Comparison of multiple methods of diagnosis of mycobacterial infection from bone marrow samples of HIV positive patients

• Main Author: Chosmata, Benford Ivan
• Format: Others
• Language: en
• Published: 2011
• Subjects: diagnosis; mycobacterium infections; HIV
• Online Access: http://hdl.handle.net/10539/9053

MMed, Haematology, Faculty of Health Sciences, University of the Witwatersrand === Background: Mycobacterium tuberculosis (MTB) infection remains a serious public health challenge in sub-Saharan Africa. Rapid and early diagnosis is critical in the successful control of this eminently treatable infection. This study compared the diagnostic usefulness of culture, bone marrow trephine biopsy granulomata, bone marrow trephine biopsy Ziehl-Neelsen (ZN) stain and bone marrow mycobacterial polymerase chain reaction (PCR) in establishing the diagnosis of mycobacterial infection in HIV infected patients. Materials and methods: The trephine biopsies of HIV positive patients done for the investigation of suspected tuberculosis were reviewed for granulomata and stained with ZN stain. The corresponding bone marrow aspirates were subjected to DNA real-time PCR analyses using LightCyler TB Kit® (Roche Diagnostic). Culture results were used as diagnostic gold standard. Results: Of the 60 patients studied, 24 were culture negative. Of the 34 culture positive, 62% were Mycobacterium tuberculosis and 38% were Mycobacterium avium intracellulare. Using the culture method as a gold standard, the sensitivities and specificities were 97% and 23% for bone marrow trephine biopsy granulomata, 65% and 58% for bone marrow trephine biopsy ZN staining and 50% and 73% for bone marrow aspirate PCR analysis respectively. Ninety-seven percent of all trephine biopsies with positive ZN stain had granulomata. Conclusion: The presence of granulomata in bone marrow trephine biopsies of HIV infected patients appear to have a high diagnostic yield whilst mycobacterial PCR has the lowest yield but highest specificity. These results should be confirmed in a prospective case controlled study because the sample size in this study was small, and the study was a retrospective one.