Tissue identification with micro-magnetic resonance imaging in a caprine spinal fusion model

Abstract

Nonunion is a major complication of spinal interbody fusion. Currently X-ray and computed tomography (CT) are used for evaluating the spinal fusion process. However, both imaging modalities have limitations in judgment of the early stages of this fusion process, as they only visualize mineralized bone. Magnetic resonance imaging (MRI) could be of great value as it is able to discriminate between different types of tissue. A feasibility study was performed in nine animals from a goat spinal fusion study, to evaluate the detection capacity of different tissues with micro-MRI. In this study bioresorbable polylactic acid cages were used. Six- and 12-months follow-up specimens were scanned in a 6.3 T micro-MRI scanner. After scanning, the specimens were processed for histology. Different types of tissue as well as the degradable cage material were identified in the fusion zone and designated as regions of interest (ROIs). Subsequently, the location of these ROIs was determined on the corresponding micro-MRI image, and average signal intensities of every individual ROI were measured. An excellent match was seen between the histological sections and micro-MRI images. The micro-MRI images showed quantifiable differences in signal intensity between bone with adipose marrow, bone with hematopoietic marrow, fibrocartilage, fibrous tissue, and degradable implant material. In time the signal intensity of bone with adipose marrow, bone with hematopoietic red marrow, and of fibrous tissue remained relatively constant. On the other hand, the signal intensity of the degradable implant material and the fibrocartilage changed significantly in time, indicating change of structure and composition. In conclusion, in our model using bioresorbable cages the MRI provides us with detailed information about the early fusion process and may therefore, allow early diagnosis of non-union.