|Year : 2018 | Volume
| Issue : 3 | Page : 357-363
Diagnostic accuracy of Xpert MTB/RIF assay in extrapulmonary tuberculosis
Sheetal Bankar1, Reena Set1, Disha Sharma1, Daksha Shah2, Jayanthi Shastri1
1 Department of Microbiology, T.N. Medical College and B Y L Nair Ch. Hospital, Mumbai, Maharashtra, India
2 City TB Officer, Mumbai District Tuberculosis Control Society, Mumbai, Maharashtra, India
|Date of Web Publication||14-Nov-2018|
Dr. Reena Set
Department of Microbiology, T.N. Medical College and B Y L Nair Ch. Hospital, Mumbai, Maharashtra
Source of Support: None, Conflict of Interest: None
Introduction: The WHO endorsed Xpert Mycobacterium tuberculosis/rifampicin (MTB/RIF) assay, has been evaluated for pulmonary TB in a number of studies but very few have investigated it for extrapulmonary specimens. The present study evaluates the performance of Xpert MTB/RIF assay in the diagnosis of extrapulmonary TB (EPTB). Aim and Objectives: The aim of the study is to determine sensitivity and specificity of Xpert MTB/RIF assay for diagnosis of EPTB and RIF resistance in comparison to culture on Lowenstein–Jensen (LJ) medium and proportion method (PM), respectively. Materials and Methods: A total of 738 specimens from clinically suspected cases of EPTB were subjected to Ziehl–Neelsen staining, Xpert MTB/RIF assay and culture on LJ medium. PM was done on MTB isolates. Results: The sensitivity, specificity of Xpert MTB/RIF assay for diagnosis of EPTB were 84.91% (95% confidence interval [CI] 72.41%–93.25%) and 86.72% (95% CI 83.94%–89.17%) and for RIF resistance detection were 60.00% (95% CI 32.29%–83.66%) and 94.74% (95% CI 73.97%–99.87%), respectively. Among culture-positive cases, the sensitivity of Xpert MTB/RIF assay was 94.12% in smear positive and 80.56% in smear-negative cases. Xpert MTB/RIF showed maximum sensitivity of MTB detection from lymph node specimens (100% [95% CI 54.07%–100.00%]) and other body fluids (100% [95% CI 15.81%–100.00%]). Conclusion: The present study establishes Xpert MTB/RIF assay as a promising tool in the rapid diagnosis of EPTB and detection of RIF resistance.
Keywords: Extrapulmonary tuberculosis, rifampicin resistance, Xpert MTB/RIF assay
|How to cite this article:|
Bankar S, Set R, Sharma D, Shah D, Shastri J. Diagnostic accuracy of Xpert MTB/RIF assay in extrapulmonary tuberculosis. Indian J Med Microbiol 2018;36:357-63
|How to cite this URL:|
Bankar S, Set R, Sharma D, Shah D, Shastri J. Diagnostic accuracy of Xpert MTB/RIF assay in extrapulmonary tuberculosis. Indian J Med Microbiol [serial online] 2018 [cited 2019 Nov 18];36:357-63. Available from: http://www.ijmm.org/text.asp?2018/36/3/357/245383
| ~ Introduction|| |
Tuberculosis (TB) is a major global health problem. The South East Asian (SEA) Region bears an inordinately high share of the global TB burden. The SEA region accounts for nearly half (45.6%) of the global burden in terms of TB incidence. Of this, India (23%) alone accounts for one-fourth of the TB burden worldwide. According to the WHO, there were an estimated 10.4 million new TB cases worldwide in 2015. India alone accounted for 2.8 million TB cases. Extrapulmonary TB (EPTB) constitutes about 15%–20% of all cases of TB. Out of 580,000 new cases of multidrug-resistant TB (MDR-TB) globally, 1.3 lakh MDR/Rifampicin resistance (RR) TB patients were in India.
