Background/Purpose : Understanding the effect of therapies in the vertebral compartments is relevant to bone biology and clinical practice. We developed an improved technique using cortical shell segmentation-based layering of high resolution quantitative computed tomography (HR-QCT) scans of T12 vertebral bodies to evaluate compartment-specific changes in BMD and microstructure.

Methods: In an international, randomized, phase 2 study (McClung et al., N Engl J Med. 2014), postmenopausal women with low BMD supplemented with calcium and vitamin D received subcutaneous (SC) romosozumab (210 mg monthly), teriparatide (20 mcg SC daily), or placebo (PBO). A subset of these women underwent HR-QCT scanning of T12 (N=11 romosozumab, 12 teriparatide, 8 PBO) at baseline and month 12. For cortical HR-QCT analysis (blinded to treatment assignment), adjacent 200 micron thin films of the cortical shell were evaluated to determine changes from the outer soft tissue bordering the vertebrae to the medullary spongiosa (Figure). In addition to the standard cancellous compartment variables previously described (Graeff et al., J Bone Miner Res. 2007), this improved method allows accurate determination of cortical variables in the subregions, including apparent and corrected (deconvolved) cortical thickness, bone mineral content (BMC), and BMD. Changes in cortical thickness are modeled assuming an average 50% mineralization of newly added matrix.

Results: At baseline, mean (SD) apparent cortical thickness of 1.37 (0.13) mm was corrected to a cortical thickness of 0.29 (0.05) mm. At month 12, romosozumab significantly improved cortical BMC and BMD from baseline and in comparison to teriparatide or PBO (all P≤0.0003; Table). These gains were attained by both endosteal and periosteal bone matrix apposition. Improvements in cancellous BMD were similar between romosozumab and teriparatide.

Conclusion: Using HR-QCT scans of the spine, it is possible to evaluate changes across the cortical shell of the vertebral bodies and determine alterations in the endosteal and periosteal regions. The anatomical location and magnitude of these changes could impact changes in bone strength and thus affect fracture risk. Romosozumab administration was associated with significant increases in cortical thickness and improvement in all measured cortical parameters at 12 months compared with teriparatide or PBO. The clinical effect of romosozumab to reduce fractures is being evaluated in an ongoing phase 3 clinical program.