Prospective evaluation of patellar tendon graft fixation in anterior cruciate ligament reconstruction comparing composite bioabsorbable and allograft interference screws

Purpose: Bioabsorbable interference screws have proved to be biologically safe and to provide adequate graft fixation. Metallic interference screws have therefore been continuously replaced over the years. However, degradation times are highly variable, and bony replacement of the screw does not always occur in human beings. Composite interference screws have recently been introduced to enhance bony integration. We evaluated 2 different composite interference screws and compared them with an allograft interference screw over a 2-year period after anterior cruciate ligament (ACL) reconstruction with an autologous bone-patellar tendon-bone graft. Methods: Three groups of patients were prospectively evaluated. Poly-L-lactid acid/hydroxyapatite composite screws (BioRCI-HA; Smith & Nephew, Andover, MA) (group A), poly-L-lactid acid/beta-tricalcium phosphate composite screws (Bilok; Atlantech, Radevormwald, Germany) (group B), or allograft interference screws (CorlS; Regeneration Technologies, Alachua, FL) (group Q were used for tibial fixation of a patellar tendon autograft in ACL reconstruction. Each group consisted of 20 patients (mean age, 32.2 +/- 10.9 years in group A, 32.3 +/- 10.6 years in group B, and 31.1 +/- 6.6 years in group Q with no significant (P < .05) differences in age, sex, and time of follow-up. Subjective and clinical International Knee Documentation Committee parameters were evaluated preoperatively and at 3, 12, and 24 months postoperatively; computed tomography scans were obtained postoperatively and at 3, 12, and 24 months; and magnetic resonance imaging (MRI) scans were obtained at 3 and 24 months postoperatively. Results: Screw breakage during insertion occurred in 2 cases (1 in group A and 1 in group B). There was no significant (P < .05) difference in subjective and clinical results at any time of follow-up. No inflammatory response could be detected on MRI in arty of the patients in all groups. Computed tomography scans documented complete bone block incorporation at 3 months in all groups. Both composite interference screws showed signs of degradation but were still clearly visible 24 months after ACL reconstruction. At 12 months, a sclerotic rim around the screw cavity could be outlined in group B only. No bony replacement had taken place up to 24 months postoperatively in group A or B. In group C the allograft bone screw was completely integrated and barely visible after 24 months. Conclusions: All 3 screw types provided adequate graft fixation and were associated with excellent clinical results and no inflammatory response on MRI. Ultimately, a resorbable screw has to be replaced by bone to facilitate revision surgery. Formation of a sclerotic rim around the outline of the screw makes later bony replacement (group B) unlikely. After 24 months, complete bony integration had only taken place when allograft screws were used. Clinical Relevance: At 24 months, no advantage of composite screw materials over conventional bioabsorbable screws could be detected. If composite materials will be of any advantage with respect to bony replacement has to be observed with longer-term follow-up. The allograft bone screw was completely incorporated and replaced by cancellous bone after 24 months. Unfortunately, the screw's more complicated handling, higher cost, and limited availability impair the possibilities for its standard clinical use.