Strategic and hands-on MIDD for clinical development
A model-informed plan for the clinical programme, and the analyses that carry it out, documented in line with ICH M15.
Pharmacometrics and MIDD consulting
Model-informed decisions you can defend
MIDD for pharma and biotech: clinical development, preclinical and translational PK/PD, PBPK, and virtual trials for complex generics and biosimilars, plus NCA and toxicokinetic analysis with submission-ready TFLs. Strategy and hands-on execution, from the first question to the regulatory submission.
15+ years
At Director level in industry and CROs
50+ clients
Pharma and biotech, worldwide
End to end
Preclinical through filing, no handoff
What we offer
Four service areas
We start from your question, design the MIDD strategy, execute it and deliver toward submission, for internal and regulatory decisions.
Preclinical and translational PK/PD modeling
From animal data to a human dose you can defend.
PBPK modeling
Predicting exposure where it has not been measured, to decide which clinical studies are still needed.
Virtual BE and PK-similarity trials
Bioequivalence for complex generics and PK similarity for biosimilars, simulated before the pivotal study is committed.
Core PK services
PK and TK analysis
Non-compartmental analysis and toxicokinetics for preclinical and clinical studies, delivered as a standalone package or as the base of a modeling programme.
Quality
Built for regulatory scrutiny
Every analysis is done to the standard a regulator expects, from the first day of the project.
Industry-standard tools
The platforms regulators know and expect to see in a submission.
GxP-compliant environment
Controlled, documented and version-controlled, so every number can be traced and reproduced.
Submission-ready by default
Datasets, code, models and reports are built to submission standard from the start.
Who you work with
You talk to the modeller, and the team scales to the project
Dr Raphael Bilgraer
Founder and Principal Modeller, Pharm.D, Ph.D
15+ years of experience at Director level in the pharmaceutical industry and at CROs. Pharm.D and Ph.D in Biomedical Sciences, Paris University. Every engagement is led personally, with additional pharmacokineticists and clinical pharmacology experts brought in as the project requires.
MIDD in practice
Where a model changes the decision
Illustrative applications drawn from regulatory guidance and the literature, not accounts of client work.
Setting the first dose in humans
The situation
The animal model works and GLP toxicology is done. Now the starting dose has to be defended in an IND or CTA.
What we do
- Predict human PK by allometry and by PBPK, then reconcile
- Find the driver of effect: peak, AUC, or time above threshold
- Translate on exposure, not dose
What you get
- Predicted human profile, assumptions stated
- Starting dose and escalation, with the rationale
- A Phase I sampling schedule that informs the model
Regulatory basis
The starting dose is among the first things a reviewer scrutinises. A rationale tied to the toxicology exposures is far easier to defend than an unexplained scaling factor.
Choosing the Phase 2 dose
The situation
First-in-human is complete. The single and multiple ascending dose data are in, and the Phase 2 dose now has to be chosen and justified.
What we do
- Characterise the PK profile across the dose range
- Find the sources of variability that actually matter
- Simulate candidate doses against the exposure target
What you get
- A population PK model, covariates documented
- The recommended Phase 2 dose, with the simulations behind it
- A Phase 2 sampling plan that keeps the model learning
Regulatory basis
FDA's population PK guidance describes using population analyses to identify the factors that explain variability in exposure and to inform dosing. Applied to the first-in-human data, it gives the Phase 2 dose a quantitative basis rather than resting on tolerability alone.
Have a question a model could answer?
Tell us about the molecule, the stage you are at, and the decision you need to make. We will come back with an honest read on what modeling can and cannot settle.