Diagnostic and quantitative performance of the GeneXpert Mycobacterium tuberculosis/resistance to rifampin assay on pulmonary and extrapulmonary tuberculosis samples

Tummidi Santosh; Prashant Sood; Ashok Rattan; Deepa Dave; Nagiredla Puneeta

doi:10.4103/jacm.jacm_3_23

ORIGINAL ARTICLE

Year : 2023 | Volume : 25 | Issue : 1 | Page : 16-23

Diagnostic and quantitative performance of the GeneXpert Mycobacterium tuberculosis/resistance to rifampin assay on pulmonary and extrapulmonary tuberculosis samples

Tummidi Santosh¹, Prashant Sood², Ashok Rattan³, Deepa Dave⁴, Nagiredla Puneeta⁵
¹ Department of Pathology and Laboratory Medicine, All India Institute of Medical Sciences, Kalyani, West Bengal, India
² Department of Microbiology, All India Institute of Medical Sciences, Bilaspur, Himachal Pradesh, India
³ Chairman Medical Committee and Quality, Redcliffe Noida, India
⁴ National Director Operations, Pathkind Labs, Gurugram, India
⁵ Medical Officer, Vikash Multi-Speciality Hospital, Bargarh, Odisha, India

Date of Submission	18-Jan-2023
Date of Decision	08-Mar-2023
Date of Acceptance	06-Apr-2023
Date of Web Publication	1-Jun-2023

Correspondence Address:
Tummidi Santosh
Department of Pathology and Laboratory Medicine, All India Institute of Medical Sciences, Kalyani, West Bengal
India

Source of Support: None, Conflict of Interest: None

Check

DOI: 10.4103/jacm.jacm_3_23

Abstract

INTRODUCTION: No study has simultaneously assessed the diagnostic and quantitative performance of the GeneXpert Mycobacterium tuberculosis/resistance to rifampicin (MTB/RIF) assay on pulmonary and extrapulmonary samples processed in an Indian laboratory. We examined this to assess if GeneXpert can complement existing diagnostic facilities.
METHODS: The performance of GeneXpert and acid-fast microscopy were evaluated against culture in 455 pulmonary and 69 extrapulmonary specimens. Their sensitivity, specificity, predictive values and area under the curve (AUC) were estimated. GeneXpert cycle threshold (Ct) cut-offs were also computed to assess their utility in predicting smear positivity and time to culture positivity (TTCP).
RESULTS: GeneXpert was significantly more sensitive (100% vs. 84.4%; P < 0.001) and nearly as specific (94.8% vs. 96.7%; P = 0.344), with a greater AUC (0.97 vs. 0.91; P = 0.002) than acid-fast microscopy. It was sensitive and specific for pulmonary (100%; 94.1%) and extrapulmonary (100%; 96.5%) samples, and 100% sensitive for sputum, bronchoalveolar lavage (BAL), lymph node aspirates, pleural fluid, pus and urine. It detected smear-negative specimens with 100% sensitivity. GeneXpert Ct was inversely correlated with sputum smear grading (ρ = −0.66) and positively with TTCP (ρ =0.70). Mean Ct cut-offs of 21.4, 20.1 and 24.6 predicted smear positivity with maximum sensitivity and specificity in sputum, BAL and extrapulmonary samples, respectively. Rifampicin resistance was seen in 17% of samples.
CONCLUSIONS: GeneXpert can be a reliable first-line diagnostic for both pulmonary and extrapulmonary samples. It can effectively detect smear-negative and paucibacillary patients, and estimate smear positivity and bacillary burden. This can help Indian laboratories to detect cases early and interrupt community transmission.

