Comparison of the Genedia MTB Detection Kit and the Cobas TaqMan MTB Assay for Detection of Mycobacterium tuberculosis in Respiratory Specimens

Abstract

The performance of the Genedia MTB detection kit was compared with that of the Cobas TaqMan MTB test using respiratory specimens. The Genedia and Cobas assays showed comparable sensitivities (81.8% and 78.8%, respectively) and specificities (99.8% and 99.5%, respectively), while the Genedia assay produced fewer invalid results and required less turnaround time and labor.

TEXT

Tuberculosis is a worldwide public health concern. According to the 2013 World Health Organization Global Tuberculosis Report, the country of South Korea has an intermediate level of tuberculosis burden, with an incidence rate of 108 per 100,000 inhabitants (1). Rapid diagnoses and drug treatments that prevent person-to-person transmission remain the top priorities in tuberculosis control (2, 3). Therefore, direct detection of Mycobacterium tuberculosis complex DNA using nucleic acid amplification tests has become an important aspect of rapid diagnosis (4).

The Genedia MTB detection kit (Genedia assay; Green Cross Medical Science Corp., Chungbuk, Republic of Korea) was recently developed for rapid molecular detection of the M. tuberculosis complex. The Genedia assay is a real-time PCR method targeting IS6110 with TaqMan hydrolysis probes and has been approved by the Korean Ministry of Food and Drug Safety. We compared the performance of this kit to that of the Cobas TaqMan MTB test (Cobas assay; Roche Diagnostics, Basel, Switzerland). The Cobas assay targets the 16S rRNA region and is one of the most widely utilized real-time PCR assays for M. tuberculosis complex detection outside the United States.

This study was conducted at a tertiary-care hospital in Seoul, Republic of Korea, and was approved by the Institutional Review Board of the Samsung Medical Center. In total, 629 consecutive respiratory specimens were prospectively collected from patients with suspected pulmonary tuberculosis between November 2013 and August 2014. All specimens were examined by direct microscopy, mycobacterial cultures, and Cobas and Genedia assays, simultaneously.

The respiratory specimens were processed with NALC-NaOH (2% N-acetyl-l-cysteine–sodium hydroxide), followed by centrifugation at 3,000 × g for 20 min. An acid-fast bacillus (AFB) smear procedure was performed with an auramine-rhodamine fluorescent stain, followed by confirmation with Ziehl-Neelsen staining. Staining results were graded according to the American Thoracic Society/Centers for Disease Control and Prevention guidelines (5). Specimens were defined as smear positive when the AFB smear grades were between 1 and 4. All specimens were cultured on both solid and liquid media for 6 weeks. Decontaminated samples were inoculated into a mycobacterial growth indicator tube (MGIT 960 system; Becton Dickinson, Sparks, MD) and on 3% Ogawa agar (Shinyang, Seoul, Republic of Korea). Positive cultures were confirmed both by the presence of cord formation and by MPT64 antigen testing (SD Bioline TB Ag MPT64 Rapid test; Standard Diagnostics Inc., Yongin-si, South Korea), a rapid immunochromatographic assay. When either of these tests yielded a negative result, an rpoB-specific PCR test using an MTB-ID V3 kit (YD Diagnostics, Yongin-si, South Korea) was performed on isolated colonies or liquid cultures to differentiate between M. tuberculosis and nontuberculous mycobacteria.

The Cobas assay was conducted according to the manufacturer's instructions, as described previously (6). For the Genedia assay, 100-μl aliquots of the decontaminated samples were resuspended in 100 μl of DNA extraction buffer. The mixtures were boiled at 100°C for 10 min. After centrifugation, the supernatants were analyzed. Real-time PCRs were performed on a 7500 Fast real-time PCR system (Applied Biosystems, Foster City, CA, USA). The PCR was performed in a total volume of 20 μl (10 μl of PCR mixture, 2 μl of primer mixture, 2 μl of distilled water, 1 μl of internal control, and 5 μl of template DNA). Thermocycling conditions included a step at 95°C for 15 min, followed by 40 cycles of 15 s at 95°C and 30 s at 60°C. For both PCR assays, internal-control DNA was used to detect any PCR inhibitors that might have been present in specimens. When the internal control was not amplified, the test was considered to be invalid; the specimen was then diluted 1:10 and retested.

Of 629 respiratory specimens, 474 were sputum samples, 123 were bronchial wash fluid samples, and 32 were bronchoalveolar lavage fluid samples. After excluding nine samples with contaminated culture results, 5.3% (33/620) were discovered to be positive for M. tuberculosis by mycobacterial culture. Of these, 26 were positive according to both the Cobas and Genedia assays, while one was positive by only the Genedia assay. The six remaining specimens were negative for M. tuberculosis according to both PCR assays. Of the 587 (94.7%) specimens that were culture negative for M. tuberculosis, 65 nontuberculous mycobacterium isolates were identified whereas the remaining 522 specimens showed no growth. Among these culture-negative cases, one showed positive results on both PCR tests, while two were positive only according to the Cobas assay.

Based on the results of concurrent cultures, the overall sensitivity and specificity were 81.8% (95% confidence interval [CI], 63.9% to 92.4%) and 99.8% (95% CI, 98.9% to 99.9%) for the Genedia assay and 78.8% (95% CI, 60.6% to 90.4%), and 99.5% (95% CI, 98.4% to 99.9%) for the Cobas assay, respectively (Table 1). The sensitivities in smear-positive specimens were >95%, while the sensitivity in smear-negative specimens was 40% for both assays.

TABLE 1.

