Percent spinal canal compromise on MRI utilized for predicting the need for surgical treatment in single-level lumbar intervertebral disc herniation.

BACKGROUND CONTEXT: There is limited information describing the correlation between the initial quantitative measurements on magnetic resonance imaging (MRI) scans of disc herniation area, canal cross-section areas, percent canal compromise, and disc herniation location to the need for surgery.
PURPOSE: Our aim is to determine if the size of disc herniation area, canal cross-section area, percent canal compromise, and disc herniation location taken from MRI images of patients with symptomatic single-level lumbar herniated intervertebral discs upon initial presentation to a spine surgeon, were predictive of the need for surgical treatment. STUDY DESIGN/
SETTING: This is a retrospective case matched study of patient MRI images in the senior author's private practice. PATIENT SAMPLE: From a pool of 332 patients with sciatica caused by lumbar intervertebral disc herniations at our institution, 65 patients had surgery, of which MRI images were available and analyzed on 44 patients. Forty-four additional patients were randomly selected from the remaining 267 original group as nonoperative controls.
METHODS: The axial MRI image showing the largest canal compromise by the herniated disc was selected for measurements. Using T1- and T2-weighted images, the areas of interest were digitally scanned at high resolution. The canal area and disc herniation area measurement were calculated from the total number of pixels per cross-sectional area, multiplied by a scan correction factor, mm(2) /pixel. Disc herniation locations were classified into either central or paracentral. The percent canal compromise was obtained by disc herniation area divided by canal cross-section area and multiplied by 100.
RESULTS: The surgical group's overall mean herniated disc area was 219.6 square millimeter (mm(2)), 179.8 at L4-5, and 267.4 at L5-S1. The nonoperative group's overall mean herniated disc area was 178.4 mm(2), 135.1 at L2-3, 160.3 at L4-5, and 207.4 at L5-S1. The surgical group's overall mean canal cross-sectional area was 471.8 mm(2), 418.6 at L4-5, and 535.6 at L5-S1. The nonoperative group's overall mean canal cross-sectional area was 541.3 mm(2), 518.1 at L2-3, 446.8 at L4-5, and 669.9 at L5-S1. The overall percent canal compromise ratio in the surgery group was 46.7%, 44.1% at L4-5, and 49.8% at L5-S1. The overall percent canal compromise in the nonoperative group was 34.2%, 34.1% at L2-3, 36.1% at L4-5, and 31.8% at L5-S1. The percent canal compromise in central herniations at L4-5 level was 53.0% in the surgical group, and 32.8% in the nonoperative group; at the L5-S1 level surgical group percent canal compromise was 64.1% and in the nonoperative group canal compromise was 27%. L4-L5 level paracentral herniations canal compromise was 36.7% in the surgical group compared with 42.5% canal compromise in the nonoperative group. At the L5-S1 level the canal compromise was 45% in the surgical group and 34.8% in the nonoperative group.
CONCLUSIONS: Our findings show a trend for patients treated with surgery to have larger disc herniation areas and smaller canal cross-section areas, corresponding to larger percent canal compromise than the nonoperative group. Centrally located herniations followed this trend closely at all levels studied. However, the paracentral herniation at the L4-5 level does not follow this trend, possibly because paracentral disc herniation clinical course is determined more by herniation location rather than the overall herniation size.