Diagnostic Accuracy of BD MAX MDR-TB Assay Performed on Bronchoscopy Specimens in Patients with Suspected Pulmonary Tuberculosis
Article information
Abstract
Background
Several novel molecular platforms using nucleic acid amplification tests have been developed for the diagnosis of pulmonary tuberculosis (PTB) and rapid detection of isoniazid and rifampin resistance. Among them, the BD MAX MDR-TB assay (BD MAX) has shown high sensitivity and specificity; however, its diagnostic accuracy performed on bronchoscopy specimens has not been reported.
Methods
We retrospectively reviewed the medical records of patients with suspected PTB who underwent bronchoscopy. Patients who underwent BD MAX testing of bronchoscopy specimens were included in the final analysis. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for PTB diagnosis were calculated using a positive culture of Mycobacterium tuberculosis as the reference standard.
Results
Of 114 patients, 34 had culture-confirmed PTB. The sensitivity, specificity, PPV, and NPV of BD MAX performed on bronchoscopy specimens for the diagnosis of PTB were 79.4%, 88.8%, 75.0%, and 91.0%, respectively. The sensitivity of BD MAX was superior to that of acid-fast bacillus smear (79.4% vs. 38.2%, p<0.001).
Conclusion
BD MAX performed on bronchoscopy specimens showed high accuracy for diagnosing PTB. BD MAX can be performed on bronchoscopy specimens in patients with suspected PTB.
Introduction
Tuberculosis (TB) remains a major global health challenge and an important cause of death worldwide. In 2022, 7.5 million people were diagnosed with TB, and 1.13 million people died of TB worldwide [1]. Timely and accurate diagnosis is fundamental for curbing its spread and ensuring effective patient management. In this context, isoniazid (INH)-resistant TB (Hr-TB) and multidrug-resistant/rifampin (RIF)-resistant TB (MDR/RR-TB) pose unique challenges to public health. The detection of Hr-TB and MDR/RR-TB is important, as it not only guides treatment decisions but also helps limit further transmission of drug-resistant TB within the community.
Because acid-fast bacillus (AFB) culture, the gold standard test for diagnosing TB, takes several months, molecular biological techniques using polymerase chain reaction (PCR) technology to detect specific genetic markers associated with TB diagnosis and drug resistance have been developed to rapidly diagnose TB and determine resistance. Among them, the Xpert MTB/RIF assay (Xpert, Cepheid, Sunnyvale, CA, USA), which can diagnose TB and determine RIF resistance, has been widely used because of its high diagnostic accuracy that enables rapid treatment in both RIF-susceptible and RIF-resistant TB [2]; however, it has the disadvantage of not being able to determine INH resistance. Recently, several moderate complexity automated nucleic acid amplification techniques (MCNAATs) including BD MAX MDR-TB assay (BD MAX, Becton, Dickinson and Company, Franklin Lakes, NJ, USA), Abbott real-time MTB-RIF/INH (Abbott, Chicago, IL, USA), FluoroType MTBDR (Bruker-Hain Lifescience, Nehren, Germany), and cobas MTB-RIF/INH (Roche Diagnostics, Basel, Switzerland) have been developed [3]. MC-NAATs can simultaneously diagnose TB and resistance to INH and RIF and have the advantage of being largely automated following the sample preparation step, thus offering rapid and accurate results, especially when high volumes of tests are required. Although the World Health Organization recommends that these MC-NAATs be used on respiratory samples for the detection of pulmonary tuberculosis (PTB) [3], clinical experience with MC-NAATs is considerably more limited than that with Xpert.
Bronchoscopy specimens are useful for diagnosing PTB. These samples not only improve respiratory specimen abundance, particularly in patients with suspected PTB who have a small sputum or cannot produce it, but also offer the advantage of direct access to the lower respiratory tract where Mycobacterium tuberculosis is localized, thereby improving the sensitivity for diagnosing PTB, especially for diagnosing paucibacillary PTB that is difficult to diagnose using sputum specimens [4]. It has been proven that testing bronchoscopy specimens with Xpert, which performs molecular biological tests on unprocessed specimens, provides excellent sensitivity for PTB diagnosis [5,6]. However, unlike Xpert, little research has been conducted on the performance of MC-NAATs using bronchoscopy specimens.
As the demand for efficient and accurate diagnostic tools for TB increases, a comprehensive assessment of the diagnostic accuracy of MC-NAATs using bronchoscopy specimens from patients with suspected PTB is needed. This study aimed to evaluate the diagnostic accuracy of BD MAX, an MC-NAAT, using bronchoscopy specimens so as to enhance our understanding of this diagnostic arsenal and influence clinical practice.
