Baseline MRI findings as predictors for progression of radiographic JSN in the same joint
| Presence of MRI features | OR (95% CI) of radiographic progression (JSN) (in joints with baseline JSN=0–2) | ||
|---|---|---|---|
| MRI features: | n (%) | Crude associations | Adjusted associations* |
| Synovitis | |||
| Grade 1 | 201/444 (45.3) | 1.52 (0.74 to 3.15); p=0.26 | 1.32 (0.54 to 3.22); p=0.55 |
| Grades 2 and 3† | 99/444 (22.3) | 3.43 (1.59 to 7.43); p=0.002 | 3.52 (1.29 to 9.59); p=0.01 |
| Flexor tenosynovitis | 79/430 (18.4) | 1.03 (0.58 to 1.86); p=0.91 | 1.02 (0.53 to 2.00); p=0.94 |
| Focal erosions | 217/498 (43.6) | 1.14 (0.70 to 1.84); p=0.60 | 0.85 (0.49 to 1.46); p=0.55 |
| Subchondral bone attrition | 45/499 (9.0) | 1.19 (0.53 to 2.68); p=0.67 | 1.24 (0.50 to 3.08); p=0.65 |
| Cysts | 20/498 (4.0) | 2.53 (0.93 to 6.91); p=0.07 | 2.98 (1.04 to 8.53); p=0.04 |
| Osteophytes (grades 2 and 3)‡ | 106/501 (21.2) | 1.20 (0.61 to 2.38); p=0.60 | 1.27 (0.61 to 2.62); p=0.52 |
| JSN | |||
| Grade 1 | 182/501 (36.3) | 2.68 (1.32 to 5.41); p=0.006 | 2.70 (1.22 to 5.98); p=0.01 |
| Grade 2 | 122/501 (24.4) | 4.90 (2.35 to 10.19); p<0.001 | 4.60 (2.02 to 10.49); p<0.001 |
| Grade 3 | 20/501 (4.0) | 13.22 (4.04 to 43.20); p<0.001 | 11.05 (3.22 to 37.90); p<0.001 |
| Bone marrow lesions | 52/506 (10.3) | 2.32 (1.19 to 4.54); p=0.01 | 2.73 (1.29 to 5.78); p=0.009 |
| Malalignment | 26/502 (5.2) | 0.91 (0.37 to 2.24); p=0.83 | 1.09 (0.46 to 2.60); p=0.84 |
| Collateral ligament discontinuity | 252/501 (50.3) | 1.01 (0.65 to 1.55); p=0.98 | 1.04 (0.65 to 1.68); p=0.87 |
Generalised Estimating Equations with separate models for each MRI feature. Statistically significant associations are shown in italics.
*Adjusted for age, sex, BMI and follow-up time.
†Few joints with synovitis grade 3 (n=8).
‡Osteophyte grades 2 and 3 versus grade 0 (n=74) and grade 1 (n=321) as reference.
BMI, Body Mass Index; JSN, joint space narrowing.