The human T-lymphotropic virus type 1 (HTLV-1) was the first human retrovirus identified. It has been clearly implicated in the pathogenesis of adult T-cell leukemia/lymphoma and myelopathy/tropical spastic paraparesis.1-4 Other infectious, inflammatory, or immune-related diseases (uveitis, infective dermatitis, polymyositis, arthropathy, Sjögren syndrome, interstitial lymphocytic pneumonia) are suspected to be associated with infection with HTLV-1.5,6 A higher prevalence of HTLV-1 infections has been described in patients with tuberculosis, as well as an increased mortality rate in HTLV-1- and tuberculosis-coinfected patients.7,8 However, studies addressing this question have been small and failed to provide conclusive evidence of an association.9,10 The case-control study reported here investigated whether infection with HTLV-1 is associated with increased risk of tuberculosis. The study was conducted in Salvador, a city known to have a relatively high prevalence of HTLV-1 infection (1.76% in the overall population)11 and a yearly incidence of tuberculosis varying around 120 cases per 100,000 inhabitants.12
PARTICIPANTS AND METHODS
Study Site and Population
Salvador is the capital city of the state of Bahia, in northeastern Brazil, and is characterized by marked socioeconomic differences. The majority of the estimated population of approximately 2.5 million of inhabitants is of African or mixed African/Portuguese decent. Five of the 12 health districts in Salvador were selected to be included in the study because they had a high incidence of tuberculosis. The estimated population in these 5 districts was 1,230,000 inhabitants, just under half of the total population of Salvador.
The study was approved by the institutional review board of the Hospital Santa Isabel and case and control subjects gave written informed consent before entering the study.
Case subjects were recruited from those diagnosed by the Tuberculosis Control Program. The inclusion criteria were a first diagnosis of tuberculosis according to the World Health Organization criteria,13 confirmed by a senior pulmonologist; duration of treatment not exceeding 6 months prior to recruitment; and age 15 years or older. Exclusion criteria were having had a previous diagnosis of tuberculosis; pregnancy; and presence of HIV infection or any other condition leading to immunosuppression, including suspected or confirmed cancer and collagenosis and treatment with systemic corticosteroids.
Control subjects were selected from the outpatient clinic of the Hospital Santa Isabel, matched to case subjects by sex and age (in 10-year age bands). This hospital is part of the National Public Health Service, and people living in its catchments have similar socioeconomic and demographic characteristics as those living in the areas where the patients with tuberculosis were recruited. The inclusion criteria included absence of tuberculosis; residence in the city of Salvador; and age 15 years or older. The exclusion criteria included pregnancy and presence of HIV infection or other conditions associated with a state of immunosuppression including suspected or confirmed cancer, collagenosis, and treatment with systemic corticosteroids.
Demographic information (age, sex), socioeconomic status (income and level of education), lifelong number of sexual partners, recreational drug use, and history of blood transfusion were collected by standard questionnaire, which was applied to all individuals participating in the study. Ten milliliters of blood were withdrawn from each individual in ethylene diamine tetra-acetic acid-containing sterile tubes. The plasma was separated through centrifugation and both plasma and blood cells were stored at −20°C.