Materials and Methods
1. Study design and participants
We conducted a retrospective cohort study of patients who underwent bronchoscopy for suspected PTB at a secondary referral hospital. All patients who underwent bronchoscopy between March 2020 and February 2024 were screened for eligibility. Patients who underwent bronchoscopy for reasons other than suspected PTB were excluded. Patients who started taking TB medication before bronchoscopy and those who did not undergo BD MAX testing of bronchoscopy specimens were also excluded from the analysis.
This study was approved by the Institutional Review Board (IRB) of Wonkwang University Sanbon Hospital (IRB No. WMCSB 202404-39-240407). The requirement for informed consent was waived by the IRB because of the retrospective design of the study and the use of medical records only.
2. Data collection
Patient age, sex, height, and body weight were collected from their medical records. History of TB treatment, smoking status, whether bronchioloalveolar lavage (BAL) was performed, whether a sputum test was available, the presence of comorbidities, symptoms, and radiographic findings were also recorded. The final diagnoses were classified into several categories.
The decision to perform a bronchoscopy was based on the clinician’s judgment regarding the degree of suspicion of PTB based on the patient’s symptoms and radiographic findings. All bronchoscopy procedures were performed by pulmonologists using a flexible bronchoscope with a 5.5-mm diameter (EB-1570K, Pentax, Tokyo, Japan). Respiratory samples were obtained from the bronchi of the lung subsegments with abnormal pulmonary lesions suggestive of PTB and were suitable for specimen collection. After the instillation of 30 to 50 mL of normal saline, bronchial washing fluid was aspirated until 20 to 30 mL amount of aspirate was recovered in the collection bottle. The specimens obtained were divided appropriately for AFB staining, AFB culture, TB-PCR, and BD MAX testing.
3. AFB smear and culture, TB-PCR, and BD MAX
All bronchoscopy specimens were stained using the Ziehl-Neelsen method for the detection of AFB-positive bacteria. The specimens were also inoculated in 3% Ogawa agar (Shinyang Chemical, Seoul, Korea) and BBL mycobacteria growth indicator tubes (MGIT, Becton Dickinson), and then cultured in a 37°C incubator and MGIT 960 system (Becton, Dickinson and Company) for 8 and 6 weeks, respectively. All positive cultures were identified at the species level using the AdvanSure Mycobacterium GenoBlot Assay (LG Chemistry, Seoul, Korea).
TB-PCR was performed using the AdvanSure TB/NTM real-time PCR assay (LG Chemistry) according to the manufacturer’s protocol: a pretreatment solution was added to the specimen, centrifugation was performed, and the supernatant was removed. This process was repeated until the sample lost its viscosity. DNA was then extracted from the precipitate, and the IS6110 region of TB and the internal transcribed spacer region of mycobacteria were detected using a quantitative PCR detection system.
BD MAX, an MC-NAAT, is a multiplexed real-time PCR NAAT. To detect M. tuberculosis, it targets the IS6110 and IS1081 regions. To detect RIF and INH resistances, it targets the RIF resistance-determining region (codons 426-452 in M. tuberculosis) of rpoB gene, the inhA promoter region and the 315 codon of the katG gene. It shares the advantage of Xpert in that samples can be entered into the PCR detection system without a preprocessing step such as centrifugation. Its limit of detection reported by the company is 0.5 colony-forming unit (cfu)/mL for M. tuberculosis detection, and 6 cfu/mL for drug resistance detection, which is lower than that of Xpert (112.6 cfu/mL) and Xpert MTB/RIF Ultra assay (Cepheid) (15.6 cfu/mL) [7]. In this study, BD MAX was performed according to the manufacturer's protocol: concentrated specimen samples were added to the sample treatment reagent and shaken manually. After incubation, the samples were transferred to a BD MAX system for PCR amplification and detection of M. tuberculosis and drug resistance.
4. Statistical analysis
We analyzed the performance of BD MAX using two reference standards: when cases with positive AFB cultures performed on bronchoscopy specimens were considered PTB and when both culture-confirmed and clinically diagnosed cases were considered PTB (composite reference standard [CRS]). We calculated the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the AFB smear, TB-PCR, and BD MAX for TB detection using both reference standards. The sensitivity of BD MAX was compared with that of AFB smear and TB-PCR using the McNemar’s test. We also compared the clinical characteristics of BD MAX positive- and BD MAX false-negative patients among the patients with culture-confirmed PTB.
All statistical analyses were performed using the R software version 4.4.0 (R Foundation for Statistical Computing, Vienna, Austria). All statistical tests were two-sided, and differences with p<0.05 were considered statistically significant.
