Corresponding Author Andréa R. B. Rossit, Microbiology and Parasitical Disease Department, Biomedical Institute, Fluminense Federal University, Rua Hernani Pires de Melo, 101, São Domingos, 24210 130 Niterói, RJ, Brazil. E mail: email@example.com
Objective To evaluate retrospectively the microbiological profile of Mycobacterium species isolated from HIV-infected patients attending the HIV/TB reference health care units in São José do Rio Preto, Brazil.
Method Retrospective evaluation of all HIV-1 positive patients whose IAL-SJRP laboratorial analysis was positive for Mycobacterium sp. after diagnosis of HIV Infection, from January 2000 to December 2006.
Results Of 198 patients, acid-fast staining detected mycobacteria early in 41%. Culture revealed 52.5% to be infected with Mycobacterium tuberculosis (MT). 42.4% had non-tuberculous mycobacteria (NTM) and 5.1% had MT/NTM positive cultures. Eleven per cent of MT strains were resistant to at least one of the antimycobacterial drugs and 3.1% were multidrug resistant. 39.4% of isolated mycobacteria were NTM species.
Conclusion Our data may serve as a starting point for further comparisons with other Brazilian regions and other developing countries. The data may provide important clues to the future understanding, prevention and control of such co-infections around the world.
Objectif: Evaluer rétrospectivement le profil microbiologique des espèces de Mycobacterium isolées de patients infectés par le VIH fréquentant des unités de soins de santé de référence pour la tuberculose/VIH à São José do Rio Preto, au Brésil.
Méthode: Evaluation rétrospective de tous les patients VIH-1 positifs dont l’analyse IAL-SJRP de laboratoire était positive pour Mycobacterium sp, de janvier 2000 à décembre 2006.
Résultats: Sur 198 patients, la coloration acido-résistante a permis de détecter des mycobactéries chez 41% d’eux. La culture a révélé 52,5% infectés par Mycobacterium tuberculosis (MT). 42,4% avaient des mycobactéries non-tuberculeuses (NTM) et 5,1% avaient des cultures positives pour MT et NTM. 11% des souches MT étaient résistantes à au moins un des médicaments antituberculeux, 3,1%étaient multirésistantes. 39,4% des mycobactéries isolées étaient des espèces NTM.
Conclusion: Nos données pourraient servir de point de départ pour des comparaisons avec d’autres régions du Brésil et d’autres pays en développement. Les données pourraient fournir d’importants indices à la compréhension, prévention et au contrôle de telles coinfections à travers le monde.
Objetivo: Evaluar de forma retrospectiva el perfil microbiológico de las especies de Mycobacterium aisladas de pacientes infectados con VIH atendidos en los centros sanitarios de referencia para VIH/TB en São José do Rio Preto, Brasil.
Método: Evaluación retrospectiva de todos los pacientes VIH-1 positivos cuyos análisis realizados en el laboratorio IAL-SJRP eran positivos para Mycobacterium sp. tras habérseles diagnosticado la infección por VIH entre Enero 2000 y Diciembre 2006.
Resultados: Las micobacterias se detectaron mediante tinción ácido-resistente en 41% de 198 pacientes. Los cultivos revelaron que un 52.5% estaban infectados con Mycobacterium tuberculosis (MT). 42.4% tenían una micobacteria no tuberculosa (MNT), 5.1% tenía cultivos MT/MNT positivos. Un 11% de las cepas de MT eran resistentes a al menos uno de los medicamentos antimicobacterianos. Un 39.4% de las micobacterias aisladas eran especies MNT.
Conclusiones: Nuestros datos podrían servir como punto de partida para futuras comparaciones con otras regiones del Brasil y de otros países en vías de desarrollo. Los datos podrían proveer claves importantes para la comprensión futura, prevención y control de este tipo de coinfecciones alrededor del mundo.
Tuberculosis (TB) and AIDS represent major public health failures. Depending on the geographic area, the approximately 40 million people living with HIV are 6–50 times more likely to develop active TB than HIV-uninfected people (Perkins & Cunningham 2007). In terms of prevalence of TB-co-infected HIV cases, Brazil (17%) occupies a middle position between African countries such as Zimbabwe (68%), South Africa (60%) and Mozambique (48%), and Asian countries such as Cambodia (13%), Thailand (8.5%), India (5.2%) and the Russian Federation (6.8%) (WHO 2006). Non-tuberculous mycobacteria (NTM) are ubiquitous environmental organisms that have been implicated in progressive pulmonary and non-pulmonary diseases. NTM isolation rates differ greatly between the distinct Brazilian regions, ranging from 5.8% to 59.6% in HIV-seropositive populations (Barreto et al. 1993; Fandinho et al. 1997; Conde et al. 1999; Ferreira et al. 2002; Hadad et al. 2003; Ueki et al. 2005).
