Lumbar extraforaminal entrapment: performance characteristics of detecting the foraminal spinal angle using oblique coronal MRI. A multicenter study.

BACKGROUND CONTEXT

Previous conventional magnetic resonance imaging reports on extraforaminal entrapment (e-FE) on L5-S1 have been problematic because of their complexity or lack of sensitivity and specificity. In this study, we propose a simple diagnostic method for e-FE.

PURPOSE

The purpose of this study was to determine the sensitivity and specificity of using the difference in the foraminal spinal nerve (FSN) angle of the L5 nerve, as determined by oblique coronal T2-weighted imaging (OC-T2WI), for diagnosing L5-S1 unilateral e-FE.

STUDY DESIGN

The study design involves diagnostic accuracy with retrospective case-control study.

PATIENT SAMPLE

Seventy consecutive patients with unilateral L5 radiculopathy who underwent unilateral L5-S1 extraspinal canal decompression for e-FE or 4/5 intraspinal canal decompression for lumbar spinal canal stenosis between 2009 and 2013 were included.

OUTCOME MEASURES

The Japanese Orthopedic Association score, Visual Analog Scale score for leg pain, and OC-T2WI for the FSN angle of the L5 nerve were examined.

METHODS

The 70 patients were divided into two groups: Group A (n=21) with unilateral L5-S1 e-FE and Group B (n=49) with intraspinal canal L4-L5. Group C (n=44) comprised the control group, which included only patients with back pain without leg radiculopathy. All patients underwent OC-T2WI, and the differences in the FSN angle of the fifth lumbar spinal nerve between the symptomatic and asymptomatic sides (ΔFSN angle) were examined and compared among the groups.

RESULTS

There were no significant differences in the patient characteristics among the three groups. The ΔFSN angle was 17° in Group A, 4.8° in Group B, and 6.4° in Group C, and the laterality was significantly larger in Group A than in the other two groups. A receiver-operating characteristic curve showed areas under the curve between groups A and B and between groups A and C of 0.93 and 0.97, respectively. In addition, the cutoff value of the ΔFSN angle (10°) indicated diagnostic accuracies of 94% and 91% (sensitivity and specificity) and of 93% and 95%, respectively.

CONCLUSIONS

Determining differences in the FSN angle between the symptomatic and asymptomatic sides of greater than 10° via OC-T2WI represented a simple, readily available, and complementary diagnostic method for lumbar e-FE.