Logo Logo
Help
Contact
Switch language to German
Development and evaluation of urine based rapid molecular diagnostic test for pulmonary tuberculosis with potential for point of care: Cape Town Cohort
Development and evaluation of urine based rapid molecular diagnostic test for pulmonary tuberculosis with potential for point of care: Cape Town Cohort
Background: Current sputum-based diagnostic tests for tuberculosis (TB) do not serve all target populations, making optimal diagnosis difficult. TB diagnosis among scarce sputum patients is challenging for timely diagnosis and treatment initiation. Thus, an alternative non-sputum-based diagnostic test is urgently needed. Mycobacterium tuber-culosis (MTB)-specific transrenal (Tr) DNA from urine is a potential target for molecular testing to close this gap in TB diagnosis. The objective of this cross-sectional study was to evaluate the accuracy of a novel, molecular test using a portable fully-automatic analyser with the potential for usage in limited resource settings at the point of care. Materials/Methods: Spot urine, blood and sputum samples from 428 adults with sus-pected pulmonary TB (164 HIV positive, 263 HIV-negative) were collected at three clini-cal sites in Cape Town, South Africa. Tr-DNA was isolated from 4 ml of EDTA urine using an in-house method optimised for DNA fragments larger than 38 bp. A rapid double-stranded primer-based real-time PCR method targeting an MTB-specific direct repeat region was applied for Tr-DNA identification. The eluate was tested in triplicate using an automated molecular analyser with assay process controls included in each test run. The Tr-DNA-based test had a short time to result of 45 minutes including DNA isolation. Results: Liquid cultures were performed on 412 of the 428 TB suspects. Of these, 175/412 (42.5%) were microbiologically positive. Using liquid culture as gold standard, the Tr-DNA test showed sensitivity of 42.86% (95% CI: 35.42 – 50.54) and specificity of 86.61% (95% CI: 83.86 – 92.36). The Tr-DNA test had positive predictive value of 73.53% and negative predictive value of 67.74%. Among HIV-infected TB patients, the sensitivity and specificity were 45.24% and 89.04%, respectively. Combination of smear microscopy and Tr-DNA increased the sensitivity to 83.82% (smear microscopy alone was 75.14%); the specificity remained unchanged (96.61%). Conclusions: This multi-centre proof of concept study indicates that Tr-DNA has a high specificity and modest sensitivity. Furthermore, future study should be performed on larger cohort not restricted to pulmonary TB using fresh urine samples with patient follow up. Although unsuitable for implementation as a stand-alone test, it may have the potential, in combination with smear microscopy, to aid TB diagnosis in HIV-endemic regions where sputum scarce and extra-pulmonary TB are common.
Pulmonary tuberculosis, urine, transrenal DNA, diagnosis, PCR, HIV
Patel, Krutarth
2017
English
Universitätsbibliothek der Ludwig-Maximilians-Universität München
Patel, Krutarth (2017): Development and evaluation of urine based rapid molecular diagnostic test for pulmonary tuberculosis with potential for point of care: Cape Town Cohort. Dissertation, LMU München: Faculty of Medicine
[img]
Preview
PDF
Patel_Krutarth.pdf

3MB

Abstract

Background: Current sputum-based diagnostic tests for tuberculosis (TB) do not serve all target populations, making optimal diagnosis difficult. TB diagnosis among scarce sputum patients is challenging for timely diagnosis and treatment initiation. Thus, an alternative non-sputum-based diagnostic test is urgently needed. Mycobacterium tuber-culosis (MTB)-specific transrenal (Tr) DNA from urine is a potential target for molecular testing to close this gap in TB diagnosis. The objective of this cross-sectional study was to evaluate the accuracy of a novel, molecular test using a portable fully-automatic analyser with the potential for usage in limited resource settings at the point of care. Materials/Methods: Spot urine, blood and sputum samples from 428 adults with sus-pected pulmonary TB (164 HIV positive, 263 HIV-negative) were collected at three clini-cal sites in Cape Town, South Africa. Tr-DNA was isolated from 4 ml of EDTA urine using an in-house method optimised for DNA fragments larger than 38 bp. A rapid double-stranded primer-based real-time PCR method targeting an MTB-specific direct repeat region was applied for Tr-DNA identification. The eluate was tested in triplicate using an automated molecular analyser with assay process controls included in each test run. The Tr-DNA-based test had a short time to result of 45 minutes including DNA isolation. Results: Liquid cultures were performed on 412 of the 428 TB suspects. Of these, 175/412 (42.5%) were microbiologically positive. Using liquid culture as gold standard, the Tr-DNA test showed sensitivity of 42.86% (95% CI: 35.42 – 50.54) and specificity of 86.61% (95% CI: 83.86 – 92.36). The Tr-DNA test had positive predictive value of 73.53% and negative predictive value of 67.74%. Among HIV-infected TB patients, the sensitivity and specificity were 45.24% and 89.04%, respectively. Combination of smear microscopy and Tr-DNA increased the sensitivity to 83.82% (smear microscopy alone was 75.14%); the specificity remained unchanged (96.61%). Conclusions: This multi-centre proof of concept study indicates that Tr-DNA has a high specificity and modest sensitivity. Furthermore, future study should be performed on larger cohort not restricted to pulmonary TB using fresh urine samples with patient follow up. Although unsuitable for implementation as a stand-alone test, it may have the potential, in combination with smear microscopy, to aid TB diagnosis in HIV-endemic regions where sputum scarce and extra-pulmonary TB are common.