Southeastern Brazil has 79.7% of all notified AIDS cases and, associated with the southern region, the highest mortality rates (8.7/100 000 inhabitants) in the country (Brasil 2002). São José do Rio Preto city, the main city of the Northwestern region of São Paulo State, presents a high proportion of TB/HIV-co-infected individuals (Centro de Vigilância Epidemiológica 2006); hence it was assigned priority status by the National Programs of Tuberculosis and AIDS Control. The purpose of the present study was to evaluate the microbiological profile of Mycobacterium species isolated from HIV-infected patients attending the HIV/TB reference health care units from São José do Rio Preto, Brazil.
Material and methods
This was a retrospective analysis, conducted by the staff of the Adolfo Lutz Institute (IAL-SJRP) and Center for Microorganisms Investigation, Faculty of Medicine from São José do Rio Preto, Southeastern Brazil, Northwestern São Paulo State region. It was conducted according to a protocol approved by the Research Board of the Faculty of Medicine from São José do Rio Preto (CEP-FAMERP number 279/2006). We evaluated all HIV-1-positive patients whose IAL-SJRP laboratorial analysis was positive for Mycobacterium sp. after diagnosis of HIV infection, from January 2000 to December 2006. The IAL-SJRP represents the regional tuberculosis diagnostic centre in the São Paulo State Health Regional Department XV (DRS XV), which includes 101 municipalities. The 198 patients included were at least 18 years old and attended the Infectious and Parasitical Diseases Service from Hospital de Base (DIP-HB), Faculty of Medicine from São José do Rio Preto Foundation (n = 92) and the DST/AIDS Specialized Attendance Service (SAE) from the same city (n = 106). The HB is a tertiary care hospital catering to a large population from the Northwestern region of São Paulo State, Brazil, representing the regional centre of AIDS and tuberculosis treatment. The multidisciplinary assistance to the HIV/AIDS population in SJRP is located at the SAE, which includes a prevention and surveillance programme for HIV infection and for TB patients’ contacts.
The Clinical specimens were processed according to isolation techniques standardized in the guidelines of the Brazilian National Ministry of Health (Brasil, 2005). Acid-fast stain of each clinical material was performed using the Ziehl-Nielsen method and Mycobacterium sp.-positive cultures by the automated BacT/Alert system (BioMérieux, France) were sent to the IAL Central Laboratory for identification of the species through phenotyping (Kent & Kubica 1985; Morris & Harrison 2002) and molecular methods (Telenti et al. 1993; Brunello et al. 2001; Chimara et al. 2004). Briefly, the cultures were submitted to screening using macroscopic (morphology and pigmentation) and microscopic analyses (presence of cord factor). The strains were identified through phenotyping methods described previously (Ueki et al. 2005), which included the analysis of the time of growth using tests with picric acid, sodium nitrite, sodium chloride and common agar, as well as growth temperature (26, 37 and 45°C), photochromogenic test, biochemical tests (arylsulphatase – 3 and 15 days; β-galactosidase; nitrate reductase, catalase and urease activity; use of sugars; and Tween 80 hydrolysis test) and the test of growth in the presence of the p-nitrobenzoic acid. Molecular identification was carried out using the polymerase chain reaction and restriction enzyme analysis (PRA) method, based on the amplification of a 441-bp fragment of the hsp65 gene and on the analysis of the restriction profiles generated by the enzymes BstEII and HaeIII (Nusieve, FMC Bioproducts, Rockland, Maine, USA) as described by Chimara et al. 2008, with modifications.
All M. tuberculosis strains were sent to Adolfo Lutz Institute – Central Laboratory in São Paulo, SP (IAL-São Paulo) to investigate their drug susceptibility by the Bactec MGIT 960 Becton & Dickinson-BD system, according to the manufacturer’s instructions, to four drugs: isoniazid (INH), rifampicin (RMP), streptomycin (SM) and ethambutol (ETH). Multidrug-resistant tuberculosis (MDR-TB) was defined as Mycobacterium tuberculosis (MT) strains showing simultaneous resistance at least to INH and RMP (WHO 2004). Records pertaining to clinical and epidemiological data such as gender, age, prison record, therapy, immune status, chest X-ray, hospitalization and others were obtained from medical files. Statistical analysis was performed using the Minitab software (version 12.22). To obtain the independence among proportions, we applied the chi-square test or the Fisher’s exact test. The adopted significance level for statistical inference was 5%.
There were 198 HIV-seropositive individuals aged 20–70 years (mean age, 35.9 years; SD ±8.09) included in this study. As summarized in Table 1, 74.3% were men of whom 60% belonged to the ‘white’ ethnic group. Approximately half of all the individuals who had a Mycobacterium sp.-positive diagnosis were between 30 and 39 years old. Most of them were employed in positions that do not require specific qualification or schooling, such as construction and domestic service. Sexual behaviour was the most common (48.9%) risk factor for TB and/or HIV, followed by drug abuse (30.80%), detention (28.28%) and alcoholism (27.78%). There was a significant positive correlation between MT isolation and prison record (Chi-square test; P = 0.021).
