Results from an investigation of 2,753 cases of cryptococcal infection in a South African province with high HIV/AIDS prevalence suggest that interventions like HIV counselling and testing, early detection, and prevention/treatment of cryptococcosis with antifungal drugs could considerably reduce the burden of morbidity and mortality in resource-limited settings. Researchers warn that risk of immune reconstitution inflammatory syndrome (IRIS) is high upon introduction of antiretroviral treatment in HIV-positive individuals who have previously undergone treatment for cryptococcosis.
Cryptococcus is one of the main causal agents of opportunistic infections in HIV-positive individuals. When the pathogen reaches the cerebro-spinal fluid, it causes meningitis, which is a major cause of morbidity and mortality, and a burden to the healthcare sector. In 1998 a cohort study of HIV-positive South African miners attributed to the infection 37% of all deaths in the cohort. Results of similar investigations in Rwanda, Zimbabwe, and South Africa identified cryptococcal infection as being the leading cause of meningitis among HIV-positive patients, with in-hospital mortality ranging from 43 to 64%.
In order to quantify the burden of the infection in Gauteng Province (pop. 8.8 million – October 2001 census), a prospective surveillance study involving academic, governmental and private institutions was carried out from 1 March 2002 to 29 February 2004. At the time (2002), it was estimated that 1,449,899 individuals in the province were living with HIV/AIDS, 96.8% of whom were adults aged 18-64 years.
For the purpose of the study, an incident case of cryptococcosis was defined as a first positive test to any Cryptococcus species by either: culture, histopathology, India-ink or cryptococcal latex agglutination test on any body fluid, in a Gauteng resident. Cases were identified prospectively by means of regular weekly visits and monthly electronic audits of enrolled laboratories, with the Gauteng Cryptococcal Surveillance Initiative Core Committee managing all aspects of the surveillance.
Univariate and multivariate statistical analysis of the data was performed and odds ratios on three possible outcomes were calculated: hospital death within two days of admission, death in hospital more than two days after admission, and survival until discharge or transfer. General socio-economic variables and meningitis or HIV/AIDS-specific variables were included in the multivariate model.
Over the study period 2,753 incident cases of cryptococcosis were identified. The overall incidence rate was 15.6/100,000, (males = 15.2/100,000 (49% of cases), females = 15.8/100,000). Children represented 0.9% of all cases (n=24) and the incidence rate in HIV-infected children under 14 years of age was 38/100,000. Ethnic group-specific incidence rates showed the highest level of disease occurred in blacks (98% of patients), with an incidence rate of 20.7/100,000 (mixed race = 2.7/100,000, Indian/Asian = 0.7/100,000, white = 0.4/100,000). Among HIV-positive individuals incidence rates were 95/100,000, rising to 1400/100,000 among persons living with AIDS.
Of the 2,424 incident cases for whom complete information was available, 178 died within two days of admission, 529 after day two (total in-hospital mortality = 27%), and 1717 patients were discharged or transferred. Almost all patients (97%) presented with some signs and symptoms suggestive of meningitis, such as: headache (n=2147 – 78%), neck stiffness (1900 – 69%), fever (1514 – 55%), nausea and vomiting (1129 – 41%), altered mental status (853 – 31%), seizures (248 – 9%), coma (83 – 3%), sixth cranial nerve palsy (28 – 1%). For all patients, other than suspected or confirmed HIV infection, no other possible risk factor for cryptococcal infection could be identified.
A high proportion of patients were admitted to hospital with one or more critical conditions such as coma or advanced systemic illness manifested in wasting. Many (36%) died or left hospital without being offered HIV testing and counselling, or without being offered prophylaxis with fluconazole, indicating that although cryptococcosis is preventable and prophylaxis/treatment is free, health services were not yet geared to cope with the high case management burden, or to identify less severe forms of cryptococcosis.
Besides revealing that cryptococcosis is a cause of high mortality in Gauteng, graphic representation of the results show the very close overlap between age-adjusted curves for HIV mortality rates and for cryptococcosis incidence rates, suggesting that appropriate treatment and prevention of cryptococcosis would significantly reduce HIV/AIDS-related morbidity and mortality.
Since 2002 fluconazole has been made freely available to hospitals in Gauteng, and study findings confirm that appropriate antifungal medication for Cryptococcus – such as fluconazole or amphotericin B – is significantly protective from death within 2 days of admission. Fluconazole primary prophylaxis is an evidence-based intervention, which has proved very effective in a variety of settings. The researchers recommend the scaling up of the intervention along with an effort to fill the gaps in detection, counselling and treatment still present in the care of HIV-infected South African patients.
Special caution is however recommended for physicians initiating patients on antiretroviral treatment in this or similar settings, as the risk of death due to IRIS in patients previously treated for cryptococcosis is high, due to unmasking of latent cryptococcosis or to the development of paradoxical reactions resembling exacerbations of the primary infection.
McCarthy K et al. Population-based surveillance for cryptococcosys in an antiretroviral-naïve South African province with a high seroprevalence. AIDS 20: 2199 – 2206, 2006.