INTRODUCTION — Forms of central nervous system (CNS) infection due to Mycobacterium tuberculosis include meningitis, tuberculoma, and spinal arachnoiditis.
Issues related to management of CNS tuberculosis (TB) will be reviewed here. The epidemiology and pathogenesis, clinical manifestations, and diagnosis of CNS TB are discussed separately. (See "Central nervous system tuberculosis: An overview" and "Tuberculous meningitis: Clinical manifestations and diagnosis".)
Issues related to pulmonary TB and miliary TB are discussed separately. (See "Clinical manifestations and complications of pulmonary tuberculosis" and "Diagnosis of pulmonary tuberculosis in adults" and "Epidemiology and pathology of miliary and extrapulmonary tuberculosis".)
Issues related to treatment of TB are discussed separately. (See "Treatment of drug-susceptible pulmonary tuberculosis in nonpregnant adults without HIV infection" and "Treatment of drug-resistant pulmonary tuberculosis in adults".)
Issues related to treatment of TB in patients with HIV infection are discussed separately. (See "Treatment of drug-susceptible pulmonary tuberculosis in nonpregnant adults with HIV infection: Initiation of therapy" and "Treatment of pulmonary tuberculosis in adults with HIV infection: Follow-up after initiation of therapy".)
TUBERCULOUS MENINGITIS
Treatment — Treatment of tuberculous meningitis consists of prompt administration of antituberculous therapy, together with glucocorticoids. For patients with hydrocephalus, surgical consultation is warranted. Issues related to management of hyponatremia are discussed separately. (See "Overview of the treatment of hyponatremia in adults".)
Empiric treatment should not be delayed, given the high mortality and complication rate of untreated infection (see 'Outcomes' below). Initiation of empiric treatment is warranted for patients with relevant epidemiologic factors (history of prior TB infection or disease, known or possible TB exposure, and/or past or present residence in or travel to an area where TB is endemic), clinical manifestations (subacute presentation of stiff neck, headache, fever and vomiting), pertinent radiographic findings (such as hydrocephalus, basilar meningeal thickening, edema and/or tuberculomas), and typical cerebrospinal fluid (CSF) findings (lymphocytic pleocytosis, elevated protein concentration, and low glucose concentration).
The likelihood of a favorable clinical outcome is increased with early initiation of therapy; treatment delay (even for a few days) may be associated with increased likelihood of unfavorable outcome [1]. (See 'Outcomes' below.)
Antituberculous therapy — The principles of treatment of tuberculous meningitis are derived from those governing the management of pulmonary TB. The treatment regimens outlined below reflect guidance from United States Centers for Disease Control and Prevention, British Society guidelines, and World Health Organization (WHO) guidelines for treatment of CNS TB [2-4]. The evidence for treatment of pulmonary TB is discussed separately. (See "Treatment of drug-susceptible pulmonary tuberculosis in nonpregnant adults without HIV infection", section on 'Antituberculous therapy'.)
Drug susceptible disease
Traditional regimen — In general, treatment of CNS TB consists of an initial intensive phase (4 drugs administered for 2 months) followed by a prolonged continuation phase (usually 2 drugs administered for an additional 7 to 10 months), for a total treatment duration of 9 to 12 months. Ideally all patients with TB should be treated with clinical case management and directly observed therapy. (See "Adherence to tuberculosis treatment".)
●Intensive phase − Empiric intensive phase drug regimens consist of the following:
•Adults − In adults, for empiric treatment of CNS TB (not known or suspected to be drug resistant), the intensive-phase four-drug regimen consists of isoniazid, rifampin, pyrazinamide, and a fourth agent administered daily for two months [2-4].
In children, the intensive phase four-drug regimen consists of isoniazid, rifampin, pyrazinamide, and either ethionamide or streptomycin (in place of ethambutol, given difficulty associated with monitoring for ethambutol-associated optic neuritis) administered daily for two months [5].
