Advances in tuberculosis biomarkers: unravelling risk factors, active disease and treatment success

Abstract

Tuberculosis (TB) is a major global health threat and demands improved diagnostic and treatment monitoring methods. Conventional diagnostics, such as sputum smear microscopy and culture, are limited by slow results and low sensitivity, particularly in certain patient groups. Recent advances in biomarker research offer promising solutions in three key areas: risk of disease, diagnosis of active disease and monitoring of treatment response. For risk assessment, novel genetic signatures and metabolites show potential in predicting the progression from TB infection to active TB. A 16-gene signature, for example, predicts this progression with significant accuracy. In diagnosing active TB, RNA-based transcriptomic signatures provide higher diagnostic accuracy than traditional methods. These signatures, such as a three-gene RNA sequence, effectively differentiate active TB from other diseases and infections, addressing issues of specificity and sensitivity. Monitoring treatment response is crucial, given the varying response rates in treating TB. Emerging biomarkers focus on bacterial burden and host response. They offer more precise and timely assessments of treatment efficacy, enhance personalised treatment approaches and potentially improve patient outcomes. These advancements in biomarkers for TB risk, diagnosis and treatment response represent significant steps towards more effective TB management and control, aligning with global efforts to decrease the burden of TB. Here we aim to highlight several promising biomarkers used to predict risk of disease progression, active TB disease and treatment success.

Extract

Tuberculosis (TB) continues to be one of the deadliest infectious diseases worldwide, with 10.6 million individuals developing the disease and ∼1.3 million deaths in 2022 [1]. The limitations of conventional diagnostic methods (e.g. sputum smear microscopy, PCR and culture), such as their slow turnaround times, lack of sensitivity in certain cohorts and technical challenges, have been well recognised [2, 3]. These limitations emphasise the need for novel biomarkers to advance the global TB research agenda [4].

Shareable abstract

Recent biomarker developments in TB offer immense opportunities for predicting disease risk, diagnosing active cases, and monitoring treatment response. They are pivotal in improving TB management and contribute to worldwide TB elimination efforts. https://bit.ly/3MHMora

Introduction

Tuberculosis (TB) continues to be one of the deadliest infectious diseases worldwide, with 10.6 million individuals developing the disease and ∼1.3 million deaths in 2022 [1]. The limitations of conventional diagnostic methods (e.g. sputum smear microscopy, PCR and culture), such as their slow turnaround times, lack of sensitivity in certain cohorts and technical challenges, have been well recognised [2, 3]. These limitations emphasise the need for novel biomarkers to advance the global TB research agenda [4].

In the field of TB, biomarkers serve a multitude of purposes. They play an important role in the early detection of TB infection (TBI), stratification of individuals based on the risk of disease progression, diagnosis of active disease, monitoring of treatment response, and assessment of disease relapse or cure (figure 1) [5]. However, this viewpoint focuses on three key areas that have a substantial impact on disease control and patient management: risk of disease, active disease and treatment response (figure 1, shown in green/bold text).

FIGURE 1.

Clinical stages of tuberculosis. Green/bold text shows the focus areas of this viewpoint. TB: tuberculosis; TBI: tuberculosis infection.

Biomarkers for risk of disease can identify individuals likely to progress from TBI to active disease and permit early intervention [6]. Biomarkers for active disease can support the diagnostic process and distinguish TB from other diseases with similar presentations. Biomarkers of treatment response help to monitor therapeutic effectiveness, enable personalised treatment regimens, and potentially improve patient outcomes [5, 7, 8]. This viewpoint article reflects the authors’ opinion on the highlights of recent advancements and discusses implications for TB control and patient management.

Risk of disease

Biomarkers that predict the risk of developing active TB from TBI are of great importance for several reasons: 1) identifying individuals likely to progress to active disease facilitates early intervention and thereby improves prognosis and minimises risk of transmission, 2) allocating limited resources to those individuals at highest risk increases cost-effectiveness of interventions [9, 10], and 3) increased understanding of who is likely to progress to active disease can inform public health strategies.

Recent studies have identified genetic signatures, cytokines and metabolites as promising markers of progression to active disease. Notably, a 16-gene signature was identified that can predict progression to active TB within a year with a 66% predictive value [11]. Another study identified a four-gene signature (RAB24, DPP4, STAT1, APOL1) that was able to predict the progression of the disease up to 2 years before onset and has been validated in multiple cohorts across Africa [12]. Yet another recent study identified several cytokines, chemokines and growth factors measured in the supernatant of QuantiFERON samples collected from household contacts of TB patients that proved to be promising markers to predict short-term risk of progression to active TB [13]. Finally, a metabolic biosignature was found to be able to differentiate between individuals that do or do not progress to active disease with 69% sensitivity and 75% specificity in HIV-negative individuals in sub-Saharan Africa [14]. These promising new tools need extensive operational research to understand how they may be implemented to maximise patient benefit. Unfortunately, there is currently no gold standard for diagnosing TBI [15]. This is a significant barrier to understanding how to implement these tests in research and daily clinical practice.

