Evaluation of critical parameters in the hollow-fibre system for tuberculosis: A case study of moxifloxacin.

AIMS: The hollow­fibre system for tuberculosis (HFS­TB) is a preclinical model qualified by the European Medicines Agency to underpin the anti­TB drug development process. It can mimic in vivo pharmacokinetic (PK)­pharmacodynamic (PD) attributes of selected antimicrobials, which could feed into in silico models to inform the design of clinical trials. However, historical data and published protocols are insufficient and omit key information to allow experiments to be reproducible. Therefore, in this work, we aim to optimize and standardize various HFS­TB operational procedures. METHODS: First, we characterized bacterial growth dynamics with different types of hollow­fibre cartridges, Mycobacterium tuberculosis strains and media. Second, we mimicked a moxifloxacin PK profile within hollow­fibre cartridges, in order to check drug­fibres compatibility. Lastly, we mimicked the moxifloxacin total plasma PK profile in human after once daily oral dose of 400 mg to assess PK­PD after different sampling methods, strains, cartridge size and bacterial adaptation periods before drug infusion into the system. RESULTS: We found that final bacterial load inside the HFS­TB was contingent on the studied variables. Besides, we demonstrated that drug­fibres compatibility tests are critical preliminary HFS­TB assays, which need to be properly reported. Lastly, we uncovered that the sampling method and bacterial adaptation period before drug infusion significantly impact actual experimental conclusions. CONCLUSION: Our data contribute to the necessary standardization of HFS­TB experiments, draw attention to multiple aspects of this preclinical model that should be considered when reporting novel results and warn about critical parameters in the HFS­TB currently overlooked.

Risk Factors for COVID-19 and Respiratory Tract Infections during the Coronavirus Pandemic.

(1) Background: Some individuals are more susceptible to developing respiratory tract infections (RTIs) or coronavirus disease (COVID-19) than others. The aim of this work was to identify risk factors for symptomatic RTIs including COVID-19 and symptomatic COVID-19 during the coronavirus pandemic by using infection incidence, participant baseline, and regional COVID-19 burden data. (2) Methods: Data from a prospective study of 1000 frontline healthcare workers randomized to Bacillus Calmette-Guérin vaccination or placebo, and followed for one year, was analyzed. Parametric time-to-event analysis was performed to identify the risk factors associated with (a) non-specific symptomatic respiratory tract infections including COVID-19 (RTIs+COVID-19) and (b) symptomatic RTIs confirmed as COVID-19 using a polymerase chain reaction or antigen test (COVID-19). (3) Results: Job description of doctor or nurse (median hazard ratio [HR] 1.541 and 95% confidence interval [CI] 1.299-1.822), the reported COVID-19 burden (median HR 1.361 and 95% CI 1.260-1.469 for 1.4 COVID-19 cases per 10,000 capita), or a BMI > 30 kg/m2 (median HR 1.238 and 95% CI 1.132-1.336 for BMI of 35.4 kg/m2) increased the probability of RTIs+COVID-19, while positive SARS-CoV-2 serology at enrollment (median HR 0.583 and 95% CI 0.449-0.764) had the opposite effect. The reported COVID-19 burden (median HR 2.372 and 95% CI 2.116-2.662 for 1.4 COVID-19 cases per 10,000 capita) and a job description of doctor or nurse (median HR 1.679 and 95% CI 1.253-2.256) increased the probability of developing COVID-19, while smoking (median HR 0.428 and 95% CI 0.284-0.648) and positive SARS-CoV-2 serology at enrollment (median HR 0.076 and 95% CI 0.026-0.212) decreased it. (4) Conclusions: Nurses and doctors with obesity had the highest probability of developing RTIs including COVID-19. Non-smoking nurses and doctors had the highest probability of developing COVID-19 specifically. The reported COVID-19 burden increased the event probability, while positive SARS-CoV-2 IgG serology at enrollment decreased the probability of RTIs including COVID-19, and COVID-19 specifically.

