Background/Purpose: Synovitis is common in knees with OA, and is
associated with structural progression. Intra-articular corticosteroids are
widely used and could reduce knee OA cartilage damage associated with synovitis,
but might also have adverse effects on cartilage health and periarticular bone.
Our objective was to test the potential for disease modification of synovitic
knee OA by triamcinolone hexacetonide (THA) in a clinical trial with
comprehensive measurement of effects on cartilage and subchondral bone using
MRI and DXA.

Methods: This was a 2-year NIH-funded, randomized,
placebo-controlled, double-blind, clinical trial of intra-articular THA (40 mg)
versus saline in people with symptomatic OA (KL grade 2 or 3) with ultrasonic
evidence of synovitis. Randomization was blocked and stratified by gender and KL
grade. We administered the intervention every 12 weeks over 2 years (8 doses). Assessments
included WOMAC pain and function at each visit, annual knee MRI (Philips
Achieva X-Series 3.0 Tesla using cartilage volume and morphology sequences) and
knee/hip DXA scans. We evaluated MRI cartilage damage using a validated
quantitative Cartilage Damage Index (primary structural outcome), area of
denudation, and a semi-quantitative feature score (fissures, delamination,
superficial fibrillation, signal change). We measured bone marrow lesion and
synovial fluid volume quantitatively. Readers for quantitative structural
measurements had good reliability (ICCs > 0.81). We compared change in
outcomes for the groups over the observation period using regression adjusting
for randomization stratification factors and malalignment. Skewed outcomes were
log-transformed. We did additional repeated measures analyses using random
effects mixed models. 

Results: Of the 140 participants, 54% were women, 35% were
nonwhite; mean BMI was 31.2 Kg/m2. Completion rates were high (≥90%) in
both groups. There was no significant difference between groups in mean change
in WOMAC pain (THA -2.2, placebo -2.8; p=0.3) or function (THA -7.1, placebo
-9.2; p=0.4), chair stand (THA -1.1, placebo -1.6; p=0.8), or walk time (THA
-0.5, placebo -0.03; p=0.5). There were also no significant differences in any
of the quantitative or semiquantative structural endpoints, except for CDI (greater
loss in the THA group; Table), and progression of fibrillation (THA 11%,
placebo 24%; p=0.04). Rates of hypertension and hyperglycemia were low (3%
overall for each) and did not differ significantly.

Conclusion: Intra-articular corticosteroids administered every 3
months over 2 years appear relatively safe but do not significantly reduce the
progression of structural damage or patient outcomes over the long term. The
greater rate of loss of cartilage thickness detected by the CDI in the treated
group was small in magnitude and of uncertain clinical significance.