How Prevalent is Spinal Epidural Lipomatosis on Magnetic Resonance Imaging?

Journal Published on February 8, 2018

Originally published on The Spine Journal

Key findings

2.5% was the overall spinal epidural lipomatosis (SEL) prevalence rate according to a Mass General study
Incidental SEL prevalence was 0.6%, while SEL with spine-related symptoms and symptomatic SEL were 1.8% and 0.1%, respectively
Higher modified Charlson comorbidity index was identified as a new risk factor for SEL
Cushing’s syndrome or disease was the only predictor of symptomatic SEL

The prevalence of spinal epidural lipomatosis (SEL) is not well understood. To provide clarity, a Massachusetts General Hospital research team led the first large-scale study, published in The Spine Journal, consisting of 28,902 patients, to determine the prevalence of SEL out of those receiving spine-dedicated magnetic resonance imaging (MRI). They set out to identify the overall rate and rates for various sub-groups, including incidental SEL (SEL without spine-related symptoms), SEL with spine-related symptoms, and symptomatic SEL. Factors associated with overall SEL and sub-groups, and differences between SEL subgroups, were evaluated.

2.5%

Overall SEL prevalence found in the study

Methodology

The team retrospectively reviewed radiology reports of the target patient population from January 2004 to January 2015 at two tertiary care centers. Records were scanned for “spine,” and synonyms, abbreviations and truncations of “epidural lipomatosis,” including the terms “epidural fat,” “adipose tissue” and “EL." Included were 7,955 cervical, 1,404 thoracic and 12,621 lumbosacral MRI scans, and 6,922 MR images of multiple segments. The researchers created a reference group by selecting the chronologically first MRI scan without SEL.

Spine-related symptoms included radiculopathy or myelopathy (neck or back pain radiating to the arm or leg), abnormal reflexes, and numbness, sciatica or neurogenic claudication, including cauda equine syndrome. Given that SEL is non-specific, non-radiating back pain was not considered a symptom.

Analysis of variance (ANOVA) was used to find continuous variables among SEL sub groups. These included:

Age
Gender
Race
Body mass index (BMI)
Comorbidity status
Presence of Cushing’s syndrome or disease
Hypothyroidism
Hyperprolactinemia
Systemic, epidural or injected corticosteroid use

Patients had a mean age of 55 years, with slightly more women (51%) and far more Caucasians (84%). Only 12% had diagnosed hypothyroidism, and 0.4% and 0.5% had Cushing’s syndrome or hyperprolactinemia, respectively. Prior use of systemic corticosteroids (1.1%) or epidural corticosteroid injections (5.3%) was not common.

Results

After calculating the prevalence rate as a percentage of the total number of patients, researchers identified overall SEL to be 2.5%. Prevalence of incidental SEL was 0.6%, while SEL with spine-related symptoms and symptomatic SEL were 1.8% and 0.1%, respectively.

Out of total patient records, 72% showed spine-related symptoms, with the most common being radiculopathy or myelopathy (67%). Five percent had SEL-specific symptoms, and 23% had no symptoms at all.

After multivariate analysis, the following were found to be independently associated with SEL:

Being male
Classified as Black or African American race
Older age
BMI >30
Systemic and epidural corticosteroid use

Newly noted was that higher modified Charlson comorbidity index is an independent predictor of SEL.

Cushing’s syndrome or disease was the only predictor of symptomatic SEL.

For all patients with SEL, 57% had it in multiple spinal regions, with 35% having it in the lumbar region, the most frequent area affected. Further, 82% evidenced further spinal pathology.

Overall, the team’s results revealed characteristics that should flag suspicion and might warrant diagnostic investigation into possible SEL.