Subcutaneous Marzeptacog Alfa (Activated) for On-Demand Treatment of Bleeding Events in Subjects With Hemophilia A or B With Inhibitors.

Marzeptacog alfa (MarzAA) is under development for subcutaneous treatment of episodic bleeds in patients with hemophilia A/B and was studied in a phase III trial evaluating MarzAA compared with standard-of-care (SoC) for on-demand use. The work presented here aimed to evaluate MarzAA and SoC treatment of bleeding events on a standardized four-point efficacy scale (poor, fair, good, and excellent). Two continuous-time Markov modeling approaches were explored; a four-state model analyzing all four categories of bleeding improvement and a two-state model analyzing a binarized outcome (treatment failure (poor/fair), and treatment success (good/excellent)). Different covariates impacting improvement of bleeding episodes as well as a putative relationship between MarzAA exposure and improvement of bleeding episodes were evaluated. In the final four-state model, higher baseline diastolic blood pressure and higher age (> 33 years of age) were found to negatively and positively impact improvement of bleeding condition, respectively. Bleeding events occurring in knees and ankles were found to improve faster than bleeding events at other locations. The covariate effects had most impact on early treatment success (≤ 3 hours) whereas at later timepoints (> 12 hours), treatment success was similar for all patients indicating that these covariates might be clinically relevant for early treatment response. A statistically significant relationship between MarzAA zero-order absorption and improvement of bleedings (P < 0.05) were identified albeit with low precision. No statistically significant difference in treatment response between MarzAA and intravenous SoC was identified, indicating the potential of MarzAA for treatment of episodic bleeding events with a favorable subcutaneous administration route.

Derivation and Clinical Utility of Safety Targets for Linezolid-Related Adverse Events in Drug-Resistant Tuberculosis Treatment.

Long-term usage of linezolid can result in adverse events such as peripheral neuropathy, anemia and thrombocytopenia. Therapeutic drug monitoring data from 75 drug-resistant tuberculosis patients treated with linezolid were analyzed using a time-to-event (TTE) approach for peripheral neuropathy and anemia and indirect response modelling for thrombocytopenia. Different time-varying linezolid pharmacokinetic exposure indices (AUC0-24h,ss, Cav, Cmax and Cmin) and patient characteristics were investigated as risk factors. A treatment duration shorter than 3 months was considered dropout and was modelled using a TTE approach. An exposure-response relationship between linezolid Cmin and both peripheral neuropathy and anemia was found. The exposure index which best described the development of thrombocytopenia was AUC0-24h. The final TTE dropout model indicated an association between linezolid Cmin and dropout. New safety targets for each adverse event were proposed which can be used for individualized linezolid dosing. According to the model predictions at 6 months of treatment, a Cmin of 0.11 mg/L and 1.4 mg/L should not be exceeded to keep the cumulative probability to develop anemia and peripheral neuropathy below 20%. The AUC0-24h should be below 111 h·mg/L or 270 h·mg/L to prevent thrombocytopenia and severe thrombocytopenia, respectively. A clinical utility assessment showed that the currently recommended dose of 600 mg once daily is safer compared to a 300 mg BID dosing strategy considering all four safety endpoints.

Reproducibility in pharmacometrics applied in a phase III trial of BCG-vaccination for COVID-19.

