Secondary Logo

Journal Logo

Clinical Features and Neuroimaging Findings in Patients with Cerebral Fat Embolism

Han, Yang-Tong1; Tang, Jing2; Gao, Zhi-Qiang3; Hu, Hong-Tao1,

doi: 10.4103/0366-6999.178963
Clinical Practice

1Department of Neurology, Beijing Jishuitan Hospital, Beijing 100035, China

2Department of Orthopedic Osteology, Beijing Jishuitan Hospital, Beijing 100035, China

3Department of Trauma Osteology, Beijing Jishuitan Hospital, Beijing 100035, China

Address for correspondence: Dr. Hong-Tao Hu, Department of Neurology, Beijing Jishuitan Hospital, Beijing 100035, China E-Mail:

Received November 20, 2015

Back to Top | Article Outline


Cerebral fat embolism (CFE) is a rare but potentially lethal disease that is usually complicated by long bone fracture, which can develop over 12–72 h. Since the main form of treatment for CFE is supportive, early diagnosis, prevention, and appropriate treatment can reduce mortality and the development of complications. However, due to symptom heterogeneity and imprecise diagnostic methods, CFE may be masked by other clinical conditions that make early detection challenging. Given its low incidence, most clinical studies of CFE are single cases or small case series. To characterize the features of CFE, clinical manifestations, laboratory examinations, and neuroimaging data were evaluated in a large sample of patients with CFE.

Back to Top | Article Outline


The current study was a retrospective review of CFE at the Department of Neurology, Jishuitan Hospital (Beijing, China) from 2003 to 2014. The study was approved by the Ethics Committee of Jishuitan Hospital and conducted according to the principles of the Declaration of Helsinki. A diagnosis of CFE was made based on Gurd's criteria in conjunction with neuroimaging results and initial presentation of neurological symptoms.[1] Patients whose initial neurological symptoms were subsequent to respiratory manifestations, as well as those who were diagnosed with fat embolism syndrome but showed no neurological symptoms, were excluded from the study. On the other hand, patients whose initial symptoms were neurological manifestations with or without subsequent respiratory manifestations were included. Patients with infections, tumors, subdural hematomata, acute cerebrovascular diseases, visceral injuries, or inflammatory diseases were also excluded from the study. Demographic data, clinical manifestations, treatment, prognosis, laboratory examinations, and neuroimaging results were all evaluated.

Back to Top | Article Outline


The study included 10 male and 11 female patients ranging in age from 24 to 71 years (mean age: 34.9 ± 10.6 years). Detailed demographic data, clinical manifestations, and results from laboratory examinations are presented in Table 1. Patent foramen ovale was detected in three of 15 patients who underwent echocardiographic examination. Pulmonary arteriovenous malformations (AVMs) were absent in all nine patients who received chest computed tomography (CT) examination. Altogether, 18 patients recovered without developing neurological sequelae within 1 month of injury and three remained in a coma at the 3-month follow-up.

Table 1

Table 1

All patients underwent cerebral CT examination and nine were examined by magnetic resonance imaging (MRI). Five of the nine patients had abnormal multiple scattered long T1 and T2 signals in the bilateral corona radiata and centrum semiovale and four had abnormal multiple stippled long T1 and T2 signals in the bilateral basal ganglia and centrum semiovale. The fluid-attenuated inversion recovery sequence MRI showed hyperintense lesions in nine patients. The diffusion-weighted image (DWI) showed multiple scattered bright spots on a dark background (i.e., a starfield pattern) in four patients [Figure 1].

Figure 1

Figure 1

Back to Top | Article Outline


The pathogenesis of CFE is still controversial and involves a variety of factors, with no single theory explaining all of the pathophysiological features of CFE. For instance, the infloating theory suggests that fat globules are physically forced into the venous system, resulting in multisystem dysfunction due to the physical obstruction of vessels. On the other hand, the free fatty acid theory proposes that excessive mobilization of free fatty acids leads to severe vasculitis that is secondary to its histotoxic effects.[2] CFE can be present without pulmonary effects; potentially due to the fact that smaller globules may traverse the pulmonary microvasculature and reach the systemic circulation; alternatively, CFE may be caused by fat globules via the patent foramen ovale.[3] Three patients went into a coma shortly after the fracture and remained comatose at the 3-month follow-up. Patent foramen ovale was detected in two of the three coma patients. This suggests that CFE severity and outcome may be linked to the time interval from injury to the development of CFE, i.e., a shorter interval results in more severe CFE and a worse prognosis. A short onset time after injury may lead directly to CFE from fat globules via the patent foramen ovale or pulmonary AVM. In contrast, a long onset time could support the free fatty acid theory of CFE.