Keywords: Cycle threshold, extrapulmonary, GeneXpert Mycobacterium tuberculosis/resistance to rifampin, pulmonary, tuberculosis

How to cite this article:
Santosh T, Sood P, Rattan A, Dave D, Puneeta N. Diagnostic and quantitative performance of the GeneXpert Mycobacterium tuberculosis/resistance to rifampin assay on pulmonary and extrapulmonary tuberculosis samples. J Acad Clin Microbiol 2023;25:16-23

How to cite this URL:
Santosh T, Sood P, Rattan A, Dave D, Puneeta N. Diagnostic and quantitative performance of the GeneXpert Mycobacterium tuberculosis/resistance to rifampin assay on pulmonary and extrapulmonary tuberculosis samples. J Acad Clin Microbiol [serial online] 2023 [cited 2023 Sep 30];25:16-23. Available from: https://www.jacmjournal.org/text.asp?2023/25/1/16/378069

Introduction

India carries the world's largest burden of tuberculosis with 6.4 million cases and 1.4 million annual deaths.^[1] Successfully tackling this massive health-care challenge requires rapid and reliable detection of cases for early isolation and treatment, and breaking community transmission. The WHO endorsed the GeneXpert Mycobacterium tuberculosis/resistance to rifampin (MTB/RIF) assay in 2010 and 2013 for the initial diagnosis of pulmonary and extrapulmonary tuberculosis in adults and children.^[2],[3],[4] This paved the way for the adoption of a rapid test which detects tuberculosis better than conventional acid-fast microscopy. GeneXpert's diagnostic superiority has been demonstrated in several pulmonary and extrapulmonary patient cohorts from across the globe.^{[3],[4],[5],[6],[7],[8],[9],[10],[11]} Some studies have also examined its quantitative performance at predicting smear positivity and mycobacterial burden.^{[2],[12],[13],[14],[15]}

It is well established that acid-fast microscopy and mycobacterial culture are operator-dependent and are affected by laboratory quality control and strain fitness.^{[12],[13],[16]} In contrast, GeneXpert is a robust, automated system that can largely overcome these sources of variation. With the growing adoption of GeneXpert in Indian laboratories, harnessing its quantitative capabilities in addition to its diagnostic prowess, for both pulmonary and extrapulmonary samples, can maximise its utility and complement existing diagnostic facilities. Maximising the utilisation of these diagnostic resources in a country with a burgeoning tuberculosis population and unequal healthcare access can help combat tuberculosis better. However, to the best of our knowledge, no study has simultaneously examined GeneXpert's diagnostic and quantitative performance for both pulmonary and extrapulmonary samples in a cohort of Indian samples. We undertook this study to address this.

Methods

This was a retrospective analysis of diagnostic microbiology data of all consecutive samples of presumptive tuberculosis patients received at our laboratory in Northen India from September 2017 to August 2018. The samples included pulmonary and extrapulmonary specimens from various body sites. All samples were collected after informed patient consent and as per institutional ethical guidelines. De-identified data were utilised for this analysis and the study results presented here did not influence the original diagnosis or management of the patients in any way.

Microbiological methods

The tuberculosis specimens received at the laboratory were decontaminated using the NALC-NaOH method except when they were collected from a sterile site such as pericardial fluid, synovial fluid, cerebrospinal fluid or peritoneal fluid.^{[3],[12],[16]} These latter samples, unless contaminated with blood, were directly concentrated by centrifugation and used further. Fine-needle aspirates (FNA) were raised to 2 ml by adding phosphate-buffered saline. Processed samples were divided into two equal parts, with one part being used for acid-fast microscopy and BACTEC MGIT 960 liquid cultures (Becton Dickinson, NJ, USA), and the second part for the GeneXpert MTB/RIF real-time polymerase chain assay. Acid-fast microscopy was carried out using Ziehl–Neelsen (ZN) staining and the results were graded using the International Union Against Tuberculosis and Lung Disease Scale.^[12] Time to culture positivity (TTCP) was recorded as the number of days that elapsed from culture inoculation to detection of a positive Mycobacteria Growth Indicator Tube (MGIT) result. The GeneXpert assays were carried out as per the manufacturer's instructions. The GeneXpert cycle threshold (Ct) was measured as the average number of polymerase chain reaction cycles after which each of the GeneXpert probes classified a sample as positive. Only samples with valid GeneXpert, culture and microscopy results were included in the analysis.

