Numeric and Morphological Verification of Lumbosacral Segments in 8280 Consecutive Patients

STUDY DESIGN: An analysis of imaging data. OBJECTIVE: To investigate concurrent numeric and morphological variations of presacral vertebrae and to propose a modified designation for the lumbosacral transitional vertebra (LSTV). SUMMARY OF BACKGROUND DATA: During the assessment of the lumbosacral vertebra, variations from typical anatomy (numeric, morphological, or both) may confuse the practitioner, potentially leading to significant clinical errors. Common practice, which involves counting cephalad from the presumed fifth lumbar vertebra, may result in inaccurate localization of lumbosacral levels. METHODS: The study group was composed of 8280 consecutive patients who underwent both lumbar magnetic resonance imaging with cervicothoracic scanning and lumbar radiographical examinations. The presacral vertebral number was verified by counting caudally from C2, with cross-referencing of cervicothoracic and lumbar sagittal scans on a picture archiving and communication system workstation. After correlating the numbering on the magnetic resonance images with those on the radiographs, the lumbosacral junction was classified according to the Castellvi's method. RESULTS: Of the 8280 consecutive patients, 214 (2.6%) had 4 lumbar vertebrae (L4), 7384 (89.2%) had 5 lumbar vertebrae (L5), and 682 (8.2%) had 6 lumbar vertebrae (L6). Overall, 877 (10.6%) patients had LSTV of types II, III, or IV, including 439 (5.3%) with sacralized L5 vertebra and 438 (5.3%) with lumbarized S1 vertebra. The most common LSTV was L5-type vertebra with a unilateral type II transition, designated as L5IIa, in 222 (2.7%) patients. The second most common LSTV was L6-type vertebra with a bilateral type III transition in 174 (2.1%) patients that was designated as L6IIIb. Only 6945 (83.9%) of the population were modal type, with 5 lumbar vertebrae without transitional vertebra. All the 214 (2.6%) L4-type and 244 (2.9%) of the 682 L6-type patients presented with no transitional vertebra, looking like a modal L5-type patient. CONCLUSION: Spine physicians and radiologists should consider the possibility of both numeric and morphological variations when evaluating lumbosacral spine images.