Isoniazid, rifampin, and pyrazinamide are bactericidal, penetrate inflamed meninges, and achieve CSF levels that exceed the minimum inhibitory concentration for sensitive strains. Isoniazid has excellent CNS penetration and is more active against rapidly dividing than against semi-dormant organisms. Rifampin is active against both rapidly dividing organisms and semi-dormant organisms, despite its relatively poor CNS penetration [6]. Pyrazinamide readily penetrates the CNS and is highly active against intracellular mycobacteria.
For selection of the fourth drug, there is a lack of consensus among authorities, and there are no data from controlled trials. Options include ethambutol, streptomycin, levofloxacin, or ethionamide [6]. The WHO guidelines favor streptomycin over ethambutol [7]. Drug doses are shown in the tables (table 1 and table 2).
Use of higher-dose rifampin and levofloxacin during the intensive phase has not been associated with improved survival. In a randomized trial including 817 patients with tuberculous meningitis (of whom 349 were had HIV infection), patients were randomly assigned to treatment with a standard regimen or an intensified regimen that included higher-dose rifampin (15 mg/kg per day) and levofloxacin (20 mg/kg per day) during the intensive phase; during 9 months of follow-up, no survival benefit was observed between the groups (survival probability approximately 75 percent) [8]. Use of even higher rifampin dosing (>30 mg/kg) warrants further study; this may improve CSF penetration and may overcome the issue of isoniazid monoresistance [9].
•Children − In children, monitoring for ethambutol-associated optic neuritis can be difficult; therefore, we are in agreement with the American Academy of Pediatrics which recommends the substitution of either ethionamide or an aminoglycoside (such as streptomycin) in place of ethambutol in the initial empiric treatment regimen, given difficulty associated with monitoring for ethambutol-associated optic neuritis [5]. Pediatric dosing of antituberculous agents is summarized separately. (See "Tuberculosis disease in children".)
●Continuation phase − In the setting of infection known or presumed to be caused by susceptible strains, the continuation phase consists of isoniazid and rifampin (given daily), continued for 7 to 10 months [2-4,7]. The optimal duration of treatment is uncertain; in general, an extended period of treatment is administered because tuberculous meningitis is associated with high rates of disability and deaths. In one review and meta-analysis comparing six months versus longer treatment duration for tuberculous meningitis with at least isoniazid, rifampin, and pyrazinamide, there was no difference in two-year relapse rate between groups (1.5 versus 0 percent) [10], suggesting six months might be sufficient.
The treatment regimen should be tailored to the drug susceptibility data (when available) and clinical response.
Following initiation of antituberculous therapy, paradoxical worsening of clinical manifestations may occur. (See 'Paradoxical worsening' below.)
Shortened regimen — While the WHO stated in 2021 that an alternative shortened regimen (six-month regimen of daily isoniazid, rifampin, pyrazinamide, and ethionamide) may be used as an alternative treatment approach for children and adolescents [11], we continue to favor use of the traditional regimen pending further data. The shortened regimen should be used only for patients without HIV infection and with no evidence of drug-resistant TB [12].
A six-month regimen of daily isoniazid, rifampin, pyrazinamide, and ethionamide has been evaluated as an alternative option for the treatment of children and adolescents with tuberculous meningitis [11]. In a systematic review and meta-analysis including three studies and more than 1000 participants, lower mortality (8 versus 24 percent) and higher treatment success rate (83 versus 75 percent) were observed with the shortened regimen compared with the traditional regimen [11].
Drug resistant infection — The prevalence of M. tuberculosis strains resistant to one or more first-line drugs is increasing [13]. Risk factors for drug-resistant TB are summarized in the table (table 3). The presence of risk factors should prompt susceptibility testing (by molecular and conventional methods) on diagnostic specimens. (See "Diagnosis of pulmonary tuberculosis in adults", section on 'Drug susceptibility testing' and "Diagnosis of pulmonary tuberculosis in adults", section on 'Molecular testing'.)
The optimal approach to treatment of drug-resistant tuberculous meningitis is uncertain:
●For treatment of isoniazid-resistant tuberculous meningitis, we favor intensive phase treatment with daily rifampin, ethambutol, pyrazinamide, and a fluoroquinolone for 2 months, followed by a continuation phase of rifampin, pyrazinamide, and a fluoroquinolone for at least 10 months [3,14]. This approach is derived from management of drug-resistant pulmonary TB, the evidence for which is discussed separately. (See "Treatment of drug-resistant pulmonary tuberculosis in adults".)
