Central nervous system (CNS) infections are common across all ages in Sub-Saharan Africa in people with or without HIV-infection. In persons with HIV, cryptococcal meningitis has historically been the second most common AIDS-defining illness in Africa and the most common cause of adult meningitis in Sub-Saharan Africa overall. The next most common cause of meningitis is likely TB meningitis, although CSF diagnostics are challenging. With the widespread availability of antiretroviral therapy (ART), long term survival in persons living with AIDS and CNS infections should be possible, but delayed or inaccurate diagnoses and limited therapeutic options contribute to poor outcomes. Furthermore, with the ‘test and treat’ strategy of immediate ART initiation, more people are presenting with CNS infections unmasked after starting ART, yet their outcomes are unclear. We propose to continue a prospective cohort study of 1200 new HIV-infected persons presenting with suspected CNS infection in Kampala and Mbarara, Uganda. We will use point-of-care and molecular diagnostics to rapidly determine the etiologies of CNS infections, with a specific focus on optimizing and validating new diagnostic tests, such as a semi-quantitative cryptococcal antigen (CrAg-SQ) lateral flow assay and the Xpert MTB/RIF Ultra for TB meningitis. Second, we propose to conduct phase II trial investigating the microbiologic effects of a novel oral antifungal agent (APX001) that inhibits fungal GWT1 enzyme blocking fungal mannoprotein transport to the cell wall surface and is synergistic with fluconazole. Finally, we will measure neurocognitive performance to investigate the effect of recent ART initiation on neurologic outcomes. Specific Aims 1. Determine the etiology of CNS infections in Sub-Saharan Africa among HIV-infected adults through use of a stepwise diagnostic algorithm coupled with next-generation sequencing and post-mortem exams. 2. Determine in HIV-related cryptococcal meningitis if oral APX001, a novel fungal GWT1 (GPI-anchored wall transport protein 1) inhibitor, achieves with concomitant fluconazole a non-inferior rate of CSF Cryptococcus clearance as compared with IV amphotericin B deoxycholate and fluconazole. 3. Determine if neurocognitive outcomes in HIV-infected persons presenting with CNS infections unmasked after recent ART initiation are worse than in ART-naïve persons presenting with CNS infections. Hypotheses: 1. We hypothesize in the ART era that cryptococcal and TB meningitis remain the two most common etiologies of meningitis in an HIV-infected population despite the ‘test and treat’ ART strategy. 2. We hypothesize that APX001, a novel broad spectrum oral antifungal agent is well tolerated and will have a non-inferior rate of CSF sterilization compared with IV amphotericin B in cryptococcal meningitis. 3. We hypothesize neurocognitive outcomes are worse in CNS infections unmasked on ART vs. ART-naïve.