Triage test to diagnose presumptive pulmonary tuberculosis

Tuberculosis remains a global public health challenge, with a substantial number of undiagnosed cases contributing to ongoing transmission and increasing disease burden. A cost-effective triage test for individuals displaying tuberculosis symptoms or at risk of disease could offer an affordable solution. In this regard, non-sputum-based tests that have a minimum of 90% sensitivity and 70% specificity have been prioritised by WHO to rule out tuberculosis and identify the individuals that require further testing. Such tests might reduce treatment delays, overall costs, and the infrastructure burden associated with existing tuberculosis diagnostics. Studies have shown that host blood transcriptomic signatures can accurately discriminate active tuberculosis from other illnesses.1Sweeney TE Braviak L Tato CM Khatri P Genome-wide expression for diagnosis of pulmonary tuberculosis: a multicohort analysis.Lancet Respir Med. 2016; 4: 213-224Summary Full Text Full Text PDF PubMed Google Scholar In The Lancet Global Health, Ankur Gupta-Wright and colleagues conducted the largest prospective multicountry study done to date, comprising 1499 adults, to evaluate the diagnostic accuracy of the late-prototype Xpert MTB Host Response (Xpert HR) assay (Cepheid, Sunnyvale, CA, USA) as a triage test for pulmonary tuberculosis.2Gupta-Wright A Ha H Abdulgadar S et al.Evaluation of the Xpert MTB Host Response assay for the triage of patients with presumed pulmonary tuberculosis: a prospective diagnostic accuracy study in Viet Nam, India, the Philippines, Uganda, and South Africa.Lancet Glob Health. 2024; 12: e226-e234Google Scholar Xpert HR is the first cartridge-based automated molecular test designed to quantify host gene expression.3Södersten E Ongarello S Mantsoki A et al.Diagnostic accuracy study of a novel blood-based assay for identification of tuberculosis in people living with HIV.J Clin Microbiol. 2021; 59: e01643-e01720Crossref PubMed Scopus (28) Google Scholar The early set of host transcriptomic markers reported by Sweeny and colleagues1Sweeney TE Braviak L Tato CM Khatri P Genome-wide expression for diagnosis of pulmonary tuberculosis: a multicohort analysis.Lancet Respir Med. 2016; 4: 213-224Summary Full Text Full Text PDF PubMed Google Scholar through multicohort analysis in ten countries was later used to develop an Xpert HR prototype.2Gupta-Wright A Ha H Abdulgadar S et al.Evaluation of the Xpert MTB Host Response assay for the triage of patients with presumed pulmonary tuberculosis: a prospective diagnostic accuracy study in Viet Nam, India, the Philippines, Uganda, and South Africa.Lancet Glob Health. 2024; 12: e226-e234Google Scholar This test quantifies mRNA levels to derive a TB Score, which distinguishes active tuberculosis from other conditions. Gupta-Wright and colleagues reported 90·3% (95% CI 86·5–93·3) sensitivity and 62·6% specificity (59·7–65·3) for Xpert HR, and an area under the receiver operating characteristic curve (AUC) of 0·89 (95% CI 0·86–0·91) against the microbiological reference standard. Although these findings fell below the WHO minimum specificity targets for a non-sputum-based triage test, the assay had a high negative predictive value (95·8%; 95% CI 94·1–97·1) making it a potential rule-out test. The test reported similar sensitives across five countries (Viet Nam, India, Philippines, Uganda, and South Africa), by sex, and within subgroups. The results further indicate that assuming a tuberculosis prevalence of 10%, using Xpert HR as a triage test could have decreased the need for confirmatory tuberculosis testing by 57·3% (95% CI 54·2–60·4). Prospective evaluations of the Xpert HR prototype by other researchers showed mixed results; Sutherland and colleagues reported promising diagnostic accuracy (AUC 0·88, sensitivity 80%, specificity 94%),4Sutherland JS van der Spuy G Gindeh A et al.Diagnostic accuracy of the Cepheid 3-gene host response fingerstick blood test in a prospective, multi-site study: interim results.Clin Infect Dis. 2022; 74: 2136-2141Crossref PubMed Google Scholar Li and colleagues observed similar results (AUC 0·91, sensitivity 90%, specificity 70%),5Li M Qiu Y Guo M et al.Evaluation of the Cepheid 3-gene host response blood test for tuberculosis diagnosis and treatment response monitoring in a primary-level clinic in rural China.J Clin Microbiol. 2023; 61: e0091123Crossref Scopus (0) Google Scholar and Moreira and colleagues reported 90% sensitivity and modest specificity (53%).6Moreira FMF Verma R Pereira Dos Santos PC et al.Blood-based host biomarker diagnostics in active case finding for pulmonary tuberculosis: a diagnostic case-control study.EClinicalMedicine. 