Results
1. Study population
During the study period, 160 patients underwent bronchoscopy in our institution, of whom 39 underwent bronchoscopy for reasons other than suspected PTB. BD MAX was not performed on the bronchoscopy specimens of five patients. Two patients were excluded from the analysis because they had been taking anti-TB drugs before bronchoscopy. Finally, 114 patients who underwent BD MAX using bronchoscopy specimens owing to suspected PTB were included in the analysis (Figure 1).
Inclusion flow chart. PTB: pulmonary tuberculosis; BD MAX: BD MAX MDR-TB assay; TB: tuberculosis; AFB: acid-fast bacillus.
Of the 114 patients, 34 were diagnosed with PTB based on positive AFB culture for TB. There were no significant differences in the baseline characteristics between patients diagnosed with TB and those without TB (Table 1). Patients with culture-confirmed TB more frequently exhibited symptoms of weight loss, and nodules and mediastinal lymph node enlargement were more commonly observed on radiography (Table 2).
Baseline characteristics of the patients
Symptoms and radiographic findings of the patients
Of the 80 patients with negative AFB culture for TB, 10 were clinically diagnosed with PTB based on their symptoms and radiographic findings and treated for PTB (Table 3). Of the nine patients with false-positive BD MAX results, two were diagnosed with clinical TB, and a further three, three, and one patients were finally diagnosed with pneumonia, fibrotic sequelae of previous inflammation, and non-tuberculous mycobacterial lung disease, respectively.
Final diagnosis of patients with negative acidfast bacilli culture for TB
2. Diagnostic accuracy of AFB smear, AdvanSure TB-PCR, and BD MAX MDR-TB assay
Of the 34 patients with culture-positive PTB, BD MAX performed on bronchoscopy specimens showed positive results in 27. Of the 80 patients with negative AFB cultures for TB, nine had positive BD MAX results. The sensitivity, specificity, PPV and NPV of BD MAX were 79.4%, 88.8%, 75.0%, and 91.0%, respectively. The sensitivity of AFB smear and AdvanSure TB-PCR was 38.2% and 70.6%, respectively. The sensitivity of BD MAX was significantly higher than that of AFB smear (p<0.001) and higher than that of AdvanSure TB-PCR although this difference was not statistically different (p=0.248). When CRS was applied, the sensitivity of BD MAX decreased to 65.9%, but was still significantly higher than that of AFB smear (Table 4).
Diagnostic accuracy of AFB smear, AdvanSure TB-PCR, and BD MAX MDR-TB assays performed on bronchoscopy specimens for the diagnosis of pulmonary tuberculosis
Of the 11 patients with a positive BD MAX result and negative AdvanSure TB-PCR result, three had culture-positive PTB, and two had clinical PTB and were treated accordingly. No patients had a negative BD MAX result and positive AdvanSure TB-PCR result.
3. Clinical characteristics of patients with false-negative and false-positive BD MAX results
The seven patients with false-negative BD MAX results were younger and their body weight was heavier than those with accurate BD MAX results. However, there were no differences between the two groups in terms of smoking status, history of TB, or comorbidities. Patients with false-negative BD MAX results had more PTB lesions confined to one lobe, but this difference was not statistically significant (p=0.074) (Table 5). Of the nine patients with false-positive BD MAX results, three had a history of TB treatment.
Comparison of patients with accurate vs. false-negative BD MAX MDR-TB assay results
Discussion
As AFB culture and phenotypic drug sensitivity testing (pDST) is time-consuming, technologies have been developed that not only diagnose TB more rapidly but also determine drug resistance early. Although the gold standard for diagnosing TB is still a positive AFB culture result and additional pDST testing is required, NAATs are clinically useful because they are fast and accurate. MC-NAATs, whose use has been increasing recently, have the advantage of being able to diagnose Hr-TB and MDR/RR-TB at the same time as diagnosing TB, with comparable diagnostic accuracy [8,9], thus enabling early initiation of appropriate treatment. To our knowledge, this is the first study to investigate the performance of MC-NAAT using bronchoscopy specimens.
Among the MC-NAATs, the effectiveness of BD MAX has been explored in several studies. BD MAX showed excellent performance in the detection of TB and INH/RIF resistance in a large-scale study [10], and appeared to be more convenient to use than other MC-NAATs [11]. It can be applied not only to pulmonary samples but also to extrapulmonary samples [12,13] and has been shown to shorten the turnaround time to initiate TB treatment with high accuracy when used in clinical settings [14]. In this study, we demonstrated the performance of BD MAX using bronchoscopy specimens.
Several studies have explored the diagnostic accuracy of Xpert, an assay that can diagnose TB and RIF resistance simultaneously using bronchoscopy specimens. According to meta-analyses, the sensitivity and specificity of Xpert are 87%–88% and 92%–94%, respectively [5,6]. Studies using bronchoscopy specimens have consistently found that the sensitivity and specificity of Xpert are superior to those of AFB smears, suggesting that NAATs are useful for diagnosing paucibacillary PTB, which is difficult to diagnose using AFB smear [15-19]. In this study, the sensitivity of BD MAX performed on bronchoscopy specimens was superior to that of AFB smear, which is consistent with the results of previous studies using Xpert.
