Browse the corpus

Trial of High-Dose Oral Rifampin in Adults with Tuberculous Meningitis. BACKGROUND: Tuberculous meningitis is often lethal, and many survivors have disabilities despite antimicrobial treatment and adjunctive glucocorticoid therapy. Standard-dose rifampin has limited central nervous system penetration. Whether high-dose rifampin could improve survival outcomes is unknown. METHODS: We performed a double-blind, randomized, placebo-controlled clinical trial involving adults with tuberculous meningitis in Indonesia, South Africa, and Uganda. We assigned persons with and those without human immunodeficiency virus (HIV) coinfection to receive standard daily isoniazid, rifampin (at a dose of 10 mg per kilogram of body weight), ethambutol, and pyrazinamide plus either additional rifampin (for a cumulative dose of 35 mg per kilogram; high-dose group) or matched placebo (standard-dose group) for 8 weeks; participants in both groups received standard therapy for the remainder of the 9-to-12-month treatment course. The primary outcome was 6-month mortality. RESULTS: A total of 499 participants were included in the intention-to-treat population (249 randomly assigned to the high-dose group and 250 to the standard-dose group), of whom 304 (60.9%) were persons living with HIV and 428 (85.8%) had definite or probable tuberculous meningitis. During 6 months of follow-up, 109 participants (Kaplan-Meier estimate, 44.6%) in the high-dose group and 100 participants (Kaplan-Meier estimate, 40.7%) in the standard-dose group died (hazard ratio, 1.17; 95% confidence interval, 0.89 to 1.54; P = 0.25). Among the participants who died within 6 months, the median time to death was 13 days (interquartile range, 4 to 39) in the high-dose group and 24 days (interquartile range, 6 to 56) in the standard-dose group. Drug-induced liver injury occurred in 8.0% of the participants in the high-dose group and in 4.4% of those in the standard-dose group, but no deaths from drug-induced liver injury occurred. CONCLUSIONS: Among persons with tuberculous meningitis, no evidence of beneficial effect from high-dose rifampin was observed, and the potential for a harmful effect cannot be ruled out. (Funded by the U.K. Medical Research Council and others; ISRCTN Registry number, ISRCTN15668391.).

The “High Dose Oral Rifampicin to Improve Outcomes from Adult Tuberculous Meningitis” (HARVEST) trial was a double-blind, placebo-controlled, randomized phase III trial with parallel group design, conducted in Indonesia, South Africa, and Uganda (ISRCTN15668391).11 Research ethics committees and relevant regulatory approvals occurred at all sites.

n to Improve Outcomes from Adult Tuberculous Meningitis” (HARVEST) trial was a double-blind, placebo-controlled, randomized phase III trial with parallel group design, conducted in Indonesia, South Africa, and Uganda (ISRCTN15668391).11 Research ethics committees and relevant regulatory approvals occurred at all sites. The trial aimed to enroll 500 eligible participants aged ≥18 years with tuberculous meningitis at 9 hospitals. Eligibility criteria included microbiologically-confirmed (positive CSF Xpert MTB/Rif Ultra (Cepheid, CA),12,13,14 mycobacterial growth indicator tube culture (Becton Dickinson, NJ), or smear microscopy) or suspected tuberculous meningitis based on abnormal CSF parameters and clinical findings with antituberculous treatment planned. Written informed consent was obtained from all participants, or surrogates for people lacking capacity to consent. Exclusion criteria included hypersensitivity, receipt of >5 doses of antituberculous therapy, confirmed untreated neuroinfection other than tuberculosis, HIV protease inhibitor use, corticosteroid contraindication, pregnancy or breastfeeding, or inability to follow-up. As meningitis is a medical emergency and similar to other meningitis trials,16 participants could be randomized immediately and withdrawn within 72 hours for baseline late exclusion criteria (alanine transaminase >5 times the upper limit of normal or glomerular filtration rate <30 ml/min/1.73m2). Participants could be withdrawn within 15 days for rifampicin resistance or confirmed alternative neuroinfection with decision to stop antituberculous therapy. Full details of study conduct can be found in the protocol at nejm.org.

