Comparison of tibial anatomical-mechanical axis angle between predisposed dogs and dogs at low risk for cranial cruciate ligament rupture.

Abstract

The purpose of this prospective, radiographic, descriptive study was to compare measurements of tibial anatomical-mechanical axis angle (AMA-angle), tibial plateau angle (TPA), relative tibial tuberosity width (rTTW) and Z-angle from mediolateral radiographs of the tibia between two canine breeds (72 dogs) not predisposed to cranial cruciate ligament rupture (CCLR) and those from a consecutive series of 185 large dogs and 17 West Highland white terriers (WHWT) diagnosed with unilateral, surgically confirmed CCLR. Correlations among these measurements were determined, and levels of inter- and intra-observer variability among and within three observers for each measurement were established using Kendall's coefficient of concordance. Breed had a significant effect on AMA-angle. The median AMA-angle of the subject population of large dogs affected by CCLR was 2.80&#xb0; (range 1.09&#xb0;-5.21&#xb0;); for the WHWT, it was 6.34&#xb0; (range 5.68&#xb0;-8.88&#xb0;); and for the clinically normal dogs, it was 0.74&#xb0; (range 0.00&#xb0;-5.40&#xb0;). In the CCLR group, AMA-angle and TPA were strongly correlated (r=0.745; p<0.0001). A receiver operating characteristic (ROC) curve analysis showed that an AMA-angle higher than 1.87&#xb0; had a sensitivity of 0.941 (95% confidence interval [CI]: 0.898-0.966) and a specificity of 0.965 (95% CI: 0.919-0.987) for predicting CCLR and was more accurate than TPA, rTTW and Z-angle at predicting CCLR (p<0.0001). Good inter- and intra-observer agreement was found for all measurements. The highly significant difference in AMA-angle found between clinically normal dogs and dogs with CCL injury suggests that AMA-angle magnitude may be a clinically relevant predisposing factor for the development of canine CCLR.