Simulations of site-specific target-mediated pharmacokinetic models for guiding the development of bispecific antibodies

Mechanistic computational modeling of monospecific and bispecific antibodies targeting interleukin-6/8 receptors

The spread of cancer from organ to organ (metastasis) is responsible for the vast majority of cancer deaths; however, most current anti-cancer drugs are designed to arrest or reverse tumor growth without directly addressing disease spread. It was recently discovered that tumor cell-secreted interleukin-6 (IL-6) and interleukin-8 (IL-8) synergize to enhance cancer metastasis in a cell-density dependent manner, and blockade of the IL-6 and IL-8 receptors (IL-6R and IL-8R) with a novel bispecific antibody, BS1, significantly reduced metastatic burden in multiple preclinical mouse models of cancer. Bispecific antibodies (BsAbs), which combine two different antigen-binding sites into one molecule, are a promising modality for drug development due to their enhanced avidity and dual targeting effects. However, while BsAbs have tremendous therapeutic potential, elucidating the mechanisms underlying their binding and inhibition will be critical for maximizing the efficacy of new BsAb treatments. Here, we describe a quantitative, computational model of the BS1 BsAb, exhibiting how modeling multivalent binding provides key insights into antibody affinity and avidity effects and can guide therapeutic design. We present detailed simulations of the monovalent and bivalent binding interactions between different antibody constructs and the IL-6 and IL-8 receptors to establish how antibody properties and system conditions impact the formation of binary (antibody-receptor) and ternary (receptor-antibody-receptor) complexes. Model results demonstrate how the balance of these complex types drives receptor inhibition, providing important and generalizable predictions for effective therapeutic design.

Minimally sufficient experimental design using identifiability analysis

Mathematical models are increasingly being developed and calibrated in tandem with data collection, empowering scientists to intervene in real time based on quantitative model predictions. Well-designed experiments can help augment the predictive power of a mathematical model but the question of when to collect data to maximize its utility for a model is non-trivial. Here we define data as model-informative if it results in a unique parametrization, assessed through the lens of practical identifiability. The framework we propose identifies an optimal experimental design (how much data to collect and when to collect it) that ensures parameter identifiability (permitting confidence in model predictions), while minimizing experimental time and costs. We demonstrate the power of the method by applying it to a modified version of a classic site-of-action pharmacokinetic/pharmacodynamic model that describes distribution of a drug into the tumor microenvironment (TME), where its efficacy is dependent on the level of target occupancy in the TME. In this context, we identify a minimal set of time points when data needs to be collected that robustly ensures practical identifiability of model parameters. The proposed methodology can be applied broadly to any mathematical model, allowing for the identification of a minimally sufficient experimental design that collects the most informative data.

A Bispecific Modeling Framework Enables the Prediction of Efficacy, Toxicity, and Optimal Molecular Design of Bispecific Antibodies Targeting MerTK

Inhibiting MerTK on macrophages is a promising therapeutic strategy for augmenting anti-tumor immunity. However, blocking MerTK on retinal pigment epithelial cells (RPEs) results in retinal toxicity. Bispecific antibodies (bsAbs) containing an anti-MerTK therapeutic and anti-PD-L1 targeting arm were developed to reduce drug binding to MerTK on RPEs, since PD-L1 is overexpressed on macrophages but not RPEs. In this study, we present a modeling framework using in vitro receptor occupancy (RO) and pharmacokinetics (PK) data to predict efficacy, toxicity, and therapeutic index (TI) of anti-MerTK bsAbs. We first used simulations and in vitro RO data of anti-MerTK monospecific antibody (msAb) to estimate the required MerTK RO for in vivo efficacy and toxicity. Using these estimated RO thresholds, we employed our model to predict the efficacious and toxic doses for anti-MerTK bsAbs with varying affinities for MerTK. Our model predicted the highest TI for the anti-MerTK/PD-L1 bsAb with an attenuated MerTK binding arm, which was consistent with in vivo efficacy and toxicity observations. Subsequently, we used the model, in combination with sensitivity analysis and parameter scans, to suggest an optimal molecular design of anti-MerTK bsAb with the highest predicted TI in humans. Our prediction revealed that this optimized anti-MerTK bsAb should contain a MerTK therapeutic arm with relatively low affinity, along with a high affinity targeting arm that can bind to a low abundance target with slow turnover rate. Overall, these results demonstrated that our modeling framework can guide the rational design of bsAbs.

