The prevalence of transitional vertebrae in the lumbar spine

Abstract Background context Lumbosacral transitional vertebrae (LSTVs) are a congenital vertebral anomaly of the L5–S1 junction in the spine. This alteration may contribute to incorrect identification of a vertebral segment, leading to wrong-level spine surgery and poor correlation with clinical symptoms. Although several studies describe the occurrence of this anomaly in back pain populations, investigation of the prevalence in the American general population is lacking. Purpose To establish the prevalence rates for LSTVs in the general population. Study design Retrospective review. Patient sample Consecutive kidney-urinary bladder (KUB) radiographs of subjects from the past 2 years (2008–2009). Outcome measures Clinical demographics, number of lumbar vertebrae, L5–S1 transverse process (TP) height, and rib length. Methods Consecutive adult KUB studies of adult subjects were queried with clear visibility of the last rib’s vertebral body articulation, all lumbar TPs, and complete sacral wings. Exclusion criteria consisted of any radiologic evidence of previous lumbosacral surgery that would obstruct our measurements. A total of 1,100 abdominal films were reviewed, and 211 were identified as being adequate for the measurement of the desired parameters. Results Two hundred eleven subjects were identified as eligible for the study, and 75 (35.6%) were classified as positive for a transitional lumbosacral vertebra. The most common anatomical variant was the Castellvi Type IA (14.7%). The average age at the time of the KUB study was 59.8 years (18–95 years). One hundred ninety-seven subjects (93.4%) presented five lumbar (nonribbed) vertebrae, and only 14 (6.6%) had six lumbar vertebrae. Conclusions The significance of lumbosacral transitional level changes to the establishment of pain, degenerative changes, stenosis, and disc disease have been well documented in symptomatic patients. Although LSTV’s role in low back pain remains controversial, our study has shown that, when the same criteria are used for classification, prevalence among the general population and symptomatic patients may be similar.