Компаративна студија на дијагностичките постапки за молекуларна брза директна детекција на mycobacterium tuberculosis complex и резистенција кон рифампицин со класичните микробиолошки методи кај белодробна туберкулоза

Abstract

Background. Tuberculosis (TB) remains the global public health concern, being one of the top 10 causes of death and the leading cause from a single infectious agent. The emergence and spread of resistant forms of TB (multidrug-resistant and extensively-drug resistant, MDR-TB and XDR-TB respectively) multiply the problem. The challenge of early and accurate diagnosis of TB took advantage several years ago, by World Health Organization – endorsement of molecular assays for simultaneous detection of Mycobacterium tuberculosis complex (MTBC) and resistance to anti-tuberculous drugs. Aim. The aim of the study was to assess the performance of GeneXpert MTB/RIF assay (Cepheid, Sunnyvale, CA, USA; Xpert), which possesses a capacity for detection both MTBC and rifampicin (RIF)-resistance, as a diagnostic tool for pulmonary TB. This is a first study designed to evaluate the performance of a molecular assay for TB in the Republic of Macedonia. Material and methods. The study took place in the Institute of lung disease and tuberculosis in Skopje, Republic of Macedonia, and covered a two-year period (2016-2017). We examined 185 respiratory samples by Xpert and conventional methods, including fluorescent smear microscopy (auramine-O staining) for acid-fast-bacilli (AFB), solid and liquid culture medium (Löwenstein-Jensen and BACTEC MGIT 960, respectively), and phenotypic drug susceptibility test (pDST, proportion method on Löwenstein-Jensen medium). Additional molecular diagnostics was conducted in samples with confirmed rifampicin-resistance (MTBDRplus, MTBDRsl, Hain Lifescience GmbH, Nehren, Germany). Statistical evaluation of Xpert, smear and culture tests and their comparisons were performed using XLSTAT software add-on for Excel 2016. Results. Out of 185 respiratory samples, 167 (90,28%) were Xpert-positive for MTBC and 91 (49,19%) were AFB positive. Among culture-positive samples, Xpert showed a sensitivity of 97,67% and 86,36% in smear-positive and in smear-negative, respectively. Xpert identified an additional 57 smear-negative culture-positive samples, increasing the TB detection rate by 36,18% in culture-positive and by 41,08% in the whole cochort compared to smear microscopy (Xpert added value, Δ Xpert vs smear). Compared to culture, Xpert showed a sensitivity of 92,76% (95% CI; 87,42-96,33%) and specificity of 22,58% (95% CI; 9,59-41,10%). In comparison with any conventional microbiological test, we found a gradual increase of sensitivity for Xpert, being the highest vs smear-microscopy (93, 82%; 95% CI; 89,12-96,60%) with the negative predictive value of 0,389; specificity was 100%, with a positive predictive value of 1. 152/185 samples (82,16%) yielded valuable culture for further pDST. In 151/152 (99,34%) samples the RIF-susceptibility result of Xpert was concordant with pDST, including one RIF-resistant sample. Only one sample out of 152 (0,66%) showed discordant result – being RIF-susceptible by Xpert and RIF-resistant by pDST. Further molecular investigation (MTBDRplus, MTBDRsl) of both samples confirmed the existence of two MDR-TB strains of MTBC, allowing to determine the location of the mutation in one of them (D516V in rifampicin-resistance-determining-region, RRDR of rpoB gene). The quantitative Xpert results (Ct-values) were subject for advanced statistical analyses. Ct-values significantly correlate with smear positivity status (Linear regression test, p<000.1), culture time-to-positivity (ANCOVA test, p<0,0001), as well as both solid culture positivity status and overall conventional microbiology diagnostic procedures for TB (Logistic regression test, Wald; p<0,0001). No correlation was observed regarding the liquid culture. There was a significant correlation of Xpert with erythrocyte sedimentation rate as well as the appearance of caverns on chest X-ray, and no difference in diabetic patients. Similarly, there was no significant difference for Xpert in new and previously treated patients, neither qualitative nor quantitative for Xpert Ct- values. Discussion and conclusions. Xpert showed high sensitivity for MTBC-detection in various subgroups of samples and overall, being similar and even better to other investigator’s results obtained in a low-incidence setting. The low specificity of Xpert vs culture is due to its comparison with imperfect “gold standard”, as approximately 20% of cases of pulmonary tuberculosis remain culture negative; in comparison with conventional microbiology diagnostics as a whole, its specificity achieves 100%. Regarding RIF-resistance, a group of two MDR-TB strains is not sufficient for advanced statistical analysis, but indicates the need of prompt resolving of clinical relevance of mutations discovered (i.e. starting an appropriate anti-tuberculosis regimen), and anticipating the possibility of pDST discordance, including presence of silent as well as low-level resistance-conferring mutations (“disputed” mutations). The good correlation of Xpert Ct-values as a measure of bacterial load, with smear and solid culture positivity, along with high sensitivity, support the intention of future replacement of microscopy with Xpert. All things considered, Xpert offers an undoubted advantage in early diagnosis of pulmonary TB, particularly in AFB-negative cases. A dose of cautiousness is needed when considering RIF-resistance results and should be improved by introducing the genome sequencing methods. This would also help in creating the MTBC-mutation database on the territory of our country, which has started with the revealing the first one in this study. Finally, decision-making regarding TB-relapse in previously treated patients should not be based exclusively on Xpert result per se, but rather on an individual, comprehensive, clinical and radiological assessment.