Semi-mechanistic Model Research Articles

Predicting Drug-Drug Interactions Involving Rifampicin Using a Semi-mechanistic Hepatic Compartmental Model.

To develop a semi-mechanistic hepatic compartmental model to predict the effects of rifampicin, a known inducer of CYP3A4 enzyme, on the metabolism of five drugs, in the hope of informing dose adjustments to avoid potential drug-drug interactions. A search was conducted for DDI studies on the interactions between rifampicin and CYP substrates that met specific criteria, including the availability of plasma concentration-time profiles, physical and absorption parameters, pharmacokinetic parameters, and the use of healthy subjects at therapeutic doses. The semi-mechanistic model utilized in this study was improved from its predecessors, incorporating additional parameters such as population data (specifically for Chinese and Caucasians), virtual individuals, gender distribution, age range, dosing time points, and coefficients of variation. Optimal parameters were identified for our semi-mechanistic model by validating it with clinical data, resulting in a maximum difference of approximately 2-fold between simulated and observed values. PK data of healthy subjects were used for most CYP3A4 substrates, except for gilteritinib, which showed no significant difference between patients and healthy subjects. Dose adjustment of gilteritinib co-administered with rifampicin required a 3-fold increase of the initial dose, while other substrates were further tuned to achieve the desired drug exposure. The pharmacokinetic parameters AUCR and CmaxR of drugs metabolized by CYP3A4, when influenced by Rifampicin, were predicted by the semi-mechanistic model to be approximately twice the empirically observed values, which suggests that the semi-mechanistic model was able to reasonably simulate the effect. The doses of four drugs adjusted via simulation to reduce rifampicin interaction.

Abstract 7176: Optimization of T cell co-stimulatory agonists: A semi-mechanistic PKPD model integrating drug properties and tumor-immune interactions

Abstract Background: Despite the success of some immune checkpoint inhibitors in the treatment of cancer, T-cell co-stimulation has faced limited clinical success due to the intricacies around the optimal engagement of agonistic antibodies to co-stimulatory receptors, and the absence of biomarkers for patient selection. Understanding agonist-binding characteristics, pharmacokinetics, T cell subpopulation activation and differentiation kinetics, T cell life span in the tumor microenvironment, and the impact of T cell/tumor interactions on tumor growth kinetics is crucial for the positive clinical outcomes of T cell agonists. Currently there is not a suitable model to describe these complex interactions. Methods: We investigated the pharmacokinetic (PK), pharmacodynamics (PD), and anti-tumor efficacy of an anti-TNFR2 agonist, HFB3-1, and an anti-OX40 agonist, HFB10-1, in syngeneic tumor models. Flow cytometry was employed to profile post-treatment tumor infiltrating lymphocytes (TILs) in dissociated tumor tissues. In vitro binding affinity was determined using Biolayer Interferometry. A semi-mechanistic model, integrating PK, tumor growth, and immune interaction networks (interactions among Teff, Treg, and tumor cells), was developed in Berkeley Madonna (version 10.5.1). This was subsequently adapted to a human tumor growth model with T cell interaction networks to assess the impact of doses and dosing regimens on tumor growth inhibition. Results: A semi-mechanistic PKPD model was successfully developed incorporating drug properties including binding affinity, agonistic effect, ADCC activity, and pharmacokinetics, and tumor growth characteristics including tumor growth kinetics, TIL composition, interaction, proliferation, and turnover. This model was applied to investigate the effect of FcγR, dose, and dosing frequency on in vivo efficacy. Following incorporation of human PK and human tumor microenvironment properties into the model, tumor growth characteristics were simulated under varying human dosing regimens. Simulation results indicated that lower and less frequent doses optimize T cell co-stimulation for optimal anti-tumor activity. Conclusion: Our semi-mechanistic PKPD model elucidates the intricate interactions between T cell co-stimulatory agonists and the tumor-immune cells in the tumor microenvironment. The model advocates for pulsatile agonism as the optimal anti-tumor activity for T cell co-stimulation agents. This model serves as a valuable tool for guiding dose and dose regimen optimization for T cell co-stimulatory agonists in clinical development. Citation Format: Jinping Gan, William Hedrich, Yun-Yueh Lu, Wenhua Xu, Robert H. Andtbacka, Francisco Adrian, Liang Schweizer. Optimization of T cell co-stimulatory agonists: A semi-mechanistic PKPD model integrating drug properties and tumor-immune interactions [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 7176.

