Predictive Modeling of Drug-Related Adverse Events with Real-World Data: A Case Study of Linezolid Hematologic Outcomes.

Electronic health records (EHRs) provide meaningful knowledge of drug-related adverse events (AEs) that are not captured in standard drug development and postmarketing surveillance. Using variables obtained from EHR data in the University of California San Francisco de-identified Clinical Data Warehouse, we aimed to evaluate the potential of machine learning to predict two hematological AEs, thrombocytopenia and anemia, in a cohort of patients treated with linezolid for 3 or more days. Features for model input were extracted at linezolid initiation (index), and outcomes were characterized from index to 14 days post-treatment. Random forest classification (RFC) was used for AE prediction, and reduced feature models were evaluated using cumulative importance (cImp) for feature selection. Grade 3+ thrombocytopenia and anemia occurred in 31% of 2,171 and 56% of 2,170 evaluable patients, respectively. Of the total 53 features, as few as 7 contributed at least 50% cImp, resulting in prediction accuracies of 70% or higher and area under the receiver operating characteristic curves of 0.886 for grade 3+ thrombocytopenia and 0.759 for grade 3+ anemia. Sensitivity analyses in strictly defined patient subgroups revealed similarly high predictive performance in full and reduced feature models. A logistic regression model with the same 50% cImp features showed similar predictive performance as RFC and good concordance with RFC probability predictions after isotonic calibration, adding interpretability. Collectively, this work demonstrates potential for machine learning prediction of AE risk in real-world patients using few variables regularly available in EHRs, which may aid in clinical decision making and/or monitoring.

Predicted Pharmacokinetic Interactions Between Hormonal Contraception and Single or Intermittently Dosed Rifampicin.

The scale-up of rifampicin-based prevention regimens is an essential part of the global leprosy strategy. Daily rifampicin may reduce the effectiveness of the oral contraceptive pill (OCP), but little is known about the effects of rifampicin at the less frequent dosing intervals used for leprosy prophylaxis. As many women of reproductive age rely on OCP for family planning, evaluating the interaction with less-than-daily rifampicin regimens would enhance the scalability and acceptability of leprosy prophylaxis. Using a semi-mechanistic pharmacokinetic model of rifampicin induction, we simulated predicted changes in OCP clearance when coadministered with varying rifampicin dosing schedules. Rifampicin given as a single dose (600 or 1200 mg) or 600 mg every 4 weeks was not predicted to result in a clinically relevant interaction with OCP, defined as a >25% increase in clearance. Simulations of daily rifampicin were predicted to increase OCP clearance within the range of observed changes previously reported in the literature. Therefore, our findings suggest that OCP efficacy will be maintained when coadministered with rifampicin-based leprosy prophylaxis regimens of 600 mg once, 1200 mg once, and 600 mg every 4 weeks. This work provides reassurance to stakeholders that leprosy prophylaxis can be used with OCP without any additional recommendations for contraception prevention.

QT Interval Prolongation with One or More QT-Prolonging Agents Used as Part of a Multidrug Regimen for Rifampicin-Resistant Tuberculosis Treatment: Findings from Two Pediatric Studies.

Rifampicin-resistant tuberculosis (RR-TB) involves treatment with many drugs that can prolong the QT interval; this risk may increase when multiple QT-prolonging drugs are used together. We assessed QT interval prolongation in children with RR-TB receiving one or more QT-prolonging drugs. Data were obtained from two prospective observational studies in Cape Town, South Africa. Electrocardiograms were performed before and after drug administration of clofazimine (CFZ), levofloxacin (LFX), moxifloxacin (MFX), bedaquiline (BDQ), and delamanid. The change in Fridericia-corrected QT (QTcF) was modeled. Drug and other covariate effects were quantified. A total of 88 children with a median (2.5th-to-97.5th range) age of 3.9 (0.5 to 15.7) years were included, of whom 55 (62.5%) were under 5 years of age. A QTcF interval of >450 ms was observed in 7 patient-visits: regimens were CFZ+MFX (n = 3), CFZ+BDQ+LFX (n = 2), CFZ alone (n = 1), and MFX alone (n = 1). There were no events with a QTcF interval of >500 ms. In a multivariate analysis, CFZ+MFX was associated with a 13.0-ms increase in change in QTcF (P < 0.001) and in maximum QTcF (P = 0.0166) compared to those when other MFX- or LFX-based regimens were used. In conclusion, we found a low risk of QTcF interval prolongation in children with RR-TB who received at least one QT-prolonging drug. Greater increases in maximum QTcF and ΔQTcF were observed when MFX and CFZ were used together. Future studies characterizing exposure-QTcF responses in children will be helpful to ensure safety with higher doses if required for effective treatment of RR-TB.

