-2b, approved by the FDA) and Pegasys (pegylated interferon alfa-2a, submitted for FDA approval), although a comparison study of the two versions has not been conducted.
The 2002 statement also supports treating injection drug users, reversing the 1997 stance of withholding treatment until at least six months of abstinence from drugs. The original statement was guided more by assumption than evidence, which was scant. A letter signed by a broad group of activists, service providers, doctors and researchers supported this revision in the new statement.
Drug users have not always fared so well in the efforts of hepatitis advocates; some established groups have attempted to down play the prevalence of hepatitis C infection among IDUs-as high as to 70-90%, according to some studies-out of concern for associating the stigma of drug use with people suffering from hepatitis C infection. However, an explosion of interest and concern regarding this population in recent years among service providers has led to widespread initiatives promoting education and testing, particularly in HIV outreach and needle exchange programs.
When Is the Optimal Time to Initiate Hepatitis C Treatment in Co-Infected Patients?
Estimates of HCV prevalence among people with HIV in the United States range from 30% to 50%, with even higher rates among people infected through injection drug use. In the U.S., end stage liver disease has become a leading cause of death in people with HIV.
The current state of hepatitis C knowledge — and its limitations — is reflected in the guidance on HIV/HCV co-infection. The consensus statement notes that based on available data, hepatitis C treatment appears safe and reasonably effective in people with HIV. The panel calls for more research into natural history and pathogenesis, treatment safety and outcomes, and optimal dosing and duration of therapy. However, the recommendations fail to address — or, in the current version, acknowledge — the most urgent and vexing question among co-infected people and their doctors in recent years: which should be treated first, HIV or HCV — or both, simultaneously? And when is the optimal time to initiate hepatitis C treatment in co-infected patients? Research indicates that people with a CD4 count below 200 have a significantly increased risk of progression to cirrhosis, while response rates to hepatitis C treatment improve with higher CD4 levels. A liver disease specialist looking to this document for guidance on treating a co-infected patients would not find adequate reference to the influence of CD4 counts on disease progression and treatment response rates.
Hepatitis C Treatment, HIV Treatment and the Risks of Hepatotoxicity From Antiretroviral Medications
Major concerns for co-infected persons and their doctors include balancing decisions about whether and when to start hepatitis C treatment with the risks of hepatotoxicity from antiretroviral medications used to treat HIV. The clinical relevance of antiretroviral hepatotoxicity to risk of hepatitis C disease progression has not been established; however, the incidence of HAART-related hepatotoxicity is significantly higher in people co-infected with hepatitis C (and/or with hepatitis B). Certain medications, including nevirapine and ritonavir, have been linked to a higher rate of hepatotoxicity, but it’s not clear how that should inform decisions about treatment regimens, although frequent monitoring of ALT levels and for symptoms is clearly necessary.
Does Hepatitis C Infection Affect Immune Reconstitution?
A great deal of uncertainty and guesswork still governs the thinking among both patients and clinicians in the treatment of HIV/HCV co-infection. A rapidly increasing number of studies explore a range of questions related to treatment strategies, but they often yield conflicting results, use small sample sizes, have limited follow-up, or lack randomized controls. Comparisons between studies are made difficult due to inconsistent analysis and reporting of variables that may influence results. For instance, as Michael Marco reported in the May 2001 TAGline (volume 8, issue 4), while HIV infection’s tendency to accelerate hepatitis C disease progression has been well documented, it’s less clear whether hepatitis C infection worsens HIV disease.
More recently, a number of studies have examined the effect of hepatitis C infection on immune reconstitution following the initiation of HAART. The touchstone for this debate is the Swiss HIV Cohort Study, which reported that co-infected patients showed a blunted immunological recovery on treatment which could not be attributed to poorer virologic response. A subset analysis of 56 co-infected subjects matched for baseline HIV viral load and CD4 count as well as age and sex found no correlation between HCV viral load and immunologic response, although genotype 3a was significantly associated with CD4 increases below 50 in the first year on potent antiretroviral therapy.
The interpretation of the subset analysis is complicated by the fact that 25% of the co-infected subjects had no detectable HCV viremia, implying a resolved infection. Some subsequent studies have found that despite comparable virologic responses, co-infected patients exhibit a diminished immunologic response to treatment as measured by gains in CD4 cells; others, however, report no immunologic differences.
This question has potential implications for treatment decisions, but a conclusive answer does not seem forthcoming. Some of the confusion relates to differences in study design. Three studies presented this year at the 9th Conference on Retroviruses and Opportunistic Infections illustrate this theme.
The ACTG 383 study (which added a retrospective component to a prospective study when the original trial failed to meet its target enrollment of 60 patients) found no differences in immunological response to HAART, following subjects out to 48 weeks; however, a Spanish prospective cohort study following clinic patients for 24 months found that impaired immunological recovery among co-infected patients becomes evident after the first year of HAART. A Thai analysis from the HIV-NAT trial also reported diminished immune reconstitution among HIV/HCV co-infected patients, although at 48 weeks there were no significant differences in risk of HIV disease progression.
A report from an observational study of the Frankfurt HIV Clinic Cohort also found that co-infected patients experienced CD4 cell increases of lesser magnitude, but qualified by noting that in their sample hepatitis C infection was closely linked to a history of injection drug use, which could be associated with another variable-the authors suggested adherence-that accounted for impaired immunologic response.
How Should These Divergent Results Be Interpreted?
Co-infection studies — particularly those involving HAART — pose a potentially overwhelming array of confounding variables. In hepatitis C research, age, sex, degree of fibrosis, HCV viral load, race, genotype, and alcohol use can influence disease progression and/or response to treatment; various biochemical markers as well as quasispecies diversity may also play a role. HIV studies can incorporate variables including baseline CD4 count and viral load, nadir CD4 count, history of antiretroviral treatment, specific drug classes and regimens, adherence, and other more esoteric virological and host immunological factors such as phenotype and immune activation. None of the studies listed above considers all of these factors, nor could they; the resolution of such debates lies in identifying factors most likely to influence outcomes.
These lines of inquiry may ultimately lead us towards a less monolithic view of co-infection, as understanding of the interaction between the two viruses (and potentially hepatitis B virus and hepatitis G virus) and host responses continues to improve. It would be helpful to explore the possibility of identifying the varying patterns of co-infection based on rate of progression. Accelerated disease progression of one virus, or both might for instance, characterize a particular co-infection pattern, or neither until CD4 counts fall below a certain threshold. Such a classification system would have significant relevance to treatment decisions, but would ultimately require long-range cohort studies to better characterize the natural history of co-infection and the long-term clinical impact of various interventions.