A hierarchical Bayesian network-based semi-mechanistic model for handling data variabilities in dynamical process systems

This work presents a hierarchical Bayesian network-based model for handling dynamic nature, data variabilities, and information fusion in a dynamical process system. The overall dynamics has been captured by physics-based models while uncertainty and local variabilities in the model parameters have been accounted for by using a hierarchical Bayesian network formalism. The resulting semi-mechanistic model has been investigated in two cases, a simulated case of an enzymatic reaction system and a real-world scenario of pandemic risk forecasting. The first case study on the enzymatic reaction system shows effectiveness of the proposed semi-mechanistic framework in a closed system with limited data from different sources. The second case illustrates scope of the proposed semi-mechanistic framework to open systems by investigating transmission of disease under evolving conditions of the advent of multiple waves of a pandemic outbreak. The findings have been presented in terms of the most probable values and credible intervals that account for data variabilities in different scenarios. The model is benchmarked and validated for the fatality forecast of the Coronavirus Disease of 2019 by the observed data and benchmark models reported by the Center for Disease Prevention and Control of the United States [https://covid.cdc.gov/covid-data-tracker/#datatracker-home].

Population pharmacokinetics, enzyme occupancy, and pharmacodynamic modeling of soticlestat in patients with developmental and epileptic encephalopathies.

Soticlestat (TAK-935) is a first-in-class, selective inhibitor of cholesterol 24-hydroxylase (CH24H) under phase III development for the treatment of the developmental and epileptic encephalopathies (DEEs), Dravet syndrome (DS), and Lennox-Gastaut syndrome (LGS). A previous model characterized the pharmacokinetics (PKs), CH24H enzyme occupancy (EO), and pharmacodynamics (PDs) of soticlestat in healthy volunteers. The present study extended this original model for patients with DEEs and investigated sources of variability. Model-based simulations were carried out to optimize dosing strategies for use in clinical trials. Data from eight phase I and II trials of healthy volunteers or patients with DEEs receiving oral soticlestat 15-1350 mg were included, encompassing 218 individuals for population PK (PopPK) analyses and 306 individuals for PK/PD analyses. Dosing strategies were identified through model-based simulations. The final mixed-effect PopPK/EO/PD model consisted of a two-compartment PK model and an effect-site compartment in the PK/EO model; soticlestat concentrations at the effect site were linked to 24S-hydroxycholesterol plasma concentrations using a semimechanistic inhibitory indirect response model. Covariates were included to account for sources of variability. Pediatric dosing strategies were developed for four body weight bands (10 to <15, 15 to <30, 30 to <45, and 45-100 kg) to account for covariate effects by body weight. The final PopPK and PK/EO/PD models accurately described PK, EO, and PD profiles of soticlestat in healthy volunteers and patients with DEEs. Covariate analyses and model-based simulations facilitated optimization of phase III trial dosing strategies for patients with DS or LGS.

Population pharmacokinetics of the GIP/GLP receptor agonist tirzepatide.

Tirzepatide is a first-in-class glucose-dependent insulinotropic polypeptide and glucagon-like peptide-1 receptor agonist approved as for the treatment of type 2 diabetes mellitus. A population-based pharmacokinetic (PK) model was developed from 19 pooled studies. Tirzepatide pharmacokinetics were well-described by a two-compartment model with first order absorption and elimination. The tirzepatide population PK model utilized a semimechanistic allometry model to describe the relationship between body size and tirzepatide PK. The half-life of tirzepatide was ~5 days and enabled sustained exposure with once-weekly subcutaneous dosing. The covariate analysis suggested that adjustment of the dose regimen based on demographics or subpopulations was unnecessary. The tirzepatide PK model can be used to predict tirzepatide exposure for various scenarios or populations.

The public health impact and cost-effectiveness of the R21/Matrix-M malaria vaccine: a mathematical modelling study

