Tuberculous Meningitis Across the Lifespan

Abstract

Mycobacterium tuberculosis most commonly causes lung disease but can infect the central nervous system (CNS) causing tuberculous meningitis—the most lethal form of tuberculosis. In 2023, 10.8 million people developed tuberculosis [1]. A recent burden of disease study in adults estimated the global incidence of tuberculous meningitis was between 129 000 and 199 000 cases, with up to 70 000 (35%–54%) of these cases coinfected with human immunodeficiency virus (HIV) [2]. The model-estimated mortality rate was 27% [2]. A 2020 meta-analysis of treatment outcomes in adult tuberculous meningitis found a pooled mortality rate of 23%, with much higher mortality in patients with HIV (57%) compared to those who were HIV-negative (16%), and neurological disability was reported in 32% of survivors [3]. In young children with tuberculosis, tuberculous meningitis occurs in around 4% of those aged <5 years and in around 2% of children between 5 and 9 years. A meta-analysis of childhood tuberculous meningitis reported a risk of mortality of 19%, with a risk of neurological disability of 54% in survivors [4].

The differences in presentation and outcome of tuberculous meningitis between adults and children are poorly understood and rarely described. Most studies focus on either age group or pool the data across ages. In a rare comparative report of adults and children with tuberculous meningitis from a single center [5], differences were reported in presentation and outcomes: children presented with earlier stages of tuberculous meningitis and commonly with vomiting, reported more known tuberculosis contacts, were more likely to have confirmed tuberculous meningitis, and had more neurological sequalae. Adults presented more commonly with signs of confusion, history of tuberculosis, and comorbidities. These findings highlight important age-related differences, such as stage of brain maturation and general health status, the ability to verbally express disease symptoms and a history of pulmonary tuberculosis to raise the suspicion of tuberculous meningitis early versus the diagnostic challenge of nonspecific symptomatology. Differences between adult and pediatric tuberculous meningitis is not unexpected given that their considerable physiological differences may contribute to unique pathogenetic and pathophysiological mechanisms of tuberculous meningitis. In this article we suggest possible age-specific considerations that could be relevant to tuberculous meningitis across the lifespan and describe clinical characteristics, management, and outcome from research on adult and pediatric tuberculous meningitis over the last 5 years.

PATHOGENESIS

M. tuberculosis is transmitted within aerosolized droplets. Inhaled bacteria infect alveolar macrophages, neutrophils, and dendritic cells [6, 7]. These cells release chemokines and cytokines that initiate an inflammatory process and stimulate activation of T-helper (Th) cells to assist in the encapsulation of the bacilli in granulomas [8]. M. tuberculosis is often contained in the lungs after inhalation of aerosolized droplets, but in cases where the host-immune response is inadequate, dissemination may occur [8].

Several factors may impair the immune response in tuberculous meningitis—comorbidities like diabetes and HIV are seen commonly in adults; malnutrition is seen in children and adults alike and is associated with reduced T-cell and macrophage counts [9, 10].

The route by which M. tuberculosis reaches the brain remains unclear. However, experimental research suggests that M. tuberculosis can enter the brain from the blood by rearranging the actin cytoskeleton of brain endothelial cells lining the blood-brain barrier (BBB), possibly allowing the migration of bacilli across it [8, 11]. Bacteria may also enter the brain via a “Trojan horse” mechanism, where M. tuberculosis passes the BBB inside infected macrophages [12]. Rich and McCordick were first to investigate the pathogenesis of tuberculous meningitis and found concentrated lesions containing bacilli adjacent to the meninges or in the subpial cortex in postmortem tissue of children with tuberculous meningitis [7]. They called these caseating lesions “Rich foci”; rupture of these lesions was thought to cause tuberculous meningitis by initiating an inflammatory response [6, 7].

As the resident macrophages of the brain, microglia are thought to recognize M. tuberculosis upon entry to the CNS and initiate the immune response [13, 14]. Proinflammatory cytokines, such as tumor necrosis factor-α (TNF-α) and interferon-γ (INF-γ), are released and initiate neuroinflammation, as well as chemokines which recruit peripheral leukocytes to the CNS to contribute to the local immune response [13, 14]. Astrocytes play an important role in BBB maintenance, but also participate in neuroinflammation [15].

Although the pathogenesis of tuberculous meningitis appears to be consistent across the lifespan, particularities of the developing immune system alongside the developing brain introduce unique considerations for the neuroinflammatory response in children with tuberculous meningitis, while the higher incidence of HIV coinfection and other comorbidities in adults increases the risk of tuberculous meningitis and poor outcomes (Figure 1) [8].