Mechanistic computational modeling of monospecific and bispecific antibodies targeting interleukin-6/8 receptors

The spread of cancer from organ to organ (metastasis) is responsible for the vast majority of cancer deaths; however, most current anti-cancer drugs are designed to arrest or reverse tumor growth without directly addressing disease spread. It was recently discovered that tumor cell-secreted interleukin-6 (IL-6) and interleukin-8 (IL-8) synergize to enhance cancer metastasis in a cell-density dependent manner, and blockade of the IL-6 and IL-8 receptors (IL-6R and IL-8R) with a novel bispecific antibody, BS1, significantly reduced metastatic burden in multiple preclinical mouse models of cancer. Bispecific antibodies (BsAbs), which combine two different antigen-binding sites into one molecule, are a promising modality for drug development due to their enhanced avidity and dual targeting effects. However, while BsAbs have tremendous therapeutic potential, elucidating the mechanisms underlying their binding and inhibition will be critical for maximizing the efficacy of new BsAb treatments. Here, we describe a quantitative, computational model of the BS1 BsAb, exhibiting how modeling multivalent binding provides key insights into antibody affinity and avidity effects and can guide therapeutic design. We present detailed simulations of the monovalent and bivalent binding interactions between different antibody constructs and the IL-6 and IL-8 receptors to establish how antibody properties and system conditions impact the formation of binary (antibody-receptor) and ternary (receptor-antibody-receptor) complexes. Model results demonstrate how the balance of these complex types drives receptor inhibition, providing important and generalizable predictions for effective therapeutic design.

Mechanism-based pharmacokinetic and pharmacodynamic modeling for bispecific antibodies: challenges and opportunities

INTRODUCTION

Unlike conventional antibodies, bispecific antibodies (bsAbs) are engineered antibody- or antibody fragment-based molecules that can simultaneously recognize two different epitopes or antigens. Over the past decade, there has been an explosion of bsAbs being developed across therapeutic areas. Development of bsAbs presents unique challenges and mechanism-based pharmacokinetic/pharmacodynamic (PK/PD) modeling has served as a powerful tool to optimize their development and realize their clinical utility.

AREAS COVERED

In this review, the guiding principles and case examples of how fit-for-purpose, mechanism-based PK/PD models have been applied to answer questions commonly encountered in bsAb development are presented. Such models characterize the key pharmacological elements of bsAbs, and they can be utilized for model-informed drug development. We also include the discussion of challenges, knowledge gaps and future direction for such models.

EXPERT OPINION

Mechanistic PK/PD modeling is a powerful tool to support the development of bsAbs. These models can be extrapolated to predict treatment outcomes based on mechanisms of action (MoA) and clinical observations to form positive learn-and-confirm cycles during drug development, due to their abilities to differentiate system- and drug-specific parameters. Meanwhile, the models should keep being adapted according to novel drug design and MoA, providing continuous opportunities for model-informed drug development.

Nonclinical Pharmacokinetics and Pharmacodynamics Characterization of Anti-CD79b/CD3 T Cell-Dependent Bispecific Antibody Using a Surrogate Molecule: A Potential Therapeutic Agent for B Cell Malignancies

The T cell-dependent bispecific (TDB) antibody, anti-CD79b/CD3, targets CD79b and CD3 cell-surface receptors expressed on B cells and T cells, respectively. Since the anti-CD79b arm of this TDB binds only to human CD79b, a surrogate TDB that binds to cynomolgus monkey CD79b (cyCD79b) was used for preclinical characterization. To evaluate the impact of CD3 binding affinity on the TDB pharmacokinetics (PK), we utilized non-tumor-targeting bispecific anti-gD/CD3 antibodies composed of a low/high CD3 affinity arm along with a monospecific anti-gD arm as controls in monkeys and mice. An integrated PKPD model was developed to characterize PK and pharmacodynamics (PD). This study revealed the impact of CD3 binding affinity on anti-cyCD79b/CD3 PK. The surrogate anti-cyCD79b/CD3 TDB was highly effective in killing CD79b-expressing B cells and exhibited nonlinear PK in monkeys, consistent with target-mediated clearance. A dose-dependent decrease in B cell counts in peripheral blood was observed, as expected. Modeling indicated that anti-cyCD79b/CD3 TDB’s rapid and target-mediated clearance may be attributed to faster internalization of CD79b, in addition to enhanced CD3 binding. The model yielded unbiased and precise curve fits. These findings highlight the complex interaction between TDBs and their targets and may be applicable to the development of other biotherapeutics.

