Chi, John H.
Sciatica is an almost universal condition that is fortunately self-limited in most episodes, but can often become a vexing problem leading to a variety of treatments, including surgery. Disc herniation is a common cause of sciatica and resolution of the herniation, either spontaneously or surgically, has always been thought of as the reason for improved symptoms. However, prior studies have demonstrated that disc herniations can be present in asymptomatic patients to an equal degree as in symptomatic ones, and clinical trial data has suggested that despite surgical removal of a disc herniation, clinical symptoms improve on similar scales as with no surgery, albeit at a faster pace with surgery. Barzouhi et al (Magnetic resonance imaging in follow-up assessment of sciatica. NEJM. 2013;368(11):999-1007) report results showing that magnetic resonance imaging (MRI) findings 1 year after treatment of sciatica cannot differentiate patients with good and bad results from their treatment, casting question on the role of disc herniation in causing sciatica and opening the door for other reasons.
Participants in a prior multicenter, randomized trial were used for this study and had to have both 6 to 12 weeks of sciatica and MRI confirmed disc herniation that was concordant. The 2 cohorts included group 1 with early surgery and group 2 with prolonged conservative care. Patients were allowed to cross over as needed. Outcome measures were assessed at baseline and at several points up to 1 year from enrollment using the Roland Disability Questionnaire for Sciatica and Likert self-rating scale. Favorable outcome was then dichotomized as complete or near-complete improvement or not. MRIs were then performed at 1 year and blinded raters would rate the index level using standardized definitions as herniated, bulging or normal, each on 4 point “confidence” scale.
A total of 283 patients were enrolled in this study, 170 receiving surgery and 97 not after cross over (56 patients). In general, patients with surgery had significantly higher rates of disc herniation resolution and improvement in root compression radiographically than patients without surgery (79% vs 40%, P <.001). However, when looking at clinical responders and non-responders, the presence or absence of a disc herniation at 1 year after treatment had no predictive value. Disc herniations were still visible in 35% of patients with favorable outcome and in 33% of those with unfavorable outcomes. Of those patients, 85% of patients with a disc herniation reported favorable outcome, while 83% of those with no disc herniation reported the same. Adjustment for assigned group and received treatment as well as other multivariate analyses did not reveal a significantly different odds ratio for MRI findings and favorable outcome at 1 year. Additionally, the presence of epidural scarring also did not seem to differ in patients with good and bad outcome.
Though the results of this study suggest that clinical improvement can occur with or without a disc herniation being present, it is important to recognize that several limitations exist making interpretation difficult. First, favorable and unfavorable outcomes were dichotomized, which in real life is not necessarily the case. Shades of improvement can be clinically relevant for patients. Second, other anatomical parameters such as neuroforamen height, facet hypertrophy, and lumbar lordosis were not factored in, thus could be confounding the results of surgery on outcome. Disc decompression may, in fact, not be as desirable as decompression with spinal restoration. And third, MRI findings and clinical results were only correlated at one time point only, making it difficult to understand the relationship of MRI and clinical outcomes at different time points, such as at the maximal time of pain relief.
Ultimately, sciatica is most certainly a multifactorial problem with disc herniation causing mechanical compression being only one factor. The pressure of compression and pattern of disc failure likely also affect outcome and deserve further attention as well.
Figure. Statistical ...Image Tools