Pharmacokinetics and Safety of Bedaquiline in Human Immunodeficiency Virus (HIV)-Positive and Negative Older Children and Adolescents With Rifampicin-Resistant Tuberculosis.

BACKGROUND: Pharmacokinetic data for bedaquiline in children are limited. We described the pharmacokinetics and safety of bedaquiline in South African children and adolescents receiving treatment for multidrug/rifampicin-resistant tuberculosis (MDR/RR-TB) in routine care. METHODS: In this observational cohort study, children aged 6-17 years receiving bedaquiline at recommended doses as part of MDR/RR-TB treatment underwent semi-intensive pharmacokinetic sampling. Bedaquiline and the M2 metabolite plasma concentrations were quantified, and nonlinear mixed-effects modeling performed. Pediatric data were described using a pre-established model of bedaquiline pharmacokinetics in adults. The exposure reference was 187 µg ⋅ h/mL, the median weekly area under the curve (AUC) of adults at week 24 of treatment with bedaquiline. Safety was assessed through monthly clinical, blood and electrocardiogram monitoring, and treatment outcomes described. RESULTS: Fifteen children (3 human immunodeficiency virus [HIV]-positive) with median age 13.3 years (range 6.5-16.3) were included. A bedaquiline pharmacokinetic model was adapted to be allometrically scaled in clearance and volume, centered in the median child population weight. Bedaquiline bioavailability was 57% of that in adults. Overall bedaquiline exposures were below target, and AUC reference attainment was achieved in only 3 (20%) children. Ten children experienced 27 adverse events at least possibly related to bedaquiline; no adverse events led to bedaquiline withdrawal. Two adverse events (arthritis and arthralgia) were considered severe, and 2 children had mild QT interval corrected for heart rate using Fridericia's formula (QT) prolongation. CONCLUSIONS: The evaluated doses of bedaquiline in children ≥ 6 years of age were safe but achieved slightly lower plasma concentrations compared to adults receiving the recommended dose, possibly due to delayed food intake relative to bedaquiline administration.

Polypharmacy Is Associated With Falls in Women With and Without HIV.

BACKGROUND: Aging in people with HIV is associated with increased risk of developing synergistic conditions such as neurocognitive impairment, polypharmacy, and falls. We assessed associations between polypharmacy (use of 5 or more non-ART medications), use of neurocognitive adverse effects (NCAE) medications, and odds of falls in women with HIV (WWH) and without HIV (HIV-). METHODS: Self-reported falls and medication use data were contributed semiannually by 1872 (1315 WWH and 557 HIV-) Women's Interagency HIV Study participants between 2014 and 2016. Polypharmacy and NCAE medication use were evaluated separately and jointly in multivariable models to assess their independent contributions to single and multiple falls risk. RESULTS: The proportion of women who reported any fall was similar by HIV status (19%). WWH reported both greater polypharmacy (51% vs. 41%; P < 0.001) and NCAE medication use (44% vs. 37%; P = 0.01) than HIV- women. Polypharmacy conferred elevated odds of single fall [adjusted odds ratio (aOR) 1.67, 95% CI: 1.36 to 2.06; P < 0.001] and multiple falls (aOR 2.31, 95% CI: 1.83 to 2.93; P < 0.001); the results for NCAE medications and falls were similar. Both polypharmacy and number of NCAE medications remained strongly and independently associated with falls in multivariable models adjusted for HIV serostatus, study site, sociodemographics, clinical characteristics, and substance use. CONCLUSIONS: Polypharmacy and NCAE medication use were greater among WWH compared with HIV-, and both were independently and incrementally related to falls. Deprescribing and avoidance of medications with NCAEs may be an important consideration for reducing fall risk among WWH and sociodemographically similar women without HIV.

