Original Research

Comparative evaluation of the Xpert MTB/XDR and FluoroType XDR assays for detecting fluoroquinolone resistance in multidrug-resistant tuberculosis clinical isolates

Kabelo G. Kaapu, Mmule K. Mathebe, Ndivhuho A. Makhado, Molebogeng R. Lekalakala-Mokaba, Ivy Rukasha

Received: 04 Nov. 2025; Accepted: 28 May 2026; Published: 21 Sept. 2026

Copyright: © 2026. The Authors. Licensee: AOSIS.
This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license (https://creativecommons.org/licenses/by/4.0/).

Abstract

Background: Accurate detection of fluoroquinolone (FLQ) resistance in multidrug-resistant tuberculosis (MDR-TB) is essential for effective treatment.

Aim: To compare Xpert MTB/XDR and FluoroType XDR for detecting FLQ resistance in MDR-TB.

Setting: National Health Laboratory Service diagnostic laboratory, Polokwane, South Africa.

Methods: Seventy-eight MDR-TB isolates collected from January 2024 to July 2024 were tested using Xpert MTB/XDR, FluoroType XDR and GenoType MTBDRsl line probe assay (LPA) as the reference standard. Whole-genome sequencing (WGS) data from 35 isolates were used to resolve discordant results. Diagnostic accuracy and agreement were assessed using sensitivity, specificity and Cohen’s kappa.

Results: Xpert MTB/XDR showed 97.7% sensitivity (95% confidence interval [CI]: 87.9–99.8) and 100.0% specificity (95% CI: 89.7–100.0). FluoroType XDR showed 100.0% sensitivity (95% CI: 90.6–100.0) and 97.2% specificity (95% CI: 85.5–99.9), with five indeterminate results. Compared directly with Xpert, FluoroType showed 94.9% sensitivity and 100.0% specificity. Agreement was strong (κ = 0.95; 95% CI: 0.88–1.00). WGS resolved two discordant cases: one resistant isolate with a *gyrA* Asp94Gly mutation and one susceptible isolate without resistance-associated mutations.

Conclusion: Both assays performed well for FLQ resistance detection. FluoroType XDR showed higher sensitivity but more indeterminate results. Strong assay agreement and WGS resolution of discordances support their use in MDR-TB diagnostics.

Contribution: This study demonstrates strong concordance between Xpert MTB/XDR and FluoroType XDR for FLQ resistance detection and highlights the value of WGS in resolving discordant results.

Keywords: multidrug-resistant tuberculosis; fluoroquinolone resistance; Xpert MTB/XDR; FluoroType XDR; whole-genome sequencing; molecular diagnostics.

Introduction

Tuberculosis (TB) remains a leading global infectious disease, despite decades of coordinated control efforts.¹ According to the World Health Organization (WHO), an estimated 10.8 million people developed TB in 2023, a slight increase from 10.6 million in 2021 and 10.7 million in 2022, indicating a sustained upward trend since the disruption of TB services due to the coronavirus disease 2019 (COVID-19) pandemic.¹ It is alarming that TB has once again regained its position as the leading cause of death from a single infectious agent, with 1.25 million deaths reported in 2023, up from 1.2 million in 2022.¹ Of increasing concern is the rise in multidrug-resistant TB (MDR-TB), defined as resistance to at least rifampicin (RIF) and isoniazid (INH), which continues to complicate and undermine treatment success, especially in high-burden resource-limited settings.²

There is a growing concern regarding the increasing prevalence of fluoroquinolone (FLQ) resistance within MDR-TB treatment regimens. Fluoroquinolones, especially levofloxacin and moxifloxacin, are essential to current all-oral MDR-TB regimens recommended by WHO.^3,4 However, resistance to FLQs is increasingly reported, even among individuals with no prior TB treatment, indicating community-level transmission of resistant strains.⁵ Fluoroquinolone resistance significantly undermines treatment efficacy, increases the risk of progression to extensively drug-resistant TB (XDR-TB), and is linked to high mortality rates, with some South African cohorts reporting XDR-TB death rates exceeding 60%.²

