Modelling and simulation (M&S) plays a crucial role in drug development. It allows information from diverse sources – pre-clinical and in-vitro data, early and late-stage clinical trial data, biomarkers, and efficacy outcomes – to be included with a dataset. Modelling enables researchers to visualize complex scenarios and determine what will happen if a parameter is changed in a clinical trial. It can optimize clinical study design, enable unnecessary studies to be waived, and inform risk: benefits in “what if” scenarios with specific patient populations or drug-drug interactions (DDI).
M&S is used in all phases of drug development in the EU, US, and UK starting from pre-clinical pharmacokinetic (PK) and pharmacokinetic/pharmacodynamics (PK/PD) studies, through clinical trials to post-marketing commitments. It is generally used to support the clinical pharmacology files and labelling for new drug applications. M&S supports early dose selection based on pre-clinical data, PK/PD in-vitro animal data and scaling approaches and helps to set the first-in-human dose. M&S also shows the impact of intrinsic factors, such as age, body weight, or organ function on PK, PK/PD, and dose-exposure-response relationships.
Dose-exposure-response analysis is used to support safety and efficacy evaluations. The target exposure range – the concentration range that produces efficacy at an appropriate safety level – is a particularly important part of the clinical pharmacology evaluation. Researchers try to keep all clinical trial participants within that dose exposure range. If the concentration is too high, they will lower the dose to try to bring it into that range.
Using physiologically based PK (PBPK) and population PK (PopPK) approaches in parallel for model-informed drug development (MIDD) is proving especially valuable. But it is important to start populating the models early using pre-clinical data, and then adding new clinical data throughout the development process to strengthen and refine the models. The resulting models can be used to run simulations that answer questions regarding populations and situations that are difficult to study, and optimize clinical study design, in addition to generating evidence.
M&S can also help to address major issues relating to clinical pharmacology, PK in the target population, and dose-exposure-response analyses. These issues arising during the assessment of a marketing authorisation application (MAA) could potentially lead to a negative opinion, rejection, or delays in product approval.1 By employing high quality M&S, and seeking scientific advice during the development process, sponsors can mitigate many issues.
M&S has become an accepted part of the new drug development process. M&S used to be a supplemental part of the dossier, but regulators now expect it to be included in the files. In fact, MAAs are rarely submitted without modelling being used to describe the PK of a new medicine.
Research at academic institutions and software companies played a significant role in advancing MIDD. Some agencies, such as the U.S. Food and Drug Administration (FDA), have also sponsored related research. In addition, several Innovative Medicines Initiative projects in the EU are focused on applications of MIDD.
During the past decade, the European Medicines Agency (EMA) has drafted many guidelines that discuss M&S approaches, published papers on MIDD, created an M&S working party (MSWP), and hosted several MIDD-centric workshops. EMA has promoted the use of MIDD in dose finding during its workshops and locally in the agencies. It has produced guidance documents on the use of MIDD approaches in paediatric drug development, DDI risk assessment, renal and hepatic impairment, obesity, and pharmacogenetics. It also has clinical guidelines that reference MIDD approaches to support dosing for oncology, HIV, and antibiotic use.
In its “Regulatory Science to 2025” strategy document, the EMA mentions modelling 37 times and specifically references its desire to “Optimize capabilities in modelling, simulation and extrapolation” to drive collaborative evidence generation and improve the scientific quality of evaluations.2
While the FDA and EMA have both embraced M&S, they take slightly different approaches. For example, the FDA always requests that sponsors provide raw data with their new drug applications (NDAs), and its internal team of pharmacometricians can reanalyse it. But the EMA generally does not currently perform its own M&S. However, the EMA did announce a pilot study in July 2022 for which it is requesting that raw data be submitted for possible
re-analysis. But the Agency did not state whether it plans to conduct M&S or statistical analysis of the data.