Active disease

Active TB diagnosis relies on various tests. Tuberculin skin test and interferon-γ release assays (IGRA) are often used at the beginning of the diagnostic cascade but cannot differentiate between TBI and active TB [16]. Tests indicating active disease include urinary lipoarabinomannan (LAM) tests, nucleic acid amplification tests, and solid or liquid culture, often supplemented by radiology and clinical presentation. However, active TB is mostly diagnosed by a combination of these as there is no gold standard test. Urinary LAM tests lack sensitivity, especially among non-HIV-infected patients [17]. Nucleic acid amplification tests like GeneXpert, although rapid, might not detect all cases of active TB due to their lower sensitivity in patients with paucibacillary disease [18], and they require stable laboratory conditions. However, new generations of nucleic acid amplification test are under development with the possibility of improved performance in detecting active TB, as evidenced by improvements with the GeneXpert ultra cartridge when compared to GeneXpert MTB/RIF [19]. Liquid culture, the best option at present, is time-consuming and less effective in identifying TB in individuals unable to produce sputum or with paucibacillary TB [20].

RNA-based transcriptomic signatures could help to solve these problems: a three-gene RNA signature of GBP5, DUSP3 and KLF2 was developed in 2016 and can identify active TB and differentiate it from other diseases with similar clinical presentations as well as from TBI with high diagnostic accuracy (area under the curve of active TB versus healthy control: 0.90; active TB versus TBI: 0.88; and active TB versus other diseases: 0.84) [21]. While these results are very promising, the signature remains to be validated in a prospective cohort. Thereby, it addresses the specificity and sensitivity issues that come with currently established methods and enables a timely start of treatment, improves patient outcomes and reduces transmission. However, in addition to diagnosis, the determination of drug susceptibility is crucial for adequate TB treatment, and this would need to be incorporated in novel biomarkers as well.

Treatment response

According to the 2022 World Health Organization report ∼86% of drug-susceptible TB is successfully treated, this percentage drops dramatically to ∼60% in people with drug-resistant TB [1]. Monitoring of treatment response is, therefore, of crucial importance both on an individual level and for TB control efforts. In addition to treatment in programmatic settings, biomarkers of treatment response are of great value in clinical trials where they can inform on the efficacy of novel regimens, and potentially even cure, at an early stage [22]. Currently, a combination of smear microscopy and sputum culture conversion is recommended to monitor treatment response [23]. However, both these methods have important limitations, including the availability of sputum, the operator-dependency of the outcomes and in the case of smear-microscopy, low sensitivity and specificity due to the inability to distinguish live from dead bacteria [3, 24]. Many novel biomarkers are currently being developed that may provide alternative, better methods of monitoring treatment response. Broadly speaking, these biomarkers can be split into two main categories, those based on bacterial burden, often sharing the same downside of being reliant on the availability of sputum, and those based on host characterisation.

Biomarkers based on bacterial burden rely on the measurement of the number of viable bacteria and the decrease thereof as a proxy for treatment efficacy. While routine solid and liquid culture can be used to infer bacterial burden, these methods are both time-consuming and prone to contamination, which results in the loss of signal. Therefore, there is a need for simpler, faster assays for quantifying bacterial burden. A few promising biomarkers of bacterial burden in development are the measurement of LAM in both sputum or urine, the mycobacterial load assay (MBLA) and the measurement of Mycobacterium tuberculosis complex protein 64 (MPT64) [25–27]. These biomarkers all measure the concentration of TB-specific molecules including LAM from the cell wall, M. tuberculosis 16S ribosomal RNA (MBLA) and the TB-specific antigen MPT64 as a proxy for bacterial burden. The TB MBLA has been shown to be specific and not compromised by contamination of the sample by oral or other bacteria. Quantitative results correlate well with both solid and liquid culture methods [26]. It can be applied to other specimens including stool [28]. It can be used to distinguish the response to different regimens and can be used to better predict outcome [29, 30].

Biomarkers based on host characterisation relate to the concept of cure and aim to measure host characteristics that correlate with disease severity and thereby treatment response and success. Many host-related biomarkers in development are aimed at characterising the immunological response. Whole-blood transcriptomic profiling has yielded promising transcriptional signatures that correlate with treatment outcome, and multiple assays based on these signatures are in development including the Cepheid host response cartridge, the RISK6 transcriptomic signature and the TB22 assay [31–33]. In addition, the host antibody profile changes throughout treatment and may be used to monitor the response to treatment and the likelihood of treatment success [34].

Conclusion

TB remains a global health challenge and necessitates improved diagnostic and prognostic approaches. The limitations of traditional diagnostic methods highlight the urgent need for innovative biomarkers. Recent advances have been made in identifying biomarkers for risk of disease, active disease, and treatment response. Such advances not only enable timely treatment initiation but also enhance patient outcomes and aid in controlling disease transmission. The emerging biomarkers offer promising alternatives to traditional methods for monitoring treatment efficacy and predicting treatment outcomes. Together, these innovative biomarkers and assays have the potential to significantly contribute to achieving the goal of decreasing the global TB burden. Currently, much of the work on biomarkers is undertaken retrospectively or is nested within other studies that are not powered for the evaluation of biomarkers. Therefore, further studies aimed specifically at determining the usability of biomarkers in clinical practice are needed before biomarkers can be routinely implemented.