The TB-PCR method has been used to detect TB bacilli even prior to the introduction of Xpert. However, the conventional TB-PCR method has a disadvantage in that the procedure is complicated and requires skilled personnel, and it may be difficult to diagnose TB patients in the paucibacillary state because TB bacilli may be lost during the sample preprocessing steps, including decontamination and concentration. Xpert, which does not have this shortcoming because it uses respiratory samples directly without preprocessing steps, has been shown to have higher sensitivity than conventional TB-PCR, especially when performed with bronchoscopy specimens considered to be in the paucibacillary state [20-22], and some studies have reported statistically significant differences [23,24]. In this study, three patients had negative AdvanSure TB-PCR results and positive BD MAX results, indicating a higher sensitivity of BD MAX, although the difference was not statistically significant. It can be assumed that BD MAX, which does not require preprocessing steps such as decontamination and concentration, also provides additional diagnostic advantages compared with conventional TB-PCR.
The sensitivity of BD MAX in this study was 79.4%. In a large-scale study that analyzed the accuracy of BD MAX using sputum samples from patients with presumptive PTB [10], the sensitivity of BD MAX was 93%; however, when the subjects were limited to smear-negative samples, the sensitivity was 81%, which was similar to that in this study.
However, the sensitivity of BD MAX in our study was lower than that of Xpert, which was reported in previous studies using bronchoscopy, and there seem to be several reasons for this. First, 61.4% of the total patients, and 73.5% of patients with culture-confirmed PTB in our study, had insufficient sputum to perform a sputum test, which is higher than the rate of 28.4%–71.2% of sputum-scarce patients reported in other studies [15-18]. The proportion of asymptomatic patients among the patients in our study was 30.7% of the total and 32.4% of patients with culture-confirmed PTB, which is higher than the rate of 6.8%–24.0% of asymptomatic patients in other studies [18,19]. This indicates that the severity of PTB in the patients included in our study was lower than that in other studies, which may have resulted in the lower sensitivity of BD MAX in this study. It was implicated that patients with Xpert-negative TB have a lower mycobacterial burden, and thus, less advanced TB disease [25]. In our study, patients with BD MAX-negative PTB tended to have PTB lesions confined to one lobe, which is consistent with a previous study conducted using Xpert [15]. Additionally, only one patient underwent BAL in this study, which is a lower number compared to other studies in which BAL was performed in 20.5%–100% of patients [15-19]. Considering a study that reported a higher sensitivity of BAL than bronchial washing [26], this may have been a factor in the low sensitivity of BD MAX in this study.
The specificity of BD MAX in this study was 88.8%, which was lower than that of AdvanSure TB-PCR. However, among the nine patients with false-positive results, two were diagnosed with PTB based on clinical judgment and treated as PTB, indicating that BD MAX may identify additional cases of PTB not detected on AFB culture. Up to 40% of the patients diagnosed with TB in South Korea have negative AFB culture results [27]. When diagnosing PTB, an optimal decision should be made by comprehensively considering various factors, such as patient symptoms and radiographic findings, as well as their test results, including conventional TB-PCR, Xpert, BD MAX, and AFB culture.
Our study had several limitations. First, because our study had a retrospective design, it is susceptible to bias. Second, we could not compare BD MAX with Xpert, because our hospital does not use Xpert. However, one study has shown that BD MAX has a lower limit of detection for M. tuberculosis than that of Xpert [8]. If a sufficiently large number of patients is studied, the sensitivity of BD MAX may be clinically equivalent to or superior to that of Xpert. Third, we did not present the results of determining INH/RIF resistance using BD MAX because only a few of the 34 culture-confirmed PTB patients included in this study had INH or RIF resistance, making it difficult to judge the utility of the test. Further studies are required to overcome these limitations.
In conclusion, BD MAX performed with bronchoscopy specimens showed accurate diagnostic performance for PTB diagnosis. Performing BD MAX on bronchoscopy specimens is useful for diagnosing PTB.
Notes
Authors’ Contributions
Conceptualization: Ko SJ. Methodology: all authors. Formal analysis: Ko SJ. Data curation: Ko SJ. Investigation: Ko SJ. Writing - original draft preparation: Ko SJ. Writing - review and editing: all authors. Approval of final manuscript: all authors.
Conflicts of Interest
No potential conflict of interest relevant to this article was reported.
Funding
No funding to declare.
Acknowledgements
We would like to thank Sang Hee Lee for performing bronchoscopy and treating a few of the patients included in this study.