ase >5 times the upper limit of normal or glomerular filtration rate <30 ml/min/1.73m2). Participants could be withdrawn within 15 days for rifampicin resistance or confirmed alternative neuroinfection with decision to stop antituberculous therapy. Full details of study conduct can be found in the protocol at nejm.org. Participants were classified by the uniform case definition as definite, probable, or possible tuberculous meningitis.15 We randomized participants in a 1:1 ratio to either high-dose or standard-dose oral rifampicin, using permutated blocks with sizes of two or four. Randomization was stratified by trial centers, HIV status, and tuberculous meningitis severity (Medical Research Council grade one versus grades two or three). Both participants and investigators were blinded to treatment allocation, with identical-appearing intervention and placebo tablets.

with sizes of two or four. Randomization was stratified by trial centers, HIV status, and tuberculous meningitis severity (Medical Research Council grade one versus grades two or three). Both participants and investigators were blinded to treatment allocation, with identical-appearing intervention and placebo tablets. Participants in the intervention arm received high-dose oral rifampicin of ~35 mg per kilogram daily, administered for 8 weeks. This dose was achieved by providing extra rifampicin tablets (900 mg if ≤38 kilograms, otherwise 1200 mg) in addition to the standard tuberculosis fixed-dose combination tablets that included rifampicin (~10 mg per kilogram), isoniazid, pyrazinamide, and ethambutol.17 Participants in the control arm received the same standard tuberculosis fixed-dose combination tablets that included rifampicin (~10 mg per kilogram) with the same number of placebo tablets as the intervention group. Participants received adjunctive corticosteroids as per international guidelines,17 with clinician discretion (Supplementary Methods). The primary outcome was 6-month survival. Secondary outcomes included 12-month survival, improvement in functional status as measured by the modified Rankin scale at week 24; Liverpool outcome score at 2-weeks and 6-months; neurocognitive performance at 2 and 12 months;18,19 hospitalization duration; incidence of all-cause treatment discontinuation for ≥5 days in the first 8-weeks; re-hospitalization due to neurological decline; and incidence of grade ≥3 or serious adverse events, including hepatotoxicity.

e score at 2-weeks and 6-months; neurocognitive performance at 2 and 12 months;18,19 hospitalization duration; incidence of all-cause treatment discontinuation for ≥5 days in the first 8-weeks; re-hospitalization due to neurological decline; and incidence of grade ≥3 or serious adverse events, including hepatotoxicity. Following randomization, participants were monitored daily until hospital discharge. Following hospital discharge, clinical and laboratory assessments occurred at 2, 4, and 8-weeks, with additional follow-up visits at weeks-12, 18, 24, 36, and 52. An independent data safety monitoring board reviewed interim data at least annually. Assuming 50% survival in the control group, an overall sample size of 500 participants (250 per group) provided 80% power with a two-sided 0.05 significance level to detect a hazard ratio for death of 0.68 (approximately 13% absolute survival improvement) while accounting for ≤5% loss-to-follow up.

Following randomization, participants were monitored daily until hospital discharge. Following hospital discharge, clinical and laboratory assessments occurred at 2, 4, and 8-weeks, with additional follow-up visits at weeks-12, 18, 24, 36, and 52. An independent data safety monitoring board reviewed interim data at least annually. Assuming 50% survival in the control group, an overall sample size of 500 participants (250 per group) provided 80% power with a two-sided 0.05 significance level to detect a hazard ratio for death of 0.68 (approximately 13% absolute survival improvement) while accounting for ≤5% loss-to-follow up. The primary analysis was based on the intention-to-treat population, which was pre-specified to include all randomized participants who were alive at the time of randomization, excluding study-sponsored withdrawals including late exclusion criteria. Survival distributions and 6-month mortality were estimated via Kaplan-Meier estimators. The treatment effect of high-dose rifampicin vs standard-dose rifampicin was estimated via hazard ratio from an unadjusted Cox regression. Adjusted analyses were conducted with pre-specified covariates, including randomization strata, age, gender, baseline Glasgow coma scale (<15 or 15) and tuberculous meningitis case definition. Additional statistical details are included in Supplementary Methods. Predefined subgroup analyses examined treatment effect across key clinical and demographic subgroups: baseline Glasgow coma scale score, severity grade, HIV status, antiretroviral status, and clinical trial site. Subgroup analyses were not adjusted for multiplicity.

stical details are included in Supplementary Methods. Predefined subgroup analyses examined treatment effect across key clinical and demographic subgroups: baseline Glasgow coma scale score, severity grade, HIV status, antiretroviral status, and clinical trial site. Subgroup analyses were not adjusted for multiplicity. For secondary endpoints, the modified Rankin Scale at week-24 was compared between the treatment arms using a cumulative logistic regression with proportional odds assumption. For safety endpoints, the total number of events as well as the total number and percentage of participants reporting at least one event were summarized by arms. Fine-Gray model accounted for death as a competing risk for rehospitalization.