Large clinical trials often generate complex and large datasets which need to be presented frequently throughout the trial for interim analysis or to inform a data safety monitory board (DSMB). In addition, reliable and traceability are required to ensure reproducibility in pharmacometric data analysis. A reproducible pharmacometric analysis workflow was developed during a large clinical trial involving 1000 participants over one year testing Bacillus Calmette-Guérin (BCG) (re)vaccination in coronavirus disease 2019 (COVID-19) morbidity and mortality in frontline health care workers. The workflow was designed to review data iteratively during the trial, compile frequent reports to the DSMB, and prepare for rapid pharmacometric analysis. Clinical trial datasets (n = 41) were transferred iteratively throughout the trial for review. An RMarkdown based pharmacometric processing script was written to automatically generate reports for evaluation by the DSMB. Reports were compiled, reviewed, and sent to the DSMB on average three days after the data cut-off, reflecting the trial progress in real-time. The script was also utilized to prepare for the trial pharmacometric analyses. The same source data was used to create analysis datasets in NONMEM format and to support model script development. The primary endpoint analysis was completed three days after data lock and unblinding, and the secondary endpoint analyses two weeks later. The constructive collaboration between clinical, data management, and pharmacometric teams enabled this efficient, timely, and reproducible pharmacometrics workflow.

Seasonal influence on respiratory tract infection severity including COVID-19 quantified through Markov Chain modeling.

Respiratory tract infections (RTIs) are a burden to global health, but their characterization is complicated by the influence of seasonality on incidence and severity. The Re-BCG-CoV-19 trial (NCT04379336) assessed BCG (re)vaccination for protection from coronavirus disease 2019 (COVID-19) and recorded 958 RTIs in 574 individuals followed over 1 year. We characterized the probability of RTI occurrence and severity using a Markov model with health scores (HSs) for four states of symptom severity. Covariate analysis on the transition probability between HSs explored the influence of demographics, medical history, severe acute respiratory syndrome-coronavirus 2 (SARS-CoV-2), or influenza vaccinations, which became available during the trial, SARS-CoV-2 serology, and epidemiology-informed seasonal influence of infection pressure represented as regional COVID-19 pandemic waves, as well as BCG (re)vaccination. The infection pressure reflecting the pandemic waves increased the risk of RTI symptom development, whereas the presence of SARS-CoV-2 antibodies protected against RTI symptom development and increased the probability of symptom relief. Higher probability of symptom relief was also found in participants with African ethnicity and with male biological gender. SARS-CoV-2 or influenza vaccination reduced the probability of transitioning from mild to healthy symptoms. Model diagnostics over calendar-time indicated that COVID-19 cases were under-reported during the first wave by an estimated 2.76-fold. This trial was performed during the initial phase of the COVID-19 pandemic in South Africa and the results reflect that situation. Using this unique clinical dataset of prospectively studied RTIs over the course of 1 year, our Markov Chain model was able to capture risk factors for RTI development and severity, including epidemiology-informed infection pressure.

Model-Informed Support of Dose Selection for Prophylactic Treatment with Dalcinonacog Alfa in Adult and Paediatric Hemophilia B Patients.

INTRODUCTION: Dalcinonacog alfa (DalcA), a novel subcutaneously administered recombinant human factor IX (FIX) variant is being developed for adult and paediatric patients with hemophilia B (HB). DalcA has been shown to raise FIX to clinically meaningful levels in adults with HB. This work aimed to support dosing regimen selection in adults and perform first-in-paediatric dose extrapolations using a model-based pharmacokinetic (PK) approach. METHODS: A population PK model was built using adult data from two clinical trials (NCT03186677, NCT03995784). With allometry in the model, clinical trial simulations were performed to study alternative dosing regimens in adults and children. Steady-state trough levels and the time-to-reach target were derived to inform dose selection. RESULTS: Almost 90% of the adults were predicted to achieve desirable FIX levels, i.e. 10% FIX activity, following daily 100 IU/kg dosing, with 90% of the subjects reaching target within 1.6-7.1 days. No every-other-day regimen met the target. A dose of 125 IU/kg resulted in adequate FIX levels down to 6 years, whereas a 150 IU/kg dose was needed below 6 down to 2 years of age. For subjects down to 6 years that did not reach target with 125 IU/kg, a dose escalation to 150 IU/kg was appropriate. The children below 6 to 2 years were shown to need a dose escalation to 200 IU/kg if 150 IU/kg given daily was insufficient. CONCLUSION: This study supported the adult dose selection for DalcA in the presence of sparse data and enabled first-in-paediatric dose selection to achieve FIX levels that reduce risk of spontaneous bleeds.