Many M&S approaches are used in paediatric drug development. Underscoring its importance, the EMA published a reflection paper in 2018, which outlined scenarios under which sponsors could use M&S to extrapolate efficacy from adults to children when developing medicines for paediatric patients.3 Most importantly, M&S is used to extrapolate from adult efficacy and safety data to determine the initial paediatric drug doses for clinical studies. M&S is also employed to optimize the blood sampling times and minimize the number of blood samples that need to be collected from paediatric patients. M&S also helps to determine how many paediatric patients need to be recruited in each age group or body weight category for a specific clinical study. Those results are included in the Paediatric Investigational Plan, which is submitted to the Paediatric Committee (PDCO), the EMA’s scientific committee responsible for activities regarding medicines for children.
All the available data are used to create and refine a dynamic paediatric model, which shows the effect of body weight or other covariates on PK parameters and dosing. M&S enables all the available data to be used in a much more efficient way.
Furthermore, by using PBPK modelling, researchers can progress beyond scaling based on body weight and factor in the paediatric patient’s maturing physiology and enzyme levels as well. Literature data can be employed to support the maturation of different enzymes in paediatric patients.
The International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) strives to achieve greater harmonisation worldwide to ensure that safe, effective, and high-quality medicines are developed registered and maintained in the most resource efficient manner whilst meeting high standards.
ICH has a rich history of collaboration. ICH members have worked closely together for many years to develop comprehensive guidelines for PopPK and exposure-response analyses in multi-national clinical trials. Those guidelines include E5 which focuses on “Ethnic Factors in the Acceptability of Foreign Clinical Data,” E7 for “Studies in Support of Special Populations: Geriatrics,” E14 for “The Clinical Evaluation of QT/QTc Interval Prolongation and Proarrhythmic Potential for Non-Antiarrhythmic Drugs,” and E17 which outlines “General Principles for Planning and design of Multi-Regional Clinical Trials.”
In April 2022, ICH also released a draft guideline for “Paediatric Extrapolation” (E11a).4 Even more recently, ICH released draft drug interaction guidance, which provides a harmonized approach to drug interactions and references PBPK, in addition to PK, for the first time.5 In fact, PBPK is mentioned about 50 times in the extrapolation guideline, and it is referenced in the drug interaction guidance as well.
In 2021, ICH founded a MIDD Discussion Group in Europe. In March 2022, that group released a roadmap, outlining its plans. It is expected to be a particularly important platform, which regulators can use to advance the field. It enables staff from different agencies to contribute their expertise, exchange, and challenge ideas, and develop new guidance on MIDD principles.
A clinical trials regulation (Regulation (EU) No 536/2014) was also introduced in early 2022, which describes harmonized ways to conduct assessments of clinical trials. It is hoped that this will lead to harmonized assessments of M&S methodologies as well.
Several EMA working groups have also helped to advance the field. They include the MSWP, PDCO, and the PK Working Party. Recently, EMA announced the formation of a Methodology Working Party, which will include experts in biostatistics, M&S, extrapolation, PK, real-world evidence, etc. The working party will support both scientific advice and scientific committees, write guidelines and provide assessor training.
The working parties participate in cluster meetings with experts from other global regulatory agencies, including the FDA, Health Canada, Japan’s Pharmaceuticals and Medical Devices Agency, and Australia’s Therapeutic Goods Administration. The MSWP, for example, usually meets quarterly, bringing together pharmacometricians, and other clinical pharmacology assessors, to share information and discuss common issues. At the World Conference on Pharmacometrics earlier this year, representatives from the EMA and FDA met with colleagues from the African agencies, so there is scope for harmonization there as well.
A lot of cross-agency reliance models have also been implemented. For example, the regulatory agencies in Australia, Canada, Singapore, Switzerland, and the UK are members of the Access Consortium and participate in joint assessment work. Sponsors can submit a new drug application to multiple agencies in the Access Consortium. Then, the members prepare a work-sharing agreement to assess the new medicines.
The FDA has introduced a similar model, called Project Orbis, which provides a framework for the concurrent submission and review of oncology products among international partners. Project Orbis includes all the agencies in the Access Consortium plus the Brazilian agency.
As the EU is comprised of 27 member states, it focuses on harmonization internally as well as externally with other global regulatory agencies. The EMA MSWP is helping to share knowledge between member states and harmonize the new drug assessment. This cooperation between member states also reduces duplication of efforts and helps to ensure that vulnerable, rare patient populations are not asked to participate in multiple study designs.
