Diagnosis and management of extrapulmonary tuberculosis in low-resource settings, a study from Zanzibar, Tanzania

Abstract

Tuberculosis (TB) is still a major global public health concern. Of the 7 million incident TB cases recognized by the World Health Organization in 2018, 15% were extrapulmonary tuberculosis (EPTB) cases. Diagnosing EPTB remains a challenge despite continuing efforts and progress in the development of new TB diagnostic tools. Signs and symptoms of EPTB are often non-specific and similar in a range of other conditions. Laboratory confirmation of extrapulmonary disease is constrained by the difficulty of obtaining appropriate biological specimens, conventional diagnostic methods with low sensitivity and lack of an accurate, rapid, low-cost diagnostic test for EPTB. In addition, laboratory facilities are costly and often limited in low-resource settings. This may lead to diagnostic delay in initiating antituberculosis (anti-TB) treatment, on the other hand, starting empirical treatment, without laboratory confirmation, emphasises the need of close monitoring of treatment response.

The overall aim of this thesis was to improve the diagnosis and management of EPTB cases in a low-resource setting. Keeping the primary objective in mind, the secondary aims were to implement and assess the performance of a new diagnostic test based on immunochemical detection of the Mycobacterium tuberculosis complex specific antigen MTP64 (MPT64 test), for diagnosing EPTB in the routine diagnostics at Mnazi Mmoja Hospital (MMH) in Zanzibar. Further, to evaluate the health care seeking pathways and the diagnostic delays experienced by presumptive EPTB patients, identify factors associated with longer diagnostic delay, and to assess the impact of anti-TB treatment on self-rated health among EPTB cases. Next, to describe the clinical presentation of EPTB, and follow the study participants during anti-TB treatment to assess clinical parameters, independent of laboratory investigations, which could aid in the monitoring of treatment response among EPTB cases.

Presumptive EPTB patients of all ages were prospectively enrolled at MMH for thirteen months from august 2014. At inclusion, data were collected in a face-to face interview using the semi-structured study questionnaire, the results from a full physical examination were recorded and the adult patients initiating anti-TB treatment answered the EQ-5D-3L to evaluate the study participant`s self-rated health before treatment. Further, a biological specimen was collected from the site of assumptive EPTB infection and subjected to the MPT64 test, GeneXpert® MTB/RIF assay and routine laboratory diagnostics. The included patients initiating anti-TB treatment were assessed after the intensive phase of treatment and at treatment completion. The adult study participants again reported their self-rated health using the EQ-5D-3L after completing anti-TB treatment. Utilizing a predefined composite reference standard, the patients were classified as TB (confirmed TB, probable TB or possible TB) or non-TB cases to assess the diagnostic tests performances and other variables.

We included 132 patients (median age 27 years, interquartile range 8-41 years), who were defined in accordance with the composite reference standard as TB cases (n=64 in paper I and III; n=69 in paper II) and non-TB cases (n=62 in paper I and III, n=63 in paper II). Six patients were classified as uncategorized cases in paper I and III. A higher proportion of positive test results was found for the MPT64 test in TB cases (45/69, 65%) as compared to ZN staining (8/69, 12%), culture (8/60, 13%) and the GeneXpert® MTB/RIF assay (6/38, 16%). The MPT64 test showed an overall sensitivity and specificity of 69% and 95%, respectively, with the best test performance demonstrated among TB lymphadenitis cases and in paediatric TB. Many EPTB cases experienced a delay exceeding two months from symptom onset until treatment was initiated, with health system delay as the main contributor to overall delay. The majority of adult TB cases described reduced work capacity with a median of 60 days due to the ongoing illness and using the EQ-5D instrument, a significantly improved self-perceived health status was noted after as compared to before anti-TB treatment. We further evaluated three clinical parameters, weight gain, clinical improvement and regression of objective findings during anti-TB treatment and found that most of the TB cases fulfilled ≥2 parameters after the intensive phase of anti-TB treatment.

The MPT64 test is implementable in the routine diagnostic laboratory in this low-resource setting and has the potential to improve the diagnosis of EPTB, especially for lymph node TB and paediatric TB in this and similar settings. With many EPTB patients experiencing long delays in the initiation of treatment together with the reported reduced work capacity among the adult TB patients and improvement of self-reported health status after treatment, reducing the diagnostic delay and timely initiation of appropriate treatment can have crucial impact on the economic loss and morbidity of the affected patients. We propose that a combination of only clinical parameters can be incorporated in a simple assessment tool to aid health care workers in low-resource settings to monitor treatment response among EPTB patients. The findings from this study can be used to improve EPTB patient management in the current setting, but larger and more studies in various routine diagnostic settings are needed to expand the knowledge base regarding the MPT64 test and further evaluate and validate various simple clinical parameters to be incorporated as the suggested easy treatment response assessment tool.