Clinical Pharmacology Portfolio

Areas of practice

The recurring problems this portfolio covers — the questions that decide whether a program moves forward, at what dose, and with what evidence.

DOSE SELECTION

First-in-human starting dose & escalation

Integrating NOAEL- and MABEL-based approaches, safety margins, and pharmacologically active exposure ranges into starting doses and escalation schemes that protect subjects without stalling the program.

MODELING & SIMULATION

Population PK & exposure–response

Building popPK and ER models that turn sparse early-phase data into dose-range decisions, covariate strategies, and defensible Phase 2/3 dose justifications.

DRUG INTERACTIONS

PBPK-based DDI risk assessment

Using mechanistic static and PBPK models to predict interaction magnitude, design (or waive) clinical DDI studies, and draft label language regulators can accept.

TRANSLATIONAL

Translational & biomarker strategy

Linking nonclinical PK/PD to first human readouts: target-engagement biomarkers, proof-of-mechanism design, and go/no-go criteria set before the data arrive.

SPECIAL POPULATIONS

Organ impairment, food effect & formulation bridging

Planning hepatic/renal impairment, food-effect, and bioequivalence strategies — and knowing when a model can answer the question instead of another study.

REGULATORY

Regulatory interaction & dose justification

Preparing clinical pharmacology packages for agency meetings and submissions: the dose rationale, the DDI plan, and the modeling evidence, written to be reviewed.

Case studies

The four studies below are worked examples — realistic clinical pharmacology decision problems written out in full, so you can see the format, depth, and sanitization standard this page uses. Program identities, compounds, doses, and timelines are removed; the pharmacology reasoning is intact. Replace each example with your own case and it drops straight into the same structure — and keep the “Withheld” line, which tells readers exactly what was taken out.

Choosing a first-in-human starting dose when the toxicology margin was thin

EXAMPLEFIRST-IN-HUMANSMALL MOLECULE

Situation

A novel small molecule entered first-in-human planning with a narrow gap between the exposure expected to be pharmacologically active and the exposure at the nonclinical no-observed-adverse-effect level. A conventional NOAEL-based starting dose, divided by a standard safety factor, risked landing well below any informative exposure — while a more aggressive start was hard to defend.

Approach

Both anchors were built in parallel: a NOAEL-based maximum recommended starting dose and a MABEL estimate from in-vitro potency and receptor-occupancy modeling, translated to human exposure with predicted PK. The escalation scheme used sentinel dosing, exposure-triggered stopping rules, and cohort spacing tied to emerging half-life rather than fixed calendar intervals.

Decision & outcome

The MABEL-anchored dose was selected as the start, with the NOAEL ceiling kept as a hard cap for escalation. The single-ascending-dose study reached pharmacologically relevant exposures without dose-limiting findings, and the observed PK matched predictions closely enough that the multiple-dose design needed no structural changes.

Withheld:sponsorcompound & targetindicationdose levelsexposure valuesstudy dates

When the drug-interaction risk lived in a metabolite, not the parent

EXAMPLEDRUG INTERACTIONSPBPK

Situation

Standard in-vitro screening cleared the parent compound as a low DDI perpetrator risk. But a major circulating metabolite — present at higher exposure than parent — showed time-dependent inhibition of a key metabolic pathway in vitro, raising the possibility of a clinically relevant interaction the parent-only assessment had missed.

Approach

A combined parent–metabolite PBPK model was built and verified against single- and multiple-dose clinical PK, including the metabolite-to-parent exposure ratio across the dose range. Simulations tested the interaction against sensitive index substrates across plausible dose regimens and in virtual populations with reduced pathway activity.

Decision & outcome

The model predicted an interaction below the threshold that would require dose adjustment, but large enough to characterize. Rather than a dedicated cocktail study, a focused clinical assessment was embedded in an ongoing trial, and the prediction informed draft label wording — a plan the regulatory feedback accepted without requesting an additional standalone study.

