Management of Mycobacterium bovis Meningitis: The Case for Adjunctive Empirical Therapy Authors: Morgan Birabaharan, Annie N Cowell Published: 2026 ABSTRACT AbstractTuberculous meningitis is a devastating manifestation of tuberculosis, disproportionately affecting children and persons with HIV. Despite decades of clinical experience, diagnosis is often delayed and outcomes remain poor. At the US–Mexico border, management is further complicated by the high prevalence of Mycobacterium bovis, a member of the Mycobacterium tuberculosis complex intrinsically resistant to pyrazinamide, one of the few first-line agents with reliable central nervous system penetration. We present a case of M. bovis meningitis in a patient residing near the US–Mexico border, highlighting the diagnostic and therapeutic challenges posed by intrinsic pyrazinamide resistance. Drawing from this case, we review evidence supporting the early use of adjunctive CNS-penetrative antimicrobials during suspected tuberculosis meningitis in regions where M. bovis is endemic. Tuberculosis meningitis (TB meningitis) is the most feared complication of tuberculosis [1–3]. In the absence of therapy, it is almost uniformly fatal, and even with appropriate antimicrobial therapy persons with TB meningitis mortality rates remain as high as 50% and often with lifelong neurological disability [1–3]. Prompt diagnosis and early initiation of effective antituberculosis therapy are therefore critical determinants of outcome [1–3]. The importance of early use of active antimicrobials is underscored by outcomes in drug-resistant TB meningitis, where delayed addition or substitution of effective agents is associated with increased mortality and worse neurologic sequelae compared with drug-sensitive disease [2–4]. However, TB meningitis is a paucibacillary syndrome, and there can often be a significant delay between initial diagnosis by smear or molecular testing and the availability of culture-based drug susceptibility results to guide therapy. In the absence of early susceptibility data, clinicians rely heavily on epidemiologic risk factors, such as prior TB treatment, known exposure to drug-resistant TB, or residence in regions with high multidrug-resistant TB prevalence, to assess the likelihood of resistance. In certain settings, the risk assessment should also include consideration of Mycobacterium bovis infection. M. bovis is a member of the M. tuberculosis complex and clinically, M. bovis causes presentations indistinguishable from M. tuberculosis infection, but it is intrinsically resistance to pyrazinamide [5]. This distinction is particularly consequential in TB meningitis, as pyrazinamide and isoniazid are the only first-line agents with reliable central nervous system (CNS) penetration within the standard rifampin, isoniazid, pyrazinamide, and ethambutol (RIPE) regimen. Here, we present a case of disseminated M. bovis infection complicated by meningitis and highlight the therapeutic challenges posed by intrinsic pyrazinamide resistance. In this patient, clinical improvement appeared to follow the early addition of adjunctive, CNS-penetrant antimicrobials to standard therapy prior to species confirmation. We use this case to review the available evidence supporting consideration of empiric adjunctive therapy in suspected TB meningitis occurring in regions where M. bovis is endemic. CASE A 43-year-old woman born in Mexico was brought to the hospital by her family after 2 days of progressively worsening altered mental status. Her medical history was notable for a prior oophorectomy 3 years earlier for a left adnexal mass causing pelvic discomfort, which had been complicated by bowel perforations, multiple resections, and persistent enterocutaneous fistulas. She reported a childhood history of consuming unpasteurized dairy products. Brain magnetic resonance imaging on admission revealed multiple rim-enhancing lesions (Figure 1) and a lumbar puncture demonstrated a lymphocytic pleocytosis (51 WBC w/75% lymphocytes). A cerebrospinal fluid (CSF) fluid sample tested by the Xpert MTB/RIF assay detected M. tuberculosis complex without rifampin resistance. Empiric RIPE therapy was initiated via nasogastric tube with rifampin 900 mg daily (13.7 mg/kg), isoniazid 300 mg daily (4.6 mg/kg), pyrazinamide 2000 mg daily (30.4 mg/kg), and ethambutol 1600 mg daily (24.4 mg/kg), along with adjunctive dexamethasone. She tested negative twice for HIV. Representative postcontrast T1-FLAIR images demonstrates rim-enhancing CNS lesions (white arrows). Her clinical course initially waxed and waned and was complicated by elevated intracranial pressure requiring placement of both an external ventricular drain and a lumbar drain. Approximately 1 month after initiation of RIPE therapy, she experienced clinical deterioration with recurrent low-grade fevers, worsening headaches, and progressive blurry vision. At that time, concern arose for suboptimal response to standard therapy, although microbiologic confirmation was unavailable due to persistent negatively cultures. Therapeutic drug monitoring demonstrated serum concentrations within expected ranges. Given the regional endemicity of M. bovis and concern for intrinsic pyrazinamide resistance, empiric levofloxacin was added to the regimen. Following this change, the patient demonstrated gradual clinical improvement. Two months after CSF culture collection, the isolate was confirmed to be susceptible to rifampin, isoniazid, and