Beyond the duality: the spectrum between active and latent tuberculosis

The spectrum of TB

TB is a complex disease that exists along a dynamic spectrum, ranging from complete bacterial clearance to active, symptomatic disease (for review7). Among those outcomes, recent research increasingly points to a wide diversity of TB states that are often challenging to distinguish from one another.

From clinical diversity to diagnostic challenges

Following exposure to Mtb, two broad outcomes are possible: elimination of the pathogen by the immune system or persistence of infection. Elimination of Mtb can occur either through innate immune response, leading to a negative tuberculin skin test (TST) or interferon-gamma release assay (IGRA), or through adaptive immune response,8 where TST and IGRA outcomes depend on memory T cell activation.7 Some individuals, called “resisters,” appear to be resistant to Mtb infection and remain TST/IGRA negative despite multiple exposures.9 Recent studies have shown that they develop Mtb-specific adaptive immune responses lacking IFN production, with CD4+ T cells displaying Th17- and regulatory-like features.10 Additionally, individuals called “reverters” show transient TST or IGRA positivity before reverting to test negativity over time, possibly indicating self-clearance of infection.11,12

If not cleared, Mtb persists in the body, mostly in the lungs but also in other tissues such as the bone marrow13 and adipose tissue.14 Most people control but do not eliminate the infection, leading to LTBI. In clinical and epidemiological studies, LTBI is generally inferred from a positive TST or IGRA in the absence of clinical evidence of active disease; however, these tests detect immune sensitization to Mtb antigens and cannot distinguish LTBI from other states within the TB spectrum.15 Within two years after infection, 5 to 15% of people will spontaneously progress to the active, symptomatic, and transmissible state, termed primary TB, while others remain at risk of reactivation, with an estimated 10% lifetime risk.16 Active TB causes cough, fever and weight loss, and is diagnosed via sputum smear, culture or molecular tests, as immune suppression or disease-induced anergy can cause false-negative TST/IGRA results.17 Additionally, some patients exhibit radiological evidence of TB and culture-positive disease without reporting symptoms, a condition referred to as subclinical TB.18 This form is increasingly recognized as an important contributor to transmission, despite typically lower bacterial loads, and may account for a large proportion of undiagnosed and untreated cases.19

Given the diverse manifestations of Mtb infection, distinguishing between disease states is critical for both clinical management and research. However, current diagnostic tools face significant limitations in accurately defining infection and disease state, making it difficult to differentiate latent infection from active disease or to confirm bacterial clearance. TST measures delayed-type hypersensitivity to mycobacterial antigens but cross-reacts with non-tuberculous mycobacteria and the current vaccinal strain Mycobacterium bovis Bacillus Calmette-Guérin (BCG), and therefore cannot differentiate latent from active TB.20 IGRA, introduced in the 2000s, quantifies IFN-γ responses to Mtb-specific antigens and avoids BCG interference, yet likewise fails to distinguish infection states. The diagnostic work-up for active TB has traditionally relied on chest radiography alongside bacteriological testing by smear microscopy and culture,7 while World Health Organization (WHO)-endorsed rapid molecular tests have increasingly become part of standardized diagnostic algorithms in recent years.21 Since 2014, WHO has been recommending LTBI screening in selected high-risk groups22 (people living with HIV, close contact with patients with active pulmonary TB, individuals at increased risk of progression due to immunosuppressive conditions or therapies). Altogether, diagnostic challenges persist across the TB spectrum, particularly for non-active and early disease states. Coinfections can further obscure the clinical picture by influencing the progression of TB.

Geographic distribution in TB manifestations

The complexity of TB and its diverse manifestations makes it challenging to fully grasp the global impact of the disease. TB burden is unevenly distributed across regions, and the majority of cases occur in low- and middle-income countries, where poverty, malnutrition, and limited healthcare access contribute to higher transmission rates.7 In 2023, most declared TB cases were in the WHO regions of South-East Asia (45%), Africa (24%) and the Western Pacific (17%), with lower shares in the Eastern Mediterranean (8.6%), the Americas (3.2%) and Europe (2.1%).23 This geographic disparity is also reflected in LTBI, which affects an estimated quarter of the global population. 80% of cases are located in South-East Asia (31%), the Pacific region (28%) and Africa (22%), while Europe accounts for 14% of cases.24 However, as mentioned before, only a small proportion of them will suffer from active disease during their lifetime. The large majority will never be diagnosed or experience symptoms, but many of them will encounter other pathogens throughout their life. It is therefore crucial to understand how LTBI affects the outcome of secondary infections, and conversely, whether these infections can impact the control of Mtb. Considering geographical context is essential, as regional disparities in global health, immune status, circulating Mtb lineages, and exposure to distinct coinfections critically shape patient status and must therefore be considered a unifying framework throughout the subsequent sections.

Immune factors in pulmonary TB control

Mtb infection is characterized by complex and finely balanced immune mechanisms. This section briefly summarizes these key concepts to support the interpretation of the coinfection scenarios discussed in Section “bridging scales: from population patterns to animal models and cellular mechanisms of coinfection”. After inhaling infectious aerosols, Mtb reaches the pulmonary alveoli.25 There, pathogen-associated molecular patterns on the bacterial surface are sensed by host immune cells through pattern-recognition receptors, triggering phagocytosis primarily by macrophages, but also by dendritic cells and neutrophils.26,27 Toll-like receptors (TLRs) contribute to bacterial sensing and to the initiation of inflammatory and adaptive immune responses. Among them, TLR2 contributes to T cell co-stimulation by activating antigen-presenting cells, thereby promoting the production of IFN-γ and TNF-α.28 Similarly, TLR9 recognizes unmethylated CpG motifs in bacterial DNA, enhancing the activation of dendritic cells and promoting a Th1-biased immune response that is critical for controlling Mtb infection.29

These early sensing events not only initiate antimicrobial responses but also shape the recruitment and spatial organization of immune cells in the lung, ultimately leading to granuloma formation. In pulmonary TB, these structured aggregates of immune cells form in the lungs and represent a central feature of the host response to Mtb.4 They reflect the duality of Mtb infection: while granulomas can help contain the bacterium and prevent dissemination, they can also provide niches for bacterial persistence or replication.

