Drug Formulation

How AI addresses common pain points experienced in the formulation stage of the drug development process.

%20(4).png)

Introduction

Drug formulation is one of the most critical parts of pharmaceutical development, helping to determine the very best way to deliver an active ingredient or molecule to patients. The drug formulation process seeks to determine the right combination of inactive substances and active pharmaceutical ingredients (API), evaluate a drug’s scalability for manufacturing, establish the most effective protocols for treatment, and identify the right form for the drug itself (e.g. tablet, capsule, oral suspension, injection, etc). Only then will a patient-ready end product be achieved.

Situation

Finding the ideal drug formulation is typically a complex, slow, laborious and expensive part of early phase R&D. A variety of pain points combine to create considerable challenges for pharma researchers, who have traditionally lacked an integrated approach for de-risking their early R&D decisions in bringing a new drug to market.

Common pain points in the drug formulation process include:

• Solubility – In order to achieve the desired concentration of a drug, its solubility in water must be determined. Low solubility is often a problem in formulation development, limiting bioavailability, and requiring a variety of techniques to enhance solubility including physical and chemical modifications.
• Stability – New pharmaceuticals must be shelf stable within predetermined limits, meaning they retain their chemical, physical, microbiological and biopharmaceutical properties over a given period, while also maintaining their safety and efficacy. Because any changes or degradations that occur over time can impact efficacy and/or lead to adverse effects, a drug’s stability must be carefully studied.
• Dose Range– Researchers need to identify the ideal dose range and dose formulation for a new drug to achieve its desired therapeutic effect, defining both the no-effect dose and the mean effective and maximal effective doses. To do so, researchers typically engage in numerous dose-finding studies, including: dose escalation, single ascending, and multiple dosing studies.
• API Availability– In order to determine the right formulation for a drug, a significant amount of API is necessary for testing. However, in the early stages of drug development, a very limited supply of API for formulation studies is typically available due to either complexity or the cost of manufacturing.

While not all drugs face every one of these challenges during their development, most will face multiple open questions around a drug’s stability, absorption, solubility, bioavailability, and/or pharmacokinetics. And, typically, difficulties in being able to make early predictions around the right composition or combination of APIs and the materials necessary for formulation take a heavy toll on performance-related parameters. Moreover, dealing with a variety of unknowns around composition can be challenging for researchers to deal with simultaneously, especially when using traditional methods which tend to be both highly linear and iterative.

The VeriSIM Life advantage

Today, there’s good news for pharma researchers embarking on the drug formulation process. Innovations in AI-driven technologies have allowed for the streamlining and de-risking of many aspects of early-stage drug development, including the drug formulation process.

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, and VSL’s associated services, can be used to problem-solve some of the most significant challenges facing early-stage drug R&D – meeting development timelines faster, at lower costs, and with better results than traditional approaches.

The BIOiSIM AI-driven technology framework supports the drug formulation process by addressing key pain points in the journey:

• Predicts drug solubility and model variations for individuals, including interactions between API and carrier, solubility parameters, and by simulating amorphous solid dispersions (ASD) formation and dissolution mechanisms.
• Predicts drug stability faster, and at a significantly lower cost, than human and animal studies, using molecular dynamics simulations which model and predict ASD physical stability.
• Uses advanced modeling software as a starting point for pinpointing the most appropriate dose for human consumption, helping elucidate the relationship between compound exposure and therapeutic effect, including immune response, therapeutic window, and safety and efficacy for compounds in Phase-1 trials.
• Reduces the need for costly, resource-intensive formulation studies on humans and animals, predicting computationally how the API interacts with virtual versions of animal and human subjects.
• Uses state-of-the-art quantum and molecular mechanics simulations to predict changes in the oral bioavailability of API due to formulations and co-administered food containing metal ions.
• Utilizes robust mechanistic models of intestinal transit to predict bioavailability and drug exposure for formulations designed to be absorbed in different parts of the gastrointestinal tract.
• Machine learning platform capacities can be implemented as for oral drug intake as for inhalation, transdermal, and topical routes of administration.

BIOiSIM, and its groundbreaking Translational Index™️ technology

which advances only the most promising drug candidates through R&D to investigational new drug (IND) application, offers actionable insights of unprecedented value to the drug development industry.

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/gene therapy.

AtlasGEN™️ Novel Drug Designer benefits

The BIOiSIM® platform features a robust data lake foundation, integrating:

Vast chemical search space: The ability to generate target engaging compound hits depends largely on the size of the molecular search library. The AtlasGEN search space is supported by more than 1012 compounds.

Super fast discovery: AtlasGEN accurately predicts protein-ligand binding affinity by combining geometric conformer analysis with machine learning based on experimental data X% more efficiently than molecular docking-based approaches.

Pre-validated hits: Iterative ranking of target engaging hits by their Translational Index values reduces the time and cost associated with other computational hit discovery approaches, while compressing the hit-to-lead refinement and optimization process by pre-validating hits for translatability. This unique integration dramatically reduces the number of compounds to evaluate in subsequent experimental research.

Proof of Value

Using Molecular Simulation and Statistical Learning Methods in Low-Solubility Drug Formulation Design

Challenge

Using Molecular Simulation and Statistical Learning Methods in Low-Solubility Drug Formulation Design

Predicting the API solubility with various carriers in the API–carrier mixture and the principal API–carrier non-bonding interactions are critical factors for rational drug development and formulation decisions. However, experimental determination of these interactions, including solubility and dissolution mechanisms, are time-consuming, costly and reliant on trial and error.

Solution

To streamline the formulation design process, molecular modeling has been applied to simulate amorphous solid dispersions (ASD) properties and mechanisms in order to predict the API solubility of various carriers.

Methods

• Quantum mechanical methods elucidate the strength of API–carrier non-bonding interactions
• Molecular dynamics simulations model and predict ASD physical stability, solubility, and dissolution mechanisms
• Statistical learning models predict a variety of drug formulation properties to help predict ASD solubility
• Other computational applications help accelerate lead compound development before clinical trial

Outcome

In silico research has demonstrated the viability of rational formulation design of low-solubility drugs. Pertinent theoretical groundwork, including modeling applications and limitations, have shown the prospective clinical benefit of accelerated ASD formulation.

ML methods have the potential to provide the next transformative leap forward toward rapid polymer screening and formulation design.