The R21/Matrix-M vaccine has demonstrated high efficacy against Plasmodium falciparum clinical malaria in children in sub-Saharan Africa. Using trial data, we aimed to estimate the public health impact and cost-effectiveness of vaccine introduction across sub-Saharan Africa. We fitted a semi-mechanistic model of the relationship between anti-circumsporozoite protein antibody titres and vaccine efficacy to data from 3 years of follow-up in the phase 2b trial of R21/Matrix-M in Nanoro, Burkina Faso. We validated the model by comparing predicted vaccine efficacy to that observed over 12-18 months in the phase 3 trial. Integrating this framework within a mathematical transmission model, we estimated the cases, malaria deaths, and disability-adjusted life-years (DALYs) averted and cost-effectiveness over a 15-year time horizon across a range of transmission settings in sub-Saharan Africa. Cost-effectiveness was estimated incorporating the cost of vaccine introduction (dose, consumables, and delivery) relative to existing interventions at baseline. We report estimates at a median of 20% parasite prevalence in children aged 2-10 years (PfPR2-10) and ranges from 3% to 65% PfPR2-10. Anti-circumsporozoite protein antibody titres were found to satisfy the criteria for a surrogate of protection for vaccine efficacy against clinical malaria. Age-based implementation of a four-dose regimen of R21/Matrix-M vaccine was estimated to avert 181 825 (range 38 815-333 491) clinical cases per 100 000 fully vaccinated children in perennial settings and 202 017 (29 868-405 702) clinical cases per 100 000 fully vaccinated children in seasonal settings. Similar estimates were obtained for seasonal or hybrid implementation. Under an assumed vaccine dose price of US$3, the incremental cost per clinical case averted was $7 (range 4-48) in perennial settings and $6 (3-63) in seasonal settings and the incremental cost per DALY averted was $34 (29-139) in perennial settings and $30 (22-172) in seasonal settings, with lower cost-effectiveness ratios in settings with higher PfPR2-10. Introduction of the R21/Matrix-M malaria vaccine could have a substantial public health benefit across sub-Saharan Africa. The Wellcome Trust, the Bill & Melinda Gates Foundation, the UK Medical Research Council, the European and Developing Countries Clinical Trials Partnership 2 and 3, the NIHR Oxford Biomedical Research Centre, and the Serum Institute of India, Open Philanthropy.

Numerical Model of Simultaneous Multi-Regime Boiling Quenching of Metals

This work presents a heat transfer and boiling model that computes the evolution of the temperature field in a representative steel workpiece quenched from 850 or 930 °C by immersion in water flowing at average velocities of 0.2 or 0.6 m/s, respectively. Under these conditions, all three boiling regimes were present during cooling: stable vapor film, nucleate boiling, and single-phase convection. The model was based on the numerical solution of the heat conduction equation coupled to the solution of the energy and momentum equations for water. The mixture phase approach was adopted using the Lee model to compute the rates of water evaporation–condensation. Heat flux at the wall was calculated for all regimes using a single semi-mechanistic model. Therefore, the evolution of boiling regimes at every position on the wall surface was automatically determined. Predictions were validated using laboratory results, namely: (a) videorecording the upward motion of the wetting front along the workpiece wall surface; and (b) cooling curves obtained with embedded thermocouples in the steel probe. Wall heat flux calculations were used to determine the importance of the simultaneous presence of all three boiling regimes on the heat flux distribution. It was found that this simultaneous presence leads to high heat flux variations that should be avoided in production lines. In addition, it was determined that the corresponding inverse heat conduction problem to estimate the active heat transfer boundary condition must be set-up for 2D heat flow.

Modelling the effect of context-specific greenhouse gas and nitrogen emission mitigation options in key European dairy farming systems

Understanding the environmental consequences associated with dairy cattle production systems is crucial for the implementation of targeted strategies for emission reduction. However, few studies have modelled the effect of tailored emission mitigation options across key European dairy production systems. Here, we assess the single and combined effect of six emission mitigation practises on selected case studies across Europe through the Sustainable and Integrated Management System for Dairy Production model. This semi-mechanistic model accounts for the interacting flows from a whole-farm perspective simulating the environmental losses in response to different management strategies and site-specific conditions. The results show how reducing the crude protein content of the purchased fraction of the diet was an adequate strategy to reduce the greenhouse gas and nitrogen emission intensity in all systems. Furthermore, implementing an anaerobic digestion plant reduced the greenhouse gas emissions in all tested case studies while increasing the nitrogen emissions intensity, particularly when slurry was applied using broadcast. Regarding the productivity increase, contrasting effects were observed amongst the case studies modelled. Moreover, shallow slurry injection effectively mitigated the intensity of nitrogen losses from the fields due to strong reductions in ammonia volatilisation. When substituting urea with ammonium nitrate as mineral fertiliser, site-specific conditions affected the mitigation potential observed, discouraging its application on sandy-loam soils. Rigid slurry covers effectively reduced the storage-related nitrogen emissions intensity while showing a minor effect on total greenhouse gas emission intensity. In addition, our results provide novel evidence regarding the advantages of cumulative implementation of adapted mitigation options to offset the negative trade-offs of single-option applications (i.e. slurry covers or anaerobic digestion and slurry injection). Through this study, we contribute to a better understanding of the effect of emission mitigation options across dairy production systems in Europe, thus facilitating the adoption of tailored and context-specific emission reduction strategies.

Near-term forecasting of Covid-19 cases and hospitalisations in Aotearoa New Zealand.

