Brief Report: Higher Fentanyl Exposures Require Higher Doses of Naloxone for Successful Reversals in a Quantitative Systems Pharmacology Model

Background: Previously, we reported on an opioid receptor quantitative systems pharmacology (QSP) model to evaluate naloxone dosing. Methods: In this study we extended our model to include higher systemic levels of fentanyl (up to 100 ng/ml) and the newly approved 8mg IN naloxone dose (equivalent to 4 mg)Results : As expected, at the lower peak fentanyl concentrations (25 ng/ml and 50 ng/ml), the simulations predicted that 2 mg, 4 mg, 5 mg, and 10 mg IM doses of naloxone displaced fentanyl and reached below the 50% receptor occupancy within 10 minutes. However, at the concentration of 75 ng/ml, the simulation predicted that the 2 mg dose of naloxone failed to reach below the 50% occupancy within 10 minutes. Interestingly, at the highest peak concentration of fentanyl studied (100 ng/ml), the model predicted that the 4 mg of naloxone IM (equivalent to 8 mg IN) failed to reach below the threshold of 50 % occupancy within 10 minutes or even within 15 minutes (Data not shown). In contrast, the model predicted successful reversals when 5 and 10 mg IM doses were utilized. Conclusion:These results support the notion that acutely administered higher doses of naloxone are needed for rapid and adequate clinical reversal, particularly when higher systemic exposure of the potent synthetic opioids occur.

Computational Models for Diagnosing and Treating Endometriosis

Endometriosis is a common but poorly understood disease. Symptoms can begin early in adolescence, with menarche, and can be debilitating. Despite this, people often suffer several years before being correctly diagnosed and adequately treated. Endometriosis involves the inappropriate growth of endometrial-like tissue (including epithelial cells, stromal fibroblasts, vascular cells, and immune cells) outside of the uterus. Computational models can aid in understanding the mechanisms by which immune, hormone, and vascular disruptions manifest in endometriosis and complicate treatment. In this review, we illustrate how three computational modeling approaches (regression, pharmacokinetics/pharmacodynamics, and quantitative systems pharmacology) have been used to improve the diagnosis and treatment of endometriosis. As we explore these approaches and their differing detail of biological mechanisms, we consider how each approach can answer different questions about endometriosis. We summarize the mathematics involved, and we use published examples of each approach to compare how researchers: (1) shape the scope of each model, (2) incorporate experimental and clinical data, and (3) generate clinically useful predictions and insight. Lastly, we discuss the benefits and limitations of each modeling approach and how we can combine these approaches to further understand, diagnose, and treat endometriosis.

A Quantitative Systems Pharmacology Model of the Pathophysiology and Treatment of COVID-19 Predicts Optimal Timing of Pharmacological Interventions

A quantitative systems pharmacology (QSP) model of the pathogenesis and treatment of SARS-CoV-2 infection can streamline and accelerate the development of novel medicines to treat COVID-19. Simulation of clinical trials allows in silico exploration of the uncertainties of clinical trial design and can rapidly inform their protocols. We previously published a preliminary model of the immune response to SARS-CoV-2 infection. To further our understanding of COVID-19 and treatment we significantly updated the model by matching a curated dataset spanning viral load and immune responses in plasma and lung. We identified a population of parameter sets to generate heterogeneity in pathophysiology and treatment and tested this model against published reports from interventional SARS-CoV-2 targeting Ab and anti-viral trials. Upon generation and selection of a virtual population, we match both the placebo and treated responses in viral load in these trials. We extended the model to predict the rate of hospitalization or death within a population. Via comparison of the in silico predictions with clinical data, we hypothesize that the immune response to virus is log-linear over a wide range of viral load. To validate this approach, we show the model matches a published subgroup analysis, sorted by baseline viral load, of patients treated with neutralizing Abs. By simulating intervention at different timepoints post infection, the model predicts efficacy is not sensitive to interventions within five days of symptom onset, but efficacy is dramatically reduced if more than five days pass post-symptom onset prior to treatment.

Quantitative Systems Pharmacology Model of Sickle Cell Disease and Response to Gene Editing Therapy to Support Clinical Development of SAR445136 (BIVV003)

