The Establishment And Preliminary Application Of The Method For Detection Of Tuberculosis From Sputum Based On Variable Number Of Tandem Repeat

Objective:The conventional nucleic acid amplification(NAA) tests for diagnosis of tuberculosis(TB) were widely applied by clinical with the character of simple,quick and high sensitive,but for lack of effective quality guarantee measures lead to some false positive result which was its main drawback. In this study,mycobacterial interspersed repetitive unit–variable number of tandem repeat(MIRU-VNTR) was first applied to detect TB as a NAA method from sputum samples for quality control,and clinical application of MIRU-VNTR was evaluated by the performance of diagnosis of TB.Methods:According to the amplification conditions and clinical detection needs, MTUB21, MUTB04, QUB-18, QUB-26, QUB-11 b, MIRU31, MIRU10 and MIRU26 were selected as test targets. 130 sputum samples from patients with pulmonary tuberculosis and 200 specimens from other lung diseases were detected by VNTR and fluorescent quantitative reverse transcription polymerase chain reaction(FQ-PCR). The results of VNTR and FQ-PCR were compared with Lowenstein-Jensen(L-J)culture and clinical diagnosis.Results:Regarding the results of L-J culture as the golden standard, the sensitivity and specificity of VNTR was 93.1%(108/116) and 97.7%(209/214), respectively. The sensitivity and specificity of FQ-PCR was 94.0%(109/116) and 96.7%(207/214). The difference of accuracy was not statistically significant by chi-square test(χ2 =0.352, P=0.569 ﹥ 0.05). Using the clinical diagnosis as the golden standard, the sensitivity and specificity of VNTR was 86.9%(113/130) and 100%(200/00), respectively. The sensitivity and specificity of FQ-PCR was 87.7%(114/130) and 99.0%(198/200). The difference of accuracy was not statistically significant by chi-square test(χ2 =0.030,P=0.862﹥0.05), this method based on VNTR had no false-positive result in the research. Of 113 VNTR positive samples,the molecular code of 98.2%(111/113) was different from others, only 2 samples had the same code(5-4-6-8-5-5-3-8). According to repeated testing, the 2 samples were TB not caused due to cross contamination.Conclusion:Compared with conventional NAA tests,The method provided us a approach to separate one positive sample from other positive one in the same batch and control the false positive result by giving each sample a different identity number. The research suggested that MIRU-VNTR provided us a promising application prospect for diagnosis of clinical TB.