Perioperative support of a patient with fat embolism syndrome with extracorporeal membraneoxygenation

Abstract

Fat embolism syndrome (FES) is a serious complication of trauma that can result in multiorgan failure, including the acute respiratory distress syndrome. Occasionally, the severity of respiratory failure associated with FES warrants support with venovenous extracorporeal membrane oxygenation (VV-ECMO), a therapy with widespread use but inconclusive evidence. Early definitive fracture fixation is the mainstay of preventing further fat embolism and ongoing organ dysfunction, but poses significant risks to the maintenance of the extracorporeal circuit. We describe a rare case of a patient who required VV-ECMO for respiratory support prior to fracture fixation. The risks of intraoperative fat embolisation causing sudden circuit failure were managed by having a spare circuit available outside the operating room with readiness for an emergency circuit change. Postoperative fat deposition in the oxygenator was managed by a circuit change. Our case is the first to describe preoperative initiation of VV-ECMO for FES and highlights why this therapy should not delay definitive fracture fixation and how it can be safely managed in this setting.

Background

Fat embolism syndrome (FES) is a serious complication of trauma. Early fixation is important in reducing embolic load and ongoing complications.¹ However, patients can develop marked respiratory and haemodynamic compromise which complicates the feasibility of early definitive operative management. Respiratory complications include the development of acute respiratory distress syndrome (ARDS).

The best treatment for severe ARDS remains controversial but there is increasing evidence for the use of venovenous extracorporeal membrane oxygenation (VV-ECMO) as a supportive treatment.² Despite the benefits of extracorporeal membrane oxygenation (ECMO), it is a resource intensive intervention that can be associated with significant complications.

Limited literature exists reporting the institution of VV-ECMO for ARDS secondary to FES, and this only describes ECMO following fracture fixation. We present an interesting and rare case of a patient with long bone fractures following trauma who required urgent VV-ECMO prior to fracture fixation.

Case presentation

Our patient is a 24-year-old man who works as a bar manager in a rural town of New Zealand. He has no medical history and is a non-smoker. He was involved in a side-on motor vehicle accident at 50 km/hour while riding his scooter. There was no loss of consciousness and he was transferred to the nearest trauma referral centre. Following his assessment in the emergency department, he was noted to have the following right-sided fractures; a closed proximal third femoral fracture (figure 1), distal third tibial and fibular fractures (figure 2) and a patella fracture. His right leg was neurovascularly intact. A secondary survey revealed no other injuries. Imaging at the time included standard trauma plain film radiographs and imaging of the affected limb.

Figure 1.

Plain film radiograph of complex proximal femoral fracture.

Figure 2.

Plain film radiograph of tibia and fibula fractures.

The decision was made to not operate on the leg fractures immediately, as it was the middle of the night. He was admitted to an orthopaedic ward, with operative management planned for the following day. His right leg was placed in skin traction overnight.

He developed progressive hypoxia the following morning (day 1). Pulse oximetry on the ward demonstrated oxygen saturations of 68%–72% on room air, and an arterial blood gas at this time showed a PaO₂ of 5.84 kPa. He was admitted to the intensive care unit (ICU) and his oxygen saturations improved to 92% following institution of high-flow nasal prong oxygen with 40 L at FiO₂ of 40%. A chest radiograph was normal at this stage. Given his respiratory deterioration, there was now a reluctance to proceed with his fracture fixation, and his surgery was further delayed. He continued to deteriorate over the course of the morning with worsening hypoxia. Non-invasive ventilation was instituted at this point.

He had worsening tachypnoea and tachycardia with a respiratory rate in the 50’s and reduced Glasgow Coma Scale (GCS) of 12/15. He was febrile with a temperature >38°C for the duration of his admission. Due to the ongoing deterioration, he was intubated and ventilated. At this stage, he developed a requirement for norepinephrine at 0.05 μg/kg/min to maintain a mean arterial pressure of 70 mm Hg. Renal function was normal at this time. He had developed a cumulative fluid balance of positive 3 L over the preceding 36 hours. A full tertiary survey revealed no other issues.

Investigations

Multiple chest radiographs showed progressive bilateral pulmonary infiltrate in keeping with ARDS (day 2) (figure 3), culminating in extensive bilateral diffuse opacifications of both lung fields post VV-ECMO institution (day 3) (figure 4). A CT pulmonary angiogram (CTPA) performed on day 1 following his respiratory deterioration showed no filling defects consistent with thrombotic pulmonary embolism. There was, however, widespread multifocal ground-glass opacity throughout both lung fields. This was felt to be in keeping with fat embolism. There was no evidence of pulmonary contusion.

Figure 3.

