Severe and Norwegian scabies are strongly associated with retroviral (HIV-1/HTLV-1) infection in Bahia, Brazil

Severe scabies has been associated with HTLV infection. To evaluate the impact of HTLV-I/HIV-1 co-infection on the clinical presentation of scabies, we reviewed 91 cases of scabies in Bahia, Brazil, during a 3 year period. Infections by HIV-1 (50%), HTLV-I (32%), and both (20%) were highly prevalent. Crusted or severe scabies were strongly associated with HTLV-I and, to a lesser degree, with HIV-1 infection. Co-infected patients had a higher risk of death (P= 0.01). Severe forms of scabies were highly predictive of double retroviral infection. Crusted (Norwegian) scabies has been associated with HTLV-I/II infection in some studies [1]. Bahia, a northeastern state in Brazil, is an endemic area for HTLV-I, which infects approximately 2% of the state's general population [2]. A preliminary study conducted by our group showed a strong association between crusted (Norwegian) scabies and retroviral (HIV-1/HTLV-I) infection among patients treated at the Infectious Diseases Clinic of the Universidade Federal da Bahia Hospital (HUPES) [3]. To evaluate the association between retroviral infection and a clinical presentation of scabies, we reviewed the medical records of patients treated at the Dermatology and Infectious Diseases Clinics of HUPES from January, 1996 to January 1999, and patients admitted to the hospital as a result of severe forms of scabies. Serological diagnoses of retroviral infection (HTLV-I or HIV-1) were assessed by cross-referencing patients’ medical records with the serology databank of the Laboratory of Human Retrovirology at the same hospital. Severe scabies was defined as disease affecting at least 80% of the body surface, but not fulfilling the criteria for crusted scabies. Demographic data and, when available, CD4 cell counts/HIV viral loads were recorded. When more than one result was available for a specific patient, we considered the result nearest to the date when scabies was first diagnosed. Infection by HIV and/or HTLV-I was confirmed by a positive Western blot test. A total of 91 patients (39 men; 52 women) were diagnosed as having scabies during the study period. The mean age was 35.5 ± 12.9 years for men, and 37.0 ± 20 years for women patients. Patients infected by HTLV-I were significantly older than those individuals not harboring the virus (42.3 ± 12 versus 32.3 ± 17 years for HTLV-I-positive and negative patients, respectively;P = 0.004, Kruskal–Wallis test). There was no detectable difference in mean age for HIV-positive and negative subjects (P = 0.3). Also, the mean age was similar for men (35.6 ± 13 years) and women (35.6 ± 19 years). Infection by HIV-1 was diagnosed in 45 (50%) patients, whereas antibodies against HTLV-I were detected in 29 (32%) individuals. Eighteen patients (20%) were positive for both viruses. Among patients infected by HIV-1, 44% were under antiretroviral therapy at the time scabies was diagnosed. A total of 58 subjects (64%) fulfilled the criteria for crusted scabies (23 patients), or severe scabies (35 patients). Viral load results were available for 22 subjects, and no difference was detected in the mean viral load for groups of patients, when compared according to the severity of skin disease with which they presented (P = 0.4). A similar result was obtained when comparing the mean CD4 cell count (28 patients), and the type of scabies (P = 0.4). Table 1 summarizes the serology results according to the severity of scabies.Table 1: Characteristics of patients according to serology for HIV-1 and HTLV-I.We detected a significant association between HIV-1 infection [odds ratio (OR) 3.28; 95% confidence interval (CI) 2.00–5.37;P < 0.0001, Yates corrected), or patients presenting with double viral infection (OR 30; 95% CI 5.28–205.8;P < 0.0001, Yates corrected) and male sex. However, this association was not observed for patients infected only by HTLV-I (P = 1.0). A univariate analysis of the type of exposure and retroviral infection also showed a strong association between infection by either HIV-1 (OR 21.5; 95% CI 2.71–45.65;P = 0.0004) or HTLV-I (OR 4.74; 95% CI 1.34–17.29;P = 0.01) and illicit intravenous drug use, when considering the entire population studied. This association remained significant when considering only those patients not infected with HTLV-I (P = 0.005, Fisher exact test), but not for those patients who tested negative for HIV-1 (P = 0.2). Patients acquiring retroviral infection through sexual exposure were less likely to present with crusted scabies than those infected by the parenteral route (OR 0.27; 95% CI 0.6–0.46;P = 0.00002, Fisher exact test). This association was not detected for patients with non-crusted forms of the disease (P = 0.1). Five patients died during the study period. All were diagnosed with the double infection, and presented with crusted scabies (P = 0.01). There was no detectable association between HIV single infection and crusted scabies (no cases among HIV-positive/HTLV-negative patients). On the other hand, all patients presenting with crusted scabies, and having available information about HTLV serology (21/23) tested positive for that agent (P < 10−7). This association was significant even when considering only patients presenting with HTLV-I single infection (P = 0.02). In addition, when crusted scabies cases were excluded, severe scabies cases remained associated with HTLV-I single infection (OR 3.0; 95% CI 1.85–4.86;P = 0.01), and showed a trend towards association with infection by HIV-1 (OR 2.95; 95% CI 0.9–9.65;P = 0.08). However, double retroviral infection was not associated with non-crusted forms of scabies (P = 0.3). Some previously published data suggest that crusted scabies could be a predictive marker for adult T cell leukaemia/lymphoma [4]. In the present study, even after 2 years or more of follow-up for the majority of patients, we did not see any evidence of adult T cell leukaemia/lymphoma in this group of patients. On the other hand, patients presenting with severe scabies had a lower risk of death when compared with HIV-infected patients presenting with crusted scabies (OR 0.75; 95% CI 0.58–0.97;P = 0.01). We concluded that severe forms of scabies are strongly associated with HTLV-I infection in Bahia. In addition, co-infection by HIV-1 seems to be a risk factor for the increased severity of scabies. Patients infected only by HIV-1 showed a trend towards presenting with severe but not crusted scabies. The crusted form of the disease was highly predictive of double retroviral infection and seemed to be associated with profound immunodeficiency, because HIV-1/HTLV-I co-infected patients were more likely to die during the study period. Our data suggest that co-infection by HIV-1/HTLV is associated with a deeper degree of immunodeficiency, which increases the risk of developing severe forms of scabies. Carlos Britesa,b Mônica Weylla Célia Pedrosob Roberto Badaróa,b