Plasma samples were screened for antibodies to HTLV-1 and 2 (HTLV-I/HTLV-II Ab-Capture ELISA Test System, Ortho Clinical Diagnostic Inc., Raritan, NJ) and HIV-1 and 2 (HIV-1/HIV-2 Ab-Capture ELISA Test System, Ortho Clinical Diagnostic). All samples with positive results for HTLV or HIV in the 1st test were tested again in duplicate. In case of positive results, on the 2nd test the samples were submitted for serologic confirmation. Confirmation of HIV-1 infection was performed by indirect immunofluorescence assay (Biomanguinhos Fiocruz, Rio de Janeiro Brazil) following the manufacturer's recommendations. For HTLV the confirmation and discrimination between HTLV-1 and HTLV-2 were performed using Western blot analysis (HTLV Blot 2.4; Genelabs, Singapore). Polymerase chain reaction analysis was performed in samples with undetermined HTLV serologic status as described elsewhere.14
The main dependent variable was being a case of tuberculosis and the main exposure of interest was seropositivity for HTLV-1. Univariable analysis was explored using the Pearson χ2 test with no continuity correction, with Yates correction, or Fisher exact test; and for ordered or asymmetric, continuous variables, Mann-Whitney U test. Multivariable analysis was used to investigate the association between HTLV seropositivity and tuberculosis, adjusting for potential confounding variables, using unconditional logistic regression keeping age and sex in the model. This analytical approach is recommended for frequency matching and is equivalent to a conditional logistical regression in which the matching sets are large and include all the cases and all the controls on the same category of the matching variables, in this case of the same sex and age group. The final model was selected by incorporating variables based on the “likelihood ratio” with α = 0.05. We present odds ratios (ORs) and 95% CIs. All analysis were done using the statistical software STATA 7.0 (Statacorp., College Station, TX) and SPSS version 11.0 (SPSS, Inc., Chicago, IL).
A total of 753 individuals were enrolled in this study between September 2001 and October 2002. There were 389 case subjects and 378 control subjects; 14 individuals in the case group were excluded because they were infected with HIV (3 of whom were also infected with HTLV-1). The median age was 34 years (range, 15-84 years) in cases and 35 years (range, 15-84 years) in controls. Cases had lower levels of education and income and more frequent history of using intravenous illicit drugs than controls (Table 1). The majority of cases (59%) were recruited within 3 months of the start of their treatment for tuberculosis, and 20% at the time of diagnosis (data not shown). No individual in the control group was excluded because of infection with HIV.
Sixteen of 375 individuals in the case group were infected with HTLV-1, compared with 5 of 378 individuals in the control group. Giving a crude OR 3.31 (95% CI, 1.20-9.13); this remained substantial and statistically significant (OR 3.01; 95% CI, 1.06-8.58) after controlling for age, sex, education, income, ethnicity, sexual history, and history of blood transfusion and intravenous drug use (Table 2).
We report here a statistically significant increase in the risk of tuberculosis in subjects infected with HTLV-1. Previous studies have investigated this association, but their results have been inconclusive: Matsuzaki et al15 described an increased reported history of tuberculosis in 2847 subjects infected with HTLV-1 in Japan. A case-control study in Senegal found no significant association between tuberculosis and HTLV-1,9 but the power of the study was small.5 A study comparing 154 HTLV-1-infected individuals with 799 uninfected blood donors concerning their history of infectious diseases, including tuberculosis, showed an OR of 3.3 (95% CI, 0.8-14.2).10
The prevalence of infection with HTLV-1 reported here (4.27%) is similar to that described in outpatients with tuberculosis from Nigeria (3.6%).16 The prevalence found in cases of tuberculosis in the present study was lower, when compared with previous surveys in cases of tuberculosis in the same area.7,8 In both previous papers, hospitalized cases of tuberculosis were studied, as opposed to the study reported here, which included outpatients with a 1st episode of tuberculosis. HTLV-1 has been shown to be associated with a worse course of tuberculosis, with lethality up to 25% in those infected with HTLV-1 compared with 8% in those not infected.8 This would lead to a higher prevalence in hospitalized cases.
The association we found between HTLV-1 and tuberculosis is biologically plausible, as a consequence of the immunologic deficiencies associated with infection by HTLV-1,17 which would facilitate multiplication of the Mycobacterium and progression to clinical disease.8,18 There is evidence that HTLV-1 infection could lead to immunologic modulation with significant effects on infection with helminths such as Strongyloides stercoralis and Schistosoma mansoni.19,20 It has been established that HTLV-1 induces T-lymphocyte spontaneous proliferation in vitro21,22 that is due predominantly to a memory CD4CD45RO+ cell subset.23 Impairment of the T-cell function is likely to occur: T lymphocytes from HTLV-1-infected individuals vaccinated with bacillus Calmette-Guerin fail to respond to purified protein derivative stimulation (PPD) in vitro,24 even in the presence of a strong T-helper type 1 response with high levels of interferon-γ.25 This is consistent with the evidence that asymptomatic infection with HTLV-1 is associated with reduced skin response to tuberculin (purified protein derivative).26-31 This has been interpreted as reflecting reduced cellular immunity, which potentially could lead to a higher rate of active tuberculosis in those with tuberculosis infection.5 Consistent with this hypothesis is the fact that the prevalence of HTLV-1 is higher in patients with pulmonary disease caused by Mycobacterium avium32; however, the fact that delayed hypersensitivity skin testing to mumps and Candida albicans antigens is the same among HTLV-1 and seronegative individuals33 suggests that the picture is more complex.