Figure 1.

Factors contributing to a vulnerability to tuberculous meningitis across the lifespan. Children have immature immune systems, making them more susceptible to developing active tuberculosis and dissemination of tuberculosis to the brain. The diagnosis of tuberculous meningitis is difficult, often resulting in delayed and severe presentation of the disease. This, together with the inflammatory process in the brain and risk of ischemia, lead to poor neurodevelopmental and neurocognitive outcomes. Adults with tuberculous meningitis are frequently coinfected with human immunodeficiency virus (HIV). The compromised immune system caused by HIV increases the risk of active tuberculosis disease and dissemination. The initiation of antiretroviral therapy to treat HIV puts tuberculous meningitis patients at risk to develop an immune reconstitution inflammatory syndrome (IRIS). Tuberculous meningitis IRIS, together with the ongoing inflammatory process in the brain, and the risk of ischemia leads to unfavorable outcomes. Older people are also at increased risk of late presentation and poor tuberculous meningitis outcomes. Created in BioRender. Rohlwink, U. (2025) https://BioRender.com/b86q069

Open in new tabDownload slide

The Role of an Immature Immune System in Tuberculous Meningitis Pathogenesis

Early childhood (<5 years) represents a vulnerable time window where the immune system is undergoing development. In young age, the peripheral immune system consists mostly of naive cell populations that are unable to mount effective responses against many pathogens. This leaves children vulnerable to systemic infections and disseminated forms thereof [10]. Newborns rely on innate immune responses as adaptive immunity remains immature due to lack of antigenic exposure. Reduced numbers of effector T cells and cytokine secretion increases the susceptibility to infections. Adaptive immune responses develop after the age of 2 years with immune maturity reached around 7–8 years [16]. Other contributing factors include maternal health, nutrition, vaccination, and sex hormones [10, 17].

Paucibacillary tuberculosis disease and greater mediastinal lymph node involvement is reported among children aged <10 years [18]. The immune response to M. tuberculosis appears to improve during middle childhood (5–10 years) when immune cell numbers and function improve, and children become less susceptible to tuberculosis infection [18].

Young children are known to be vulnerable to disseminated forms of tuberculosis, like miliary tuberculosis, due to reduced function of their immune cells [18]. The Bacille Calmette-Guérin vaccine protect against tuberculous meningitis in young children by inducing an effective adaptive immune response, particularly memory CD4⁺ T cells, and promoting secretion of Th1 cytokines such as IFN-γ [19, 20]. In a study comparing pulmonary tuberculosis and tuberculous meningitis in children, a greater degree of impaired T-cell responses, decreased abundance in genes involved in T-cell receptor signaling, and T-cell proliferation and activation were found in patients with tuberculous meningitis [21]. In a separate study, transcriptomic data revealed reduced peripheral T-cell responses in pediatric tuberculous meningitis patients [22]. This study compared transcriptomic profiles in ventricular cerebrospinal fluid (CSF) and lumbar CSF, and found neuronal excitotoxicity to be upregulated in ventricular CSF whereas lumbar CSF represented protein translation and cytokine signaling [22]. This highlights the importance of studying site-of-disease samples and their value in understanding the local immune response in the brain.

In the brain, microglia initiate differentiation of other CNS populations and can promote or inhibit neurogenesis via neurotrophic factors [23, 24]. These functions are essential to the development of neural networks. Under pathological conditions like tuberculous meningitis, microglia serve as the innate macrophages of the brain and are responsible for eliciting an inflammatory response. Single-cell ribonucleic acid sequencing data on lumbar CSF of pediatric patients with tuberculous meningitis showed enrichment of highly activated microglial-like cells and other lymphocyte subpopulations (CD4 and CD8 T cells, plasma B cells). Furthermore, these microglial-like cells exhibited complement activation, expressed proinflammatory genes, and were shown to interact mostly with tissue-resident CD4 T cells [25].

While activation of the microglia is essential to pathogen clearance, changes in numerous metabolic pathways associated with neuroinflammation may negatively impact brain recovery and maturation. Activated microglia contribute to alterations in glucose, lipid, fatty acid, and amino acid metabolism, which may contribute to secondary brain injury and disturbed function of neurons and astrocytes [26]. In a pediatric tuberculous meningitis study, markers of neuronal injury and astrocyte activation were elevated in CSF and associated with poor prognosis [27]. Astrocytes constitute 30% of the CNS and play a crucial role in synaptic development during neurodevelopment but also contribute to the neuroinflammatory process [28]. Therefore, understanding the impact of microglial and astrocyte activation on overall brain metabolism and function in the context of both neuroinflammation and neurodevelopment in children with tuberculous meningitis warrants further investigation.

