Using Imaging, Artificial Intelligence and Genetics to Predict Drug Effectiveness in Personalized Medicine for Patients with Alzheimer’s Disease
Doctors are working towards delivering personalized medicine, an approach in which doctors tailor therapy according to each patient's unique physiology. Personalized medicine is working well in many fields of medicine, including oncology cancer care and cardiology. The approach is not as widely used in other fields, such as neurology. Now scientists think they can deliver personalized medicine to treat patients with neurological disease.
About Alzheimer's Disease as a Neurological Condition
About 5.7 million people in the United States have Alzheimer's disease, according to the Alzheimer's Association. The National Institutes on Aging point out that Alzheimer's disease, which is a neurological condition, is the most common cause of dementia. Alzheimer's disease is an irreversible, progressive brain disease. The condition makes changes to the brain that slowly destroys a person's thinking skills and memory. In time, the individual loses the ability to carry out even the most basic tasks of daily living.
Like other diseases, Alzheimer's disease and other neurological conditions often progress and get worse over time. Certain biological factors, such as inflammation, reflect the state of the disease. In addition to inflammation, many neurological diseases involve proteins, known as brain amyloid and tau. Doctors can target these specific biological factors to treat diseases.
A person's genetics can affect how a disease progresses. Doctors refer to this as gene expression, which is the process in which simply having a gene can influence what cells will do. In other words, gene expression can affect how a disease will get worse. People with similar genes can expect similar disease outcomes.
Using Personalized Medicine in Alzheimer's Patients
Many doctors now use personalized medicine to determine the medical treatment for their patients, and to establish the correct dosage of medicine. Finding a treatment for Alzheimer's disease has eluded researchers so far, partly because of the individualized nature of many neurological conditions like Alzheimer's disease.
A team of scientists from Montreal Neurological Institute and Hospital (The Neuro) of McGill University and the Ludmer Centre for Neuroinformatics and Mental Health say they have developed a technique that predicts the effectiveness of targeting those biological factors associated with the progression of a patient's disease. They refer to their technique as personalized Therapeutic Intervention Fingerprint (pTIF).
The researchers used artificial intelligence (AI) and advanced brain modeling to analyze positron emission tomography (PET) and magnetic resonance imaging (MRI) from 331 participants with Alzheimer's disease and people without the disease. They published their results in the medical journal Neuroimage on June 14, 2018.
The lead author of the study, Yasser Iturria-Medina, and colleagues used the information to categorize participants into TIF subtypes according to the treatments that would be more beneficial.
The researchers verified the relevance of the subtypes by comparing their results to the participant's individual genetic profiles. They found that participants in the same pTIF groups all had similar gene expression. This means that all the participants in each pTIF group could expect the same progression of disease.
Drugs to control neurological diseases have to do two jobs: modify gene expression and manage the amyloid, tau and inflammation. Because gene expression and biological factors are highly individual, drugs tailored to pTIF subtypes would be much more effective than those medications designed to treat all patients with a certain disease, such as Alzheimer's disease.
This study is important because it is the first to pinpoint the link between the dynamics of the brain, predicted responses to drug treatment, and the brain changes that occur in patients with neurological diseases. Medical professionals could someday use pTIF subtypes to design drugs to fit each patient's unique gene expression profile and biological factors to provide the best treatment possible for Alzheimer's disease and other neurological problems. When used as a method to select participants, pTIF subtypes could improve the effectiveness of a drug and even reduce the cost of clinical drug trials.
"In keeping with the tenets of personalized medicine, the introduced framework could lead to more effective medical care, decreased undesired secondary effects, and substantial reduction of pharmaceutical/clinical costs associated with clinical trials, thereby accelerating the creation-evaluation cycle of new therapeutic agents," says the lead author in a press release. "Our future work will focus on applying the pTIF to other neurological disorders, extensively validating it, and, importantly, making the resulting analytic tools available to the international community, via open-access platforms."
Imaging techniques, such as PET scans and MRIs, will be an important part of creating new approaches to personalized medicine. Imaging helps researchers categorize patients according to the biological factors, such as the brain changes that often characterize many neurological conditions.
For more information on using imaging and other new techniques to understand conditions like Alzheimer's disease, consult with your doctor.