Plain film chest anteroposterior (AP) radiograph showing changes consistent with ARDS (day 2). ARDS, acute respiratory distress syndrome.

Figure 4.

Plain film chest anteroposterior (AP) radiograph showing white out of lungs post VV-ECMO institution (day 4). VV-ECMO, venovenous extracorporeal membrane oxygenation.

A CT head scan on day 1 following his neurological worsening showed no acute cerebral infarct or intracranial bleed. This was repeated on day 4 following transfer to a tertiary centre with findings of tiny haemorrhages in the corpus callosum which could be consistent with FES.

Thrombocytopenia was noted with a platelet count of 70×10⁹/L.

Differential diagnosis

The most likely diagnosis was fat embolism with subsequent ARDS. Less likely diagnoses included aspiration pneumonia or fluid overload. A CTPA had ruled out pulmonary embolus as well as pulmonary contusions or pneumothorax as a result of trauma.

The diagnosis of FES is clinically challenging. The radiological findings are coupled with the overall clinical picture. The two main criteria for diagnosing FES are Gurd and Wilson³ and Schonfelds et al.⁴ These are further described in table 1.

Table 1.

The two major diagnostic criteria for FES

Gurds (at least 1 major and 4 minor suggests FES)		Schonfelds (total score >5 suggests FES)
Major	Minor	Clinical finding	Score
Axillary or subconjunctival petechiae	Tachycardia (>110 BPM)	Petechiae	5
Hypoxaemia (Pa <8 on FiO240%)	Fever (>38.5°C)	Chest radiograph changes (diffuse alveolar infiltrates)	4
Central nervous system depression disproportionate to hypoxaemia	Emboli present in retina on funduscopy	Hypoxaemia (PaO2 <9.3)	3
Pulmonary oedema	Fat present in urine	Fever (>38°C)	1
	A sudden inexplicable drop in haematocrit or platelet values	Tachycardia (>120 beats/min)	1
	Increasing Erythrocyte Sed-imentation Rate	Tachypnoea (>30 beats/min)	1
	Fat globules in sputum

Positive findings are highlighted in bold. He met both criteria, further increasing the likelihood of FES. Of note, there was absence of petechiae despite thorough clinical review. Petechiae are often thought to be pathognomonic of FES, however, are not seen in up to 40% of cases.¹⁸

FES, fat embolism syndrome.

Treatment

He continued to have worsening type 1 respiratory failure with conventional lung protective ventilation. He was therefore referred to the regional ECMO service, located ~200 km away. A retrieval team arrived by helicopter shortly after to institute VV-ECMO. VV-ECMO was chosen over venoarterial (VA)-ECMO as cardiac function was relatively unaffected aside from some right ventricular strain from the respiratory pathology. A transoesophageal echocardiogram (TOE) prior to cannulation revealed normal left ventricular size and function, a mildly dilated right ventricle with normal systolic function and no significant valvular abnormalities or pericardial effusion. VV-ECMO cannulation, guided by TOE, was performed with a 25 French multistage left femoral vein drainage cannula sitting with its tip in the inferior vena cava at the level of the hepatic vein. The return cannula was a 19 French single stage cannula situated in the right internal jugular vein with its tip in the proximal superior vena cava (figure 4). He remained stable on flows of 5 L/min with 100% FiO₂ and 6 L/min sweep gas flow throughout transfer.

There was no patent foramen ovale on a bubble study. The circuit was run heparin free given his low platelet count and recent trauma. Prior to transfer, his right leg was taped to his left leg as an immobilising measure, as other available Thomas splints would not fit on the ECMO transfer stretcher. He was transferred to the ECMO referral centre via helicopter.

Plans were made to fix the patient’s fracture as early as possible following transfer. On his first day in the tertiary centre (day 3 following injury), he proceeded to fracture fixation with tension band wiring of his patella fracture and intramedullary nailing of his tibial and femoral fractures.

Further intraoperative fat embolisation posed a threat to the function of the oxygenator. During the surgery, a spare ECMO circuit remained immediately available outside the operating theatre, and a plan had been formulated for rapid exchange of the circuit should this become an issue. The preferred patient positioning for this procedure was supine on a Jackson table with a bump under the ipsilateral hip. This was not possible due to likely kinking of the drainage cannula. The surgeons were able to modify their approach by using a sloppy lateral position instead. Although the circuit continued to be run heparin free, the potential for more blood loss than usual was anticipated.

Despite our concerns, the surgery was performed without incident. Specifically, there was only 500 mL of blood loss and there were no circuit issues intraoperatively. Postoperatively we continued to target a platelet count of >100 initially, and this required an ongoing substantial platelet transfusion.