2021; 33: 100776Summary Full Text Full Text PDF PubMed Scopus (0) Google Scholar Despite proof of principle for host markers as a triage test, the smaller cohort sizes than in Gupta-Wright and colleagues' study and the focus on particular populations and groups might have introduced bias and warrant further evaluation. In this regard, the prospective evaluation of Xpert HR by Gupta-Wright and colleagues has many strengths, which include the large cohort size, geographically diverse sites (including important subpopulations), and representative samples (including patients with comorbidities such as HIV and diabetes). Although Gupta-Wright and colleagues present a robust prospective analysis, the study has notable limitations. The positive predictive value observed for the Xpert HR test was modest (40·1%, 95% CI 36·8–44·1), resulting in a high likelihood of false positives. The study reported approximately 20% lower specificity in individuals living with HIV compared with those without HIV, raising concerns about using this test in populations with a higher prevalence of other infections that can trigger a similar host response. We recommend that future studies compare host immune responses in various infections that can trigger similar responses. Such analyses could also explain the lower sensitivity of Xpert HR tests observed by Gupta-Wright and colleagues in the Philippines compared with other countries, for which they suggest a different cutoff to achieve similar performance. The study could also have benefitted from including patients with extrapulmonary tuberculosis and control groups to compare gene expression between people with presumptive tuberculosis and healthy controls. Perumal and colleagues validated a quantitative PCR assay targeting different host mRNA signature panels using samples from pulmonary tuberculosis, extrapulmonary tuberculosis, and healthy controls. They observed that two of their panels met the minimum criteria for a triage test (AUC 0·94 and 0·96, sensitivity 90·2% and 90·9%, and specificity 88·8% and 89·9%).7Perumal P Abdullatif MB Garlant HN et al.Validation of differentially expressed immune biomarkers in latent and active tuberculosis by real-time PCR.Front Immunol. 2021; 11: 612564Crossref Scopus (12) Google Scholar Studies have shown that triage tests can substantially improve cost-effective community-based detection of tuberculosis.8Brümmer LE Thompson RR Malhotra A et al.Cost effectiveness of low-complexity screening tests in community-based case-finding for tuberculosis..Clin Infect Dis. 2023; (published online Aug 25. https://doi.org/10.1093/cid/ciad501)Crossref Google Scholar Although the Xpert HR assay cost will probably exceed WHO's recommended US$2 per test, the study by Gupta-Wright and colleagues emphasises the need for further investigations, including cost-effectiveness modelling and assessing usability and acceptability. Additionally, the study would benefit from a comparative analysis with the currently available triage tests that are under evaluation. One such example is host protein markers that are suitable for point-of-care settings and could meet cost targets. Garlant and colleagues9Garlant HN Ellappan K Hewitt M et al.Evaluation of host protein biomarkers by ELISA from whole lysed peripheral blood for development of diagnostic tests for active tuberculosis.Front Immunol. 2022; 13: 854327Crossref Scopus (3) Google Scholar and Koeppel and colleagues10Koeppel L Denkinger CM Wyss R et al.Diagnostic performance of host protein signatures as a triage test for active pulmonary TB.J Clin Microbiol. 2023; 61: e0026423Crossref Scopus (0) Google Scholar validated protein biomarkers that met the minimum WHO criteria for a triage test in adults. The goal of introducing triage tests is to increase active screening capacity and reduce the number of patients requiring expensive confirmatory testing. The evaluation of the Xpert HR assay in a large cohort representing different populations holds considerable merits. The availability of this assay on a globally available GeneXpert platform further reduces the cost of initial infrastructure building. Future research should focus on refining the assay's performance by including other potential markers, assessing its cost-effectiveness, and determining its acceptability in diverse clinical settings. KEK has a patent (WO2015170108A1) on biomarkers and combinations thereof for diagnosing tuberculosis. All other authors declare no competing interests. Evaluation of the Xpert MTB Host Response assay for the triage of patients with presumed pulmonary tuberculosis: a prospective diagnostic accuracy study in Viet Nam, India, the Philippines, Uganda, and South AfricaXpert HR did not meet WHO minimum specificity targets for a non-sputum-based triage test for pulmonary tuberculosis. Despite promise as a rule-out test that could reduce confirmatory sputum testing, further cost-effectiveness modelling and data on acceptability and usability are needed to inform policy recommendations. Full-Text PDF Open Access