Dick Menzies

Background: People receiving immunosuppressive medications are at increased risk of tuberculosis. Tuberculosis preventive treatment (TPT) can reduce risk; however, its safety among these populations is uncertain. We conducted a meta-analysis to evaluate adverse events (AEs) associated with TPT among people receiving or about to initiate immunosuppressive medications. Methods: We searched Medline, Cochrane, Health Star, and Embase (1952–1 May 2025), for studies reporting AEs during TPT among people receiving or about to initiate immunosuppressive medications. We used random-effects generalized linear mixed models with Hartung–Knapp 95% confidence intervals (CIs) to estimate our primary outcome: pooled proportion of AE-related discontinuation, stratified by TPT regimen. Results: We included 27 studies (3184 participants). Populations included biologics/steroids recipients and solid organ/hematopoietic transplant candidates/recipients. For the primary outcome (25 studies, 2802 participants), TPT regimens included 6- to 12-month mono-isoniazid regimens (1511 participants), 4-month mono-rifamycin regimens (228 participants), 3-month isoniazid-rifamycin combination regimens (990 participants), and 6- to 9-month fluoroquinolone-based regimens (73 participants). Only mono-isoniazid regimens were provided while receiving immunosuppressive medications. Pooled AE-related discontinuation was 5.3% (95% CI, 3.1%–8.9%; I2 = 65.4%) for mono-isoniazid regimens, 4.4% (95% CI, .1%–75.4%; I2 = 86.3%) for mono-rifamycin regimens, 2.0% (1.0%–4.1%; I2 = 0.0%) for isoniazid-rifamycin combination regimens, and 6.9% (95% CI, 0–92.3%; I2 = 65.4%) for fluoroquinolone-based regimens. Proportions were similar when stratified by biologics/steroids and transplant populations. Conclusions: Overall, AE-related TPT discontinuation was similar to that in the broader TPT literature. AE-related discontinuation of TPT was lowest with rifamycin-containing regimens, but these were only provided prior to initiating immunosuppressive medications. Data on the safety of non-isoniazid regimens among people receiving immunosuppressive medications are needed.

SUMMARY: BACKGROUND: We evaluated the safety, acceptability, and cost-effectiveness of all published studies on fluoroquinolone (FQ)-based TB preventive treatment (TPT) regimens in all settings and populations. METHODS: We searched for studies published between February 2014 and September 2023 using PubMed, Embase, Turning Research Into Practice (TRIP), Global Health Library, and the Cochrane Central Register of Controlled Trials. Meta-analysis was conducted using random effects models. RESULTS: Twenty-two observational and modelling studies evaluating FQ-based TPT were included, assessing outcomes of safety (n = 16), acceptability (n = 9), and cost-effectiveness (n = 2). Only 2% (95% confidence interval [CI]: 0%–4%) of children and adolescents and 8% (95% CI: 2%–14%) of adults discontinued FQ monotherapy due to drug-related adverse events (AEs). Compared to FQ monotherapy, rates of drug-related AEs of any severity were non-significantly higher with multidrug FQ regimens without pyrazinamide (PZA). Pooled discontinuation rates were 50% (95% CI: 34%–67%) among children and 57% (95% CI: 16%–98%) among adult contacts treated with PZA-containing FQ-based regimens. Pooled acceptance rates exceeded 80% among all groups. FQ-based TPT was cost-effective to avert TB-related deaths and loss of quality-adjusted life-years. CONCLUSION: These results complement the evidence from two recent randomised trials. FQ monotherapy is a safe, acceptable, and cost-effective TPT regimen for contacts exposed to multidrug-resistant or rifampicin-resistant TB disease.

Background: The World Health Organization (WHO) currently endorses several commercial interferon-gamma release assays (IGRAs); policy updates are needed given the development of new tests. We conducted an updated systematic review and meta-analysis to compare sensitivity, specificity, and agreement of newly commercialized IGRAs with WHO-endorsed IGRAs. Methods: Databases (Medline, Embase, Web of Science, Cochrane Central Register of Controlled Trials, and International Clinical Trials Registry Platform) were searched for studies published from 1 February 2021 to 30 September 2024, then again up to 16 December 2025. We included cross-sectional, cohort, and case-control studies and assessed risk of bias using QUADAS-2. The primary outcomes were pooled differences in sensitivity, specificity, and agreement between the new IGRAs and WHO-endorsed IGRAs. Results: From 3039 records screened, 39 reports (36 published) describing 8 new IGRAs met eligibility: STANDARD E TB-Feron, LIAISON XL Analyzer (chemiluminescence immunoassay [CLIA]), Ichroma IGRA-TB, STANDARD F TB-Feron, VIDAS TB-IGRA, ASACIR.TB, Wantai CLIA, and AdvanSureI3 TB-IGRA. STANDARD E TB-Feron had nonsignificantly lower sensitivity (−2.8% [95% confidence interval {CI}, −11.8% to 6.2%]) and specificity (−3.6% [95% CI, −10.0% to 2.7%]), with kappa of 0.802 (95% CI, .729–.874]), compared to QFT-Plus. For LIAISON XL Analyzer, the specificity was 0.1% higher (95% CI, −6.2% to 6.4%) than QFT-Plus/QFT-GIT, with a pooled kappa of 0.846 (95% CI, .814–.878). Conclusions: Two of the 8 new tests reviewed showed promising accuracy and high agreement with WHO-endorsed IGRAs. The overall certainty of evidence remains low due to risk of bias and imprecision. High-quality, independent studies in diverse populations and settings are needed to inform future WHO recommendations for tuberculosis infection tests. Clinical Trials Registration. PROSPERO (CRD42024544404).