The diagnosis of EPTB is challenging due to the paucibacillary nature of specimens, difficulty in obtaining specimens from deep-seated organs and inability to get an additional specimen. As a result, failure to diagnose and treat affected patients leads to increased morbidity and mortality, development of secondary resistance (including extensively drug-resistant TB) and ongoing transmission of disease.
Microscopy is often negative in extrapulmonary specimens; in addition, it is impossible to differentiate non-tuberculous mycobacteria (NTM) from Mycobacterium tuberculosis (MTB). Culture through the gold standard for TB diagnosis often leads to considerable delays, compromising patient care and outcome. Furthermore, the lack of diagnostic infrastructure substantially aggravates the problem of EPTB diagnosis. Not only the rapid detection of TB is essential but also the early determination of MDR/RR status is a necessity.
With the increase in the number of drug-resistant TB patients around the world, appropriate diagnosis and treatment of different presentations of the disease need a special attention. Hence, rapid detection of TB and early determination of MDR/RR status should be the priority.
As TB is a curable disease, the role of accurate and rapid diagnosis and treatment is vital for improving the global picture. The WHO endorsed Xpert MTB/rifampicin (RIF) assay provides a diagnosis of TB and RIF resistance (RR) within 2 h. It is not prone to cross-contamination, requires minimal biosafety facilities and minimal training. These characteristics make it potentially an attractive tool for extrapulmonary specimens. The introduction of Xpert MTB/RIF assay has helped in bridging the gap between diagnosis and treatment.
Studies have been conducted in EPTB samples. However, biopsy and tissue samples had not been included in Singh et al.'s study. Vadwai et al.'s study had very few cerebrospinal fluid (CSF) and body fluids samples. Although Sharma et al.'s study had a large sample size, they had not elaborated on the sample-wise sensitivity and specificity.,,
Therefore, this study was envisaged to evaluate the sample-wise diagnostic accuracy of Xpert MTB/RIF assay for diagnosis of EPTB and RR TB in our tertiary care setting.
| ~ Materials and Methods|| |
This was a single-centric, clinical microbiology laboratory-based prospective study conducted in the Department of Microbiology of a tertiary care setting in Mumbai, India. As per Institutional policy ethical approval was not required.
A total of 738 consecutive EPTB samples from January 2016 to February 2017, were analysed in the study. The EPTB specimens comprised of pus (n = 132), biopsy (n = 23), tissues (n = 05), aspirates (vitreous and body lesions) (n = 06), lymph nodes (n = 70), CSF (n = 100), bronchoalveolar lavage (n = 16), pleural fluid (n = 304) and other fluids (n = 82) (peritoneal, pericardial, synovial and ascitic fluids) [Figure 1]. Pus specimens included breast abscess, liver abscess, paravertebral abscess and psoas abscesses. Lymph nodes were sampled from cervical and mediastinal nodes.
Processing of samples
The samples collected by invasive procedures were processed directly. The sterile body fluids were centrifuged, and pellet was used as inoculum.
Acid-fast bacilli smears
The processed specimens were used for making smears for all the samples. Samples were stained with Ziehl–Neelsen (Z-N) method and examined with a light microscope.
Samples were inoculated in Lowenstein–Jensen (LJ) medium. Positive cultures were confirmed by secondary smears, culture on Para-nitrobenzoic acid, niacin and nitrate test.
Drug susceptibility testing
Drug susceptibility testing (DST) was performed for the all the positive cultures. RR was tested using economic variant of proportion method (PM) according to standard protocol.
Xpert MTB/RIF assay
The Xpert assay was performed according to manufacturer's instructions (Cepheid, Sunnyvale, CA, USA) using the latest version of G4 cartridges. In the Xpert assay, sample reagent was added at a 2:1 ratio to clinical specimens and it was incubated for 15 min at room temperature with intermittent shaking. Following which, 2 ml of the inactivated material was transferred to the cartridge. The cartridges were inserted into the test device, and the results were generated after 90 min. Repeat GeneXpert MTB/RIF tests were run for results that gave invalid, error and indeterminate results.
Bias was minimised through blinding since technicians performing conventional method, and Xpert MTB/RIF assay was unaware of the results of other tests.