Accurate and timely detection of FLQ resistance is critical for effective MDR-TB treatment and TB control. In fact, FLQ resistance is now a defining feature in the WHO’s classification of pre-XDR and XDR-TB. Molecular diagnostics such as Xpert MTB/XDR and FluoroType XDR can detect resistance-associated mutations in gyrA and gyrB genes, offering rapid alternatives to conventional phenotypic methods.⁶

Among available molecular tests, the Xpert MTB/XDR assay, run on the GeneXpert platform, is the most widely used globally. It is simple to operate, requires minimal training, and delivers results in under 2 h, making it well suited for decentralised, high-burden settings. The Xpert MTB/XDR-TB assay is WHO-endorsed for detecting resistance to INH, FLQs, and other second-line drugs and shows strong diagnostic performance. However, it may fail to detect certain low-level mutations such as S91P or D94A.^7,8

In contrast, the FluoroType XDR assay, used on the FluoroCycler XT platform, is designed for centralised laboratories with stronger infrastructure. The assay detects a broader range of mutations in gyrA and gyrB, even in samples with low bacterial load. However, it can produce indeterminate results, which may limit its use in settings with limited technical capacity.^9,10

This study selected Xpert MTB/XDR as the primary comparator of the FluoroType XDR due to its widespread use in TB diagnostic programmes and its practical relevance in both low- and high-resource settings. To strengthen diagnostic comparisons, we included the GenoType MTBDRsl assay as a reference standard, given its well-established role in detecting second-line drug (SLID) resistance. Whole-genome sequencing (WGS) was further employed to resolve discordant results, ensuring robust validation of FLQ resistance detection. Together, these molecular tools provide a comprehensive framework for evaluating diagnostic accuracy and guiding individualised treatment strategies.

This study focuses on FLQ resistance for three main reasons. Firstly, FLQ resistance is the most clinically significant second-line resistance and directly influences patient eligibility for key MDR-TB regimens. Secondly, both molecular assays evaluated in this study detect mutations in the same genetic targets, making direct performance comparisons meaningful. Thirdly, limiting the scope to FLQ resistance allows for a focused, high-resolution evaluation of diagnostic performance in a high-burden, resource-limited setting.

The study setting, Limpopo province in South Africa, adds further significance to this work. As a predominantly rural, high-burden province with limited diagnostic capacity and direct borders with Zimbabwe, Mozambique, and Botswana, Limpopo faces a heightened risk of imported drug-resistant TB strains.^11,12 Evaluating the performance of rapid molecular diagnostics in this context is essential to improve timely detection and effective treatment of FLQ-resistant MDR-TB. Thus, the aim of the study was to assess the accuracy and agreement of Xpert MTB/XDR and FluoroType XDR assays in detecting FLQ resistance in MDR-TB clinical isolates.

Research methods and design

Study design and samples

This was a cross-sectional diagnostic accuracy study aimed at comparing the performance of the Xpert MTB/XDR and FluoroType XDR assays for detecting FLQ resistance in MDR-TB isolates. We utilised 78 cryopreserved isolates collected from the National Health Laboratory Service (NHLS) Polokwane TB repository between January 2024 and July 2024. These isolates were previously confirmed to be MDR-TB using line probe assay (LPA) according to the manufacturer (Hain Lifescience GmbH, Nehren, Germany): MTBDRplus for first-line drugs INH and RIF and MTBDRsl for second-line drugs (SLID and FLQ). The demographic information was obtained from the laboratory information system. All isolates were subcultured in Mycobacterial Growth Indicator Tube (MGIT) tubes containing Middlebrook 7H9 broth and incubated in the BACTEC MGIT 960 system (Becton, Dickson and Company [BD], Franklin Lakes, NJ, United States [US]) until flagged as positive. The positive MGIT tubes were then tested using three assays: Xpert MTB/XDR, FluoroType XDR, and GenoType MTBDRsl, with the latter used as the primary reference. Of the 78 MDR-TB isolates, 35 had WGS data already available and were used to resolve any discordant results.

