Spinal epidural lipomatosis (SEL) is a condition characterized by the over-deposition of unencapsulated epidural fat in the spinal canal, usually in the lumbosacral tract. The non-invasive diagnosis of SEL can be made with Magnetic Resonance Imaging (MRI) or Computed Tomography (CT) only.

SEL, a pathologic condition, may lead to the narrowing of the spinal canal and compression of surrounding neural structures. SEL can be asymptomatic in mild–moderate disease and usually becomes symptomatic in moderate–severe cases. The clinical presentation typically involves progressive lower back pain radiating to the lower limbs, spinal claudication, radiculopathy, myelopathy, and cauda equine syndrome ^[1][2][1,2].

In most cases, symptoms develop over months to years but may develop acutely in some rare cases ^[3][4][5][6][3,4,5,6]. For unclear reasons, males are affected more commonly than females (M:F = 3:1) [7]. Moreover, more than 75% of all reported patients are obese [8]. Exceptionally, children can suffer from neurological symptoms associated with SEL, and this is often accompanied with a long history of steroid treatment ^{[9][10][11][12]}[9,10,11,12].

A previous large population study based on MRI reports including 28,902 subjects reported a prevalence of SEL of approximately 2.5%. In this series, the main factors associated with the disease included being male, obese, and having a history of systemic corticosteroid use [13]. Recent work has focused attention on the ratio of radiologists’ awareness and the detection of this condition, highlighting a fairly significant underestimation of SEL in radiologists’ reports and suggesting that the actual prevalence of SEL has been underestimated so far [14]. The prevalence of SEL in this latter research study, which includes 450 subjects, was notably higher (16.7%) [14]. Another previous large series focused on lumbar MRIs and including 2528 patients found quite similar results (SEL prevalence = 20.8%) [15].

Historically, steroid excess (exogenous or endogenous) was considered the most common etiological factor associated with SEL ^[16][17][18][16,17,18]. Indeed, the first case of steroid-induced lipomatosis was reported by Lee et al. in 1975 [19].

Because obesity in adults is usually accompanied by common non-communicable diseases, such as dyslipidemia, Type 2 diabetes, hypertension, and arteriosclerosis, the high rate of obesity in patients with idiopathic SEL suggests that non-communicable diseases and SEL are causally related. The primary line of treatment is to reduce or eliminate the etiologic factors. When a clear etiologic factor cannot be ruled out, weight loss has been proposed as a conservative treatment. However, surgical decompression is the treatment of choice, especially for acute or severe cases ^[20][21][20,21]. Moreover, in patients with worsening or persistent neurologic symptoms, surgery should be suggested ^[1][21][22][1,21,22].

When SEL is a concause of neurological symptoms, together with other conditions (i.e., thickened yellow ligament and disc herniation), a negative prognostic impact on the surgical outcome of posterior decompression surgery has been reported [23], giving paramount importance to its correct reporting by radiologists.

2. SEL’s Etiology

In the following subsections, the recognized causes of SEL are reported and discussed, with reference to the available literature.