Statistical analysis

Descriptive statistics were used to describe the study cohort. The performance of GeneXpert was assessed using mycobacterial culture as the reference standard. Assay sensitivity, specificity, positive predictive value, negative predictive value, accuracy, area under the receiver operating characteristic curve area Under the Curve; (AUC) and kappa were computed for assessing GeneXpert's overall performance and in different sample subsets. The sensitivity and specificity of GeneXpert and acid-fast microscopy were compared using McNemar's test. Correlations of bacillary burden measured by the GeneXpert Ct, TTCP and acid-fast microscopy grading were calculated using Spearman's correlation (ρ). Optimal cut-offs of GeneXpert Ct were computed for predicting smear positivity and TTCP. Given the morbidity and mortality associated with tuberculosis, the optimal cut-offs were calculated by maximising their sensitivity and specificity, so as to minimise misdiagnoses and maximise the correct identification of smear-positive patients and their bacillary burden. Bias-corrected and accelerated confidence intervals (CI) for Spearman's ρ and optimal Ct cut-offs were computed from the distribution of 1000 bootstrapped samples. Statistical analyses were carried out using SPSS software v23.0(IBM Corp. Released 2015, Armonk, NY) and R v3.6.3. Optimal Ct cut-offs were computed using the optimal cutpoints v. 1.1-4 package.^[17] A P < 0.05 was taken as significant.

Results

Microbiology data of 524 presumptive tuberculosis samples with valid test results were included in this study. Of these 86.8% (455) were pulmonary and 13.2% (69) were extrapulmonary. Pulmonary samples included sputum (96.3%; 438/455) and bronchoalveolar lavage (BAL) (3.7%; 17/455). Extrapulmonary samples came from various body sites [Figure 1] and [Supplementary Figure 1]. While all 524 samples underwent acid-fast microscopy, 49.0% pulmonary and 92.8% extrapulmonary samples underwent culture and 60.2% pulmonary and 97.1% extrapulmonary samples were tested by GeneXpert. The culture was positive in 26.8% (77/287) samples overall, and smear in 30.0% (157/524). We did not have access to the clinical records of the patients and hence are unable to clarify why culture and/or GeneXpert were not requested for many of the patients. For further analysis, we used data with complete microscopy, culture and GeneXpert results.

Figure 1: (a) The distribution of pulmonary and extrapulmonary samples by site and their microbiology results. While a large majority of samples were sputum, common extrapulmonary samples included urine, pleural fluid and endometrial tissue. (b) The diagnostic and quantitative performance of GeneXpert were compared in subsets of samples with complete microscopy, GeneXpert and culture results. Please see Supplementary Figure 1 for a full break-up of results by sample subtypes. CSF: Cerebrospinal fluid; BAL: Bronchoalveolar lavage; ZN: Ziehl–Neelsen

Click here to view

Diagnostic performance of GeneXpert

With culture as the gold standard, GeneXpert correctly identified 96.2% of the samples. It reported an overall sensitivity of 100% (95% CI: 95.3–100) and specificity of 94.8% (95% CI: 90.8–97.4) [Table 1]. It was 100% sensitive for sputum, BAL, lymph node aspirates, pleural fluid, pus and urine. Overall, it performed best with sputum, reporting 100% (95% CI: 93.9–100) sensitivity, 95.2% (95% CI: 90.4–98.1) specificity and a predictive accuracy of 0.98 (95% CI: 0.95–0.99). However, its specificity (60.0%; 95% CI: 14.7–94.7) and predictive accuracy (AUC: 0.80; 95% CI: 0.53–0.95) dropped for BAL. For extrapulmonary samples, GeneXpert reported overall 100% (95% CI: 59.0–100.0) sensitivity and 96.5% (95% CI: 87.9–99.6) specificity, comparable to pulmonary samples, but with wider confidence intervals and lower positive predictive value (77.8% vs. 88.6%). It performed better with urine and pleural fluid, demonstrating 100% sensitivity and specificity but with wider confidence intervals due to small sample sizes [Table 1]. It showed the lowest concordance rates for pus (88.9%) and ascitic fluid (75.0%), with lower positive predictive value (75.0%). Overall, GeneXpert reported 100% negative predictive values for both pulmonary and extrapulmonary samples. We did not perform an in-depth performance assessment of extrapulmonary subsets with small numbers.