Bispecific antibodies: A guide to model informed drug discovery and development

Affinity (KD) optimization of monoclonal antibodies is one of the factors that impacts the stoichiometric binding and the corresponding efficacy of a drug. This impacts the dose and the dosing regimen, making the optimum KD a critical component of drug discovery and development. Its importance is further enhanced for bispecific antibodies, where affinity of the drug needs to be optimized with respect to two targets. Mathematical modeling can have critical impact on lead compound optimization. Here we build on previous work of using mathematical models to facilitate lead compound selection, expanding analysis from two membrane bound targets to soluble targets as well. Our analysis reveals the importance of three factors for lead compound optimization: drug affinity to both targets, target turnover rates, and target distribution throughout the body. We describe a method that leverages this information to help make early stage decisions on whether to optimize affinity, and if so, which arm of the bispecific should be optimized. We apply the proposed approach to a variety of scenarios and illustrate the ability to make improved decisions in each case. We integrate results to develop a bispecific antibody KD optimization guide that can be used to improve resource allocation for lead compound selection, accelerating advancement of better compounds. We conclude with a discussion of possible ways to assess the necessary levels of target engagement for affecting disease as part of an integrative approach for model-informed drug discovery and development.

Therapeutic drug monitoring of anti-tumour necrosis factor-α agents in inflammatory bowel disease: Limits and improvements

AIMS

Since the publication of the American Gastroenterological Association's recommendations in 2017, there have been no significant changes in the biological monitoring recommendations in inflammatory bowel disease. Possible limitations are the lack of evidence to recommend proactive therapeutic drug monitoring (pTDM) over reactive TDM (rTDM), and the limited information about individualized dosing methods. This article aims to review the TDM strategy updates and the use of individualized dosing methods.

METHODS

For the analysis of the TDM strategies and individualized dosing method, a search was carried out in PubMed and Cochrane Central. In the TDM case, since August 2017.

RESULTS

A total of 263 publications were found, but only 7 related to proactive TDM. Five of these publications directly compared pTDM vs rTDM and 2 were randomized clinical trials. Six studies found benefits of pTDM and 1 found no differences. Regarding the individualized dosing method, 229 distinct results were found. Population pharmacokinetics was the most widely used method to develop individual dosage models and to analyse the influence of factors on drug concentrations (albumin concentration, weight, presence of anti-drug antibodies etc).

CONCLUSION

We have found no major changes in TDM strategies. There is a growing trend towards the use of pTDM because it has shown a longer duration of treatment response, lower rates of discontinuation and relapses. However, the available evidence is limited and of low quality. Despite the common use of population pharmacokinetic methods to analyse pharmacokinetic factors, they are not commonly used for personalized dosing.

Informing Development of Bispecific Antibodies Using Physiologically Based Pharmacokinetic-Pharmacodynamic Models: Current Capabilities and Future Opportunities

Antibody therapeutics continue to represent a significant portion of the biotherapeutic pipeline, with growing promise for bispecific antibodies (BsAbs). BsAbs can target 2 different antigens at the same time, such as simultaneously binding tumor-cell receptors and recruiting cytotoxic immune cells. This simultaneous engagement of 2 targets can be potentially advantageous, as it may overcome disadvantages posed by a monotherapy approach, like the development of resistance to treatment. Combination therapy approaches that modulate 2 targets simultaneously offer similar advantages, but BsAbs are more efficient to develop. Unlike combination approaches, BsAbs can facilitate spatial proximity of targets that may be necessary to induce the desired effect. Successful development of BsAbs requires understanding antibody formatting and optimizing activity for both targets prior to clinical trials. To realize maximal efficacy, special attention is required to fully define pharmacokinetic (PK)/pharmacodynamic (PD) relationships enabling selection of dose and regimen. The application of physiologically based pharmacokinetics (PBPK) has been evolving to inform the development of novel treatment modalities such as bispecifics owing to the increase in our understanding of pharmacology, utility of multiscale models, and emerging clinical data. In this review, we discuss components of PBPK models to describe the PK characteristics of BsAbs and expand the discussion to integration of PBPK and PD models to inform development of BsAbs. A framework that can be adopted to build PBPK-PD models to inform the development of BsAbs is also proposed. We conclude with examples that highlight the application of PBPK-PD and share perspectives on future opportunities for this emerging quantitative tool.

