Seventy-two patients with clinical or histological findings compatible with extrapulmonary tuberculosis were studied. Patients attended four public hospitals in the Mexican cities of Durango, Zacatecas and Guadalajara. Durango City is located in north central Mexico, while Zacatecas and Guadalajara Cities are located in central Mexico. Specimens from the patients were prepared as either paraffin-embedded tissue or cultures of body fluids. For paraffin-embedded tissues, we studied recent and stored specimens obtained from the pathology departments of two hospitals from 2002 to 2008. In total, tissues of 42 patients were studied: twelve of them were obtained in Durango City and 30 in Zacatecas City. For body fluids, we studied recent and stored cultures in Löwenstein Jensen medium from urine, pericardial fluid, gastric aspirate, synovial fluid and cerebrospinal fluid obtained from 30 patients. These specimens were obtained in two hospitals in Guadalajara City from 2007 to 2008. The studied specimens were obtained only from the participating hospitals and are not representative of specimens from all hospitals in these cities. DNA was extracted from all specimens using a commercially available kit (QIAamp DNA Mini Kit, QIAGEN. Germany) following the instructions of the manufacturer. Prior to extraction, paraffin-embedded tissues were treated with xylene and alcohol to remove the paraffin, and bacterial cultures were heated at 95°C for 10 minutes.

DNA extracted from the 72 specimens was amplified by PCR using the INNO-LiPA MYCOBACTERIA v2 Amp. (Innogenetics NV, Gent, Belgium) kit according to the manufacturers' instructions. PCR parameters were: denaturation at 95°C for 1 min followed by 40 cycles of denaturation at 95°C for 30 sec, annealing at 62°C for 30 sec, and extension at 72°C for 30 sec. Electrophoresis of amplified products was performed in a 2% agarose gel. The gel was then stained with ethidium bromide and visualized by ultraviolet transillumination. In addition, amplified products were analyzed by line probe assay (INNO-LiPA MYCOBACTERIA v2. Innogenetics NV, Gent, Belgium) following the manufacturers' instructions. This DNA probe test targets the 16S–23S ribosomal RNA spacer region and detects and identifies the genus Mycobacterium and 16 different mycobacterial species. As a quality control for amplifications, we included water samples as negative controls, and M. tuberculosis complex DNA as positive controls in each run.

Amplification of the rifampicin resistance region of the gene encoding for the β-subunit of the RNA polymerase (rpoB) in DNA samples positive for M. tuberculosis complex by INNO-LiPA MYCOBACTERIA v2 was performed using the INNO-LiPA Rif.TB Amplification kit (Innogenetics NV, Gent, Belgium). The PCR parameters were denaturation at 95°C for 5 min followed by 30 cycles of denaturation at 95°C for 1 min, annealing at 55°C for 1 min, extension at 72°C for 1 min, and an elongation at 72°C for 10 min. After electrophoresis of the amplified products in 2% agarose gel, the gel was stained with ethidium bromide and visualized by ultraviolet transillumination. In addition, amplification products were analyzed by line probe assay (INNO-LiPA Rif.TB. Innogenetics NV, Gent, Belgium) for detection of rifampicin resistance. This assay was performed following the manufacturers' instructions. As a quality control for amplifications, we included water samples as negative controls, and M. tuberculosis complex DNA as positive controls in each run.

As shown in Tables 1 and 2, only 26 (36.1%) and 23 (31.9%) of the 72 DNA extracts were PCR positive for the 16S–23S ribosomal RNA spacer region of Mycobacterium species and LiPA hybridization, respectively. The low rate of positive samples could be explained by: 1) No Mycobacterium species were present in the samples: pathology diagnosis is not conclusive of tuberculosis; 2) DNA of Mycobacterium was present in the tissues but at levels that were undetectable by a single-round PCR assay; 3) DNA preservation in some samples was sub-optimal; and 4) the presence of PCR inhibitors. Most samples tested were archival and a comparable rate of positive PCR results in archival tissue samples was reported in a previous study 14. Moreover, our study was based in a single determination and it is possible that repeating assays could increase the rate of positive samples.

Table 1 shows the Mycobacterium species identified in paraffin-embedded tissues, and Table 2 shows results obtained for body fluids. In paraffin-embedded tissues, M. tuberculosis complex and M. genus were found in 5 and 3 lymph nodes, respectively. The predominant M. tuberculosis complex infection in lymph nodes found in our patients is consistent with results in patients with lymphadenitis reported by other researchers 41516. With respect to other tissues, M. genus was found in specimens of brain and vagina. Hybridization with M. genus probes indicates that amplified DNA belongs to mycobacteria but it is not one of the 16 Mycobacterium species included in the kit. In body fluids, M. tuberculosis complex and nontuberculous mycobacteria were found in 1 (10%) and 9 (90%) of urine samples with positive hybridization, respectively. The frequency of M. tuberculosis complex infection in these patients is lower than that reported in a study performed in Turkey, where 12 (70.5%) of 17 strains isolated from urine cultures of patients with suspected urinary tuberculosis were identified as M. tuberculosis complex 17. The finding of a high frequency of nontuberculous mycobacteria in urine may indicate contamination by mycobacteria in the environment. Nevertheless, all urine samples were obtained from symptomatic subjects with clinical diagnosis of renal tuberculosis and had obtained clinical improvement after anti-tuberculosis therapy. Many nontuberculous mycobacteria have recently been recognized as pathogenic 18. The species of nontuberculous mycobacteria found in the urine of our patients have been found in symptomatic patients in several studies: 1) M. fortuitum was found in 2 of 5 renal transplant recipients with urinary symptoms 19; 2) a case report of a woman suffering from urinary complaints describes that M. tuberculosis and other pathogens were not detected, but that M. gordonae was repeatedly isolated from urine. The patient responded to a standard anti-tuberculosis regimen 20; and 3) positive urine culture for M. kansasii in patients with persistent fever who were suffering from hairy cell leukemia and disseminated atypical mycobacterial infection 21. On the other hand, M. smegmatis has been isolated from skin and soft tissue infections 22, but to the best of our knowledge there is not any report of its isolation in urine from patients with urinary disease. The finding of nontuberculous mycobacteria in any specimen should always be correlated with clinical data of the patients to make appropriated therapeutic decisions. Since most Mycobacterium species found in extrapulmonary specimens in our symptomatic patients did not belong to the M. tuberculosis complex, our results emphasize the importance of performing identification of mycobacterial species. Discrimination of infecting Mycobacterium species is important since treatment differs 23. PCR and LiPA technology has been successfully used in identifying both Mycobacterium species and rifampicin resistance in M. tuberculosis isolates in either Mycobacterium grown in culture or directly in some clinical samples 24252627. However, this study is the first to identify and classify Mycobacterium species and to detect genotypic rifampicin resistance from paraffin-embedded tissues by using LiPA technology. In addition, this is the first report of Mycobacterium species diversity detected by molecular methods in symptomatic patients in Mexico suffering from extrapulmonary disease. Our results indicate that LiPA can be used successfully in analyzing DNA extracts of paraffin-embedded tissues.

Two of the seven M. tuberculosis complex samples were positive for amplification of the rpoB gene and showed hybridization patterns compatible with wild type M. tuberculosis. Samples negative for LiPA were not further analyzed. The two positive samples were obtained from lymph nodes and were considered sensitive to rifampicin. These results further provide molecular characterization of the M. tuberculosis complex isolates. Patients suffering from extrapulmonary tuberculosis have a delay in diagnosis 28. Therefore, fast molecular analysis of Mycobacterium using LiPA technology provides an aid for rapid diagnosis and the opportunity to take optimal preventive and treatment measures.