Reference standard comparison
Xpert MTB/RIF assay results were compared with that of LJ culture. Furthermore, the RR detection was compared with DST on LJ medium by PM.
The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of Xpert assay were compared to TB culture on LJ medium as a reference standard using contingency two-by-two tables. P < 0.05 was considered as statistically significant. All statistical analyses were performed using MedCalc statistical software version 17.9.6.
| ~ Results|| |
Out of total 769 samples, 31 (4.03%) samples were excluded from the study (contaminated culture – 23 [2.99%], 2 [0.26%] invalid result and 2 [0.26%] error on Genexpert and 4 [0.52%] non-tuberculous mycobacteria). Of the 738 samples included in the study, culture detected MTB in 53 samples, compared to 136 by Xpert MTB/RIF assay. The positivity rate in various specimens by different methods is depicted in [Table 1].
Of the 23 specimens whose cultures got contaminated, GeneXpert detected RIF sensitive MTB in 4 specimens, which would have got undetected but for Xpert assay. Of the two invalid results on GeneXpert one for pleural fluid and other for LN biopsy, repeat test on pleural fluid did not detect MTB whereas repeat test on LN biopsy could not be done, as specimen was used up for 1st test and could not obtain 2nd specimen. Repeat tests were performed for results that gave errors, as one was pus sample and other was pleural fluid. In both the samples, MTB was not detected. All the four NTMs, gave 'MTB not detected' result by Xpert MTB/RIF assay.
On comparison with culture, ZN smear microscopy showed sensitivity and specificity of 32.08% (95% confidence interval [CI] 19.92%–46.32%) and 94.01% (95% CI 91.97%–95.67%) respectively [Table 2].
|Table 2: Overall sensitivity and specificity of Ziehl-Neelsen smear with reference to Lowenstein-Jensen culture (n=738)|
Click here to view
On comparison with culture, overall sensitivity, specificity, PPV and NPV of Xpert MTB/RIF assay was 84.91% (95% CI 72.41%–93.25%), 86.72% (95% CI 83.94%–89.17%), 33.09% (95% CI 28.36%–38.18%) and 98.67% (95% CI 97.51%–99.29%), respectively, the positive and negative likelihood ratio was 6.39 (95% CI 5.12–7.98) and 0.17 (95% CI 0.09–0.33), respectively. The diagnostic accuracy of Xpert MTB/RIF assay was 86.59% (95% CI 83.94%–88.86%) [Table 3].
|Table 3: Overall sensitivity and specificity of Xpert MTB/RIF assay with reference to Lowenstein-Jensen culture (n=738)|
Click here to view
Among the diverse clinical samples, Xpert MTB/RIF assay could detect 10.57% more positivity compared to smear microscopy and 11.25% more compared to culture [Table 1]. The sensitivity and specificity of Xpert MTB/RIF assay varied markedly among different clinical samples. The sensitivity was highest for lymph node and other body fluids followed by pus, CSF and pleural fluid samples. Tuberculous meningitis being a challenging diagnosis with high mortality, Xpert MTB/RIF assay offers a diagnostic advantage with a sensitivity of 71.43% (95 CI 29.04%–96.33%) [Table 4].
|Table 4: Sensitivity and specificity of Xpert MTB/RIF assay for various specimens with Lowenstein-Jensen culture method as reference standard|
Click here to view
Among culture-positive specimens, Xpert MTB/RIF assay showed a sensitivity of 94.12% (95% CI 71.31%–99.85%) in smear-positive cases and 80.56% (95% CI 63.98%–91.81%) in smear-negative cases. Although the sensitivity was low in smear-negative samples, statistically, no significant difference was seen (P = 0.41) [Table 5].