Standards for model-based early bactericidal activity analysis and sample size determination in tuberculosis drug development.

Background: A critical step in tuberculosis (TB) drug development is the Phase 2a early bactericidal activity (EBA) study which informs if a new drug or treatment has short-term activity in humans. The aim of this work was to present a standardized pharmacometric model-based early bactericidal activity analysis workflow and determine sample sizes needed to detect early bactericidal activity or a difference between treatment arms. Methods: Seven different steps were identified and developed for a standardized pharmacometric model-based early bactericidal activity analysis approach. Non-linear mixed effects modeling was applied and different scenarios were explored for the sample size calculations. The sample sizes needed to detect early bactericidal activity given different TTP slopes and associated variability was assessed. In addition, the sample sizes needed to detect effect differences between two treatments given the impact of different TTP slopes, variability in TTP slope and effect differences were evaluated. Results: The presented early bactericidal activity analysis approach incorporates estimate of early bactericidal activity with uncertainty through the model-based estimate of TTP slope, variability in TTP slope, impact of covariates and pharmacokinetics on drug efficacy. Further it allows for treatment comparison or dose optimization in Phase 2a. To detect early bactericidal activity with 80% power and at a 5% significance level, 13 and 8 participants/arm were required for a treatment with a TTP-EBA0-14 as low as 11 h when accounting for variability in pharmacokinetics and when variability in TTP slope was 104% [coefficient of variation (CV)] and 22%, respectively. Higher sample sizes are required for smaller early bactericidal activity and when pharmacokinetics is not accounted for. Based on sample size determinations to detect a difference between two groups, TTP slope, variability in TTP slope and effect difference between two treatment arms needs to be considered. Conclusion: In conclusion, a robust standardized pharmacometric model-based EBA analysis approach was established in close collaboration between microbiologists, clinicians and pharmacometricians. The work illustrates the importance of accounting for covariates and drug exposure in EBA analysis in order to increase the power of detecting early bactericidal activity for a single treatment arm as well as differences in EBA between treatments arms in Phase 2a trials of TB drug development.

Model-informed pediatric dose selection of marzeptacog alfa (activated): An exposure matching strategy.

Marzeptacog alfa (activated) (MarzAA) is an activated recombinant human rFVII variant intended for subcutaneous (s.c.) administration to treat or prevent bleeding in individuals with hemophilia A (HA) or B (HB) with inhibitors, and other rare bleeding disorders. The s.c. administration provides benefits over i.v. injections. The objective of the study was to support the first-in-pediatric dose selection for s.c. MarzAA to treat episodic bleeding episodes in children up through 11 years in a registrational phase III trial. Assuming the same exposure-response relationship as in adults, an exposure matching strategy was used with a population pharmacokinetics model. A sensitivity analysis evaluating the impact of doubling in absorption rate and age-dependent allometric exponents on dose selection was performed. Subsequently, the probability of trial success, defined as the number of successful trials for a given pediatric dose divided by the number of simulated trials (n = 1000) was studied. A successful trial was defined as outcome where four, three, or two out of 24 pediatric subjects per trial were allowed to fall outside the adult exposures after s.c. administration of 60 µg/kg. A dose of 60 µg/kg in children with HA/HB was supported by the clinical trial simulations to match exposures in adults. The sensitivity analyses further supported selection of the 60 µg/kg dose level in all age groups. Moreover, the probability of trial success evaluations given a plausible design confirmed the potential of a 60 µg/kg dose level. Taken together, this work demonstrates the utility of model-informed drug development and could be helpful for other pediatric development programs for rare diseases.

Model-based analysis of bactericidal activity and a new dosing strategy for optimised-dose rifampicin.

