Whether HIV infection increases the risk of Buruli ulcer has been little investigated. Cases of Buruli ulcer/HIV coinfection have been reported with typical Buruli ulcer clinical presentation [1–3] or severe lesions [4–6]. HIV seroprevalence has been reported in Buruli ulcer cases series [4,7–9], but in only one study with a comparison group. In Ghana, 6/116 (5.2%) Buruli ulcer patients aged 2–53 years were HIV-positive compared with 1/116 (0.9%) individually age-matched neighbourhood controls .
We investigated the association between Buruli ulcer and HIV in a case–control study by comparing the HIV-1/2 seroprevalence in a series of Buruli ulcer patients with controls selected from within the community in southern Benin.
Between January 2002 and August 2003, we recruited 426 Buruli ulcer patients at the ‘Centre Nutritionnel et Sanitaire Gbemoten of Zagnanado’ (CSNG) and at the ‘Centre de Dpistage et de Traitement de l'Ulcère de Buruli of Lalo’ (CDT/UB), in Benin. Patients younger than 18 months and those living in Cotonou and Porto-Novo cities or outside Benin were excluded. Blood samples were collected after counselling and obtaining the informed consent of the patient or their guardian. Individual refusal was not systematically recorded but, according to local staff, it was low.
The controls were community residents who were part of a larger case–control study conducted in the same region to estimate the effectiveness of BCG vaccine against Buruli ulcer. The methods employed have been detailed elsewhere . Briefly, the larger study recruited one to four individually age-matched and sex-matched neighbourhood controls for Buruli ulcer patients hospitalized at the CSNG or CDT/UB between August 2002 and August 2003. A door-to-door systematic procedure starting at the patient's house was used for control selection. Participating controls were invited to provide a blood sample and 74% agreed. In total, HIV testing was performed anonymously for 613 controls (belonging to 180 matched sets).
HIV serology used Abbott Determine HIV-1/2 (Abbott Laboratories, Dainabot Co. Ltd, Tokyo, Japan). Positive results were confirmed using either the Uni-Gold HIV test (Trinity Biotech PLC, Bray, Ireland), ImmunoComb II HIV 1&2 Bispot (Orgenics Ltd, Yavne, Israel), Genie II HIV-1/HIV-2 (Bio-Rad, Marnes-la-Coquette, France) or the Vironostika HIV Uni-Form II Ag/Ab test (bioMérieux, Marcy l'Etoile, France). Proportions were compared using Pearson's chi-square or Fisher's exact tests. Confidence intervals (CIs) of crude odds ratios (ORs) were calculated using exact methods. ORs were adjusted for age, sex and region by logistic regression in STATA, version 9.0 (Stata Corp, College Station, Texas, USA). CI calculations used robust standard errors to allow for the clustering of controls in the matched sets. The Beninese Ministry of Public Health approved this study.
Eleven out of 13 Buruli ulcer patients positive for the Determine HIV-1/2 were confirmed by a second test. The prevalence of HIV infection was 2.6% (11/426) among all patients, 0.4% (1/246) among those aged less than 15 years, 6.0% (8/133) among those aged 15–49 years and 4.3% (2/47) among those aged 50 years or more. Two hundred and fifty-eight (60.6%) Buruli ulcer patients were confirmed by Ziehl-Neelsen, culture and/or IS2404-PCR  including six of those who were HIV-positive. Of the 415 HIV-negative patients, 35.3% presented with an ulcer, 28.7% with a plaque, 30.7% mixed form and 5.3% another lesion (nodule, oedema or osteomyelitis). The HIV-positive patients were more likely to have plaques (6/11, 54.5%; P = 0.06). Other HIV-positive patients presented with ulcer (1/11, 9.1%) or mixed form (4/11, 36.4%). Thirty (7.2%) HIV-negative patients had bone involvement (24 with and six without active skin lesions) compared with one (9.1%) HIV-positive patient (P = 0.6). Five HIV-positive patients were women. All were new incident Buruli ulcer patients. Buruli ulcer lesions among the nine patients with otherwise asymptomatic HIV infection evolved favourably after surgery. Two patients with clinical AIDS presented with severe mixed forms (ulcer with plaque or oedema on the same anatomical site). Both arrived late for Buruli ulcer treatment (8 and 24 months after Buruli ulcer appearance) and died during hospitalization.
The age distribution was similar in cases and controls. Compared with cases, controls were less likely to be women (42.9 versus 49.8%; P = 0.03) and to come from the Zou and Oueme-Plateau regions (73.4 versus 82.2%; P < 0.001). Three of the 613 controls tested HIV-positive using Determine HIV-1/2, two of whom (women, aged 55 and 6 years) were confirmed by a second test. The overall prevalence of HIV infection among controls was 0.3% [0.3% (1/326) among controls younger than 15 years, 0% (0/222) among those aged 15–49 years and 1.5% (1/65) among those aged 50 years or more). The crude OR for the association between HIV and Buruli ulcer was 8.1 (exact 95% CI = 1.8–75; P = 0.003). Table 1 presents the crude and adjusted ORs overall and when analysis was restricted to confirmed cases, when analysis included only the first control of each matched set, and when excluding the cases with clinical AIDS.