Table 1: GeneXpert Mycobacterium tuberculosis/resistance to rifampin mean cycle threshold cut-offs for various sample subsets, sputum Ziehl–Neelsen smear grading and sputum time to culture positivity

Click here to view

GeneXpert performed much better than acid-fast microscopy. Overall, GeneXpert was significantly more sensitive (100% vs. 84.4%; P < 0.001) and nearly as specific (94.8% vs. 96.7%; P = 0.344), with a greater AUC (0.97 vs. 0.91; P = 0.002), than acid-fast microscopy. It showed higher accuracy (96.2% vs. 93.4%; P = 0.003) and agreement (κ: 0.91 vs. 0.83) with culture than microscopy. This superior performance was consistent across pulmonary and extrapulmonary samples [Table 2].

Table 2: The performance of GeneXpert Mycobacterium tuberculosis/resistance to rifampin assay and Ziehl–Neelsen acid-fast microscopy with reference to mycobacterial liquid cultures

Click here to view

GeneXpert was 100% sensitive in detecting smear-cum-culture positive samples. Twelve samples (8 sputa, 2 BAL and 2 pus) were smear-negative but culture-positive. GeneXpert detected these samples also with 100% sensitivity classifying them with low to very low bacillary burden. These samples yielded a positive culture in median 31.5 days (range: 14–42 days). Seven samples (5 sputa, 1 pus, 1 lymph node FNA) were smear-positive but culture-negative. GeneXpert reported 57.1% of these samples as positive, classifying them with medium to a low bacillary burden, while microscopy graded them from 2+ to scanty. GeneXpert reliably ruled out 96.6% of smear-cum-culture negative samples. However, it reported 7 of these samples (4 sputa, 2 BAL, 1 ascitic fluid) as positive, classifying them with low to very low bacillary burden.

Out of the total 524 samples analysed in this study, 341 were tested with GeneXpert. Of these, 41.6% (142/341) came positive for tuberculosis. Rifampicin resistance was detected in 40.9% (58/142) of these positive samples. Rifampicin resistance was high in ascitic fluid (3/3), pleural fluid (1/1) and pus (2/4) samples, and comparatively lower in BAL (46.2%; 6/13) and sputum (39.3%; 46/117).

Quantitative performance of GeneXpert

Acid-fast smear positivity and grading are used to estimate bacillary burden in tuberculosis samples at the time of diagnosis.^[2] However, these measures have low sensitivity and tend to vary with the training, equipment and workload available in a laboratory.^[12] Hence, we assessed the quantitative ability of GeneXpert to assay these roles for various sample subsets, and more specifically for sputum samples, which was the largest subset in our cohort.

GeneXpert Ct was inversely correlated with acid-fast smear grading (ρ = −0.67; 95% CI: −0.75–−0.57; P < 0.001) for all samples put together, and for sputum samples (ρ = −0.66; 95% CI: −0.77–−0.53; P < 0.001) independently [Figure 2]a. The mean Ct of pulmonary samples (18.3; standard deviation [SD]: ±5.5) was significantly lower than that of extrapulmonary samples (23.3; SD: ±4.3) (P = 0.003). The mean Ct of smear-positive sputum samples (16.9; SD: ±4.8) was also significantly lower than smear-negative samples (24.4; SD: ±4.5) (P < 0.001). Furthermore, mean Ct values showed considerable variation and overlap across different sputum smear grades [Figure 2]a. GeneXpert provides semi-quantitative classification of samples based on their Ct values. We examined how well these categories corroborated with the sputum smear grading and found considerable variation at both ends of the grading spectrum. We found that 79.6% of the samples labelled “high” by GeneXpert had been graded ≥2+ on smear; 97.8% of “medium” samples were graded ≥1+; 92.0% of “low” samples were ≤1+; and 88.9% of “very low” samples had been graded ≤scanty.

