Abstract:

Objective To evaluate the clinical application of domestic real-time fluorescence quantitative PCR (FQ-PCR) reagent and GeneXpert MTB/RIF (GeneXpert) in the detection of Mycobacterium tuberculosis (MTB) in sputum specimens, and compare the detection efficacy of the two methods. Methods From January 2017 to December 2018, a total of 210 sputum specimens of suspected pulmonary tuberculosis cases were collected from Shenzhen Nanshan Center for Chronic Disease Control for smear microscopy, MGIT (Mycobacteria growth indicator tube) 960 culture, GeneXpert and FQ-PCR. Using MGIT 960 result as a reference standard, the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of GeneXpert and FQ-PCR methods were calculated and compared. The consistency of GeneXpert, FQ-PCR and MGIT 960 as well as the consistency between both techniques were compared. According to the grading reporting standard of smear microscopic examination results, the specimens were divided into six groups with progressively increasing bacterial load (negative, scanty, +, ++, +++, ++++), and the Ct mean value of each group detected by FQ-PCR and GeneXpert was compared. Results The sensitivity, specificity, PPV and NPV of GeneXpert were 83.7% (123/147), 87.9% (51/58), 94.6% (123/130), and 68.0% (51/57), respectively, while the sensitivity, specificity, PPV and NPV of FQ-PCR were 83.7% (123/147), 89.8% (53/59), 95.3% (123/129), and 68.8% (53/77), respectively. Kappa test was performed between FQ-PCR, GeneXpert and MGIT 960 results, and the consistency was 85.4% (176/206) between FQ-PCR and MGIT 960 results and 84.9% (174/205) between GeneXpert and MGIT 960 results, with the Kappa value of 0.70 and 0.66, respectively; and the consistency between GeneXpert and FQ-PCR was 92.8% (194/209) (Kappa=0.85). The mean Ct value of negative group and scanty group detected by FQ-PCR was (33.87±5.00) and (27.29±1.30) cycles, respectively, and the difference was statistically significant (t=5.56, P<0.001); while which by GeneXpert was (33.32±6.05) cycles and (23.99±3.36) cycles, respectively, and the difference was also statistically significant (t=5.19, P<0.001). There was no significant difference in mean Ct values between FQ-PCR and GeneXpert when smear negative samples were detected ((33.87±5.00) cycles vs (33.32±6.05) cycles, t=0.32, P=0.750). However, when the smear was positive (scanty, +, ++, +++, ++++), the mean Ct values of FQ-PCR were significantly higher than that of GeneXpert ((27.29±1.30) cycles vs (23.99±3.36) cycles, t=2.60, P=0.030; (27.95±2.85) cycles vs (23.10±4.05) cycles, t=4.71, P<0.001; (25.88±3.62) cycles vs (20.22±2.81) cycles, t=6.08, P<0.001; (24.79±2.46) cycles vs (20.31±4.16) cycles, t=3.85, P<0.001; and (22.38±2.72) cycles vs (16.48±2.78) cycles, t=9.02, P<0.001, respectively). Conclusion The domestic FQ-PCR method has good consistency with the GeneXpert results, which has high clinical application value for large-scale screening of tuberculosis and reducing the detection cost.

Key words: Mycobacterium tuberculosis, Nucleic acid amplification techniques, Polymerase chain reaction, Comparative study