Background: High-dose rifampicin could potentially shorten tuberculosis preventive therapy (TPT) and improve outcomes. We aimed to characterize population pharmacokinetics of standard- and high-dose rifampicin for TPT among individuals with tuberculosis infection. Methods: Intensive and sparse pharmacokinetic substudies were conducted in Indonesia, Canada, and Vietnam within the 2R2 randomized trial, which compared 2 months of high-dose rifampicin at 20 mg/kg/day (2R20) or 30 mg/kg/day (2R30) with 4 months of standard-dose rifampicin at 10 mg/kg/day (4R10) in adults and adolescents aged ≥ 10 years. Venous blood samples were collected after 4 weeks of treatment. Rifampicin pharmacokinetics were analyzed using nonlinear mixed-effects modeling in NONMEM. Results: Among 1368 trial participants, 440 were included in model development (51 intensive and 389 sparse sampling), with 191 (43%) assigned to 4R10, 159 (36%) to 2R20, and 90 (20%) to 2R30. A 1-compartment model with saturable hepatic extraction and transit-compartment absorption best described rifampicin pharmacokinetics. All disposition parameters were allometrically scaled using fat-free mass. Country-specific differences, particularly variation in drug formulation, were associated with lower bioavailability in Canada {−21.8% (95% confidence interval [CI] −27.9 to −18.0%)} and Vietnam (−12.3% [95% CI −17.7 to −7.9%]) compared with Indonesia. The 24-hour area under the concentration–time curve increased more than proportionally with dose and was higher across treatment arms in Indonesia, followed by Vietnam and Canada. Conclusions: High-dose rifampicin for TPT resulted in greater-than-proportional increases in exposure due to nonlinear clearance at higher doses. Substantial between-country variability in exposure was observed, which may have been due to multiple factors, including differences in country-specific formulations, fat-free mass, and unmeasured confounders.

Background: Cost is an important consideration when implementing tuberculosis preventive treatment regimens (TPT). We used data from a phase 2b randomized trial of TPT to estimate overall cost and key drivers of costs for two TPT regimens. Methods: We did a post-hoc analysis of 915 participants aged ≥10 years who were randomized 1:1 to 2 rifampin-based regimens: a four-month daily regimen at 10 mg/kg (4R10) and a 2-month daily regimen at 20 mg/kg (2R20; 461 participants). We collected country-specific costs for medications, evaluations, and medical follow-ups from the three participating countries (Canada, Indonesia, and Viet Nam), and converted all costs to 2024 Canadian dollars. We report the overall costs of each regimen and cost drivers. Results: Overall, 454 participants received 4R10 and 461 participants received 2R20. We found no difference in the cost of 2R20 versus 4R10, with a cost ratio of 0.93 (95% CI: .79–1.07); this was consistent in analyses limited to only those who completed treatment and stratified by country. Costs for medications and the baseline visit accounted for 68%, 49%, and 55% of all costs in Canada, Indonesia, and Viet Nam, respectively. Corresponding costs of routine follow-up visits accounted for approximately 26%, 45%, and 42% of all costs. In all countries, a minority of costs (<10%) were due to additional visits or evaluations not specified in the protocol. Conclusions: Most costs associated with TPT are due to medications and the baseline treatment initiation visit. TPT regimens requiring fewer follow-up visits may reduce overall cost, but the magnitude of this reduction varies by country.

