Comparative assessment of viral dynamic models for SARS-CoV-2 for pharmacodynamic assessment in early treatment trials.

Pharmacometric analyses of time series viral load data may detect drug effects with greater power than approaches using single time points. Because SARS-CoV-2 viral load rapidly rises and then falls, viral dynamic models have been used. We compared different modelling approaches when analysing Phase II-type viral dynamic data. Using two SARS-CoV-2 datasets of viral load starting within 7 days of symptoms, we fitted the slope-intercept exponential decay (SI), reduced target cell limited (rTCL), target cell limited (TCL) and TCL with eclipse phase (TCLE) models using nlmixr. Model performance was assessed via Bayesian information criterion (BIC), visual predictive checks (VPCs), goodness-of-fit plots, and parameter precision. The most complex (TCLE) model had the highest BIC for both datasets. The estimated viral decline rate was similar for all models except the TCL model for dataset A with a higher rate (median [range] day-1 : dataset A; 0.63 [0.56-1.84]; dataset B: 0.81 [0.74-0.85]). Our findings suggest simple models should be considered during pharmacodynamic model development.

Application of the hollow fibre infection model (HFIM) in antimicrobial development: a systematic review and recommendations of reporting.

OBJECTIVES: This systematic review focuses on the use of the in vitro hollow fibre infection model (HFIM) for microbial culture. We summarize the direction of the field to date and propose best-practice principles for reporting of the applications. METHODS: Searches in six databases (MEDLINE®, EMBASE®, PubMed®, BIOSIS®, SCOPUS® and Cochrane®) up to January 2020 identified 129 studies meeting our inclusion criteria. Two reviewers independently assessed and extracted data from each publication. The quality of reporting of microbiological and technical parameters was analysed. RESULTS: Forty-seven out of 129 (36.4%) studies did not report the minimum pharmacokinetic parameters required in order to replicate the pharmacokinetic profile of HFIM experiments. Fifty-three out of 129 (41.1%) publications did not report the medium used in the HFIM. The overwhelming majority of publications did not perform any technical repeats [107/129 (82.9%)] or biological repeats [97/129 (75.2%)]. CONCLUSIONS: This review demonstrates that most publications provide insufficient data to allow for results to be evaluated, thus impairing the reproducibility of HFIM experiments. Therefore, there is a clear need for the development of laboratory standardization and improved reporting of HFIM experiments.

Revising Pediatric Vancomycin Dosing Accounting for Nephrotoxicity in a Pharmacokinetic-Pharmacodynamic Model.

This study aimed to suggest an initial pediatric vancomycin dose regimen through population pharmacokinetic-pharmacodynamic modeling. A population pharmacokinetic approach was used to analyze vancomycin concentration-time data from a large pediatric cohort. Pharmacokinetic target attainment for patients with bloodstream isolates was compared with clinical outcome using logistic regression and classification and regression trees. Change in serum creatinine during treatment was used as an indicator of acute nephrotoxicity. Probability of acute kidney injury (50% increase from baseline) or kidney failure (75% increase from baseline) was evaluated using logistic regression. An initial dosing regimen was derived, personalized by age, weight, and serum creatinine, using stochastic simulations. Data from 785 hospitalized pediatric patients (1 day to 21 years of age) with suspected Gram-positive infections were collected. Estimated (relative standard error) typical clearance, volume of distribution 1, intercompartmental clearance, and volume of distribution 2 were (standardized to 70 kg) 4.84 (2.38) liters/h, 39.9 (8.15) liters, 3.85 (17.3) liters/h, and 37.8 (10.2) liters, respectively. While cumulative vancomycin exposure correlated positively with the development of nephrotoxicity (713 patients), no clear relationship between vancomycin area under the plasma concentration-time curve and efficacy was found (102 patients). Predicted probability of acute kidney injury and kidney failure with the optimized dosing regimen at day 5 was 10 to 15% and 5 to 10%, increasing by approximately 50% on day 7 and roughly 100% on day 10 across all age groups. This study presents the first data-driven pediatric dose selection to date accounting for nephrotoxicity, and it indicates that cumulative vancomycin exposure best describes risk of acute kidney injury and acute kidney failure.

Can phenotypic data complement our understanding of antimycobacterial effects for drug combinations?

