Abdominal and thoracic ultrasound for the diagnosis of pulmonary and extra-pulmonary tuberculosis
Children as well as HIV-positive adults often suffer from extrapulmonary tuberculosis (TB). Due to the low sensitivity of routine microbiological TB tests, it is often difficult to obtain confirmation of mycobacterial disease. Point-of-care ultrasound (POC-US) is becoming increasingly established as a promising method for the diagnosis and treatment monitoring of a wide variety of infectious diseases. Previous experience with POC-US indicates that the method may help to support clinical decision making in the treatment of pulmonary and extrapulmonary TB. Using a prospective, multicentre cohort study, we therefore aim to evaluate the diagnostic accuracy of abdominal and thoracic ultrasound in adults with presumed or currently treated TB.
In 2018, 10 million cases of TB and 1.5 million deaths were recorded worldwide. Especially in children and HIV-positive people, who frequently suffer from extrapulmonary TB, it is often difficult to obtain confirmation of mycobacterial disease due to the low sensitivity of routine microbiological tests. For treatment outcome monitoring in pulmonary TB, sputum smear microscopy and/or mycobacterial culture conversion—from mycobacteria-positive to -negative—are poor indicators, and essentially no treatment monitoring tools have been validated for extrapulmonary TB. Therefore, to support clinical decision making, close the diagnostic gap of undiagnosed individuals, and avoid overtreatment, further—preferably sputum-free—tests and diagnostic algorithms are needed.
Point-of-care ultrasound (POC-US) is a diagnostic tool that can be used outside of radiology departments and by the physicians, who are caring for the patients, at the point of care. In sub-Saharan countries with high TB prevalence or incidence in HIV-positive adults, the Focused-Assessment with Sonography for HIV-associated tuberculosis (FASH) POC-US protocol has shown promise. The protocol evaluates extra-pulmonary manifestations of TB and was developed specifically for HIV-associated TB. However, studies on the diagnostic accuracy of FASH are limited. Therefore, further validation is needed to determine the diagnostic accuracy of FASH for extra-pulmonary TB in people living with HIV using a comprehensive reference standard.
Another potential application of POC-US is lung ultrasound (LUS) for differential diagnosis of TB of the lung and pleura, respectively, for each of which specific sonographic features have been described. However, no data are yet available on the specificity and sensitivity of LUS with a specific TB protocol.
Sonographic features of tuberculosis detected by FASH or LUS may be useful for assessing therapeutic response and monitoring treatment. Follow-up POC-US (by FASH and/or LUS) can be used to detect an inadequate response to treatment and to prompt further investigation, for example, of mycobacterial resistance. With sufficient evidence of the effectiveness of POC-US for TB diagnosis, it may also be useful to consider artificial intelligence (AI) for interpretation of findings to compensate for the lack of medical personnel with sonographic expertise in facilities with limited resources. Several companies are already working to enable AI interpretation of POC-US findings.
In the proposed multicentre study, we aim to evaluate the feasibility and accuracy of TB-focused POC-US modules for the diagnosis of TB compared to a comprehensive reference standard, which is independent of HIV status, in three clinical centres in Germany and India. In addition, we will evaluate the use of POC-US for treatment monitoring and the feasibility and accuracy of AI in interpreting POC-US modules. The study in Germany will primarily serve to answer exploratory questions, while the study in India will provide an evaluation of TB-focused POC-US in a representative and unbiased patient population.
Abnormal findings on lung ultrasound in a patient with culture-positive pulmonary tuberculosis.