Target-Mediated Drug Disposition Model for Bispecific Antibodies: Properties, Approximation, and Optimal Dosing Strategy

Bispecific antibodies (BsAbs) bind to two different targets, and create two binary and one ternary complex (TC). These molecules have shown promise as immuno-oncology drugs, and the TC is considered the pharmacologically active species that drives their pharmacodynamic effect. Here, we have presented a general target-mediated drug disposition (TMDD) model for these BsAbs, which bind to two different targets on different cell membranes. The model includes four different binding events for BsAbs, turnover of the targets, and internalization of the complexes. In addition, a quasi-equilibrium (QE) approximation with decreased number of binding parameters and, if necessary, reduced internalization parameters is presented. The model is further used to investigate the kinetics of BsAb and TC concentrations. Our analysis shows that larger doses of BsAbs may delay the build-up of the TC. Consequently, a method to compute the optimal dosing strategy of BsAbs, which will immediately create and maintain maximal possible TC concentration, is presented.

Comparison of tenofovir plasma and tissue exposure using a population pharmacokinetic model and bootstrap: a simulation study from observed data

Sparse tissue sampling with intensive plasma sampling creates a unique data analysis problem in determining drug exposure in clinically relevant tissues. Tissue exposure may govern drug efficacy, as many drugs exert their actions in tissues. We compared tissue area-under-the-curve (AUC) generated from bootstrapped noncompartmental analysis (NCA) methods and compartmental nonlinear mixed effect (NLME) modeling. A model of observed data after single-dose tenofovir disoproxil fumarate was used to simulate plasma and tissue concentrations for two destructive tissue sampling schemes. Two groups of 100 data sets with densely-sampled plasma and one tissue sample per individual were created. The bootstrapped NCA (SAS 9.3) used a trapezoidal method to calculate geometric mean tissue AUC per dataset. For NLME, individual post hoc estimates of tissue AUC were determined, and the geometric mean from each dataset calculated. Median normalized prediction error (NPE) and absolute normalized prediction error (ANPE) were calculated for each method from the true values of the modeled concentrations. Both methods produced similar tissue AUC estimates close to true values. Although the NLME-generated AUC estimates had larger NPEs, it had smaller ANPEs. Overall, NLME NPEs showed AUC under-prediction but improved precision and fewer outliers. The bootstrapped NCA method produced more accurate estimates but with some NPEs > 100%. In general, NLME is preferred, as it accommodates less intensive tissue sampling with reasonable results, and provides simulation capabilities for optimizing tissue distribution. However, if the main goal is an accurate AUC for the studied scenario, and relatively intense tissue sampling is feasible, the NCA bootstrap method is a reasonable, and potentially less time-intensive solution.

Assessing the Impact of Tissue Target Concentration Data on Uncertainty in In Vivo Target Coverage Predictions

Understanding pharmacological target coverage is fundamental in drug discovery and development as it helps establish a sequence of research activities, from laboratory objectives to clinical doses. To this end, we evaluated the impact of tissue target concentration data on the level of confidence in tissue coverage predictions using a site of action (SoA) model for antibodies. By fitting the model to increasing amounts of synthetic tissue data and comparing the uncertainty in SoA coverage predictions, we confirmed that, in general, uncertainty decreases with longitudinal tissue data. Furthermore, a global sensitivity analysis showed that coverage is sensitive to experimentally identifiable parameters, such as baseline target concentration in plasma and target turnover half‐life and fixing them reduces uncertainty in coverage predictions. Overall, our computational analysis indicates that measurement of baseline tissue target concentration reduces the uncertainty in coverage predictions and identifies target‐related parameters that greatly impact the confidence in coverage predictions.