OBJECTIVES: To demonstrate how phenotypic cell viability data can provide insight into antimycobacterial effects for the isoniazid/rifampicin treatment backbone. METHODS: Data from a Mycobacterium komossense hollow-fibre infection model comprising a growth control group, rifampicin at three different exposures (Cmax = 0.14, 0.4 and 1.47 mg/L with t½ = 1.57 h and τ = 8 h) and rifampicin plus isoniazid (Cmax rifampicin = 0.4 mg/L and Cmax isoniazid = 1.2 mg/L with t½ = 1.57 h and τ = 8 h) were used for this investigation. A non-linear mixed-effects modelling approach was used to fit conventional cfu data, quantified using solid-agar plating. Phenotypic proportions of respiring (alive), respiring but with damaged cell membrane (injured) and 'not respiring' (dead) cells data were quantified using flow cytometry and Sytox Green™ (Sigma-Aldrich, UK) and resazurin sodium salt staining and fitted using a multinomial logistic regression model. RESULTS: Isoniazid/rifampicin combination therapy displayed a decreasing overall antimicrobial effect with time (θTime1/2 = 438 h) on cfu data, in contrast to rifampicin monotherapy where this trend was absent. In the presence of isoniazid a phenotype associated with cell injury was displayed, whereas with rifampicin monotherapy a pattern of phenotypic cell death was observed. Bacterial killing onset time on cfu data correlated negatively (θTime50 = 28.9 h, θLAGRIF50 = 0.132 mg/L) with rifampicin concentration up to 0.165 mg/L and this coincided with a positive relationship between rifampicin concentration and the probability of phenotypic cell death. CONCLUSIONS: Cell viability data provide structured information on the pharmacodynamic interaction between isoniazid and rifampicin that complements the understanding of the antibacillary effects of this mycobacterial treatment backbone.

Pharmacokinetic-pharmacodynamic modelling to investigate in vitro synergy between colistin and fusidic acid against MDR Acinetobacter baumannii.

OBJECTIVES: The potential for synergy between colistin and fusidic acid in the treatment of MDR Acinetobacter baumannii has recently been shown. The aim of this study was to perform an extensive in vitro characterization of this effect using pharmacokinetic-pharmacodynamic modelling (PKPD) of time-kill experiments in order to estimate clinical efficacy. METHODS: For six clinical strains, 312 individual time-kill experiments were performed including 113 unique pathogen-antimicrobial combinations. A wide range of concentrations (0.25-8192 mg/L for colistin and 1-8192 mg/L for fusidic acid) were explored, alone and in combination. PKPD modelling sought to quantify synergistic effects. RESULTS: A PKPD model confirmed synergy in that colistin EC50 was found to decrease by 83% in the presence of fusidic acid, and fusidic acid maximum increase in killing rate (Emax) also increased 58% in the presence of colistin. Simulations indicated, however, that at clinically achievable free concentrations, the combination may be bacteriostatic in colistin-susceptible strains, but growth inhibition probability was <20% in a colistin-resistant strain. CONCLUSIONS: Fusidic acid may be a useful agent to add to colistin in a multidrug combination for MDR Acinetobacter baumannii.

Artemether-lumefantrine dosing for malaria treatment in young children and pregnant women: A pharmacokinetic-pharmacodynamic meta-analysis.

BACKGROUND: The fixed dose combination of artemether-lumefantrine (AL) is the most widely used treatment for uncomplicated Plasmodium falciparum malaria. Relatively lower cure rates and lumefantrine levels have been reported in young children and in pregnant women during their second and third trimester. The aim of this study was to investigate the pharmacokinetic and pharmacodynamic properties of lumefantrine and the pharmacokinetic properties of its metabolite, desbutyl-lumefantrine, in order to inform optimal dosing regimens in all patient populations. METHODS AND FINDINGS: A search in PubMed, Embase, ClinicalTrials.gov, Google Scholar, conference proceedings, and the WorldWide Antimalarial Resistance Network (WWARN) pharmacology database identified 31 relevant clinical studies published between 1 January 1990 and 31 December 2012, with 4,546 patients in whom lumefantrine concentrations were measured. Under the auspices of WWARN, relevant individual concentration-time data, clinical covariates, and outcome data from 4,122 patients were made available and pooled for the meta-analysis. The developed lumefantrine population pharmacokinetic model was used for dose optimisation through in silico simulations. Venous plasma lumefantrine concentrations 7 days after starting standard AL treatment were 24.2% and 13.4% lower in children weighing <15 kg and 15-25 kg, respectively, and 20.2% lower in pregnant women compared with non-pregnant adults. Lumefantrine exposure decreased with increasing pre-treatment parasitaemia, and the dose limitation on absorption of lumefantrine was substantial. Simulations using the lumefantrine pharmacokinetic model suggest that, in young children and pregnant women beyond the first trimester, lengthening the dose regimen (twice daily for 5 days) and, to a lesser extent, intensifying the frequency of dosing (3 times daily for 3 days) would be more efficacious than using higher individual doses in the current standard treatment regimen (twice daily for 3 days). The model was developed using venous plasma data from patients receiving intact tablets with fat, and evaluations of alternative dosing regimens were consequently only representative for venous plasma after administration of intact tablets with fat. The absence of artemether-dihydroartemisinin data limited the prediction of parasite killing rates and recrudescent infections. Thus, the suggested optimised dosing schedule was based on the pharmacokinetic endpoint of lumefantrine plasma exposure at day 7. CONCLUSIONS: Our findings suggest that revised AL dosing regimens for young children and pregnant women would improve drug exposure but would require longer or more complex schedules. These dosing regimens should be evaluated in prospective clinical studies to determine whether they would improve cure rates, demonstrate adequate safety, and thereby prolong the useful therapeutic life of this valuable antimalarial treatment.

