HOW DOES MAGNETIC RESONANCE IMAGING REPRESENT HISTOLOGIC FINDINGS IN THE EQUINE DIGIT?

Abstract

Magnetic resonance (MR) imaging is increasingly used in the diagnosis of equine foot pain, but improved understanding of how MR images represent tissue-level changes in the equine foot is required. We hypothesized that alterations in signal intensity and tissue contour would represent changes in tissue structure detected using histologic evaluation. The study objectives were to determine the significance of MR signal alterations in feet from horses with and without lameness, by comparison with histopathologic changes. Fifty-one cadaver feet from horses with a history of lameness improved by palmar digital analgesia (n = 32) or age-matched control horses with no history of lameness (n = 19) were stored frozen before undergoing MR imaging and subsequent histopathological examination at standard sites (deep digital flexor tendon, navicular bone, distal sesamoidean impar ligament, collateral sesamoidean ligament, and navicular bursa). Using MR images, signal intensity and homogeneity, size, definition of anatomic margins, and relationships with other structures were described. Alterations were graded as mild, moderate, or severe for each structure. For each anatomic site examined histologically the structures were described and scored as no changes, mild, moderate, or severe abnormalities, also taking into account adhesion formation within the navicular bursa detected on macroscopic examination. Alterations in MR signal intensity were related to changes at the tissue level detected by histologic examination. A sensitivity and specificity comparison of MR imaging with histologic examination was used to evaluate the significance of MR signal alterations for detection of moderate-to-severe lesions of the deep digital flexor tendon (DDFT), navicular bone, distal sesamoidean impar ligament (DSIL), collateral sesamoidean ligament (CSL) and navicular bursa. Agreement between the MR and histologic grading was assessed for each structure using a weighted kappa agreement. Direct comparison between histology and MR imaging for individual limbs revealed that signal alterations on MR imaging did represent tissue-level changes. These included structural damage, fibroplasia, fibrocartilaginous metaplasia, and hemosiderosis in ligaments and tendons; trabecular damage, osteonecrosis, fibroplasia, cortical defects, and increased vascularity in bone; and fibrocartilage defects. MR imaging had a high sensitivity and specificity for most structures. MR imaging had high specificity for lesions of the DDFT, CSL and navicular bursa, quite high specificity for lesions of the medulla of the navicular bone and its proximal aspect, with moderate specificity for the DSIL, and distal, dorsal and palmar aspects of the navicular bone, and was sensitive for detection of abnormalities in all structures except the dorsal aspect of the navicular bone. When MR and histologic grades alone were compared, there was good agreement between MR and histologic grades for the navicular bursa, DDFT, navicular bone medulla and CSL; moderate-to-good agreement in grades of the distal and palmar aspects of the navicular bone; fair to moderate in grades of the DSIL, and poor agreement for the dorsal and proximal aspects of the navicular bone. The results of this study support our hypothesis and indicate the potential use and limitations of MR imaging for visualization of structural changes within osseous and soft tissue structures of the equine foot.