Figure 2: Correlation of GeneXpert mean Ct with ZN smear grading and TTCP in sputum samples. (a) GeneXpert Ct was inversely correlated with ZN smear grading, and mean Ct values showed considerable overlap between successive grades. (b) In contrast, GeneXpert Ct was positively correlated with TTCP, with the mean Ct values showing considerable variation across samples with similar TTCP. Ct: Cycle threshold; ZN: Ziehl–Neelsen; TTCP: Time to culture positivity

Click here to view

In view of the inconsistency seen with GeneXpert's internal classification, we computed optimal Ct cut-offs for various sample subsets in our cohort. GeneXpert Ct cut-offs for smear positivity varied substantially across sputum (21.4), BAL (20.1), pulmonary (21.4) and extrapulmonary (24.6) samples [Table 1]. Besides these cut-offs maximised for sensitivity and specificity, a rule-out Ct cut-off for sputum smear positivity was determined at ≥26.6 (98.0% sensitivity, 95% CI: 92.9–99.8; 29.4% specificity, 95% CI: 10.3–56.0; 71.4% negative predictive value, 95% CI: 40.3–88.4). For sputum, the Ct cut-offs progressively reduced from 26.1 to 15.3 as the smear grading increased from scanty to 3+, reflecting increasing bacillary burden [Table 1]. The Ct cut-offs maximised the sensitivity and specificity for predicting each grade and came close to their respective mean Ct [Figure 2]a.

TTCP of a sample also provides an estimate of its bacillary burden. Mean TTCP for our pulmonary (20.9 days; SD: ±11.2) samples was lower than extrapulmonary (29.5 days; SD: ±10.7) (P = 0.056). GeneXpert Ct positively correlated with TTCP of all the samples put together (ρ =0.69; 95% CI: 0.57–0.78; P < 0.001) and of sputum samples (ρ =0.70; 95% CI: 0.56–0.80; P < 0.001) independently [Figure 2]b. However, considerable variation in Ct values was seen within samples with identical TTCP [Figure 2]b, possibly due to variations in sample processing and decontamination, and differences in strain fitness. Optimal Ct cut-offs for TTCP showed a progressively increasing trend from 14.4 to 25.3, as the TTCP increased from one to six weeks, reflecting decreasing bacillary burden with increasing TTCP.

Discussion

GeneXpert demonstrated high diagnostic efficacy in detecting tuberculosis in our samples. Its sensitivity (100% vs. 74.0%–100%) and specificity (95.2% vs. 86.0%–99.4%) for sputum were comparable to previous studies.^{[2],[6],[7],[9],[18],[19],[20]} It was highly sensitive (100% vs. 81.4%–90%) for BAL as well, albeit with lower specificity (60.0% vs. 93.4%–100%).^{[7],[10],[18]} This dip in specificity was due to the small sample size and GeneXpert reporting 40% (2/5) of culture-negative BAL specimens as positive. As with pulmonary samples, GeneXpert outperformed microscopy and culture with extrapulmonary specimens as well [Table 2]. When compared to previous studies, GeneXpert demonstrated high sensitivity (100% vs. 18.3%–100%) and specificity (96.5% vs. 73.3%–100%) for extrapulmonary specimens.^{[3],[5],[11],[20],[21],[22]} In particular, we noticed a high diagnostic performance for urine (sensitivity: 100% vs. 89.0%–94.6%; specificity: 100% vs. 86.5%–95.0%), pleural fluids (sensitivity: 100% vs. 43.7%–100%; specificity: 100% vs. 96.5%–100%) and pus (sensitivity: 100% vs. 86.0%–98.0%; specificity: 83.3% vs. 37.0%–58.0%) as compared to other studies.^{[3],[4],[5],[8],[21],[23]} However, since these sample subsets were small in size, their results need to be interpreted with caution and require further validation in larger cohorts.