Lumefantrine and Desbutyl-Lumefantrine Population Pharmacokinetic-Pharmacodynamic Relationships in Pregnant Women with Uncomplicated Plasmodium falciparum Malaria on the Thailand-Myanmar Border.

Artemether-lumefantrine is the most widely used antimalarial artemisinin-based combination treatment. Recent studies have suggested that day 7 plasma concentrations of the potent metabolite desbutyl-lumefantrine correlate better with treatment outcomes than those of lumefantrine. Low cure rates have been reported in pregnant women with uncomplicated falciparum malaria treated with artemether-lumefantrine in northwest Thailand. A simultaneous pharmacokinetic drug-metabolite model was developed based on dense venous and sparse capillary lumefantrine and desbutyl-lumefantrine plasma samples from 116 pregnant patients on the Thailand-Myanmar border. The best model was used to evaluate therapeutic outcomes with a time-to-event approach. Lumefantrine and desbutyl-lumefantrine concentrations, implemented in an Emax model, both predicted treatment outcomes, but lumefantrine provided better predictive power. A combined model including both lumefantrine and desbutyl-lumefantrine did not improve the model further. Simulations suggested that cure rates in pregnant women with falciparum malaria could be increased by prolonging the treatment course. (These trials were registered at controlled-trials.com [ISRCTN 86353884].).

Opposite malaria and pregnancy effect on oral bioavailability of artesunate - a population pharmacokinetic evaluation.

AIM: The aim was to compare the pharmacokinetic properties of artesunate and dihydroartemisinin in the same women: i) pregnant with acute uncomplicated malaria on day 1 and 2, ii) pregnant with convalescent malaria on day 7 and iii) in a healthy state 3 months post-partum on day 1, 2 and 7. METHODS: Non-linear mixed-effects modelling was used to compare plasma concentration-time profiles of artesunate and dihydroartemisinin over 7 days of treatment following oral and intravenous artesunate administration to pregnant women with uncomplicated Plasmodium falciparum malaria during their second or third trimesters of pregnancy. The same women were restudied 3 months after delivery when fully recovered. Non-compartmental results of the same study have been published previously. RESULTS: Twenty pregnant patients on the Thailand-Myanmar border were studied and 15 volunteered to be restudied 3 months post-partum. Malaria and pregnancy had no effect on the pharmacokinetic properties of artesunate or dihydroartemisinin after intravenous artesunate administration. However, malaria and pregnancy had opposite effects on the absorption of orally administered artesunate. Malaria increased the absolute oral bioavailability of artesunate by 87%, presumably by inhibiting first pass effect, whereas pregnancy decreased oral bioavailability by 23%. CONCLUSIONS: The population pharmacokinetic analysis demonstrated opposite effects of malaria and pregnancy on the bioavailability of orally administered artesunate. Lower drug exposures during the second and third trimesters of pregnancy may contribute to lower cure rates and thus the development of drug resistance. Dose optimization studies are required for artesunate containing artemisinin-based combination therapies (ACTs) in later pregnancy.

Population pharmacokinetics of quinine in pregnant women with uncomplicated Plasmodium falciparum malaria in Uganda.