●For treatment of multidrug-resistant tuberculous meningitis, regimen selection should be guided by the approach to treatment of pulmonary disease; the regimen should include any first-line drugs to which the isolate is susceptible, with addition of a fluoroquinolone and additional second-line drugs with good CSF penetration (such as ethionamide and cycloserine) to make a regimen including at least five effective drugs (table 2) [15]. A combination of levofloxacin, kanamycin, ethionamide, linezolid, and pyrazinamide is a favorable regimen for multidrug-resistant tuberculous meningitis; kanamycin penetration is adequate if meninges are inflamed [16].
Data on CSF penetration of bedaquiline, clofazimine and delamanid are limited [17,18]. (See "Treatment of drug-resistant pulmonary tuberculosis in adults".)
The optimal duration of therapy for treatment of drug-resistant CNS disease is uncertain. It may be advisable to extend the duration of therapy to 18 to 24 months, considering the severity of illness, the clinical response, and the patient's immune status. (See "Treatment of drug-resistant pulmonary tuberculosis in adults".)
Antiretroviral therapy — Patients with HIV infection are treated with the same antituberculous therapy as patients without HIV infection as described above. (See 'Antituberculous therapy' above.)
In patients on antituberculous therapy, administration of antiretroviral therapy (ART) requires careful attention to potential interactions.
For patients already on ART prior to diagnosis of CNS TB, modification of the ART regimen may be required in some cases. The approach to selection of ART in patients with CNS TB and HIV infection is discussed separately. (See "Selecting antiretroviral regimens for treatment-naïve persons with HIV-1: General approach" and "Treatment of drug-susceptible pulmonary tuberculosis in nonpregnant adults with HIV infection: Initiation of therapy".)
For patients with CNS TB who are ART-naïve, initiation of ART should be delayed for the first eight weeks of antituberculous therapy, regardless of CD4 count [2]. This is because initiation of ART may be complicated by the immune reconstitution inflammatory syndrome (IRIS), which can manifest as progression of TB disease or clinical deterioration in patients previously improving on antituberculous therapy. (See 'Paradoxical worsening' below and "Immune reconstitution inflammatory syndrome".)
The above approach is supported by a randomized trial including 253 patients with tuberculous meningitis and HIV infection treated with immediate or deferred ART (two months after randomization) [19]; there was no significant difference in 9-month mortality (hazard ratio [HR] 1.12, 95% CI 0.81-1.55) or time to new acquired immunodeficiency syndrome (AIDS) event or death (HR 1.16, 95% CI 0.87-1.55) between the groups; however, life-threatening adverse events occurred more frequently among those who received immediate ART than those who received deferred ART (80 versus 69 percent).
Glucocorticoids — We are in agreement with the WHO which recommends six to eight weeks of adjunctive glucocorticoid therapy for patients with tuberculous meningitis (suspected or confirmed). We administer glucocorticoids to patients with and without HIV infection. Dexamethasone or prednisolone can be given. After six to eight weeks, glucocorticoids should be tapered off [4]. Our approach is as follows:
●For patients >14 years of age − Dexamethasone 0.3 to 0.4 mg/kg/day intravenously (IV) for 2 weeks, then 0.2 mg/kg/day IV week 3, then 0.1 mg/kg/day IV week 4, then 4 mg per day orally and taper 1 mg off the daily dose each week; total duration approximately 8 weeks [20,21]. In conscious patients, oral prednisolone 0.5 mg/kg (up to 40 mg/day) may be given for 4 weeks and then tapered over the following 4 weeks [22].
●For patients ≤14 years of age − Dexamethasone 0.6 mg/kg/day IV for 4 weeks, followed by a reducing course over 4 weeks [3]. An alternative (oral) regimen consists of prednisone 2 mg/kg daily (maximum 60 mg/day), or its equivalent for 4 to 6 weeks, followed by tapering [5].
