Elbow Joint Bone Density Analysis in Labrador Retrievers and Golden Retrievers

In a recent retrospective study published in The Veterinary Journal, investigators conducted bone density (BD) analyses of healthy and diseased elbow joints and identified computed tomography (CT)-based regions of interest (ROIs) associated with elbow pathology in Labrador Retrievers and Golden Retrievers.

Radiography is the traditional imaging modality for evaluating canine elbow joints. For example, previous studies have reported the relationship between radiographic trochlear notch sclerosis and medial coronoid disease (MCD), one of the primary lesions of canine elbow dysplasia. In addition, dual-energy x-ray absorptiometry has been used to quantify BD of the medial coronoid process in healthy dogs and those with MCD. CT has been used to evaluate BD distribution in the subchondral bone plate of normal elbow joints.

Study authors collected retrospective data and grouped study animals into two separate populations. In one population, three observers independently evaluated ROIs in 20 CT elbow joint images from 20 randomly selected dogs (mean age, 22 months). Inter-observer repeatability for each ROI was determined using two 95% reference intervals (differences between measurements within the same dog and measurements from two different dogs).

For the second population, one observer assessed CT images of 80 healthy and 80 diseased elbow joints from young Labrador Retrievers and Golden Retrievers. Hounsfield units (HU) and BD were measured from these images. Using a linear fixed effect model, authors assessed the effects of age, body weight, and breed on HU and BD in normal elbow joints; they also compared HU and BD measurements of healthy and diseased elbow joints (MCD and other elbow pathologies).

CT images were sagittal and sagittal oblique views of the elbow joint. ROIs were the humerus (Hum, Hum1, Hum2), radial head (RH), medial coronoid process base (MCPB1, MCPB2), medial coronoid process apex (MCPA1, MCPA2), trochlear notch of the ulna (TrN), and caudal ulna (CdU).

Authors used receiver operating characteristic (ROC) curves to determine the optimal HU and BD cut-off values between healthy joints and those with MCD. ROC curves were also used to calculate sensitivity and specificity for each ROI.

Hum, Hum1, MCPB2, and MCPA2 were the ROIs with the highest inter-observer repeatability.

Authors reported a significant effect of age on HU and BD values for several ROIs, including MCPB1, MCPA2, and CdU. Breed also had a significant effect on HU and BD values; HU and BD values for most ROIs were significantly higher in Labrador Retrievers than Golden Retrievers. Body weight did not significantly affect any ROI.

For most ROIs, HU and BD values were significantly higher in elbow joints with MCD than healthy elbow joints and those with other pathologies. For the humeral ROIs, these values were significantly higher in healthy elbow joints than those with other pathologies.

ROIs with the highest areas under the curve, sensitivity, and specificity were MCPB1, MCPB2, MCPA1, and MCPA2.

Study results suggest a relationship between increased or decreased HU and BD values and elbow joint pathology; in particular, increased HU and BD values in the medial coronoid base and apex could be associated with MCD. Authors noted that MCPB2 could be a very promising ROI, given that it demonstrated high inter-observer repeatability and was not significantly affected by age or breed.

Authors suggest further evaluation to determine whether this study’s ROIs results could be applied to detection of non-MCD elbow joint pathologies.

Dr. JoAnna Pendergrass received her doctorate in veterinary medicine from the Virginia-Maryland College of Veterinary Medicine. Following veterinary school, she completed a postdoctoral fellowship at Emory University’s Yerkes National Primate Research Center. Dr. Pendergrass is the founder and owner of JPen Communications, LLC.