Pharmacometrics & FDA's MIDD Program

Pharmacometrics is the quantitative discipline that uses mathematical and statistical models — built from pharmacokinetic (PK), pharmacodynamic (PD), and disease-progression data — to characterize, predict, and optimize how a drug behaves in the body and in a population of patients. It converts data that would otherwise sit in separate study reports (a Phase 1 PK study here, a Phase 2 dose-ranging study there, preclinical toxicology somewhere else) into a single quantitative framework that can answer questions no individual study was designed to answer on its own: What dose should a pediatric patient with impaired renal function receive? Can a Phase 3 program be run with two dose arms instead of four? Does a drug-drug interaction predicted from physiology require a dedicated clinical trial, or can a model substitute for one?

When pharmacometric modeling is deliberately built into a drug’s regulatory strategy — not just to support internal development decisions but to inform an FDA review division’s own risk-benefit judgment — the field calls this Model-Informed Drug Development (MIDD). MIDD is pharmacometrics applied with a regulatory audience in mind, and FDA has built formal mechanisms, described below, for sponsors to bring MIDD approaches directly to the agency.

The Four Modeling Disciplines Pharmacometrics Draws On

“Pharmacometrics” is an umbrella term. In practice, a pharmacometrician reaches for one or more of four distinct modeling approaches depending on the question being asked:

Population pharmacokinetics (PopPK). Statistical modeling of drug-concentration data pooled across many patients, used to quantify how covariates — body weight, age, renal or hepatic function, genotype — explain variability in exposure. PopPK is the standard tool for justifying a dose adjustment in a subpopulation without running a dedicated trial in that subpopulation.
Physiologically based pharmacokinetic (PBPK) modeling. A mechanistic approach that represents the body as a set of physiologically defined compartments (organs, blood flows, enzyme and transporter abundance) and combines that with a drug’s physicochemical and metabolic properties to predict absorption, distribution, metabolism, and excretion. PBPK is FDA’s preferred tool for predicting drug-drug interactions (DDIs) computationally, and it is heavily used to extrapolate dosing to pediatric patients and other populations that are difficult or unethical to study directly.
Exposure-response (E-R) modeling. Links measured drug exposure (concentration over time, or a summary metric like AUC or Cmax) to a clinical outcome — efficacy, a biomarker, or an adverse event — to locate the dose range that maximizes benefit relative to risk. E-R modeling is the analytical core of dose selection and dose optimization; it is also the methodology underpinning FDA’s oncology dose-optimization initiative, Project Optimus, which pushed back against the historical practice of carrying the maximum tolerated dose from Phase 1 straight into pivotal trials without exposure-response justification.
Quantitative systems pharmacology (QSP). The most mechanistically ambitious of the four: QSP models represent the biological pathways a disease and a drug act on — receptor binding, signaling cascades, feedback loops — as a dynamic system, rather than treating the body as a set of compartments defined only by drug movement. QSP is used earliest in development, often to select a target, predict a first-in-human starting dose, or explain why a drug’s effect might differ across patient subgroups with different underlying biology.

These four are not competing methods; a single development program routinely uses several of them at different stages, and a MIDD meeting package (see below) may combine more than one to answer a single regulatory question.

Model-Informed Drug Development: Pharmacometrics Applied to Regulatory Strategy

MIDD is not a fifth modeling method — it is the practice of using the four disciplines above (individually or combined) to directly inform a specific development or regulatory decision, and of bringing that modeling to FDA as evidence rather than treating it purely as an internal R&D tool. Typical MIDD applications include:

Justifying a dose or dosing regimen for a pivotal trial from earlier, smaller studies, rather than running a full dedicated dose-ranging trial.
Extrapolating efficacy or a dosing regimen from an adult population to a pediatric population, per the extrapolation logic described in ICH E11A.
Predicting a drug-drug interaction from a validated PBPK model in place of a dedicated clinical DDI trial.
Simulating alternative trial designs — sample size, number of dose arms, endpoint timing — before committing resources to one design, a use case closely related to the adaptive trial designs and trial design fundamentals CASRAI already covers.
Supporting an accelerated-approval or rare-disease program where enrollment is inherently limited, an area also shaped by FDA’s Rare Disease Evidence Principles.

FDA’s MIDD Paired Meeting Program

FDA operationalized MIDD as a standing meeting mechanism, distinct from the general Type A/B/C/D meeting framework and from a routine pre-IND meeting, though a sponsor typically has already engaged FDA through those channels before requesting a MIDD meeting.