Near-term forecasting of infectious disease incidence and consequent demand for acute healthcare services can support capacity planning and public health responses. Despite well-developed scenario modelling to support the Covid-19 response, Aotearoa New Zealand lacks advanced infectious disease forecasting capacity. We develop a model using Aotearoa New Zealand's unique Covid-19 data streams to predict reported Covid-19 cases, hospital admissions and hospital occupancy. The method combines a semi-mechanistic model for disease transmission to predict cases with Gaussian process regression models to predict the fraction of reported cases that will require hospital treatment. We evaluate forecast performance against out-of-sample data over the period from 2 October 2022 to 23 July 2023. Our results show that forecast performance is reasonably good over a 1-3 week time horizon, although generally deteriorates as the time horizon is lengthened. The model has been operationalised to provide weekly national and regional forecasts in real-time. This study is an important step towards development of more sophisticated situational awareness and infectious disease forecasting tools in Aotearoa New Zealand.

Modeling transpiration using solar-induced chlorophyll fluorescence and photochemical reflectance index synergistically in a closed-canopy winter wheat ecosystem

The joint use of solar-induced chlorophyll fluorescence (SIF) and the photochemical reflectance index (PRI) has been shown to improve gross primary productivity (GPP) estimation across various plant functional types. However, the utility of PRI in combination with SIF for transpiration (T) estimation has not yet been explored. Additionally, current SIF-driven transpiration models including linear models, semi-mechanical models (combination of canopy conductance, gC, derived from a SIF and vapor pressure deficit, VPD, driven linear model with the Penman-Monteith model), and hybrid models (combination of gC derived from a SIF and VPD driven machine learning model with the Penman-Monteith model) have rarely been mutually assessed. Based on concurrent remotely sensed SIF and PRI, and eddy covariance flux measurements during one growing season for a winter wheat ecosystem in northern China, we investigated the mediating effect of PRI on SIF-driven T estimation under different VPD conditions and compared the performance of linear, semi-mechanical, and hybrid models in estimating T. Our results showed that the mediating effect of PRI on T described in the SIF-driven linear, semi-mechanistic, and hybrid models was significant under high VPD conditions rather than under low VPD conditions. Specifically, based on T partitioned using an underlying water use efficiency method as a benchmark, the root mean square error (RMSE) value of the PRI-mediated linear, semi-mechanistic, and hybrid models was 28.01 W/m2, 22.25 W/m2, and 28.71 W/m2 lower, respectively, than those of the corresponding models without PRI when VPD was >1.5 kPa. Based on T partitioned using a transpiration estimation algorithm as a benchmark, these three models also exhibited a significant reduction in RMSE under high VPD conditions after considering PRI. The main rationale behind the PRI improvement is that PRI can track photosynthetic dynamics under high VPD conditions. Based on the simulation results of the Soil-Canopy-Observation of Photosynthesis and Energy fluxes (SCOPE) model, PRI can serve as an indicator for non-photochemical quenching (NPQ) within this ecosystem. Consequently, PRI can enhance the capability of SIF to characterize the energy dissipation of photosynthetically active radiation and help SIF to yield more accurate information on GPP and gc under high VPD conditions. Finally, the order of model performance in estimating T was generally hybrid model > semi-mechanistic model > linear model. Our findings show the effectiveness of PRI for improving SIF-driven transpiration estimation under high VPD conditions and provide a new hybrid model for estimating T from SIF.

VPD-based models of dead fine fuel moisture provide best estimates in a global dataset

Dead fine fuel moisture content (FM) is one of the most important determinants of fire behavior. Fire scientists have attempted to effectively estimate FM for nearly a century, but we are still lacking broad scale evaluations of the different approaches for prediction. Here we tackle this problem by taking advantage or a recently compiled global fire behavior database (BONFIRE) gathering 1603 records of 1h (i.e., <6 mm diameter or thickness) dead fuel moisture content from measurements before experimental fires. We compared the results of models routinely used by different agencies worldwide, empirical models, semi-mechanistic models and also non-linear and machine learning approaches based on either temperature and relative humidity or vapor pressure deficit (VPD). A semi-mechanistic model based on VPD showed the best performance across all FM ranges and a historical model developed in Australia (MK5) was additionally recommended for low fuel moisture estimations. We also observed significant differences in FM dynamics between vegetation types with FM in grasslands more responsive to changes in atmospheric dryness than woody ecosystems. The addition of computational complexity through machine learning is not recommended since the gain in model fit is small relative to the increase in complexity. Future research efforts should concentrate on predictions at low FM (<10 %) as this is the range most significant for fire behavior and where the poorest model performance was observed. Model predictions are available from https://hcfm.shinyapps.io/shinyfmd/.

