Use of a novel magnetic resonance imaging based modeling technique to investigate differences in tibiofemoral articular cartilage contact area In subjects with moderate knee osteoarthritis
University of Delaware
Knee osteoarthritis (OA) detrimentally impacts the lives of millions of older Americans through pain and decreased functional ability. Unfortunately, the pathomechanics and associated changes that OA patients experience are not well understood. Mechanical stress in the knee joint may play an essential role in OA; however existing literature in this area is limited. Purpose: The purpose of this study was two-fold. First, we wanted to evaluate an existing magnetic resonance imaging (MRI) based modeling method’s estimation of articular cartilage contact area in vivo. Secondly, we wanted to apply this method to a cohort of subjects with moderate knee OA and compare their medial compartment articular cartilage contact area estimates with healthy, age matched controls. Methods: In order to establish confidence in the modeling method’s ability to estimate articular cartilage contact area, imaging data on a single, healthy subject were collected and compared to existing contact area estimates in the literature. Intra-observer reliability and sensitivity studies were also performed in an attempt to further establish confidence in the method. In the second half of this study, MRIs of the knee at 0°, 15°, and 30° flexion were collected during partial-weightbearing in subjects with moderate knee OA (n = 11) and healthy, age matched controls (n = 11). Articular cartilage contact area estimates were normalized to an approximation of the surface area of the tibial plateau in order to account for joint size differences between subjects. The relationship between medial compartment articular cartilage contact area and knee flexion was investigated in each group. Results: The single healthy subject was found to have articular cartilage contact areaestimates similar to those reported in the literature. The method was found to be sensitive to changes in the cartilage tracings on the peripheries of the compartment and demonstrated an intra-observer reliability of 0.95 when assessed using Pearson’s correlation coefficient. In the second half of the study, medial compartment articular cartilage contact area in the healthy controls was found to be significantly correlated with knee flexion angle (p ≤ 0.01), while no such correlation was found in the moderate OA subjects (p = 0.34). Linear regression analysis found that the moderate knee OA subjects had higher articular cartilage contact areas than their healthy, control counterparts across all flexion angles considered. Conclusions: Confidence was established in the MRI based knee modeling method’s ability to estimate articular cartilage contact area through a series of assessments and comparison with existing literature on healthy subjects. Healthy subjects were found to have a significant correlation between medial compartment articular cartilage contact area and knee flexion, which agreed with the literature. The moderate OA subjects did not show the same relationship between medial compartment articular cartilage contact area and knee flexion. Regression analysis found that they had higher medial compartment articular cartilage contact area estimates in all instances when compared with healthy controls. Increased contact area may be a biomechanical adaptation in response to OA in order to decrease the mechanical stress applied across the painful joint.