These results suggest that HIV increases the risk of Buruli ulcer. Although our estimate of HIV prevalence among controls may have been biased by refusals, it is in line with expectations: in Benin in 2003, the estimated HIV prevalence was in the range 1.1–3.3% among adults aged 15–49 years , lower than the prevalence in the patients in that age group but higher than that in the controls. Moreover, as cases and controls were mainly from Buruli ulcer endemic remote rural areas, a lower HIV prevalence can be expected.
Including all controls in the analysis may not be optimal, as controls were more similar within each matched set but robust standard errors were used to take account of the clustering. Furthermore, although not statistically significant, the OR remained high when analysis was restricted to the first control of each set but numbers were small for this analysis.
The patients with clinical AIDS died of AIDS during the admission but, for the other HIV-positive patients, the Buruli ulcer followed a normal clinical course. In contrast with previous observations , bone involvement was not more frequent among HIV-positive patients.
In conclusion, our study suggests an association between HIV and Buruli ulcer. An increase in HIV prevalence may have an impact on Buruli ulcer incidence in endemic areas. Moreover, until more is known, in settings where counseling and treatment are available, HIV testing should be suggested to all Buruli ulcer patients.
Christian Roch Johnson, Claude Zinsou and Elisa de Biurrun Bakedano were supported by a grant from the Directorate-General for Development Cooperation (DGDC, Brussels, Belgium. Project: Buruli ulcer in Benin). Judith Glynn was supported by the UK Department of Health (Public Healthcareer Scientist Award). This work was also partly supported by the Damien Foundation (Brussels Belgium).
1. Allen S. Buruli ulcer and HIV infection [Letter]. Int J Dermatol
2. Delaporte E, Alfandari S, Piette F. Mycobacterium ulcerans associated with infection due to the human immunodeficiency virus [Letter]. Clin Infect Dis
3. Ameh EA, Dogo PM, Ahmed A, Maitama HY, Esangbedo AE, Nmadu PT. Mycobacterium ulcerans skin infection in a patient with HIV infection: is this incidental? [Letter] Trop Doct
4. Portaels F, Zinsou C, Aguiar J, Debacker M, de Biurrun Bakedano E, Guedenon A, et al
. Les atteintes osseuses dans l'ulcère de Buruli: à propos de 73 cas. Bull Seanc Acad R Sci Outre-Mer 2003; 49:161–190.
5. Johnson RC, Ifebe D, Hans-Moevi A, Kestens L, Houessou R, Guedenon A, et al. Disseminated Mycobacterium ulcerans disease in an HIV-positive patient: a case study [Letter]. AIDS
6. Toll A, Gallardo F, Ferran M, Gilaberte M, Iglesias M, Gimeno JL, et al
. Aggressive multifocal Buruli ulcer with associated osteomyelitis in an HIV-positive patient. Clin Exp Dermatol 2005; 30:649–651.
7. Asiedu K, Etuaful S. Socioeconomic implications of Buruli ulcer in Ghana: a three-year review. Am J Trop Med Hyg 1998; 59:1015–1022.
8. Darie H, Le Guyadec T, Touze JE. Epidemiological and clinical aspects of Buruli ulcer in Ivory Coast. 124 recent cases. Bull Soc Pathol Exot 1993; 86:272–276.
9. Ouattara D, Meningaud JP, Saliba F. Multifocal forms of Buruli ulcer: clinical aspects and management difficulties in 11 cases. Bull Soc Pathol Exot 2002; 95:287–291.
10. Raghunathan PL, Whitney EA, Asamoa K, Stienstra Y, Taylor TH Jr, Amofah GK, et al
. Risk factors for Buruli ulcer disease (Mycobacterium ulcerans
infection): results from a case–control study in Ghana. Clin Infect Dis 2005; 40:1445–1453.
11. Nackers F, Dramaix M, Johnson RC, Zinsou C, Robert A, de Biurrun Bakedano E, et al
. BCG vaccine effectiveness against Mycobacterium ulcerans
disease (Buruli ulcer): a case–control study in Benin. Am J Trop Med Hyg 2006; 75:768–774.
12. World Health Organization. Microbiological and histopathological methods. In: Portaels F, Johnson P, Meyers WM, editors. Buruli ulcer. Diagnosis of Mycobacterium ulcerans
disease. A manual for healthcare providers. WHO/CDS/CPE/GBUI/2001.4. WHO: Geneva, Switzerland; 2001. pp. 27–46.
13. UNAIDS/UNICEF/WHO. Epidemiological fact sheets on HIV/AIDS and sexually transmitted Infections. Benin
; 2004 Update.