Multimodal imaging tools for diagnosis of fat embolism.

It is important to consider several differential diagnoses in a patient presenting with altered sensorium following surgery. Fat embolism syndrome (FES) is a serious condition that needs to be excluded. Although criteria for diagnosis of FES are available, all patients may not satisfy them. We discuss a patient who presented with an incomplete triad of the FES, where the diagnosis was supported by transcranial doppler monitoring of microembolic signals and magnetic resonance imaging.

INTRODUCTION

Altered sensorium in the postoperative period poses diagnostic and therapeutic challenges to the treating physician. Fat embolism syndrome (FES) is an important possibility, especially in the setting of orthopedic surgeries. FES is a serious manifestation of fat embolism that involves petechial rash, deteriorating mental status, and progressive respiratory insufficiency. In the perioperative period, an incidence of fat embolism between 3.5% and 5% is noted with the strategy of early fixation of fractures.[1] We discuss the utility of transcranial doppler and MR imaging in the diagnosis of FES in our patient.

CASE REPORT

Mrs. R, an 84 year-old-lady, with no significant past medical history, presented to the emergency department with a fracture of the left femoral neck following a fall. Hemiarthroplasty was performed the following day. Post-procedure, the patient failed to regain her sensorium for more than 10 h. Examination revealed a febrile patient with a pulse rate of 112/min and a blood pressure of 110/80. General physical exam and cardiovascular exams were normal; she was intermittently tachypneic with a respiratory rate of 40/min and clear lungs. She was stuporous with minimal wincing to pain, but no verbal or motor response, deep tendon reflexes were sluggish and plantars were extensor. Fundoscopic examination was normal. The laboratory parameters are shown in Table 1. The renal, liver function tests, and the electrolytes were unremarkable. Arterial blood gas analysis did not show hypoxemia, paO2 was 102 mm Hg; and pulse oximetry showed an O2 saturation of 99%. In view of deteriorating sensorium, she was intubated and ventilated. CT scan of the brain was unremarkable. CT scan of her chest did not show evidence of pulmonary embolism, venous doppler of the lower limbs was normal. A transthoracic ECHO did not reveal a patent foramen ovale, elevated pulmonary artery pressures, or any other abnormality. Blood cultures showed no growth.

Table 1

Laboratory parameters

MRI brain done 2 days later showed a shower of emboli with multiple scattered tiny hyperintense DWI (diffusion-weighted imaging) [Figure 1a] lesions in bilateral cortical areas, deep watershed regions, basal ganglia, and the posterior fossa. Some of these lesions were hypointense on apparent diffusion coefficient (ADC) [Figure 1b] map and a few of them were isointense. Gradient Echo (GRE) showed hypointense lesions suggestive of petechial hemorrhages [Figure 1c]. There was no vascular occlusion or stenosis on MR angiography.

Figure 1

MRI brain showed restricted diffusion on diffusion-weighted (DWI) sequences (a) in bilateral cortical areas, deep watershed regions, and basal ganglia, hypointense on apparent diffusion coefficient (ADC) (b) and gradient-recalled echo (GRE) (c)

Transcranial Doppler performed for a duration of 30 min detected emboli in bilateral middle cerebral arteries (MCAs). PMD (Spencer Technologies, Inc; PMD 100 mol/L) was used along with a 2-MHz pulsed-wave transducer to generate simultaneous M-mode and spectral TCD displays. Probe fixation using a head frame (Marc 500, Spencer Technologies) was used for monitoring. The insonation depth for spectrogram recording was between 45 and 65 mm for the ipsilateral MCA.

