Multiple epidural steroid injections and body mass index linked with occurrence of epidural lipomatosis: a case series
Epidural lipomatosis (EL) is an increase of adipose tissue, normally occurring in the epidural space, sufficient to distort the thecal sac and compress neural elements. There is a lack of knowledge of risk factors, impact on patient’s symptoms, and its possible association with epidural steroid injections.
History, physical examination, patient chart, and MRI were analyzed from 856 outpatients referred for epidural steroid injections. Seventy patients with signs of EL on MRI comprised the study group. Thirty-four randomly selected patients comprised the control group. The severity of EL was determined by the MRI assessment. The impact of EL was determined by the patient’s history and physical examination. Logistic regression was used to correlate the probability of developing EL with BMI and epidural steroid injections.
EL was centered at L5 and S1 segments. The average BMI for patients with EL was significantly greater than that of control group (36.0 ± 0.9 vs. 29.2 ± 0.9, p <0.01). The probability of developing EL with increasing BMI was linear up to the BMI of 35 after which it plateaued. Triglycerides were significantly higher for the EL group as compared to controls (250 ± 30 vs. 186 ± 21 mg/dL p < 0.01). The odds of having EL were 60% after two epidural steroid injections, 90% after three epidural steroid injections and approached 100% with further injections, independent of BMI. Other risk factors considered included alcohol abuse, use of protease inhibitors, levels of stress, hypothyroidism and genetic predisposition. However there were insufficient quantities to determine statistical significance with a degree of confidence. The impact of EL on patient’s symptoms correlated with EL severity with Spearman correlation coefficient of 0.73 at p < 0.01 significance level.
The BMI and triglycerides levels were found to be significantly elevated for the EL group, pointing to an increased risk of EL occurrence in progressively more obese US population. The data also revealed a strong correlation between the number of subsequent epidural steroid injections and EL occurrence calling for caution with the use of corticosteroids.
KeywordsEpidural lipomatosis Epidural steroid injection Obesity Cortisol Sciatica
Epidural lipomatosis (EL) is an increase of adipose tissue, normally occurring in the epidural space, sufficient to distort the thecal sac and even compress neural elements. EL was first reported in 1975 with the use of corticosteroids to prevent rejection of a kidney transplant . A 2005 review of the 104 cases of EL in the literature identified four categories associated with EL: i) exogenous steroid use, ii) obesity, iii) endogenous steroid excess and iv) idiopathic . A 2008 review of the world literature found 111 cases of EL with 56% secondary to corticosteroids, exogenous or endogenous . Thoracic EL was associated with corticosteroid use and lumbar EL with obesity . Tissue surgically removed from patients with EL showed histologically normal, un-encapsulated fat in increased amounts [2, 3, 4, 5, 6, 7]. Although the existing literature clearly points to a link between obesity and steroid administration, there has not been a systematic assessment of risk factors that can lead to EL, particularly the effect of multiple steroid injections for lower back pain. We sought to address this question using an analysis of 856 patients referred to our clinic for lower back pain.
All statistical analysis was performed in SAS 9.3. BMI, triglycerides, cholesterol, and fat from sagittal measurements (total epidural and posterior subcutaneous) were compared with Wilcoxon-Mann–Whitney (WMW) and/or T-test as noted. Statistical significance was defined at p-values of less than 0.01.
Where β0 and β1 are the solved coefficients, Var is the number of ESI given or BMI, and p i is the probability of acquiring EL. The correlation was considered to be significant for variables with a p-value of less than 0.01. Logistic regression was calculated separately for BMI and ESI to ensure independence.
After above mentioned selection from an original pool of 856 patients, there were a total of 104 patients analyzed. This included 70 patients with EL and 34 without EL. There were 43 men and 61 women in the study. A binomial test of these two groups resulted in a p < 0.05, confirming that the numbers of men and women in both groups were statistically equal. The average age of the patients was 61.8 ± 2.8 years. Those with EL had an average age of 61.7 and those without, 61.8.
Identification of EL
Linear image measurements of epidural and of subcutaneous fat
Increased body mass index and triglycerides are associated with epidural lipomatosis (EL)
Those with EL had an average triglycerides value of 250 ± 30 and those without EL had an average of 186 ± 21 mg/dL. To account for the fact that triglycerides values were positively skewed in both groups, a WMW test was used. Triglycerides in the EL group were significantly elevated compared to the control group with p < 0.01. Triglycerides do not independently determine the probability of developing EL with a p > 0.01 in a logistic regression model.
