Self-reported stiffness comparable to MRI in predicting radiographic knee OA
TORONTO — Self-reported data on swelling and morning stiffness showed comparable efficacy to MRI in predicting incident radiographic knee osteoarthritis, according to findings presented here.
“There is relevant inflammation on MRI in the OA process,” Jos Runhaar, PhD, of the department of general practice at Erasmus University Medical Center in Rotterdam, Netherlands, said. “We see that inflammation is related to the incidence of radiographic OA, and progression of both radiographic and symptomatic knee arthritis.”
However, Runhaar noted that MRI remains an expensive measure. “Also, if we want to use MRI for prediction, we are not quite sure how well it predicts,” he said. “We are not quite sure if it is better than some patient-reported measures that you can use for inflammation.”
To evaluate the predictive value of MRI-assessed vs. patient-reported knee OA in a cohort of 407 high-risk women, Runhaar and colleagues assessed inflammation observed on MRI using the MOAKS scoring system. The other measures of inflammation included effusion synovitis of grade 1 or higher, or HOFFA synovitis of grade 1 or higher.
For the survey component of the study, participants answered questions from the KOOS questionnaire about swelling, inflammation and morning stiffness. Eligible participants were free of knee complaints or radiographic evidence of knee OA at baseline.
The initial follow-up period was 2.5 years, and then the researchers followed patients for another 4 years after that. The final cohort included women aged 55 years with a BMI of 31. Baseline data showed that inflammation was reported in 15%, while swelling was reported in 12% and morning stiffness in 14%.
According to study results, at 2.5 years, clinical OA occurred in 7% of the cohort, while radiographic OA — defined as a KL score of 2 or higher — occurred in 5% of patients. “We saw that inflammation on MRI was associated to the morning stiffness,” Runhaar said. He added that patient-reported swelling also carried an association with inflammation on MRI.
By 6.5 years, the rates of inflammation were 12% for clinical OA and 16% for radiographic OA. Runhaar noted that the association between inflammation and morning stiffness persisted through this time point.
Pre-study probability estimates showed that one out of 14 knees were likely to develop incident clinical OA after 2.5 years, according to Runhaar.
At 2.5 years, results showed that clinical inflammation occurred in one out of 12 knees, while swelling occurred in one out of nine knees and morning stiffness occurred in one in six.
“If we look at incident radiographic OA, we see that one out of 21 knees in the entire cohort developed incident radiographic OA,” Runhaar said. He noted that among those, inflammation on MRI occurred in one out of seven knees, swelling occurred in one out of 10, and morning stiffness occurred in one out of 22 knees.
Looking at 6.5-year data, the trends persisted, with one out of nine knees in the entire cohort developing incident radiographic OA.
Runhaar stressed that inflammation on MRI was the best predictor of incident radiographic knee OA.
“However, the same was true for self-reported swelling,” he said. “If you had to compare them based on the predictive abilities for number needed to screen, you would say that MRI wins over self-reported swelling. But when you look at morning stiffness, clearly the numbers show that the questionnaire performed better than MRI.” —by Rob Volansky
Landsmeer ML, et al. Abstract #44. Presented at: OARSI 2019 World Congress on Osteoarthritis; May 2-5; Toronto, Canada.
Disclosure: Runhaar reports no relevant financial disclosures.