Quotient Sciences Uses AI Algorithm to Select Modified-Release Formulations in Clinical Trial

AI Algorithm Reaches Pharmacokinetic Target Within Three Dosing Periods

Quotient Sciences has reported interim results from a clinical study testing whether artificial intelligence can help select modified-release drug formulations during a clinical trial.

The company said a proprietary AI algorithm successfully selected formulation compositions and doses that reached a preset pharmacokinetic target within three dosing periods.

This means the algorithm was able to learn from the clinical data and use that information to choose the next formulation and dose during the study.

The results are part of an effort by Quotient Sciences to make modified-release formulation development faster and reduce the amount of clinical testing that may be needed.

Why Modified-Release Formulation Development Can Take a Long Time

Developing a modified-release tablet is not a simple process.

A formulation needs to release the drug in the body at the right rate and produce the desired drug levels over time.

Traditionally, scientists develop several different formulations in the laboratory and then test selected formulations in people.

If the results do not match the desired pharmacokinetic profile, the formulation may need to be changed and tested again.

This process can require multiple rounds of laboratory work and clinical testing and can take months or even years.

Quotient Sciences is testing whether AI can help reduce some of this trial-and-error work.

How the AI Algorithm Was Used in the Study

The algorithm did not start the clinical study with human clinical data.

Instead, it was initially trained using in vitro drug-release data.

In vitro means testing performed outside the human body, usually in a laboratory.

The algorithm first learned how different tablet compositions affected drug release in laboratory testing.

Once the clinical study began, the algorithm was retrained after each dosing period.

It received new information from tablet dissolution results and pharmacokinetic data collected from healthy participants.

It then used that information to select the composition and dose for the next dosing period.

In simple terms, the algorithm was allowed to learn as the study progressed.

Algorithm Reached the Target in Three Dosing Periods

The main interim finding was that the algorithm reached the study's preset pharmacokinetic target within three dosing periods.

Pharmacokinetics, or PK, describes what happens to a drug in the body. It looks at factors such as how quickly a drug enters the bloodstream, how much of it is present and how long it remains in the body.

For a modified-release formulation, getting this profile right is important because the tablet is designed to release the drug gradually.

According to Quotient Sciences, the interim results met the program's preset objectives.

The clinical study is still ongoing, so the company plans to provide the complete results after dosing is finished.

Earlier Laboratory Work Helped Lead to the Clinical Study

The clinical study was based on earlier laboratory testing using the same AI algorithm.

During that work, the algorithm was tested to see whether it could learn the relationship between tablet composition and drug release.

Quotient Sciences reported that the algorithm mapped the formulation design space after screening one-third fewer formulations than conventional methods.

This finding led to the next question.

The company wanted to know whether an AI model that could learn the relationship between formulation composition and drug release in the laboratory could also learn how formulation composition affects pharmacokinetics in humans.

The current clinical study is designed to test that idea.

Scientists Kept Human Oversight in the Trial

Although AI was used to select formulations and doses, the algorithm did not operate without human control.

Quotient Sciences established limits for what the algorithm could select.

These limits included a dose cap for the first prototype.

A safety committee also reviewed and approved every formulation before it was manufactured and given to participants.

This created a human-in-the-loop system.

In simple terms, the AI could make recommendations based on the available data, but people remained responsible for safety decisions before a formulation reached the clinical trial.

Study Uses an Existing Generic Drug

The study uses a generic drug that already has an established safety record and a large amount of published information.

The drug was selected as a test case for the AI system rather than as a potential new commercial product.

Quotient Sciences said it does not plan to develop the drug as a product.

This allowed the company to focus on testing whether the AI approach could learn the relationship between formulation composition and human pharmacokinetics.

What Quotient Sciences Wanted to Find Out

The company said the study was designed around three main questions.

The first was whether the AI model could learn the relationship between formulation composition and performance in humans.

The second was how quickly the model could learn that relationship.

The third was how accurately it could use the information to select formulations and doses.

Andrew Lewis, Ph.D., chief scientific officer at Quotient Sciences, said the interim results indicate that the model was able to learn the relationship and do so quickly.

He said the findings could mean that less clinical testing may be needed when developing modified-release formulations for other drug molecules.

AI Could Support Faster Formulation Development

The potential benefit of this approach is reducing the number of formulation and clinical testing cycles.

Instead of testing many formulations one after another, an AI model could potentially learn from each experiment and use the new information to select a more informed formulation for the next stage.

This could make early drug development more efficient.

However, the current results are interim findings from one clinical study.

The full study results will provide more information about how accurately and consistently the approach works.

Quotient Sciences Is Adding AI to Translational Pharmaceutics

The work builds on Quotient Sciences' existing Translational Pharmaceutics platform.

The platform combines drug product development, manufacturing and clinical testing within an integrated development process.

The company is now adding AI-based formulation insights to this approach.

The AI-enhanced solution is designed to support model-informed drug development.

One part of the approach involves creating a digital twin that connects formulation composition with laboratory drug-release performance and human pharmacokinetics.

A digital twin in this setting is a computer-based model that represents how a formulation is expected to behave based on available data.

As more data become available, the model can potentially be updated to improve its understanding of the relationship between formulation design and drug performance.

What Happens Next?

Dosing in the current clinical study is continuing.

Quotient Sciences expects to report the full study results once the trial is completed toward the end of 2026.

The final data will provide a more complete view of how well the AI algorithm performed throughout the study.

The company also plans to use the work to further develop its AI-enhanced formulation development solution for early-stage drug development programs.

About Quotient Sciences

Quotient Sciences is a global integrated contract research, development and manufacturing organization, commonly known as a CRDMO.

The company provides services across different stages of drug development and clinical research.

Its Translational Pharmaceutics platform brings together formulation development, drug product manufacturing and clinical testing.

The platform is used by companies ranging from emerging biotechnology businesses to large pharmaceutical organizations.

Quotient Sciences has more than 20 years of experience in drug development and is now adding AI-driven formulation insights to its existing services.

The company says the goal is to help drug developers make earlier development decisions and move toward proof-of-concept studies while reducing development time, cost and risk.

Optimize Your trial insights with Clival Database.

Are you exhausted from the uncertainty of trial insights pricing? Clival Database ensures the clarity in the midst of the global scenario for clinical trials to you.

Clival Database is one of the best databases that offers an outstanding number of clinical trial data in terms of 50,000+ molecules and from primary regulatory markets as well as new entrants like Indian and Chinese markets.

With Clival, you get accurate positioning of historical sales data, patent database, company profiling, safety & efficacy, and prediction of launch of new innovative molecules helping you to align your research and driving down the cost.

To add value, we further break down our analytics for you so that improving your operational effectiveness; optimizing your clinical trials; and offering you accurate and high-quality data at lowest possible prices becomes possible.

Elevate your trial success rate with the cutting-edge insights from Clival database.

Check it out today and make more informed sourcing decisions! Learn More!