Comparison of lumbar CT and routine lumbar MRI sequences in detecting lumbosacral transitional vertebrae in patients with low back pain and the value of coronal STIR sequence on MRI

Purpose: This study aims to compare the detection rates of LSTV (lumbosacral transitional vertebrae) and its subtypes in routine lumbar MRI performed without a coronal sequence with those of lumbar spine CT, which is the gold-standard method in this regard. Materials and methods: This retrospective study evaluated 1,560 patients who presented with "back pain" between February 2016 and April 2024 and had both lumbar MRI and CT images recorded. A total of 105 patients with LSTV on CT were identified, and two observers independently reviewed the MRI images of these patients for LSTV detection, Castellvi subtype classification, extraforaminal stenosis, and detection of edema at the pseudoarticular level. Results: LSTV was identified in 9.16% of the 1.446 patients. The mean age of the patients was 60, and 72.4% were female. The MRI evaluation revealed LSTV in 64 patients (61%). A high level of agreement was observed between the observers for LSTV detection (κ=0.795, p<0.001); however, a significant discrepancy appeared in subtype classification (κ=0.0, p=0.755). When comparing CT and MRI results, the accuracy of MRI in classifying LSTV types remained low (κ=0.192, p<0.001). Both observers similarly detected edema at the pseudoarticulation level (κ=0.9576, p<0.001). Extraforaminal stenosis was identified in 20 patients on CT and 16 patients on MRI, with a high degree of agreement between the observers (κ=0.926, p<0.001). Conclusion: LSTV is often overlooked in routine MRI protocols because the focus is on disc pathology. A coronal STIR sequence is essential for detecting LSTV and identifying inflammation and stenosis at this level.