HIV and hepatitis B co-infection in Africa.

210 http://infection.thelancet.com Vol 8 April 2008 The Review by Christopher Hoff mann and Chloe Thio brings to the fore the issue of HIV and hepatitis B virus (HBV) co-infection in resource-poor settings. We support their call for further research in this fi eld. The management of HIV/HBV co-infection in highincome settings involves individualised therapy usually from an expert provider with the support of an array of diagnostic tests. The primary goal of this treatment is to reduce the risk of cirrhosis and hepatocellular carcinoma. However, the public-health approach to antiretroviral therapy (ART) aims to maximise survival at the population level. ART programmes in Africa and other resource-poor settings involve standardised treatment protocols and simplifi ed monitoring to achieve best possible use of available resources. The treatment of hepatitis B is already entwined with the roll-out of ART since lamivudine, a key drug in most fi rst-line HIV treatment combinations, is also active against HBV. Within a public-health approach, recommendations for management need to be shown as cost eff ective, using the best evidence available. It is still unclear whether routine testing for hepatitis B surface antigen (HBsAg) in patients starting ART will prove cost eff ective. In a richer African country with good laboratory infrastructure, such as South Africa, a cost of US$5 per test might be feasible and aff ordable even with 500 000 new HIV infections every year. However, there is some uncertainty as to whether HBsAg is a suffi cient marker, since occult HBV infection (where patients are HBsAg negative but have detectable HBV DNA) could be common in immunosuppressed populations of Africa and Asia. If ongoing work fi nds occult HBV to be important, it might be better for ART programmes in areas of high HBV prevalence to treat all patients similarly, therefore saving the costs of testing. A test for HBV can clearly be justifi ed when it is likely to change patient management—eg, prescribing tenofovir, another drug with dual activity against HIV and HBV, and lamivudine (or emtricitabine) in combination for HBV/HIV co-infected patients. WHO has recently recommended tenofovir in the fi rst-line treatment of HIV in place of stavudine because of better tolerability, ease of once-daily dosing, and reduced long-term toxicity. However, the incorporation of tenofovir into ART programmes is still limited by aff ordability and, in its absence, the benefi t of testing for HBV is less clear. Patients with chronic hepatitis B should perhaps have more intense monitoring, but abnormalities in liver function tests are common both before and after commencing antiretroviral therapy. Monitoring approaches based on clinical assessment may be as eff ective as ones that include laboratory tests (eg, liver function tests), and ongoing studies might provide evidence to guide such recommendations. Individuals who might benefi t from testing for HBV are those who are switching to second-line treatment. A proportion of HBV/HIV co-infected patients will get a fl are of hepatitis if they stop taking HBV-active agents. In ART programmes with fi xed treatment regimens, often the second-line treatment will not contain an HBVactive agent. In the government programme in South Africa, for example, patients switch from stavudine, lamivudine, and efavirenz or nevirapine to zidovudine, didanosine, and lopinavir-ritonavir. If such fl ares are a common fi nding, it could be that continuation of lamivudine is routinely recommended for all patients, or those that are HBsAg positive where tests are available. A disadvantage of using lamivudine as the sole HBVactive agent is that it leads to the rapid development of viral resistance. From a public-health perspective it will be important to know whether this could lead to an epidemic of resistant virus, which could also undermine vaccination eff orts. The incidence of HBV resistance reaches more than 90% after 4 years of lamivudine therapy in coinfected individuals. Resistance is most commonly characterised by mutations in the reverse transcriptase domain of HBV polymerase—the most important mutations being rtL180M and rtM204V/I. The emergence of a triple polymerase mutation (rtL173V, rtL180M, rtM204V) has more recently been documented in HIV/ HBV co-infected individuals on lamivudine therapy, and is recognised as a “vaccine-escape” mutant because of the altered structure of HBsAg and reduced binding to anti-HBs. These mutants can also escape detection by standard HBsAg tests. The resistance patterns emerging in areas of high prevalence and poor resources are not well characterised and should be monitored in the future. If resistance becomes an important issue it will support the case for tenofovir to be made more widely available. Finally, an ethical issue arises from the provision of HBV-active treatment within ART programmes. In HIV and hepatitis B co-infection in Africa