Table 1. Distribution of the socio-demographic and clinical characteristics of the 198 Mycobacterium/HIV-positive individuals
Patients (n = 198)
Age at diagnosis
Between 30 and 39 years
Occupational status (employed)
Patients with MT (Table 2) were more frequently treated with antimycobacterial drugs (P = 0.0001), while no difference was observed regarding HAART therapy. The majority of MT and NTM patients were hospitalized at DIP-HB (more than 60%, both groups; P = 0.822). Severe immunosuppression and viral load did not differ between both groups. The mean CD4+ cell count was 115 ± 167.15/mm3 (range 1–1145).
Table 2. Clinical characteristics of the 112 Mycobacterium/HIV-infected individuals according to the isolation of MT or non-tuberculous mycobacteria (NTM) available for 198 patients
MT n (%)
NTM n (%)
*Ficher’s Exact test.
Viral load (HIV-1/ml)
Fifty-one per cent of participants had two samples collected (the same or different clinical specimen), which were further examined for microbiological characteristics. In total, we analysed 287 clinical specimens of both pulmonary (225 sputum specimens; eight bronchoalveolar lavage; seven gastric lavage) and non-pulmonary sites (28 blood cultures; 11 cervical aspirate; four pleural fluid; two cerebrospinal fluid; one ascitic fluid; one bone marrow aspirate).
Information from acid-fast stain allowed early detection of mycobacteria in 41% (81/198) of the patients. Culture identified 52.5% (104/198) of the patients as MT-positive. 42.4% (84/198) had NTM, while the remaining 5.1% (10/198) had MT/NTM-positive cultures (Figure 1, Table 3). Drug susceptibility testing revealed 11.45% of MT as resistant strains – 54% to streptomycin, 18.2% to rifampicin and 27.3% to more than one antimycobacterial drug. Overall, a 3.1% prevalence of MDR-TB was detected (isoniazid/rifampicin, isoniazid/rifampicin/ethambutol and isoniazid/rifampicin/streptomycin, one strain each). All resistant strains were isolated in 2000 and represented the patient’s first TB diagnosis.
Table 3. Mycobacteria laboratorial diagnosis for the 198 HIV-infected patients who provided 287 clinical specimens during medical care at HB-SJRPreto-SAE, São José do Rio Preto, São Paulo State
Patients (n = 198)
Clinical specimens (n = 287)
*Acid fast stain results were unavailable.
MT, Mycobacterium tuberculosis; NTM, non-tuberculous mycobacteria; AST, Antimycobacterial susceptibility test (one MT strain tested per patient, n = 96).
In 25% of patients where NTM were isolated, infection was confirmed according to American Thoracic Society (ATS 2007) criteria. Mycobacterium avium Complex (MAC) was responsible for bloodstream infections in 14 patients (66.7%); two others showed MAC pulmonary disease with positive X-ray findings, severe immunosuppression and higher viral loads; three individuals had cervical abscesses by the same bacteria.
Undoubtedly, TB/HIV co-infection persists as an important public health problem (Perkins & Cunningham 2007) and the numerous reports of increasing NTM infections point to the emergence of an additional concern. São José do Rio Preto city (population 400 000) is one of the 12 cities with higher coefficients of TB/HIV cases (CVE 2006) in São Paulo State and occupies the fifth place in AIDS incidence (CVE 2008). The IAL is the main governmental reference laboratory in this municipality and region for infectious and parasitical disease diagnosis and follow-up, including mycobacteriosis, tuberculosis and AIDS. Therefore, surveys performed at IAL can be considered representative of the Northwestern São Paulo State population. The socio-demographic profile of this Mycobacterium/HIV population is comparable to the socio-demographic profile of TB/HIV-co-infected populations in other previously published reports worldwide in terms of all studied variables, as well as to that of populations from other Brazilian major regions (Kerr-Pontes et al. 1997; Murcia-Aranguren et al. 2001; Aerts & Jobim 2004; Matos et al. 2004; Silveira et al. 2006). As expected, the positive association between prison record and MT isolation (28.3%) found in this study was greater than that previously published (6.3%) for a TB/HIV-co-infected population from Rio Grande do Sul, Extreme South of Brazil (Mattos et al. 2006). This is not surprising as the incidence index of the prison population in São Paulo State is 17.8 times higher than that of the general population (Nogueira & Abrahão 2009).