BACKGROUND: Higher doses of rifampicin for tuberculosis have been shown to improve early bactericidal activity (EBA) and at the same time increase the intolerability due to high exposure at the beginning of treatment. To support dose optimisation of rifampicin, this study investigated new and innovative staggered dosing of rifampicin using clinical trial simulations to minimise tolerability problems and still achieve good efficacy. METHODS: Rifampicin population pharmacokinetics and time-to-positivity models were applied to data from patients receiving 14 days of daily 10-50 mg/kg rifampicin to characterise the exposure-response relationship. Furthermore, clinical trial simulations of rifampicin exposure were performed following four different staggered dosing scenarios. The simulated exposure after 35 mg/kg was used as a relative comparison for efficacy. Tolerability was derived from a previous model-based analysis relating exposure at day 7 and the probability of having adverse events. RESULTS: The linear relationship between rifampicin exposure and bacterial killing rate in sputum indicated that the maximum rifampicin EBA was not reached at doses up to 50 mg/kg. Clinical trial simulations of a staggered dosing strategy starting the treatment at a lower dose (20 mg/kg) for 7 days followed by a higher dose (40 mg/kg) predicted a lower initial exposure with lower probability of tolerability problems and better EBA compared with a regimen of 35 mg/kg daily. CONCLUSIONS: Staggered dosing of 20 mg/kg for 7 days followed by 40 mg/kg is predicted to reduce tolerability while maintaining exposure levels associated with better efficacy.

Implementing best practices on data generation and reporting of Mycobacterium tuberculosis in vitro assays within the ERA4TB consortium.

Tuberculosis (TB) is the historical leading cause of death by a single infectious agent. The European Regimen Accelerator for Tuberculosis (ERA4TB) is a public-private partnership of 30+ institutions with the objective to progress new anti-TB regimens into the clinic. Thus, robust and replicable results across independent laboratories are essential for reliable interpretation of treatment efficacy. A standardization workgroup unified in vitro protocols and data reporting templates. Time-kill assays provide essential input data for pharmacometric model-informed translation of single agents and regimens activity from in vitro to in vivo and the clinic. Five conditions were assessed by time-kill assays in six independent laboratories using four bacterial plating methods. Baseline bacterial burden varied between laboratories but variability was limited in net drug effect, confirming 2.5 µL equally robust as 100 µL plating. This exercise establishes the foundations of collaborative data generation, reporting, and integration within the overarching Antimicrobial Resistance Accelerator program.

Combined quantitative tuberculosis biomarker model for time-to-positivity and colony forming unit to support tuberculosis drug development.

Biomarkers are quantifiable characteristics of biological processes. In Mycobacterium tuberculosis, common biomarkers used in clinical drug development are colony forming unit (CFU) and time-to-positivity (TTP) from sputum samples. This analysis aimed to develop a combined quantitative tuberculosis biomarker model for CFU and TTP biomarkers for assessing drug efficacy in early bactericidal activity studies. Daily CFU and TTP observations in 83 previously patients with uncomplicated pulmonary tuberculosis after 7 days of different rifampicin monotherapy treatments (10-40 mg/kg) from the HIGHRIF1 study were included in this analysis. The combined quantitative tuberculosis biomarker model employed the Multistate Tuberculosis Pharmacometric model linked to a rifampicin pharmacokinetic model in order to determine drug exposure-response relationships on three bacterial sub-states using both the CFU and TTP data simultaneously. CFU was predicted from the MTP model and TTP was predicted through a time-to-event approach from the TTP model, which was linked to the MTP model through the transfer of all bacterial sub-states in the MTP model to a one bacterial TTP model. The non-linear CFU-TTP relationship over time was well predicted by the final model. The combined quantitative tuberculosis biomarker model provides an efficient approach for assessing drug efficacy informed by both CFU and TTP data in early bactericidal activity studies and to describe the relationship between CFU and TTP over time.

Early bactericidal activity studies for pulmonary tuberculosis: A systematic review of methodological aspects.

