Predictive value of semi-quantitative MRI-based scoring systems for future knee replacement: data from the osteoarthritis initiative

Nima Hafezi-Nejad, Bashir Zikria, John Eng, John A. Carrino, Shadpour Demehri

Research output: Contribution to journalArticle

Abstract

Purpose: To evaluate, in a confirmatory fashion, whether baseline and change from baseline to 24-month follow-up in cartilage damage, bone marrow lesions and meniscal damage are predictors of knee replacement (KR) in subjects with a high risk of osteoarthritis (OA), independent of the level of physical activity, symptom severity and radiographic abnormalities. Methods: Data from the Osteoarthritis Initiative’s (OAI) baseline and 24-month follow-up knee MRIs of 115 patients (age range: 45–78 years; 48 % female; BMI: 20.9–48.7) were analyzed. Cartilage, bone marrow and menisci were semi-quantitatively scored according to the Whole-Organ Magnetic Resonance Imaging Score (WORMS) and Boston-Leeds Osteoarthritis Knee Score (BLOKS) systems in all compartments. Baseline and 24-month interval changes in structural tissue damage assessed by BLOKS and WORMS were used as predictors of KR independent of clinical and radiographic parameters using Cox hazard analysis. Adjustments were performed for age, gender, BMI and physical activity (Physical Activity Scale for the Elderly: PASE), Western Ontario and McMaster Questionnaire (WOMAC) total score and radiographic Kellgren-Lawrence (KL) score. Results: BLOKS and WORMS baseline cartilage scores were predictors of KR independent of the PASE, WOMAC and KL score. One score increase in the average baseline BLOKS full-thickness cartilage defect score was associated with a [hazard ratio (95 % CI)] 13.55 (3.61–50.89) times greater risk of KR independent of the PASE, WOMAC and KL score. Net reclassification improvements (NRIs) of the additional evaluation of 24-month follow-up MRI scores and assessment of changes were not significant for prediction of KR (NRI range: − 7.23 – 24.8 %). Conclusions: The BLOKS cartilage score for full-thickness cartilage defects had the highest hazard for KR. Follow-up MRI changes in structural tissue damage, detected by BLOKS and WORMS cartilage, bone marrow or meniscus scores (up to 24 months) had no significant predictive value in addition to the baseline MRI.

Original languageEnglish (US)
Pages (from-to)1655-1662
Number of pages8
JournalSkeletal Radiology
Volume44
Issue number11
DOIs
StatePublished - Nov 26 2015

Keywords

  • Knee replacement
  • MRI
  • Net Reclassification improvement (NRI)
  • Osteoarthritis
  • Outcome prediction

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging

Fingerprint Dive into the research topics of 'Predictive value of semi-quantitative MRI-based scoring systems for future knee replacement: data from the osteoarthritis initiative'. Together they form a unique fingerprint.

  • Cite this