2.1. Excessive Amount of Corticosteroids

SEL has been frequently associated with excessive use of corticosteroids, which can have an exogenous source-like steroid injection [24] or long-term corticosteroid therapy ^{[25][26][27][28]}[25,26,27,28]. It is also associated with an endogenous source-like corticotropin syndrome from ACTH-secreting extra pituitary tumors (endogenous Cushing’s syndrome) [17] or from an ACTH-secreting pituitary gland adenoma (Cushing’s disease) ^[29][30][29,30]. Among the series that have been reported, approximately 75% of cases have been associated with exogenous steroid use [31], with great variability from case to case in terms of the doses and duration of treatments associated with this condition [32]. Mostly, moderate-to-high steroid dosages were received by patients for years before the development of symptoms. Although relatively rare, some cases of patients who developed SEL with a long-term and low-dosage steroid treatment have been described in the literature [33]. When the treatment with steroids is both long-term and high-dosage, there is a risk of developing severe SEL, even in children, as evident in an 8-year-old patient in treatment for Crohn’s disease who suffered from myelopathy due to the SEL-induced compression of the spinal cord [34]. There have been reports of SEL associated with epidural steroid injections and even inhaled steroids, but the long-term use of oral steroids is by far the most common association ^[35][36][37][35,36,37]. Koch et al. described a case of a patient with bronchial carcinoid associated with ectopic corticotropin syndrome. The patient complained of severe back pain and leg weakness among other clinical issues. The MRI showed multiple levels of excessive epidural fat, which reverted to normal after two months of treatment with the inhibitors of steroidogenesis (metyrapone and ketoconazole). At the same time, there was a complete resolution of neurological symptoms [38]. Bathia et al. reported a similar case of ACTH-secreting bronchial carcinoid in a male farmer with Type 2 diabetes mellitus. The MRI showed multiple levels of SEL of the thoracic spine, without significant disc herniation. Since the patient’s pain was improving, despite treatment with ketoconazole, he was successfully treated with decompressive hemilaminectomy and the removal of the epidural adipose tissue (no stabilization was performed due to massive osteoporosis) [39]. A 17 yo with an ACTH-secreting adenoma of the pituitary gland, which developed after trans-sphenoidal adenectomy, experienced a sudden spastic paraparesis with a sensory deficit to the level of T5. An MRI scan showed a severe SEL of the thoracic spine, which was treated conservatively with success. This case shows that it is important to rule out SEL from a myopathy caused by hypercortisolism, starting with a proper collection of the patient’s clinical history [40]. Historically, SEL was associated with exogenous steroid use as the first cause of disease, even if recent reports suggest that obesity is becoming the major risk factor. A study by Fogel et al. reports that 55.8% of SEL cases caused by exogenous steroids affect the thoracic spine, compared to 32.7% that only involve the lumbosacral region and 11.5% that affect both [2]. This differs from the endogenous steroid disease-related SEL, which affects the thoracic and lumbosacral areas relatively the same. In addition, most obesity-related SEL (69.6%) and idiopathic-related SEL (50%) result in lumbosacral involvement. On the other hand, a recent systematic review with meta-analyses revealed that obesity is the current leading cause of SEL [41].

2.2. SEL and Obesity

Obesity has always been considered one of the most common causes of SEL [2]. Nonetheless, in recent years, obesity is believed to be the most common cause of SEL. A recent meta-analysis confirmed that obesity should be considered the main risk factor for SEL development (52% of cases) [41]. Consistent with previous studies, there was a positive correlation between the degree of obesity and the severity of SEL ^{[42][43][44][45]}[42,43,44,45]. In a large series by Theyskens et al., which included MRI scans of 28,902 patients, a BMI > 30 was independently associated with SEL in a multivariate analysis [13]. Yildirim et al. reported the strong correlation of an increased BMI with SEL in a retrospective case-matched control study including 199 patients. In this research study, the authors indicated that patients with SEL had a significantly higher median BMI than control subjects (36.7 vs. 29.4 kg/m², p < 0.001) [46].

2.3. Non-Alcoholic Fatty Liver Disease

Non-alcoholic fatty liver disease (NAFLD) is nowadays the most common chronic liver disease in developed countries. The clinical presentation of NAFLD ranges from asymptomatic (only with elevated liver enzyme levels) to cirrhosis, with possible complications of liver failure and hepatocellular carcinoma. There is sufficient evidence supporting an association between NAFLD and metabolic syndrome [47]; interestingly, recent studies have also found an association between SEL and NAFLD [48].