OBJECTIVES: Oral quinine is used for the treatment of uncomplicated malaria during pregnancy, but few pharmacokinetic data are available for this population. Previous studies have reported a substantial effect of malaria on the pharmacokinetics of quinine resulting from increased α-1-acid glycoprotein levels and decreased cytochrome P450 3A4 activity. The aim of this study was to investigate the pharmacokinetic properties of oral quinine in pregnant women with uncomplicated malaria in Uganda using a population approach. METHODS: Data from 22 women in the second and third trimesters of pregnancy with uncomplicated Plasmodium falciparum malaria were analysed. Patients received quinine sulphate (10 mg of salt/kg) three times daily (0, 8 and 16 h) for 7 days. Plasma samples were collected daily and at frequent intervals after the first and last doses. A population pharmacokinetic model for quinine was developed accounting for different disposition, absorption, error and covariate models. RESULTS: Parasitaemia, as a time-varying covariate affecting relative bioavailability, and body temperature on admission as a covariate on elimination clearance, explained the higher exposure to quinine during acute malaria compared with the convalescent phase. Neither the estimated gestational age nor the trimester influenced the pharmacokinetic properties of quinine significantly. CONCLUSIONS: A population model was developed that adequately characterized quinine pharmacokinetics in pregnant Ugandan women with acute malaria. Quinine exposure was lower than previously reported in patients who were not pregnant. The measurement of free quinine concentration will be necessary to determine the therapeutic relevance of these observations.

Antibiotic Residues in UK Foods: Exploring the Exposure Pathways and Associated Health Risks.

While the use of antibiotics has been reported as extensive in the rearing of agricultural animals, insufficient information is available on the antibiotic residues in animal products and the adverse impact that consistent low-level exposure to antibiotics might have on the human body and its microbiome. The aim of this study was to estimate the antibiotic concentrations that humans are exposed to via their diet using the concentration of antibiotics in animal food products and water and an online survey on dietary intake. A total of 131 participants completed the dietary intake survey, with the majority belonging to the omnivorous diet group (76.3%). Distinct dietary trends were observed in the omnivorous and unknown groups eating animal products, with specific food types dominating each meal: pork (e.g., ham) and dairy products (e.g., milk, yoghurt) during breakfast, beef (e.g., burgers) and chicken (e.g., chicken breast) products during lunch, and fish (e.g., salmon fillet) during dinner. In total, 34 different animal-based food and drink products were tested for the presence of ten different antibiotics. Of all the products tested, over 35% exceeded the acceptable daily antibiotic intake for amoxicillin, ampicillin, and enrofloxacin.

Pharmacokinetic properties of artemether, dihydroartemisinin, lumefantrine, and quinine in pregnant women with uncomplicated plasmodium falciparum malaria in Uganda.

Pregnancy alters the pharmacokinetic properties of many drugs used in the treatment of malaria, usually resulting in lower drug exposures. This increases the risks of treatment failure, adverse outcomes for the fetus, and the development of resistance. The pharmacokinetic properties of artemether and its principal metabolite dihydroartemisinin (n = 21), quinine (n = 21), and lumefantrine (n = 26) in pregnant Ugandan women were studied. Lumefantrine pharmacokinetics in a nonpregnant control group (n = 17) were also studied. Frequently sampled patient data were evaluated with noncompartmental analysis. No significant correlation was observed between estimated gestational age and artemether, dihydroartemisinin, lumefantrine, or quinine exposures. Artemether/dihydroartemisinin and quinine exposures were generally low in these pregnant women compared to values reported previously for nonpregnant patients. Median day 7 lumefantrine concentrations were 488 (range, 30.7 to 3,550) ng/ml in pregnant women compared to 720 (339 to 2,150) ng/ml in nonpregnant women (P = 0.128). There was no statistical difference in total lumefantrine exposure or maximum concentration. More studies with appropriate control groups in larger series are needed to characterize the degree to which pregnant women are underdosed with current antimalarial dosing regimens.

Optimising intravenous salbutamol in children: a phase 2 study.