ethambutol, resistant to pyrazinamide, and subsequently speciated as M. bovis. After a 2-month hospitalization, she was discharged on rifampin, isoniazid, levofloxacin, and vitamin B6. The levofloxacin was continued because of slow clinical and radiographic response to therapy. When de-escalation to rifampin and isoniazid alone was considered to complete the treatment course, the patient elected through shared decision-making to continue all 3 agents. In the outpatient setting, her course was complicated by episodes of hepatotoxicity requiring brief interruptions in therapy. She ultimately completed a 15-month course of treatment. During therapy, her chronic enterocutaneous fistulas were noted to decrease in size. At the completion of treatment, she continued to experience residual blurry vision attributed to CNS involvement and was awaiting ophthalmologic evaluation. She additionally reported that her sister, who had also consumed unpasteurized dairy products, had been diagnosed with peritoneal tuberculosis within the past year. Retrospective review of pathology from the prior left adnexal mass demonstrated necrotizing granulomatous inflammation, AFB staining was negative. DISCUSSION Multiple cases of M. bovis meningitis have been reported in the literature, underscoring the substantial diagnostic and therapeutic challenges associated with this entity [6–9]. Our patient shared several established risk factors for M. bovis, including birth in an endemic region, consumption of unpasteurized dairy products, and extrapulmonary disease. In this context, the previously resected left adnexal mass most likely represented pelvic lymphadenopathy, a recognized manifestation of M. bovis infection [6–9]. As described in other reports, delayed recognition of these epidemiologic clues likely contributed to prolonged hospitalization, delayed optimization of therapy, and persistent neurological sequelae despite appropriate treatment [6–9]. The diagnostic delay of M. bovis infection is also partly explained by limitations of rapid molecular testing. The Xpert MTB/RIF assay, which identified M. tuberculosis complex in our patient, detects M. tuberculosis complex DNA and rpoB mutations associated with rifampin resistance, but it does not provide species-level differentiation within the complex, including distinction between M. tuberculosis and M. bovis. This distinction typically requires additional testing on cultured isolates through public health laboratories or specialized reference centers, using methods such as MALDI-TOF, whole genome or pncA sequencing. Recent disruptions in phenotypic pyrazinamide susceptibility testing have increased reliance on public health laboratories and specialized reference centers, further underscoring the diagnostic and treatment challenges posed by M. bovis infection. At present, there is no consensus on how to adapt treatment for M. bovis CNS infection. Although M. bovis accounts for a small proportion of tuberculosis cases in developed countries overall, the absence of specific guidance has important implications for US–Mexico border regions with large binational communities, such as San Diego, California, where M. bovis represents approximately 10% of tuberculosis cases, nearly 10-fold higher than the national average (Figure 2) [10]. The burden is even greater among patients with HIV-associated tuberculosis and in pediatric populations, where M. bovis may account for up to 30% of 50% of cases, respectively [11, 12]. In San Diego and other US–Mexico border communities, this disproportionate burden has been associated with Mexico-born, Hispanic ethnicity, and consumption of unpasteurized dairy products, particularly Mexican-style fresh cheeses such as queso fresco. Notably, M. bovis infection linked to queso fresco has been reported in nonborder US cities, including New York City and Baltimore. Given this broader epidemiology and evidence supporting adjunctive therapy in isoniazid-resistant tuberculosis meningitis, consideration of empiric adjunctive antimicrobials is particularly relevant when M. bovis meningitis is suspected. Cumulative incidence of human Mycobacterium bovis tuberculosis disease, by county, United States, 2006–2013. *National average incidence rate for 2006–2013 was 0.29 (Wilson 95% confidence interval, 0.26–0.29) M. bovis cases reported per 100 000 population. Reproduced from Scott C et al 2016 [10]. Standard treatment of tuberculosis meningitis mirrors pulmonary regimens, with RIPE commonly used despite marked differences in CNS penetration. Among first-line agents, only isoniazid and pyrazinamide achieve high CSF penetration (90%–95%) (Table 1) [13, 14]. Rifampin penetrates the CNS at approximately 5%–25% of plasma levels, while ethambutol has minimal CNS activity [13, 14]. These pharmacokinetic limitations have generated longstanding debate regarding the adequacy of RIPE therapy for CNS disease and whether alternative or intensified regimens should be considered, potentially guided by plasma and CSF therapeutic drug monitoring in selected cases, should be considered [14]. In M. bovis meningitis, these concerns are amplified, as intrinsic pyrazinamide resistance effectively removes one of the few reliably CNS-active first-line agents. Table 1.Usual Adult Dosing, Approximate CSF Penetration, and Potential Therapeutic Considerations for Selected Agents in Suspected Mycobacterium bovis MeningitisDrugUsual Adult Dose Used in TB/TBMApproximate