n the evolving landscape of scientific research and healthcare, technologies like artificial intelligence (AI), machine learning (ML), and biosimulation are reshaping how data is managed, processed, and utilized. In silico technologies (IST) represent a key component of this transformation, providing a groundbreaking acceleration of predictive intelligence, insights, and evidence, revolutionizing traditional research and development (R&D) by leveraging advanced computational techniques.

These technologies have demonstrated their potential to deliver rapid and cost-effective advancements, gaining significant support from regulatory bodies since 2005. Among these supporters is the US Food and Drug Administration (FDA), which has increasingly endorsed Model-Informed Drug Development (MIDD) as a critical tool. A surge of new guidance and best practices has been published to help bridge policy to practice, including Toward Good Simulation Practice (2024), which features contributions from 13 FDA experts along with leading data science and platform experts.

By integrating these digital approaches, organizations can accelerate innovation, improve data handling, and harness complex computational models to achieve breakthroughs and secure market leadership in drug approval.

This article explores the evolution from traditional in vivo (within the living organism) and in vitro (outside the living organism, in an artificial environment) methods to advanced in silico (within the computer) approaches, highlighting their significance and future potential.

WHAT: In Silico Technologies (IST)

ISTs involve using computer-based algorithms to replicate and study complex biological systems, with the term “in silico” originating from silicon, the key material in computer chips. Unlike traditional lab-based experiments, in silico approaches (ISAs) can predict the behavior of biological entities under various conditions without the need for physical experiments.

Evolution from In Vivo to In Silico: The journey of scientific experimentation began with in vivo methods, where studies were conducted within living organisms. While effective, these methods were often slow, expensive, and ethically challenging. In vitro techniques allowed experiments to be conducted in controlled laboratory environments, reducing some ethical concerns but still presenting limitations in terms of cost and scalability. The progression to in silico methods has revolutionized the field by offering a faster, more cost-effective, ethical alternative.

In the traditional drug development process, the journey from discovery to market involves several lengthy phases. The initial discovery and pre-clinical phase requires extensive laboratory work, animal testing, and early-stage trials. Following this, the clinical trial phases are time-consuming, with phase 1 typically taking 32 months, phase 2 about 39 months, and phase 3 around 40 months. After these trials and regulatory approval, the drug finally goes to market. The patent duration usually spans 20 years from the filing date, but the market exclusivity period in practice is only about eight (8) years once the drug reaches the market.

Figure 2: Traditional versus in silico drug development.

In contrast, ISTs have the potential to revolutionize drug development by significantly accelerating timelines. Current evidence shows that these approaches can reduce the time to market by several years compared to traditional methods (the exact time savings can vary depending on the specific application and drug being developed). For example, studies have demonstrated that in silico modeling can streamline early drug development phases by optimizing clinical trial design, predicting drug behavior, and supporting regulatory submissions.

Case Study Overview

The case study showcases the use of in silico models to accelerate the development of a novel peptide for neurodegenerative diseases, specifically ALS, by integrating virtual patients into clinical trials to predict disease progression. In silico modeling and quantitative systems pharmacology (QSP) were employed to optimize clinical outcomes.

Specifically, clinical optimization involved fine-tuning dose regimens and linking peptide concentrations to efficacy biomarkers, with a synthetic control arm used to improve study precision. Additionally, mechanisms for expanded indications, such as multiple sclerosis, Alzheimer’s, and Parkinson’s, were evaluated, allowing for the exploration of additional market potential without the need for lengthy clinical trials.

The strategic advantages of this approach include enhancing the value of clinical trials, derisking development plans, and increasing attractiveness to investors and pharmaceutical partners.

Application of in silico technologies in the development of this molecule provides a tangible example of how these methodologies can be applied to reduce reliance on traditional animal and human studies. For instance, during the phase 1 study, in silico pharmacokinetic (PK) and pharmacodynamic (PD) modeling was utilized to correlate drug concentrations with efficacy biomarkers.

This modeling not only optimized the dose regimen for the subsequent phase 2 study but also suggested the inclusion of a synthetic control arm, which could reduce the need for a larger control group of human subjects by augmenting it with virtual placebo patients constructed via machine learning models. The synthetic control arm, validated against real patient data, significantly enhanced the statistical power of the study while minimizing the number of actual patients required. This approach exemplifies how in silico methods can streamline drug development, reduce costs, and mitigate ethical concerns associated with extensive human or animal testing.

Perpetual Refinement

In silico approaches serve as a vital complement to in vivo and in vitro data, providing a holistic and dynamic method for model refinement and validation. The perpetual refinement cycle involves several critical steps that ensure continuous improvement and accuracy of the models. The process can be visualized in the accompanying diagram (Figure 3):

• Model Construction (Based on available data): Initially, a model is constructed using the current available data. During the pre-clinical phase of drug development, such data is derived from animal or in vitro experiments. Typical types of data collected at this stage may be drug concentrations measured in plasma or other tissues, receptor occupancy, and biomarkers of drug effect. Similarly, PK and drug-effect data are collected from human participants during clinical development. Additionally, patient outcome data is generally collected and modeled. This foundational step sets the stage for subsequent predictive capabilities and refinements.
• Prediction Phase (Extending beyond current data): Once constructed, the model is employed to make predictions that extend beyond the existing data. This can involve various scenarios such as different drugs, dosages, or populations, providing a broader applicability and testing ground for the model.
• Experimental Validation (Obtaining new data): To validate the predictions made by the model, additional experimental data is collected. Depending on the stage of development, validation data may originate from in vitro experiments, animal studies, or human clinical trials. The types of data collected are generally the same as those utilized during model construction. This step is crucial for assessing the accuracy and reliability of the model’s forecasts.
• Model Refinement (Addressing discrepancies): The final step involves refining the model based on any discrepancies identified between the predicted and observed data. This refinement is essential for improving the model’s precision and reliability, effectively bringing the cycle back to the initial construction phase with enhanced data and insights.