Smear-positive pulmonary tuberculosis cases comprise the bulk of infectious cases in an endemic region. However, up to 37%–43% of cases in certain population subsets can be smear-negative, culture-positive patients.^[24] Despite carrying a lower bacillary burden, these smear-negative patients can account for 17% of community transmission.^[24] Smear-negative cases also run the risk of being misidentified and mismanaged if microscopy is used as the sole screening method. GeneXpert has been reported to detect such smear-negative cases with 52%–79% sensitivity.^{[2],[6],[7],[9],[18],[24],[25],[26]} In our cohort GeneXpert picked up these smear-negative cases with 100% sensitivity, in both pulmonary and extrapulmonary specimens. This demonstrates that despite its initial set-up cost, GeneXpert can be a cost-effective tool in the long run. By enabling early detection, isolation and treatment of smear-negative cases, it breaks community transmission and reduces long-term healthcare costs of complications and deaths. GeneXpert also picked up 57% of smear-positive, culture-negative cases in our study. Since both GeneXpert and ZN microscopy do not require metabolically fit mycobacteria for detecting a positive signal, these cases could likely be culture-negative due to reduced strain fitness or due to batch variations in decontamination procedures.^{[3],[12],[16]} Being a closed automated system, these cases are less likely to be contaminated.^[3] The detection of these cases underscores GeneXpert's strength in detecting additional cases that might be missed by culture.^{[3],[8],[10],[18s]} However, such GeneXpert-positive, culture-negative cases should be interpreted in light of the patient's radiological and clinical findings. The capacity of GeneXpert to identify true negative cases in smear-cum-culture negative samples was also high (96.6%). The few cases reported as positive in this category could also be potential tuberculosis cases but are best correlated clinically. We had no GeneXpert negative, culture-positive cases, suggesting reliable sample processing and elimination of polymerase chain reaction inhibitors.^[10]

Besides diagnostic performance, another key goal was to examine how well GeneXpert can quantitatively predict smear positivity and bacillary burden in pulmonary and extrapulmonary specimens. Such a comprehensive analysis has not been undertaken on Indian samples processed and analysed in an Indian laboratory. Since the comparison of GeneXpert's quantitative performance with smear microscopy and culture is influenced by the training and quality control of a laboratory,^[12],[16] it became pertinent to address this issue, especially with the growing adoption of GeneXpert in Indian laboratories. All the GeneXpert results included in this study had mean Ct values <34 as previously standardised, and <36 as recommended by the manufacturer to avoid false positives.^[2] In our cohort, GeneXpert mean Ct showed a similar correlation with sputum smear grading (−0.66 vs. −0.74–−0.55) as previously seen.^{[2],[12],[14],[16]} Mean Ct values for individual smear grades also fell within similar ranges as previously reported [Figure 2]a.^[14] We identified a mean Ct cut-off of 21.4 for predicting sputum smear positivity with maximum sensitivity and specificity, which is close to the 23.6 reported by Najjingo et al.^[14] However, studies have noted that rule-out Ct cut-offs are clinically more useful since they are less affected by the inherent variations of smear grading.^[15],[16] Hence, we also computed a mean Ct cut-off for ruling-out sputum smear positivity at ≥26.6, which was a shy lower than the rule-out cut-off range of 27.7–31.8 reported in a recent meta-analysis.^[13]

Few studies have attempted to compute the Ct cut-offs for predicting smear positivity in extrapulmonary and BAL samples. In our cohort, the pooled mean Ct cut-off for extrapulmonary specimens was 24.6, reflective of their paucibacillary nature. The mean Ct cut-off for BAL on the other hand was lower at 20.1, possibly due to targeted sample retrieval and higher bacillary loads. We also examined how mean Ct values correlated with TTCP in our samples. In contrast, to previous studies, we found a good correlation of 0.70 for sputum samples as compared to previous reports varying from 0.37 to 0.64,^{[2],[14],[16]} and more similar to the 0.68 reported by Blakemore et al.^[12] The good correlation between mean Ct and TTCP possibly reflects good laboratory control of sputum processing, although bacillary burden and strain fitness also contribute to this correlation.^[12],[16] We further estimated mean Ct cut-offs for weekly TTCP. We found a steady increase in the cut-off values with an increasing number of weeks to culture positivity, as would be expected with decreasing bacillary burden in specimens taking longer to turn culture-positive. As a corollary, these results also demonstrate that monitoring mean Ct values and their cut-offs can be a reliable means of monitoring the accuracy and consistency of smear, culture and GeneXpert results in our laboratories.^[12] We did not delve deeper into extrapulmonary subsets to determine their mean Ct cut-offs due to small sample sizes.

