MRI definitions insufficient to identify risk for radiographic knee osteoarthritis
Click Here to Manage Email Alerts
Key takeaways:
- More than 50% of people with knee OA based on either of two MRI definitions did not show disease on X-ray up to 11 years later.
- Future efforts may focus on imaging that includes MRI features, symptoms and exams.
Current MRI-based definitions of knee osteoarthritis fail to sufficiently pinpoint knees that will later develop radiographic disease visible on X-ray, according to data published in Arthritis & Rheumatology.
“Early-stage knee OA is a critical focus because it presents a window of opportunity for implementing lifestyle changes and early treatment before substantial structural changes occur or chronic pain sets in,” Alison H. Chang, PT, DPT, MS, of the department of physical therapy and human movement sciences at the Northwestern University Feinberg School of Medicine, told Healio. “Unlike many other chronic diseases, there are no clear biomarkers to signal the onset of knee OA, making early detection challenging.
“Although radiography is commonly used to diagnose knee OA, its sensitivity in detecting early structural changes is limited,” she added. “MRI offers a more detailed view of joint structures. Existing MRI-based definitions of knee OA have not been examined for their utility in identifying individuals at risk for developing clinically significant OA over the long term.”
To assess the use of MRI-based OA definitions in determining which knees will develop radiographic disease within 11 years, Chang and colleagues analyzed data from the Osteoarthritis Initiative, a cohort of adults with increased risk for radiographic disease. Using two previously proposed MRI definitions of OA, the researchers examined the prevalence of MRI-defined OA among the cohort, as well as whether the presence of OA, according to either definition, could be linked to the subsequent development of radiographic and symptomatic disease.
The first MRI-based definition of OA involved assessment of cartilage damage, osteophytes, subchondral bone marrow lesions, meniscal damage or extrusion and bone attrition, according to the researchers. The second definition was simpler and only examined cartilage damage and osteophytes.
The analysis included 1,621 adults (mean age, 58.8 years) without radiographic knee OA at baseline. Using adjusted Cox proportional hazards regression models, links between MRI results and radiograph-based Kellgren-Lawrence grade — with knee OA defined as a score of two or greater — were analyzed across a follow-up period of up to 11 years.
At baseline, 17% of participants demonstrated knee OA according to the first MRI definition, and 24% had knee OA according to the second. According to the researchers, having MRI-defined knee OA at baseline was associated with later having a Kellgren-Lawrence grade of two or higher under both the first (OR = 2.94; 95% CI, 2.34-3.68) and second (OR = 2.44; 95% CI, 1.97-3.03) definitions.
However, among participants with MRI-defined knee OA at baseline, 59% of those in the first definition, and 64% of those in the second definition, never developed a Kellgren-Lawrence grade of two or more.
“This raises questions about whether the current MRI definitions of knee OA truly represent a precursor state for the disease or if additional factors are necessary for the transition to clinically significant OA,” Chang said. “This result underscores the complexity of predicting OA development and suggests that MRI alone may not capture all the necessary information to forecast disease progression.”
Identifying factors that increase the likelihood of clinically significant disease “remains a key priority,” Chang added.
“Future studies may consider developing and validating more comprehensive algorithms that integrate MRI data with clinical symptoms and physical examination findings to improve the prediction of clinically significant OA,” she said. “Additionally, machine learning and artificial intelligence methods could help enhance the precision and predictive power of these MRI-based criteria.”
Collapse
Read more about
Click Here to Manage Email Alerts