Patient demographics and MRI-based measurements predict redundant nerve roots in lumbar spinal stenosis: a retrospective database cohort comparison

Background

Up to 40% of patients diagnosed with lumbar spinal stenosis (LSS) show evidence of redundant nerve roots (RNR) of the cauda equina on their magnetic resonance images (MRI). The etiology of RNR is still unclear. Preoperative evidence of RNR is associated with a worse postsurgical outcome. Consequently, potential predictors of RNR could have a prognostic value. The aim was to test whether patient demographics and MRI-based measurements can predict RNR in LSS patients.

Methods

In a retrospective database-based cohort study the preoperative data of 300 patients, 150 with (RNR+) and 150 without (RNR-) evidence of RNR on their MRI were analyzed. Three independent researchers performed the MRI reads. Potential predictors were age, gender, body height (BH), length of lumbar spine (LLS), segmental length of lumbar spine (SLLS), lumbar spine alignment deviation (LSAD), relative LLS (rLLS), relative SLLS (rSLLS), number of stenotic levels (LSS-level), and grade of LSS severity (LLS-grade, increasing from A to D). Binomial logistic regression models were performed.

Results

RNR+ patients were 2.6 years older (p = 0.01). Weak RNR+ predictors were a two-years age increase (OR 1.06; p = 0.02), 3 cm BH decrease (OR 1.09; p = 0.01) and a 5 mm SLLS decrease (OR 1.34; p < 0.001). Strong RNR+ predictors were a 1% rLLS decrease (OR 2.17; p < 0.001), LSS-level ≥ 2 (OR 2.59; p = 0.001), LLS-grade C (OR 5.86; p = 0.02) and LLS-grade D (OR 18.4; p < 0.001). The mean rSLLS of RNR+ patients was 0.6% shorter (p < 0.001; 95% C.I. 0.4 to 0.8) indicating a disproportionate shorter lumbar spine.

Conclusions

We identified LSS severity grade and LSS levels as the strongest predictors of RNR. In addition to previous studies, we conclude that a shortened lumbar spine by degeneration is involved in the development of RNR.

Post-surgical functional recovery, lumbar lordosis, and range of motion associated with MR-detectable redundant nerve roots in lumbar spinal stenosis

OBJECTIVE

T1- and T2-weighted magnetic resonance images (MRI) can reveal lumbar redundant nerve roots (RNRs), a result of chronic compression and nerve elongation associated with pathogenesis of cauda equina claudication (CEC) in degenerative lumbar canal stenosis (DLCS). The study investigated effects of lumbar lordosis angle and range of motion on functional recovery in lumbar stenosis patents with and without RNRs.

METHODS

A retrospective study was conducted of 93 lumbar spinal stenosis patients who underwent decompressive surgery. Eligible records were assessed by 3 independent blinded radiologists for presence or absence of RNRs on sagittal T2-weighted MR (RNR and non-RNR groups), pre- and post-operative JOA score, lumbar lordosis angle, and range of motion.

RESULTS

Of 93 total patients, the RNR group (n=37, 21/37 female) and non-RNR group (n=56; 31/56 female) had similar preoperative conditions (JOA score) and were not significantly different in age (mean 64.19 ± 8.25 vs. 62.8 ± 9.41 years), symptom duration (30.92 ± 22.43 vs. 28.64 ± 17.40 months), or follow-up periods (17.35 ± 4.02 vs. 17.75 ± 4.29 mo) (all p>0.4). The non-RNR group exhibited significantly better final JOA score (p=0.015) and recovery rate (p=0.002). RNR group patients exhibited larger lumbar lordosis angles in the neutral position (p=0.009) and extension (p=0.021) and larger range of motion (p=0.008).

CONCLUSIONS

Poorer surgical outcomes in patients with RNRs indicated that elevated lumbar lordosis angle and range of motion increased risks of RNR formation, which in turn may cause poorer post-surgical recovery, this information is possibly useful in prognostic assessment of lumbar stenosis complicated by RNRs.