Currently, corticosteroids are the only host-directed immune modulator proven to increase survival from tuberculous meningitis, but they do not reduce neurological sequelae [29, 30]. Improving our understanding of the dual role of brain-derived immune cells, such as microglia and astrocytes, during inflammation and neurodevelopment could potentially aid in identifying more effective and age-appropriate host-directed therapies.

The Role of a Compromised Immune System in Tuberculous Meningitis Pathogenesis

Important causes of a compromised immune response in adults include diabetes and HIV coinfection [1]. HIV coinfection is less commonly reported in children given the widespread efficacy of prevention of mother-to-child transmission programs [31]. HIV results in reduced CD4⁺ T cells and macrophages, which play important roles in the containment of M. tuberculosis and formation of granulomas. HIV has been shown to impair TNF-α–mediated macrophage activation, which is crucial in containing tuberculosis, disrupt the BBB, and lead to more frequent dissemination of M. tuberculosis to the brain [32, 33]. Microglia are the primary antigen-presenting cells of the CNS, but also the primary brain cells infected by HIV. Literature suggests that HIV infection negatively impacts microglial ability to fight infection and contribute to regeneration and repair, and may precipitate microglial release of neurotoxic factors, which can lead to neuroinflammation, neurodegeneration, and death of surrounding neurons [34].

Antiretroviral therapy (ART) is lifesaving in patients coinfected with HIV; however, ART initiation may result in complications known as tuberculosis-associated immune reconstitution inflammatory syndrome (IRIS) in ART-naive patients [35]. These patients initially respond well to tuberculosis treatment but develop recurrent, new, or worsening symptoms, or deterioration of clinical and radiological manifestations of tuberculosis disease after the initiation of ART. These inflammatory complications are thought to result from a reconstitution of pathogen-specific immunity and occur in around 40% of adults with pulmonary tuberculosis [36]. The incidence of tuberculous meningitis IRIS amongst adults is reported to be 10% and it is known to be associated with poorer mortality and morbidity [37]. The most common symptoms associated with tuberculous meningitis IRIS include headache, neck stiffness, confusion, and new-onset seizures [37]. Higher bacterial loads in CSF increases the risk of developing tuberculous meningitis IRIS, and tuberculous meningitis IRIS has been associated with increased CSF concentrations of inflammatory mediators and elevated neutrophil counts [38].

In children, tuberculosis IRIS is less commonly described; a recent pediatric study found a tuberculosis IRIS incidence of 18.8% and tuberculous meningitis IRIS was rare [39]. A pediatric tuberculous meningitis study in patients coinfected with HIV found that only 1 patient in a cohort of 31 children developed paradoxical worsening of a tuberculoma while on ART [40]. Paradoxical worsening may occur in the absence of HIV coinfection, possibly as an excessive immune response to treatment-initiated release of bacterial antigens, but the underlying pathophysiology remains unclear [41].

Older people often become immunosenescent, involving impaired T-cell production and reactivity and reduced B-cell numbers, receptors, and proliferation [42]. Consequently, older adults may have impaired B- and T-cell functions, and reduced capacity to generate immune memory.

Neurophysiological Vulnerabilities to Excessive Neuroinflammation

An identifying feature of the neuroinflammatory response to tuberculous meningitis is the development of inflammatory exudate, particularly at the base of the brain. The exudate disrupts the flow of CSF, causing hydrocephalus and raised intracranial pressure (ICP), and causes vasculitis and occlusion of major cerebral vessels [8]. Collectively these changes in intracranial dynamics put the brain at risk of ischemia; infarcts occur commonly and drive the high rates of mortality and morbidity [27].

Given that the brain exists within a fixed bony cavity there is finite space for cerebral tissue, blood, and CSF. Should one of these components change in volume this can quickly impact on ICP [43]. The brain's compliance reflects its ability to maintain a stable ICP in the context of changes in the intracranial contents, all of which vary with age. Cerebral blood flow and volume are lowest in neonates, peak between the ages of 3 and 7 years (reaching between 140% and 175% of adult levels), and then decline to adult values [44]. Children also have a much higher cerebral metabolic rate due to evolving myelination and synaptogenesis [44]. Their vulnerability to irreversible brain damage because of ischemia due to exudative neuroinflammation is likely greater than in adults, and the functional consequences of disruption to the developing brain more considerable. Although there are no clear age-related normative values for ICP, data from studies in traumatic brain injury suggest that children are more sensitive to increases in ICP at lower thresholds than adults, and in the context of tuberculous meningitis-associated hydrocephalus this could have implications for appropriate treatment approaches, but this requires proper investigation [45].