Origin and current authorization. The program began as a pilot under PDUFA VI (fiscal years 2018–2022) and was formally continued as the MIDD Paired Meeting Program under PDUFA VII, which covers fiscal years 2023 through 2027 (October 1, 2022 – September 30, 2027).
What “paired” means. A granted request gives the sponsor two connected meetings with the review division on the same development question — an initial meeting to discuss the proposed modeling approach, and a follow-up meeting once the analysis is complete — rather than a single one-off consultation.
Volume. FDA has publicly described accepting a limited number of paired-meeting requests each quarter under PDUFA VII, with room for additional proposals when agency resources allow; sponsors should treat this as a competitive, capacity-limited program rather than a meeting they are automatically entitled to, and should confirm current capacity directly with FDA or the relevant review division before planning around a specific number.
Eligibility. A sponsor generally needs an active investigational program — a pre-IND or IND on file — and a well-defined modeling question, not an open-ended request to “discuss MIDD.”
How to request one. The request is submitted through the sponsor’s existing regulatory correspondence channel with FDA (the same electronic submission pathway used for other IND correspondence), explicitly identified as a MIDD Paired Meeting Program request, and accompanied by a meeting package that states the specific regulatory question, describes the proposed model and the data supporting it, and explains how the modeling result would change a development or regulatory decision if FDA agrees with it.
Review divisions involved. Requests are evaluated by FDA’s clinical pharmacology and quantitative-methods reviewers within the relevant center — CDER for drugs, CBER for biologics — alongside the product’s regular review division.

Program mechanics change with each PDUFA reauthorization cycle; confirm current meeting volume, submission format, and eligibility directly against FDA’s own MIDD program page and the current PDUFA commitment letter before relying on a specific number for a real submission.

Where Pharmacometrics Shows Up Across a Development Program

Stage	Typical pharmacometric use
Preclinical / first-in-human	QSP or allometric scaling to predict a safe starting dose; early PBPK to anticipate metabolism and clearance.
Dose-finding (Phase 1/2)	Exposure-response modeling to narrow a wide dose range down to one or two doses worth carrying into pivotal trials — the logic behind Project Optimus in oncology.
Trial design	Clinical trial simulation to compare candidate designs (sample size, randomization ratio, interim look timing) before locking a protocol; overlaps with interim analysis planning.
Special populations	PopPK and PBPK to justify dosing in renal/hepatic impairment, pediatrics, or pregnancy without a dedicated trial in each subgroup.
Drug-drug interactions	PBPK modeling submitted in place of, or alongside, a dedicated clinical DDI study.
Regulatory submission	A consolidated pharmacometrics/clinical pharmacology summary within the submission package, and, where a MIDD meeting was used, the record of FDA’s agreement with the modeling approach.

Pharmacometrics vs. Biostatistics: Related, Not Interchangeable

The two disciplines are frequently confused because both are quantitative and both sit inside a clinical trial’s analysis plan. Biostatistics is primarily concerned with the design and inferential analysis of a single trial — hypothesis testing, Type I/II error control, the statistical analysis plan that governs how that trial’s own data will be analyzed. Pharmacometrics is primarily concerned with characterizing the drug itself — its PK/PD behavior — using data that may be pooled across multiple studies, and is oriented toward prediction (what would happen at a dose or in a population not directly studied) rather than only toward confirmatory inference on data already collected. In practice the two functions collaborate closely: a pharmacometric model often supplies the dose-justification and simulation work that a statistical analysis plan then builds its confirmatory design around.

Frequently Asked Questions

Is pharmacometrics the same thing as clinical pharmacology?

They overlap heavily but are not identical. Clinical pharmacology is the broader discipline covering a drug’s PK, PD, and interactions in humans generally, including studies that don’t involve formal modeling. Pharmacometrics is specifically the modeling and simulation toolkit within clinical pharmacology; most pharmacometricians work inside a clinical pharmacology function.

Do I need a PhD in pharmacometrics to request a MIDD meeting?

No — the request is made by the sponsor (typically through regulatory affairs, with clinical pharmacology/pharmacometrics staff or consultants preparing the technical package), not by an individual credentialed reviewer. What FDA evaluates is the quality and regulatory relevance of the proposed modeling question, not the requester’s personal credentials.

Can a PBPK model really replace a clinical drug-drug interaction study?

In defined circumstances, yes. FDA has published guidance describing when a verified, validated PBPK model is an acceptable substitute for a dedicated clinical DDI trial, particularly for interactions the model has been shown to predict reliably for the drug class in question. It is not a universal substitute — FDA still expects clinical confirmation for higher-risk or poorly characterized interactions.

Where does quantitative systems pharmacology (QSP) fit if population PK already exists for a drug?

They answer different questions. PopPK describes how exposure varies across the population you already have data on. QSP is used to extrapolate beyond existing data — predicting behavior in a mechanism, patient subgroup, or combination that hasn’t been directly tested — because it models the underlying biology rather than only the observed concentration data.

Related CASRAI Resources

Last verified: August 31, 2026, against FDA’s publicly described MIDD Paired Meeting Program mechanics (PDUFA VII, FY2023–2027) and cross-referenced summaries from the American Society for Clinical Pharmacology and Therapeutics (ASCPT). PDUFA program mechanics are re-negotiated on each reauthorization cycle; re-confirm current meeting volume and submission requirements directly against FDA’s own program page before relying on a specific number for a live submission.

What Is Pharmacometrics? FDA’s Model-Informed Drug Development (MIDD) Program

Ask about What Is Pharmacometrics? FDA’s Model-Informed Drug Development (MIDD) Program