The above approach is supported by data from randomized trials demonstrating a mortality benefit associated with use of adjunctive glucocorticoid therapy in patients without HIV infection [7,20,23-25]. Among patients with HIV infection, data are sparse and conclusions regarding efficacy are less certain:
●In a meta-analysis of nine trials including 1337 children and adults with tuberculous meningitis, glucocorticoid therapy was associated with a lower mortality rate over 2 to 24 months follow-up (31 versus 41 percent [risk ratio 0.75, 95% CI 0.65-0.87]) [25]. Rates of disabling neurologic deficits and serious adverse events were similar among treatment groups. The efficacy of corticosteroid therapy also appeared similar across categories of disease severity. Most (93 percent) of the patients in this analysis were HIV negative.
One randomized trial in 545 adolescents and adults compared the relative efficacy in those with and without HIV infection (98 and 446 participants respectively) [20]. In a subgroup analysis, the relative risk reduction in mortality appeared similar for patients with and without HIV infection, but the analysis was underpowered to show a difference [20,25]. [26-29]
Management of complications — Clinical manifestations and diagnosis of complications of tuberculous meningitis are described separately. (See "Tuberculous meningitis: Clinical manifestations and diagnosis", section on 'Complications'.)
●Hydrocephalus − Progressively enlarging ventricles may manifest with elevated intracranial pressure, which can be life-threatening. Progressive ventriculomegaly can clinically manifest with increasing intracranial tension. Patients with progressive vision impairment and unconsciousness should be considered for CSF diversion via ventriculoperitoneal shunt or endoscopic third ventriculostomy [30-33]. (See "Evaluation and management of elevated intracranial pressure in adults" and "Hydrocephalus in children: Management and prognosis".)
●Hyponatremia − The approach to management of hyponatremia is discussed in detail separately. (See "Overview of the treatment of hyponatremia in adults".)
●Optochiasmatic arachnoiditis − Optochiasmatic arachnoiditis, encasement of optic nerves and optic chiasma by thick tuberculous exudates, is an important cause of vision loss in tuberculous meningitis. This condition may develop paradoxically, despite administration of appropriate antituberculous therapy.
The optimal to management of this complication is uncertain. Some patients regain vision with antituberculous therapy and corticosteroids. Neurosurgery may be considered in patients who do not respond to these interventions. Thalidomide, an inhibitor of tumor necrosis factor-alpha, has been associated with clinical improvement in isolated case reports [34].
Follow-up monitoring
General principles — Careful and regular clinical monitoring, including serial neurologic examinations to assess for mental status change and emerging focal deficits, is an essential component of follow-up. The risk of clinical deterioration and death is high during the intensive phase of treatment, and complications may arise or progress unexpectedly. The most challenging complications include hydrocephalus, hyponatremia, tuberculoma, and brain ischemia/infarction. (See "Tuberculous meningitis: Clinical manifestations and diagnosis", section on 'Complications' and "Central nervous system tuberculosis: An overview", section on 'Tuberculoma' and "Initial assessment and management of acute stroke".)
The approach to use of follow-up laboratory and radiographic studies should be guided by individual clinical circumstances and are contingent upon a number of factors; these include patient age, clinical stage of illness at which treatment is initiated, presence and severity of complications, initial response to therapy, and certainty of diagnosis.
Once adequate CSF specimens have been collected for diagnostic purposes, indications for follow-up lumbar puncture include:
●Diagnostic uncertainty and unsatisfactory response to therapy during the intensive phase of treatment
●Presence of atypical features of potential importance on initial CSF examination (such as neutrophilic predominance and unusually high protein concentration, which may be suggestive of emerging spinal block)
●Presence of elevated intracranial pressure in the setting of communicating hydrocephalus, for which lumbar puncture is needed for monitoring and temporary management
Follow-up radiographic imaging is useful to assess response to treatment in patients with complications such as hydrocephalus, tuberculoma, and brain infarction. Follow-up imaging should be done in patients with poor response to treatment, patients with rapid clinical deterioration, and patients with suspected paradoxical reaction.
