Artificial Neural Network vs. Pharmacometric Model for Population Prediction of Plasma Concentration in Real-World Data: A Case Study on Valproic Acid.

We compared the predictive performance of an artificial neural network to traditional pharmacometric modeling for population prediction of plasma concentrations of valproate in real-world data. We included individuals aged 65 years or older with epilepsy who redeemed their first prescription of valproate after the diagnosis of epilepsy and had at least one valproate plasma concentration measured. A long short-term memory neural network (LSTM) was developed using the training data set to fit the LSTM and the test data set to validate the model. Predictions from the LSTM were compared with those obtained from the population predictions from a pharmacometric model by Birnbaum et al. which had the best predictive performance for population predictions of valproate concentrations in Danish databases. We used the cutoff of ± 20 mg/L of prediction error to define good predictions. A total of 1,252 individuals were included in the study. The LSTM fitted using the training data set had poor predictive performance in the test data set, but better than that of the pharmacometric model. The proportion of individuals with at least one predicted concentration within ± 20 mg/L of observed concentration was largest in case of the LSTM (64.4%, 95% confidence interval (CI): 58.4-70.2%) compared with the pharmacometric model by Birnbaum et al. (49.8%, 95% CI: 47.0-52.6%). LSTM shows better predictive performance to predict valproate plasma concentrations compared with a traditional pharmacometric model in the investigated setting with real-world data in older patients with epilepsy where information on exact timepoints for both dosing and plasma concentration measurement are missing.

The PHARMACOM-EPI Framework for Integrating Pharmacometric Modelling Into Pharmacoepidemiological Research Using Real-World Data: Application to Assess Death Associated With Valproate.

In pharmacoepidemiology, it is usually expected that the observed association should be directly or indirectly related to the pharmacological effects of the drug/s under investigation. Pharmacological effects are, in turn, strongly connected to the pharmacokinetic and pharmacodynamic properties of a drug, which can be characterized and investigated using pharmacometric models. Recently, the use of pharmacometrics has been proposed to provide pharmacological substantiation of pharmacoepidemiological findings derived from real-world data. However, validated frameworks suggesting how to combine these two disciplines for the aforementioned purpose are missing. Therefore, we propose PHARMACOM-EPI, a framework that provides a structured approach on how to identify, characterize, and apply pharmacometric models with practical details on how to choose software, format dataset, handle missing covariates/dosing data, how to perform the external evaluation of pharmacometric models in real-world data, and how to provide pharmacological substantiation of pharmacoepidemiological findings. PHARMACOM-EPI was tested in a proof-of-concept study to pharmacologically substantiate death associated with valproate use in the Danish population aged ≥ 65 years. Pharmacological substantiation of death during a follow-up period of 1 year showed that in all individuals who died (n = 169) individual predictions were within the subtherapeutic range compared with 52.8% of those who did not die (n = 1,084). Of individuals who died, 66.3% (n = 112) had a cause of death possibly related to valproate and 33.7% (n = 57) with well-defined cause of death unlikely related to valproate. This proof-of-concept study showed that PHARMACOM-EPI was able to provide pharmacological substantiation for death associated with valproate use in the study population.

Utilizing drug-target-event relationships to unveil safety patterns in pharmacovigilance.

INTRODUCTION: Signal detection is the most pivotal activity of signal management to guarantee that drugs maintain a positive risk-benefit ratio during their lifetime on the market. Signal detection is based on the systematic evaluation of available data sources, which have recently been extended in order to improve timely and comprehensive signal detection of drug safety problems. AREAS COVERED: In recent years, attempts have been made to incorporate pharmacological data for the prediction of safety signals. Previous studies have shown that data on the pharmacological targets of drugs are predictive of post-marketing adverse events. However, current approaches limit such predictions to adverse events expected from the interaction of a drug with the main pharmacological target and do not take off-target interactions into consideration. EXPERT OPINION: The authors propose the application of predictive modeling techniques utilizing pharmacological data from public databases for predicting drug-target-event relationships deriving from main- and off-target binding and from which potential safety signals can be deduced. Additionally, they provide an operative procedure for the identification of clinically relevant subgroups for predicted safety signals.

Midazolam Pharmacokinetics in Obese and Non-obese Children and Adolescents.

