The objective of articular cartilage repair is to provide long-lasting improvement of the patient’s quality of life. It is also important that the technique be readily available and as harmless as possible. Cultured chondrocyte implantation provides good repair capacity, but has disadvantages such as high cost, limited availability, and the need for two procedures. Salzmann et al.  demonstrated a procedure using a membrane seeded with autologous cartilage cells, which can solve the problems of cost and availability and be performed in a single stage.
Among the 64 repairs carried out in our sample, only once did the membrane actually come loose (1.5%); the other four cases of failure were likely due to progression of the underlying degenerative process, requiring conversion to arthroplasty (6.25%). Gomoll et al. , in a study of 101 patients with up to 1 year of follow-up, used fibrin glue and sutures along the periphery only when they felt the need for greater fixation, achieving a 5% reoperation rate due to hypertrophy or delamination.
In our study, VAS scores for pain decreased from 5.92 to 2.37 in the postoperative period (P < 0.001), demonstrating clinical improvement in this aspect at 12 months of follow-up. Siclari et al. , using the same microfracture scaffold combined with platelet-rich plasma and Smart Nails® (ConMed Linvatec Italy, Milan, Italy) and fibrin glue for fixation, also reported improvement on the KOOS pain scale (from 54.1 to 93.5). Verdonk et al.  used the MaioRegen® three-layer scaffold (Fin-Ceramica Faenza SpA, Faenza, Italy) and likewise reported a reduction in VAS pain scores, from 63.1 to 22.7. Although pain may not be related to the articular cartilage lesion per se, it is interesting to note that it improves markedly after cartilage repair procedures are performed. This effect can be credited to restoration of the load-bearing property, to reduction of subchondral pressure by the microfractures, or to neovascularization of the lesion site, reducing inflammatory cytokine levels and thus relieving symptoms [9, 10].
We also observed improvements in IKDC score at 12 months postoperatively (from 33.44 to 56.33, P < 0.001). Berruto et al. , using a three-layer matrix implanted through the “press-fit” technique in 49 patients, reported an increase in IKDC score from 45.55 to 70.86 at 12 months. Delcogliano et al.  used the same technique and matrix in 19 patients (21 lesions) and demonstrated an approximate gain of 30 score points (35.7 to 67.7) in 12 months. Theoretically, a three-layer matrix mimics hyaline cartilage more closely. Even so, the clinical outcomes achieved are very similar, and may be attributable to the use of bone perforations or autogenous grafts. In the present study, an approximate 23-point improvement in the mean IKDC score was observed, noting that patients had lower baseline (pre-operative) scores than in the case series of Berruto et al.  (45 points) and similar to those reported by Delcogliano et al.  (35.7 points). This may be attributed to the more advanced average age of our patients compared to the other series (44 years, versus 37 and 33 years respectively) [11, 12].
In our cohort, the total WOMAC score improved from 53.26 to 75.93 (P < 0.001) at 12 months post-operatively. Dhollander et al.  used allogeneic chondrocyte cell cultures from fresh tissues protected with periosteum. This procedure was performed on 21 patients aged 12 to 47 years, who obtained an average 42% improvement in WOMAC. For comparison, in our study, we observed a 42.6% increase in total WOMAC scores. Although part of the cases in the present study were of focal, trauma-related lesions, a higher percentage of patients had progressive pain, probably indicative of osteoarthritis rather than traumatic aetiology; the WOMAC score could be a better tool to assess these patients, since it is a scale designed for degenerative processes.
Few studies have demonstrated improvement in quality of life after cartilage repair. The SF-36 showed an improvement in its physical domain, from 30.49 at baseline to 40.23 post-operatively (P < 0.001). Gelber et al.  reported an improvement in SF-36 scores from 53.9 to 86.6. In our series, improvement was not as marked, which again may be explained by the higher average age of our patients (44 versus 36 years). Older patients tend to have a lower SF-36 because they experience greater difficulty climbing stairs, kneeling, and in other activities of daily living. Cole et al.  reported an improvement in scores from 35.4 to 45.5 in the treatment of osteochondritis-related lesions. This more modest improvement, closer to ours, may be due to their patients having osteochondritis dissecans, which carries a less favourable prognosis. The normal range of the SF-36 physical domain has been described as 49.7 ± 9.4  in the Canadian population. Thus, it is clear that, despite improvement in patient quality of life, they are unlikely to reach the same level of quality of life experienced by the general population.
Knees with a cartilage lesion but no history of surgery and no comorbid pathologies, so-called “green knees”, exclude those most likely to have an unfavourable outcome. However, most patients (70% of cases) have combined lesions, or “red knees” [17, 18]. Martinčič et al.  followed 151 patients for an average of 10 years after knee cartilage repair. They found no statistically significant difference in outcomes between patients with and without a history of prior surgery. In our study, we chose not to exclude “red knees”, so as to ensure that our sample was representative of the patients most commonly seen in clinical practice. We found no significant differences when patients were stratified by traumatic versus non-traumatic aetiology, prior surgery, combined surgery, or lesion location.
Despite showing an improvement in scores, our patients aged > 45 years fared rather poorly compared to their younger counterparts. Turinetto et al.  demonstrated that advancing age decreases the potential for differentiation, immunomodulatory effect, and migration ability of pluripotential cells and chondrocytes, thus reducing the healing power of tissues. An alternative to improve this capacity would be to introduce cell growth factors such as SDF-1alpha , which has been demonstrated to increase recruitment of cartilage progenitor cells in laboratory studies.
Overall there were five cases of treatment failure (7.7%). Assessment showed that all patients needing conversion to total knee replacement had lesions that required greater healing potential and were over 45 years of age. The only patient whose implant came loose was also > 45 years old and had a large lesion, despite no history of trauma. Compared to those of other series, however, our failure rate was similar. Verdonk et al.  reported a 5.3% failure rate, and Delcogliano et al. , 2 out of 19 cases (10.5%).