SO WHAT: The Transformative Impact of ISTs on Drug Development

In the realm of drug development, assessing the potential toxicological risk of a drug candidate early on is crucial to reducing costs and expediting the process. Traditional approaches still rely heavily on in vivo and in vitro methods to evaluate these risks. However, in silico methods have emerged as a powerful complement, offering significant advantages in clinical trials.

In silico clinical trials have demonstrated their effectiveness in speeding up the market entry of new treatments while cutting costs. Michael Hill (Vice President, Science, Technology, and Clinical Affairs at Medtronic), in In Vivo, In Vitro, In Silico: Why Computer Modelling Is the Next Evolution of the Healthcare Sector, illustrated the concrete impact of ISTs on providing regulatory evidence:

• Accelerated Market Entry: His company’s product was released and achieved market dominance two years earlier than expected.
• Reduced Patient Enrollment: The clinical study required 256 fewer patients compared to traditional trials.
• Cost Savings: $10 million saved due to reduced patient numbers and two years of market dominance.
• Patient Treatment: 10,000 patients treated in the first two years of market dominance.

Integrating in silico methods into clinical trials optimizes the drug development process and demonstrates significant regulatory impact. Figure 4 showcases how the FDA has adopted ISTs across different aspects of product lifecycles and through its centers and functions. This adoption highlights the growing acceptance and importance of ISTs in regulatory frameworks, ensuring that new treatments are both effective and safe for public use.

NOW WHAT: Moving Forward with ISTs

As the integration of in silico methods continues, it is essential to consider the practical steps necessary to maximize their potential. Embracing these technologies involves not only understanding their current applications but also anticipating future trends and preparing the next generation of healthcare professionals and researchers. By learning from past experiences and fostering collaboration across disciplines, we can ensure that in silico modeling becomes an integral part of scientific research and medical practice.

Real-World Data (RWD) Insights to Impact Across the Value Chain

ISTs offer essential predictive intelligence to enhance performance outcomes throughout the entire drug development lifecycle, from discovery and clinical trials to development strategies. Researchers now have access to sophisticated data models that integrate digital twins and virtual patient simulations, enabling them to simulate disease progression. This empowers sponsors to leverage comprehensive, robust real-world data (RWD) with higher statistical power, facilitating better decision-making across the value chain for improved return on investment (ROI) and strategic advantage.

Figure 5: IST across the drug development value chain.

Discovery

• Predict target engagement and pharmacokinetics (PK), which is the study of how drugs move through the body, based on chemical structure
• Optimize and design novel treatments, prioritize pipelines, and enhance portfolio strategies.

Pre-Clinical

Provide a sustainable alternative to animal testing by utilizing digital twins, which are virtual representations of human biology that can simulate individual patient responses to drugs or treatments. This approach aligns with the FDA Modernization Act 2.0, passed in 2022, which supports the use of “alternative methods” or “non-animal testing methods” that could include a range of innovative technologies, such as computer modeling, organ-on-a-chip, and other methods that can potentially replace or reduce animal testing. The act encourages the adoption of these advanced technologies while ensuring safety and efficacy.

Clinical

• Maximize trial success probability through simulation-based design optimization
• Expedite study recruitment and conduct by augmenting real-world patient data with virtual patient data
• Optimize drug label model-informed dose selection.

Development

Determine drug repurposing opportunities: Focus on drug repurposing in areas like neurodegenerative diseases (e.g., Alzheimer’s, Parkinson’s) and oncology, where existing drugs have shown promise in new therapeutic roles. For instance, studies have indicated that drugs like metformin, traditionally used for diabetes, may have antitumor properties and are being investigated for cancer treatment. Additionally, the antiviral drug remdesivir, initially developed for hepatitis C, was successfully repurposed for COVID-19 treatment, demonstrating the potential for quick deployment in emerging health crises.

Next Trends

Future trends indicate a continued expansion of in silico methods in drug development. With advances in artificial intelligence (AI) and machine learning (ML), these models are expected to become even more accurate and predictive. This will likely lead to more personalized medicine approaches, where treatments are tailored to individual genetic profiles, and a reduction in the need for extensive human and animal testing. Furthermore, the integration of big data and real-world evidence into in silico models will enhance their capability to simulate complex biological systems and predict long-term outcomes, revolutionizing the landscape of clinical trials and regulatory science.

Conclusion and Future Considerations

Pharmaceutical companies are increasingly prioritizing ISTs in drug development. In silico methods are crucial for drug discovery and clinical research due to their cost efficiency, ethical benefits, and speed. With rapid technological advances, these methods are set to become fundamental, driving innovation and personalized medicine while unlocking new opportunities in the biomedical field.

To facilitate a streamlined regulatory process, regulators and policy makers should:

• Encourage the use of modeling and simulation approaches in product development
• Harmonize model acceptance and review processes
• Accept digital evidence for all health-regulated products
• Support the adoption of Good Simulation Practices (GSP)
• Provide resources and incentives.

Similarly, industry should:

• Adopt in silico tools and demonstrate benefits through use cases.
• Build awareness.
• Submit in silico evidence to regulators as supportive data.
• Collaborate in funded projects, consortia, and alliances.