The trial aimed to enroll 500 eligible participants aged ≥18 years with tuberculous meningitis at 9 hospitals. Eligibility criteria included microbiologically-confirmed (positive CSF Xpert MTB/Rif Ultra (Cepheid, CA),12,13,14 mycobacterial growth indicator tube culture (Becton Dickinson, NJ), or smear microscopy) or suspected tuberculous meningitis based on abnormal CSF parameters and clinical findings with antituberculous treatment planned. Written informed consent was obtained from all participants, or surrogates for people lacking capacity to consent. Exclusion criteria included hypersensitivity, receipt of >5 doses of antituberculous therapy, confirmed untreated neuroinfection other than tuberculosis, HIV protease inhibitor use, corticosteroid contraindication, pregnancy or breastfeeding, or inability to follow-up. As meningitis is a medical emergency and similar to other meningitis trials,16 participants could be randomized immediately and withdrawn within 72 hours for baseline late exclusion criteria (alanine transaminase >5 times the upper limit of normal or glomerular filtration rate <30 ml/min/1.73m2). Participants could be withdrawn within 15 days for rifampicin resistance or confirmed alternative neuroinfection with decision to stop antituberculous therapy. Full details of study conduct can be found in the protocol at nejm.org. Participants were classified by the uniform case definition as definite, probable, or possible tuberculous meningitis.15

We randomized participants in a 1:1 ratio to either high-dose or standard-dose oral rifampicin, using permutated blocks with sizes of two or four. Randomization was stratified by trial centers, HIV status, and tuberculous meningitis severity (Medical Research Council grade one versus grades two or three). Both participants and investigators were blinded to treatment allocation, with identical-appearing intervention and placebo tablets.

Participants in the intervention arm received high-dose oral rifampicin of ~35 mg per kilogram daily, administered for 8 weeks. This dose was achieved by providing extra rifampicin tablets (900 mg if ≤38 kilograms, otherwise 1200 mg) in addition to the standard tuberculosis fixed-dose combination tablets that included rifampicin (~10 mg per kilogram), isoniazid, pyrazinamide, and ethambutol.17 Participants in the control arm received the same standard tuberculosis fixed-dose combination tablets that included rifampicin (~10 mg per kilogram) with the same number of placebo tablets as the intervention group. Participants received adjunctive corticosteroids as per international guidelines,17 with clinician discretion (Supplementary Methods).

The primary outcome was 6-month survival. Secondary outcomes included 12-month survival, improvement in functional status as measured by the modified Rankin scale at week 24; Liverpool outcome score at 2-weeks and 6-months; neurocognitive performance at 2 and 12 months;18,19 hospitalization duration; incidence of all-cause treatment discontinuation for ≥5 days in the first 8-weeks; re-hospitalization due to neurological decline; and incidence of grade ≥3 or serious adverse events, including hepatotoxicity.

Following randomization, participants were monitored daily until hospital discharge. Following hospital discharge, clinical and laboratory assessments occurred at 2, 4, and 8-weeks, with additional follow-up visits at weeks-12, 18, 24, 36, and 52. An independent data safety monitoring board reviewed interim data at least annually.

Assuming 50% survival in the control group, an overall sample size of 500 participants (250 per group) provided 80% power with a two-sided 0.05 significance level to detect a hazard ratio for death of 0.68 (approximately 13% absolute survival improvement) while accounting for ≤5% loss-to-follow up. The primary analysis was based on the intention-to-treat population, which was pre-specified to include all randomized participants who were alive at the time of randomization, excluding study-sponsored withdrawals including late exclusion criteria. Survival distributions and 6-month mortality were estimated via Kaplan-Meier estimators. The treatment effect of high-dose rifampicin vs standard-dose rifampicin was estimated via hazard ratio from an unadjusted Cox regression. Adjusted analyses were conducted with pre-specified covariates, including randomization strata, age, gender, baseline Glasgow coma scale (<15 or 15) and tuberculous meningitis case definition. Additional statistical details are included in Supplementary Methods. Predefined subgroup analyses examined treatment effect across key clinical and demographic subgroups: baseline Glasgow coma scale score, severity grade, HIV status, antiretroviral status, and clinical trial site. Subgroup analyses were not adjusted for multiplicity.

