Introduction: Interest in the relationships between subchondral bone pathology and cartilage breakdown has been stimulated by the observations that bone marrow edema (BME) is related to both pain and bone remodeling and that the progression of cartilage lesions is greater in joints with significant BME. The hypothesis of this study is that changes in perfusion in subchondral bone bear a functional relationship to bone remodeling and cartilage degradation and are a part of a physicochemical signaling mechanism to osteoblasts. We have utilized dynamic, contrast-enhanced magnetic resonance imaging (MRI) to assess perfusion and BME in osteoarthritis (OA) and osteonecrosis (ON).

Methods: Investigation of marrow perfusion in BME was performed in both the Dunkin-Hartley guinea pig model of OA patients and cohort of 26 control patients. The human study was performed on a 1.5T magnet using a dedicated surface coil and STIR [3500/17/150 (TR/TE/TI)] and VIBE [5.50/2.89 (TR/TE); 10° (flip angle)] pulse sequences. We determined pharmacokinetic parameters of marrow perfusion according to the two compartment model of Brix, which is characterized by rate and volume transfer constants that can be derived mathematically from time-signal intensity curves.

Results: In the guinea pig model, inflow slope was similar at all ages; kel was decreased in the affected medial, but not in the normal lateral, tibia indicating reduced perfusion and outflow obstruction. Comparison of BME and perfusion metrics with morphological features (Mankin scores and subchondral bone plate thickness) at the medial tibia demonstrates that changes in perfusion dynamics precede bone remodeling and cartilage breakdown by several months. Compared to normal marrow, kinetic parameters of contrast-enhancement in areas of BME included higher initial slope (p< .001), higher A (p< .001), lower kep (p=.004), and lower kel (p< .001). In areas of BME around ON, there was significantly lower A (p=.009) and lower kel (p=.04) compared to BME adjacent to OA, but no significant difference in either initial slope (p=.06) or kep (p=.26).

Discussion: To our knowledge, these are the first reports of the use of dynamic, enhanced-MRI to characterize bone marrow perfusion in BME associated with OA and ON. In the Dunkin-Hartley guinea pig, reduced perfusion in BME temporally precede alterations in bone remodeling and appearance of cartilage lesions, and are spatially localized to bone subjacent to eventual cartilage lesions. We have also demonstrated similar perfusion kinetics associated with BME in human ON and OA. Calculations of intraosseous pressure associated with outflow obstruction and decreased perfusion are consistent with measurements made in end-stage ON and OA. Osteoblasts are known to be responsive to flow, pressure, and pO2. Increased pressure and decreased flow associated with outflow obstruction may constitute physicochemical signals to osteoblasts which result in changes in cytokine expression and contribute to trabecular remodeling and cartilage breakdown.