Fat Embolism Syndrome

Fat embolism syndrome is a leading cause of preventable morbidity and mortality in young trauma patients with long-bone fractures. It is rare in absolute terms, complicating about 1-5% of long-bone fractures, and almost entirely preventable by timely fracture stabilisation. The danger lies in the interval: the syndrome declares itself 24-72 hours after injury. The earliest signs may appear at 12 hours, but the classic presentation is on day 2 to 3, which is when the patient has to be re-examined.

Risk factors. The injuries and interventions that matter are those that expose a large marrow volume or disturb the fracture repeatedly:

• Long-bone fractures, the femur, tibia and humerus most commonly; the femur is the highest-risk single long-bone injury
• Pelvic and acetabular fractures, which expose a large marrow volume and release large volumes of fat
• Multiple fractures, the risk additive with each further bone
• Closed fractures, which hold a higher intramedullary pressure than open injuries
• Young adults, whose marrow has a high fat content
• Intramedullary nailing, where reaming and instrumentation force fat into the veins
• Repeated manipulation, since each reduction attempt embolises more fat
• Delayed fixation, beyond 24 hours

What it costs. Respiratory failure progresses to ARDS in severe cases, and the neurological injury ranges from cognitive impairment through focal deficits to coma. Recognition within the 24-72 hour window lets supportive measures start before fulminant respiratory failure or irreversible neurological injury.

From marrow to lung. Fat globules from the bone marrow enter the venous circulation through torn medullary veins at the fracture site, under raised intramedullary pressure, and travel to the pulmonary capillary bed. There they injure the lung in two ways. Mechanical obstruction of the capillaries raises pulmonary vascular resistance; lipase then releases free fatty acids from the trapped fat, and these damage the endothelium directly and inactivate surfactant, a chemical pneumonitis. The inflammatory cascade that follows produces capillary leak and an ARDS picture, ventilation-perfusion mismatch produces refractory hypoxia, and a massive embolus load strains the right heart.

Why the delay. The 24-72 hour latency reflects the time required for fat mobilisation, embolisation and the secondary inflammatory lung injury. The cascade runs in three phases:

• Embolisation, immediate to hours: fat globules enter the venous system, often subclinically
• Pulmonary phase, 24-48 hours: mechanical block plus chemical injury, giving hypoxia, tachypnoea and infiltrates
• Systemic phase, 48-72 hours: paradoxical emboli and inflammation, giving petechiae and confusion

View larger

View larger

Venous source, arterial disease. Marrow fat lodges in the pulmonary capillaries, yet the cerebral, cutaneous and retinal features are in systemic, arterial territories. For venous fat to reach them it must cross to the arterial side, and how it does so is a stock examiner question. There are two routes:

• Transpulmonary passage, the dominant route. Fat globules are small and deformable, so they pass through the pulmonary capillary bed, or through precapillary arteriovenous channels that open as pulmonary pressure rises, into the left heart and out to the systemic arteries. No intracardiac defect is required, which is why most patients with cerebral or cutaneous features have a structurally normal heart.
• Paradoxical embolism through a patent foramen ovale. A PFO is present in roughly a quarter of adults. The pulmonary fat load raises right-heart and pulmonary-artery pressure, which can open a right-to-left shunt and let venous fat bypass the lungs entirely. This is invoked particularly for florid or early cerebral fat embolism, and for cerebral involvement out of proportion to the lung injury.

The second hit at each end organ. Once globules reach a capillary bed, the free-fatty-acid injury amplifies the local damage. In the brain that means cerebral oedema and microinfarcts; in the skin, dermal capillary occlusion and leak, together with the thrombocytopenia, produce the petechial rash; in the retina, emboli may be seen on fundoscopy as Purtscher retinopathy.

Reaming pistons marrow fat into the venous system at high intramedullary pressure, and the same mechanism operates, often more dramatically, during cemented arthroplasty.

Bone cement implantation syndrome (BCIS). Pressurising PMMA cement into a prepared canal and impacting the prosthesis generates high intramedullary pressure that embolises marrow fat, cement monomer and air to the lungs. Within minutes of cementation or stem insertion this produces hypoxia, hypotension, pulmonary hypertension with right-heart strain, arrhythmia and, in severe cases, cardiac arrest.

Risk reduction. Warn the anaesthetist before cementing, and ensure euvolaemia and optimised haemodynamics. Lower the intramedullary pressure at the source:

• Lavage and dry the canal
• Use a cement restrictor
• Use a suction or venting catheter
• Insert the cement retrograde with a cement gun

Consider uncemented fixation in the very highest-risk patients, balanced against the higher periprosthetic-fracture rate of uncemented stems in osteoporotic bone.

