Thinking beyond diagnostic accuracy to evaluate tuberculosis screening tests

Tuberculosis is a leading cause of morbidity and mortality among people living with HIV.1WHOGlobal tuberculosis report 2023.https://www.who.int/publications/i/item/9789240083851Date: 2023Date accessed: February 12, 2024Google Scholar Diagnosing tuberculosis in people living with HIV is often challenging, but even tests with moderate accuracy can be helpful when deployed in an appropriate diagnostic algorithm. For example, a urine lateral flow lipoarabinomannan assay can moderately decrease all-cause mortality risk among severely ill hospitalised people living with HIV,2Gupta-Wright A Corbett EL van Oosterhout JJ et al.Rapid urine-based screening for tuberculosis in HIV-positive patients admitted to hospital in Africa (STAMP): a pragmatic, multicentre, parallel-group, double-blind, randomised controlled trial.Lancet. 2018; 392: 292-301Summary Full Text Full Text PDF PubMed Scopus (143) Google Scholar and its sensitivity is, at most, 50–60%.3WHOLateral flow urine lipoarabinomannan assay (LF-LAM) for the diagnosis of active tuberculosis in people living with HIV, 2019 update.https://www.who.int/publications/i/item/9789241550604Date: 2019Date accessed: February 12, 2024Google Scholar Tests that can readily detect tuberculosis disease and rapidly lead to further action, including treatment initiation, will greatly aid those at risk of tuberculosis, particularly people living with HIV initiating antiretroviral therapy, whose risk of incident tuberculosis is highly elevated during the first few months of the therapy.4Lawn SD Harries AD Meintjes G Getahun H Havlir DV Wood R Reducing deaths from tuberculosis in antiretroviral treatment programmes in sub-Saharan Africa.AIDS. 2012; 26: 2121-2133Crossref PubMed Scopus (0) Google Scholar Such individuals would benefit greatly from better tuberculosis screening tests. In diagnostic algorithms incorporating an initial tuberculosis screening test, non-sputum-based rapid tests with good sensitivity are preferred, even at the expense of specificity; this is because screening tests are intended to detect as many potentially disease-positive individuals as possible, including those without clinical symptoms. Subsequently, people who screen positive would undergo a confirmatory diagnostic test with known higher specificity. Using the combination of a high-sensitivity screening test and then a higher-specificity confirmatory test quickly rules out people without tuberculosis, and then genuine cases are confirmed. As screening tests alone do not produce definite diagnoses, it makes sense to evaluate their clinical utility within the broader health-care system context in which they operate. In The Lancet Global Health, Tiffeney Mann and colleagues5Mann T Gupta RK Reeve BWP et al.Blood RNA biomarkers for tuberculosis screening in people living with HIV before antiretroviral therapy initiation: a diagnostic accuracy study.Lancet Glob Health. 2024; (published online April 4.)https://doi.org/10.1016/S2214-109X(24)00029-9Google Scholar describe the prospective evaluation of seven blood-based RNA biosignatures for tuberculosis screening among a large cohort of people living with HIV in South Africa. Sensitivity and specificity for each signature were evaluated using a reference standard of two liquid cultures. For Suliman4—the RNA signature with the highest value for area under the receiver-operating characteristic curve—clinical utility was also assessed, along with that of C-reactive protein (CRP) and WHO four-symptom screen (W4SS), both of which are WHO-endorsed tuberculosis screening tests in people living with HIV.6WHOWHO consolidated guidelines on tuberculosis: module 2: screening: systematic screening for tuberculosis disease.https://www.who.int/publications/i/item/9789240022676Date: 2021Date accessed: February 12, 2024Google Scholar No RNA signatures met the diagnostic accuracy criteria specified in WHO target product profiles,7WHOA rapid biomarker-based non-sputum-based test for detecting TB. In: High priority target product profiles for new tuberculosis diagnostics: report of a consensus meeting.https://www.who.int/publications/i/item/WHO-HTM-TB-2014.18Date: 2014Date accessed: February 12, 2024Google Scholar although signatures performed better than W4SS. All signatures performed less well in asymptomatic individuals, hampering their potential as true up-front screening tests, and no subgroup analyses considered extrapulmonary tuberculosis. This is a missed opportunity, as blood-based assays present a minimally invasive and much needed option for this population. Uniquely, the authors employed decision curve analysis to assess the clinical utility of various hypothetical screening-test-guided diagnostic algorithms. In general, decision curve analysis estimates the net benefit of using a diagnostic test in a specified manner, compared with the default approaches of simply intervening on either every patient or no patients.8Vickers AJ Elkin EB Decision curve analysis: a novel method for evaluating prediction models.Med Decis Making. 