Study setting

This study was conducted at the NHLS routine diagnostic laboratory, Polokwane, South Africa. The NHLS Polokwane laboratory serves as a key provincial referral centre, providing diagnostic services to the five districts across Limpopo province, a region recognised for its high burden of drug-resistant TB. The laboratory serves as a critical referral laboratory in supporting complex diagnostic workflows and maintains a repository of Mycobacterium tuberculosis (MTB) isolates, positioning it as a central hub for TB diagnosis and surveillance in the mostly rural region. The laboratory routinely performed molecular detection tests and drug susceptibility testing for TB using the GenoType MTBDRplus assay.

Sampling technique

We used purposive sampling to select 78 MDR-TB clinical isolates from the NHLS Polokwane drug-resistant TB repository. The cryopreserved isolates were chosen based on confirmed GenoType MTBDRplus status, viability for subculture, and availability from the repository for the period of January 2024 to July 2024 after ethical approval was granted. The study used the most recent cultures, with a diverse and representative set of drug resistance statuses, to evaluate the performance of the molecular assays.

Subculture and quality control

The cryopreserved isolates were subcultured in MGIT tubes containing Middlebrook 7H9 broth, supplemented with oleic acid, albumin, dextrose, catalase (OADC), and polymyxin B, amphotericin B, nalidixic acid, trimethoprim, azlocillin (PANTA) (BD, Franklin Lakes, NJ, US), and incubated in the BACTEC MGIT 960 system until they flagged positive, following manufacturer’s instructions. The subcultures were left in the BACTEC MGIT 960 system at 37 °C for an additional 2 weeks to increase bacterial biomass. Quality control were checked for contamination before analysis by plating subcultured isolates onto blood agar plates and incubating at 37 °C and monitoring for 48 h for growth.

GenoType MTBDRsl

The GenoType MTBDRsl version 2.0 (Hain Lifescience GmbH, Nehren, Germany), the extended gold standard for second-line drug resistance detection, is an LPA used to detect resistance to second-line anti-TB drugs, including FLQs and SLIDs. It targets mutations in the gyrA and gyrB genes for FLQ resistance and rrs and eis promoter regions for injectable drug resistance. In this study, it was used as the reference standard for evaluating the performance of Xpert MTB/XDR and FluoroType XDR in detecting FLQ resistance. The test involves deoxyribonucleic acid (DNA) extraction, polymerase chain reaction (PCR) amplification, and reverse hybridisation, with banding patterns visually interpreted following the manufacturer’s instructions. Only the FLQ resistance results were analysed for comparison purposes.

FluoroType XDR (LiquidArray MTB-XDR) assay

The FluoroType XDR (Hain Lifescience GmbH, Nehren, Germany) is a real-time PCR-based assay designed to detect mutations associated with resistance to second-line anti-TB drugs, including FLQs. The assay uses automated DNA extraction followed by multiplex PCR amplification and fluorescence-based detection on the FluoroCycler XT platform. It targets key resistance-associated regions in the gyrA and gyrB genes for FLQ resistance. In this study, all testing was performed according to the manufacturer’s instructions, with results interpreted using the accompanying FluoroSoftware (Hain Lifescience GmbH, Nehren, Germany). Indeterminate results were recorded when amplification failed or when the software could not assign a clear resistance call. The software reports ‘IND’ for indeterminate results and ‘invalid’ if controls fail.

Xpert MTB/XDR assay

The Xpert MTB/XDR assay (Cepheid Inc., Sunnyvale, CA, US) is a cartridge-based, automated real-time PCR test that detects MTB complex DNA and resistance-associated mutations to several key anti-TB drugs, including FLQs, INH, ethionamide, and SLIDs. For FLQ resistance, the assay targets mutations in the gyrA and gyrB genes. Testing was performed using the Xpert MTB/XDR platform following the manufacturer’s protocol. The system provides results in under 90 min, with minimal hands-on time, and automatically interprets susceptibility based on detected mutations. No indeterminate results were recorded for this assay in this study.

