Surgically Relevant Anatomy of Marfan Spinal Deformity

Introduction: Understanding pedicle and laminar anatomy is crucial for secure fixation of spinal instrumentation. This study explores the spinal morphology in the Marfan vertebrae and compares it to vertebrae of TDP. Methods: The morphology of 134 vertebrae from T1 to L5 were measured in 16 Marfan patients with spine deformity, using MRI and CT scans. The morphology of 140 vertebrae from T1 to L5 were measured in TDP, using MRI, CT, and X-rays. These values were then compared to each other. Comparisons via paired Student’s t-test were made between the concave and convex side of the curve. Results: On average, Marfan pedicles were significantly thinner (by mean 2mm) than those in TDP (p<0.001) (Fig. 1). Thirty four percent of the total Marfan pedicles were type D (< 2mm wide) vs only 9% in TDP (p<0.001). Marfan type D pedicles were found in 36% of atypical locations (outside of the concave apex). Pedicle width was significantly correlated with both vertebral level and distance from the apex. The average thickness of the Marfan lamina was significantly thinner (by mean 1mm) than those in TDP (p<0.001). Marfan patients with dural ectasia had significantly thinner lamina in the lumbar spine than those without (p=0.023). Pedicle width and laminar thickness measurements from CT and MRI were not significantly different. X-ray pedicle width measurements correlated moderately well with both MRI and CT pedicle width measurements (R2=0.52 and 0.59, respectively). Pedicle measurements by X-ray were larger than by CT scans 82.5% of the time (by mean 1.7mm). Conclusion: Marfan spines are associated with thinning of fixation sites, most pronounced at the apex of the curve. Laminar thinning was more common on the concavity and with dural ectasia. X-rays can be useful in identifying narrow pedicles, but often overestimates the width. Therefore, this study supports obtaining preoperative routine MRI or CTs when operating on Marfan spine with pedicle screws and lamina hooks.