Microembolic signals (MES) were defined as high-amplitude, unidirectional, transient signals lasting less than 300 ms [Figure 2] and associated with an acoustic resemblance to a characteristic chirp or snap. Ten MESs each were detected in either MCAs. The intensity of solid MESs ranged from 7 to 9 dB, and the duration ranged from 40 to 60 ms. These emboli had an Relative Energy Index of MES (REIM) of 0.40. The relative energy index of MES (REIM) was obtained by multiplication of maximum duration of MES in a 3-mm gate (MaxD) with MES maximum intensity adjusted for the intensity of background blood flow (MaxI) on Power M-mode display.[2] A final diagnosis of FES was made. She received conservative management in the form of antiplatelet agents, intravenous fluids, deep vein thrombosis prophylaxis, and ventilatory support. Unfortunately, she progressively deteriorated and died after 5 days.

Figure 2

Transcranial Doppler of the middle cerebral artery showing microembolic signals

DISCUSSION

The incidence of symptomatic fat embolism in a single long bone fracture ranges between 0.5 and 3%.[3] There are two mechanisms cited for the genesis of fat embolism.[4] The first mechanism suggests that the dislodged marrow fat enters the systemic circulation during long bone fractures due to increased intramedullary pressure as seen in closed fractures or surgical manipulation even in the absence of cardiac or pulmonary shunts, bypassing the filtration mechanisms of the pulmonary capillary bed. The second proposed mechanism is that, during periods of stress, secreted catecholamines trigger metabolism of the body fat stores leading to fat embolism. MES to brain with TCD monitoring have been recorded more commonly during impaction of a cemented component or after relocation of the hip than during impaction of the acetabulum component.[5] The FES may manifest after a latent period of 12–48 h.[1] The major diagnostic criteria of FES are hypoxia, deteriorating mental status, and petechiae; minor signs include tachycardia >120 beats/min, fever usually <39° C, sudden drop in hemoglobin of >20% of the admission value, sudden drop in platelet counts of >50% of the admission value, retinal changes, jaundice, oliguria, high erythrocyte sedimentation rate (ESR >71mm/h), and fat macroglobulinemia.[67] Our patient did not meet the criteria for FES by Gurd[67] or Schonfeld.[8] However, modern imaging tools such as MR brain and transcranial doppler were not available when these criteria were formulated.

CT brain in most cases is normal. The diagnostic features of cerebral fat embolism on MRI are multiple small, nonconfluent hyperintense lesions on DWI and T2-weighted images, usually situated in the cerebral white matter and deep gray matter.[9] DWI lesions represent the cytotoxic edema which develops initially, areas of increased signal intensity on T2-weighted scans presumably reflect vasogenic edema, which develops at a later stage.[10] However, some DWI lesions are isointense on ADC suggesting a component of vasogenic edema too.[9] Our patient had MR findings suggestive of fat embolism.

TCD monitoring for cerebral emboli was performed in vivo in long bone fractures in patients with FESs.[11]

MES are high-amplitude, unidirectional, transient signals lasting <300 ms with acoustic resemblance characteristic “snaps, tonal chirps, or moans”.[12] Acoustic artifacts can be differentiated from MES by their bidirectional nature and a low frequency (<400 Hz).It is difficult to differentiate different types of solid emboli by TCD like atherosclerotic, fat, or platelet emboli. Gaseous MES are often differentiated by their bidirectionality and higher intensities (>25 dB above background).[12] However, in the appropriate clinical scenario like ours, in the absence of other sources of embolism such as cardiac or carotid disease or venous thrombosis, the likelihood of the emboli being fat emboli is quite high.

Early institution of oxygen therapy reduces the effect of hypoxemia and the systemic complications of FES.[13] Medications including steroids, heparin, alcohol, and dextran have been tried but with uncertain benefit.[7]

The mortality in patients with FES and long bone fractures ranges from 0.9% to 8.7%, but with improvement in supportive therapy, and early stabilization and definitive fixation of fractures, the overall outcomes have improved recently.[1]

CONCLUSION

Diagnosis of FES requires a high index of suspicion. Diagnosis may not be possible with clinical criteria in all cases. MR imaging of the brain and transcranial doppler monitoring of microembolism to the brain are useful techniques in supporting the diagnosis of FES. Transcranial doppler is a useful modality in diagnosing cerebral microembolism especially in situations where MR brain is not feasible due to technical or logistic reasons.