Diagnostic accuracy of Xpert MTB/RIF assay and non-molecular methods for the diagnosis of tuberculosis lymphadenitis

Abstract

Background Tuberculous lymphadenitis (TBLN) diagnosis remains a challenge in resource limited countries like Ethiopia. Most diagnostic centers in Ethiopia use smear microscopy, but it has low sensitivity in detecting tubercle bacilli in fine needle aspiration (FNA) specimens. FNA cytology (FNAC) is another widely applicable diagnostic option but it has low specificity for diagnosing TBLN. In 2014, WHO recommended Xpert MTB/RIF assay to be used in detecting TB from FNA specimen by considering the diagnostic limitations of microscopy and cytology. In Ethiopia, there is limited data on Xpert MTB/RIF performance in detecting TBLN from FNA. Therefore, this study aimed to evaluate the diagnostic performance of Xpert MTB/RIF assay and non-molecular methods (cytology, microscopy and culture) for the diagnosis of TBLN. Methods A cross-sectional study was conducted on 152 presumptive TBLN patients at St. Paul’s Hospital Millennium Medical College (SPHMMC) from December 2015 to May 2016 in Addis Ababa, Ethiopia. FNA specimens were collected from each patient. Individual patient specimens were examined by microscopy (acid fast and auramine O staining), cytology, Xpert MTB/RIF and culture. Each specimen was directly inoculated and its sediment following decontamination procedure onto two duplicate Löwenstein-Jensen (LJ) media. Composite culture (specimen positive by direct or concentrated or both culturing methods) and composite method (positive by either one of the non-molecular methods) were taken as reference methods. The data was captured and analyzed using software packages SPSS version 20 (SPSS Inc, Chicago, Illinois, USA). Sensitivity, specificity, positive predictive value, and negative predictive value were calculated. Result A total of 152 presumptive TBLN patients were enrolled in this study. Of these, 105(69%), 68(44.7%), 64(42%), 48(32%) and 33(22%) were positive for M. tuberculosis using composite method (positive by either one of the non-molecular method), composite culture, direct, and concentrated culture, respectively. TB positivity rate was 67.8%, 49.3%, 24.3%, and 14.5% using cytology, Xpert MTB/RIF, Auramine O (FM) microscopy, and Ziehl Nelson (ZN) microscopy, respectively. Using composite culture as reference, the sensitivity and specificity of Xpert MTB/RIF was 78% (95% CI: 73.7% to 82.3%) and 74% (95%CI: 69.4% to 78.6%), respectively. However, the sensitivity of Xpert MTB/RF improved from 78% to 92% using composite method as a reference. The high positivity rate observed in purulent (70%) followed by caseous (66.7%) type of aspirates by Xpert MTB/RIF. Conclusion Xpert MTB/RIF assay has both considerable sensitivity and specificity; it may be employed for better diagnosis, management and treatment of presumptive TBLN patients.