Utility of MRI for Characterizing Articular Cartilage Pathology in Dogs with Medial Coronoid Process Disease

Abstract

Objective: To determine whether assessment of morphological MRI sequences or delayed gadolinium-enhanced MRI of cartilage (dGEMRIC) would have strong correlations with arthroscopic assessment of cartilage pathology in dogs with naturally occurring medial compartment pathology of the elbow.

Methods: Dogs tentatively diagnosed with medial coronoid disease had evaluation of their affected elbows using radiography, morphological MRI sequences, and dGEMRIC MRI evaluation prior to arthroscopy. Elbow radiographs were graded 0-6 for severity of changes. Cartilage of the medial coronoid process (MCP) and humeral trochlea (HT) were scored on a 0-3 scale using anatomical MRI sequences. The T1 relaxation times for the MCP and trochlea were quantified using dGEMRIC. Cartilage pathology was graded arthroscopically using a modified Outerbridge score (MOS) by a surgeon blinded to MRI assessment. Correlations between radiography and MOS, and between MRI and MOS, were quantified.

Results: Twenty-six elbows in 14 dogs were evaluated. There were statistically significant (p < 0.05) moderate correlations between radiographic scores and MOS for the MCP (r = 0.71) and HT (0.57). There was a statistically significant moderate correlation between morphological MRI scoring and MOS for the HT (r = 0.54; p < 0.05), but not for the MCP (p > 0.05). There was a weak, but significant correlation, between the dGEMRIC value and MOS of the MCP (r = 0.41; p < 0.05), but no correlation between the dGEMRIC values and MOS for the HT (p > 0.05).

Clinical relevance: Statistically significant correlations to MOS were identified for both radiography and MRI but neither diagnostic modality provided sufficiently strong correlations to serve as a substitute for arthroscopic evaluation of the articular cartilage.

Keywords: MRI; arthroscopy; dGEMRIC; elbow dysplasia; medial coronoid process.

Figure 1

Representative sagittal plane MRI images. (A) T1-weighted anatomical reference image and (B–D) magnified delayed gadolinium-enhanced MRI of cartilage images depicting three methods for drawing regions of interest (ROIs) for cartilage scoring. (B) One ROI is on the medial coronoid process (MCP) and two on the humeral trochlea (HT). Note that the ROIs include the articular cartilage layer without substantial inclusion of adjacent subchondral bone, joint fluid, or opposing articular surface. (C) A single square ROI is used that spans the joint space and includes the articular cartilage of the medial HT and MCP. (D) Two free-form ROIs are used that trace the MCP and HT.

Figure 2

Sagittal plane MRI images and associated arthroscopic images from three patients (patients A, B, and C) with cartilage pathology of varying severity. From top to bottom within a column, the images are fat-suppressed proton density (PD)-weighted (2.5 mm thick), PD-weighted (1.5 mm thick), and T1-weigthed (1.0 mm thick) images followed by a representative arthroscopic image of the medial coronoid process and the humeral trochlea. Note that the spectrum of cartilage pathology is readily apparent with the arthroscopic images, but less apparent with the MRI images.