CLINICAL AND RADIOLOGICAL CHARACTERISTICS

The international research consensus criteria by Marais et al is commonly used in research to categories tuberculous meningitis diagnosis as definite, probable, or possible [46]. The criteria are based on typical clinical, radiological, and CSF findings. Common clinical characteristics include long duration of symptoms (≥5 days), systemic symptoms of tuberculosis, history of tuberculosis contact, and focal neurology (including a decreased level of consciousness) (Figure 2) [46]. Typical radiological findings include hydrocephalus, basal meningeal enhancement, infarcts, and tuberculomas, and CSF findings include an increased cell count (10–500 per μL) with lymphocyte predominance (>50%), increased protein (>1 g/L), and decreased glucose (<2.2 mmol/L) (Figure 2) [46]. Definite tuberculous meningitis is diagnosed based on successful isolation of M. tuberculosis from CSF using either culture, nucleic amplification tests, or microbiological staining [46]. Other diagnostic scoring tools are also available, but their performance varies across populations [47].

Figure 2.

Commonly reported characteristics of tuberculous meningitis. Clinical characteristics include fever, vomiting, a decreased level of consciousness, malnourishment, and seizures, headache, meningism, and signs of pulmonary tuberculosis. Imaging findings often show hydrocephalus, basal meningeal enhancement, and infarcts. Cerebrospinal fluid characteristics include elevated white blood cell count, lymphocyte predominance, and increased protein. Created in BioRender. Rohlwink, U. (2025) https://BioRender.com/a03y524

Open in new tabDownload slide

We have not performed a formal meta-analysis, and studies were not selected on uniform criteria, therefore no absolute conclusions regarding differences in these characteristics between adults and children can be drawn. Table 1 reports the range of reported frequencies for these key tuberculous meningitis characteristics to describe possible trends and stimulate further research.

A recent study showed that 73% of children presented with tuberculous meningitis stage III (the most severe stage, with coma) compared to 41% of adults, which may indicate that children present with more severe disease [74]. The most reported symptoms in various studies in children included fever, altered level of consciousness, and seizures (Table 1), which could reflect the higher frequency of stage III [40, 48]. Fever, headache, and meningism were frequently reported in adults (Figure 2) [49, 54, 55]. Malnutrition is a well described feature of pediatric tuberculous meningitis, and loss of weight is a strong clue for early diagnosis [75]. Basal meningeal enhancement, tuberculomas, and infarcts were often reported in pediatric and adult patients (Table 1) [40, 49–51, 54, 56]. Hydrocephalus is more common in children and the associated increase in ICP may contribute to their higher frequency of seizures, but both features may also reflect more severe disease at presentation [51, 52]. HIV coinfection has been associated with less radiological evidence of exudate and tuberculoma formation, possibly due to impaired granulomatous inflammation, although similar findings have been noted in patients with age-appropriate CD4 counts [76, 77]. CSF findings were similar across tuberculous meningitis studies and followed expected tuberculous meningitis patterns (Figure 2), although atypical findings have been reported in children and adults coinfected with HIV, including a neutrophilic predominance, normal or elevated glucose, and low protein [38, 40, 52, 56, 57].

In a study comparing younger adult tuberculous meningitis patients (18 to 59 years, n = 176) to older patients (60 to 76 years, n = 21), it was found that older patients presented with a decreased level of consciousness more frequently (67% vs 40%, P = .02), experienced significantly more peripheral nerve dysfunction, changes in cognition, and focal seizures, were more likely to present in late stages of the disease, and had poorer neurological outcome [58]. Signs and symptoms that were significantly less common in older people were headache, meningism, and vomiting [58]. Another study comparing middle-aged to older tuberculous meningitis patients found similar results with older patients (>60 years) presenting with stage II and III tuberculous meningitis [78], significantly higher mortality rate (43.6% vs 23.3%, P = .02), and more frequently presented with diabetes [78]. Diabetes increases the risk of tuberculosis by 1.5 to 2.4 times, but the association with tuberculous meningitis is mixed across studies [79]. These studies highlight the relevance of the age continuum in the manifestation of tuberculous meningitis.

