Paper 54: Long Term Clinical and Radiological Review of Matrix-Induced Autologous Chondrocyte Implantation

Abstract

Objectives: Long-term outcomes following matrix-induced autologous chondrocyte implantation (MACI) are required to better demonstrate later-stage clinical outcomes, patient satisfaction and longevity of repair tissue. This prospective study sought to present the long-term (minimum 10 year) clinical and radiological outcomes of the largest prospectively assessed MACI dataset yet to be reported and compare outcomes in patients undergoing tibiofemoral (TF) versus patellofemoral (PF) MACI. Methods: Between September 2002 and December 2012, 204 patients that underwent MACI were prospectively registered into a research program and assessed pre-surgery and at 2, 5 and 10 years. Of these, 168 patients (182 grafts) were available for clinical review at a minimum 10 years (range 10 to 16 years), of which 151 also underwent MRI. Included in the 182 MACI grafts with minimum 10-year review were: 83 medial femoral condyle (MFC), 32 lateral femoral condyle (LFC), 35 trochlea and 32 patella. Clinically, patients were evaluated with the Knee Injury and Osteoarthritis Outcome Score (KOOS), the 36-item Short Form Health Surveys (SF-36), a visual analogue scale for pain frequency (VAS-F) and severity (VAS-S), satisfaction and peak isokinetic knee extensor and flexor strength. Limb symmetry indices (LSIs) were calculated for strength measures. Grafts were scored on MRI via the magnetic resonance observation of cartilage repair tissue (MOCART) system, with a focus on tissue infill and an overall MRI graft composite score. Results: All PROMs significantly improved (p<0.0001) over the pre- and post-operative period to 2-years post-surgery (Figure 1), with no statistical change (p>0.05) from 2-years to final 10-year follow-up. While the peak isokinetic knee extensor LSI improved significantly over time (p=0.016), at final follow up the mean LSI for peak isokinetic knee flexor and extensor strength was 99.1% and 92.8%, respectively. At final follow-up, 93% of patients were satisfied with MACI to provide knee pain relief, 77% were satisfied with their ability to participate in sports and 89% were satisfied overall. A statistically significant decline was observed from 5 years post-surgery to final follow-up in the MOCART variable of subchondral lamina (p=0.002), while a non-statistical decline was observed in tissue infill (p=0.066) and the overall MRI composite score (p=0.072) (Figure 2), with the MRI-based progression of one patient over the period demonstrated in Figure 3. Of the 151 grafts reviewed via MRI at 10 years or beyond, 14 (9.3%) had failed as defined by graft de-lamination or a graft bed devoid of any repair tissue. Furthermore, of the 36 patients (out of the prospectively recruited 204) that were not available for longer term review, 7 had already proceeded to total knee arthroplasty (TKA) and 1 patient had undergone secondary MACI at the same MFC site due to earlier graft failure. Therefore, a total of 22 patients (10.8%) had essentially failed at or before the final review time. At final follow-up, patients that underwent MACI in the TF (versus PF) joint reported significantly better KOOS sub-scale scores for Pain (p=0.013, TF mean 86.5, PF mean 80.1), Sport (p<0.001, TF mean 71.5, PF mean 60.5) and Quality of Life (p=0.010, TF mean 65.8, PF mean 58.7), as well as a greater knee extensor strength LSI (p=0.002, TF mean 96.0%, PF mean 85.8%). While the TF group demonstrated better 10-year MOCART scores for tissue infill (p=0.027, TF mean 3.2, PF mean 2.9), there were no other differences in MRI-based scores including the overall MRI composite score (p=0.481, TF mean 3.0, PF mean 3.1). Conclusions: This study reports the largest prospective series of patients with long-term review, demonstrating good clinical scores, levels of patient satisfaction and graft survivorship beyond 10 years. Patients undergoing TF (versus PF) MACI reported better long-term clinical outcomes, despite largely similar MRI-based outcomes.