Abstract
Objectives:
Anterior cruciate ligament reconstruction (ACLR) restores passive knee stability but does not reduce the rate of osteoarthritis (OA). Meniscal status, requiring repair or meniscectomy, is a known component of degenerative progression. Regarding the patellofemoral joint (PFJ) specifically, traditional imaging modalities have shown some evidence that patellar tendon autograft (BPTB) is associated with an increased risk of OA. T2 relaxation time using quantitative MRI (qMRI) is an established method to identify early cartilage changes. However, T2 relaxation in the PFJ as a function of graft type and meniscal status has not been previously evaluated. We aimed to evaluate the effect of autograft type (BPTB, hamstring [HT] and quadriceps tendon [QT]) as well as concomitant meniscal surgery (repair or meniscectomy) at the time of ACLR on T2 relaxation time in PFJ cartilage. We hypothesized that BPTB grafts, concomitant meniscal repair, and meniscectomy would result in higher (worse) T2 relaxation times at 6 months after ACLR. We also evaluated the effect of abnormal joint loading during gait on the PFJ cartilage. We hypothesized that T2 relaxation times would not differ in any group after controlling for knee joint loading during walking.
Methods:
Thirty-four participants aged 15-35 years were enrolled prospectively within 1 month of ACL injury. Exclusion criteria included prior knee injury or surgery, concomitant grade III tear to other knee ligaments that was symptomatic or required surgical intervention and presence of chondral damage. Sagittal qMRIs were obtained within 1 month of injury (but prior to ACLR) and 6 months after ACLR. Articular cartilage of the PFJ was manually segmented using ITK-SNAP software (Penn Image Computing and Science Laboratory, University of Pennsylvania) and was confirmed for accuracy by a board-certified, fellowship-trained musculoskeletal radiologist (Figure 1). Mean T2 relaxation times were extracted from the trochlear and patellar cartilage. Non-physiological T2 relaxation times less than 10 ms and greater than 90 ms were excluded. Gait biomechanics were captured using two in-ground force plates (Bertec Corporation, Columbus, OH) sampled at 1,080 Hz and retroreflective markers with an 8-camera motion capture system (Qualisys, Göteborg, Sweden) sampled at 120 Hz. Data was processed using Visual 3D software (C-motion, Bethesda, MD). Using an inverse dynamic approach, joint angle and force plate data were used to calculate the knee flexion moment (KFM) impulse normalized to mass and height during the stance phase. Separate one-way ANOVAs were used to evaluate the relationship between graft type and concomitant meniscus surgery with the percent change in T2 relaxation time in PFJ cartilage from baseline to 6 months after ACLR. To determine the influence of gait biomechanics on T2 relaxation time across groups, general linear models were used with the interlimb ratio (involved/uninvolved) of KFM impulse as a covariate.
Results:
The 34 participants had a mean age of 19.0±4.7 years, and 21 (61.8%) were female. Grafts used during ACLR included 22 BPTB, 4 HT and 8 QT. Seventeen participants (50.0%) had a concomitant meniscus repair and 3 (8.8%) had a meniscectomy. Overall, T2 relaxation time increased 8.7±7.3% in the patellar cartilage (pre-ACLR: 39.1±2.6 ms, 6 months: 42.4±3.6 ms, p<0.001) and 7.1±7.3% in the trochlear cartilage (pre-ACLR: 46.3±2.6 ms, 6 months: 49.5±2.7 ms, p<0.001). The change in T2 relaxation time 6 months after ACLR did not differ based on graft type in either the patellar (p=0.729) or trochlear cartilage (p=0.525) (Figure 2). Group differences remained insignificant after controlling for knee loading (KFM impulse) during gait (patellar: p=0.594; trochlear: p=0.222). Meniscal status was also not associated with a change in T2 relaxation time in either the patellar (p=0.392) or trochlear cartilage (p=0.190) (Figure 3) although there was a trend toward higher T2 relaxation times in those with a concomitant meniscus repair or meniscectomy. Group differences remained statistically insignificant after controlling for the KFM impulse during gait (patellar: p=0.400; trochlear: p=0.197).
Conclusions:
Our findings provide insufficient evidence that graft type is associated with worse quantitative chondral changes within the PFJ at 6 months following ACLR. While meniscal status is a known risk factor for increased contact pressure in the involved compartment and subsequent degenerative progression, our data does not demonstrate statistically significant changes to the PFJ cartilage when meniscal tears were present, likely due to the limited sample size and large variation within groups. Future work with larger cohorts is needed to confirm our findings and further investigate the role that graft type and meniscal status may play within the PFJ after ACLR.