CSF PenetrationPotential Role in Suspected M. bovis TBMKey LimitationsIsoniazid5 mg/kg daily, max 300 mg90%–95%Core first-line CNS-active drugHepatotoxicity, neuropathyPyrazinamide20–25 mg/kg daily95%–100%…Not active against M. bovisRifampin10 mg/kg daily; higher-dose strategies being studied in TBM5%–25%Core sterilizing drug; essential if susceptibleDrug–drug interactions, hepatotoxicityEthambutol15–20 mg/kg daily10%–50%Initial companion drug while susceptibilities pendingLimited CNS activity; optic neuritisLinezolid600 mg daily80%–100%Adjunctive CNS-active option; used in MDR TBCytopenias, neuropathy, serotonin syndrome riskLevofloxacin750 mg daily60%–80%Adjunctive CNS-active option when M. bovis suspectedQTc prolongation, tendinopathy, dysglycemia, neuropsychiatric effectsMoxifloxacin400 mg daily70%–80%Adjunctive CNS-active option when M. bovis suspectedQTc prolongation, tendinopathy, dysglycemia, neuropsychiatric effectsCycloserine10–15 mg/kg/d, commonly 250–500 mg twice daily60%–70%Adjunctive CNS-active option when M. bovis suspectedLimited clinical data; neuropsychiatric toxicityAbbreviations: CNS, central nervous system; CSF, cerebrospinal fluid; MDR-TB, multidrug-resistant tuberculosis; TB, tuberculosis; TBM, tuberculosis meningitis. Usual Adult Dosing, Approximate CSF Penetration, and Potential Therapeutic Considerations for Selected Agents in Suspected Mycobacterium bovis Meningitis Abbreviations: CNS, central nervous system; CSF, cerebrospinal fluid; MDR-TB, multidrug-resistant tuberculosis; TB, tuberculosis; TBM, tuberculosis meningitis. Recent international guidelines for tuberculosis meningitis discuss empiric strategies for isoniazid resistance; but, do not address intrinsic pyrazinamide resistance due to M. bovis [15]. This omission is clinically significant, as available data suggest that outcomes in M. bovis meningitis may be worse than those seen with M. tuberculosis, likely reflecting delays in diagnosis and effective therapy rather than intrinsic differences in virulence [16]. In the absence of specific guidance, clinicians practicing in regions where M. bovis is endemic have empirically added CNS-penetrant agents to RIPE therapy when M. bovis infection is suspected, even prior to culture confirmation. Fluoroquinolones, which were used in our patient, are attractive candidates given their excellent CSF penetration and established efficacy in drug-resistant tuberculosis [15]. In Vietnam, early fluoroquinolone initiation in combination with RIPE and corticosteroids was associated with improved outcomes, particularly when administered before coma onset [17]. An open-label phase 2 trial in Indonesia demonstrated a survival benefit with fluoroquinolones combined plus high-dose rifampin [17, 18]. In a randomized controlled trial conducted in Vietnam, the addition of fluoroquinolone did not show benefit in drug-sensitive TB, but a possible benefit was observed in the subgroup with isoniazid-resistant disease [19]. However, potential benefits must be weighed against known risks, including increased seizure rates reported in an Indian trial [20], as well as associations with aortic pathology, neuropsychiatric effects, dysglycemia, and peripheral neuropathy [21]. Linezolid represents another potential adjunctive agent. It achieves therapeutic CSF levels and is endorsed by the World Health Organization for multidrug-resistant tuberculosis [22]. Although data specific to TB meningitis remain limited, available studies are encouraging. In a Chinese cohort, addition of linezolid to RIPE and corticosteroids was associated with more rapid recovery of consciousness and defervescence [23], and a retrospective study of 36 pediatric TBM cases demonstrated improved outcomes, defined as survival without coma or major neurologic deficit [24]. Its use, however, is constrained by hematologic and neurologic toxicities. Other second-line agents have been explored cautiously. While bedaquiline has demonstrated benefit in drug-resistant pulmonary tuberculosis, current evidence does not support its use in TB meningitis. Animal models have failed to show additive benefit in CNS disease [25], and pharmacokinetic studies suggest poor CNS penetration [26]. In contrast, cycloserine has demonstrated adequate CSF penetration across multiple studies, suggesting potential utility in CNS infection, although clinical data remain limited and further investigation is warranted [27–30]. CONCLUSION Our case highlights the challenges of managing M. bovis meningitis, where loss of pyrazinamide activity removes one of the few first-line agents with reliable CNS penetration. In such scenarios, reliance on standard RIPE therapy alone may place patients at risk for delayed clinical response and increased morbidity and mortality while awaiting species identification and susceptibility data. Early recognition of epidemiologic risk factors, including birth in endemic regions, consumption of unpasteurized dairy products, and extrapulmonary disease, should prompt clinicians to consider M. bovis in the differential diagnosis of tuberculosis. When suspicion is high, our case suggests the early consideration of adjunctive CNS-active agents, although potential toxicities must be carefully weighed. Development of clinical guidance that explicitly addresses M. bovis CNS infection would help standardize care, reduce delays in effective therapy, and improve neurologic outcomes in binational disproportionately affected by this pathogen.