A rapid method to identify lumbarization on lumbar spine radiography and computed tomography

Lumbosacral transitional vertebrae (LSTV) are very common congenital spinal anomalies that consist of sacralization of the lowest lumbar vertebra or lumbarization of the uppermost sacral vertebra. They are a main cause of wrong-level surgery.

This study aimed to explore morphological characteristics which can help to rapidly identify LSTV on lumbar spine radiography and computed tomography.

Retrospective, case-control study.

(1) Anterior superior angle (ASA) of the last presacral vertebra (PSV)-the angle between the anterior edge and superior endplate of the last PSV on a mid-sagittal lumbar CT view; (2) Posterior angle (PA) of the last PSV and the first sacral vertebra-the angle between the posterior edge of the last PSV and the posterior edge of the uppermost sacral-type segment on a mid-sagittal lumbar CT view; (3) Location of the iliac crest line (ICL) on AP whole-spine image-it was defined based on its location with respect to the transverse processes of the vertebra above the last PSV. An ICL higher than the inferior edges of the transverse process of the vertebra above the last PSV was defined as high ICL.

Data of all patients, who underwent lumbar computed tomography (CT) and whole-spine radiography for lumbar spinal stenosis or lumbar disc herniation, at a single institution from 2017 to 2023, were analyzed from the local database.

Data of 909 patients were grouped according to lumbosacral anatomy as follows: lumbarization of the first sacral vertebra (LZ), sacralization of the fifth lumbar vertebra (SZ), and normal spine (control). The anterior superior angle (ASA) of the last PSV, posterior angle (PA) of the last PSV and the first sacral vertebra, and location of the iliac crest line (ICL) on AP whole-spine imaging were measured. Receiver operating characteristic (ROC) curve analysis was performed to assess the diagnostic performance of the measurements (ASA; PA; ICL; ASA and PA; ASA, PA, and ICL). Inter- and intra-observer reliabilities of each measurement were assessed using the intraclass correlation coefficient (ICC).

LSTV were present in 105 patients (11.6%): LZ in 76 (8.4%) and SZ in 29 (3.2%). ASA of the last PSV (85.7° ± 4.1° vs. 90.6° ± 2.9°; P < 0.001) and PA of the last PSV and the first sacral vertebra (19.7° ± 14.2° vs. 36.9° ± 9.1°; P < 0.001) were significantly lower in the LZ group than the control group. The prevalence of a high ICL was significantly higher in the LZ group than the control group (36.8% vs. 0.9%; P < 0.001). Mean ASA and PA and prevalence of high ICL were 90.1° ± 2.6°, 32.6° ± 11.4°, and 6.9%, respectively, in the SZ group; corresponding values in the control group were 90.6° ± 2.9°, 36.9° ± 9.1°, and 0.9%, respectively. The differences were not significant. Among the 37 patients with a high ICL, 30 (81.1%) also had an LZ. Twenty-three of the 24 patients (95.8%) with both a high ICL and PA < 20° had an LZ. ASA, PA, and ICL combined demonstrated excellent or good performance in differentiating patients in the LZ group from those in the control group (area under the ROC curve = 0.91; 95% confidence interval, 0.85-0.96) and from those in the SZ group (area under the ROC curve = 0.84; 95% confidence interval, 0.76-0.92).

Lumbarization can be identified on lumbar spine imaging using a rapid method. When the ICL is high and both the ASA and PA are small on lumbar imaging, whole-spine radiography is recommended to definitively identify if lumbarization is present.