Comparative impact assessment of COVID-19 policy interventions in five South Asian countries using reported and estimated unreported death counts during 2020-2021.

There has been raging discussion and debate around the quality of COVID death data in South Asia. According to WHO, of the 5.5 million reported COVID-19 deaths from 2020-2021, 0.57 million (10%) were contributed by five low and middle income countries (LMIC) countries in the Global South: India, Pakistan, Bangladesh, Sri Lanka and Nepal. However, a number of excess death estimates show that the actual death toll from COVID-19 is significantly higher than the reported number of deaths. For example, the IHME and WHO both project around 14.9 million total deaths, of which 4.5-5.5 million were attributed to these five countries in 2020-2021. We focus our gaze on the COVID-19 performance of these five countries where 23.5% of the world population lives in 2020 and 2021, via a counterfactual lens and ask, to what extent the mortality of one LMIC would have been affected if it adopted the pandemic policies of another, similar country? We use a Bayesian semi-mechanistic model developed by Mishra et al. (2021) to compare both the reported and estimated total death tolls by permuting the time-varying reproduction number (Rt) across these countries over a similar time period. Our analysis shows that, in the first half of 2021, mortality in India in terms of reported deaths could have been reduced to 96 and 102 deaths per million compared to actual 170 reported deaths per million had it adopted the policies of Nepal and Pakistan respectively. In terms of total deaths, India could have averted 481 and 466 deaths per million had it adopted the policies of Bangladesh and Pakistan. On the other hand, India had a lower number of reported COVID-19 deaths per million (48 deaths per million) and a lower estimated total deaths per million (80 deaths per million) in the second half of 2021, and LMICs other than Pakistan would have lower reported mortality had they followed India's strategy. The gap between the reported and estimated total deaths highlights the varying level and extent of under-reporting of deaths across the subcontinent, and that model estimates are contingent on accuracy of the death data. Our analysis shows the importance of timely public health intervention and vaccines for lowering mortality and the need for better coverage infrastructure for the death registration system in LMICs.

Influence of a Short Course of Ritonavir Used as Booster in Antiviral Therapies Against SARS-CoV-2 on the Exposure of Atorvastatin and Rosuvastatin.

Early antiviral treatment with nirmatrelvir/ritonavir is recommended for SARS-CoV-2-infected patients at high risk for severe courses. Such patients are usually chronically ill and susceptible to adverse drug interactions caused by ritonavir. We investigated the interactions of short-term low-dose ritonavir therapy with atorvastatin and rosuvastatin, two statins commonly used in this population. We assessed exposure changes (area under the concentration-time curve (AUC∞) and maximum concentration (Cmax)) of a single dose of 10 mg atorvastatin and 10 mg rosuvastatin before and on the fifth day of ritonavir treatment (2 × 100 mg/day) in healthy volunteers and developed a semi-mechanistic pharmacokinetic model to estimate dose adjustment of atorvastatin during ritonavir treatment. By the fifth day of ritonavir treatment, the AUC∞ of atorvastatin increased 4.76-fold and Cmax 3.78-fold, and concurrently, the concentration of atorvastatin metabolites decreased to values below the lower limit of quantification. Pharmacokinetic modelling indicated that a stepwise reduction in atorvastatin dose during ritonavir treatment with a stepwise increase up to 4 days after ritonavir discontinuation can keep atorvastatin exposure within safe and effective margins. Rosuvastatin pharmacokinetics were only mildly modified; ritonavir significantly increased the Cmax 1.94-fold, while AUC∞ was unchanged. Atorvastatin doses should likely be adjusted during nirmatrelvir/ritonavir treatment. For patients on a 20-mg dose, we recommend half of the original dose. In patients taking 40 mg or more, a quarter of the dose should be taken until 2 days after discontinuation of nirmatrelvir/ritonavir. Patients receiving rosuvastatin do not need to change their treatment regimen. EudraCT number: 2021-006634-39. DRKS00027838.