He required VV-ECMO for 12 days in total. Notably on day 5, he had oxygen saturations in the low 80’s despite maximal ECMO circuit blood flow (5.8 L/min). There was significant visible fat deposition noted in the oxygenator. Despite a post oxygenator PaO₂ of oxygen of 30 kPa, a decision was made to perform a circuit change and add a second oxygenator. Oxygenation improved following this. On day 9, he developed a ventilator-associated pneumonia that required treatment with Tazocin. On day 15, a trial off sweep was successful and he was decannulated from ECMO without issue. A percutaneous tracheostomy was inserted the following day.

Outcome and follow-up

Following tracheostomy, he was transferred to a general ICU in the same institution.

He had a slow respiratory wean from mechanical ventilation and had his tracheostomy removed on day 22. He was slow to respond and interact with staff and family initially without any other focal neurological signs. This was diagnosed as a hypoactive delirium which subsequently slowly improved with conservative management. During his illness, he had also sustained an acute kidney injury requiring ongoing renal replacement therapy initially. The requirement for this ceased on day 25. He was discharged back to the community on day 36, neurologically intact with no focal neurological deficit, with normal renal function and no residual medical issues.

Discussion

Our case is the first to highlight the challenges of performing definitive fracture fixation in a patient on VV-ECMO for FES.

The pathophysiology of FES is uncertain with both mechanical and biochemical theories used to explain the clinical picture. Liberation of fat globules ranging from 10 to 40 µm in diameter into the circulation occurs in all cases of pelvic and long bone fractures, and during surgical fixation.⁵ There have been two predominant theories proposed to explain the end organ dysfunction seen with FES. In the mechanical theory, emboli from this fat result in focal ischaemia in the affected organs. In the biochemical theory, the fat emboli result in an inflammatory response which causes the symptoms. This results in diffuse pulmonary, cerebral or dermal perivascular haemorrhages and odema.⁵

Study of FES can be challenging as the diagnosis is uncertain with two different diagnostic criteria used. A large review looking at 10 years of trauma patients using the Gurd criteria has the incidence of FES at 0.9%. Ninety-six per cent of these patients had hypoxia but the level of severe respiratory failure is not quantified. Overall mortality from this study was high at 7%.⁶ Further research has shown that mechanical ventilation is required in ~50% of patients who develop FES.⁷

The classical clinical triad of FES is neurological symptoms, respiratory distress and petechial rash. A higher embolic load can manifest as obstructive shock.⁵ Our patient displayed a fairly classical presentation with a roughly 12-hour asymptomatic period follow by rapid deterioration with hypoxia and ARDS. His GCS also worsened during a similar time period. Although this could be representative of diffuse axonal injury, given the location of the small haemorrhages in the corpus callosum, it is more likely to be a manifestation of FES given his normal level of consciousness on presentation. This highlights the diagnostic difficulties that can arise in this condition especially in the setting of polytrauma.

VV-ECMO is a therapy for severe ARDS, that is widely used in tertiary centres; however, the evidence base remains controversial. It involves circulating a portion of cardiac output through an extracorporeal circuit consisting of a pump and membrane oxygenator in order to provide temporary gas exchange while the lungs recover. Blood is returned to the venous side of the circulation. Theoretical advantages of VV-ECMO include the avoidance of ventilator induced lung injury while the cause of ARDS is treated and the lungs recover. An observational study in 2009⁸ during the H1N1 influenza epidemic demonstrated that ECMO could be used safely and effectively for the treatment of severe respiratory failure due to H1N1 influenza. In the same year, the Conventional ventilatory support vs Extracorporeal membrane oxygenation for Severe Adult Respiratory failure (CESAR) randomised control trial⁹ was published. Although the CESAR trial had methodological shortcomings, it again confirmed that ECMO could be used safely and that it had a place in the management of patients with severe respiratory failure refractory to conventional ventilation. Following these important publications in 2009, and with the advent of improved technology, the use of ECMO for severe respiratory failure has rapidly expanded. More recently, the 2018 ECMO to rescue Lung Injury in severe ARDS trial² attempted to clarify the place of ECMO in the management of hypoxic respiratory failure. Patients in the control arm received good, modern conventional management of their respiratory failure including lung protective ventilation, neuromuscular blockade, prone positioning, fluid restriction and inhaled vasodilators. No statistical difference in mortality or complications was shown; however, this result must be interpreted in the context of 28% of patients in the control arm crossing over to receive VV-ECMO as a rescue therapy. In the absence of clear harm, most tertiary centres continue to offer VV-ECMO as a rescue therapy for the management of severe respiratory failure, refractory to conventional management.