|Table 5: Comparison of Xpert MTB/RIF assay results with respect to Ziehl-Neelsen smear results|
Click here to view
Of the 53 culture-positive samples, DST was performed on all the isolates. However, 16 isolates gave invalid and 37 gave valid DST results. Of 37 isolates, RR was seen in 17 isolates (45.95%). In DST, RR was maximum in pus samples 18.92% (7/37). Xpert MTB/RIF assay detected RR in 34/136 (25%) samples, 101 were RIF sensitive and one was indeterminate. RR was highest in pus samples (13/34), followed by lymph node and pleural fluid (7/34 each). The comparison between DST results by a conventional method and RR by Xpert MTB/RIF assay was done in 34 isolates (16 isolates gave invalid DST results by conventional method and 3 samples were MTB not detected by Xpert MTB/RIF assay). Of the 34 MTB isolates, Xpert MTB/RIF assay detected Rif resistance in 9 (26.47%) isolates [Table 6]. The sensitivity, specificity, PPV and NPV with 95% CI for RIF resistance detection by Xpert MTB/RIF assay was 60.00% (32.29%–83.66%), 94.74% (73.97%–99.87%), 90.00% (56.10%–98.45%) and 75.00% (61.53%–84.91%), respectively. There was no statistically significant difference between Xpert MTB/RIF assay and conventional DST method for detection of MDR TB (P = 0.31). Resistance to both RIF and isoniazid was seen in 6/37 isolates whereas 9/37 isolates were resistant to RIF and sensitive to isoniazid.
|Table 6: Detection of Rifampicin resistance by Xpert MTB/RIF assay in comparison to drug susceptibility testing on Lowenstein Jensen medium|
Click here to view
Xpert test has the advantage in providing the bacillary load in the sample. It was 'very low' in 8 samples, 'low' in 75, medium in 26 samples and high in 1 sample. Maximum specimens had low and medium bacillary load, which was detectable by Xpert assay.
| ~ Discussion|| |
With different clinical presentations, EPTB poses a diagnostic challenge. With EPTB on the rise, exploration of rapid diagnostic possibilities is the need of an hour. This study is one such effort towards evaluating the performance of Xpert MTB/RIF assay which will add valuable information to the existing literature. The study was performed on large number (769) of EPTB samples only next to studies done by Singh et al., Vadwai et al. and Tortoli et al.,,
The sensitivity and specificity of Xpert MTB/RIF assay in our study was 84.91% and 86.72%. The observed sensitivity of Xpert MTB/RIF assay is fairly consistent with other published studies in which reported sensitivities ranged from 77.3% to 100% [Table 7].,,,,,,, The heterogeneity between studies may be due to differences between patient population, diversity of clinical specimens, the quality of samples, type of EPTB, differences in sample processing and the diagnostic gold standard.
|Table 7: Summary of studies in which performance of Xpert MTB/RIF assay for extrapulmonary tuberculosis was assessed|
Click here to view
Among the culture-positive specimens, the assay showed a good sensitivity in smear-positive cases (94.12%). It showed a reasonable sensitivity in smear-negative samples as well (80.56%). This correlated well with the other published studies with sensitivity ranging from 94% to 100% for smear-positive cases [Figure 2],,, and 63% to 83% for smear-negative specimens. Statistically, no significant difference was seen in Xpert MTB/RIF assay sensitivity in smear positive and negative cases (P = 0.41). The Xpert MTB/RIF assay showed a very good specificity of 90.99% (95% CI 88.51%–93.09%) in culture-positive and smear-negative samples. High specificity suggests that Xpert MTB/RIF assay has a useful role as rapid rule – in test for diagnosis of EPTB in smear-negative specimens.
|Figure 2: Forest plot showing Sensitivity and specificity of Xpert Mycobacterium tuberculosis/rifampicin assay in smear-positive culture-positive specimens in various studies|
Click here to view
Overall, culture detected MTB in 53/738 samples (7.18%) as against Xpert MTB/RIF assay which detected MTB in 136/738 samples (18.43%). Low positivity of culture could be due to paucibacillary nature of EPTB samples and a tendency of MTB to form clumps leads to uneven distribution of bacilli. Furthermore, Xpert Assay sample reagent has better homogenisation and liquefaction efficiency.
