INTRODUCTION
Detection of lumbosacral transitional vertebrae (LSTV) is of clinical importance as failure to recognise this variation may lead to wrong level surgery. Vertebral body counting cephalad from L5 or caudad from T12 can increase the risk for surgical or procedural error when LSTV are not recognised.
Previous literature shows a variable prevalence of LSTV ranging 4-35.9% in the North American, Asian, Australian, and European populations. While most the population of New Zealand is mostly of European (NZE) heritage, NZ has a large indigenous (Māori) population (16.5%). Cook and Baker showed that Māori have variations in lumbar spine anatomy, but it is unknown if ethnicity is a factor for having LSTV.
The aim of this study was to report the prevalence of LSTV in a NZ cohort and determine the association with ethnicity.
MATERIALS AND METHODS
In a cross-sectional analysis, 336 computer tomography (CT) scans performed for the assessment of major trauma were evaluated. Scans were excluded if the imaging was incomplete or existing pathology precluded accurate assessment.Vertebral counts were confirmed counting caudad from C2. LSTV were subsequently categorized using the Castellvi classification system. Analysis comprised scans from 200 NZE and 136 MA. Proportions were compared using z-test.
RESULTS
MA were younger than NZE (32 vs. 47 years; p<0.001). There was a similar gender distribution between the groups (30% and 31% female respectively; p=0.87). The overall prevalence was 48.2%. LSTV was more common in male (54.2%), than in female (35.2%).
162 (48%) had some class of LSTV; 52% of MA and NZE 45.5% (p=0.27).
MA and NZE respectively had Castellvi: type-1a 11% vs. 4% (p=0.034); type-1b 17% vs. 9% (p=0.56); type-2a 4% vs. 9% (p=0.139); type-2b 10% vs. 11% (p=1); type-3a 0.7% vs. 0.5% (p=0.001); type-3b 8% vs. 12% (p=0.39); type-4 0.7% vs. 2% (p=0.896).
Patients with LSTV had greater prevalence of abnormal vertebral counts, L4 (8%) and L6 (8%), compared to those without LSTV 1% and 1% respectively. No significant difference for ethnicity was found in in those with abnormal lumbar vertebrae counts. T13 (5%) was more prevalent in in NZE compared to MA (0%) while both ethnic groups had T11 in 1%.
CONCLUSIONS
In this first local study, using whole spine CT permitting accurate counting, we found a higher prevalence of LSTV compared to previous studies. Whole spine imaging counting from C2 caudad remains the gold standard. Māori were more likely to have LSTV, although predominantly Castellvi type I which are of lesser clinical significance. This may simply reflect variations in skeletal size and anatomy between the ethnic groups. NZE had a greater prevalence of Castellvi types II-IV which surgeons must be cognisant of when planning surgery.