L'année où j'ai terminé mes études, un vieil ami de la famille prenait sa retraite. Il avait consacré l'essentiel de sa carrière à concevoir et déployer du matériel agricole aux États-Unis. Lorsque je lui ai demandé conseil concernant mon entrée sur le marché du travail, il m'a raconté une histoire.
Il était à la retraite depuis trois mois quand une société en Californie lui a demandé de prendre un avion pour voir avec leurs mécaniciens pourquoi un de leurs appareils ne fonctionnait pas. Il a pris l'avion, passé 5 minutes sur le site de la société, dessiné un « X » à la craie sur l'appareil et est rentré chez lui. Ils l'ont appelé le lendemain pour savoir ce qu'ils devaient faire. Il leur a indiqué le « X ». Aussi simple que soit un « X », il représentait des années d'expérience et de compréhension de la complexité de la machine et des dysfonctionnements potentiels.
Pour lui, la morale de l'histoire c'est que « c'est bien de savoir où dessiner le "X" ».
Surely this could work in the fields of engineering, manufacturing, or construction (to name a few), but how would this apply in pharmaceutical development where we are dealing with complex living species that are all drastically different in their genetic makeup? And what about the differences in the indications, modalities, and pharmacology of the drugs we test?
So where to put the "X" in pharmaceutical development? That's what we in PK/PD modeling and simulation strive to address. We offer a multitude of different services. These range from noncompartmental analysis to help parameterize and guide the interpretation of pharmacokinetic, pharmacodynamic, or toxicology studies, to fully mechanistic PK/PD models taking into account the anatomical physiology of test species, the physiochemical properties of the drug, and the metabolic profile tested in vitro. These mechanistic models mathematically describe the PK and simulate a range of outcomes to help design better studies and make better decisions.
Covance works with different partners from academia to large pharma and everywhere in between. We work across all phases of development and across a myriad of molecule types, disease states and routes of administration.
Over the next six months we will be introducing topics that we get regular questions about from across the industry and discuss how we help to offer up modeling and simulation solutions to progress their programs more efficiently.
Here are some of the topics we plan to cover:
Target Mediated Drug Disposition (TMDD) for Biologics
TMDD is a very common phenomenon observed when developing biologics. The pharmacokinetics properties of the drug are impacted by the high affinity of the pharmacological target. While in many cases this affinity to the target is an advantage to the pharmacology the drug is influencing, we will discuss utilizing TMDD PK modeling to optimize the therapeutic window, the dosing regimen, and additionally, the translation from preclinical species to the clinic.
Noncompartmental Analysis: Considerations, Efficiencies, and SEND
In the past year, the Covance modeling and simulation team has had the opportunity to support more than 170 unique clients. We see a wide variety of approaches and opinions from handling of anomalous data, to appropriate sampling schemes and parameters to characterize the PK/PD. In this post we will discuss our approach which has been informed by observation and consensus across the industry, the potential pitfalls that are often missed in Noncompartmental analysis, and how to effectively bring forward the data into SEND ready domains.
Physiologically Based PK Models for Understanding and Predicting DDI’s
With the recent updates to the FDA guidances on drug-drug interactions (DDI) there is a buzz in the industry around utilizing risk-based approaches in assessing potential DDI and further mitigating potential DDIs as we design clinical trials and develop product labeling. We will discuss the utilization of the in vitro metabolism and drug transporter data to build mechanistic models that can be utilized to simulate potential DDI with a range of perpetrator drugs.
First in Human Dose Predictions: Making Applicable to Your Drug
Non-linear pharmacokinetics, in vitro-in vivo correlations, narrow safety margins, non-translational systemic exposure, drug transporters, applicable preclinical species, target engagement, body surface area scaling and allometry, MABEL, mechanistic interspecies scaling (the list goes on)…
We can certainly all agree there are a number of factors that go into understanding your preclinical data and justifying a safe first-in-human dose. Let's try to simplify the approach. In this post we will discuss how we evaluate the sponsor programs, help to understand the limitations, and cater an approach so we can dose with confidence and design a Phase I study that captures all the endpoints.
Model-Based Drug Development: Predict, Define, Refine
There is an old adage that "All models are wrong, but some are useful". Our goal is to improve this statement. We feel that all models have limitations, but we want to do three things to ensure the majority are useful:
- Think about where the model could be useful and what the model means in the scope of the overall development of the drug
- Understand the limitations and specific client perspectives, so we can interpret appropriately
- Improve each iteration as we acquire additional information (knowledge and data)
This post will discuss our approach, and where we think we can help to find solutions that will save time, resources, and money.
Population PK/PD Analysis and Understanding Covariates
Population pharmacokinetic and pharmacodynamic analyses are used to help understand the differences and variability among a target population to aid in the determination of safe and efficacious drug administration. They are utilized in support of preclinical and clinical data and allow for integration of sparse data but also for combinations of sparse and dense data. We will share our experiences using these techniques across the phases of drug development to help investigate the sources of variability (e.g., bodyweight, metabolic functions, patient demographics, etc) and how to adjust the dosing regimens based on the covariates that impact dose-concentration-efficacy/safety relationships.