Emerging data on rifampicin pharmacokinetics and approaches to optimal dosing in children with tuberculosis.

INTRODUCTION: Despite its longstanding role in tuberculosis (TB) treatment, there continues to be emerging rifampicin research that has important implications for pediatric TB treatment and outstanding questions about its pharmacokinetics and optimal dose in children. AREAS COVERED: This review aims to summarize and discuss emerging data on the use of rifampicin for: 1) routine treatment of drug-susceptible TB; 2) special subpopulations such as children with malnutrition, HIV, or TB meningitis; 3) treatment shortening. We also highlight the implications of these new data for child-friendly rifampicin formulations and identify future research priorities. EXPERT OPINION: New data consistently show low rifampicin exposures across all pediatric populations with 10-20 mg/kg dosing. Although clinical outcomes in children are generally good, rifampicin dose optimization is needed, especially given a continued push to shorten treatment durations and for specific high-risk populations of children who have worse outcomes. A pooled analysis of existing data using applied pharmacometrics would answer many of the important questions remaining about rifampicin pharmacokinetics needed to optimize doses, especially in special populations. Targeted clinical studies in children with TB meningitis and treatment shortening with high-dose rifampicin are also priorities.

Model-Based Efficacy and Toxicity Comparisons of Moxifloxacin for Multidrug-Resistant Tuberculosis.

BACKGROUND: Moxifloxacin (MOX) is used as a first-choice drug to treat multidrug-resistant tuberculosis (MDR-TB); however, evidence-based dosing optimization should be strengthened by integrative analysis. The primary goal of this study was to evaluate MOX efficacy and toxicity using integrative model-based approaches in MDR-TB patients. METHODS: In total, 113 MDR-TB patients from 5 different clinical trials were analyzed for the development of a population pharmacokinetics (PK) model. A final population PK model was merged with a previously developed lung-lesion distribution and QT prolongation model. Monte Carlo simulation was used to calculate the probability target attainment value based on concentration. An area under the concentration-time curve (AUC)-based target was identified as the minimum inhibitory concentration (MIC) of MOX isolated from MDR-TB patients. RESULTS: The presence of human immunodeficiency virus (HIV) increased clearance by 32.7% and decreased the AUC by 27.4%, compared with HIV-negative MDR-TB patients. A daily dose of 800 mg or a 400-mg, twice-daily dose of MOX is expected to be effective in MDR-TB patients with an MIC of ≤0.25 µg/mL, regardless of PK differences resulting from the presence of HIV. The effect of MOX in HIV-positive MDR-TB patients tended to be decreased dramatically from 0.5 µg/mL, in contrast to the findings in HIV-negative patients. A regimen of twice-daily doses of 400 mg should be considered safer than an 800-mg once-daily dosing regimen, because of the narrow fluctuation of concentrations. CONCLUSIONS: Our results suggest that a 400-mg, twice-daily dose of MOX is an optimal dosing regimen for MDR-TB patients because it provides superior efficacy and safety.

Tuberculosis in children with severe acute malnutrition.

INTRODUCTION: With growing attention globally to the childhood tuberculosis epidemic after decades of neglect, and with the burden of severe acute malnutrition (SAM) remaining unacceptably high worldwide, the collision of these two diseases is an important focus for improving child health. AREAS COVERED: This review describes the clinical and public health implications of the interplay between tuberculosis and SAM, particularly for children under the age of five, and identifies priority areas for improved programmatic implementation and future research. We reviewed the literature on PubMed and other evidence known to the authors published until August 2021 relevant to this topic. EXPERT OPINION: To achieve the World Health Organization's goal of eliminating deaths from childhood tuberculosis and to improve the abysmal outcomes for children with SAM, further research is needed to 1) better understand the epidemiologic connections between child tuberculosis and SAM, 2) improve case finding of tuberculosis in children with SAM, 3) assess unique treatment considerations for tuberculosis when children also have SAM, and 4) ensure tuberculosis and SAM are strongly addressed in decentralized, integrated models of providing primary healthcare to children.

Moxifloxacin Pharmacokinetics, Cardiac Safety, and Dosing for the Treatment of Rifampicin-Resistant Tuberculosis in Children.