Background: The World Health Organization recommends evaluation of all household contacts (HHC) of index tuberculosis (TB) patients for TB disease (TBD) and TB infection (TBI). Tests to identify TBI and TBD are preferred but can be skipped in persons living with HIV and children <5 years. There is equipoise on the need for these tests in other HHC. Methods: We conducted a superiority, open label cluster-randomized trial in Benin and Brazil to compare three strategies to evaluate HHC aged 5–50 of persons newly diagnosed with drug susceptible pulmonary TBD: Standard: tuberculin skin testing (TST) for TBI and if positive, chest X-ray (CXR) to rule out TBD; rapid molecular test (RMT): same as Standard, except CXR replaced by an RMT; and No-TST: CXR for all but no TST. Randomization was computer-generated and stratified by country, in blocks of variable length. The primary outcome was TB preventive therapy (TPT) initiation among HHC considered eligible (positive TST, if done, and no evidence of TBD on CXR or RMT). Secondary outcomes were: completion of investigations to detect TBI and TBD, detection of TBD, TPT completion, severe adverse events, and societal costs. Results: Among 1,589 participating HHC enrolled from 29 January 2020, to 30 November 2022, 474 were randomized to the standard, 583 to the RMT, and 532 to the no-TST strategies; all were included in the analyses. Of 848 HHC considered eligible for TPT, 802 (94.6%) initiated TPT, with no difference between strategies (95%, 94%, and 95% for the standard, RMT, and no-TST strategies, respectively). Of the secondary outcomes, protocol-mandated investigations to detect TBI and exclude possible TBD were completed for 93.4% overall, with slight differences between arms (93%, 95%, and 93% for the standard, RMT, and no-TST strategies, respectively). Adverse events resulting in discontinuation of TPT occurred in 3 (0.4%) participants in total (with 1, 0, and 2 events among participants in the Standard, RMT, and no-TST arms, respectively). The proportion completing TPT was similar with Standard and RMT strategies but was 13% lower (95% confidence interval: 3% to 23% lower) with the No-TST strategy. Societal costs per HHC completing investigations were $61 ($56–$65) with the standard strategy, compared to $52 ($49–$55) with the RMT strategy and $74 ($72–$77) with the no-TST strategy. Conclusion: This randomized trial provides high-quality evidence that TST followed by selected use of CXR or an RMT to exclude disease can achieve high rates of TPT initiation at reasonable costs. A limitation of the trial is the potential study effect, which may have affected adherence by providers and HHCs. RMT could replace CXR in the management of HHC in resource limited settings. Registration: clinicaltrials.gov NCT04528823

Platform trials evaluate multiple experimental treatments against a common control group (and/or against each other), which often reduces the trial duration and sample size. Bayesian platform designs offer several practical advantages, including the flexible addition or removal of experimental arms using posterior probabilities and the incorporation of prior/external information. Regulatory agencies require that the operating characteristics of Bayesian designs are assessed by estimating the sampling distribution of posterior probabilities via Monte Carlo simulation. It is computationally intensive to repeat this simulation process for all design configurations considered, particularly for platform trials with complex interim decision procedures. In this paper, we propose an efficient method to assess operating characteristics and determine sample sizes as well as other design parameters for Bayesian platform trials. We prove theoretical results that allow us to model the joint sampling distribution of posterior probabilities across multiple endpoints and trial stages using simulations conducted at only two sample sizes. This work is motivated by design complexities in the SSTARLET trial, an ongoing Bayesian adaptive platform trial for tuberculosis preventive therapies (ClinicalTrials.gov ID: NCT06498414). Our proposed design method is not only computationally efficient but also capable of accommodating intricate, real-world trial constraints like those encountered in SSTARLET.

Background: Recent studies suggest that triple-dose rifampin (TDR; ≥30 mg/kg/d) may be unsafe. We updated a systematic review to investigate the safety and efficacy of TDR. Methods: We searched Embase, MEDLINE, Cochrane CENTRAL, Cochrane Database for Systematic Reviews, and clinicaltrials.gov for randomized, controlled trials from 1 January 1965 to 10 February 2024 that compared standard-dose rifampin (SDR) with TDR and/or double-dose rifampin (DDR) in human tuberculosis treatment. The primary outcome was pooled incidence rate ratio (IRR) of severe adverse events (SevAEs) between participants who received TDR and those who received SDR. Pooled relative risk (RR) of death was a key secondary outcome. Meta-analysis was performed using the inverse variance method. Heterogeneity was assessed using I2, and bias was assessed using Cochrane Risk of Bias 2. The protocol was prospectively registered (osf.io/kfn5a). Results: Of the 11 315 articles identified, 17 met inclusion criteria, enrolling 2313 SDR participants (17 studies), 2238 receiving DDR (12 studies), and 1199 receiving TDR (11 studies). Six studies had a high risk of bias. There was an increase in pooled SevAEs among participants who received TDR compared with those who received SDR (IRR, 1.48; 95% confidence interval [CI], 1.12–1.96; I2, 23%), driven by hepatic events (IRR, 1.96; 95% CI, 1.21–3.18). Death did not differ between participants who received TDR and SDR (RR, 1.19; 95% CI, .71–1.99). One limitation is that only 2 included studies were blinded. Conclusions: Regimens that used TDR were associated with an increase in SevAEs, raising concerns regarding safety of TDR regimens.