DNA extraction and whole-genome sequencing

For the WGS assay, a 1.2 mL aliquot of each MGIT culture with M. tuberculosis complex (MTBC) clumps was transferred to a 1.5 mL tube containing glass beads and heat-inactivated for 20 min at 80 °C in the Biosafety Level 3 laboratory (BSL3). Heat-inactivated aliquots were centrifuged for 5 min in the Biosafety Level 2 (BSL2) laboratory to pellet MTB cells, and 0.6 mL of the supernatant was discarded. Genomic DNA was extracted from the cultured MTB isolate using the NucliSENS easyMAG assay (BioMérieux, Marcy-l’Étoile, France) from a concentrated 0.06 mL volume of positive culture 7H9 MGIT medium. The DNA extracts were quantified using the Qubit 4 Fluorometer (Life Technologies, Carlsbad, CA, US).

Whole-genome sequencing was performed using the MiSeq (Illumina, Cambridge, United Kingdom [UK]) sequencing system. Library preparation was performed using the Nextera-XT library preparation kit (Illumina, Cambridge, UK), and sequencing was performed using the 2 × 300 base pairs (bp) MiSeq cartridge v.3 (Illumina, Cambridge, UK) with a target of 30 × –50 × paired coverage (approximately 100× coverage). Magma bioinformatics pipeline was used to detect resistance associated variants. Variant analysis focused primarily on gyrA and gyrB resistance-associated mutations relevant to FLQ resistance using reference mapping against the annotated reference genome MTB H37Rv (NC00962.3) and the quality-based variant analysis tools.¹³

Data analysis

The results were loaded into Microsoft Excel (Microsoft Corp., Redmond, WA, US) for cleaning and categorisation. To assess and compare the diagnostic accuracy of the Xpert MTB/XDR and FluoroType XDR assays for detecting FLQ resistance, we performed statistical analysis to determine the performance indices using MedCalc (Ostend, Belgium). Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated using the GenoType MTBDRsl assay as the reference standard. For the MTB isolates with conflicting results across assays, WGS was used to confirm the true resistance profile. Agreement between Xpert MTB/XDR and FluoroType XDR was evaluated using Cohen’s Kappa statistic. The genomic data were analysed using the bioinformatics tool (QIAGEN CLC Genomics Workbench) to detect resistance-associated variants.

Ethical considerations

Permission to conduct the study was obtained from the University of Limpopo’s Turfloop Research Ethics Committee (TREC/1432/2024: UG). This approval was obtained before the start of the study, and all ethical guidelines were strictly adhered to throughout the research process. Approval to access the samples was also granted by NHLS through their Academic Affairs and Research Management System (PR2345512). This study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki.

Results

Diagnostic performance of Xpert MTB/XDR and FluoroType XDR compared to GenoType MTBDRsl

Table 1 presents the diagnostic performance of the Xpert MTB/XDR and FluoroType XDR assays for detecting FLQ resistance, using GenoType MTBDRsl as the primary reference assay.

Xpert MTB/XDR demonstrated a sensitivity of 97.7% (95% confidence interval [CI]: 87.9–99.8) and specificity of 100.0% (95% CI: 89.7–100.0), with PPV of 100.0% and NPV of 97.1%.

FluoroType XDR showed a sensitivity of 100.0% (95% CI: 90.6–100.0) and specificity of 97.2% (95% CI: 85.5–99.9), with PPV of 97.4% and NPV of 100.0%.

Notably, FluoroType XDR had five indeterminate results, while Xpert MTB/XDR reported none.

These results indicate high diagnostic accuracy for both assays, with FluoroType XDR slightly outperforming in sensitivity but exhibiting a few indeterminate outcomes.