OBJECTIVE: The ß2-agonists such as salbutamol are the mainstay of asthma management. Pharmacokinetic-pharmacodynamic (PKPD) models to guide paediatric dosing are lacking. We explored the relationship between salbutamol dose, serum concentration, effectiveness and adverse effects in children by developing a PKPD model. DESIGN: A prospective cohort study of children admitted to hospital with acute asthma, who received intravenous salbutamol. SETTING: Children were recruited in two cohorts: the emergency departments of two London hospitals or those retrieved by the Children's Acute Transport Service to three London paediatric intensive care units. PATIENTS: Patients were eligible if aged 1-15 years, admitted for acute asthma and about to receive or receiving intravenous salbutamol. INTERVENTIONS: Treatment was according to local policy. Serial salbutamol plasma levels were taken. Effectiveness measurements were recorded using the Paediatric Asthma Severity Score (PASS). Toxicity measurements included lactate, pH, glucose, heart rate, blood pressure and arrhythmias. PKPD modelling was performed with non-linear mixed-effect models. MAIN OUTCOMES: Fifty-eight children were recruited with 221 salbutamol concentration measurements from 54 children. Median (range) age was 2.9 (1.1-15.2) years, and weight was 13.6 (8-57.3) kg. Ninety-five PASS measurements and 2078 toxicity measurements were obtained. RESULTS: A two-compartment PK model adequately described the time course of salbutamol-plasma concentrations. An EMAX (maximum drug effect) concentration-effect relationship described PASS and toxicity measures. PKPD simulations showed an infusion of 0.5 µg/kg/min (maximum 20 µg/min) for 4 hours after bolus achieves >90% maximal bronchodilation for 12 hours. CONCLUSIONS: A paediatric PKPD model for salbutamol is described. An infusion of 0.5 µg/kg/min after bolus achieves effective bronchodilation. Higher rates are associated with greater tachycardia and hyperglycaemia.

Population pharmacokinetics of Artemether and dihydroartemisinin in pregnant women with uncomplicated Plasmodium falciparum malaria in Uganda.

BACKGROUND: Malaria in pregnancy increases the risk of maternal anemia, abortion and low birth weight. Approximately 85.3 million pregnancies occur annually in areas with Plasmodium falciparum transmission. Pregnancy has been reported to alter the pharmacokinetic properties of many anti-malarial drugs. Reduced drug exposure increases the risk of treatment failure. The objective of this study was to evaluate the population pharmacokinetic properties of artemether and its active metabolite dihydroartemisinin in pregnant women with uncomplicated P. falciparum malaria in Uganda. METHODS: Twenty-one women with uncomplicated P. falciparum malaria in the second and third trimesters of pregnancy received the fixed oral combination of 80 mg artemether and 480 mg lumefantrine twice daily for three days. Artemether and dihydroartemisinin plasma concentrations after the last dose administration were quantified using liquid chromatography coupled to tandem mass-spectroscopy. A simultaneous drug-metabolite population pharmacokinetic model for artemether and dihydroartemisinin was developed taking into account different disposition, absorption, error and covariate models. A separate modeling approach and a non-compartmental analysis (NCA) were also performed to enable a comparison with literature values and different modeling strategies. RESULTS: The treatment was well tolerated and there were no cases of recurrent malaria. A flexible absorption model with sequential zero-order and transit-compartment absorption followed by a simultaneous one-compartment disposition model for both artemether and dihydroartemisinin provided the best fit to the data. Artemether and dihydroartemisinin exposure was lower than that reported in non-pregnant populations. An approximately four-fold higher apparent volume of distribution for dihydroartemisinin was obtained by non-compartmental analysis and separate modeling compared to that from simultaneous modeling of the drug and metabolite. This highlights a potential pitfall when analyzing drug/metabolite data with traditional approaches. CONCLUSION: The population pharmacokinetic properties of artemether and dihydroartemisinin, in pregnant women with uncomplicated P. falciparum malaria in Uganda, were described satisfactorily by a simultaneous drug-metabolite model without covariates. Concentrations of artemether and its metabolite dihydroartemisinin were relatively low in pregnancy compared to literature data. However, this should be interpreted with caution considered the limited literature available. Further studies in larger series are urgently needed for this vulnerable group.

Emergence of phenotypic and genotypic antimicrobial resistance in Mycobacterium tuberculosis.

Concentration dependency of phenotypic and genotypic isoniazid-rifampicin resistance emergence was investigated to obtain a mechanistic understanding on how anti-mycobacterial drugs facilitate the emergence of bacterial populations that survive throughout treatment. Using static kill curve experiments, observing two evolution cycles, it was demonstrated that rifampicin resistance was the result of non-specific mechanisms and not associated with accumulation of drug resistance encoding SNPs. Whereas, part of isoniazid resistance could be accounted for by accumulation of specific SNPs, which was concentration dependent. Using a Hollow Fibre Infection Model it was demonstrated that emergence of resistance did not occur at concentration-time profiles mimicking the granuloma. This study showed that disentangling and quantifying concentration dependent emergence of resistance provides an improved rational for drug and dose selection although further work to understand the underlying mechanisms is needed to improve the drug development pipeline.

