Spinal cord MRI signal changes at 1 year after cervical decompression surgery is useful for predicting midterm clinical outcome: an observational study using propensity scores

Spine J. 2018 May;18(5):755-761. doi: 10.1016/j.spinee.2017.09.004. Epub 2017 Sep 19.

Abstract

Background context: There is little information on the relationship between magnetic resonance imaging (MRI) T2-weighted high signal change (T2HSC) in the spinal cord and surgical outcome for cervical myelopathy. We therefore examined whether T2HSC regression at 1 year postoperatively reflected a 5-year prognosis after adjustment using propensity scores for potential confounding variables, which have been a disadvantage of earlier observational studies.

Purpose: The objective of this study was to clarify the usefulness of MRI signal changes for the prediction of midterm surgical outcome in patients with cervical myelopathy.

Study design/setting: This is a retrospective cohort study.

Patient sample: We recruited 137 patients with cervical myelopathy who had undergone surgery between 2007 and 2012 at a median age of 69 years (range: 39-87 years).

Outcome measures: The outcome measures were the recovery rates of the Japanese Orthopaedic Association (JOA) scores and the visual analog scale (VAS) scores for complaints at several body regions.

Materials and methods: The subjects were divided according to the spinal MRI results at 1 year post surgery into the MRI regression group (Reg+ group, 37 cases) with fading of T2HSC, or the non-regression group (Reg- group, 100 cases) with either no change or an enlargement of T2HSC. The recovery rates of JOA scores from 1 to 5 years postoperatively along with the 5-year postoperative VAS scores were compared between the groups using t test. Outcome scores were adjusted for age, sex, diagnosis, symptom duration, and preoperative JOA score by the inverse probability weighting method using propensity scores.

Results: The mean recovery rates in the Reg- group were 35.1%, 34.6%, 27.6%, 28.0%, and 30.1% from 1 to 5 years post surgery, respectively, whereas those in the Reg+ group were 52.0%, 52.0%, 51.1%, 49.0%, and 50.1%, respectively. The recovery rates in the Reg+ group were significantly higher at all observation points. At 5 years postoperatively, the VAS score for pain or numbnessin the arms or hands of the patients in the Reg+ group (24.7 mm) was significantly milder than that of the patients in the Reg- group (42.2 mm).

Conclusions: Spinal T2HSC improvement at 1 year postoperatively may predict a favorable recovery until up to 5 years after surgery.

Keywords: Cervical myelopathy; Cervical spondylotic myelopathy; Laminoplasty; Magnetic resonance imaging; Ossification of the posterior longitudinal ligament; Outcome.

Publication types

  • Observational Study

MeSH terms

  • Adult
  • Aged
  • Aged, 80 and over
  • Cervical Vertebrae / diagnostic imaging
  • Cervical Vertebrae / surgery*
  • Decompression, Surgical / adverse effects*
  • Female
  • Humans
  • Magnetic Resonance Imaging*
  • Male
  • Middle Aged
  • Postoperative Complications / diagnostic imaging*
  • Postoperative Complications / epidemiology
  • Spinal Cord Diseases / surgery*