BACKGROUND: Moxifloxacin is a recommended drug for rifampin-resistant tuberculosis (RR-TB) treatment, but there is limited pediatric pharmacokinetic and safety data, especially in young children. We characterize moxifloxacin population pharmacokinetics and QT interval prolongation and evaluate optimal dosing in children with RR-TB. METHODS: Pharmacokinetic data were pooled from 2 observational studies in South African children with RR-TB routinely treated with oral moxifloxacin once daily. The population pharmacokinetics and Fridericia-corrected QT (QTcF)-interval prolongation were characterized in NONMEM. Pharmacokinetic simulations were performed to predict expected exposure and optimal weight-banded dosing. RESULTS: Eighty-five children contributed pharmacokinetic data (median [range] age of 4.6 [0.8-15] years); 16 (19%) were aged <2 years, and 8 (9%) were living with human immunodeficiency virus (HIV). The median (range) moxifloxacin dose on pharmacokinetic sampling days was 11 mg/kg (6.1 to 17). Apparent clearance was 6.95 L/h for a typical 16-kg child. Stunting and HIV increased apparent clearance. Crushed or suspended tablets had faster absorption. The median (range) maximum change in QTcF after moxifloxacin administration was 16.3 (-27.7 to 61.3) ms. No child had QTcF ≥500 ms. The concentration-QTcF relationship was nonlinear, with a maximum drug effect (Emax) of 8.80 ms (interindividual variability = 9.75 ms). Clofazimine use increased Emax by 3.3-fold. Model-based simulations of moxifloxacin pharmacokinetics predicted that current dosing recommendations are too low in children. CONCLUSIONS: Moxifloxacin doses above 10-15 mg/kg are likely required in young children to match adult exposures but require further safety assessment, especially when coadministered with other QT-prolonging agents.

Advancing evidence-based treatment of infectious diseases in children with real-world data: Opportunities and challenges.

There is an increased interest in utilizing real-world data (RWD) for pharmaceutical research and regulatory decision-making. The development and use of pediatric medicines could benefit greatly from real-world data studies given nearly half of drugs prescribed to children are "off-label", meaning there is a lack of pediatric-specific evidence from controlled trials, while there is an abundance of data from routine clinical practice. Currently, the use of real-world data, such as data from electronic health records, is lacking in pediatric research, especially within infectious diseases. Here, we discuss opportunities and challenges for real-world data to generate evidence on the optimal treatment and management of infectious diseases in children.

Comparative Efficacy of Rifapentine Alone and in Combination with Isoniazid for Latent Tuberculosis Infection: a Translational Pharmacokinetic-Pharmacodynamic Modeling Study.

Rifapentine has facilitated treatment shortening for latent tuberculosis infection (LTBI) in combination with isoniazid once weekly for 3 months (3HP) or daily for 1 month (1HP). Our objective was to determine the optimal rifapentine dose for a 6-week monotherapy regimen (6wP) and predict clinical efficacy. Rifapentine and isoniazid pharmacokinetics were simulated in mice and humans. Mouse lung CFU data were used to characterize exposure-response relationships of 1HP, 3HP, and 6wP and translated to predict clinical efficacy. A 600-mg daily dose for 6wP delivered greater cumulative rifapentine exposure than 1HP or 3HP. The maximum regimen effect (Emax) was 0.24 day-1. The regimen potencies, measured as the concentration at 50% of Emax (EC50), were estimated to be 2.12 mg/liter for 3HP, 3.72 mg/liter for 1HP, and 4.71 mg/liter for 6wP, suggesting that isoniazid contributes little to 1HP efficacy. Clinical translation predicted that 6wP reduces bacterial loads at a higher rate than 3HP and to a greater extent than 3HP and 1HP. 6wP (600 mg daily) is predicted to result in equal or better efficacy than 1HP and 3HP for LTBI treatment without the potential added toxicity of isoniazid. Results from ongoing and future clinical studies will be required to support these findings.

Rifapentine Population Pharmacokinetics and Dosing Recommendations for Latent Tuberculosis Infection.