Improving the Drug Development Pipeline for Mycobacteria: Modelling Antibiotic Exposure in the Hollow Fibre Infection Model.

Mycobacterial infections are difficult to treat, requiring a combination of drugs and lengthy treatment times, thereby presenting a substantial burden to both the patient and health services worldwide. The limited treatment options available are under threat due to the emergence of antibiotic resistance in the pathogen, hence necessitating the development of new treatment regimens. Drug development processes are lengthy, resource intensive, and high-risk, which have contributed to market failure as demonstrated by pharmaceutical companies limiting their antimicrobial drug discovery programmes. Pre-clinical protocols evaluating treatment regimens that can mimic in vivo PK/PD attributes can underpin the drug development process. The hollow fibre infection model (HFIM) allows for the pathogen to be exposed to a single or a combination of agents at concentrations achieved in vivo-in plasma or at infection sites. Samples taken from the HFIM, depending on the analyses performed, provide information on the rate of bacterial killing and the emergence of resistance. Thereby, the HFIM is an effective means to investigate the efficacy of a drug combination. Although applicable to a wide variety of infections, the complexity of anti-mycobacterial drug discovery makes the information available from the HFIM invaluable as explored in this review.

Population Pharmacokinetics of Intraventricular Vancomycin in Neonatal Ventriculitis, A Preterm Pilot Study.

AIM: Intraventricular vancomycin is an effective treatment for neonatal ventriculitis, as the cerebrospinal fluid (CSF) vancomycin levels reach adequate concentrations to achieve microbiological cure. There is no robust data on intraventricular vancomycin pharmacokinetics in the preterm population. This pilot population pharmacokinetic modelling study examines the pharmacokinetic behaviour of intraventricular vancomycin in the preterm population of < 28 weeks gestation, to inform the feasibility of future prospective studies. METHODS: The study comprised 8 preterm infants with neonatal ventriculitis (median gestation age 25.3 weeks; range 23.9 - 27.7). Population pharmacokinetics (non-linear mixed effects modelling) were described with one- and two-compartment models to fit plasma concentrations of vancomycin. A CSF compartment was added to the plasma modelling and mass transfer examined. Three covariates (serum creatinine, ventricular index (VI) and CSF protein) were tested on the final model. Area under the curve (AUC) and average CSF concentration (C average) predictions were generated from the final model and compared with time to microbiological cure. RESULTS: A one-compartment model provided the best fit to the data. There was no appreciable transfer between plasma and CSF. None of the covariates provided a significant reduction in the objective function value (OFV). Generally, time to sterilisation with higher CSF AUC (0-24) and C average tends to be shorter, however this should be interpreted with caution as data is erratic. CONCLUSION: This pilot population pharmacokinetic analysis provides important information to warrant changes in the management of intraventricular vancomycin treatment in the preterm population, such as the current use of VI as a dosing parameter. Further study with a larger data pool is necessary to investigate the influence of VI on CSF vancomycin and ascertain dosing strategies.

Exploring a combined biomarker for tuberculosis treatment response: protocol for a prospective observational cohort study.

INTRODUCTION: An improved understanding of factors explaining tuberculosis (TB) treatment response is urgently needed to help clinicians optimise and personalise treatment and assist scientists undertaking novel treatment regimen trials. Promising outcome proxy measures, including sputum bacillary load and host immune response, are widely reported with variable results. However, they have not been studied together in combination with antibiotic exposure. The aim of this observational cohort study is to investigate which antibiotic exposures correlate with sputum bacillary load and which with the host immune response. Subsequently, we will explore if these correlations can be used to inform a candidate combined biomarker predicting cure. METHODS AND ANALYSIS: All patients aged ≥ 18, diagnosed with drug-sensitive pulmonary TB (culture or molecular test), eligible for standard anti-TB treatment, at selected London, UK TB Services, will be invited to participate in this observational cohort study (target sample size=210). Patients will be asked to give blood for host transcriptomics and antibiotic plasma exposure, in addition to standard of care sputum samples for bacillary load. Antibiotic plasma concentrations will be quantified using a validated liquid chromatograph triple quadrupole mass spectrometer (LC-MS/MS) assay and sputum bacillary load by mycobacterial growth incubator tube time to positivity. Expression from a total of 35 prespecified host blood genes will be quantified using NanoString®. Antibiotic exposure, sputum bacillary load and host blood transcriptomic time series data will be analysed using nonlinear mixed-effects models. Correlations between combinations of longitudinal biomarkers and microbiological cure at the end of treatment and remaining relapse free for 1 year thereafter will be analysed using logistic regression and Cox proportional hazard models. ETHICS AND DISSEMINATION: The observational cohort study has been approved by the UK's HRA REC (20/SW/0007). Written informed consent will be obtained. Results will be disseminated via publication, presentation and through engagement with institutes/companies developing novel anti-TB treatment combinations.