Severity of lipomatosis on patient’s symptoms
Categories of Chief Complaint
EL group (n = 70)
Neurogenic Claudication (n=28)
Radicular pain, sensory and/or motor deficit
Severity of EL and its relationship to gradation of impact
Epidural steroid injections, stress, & alcohol consumption
The number of ESI deliveries was significant in an independent logistic regression model with a p < 0.01. Absence of ESI deliveries or one ESI delivery did not increase the patient’s odds of developing EL. After two ESI the odds of developing EL was 66%. After three ESI, the odds were 98%. Four or more ESI increased the odds approaching 100%.
The average number of ESI delivered in the patients with no EL was 1.0 ± 0.0, and there were no patients who received more than one. The average number of ESI delivered was 1.8 ± 1.5 to the EL group. There were two patients in the EL group who did not receive any ESI. Their BMI were 40 and 42.
It was noted in the charts of ten patients in the EL group who self-reported stress and eight who self-reported alcohol consumption. No patients in the control group reported stress or alcohol consumption.
Definition, location, severity, impact
The definition of EL presently available, “an excessive accumulation of normally occurring fat in the epidural space” [2, 3, 4, 5, 6, 7, 9, 10, 11], does not provide a definitive boundary between the normal and excessive amount of fat in the epidural space thus we added visible deformation of the thecal sac/neural elements to provide a more definitive end point.
Severity was correlated with impact. Pinkhardt et al., found a “very insecure association (of EL) with clinical symptoms” . They made measurements in L5/S1 axial images according to Borre  retrospectively, excluding patients with deformities whereas our study included all sagittal and axial T1 MR images. The MRI findings must be put in the context of the history and physical exam.The mean BMI for the control group was under 30, but approximately 34 for EL patients indicating that there is step response beyond the threshold of obesity (Figure 5B). With a BMI of 28, the probability of EL is only 50% which hints there are other contributing factors. Other causative factors, including, but not limited to ESI and elevated triglycerides, will significantly contribute to the probability of developing EL up to a BMI of 35. The probability of developing EL is linear with increasing BMI until a value of 35, at which point the probability curve begins to plateau (Figure 6A). Once BMI is greater than 35, the other causative factors have less of a contribution toward the probability of developing EL, but may contribute to the grade of EL severity. The grade of EL severity is not correlated with BMI (Figure 5B).
The causes of epidural lipomatosis
The probability of developing EL is linear with increasing BMI between 28 and 35. BMI of 28 acts as a threshold for developing EL and at a BMI 35 its influence begins to wane. There are other factors involved which will be discussed later. Obesity, BMI over 30, is an imputed cause of EL [2, 3, 4, 10, 12, 13] which provides a milieu for the enzyme, 11β-HSD-1 to convert corticosteroids to cortisol , Diet induced obesity leads to an increase in 11β-HSD-1 which increases cortisol by the local conversion of corticosteroids to the biologically active cortisol . Corticosteroids also, an imputed cause of EL [2, 3, 4, 6, 15, 16, 17, 18, 19], fit into the schema by providing the substrate for the formation of cortisol.
Study of the physiology of cellular liporegulation in rodents sheds some light on the role of Leptin in human metabolism of fat . Leptin is required for normal liporegulation in tissues just as insulin is required for normal glucoregulation. In over-nutrition, leptin allows extra fat to be deposited in body fat without metabolic injury to non-adipose tissues. An increase in cortisol leads to leptin resistance; exacerbated by glucocorticoid administration  and ameliorated by adrenalectomy . If leptin is overwhelmed, then deposition of fat may take place in the liver, heart, muscles and pancreas causing lipotoxicity or lipoapoptosis . The liver has a margin of safety by exporting triglycerides in the form of very low density lipoproteins (VLDL); the muscle can metabolize triglycerides as a result of exercise, but the pancreas has neither of these options .
A phase IIB study in patients with Type 2 diabetes given an antagonist to the enzyme 11β-HSD-1 responsible for conversion of corticosteroids to cortisol showed increased insulin sensitivity and improvement in: glycemic control, total cholesterol, LDL-cholesterol and triglycerides . Studies are needed to show if 11β-HSD-1 and cortisol lead to EL.
