Verisimlife

Introduction

The pharmacological effect of any potential drug candidate must always be carefully vetted for safety and efficacy.

In the context of drug development, efficacy is defined as a medicine’s ability to produce a desired effect or treat a specifically indicated condition. Measured under expert supervision with a group of patients most likely to have a response to a drug, an estimated 40%–50% of all clinical phase drug development fails due to poor efficacy.

Understandably, novel approaches which can help predict the efficacy of treatments during preclinical and clinical studies are an area of great opportunity and interest to pharma researchers and industry players – potentially representing major time, cost and patient benefit.

Situation

Within the lengthy, complex and expensive landscape of drug development, poor efficacy remains the most common cause of late-phase drug development failure.

Failures in efficacy are generally attributed to two major pain points:

• Validation– The validation of a first-in-class drug’s efficacy in both disease-relevant cell models and animal models has traditionally been the ultimate measure of a drug’s success in a clinical setting. However, preclinical experiments using in vitro or isolated systems (cells, tissue preparations, organs) and animal disease models are often unreliable predictors of actual human efficacy. This impedes true validation of the molecular target’s function in earlier stages of drug development.
• Safety – Finding the right balance between safety and efficacy for a candidate drug is one of the ongoing challenges of drug discovery.

The VeriSIM Life advantage

Advancements in biomarker discovery and validation methods, spurred on by innovations in instrumentation and in silico predictive tools, are playing a crucial role in improving efficacy prediction in drug development.

VeriSIM Life (VSL)’s drug decision engine, BIOiSIM®, is a computational platform deploying advanced artificial intelligence and machine learning, a proprietary big data foundation, and state-of-the-art mechanistic models to discover novel therapies from existing molecules.

BIOiSIM® helps not only measure and predict the effects of investigational treatments in humans during clinical trials, but also successfully complete target validation across in vitro, animal, or human disease models. This provides a more comprehensive prediction of a drug’s efficacy and significantly reduces discrepancies between models.

Additionally, VSL’s groundbreaking Translational Index™️ technology helps establish a well-balanced profile between efficacy and safety for candidate drugs.

BIOiSIM further supports efficacy prediction through:

• Simulating and predicting pharmacokinetics (PK) and pharmacodynamics (PD) of compounds known to match the targeted pathway in therapeutics with an established mechanism-of-action
• Leveraging predictive simulation capabilities which guide the development of new drug combinations
• Applying machine learning to predict drug pathways and anticipated disease response in lieu of experimental data
• Predicting drug solubility, stability, and model variations for individuals.

BIOiSIM®, and its groundbreaking Translational Index™️ technology

Combining thousands of validation data sets, multi-compartmental models, and its integrated AI/ML engine, BIOiSIM® achieves superior physiological and biological relevance within three classes of therapeutics: small molecules, large molecules, and re-engineered viruses.

Proof of Value

Predicting Translatable Drug Combinations for ADCs

Challenge

Among blood cancers, acute myeloid leukemia (AML) and diffuse large B-cell lymphoma (DLBCL) are the most rapidly progressing tumors. One of the promising approaches to AML and DLBCL therapies includes the administration of antibody-drug conjugates (ADC) delivering highly toxic antitumor agent payloads linked to highly specific antibodies.

Solution

Unlike conventional systemic distributed chemotherapy, VeriSIM Life (VSL) investigated a novel drug combination strategy which combined ADCs with current therapies in order to enhance their effectiveness and reduce tumor burden.

Methods

• PKPD monotherapy digital models were designed to evaluate the efficacy of a monotherapy at an arbitrary dosing regimen.
• Tumor-specific ML models were then trained to provide concentration-dependent predictions of synergy or inhibition between two therapeutic agents.
• A matrix of dosing regimens was created for each combinatorial therapy to evaluate the predicted combinatorial efficacy at different dosing regimens.

Outcome

VeriSIM Life generated a multidimensional Translational Index score for each combinatorial therapy, producing a ranking of the partner’s combinatorial therapies from the most to least efficacious combination of therapeutic agents.