Osteoarthritis (OA) is a chronic disease that affects the joints, most commonly hands, hips, knees, feet, and spine. The disease becomes more common with advanced age and is one of the most prevalent causes of disability in older populations. Currently there are no disease-modifying drugs available, and the only curative treatment is total joint replacement.

The cause of pain in OA is not well understood, but it is known that bone marrow lesions (BMLs) identified in subchondral bone by MRI are an important determinant of pain. Abnormal blood vessel growth may be responsible MRI signature of BMLs, and the commonality between pathways for angiogenesis and neurogenesis suggests this pathologic process may be the source of pain in OA. The objective of this study was to characterize the histologic nature of subchondral cysts identified by micro computed tomography (μCT) which had been registered with MRI images in which marrow lesions were identified.

Femoral heads were collected from 10 patients (6 females and 4 females;​ age 29- 80) who underwent total hip arthroplasty. All patients had MRIs performed within 6 months prior to surgery. The heads were fixed and scanned with μCT to identify cysts in the subchondral bone. A block of the femoral head containing the cyst of interest was resected and processed for histologic analysis. A two-dimensional image from the μCT that corresponded to the histologic slice was matched with a coronal view from the MRI. The primary compressive group was reliably identified on μCT images and served as good indicator for orienting the CT to match with the MRI. The subchondral cysts that were matched to MRI all consisted of predominantly fibrous bone marrow and frequently had a large number of blood vessels within the tissue. Three of the eight cysts had cartilage intrusions that were located mostly within peripheral trabecular bone, though one cyst contained a nodule of cartilage surrounded by organized fibers with the texture of granulation tissue. The process of image registration was mostly performed manually, but the development of this process will contribute to a more refined, semi-automated process in the future. The ability to correlate the histopathology of CT-identified lesions with a signature pattern on MRI will be an important tool for better characterizing the nature of BMLs and understanding the pathogenesis of OA.