Scientists have discovered two new subtypes multiple sclerosis using artificial intelligence, paving the way to personalized treatment and better outcomes for patients.
The disease affects millions of people worldwide, but treatments are largely chosen based on symptoms and may be ineffective because they do not address the underlying biology of the patient.
Now scientists have discovered two new biological branches of multiple sclerosis using artificial intelligence, a simple blood test and MRI. Experts said the “exciting” breakthrough could revolutionize treatment of the disease worldwide.
In a study of 600 patients conducted by University College London (UCL) and Queen Square Analytics, researchers looked at blood levels of a special protein called serum neurofilament light chain (sNfL). The protein can help determine the level of damage to nerve cells and signal disease activity.
The results of the SNfL and brain scans of the patients were interpreted using the SuStaIn machine learning model. Results, published in the medical journal Brainidentified two different types of MS: early SNFL and late SNFL.
In the first subtype, patients in the early stages of the disease had high levels of sNfL with visible damage in a part of the brain called the corpus callosum. They also rapidly developed brain lesions. According to scientists, this type seems more aggressive and active.
In the second subtype, patients experienced brain shrinkage in areas such as the limbic cortex and deep gray matter before sNfL levels increased. This type appears to be slower, with obvious damage occurring later.
The researchers say the breakthrough will allow doctors to more accurately understand which patients are at higher risk for various complications, paving the way for more personalized treatments.
Lead author of the study, Dr Arman Eshagi from UCL, said: “MS is not one disease and current subtypes cannot describe the underlying tissue changes we need to know to treat it.
“Using an artificial intelligence model combined with a highly available blood marker from MRI, we have been able to demonstrate for the first time two clear biological patterns of multiple sclerosis. This will help clinicians understand where a person is in the disease pathway and who may need closer monitoring or earlier targeted treatment.”
In the future, when an artificial intelligence tool suggests that a patient has early-stage SNFL-MS, they may be eligible for more effective treatment and closer monitoring, Eshagi said.
In contrast, people with advanced SNF may be offered different types of treatment, such as personalized therapies to protect brain cells or neurons. “The innovations will therefore be twofold: transform clinical and neurological examinations, which have remained unchanged for centuries, with the help of artificial intelligence algorithms and provide personalized treatment based on the disease profile.”
Caitlin Astbury, senior research communications manager at the charity MS Society, said: “This is an exciting development in our understanding of multiple sclerosis.
“This study used machine learning to examine MRI data and biomarkers in people with relapsing-remitting and secondary progressive multiple sclerosis. By combining this data, they were able to identify two new biological subtypes of multiple sclerosis.
“In recent years, we have gained a better understanding of the biology of the disease. But currently, definitions are based on the clinical symptoms a person experiences. Multiple sclerosis is a complex disease, and these categories often do not accurately reflect what is happening in the body, which can make effective treatment difficult.”
There are about 20 treatment options for people with relapsing multiple sclerosis, and some are starting to emerge for progressive multiple sclerosis, but for many there are no options, Astbury says. “The more we learn about this disease, the more likely we will be able to find treatments that can stop the progression of the disease.
“This study adds to a growing body of evidence supporting a move away from existing descriptions of multiple sclerosis (such as 'relapsing' and 'progressive') and towards terms that reflect the underlying biology of the disease. This could help identify people at increased risk of progression – and allow people to be offered more personalized treatments.”
