2015 ISAKOS Biennial Congress ePoster #1285

Predictive Factors and Odds of Re-Injury at 2 Years: A Randomized Clinical Trial Comparing Three Graft Types for ACL Reconstruction

Nick Mohtadi, MD, MSc, FRCSC, Calgary, AB CANADA
Denise S. Chan, MBT, MSc, Calgary, AB CANADA
Rhamona Barber, BKin, CAT(C), Calgary, AB CANADA
Elizabeth Oddone Paolucci, PhD, Calgary, AB CANADA

University of Calgary Sport Medicine Centre, Calgary, Alberta, CANADA

FDA Status Not Applicable

Summary: The predictive factors and odds for traumatic re-rupture and re-injury, and atraumatic graft failure are evaluated in patients with patellar tendon, quadruple-stranded or double-bundle hamstring ACL reconstructions in this double-blind RCT at a minimum 2-years post-operatively.




This large double-blind randomized clinical trial (RCT) evaluates quality of life and identifies the predictive factors and odds of traumatic re-rupture and re-injury, and atraumatic graft failure at 2 years follow-up.


Three-hundred-and-thirty patients (183 males) aged 14-50 (median 27.0) years were randomly (computer-generated, varied block sizes) and intra-operatively allocated to ACL reconstruction with one anatomic autograft (n=110/group): 1) Patellar Tendon (PT; mean 29.2 years), 2) Quadruple-stranded Hamstring Tendon (HT; mean 29.0 years), or 3) Double-Bundle using hamstring tendons (DB; mean 28.8 years). Patients and the independent trained assessor were blinded to treatment allocation.

The disease-specific Anterior Cruciate Ligament Quality of Life (ACL-QOL) and the Tegner Activity Level were assessed at baseline, 3 and 6 months, 1 and 2 years post-operatively. At 2 years, the proportion of complete traumatic re-ruptures, traumatic re-injuries (which includes partial and complete re-ruptures), and atraumatic graft failures were determined for each technique.

An analysis of variance for repeated measures and a Bonferroni post-hoc method for multiple comparisons were used for the mean ACL-QOL scores. Chi-square proportional analyses were used for the re-injury and graft failure data.

Logistic regression was used to assess 5 independent predictors of re-rupture, re-injury or failure, determined a priori: chronicity, graft type, age at surgery, sex, and Tegner activity level. Odds ratios were also determined for each of these factors. A 5% significance level was used for all analyses.


The baseline characteristics between the groups were not statistically different. At 2 years, 322 patients completed follow-up. Mean ACL-QOL scores were not different between groups (p=0.59): PT=85 (SD 17, 95%CI 81–88); HT=83 (SD 18, 95%CI 79–86); DB=82 (SD 18, 95%CI 79–86).

More complete traumatic re-ruptures occurred in the hamstring and double-bundle groups compared to patellar tendon reconstructions: PT=3; HT=7; DB=7 (p=0.37). Traumatic re-injuries were statistically different between groups: PT=3; HT=12; DB=11 (p=0.05). Atraumatic graft failures were not statistically different between groups: PT=16; HT=17; DB=20 (p=0.75).

Logistic regression analysis identified younger age as a significant predictor of complete traumatic re-ruptures and traumatic re-injuries (p<0.01). Males, higher activity level, and patients with HT, DB and acute reconstructions had greater odds of traumatic re-injury. None of these factors were statistically significant. No factors were significant in predicting atraumatic failures.


In this double-blind RCT, there was no difference in quality of life outcome between the ACL reconstruction techniques. More complete traumatic re-ruptures and traumatic graft re-injuries occurred in the HT and DB groups. This trial suggests that younger age is a predictor of complete traumatic re-rupture and traumatic re-injury, irrespective of graft type.