Twin ambitions

A new vision and the digitalisation of clinical trial design

A single poorly designed clinical trial can cost pharmaceutical companies hundreds of millions of pounds, years of development time or, in some cases, prevent a new treatment from ever reaching patients.

‘Virtual twins’ aim to prevent avoidable trial failures by allowing sponsors to test, refine and optimise trial designs before they begin. This reduces risk, accelerates timelines and supports innovation across the pharmaceutical ecosystem.

The concept of a virtual twin is rooted in product life cycle management, historically used to design complex products such as aircraft.

Previously, development required building physical prototypes for testing, with limited ability to pressure test designs early. In aviation, for example, testing a newly designed plane required building it first – an expensive and time-consuming process.

Virtual twins transformed this by creating sufficiently rich in silico models, enabling testing and visualisation before production.

Similarly, a virtual twin in clinical research is a digital model of an entire trial. It allows earlier testing of design choices, such as patient numbers or visit schedules, rather than discovering preventable causes of delay, cost or failure once the trial is underway.

The power of virtual twins rests on sophisticated data architecture with two essential components.

First, the underlying data structure contains all the information needed to run a clinical trial. What patients need to be enrolled? What materials are required?

What skills are needed? How will data be collected and processed?

Many parameters are coupled. If an endpoint is adjusted, it affects activities, data collection, cost, operational complexity and participant burden. A virtual twin shows how changing one parameter affects the whole trial.

Second, these models are underpinned by historical data from hundreds of thousands of trials across many therapeutic areas.

In a lung cancer trial, for example, the system recognises that endpoints relate to tumour growth, implying specific imaging requirements.

Sponsors know the cost and burden of CT scans and can use this to understand trial expense or how design choices may affect enrolment and retention.

Historical data helps designers build real-world expectations and configure processes more quickly.

On a practical level, the benefits of virtualisation in clinical trials are threefold.

First, trial designers can pressure test every design decision. This helps identify and mitigate challenges that could harm participants or risk trial success before they materialise.

Second, virtual twins democratise trial design. Traditionally, only the designer had a holistic view. Virtualisation allows more stakeholders – patient groups, clinical teams and others – to input on feasibility, objectives and regulatory acceptability. This strengthens design and helps identify challenges earlier.

Third, virtual twins streamline document management and change control. Traditional methods require manual drafting and updating of every component.
Virtual twins automate this, reducing administrative burden and lowering the risk of human error.

While the above advantages benefit patients indirectly, perhaps the most profound impact lies in optimising the patient experience.

Because systems contain historical data and linkages, virtual twins help sponsors design trials that are easier to enrol in, more likely to achieve compliance and less prone to drop-outs. This includes considering visit frequency, procedures, discomfort, inconvenience and protocol adherence challenges.

This can generate actionable recommendations, such as consolidating clinic visits or incorporating wearables to collect data at home.

There is a growing desire to embed patient perspectives throughout trial design. Virtual twins allow designers to incorporate patient advocate recommendations directly, translating insights into concrete design modifications.

Looking ahead five to ten years, the capabilities and value of virtual twins will expand. Drug development is inherently complex, with countless potential failure points.

AI-powered virtual twins can identify where a drug is likely to fail before it happens. Future systems may run millions of simulations at once, identifying the pathway with the highest chance of success. In this scenario, virtual twins will further strengthen and accelerate drug development.

The emergence of virtual twins fundamentally shifts how clinical trials are conceived, designed and executed. By enabling comprehensive pre-production testing, democratising design collaboration, automating documentation and optimising patient experience, virtual twins address challenges that have constrained clinical research for decades.

The adoption of digital innovation during clinical research has the potential to unlock further innovation across the drug development life cycle, bringing life-saving treatments to patients sooner.