Skip to main content

Use Cases

Where virtual patient modeling fits in clinical development

The platform applies at specific points in Phase I-II development where population heterogeneity could determine whether a trial reads out. Three distinct use cases, each with a different decision it supports.

Use Case 1

Target prioritization across a pipeline with shared indication

When two or more candidate compounds target the same indication via different mechanisms, heterogeneity in the development population creates differential risk. Simulation identifies which compound is best matched to the population you can actually enroll.

Development context

Two anti-inflammatory candidates targeting the same Phase II population

A translational team has two candidates, both advancing toward Phase II in a moderate-to-severe inflammatory indication. Available biomarker data suggests different mechanism-of-action profiles. The same trial population cannot support both programs simultaneously; one must go first.

Subgroup failure risk score for each candidate against the same virtual cohort
Biomarker strata where each candidate shows divergent predicted response
Recommended lead compound ranking with uncertainty bounds
Development context

Post-Phase I portfolio review before IND filing

Before filing the IND for a Phase II program, the development team wants to understand whether the intended trial population will produce a readable efficacy signal. Available Phase I pharmacokinetic and biomarker data provides the basis for population heterogeneity modeling.

Predicted responder subgroup definition based on Phase I biomarker data
Enrichment criteria options with projected power improvement estimates
Sensitivity analysis over the key biomarker assumptions in the model

Use Case 2

Enrichment strategy development before protocol lock

Protocol inclusion/exclusion criteria are often set based on prior Phase II experience or investigator intuition. Simulation gives the study team a quantitative basis for evaluating enrichment options before the protocol is submitted.

Challenge

A broad eligibility population diluting predicted response signal

The planned trial population for a CNS indication is broad, consistent with regulatory guidance on inclusivity, but preclinical data suggests a responder subgroup defined by a specific receptor expression profile. Narrowing eligibility may improve signal but reduce enrollment speed. The study team needs a quantified estimate of the trade-off.

Predicted trial success probability under broad versus restricted eligibility definitions
Estimated screening-to-enrolment ratio under candidate enrichment criteria
Biomarker-defined protocol language for inclusion/exclusion review
Challenge

Identifying the non-responder subgroup to exclude rather than enrich for

Rather than defining who should be included, the clinical team wants to understand which patient population should be excluded from a Phase II immunotherapy trial. The compound shows broad activity, but a comorbidity-linked subgroup appears to drive efficacy signal collapse in the planned population model.

Comorbidity and concomitant medication axes driving predicted non-response
Exclusion criteria formulation with predicted improvement in trial response signal
Estimated fraction of intended population affected by proposed exclusions

Use Case 3

Rare disease site selection using simulation-first feasibility

In rare disease programs, the eligible patient population may number in the hundreds nationally. Determining which sites can actually enroll the required patient profile is a critical pre-trial planning step that conventional feasibility surveys handle poorly.

Challenge

A gene-defined subtype with inconsistent diagnosis rates across potential sites

A rare metabolic disorder trial requires patients with a specific gene variant confirmed by an assay not uniformly available across sites. Site feasibility questionnaires consistently over-estimate the number of eligible patients in the pipeline. The sponsor needs a model-based estimate of site-level patient availability before committing to site contracts.

Simulated eligible patient count distributions by regional epidemiology
Diagnosis rate adjustment model incorporating assay availability at candidate sites
Site prioritization ranking with confidence intervals on enrollment projections
Challenge

Pediatric indication with age-banded enrollment constraints

A pediatric rare disease program requires enrollment across three distinct age bands, each with different eligibility criteria and different concentrations in specific site types. Historical enrollment in adult studies at candidate sites provides limited signal for pediatric feasibility. Simulation of the target population by age band supports a data-driven site shortlist.

Age-stratified patient availability models for each site type under consideration
Enrollment timeline projections under optimal versus conservative site selection scenarios
Minimum number of sites required to hit enrollment targets with 80% probability

Get Started

Tell us about your development program

We work with a small number of programs at a time. If one of these use cases maps to a decision your team is facing before Phase II, we want to hear from you.

Request Early Access Learn how the platform works

Common questions from development teams

The platform has been applied most extensively in oncology and immunology, where published biomarker-response relationships provide the densest literature prior for conditioning the response surface model. We are extending coverage to CNS and rare metabolic disorders. If your indication falls outside these areas, we assess feasibility based on available biomarker data quality during the intake conversation.
The minimum is aggregate statistics from 200 real patient records. This provides enough coverage to fit a Bayesian network with meaningful biomarker-level conditioning. Below this threshold, uncertainty bounds in the output expand substantially and enrichment recommendations become less specific. We report confidence intervals explicitly in all outputs so development teams can assess the practical utility given their data situation.
Our output reports are designed to be included as supporting documentation in regulatory submissions where modeling and simulation evidence for trial design decisions is appropriate. Agencies including FDA and EMA have issued guidance on the use of M&S in clinical development. We provide a full methodology document with each report suitable for inclusion in an IND or clinical study report appendix.
Typical turnaround is 5 to 10 business days from intake of a complete data package. Complex programs with multiple candidate compounds or sparse input datasets take longer. We discuss timeline expectations during the access request intake and confirm before work begins. We do not guarantee turnaround times that would require us to cut corners on uncertainty quantification.