Population pharmacokinetics and pharmacodynamics of investigational regimens' drugs in the TB-PRACTECAL clinical trial (the PRACTECAL-PKPD study): a prospective nested study protocol in a randomised controlled trial.

INTRODUCTION: Drug-resistant tuberculosis (TB) remains a global health threat, with little over 50% of patients successfully treated. Novel regimens like the ones being studied in the TB-PRACTECAL trial are urgently needed. Understanding anti-TB drug exposures could explain the success or failure of these trial regimens. We aim to study the relationship between the patients' exposure to anti-TB drugs in TB-PRACTECAL investigational regimens and their treatment outcomes. METHODS AND ANALYSIS: Adults with multidrug-resistant TB randomised to investigational regimens in TB-PRACTECAL will be recruited to a nested pharmacokinetic-pharmacodynamic (PKPD) study. Venous blood samples will be collected at 0, 2 and 23 hours postdose on day 1 and 0, 6.5 and 23 hours postdose during week 8 to quantify drug concentrations in plasma. Trough samples will be collected during week 12, 16, 20 and 24 visits. Opportunistic samples will be collected during weeks 32 and 72. Drug concentrations will be quantified using liquid chromatography-tandem mass spectrometry. Sputum samples will be collected at baseline, monthly to week 24 and then every 2 months to week 108 for MICs and bacillary load quantification. Full blood count, urea and electrolytes, liver function tests, lipase, ECGs and ophthalmology examinations will be conducted at least monthly during treatment.PK and PKPD models will be developed for each drug with nonlinear mixed effects methods. Optimal dosing will be investigated using Monte-Carlo simulations. ETHICS AND DISSEMINATION: The study has been approved by the Médecins sans Frontières (MSF) Ethics Review Board, the LSHTM Ethics Committee, the Belarus RSPCPT ethics committee and PharmaEthics and the University of Witwatersrand Human Research ethics committee in South Africa. Written informed consent will be obtained from all participants. The study results will be shared with public health authorities, presented at scientific conferences and published in a peer-reviewed journal. TRIAL REGISTRATION NUMBER: NCT04081077; Pre-results.

An automated approach to identify scientific publications reporting pharmacokinetic parameters.

Pharmacokinetic (PK) predictions of new chemical entities are aided by prior knowledge from other compounds. The development of robust algorithms that improve preclinical and clinical phases of drug development remains constrained by the need to search, curate and standardise PK information across the constantly-growing scientific literature. The lack of centralised, up-to-date and comprehensive repositories of PK data represents a significant limitation in the drug development pipeline.In this work, we propose a machine learning approach to automatically identify and characterise scientific publications reporting PK parameters from in vivo data, providing a centralised repository of PK literature. A dataset of 4,792 PubMed publications was labelled by field experts depending on whether in vivo PK parameters were estimated in the study. Different classification pipelines were compared using a bootstrap approach and the best-performing architecture was used to develop a comprehensive and automatically-updated repository of PK publications. The best-performing architecture encoded documents using unigram features and mean pooling of BioBERT embeddings obtaining an F1 score of 83.8% on the test set. The pipeline retrieved over 121K PubMed publications in which in vivo PK parameters were estimated and it was scheduled to perform weekly updates on newly published articles. All the relevant documents were released through a publicly available web interface (https://app.pkpdai.com) and characterised by the drugs, species and conditions mentioned in the abstract, to facilitate the subsequent search of relevant PK data. This automated, open-access repository can be used to accelerate the search and comparison of PK results, curate ADME datasets, and facilitate subsequent text mining tasks in the PK domain.