With 60% of the world's population, Asia is home to one of the fastest growing HIV epidemics in the world [1,2]. Most HIV-infected patients here present with an opportunistic infection, most commonly tuberculosis or cryptococcal meningitis (CM), rather than with early disease [3,4], and warrant antiretroviral therapy (ART) by virtue of their usually severe immunosuppression. The optimal time to initiate ART in patients with HIV-associated central nervous system (CNS) opportunistic infections (OI) is unknown. There are concerns that immediate ART may worsen rather than improve outcomes because of combined drug interactions and toxicities or immune reconstitution inflammatory syndrome occurring within the confined space of the skull. However, mortality in HIV-associated tuberculous meningitis (TBM) and CM is unacceptably high [5,6], and delaying ART may result in an increased incidence of HIV-related deaths. Currently, there are no randomized controlled trial data to guide the decision as to the optimal time to commence ART. Clinical guidelines exist for tuberculosis [7–9], but even these acknowledge the limited evidence on which they are based. We conducted a survey of HIV physicians around the world to determine their opinions about initiating ART in HIV-infected patients presenting with CNS OI. A questionnaire containing two case histories of patients who had presented to our unit with HIV-associated TBM and CM was e-mailed to 20 HIV physicians around the world, of whom 12 responded within a specified time period. In the TBM case, all 12 physicians commenced treatment with antituberculous chemotherapy (ATC), the majority (11/12) with a standard four-drug regimen. Five physicians also commenced other drugs for potential drug-resistant tuberculosis. In terms of timing of the initiation of ART, there was a range of responses (2 weeks to 12 months), with 2 months being the most popular response. The majority of respondents chose a combination of zidovudine, lamivudine and efavirenz both in their own country (10/12) and in a developing country setting (5/12). Eleven physicians commenced other medications, such as adjunctive corticosteroids (8/12) or prophylactic cotrimoxazole (6/12). In the CM case, the majority of physicians (10/12) gave amphotericin and flucytosine as the initial antifungal regimen. With regard to the timing of ART, there was a range of responses (0–10 weeks), with 2 weeks being the most common response. Again, the most popular ART regimen in the physicians’ own country was zidovudine, lamivudine and efavirenz (9/12), whereas zidovudine, lamivudine and nevirapine was the most popular regimen (5/12) in a developing country setting. The majority of physicians (9/12) also gave prophylactic cotrimoxazole. Our survey showed that there was good consensus among the respondents in terms of the choice of initial ATC and antifungal therapy. This is hardly surprising as there are a limited number of drugs available and established clinical guidelines to support their decisions. There was also a good consensus with regard to the initial choice of ART in the physicians’ own countries. In the TBM patient this may have been guided by a wish to avoid drugs that interact with rifampicin. In the CM case the uniformity of response was more surprising, given the fewer interactions and wider choice of drugs. In a developing country setting, the ART regimens were more varied, with an increase in the number of nevirapine-containing regimens, presumably reflecting cost concerns. In terms of the optimal timing of the initiation of ART, there was a wide range of responses. In the TBM case, the earliest time that anyone would consider starting ART was at 2 weeks. The majority of physicians would, however, commence ART at 2 months, which coincides with the start of the continuation phase of ATC, when drugs are usually rationalized to dual therapy. The three most common factors influencing the timing of ART were the clinical response to ATC, concerns regarding drug side-effects/interactions, and concerns about immune reconstitution inflammatory syndrome. One respondent gave several answers, ranging from 2 weeks to 12 months, stating ‘the optimal time is uncertain’. The timing of the initiation of ART in the CM case also showed a wide variation of responses, with 2 weeks being the most common response. In some ways this case was less of a management problem than the TBM case, as there are fewer drug interactions and the treatment duration is considerably shorter. There is thus considerable variation in practice, even among experienced HIV physicians, in the management of HIV-infected patients presenting with CNS OI. This reflects the lack of evidence, and prospective data from randomized controlled trials with clinical endpoints are urgently needed. Sponsorship: The Wellcome Trust and the British Infection Society supported this work. M.E. Torok is a Wellcome Trust clinical research fellow, J.N. Day is a British Infection Society fellow, and J.J. Farrar is a Wellcome Trust senior fellow.
Read full abstract
Nov 1, 2005
Christian Manzardo + 5
Open Access