We are aware that the main limitation of the present study is the possibility of incomplete control of confounding. Tuberculosis and HTLV-1 have similar risk factors: poverty, low level of education, and intravenous drug use. However, control for these potential confounding variables hardly changed the OR, suggesting that the association is a causal effect. In conclusion, our results suggest that infection by HTLV-1 is significantly associated with a 1st diagnosis of tuberculosis. This would have implications for public health in areas where the prevalence of HTLV-1 is high. We recommend that HTLV-1-infected individuals be monitored for symptoms of tuberculosis, allowing for an early diagnosis and treatment, and that cases of tuberculosis be screened for HTLV-1 in areas of high prevalence of both diseases. Consideration should be given to the potential benefit of chemoprophylaxis in subjects with HTLV-1 infection and strong reaction to purified protein derivative to prevent development of tuberculosis, similar to the recommendations for HIV coinfection.34
The authors thank Dr. Fernanda Grassi and Mr. Noilson Gonçalves for critical review of this manuscript and technical assistance, respectively.
1. Poiesz BJ, Ruscetti FW, Gazdar AF, et al. Detection and isolation of type C retrovirus particles from fresh and cultured lymphocytes of a patient with cutaneous T-cell lymphoma. Proc Natl Acad Sci USA
2. Yoshida M, Miyoshi I, Hinuma Y. Isolation and characterization of retrovirus from cell lines of human adult T-cell leukemia and its implication in the disease. Proc Natl Acad Sci USA
3. Gessain A, Barin F, Vernant JC, et al. Antibodies to human T-lymphotropic virus type-I in patients with tropical spastic paraparesis. Lancet
4. Osame M, Usuku K, Izumo S, et al. HTLV-I associated myelopathy, a new clinical entity. Lancet
5. Marsh BJ. Infectious complications of human T cell leukemia/lymphoma virus type I infection. Clin Infect Dis
6. Manns A, Hisada M, La Grenade L. Human T-lymphotropic virus type I infection. Lancet
7. Moreira ED Jr, Ribeiro TT, Swanson P, et al. Seroepidemiology of human T-cell lymphotropic virus type I/II in northeastern Brazil
. J Acquir Immune Defic Syndr
8. Pedral-Sampaio DB, Martins Netto E, Pedrosa C, et al. Co-infection of tuberculosis
and HIV/HTLV retroviruses: frequency and prognosis among patients admitted in a Brazilian hospital. Braz J Infect Dis
9. Kaplan JE, Camara T, Hanne A. Low prevalence of human T-lymphotropic virus type I among patients with tuberculosis
in Senegal. J Acquir Immune Defic Syndr
10. Murphy EL, Glynn SA, Fridey J, et al. Increased prevalence of infectious diseases and other adverse outcomes in human T lymphotropic virus types I- and II-infected blood donors. Retrovirus Epidemiology Donor Study (REDS) Study Group. J Infect Dis
11. Dourado I, Alcantara LC, Barreto ML, et al. HTLV-I in the general population of Salvador, Brazil
: a city with African ethnic and sociodemographic characteristics. J Acquir Immune Defic Syndr
12. Xavier M. Tuberculose na Cidade de Salvador: Perfil epidemiológico na decada de 1990 e Tendências das Taxas de Incidência e Mortalidade Entre 1981 a 2000
[dissertation]. Salvador, Brasil: Federal University of Bahia; 2002.
