Causes of HIV-related CNS infection in Cameroon, Malawi, and Tanzania: epidemiological findings from the DREAMM HIV-related CNS implementation study.

Background

CNS infections cause approximately a third of HIV-related deaths. The Driving Reduced AIDS-Associated Meningo-encephalitis Mortality DREAMM study aimed to prospectively diagnose the aetiology of HIV-related CNS infection in five public hospitals in Cameroon, Malawi, and Tanzania.

Methods

DREAMM was a multicentre, hybrid type-2 implementation science project. Adults (aged ≥18 years) presenting with a first episode of suspected CNS infection, who were HIV seropositive or willing to have an HIV test, were eligible for recruitment. Following implementation of the DREAMM model of care, we measured the prevalence of cryptococcal meningitis, tuberculous meningitis, bacterial meningitis, and cerebral toxoplasmosis and did a χ2 test to assess whether prevalence differed between countries. We also reported disease-specific mortality and Toxoplasma gondii seroprevalence.

Findings

Of 356 participants with suspected CNS infection analysed at baseline, 269 (76%) were diagnosed as having a CNS infection. Of these, 202 (75%) had a confirmed diagnosis. Between Cameroon, Malawi, and Tanzania, the prevalence of the four main types of CNS infection differed (cryptococcal meningitis p=0·0014, bacterial meningitis p=0·0043, CNS tuberculosis p<0·0001, and toxoplasmosis p<0·0001). Cryptococcal meningitis (148 [55%] of 269) was the leading cause overall. The next most common causes were CNS tuberculosis in Tanzania (29 [29%] of 99) and bacterial meningitis in Malawi (15 [19%] of 80). In Cameroon, cerebral toxoplasmosis (39 [43%] of 90) was the leading cause followed by cryptococcal meningitis (36 [40%] of 90). For cryptococcal meningitis, all-cause 2-week mortality was 23% (34 of 147) and all-cause 10-week mortality was 45% (66 of 146).

Interpretation

Within the study population, the aetiology of HIV-related CNS infection varied substantially between Malawi, Cameroon, and Tanzania. Additional prospective epidemiological data are needed to inform HIV programmes. 2-week cryptococcal meningitis mortality outcomes were similar to those of clinical trials. However, new interventions are urgently needed to sustain mortality reductions following hospital discharge.