MANAGEMENT

Management of tuberculous meningitis involves antituberculous antibiotics, steroids, hydrocephalus management, and supportive care. Table 2 shows differences in the World Health Organization recommended first-line drug regimen for adults and children [80, 81].

Physiological differences between children and adults lead to variation in drug pharmacokinetics. Infants aged <2 years have delayed gastric emptying, affecting absorption [82]. Distribution is affected by proportions of total body water and drug-protein binding affinities, which changes with age [82]. Levels of cytochrome P450-dependent enzymes in children aged <10 years are 30%–60% of those in adults, and infants <1 year have a decreased glomerular filtration rate and slowed excretion, leading to lower drug concentrations in children [82].

OUTCOMES AND COMPLICATIONS

Although this review cannot formally compare the differential risk of poor outcome between adults and children, poor outcomes in survivors appear to be more common in pediatric studies. This is consistent with the meta-analyses on treatment outcomes in adults and pediatric tuberculous meningitis, which found that neurological disability was present in 32% of adult survivors [3] but 54% of pediatric survivors [4]. In these studies, the mortality rates were similar, 23% in adults and 19% in children, but adult deaths were influenced by HIV coinfection (57% in patients with coinfection vs 16% in patients without HIV infection). Outcomes appear to be worse in children aged <2 years and adults >60 years, and factors that commonly contribute to poor outcome across the lifespan include brain infarcts and proxy markers of severe disease (like hydrocephalus, focal neurology, Glasgow Coma Scale, and perturbations in CSF parameters) [50, 52, 58, 83–86]. Morbidity may improve with time [87].

Neurocognitive impairment and neurodevelopmental deficits are well reported in pediatric tuberculous meningitis, even when brain scans appeared normal [88]. This may be influenced by age (greater vulnerability in younger children), likely reflecting critical stages of brain maturation and plasticity [89]. Data on neurocognitive outcomes in adults are scarce. It seems that over half of surviving adult patients suffer cognitive impairments 6 months to 1 year after diagnosis [90–92]. Similar to children, the main cognitive areas affected in adults are attention, executive function, working memory, and learning memory [93], with children also experiencing learning, attention, and behavior disorders [90]. In adults and children, the cause of these impairments is likely cerebrovascular complications and infarction [90, 94, 95]. The long-term consequences of these deficits are challenging for all patients, but the lifelong implications for children are greatest. The cumulative impact of the enduring need for special schooling and rehabilitation that frequently outstrips access and resources can present a much heavier burden for patient and family. There is a lack of standardization in the measurement of treatment outcomes; however, numerous age-appropriate tools have been recommended [3, 4, 90].

RESEARCH PRIORITIES AND FUTURE DIRECTIONS

Data from single centers that treat adults and children in a standardized way are needed to compare patient characteristics and outcome. A systems biology approach to site-of-disease immune response (human and preclinical) could reveal age and brain-specific responses to tuberculous meningitis in the developing, developed, and degenerating brain. Key age-related considerations are summarised in Table 3.

CONCLUSION

Tuberculous meningitis is a devastating disease for children and adults. The major disease burden appears to rest on the very young, older people, and immune-compromised people. Contributing factors include an immature or impaired immune response, and neurological and systemic vulnerability, all of which vary across the lifespan (Table 3). A better understanding of these peculiarities of age could aid in the identification of novel and targeted, age-appropriate, host-directed treatment approaches and thresholds. Equipped with this knowledge perhaps the future holds promise for those at greatest risk.

Notes

Author contributions. U. R. and G. T. planned the manuscript. R. L., G. S., T. A., G. T., and U. R. wrote the manuscript. R. L., G. T., and U. R. edited the manuscript. G. T. and U. R. reviewed the final versions of the manuscript.

Financial support. This work was supported by the National Research Foundation (South African Research Chairs Initiative Postdoctoral Scholarship to R. L.); the Oppenheimer Memorial Trust (postdoctoral scholarship to R. L.); Carnegie Corporation of New York (PhD fellowship to G. S.); Science for Africa Foundation: Developing Excellence in Leadership, Training and Science Foundation in the African Leadership for Measuring Brain Health in Children and Adolescents network (fellowship to T. A.); the Wellcome Trust (grant numbers 225167/A/22/Z to G. T. and 224176/Z/21/Z to U. R.); and the National Institutes of Health (grant numbers R01AI145781 and R01AI165721 to G. T.).

References

Author notes