Performances of the Genedia MTB detection kit and Cobas TaqMan MTB test, stratified by AFB smear status^a

Smear result	Genedia MTB detection kit								Cobas TaqMan MTB test
	Sensitivity		Specificity		PPV		NPV		Sensitivity		Specificity		PPV		NPV
	No./total no.	% (95% CI)	No./total no.	% (95% CI)	No./total no.	% (95% CI)	No./total no.	% (95% CI)	No./total no.	% (95% CI)	No./total no.	% (95% CI)	No./total no.	% (95% CI)	No./total no.	% (95% CI)
All (n = 620)	27/33	81.8 (63.9–92.4)	586/587	99.8 (98.9–99.9)	27/28	96.4 (79.8–99.8)	586/592	99.0 (97.7–99.6)	26/33	78.8 (60.6–90.4)	584/587	99.5 (98.4–99.9)	26/29	89.7 (71.5–97.3)	584/591	98.8 (97.5–99.5)
Positive (n = 32)	23/23	100 (82.2–100)	8/9	88.9 (50.7–99.4)	23/24	95.8 (76.9–99.8)	8/8	100 (59.8–100)	22/23	95.7 (76.0–99.8)	8/9	88.9 (50.7–99.4)	22/23	95.7 (76.0–99.8)	8/9	88.9 (50.7–99.4)
Negative (n = 588)	4/10	40.0 (13.7–72.6)	578/578	100 (99.2–100)	4/4	100 (39.6–100)	578/584	99.0 (97.7–99.6)	4/10	40.0 (13.7–72.6)	576/578	99.7 (98.6–99.9)	4/6	66.7 (24.1–94.0)	576/582	99.0 (97.7–99.6)

^aAbbreviations: CI, confidence interval; NPV, negative predictive value; PPV, positive predictive value.

Of a total of 629 specimens, including samples with contaminated culture results, 29 and 30 samples were characterized as M. tuberculosis positive by the Genedia assay and the Cobas assay, respectively (Table 2). The percentage of positive agreement between the two assays was 93.3% (95% CI, 76.5 to 98.8), and the percentage of negative agreement was 99.8% (95% CI, 98.9 to 99.9). The kappa value was 0.95 (95% CI, 0.89 to 1.00).

TABLE 2.

Comparison of the Cobas TaqMan MTB test and Genedia MTB detection kit

Cobas TaqMan MTB test result	No. of samples with indicated Genedia MTB detection kit result		Totala
	Positive	Negative
Positive	28	2	30
Negative	1	598	599
Total	29	600	629

^aNine samples (one sample with a positive result and eight samples with negative results) with contaminated culture results were included in the analysis. Positive percent agreement, 93.3% (95% CI, 76.5 to 98.8); negative percent agreement, 99.8% (95% CI, 98.9 to 99.9); Kappa value, 0.95 (95% CI, 0.89 to 1.00).

No invalid results were detected by the Genedia assay, while eight specimens showed invalid results after the Cobas assay due to failed amplification of internal controls. In analytical specificity tests using purified DNA samples that were extracted from pure cultures of 38 reference strains (25 mycobacterial and 13 other strains), no cross-reactivity was found by the Genedia assay (see Table S1 in the supplemental material). The detection limit of the Genedia assay was measured using 1:10 serial dilutions of DNA from a cultured colony of M. tuberculosis ATCC 27294 (H37Rv strain). The measured limit for M. tuberculosis was 3.0 × 10⁻⁵ ng/μl (equal to 6.3 copies/μl). The total processing time (3 h) and hands-on time (<1 h) required to perform the Genedia assay were both shorter than the times required for the Cobas assay (5 h and 1 to 1.5 h, respectively) (Table 3).

TABLE 3.

Hands-on time and total processing time for the Cobas TaqMan MTB test and Genedia MTB detection kit

Parameter	Cobas TaqMan MTB test value	Genedia MTB detection kit value
Hands-on time	80 min	50 min
Specimen and control prepn for DNA extraction	60 min	30 min
Reagent preparation and loading for real-time PCR	20 min	20 min
Total processing time	∼5 h	∼3 h
Specimen and control prepn	60 min	30 min
DNA extraction	65 min	30 min
Reagent preparation and loading for real-time PCR	20 min	20 min
Real-time PCR	150 min	90 min

According to previous studies, the sensitivity and specificity ranges of the Cobas assay are 71.4% to 97.2% and 95.8% to 100%, respectively (6–13). The sensitivity (78.8%) and specificity (99.5%) of the Cobas assay in this study were also quite similar to those seen in our previous studies (6, 7). The performance of the Genedia assay was comparable to that of the Cobas assay, with a kappa value of 0.95. Six specimens with false-negative results produced by both the Cobas and Genedia assays were also smear negative, while one result that was false negative in the Cobas assay was smear positive with a 1+ grade. One sample that was false positive by both PCR assays was smear positive with a 2+ grade, and this could be categorized as a true-positive sample since the specimen was collected from a patient with pulmonary tuberculosis who was receiving tuberculosis medication. There were two false-positive specimens misidentified by the Cobas assay that were smear negative. The low (40.0%) sensitivity with smear-negative specimens is a limitation of these assays (10).

Note that there were no invalid results produced by the Genedia assay. In addition, this assay provides the advantages of both workload reduction and flexibility of choice in real-time PCR platforms. However, the present study had some limitations. First, we could not evaluate the IS6110 copy numbers in the M. tuberculosis isolates. Second, only respiratory specimens were evaluated; thus, further studies will be required to evaluate the performance of these assays using other types of specimens.

In conclusion, the Genedia assay is a rapid identification tool with a high degree of specificity for the detection of the M. tuberculosis complex in respiratory specimens. The assay shows performance comparable to that of the Cobas assay but generates fewer invalid results and requires less turnaround time and labor.