From March 12, 2021, to July 31, 2024, we randomly assigned 529 adult patients with tuberculous meningitis to receive either high-dose or standard-dose oral rifampicin. Of these participants, 30 (5.6%) were excluded by baseline late exclusion criteria, leaving 499 participants (249 in the high-dose group and 250 in the standard-dose group) in the intention-to-treat population (Figure 1). Study visit attendance through month 6 of follow-up was 96%, and adherence to study treatment was good, with 6% missing ≥2 days of study drug or matched-placebo through 8 weeks. Participant characteristics were similar between the trial groups (Table 1). The median age was 37 years (interquartile range, 28 to 45), and 222 of 499 (44%) were female. Overall, 306 (61%) participants had Medical Research Council grade 2 severity, and 218 (43.7%) had microbiologically-confirmed definite tuberculous meningitis. Overall, 304 of 499 (61%) were persons living with HIV; 125 of 304 (41%) were receiving antiretroviral therapy, and 115 of 226 (51%) had CD4 counts <100 cells/mm3. At randomization, 348 of 499 (70%) had received antituberculous treatment for a median of 3 (interquartile range, 1 to 4) days. Overall, 473 of 499 (95%) were prescribed corticosteroids.

persons living with HIV; 125 of 304 (41%) were receiving antiretroviral therapy, and 115 of 226 (51%) had CD4 counts <100 cells/mm3. At randomization, 348 of 499 (70%) had received antituberculous treatment for a median of 3 (interquartile range, 1 to 4) days. Overall, 473 of 499 (95%) were prescribed corticosteroids. Death occurred in 109 of 249 (Kaplan-Meier estimate: 44.6%) participants within six months in the high-dose rifampicin group and in 100 of 250 (Kaplan-Meier estimate: 40.7%) in the standard-dose group (Hazard Ratio, 1.17; 95% confidence interval [CI], 0.89 to 1.54; P=0.25; Figure 2). Similar results were observed in the adjusted analysis with pre-specified baseline covariates or using per-protocol population. Among participants who died within 6 months, the median time to death in the high-dose group was 13 (interquartile range, 4 to 39) days and 24.5 (interquartile range, 6 to 56) days in the standard-dose group. None of the subgroup analyses showed evidence of beneficial effects of high-dose rifampin. (Figure 3). We observed higher risk for mortality in the high-dose group compared to the standard-dose group among those receiving antiretroviral therapy at the time of presentation (Hazard Ratio, 2.01; 95% CI, 1.07 to 3.78). Among participants with CSF white cells <5 cells/mm3, we observed higher risk for mortality in the high-dose group compared to the standard-dose group (Hazard Ratio, 2.01; 95% CI, 1.14 to 3.54) whereas the hazard ratio among those with ≥5 cells/mm3 was 0.98 (95% CI, 0.72 to 1.35).

rd Ratio, 2.01; 95% CI, 1.07 to 3.78). Among participants with CSF white cells <5 cells/mm3, we observed higher risk for mortality in the high-dose group compared to the standard-dose group (Hazard Ratio, 2.01; 95% CI, 1.14 to 3.54) whereas the hazard ratio among those with ≥5 cells/mm3 was 0.98 (95% CI, 0.72 to 1.35). No evidence of benefit was observed for any secondary endpoints. First, in assessing functional and neurological disability, we observed no difference in the modified Rankin score (Odds Ratio, 0.80; 95% CI, 0.58 to 1.11) or Liverpool outcome score at 24-weeks (Table S4). Among participants who had baseline GCS <15, 48 of 169 (28%) had normalized their mental status by day 20 in the high-dose rifampin group, and 65 of 163 (39.9%) in the standard dose group achieved this. The median of Liverpool outcome score was 53 (interquartile rage: 1 to 72) among 204 participants in the high-dose rifampicin group and was higher in the standard-dose group with median 63 (interquartile range 41 to 73) among 202 participants at Week 2 (Table S5B). 12-month mortality did not differ between the high-dose (Kaplan-Meier estimate: 47.0%) and standard-dose groups (Kaplan-Meier estimate: 43.7%) (Hazard Ratio, 1.14; 95% CI 0.88 to 1.49, Figure S12).