Antituberculous drugs are associated with a broad array of adverse effects. Hepatotoxicity is an important adverse effect that warrants careful clinical attention. Issues related to clinical monitoring during antituberculous therapy are discussed separately. (See "Treatment of drug-susceptible pulmonary tuberculosis in nonpregnant adults without HIV infection", section on 'Regimen adjustments for drug intolerance' and "Antituberculous drugs: An overview", section on 'Clinical and laboratory monitoring'.)
For patients who initiate empiric antituberculous therapy in the setting of suspected CNS TB but in the absence of definitive diagnosis, the course of treatment should be completed unless an alternative diagnosis is established [3]. The clinical response to treatment, whether favorable or unfavorable, should not be used to determine whether to continue therapy.
Paradoxical worsening — After initiation of antituberculous therapy, paradoxical worsening of clinical manifestations may occur (image 1). The mechanism of paradoxical reaction is not known; it has been hypothesized to occur because of an exaggerated immune response to mycobacterial antigens.
In one study including 141 patients with tuberculous meningitis (of whom 13 had HIV infection), paradoxical worsening was observed in approximately one-third of cases; in most cases such findings were observed within three months of starting antituberculous therapy [35]. Risk factors for paradoxical worsening included female sex, concomitant HIV infection, and a shorter duration of illness prior to clinical presentation.
Among patients with HIV infection, paradoxical worsening typically occurs after initiation of ART and is known as IRIS. The risk of IRIS is increased in patients with an initial CD4 count below 100/microL and in patients with a rapid fall in viral load and rise in CD4 count [3].
Manifestations of CNS disease progression in the setting of IRIS include meningitis, intracranial tuberculoma, brain abscess, radiculomyelitis, and spinal epidural abscess [36-39]. (See "Immune reconstitution inflammatory syndrome", section on 'Tuberculosis'.)
The diagnosis of IRIS is one of exclusion. Patients should be evaluated with radiographic imaging and CSF examination if feasible; CSF examination may demonstrate neutrophilic predominance [35,40,41]. Additional diagnostic considerations include nonadherence to therapy, drug-resistant infection, and other opportunistic infections. (See "Tuberculous meningitis: Clinical manifestations and diagnosis".)
Management of TB-associated IRIS is discussed separately. (See "Treatment of pulmonary tuberculosis in adults with HIV infection: Follow-up after initiation of therapy", section on 'Immune reconstitution inflammatory syndrome'.)
TUBERCULOMA — The clinical manifestations and diagnosis of tuberculoma are discussed separately. (See "Central nervous system tuberculosis: An overview", section on 'Tuberculoma'.)
Clinical approach — Treatment of tuberculoma consists of antituberculous therapy; the approach is the same as for treatment of tuberculous meningitis, as discussed above. (See 'Antituberculous therapy' above.)
In TB-endemic areas, intracranial tuberculoma may persist following comprehensive antituberculous therapy; in some cases they persist for several years. It is uncertain whether persistent tuberculoma represents active disease or revascularization of a healing lesion; most experts believe that persistent tuberculoma does not represent treatment failure and that prolonged antituberculous therapy in an otherwise asymptomatic patient is not needed [42].
No trials have evaluated whether patients with tuberculoma benefit from adjunctive glucocorticoids. We favor their use for patients with cerebral edema (particularly when edema is out of proportion to mass effect in the setting of associated altered mental status or focal neurologic deficits), increased intracranial pressure, and/or presence of concomitant meningitis. (See 'Glucocorticoids' above.)
Surgical consultation is warranted for patients with obstructive hydrocephalus or brainstem compression.
Follow-up monitoring — There are no guidelines or clinical series data upon which to base the approach to frequency of follow-up radiographic imaging in patients with tuberculoma. Important considerations include the location and size of the tuberculoma, the degree of associated cerebral edema, the character of neurologic signs, and clinical progress on therapy. In patients whose lesion is not in a critical location, with minimal to absent ongoing neurologic signs, we obtain follow-up imaging at two weeks, one month, three months, and six months of therapy.
Additional issues related to follow-up monitoring are discussed above. (See 'Follow-up monitoring' above.)