There is sufficient evidence that early definitive long bone fracture fixation is important in the prevention of respiratory complications such as fat embolism, ARDS and pneumonia.^{10 11} Specifically, early definitive therapy is the most important factor in preventing further fat embolisation in established FES.¹ However, the approach of ‘damage control orthopaedics’ including traction and external fixation is often advocated for the unstable patient.^{11 12} Instability can result from the rapid evolution of respiratory failure prior to fracture fixation. Further respiratory compromise from fat embolism intraoperatively is also a risk.

As experience with VV-ECMO has increased, so has the range of conditions being managed with VV-ECMO. VV-ECMO for patients with respiratory failure associated with trauma (such as pulmonary contusion, aspiration, transfusion related lung injury and FES) was once avoided. With increasing experience and improved circuit technology, however, clinicians are now more comfortable running VV-ECMO circuits on no or minimal anticoagulation for a period of time, allowing it to be used in this setting. This provides the potential option of early fracture fixation on VV-ECMO in the setting of severe respiratory failure.

The use of ECMO in ARDS secondary to FES is associated with specific complications. Circulating fat globules may cause obstruction of the oxygenator, potentially resulting in circuit failure. Our literature review found only three studies reporting on VV-ECMO for FES. We searched Embase, PubMed and Google Scholar using the terms ‘ECMO’ and ‘FES’ (including full terms).

Webb and colleagues in 2004¹³ describe successful VV-ECMO following fixation of ulnar and femoral fractures, while Valchanov and colleagues describe a similar scenario in a 32-year old with traumatic lower limb amputations following surgery.¹⁴ Both reports identify fat embolism to the ECMO oxygenator as a potential concern, but fortunately did not encounter it. A third study¹⁵ reports briefly on a case of donor acquired fat embolism following lung transplant (where the donor died of traumatic injuries) but does not focus on the specifics of the ECMO therapy and no mention of this potential complication is made. There is only one report of fat obstruction of the oxygenator; however, this occurred during cannulation for VA-ECMO where subcutaneous tissue was accidentally introduced into the circulation in an obese patient.¹⁶ In this case, the complication was immediately obvious and an immediate oxygenator exchange was performed.

In our case, the benefits of early operative fixation were thought to clearly outweigh the risks. Fracture fixation carried the risk of further lung injury from embolisation of fat, so it was important to sustain this insult early in his course of ECMO, maximising time to recovery postoperatively. Early fracture fixation is the single most important factor in preventing further fat embolisation¹ and leaving the fractures unfixed also allows ongoing release of inflammatory mediators which can further worsen ARDS and multiorgan failure.¹⁷ Additionally, getting the patient out of traction facilitates basic nursing cares which are required with patients on ECMO. We made plans for potential pitfalls. We were concerned about the risk of further fat embolisation during operative nailing. To this end, a spare ECMO circuit was available during the procedure and a plan was in place to perform an oxygenator exchange if needed. Compromise of the femoral access cannula was avoided by modification to a lateral approach by the surgical team. Postoperatively in the ICU, we were also prepared for an oxygenator exchange if required and indeed this was carried out on day 5, although on a non-emergent basis.

We are not the first to describe fixation of fractures while on ECMO.¹⁷ Carraro and colleagues published a case report of a patient with polytrauma complicated by respiratory failure who had fracture fixation while on VV-ECMO. The cause of respiratory failure is not clear in this case report; however, the rationale for early femoral fracture fixation was similar to our case. Although intraoperative fat embolism and oxygenator failure would have been a risk in this patient, it does not appear that this complication was anticipated. No postoperative issues with fat embolism to the oxygenator are reported in this case.

We believe our case has presented a unique discussion of the competing risks and benefits of providing ECMO support to patients with FES who are yet to undergo definitive fracture fixation. Preparation for emergency circuit change and pre-emptive change at signs of fat accumulation and decreasing maximal flow rates attenuates the risks of catastrophic circuit failure both intraoperatively and postoperatively. The risks of operating are outweighed by the risks of delaying ‘source control’ of embolism. We believe this is a viable therapy in this population of patients.

Learning points.

• Fat embolism syndrome (FES) is a complex syndrome with both uncertain diagnostic criteria and pathophysiology. It is rare but has a significant associated mortality.

• FES can occur prior to fracture fixation and be associated with severe acute respiratory distress syndrome and support with venovenous extracorporeal membrane oxygenation (VV-ECMO) is a viable therapy.

• VV-ECMO therapy should not delay early definitive fracture fixation.

• The risks of extracorporeal membrane oxygenation circuit failure due to fat deposition can be attenuated by preparedness for emergency oxygenator exchange and pre-emptive circuit change at signs of significant fat accumulation.