Data license + evidence platform
Cancer digital twins for oncology development.
Evinexus provides licensed, protocol-aligned oncology cohorts and a platform for examining outcomes and study scenarios.
Protocol-aligned cohort
Advanced solid tumor · 2L+
N = 268 · median age 55 · 47% male
Covariate balance, before and after weighting
Foundations
What is a digital twin?
An Evinexus digital twin is an RWD-derived representation of a patient or cohort, aligned to a defined oncology research question and used to estimate expected outcomes under specified conditions.
For a sponsor, the deliverable is a patient-level cohort built around a specified population, endpoint, and intended use.
Methodology
From research question to digital twin cohort
A reproducible workflow turning standardized real-world and trial data into validated, physician-curated digital twin cohorts.
Pipeline: a research question enters Phase I and a validated digital twin cohort leaves after Phase III, moving through three ordered phases and one sealed proprietary processing step.
- Phase I · Define & ingest
- Step 01: Clinical question. Indication, comparator, endpoints, target estimand.
- Step 02: Standardized inputs. Harmonize EHR, registries, trials and omics to OMOP CDM.
- Evinexus proprietary: Digital twin generation & validation. Standardized inputs run through Evinexus's proprietary digital-twin generation and validation modeling.
- Phase II · Curate & scope
- Step 03: Physician curation. Human-in-the-loop review of plausibility and phenotypes, with sign-off.
- Step 04: Eligibility criteria. Apply protocol inclusion and exclusion to match the target trial.
- Phase III · Deliver
- Step 05: Generate twins. Patient-level twins for external or hybrid control arms.
- Step 06: QC & delivery. Disclosure control, audit trail and sign-off; CDISC ADaM datasets.
Cross-cutting governance — applied at every step
- Human-in-the-loop: Physician oversight gates every output.
- Validated & calibrated: Fidelity and calibration quantified.
- Privacy by design: Disclosure control against re-identification.
- Reproducible: Versioned provenance and audit trail.
Product components
Two connected product layers.
Digital Twin Data
Licensed cohorts aligned to the study population, endpoints, provenance requirements, and intended use.
View data pageEvidence Platform
Tools for reviewing cohort characteristics, outcomes, safety, and study scenarios.
View platform pageEvidence Platform
Examine the cohort, outcomes, and assumptions together.
Select an endpoint to update the survival curves, arm-level statistics, and effect estimate for the same matched cohort.
Illustrative example — clinical outcomes
Overall survival
Number at risk
| Arm | Events / N | Median mo (95% CI) | 1-yr rate (95% CI) |
|---|---|---|---|
| Treatment A | 620 / 888 | 16.8 (15.5–19.1) | 69.9% (66.8–72.9%) |
| Treatment B | 590 / 896 | 27.4 (25.9–30.1) | 82.7% (80.1–85.0%) |
HR (95% CI): 0.60 (0.54–0.67), p=<0.001
Illustrative example, not a validated result.
Method & delivery
Designed for sponsor review and reproducibility.
Over 20,000 digital twins, capturing 20+ types of cancer.
Target-trial emulation
Target-trial-emulation design aligned to the defined study question.
Delivered as CDISC ADaM
Standardized CDISC ADaM delivery with supporting metadata.
Versioned and auditable
Versioned provenance and documented quality review.
Supporting capabilities
The work behind the comparator.
Medical coding, data labeling, algorithm development and validation, and clinical annotation make up the versioned, auditable provenance behind the values in our digital twins. Evinexus can provide these additional services to support your study. Contact us to learn more.
- Medical Coding & Phenotype Definition
- Algorithm Development & Validation
- Data Labeling & Annotation for AI & RWE
- Study Design, Analytics & Scientific Communication
- Patient Profile Curation
Short FAQ
Digital twins, RCTs, and real-world data
Do digital twins replace randomized controlled trials?
No. A traditional randomized controlled trial remains the confirmatory standard. Depending on the research question and available evidence, digital twins may support trial planning, prognostic adjustment, or an external- or hybrid-control strategy.
How are digital twins different from real-world data?
Real-world data are an input. Evinexus digital twins are question-specific patient or cohort representations constructed from de-identified, multimodal real-world data. For regulatory external-control analyses, the underlying real-world data and observed outcomes provide the evidence base.
When would a sponsor use each design?
The choice depends on whether randomization is ethical and feasible, whether comparable external data are available, and whether the endpoint, expected treatment effect, and standard of care support the use of external evidence. A traditional RCT uses a full concurrent randomized control. A hybrid design combines a smaller randomized control with external evidence. A fully external-control design may be considered when the evidence and regulatory context justify proceeding without a concurrent randomized control.
Start with a feasibility review.
Evinexus reviews the target population, required covariates, endpoints, and follow-up before proposing a cohort build.