Abstract
Abstract Background Tuberculosis diagnosis requires rapid and sensitive diagnostic methods. The aim was to evaluate the performance of Anyplex MTB/NTM and Xpert MTB/RIF assays for the detection of Mycobacterium tuberculosis complex (MTC) in respiratory and non-respiratory samples from patients suspected of having tuberculosis in a third-level hospital in Mexico. Methods Samples (n=436) from January 2020 to December 2023 in a hospital from Monterrey, Mexico underwent paired microbiological (MGIT or Lowenstein-Jensen culture) and molecular assays (Anyplex™ MTB/NTM Real-time Detection or Xpert MTB/RIF™) for mycobacterial detection. Results Of the 279 (63.8%) respiratory samples (bronchoalveolar lavage 71.3%, sputum 20.7% and tracheal aspirate 7.8%), positive culture and detected PCR were found in 26/201 samples by Anyplex and in 20/78 samples by Xpert whereas negative culture and detected PCR were found in 32 and 7 samples, respectively. Anyplex showed 65% sensitivity, 80.1% specificity, 26.5% positive predictive value (PPV), and 90.2% negative predictive value (NPV) whereas Xpert showed 100% sensitivity, 87.9% specificity, 74% PPV and 100% NPV. Of the 157 (36.1%) non-respiratory samples (cerebrospinal fluid 37.3%, soft tissues 21.5%, pleural 22.1%, lymph nodes 11.3%, and synovial 7.5%), positive culture and detected PCR were found in 14/118 samples by Anyplex and in 4/39 by Xpert whereas negative culture and detected PCR were found in 17 and 4 samples, respectively. Anyplex showed 73.6% sensitivity, 82.8% specificity, 45.1% PPV, 94.2% NPV whereas Xpert showed 66.6% sensitivity, 87.8% specificity, 50% PPV and 93.5% NPV. Conclusion Xpert assays showed higher sensitivity than Anyplex for the detection of MTC in respiratory samples and showed higher specificity compared to Anyplex in both respiratory and non- respiratory samples. Xpert showed higher PPV for respiratory samples, indicating its superiority disregarding true positive cases. Disclosures All Authors: No reported disclosures
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call