Long-term T2 and Qualitative MRI Morphology After First-Generation Knee Autologous Chondrocyte Implantation

Abstract

Background: There are several reports on long-term clinical outcomes after autologous chondrocyte implantation (ACI) for knee cartilage defect treatment. Few published articles have evaluated defect quality using quantitative magnetic resonance (MR) imaging techniques. Purpose: To evaluate clinical outcomes and the quality of repair tissue (RT) after first-generation periosteum-covered ACI (ACI-P) using qualitative MR outcomes and T2-weighted relaxation times. Study Design: Case series; Level of evidence, 4. Methods: All patients (n = 86) who underwent knee joint ACI-P (from 1997 through 2001) with a postoperative follow-up of at least 10 years were invited for clinical and MR evaluation. Clinical outcomes analysis included pre- and postoperative Lysholm and numeric analog scale (NAS) for pain (10 = worst, 0 = best). Radiographic analysis included postoperative T2-weighted mapping of the RT, RT-associated regions, and healthy control cartilage; MOCART (magnetic resonance observation of cartilage repair tissue) score; a modified Knee Osteoarthritis Scoring System (mKOSS; 0 = best, 15 = worst) score; as well as numeric grading for subjective RT and whole knee joint evaluation (1 = best, 6 = worst). Results: A total of 70 patients (45 male, 25 female; mean age, 33.3 ± 10.2 years; 81% follow-up rate) with 82 defects were available for follow-up at an average 10.9 ± 1.1 years postoperatively, with MR analysis for 59 patients with 71 transplant sites (average defect size, 6.5 ± 4.0 cm2). Final Lysholm (71.0 ± 17.4) and NAS (7.2 ± 1.9) scores improved significantly when compared with preoperative scores (Lysholm: 42.0 ± 22.5; NAS: 2.1 ± 2.1; P < .01 for both). Average transplant T2 was 35.2 ± 11.3 ms and thereby significantly lower (P = .005) when compared to the intraknee healthy femur T2 (39.7 ± 6.8 ms). The MOCART was 44.9 ± 23.6 and mKOSS was 4.8 ± 3.2. RT subjective grading was 3.3 ± 1.4, while it was 2.3 ± 0.7 for whole joint evaluation. The RT T2 significantly correlated with postoperative NAS (P = .04; r = −0.28); it also correlated with the healthy femur T2 (P = .004; r = 0.4). The MOCART significantly correlated with the mKOSS (P < .001). Conclusion: The MRI outcome is imperfect in this collective of patients. There is only weak correlation of quantitative imaging data and clinical function. Qualitative imaging data are much better correlated to functional outcomes.