A prospective analysis of magnetic resonance imaging findings in patients with sciatica and lumbar disc herniation. Correlation of outcomes with disc fragment and canal morphology.

STUDY
DESIGN: A two-part observational study. In Part 1, consecutive lumbar magnetic resonance imaging scans in patients with sciatica meeting enrollment criteria were prospectively and blindly analyzed by a set protocol. In Part 2, further clinical findings at the time of the magnetic resonance imaging were obtained by retrospective chart review and outcome assessment done at follow-up of more than 2 years.
OBJECTIVES: To determine the quantitative magnetic resonance findings of disc and canal measures in patients with sciatica and to analyze the predictive values of these magnetic resonance imaging and clinical variables on outcomes.
METHODS: Part 1: Consecutive patients with a primary diagnosis of sciatica who came to lumbar magnetic resonance imaging were enrolled, and magnetic resonance imaging dimensions of discs and canal at the herniated level were collected. Part 2: Of 186 patients in Part 1, 135 were followed for more than 2 years; 87 were treated conservatively, and 48 were treated surgically. Outcomes were judged on satisfaction, activity level, medication intake, and reported pain at follow-up (mean, 2.6 years).
RESULTS: Part 1: Wide ranges of disc and canal measurements were seen in all parameters. Significant differences in all magnetic resonance parameters were noted between male and female patients. Men had proportionately greater canal compromise by the affected disc than women. Positive sciatic tension signs and short duration of symptoms correlated with large disc herniation. Right-sided symptomatic herniations were usually larger than left. Part 2: At follow-up, predictors of outcome were determined independently for the surgery and the nonoperative groups. In the nonoperative group, a shorter duration of sciatica was the most significant predictor of a good outcome (P = 0.0018). Moreover, a duration of symptoms less than 6 months, no involvement with litigation, and younger age were also correlated with a favorable outcome. The only magnetic resonance parameter associated with good outcome was a small ratio of disc hemiarea to remaining canal hemiarea (P = 0.045). For the surgical group, a larger anteroposterior disc length was the most significant independent predictor of a positive outcome (P < 0.0001). Larger ratios of disc area to canal area are also significantly associated with good outcomes (P < 0.0001), as are large disc areas and small remaining canal areas. Large right-left canal widths and small disc widths are also identified as predictors of a favorable outcome. Of the clinical parameters, concurrent medical illness, workers' compensation involvement, and female gender appear to be the most significantly correlated with poor outcome. All fair or poor surgical outcomes were in patients with smaller (< 6 mm) disc herniations.
CONCLUSIONS: Quantitative measurements by magnetic resonance imaging of disc and canal morphology of 188 patients with sciatica indicate a wide range of herniation and canal sizes, with significant differences between men and women. In a cohort of 135 patients followed for more than 2 years, demographic and clinical features appeared to predict outcomes of nonoperative treatment, whereas morphometric features of disc herniation and the spinal canal seen on magnetic resonance imaging were much more powerful predictors of surgical outcomes.