Background: Cryptococcosis is a systemic and opportunistic mycosis. Decreased lymphocytic immunity, as may result from HIV infection, is the main risk factor. Although HIV and cryptococcal species have historically coexisted in the Democratic Republic of Congo (DRC), data on this potentially lethal coinfection remain scarce. In this thesis, we analyse the epidemiological, clinical, biological and genetic profiles of cryptococcosis in people living with HIV (PLHIV) in Kinshasa.
Materials and methods: A systematic review and meta-analysis, two cross-sectional studies in three Kinshasa clinics, and an environmental study were conducted between 2019 and 2021. In addition to collecting clinical and biological data from the patients, we characterised Cryptococcus strains using multiplex serotyping PCR, ITS sequencing and Multi-Locus Sequence Typing (MLST). MALDI-TOF MS was used for isolate identification and the broth microdilution technique (according to the EUCAST E. Def 7.3.1 protocol) was used to determine the strains’ susceptibility to common antifungal agents.
Results: The hospital prevalence of neuromeningeal cryptococcosis (NMC) is estimated at 8.8% in the period prior to the adoption of the “Test and Treat� approach to HIV management, with a mortality rate of 34.8%. In the current era of consolidation of the availability of highly active antiretroviral therapies, a higher prevalence of 23.7% was evaluated in the Kinhasa clinics supported by Médecins Sans Frontières (MSF). Three different cryptococcal species are implicated in NMC in PLHIV in Kinshasa, namely Cryptococcus neoformans (Cn) (79.3%), Cryptococcus/Cutaneotrichosporon curvatus (Cc) (17.2%) and Cryptococcus/Papiliotrema laurentii (Cl) (3.5%). When comparing species, the clinical presentation of patients was more severe in Cn meningitis than in Cc/Cl. Sensitivity of conventional diasgnostic tests on cerebrospinal fluid (CSF) [Indian ink staining and detection of cryptococcal antigens (CrAg)] was higher for detecting Cn than for Cc/Cl. Of the Cn identified, seven different sequences type (ST) were distinguished by MLST, including one ST (ST659) not previously identified [ST93 (n = 15), ST5 (n = 2), ST53 (n = 1), ST31 (n = 1), ST4 (n = 1), ST69 (n = 1), and ST659 (n = 2)]. Although the majority of patients received triple antifungal therapy, the medical outcome was poor in 31.8% of cases, with no difference between Cn and Cc/Cl infection. Overall, only the Cl strain was resistant to amphotericin B and 5-flucytosine (5FC), a Cn sample was resistant to 5FC, and two Cc and two Cn strains were resistant to FCZ. Among asymptomatic patients with less than 200 CD4/μL tested for CrAg, approximately 18% were positive in serum, and 7.2% (of all included patients) were positive in CSF. Genotypic comparison of isolates from symptomatic and asymptomatic NMC patients showed that in both presentations, ST93 predominates in our setting, although a relatively less pathogenic ST was identified in one asymptomatic patient. We could not demonstrate an association between environmental Cryptococcus strains (isolated in and near patient’s homes) and the isolates involved in clinical infections, due to the lack of Cryptococcus isolation from these environments.
Conclusions: The burden and characteristics of cryptococcosis in Congolese PLHIV as described in this study provide a basis for better understanding and managing this opportunistic fungal disease. Furthermore, these data suggest the need for studies covering the whole of the DRC and the organisation of a national invasive fungal diseases’ management programme.