Effective control of Mtb and prevention of reactivation depend on a balanced, coordinated immune response involving both innate and adaptive immunity. Macrophages are central to this control, activated primarily by IFN-γ to produce reactive nitrogen and oxygen intermediates and other antimicrobial mechanisms that limit bacterial replication.30,31 IFN-γ also induces chemokines like CXCL9 and CXCL10 that participate in adaptive cell recruitment to the granuloma site.32 TNF-α complements this by promoting macrophage activation and maintaining granuloma integrity,33 while it can promote CCL2 production. This chemokine contributes to monocyte recruitment to the site of infection, coordinating the cellular architecture of the granuloma.34 However, elevated levels of CCL2 have been observed in severe forms of TB, reflecting a more intense immune response associated with disease severity.35

In addition to macrophages, several innate immune cell populations, including dendritic cells,36 natural killer (NK) cells37 and neutrophils,38 contribute to early containment, organization of granulomas, as well as antigen presentation that primes the adaptive response.39 The role of eosinophils in TB is still poorly described, although Mtb-induced eosinophil recruitment has been shown to contribute in some settings beneficially to infection control.40

CD4+ T cells play an essential role in sustaining latency, not only through IFN-γ secretion but also by orchestrating immune interactions within the granuloma. Among these, Th1 cells drive macrophage activation, while Th17 cells contribute to mucosal immunity and granuloma structure through IL-17 signaling.41 In murine models, the absence of CD4+ T cells can lead to a loss of infection control, even when CD8+ T cells produce compensatory IFN-γ, suggesting additional regulatory and structural roles for CD4+ T cells.42 CD8+ T cells and unconventional T cells, such as γδ T cells and MAIT cells, also contribute to Mtb control through cytotoxic activity and rapid cytokine production, reinforcing the early containment of infection (for review, see43). Although IFN-γ and TNF-α levels are often elevated in active TB due to increased antigenic load,44 their sustained yet tightly regulated production is essential during latency.45 On the other hand, the anti-inflammatory cytokine IL-10 can suppress protective immunity by inhibiting macrophage activation, phagosome maturation, and Th1 responses.46 Dysregulation, whether through cytokine depletion, excessive inflammation or alternatively activated immune cells, can also lead to immunopathology or reactivation.47 Overall, maintaining TB control requires the sustained coordination of both innate and adaptive immune mechanisms to balance bacterial suppression with tissue preservation. In this context, the inflammatory environment induced by TB can influence the outcome of a secondary infection, while conversely, a secondary pathogen may disrupt the immune equilibrium required to contain the mycobacterial burden. The roles of the immune cell populations involved in TB, together with their key activating cytokines and secreted mediators, are detailed in Table S1.

Modeling TB heterogeneity

While human studies are the most relevant for advancing TB research, they are also the most challenging. The interpretation of results is influenced by a variety of factors, including exposure duration and frequency, bacterial strain, inoculum size, disease severity, and the presence of comorbidities.48 Importantly, as discussed in Subsection “the spectrum of TB”, patients are often classified using imperfect or indirect criteria, making it challenging to accurately position individuals along the TB disease spectrum. This diagnostic uncertainty further complicates the interpretation and comparison of human studies across cohorts and settings.

Murine models are widely used due to their practicality, but they do not fully replicate human TB pathology, lacking features such as fully organized granulomas, caseous necrosis, and pronounced hypoxia.49 Chronic infection in mice has been employed to model LTBI, though the often higher and more sustained bacterial burden does not accurately reflect human latency.50 Nonetheless, these models are crucial for studying the immunological mechanisms and TB progression. Older studies suggest that during this persistent phase, the bacteria may enter a quiescent state, characterized by markedly reduced replication.51 Additionally, the chronic state that mice develop, sometimes induced by a low-dose infection protocol,52 has been utilized by several research groups as a model for LTBI.

Non-human primates, particularly macaques, better recapitulate the full spectrum of TB, including structured granulomas, but their use remains limited,53 especially in coinfection studies.

Other animal models, such as rabbits, guinea pigs, or zebrafish (often infected with Mycobacterium marinum), are also used in TB research (for review, see54), though not in the context of coinfection in the studies cited.

Overall, while animal models have significantly advanced our understanding of Mtb infection persistence and activation, human data still primarily rely on blood samples and clinical or imaging data, rather than direct lung analysis, making it difficult to fully capture the complexity of LTBI. When studying coinfections, it is crucial to develop methods to correlate findings from murine and non-human primate models with epidemiological data from human populations to bridge the gap between experimental and real-world outcomes.

In summary, this first section addresses a particularly challenging question, as TB is not a binary condition but spans a continuum of infection states that are difficult to define, diagnose, and compare across studies and populations. The position of an individual along the TB spectrum is shaped by a finely tuned and context-dependent immune equilibrium. This balance is influenced by numerous factors, including comorbidities, pathogen characteristics, and geographic and socio-epidemiological contexts. These layers of complexity not only complicate the interpretation of human data but also limit the translational relevance of experimental models. Under these conditions, understanding the bidirectional interactions between secondary infections and distinct TB states represents a major conceptual and methodological challenge, but also a critical step toward a more integrated view of TB pathogenesis and immune regulation.