Despite the new findings reported here, this study has its limitations. A key limitation is the small number of extrapulmonary samples that were available in the year-long dataset. This prevented us from comprehensively evaluating GeneXpert's quantitative performance in different extrapulmonary subsets. The lack of access to clinical data also limited our comparison of GeneXpert's performance to culture as the sole reference standard. Studies have demonstrated that a composite reference standard comprising patients' clinical, radiology, microbiology and histopathology data, correlates better with GeneXpert's performance.^[3],[10] Unfortunately, the lack of clinical data also prevented us from identifying HIV-positive patients, which is important to monitor differences in GeneXpert's performance in these immunocompromised patients.^[2],[6],[16]

Conclusions

Notwithstanding its limitations, this study demonstrates the diagnostic and quantitative efficacy of GeneXpert MTB/RIF assay in pulmonary and extrapulmonary tuberculosis samples. GeneXpert can be effectively utilised to detect smear-negative and paucibacillary patients, estimate smear positivity, bacillary burden, monitor smear and culture performance. India laboratories equipped with GeneXpert should harness its diagnostic and quantitative strengths to complement existing diagnostic facilities, maximise early detection, contact tracing and treatment of infectious cases, and help disrupt community transmission. Our results on BAL and extrapulmonary samples show promise, and we intend to validate these further in well-powered cohorts with clinical data.

Author contributions (as per CRediT taxonomy)

Santosh Tummidi: Conceptualization (lead), investigations, data curation, writing – review and editing; Prashant Sood: Conceptualization (co-lead), formal analysis, data curation, writing – original draft, visualization; Ashok Rattan: Resources, supervision, project administration, writing – review and editing; Deepa Dave: Resources, supervision, project administration, writing – review and editing, Puneeta Nagiredla: Data curation, writing – review and editing.

Acknowledgements

Part of this work was presented at the MELAP National Conference 2020: Diagnostic Week, held online from 21^st to 27^th September 2020, from New Delhi, India.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

