Long-term degradation of poly-lactic co-glycolide/β-tricalcium phosphate biocomposite anchors in arthroscopic bankart repair: a prospective study.

PURPOSE: To evaluate, using magnetic resonance (MR), the biological efficacy of anchors made of 30% β-tricalcium phosphate and 70% poly-lactic co-glycolide (PLGA) used for the repair of Bankart lesions after shoulder instability.

METHODS: Twenty consecutive patients who were candidates for surgical treatment for unidirectional, post-traumatic shoulder instability were treated arthroscopically with anchors made of 70% PLGA plus 30% β-tricalcium phosphate preloaded with OrthoCord suture (DePuy Mitek, Raynham, MA). Fifteen of them were evaluated by MR at least 16 months after the intervention. A second evaluation was performed at least 12 months after the first evaluation in the patients in whom implanted anchors were still visible at the first evaluation (n = 5) with a low-intensity signal in all sequences. Two radiologists, with different amounts of experience (15 and 3 years), separately evaluated the MR patterns of the trabecular glenoid bone, the walls of the bone tunnel, and the signal from the anchors. The following parameters were considered in the MR evaluation: integrity of the tunnel edge (grade 0 to 2), intensity of the signal from the anchor site (grade 1 to 3), and presence of cystic lesions. The normal signal from the glenoid trabecular bone has been used as the reference parameter. The anchors were considered independent variables, and thus each one was analyzed individually, even in the same patient. At the final clinical follow-up, a Rowe questionnaire was filled out for each patient.

RESULTS: Overall, 44 anchors were evaluated (33 anchors at the first follow-up and 11 anchors at the second follow-up). The mean follow-up period was 28.6 months. With the exception of 2 patients (10%), none of the patients had any episodes of dislocation, having satisfactory postoperative results. No cystic lesions were detected by MR imaging. The interobserver concordance between the 2 radiologists calculated with the Cohen κ was substantial (κ = 0.780 and κ = 0.791 for integrity of tunnel edge and for intensity of signal from anchor site, respectively). Both the integrity of the tunnel border and the intensity of the signal at the site of the anchors that had been implanted more than 24 months before the evaluation were significantly different from those of anchors implanted less than 24 months before the evaluation (tunnel border grade of 0 in 41%, 1 in 50%, and 2 in 9% v 0 in 4.5%, 1 in 50%, and 2 in 45.5% [P = .003]; anchor signal grade of 1 in 41%, 2 in 45.5%, and 3 in 13.5% v 1 in 13.5%, 2 in 41%, and 3 in 45.5% [P = .03]). Analysis of the linear contrasts (analysis of variance) showed a linear increase in the mean values for time to increased tunnel border grade (grade 0, 22 ± 4 months; grade 1, 27 ± 8 months; and grade 2, 29 ± 5 months [P = .02]) and grade of intensity of the signal in the anchor site (grade 1, 24 ± 6 months; grade 2, 26 ± 7 months; and grade 3, 29 ± 7 months [P = .05]).

CONCLUSIONS: Anchors made of 30% β-tricalcium phosphate and 70% PLGA showed excellent biological efficacy, without causing significant cystic lesions, producing gradual changes in the MR signal that seems to become equivalent to that of the glenoid trabecular bone at a mean of 29 months after implantation.

LEVEL OF EVIDENCE: Level IV, therapeutic case series.