SPINAL ARACHNOIDITIS — The clinical manifestations and diagnosis of spinal arachnoiditis are discussed separately. (See "Central nervous system tuberculosis: An overview", section on 'Spinal arachnoiditis'.)
Clinical approach — Treatment of spinal arachnoiditis consists of antituberculous therapy, as discussed above. (See 'Antituberculous therapy' above.)
In the setting of spinal arachnoiditis (particularly in the setting of multilevel disease), an extension of the treatment duration up to 18 months may be warranted, tailored to individual clinical circumstances.
No trials have evaluated whether patients with spinal arachnoiditis benefit from adjunctive glucocorticoids. We favor their use in patients with spinal block (cerebrospinal fluid protein ≥500 mg/dL) and in patients with acute cord compression. Patients with concomitant meningitis should be treated as well. (See 'Glucocorticoids' above.)
Surgical intervention (excision, decompression) plays a limited but important role in select patients with certain forms of spinal TB who are not responding satisfactorily to medical treatment [43]. These include intramedullary tuberculoma, extramedullary intradural tuberculoma, spinal tuberculous abscess, or local compression due to spinal arachnoiditis.
Additional issues related to surgery for patients with spinal TB are discussed separately. (See "Bone and joint tuberculosis", section on 'Surgery'.)
Follow-up monitoring — Issues related to follow-up monitoring are discussed above. (See 'Follow-up monitoring' above.)
OUTCOMES — Mortality associated with tuberculous meningitis is as high as 55 to 75 percent, especially in patients with HIV infection and those with advanced disease on presentation [44,45]. In a meta-analysis including 6 studies and 663 patients hospitalized with tuberculous meningitis, the mortality rate was 42 percent [46]. Late sequelae of tuberculous meningitis include cranial nerve palsies, gait disturbance, hemiplegia, blindness, deafness, learning disabilities, dementia, and syndromes of hypothalamic and pituitary dysfunction [1].
Among patients with tuberculous meningitis, the clinical outcome depends on the extent of disease at the time of treatment initiation. In one study including 48 patients with tuberculous meningitis admitted to intensive care, the mortality rate was 65 percent; factors associated with poor outcome included clinical stage at admission and delay in treatment initiation ≥3 days [47]. In addition, 4-drug antituberculous therapy was administered in 10 percent of cases; most patients were treated with ≤2 antituberculous agents. In another study including 52 patients with tuberculous meningitis, the mortality rate among patients who started treatment prior to clinical stage-1 or early stage-2 disease was less than 10 percent; of 11 patients with stage-3 disease, 6 died or suffered severe neurologic sequelae [48].
Radiographic findings may be useful predictors of outcome in some circumstances; patients with a normal computed tomography scan often recover with appropriate antituberculous therapy, whereas presence of hydrocephalus and marked basilar enhancement carries a poor prognosis. Tuberculomas generally resolve with antituberculous therapy; in some cases, they may heal with calcification.
A bedside score (MASH-P) including five variables (modified Barthel index score (table 4) [M], age [A], stage of tuberculous meningitis [S], presence of hydrocephalus [H], and papilledema [P]), may be useful to predict probability of death in immunocompetent patients with tuberculous meningitis [49]. The score ranges from 0 to 10, with a higher score correlating with higher probability of death; those with a score of 0 had 1.7 percent risk of death, while those with score of 10 had 65 percent risk of death.
Factors that may be associated with better prognosis among adults with tuberculous meningitis include higher Glasgow Coma Scale and a higher plasma sodium concentration [50].
Drug resistance has been associated with diminished prognosis among patients with CNS TB. One study in Vietnam including 180 adults with tuberculous meningitis combined isoniazid and rifampin resistance (observed in 5 percent of cases) was strongly predictive of death (relative risk of death 11.6, 95% CI 5.2-26.3) [51].
Concomitant Strongyloides stercoralis infestation may modulate the inflammatory response to M. tuberculosis infection. In one study including 668 patients with tuberculous meningitis, 9.4 percent of patients tested positive for S. stercoralis (via serology, stool microscopy, or and/or stool polymerase chain reaction) [52]. Concomitant S. stercoralis infection was associated with lower CSF neutrophil counts and lower CSF proinflammatory cytokines.