A milestone in the development of novel antituberculosis drugs is the demonstration of early bactericidal activity (EBA) in a phase IIa clinical trial. The significant variability in measurements of bacterial load complicates data analysis in these trials. A systematic review and evaluation of methods for determination of EBA in pulmonary tuberculosis studies was undertaken. Bacterial load quantification biomarkers, reporting intervals, calculation methods, statistical testing, and handling of negative culture results were extracted. In total, 79 studies were identified in which EBA was determined. Colony-forming units on solid culture media and/or time-to-positivity in liquid media were the biomarkers used most often, reported in 72 (91%) and 34 (43%) studies, respectively. Twenty-two different reporting intervals were presented, and 12 different calculation methods for EBA were identified. Statistical testing for a significant EBA compared with no change was performed in 54 (68%) studies, and between-group testing was performed in 32 (41%) studies. Negative culture result handling was discussed in 34 (43%) studies. Notable variation was found in the analysis methods and reporting of EBA studies. A standardized and clearly reported analysis method, accounting for different levels of variability in the data, could aid the generalization of study results and facilitate comparison between drugs/regimens.

Pharmacokinetics and pharmacodynamics of anti-tuberculosis drugs: An evaluation of in vitro, in vivo methodologies and human studies.

There has been an increased interest in pharmacokinetics and pharmacodynamics (PKPD) of anti-tuberculosis drugs. A better understanding of the relationship between drug exposure, antimicrobial kill and acquired drug resistance is essential not only to optimize current treatment regimens but also to design appropriately dosed regimens with new anti-tuberculosis drugs. Although the interest in PKPD has resulted in an increased number of studies, the actual bench-to-bedside translation is somewhat limited. One of the reasons could be differences in methodologies and outcome assessments that makes it difficult to compare the studies. In this paper we summarize most relevant in vitro, in vivo, in silico and human PKPD studies performed to optimize the drug dose and regimens for treatment of tuberculosis. The in vitro assessment focuses on MIC determination, static time-kill kinetics, and dynamic hollow fibre infection models to investigate acquisition of resistance and killing of Mycobacterium tuberculosis populations in various metabolic states. The in vivo assessment focuses on the various animal models, routes of infection, PK at the site of infection, PD read-outs, biomarkers and differences in treatment outcome evaluation (relapse and death). For human PKPD we focus on early bactericidal activity studies and inclusion of PK and therapeutic drug monitoring in clinical trials. Modelling and simulation approaches that are used to evaluate and link the different data types will be discussed. We also describe the concept of different studies, study design, importance of uniform reporting including microbiological and clinical outcome assessments, and modelling approaches. We aim to encourage researchers to consider methods of assessing and reporting PKPD of anti-tuberculosis drugs when designing studies. This will improve appropriate comparison between studies and accelerate the progress in the field.

Study protocol for locoregional precision treatment of hepatocellular carcinoma with transarterial chemoembolisation (TACTida), a clinical study: idarubicin dose selection, tissue response and survival.

INTRODUCTION: Hepatocellular carcinoma (HCC) is a common cause of cancer-related death, often detected in the intermediate stage. The standard of care for intermediate-stage HCC is transarterial chemoembolisation (TACE), where idarubicin (IDA) is a promising drug. Despite the fact that TACE has been used for several decades, treatment success is unpredictable. This clinical trial has been designed believing that further improvement might be achieved by increasing the understanding of interactions between local pharmacology, tumour targeting, HCC pathophysiology, metabolomics and molecular mechanisms of drug resistance. METHODS AND ANALYSIS: The study population of this single-centre clinical trial consists of adults with intermediate-stage HCC. Each tumour site will receive TACE with two different IDA doses, 10 and 15 mg, on separate occasions. Before and after each patient's first TACE blood samples, tissue and liquid biopsies, and positron emission tomography (PET)/MRI will be performed. Blood samples will be used for pharmacokinetics (PK) and liver function evaluation. Tissue biopsies will be used for histopathology analyses, and culturing of primary organoids of tumour and non-tumour tissue to measure cell viability, drug response, multiomics and gene expression. Multiomics analyses will also be performed on liquid biopsies. PET/MRI will be used to evaluate tumour viability and liver metabolism. The two doses of IDA will be compared regarding PK, antitumour effects and safety. Imaging, molecular biology and multiomics data will be used to identify HCC phenotypes and their relation to drug uptake and metabolism, treatment response and survival. ETHICS AND DISSEMINATION: Participants give informed consent. Personal data are deidentified. A patient will be withdrawn from the study if considered medically necessary, or if it is the wish of the patient. The study has been approved by the Swedish Ethical Review Authority (Dnr. 2021-01928) and by the Medical Product Agency, Uppsala, Sweden. TRIAL REGISTRATION NUMBER: EudraCT number: 2021-001257-31.

