The evidence landscape for drug development is rapidly evolving. While randomized controlled trials (RCTs) remain the cornerstone of regulatory and health technology assessment (HTA) decision-making, advances in computational science are creating entirely new forms of evidence. Among the most promising developments are in-silico data and model-informed drug development (MIDD) approaches, which use mathematical models, simulation, and virtual patient populations to generate insights that complement traditional clinical evidence.

As healthcare systems seek faster access to innovation while maintaining rigorous standards for demonstrating value, these novel evidence sources may play an increasingly important role in HTA.

What are in-silico data and model-informed drug development?

The term in-silico refers to experiments or analyses conducted using computer-based simulations rather than laboratory (in vitro) or human (in vivo) studies. In drug development, in-silico methods can integrate biological, clinical, epidemiological, and real-world data to simulate disease progression, treatment response, and patient outcomes.

MIDD extends this concept by using quantitative models to support decisions throughout the product lifecycle. These models may incorporate pharmacokinetic and pharmacodynamic relationships, disease progression modeling, systems biology, machine learning algorithms, or virtual patient simulations. Rather than replacing clinical trials, MIDD is designed to improve study design, optimize dosing strategies, predict outcomes in underrepresented populations, and address evidence gaps that would be difficult or impractical to investigate empirically.

Regulatory agencies, including the U.S. Food and Drug Administration (FDA) and the European Medicines Agency (EMA), have increasingly recognized the value of MIDD in supporting regulatory decision-making, particularly in rare diseases, pediatric populations, and situations where conventional trials face ethical or operational constraints.

The publication of the latest guidance from ICH on MIDD and its role (ICH M15) signals the increasing interest from regulators in this novel type of evidence. The UK in-silico Centre of Excellence in Regulatory Science and Innovation (CEiRSI), established with support from the UK regulator (MHRA) and NICE, has also published a risk-informed framework for in-silico regulatory evaluation of medical devices. A recent assessment by the Office of Health Economics (OHE) of the impact of establishing this center on the UK economy has counted, amongst its benefits, enabling faster access to treatments for more diverse patient populations and reducing reliance on animal and clinical testing while supporting more sustainable healthcare innovation.

Why does this matter for HTA?

HTA agencies are frequently required to make reimbursement and coverage decisions based on incomplete evidence. Uncertainty often exists around long-term outcomes, treatment effectiveness in routine clinical practice, or impacts on patient populations that were not adequately represented in clinical trials.

In-silico evidence can help address several of these challenges.

1. First, simulation models can generate estimates of long-term treatment effects beyond the duration of clinical studies. This is particularly relevant for therapies intended to modify disease progression, where meaningful outcomes may take many years to observe.
2. Second, virtual patient populations can explore treatment effects in subgroups that are often underrepresented in clinical trials, including older adults, patients with multiple comorbidities, and individuals from diverse demographic backgrounds. Such analyses may provide valuable insights into heterogeneity of treatment effects, a growing area of focus for HTA bodies.
3. Third, in-silico approaches can strengthen economic evaluations. Many cost-effectiveness models already rely on assumptions regarding disease progression and treatment pathways. Mechanistic and disease-simulation models may improve the scientific credibility of these assumptions and reduce the economic model structural uncertainty.

The opportunities and challenges ahead

Despite their promise, in-silico data are unlikely to become a standalone source of evidence for HTA in the near future. Decision-makers will continue to require transparency, validation, and clear demonstrations that model predictions align with observed clinical outcomes.

Questions remain regarding model credibility, reproducibility, uncertainty characterization, and the standards needed to evaluate computational evidence consistently across jurisdictions. HTA agencies will need robust methodological frameworks to assess when and how in-silico evidence should influence reimbursement decisions.

Nevertheless, the direction of travel is clear. As precision medicine, gene therapies, and increasingly targeted interventions challenge the limits of traditional evidence generation, in-silico data and MIDD evidence will complement clinical and real-world evidence. For HTA, the opportunity is not simply to adopt new forms of evidence, but to develop frameworks that can harness their potential while maintaining the rigor and transparency that healthcare decision-making demands.

The NICE HTA Lab report

Recent work by the NICE Health Technology Assessment Innovation Laboratory (HTA Lab) highlights both the potential and the challenges of incorporating in-silico evidence into HTA decision-making. In its report, Appendix E: Principles and Considerations for Use of In-silico Data in HTA Decision Making, NICE concludes that in-silico trial (IST) evidence should generally be viewed as a complementary source of evidence rather than a replacement for clinical studies. The report proposes a series of principles covering the positioning of in-silico evidence within the broader evidence base, expected evidence standards, transparency and reporting requirements, risk-based credibility assessment, and ethical considerations. Importantly, NICE emphasizes that the value of in-silico data depends on the model being fit for purpose, scientifically validated, transparent, reproducible, and linked to empirical evidence wherever possible.

The report also highlights the potential for in-silico approaches to address evidence gaps in rare diseases, small patient populations, and long-term outcome assessment, while supporting analyses of population heterogeneity and treatment effects in understudied groups.

For HTA stakeholders, the report signals a growing recognition that computational evidence may contribute meaningfully to future evaluations, provided that its credibility, uncertainty, and relevance to decision-making are assessed using robust and proportionate frameworks.

Final takeaways

The future of evidence generation will likely be a hybrid one, where clinical trials, real-world data, and computational models work together to provide a more comprehensive understanding of value. HTA stakeholders should begin preparing now for a world in which virtual evidence plays an increasingly important role in informing real-world decisions.

Interested in learning more?

Dalia Dawoud will be attending ISPOR Europe in November. Stop by Booth #1315 or click below to book a meeting:

Share Via