Sickle cell disease (SCD) is caused by a mutation in the β-globin gene that produces abnormal hemoglobin (HbS), leading to clinical manifestations such as painful vaso-occlusive crises, anemia, and shortened lifespan due to organ damage. SAR445136 (BIVV003) is a zinc finger nuclease ex vivo gene editing therapy in Ph1/2 clinical development for treatment of SCD (PRECIZN-1; NCT03653247). SAR445136 targets the erythroid specific enhancer (ESE) region of the transcription factor BCL11A, which controls the switch from fetal hemoglobin (HbF) to adult forms (HbA in healthy subjects and HbS in SCD subjects). By expressing increased levels of HbF, SAR445136-edited cell progeny exhibit reduced HbS polymerization which is expected to ameliorate RBC sickling and the SCD phenotype. To better understand the dynamics and variability of clinical response to SAR445136, we developed a Quantitative Systems Pharmacology (QSP) model of SCD to describe both key elements of disease biology and the mechanism of action of SAR445136. The QSP approach provides a cohesive representation of the key disease processes in SCD by leveraging additional data sources (e.g. published and internal, clinical and preclinical) to complement data from ongoing clinical trials. The QSP model is applied to help assess mechanism-related questions, such as the observed inter-patient variability with respect to SAR445136 cellular dose, indels, and induced HbF and F cells. To explore the clinical factors that could influence the response to SAR445136, we centered the structure of the QSP model on a realistic representation of erythropoiesis that can describe hematopoietic stem and progenitor cells and erythroid progenitors in the bone marrow and the periphery, including regulation by cytokines EPO, IL-3, and stem cell factor (SCF) (Figure 1A). We first confirmed that the model recapitulates published bone marrow aspirate and blood cell sorting data from healthy individuals (Figure 1B). Next, we modified the model to describe stress erythropoiesis in SCD by incorporating published data on clinical, natural history, and in vitro assessments of SCD progenitor cells, and the resulting reticulocyte and erythrocyte levels (Figure 1B). The updated model describes key features of the SCD disease state, including reduced lifespan of HbS erythrocytes, elevated plasma reticulocytes (Steinberg MH, et al. Blood. 1997;89: 1078-1088), and altered levels of erythroid progenitors (Hoss SE, et al. Haematologica. 2020 Aug 27. doi: 10.3324/haematol.2020.265462) and cytokines, as well as the protective effects of endogenous HbF. Finally, we applied the SCD erythropoiesis model to describe the mechanism of action of SAR445136 by representing ablation followed by the introduction of CD34+ cells containing indels with enhanced HbF expression. The model captures the SAR445136 clinical data and simulates the therapeutic effects of increased total Hb and increased proportion of HbF due to the progeny of SAR445136 cells (Figure 1C). The model provides a quantitative framework for evaluating the effects of treatment parameters including cellular dose, mobilization and engraftment variability, indel quality (i.e., variable editing and HbF expression in erythroid progeny [Lessard S, et al. Blood. 2019;134(Supplement_1):97]) and assessing the pan-cellularity of the response. Due to its mechanistic structure, the model also enables exploration of inter-patient variability in terms of cytokine effects on erythropoiesis. In summary, the QSP model is a computational tool to provide mechanistic insight into emerging SAR445136 clinical data in the context of current understanding of SCD disease biology. The model is intended to provide both qualitative and quantitative support for the clinical development and competitive differentiation of SAR445136. For future development, the model can be expanded to include additional data representing the SCD bone marrow microenvironment to further explore patient heterogeneity. Figure 1 Figure 1. Disclosures Kaddi: Sanofi: Current Employment. Holz: Sanofi: Current Employment. Tao: Sanofi: Current Employment. Galeon: Sanofi: Current Employment. Reiner: Sanofi: Current Employment. Rendo: Sanofi: Current Employment, Other: May hold shares and/or stock options . Zaph: Sanofi: Current Employment.

Stably Accelerating Stiff Quantitative Systems Pharmacology Models: Continuous-Time Echo State Networks as Implicit Machine Learning

Quantitative systems pharmacology (QsP) may need to change in order to accommodate machine learning (ML), but ML may need to change to work for QsP. Here we investigate the use of neural network surrogates of stiff QsP models. This technique reduces and accelerates QsP models by training ML approximations on simulations. We describe how common neural network methodologies, such as residual neural networks, recurrent neural networks, and physics/biologically-informed neural networks, are fundamentally related to explicit solvers of ordinary differential equations (ODEs). Similar to how explicit ODE solvers are unstable on stiff QsP models, we demonstrate how these ML architectures see similar training instabilities. To address this issue, we showcase methods from scientific machine learning (SciML) which combine techniques from mechanistic modeling with traditional deep learning. We describe the continuous-time echo state network (CTESN) as the implicit analogue of ML architectures and showcase its ability to accurately train and predict on these stiff models where other methods fail. We demonstrate the CTESN’s ability to surrogatize a production QsP model, a >1,000 ODE chemical reaction system from the SBML Biomodels repository, and a reaction-diffusion partial differential equation. We showcase the ability to accelerate QsP simulations by up to 56x against the optimized DifferentialEquations.jl solvers while achieving <5% relative error in all of the examples. This shows how incorporating the numerical properties of QsP methods into ML can improve the intersection, and thus presents a potential method for accelerating repeated calculations such as global sensitivity analysis and virtual populations.