The adoption of MIDD approaches in the EU is expected to continue to grow as assessors became more familiar with PopPK, PBPK and quantitative systems pharmacology (QSP) approaches and other emerging methodologies. The planned drafting of the ICH MIDD guidance documents will also help to progress MIDD science and increase regulatory acceptance. Understanding the relationship between PK and PD has always been at the heart of clinical pharmacology yet an even greater emphasis is expected to be placed on exploring doseexposure-response relationships in future clinical trials.
Using PBPK and PopPK approaches for formulation bridging is another area that holds enormous potential. If a sponsor wants to change a product’s formulation, such as switching an injection device from vials to an auto-injector, between Phase 2 and Phase 3 clinical trials, or just prior to market introduction, when the similarity criteria are stricter, M&S can support bridging. It is also invaluable with generics and biosimilars. These forms of modelling are particularly useful in instances when it is complicated to perform conventional studies in healthy volunteers, such as when the product has a very long half-life.
Both PBPK and PopPK methods should also be used together when assessing and investigating a medicine in special populations. For example, patients with renal impairment can enroll in a clinical study if they are given an appropriate dose determined using PBPK modelling. Then, the PK data generated from the study can be analysed, using PopPK, and a learn and confirm approach, to create more clinical data for this population for which there is usually only minimal PK data available.
QSP is another application that has been steadily growing in popularity and has gained traction. It enables biological systems to be combined with mechanistic models to predict drug safety and identify critical biomarkers. It was used heavily during the pandemic for vaccine development. A mechanistic model of the human immune system, which was originally created to determine whether a biological therapeutic would create an unwanted immunogenic response, was repurposed as a COVID-19 vaccine model. While the creators of the initial model were hoping to see low levels of immunogenicity for their candidate biologicals, those working on the revised model wanted vaccine candidates that promoted high levels of immunogenicity.
Model-based meta-analysis, which allows researchers to leverage peer-reviewed clinical trial data from studies of similar medicines which are then combined with a pharmacology model, is also proving to be a powerful approach for the development of medicines. Data from the literature and other sources are combined in a PK/PD model, to enable researchers to bridge across different studies and conduct indirect comparisons with either the standard of care or a competitor drug.
M&S continues to be used extensively in paediatric studies in orphan indications when there are too few patients, in situations where there are only indirect biomarkers, or no biomarkers, such as with vaccines or gene therapies, and in new and growing fields such as biologics, bi-specific antibodies, and oligonucleotides. PBPK modelling has also recently been proposed for assessing mother and child drug exposure during pregnancy and lactation. This information is vitally important if women who are pregnant and lactating need to be included in a clinical study. These data can be used to inform both the study design and risk assessment. This approach also aligns closely with one of the M&S objectives outlined in the EMA strategy plan.2
The EMA Regulatory Science to 2025 strategy also mentions other fields where they will be optimising capabilities, including the use of artificial intelligence in M&S, real world data and M&S in special populations, maternal-foetal health, enhancing collaboration with external parties on M&S for example through IMI research projects, and developing data sharing platforms.
M&S can be employed to confirm a drug’s early dose selection and dose-exposure-response, inform study design and risk assessments, determine the potential for DDIs, and help bridge between adult and paediatric populations and between different drug formulations.
Global regulatory agencies, such as the FDA and EMA, recognize the value provided by M&S during drug development and assessment and are working diligently with the ICH and focused working parties to develop harmonised guidance to optimise its use.
References
Eva Gil Berglund, PhD, is a Certara Senior Director in Clinical Pharmacology and Regulatory Strategy with 20+ years of experience as a Clinical Pharmacology Reviewer in the Swedish Medical Products Agency working under the EMA umbrella. She has been a member of the Pharmacokinetic Working Party, contributing to many guidelines.
Justin Hay is a Senior Director at Certara with 15+ years of clinical pharmacology experience. Career highlights include working as a Senior Pharmacokinetics Assessor and Deputy Unit Manager at the Medicines and Healthcare products Regulatory Agency, UK. He has also been a member of the EMA's Modelling and Simulation Working Party.
