A novel diagnostic parameter, foraminal stenotic ratio using three-dimensional magnetic resonance imaging, as a discriminator for surgery in symptomatic lumbar foraminal stenosis.

BACKGROUND CONTEXT:No previous studies have reported the radiological features of patients requiring surgery in symptomatic lumbar foraminal stenosis (LFS). PURPOSE:This study aims to investigate the diagnostic accuracy of a novel technique, foraminal stenotic ratio (FSR), using three-dimensional magnetic resonance imaging for LFS at L5-S by comparing patients requiring surgery, patients with successful conservative treatment, and asymptomatic patients. STUDY DESIGN:This is a retrospective radiological comparative study. PATIENT SAMPLE:We assessed the magnetic resonance imaging (MRI) results of 84 patients (168 L5-S foramina) aged ≥40 years without L4-L5 lumbar spinal stenosis. The foramina were divided into three groups following standardized treatment: stenosis requiring surgery (20 foramina), stenosis with successful conservative treatment (26 foramina), and asymptomatic stenotic foramen (122 foramina). OUTCOME MEASURES:Foraminal stenotic ratio was defined as the ratio of the length of the stenosis to the length of the foramen on the reconstructed oblique coronal image, referring to perineural fat obliterations in whole oblique sagittal images. We also evaluated the foraminal nerve angle and the minimum nerve diameter on reconstructed images, and the Lee classification on conventional T1 images. MATERIALS AND METHODS:The differences in each MRI parameter between the groups were investigated. To predict which patients require surgery, receiver operating characteristic (ROC) curves were plotted after calculating the area under the ROC curve. RESULTS:The FSR showed a stepwise increase when comparing asymptomatic, conservative, and surgical groups (mean, 8.6%, 38.5%, 54.9%, respectively). Only FSR was significantly different between the surgical and conservative groups (p=.002), whereas all parameters were significantly different comparing the symptomatic and asymptomatic groups. The ROC curve showed that the area under the curve for FSR was 0.742, and the optimal cutoff value for FSR for predicting a surgical requirement in symptomatic patients was 50% (sensitivity, 75%; specificity, 80.7%). CONCLUSIONS:The FSR determined LFS requiring surgery among symptomatic patients, with moderate accuracy. Foramina occupied ≥50% by fat obliteration were likely to fail conservative treatment, with a positive predictive value of 75%.