Characterizing a design space for a twin-screw wet granulation process: A case study of extended-release tablets

Twin-screw wet granulation is an emerging continuous manufacturing technology for solid oral dosage forms. This technology has been successfully employed for the commercial manufacture of immediate-released tablets. However, the higher polymer content in extended-release (ER) formulations may present challenges in developing and operating within a desired design space. The work described here used a systematic approach for defining the optimum design space by understanding the effects of the screw design, operating parameters, and their interactions on the critical characteristics of granules and ER tablets. The impacts of screw speed, powder feeding rate, and the number of kneading (KEs) and sizing elements on granules and tablets characteristics were investigated by employing a definitive screening design. A semi-mechanistic model was used to calculate the residence time distribution parameters and validated using the tracers. The results showed that an increase in screw speed decreased the mean residence time of the material within the barrel, while an increase in the powder feeding rate or number of KEs did the opposite and increased the barrel residence time. Screw design and operating parameters affected the flow and bulk characteristics of granules. The screw speed was the most significant factor impacting the tablet's breaking strength. The dissolution profiles revealed that granule characteristics mainly influenced the early phase of drug release. This study demonstrated that a simultaneous optimization of both operating and screw design parameters was beneficial in producing ER granules and tablets of desired performance characteristics while mitigating any failure risks, such as swelling during processing.

Population Pharmacokinetic-Pharmacodynamic Modeling of Neutrophil and Platelet Count for Lower-Intensity Therapy of CPX-351 Combined with Venetoclax in Acute Myeloid Leukemia

Background: Improved survival outcomes and a comparable safety profile with CPX-351 (a dual-drug liposomal encapsulation of daunorubicin and cytarabine in a synergistic 1:5 molar ratio) vs conventional 7+3 chemotherapy in a pivotal phase 3 trial led to the approval of CPX-351 for newly diagnosed therapy-related acute myeloid leukemia (tAML) or AML with myelodysplasia-related changes (AML MRC) in adults and pediatric (aged ≥1 year) patients in the United States and in adults in Europe. The approved induction dose is 1-2 cycles of CPX-351 100 units/m 2 (daunorubicin 44 mg/m 2 and cytarabine 100 mg/m 2) as a 90-minute intravenous infusion on days 1, 3, and 5 (days 1 and 3 for second induction). In this study, we developed a population pharmacokinetic-pharmacodynamic (PK-PD) model of absolute neutrophil count (ANC) and platelet count to determine the optimal dose for a lower-intensity therapy (LiT) of CPX-351 combined with venetoclax. Methods: The CPX-351 population PK-PD analysis was based on data collected from 3 clinical studies (phase 1 study [NCT00389428; J Clin Oncol 2011; 29(8):979-85];phase 2 study [NCT02238925; Cancer Chemother Pharmacol 2019; 84(1):163-73];phase 3 study [NCT01696084; J Clin Oncol 2018; 36(26):2684-92]) in patients with newly diagnosed high-risk/secondary AML. Venetoclax population PK-PD analysis was based on data from a comprehensive safety analysis of venetoclax monotherapy ( Clin Cancer Res 2018; 24(18):4371-9). ANC and platelet data used in the analysis were from 94 patients for CPX-351 and 272 patients for venetoclax. Generalized semi-mechanistic population PK-PD analyses were performed using non-linear mixed-effects modeling to characterize the time-course of ANC and platelet count following administration of CPX-351 and venetoclax monotherapy. The majority of system parameters were shared across the 2 therapies and a limited number of treatment-specific parameters were estimated by fitting CPX-351 and venetoclax PK time-courses simultaneously. An R Shiny app was developed for the simulation of different dosing regimens with favorable safety profiles to guide optimal dose selection for LiT of CPX-351 combined with venetoclax, assuming the 2 drug effects on ANC and platelet count were additive. Results: The generalized PK-PD model included multi-compartment transit models for ANC and platelet count with separate semi-mechanistic compartments to represent proliferating, maturing, and circulating cells. The model fitted CPX-351 and venetoclax PK-PD data with shared system parameters except for the drug-specific parameter IC 50 (concentration producing 50% of maximal inhibition of cell proliferation). For ANC, the estimated IC 50 values for CPX-351 and venetoclax were 295 μM (relative standard error [RSE]: 8.96%) and 120 μM (RSE: 18.0%), respectively; the shared mean transit time (MTT) estimate for ANC maturation was 66.6 h (RSE: 4.95%), and shared maximal inhibition of cell proliferation (I max) was fixed to 1. For platelet count, the estimated IC 50 values for CPX-351 and venetoclax were 0.0109 μM (RSE: 50.2%) and 16.5 μM (RSE: 8.73%); the shared MTT estimate for platelet maturation was 87.2 h (RSE: 9.01%) and shared I max estimate was 0.313 (RSE: 11.2%). Simulations with different doses of CPX-351 in combination with 400 mg venetoclax were performed. The simulated mean nadir ANC were 2.1, 1.5, 1.1, and 0.6 × 10 9/L for CPX-351 doses of 20, 40, 60, and 100 units/m 2, respectively, when administered on days 1 and 3 and combined with venetoclax 400 mg daily for 28 days. The simulated mean nadir platelet counts were 31.0, 30.3, 30.0, and 29.7 × 10 9/L for CPX-351 doses of 20, 40, 60, and 100 units/m 2, respectively, when administered on days 1 and 3 and combined with venetoclax 400 mg daily for 28 days. Model-based simulated outcomes of ANC and platelet count were consistent with the observed trial data. C onclusions: A generalized semi-mechanistic population PK-PD model was developed to adequately describe the time-courses of ANC and platelet counts in patients with AML following the administration of CPX-351 and venetoclax monotherapy and was used to infer the impact on ANC and platelet count when given in combination. Model-based simulations supported initial dose selection in the phase 1b study of combined CPX-351 LiT plus venetoclax as first-line treatment for patients with AML who are unfit for intensive chemotherapy (NCT04038437).

