Machine Learning is Improving Treatment for Alkaptonuria

According to a story from mdmag.com, a recent proof-of-concept style study suggests that a certain machine learning tool could be of benefit to patients living with the rare condition alkaptonuria. Machine learning could help provide crucial information for maximizing quality of life in the treatment of these patients. The trial was conducted by the University of Siena’s Department of Information Engineering and Mathematics.

About Alkaptonuria

Alkaptonuria is an inherited genetic disorder in which the body cannot process tyrosine and phenylalanine, which are amino acids. Early in life, affected individuals may not display any symptoms, but their urine may be an unusually dark color, typically brown or black. As the patient ages, they eventually begin to experience pain in their weight bearing joints, particularly along the spine, knees, and hips. Other symptoms include weakening of bones, heart valve problems, hearing loss, and a greater likelihood of kidney stones, gall stones, and similar deposits. Symptoms first appear at age 30, and many people will need joint replacement surgery in their fifth decade. The disease does not appear to affect life expectancy, but it can cause debilitating pain and can drastically affect quality of life. While pain and other symptoms can be treated, there is no treatment that can reverse or halt the process that causes the condition. To learn more about alkaptonuria, click here.

About the Study and Alkaptonuria Research

Researching the disorder can be a challenge because there hasn’t been a standard method developed for assessing the severity of an individual case or its response to treatment. Research has failed to make connections between symptom variations and mutations or explain differences in presentation. The study included data from 129 alkaptonuria patients taken from a database called ApreciseKUre.

Six biomarkers were determined to have direct correlation with patient health. These included body mass index (BMI), serum amyloid A, s-thiolated proteins, chitotriosidase, and advanced oxidation protein products. These biomarkers were most associated with physical symptoms such as osteoarthritis and knee injury. No correlation was found between the biomarkers and patients’ mental health.

These findings serve as a starting point for more in-depth research and suggest that database analysis and machine learning will be critical for more effective evaluation of treatment impacts and ultimately the development of more targeted treatments for alkaptonuria.

Check out the original study in the Orphanet Journal of Rare Diseases here.


Share this post

Share on facebook
Share on google
Share on twitter
Share on linkedin
Share on pinterest
Share on print
Share on email
Close Menu