Challenges in diagnosis and treatment of extrapulmonary tuberculosis

Abstract

The infectious disease tuberculosis (TB) is a global health problem. Diagnosis of extrapulmonary TB is often challenging due to non-specific symptoms and findings, and the lack of sensitive diagnostic tests. This leads to both under- and overdiagnosis and extensive use of empirical TB treatment, which increases the risk of placing individuals without TB or individuals with multi-drug resistant TB on prolonged, incorrect treatment. New tools for diagnosis and assessment of treatment response of extrapulmonary TB are needed to improve TB care. The overall aim of this thesis was to investigate new methods for diagnosis and treatment monitoring of extrapulmonary TB. Specific aims included to assess the performance of the MPT64 antigen detection test (MPT64 test) and the Xpert Ultra assay for the diagnosis of extrapulmonary TB in a low TB prevalent setting. Further, to reproduce an anti-MPT64 antibody for large-scale use of the MPT64 test. Finally, to study the potential of IP-10 measured in dried plasma or blood spots as a biomarker for treatment response in patients with extrapulmonary TB. Extrapulmonary samples received for TB diagnostics at the Microbiology and/or Pathology laboratories in a clinical setting in Norway were collected and subjected to the MPT64 test and/or Xpert Ultra. The performance of the new tests was compared to results of routine TB diagnostics. New anti-MPT64 antibodies were developed in mice and rabbits by use of different antigen-adjuvant combinations and screening in immunohistochemistry. The best performing individual antibodies were pooled, and the performance of the final pooled anti-MPT64 antibody was evaluated in patients samples. IP-10 dynamics during TB treatment were studied in extrapulmonary TB patients at a tertiary care hospital on Zanzibar, Tanzania. Plasma, dried plasma spots and dried blood spots were collected at baseline, 2 months of treatment and end of treatment. IP-10 levels were measured by ELISA and compared to clinical improvement. In paper I, we found that the sensitivity of the MPT64 test for diagnosing extrapulmonary TB was lower compared to results obtained in previous studies conducted in TB endemic low-resource settings. The best test performance in our setting was demonstrated in formalin-fixed biopsies, where the MPT64 test showed an excellent specificity and a sensitivity in the same range as PCR-based tests, but lower than culture. The results of paper II indicate that it is possible to reproduce a functional polyclonal anti-MPT64 antibody for large-scale use of the MPT64 test, but that careful selection of the antigen-adjuvant combination for immunisation and comprehensive screening strategies are nessecary to obtain high-performance antibodies. In paper III, we found that the overall sensitivity and specificty of Xpert Ultra for diagnosing extrapulmonary TB in low-TB prevalent high-income setting were high and close to the performance of culture. In paper IV, we report that a significant decline in IP-10 levels in plasma, dried plasma spots and dried blood spots during treatment was observed in extrapulmonary TB patients, and the decline was significant already after two months in HIV-negative patients. The correlation between IP-10 measured in plasma and dried plasma spots or dried blood spots was high. To conclude, potential new tools for improved diagnosis and treatment monitoring of extrapulmonary TB have been identified. Our results indicate that Xpert Ultra can contribute to a more rapid diagnosis of extrapulmonary TB in our setting, and that IP-10 measured in dried blood spots is a promising marker for response to treatment in extrapulmonary TB patients. However, further studies with larger sample sizes that include relevant negative controls and longer follow-up of patients to evaluate the clinical impact of these new tools are needed. The utility of the MPT64 test is likely limited in settings with well-functioning culture facilities, but the test can contribute to strengthen the TB diagnosis in the event that samples have not been subjected to culture.