BACKGROUND: Midazolam is a first-line drug for the treatment of status epilepticus, both by buccal and intravenous administration. In children and adolescents with obesity, midazolam pharmacokinetics may be altered, and the current dosing guidelines may therefore be insufficient. OBJECTIVE: The objective of this study was to investigate the pharmacokinetics of midazolam, after intravenous administration, in obese and non-obese adolescents aged 11-18 years. METHODS: All trial participants received a 1-µg midazolam microdose as an intravenous bolus. 13 blood samples were collected per participant at pre-specified timepoints. Plasma concentration-time data were fitted to pharmacokinetic models using non-linear mixed-effects modeling. Covariates such as weight, age, and body mass index standard deviation score were tested to explain the inter-individual variability associated with the pharmacokinetic parameters. RESULTS: Sixty-seven adolescents were included in the analysis. The pharmacokinetics of midazolam was best described with a two-compartment model. The rate of distribution was faster, and the peripheral volume of distribution was larger in adolescents with a high body mass index standard deviation score compared with adolescents with a lower standard deviation score. Simulations revealed that long-term infusions based on total body weight could lead to high plasma concentrations in adolescents with obesity. Furthermore, simulated plasma concentrations after a fixed buccal dose indicated that adolescents with obesity may be at risk of sub-therapeutic midazolam plasma concentrations. CONCLUSIONS: The body mass index standard deviation score was shown to have a significant influence on the peripheral volume of distribution and the inter-compartmental clearance of midazolam. The current dosing guidelines for status epilepticus, where the midazolam dose is adjusted to total body weight or age, may lead to supra- and sub-therapeutic plasma concentrations, respectively, in adolescents with obesity. TRIAL REGISTRATION: EudraCT: 2014-004554-34.

Absorption of Bupivacaine after Administration of a Lozenge as Topical Treatment for Pain from Oral Mucositis.

The aim was to investigate systemic exposure after administration of a novel bupivacaine lozenge in healthy individuals with normal mucosa and in head and neck cancer (HNC) patients with oral mucositis. A lozenge containing 5, 10, 25 and 50 mg bupivacaine, respectively, was administered as single dose to 10 healthy individuals, and a lozenge containing 25 mg bupivacaine was administered as single dose to 10 HNC patients with oral mucositis and as multiple doses to five patients with HNC. Blood samples were collected for 6 hr from the healthy individuals and 3 hr from the patients with HNC, respectively, after administration. The plasma concentration-time profiles of bupivacaine were fitted to pharmacokinetic models using nonlinear mixed-effects modelling, evaluating demographics and health status as covariates. The population pharmacokinetics (PK) of bupivacaine lozenge was best described by a two-compartment distribution model with absorption transit compartments. All the observed plasma concentrations were well below the bupivacaine concentrations (2000-2250 ng/ml) which have caused toxic symptoms. The PK model suggested that relative bioavailability was two times higher in HNC patients with oral mucositis grade 1-2 and three times higher in HNC patients with oral mucositis grade 3-4 than in the healthy individuals. Simulations showed that the plasma concentrations would be below the toxic limit after repeated dosing every second hour with 25 mg bupivacaine for five days. The 25-mg bupivacaine lozenges were safe without systemic toxic levels of bupivacaine or risk of side effects. Based on PK simulations of repeated doses of 25 mg every two hours for 16 hr a day, the lozenges can be administered with minimum risk of exceeding the toxic limit.

Pharmacokinetic models of morphine and its metabolites in neonates:: Systematic comparisons of models from the literature, and development of a new meta-model.

Morphine is commonly used for pain management in preterm neonates. The aims of this study were to compare published models of neonatal pharmacokinetics of morphine and its metabolites with a new dataset, and to combine the characteristics of the best predictive models to design a meta-model for morphine and its metabolites in preterm neonates. Moreover, the concentration-analgesia relationship for morphine in this clinical setting was also investigated. A population of 30 preterm neonates (gestational age: 23-32weeks) received a loading dose of morphine (50-100µg/kg), followed by a continuous infusion (5-10µg/kg/h) until analgesia was no longer required. Pain was assessed using the Premature Infant Pain Profile. Five published population models were compared using numerical and graphical tests of goodness-of-fit and predictive performance. Population modelling was conducted using NONMEM® and the $PRIOR subroutine to describe the time-course of plasma concentrations of morphine, morphine-3-glucuronide, and morphine-6-glucuronide, and the concentration-analgesia relationship for morphine. No published model adequately described morphine concentrations in this new dataset. Previously published population pharmacokinetic models of morphine, morphine-3-glucuronide, and morphine-6-glucuronide were combined into a meta-model. The meta-model provided an adequate description of the time-course of morphine and the concentrations of its metabolites in preterm neonates. Allometric weight scaling was applied to all clearance and volume terms. Maturation of morphine clearance was described as a function of postmenstrual age, while maturation of metabolite elimination was described as a function of postnatal age. A clear relationship between morphine concentrations and pain score was not established.

