10% of HIV-related deaths in Brazil and Mozambique caused by cryptococcal infection

07 June 2019

Failure to diagnose and late-presenters to care continue to contribute to cryptococcal-related mortality in people living with HIV in low-resourced contexts.


Cryptococcal meningitis was responsible for one in ten deaths among a cohort of people living with HIV from Brazil and Mozambique. The figures are in-line with World Health Organization (WHO) estimates that blame 15% of AIDS-related deaths on cryptococcal infection – three quarters of which are in sub-Saharan Africa.

Alarmingly, the research also revealed that clinical suspicion of cryptococcal infection was very low (41%) among the people living with HIV who died, suggesting that diagnosis and patient management were inadequate, supporting the need for cryptococcal screening for early detection of the disease. This is the first time data on the burden of cryptococcal infection has been presented from Mozambique.

Cryptococcal meningitis is caused by a fungal infection that has spread to the brain and spinal cord causing meningitis. It commonly presents among those with severely compromised immune systems, such as people living with HIV who present late to care or who are not on effective antiretroviral treatment (ART).

Since effective ART, cryptococcal meningitis incidence has decreased dramatically worldwide. But among those who do get cryptococcal infection, outcomes are poor without intervention.

The cryptococcal antigen (CrAg) can be screened for. It shows up on a test weeks before the onset of meningitis. Testing positive for the antigen means people can be pre-emptively treated using fluconazole. CrAg screening and pre-emptive treatment for patients with advanced HIV has been shown to reduce deaths by 28% in Tanzania and Zambia.

Screening for cryptococcal meningitis in people living with HIV with a CD4 count ≤100 cells per copy is recommended by the World Health Organization and many national programmes, but it is poorly implemented in low-resourced contexts.

In the study, investigators looked at cryptococcus-associated mortality in a series of 284 autopsies performed in two hospitals located in high prevalence HIV areas, Mozambique (Maputo) and the Brazilian Amazonia (Manau). They also analysed the clinical presentation, management, and histopathological (studying the person’s tissues) and microbiological (looking for microbes in the person’s body) findings of 17 cases of fatal cryptococcal infection.

Complete diagnostic autopsies (CDAs) were performed on 223 people from Mozambique and 61 from Brazil. Of note, around half of the deaths of women (57 of 112) in the Mozambique group were maternal deaths, i.e. women dying during pregnancy, partum, post-partum or within 42 days of termination of pregnancy. In Brazil, 21 females were included with one maternal death.

In their autopsies, 163 (57%) tested positive for HIV, and a total of 17 (6%) fatal cryptococcal infections were confirmed – there was just one case of non-HIV related cryptococcal death in a child. In Mozambique, 10% of HIV deaths were attributed to cryptococcal infection, and 11% of maternal deaths. In Brazil, 13% of deaths were attributed to cryptococcal infection. This indicates a high fatality rate among these populations, suggesting that cases are going undiagnosed.

Among the 16 people living with HIV who died of cryptococcal infection, 13 (81%) were known to be living with HIV, while HIV status was unknown in the three other cases. Four out of the 16 (25%) HIV-positive patients were on ART, but data on how long they were on ART was only available for one patient.

In all cases of cryptococcal infection, this diagnosis was only considered as the first option in four out of 17 (23%) of the confirmed cases who presented to hospital. Mortality rates were extremely high in this cohort compared to data from elsewhere. Seven of the patients (41%) died within 72 hours of admission, and 12 out of the 16 HIV-positive patients (75%) died within one week of admission – other settings have indicated a mean of two weeks mortality from hospital admission.

In their discussion, the authors theorised that these high fatality rates may be the result of poor clinical suspicion of cryptococcal infection and therefore, a lack of prescribed antifungal treatment in more than half of the patients. Patients could have simply presented too late to care – considering seven of the patients died within 72 hours.

Of interest in this study is the high proportion of cases of C. gattii (33%) strain, compared to (66.6%) C. neoformans var. grubii. Over a two-year period in South Africa, C. gattii represented just 2.4% of cryptococcal strains, and it is more commonly associated with people who are immunocompetent people, as opposed to the immunocompromised. While their results should be interpreted with caution, larger studies are needed to understand how C. gattii strain interacts with people living with HIV.

In their conclusion, the authors highlight the substantial mortality burden associated with cryptococcal infection, they call for “supporting current recommendations of CrAg screening and pre-emptive therapy, which is even more relevant in settings in which insufficient pre-mortem clinical suspicion is given to this highly prevalent and life-threatening opportunistic infection.”


Avert.org is helping to prevent the spread of HIV and improve sexual health by giving people trusted, up-to date information.

We provide all this for FREE, but it takes time and money to keep Avert.org going.

Can you support us and protect our future?

Every contribution helps, no matter how small.

Photo credit:

Written by Caitlin Mahon

Content Specialist - HIV & Sexual Health