her in the standard-dose group with median 63 (interquartile range 41 to 73) among 202 participants at Week 2 (Table S5B). 12-month mortality did not differ between the high-dose (Kaplan-Meier estimate: 47.0%) and standard-dose groups (Kaplan-Meier estimate: 43.7%) (Hazard Ratio, 1.14; 95% CI 0.88 to 1.49, Figure S12). There was no evidence for difference in safety between study arms. Study drug discontinuation for more than 5 days, mostly due to hepatotoxicity, was similar between the two groups (Table S6). Overall, 217 serious adverse events occurred among 185 participants. There was no evidence of difference in any of the clinical grade 3-5 adverse events between groups, except for aspiration pneumonia which occurred in 13 of 249 (5.2%) high-dose participants compared with 4 of 250 (1.6%) standard-dose participants. No difference was observed in grade 3-4 laboratory abnormalities between groups, except total bilirubin for which elevations were more frequent in high-dose participants (24 of 249 [9.6%] compared with standard-dose participants (9 of 250 [3.6%]) (Table S7). Drug-induced liver injury occurred in 20 of 249 (8.0%) participants in the high-dose group and 11 of 250 (4.4%) in the standard-dose group. The number of participants with an alanine transaminase measurement ≥5 times the upper limit of normal was similar in the two groups (Table S8). Blinded study drug was interrupted for >5 days due to drug-induced liver injury in 6 of 249 (2.4%) high-dose participants vs 4 of 250 (1.6%) standard-dose participants. No participants died of drug-induced liver injury.

ransaminase measurement ≥5 times the upper limit of normal was similar in the two groups (Table S8). Blinded study drug was interrupted for >5 days due to drug-induced liver injury in 6 of 249 (2.4%) high-dose participants vs 4 of 250 (1.6%) standard-dose participants. No participants died of drug-induced liver injury. The initial duration of hospitalization did not differ. Within the first six months, there were 101 re-hospitalizations, with stroke, other infections, and tuberculosis paradoxical reactions or immune reconstitution inflammatory syndrome as the main reasons (Table S9).

From March 12, 2021, to July 31, 2024, we randomly assigned 529 adult patients with tuberculous meningitis to receive either high-dose or standard-dose oral rifampicin. Of these participants, 30 (5.6%) were excluded by baseline late exclusion criteria, leaving 499 participants (249 in the high-dose group and 250 in the standard-dose group) in the intention-to-treat population (Figure 1). Study visit attendance through month 6 of follow-up was 96%, and adherence to study treatment was good, with 6% missing ≥2 days of study drug or matched-placebo through 8 weeks.

Participant characteristics were similar between the trial groups (Table 1). The median age was 37 years (interquartile range, 28 to 45), and 222 of 499 (44%) were female. Overall, 306 (61%) participants had Medical Research Council grade 2 severity, and 218 (43.7%) had microbiologically-confirmed definite tuberculous meningitis. Overall, 304 of 499 (61%) were persons living with HIV; 125 of 304 (41%) were receiving antiretroviral therapy, and 115 of 226 (51%) had CD4 counts <100 cells/mm3. At randomization, 348 of 499 (70%) had received antituberculous treatment for a median of 3 (interquartile range, 1 to 4) days. Overall, 473 of 499 (95%) were prescribed corticosteroids.

Death occurred in 109 of 249 (Kaplan-Meier estimate: 44.6%) participants within six months in the high-dose rifampicin group and in 100 of 250 (Kaplan-Meier estimate: 40.7%) in the standard-dose group (Hazard Ratio, 1.17; 95% confidence interval [CI], 0.89 to 1.54; P=0.25; Figure 2). Similar results were observed in the adjusted analysis with pre-specified baseline covariates or using per-protocol population. Among participants who died within 6 months, the median time to death in the high-dose group was 13 (interquartile range, 4 to 39) days and 24.5 (interquartile range, 6 to 56) days in the standard-dose group.