Gurd criteria. The diagnosis is clinical: one major plus four minor criteria. Of the major criteria, petechiae are the most specific finding.

Severity. The picture runs from subclinical to fulminant, and severity correlates with the embolus load and the host inflammatory response. Early fixation dramatically reduces progression to fulminant disease.

Other criteria. The Gurd (and Gurd and Wilson) criteria are the most widely used, but they are not the only system, and examiners often ask what else exists. All are clinical, all overlap, and none has been validated as superior; they are best understood as different ways of weighting the same features. If asked for one alternative, the Lindeque criteria are the most useful answer: they make the point that a sustained, unexplained respiratory deterioration after a long-bone fracture can itself be sufficient to diagnose fat embolism syndrome, even before petechiae or confusion appear.

View larger

History. Establish the injury pattern (long-bone, pelvic or multiple fractures), the exact time of injury and of any surgery, and the precise onset of hypoxia, confusion or rash. Then the risk modifiers: age, closed versus open fracture, and whether the canal was reamed.

Examination. Reassess every multiply-injured patient on day 2 to 3 for new hypoxia or confusion, and look at four things:

• Respiratory: tachypnoea, hypoxia, bilateral crackles
• Skin: a diligent search for axillary or conjunctival petechiae
• Neurological: GCS, level of confusion, focal deficits
• Fracture sites: stability, swelling, any ongoing manipulation

The rash. Examine the axillae, upper chest and conjunctivae under good light. The petechiae are fine, 1-2 mm in diameter, often in crops, and may be transient, which is why they are missed unless deliberately sought; they typically resolve within 5-7 days. With hypoxia and neurological change after a long-bone fracture the rash makes the diagnosis virtually certain, but its absence does not exclude it. In any long-bone-fracture patient who becomes hypoxic or confused on day 2 or 3, examine the skin and eyes before attributing the change to atelectasis or head injury.

Differential diagnosis. Post-traumatic hypoxia and confusion have several causes, and the latency and the rash are what separate them.

What investigations are for. The diagnosis is clinical, on the Gurd criteria. Investigations quantify the hypoxia, support the minor criteria and exclude the treatable differentials; imaging supports the diagnosis but is never pathognomonic. Do not delay supportive care while awaiting advanced imaging, and base fixation decisions on fracture pattern and physiological stability, not on imaging confirmation of emboli.

Arterial blood gas, immediately, on any suspicion of hypoxia or tachypnoea. It measures the hypoxia against the Gurd threshold, and shows a widened A-a gradient and, initially, a respiratory alkalosis; it quantifies severity and guides oxygen therapy.

Chest radiograph, PA or AP erect. The classic finding is bilateral diffuse infiltrates resembling ARDS or pulmonary oedema; the "snowstorm" appearance is classic but uncommon. The film may be normal early, and its changes lag behind the clinical hypoxia.

View larger

View larger

Laboratory tests. Full blood count for the thrombocytopenia, coagulation, renal and liver function, and lipase if available. Fat globules in the urine or sputum have low sensitivity. These results support the minor Gurd criteria rather than make the diagnosis.

CT of the chest. High-resolution CT may demonstrate ground-glass opacities and centrilobular nodules. CT pulmonary angiography is ordered in the severe case to exclude thromboembolic pulmonary embolism, not to find fat: fat emboli are not visible on CTPA, although a filling defect of negative attenuation, as in the figure below, favours macroscopic fat over thrombus. Both rule out treatable differentials and guide ICU escalation.

View larger

View larger

Echocardiography, in the severe case, looks for right ventricular strain, right ventricular dilatation suggesting a massive embolus load, or a patent foramen ovale facilitating paradoxical emboli.

Brain imaging. CT or MRI is indicated for focal signs or persistent coma once head injury has been excluded. MRI can reveal diffuse white-matter hyperintensities on T2-weighted sequences consistent with embolic microinfarcts, which supports the diagnosis and helps prognostication; the diffusion-weighted pattern is described in the next section.

Presentation. A subset of patients present with neurological features predominating, sometimes with little or no respiratory compromise. Consciousness fluctuates, with confusion, drowsiness, seizures or focal deficits and occasionally coma, and the neurological phase may precede or occur without the pulmonary phase. It is under-recognised because the chest can look well, and it is easily mistaken for traumatic brain injury, sedation, hyponatraemia or sepsis, so it must be actively considered in a long-bone-fracture patient who becomes confused or drowsy on day 2 to 3.