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Diagnosis and treatment of tuberculosis".)
SUMMARY AND RECOMMENDATIONS
●General principles – Prompt antituberculous therapy is essential in the treatment of central nervous system (CNS) tuberculosis (TB). Because of the high associated mortality, treatment should be initiated when the diagnosis is suspected and not deferred until a diagnosis is established. (See 'Treatment' above and 'Outcomes' above.)
●Antituberculous therapy – In general, treatment of CNS TB consists of an intensive phase (4 drugs administered for 2 months) followed by a continuation phase (2 drugs administered for an additional 7 to 10 months), for a total treatment duration of 9 to 12 months (table 1). (See 'Antituberculous therapy' above.)
•Intensive phase
-Adults – For adults, the intensive phase consists of four drugs (isoniazid, rifampin, pyrazinamide, and ethambutol) administered for two months. This approach is derived from management of pulmonary TB, the evidence for which is discussed separately. (See "Treatment of drug-susceptible pulmonary tuberculosis in nonpregnant adults without HIV infection", section on 'Antituberculous therapy' and 'Traditional regimen' above.)
-Children – For children, the intensive phase consists of four drugs (isoniazid, rifampin, pyrazinamide, and either ethionamide or an aminoglycoside [in place of ethambutol, given difficulty associated with monitoring for ethambutol-associated optic neuritis]) administered for two months.
•Continuation phase – For adults and children, the continuation phase consists of 2 drugs (isoniazid and rifampin); we suggest a duration of 7 to 10 months over a shorter duration (Grade 2C). (See 'Traditional regimen' above.)
●Drug-resistant infection – For patients with isoniazid-resistant CNS TB, we favor treatment with daily rifampin, ethambutol, pyrazinamide, and a fluoroquinolone. This approach is derived from management of drug-resistant pulmonary TB, the evidence for which is discussed separately. In addition, the duration of therapy should be extended to 18 to 24 months, taking into account the severity of illness, clinical response to therapy, and the patient's immune status. (See 'Drug resistant infection' above and "Treatment of drug-resistant pulmonary tuberculosis in adults".)
●Glucocorticoids – The approach to use of glucocorticoids depends on the clinical presentation:
•Meningitis
-Patients without HIV infection – For patients with established or suspected tuberculous meningitis in the absence of HIV infection, we recommend adjunctive glucocorticoid therapy (Grade 1A). Clinical trial data have found that such treatment improves short-term survival.
-Patients with HIV infection – For patients with HIV infection, we suggest adjunctive glucocorticoid therapy (Grade 2C). Data showing a survival benefit in patients with HIV infection are limited. (See 'Glucocorticoids' above.)
•Tuberculoma – For patients with tuberculoma, we suggest adjunctive glucocorticoid therapy for patients with cerebral edema (particularly when edema is out of proportion to mass effect in the setting of associated altered mental status or focal neurologic deficits), and/or elevated intracranial pressure (Grade 2C). (See 'Tuberculoma' above.)
•Spinal arachnoiditis – For patients with spinal arachnoiditis, we suggest adjunctive glucocorticoid therapy for patients with spinal block (cerebrospinal fluid protein ≥500 mg/dL) and/or patients with acute cord compression (Grade 2C). (See 'Spinal arachnoiditis' above.)
●Antiretroviral therapy – For patients with HIV infection and CNS TB who are not already on antiretroviral therapy (ART), we suggest deferral of ART until eight weeks after starting TB treatment, regardless of CD4 count (Grade 2C). (See 'Antiretroviral therapy' above.)
●Paradoxical worsening – Development of immune reconstitution inflammatory syndrome (IRIS) occurs in approximately one-third of patients with CNS TB; it typically manifests with paradoxical worsening within three months of antituberculous therapy and may cause severe or fatal neurologic complications. The management of TB-associated IRIS is discussed separately. (See 'Paradoxical worsening' above and "Treatment of pulmonary tuberculosis in adults with HIV infection: Follow-up after initiation of therapy", section on 'Immune reconstitution inflammatory syndrome'.)