None of the subgroup analyses showed evidence of beneficial effects of high-dose rifampin. (Figure 3). We observed higher risk for mortality in the high-dose group compared to the standard-dose group among those receiving antiretroviral therapy at the time of presentation (Hazard Ratio, 2.01; 95% CI, 1.07 to 3.78). Among participants with CSF white cells <5 cells/mm3, we observed higher risk for mortality in the high-dose group compared to the standard-dose group (Hazard Ratio, 2.01; 95% CI, 1.14 to 3.54) whereas the hazard ratio among those with ≥5 cells/mm3 was 0.98 (95% CI, 0.72 to 1.35).

No evidence of benefit was observed for any secondary endpoints. First, in assessing functional and neurological disability, we observed no difference in the modified Rankin score (Odds Ratio, 0.80; 95% CI, 0.58 to 1.11) or Liverpool outcome score at 24-weeks (Table S4). Among participants who had baseline GCS <15, 48 of 169 (28%) had normalized their mental status by day 20 in the high-dose rifampin group, and 65 of 163 (39.9%) in the standard dose group achieved this. The median of Liverpool outcome score was 53 (interquartile rage: 1 to 72) among 204 participants in the high-dose rifampicin group and was higher in the standard-dose group with median 63 (interquartile range 41 to 73) among 202 participants at Week 2 (Table S5B). 12-month mortality did not differ between the high-dose (Kaplan-Meier estimate: 47.0%) and standard-dose groups (Kaplan-Meier estimate: 43.7%) (Hazard Ratio, 1.14; 95% CI 0.88 to 1.49, Figure S12).

This randomized double-blind trial demonstrated no benefit of high-dose rifampicin as prescribed among adults with tuberculous meningitis with respect to 6-month survival or any prespecified secondary endpoints or subgroups. Although no excess drug toxicity occurred, higher early mortality and slower normalization of mental status were observed in the intervention arm. Two a priori subgroups had two-fold higher hazards of death with high-dose rifampicin as compared to standard-dose rifampicin therapy; those with CSF white cells <5 per mm3 and those with HIV receiving antiretroviral therapy at baseline.

y and slower normalization of mental status were observed in the intervention arm. Two a priori subgroups had two-fold higher hazards of death with high-dose rifampicin as compared to standard-dose rifampicin therapy; those with CSF white cells <5 per mm3 and those with HIV receiving antiretroviral therapy at baseline. The lack of benefit could have several possible explanations. First, lower exposure of adjunctive corticosteroids in the high-dose arm may have occurred due to greater induction of hepatic metabolism. One study has shown that rifampicin dosed at 40 mg per kilogram reduces exposure of midazolam, used as a probe drug for hepatic cytochrome p450 CYP3A which also metabolizes dexamethasone and prednisone, by 38% compared to standard-dose rifampicin.20 Measurement of corticosteroids in blood from study participants may shed further light on this. In support of this hypothesis, we observed high-dose rifampicin had a greater absolute 6-month mortality in HIV-negative participants (5.9% lower survival than standard-dose rifampicin) than in those with HIV (2.2% lower survival than standard-dose rifampicin). This is consistent with corticosteroids having a greater impact on tuberculous meningitis outcomes in patients without HIV (~7.2% absolute mortality reduction) than in those with HIV (~4.9% reduction).3,21 Thus, increased corticosteroid metabolism could have had a greater detrimental impact in participants without HIV.3 But, steroid metabolism alone would not explain the lack of survival benefit in persons with HIV.

in patients without HIV (~7.2% absolute mortality reduction) than in those with HIV (~4.9% reduction).3,21 Thus, increased corticosteroid metabolism could have had a greater detrimental impact in participants without HIV.3 But, steroid metabolism alone would not explain the lack of survival benefit in persons with HIV. Second, more rapid killing of mycobacteria with higher-dose rifampicin might theoretically lead to a stronger, and detrimental, immunologic reaction. This could potentially explain the lower survival with higher-dose rifampicin in those on antiretroviral therapy compared to those not receiving antiretroviral therapy (18.9% lower survival versus 8.4% higher survival), as those on antiretroviral therapy would have the ability to generate a more robust immunologic response. A dysregulated response in the central nervous system can be detrimental, as previously shown for earlier antiretroviral therapy after cryptococcal meningitis.22,23 Somewhat surprisingly, participants with CSF <5 white blood cells/mm3 at baseline, which might represent a dysfunctional or anergic immune response,24 had 22.7% higher mortality with high-dose rifampicin than standard-dose. The majority of participants with CSF white cells <5 cells/mm3 were living with HIV, but 30% were HIV-negative. Our ongoing immunological and multi-omics studies may shed light on this.