HPA axis differential flatness and Liouvillian study for higher resiliency investigations

Abstract In this paper, we study several existing quantitative models of the hypothalamic–pituitary–adrenal (HPA) axis from a control systems theory viewpoint, that is, we suppose that we can act on the dynamics of the HPA axis throughout some parameters, which are the system inputs. In particular, we will focus on flatness and Liouvillian properties of the considered control systems of the HPA axis. We first study the minimal three-dimensional model of Bangsgaard and Ottesen (2017, Math. Biosci., 287:24–35) and the semi-mechanistic four-dimensional model of Gupta et al. (2007, Theor. Biol. Medical Model., 4(1):8) which are shown to be flat, and then, we consider the more involved and important model proposed in Rao &amp; Androulakis (2019, Sci. Rep., 9(1):11212; 2020, IFAC-PapersOnLine, 53(2):15858–15863), with seven states, for which we prove that for the nominal values of the parameters involved in the model, flatness no longer holds. The more involved model satisfies however a similar but weaker property than flatness: it is a Liouvillian system.

Predicting nitrogen excretion of cattle kept under tropical and subtropical conditions using semimechanistic models.

The present study aims at evaluating whether current semimechanistic models developed for temperate cattle systems can be adopted for cattle under (sub-) tropical husbandry systems to adequately (accurately and precisely) predict total nitrogen (TN), urine nitrogen (UN), faecal nitrogen (FN) excretion and its partition into different FN fractions. Selected models were built based on the feeding recommendations for ruminants of the British (Model A), German (Model G) and French (INRA; Model I) system. Model evaluation was conducted using eight nitrogen balance studies performed in El Salvador, Kenya and Peru (n = 392 individual observations including lactating cows, heifers and steers). Concordance correlation coefficient, root mean square errors (RMSE), and mean biases were estimated to evaluate the models' adequacy in predicting nitrogen excretion. Input variables causing greatest variation in nitrogen excretion prediction were identified by a sensitivity analysis and adjusted. Model G was able to adequately (i.e., RMSE of <25% of observed mean, systematic error of <5% of the mean square error) predict TN excretion through a compensation between overestimation of UN excretion and underestimation of FN excretion. None of the models were able to adequately predict UN, FN, and different FN fractions. Model I adequately predicted FN (RMSE = 18%) when duodenal microbial crude protein flow was increased, and the intercept used to predict FN excretion was reduced from 4.30 to 3.82 g of nitrogen per kilogram of dry matter intake. These adjustments, however, were not sufficient to predict adequately UN excretion (RMSE = 38%), individual FN fractions (RMSE > 56%), and TN (RMSE = 22%) excretion, by Model I.

Population pharmacokinetics of mirvetuximab soravtansine in patients with folate receptor-α positive ovarian cancer: The antibody-drug conjugate, payload and metabolite.

Mirvetuximab soravtansine is a first-in-class antibody-drug conjugate recently approved for the treatment of folate receptor-α positive ovarian cancer. The aim of this study was to develop a population pharmacokinetic model to describe the concentration-time profiles of mirvetuximab soravtansine, the payload (DM4) and a metabolite (S-methyl-DM4). Mirvetuximab soravtansine was administered intravenously from 0.15 to 7mg/kg to 543 patients with predominantly platinum-resistant ovarian cancer in 3 clinical studies, and the plasma drug concentrations were analysed using a nonlinear mixed-effects modelling approach. Stepwise covariate modelling was performed to identify covariates. We developed a semi-mechanistic population pharmacokinetic model that included linear and nonlinear routes for the elimination of mirvetuximab soravtansine and a target compartment for the formation and disposition of the payload and metabolite in tumour cells. The clearance and volume of the central compartment were 0.0153 L/h and 2.63 L for mirvetuximab soravtansine, 8.83 L/h and 3.67 L for DM4, and 2.04 L/h and 6.3L for S-methyl-DM4, respectively. Body weight, serum albumin and age were identified as statistically significant covariates. Exposures in patients with renal or hepatic impairment and who used concomitant cytochrome P450 (CYP) 3A4 inhibitors were estimated. There is no need for dose adjustment due to covariate effects for mirvetuximab soravtansine administered at the recommended dose of 6mg/kg based on adjusted ideal body weight. Dose adjustment is not required for patients with mild or moderate renal impairment, mild hepatic impairment, or when concomitant weak and moderate CYP3A4 inhibitors are used.

