Blood transcriptomic biomarkers for tuberculosis screening: time to redefine our target populations?

Abstract

The seminal observation that blood transcriptomic changes can predate diagnosis of active tuberculosis set the field on a course to evaluate their potential impact as screening tools for early detection of disease.1Zak DE Penn-Nicholson A Scriba TJ et al.A blood RNA signature for tuberculosis disease risk: a prospective cohort study.Lancet. 2016; 387: 2312-2322Summary Full Text Full Text PDF PubMed Scopus (404) Google Scholar Early detection would allow prompt treatment initiation, which could permit shorter courses of therapy and reduce morbidity, mortality, and onward transmission of infection. In the study published by The Lancet Global Health, Simon C Mendelsohn and colleagues2Mendelsohn SC Fiore-Gartland A Penn-Nicholson A et al.Validation of a host blood transcriptomic biomarker for pulmonary tuberculosis in people living with HIV: a prospective diagnostic and prognostic accuracy study.Lancet Glob Health. 2021; (published online April 13.)https://doi.org/10.1016/S2214-109X(21)00045-0Summary Full Text Full Text PDF PubMed Scopus (7) Google Scholar make an important contribution to the field by assessing the diagnostic accuracy and prognostic ability of an 11-gene transcriptomic signature (RISK11) among people living with HIV in South Africa. Mendelsohn and colleagues2Mendelsohn SC Fiore-Gartland A Penn-Nicholson A et al.Validation of a host blood transcriptomic biomarker for pulmonary tuberculosis in people living with HIV: a prospective diagnostic and prognostic accuracy study.Lancet Glob Health. 2021; (published online April 13.)https://doi.org/10.1016/S2214-109X(21)00045-0Summary Full Text Full Text PDF PubMed Scopus (7) Google Scholar use an observational cohort study design that overcomes the inherent risk of spectrum bias present in case-control studies, in which the inclusion of more extreme phenotypes can lead to overly optimistic assessments of diagnostic accuracy.3Usher-Smith JA Sharp SJ Griffin SJ The spectrum effect in tests for risk prediction, screening, and diagnosis.BMJ. 2016; 353i3139Crossref PubMed Scopus (101) Google Scholar In addition, they use real-time PCR technology with predefined test thresholds amenable to development for clinical diagnostic tests. The study is well conducted, using robust outcome definitions and comprehensive statistical analyses. Nonetheless, such studies can be challenging. Despite the focus on people living with HIV, who represent an important target group for early tuberculosis detection and prevention efforts, and the inclusion of 820 individuals with valid RISK11 results, there were only eight prevalent and eight incident primary outcome tuberculosis events. The authors found that RISK11's performance approached (albeit did not meet) WHO's target product profile benchmarks for screening and prognostic tests,4WHOHigh-priority target product profiles for new tuberculosis diagnostics. Report of a consensus meeting.https://www.who.int/tb/publications/tpp_report/en/Date: 2014Date accessed: May 21, 2019Google Scholar but the low event rate led to wide CIs in the performance metrics. So where does the field go next? Several key questions should be answered before blood transcriptomic biomarkers can be considered for programmatic implementation. First, a wide range of emerging transcriptomic, proteomic, metabolomic, and microbiological biomarkers have been proposed to facilitate early tuberculosis diagnosis.5MacLean E Broger T Yerlikaya S Fernandez-Carballo BL Pai M Denkinger CM A systematic review of biomarkers to detect active tuberculosis.Nat Microbiol. 2019; 4: 748-758Crossref PubMed Scopus (67) Google Scholar The study by Mendelsohn and colleagues2Mendelsohn SC Fiore-Gartland A Penn-Nicholson A et al.Validation of a host blood transcriptomic biomarker for pulmonary tuberculosis in people living with HIV: a prospective diagnostic and prognostic accuracy study.Lancet Glob Health. 2021; (published online April 13.)https://doi.org/10.1016/S2214-109X(21)00045-0Summary Full Text Full Text PDF PubMed Scopus (7) Google Scholar suggests that RISK11 might perform better than alternative triage strategies (ie, WHO's symptom screen or the Alere Determine TB lipoarabinomannan [LAM] assay) and an interferon-γ release assay as a prognostic test (QuantiFERON TB Gold-Plus). However, four other candidate blood transcriptomic signatures have consistently performed well in systematic head-to-head assessments for prevalent and incident tuberculosis,6Gupta RK Turner CT Venturini C et al.Concise whole blood transcriptional signatures for incipient tuberculosis: a systematic review and patient-level pooled meta-analysis.Lancet Respir Med. 2020; 8: 395-406Summary Full Text Full Text PDF PubMed Scopus (48) Google Scholar, 7Turner CT Gupta RK Tsaliki E et al.Blood transcriptional biomarkers for active pulmonary tuberculosis in a high-burden setting: a prospective, observational, diagnostic accuracy study.Lancet Respir Med. 2020; 8: 407-419Summary Full Text Full Text PDF PubMed Scopus (37) Google Scholar and a point-of-care assay of C-reactive protein8Yoon C Semitala FC Atuhumuza E et al.Point-of-care C-reactive protein-based tuberculosis screening for people living with HIV: a diagnostic accuracy study.Lancet Infect Dis. 2017; 17: 1285-1292Summary Full Text Full Text PDF PubMed Scopus (55) Google Scholar and a urine Fujifilm SILVAMP TB LAM assay9Broger T Sossen