Withheld:sponsorcompound & metaboliteenzyme pathwayinhibition constantspredicted AUC ratios

An exposure–response analysis that narrowed a Phase 2 dose range

EXAMPLEEXPOSURE–RESPONSEPOPULATION PK

Situation

An early-phase program carried a wide, cautiously chosen dose range into Phase 2 planning. Efficacy signals were encouraging but noisy, a dose-related tolerability trend was emerging, and the team needed a principled way to drop the doses that were unlikely to help and keep the ones that might.

Approach

Individual exposures were derived from a population PK model and related to both efficacy biomarkers and safety events. The analysis separated the two relationships: efficacy exposure–response that plateaued, against a safety relationship that kept climbing — with body weight and organ function tested as covariates that might otherwise masquerade as dose effects.

Decision & outcome

The upper part of the range offered no predicted efficacy gain over the middle doses while carrying the steepest safety slope, so the Phase 2 design concentrated on the lower two doses with an exposure-based monitoring rule. The analysis became the quantitative core of the dose-justification document for the regulatory interaction that followed.

Withheld:sponsorindication & endpointdose levelsexposure metricsevent rates

Bridging a formulation change with a model instead of a second bioequivalence study

EXAMPLEFORMULATION BRIDGINGABSORPTION MODELING

Situation

A manufacturing-driven formulation change arrived after early clinical work was complete. A conventional bioequivalence study was possible but would have consumed a slot in the critical path — and the compound's absorption was dissolution-limited, making the outcome genuinely uncertain rather than a formality.

Approach

A mechanistic absorption model was built from solubility, dissolution, and permeability data and qualified against the existing clinical PK of both the old formulation and a food-effect arm. Virtual bioequivalence trials then stress-tested the new formulation across realistic variability in gastric pH, transit time, and prandial state.

Decision & outcome

Simulations predicted the exposure difference would stay within the conventional equivalence window under nearly all tested conditions, with food state as the dominant sensitivity. The program proceeded with a small, targeted confirmatory study under fed conditions instead of a full two-arm program, keeping the development timeline intact.

Withheld:sponsorcompoundformulation detailsdissolution datapredicted exposure ratios

See the sanitization in action

The same paragraph, before and after. Toggle to compare how program-specific detail is converted into publishable science.

Experience

Roles are described by scope and function. Employer names, program names, and dates are added only where they are already public.

[YYYY – present]

[Role title] — Clinical Pharmacology, [organization type]

[Scope: development stage, modalities, team size, and the decisions this role owned — e.g., first-in-human dose strategy across early-phase programs.]

[YYYY – YYYY]

[Role title] — Quantitative Pharmacology, [organization type]

[Scope: modeling responsibilities, regulatory interactions supported, and therapeutic areas — described at program-type level.]

[YYYY – YYYY]

[Role title] — [function], [organization type]

[Scope: earlier-career experience — clinical study support, PK analysis, and first modeling work.]

EDUCATION

[Degree, field — institution]

[Thesis area, fellowships, certifications, and professional society roles.]

[highlighted fields] mark where your details go — send them over and they drop straight in.

How these case studies are shared

Public portfolios and sponsor confidentiality can coexist — with a protocol. Every case study on this page passes all five rules before it is published.

Generalize the program, keep the science

Compounds become modality classes ("a small molecule"), indications become development stages, and doses become relative ranges. The pharmacology — margins, model structure, decision logic — survives intact because it belongs to the discipline, not the asset.

Numbers become relationships

Absolute exposures, dose levels, and event rates are replaced by comparisons: "below the no-effect exposure," "plateaued while risk kept climbing." Readers learn how the decision was made without learning anything that identifies or advantages a program.

Nothing unpublished about an active program

Work is only described once its key facts are already public — a disclosure, a publication, a registry entry — or it is held back entirely. If a detail could move a competitor's decision and isn't public, it isn't here.

No attribution, no timelines

Sponsors, employers, collaborators, and calendar dates are removed or blurred. A case study should be impossible to attach to a company by triangulating the details that remain.

Employer policy wins

Where an employer's disclosure or publication policy applies, it governs. When in doubt, the case study is delayed, generalized further, or not published at all.

Interested in the full conversation?

Case studies here are deliberately kept at the level of the decision. The full walk-throughs — data, alternatives considered, and what was argued in the room — belong in interviews and professional discussions, not on a web page. Get in touch.

Email LinkedIn