Editorial Commentary: Cryptococcosis in AIDS: New Data but Questions Remain.

Abstract

Cryptococcal meningitis (CM) is a common and often devastating disease in areas of high human immunodeficiency virus (HIV) prevalence. In sub-Saharan Africa, CM has become the leading cause of adult meningitis [1–3]. Without treatment, CM is fatal. Treatment mortality rates remain high, up to 70% at 3 months in low-resource settings, even when antiretroviral therapy (ART) is available [4]. Almost all HIV-associated CM could be prevented if HIV-infected patients were diagnosed and started on ART before their CD4+ cell counts fell to <100 cells/µL. Unfortunately, this is not the reality, with ongoing HIV transmission and late presentation commonplace [5]. The number of patients at high risk of developing CM is not decreasing, and other interventions must be considered. Azole prophylaxis for all patients is neither practical nor desirable. Cryptococcal antigenemia has previously been shown to precede clinical cryptococcal disease and to be a sensitive predictor for impending CM without treatment [6]. In this issue of Clinical Infectious Diseases, Longley et al address early mortality among those presenting with advanced HIV by stratifying them into those with and those without detectable cryptococcal antigen (CrAg) using the latex agglutination assay (LA) and the more sensitive lateral flow assay (LFA) in blood, serum, and plasma [7]. In this prospective study, they used antifungals to treat those found to be CrAg positive.ThosealreadyfoundtohaveCMupon cerebrospinal fluid (CSF) examination were treated with amphotericin; those without proven CM but who were CrAg positive were treated with oral fluconazole. All patients were to be put on ART within 2–4 weeks after CrAg screening. The authors and study team must be commended for running a prospective study and field testing the LFA. Although some of the numbers are small, the results are important. The authors found that LFA testing of blood, serum, or plasma yields equivalent results and is more sensitive, identifying 4-fold more than the traditional LA. Even with the more sensitive LFA, therewas a lower incidence of cryptococcal antigenemia in this antiretroviral-naive, low-CD4 study population than expected (4.3%), which, as the authors state, may represent greater rollout of ART in South Africa and earlier diagnosis of HIV. The authors show that all-cause mortality and cryptococcal-specific mortality were 2-fold higher at 1 year among those screening positive for cryptococcal antigen. Since the number of CM deaths was small, the observable difference in mortality is most likely due to the fact that CrAg positivity is a marker for severe immunocompromise. There were no other clinical or laboratory predictors for a positive serum LFA or for those with positive CSF assays, justifying the screening protocol and use of the more sensitive assay. Upon further review, if those with proven CM are censored from the group that screened positive, mortality rates do not appear markedly different compared with rates for those who screened negative. The success of the study is demonstrated by the fact that the 12-month mortality for those screening positive for CrAg without proven CM was 17%. This approaches the rate for those who were CrAg negative, which was 12%. The inference here is that without antifungals, those screening positive for CrAg would have had higher rates of CM and death when compared with historical control groups. Seven patients were found to be negative by LA but positive by LFA retrospectively and could be considered a very small quasi-control group, albeit a group that may have had lower CrAg titers. Initially they were treated as part of the antigen-negative group, receiving no antifungal therapy. A Subsequent protocol amendment led to this group of patients being considered as antigen positive. It is interestingandsomewhatunexpectedthat