Field evaluation of a point-of-care triage test for active tuberculosis (TriageTB)

Abstract

Background: To improve tuberculosis (TB) diagnosis, the World Health Organisation (WHO) has called for a non-sputum based triage test to focus TB testing on people with a high likelihood of having active pulmonary tuberculosis (TB). Various host or pathogen biomarker-based testing devices are in design stage and require validity assessment. Host biomarkers have shown promise to accurately rule out active TB, but further research is required to determine generalisability. The TriageTB diagnostic test study aims to assess the accuracy of diagnostic test candidates, as well as field-test, finalise the design and biomarker signature, and validate a point-of-care multi-biomarker test (MBT).

Methods: This observational diagnostic study will evaluate sensitivity and specificity of biomarker-based diagnostic candidates including the MBT and Xpert® TB Fingerstick cartridge compared with a gold-standard composite TB outcome classification defined by symptoms, sputum GeneXpert® Ultra, smear and culture, radiological features, response to TB therapy and presence of an alternative diagnosis. The study will be conducted in research sites in South Africa, Uganda, The Gambia and Vietnam which all have high TB prevalence. The two-phase design allows for finalisation of the MBT in Phase 1 in which candidate host proteins will be evaluated on stored serum from Asia, South Africa and South America and on fingerstick blood from 50 newly recruited participants per site. The MBT test will then be locked down and validated in Phase 2 on 250 participants per site.

Discussion: By targeting confirmatory TB testing to those with a positive triage test, 75% of negative GXPU may be avoided, thereby reducing diagnostic costs and patient losses during the care cascade. This study builds on previous biomarker research and aims to identify a point-of-care test meeting or exceeding the minimum World Health Organisation target product profile of a 90% sensitivity and 70% specificity. Streamlining TB testing by identifying individuals with a high likelihood of TB should improve TB resources use and, in so doing, improve TB care.

Trial registration: NCT04232618 (clinicaltrials.gov) Date of registration: 16 January 2020.

Keywords: Biomarker; Biosignature; Non-sputum based screening; TB; Tuberculosis.

Fig. 1

Schematic of study design

Fig. 2

UCP-LF formats. Figure legend: Various UCP-LF multi-biomarker test (MBT) formats with distinctive biomarker signatures have been built and evaluated in previous studies [21]. While the structure was initially limited by restrictions of the portable UCP-strip readers, the introduction of a portable reader capable of 2D scan (ESE Quant LR3 version) for use in the TriageTB study allowed the application of the parallel MBT format. The use of parallel strips, each specific for a single biomarker, minimizes potential cross reactivity and each strip has standard dimensions, important for production. Available standard cassettes can hold 3 parallel strips. In this study, for CRP, IP-10 and SAA1

Fig. 3

Evaluation of fingerstick blood application. Figure legend: During Phase 1B of Triage the basic 3-marker MBT strip with CRP, IP-10 and SAA1 will be evaluated and compared to serum. The application of lysed and whole FSB will be compared. For the whole fingerstick blood testing strips were provided in a cassette and run horizontally immediately after sampling; for the lysed fingerstick blood and serum testing strips were provided with microtiter plates to run vertically in the laboratory. The ESE Quant LR3 version will be distributed to all sites for reading

Fig. 4

Composite Outcome Classification Algorithm. Figure Legend: The gold standard composite diagnostic algorithm comprises symptoms, CXR findings, GeneXpert® Ultra, TB microscopy, TB culture results and treatment response