Diagnostic accuracy of point-of-care triage tests for pulmonary tuberculosis using host blood protein biomarkers: a systematic review and meta-analysis.

BACKGROUND

Limited access to accurate and accessible tuberculosis (TB) diagnostics remains a critical barrier to timely diagnosis and care in high burden low- and middle-income countries. Point-of-care (POC) TB Triage tests (POC-TTTs) defined as a test performed near to a patient or at the site of patient care without need for specialized expertise or infrastructure, may bridge this gap. We systematically reviewed and meta-analyzed studies on host blood protein biomarkers for POC-TTTs including C-reactive protein (CRP), 3-gene host response (3-gene HR), monocyte-to-lymphocyte ratio (MLR), interferon-γ induced protein 10 (IP-10), hemoglobin, neutrophil-to-lymphocyte ratio (NLR), tumor necrosis factor alpha (TNF-a) and interleukin 6 (IL-6) for their accuracy for screening of pulmonary TB (PTB).

METHODS

A literature search was conducted in PubMed, EMBASE and in Web of Science from 1990 to 31st January 2025. The review included studies that used unstimulated blood of presumptive TB patients who were screened with a POC device to quantify biomarkers for PTB diagnosis. Sputum mycobacterial culture or GeneXpert MTB/Rif or Ultra were used as a reference standard. Risk of bias was assessed using QUADAS-2 tool. Random effect analysis was performed using the Hartung-Knapp-Sidik-Jonkman (HKSJ) method to calculate summary estimates with their 95% confidence intervals (CI). Heterogeneity was tested and quantified using Cochran's Q and Higgin's I test. Egger's linear regression test was used to assess small study effect. The systematic review protocol was registered in PROSPERO with an ID of CRD42023483281.

FINDINGS

We identified 282, 21, 28, 137, 132, 152, 100 and 77 studies from which 10, 6, 4, 2, 0, 0, 0 and 1 study(s) were included for CRP, 3-gene HR, MLR, hemoglobin, IP-10, TNF-a, IL-6 and NLR index tests respectively. The meta-analysis pooled sensitivity (95% CI) was 74% (58-85), 79% (59-90), 64% (15-95), 75% (18-98) and the pooled specificity was 68% (52-80), 85% (68-94), 69% (30-92) and 71% (18-100) for CRP, 3-gene HR, MLR, and hemoglobin respectively. Diagnostic odds ratios ranged from 3.70 (MLR) to 20.93 (3-gene HR) while higgin's I value ranged from 87.4% (MLR) to 99.1% (hemoglobin). Meta-analysis was not performed on NLR.

INTERPRETATION

None of the POC-TTT met the WHO target product profile minimum requirements for a TB triage test of 90% sensitivity and 70% specificity when a POC device was used for screening in a typical setting studied. Further research, specifically focusing on head-to-head comparisons and combination tests are recommended.

FUNDING

Funding was received from the Mr. Willem Bakhuys Roozeboom Foundation. Additional support was given to KK through the EDCTP grant RIA2020I-3305.