Rationale: Rifapentine has been investigated at various doses, frequencies, and dosing algorithms, but clarity on the optimal dosing approach is lacking.Objectives: To characterize rifapentine population pharmacokinetics, including autoinduction, and determine optimal dosing strategies for short-course rifapentine-based regimens for latent tuberculosis infection.Methods: Rifapentine pharmacokinetic studies were identified though a systematic review of literature. Individual plasma concentrations were pooled, and nonlinear mixed-effects modeling was performed. A subset of data was reserved for external validation. Simulations were performed under various dosing conditions, including current weight-based methods; and alternative methods driven by identified covariates.Measurements and Main Results: We identified nine clinical studies with a total of 863 participants with pharmacokinetic data (n = 4,301 plasma samples). Rifapentine population pharmacokinetics were described successfully with a one-compartment distribution model. Autoinduction of clearance was driven by rifapentine plasma concentrations. The maximum effect was a 72% increase in clearance and was reached after 21 days. Drug bioavailability decreased by 27% with HIV infection, decreased by 28% with fasting, and increased by 49% with a high-fat meal. Body weight was not a clinically relevant predictor of clearance. Pharmacokinetic simulations showed that current weight-based dosing leads to lower exposures in individuals with low weight, which can be overcome with flat dosing. In HIV-positive patients, 30% higher doses are required to match drug exposure in HIV-negative patients.Conclusions: Weight-based dosing of rifapentine should be removed from clinical guidelines, and higher doses for HIV-positive patients should be considered to provide equivalent efficacy.

Alternative dosing guidelines to improve outcomes in childhood tuberculosis: a mathematical modelling study.

BACKGROUND: Malnourished and young children are particularly susceptible to severe forms of tuberculosis and poor treatment response. WHO dosing guidelines for drugs for tuberculosis treatment are based only on weight, which might lead to systematic underdosing and poor outcomes in these children. We aimed to assess and quantify the population effect of WHO guidelines for drug-susceptible tuberculosis in children in the 20 countries with the highest disease burden. METHODS: We used an integrated model that linked country-specific demographic data at the individual level from the 20 countries with the highest disease burden to pharmacokinetic, outcome, and epidemiological models. We estimated tuberculosis treatment outcomes in children younger than 5 years following WHO guidelines (children are dosed by weight bands corresponding to the number of fixed-dose combination tablets [75 mg rifampicin, 50 mg isoniazid, 150 mg pyrazinamide]) and two alternative dosing strategies: one based on a proposed algorithm that uses age, weight, and available formulations, in which underweight children would receive the same drug doses as would normal weight children of the same age; and another based on an individualised algorithm without dose limitations, in which derived doses results in target exposure attainment for the typical child. FINDINGS: We estimated that 57 234 (43%) of 133 302 children younger than 5 years who were treated for tuberculosis in 2017 were underdosed with WHO dosing and only 47% of children would reach the rifampicin exposure target. Underdosing and subtherapeutic exposures were more common among malnourished children than among age-matched healthy children. The proposed dosing approach improved estimated rifampicin target exposure attainment to 62% and equalised outcomes by nutritional status. An estimated third of unfavourable treatment outcomes might be resolved with this dosing strategy, saving the lives of a minimum of 2423 children in these countries annually. With individualised dosing approaches, almost all children could achieve adequate exposure for cure. INTERPRETATION: This work shows that a simple change in dosing procedure to include age and nutritional status, requiring no additional measurements or new drug formulations, is one approach to improve tuberculosis treatment outcomes in children, especially malnourished children who are at high risk of mortality. FUNDING: Eunice Kennedy Shriver National Institute of Child Health and Human Development and UK Medical Research Council.

Use of Nonantiretroviral Medications That May Impact Neurocognition: Patterns and Predictors in a Large, Long-Term HIV Cohort Study.

