Home Cellular health An artificial intelligence model could predict side effects resulting from new combination therapies

An artificial intelligence model could predict side effects resulting from new combination therapies


Preliminary data from an artificial intelligence model could potentially predict side effects resulting from new combination therapies, according to findings presented at the AACR’s 2022 Annual Meeting, held April 8-13.

Clinicians are challenged by the real world problem that new combination therapies could lead to unpredictable results. Our approach can help us understand the relationship between the effects of different drugs in relation to the disease context.”

Bart Westerman, PhD, lead study author and associate professor at Cancer Center Amsterdam

Many types of cancer are increasingly being treated with combination therapies, through which clinicians attempt to maximize efficacy and minimize the risk of treatment resistance. However, these combination therapies can add multiple drugs at once to a patient’s already complicated medication list. Clinical trials that test new drugs or combinations rarely consider other drugs a patient may be taking outside of the drug regimen being tested.

“Patients seeking treatment typically use four to six drugs a day, making it difficult to decide whether a new combination therapy would risk their health,” Westerman said. “It can be difficult to assess whether the positive effect of a combination therapy will justify its negative side effects for a certain patient.”

Westerman and his colleagues, including graduate student Aslı Küçükosmanoğlu, who presented the study, sought to use machine learning to better predict adverse events resulting from new drug combinations. They collected data from the US Food and Drug Administration’s Adverse Event Reporting System (FAERS), a database containing more than 15 million adverse event records. Using a method called dimensionality reduction, they grouped together events that occur frequently to simplify analysis and strengthen associations between a drug and its side effect profile.

The researchers then fed the data into a convolutional neural network algorithm, a type of machine learning that mimics the way human brains make associations between data. Adverse events from individual therapies were then used to train the algorithm, which identified common patterns between the drugs and their side effects. Recognized patterns have been encoded in a so-called “latent space” which simplifies calculations by representing each adverse event profile as a string of 225 numbers between 0 and 1, which can be decoded back to the profile of origin.

To test their model, the researchers fed their model adverse event profiles of combination therapies, called an “adverse event atlas”, to see if it could recognize these new profiles and decode them correctly using the descriptors of latent space. This showed that the model could recognize these new patterns, demonstrating that the measured combination profiles could be converted back to those of each drug in the combination therapy.

This, Westerman said, demonstrated that the adverse effects of combination therapy could be easily predicted. “We were able to determine the sum of the individual therapeutic effects through a simple algebraic calculation of the latent space descriptors,” he explained. “Because this approach reduces noise in the data because the algorithm is trained to recognize global patterns, it can accurately capture the side effects of combination therapies.”

Westerman and colleagues then validated their model by comparing the predicted adverse event profiles of combination therapies to those observed clinically. Using data from FAERS and the US Clinical Trials Database, the researchers showed that the model could accurately recapitulate adverse event profiles for some commonly used combination therapies.

A complicating factor of combination therapies is the new, potentially unforeseen side effects that can occur when drugs are combined. Using additive patterns as identified by the model, researchers were able to differentiate additive side effects from synergistic side effects of drug combinations. This, Westerman said, can help them better understand what might happen when complex adverse event patterns intertwine.

The researchers are developing a statistical approach to quantify the accuracy of their model. “Since the landscape of drug interactions is very complex and involves many molecular, macromolecular, cellular and organic processes, our approach is unlikely to lead to black-and-white decisions,” Westerman said. “The Adverse Events Atlas is still in the proof-of-concept phase, but the most important discovery is that we were able to get snapshots of the interaction of drugs, diseases and the human body as described by millions of patients.”

Limitations of this study include potential difficulties in comparing these data with more sparse data, as well as the limited application of the model to clinical practice until further validation is provided.


American Association for Cancer Research