Phase III dose selection of marzeptacog alfa (activated) informed by population pharmacokinetic modeling: A novel hemostatic drug.

Marzeptacog alfa (activated) (MarzAA) is an activated recombinant human FVII (rFVIIa) variant developed as subcutaneous (s.c.) administration for the treatment or prevention of bleeding episodes in patients with hemophilia A (HA) or hemophilia B (HB) with inhibitors and other rare bleeding disorders. Population pharmacokinetic (PK) modeling was applied for dose selection for a pivotal phase III clinical trial evaluating s.c. MarzAA for episodic treatment of spontaneous or traumatic bleeding episodes. The population PK model used MarzAA intravenous and s.c. data from previously completed clinical trials in patients with HA/HB with or without inhibitors. Based on the model, clinical trial simulations were performed to predict MarzAA exposure after different dosing regimens. The exposure target was identified using an exposure-matching strategy with a wild-type rFVIIa but adjusting for the difference in potency between the two compounds. Simulations demonstrated a sufficient absorption rate and prolonged exposure following a single 60 µg/kg dose leading to 51% and 70% of the population reaching levels above the target after 3 and 6 h, respectively. According to the phase III protocol, if a second dose was required after 3 h because of a lack of efficacy, 90% of the population was observed to be above target 6 h after the initial dose. The model-informed drug development approach integrated information from several trials and guided dose selection in the pivotal phase III clinical trial for episodic treatment of an acute bleeding event in individuals with HA or HB with inhibitors without the execution of a phase II trial for that indication.

Machine Learning and Pharmacometrics for Prediction of Pharmacokinetic Data: Differences, Similarities and Challenges Illustrated with Rifampicin.

Pharmacometrics (PM) and machine learning (ML) are both valuable for drug development to characterize pharmacokinetics (PK) and pharmacodynamics (PD). Pharmacokinetic/pharmacodynamic (PKPD) analysis using PM provides mechanistic insight into biological processes but is time- and labor-intensive. In contrast, ML models are much quicker trained, but offer less mechanistic insights. The opportunity of using ML predictions of drug PK as input for a PKPD model could strongly accelerate analysis efforts. Here exemplified by rifampicin, a widely used antibiotic, we explore the ability of different ML algorithms to predict drug PK. Based on simulated data, we trained linear regressions (LASSO), Gradient Boosting Machines, XGBoost and Random Forest to predict the plasma concentration-time series and rifampicin area under the concentration-versus-time curve from 0-24 h (AUC0-24h) after repeated dosing. XGBoost performed best for prediction of the entire PK series (R2: 0.84, root mean square error (RMSE): 6.9 mg/L, mean absolute error (MAE): 4.0 mg/L) for the scenario with the largest data size. For AUC0-24h prediction, LASSO showed the highest performance (R2: 0.97, RMSE: 29.1 h·mg/L, MAE: 18.8 h·mg/L). Increasing the number of plasma concentrations per patient (0, 2 or 6 concentrations per occasion) improved model performance. For example, for AUC0-24h prediction using LASSO, the R2 was 0.41, 0.69 and 0.97 when using predictors only (no plasma concentrations), 2 or 6 plasma concentrations per occasion as input, respectively. Run times for the ML models ranged from 1.0 s to 8 min, while the run time for the PM model was more than 3 h. Furthermore, building a PM model is more time- and labor-intensive compared with ML. ML predictions of drug PK could thus be used as input into a PKPD model, enabling time-efficient analysis.