A review of morphine and morphine-6-glucuronide's pharmacokinetic-pharmacodynamic relationships in experimental and clinical pain.

Morphine is a widely used opioid for treatment of moderate to severe pain, but large interindividual variability in patient response and no clear guidance on how to optimise morphine dosage regimen complicates treatment strategy for clinicians. Population pharmacokinetic-pharmacodynamic models can be used to quantify dose-response relationships for the population as well as interindividual and interoccasion variability. Additionally, relevant covariates for population subgroups that deviate from the typical population can be identified and help clinicians in dose optimisation. This review provides a detailed overview of the published human population pharmacokinetic-pharmacodynamic studies for morphine analgesia in addition to basic drug disposition and pharmacological properties of morphine and its analgesic active metabolite, morphine-6-glucuronide, that may help identify future covariates. Furthermore, based on simulations from key pharmacokinetic-pharmacodynamic models, the contribution of morphine-6-glucuronide to the analgesic response in patients with renal insufficiency was investigated. Simulations were also used to examine the impact of effect-site equilibration half-life on time course of response. Lastly, the impact of study design on the likelihood of determining accurate pharmacodynamic parameters for morphine response was evaluated.

Modelling concentration-analgesia relationships for morphine to evaluate experimental pain models.

The aim of this study was to develop population pharmacokinetic-pharmacodynamic models for morphine in experimental pain induced by skin heat and muscle pressure, and to evaluate the experimental pain models with regard to assessment of morphine pharmacodynamics. In a randomised, double-blind, placebo-controlled, crossover study, 39 healthy volunteers received an oral dose of 30mg morphine hydrochloride or placebo. Non-linear mixed effects modelling was used to describe the plasma concentrations of morphine and metabolites, and the analgesic effect of morphine on experimental pain in skin and muscle. Baseline pain metrics varied between individuals and occasions, and were described with interindividual and interoccasion variability. Placebo-response did not change with time. For both pain metrics, morphine effect was proportional to baseline pain and was described with a linear model with interindividual variability on drug effect slope and linked to an effect compartment for muscle pressure. The models indicate that a steady-state morphine concentration of 21ng/ml causes 33% and 0.84% increases in stimulus intensity from baseline for muscle pressure and skin heat, respectively. The population pharmacokinetic-pharmacodynamic models developed in this study indicate that mechanical stimulation of muscle is a more clinically relevant pain stimulus for the assessment of morphine pharmacodynamics than thermal stimulation of skin.

Sensitivity of quantitative sensory models to morphine analgesia in humans.

INTRODUCTION: Opioid analgesia can be explored with quantitative sensory testing, but most investigations have used models of phasic pain, and such brief stimuli may be limited in the ability to faithfully simulate natural and clinical painful experiences. Therefore, identification of appropriate experimental pain models is critical for our understanding of opioid effects with the potential to improve treatment. OBJECTIVES: The aim was to explore and compare various pain models to morphine analgesia in healthy volunteers. METHODS: The study was a double-blind, randomized, two-way crossover study. Thirty-nine healthy participants were included and received morphine 30 mg (2 mg/mL) as oral solution or placebo. To cover both tonic and phasic stimulations, a comprehensive multi-modal, multi-tissue pain-testing program was performed. RESULTS: Tonic experimental pain models were sensitive to morphine analgesia compared to placebo: muscle pressure (F=4.87, P=0.03), bone pressure (F=3.98, P=0.05), rectal pressure (F=4.25, P=0.04), and the cold pressor test (F=25.3, P<0.001). Compared to placebo, morphine increased tolerance to muscle stimulation by 14.07%; bone stimulation by 9.72%; rectal mechanical stimulation by 20.40%, and reduced pain reported during the cold pressor test by 9.14%. In contrast, the more phasic experimental pain models were not sensitive to morphine analgesia: skin heat, rectal electrical stimulation, or rectal heat stimulation (all P>0.05). CONCLUSION: Pain models with deep tonic stimulation including C fiber activation and and/or endogenous pain modulation were more sensitive to morphine analgesia. To avoid false negative results in future studies, we recommend inclusion of reproducible tonic pain models in deep tissues, mimicking clinical pain to a higher degree.