This study shows that Xpert MTB/RIF assay has a better diagnostic potential when compared to culture with a good sensitivity of 96%–100% for specimens such as lymph nodes and other fluids (i.e., ascitic, pericardial, synovial and peritoneal fluids) and pus when compared with other published studies.,,,,,, In our study, low sensitivity was seen with pleural fluids. In this study, CSF showed a moderate sensitivity of 71.43% (95% CI 29%–96.33%) that was similar to other studies. Our study showed a higher sensitivity for CSF samples as compared with a study by Vadwai et al. which showed a lower sensitivity of 29% [Figure 3].
|Figure 3: Forest plot showing sensitivity and specificity of Xpert Mycobacterium tuberculosis/rifampicin assay in various studies|
Click here to view
In this study, the sensitivity and specificity for RR detection by Xpert MTB/RIF assay was 60% (95% CI 32.29%–83.66%) and specificity of 94.74% (95% CI 33.2%–89.65%) when compared to DST by conventional method. The sensitivity was low when compared to other studies.,, which could be due to lower sample size (34) in our study. Statistically, no significant difference was seen in RR detection by Xpert MTB/RIF assay and DST (P = 0.31).
Six false-negative RR was detected by Xpert MTB/RIF assay in the present study. These discordant results might be due to mutation outside the RR determining the region, which needs to be solved by sequencing of rpoB gene. One false-positive result was given by the assay. This discrepancy could be due to mixed TB infections. Mixed MTB infections have been suggested to be responsible for false negative and positive results for RIF by Xpert. Heteroresistance defined as the presence of both sensitive and resistant MTB populations is often suggested to be responsible for discordant DST results.
One limitation of the present study was that the culture was done on solid medium. The pre-processing of specimens with NALC NaOH method was not done, which could have resulted in less contamination in culture. Another limitation of the study was that since this was primarily a laboratory-based study, the clinical and radiological findings could not be obtained. The composite gold standard (treatment, radiology and clinical response) would have improved the specificity of the study. However, despite the limitations, the results of the study remain significant since it shows the promising diagnostic potential of the test.
| ~ Conclusion|| |
Overall, the Xpert MTB/RIF assay has a considerably good sensitivity and specificity. Its performance varies with the clinical specimens. It is a valuable diagnostic tool for smear-negative EPTB cases. Undoubtedly, with its short turnaround time along with detection of both MTB and MDR status, Xpert MTB/RIF assay perfectly fits the requirements of Indian health-care settings.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| ~ References|| |
Bending the Curve – Ending TB: Annual Report 2017. Regional office for South East Asia. World Health Organization; March, 2017.
Global Tuberculosis Report, 2016. Geneva: World Health Organization; 2016.
Sharma SK, Mohan A. Extrapulmonary tuberculosis. Indian J Med Res 2004;120:316-53.
Iram S, Zeenat A, Hussain S, Wasim Yusuf N, Aslam M. Rapid diagnosis of tuberculosis using Xpert MTB/RIF assay – Report from a developing country. Pak J Med Sci 2015;31:105-10.
Lawn SD, Zumla AI. Diagnosis of extrapulmonary tuberculosis using the Xpert(®) MTB/RIF assay. Expert Rev Anti Infect Ther 2012;10:631-5.
Mirsaeidi SM, Tabarsi P, Edrissian MO, Amiri M, Farnia P, Mansouri SD, et al.
Primary multi-drug resistant tuberculosis presented as lymphadenitis in a patient without HIV infection. Monaldi Arch Chest Dis 2004;61:244-7.
World Health Organization, Rapid Implementation of the Xpert MTB/RIF Diagnostic Test. World Health Organization; 2011.
Singh UB, Pandey P, Mehta G, Bhatnagar AK, Mohan A, Goyal V, et al.
Genotypic, phenotypic and clinical validation of GeneXpert in extra-pulmonary and pulmonary tuberculosis in India. PLoS One 2016;11:e0149258.
Vadwai V, Boehme C, Nabeta P, Shetty A, Alland D, Rodrigues C, et al
. Xpert MTB/RIF: A new pillar in diagnosis of extrapulmonary tuberculosis? J Clin Microbiol 2011;49:2540-5.