Pharmacometrics in tuberculosis: progress and opportunities.

Tuberculosis (TB) remains one of the leading causes of death by a communicable agent, infecting up to one-quarter of the world's population, predominantly in disadvantaged communities. Pharmacometrics employ quantitative mathematical models to describe the relationships between pharmacokinetics and pharmacodynamics, and to predict drug doses, exposures and responses. Pharmacometric approaches have provided a scientific basis for improved dosing of anti-TB drugs and concomitantly administered antiretrovirals at the population level. The development of modelling frameworks including physiologically based pharmacokinetics, quantitative systems pharmacology and machine learning provides an opportunity to extend the role of pharmacometrics to in-silico quantification of drug-drug interactions, prediction of doses for special populations, dose optimization and individualization, and understanding the complex exposure-response relationships of multi-drug regimens in terms of both efficacy and safety, informing regimen design for future study. This short, clinically focused review explores what has been done, and what opportunities exist for pharmacometrics to impact TB pharmacotherapy.

Safety and efficacy of BCG re-vaccination in relation to COVID-19 morbidity in healthcare workers: A double-blind, randomised, controlled, phase 3 trial.

Background: BCG vaccination prevents severe childhood tuberculosis (TB) and was introduced in South Africa in the 1950s. It is hypothesised that BCG trains the innate immune system by inducing epigenetic and functional reprogramming, thus providing non-specific protection from respiratory tract infections. We evaluated BCG for reduction of morbidity and mortality due to COVID-19 in healthcare workers in South Africa. Methods: This randomised, double-blind, placebo-controlled trial recruited healthcare workers at three facilities in the Western Cape, South Africa, unless unwell, pregnant, breastfeeding, immunocompromised, hypersensitivity to BCG, or undergoing experimental COVID-19 treatment. Participants received BCG or saline intradermally (1:1) and were contacted once every 4 weeks for 1 year. COVID-19 testing was guided by symptoms. Hospitalisation, COVID-19, and respiratory tract infections were assessed with Cox proportional hazard modelling and time-to-event analyses, and event severity with post hoc Markovian analysis. This study is registered with ClinicalTrials.gov, NCT04379336. Findings: Between May 4 and Oct 23, 2020, we enrolled 1000 healthcare workers with a median age of 39 years (IQR 30-49), 70·4% were female, 16·5% nurses, 14·4% medical doctors, 48·5% had latent TB, and 15·3% had evidence of prior SARS-CoV-2 exposure. Hospitalisation due to COVID-19 occurred in 15 participants (1·5%); ten (66·7%) in the BCG group and five (33·3%) in the placebo group, hazard ratio (HR) 2·0 (95% CI 0·69-5·9, p = 0·20), indicating no statistically significant protection. Similarly, BCG had no statistically significant effect on COVID-19 (p = 0·63, HR = 1·08, 95% CI 0·82-1·42). Two participants (0·2%) died from COVID-19 and two (0·2%) from other reasons, all in the placebo group. Interpretation: BCG did not protect healthcare workers from SARS-CoV-2 infection or related severe COVID-19 disease and hospitalisation. Funding: Funding provided by EDCTP, grant number RIA2020EF-2968. Additional funding provided by private donors including: Mediclinic, Calavera Capital (Pty) Ltd, Thys Du Toit, Louis Stassen, The Ryan Foundation, and Dream World Investments 401 (Pty) Ltd. The computations were enabled by resources in project SNIC 2020-5-524 provided by the Swedish National Infrastructure for Computing (SNIC) at UPPMAX, partially funded by the Swedish Research Council through grant agreement No. 2018-05,973.