Session Information
Session Type: Abstract Submissions (ACR)
Background/Purpose:
The cartilage damage index (CDI) is a parsimonious method of measuring articular cartilage approximating regional cartilage volume based on 18 locations within the knee, 9 in the medial tibia and femur (figure). Understanding locations that are simultaneously pathologic may better clarify the pathophysiology of knee osteoarthritis (OA). Therefore, we conducted an exploratory factor analysis of the CDI.
Figure. Spatial diagram of 9 of the femoral CDI points. Each red dot in the grid represents a separate location where CDI of articular cartilage is measured.
Methods:
We included a convenience sample of 200 participants of the progression cohort of the Osteoarthritis Initiative (OAI) with 3T knee MRIs, PA radiographs, and WOMAC pain scores (0 – 20) at the OAI 24- month visit. Long limb films obtained at the OAI 12-, 24-, or 36- month visits were used to measure Hip-Knee-Ankle (HKA) angle using a semi-automated program (negative = varus). One assessor performed CDI measurements on MRI sagittal DESS sequences. Medial joint space narrowing (JSN) (0-3) and Kellgren-Lawrence (KL) score (0-4) were centrally assessed. An exploratory factor analysis was conducted on the 18 CDI points. To assess the characteristics of factors resulting from that analysis, we performed Spearman’s correlations with the full CDI measure, the separate factors, medial JSN (construct we are trying to capture), KL score (OA severity), HKA (risk factor for medial OA), and WOMAC score (OA symptoms).
Results:
Mean age was 62.5 (9.5) years, mean BMI was 30.0 (4.5) kg/m2, 47% were female. 3 factors resulting from the factor analysis were: (1) anterior and weight bearing portions of the femur and tibia (AntWB) (2) posterior tibia (PostTib) and (3) posterior femur (PostFem), explaining 75%, 11%, and 9% of the data variation.
|
CDI |
AntWB CDI |
PostTib CDI |
PostFem CDI |
Medial JSN |
KL Score |
HKA |
WOMAC |
CDI |
1 |
0.97832 |
0.67338 |
0.68628 |
-0.50802 |
-0.28166 |
0.28489 |
-0.14135 |
|
<.0001 |
<.0001 |
<.0001 |
<.0001 |
<.0001 |
<.0001 |
0.0459 |
|
AntWB CDI |
|
1 |
0.61494 |
0.56546 |
-0.57167 |
-0.35091 |
0.32417 |
-0.1724 |
|
|
<.0001 |
<.0001 |
<.0001 |
<.0001 |
<.0001 |
0.0146 |
|
PostTib CDI |
|
|
1 |
0.45342 |
-0.1569 |
-0.10798 |
-0.04208 |
-0.04712 |
|
|
|
<.0001 |
0.0269 |
0.129 |
0.5581 |
0.5076 |
|
PostFem CDI |
|
|
|
1 |
-0.1293 |
0.12819 |
0.11757 |
0.02053 |
|
|
|
|
0.0687 |
0.0712 |
0.1008 |
0.7729 |
|
Medial JSN |
|
|
|
|
1 |
0.56874 |
-0.52189 |
0.09234 |
|
|
|
|
|
<.0001 |
<.0001 |
0.1946 |
|
KL Score |
|
|
|
|
|
1 |
-0.19741 |
0.29629 |
|
|
|
|
|
|
0.0057 |
<.0001 |
|
HKA |
|
|
|
|
|
|
1 |
0.02272 |
|
|
|
|
|
|
|
0.7519 |
|
WOMAC |
|
|
|
|
|
|
|
1 |
|
|
|
|
|
|
|
|
Table. Spearman’s correlations of CDI, anterior weight-bearing (Ant WB) CDI, posterior tibia (PostTib) CDI, posterior femur (PostFem) CDI, Medial JSN, KL Score, HKA, and WOMAC scores.
Conclusion:
These findings suggest 3 patterns of cartilage damage occur in the medial tibia and femur. While PostTib and PostFem CDI are correlated with the full CDI, they are at best weakly correlated with medial JSN and not correlated with KL score, static alignment or pain. The AntWB CDI pattern contributes most of the variation in the 18 points and shows construct validity, correlating with medial JSN, KL score, static alignment, and pain, better than the full CDI measurement. Future studies confirming these findings are warranted.
Disclosure:
G. H. Lo,
None;
L. L. Price,
None;
M. Zhang,
None;
J. B. Driban,
None;
D. Harper,
None;
T. E. McAlindon,
None.
« Back to 2013 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/exploratory-factor-analysis-of-a-parsimonious-medial-cartilage-damage-index-reveals-one-factor-associated-with-radiographic-severity-and-symptoms-data-from-the-osteoarthritis-initiative/