13. Falzon D. Revised international definitions in tuberculosis
control: comments from the Aral Sea Area tuberculosis
programme. Int J Tuberc Lung Dis
14. Bittencourt AL, Dourado I, Filho PB, et al. Human T-cell lymphotropic virus type 1 infection among pregnant women in northeastern Brazil
. J Acquir Immune Defic Syndr
15. Matsuzaki T, Otose H, Hashimoto K. Diseases among men living in human T-lymphotropic virus type I endemic areas in Japan. Intern Med
16. Olaleye DO, Bernstein L, Sheng Z, et al. Type-specific immune response to human T cell lymphotropic virus (HTLV) type I and type II infections in Nigeria. Am J Trop Med Hyg
17. Goon PK, Bangham CR. Interference with immune function by HTLV-1. Clin Exp Immunol
18. Hollsberg P. Mechanisms of T-cell activation by human T-cell lymphotropic virus type I. Microbiol Mol Biol Rev
19. Porto AF, Neva FA, Bittencourt H, et al. HTLV-1 decrease Th2 type of immune response in patients with strongyloidiasis. Parasite Immunol
20. Porto AF, Santos SB, Alcantara L, et al. HTLV-1 modifies the clinical and immunological response to schistosomiasis. Clin Exp Immunol
21. Kramer A, Jacobson JF, Reuben EL, et al. Spontaneous lymphocyte proliferation in symptom free HTLV-1 positive Jamaicans. Lancet
22. Popovic M, Flomenberg N, Volkman D, et al. Alteration of T-cell functions by infection with HTLV-I or HTLV-II. Science
23. Prince HE, York J, Owen SM, et al. Spontaneous proliferation of memory(CD45RO+) and naïve(CD45RO−) subsets of CD4 cells and CD8 cells in human T lymphotropic virus (HTLV) infection: distinctive patterns for HTLV-I versus HTLV-II. Clin Exp Immunol
24. Suzuki M, Dezzuti CS, Okayama A, et al. Modulation of T-cell response to recall antigens in human T-cell leukemia virus type-1 infected individuals. Clin Diagn Lab Immunol
25. Carvalho EM, Bacellar O, Porto AF, et al. Cytokine profile and immunomodulation in asymptomatic human T-lymphotropic virus type-1 infected blood donors. J Acquir Immune Defic Syndr
26. Tachibana N, Okayama A, Ishizaki J, et al. Suppression of tuberculin skin reaction in healthy HTLV-I carriers from Japan. Int J Cancer
27. Murai K, Tachibana N, Shioiri S, et al. Suppression of delayed-type hypersensitivity to PPD and PHA in elderly HTLV-I carriers. J Acquir Immune Defic Syndr
28. Welles SL, Tachibana N, Okayama A, et al. Decreased reactivity to PPD among HTLV-I carriers in relation to virus and hematologic status. Int J Cancer
29. Hisada M, Shima T, Okayama A, et al. Suppression of skin reactivity to purified protein derivative by hepatitis C virus among HTLV-I carriers in Japan. J Acquir Immune Defic Syndr Hum Retrovirol
30. Hisada M, Stuver SO, Okayama A, et al. Gender difference in skin reactivity to purified protein derivative among carriers of HTLV-I in Japan. J Acquir Immune Defic Syndr
31. Suzuki M, Dezzutti CS, Okayama A, et al. Modulation of T-cell responses to a recall antigen in human T-cell leukemia virus type 1-infected individuals. Clin Diagn Lab Immunol
32. Matsuyama W, Mizoguchi A, Iwami F, et al. Clinical investigation of pulmonary Mycobacterium avium complex infection in human T lymphotrophic virus type I carriers. Thorax
33. Murphy EL, Wu Y, Ownby HE, et al. Delayed hypersensitivity skin testing to mumps and Candida albicans antigens is normal in middle-age HTLV-I and II infected U.S. cohorts. AIDS Res Hum Retroviruses
34. Pape JW, Jean SS, Ho JL, et al. Effect of isoniazid prophylaxis on incidence of active tuberculosis
and progression of HIV infection. Lancet