MRI is the key test. Diffusion-weighted imaging shows numerous small foci of restricted diffusion scattered through the white matter, the starfield pattern, reflecting microembolic infarcts, with T2 and FLAIR white-matter hyperintensities. The diffusion pattern is more specific than the T2 changes alone and helps separate cerebral fat embolism from diffuse axonal injury.

View larger

View larger

View larger

Prognosis. Despite a dramatic presentation, neurological recovery with supportive care is often good, which is the argument for aggressive supportive management rather than nihilism.

Prevention through early fixation. The goal is to stabilise the fracture before ongoing embolisation from a mobile fracture site adds to the load. Multiple studies demonstrate that definitive fixation within 24 hours of injury significantly reduces the incidence of the syndrome compared with delayed fixation, by over 50%. Timing of surgery is the modifiable risk factor. The protocol runs in three steps:

• Pre-hospital and emergency department, 0-6 hours. Splint fractures to reduce motion and further embolisation; avoid excessive manipulation, log-roll carefully and limit reduction attempts; book theatre within 24 hours if the patient is physiologically stable
• Damage control, first 24 hours. External fixation or traction for the unstable patient; correct acidosis, hypothermia and coagulopathy; serial observations for early signs of the syndrome
• Definitive fixation, within 24 hours. Early total care, converting to an intramedullary nail or plate when safe; unreamed nails or gentle reaming in high-risk patients; continued vigilance for 72 hours after surgery

In the polytrauma patient too unstable for early total care, damage-control external fixation followed by conversion within 5-7 days still lowers the risk compared with prolonged traction.

Reaming. Reaming raises intramedullary pressure and produces a transient shower of emboli. Whether that becomes clinically important fat embolism syndrome is another matter: the Controversies section below records that large trials show no significant difference between reamed and unreamed nails when fixation occurs early, and that the choice of nail is dictated by fracture pattern and surgeon preference. The advice here is therefore graded: ream gently or use an unreamed nail in the high-risk patient, and avoid reaming where possible once the syndrome is established.

View larger

View larger

Supportive care in established FES. No pharmacological agent, whether steroids, heparin or statins, has proven benefit in altering the course of established fat embolism syndrome, so treatment is supportive: maintain oxygenation, prevent secondary organ injury, and support the patient through the inflammatory phase. Deterioration to ARDS can be rapid, so observe closely with serial blood gases and clinical examination, rule out pulmonary embolism, pneumonia and cardiac events, and run the care jointly between orthopaedics, intensive care and anaesthetics.

Airway and breathing. Target an SpO2 over 92%, with high-flow nasal oxygen or non-invasive ventilation initially. Intubate for refractory hypoxia, fatigue, a GCS under 8 or impending respiratory arrest. Ventilate with low tidal volumes (6 mL/kg) and titrated PEEP, avoiding over-distension.

Circulation. Balanced resuscitation, avoiding the overload that worsens pulmonary oedema, and avoiding hypotension and hypoxia, which cause secondary injury. Vasopressors are for a septic-like inflammatory response or right heart failure. Monitor lactate, urine output and central venous pressure.

Neurological. Check the GCS frequently; deterioration may indicate worsening embolisation or cerebral oedema. Seizure prophylaxis is not routine unless seizures occur. Maintain normocapnia and normoglycaemia to avoid secondary brain injury.

Fracture care. Complete stabilisation once the patient is physiologically optimised, using unreamed implants where possible, with daily liaison between the orthopaedic and intensive care teams.

Mortality rises sharply once the syndrome progresses to fulminant ARDS or coma, and once it is established management is supportive only. The single most effective intervention against the complications below is early definitive fixation, whenever the patient's physiological state permits.

What to expect. With fixation within 24 hours most patients have minimal long-term sequelae. Moderate disease treated with oxygen, monitoring and fracture fixation recovers fully in 70-80%, with possible mild cognitive or respiratory limitation. Fulminant disease needing ventilation and advanced therapies carries the mortality given in the severity table, and its survivors may be left with deficits and often need neurorehabilitation.

Prognostic factors. The timing of fracture fixation is the only evidence-based intervention that meaningfully alters incidence, and therefore outcome. Beyond that:

• Best prognosis: early recognition, prompt supportive care, fracture stabilisation achieved, a young patient without comorbidities
• Poor prognosis: diagnosis delayed beyond 72 hours, progression to refractory ARDS, pre-existing lung disease, a coexisting massive head injury