2006; 26: 565-574Crossref PubMed Scopus (2982) Google Scholar In the diagnostic algorithm described in the study, individuals with a positive tuberculosis screening test would be referred for confirmatory tuberculosis diagnostic testing via liquid culture. Decision curve analysis compared the use of Suliman4, W4SS, or CRP as screening tests to guide culture testing, versus the strategies of providing culture to all people living with HIV and to none.5Mann T Gupta RK Reeve BWP et al.Blood RNA biomarkers for tuberculosis screening in people living with HIV before antiretroviral therapy initiation: a diagnostic accuracy study.Lancet Glob Health. 2024; (published online April 4.)https://doi.org/10.1016/S2214-109X(24)00029-9Google Scholar The Suliman4 RNA signature had the highest clinical net benefit over the largest range of probabilities of tuberculosis that trigger conducting confirmatory testing (termed threshold probabilities in figure 3 of Mann and colleagues5Mann T Gupta RK Reeve BWP et al.Blood RNA biomarkers for tuberculosis screening in people living with HIV before antiretroviral therapy initiation: a diagnostic accuracy study.Lancet Glob Health. 2024; (published online April 4.)https://doi.org/10.1016/S2214-109X(24)00029-9Google Scholar). Compared with a strategy of culture for all people living with HIV, using Suliman4 to guide confirmatory testing would be considered beneficial if the number of culture tests needed to find one true tuberculosis case did not exceed 24. In highly tuberculosis-endemic settings, this number needed to test is reasonable, indicating the possible clinical value of Suliman4 as a screening test. Ultimately, Mann and colleagues favour implementing CRP-guided tuberculosis screening for people living with HIV over any examined RNA signature, because fast and inexpensive CRP assays already exist—and such amenable operational characteristics are critical for screening tests. As illustrated in the study by Mann and colleagues,5Mann T Gupta RK Reeve BWP et al.Blood RNA biomarkers for tuberculosis screening in people living with HIV before antiretroviral therapy initiation: a diagnostic accuracy study.Lancet Glob Health. 2024; (published online April 4.)https://doi.org/10.1016/S2214-109X(24)00029-9Google Scholar an optimal high-sensitivity screening tool for tuberculosis disease among people living with HIV remains elusive. Decision curve analysis is an intriguing method for assessing tuberculosis screening tests because it inherently recognises that a screening test does not operate in isolation; rather, the potential benefit of a screening test lies in how appropriately it guides the subsequent step in the diagnostic algorithm—ie, confirmatory testing. For novel diagnostic assays, the essential first step is determining diagnostic accuracy, but other features such as clinical utility, cost-effectiveness, impact on clinical and patient outcomes, and generalisability to subpopulations should then be evaluated.9Ochodo EA Naidoo S Schumacher S et al.Improving the design of studies evaluating the impact of diagnostic tests for tuberculosis on health outcomes: a qualitative study of perspectives of diverse stakeholders.Wellcome Open Res. 2019; 4: 183Crossref PubMed Scopus (3) Google Scholar, 10Schumacher SG Sohn H Qin ZZ et al.Impact of molecular diagnostics for tuberculosis on patient-important outcomes: a systematic review of study methodologies.PLoS One. 2016; 11e0151073Crossref Scopus (35) Google Scholar Analysing how new tuberculosis screening tests will fit within existing health-care systems before they are rolled out will aid the design of more effective diagnostic algorithms and lead to identifying more people with tuberculosis. We declare no competing interests. Blood RNA biomarkers for tuberculosis screening in people living with HIV before antiretroviral therapy initiation: a diagnostic accuracy studyRNA biomarkers showed better clinical utility to guide confirmatory tuberculosis testing for people living with HIV before ART initiation than symptom-based screening, but their performance did not exceed that of CRP and fell short of WHO recommended targets. Interferon-independent approaches might be required to improve accuracy of host-response biomarkers to support tuberculosis screening before ART initiation. Full-Text PDF Open Access