might represent a dysfunctional or anergic immune response,24 had 22.7% higher mortality with high-dose rifampicin than standard-dose. The majority of participants with CSF white cells <5 cells/mm3 were living with HIV, but 30% were HIV-negative. Our ongoing immunological and multi-omics studies may shed light on this. The rifampicin dosing choice is unlikely to explain the lack of effect. A previous phase III trial combining a more modest increased rifampicin dose of 15 mg per kilogram with levofloxacin showed no benefit in Vietnam.2 Subsequent, model-based pharmacokinetic-pharmacodynamic meta-analyses identified daily dosing of >30 mg per kilogram as optimal.5 High-dose rifampicin is being evaluated in four out of five ongoing tuberculous meningitis clinical trials, involving >1,800 adults and children.25 We believe our choice of testing 35 mg per kilogram dosing was reasonable and is close to the 40 mg per kilogram maximal tolerable dose.26 No excess toxicity associated with high-dose rifampicin occurred except for non-fatal drug-induced hepatotoxicity and transient benign hyperbilirubinemia likely caused by decreased bile efflux of conjugated bilirubin with high-dose rifampicin.26 Difference in hepatic transaminase elevation was not observed.

ble dose.26 No excess toxicity associated with high-dose rifampicin occurred except for non-fatal drug-induced hepatotoxicity and transient benign hyperbilirubinemia likely caused by decreased bile efflux of conjugated bilirubin with high-dose rifampicin.26 Difference in hepatic transaminase elevation was not observed. Our trial has several strengths. We examined a blinded, single intervention, enrolling adults with and without HIV across the spectrum of disease and in three countries on two continents. As such, our results are highly generalizable. Retention of participants, adherence to study drugs, and ascertainment of all trial endpoints were almost complete. The major limitation common to all tuberculous meningitis studies is the potential for misdiagnosis. Microbiologic confirmation of tuberculous meningitis is difficult, and any interventional therapy would not benefit patients who did not truly have tuberculosis. However, randomization should distribute any misdiagnoses equally between study arms. Overall, 80% had definite or probable tuberculous meningitis, by uniform case definition,15 and even those with definite tuberculous meningitis did not have a favorable trend for survival (Hazard Ratio 0.99). An additional limitation to interpreting our results is the limited follow-up CSF sampling over time. Future trials should prioritize longitudinal CSF sampling for biomarkers to help study possible underlying mechanisms for the differential effects related to CSF cell counts and antiretroviral therapy status we observed in this trial.

mitation to interpreting our results is the limited follow-up CSF sampling over time. Future trials should prioritize longitudinal CSF sampling for biomarkers to help study possible underlying mechanisms for the differential effects related to CSF cell counts and antiretroviral therapy status we observed in this trial. Our results raise the question of how the dismal outcomes can be improved. Rifampicin-free regimens consisting of potent antimicrobials with better brain penetration should be studied.27 Caution is warranted from any preclinical model, however, given that well-controlled murine model studies had shown clear benefit of high-dose rifampicin,28 underlining the complexity of this disease in humans and the importance of randomized clinical trials. A second strategy might be host-directed strategies to target immunopathology.1,29 Adjunctive aspirin is being evaluated in several trials.25 Tumor necrosis factor antagonist infliximab has benefit for tuberculous meningitis patients with corticosteroid-refractory inflammatory events later in the course of antituberculous treatment,30 but as deaths and irreversible neurological damage mostly occur in the first few weeks, trials should evaluate infliximab at treatment initiation. Finally, interventions that shorten the complex and lengthy patient pathways31 or lead to earlier diagnosis will likely improve outcomes.32,33

f antituberculous treatment,30 but as deaths and irreversible neurological damage mostly occur in the first few weeks, trials should evaluate infliximab at treatment initiation. Finally, interventions that shorten the complex and lengthy patient pathways31 or lead to earlier diagnosis will likely improve outcomes.32,33 Although high-dose rifampicin did not benefit adults with tuberculous meningitis, this trial demonstrates that high-quality multinational trials can be conducted successfully, despite the complexities of care. Our findings should stimulate more research for this neglected disease. Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.