Hypertriglyceridemia may develop due to overeating and/or under-exercising [23, 24]. Obesogens can alter regulation of energy balance to favor weight gain and obesity . Also, an increase in adipose tissue cortisol can lead to lipolysis with increased triglycerides .
EL and ESI
Reports have related the development of EL to the administration of epidural steroid injections [27, 28, 29, 30]. Four patients were refused ESI because the EL was severe and structural defects were minimal. Although the probability of developing EL is over 90% after four ESI, the severity and thus symptomatic impact of EL may not be significant. The incidence of EL was not assessed in this study. The corticosteroid from the ESI provides the substrate for the formation of cortisol.
The response to stress, as formulated by Selye , causes the release of corticotrophin by the pituitary; cortisol, epinephrine and other hormones by the adrenal gland. The cortisol, thus produced, augments the cortisol from the adipose tissue, ultimately promoting the metabolic syndrome [14, 15, 20, 32, 33, 34]. The case-controlled study by Brunner et al. provides evidence that chronic stress may be a cause of metabolic syndrome . Only our patients that developed EL reported stress.
Studies of alcohol abuse on metabolic syndrome are cogent since EL and metabolic syndrome share many components. Some studies have shown a beneficial effect of alcohol on the development of metabolic syndrome. However Baik found an increased risk of metabolic syndrome was associated with “obese, heavy liquor drinking persons” . Alcohol feeding of patients on a metabolic ward increased plasma triglycerides, the obese responding with a 45% increase in the production of VLDL-TG . Thus, the caveat: mild or moderate alcohol consumption may be protective, but only in persons who are not obese. Only patients in the EL group reported alcohol consumption.
The development of EL is multi-factorial, the risk factors being intertwined and additive. Obesity is the principal risk factor associated with lumbar epidural lipomatosis. Lumbar EL may worsen after the first ESI injection. The symptomatic impact of EL is correlated with its severity. EL may be the principal cause of sciatica and secondary cause of neurogenic claudication. The present trend in obesity portends a marked increase in epidural lipomatosis . Weight loss programs have reversed EL and its symptoms, including neurogenic claudication and cauda equina compression [39, 40, 41, 42, 43]. The density of cases is sufficient to formulate controlled studies.
The authors would like to thank Christina Jaimes, M.S., R.D. for introducing the role of leptin; Stephen Lipson, M.D. for identifying epidural lipomatosis on an MR image; Stephen Lipson, M.D. and Jerry Knirk, M.D. for the many hours they spent discussing patients and reviewing MR images; Narine Sarvazyan, Ph.D. for editorial assistance; and Christopher Absi for graphical aid. The authors did not receive financial aid or funding.
- 5.Flores A, VKH S, Dickman CA: Idiopathic spinal epidural lipomatosis: report of two cases and review of literature. BNI Quarterly. 1995, 11: 22-25.Google Scholar
- 22.Rosenstock J, Banarer S, Fonseca VA, Inzucchi SE, Sun W, Yao W, Hollis G, Flores R, Levy R, Williams WV, Seckl JR, Huber R: The 11-beta-hydroxysteroid dehydrogenase type 1 inhibitor INCB13739 improves hyperglycemia in patients with type 2 diabetes inadequately controlled by metformin monotherapy. Diabetes Care. 2010, 33 (7): 1516-1522.CrossRefPubMedPubMedCentralGoogle Scholar
- 26.Tomlinson JW, Sherlock M, Hughes B, Hughes SV, Kilvington F, Bartlett W, Courtney R, Rejto P, Carley W, Stewart PM: Inhibition of 11beta-hydroxysteroid dehydrogenase type 1 activity in vivo limits glucocorticoid exposure to human adipose tissue and decreases lipolysis. J Clin Endocrinol Metab. 2007, 92 (3): 857-864.CrossRefPubMedGoogle Scholar
- 35.Brunner EJ, Hemingway H, Walker BR, Page M, Clarke P, Juneja M, Shipley MJ, Kumari M, Andrew R, Seckl JR, Papadopoulos A, Checkley S, Rumley A, Lowe GD, Stansfeld SA, Marmot MG: Adrenocortical, autonomic, and inflammatory causes of the metabolic syndrome: nested case–control study. Circulation. 2002, 106 (21): 2659-2665.CrossRefPubMedGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2253/14/70/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.