Correspondence

Misdiagnosis of non-tuberculous mycobacteria as tuberculous by the GeneXpert MTB/RIF Ultra: Fact or fiction?

Christoffel J. Opperman, Zethembe Ntshangase, Minkie Gumede, Sarishna Singh, Yonas Ghebrekristos, Rob Warren, Wynand Goosen

Received: 18 Sept. 2023; Accepted: 12 Dec. 2023; Published: 29 Mar. 2024

Copyright: © 2024. The Author(s). Licensee: AOSIS.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Introduction

GeneXpert detects Mycobacterium tuberculosis complex (MTBC) at low bacterial loads, exhibiting high sensitivity and specificity for smear-negative samples.¹ It was reported that GeneXpert^® MTB/RIF (Sunnyvale, CA, United States [US]) incorrectly identified several American Type Culture Collection group (ATCC) non-tuberculous mycobacteria (NTM) (Mycobacterium abscessus, Mycobacterium marinum, Mycobacterium smegmatis, Mycobacterium phlei and Mycobacterium aurum) as MTBC when high bacterial loads [10⁶ colony forming units/millilitre (CFU/mL)] were present.² In these studies, high (> 30) cycle threshold values for probe E resulted in M. abscessus and M. smegmatis flagging as rifampicin resistant.² Similar findings were noted between GeneXpert^® MTB/RIF probe A and M. intracellulare that produced positive signals, most likely because of a mismatch between the probe and the deoxyribonucleic acid (DNA) target.³ The outcomes and prognosis for a patient receiving anti-tuberculosis (TB) regimens when harbouring a clinically significant NTM is potentially poor, with a high risk of treatment failure.² Hence, this study aimed to investigate whether the GeneXpert^®MTB/RIF Ultra (Ultra, Sunnyvale, CA, USA) assay could yield true-negative results on high bacterial load dilutions of NTM commonly occurring in clinical samples. In addition, a secondary objective assessed the presence of the four probes covering the 81-base pair core region within the ribonucleic acid (RNA) polymerase β-subunit gene (rpoB), as identified by the Xpert Ultra assay, among NTM and among NTM and a selection of other bacteria.

Methods

A total of 12 NTM species obtained from clinical samples, and 11 other bacteria from an ATCC group were included in the study (Table 1). Pure NTM organisms were cultured on Löwenstein–Jensen media (Thermo Scientific™) slants at 35 °C ± 2 °C between 1 week and 4 weeks, dependent on the different NTM growth rates. A Ziehl-Neelsen stain was done on all NTM cultures to check for acid-fast bacilli and rule out contamination. Identification of NTM species was confirmed with Genotype^® Mycobacterium CM and AS (Bruker, Billerica, MA, US) line probe assays (LPA). Bacteria, distinct from NTM, encompassed in the investigation were cultured on 2% blood agar and boiled blood agar, with an incubation period of 24 h - 48 h (Table 1). The NTM and other bacteria were collected from the solid media and homogenised in sterile water to reach the desired 0.33 McFarland, representing 1 × 10⁸ CFU/mL. Xpert Ultra testing was performed in duplicate according to the manufacturer’s instructions.⁴

Results

Mycobacterium tuberculosis complex was not detected in any NTM or other bacterial samples evaluated in the cohort selected (Table 1). Furthermore, the rpoB2 probe signal was noticed in 10/12 NTM samples and 0/11 from other bacteria. From the 10 NTM with positive rpoB probe signals, 6 showed an isolated rpoB2 positive probe, while combinations ([rpoB1 + rpoB2]; [rpoB2 + rpoB3]; [rpoB2 + rpoB4]; [rpoB1 + rpoB2 + rpoB4]) together with other rpoB probes were positive in 4 isolates (Table 1).

Discussion

Non-tuberculous mycobacteria infection is an emerging disease with complex and extended antibiotic treatment regimens.⁵ A 5-year all-cause mortality study of > 9000 patients with NTM pulmonary disease estimated a case fatality rate of 27% (95% confidence interval: 21.3–37.8).⁶ Clinical and radiological features of pulmonary NTM disease can appear similar to MTBC disease.⁷ Therefore, accurate diagnosis of NTM disease remains essential. Xpert Ultra is a front-line diagnostic test for identifying TB in many countries, such as South Africa. The current study confirms the specificity of Xpert Ultra to accurately distinguish between NTM and MTBC, providing confidence in diagnosing TB and starting treatment if clinically indicated.

The current South African TB diagnostic algorithm places its primary emphasis on detecting MTBC at the time of a patient’s initial presentation. Consequently, it may overlook the presence of NTM unless cultures are specifically conducted for further LPA analysis. An Xpert Ultra MTBC-negative sample with a positive rpoB gene signal could suggest an NTM (Table 1). However, the absence of positive rpoB probes does not preclude the presence of NTM. In circumstances where MTBC is not present, the utilisation and interpretation of rpoB signals on Xpert Ultra remain unclear. This could be relevant on a case-by-case basis in recurrent Xpert Ultra-negative patients with constitutional symptoms and radiological chest X-ray changes, such as cavities, a scenario sometimes encountered in the clinical setting. Additional investigation is imperative with a more comprehensive exploration of these intriguing findings within the realm of NTM/MTBC diagnostics. It remains uncertain whether larger NTM cohorts will demonstrate similar results or if these findings hold any relevance within existing diagnostic algorithms. Further research is essential to shed more light on this matter. In addition, we acknowledge that future investigations should incorporate sputum samples to better reflect the in vivo environment. Lastly, the question of ‘… are we doing everything with the data we have …’ was asked at the PathRed 2023 conference in Johannesburg, South Africa.⁹ The current approach suggested in this study allows us to do more with the data we already have, providing the opportunity for deeper data mining of our TB diagnostic results.

Acknowledgements

We thank the public health laboratory, Green Point complex, National Health Laboratory Service for providing the American Type Culture Collection cultures.

The authors have declared that no competing interest exists.

C.J.O. conceptualised the project and drafted the manuscript. Z.N., M.G. and C.J.O. performed the laboratory experiments. W.G. supervised the sequencing of the genes. C.J.O., Z.N., M.G., W.G., S.S., Y.G. and R.W. performed data curation, formal analysis, edited, critically revised, and approved the final version.

Ethical approval was obtained from the Human Research Ethics Committee of Cape Town University (HREC reference number: 368/2023). Institutional approval to investigate non-tuberculous mycobacteria samples was acquired from the NHLS (SR3752435).

C.J.O. received funding from the National Health Laboratory Service (NHLS) Research Trust Development Grant (Reference: PR2232714) and Harry Crossley Foundation. R.W. acknowledges the funding from the South African Medical Research Council. W.G. recognises funding received from the Wellcome Foundation (Grant # 222941/Z/21/Z).

The data supporting this study’s findings are available from the corresponding author, [C.J.O], upon reasonable request upon NHLS institutional approval from a controlled access repository.

The views and opinions expressed in this article are those of the authors, and do not necessarily reflect official policy or position of any affiliated agency of the authors and the publisher.

References