Background

Recently, primary anterior cruciate ligament repair (ACLrep) has been proposed as an alternative to ACL reconstruction (ACLRec) in proximal acute ACL tears. Some authors reported promising early clinical and radiological outcomes after ACL repair but very few papers compared the results of ACLrep and ACLRec in acute tears. The aim of this study is to compare the healing of a repaired ACL and the healing of a reconstructed ACL as evaluated by MRI at 1 year follow-up.

Hypothesis. Our hypothesis is that healing of an acutely repaired ACL is not inferior to a reconstructed ACL as evaluated by MRI.

Methods

A total of 60 patients who underwent early ACL surgery (injury- surgery interval < 14 days) were prospectively enrolled. Patients with proximal ACL tears (type I and type II with good tissue quality according to Sherman classification) underwent primary ACL repair with trans-osseus pullout technique. Patients with type III and IV and/or a fair tissue quality underwent ACLR with doubled gracilis and semitendinosus tendons (DGST) graft. Patients were re-evaluated after 1 year and MRI appearance of the operated ACL was evaluated. 1.5 T MRI was used to evaluate graft status and graft integration at 1-year post-operatively Two criteria have been used to evaluate graft incorporation at MRI: signal-to-noise quotient (SNQ) and graft maturity (water content of the graft based on the Howell scale).

Results

A sample size calculation demonstrated the need for 30 patients in each group for a Power of 95% (with a 97,5% CI) to show the noninferiority of the repair group considering at least +1 of SNQ difference. Radiological parameters were evaluated by an expert radiologist and an orthopedic surgeon and Chronbac’s Alpha was made to assess inter observer (0,86) and intraobserver (0,82) reliability. There were no significant differences in gender distribution, age, preoperative Tegner and BMI between groups. The mean signal-to-noise quotient (SNQ) was 2.21 ± 2.45 for ACLrep and 2.52 ± 1.69 for ACLrec (p=0.266). The unilateral 95% CI for SNQ in ACLrep was 2,21 (to 3,08) (< to 4,21 limit for noninferiority) and ACLrec was 2,52 (to 3,12), p=0,128. Hence, there are non-statistically differences between the two groups according to the SNQ and the noninferiority was proven. For Howell scale in ACLrep 21 patients were assigned to grade 1, 7 to grade 2, 2 to grade 3 and 0 to grade 4 while in ACLrec 20 patients were assigned to grade 1, 6 to grade 2, 4 to grade 3 and 0 to grade 4. Howell scale showed no differences between groups(p=0.769).

Conclusion. 1 year post-operatively the MRI appearance of a repaired ACL showed noninferior maturation and healing compared with a hamstrings reconstructed ACL.