Of the 21 patients, 11 had multiple (≥3) long bone fractures, suggesting that multiple (≥3) long bone fractures may constitute a risk factor for CFE. These patients may benefit from early fixation and operation. We also found that three patients had concurrent bilateral hip replacement arthroplasty, suggesting that simultaneous bilateral replacement arthroplasty might be a risk factor for CFE. Therefore, a longer time interval between the two operations of bilateral replacement arthroplasty might decrease the occurrence of CFE.

CFE typically presents with drowsiness, somnolence, and delirium, but seldom with focal neurological deficits, distinguishing it from cerebrovascular diseases. However, some patients in the present study showed focal neurological deficits (33% and 23% with ataxia and hemiparesis, respectively) that lasted for <1 week. The higher frequency of focal neurological deficits observed in this study as compared to previous studies might be due to several factors. The deficits were generally brief and mild, and may not have been obvious due to the presence of preexisting conditions that made it difficult for multiple fracture patients to perform specific tests of neurological function. The heterogeneity in disease severity among subjects and the small sample size in some studies may further contribute to the difference.

In this study, platelet count decreased in 12 patients (12/21, 57%) while urinary fat drops were detected in 9 patients (9/21, 43%). As CFE improved, fat globules disappeared from the urine and the platelet count returned to the normal range. The petechial rash is typically detected around the neck and in the axillary area and upper body, appearing within 24–36 h and spontaneously resolving within 1 week. It is thought to result from the occlusion of dermal capillaries and increased capillary fragility. The specific distribution of petechial rash is related to floating fat globules in the aortic arch. The presence of a petechial rash and a progressive decrease in blood platelet counts are valuable early indicators of CFE, but it should be noted that these symptoms are not specific to CFE.

CT lacks sensitivity and is usually unable to detect CFE. However, it may be a valuable method for excluding alternative clinical conditions. In contrast, MRI can reveal characteristic cerebral lesions in the centrum semiovale, subcortical white matter, and ganglionic regions in acute CFE. Some studies have suggested that the starfield pattern on DWIs can serve as an early and more sensitive indicator of CFE in the context of long bone fractures. The rationale for this is based on the fact that abnormalities in signal intensity on T2-weighted image reflect vasogenic edema, which develops at a later stage, whereas the DWI reveals cytotoxic edema, which develops immediately. Recent reports also indicate that T2* sequences can serve as a pathognomonic feature of CFE, due to their sensitivity to blood products, changes in iron content, and the unique distribution in white matter.[34]

Back to Top | Article Outline

Financial support and sponsorship


Back to Top | Article Outline

Conflicts of interest

There are no conflicts of interest.

Back to Top | Article Outline


1. Mijalski C, Lovett A, Mahajan R, Sundararajan S, Silverman S, Feske S. Cerebral fat embolism: A case of rapid-onset coma Stroke. 2015;46:e251–3 doi: 10.1161/STROKEAHA.115.011440
2. Takada M, Chiba S, Nagai T, Takeshita H, Kanno S, Ikawa T, et al Inflammatory responses to neutral fat and fatty acids in multiple organs in a rat model of fat embolism syndrome Forensic Sci Int. 2015;254:126–32 doi: 10.1016/j.forsciint.2015.07.011
3. Aman J, van Koppenhagen L, Snoek AM, van der Hoeven JG, van der Lely AJ. Cerebral fat embolism after bone fractures Lancet. 2015;386:e16 doi: 10.1016/S0140-6736(15)60064-2
4. Kuo KH, Pan YJ, Lai YJ, Cheung WK, Chang FC, Jarosz J. Dynamic MR imaging patterns of cerebral fat embolism: A systematic review with illustrative cases AJNR Am J Neuroradiol. 2014;35:1052–7 doi: 10.3174/ajnr.A3605

Edited by: Li-Min Chen


Cerebral Fat Embolism; Clinical Features; Neuroimaging Findings

© 2016 Chinese Medical Association