Explaining in-vitro to in-vivo efficacy correlations in oncology pre-clinical development via a semi-mechanistic mathematical model

In-vitro to in-vivo correlations (IVIVC), relating in-vitro parameters like IC50 to in-vivo drug exposure in plasma and tumour growth, are widely used in oncology for experimental design and dose decisions. However, they lack a deeper understanding of the underlying mechanisms. Our paper therefore focuses on linking empirical IVIVC relations for small-molecule kinase inhibitors with a semi-mechanistic tumour-growth model. We develop an approach incorporating parameters like the compound’s peak-trough ratio (PTR), Hill coefficient of in-vitro dose-response curves, and xenograft-specific properties. This leads to formulas for determining efficacious doses for tumor stasis under linear pharmacokinetics equivalent to traditional empirical IVIVC relations, but enabling more systematic analysis. Our findings reveal that in-vivo xenograft-specific parameters, specifically the growth rate (g) and decay rate (d), along with the average exposure, are generally more significant determinants of tumor stasis and effective dose than the compound’s peak-trough ratio. However, as the Hill coefficient increases, the dependency of tumor stasis on the PTR becomes more pronounced, indicating that the compound is more influenced by its maximum or trough values rather than the average exposure. Furthermore, we discuss the translation of our method to predict population dose ranges in clinical studies and propose a resistance mechanism that solely relies on specific in-vivo xenograft parameters instead of IC50 exposure coverage. In summary, our study aims to provide a more mechanistic understanding of IVIVC relations, emphasizing the importance of xenograft-specific parameters and PTR on tumor stasis.

Effectiveness of dapagliflozin as an insulin adjunct in type 1 diabetes: a semi-mechanistic exposure-response model.

Introduction: Dapagliflozin-induced improvement of glycemic control in patients with inadequately controlled type 1 diabetes (T1D) is complicated by the delicate balance between blood glucose and exogenous insulin. In this work, we developed a semi-mechanistic population exposure-response model using pooled patient-level data to characterize the joint effect of dapagliflozin and insulin on average daily glucose concentrations and glycated hemoglobin (HbA1c) levels in patients with T1D. Methods: A non-linear mixed-effects model was developed in Monolix (Lixoft, France) and R software (R Project, www.r-project.org) using pooled patient-level data from phase 2 and phase 3 trials (NCT01498185, NCT02460978, NCT02268214). Results: Because of the apparent lack of association between bolus insulin dose and glucose concentrations measured by continuous glucose monitoring the model was able to capture the quantitative link between basal, but not bolus, insulin dose and plasma glucose. Even so, this association remained flat, with a 50% decrease in the basal insulin dose from pretreatment level, resulting in ∼5% increase in glucose exposure. Therefore, dapagliflozin efficacy was not significantly affected by the insulin dose adjustment, with 24-week HbA1c reduction on 10-mg dapagliflozin treatment changing from -0.5 [95% CI: -0.55, -0.45] to -0.42 [95%CI: -0.48, -0.36] after adjustment. At the same time, the analysis revealed ∼2-fold steeper slope of glucose-HbA1c relationship in dapagliflozin-treated patients vs. control group, suggesting the presence of additional dapagliflozin treatment-related benefits, not explained by the dapagliflozin-mediated ∼4% increase in plasma hemoglobin levels. Finally, the efficacy of 5 and 10-mg doses, represented by the mean HbA1c reduction at week 24 of dapagliflozin treatment, was shown to be notably greater than the 1- and 2.5-mg doses. Discussion: This research is an attempt to deconvolute and reconstruct dapagliflozin-HbA1c dose-response relationship in T1D by accounting for the drug's action on both daily insulin dose and plasma glucose on a subject-level. While the model is able to adequately capture the observed data, it also revealed that the variability in CGM is poorly approximated by the variability in insulin dose alone. Furthermore, the slope of CGM/HbA1c relationship may differ depending on the population and treatment scenarios. As such, a deeper dive into the physiological mechanisms is required to better quantify the intricate network of glycemic response under dapagliflozin treatment.