References

1.	World Health Organization, editor. Global Tuberculosis Report. France: World Health Organization; 2022. p. 2.
2.	Hanrahan CF, Theron G, Bassett J, Dheda K, Scott L, Stevens W, et al. Xpert MTB/RIF as a measure of sputum bacillary burden. Variation by HIV status and immunosuppression. Am J Respir Crit Care Med 2014;189:1426-34.
3.	Tadesse M, Abebe G, Bekele A, Bezabih M, Yilma D, Apers L, et al. Xpert MTB/RIF assay for the diagnosis of extrapulmonary tuberculosis: A diagnostic evaluation study. Clin Microbiol Infect 2019;25:1000-5.
4.	Saeed M, Ahmad M, Iram S, Riaz S, Akhtar M, Aslam M. GeneXpert technology. A breakthrough for the diagnosis of tuberculous pericarditis and pleuritis in less than 2 hours. Saudi Med J 2017;38:699-705.
5.	Vadwai V, Boehme C, Nabeta P, Shetty A, Alland D, Rodrigues C. Xpert MTB/RIF: A new pillar in diagnosis of extrapulmonary tuberculosis? J Clin Microbiol 2011;49:2540-5.
6.	Theron G, Peter J, van Zyl-Smit R, Mishra H, Streicher E, Murray S, et al. Evaluation of the Xpert MTB/RIF assay for the diagnosis of pulmonary tuberculosis in a high HIV prevalence setting. Am J Respir Crit Care Med 2011;184:132-40.
7.	Sharma SK, Kohli M, Yadav RN, Chaubey J, Bhasin D, Sreenivas V, et al. Evaluating the diagnostic accuracy of Xpert MTB/RIF assay in pulmonary tuberculosis. PLoS One 2015;10:e0141011.
8.	Pang Y, Shang Y, Lu J, Liang Q, Dong L, Li Y, et al. GeneXpert MTB/RIF assay in the diagnosis of urinary tuberculosis from urine specimens. Sci Rep 2017;7:6181.
9.	Luetkemeyer AF, Firnhaber C, Kendall MA, Wu X, Mazurek GH, Benator DA, et al. Evaluation of Xpert MTB/RIF versus AFB smear and culture to identify pulmonary tuberculosis in patients with suspected tuberculosis from low and higher prevalence settings. Clin Infect Dis 2016;62:1081-8.
10.	Lu Y, Zhu Y, Shen N, Tian L, Sun Z. Evaluating the diagnostic accuracy of the Xpert MTB/RIF assay on bronchoalveolar lavage fluid: A retrospective study. Int J Infect Dis 2018;71:14-9.
11.	Liu R, Li J, Tan Y, Shang Y, Li Y, Su B, et al. Multicenter evaluation of the acid-fast Bacillus smear, mycobacterial culture, Xpert MTB/RIF assay, and adenosine deaminase for the diagnosis of tuberculous peritonitis in China. Int J Infect Dis 2020;90:119-24.
12.	Blakemore R, Nabeta P, Davidow AL, Vadwai V, Tahirli R, Munsamy V, et al. A multisite assessment of the quantitative capabilities of the Xpert MTB/RIF assay. Am J Respir Crit Care Med 2011;184:1076-84.
13.	Lange B, Khan P, Kalmambetova G, Al-Darraji HA, Alland D, Antonenka U, et al. Diagnostic accuracy of the Xpert(®) MTB/RIF cycle threshold level to predict smear positivity: A meta-analysis. Int J Tuberc Lung Dis 2017;21:493-502.
14.	Najjingo I, Muttamba W, Kirenga BJ, Nalunjogi J, Bakesiima R, Olweny F, et al. Comparison of GeneXpert cycle threshold values with smear microscopy and culture as a measure of mycobacterial burden in five regional referral hospitals of Uganda- a cross-sectional study. PLoS One 2019;14:e0216901.
15.	Theron G, Pinto L, Peter J, Mishra HK, Mishra HK, van Zyl-Smit R, et al. The use of an automated quantitative polymerase chain reaction (Xpert MTB/RIF) to predict the sputum smear status of tuberculosis patients. Clin Infect Dis 2012;54:384-8.
16.	Beynon F, Theron G, Respeito D, Mambuque E, Saavedra B, Bulo H, et al. Correlation of Xpert MTB/RIF with measures to assess Mycobacterium tuberculosis bacillary burden in high HIV burden areas of Southern Africa. Sci Rep 2018;8:5201.
17.	Lopez-Raton M, Rodriguez-Alvarez MX, Cadarso-Suarez C, Gude-Sampedro F. OptimalCutpoints: An R package for selecting optimal cutpoints in diagnostic tests. J Stat 2014;61:1-36.
18.	Agrawal M, Bajaj A, Bhatia V, Dutt S. Comparative study of GeneXpert with ZN stain and culture in samples of suspected pulmonary tuberculosis. J Clin Diagn Res 2016;10:C09-12.
19.	Aricha SA, Kingwara L, Mwirigi NW, Chaba L, Kiptai T, Wahogo J, et al. Comparison of GeneXpert and line probe assay for detection of Mycobacterium tuberculosis and rifampicin-mono resistance at the national tuberculosis reference laboratory, Kenya. BMC Infect Dis 2019;19:852.
20.	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.
21.	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.
22.	Ghariani A, Jaouadi T, Smaoui S, Mehiri E, Marouane C, Kammoun S, et al. Diagnosis of lymph node tuberculosis using the GeneXpert MTB/RIF in Tunisia. Int J Mycobacteriol 2015;4:270-5. [Full text]
23.	Chen K, Malik AA, Nantasenamat C, Ahmed S, Chaudhary O, Sun C, et al. Clinical validation of urine-based Xpert^® MTB/RIF assay for the diagnosis of urogenital tuberculosis: A systematic review and meta-analysis. Int J Infect Dis 2020;95:15-21.
24.	Baker BJ, Holtom PD. Additional benefits of GeneXpert MTB/RIF assay for the evaluation of pulmonary tuberculosis among inpatients. Clin Infect Dis 2015;60:1287-8.
25.	Tadesse M, Aragaw D, Rigouts L, Abebe G. Increased detection of smear-negative pulmonary tuberculosis by GeneXpert MTB/RIF^® assay after bleach concentration. Int J Mycobacteriol 2016;5:211-8. [Full text]
26.	Walusimbi S, Bwanga F, De Costa A, Haile M, Joloba M, Hoffner S. Meta-analysis to compare the accuracy of GeneXpert, MODS and the WHO 2007 algorithm for diagnosis of smear-negative pulmonary tuberculosis. BMC Infect Dis 2013;13:507.