Comparative performance between FluoroType XDR and Xpert MTB/XDR assays

Table 2 compares the performance of FluoroType XDR against Xpert MTB/XDR as a comparator. FluoroType XDR identified 37 FLQ-resistant and 34 FLQ-sensitive cases in concordance with Xpert but reported 2 FLQ-sensitive results where Xpert indicated resistance.

Based on this comparison, FluoroType XDR achieved a sensitivity of 94.9% (95% CI: 82.7–98.5) and specificity of 100.0% (95% CI: 89.7–100.0). The PPV was 100.0%, and NPV was 94.4%.

Despite slight discordance, FluoroType XDR showed excellent agreement with Xpert MTB/XDR, as further supported by high predictive values.

Resolution of discordant fluoroquinolone resistance results using whole-genome sequencing

Table 3 details two cases with discordant FLQ resistance results across molecular assays (Xpert MTB/XDR, FluoroType XDR, and MTBDRsl), and their resolution using WGS. Not all isolates had corresponding WGS results; thus, not all possible discrepancies were resolved.

For example, case OM6733 was identified as resistant by Xpert but as sensitive by both FluoroType and MTBDRsl; WGS confirmed the absence of resistance-associated mutations, and the isolate was reclassified as sensitive.

Conversely, case NP2143 was reported as sensitive by FluoroType but resistant by Xpert and MTBDRplus; WGS detected a gyrA Asp94Gly mutation, confirming resistance.

This table underscores the utility of WGS in clarifying assay discrepancies and highlights the clinical significance of discordant molecular results.

Concordance between FluoroType XDR and Xpert MTB/XDR assay using Cohen’s Kappa statistics

Agreement between the FluoroType XDR and Xpert MTB/XDR assays in detecting FLQ resistance was assessed using Cohen’s Kappa statistic. A total of 73 results were interpretable by both assays, excluding indeterminate outcomes. Among these, 37 samples were classified as FLQ-resistant by both assays, and 34 as FLQ-sensitive, while two samples were reported resistant by Xpert MTB/XDR but as sensitive by FluoroType XDR. No samples were found to be resistant by FluoroType and sensitive by Xpert.

Agreement between FluoroType XDR and Xpert MTB/XDR was almost perfect (κ = 0.95, 95% CI: 0.88–1.00; p < 0.001) (Table 4). This strong concordance highlights the FluoroType XDR assay as a reliable alternative to Xpert MTB/XDR in detecting FLQ resistance among MDR-TB clinical isolates.

Comparison of operational characteristics and time to result between the three assays

Table 5 summarises the operational characteristics and turnaround times of three molecular assays for FLQ resistance detection. Xpert MTB/XDR demonstrated the shortest time to result (~90 min), minimal hands-on time, low skill requirements, and high suitability for decentralised laboratories, making it the most user-friendly option. FluoroType XDR, while providing results in 3 h – 4 h, requires high technical skill and specialised infrastructure and entails greater hands-on time, limiting its use to well-equipped central laboratories. GenoType MTBDRsl offers moderate performance, with results available within one working day but also demands considerable technical expertise and laboratory infrastructure. Cost considerations ranged from low–moderate for Xpert MTB/XDR, moderate for GenoType MTBDRsl, and high for FluoroType XDR.

Discussion

This study evaluated the diagnostic performance of two molecular assays, the Xpert MTB/XDR and the FluoroType XDR, for detecting FLQ resistance in MDR-TB clinical isolates from the Limpopo Province of South Africa. To our knowledge, there are currently no published studies that directly evaluated the FluoroType XDR assay for detecting FLQ resistance, making this one of the first comparative analyses of its kind. The inclusion of FLQs in the bedaquiline-pretomanid-linezolid including moxifloxacin (BPaL/M) (bedaquiline-pretomanid-linezolid including levofloxacin [BPaL/L] for South Africa) regimen for drug-resistant TB highlights the need for rapid and accurate FLQ resistance detection.^3,4 Our study is therefore timely and necessary, as these diagnostic tools can guide prompt treatment decisions, ensuring patients receive the correct regimen sooner, thereby improving outcomes and helping prevent the spread of drug-resistant TB, particularly in high-burden, resource-limited settings like Limpopo.

