DELAYED GADOLINIUM-ENHACED MAGNETIC RESONANCE IMAGING OF CARTILAGE IS A PREDICTOR OF PERIACETABULAR OSTEOTOMY FAILURE IN PATIENTS 40 YEARS AND OLDER

Pelvic osteotomies for hip dysplasia results can be variable and depend on the amount of preexisting arthritis. Delayed gadolinium-enhanced magnetic resonance imaging of cartilage (dGEMRIC) is a technique designed to measure early arthritis, and could be used to select hips that would benefit from a joint-preserving reconstructive procedure. Our objective was to investigate the role of preoperative dGEMRIC in predicting the success of PAO in patients 40 and above. We hypothesized that patients who failed had lower preoperative dGEMRIC index compared to those who did not. Following IRB approval, patients 40 or older who underwent PAO between 1990–2013 and had preoperative dGEMRIC scan and minimum follow-up of 4 years were identified. Patients with prior hip surgeries or any pathologies were removed leading to a total of 70 patients (Age: 44.2 ± 2.9 years old, BMI: 25.7 ± 4.5 Kg/m 2 ). We only included the first hip undergoing PAO for those with bilateral PAO. Out of 70, 19 had failure defined by the need for total hip replacement or WOMAC pain score of 10 and above within 10 years after index PAO surgery. Articular cartilage was segmented on the 3D pre-operative dGEMRIC scan. The average thickness and dGEMRIC index across the whole articular surface were analyzed. Failed hips had a lower dGEMRIC index by 115 ± 20 ms (P<0.001). All but one failed hips had a dGMERIC index of 400 or less (range: 313 – 479 ms), while all survived hips had a dGMERIC index of greater than 400 (range: 403 – 691 ms). Similar trends were observed when comparing the dGEMRIC index within the 6 subgroups (P<0.01). There were no differences in cartilage thickness (combined femoral head and acetabular cartilage) between the failed and survived hips (p>0.2). Patients with a high dGMERIC index (indicating high GAG content) may have a higher chance of successful outcomes following PAO. Current efforts are underway to develop a multi-modal predictive model to evaluate risk of failure after PAO.