B du Toit E et al.Novel lipoarabinomannan point-of-care tuberculosis test for people with HIV: a diagnostic accuracy study.Lancet Infect Dis. 2019; 19: 852-861Summary Full Text Full Text PDF PubMed Scopus (75) Google Scholar have shown strong potential for tuberculosis screening among people with HIV. Head-to-head assessments of these promising biomarkers will be an important extension of the study by Mendelsohn and colleagues and could identify the best performing candidates. Such comparisons should also include health economic considerations. For example, the current cost of a blood transcriptomic test is likely to far exceed the minimum WHO target of less than US$2 for a triage test;4WHOHigh-priority target product profiles for new tuberculosis diagnostics. Report of a consensus meeting.https://www.who.int/tb/publications/tpp_report/en/Date: 2014Date accessed: May 21, 2019Google Scholar cheaper assays (eg, point-of-care C-reactive protein) might have a competitive advantage in this regard. Future approaches could also consider integrating biomarker results into multivariable prediction tools to improve discrimination between disease states.10Gupta RK Calderwood CJ Yavlinsky A et al.Discovery and validation of a personalized risk predictor for incident tuberculosis in low transmission settings.Nat Med. 2020; 26: 1941-1949Crossref PubMed Scopus (13) Google Scholar Second, how individuals with a positive blood transcriptomic biomarker test should be treated remains to be determined. Thus far, clinical trials have dichotomised the management of tuberculosis into preventive treatment for latent infection (with fewer agents for shorter periods of time) and treatment of tuberculosis disease (with more agents for longer periods of time). Subclinical or incipient disease that might be detected by blood transcriptomic biomarkers falls into the gap between these two ends of the spectrum. Early data from randomised controlled trials have long been presented to show that paucibacillary (smear-negative) disease might be adequately treated with truncated antimicrobial regimens,11Fox W Whither short-course chemotherapy?.Br J Dis Chest. 1981; 75: 331-357Summary Full Text PDF PubMed Scopus (135) Google Scholar, 12Sputum-smear-negative pulmonary tuberculosis controlled trial of 3-month and 2-month regimens of chemotherapy.Lancet. 1979; 313: 1361-1363Summary Scopus (34) Google Scholar but 3 months of weekly rifapentine and isoniazid did not reduce the incidence of tuberculosis among HIV-uninfected, RISK11-positive people in the general population in the CORTIS trial.13Scriba TJ Fiore-Gartland A Penn-Nicholson A et al.Biomarker-guided tuberculosis preventive therapy (CORTIS): a randomised controlled trial.Lancet Infect Dis. 2021; 21: 354-365Summary Full Text Full Text PDF PubMed Scopus (10) Google Scholar Finally, it remains unclear whether the diagnostic accuracy or prognostic ability achieved by RISK11 among ambulatory people living with HIV is sufficient to translate into clinical utility, a reduction in tuberculosis incidence, or cost-effectiveness. Notably, Mendelsohn and colleagues found that the positive predictive values of RISK11 were only 2·5% for primary prevalent tuberculosis and 3·2% for primary incident tuberculosis. These values reflect a combination of low event rates and imperfect test specificity, underlined by the fact that 35% of people were RISK11-positive at baseline. Previous data have also shown that test sensitivity wanes markedly after a 3–6-month time horizon, suggesting that blood transcriptomic biomarkers reflect only short-term risk of disease.6Gupta RK Turner CT Venturini C et al.Concise whole blood transcriptional signatures for incipient tuberculosis: a systematic review and patient-level pooled meta-analysis.Lancet Respir Med. 2020; 8: 395-406Summary Full Text Full Text PDF PubMed Scopus (48) Google Scholar, 13Scriba TJ Fiore-Gartland A Penn-Nicholson A et al.Biomarker-guided tuberculosis preventive therapy (CORTIS): a randomised controlled trial.Lancet Infect Dis. 2021; 21: 354-365Summary Full Text Full Text PDF PubMed Scopus (10) Google Scholar Taken together, this evidence suggests that the optimal implementation of these biomarkers will require careful selection of target populations who are at the highest risk of disease in the short term to maximise positive predictive values and reduce the number needed to treat with tuberculosis therapy. Such target groups could include recent case contacts and, potentially, people with HIV during intensified screening for subclinical disease before antiretroviral therapy initiation.14Lawn SD Kranzer K Edwards DJ McNally M Bekker L-G Wood R Tuberculosis during the first year of antiretroviral therapy in a South African cohort using an intensive pretreatment screening strategy.AIDS. 2010; 24: 1323-1328Crossref PubMed Scopus (94) Google Scholar MN has a patent pending for a method for detecting active tuberculosis (WO2017/050483). RKG declares no competing interests. Validation of a host blood transcriptomic biomarker for pulmonary tuberculosis in people living with HIV: a prospective diagnostic and prognostic accuracy studyRISK11 identified prevalent tuberculosis and predicted risk of progression to incident tuberculosis within 15 months in ambulant people living with HIV. RISK11's performance approached, but did not meet, WHO's target product profile benchmarks for screening and prognostic tests for tuberculosis. Full-Text PDF Open Access