Both assays demonstrated high diagnostic accuracy, meeting or exceeding the WHO’s minimum performance targets for molecular TB diagnostics.¹ Xpert MTB/XDR showed a sensitivity of 97.7% and specificity of 100%, while FluoroType XDR achieved 100% sensitivity and 97.2% specificity. These results indicate that both assays demonstrated strong performance in detecting FLQ-resistant MDR-TB isolates, supporting their use in routine diagnostics without the delays associated with phenotypic methods.

Our findings are in line with WHO’s diagnostic performance benchmarks, which recommend a minimum sensitivity and specificity of 90% for detecting FLQ resistance. Xpert MTB/XDR exceeded both thresholds, while FluoroType XDR achieved complete (100%) sensitivity. This reinforces the clinical reliability of both platforms in supporting timely and effective MDR-TB management.

The slightly higher values reported in our study may reflect differences in study design. We used archived MGIT culture isolates, which typically contain higher bacillary loads than direct sputum samples, potentially improving assay performance.¹⁴ This suggests that both assays may perform even better in reference laboratory settings where cultured isolates are tested.

In contrast, there are currently no studies that have evaluated the performance of the FluoroType XDR assay specifically for FLQ resistance. Our study adds new evidence, demonstrating FluoroType XDR’s ability to detect a wide range of mutations in gyrA and gyrB. However, its tendency to produce indeterminate results and requirement for more advanced infrastructure may limit its deployment in decentralised or low-resource laboratories.

While both assays demonstrated strong performance, some practical differences emerged. FluoroType XDR, despite exhibiting superior sensitivity, yielded five indeterminate results. In contrast, the Xpert MTB/XDR assay produced no indeterminate outcomes. This distinction is critical in real-world clinical settings, particularly in resource-limited environments where inconclusive results may delay timely treatment initiation. However, in high-capacity laboratories with the infrastructure to repeat or confirm testing, the higher sensitivity of FluoroType XDR could offer a clinical advantage, especially when there is a strong suspicion of FLQ resistance. Notably, indeterminate results were excluded from the calculation of performance indices because they do not yield definitive positive or negative outcomes and therefore cannot be classified into true positives, true negatives, false positives, or false negatives. Nevertheless, the exclusion of indeterminate outcomes may slightly overestimate the performance and does not diminish their clinical significance.

Agreement between the two assays was excellent, with a Cohen’s Kappa of 0.95, indicating near-perfect concordance. Still, two discordant results underscore the limitations of molecular diagnostics. In one case, Xpert MTB/XDR identified resistance that was not confirmed by WGS, likely a false positive. In the other case, FluoroType XDR missed a resistant isolate carrying the gyrA Asp94Gly mutation, which WGS detected. These findings emphasise the value of WGS as a reference standard, especially for resolving ambiguous results and confirming resistance profiles in high-stakes cases.

In this study, five indeterminate outcomes were observed with the FluoroType XDR assay, whereas no indeterminate results were reported with Xpert MTB/XDR. Indeterminate molecular results may arise from low DNA concentration, suboptimal amplification, mixed bacterial populations, or mutations outside the assay’s defined probe regions. While these results were excluded from performance calculations, their occurrence is clinically relevant, as indeterminate findings can delay treatment initiation or necessitate repeat testing, particularly in resource-limited settings. In centralised laboratories with adequate infrastructure, indeterminate results may be resolved through repeat testing or confirmatory methods such as LPAs or WGS. However, in decentralised settings, a higher rate of indeterminate outcomes may reduce the operational utility of an assay despite high analytical sensitivity. These findings highlight the need to balance diagnostic sensitivity with interpretability of results when selecting molecular tools for programmatic MDR-TB management.

