Highly active antiretroviral therapy decreases the incidence of visceral leishmaniasis in HIV-infected individuals.

Abstract

In Mediterranean countries visceral leishmaniasis (VL) is a zoonosis caused by the protozoon Leishmania infantum. In the early 1990s, the incidence of VL increased in southern European countries, probably because of the increased number of cases that occurred in patients with HIV infection [1]. Recently, highly active antiretroviral therapy (HAART), using a combination of protease inhibitors and reverse transcriptase inhibitors [2], has been demonstrated not only to have a powerful activity on HIV, as shown by the reduction in plasma HIV RNA and the increase in peripheral CD4 cells, but also to cause a drop in the incidence of opportunistic infections and deaths [3,4]. However, no data are presently available on the effect, if any, of HAART on the incidence of VL in HIV-infected individuals. Therefore, the aim of this study was to assess the incidence of VL, before and after the introduction of HAART, as standard therapy in HIV-infected patients in a large cohort of Italian HIV-infected subjects. In particular, we performed a 9 year (1991–1999) retrospective cohort study considering all cases of VL admitted to the Department of Infectious Diseases of a large university hospital in Rome, Italy. Over the study period, we observed 36 cases of leishmaniosis, 20 cases (55%) in HIV-infected patients and 16 (35%) in immunocompetent individuals. For a better evaluation of the differences in VL incidence in HIV-infected patients attributed to HAART, two periods of time were compared: January 1991–December 1995 (period A), and January 1997–December 1999 (period B). HAART became the standard of care in the majority of our patients after September 1996, when protease inhibitors became available in Italy, and for this reason we have not considered, in the statistical analysis, cases of VL occurring in 1996. In particular, we observed 18 episodes of VL in period A and 13 in period B. HIV infection was present in 78% (14/18) of the total VL cases observed in period A and in 15% (2/13) in period B (P < 0.01; odds ratio 19.25; 95% confidence interval 2.37–220.18). Comparing the incidence of VL in HIV-infected patients in periods A and B, a statistically significant reduction in the incidence of VL, from 0.7 episodes per 100 person/years to 0.13 episodes per 100 person/years, was observed (P < 0.01; odds ratio 5.25; 95% confidence interval 1.20–47.65). The majority of HIV-infected patients with VL were men (72%) and in the C3 category of HIV infection (75%) without any significant statistical differences between period A and period B. The mean number of peripheral CD4 cells was 54/mm3 (range 1–207/mm3). The mean level of plasma HIV RNA, available in patients observed in period B only, was 105 000 copies/ml. The analysis of risk factors for VL, including the geographical area of origin of the patients, did not differ significantly between the two study periods. In conclusion, our study confirms that HIV is a relevant predisposing condition for VL, in Italy as well as in other Mediterranean countries [5,6], and indicates, as a novel observation, a significant decrease in the incidence of VL in HIV-infected patients after the introduction of HAART. The reduction in the incidence of VL is probably the consequence of the well-known immune restoration induced by HAART [2], as witnessed by the increase in median peripheral CD4 cell counts and the decrease in HIV viraemia (data not shown). Such a reduced incidence of VL in HIV-infected patients could have a substantial impact on future morbidity and health costs in Mediterranean countries as well as in other geographical areas where VL is endemic. Mario Tumbarello Evelina Tacconelli Silvia Bertagnolio Roberto Cauda