Pharmacokinetics and Tolerability of a Dual Variable Domain Immunoglobulin ABT-981 Against IL-1α and IL-1β in Healthy Subjects and Patients With Osteoarthritis of the Knee

The interleukin (IL)-1 family of proinflammatory cytokines are thought to play a significant role in the structural progression of osteoarthritis and its associated symptoms. IL-1α and IL-1β are 2 distinct cytokines found in the cartilage, synovial membrane, and synovial fluid of patients with osteoarthritis. The aim of these studies was to evaluate the pharmacokinetics of ABT-981, a dual variable domain immunoglobulin (DVD-Ig) capable of simultaneously binding IL-1α and IL-1β, in healthy subjects and patients with osteoarthritis of the knee. Fifty-six healthy adult subjects were randomized to receive single doses of ABT-981 intravenously (0.3, 1, 3, or 10 mg/kg), subcutaneously (0.3, 1, 3 mg/kg), or matching placebo in a 3:1 ratio. Thirty-six patients with osteoarthritis of the knee were randomized to receive 4 subcutaneous ABT-981 doses of 0.3, 1, or 3 mg/kg administered every 2 weeks, 3 subcutaneous doses of ABT-981 3 mg/kg every 4 weeks, or matching placebo in a 7:2 active:placebo ratio. ABT-981 behaved similarly to conventional monoclonal antibodies following single or multiple doses with mean maximum serum concentrations 2 to 9 days after subcutaneous doses, mean terminal half-lives of 10 to 14 days, and an absolute subcutaneous bioavailability of 46%. Exposure of ABT-981 was approximately linear following single or multiple doses every 2 weeks with monoexponential decline of terminal-phase concentrations. The most common adverse events associated with ABT-981 were diarrhea and headache in healthy subjects and injection site erythema in subjects with osteoarthritis of the knee. Decreased absolute neutrophil counts were observed in response to ABT-981 administration.

The application of mathematical modelling to the design of bispecific monoclonal antibodies

Targeting multiple receptors with bispecific antibodies is a novel approach that may prevent the development of resistance to cancer treatments. Despite the initial promise, full clinical benefit of this technology has yet to be realized. We hypothesized that in order to optimally exploit bispecific antibody technology, thorough fundamental knowledge of their pharmacological properties compared to that of single agent combinations was needed. Therefore, we developed a mathematical model for the binding of bispecific antibodies to their targets that accounts for the spatial distribution of the binding receptors and the kinetics of binding, and is scalable for increasing valency. The model provided an adequate description of internal and literature-reported in vitro data on bispecific binding. Simulations of in vitro binding with the model indicated that bispecific antibodies are not always superior in their binding potency to combination of antibodies, and the affinity of bispecific arms must be optimized for maximum binding potency. Our results suggest that this tool can be used for the design and development of the next generation of anti-cancer bispecific compounds.

Mathematical modeling of receptor occupancy data: A valuable technology for biotherapeutic drug development

BACKGROUND

In drug development, in vivo assessment of target engagement provides confidence when testing the drug's mechanism of action and improves the likelihood of clinical success. For biologics, receptor occupancy (RO) determined from circulating cells can provide evidence of target engagement. Integrating this information with mathematical modeling can further enhance the understanding of drug-target interactions and the biological factors that are critical to the successful modulation of the target and ultimately the disease state.

METHODS

This mini-review presents two specific types of mathematical models used to describe antibody-receptor systems and highlights how experimental data can inform the model parameters. Simulations are used to illustrate how various mechanisms influence RO, PK and total cellular receptor profiles.

RESULTS

The simulations demonstrate the effect antibody-receptor internalization, affinity and receptor turnover have on commonly acquired data in drug development.

CONCLUSIONS

Integrating RO data with mathematical models such as the two presented here (target-mediated drug disposition and site-of-action models) can provide a more comprehensive view of the biological system, which can be used to test hypotheses, extrapolate preclinical findings to humans and impact clinical study designs and risk assessments for the successful development of biotherapeutics.

Pharmacokinetic and pharmacodynamic considerations for the next generation protein therapeutics

Increasingly sophisticated protein engineering efforts have been undertaken lately to generate protein therapeutics with desired properties. This has resulted in the discovery of the next generation of protein therapeutics, which include: engineered antibodies, immunoconjugates, bi/multi-specific proteins, antibody mimetic novel scaffolds, and engineered ligands/receptors. These novel protein therapeutics possess unique physicochemical properties and act via a unique mechanism-of-action, which collectively makes their pharmacokinetics (PK) and pharmacodynamics (PD) different than other established biological molecules. Consequently, in order to support the discovery and development of these next generation molecules, it becomes important to understand the determinants controlling their PK/PD. This review discusses the determinants that a PK/PD scientist should consider during the design and development of next generation protein therapeutics. In addition, the role of systems PK/PD models in enabling rational development of the next generation protein therapeutics is emphasized.