In addition to diagnostic performance, operational factors such as cost, infrastructure requirements, and turnaround time are important considerations for programmatic implementation. Xpert MTB/XDR offers a rapid time to result of approximately 90 min with minimal hands-on time and is well suited for decentralised, high-burden settings. In contrast, FluoroType XDR requires more complex laboratory infrastructure, automated extraction, and longer processing times, which may limit its use to central or reference laboratories. The GenoType MTBDRsl assay, while well established, involves multiple manual steps and typically requires a full working day to generate results. Although a formal cost-effectiveness analysis was beyond the scope of this study, these operational differences are critical when selecting diagnostic tools for MDR-TB management in resource-limited settings.

Clinically, the ability to rapidly and accurately detect FLQ resistance is vital for initiating effective second-line treatment regimens without delay. Xpert MTB/XDR, with its rapid turnaround and lack of indeterminate results, is well-suited for frontline use. FluoroType XDR, with broader mutation detection and higher sensitivity, may serve as a confirmatory test in reference laboratories, particularly when resources and expertise permit. With that said, our study had limitations. The sample size was relatively small, and the focus was specifically on FLQ resistance, as this is the common drug tested for resistance in both assays that forms part of the regimen for MDR-TB treatment.¹⁵ Further studies using larger, more diverse samples, such as sputum, are needed to fully map where each assay fits best in the TB diagnostic landscape.

Conclusion

This study showed that both the Xpert MTB/XDR and FluoroType XDR assays are highly effective in detecting FLQ resistance in MDR-TB clinical isolates, each with distinct strengths. Xpert MTB/XDR was notable for its rapid turnaround and zero indeterminate results, making it ideal for frontline diagnostic settings. FluoroType XDR, with its slightly higher sensitivity and broader mutation coverage, may offer added value as a confirmatory test in well-equipped laboratories.

Given the growing importance of FLQ resistance in guiding MDR-TB treatment, especially with the increasing adoption of BPaL-based regimens, early and accurate detection is critical. This study underscores that integrating rapid molecular diagnostics with WGS for discordant or unclear cases can significantly enhance patient care and inform national TB control strategies. The near-perfect agreement between the two assays, validated by WGS, confirms their diagnostic reliability. Importantly, this work highlights the essential role of WGS in clarifying discrepant results, ensuring the most accurate treatment decisions.

As drug-resistant TB remains a pressing global health threat, expanding access to these molecular diagnostic tools and leveraging their complementary strengths can improve treatment outcomes and help curb the spread of resistant strains.

Acknowledgements

This article is based on a conference paper originally presented at PATHRED, held in Johannesburg, on 02–05 October 2025. The conference paper, titled ‘Comparative evaluation of the Xpert MTB/XDR and FluoroType XDR assays for detecting fluoroquinolone resistance in clinical multidrug-resistant tuberculosis isolates’, was subsequently expanded and revised for this journal publication. This republication is done with permission from the conference organisers.

Competing interests

The authors declare that they have no financial or personal relationships that may have inappropriately influenced them in writing this article.

CRediT authorship contribution

Kabelo G. Kaapu: Conceptualisation, Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Validation, Writing – original draft. Mmule K. Mathebe: Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Writing – original draft. Ndivhuho A. Makhado: Formal analysis, Investigation, Methodology, Validation, Writing – review & editing. Molebogeng R. Lekalakala-Mokaba: Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Writing – review & editing. Ivy Rukasha: Conceptualisation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Validation, Writing – review & editing. All authors reviewed the article, contributed to the discussion of results, approved the final version for submission and publication, and take responsibility for the integrity of its findings.

Funding information

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Data availability

Data sharing is not applicable to this article as no new data were created or analysed in this study.

Disclaimer

The views and opinions expressed in this article are those of the authors and are the product of professional research. It does not necessarily reflect the official policy or position of any affiliated institution, funder, agency, or that of the publisher. The authors are responsible for this article’s results, findings, and content.

References