2.4. Miscellaneous

Other possible risk factors for SEL development have been suggested in recent years, but their role remains controversial and needs further investigation. The principal difficulty in defining their real role in the development of SEL is the contemporary presence of well-known risk factors such as obesity and exogenous steroid treatment. More studies are needed to assess if they can be addressed as risk factors or only associated factors. SEL has been described as a possible manifestation of highly active antiretroviral therapy (HAART)-associated lipodystrophy in HIV-positive patients ^{[49][50][51][52]}[49,50,51,52]. Androgen deprivation therapy for prostate cancer has been suggested as a possible risk factor for SEL in several case reports ^{[53][54][55][56]}[53,54,55,56]. Three case reports have described an association between SEL and scoliosis ^[57][58][59][57,58,59]. The rapid progression of SEL has been reported after spinal surgery ^[4][5][60][4,5,60]. In the study by Okada et al., SEL was significantly associated with diffuse idiopathic skeletal hyperostosis (relative risk (RR), 2.6; 95% confidential interval (CI), 1.3–5.1; p < 0.01) [61]. Finally, some authors have suggested a possible association between SEL and spinal kyphotic deformities, such as congenital, Scheuermann, and tuberculotic kyphosis ^[62][63][62,63]; Paget disease ^[64][65][64,65]; hypothyroidism [66]; chronic alcoholism [67]; and Type 1 diabetes mellitus [68].

2.5. Idiopathic

Idiopathic cases of SEL have been reported in the literature since 1982 [69]. Haddad et al. first hypothesized that idiopathic SEL was a byproduct of obesity, with the gradual overgrowth of epidural fat resulting in the compression of the spinal cord and nerves [70]. Nevertheless, cases of SEL in underweight patients suggest another kind of pathogenetic mechanism for idiopathic SEL [31]. The recognition of idiopathic SEL is of paramount importance, and in absence of risk factors, the disease cannot be clinically suspected. Patients may remain undiagnosed for a long period. As a result, the longer duration of nerve root constriction may cause a relatively lower recovery rate [71]. There is disagreement in the literature regarding the definition of idiopathic SEL. Indeed, some authors refer to idiopathic diseases to indicate the SEL of an unknown cause, while others use this term to describe SEL associated with obesity or other causes that nowadays are clearly risk factors for disease development. Indeed, the term “idiopathic SEL” should be used only in patients without any other recognized risk factor (e.g., metabolic syndrome, obesity, and increased corticosteroids). In the literature review and meta-analysis by Fogel et al., idiopathic SEL has been reported to account for 17% of the cases. This group includes those patients who did not take exogenous steroids, were not obese, and did not have an underlying endogenous steroid hormonal disease ^[2][72][73][2,72,73]. Idiopathic SEL is also reported in the pediatric age, with few reports in the literature at this regard ^[74][75][74,75]. Importantly, the grading of SEL (including a cut-off for diagnosis) was tested only in adult patients, while pediatric-age patients’ range of normality in terms of amount of epidural fat is still missing [15]. Indeed, the diagnosis at this age is purely qualitative depending on the MRI or CT evaluation. Racial differences may become more prevalent when discussing the pathogenesis of SEL. Yoo et al. reported an incidence of up to 68.8% of idiopathic SEL in the Korean population [76], while Fogel et al. reported 17% for the same SEL category in Western countries [2]. This difference may suggest a role of genetic variation in the pathogenesis of SEL.

2.6. Epidemiological Analysis of SEL’s Etiologies

Historically, SEL was associated with exogenous steroid use as the first cause of disease [2], even if recent reports suggest that obesity is becoming the major risk factor [77]. A study by Fogel et al. reported that 55.8% of SEL cases caused by exogenous steroids affect the thoracic spine, compared to 32.7% that only involve the lumbosacral region and 11.5% that affect both [2]. This differs from the endogenous steroid disease-related SEL, which affects the thoracic and lumbosacral areas relatively the same. In addition, the majority of obesity-related SEL (69.6%) and idiopathic-related SEL (50%) result in lumbosacral involvement. On the other hand, recently, obesity has been recognized as the main cause of SEL. Indeed, a recent systematic review and meta-analysis (time period considered 1990–2020) revealed that obesity is currently the main cause of SEL, accounting for more than a half of patients [77].