Sharma SK, Kohli M, Chaubey J, Yadav RN, Sharma R, Singh BK, et al.
Evaluation of xpert MTB/RIF assay performance in diagnosing extrapulmonary tuberculosis among adults in a tertiary care centre in India. Eur Respir J 2014;44:1090-3.
Revised National Tuberculosis Control Programme (RNTCP) Training Manual for Mycobacterium tuberculosis
Culture and Drug Susceptibility Testing. Central TB Division; 2009.
World health Organization. Standard Operating Procedure [SOP]: Specimen Processing of CSF, Lymph Nodes and Other Tissues for Xpert MTB/RIF. World health Organization; 2014.
MedCalc Statistical Software version 17.9.6 MedCalc Software Bvba, Ostend, Belgium. Available from: http://www.medcalc.org; 2017
. [Last accessed on 2017 Oct 25].
Tortoli E, Russo C, Piersimoni C, Mazzola E, Dal Monte P, Pascarella M, et al
. Clinical validation of Xpert MTB/RIF for the diagnosis of extrapulmonary tuberculosis. Eur Respir J 2012;40:442-7.
Hillemann D, Rüsch-Gerdes S, Boehme C, Richter E. Rapid molecular detection of extrapulmonary tuberculosis by the automated GeneXpert MTB/RIF system. J Clin Microbiol 2011;49:1202-5.
Zeka AN, Tasbakan S, Cavusoglu C. Evaluation of the GeneXpert MTB/RIF assay for rapid diagnosis of tuberculosis and detection of rifampin resistance in pulmonary and extrapulmonary specimens. J Clin Microbiol 2011;49:4138-41.
Zmak L, Jankovic M, Jankovic VK. Evaluation of xpert MTB/RIF assay for rapid molecular diagnosis of tuberculosis in a two-year period in Croatia. Int J Mycobacteriol 2013;2:179-82. [Full text]
Malbruny B, Le Marrec G, Courageux K, Leclercq R, Cattoir V. Rapid and efficient detection of Mycobacterium tuberculosis
in respiratory and non-respiratory samples. Int J Tuberc Lung Dis 2011;15:553-5.
Maynard-Smith L, Larke N, Peters JA, Lawn SD. Diagnostic accuracy of the xpert MTB/RIF assay for extrapulmonary and pulmonary tuberculosis when testing non-respiratory samples: A systematic review. BMC Infect Dis 2014;14:709.
Marouane C, Smaoui S, Kammoun S, Slim L, Messadi-Akrout F. Evaluation of GeneXpert MTB/RIF for the detection of Mycobacterium tuberculosis
and resistance to rifampin in extra-pulmonary specimens. Int J Mycobacteriol 2014;4:101.
Scott LE, McCarthy K, Gous N, Nduna M, Van Rie A, Sanne I, et al
. Comparison of xpert MTB/RIF with other nucleic acid technologies for diagnosing pulmonary tuberculosis in a high HIV prevalence setting: A prospective study. PLoS Med 2011;8:e1001061.
Tadesse M, Abebe G, Abdissa K, Aragaw D, Abdella K, Bekele A, et al.
GeneXpert MTB/RIF assay for the diagnosis of tuberculous lymphadenitis on concentrated fine needle aspirates in high tuberculosis burden settings. PLoS One 2015;10:e0137471.
Zetola NM, Shin SS, Tumedi KA, Moeti K, Ncube R, Nicol M, et al
. Mixed Mycobacterium tuberculosis
complex infections and false-negative results for rifampin resistance by geneXpert MTB/RIF are associated with poor clinical outcomes. J Clin Microbiol 2014;52:2422-9.
Van Deun A, Barrera L, Bastian I, Fattorini L, Hoffmann H, Kam KM, et al. Mycobacterium tuberculosis
strains with highly discordant rifampin susceptibility test results. J Clin Microbiol 2009;47:3501-6.
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6], [Table 7]