In 1996, Brazil took the pioneering decision to provide free HAART to all individuals with AIDS, which greatly contributed to increasing the survival of this population (Arruda et al. 2004; Goldberg et al. 2005). However, even with universal access to HAART, AIDS-related diseases are still major causes of death in some regions, especially pulmonary tuberculosis infection (BocK et al. 2007). In our study, there was no difference between MT or NTM infection prevalence among HAART-treated patients, although it was previously suggested that the introduction of HAART increases the relative frequency of NTM, while reduces MT detection (Murcia-Aranguren et al. 2001). On the other hand, as expected, NTM-infected patients were inadequately treated compared with the TB group because antimycobacterial drugs were not prescribed to approximately 17.8%. The impact of nosocomial transmission as an important source of infection due to the number of airborne expelled bacilli by TB/HIV-co-infected individuals during the period in-between hospital admittance and TB diagnosis is unknown (CDC, 2002). More than half of our patients were hospitalized at HB when MT diagnosis was established. This points to the need of establishing measures of local hospital mycobacteria surveillance, because HB as the regional reference centre for HIV/AIDS diagnosis and treatment deals with a great number of potentially TB-co-infected patients. Although tuberculosis is the most common opportunistic infection in HIV-seropositive persons and CD4+ T lymphocytes are most important in the protective response against M. tuberculosis (Rozman et al. 2007; Davoudi et al. 2008), CD4+ lymphocyte cell counts do not always correlate with the chance of developing tuberculosis, as observed for other infectious diseases (Rozman et al. 2007). Besides, the highest level of viral load can increase the susceptibility to disease (Abdool Karim et al. 2010). Additional studies focusing on patient follow-up after initial HIV and/or Mycobacterium diagnosis, in different populations, would be able to confirm such correlation.
The IAL-SJRPreto mycobacteria positivity rate for the acid-fast stain (approximately 37%) is higher than reported before in a survey conducted in an AIDS reference centre, Rio de Janeiro, Southeast Brazil, in the pre-HAART era (29.7%) among MT- and NTM-positive cultures from respiratory specimens (Conde et al. 1999). Regarding drug sensitivity, lower frequencies of MT resistance (11.5%) and MDR-TB (3.1%) were detected among the Northwestern São Paulo State HIV-infected patients than in other TB/HIV populations from Southeastern Brazil. Recently, Rozman et al. 2007 evaluated an HIV+ population from five municipalities of the São Paulo State littoral area where HIV prevalence is high, describing a resistance frequency of 18.5% and a high frequency of MDR-TB (10.6%). Similar frequencies, 16% and 8%, respectively, were found in an AIDS reference centre, Rio de Janeiro, RJ (Fandinho et al. 1997). These numbers are higher than those globally reported for the Brazilian population (1.3% and 9%, respectively) as well as for those from the different regions within São Paulo State, both independently of HIV status (Pereira Pinto et al. 1996; CVE 1998). Taken together, these data reinforce the impact of this co-infection on MT drug resistance in the most heavily populated Brazilian region.
Despite the fact that NTM infection prevalence increased as a global phenomenon in the post-HIV era, much remains to be determined in terms of well-defined diagnosis criteria and thus establishment of ‘disease’ and subsequent treatment (American Thoracic Society Documents 2007). It seems clear that the prevalence of pulmonary NTM varies by geographical region (Ferreira et al. 2002), although the majority of epidemiological studies have major limitations. For example, it is not mandatory to report NTM isolation and/or disease in many jurisdictions, the incomplete number of samples and erroneous identification by conventional methods, leading to underestimation of the diversity of the Mycobacterium genus (Ueki et al. 2005). In the present study, 42.4% of patients presented NTM. However, the recognition of the pathogenicity of those microorganisms is not an easy task because of the difficulty in associating clinical disease with positive cultures (Ueki et al. 2005). Furthermore, contamination does occur in the laboratory, especially during mycobacteria liquid culture (Perkins & Cunningham 2007), a limitation we also acknowledge in the present study. Nevertheless, 10.6% (21/198) of all HIV-seropositive patients met the ATS criteria for mycobacteriosis. Our results suggest that particular attention must be given to NTM single isolation in clinical specimens from HIV/AIDS individuals before assuming culture contamination and/or colonization by environmental mycobacteria. The necessity of NTM routine investigation by standardized methodology is therefore noticed.
A potential limitation of our study lies in the fact that a retrospective analysis may have been influenced by confounding factors, in particular incomplete patient records. Although the cases reported here are representative of the TB/HIV-co-infected community from Northwestern São Paulo State region, our data can be viewed as a starting point for further comparisons with the diverse Brazilian regions and with other developing countries. Furthermore, they provide important clues to the future understanding, prevention and control of such co-infections around the world.
We thank José Antônio Cordeiro from the Department of Epidemiology and Community Health/FAMERP for his assistance in statistical analysis, and the staff of the Adolfo Lutz Institute for the laboratory and administrative support. TC received a grant from the Conselho Nacional de Desenvolvimento e Pesquisa, PIBIC/CNPQ.