BACKGROUND: Neurocognitive impairment is a frequent and often disabling comorbidity of HIV infection. In addition to antiretroviral therapies, individuals with HIV infection may commonly use nonantiretroviral medications that are known to cause neurocognitive adverse effects (NC-AE). The contribution of NC-AE to neurocognitive impairment is rarely considered in the context of HIV and could explain part of the variability in neurocognitive performance among individuals with HIV. SETTING: Women's Interagency HIV Study, a prospective, multisite, observational study of US women with and without HIV. METHODS: After a literature review, 79 medications (excluding statins) with NC-AE were identified and reported by Women's Interagency HIV Study participants. We examined factors associated with self-reported use of these medications over a 10-year period. Generalized estimating equations for binary outcomes were used to assess sociodemographic, behavioral, and clinical characteristics associated with NC-AE medication use. RESULTS: Three thousand three hundred women (71% with HIV) and data from ∼42,000 visits were studied. HIV infection was associated with NC-AE medication use (odds ratio = 1.52; 95% confidence interval: 1.35 to 1.71). After adjustment for HIV infection status, other predictors of NC-AE medication use included having health insurance, elevated depressive symptoms, prior clinical AIDS, noninjection recreational drug use, and an annual household income of <$12,000 (Ps < 0.004). NC-AE medication use was less likely among women who drank 1-7 or 8-12 alcoholic drinks/week (vs. abstaining) (P < 0.04). CONCLUSIONS: HIV infection was associated with NC-AE medication use, which may influence determinations of HIV-associated neurocognitive impairment. Providers should consider the impact of NC-AE medications when evaluating patients with HIV and concurrent neurocognitive symptoms.

Cognitive Burden of Common Non-antiretroviral Medications in HIV-Infected Women.

OBJECTIVE: The aging HIV population has increased comorbidity burden and consequently non-antiretroviral medication utilization. Many non-antiretroviral medications have known neurocognitive-adverse effects ("NC-AE medications"). We assessed the cognitive effects of NC-AE medications in HIV+ and HIV- women. METHODS: One thousand five hundred fifty-eight participants (1037 HIV+; mean age 46) from the Women's Interagency HIV Study completed a neuropsychological test battery between 2009 and 2011. The total number of NC-AE medications and subgroups (eg, anticholinergics) were calculated based on self-report. Generalized linear models for non-normal data were used to examine the cognitive burden of medications and factors that exacerbate these effects. RESULTS: HIV+ women reported taking more NC-AE medications vs. HIV- women (P < 0.05). NC-AE medication use altogether was not associated with cognitive performance. However, among NC-AE medication subgroups, anticholinergic-acting medications, but not opioids or anxiolytics/anticonvulsants, were negatively associated with performance. HIV status moderated the association between these NC-AE medication subgroups and performance (P's < 0.05). HIV-serostatus differences (HIV- < HIV+) in global, learning, fluency, and motor function were greatest among women taking >1 anticholinergic medications. HIV-serostatus differences in performance on learning and psychomotor speed were also greatest among women taking 1 or more anxiolytics/anticonvulsants and 1 or more opioids, respectively. CONCLUSIONS: HIV+ women have increased cognitive vulnerabilities to anticholinergic, anxiolytic/anticonvulsant, and opioid medications. Potential synergy between these medications and HIV may explain some HIV-related cognitive impairments. It may be important clinically to consider these specific types of medications as a contributor to impaired cognitive performance in HIV+ women and assess the cost/benefit of treatment dosage for underlying conditions.

Interaction of N-acetylcysteine and cysteine in human plasma.

N-acetyl-L-cysteine (NAC), a well-known antioxidant, has been successfully used as adjuvant therapy for late-stage childhood cerebral adrenoleukodystrophy (c-ALD); however, the mechanisms of NAC action are poorly understood. Previous research indicates that NAC serves as a precursor to L-cysteine (Cys), the rate-limiting substrate in the biosynthesis of glutathione (GSH), a potent, endogenous antioxidant. We hypothesized that NAC acts by liberating protein-bound Cys in plasma in an NAC concentration-dependent manner, which increases unbound Cys available for GSH biosynthesis. Human plasma was incubated for 1 h with varying, clinically relevant concentrations of NAC (0-1000 µg/mL). The effect of this interaction over time was evaluated by incubating plasma for 5-90 min with 100 µg/mL NAC. Unbound and bound Cys and NAC were separated by ultrafiltration, and concentrations were measured using high-performance liquid chromatography-mass spectrometry. Significant increases in unbound Cys were observed with increasing NAC concentrations. Also, Cys plasma protein binding decreased from 85% (10 µg/mL NAC) to approximately 0% (1000 µg/mL). Total endogenous Cys was 66% unbound at 5 min after incubation. These results demonstrate that NAC liberates endogenous, protein-bound Cys in human plasma at clinically relevant NAC concentrations. A greater understanding of NAC actions will aid in the optimization of NAC therapy including its use in c-ALD.