lable at ScienceDirect

Current Anaesthesia & Critical Care 21 (2010) 277e281

Contents lists avai

Current Anaesthesia & Critical Care

journal homepage: www.elsevier .com/locate/cacc

POINTS OF VIEW

Fat embolism e An update

Pratik Sinha, Nick Bunker*, Neil SoniChelsea and Westminster Hospital, 369 Fulham Road, London SW10 9NH, UK

Keywords:IncidenceAetiologyPathophysiologyDiagnosisTreatment

* Corresponding author.E-mail address: nickbunker@doctors.net.uk (N. Bu

0953-7112/$ e see front matter � 2010 Elsevier Ltd.doi:10.1016/j.cacc.2010.03.003

s u m m a r y

Fat embolism syndrome is an unexpected and alarming complication that is difficult to actively prevent,hard to diagnose with confidence and has limitations in effective treatment modalities. The syndrome isa melange of respiratory, haematological, neurological and cutaneous symptoms and signs associatedwith trauma and other disparate surgical and medical conditions. The pathogenesis is still debated. It isclear that fat emboli are quite common yet the clinical syndrome is rare. Diagnosis is by patternrecognition as befits a syndrome, but the recently defined features on MRI could now be used to increasethe probability of the diagnosis. Various therapeutic options have been tried and failed. At presentsteroids have a single meta-analysis suggesting benefit but it is in the trauma population where they maybe contra indicated for other reasons, i.e. infection, so their place is ill defined. Supportive treatment isthe mainstay.

� 2010 Elsevier Ltd. All rights reserved.

1. Definition

It is important to differentiate fat in the circulation and thesyndrome that is sometimes seen as a consequence. Fat in thecirculation is probably quite common and embolisation can occur.Fat embolism syndrome (FES) is when there is fat in the circulationand it is associated with an identifiable clinical pattern of symp-toms and signs.

The classic description of the syndrome is the combination ofrespiratory failure associated with neurological disturbance anda petechial rash often in the absence of any obvious cause.

2. Pathophysiology

The mechanism of fat embolism syndrome is not clear. It is clearthat fat frequently gets into the circulation and can be demon-strated on both sides of the circulation. It is clear that fat can causeembolisation. Yet the syndrome is relatively uncommon. Asa syndrome the range of effects implies diffuse damage at morethan one site. This could be due to embolisation or may be asso-ciated with a different secondary mechanism. It is hard to tie all theeffects seen in FES to embolus as a sole mechanism. Several theorieshave been postulated but the two main contenders are themechanical and biochemical theories.

nker).

All rights reserved.

2.1. Fat embolus theory: mechanical

Fat globules can be physically forced into the venous systemduring trauma. In surgical situations high pressures are exerted inthe marrow and this may force fat into the blood stream. Thenormal marrow pressure is 30e50 mmHg but this can be dramat-ically increased (up to 800 mmHg) during intra-medullary reamingand insertion of intra-medullary devices.1 The presence of fat microemboli can be clearly shown with ultrasonography in a relativelyhigh percentage of patients albeit they are micro emboli and insmall numbers.2,3 Micro emboli seem to be seen most duringmanipulation of the intra-medullary cavity. Circumstantialevidence for the importance of intra-medullary fat comes from theobservation that in previously reamed femoral cavities during redoreplacements the incidence is low.4 The use of cement is alsoassociated with fat embolic phenomena but it is also seen incement-less prosthesis.5 Various models of fat injection or sub-jecting the medullary marrow to high pressures can all be shown toproduce emboli and to result in cardiorespiratory problems. It hasalso been shown that the use of external fixation in trauma isassociated with a lower incidence of FES than intra-medullaryfixation.6

2.2. Free fatty acid theory: biochemical

This is in two parts. The first is the theory that trauma results inthe release of lipases into the plasma that destabilise circulating fatmolecules resulting in their saponification and de-emulsification.In animal studies exogenous fat in the circulation has been shown

P. Sinha et al. / Current Anaesthesia & Critical Care 21 (2010) 277e281278

to cause the release of phospholipase A2, methylguanidine andproinflammatory cytokines. The significance of this phenomenonin vivo is unknown. In the same model fat droplets and fibrinthrombi were found in several organs.7

A more appealing biochemical theory invokes the histotoxiceffects of free fatty acids (FFA). It is known that FFA’s can causesevere vasculitis in animal models and in the lung this can disruptpulmonary architecture fairly rapidly. If the free fatty acids are nearthe endothelium they can cause disruption of the integrity of theendothelial membrane and capillary wall leading to haemorrhageand oedema. Although bone marrow is predominantly neutral fat,in the acute inflammatory milieu it is possible that FFA activity maybe important.8e10 If in vivo there is hydrolysis of neutral fats toFFA’s it could explain the interval before the probable time of fatrelease and the onset of signs and symptoms. In the clinical settingfollowing long bone surgery FFA levels were found to be signifi-cantly higher in patients with multiple injuries and associated withsignificantly lower oxygen tension.11

2.3. Thrombocytopaenia

Part of the syndrome is often a falling platelet count. In thetrauma populationwith hypovolaemia and coagulopathy theremaybe an association between sluggish, hypoxaemic circulation andpossible ‘sludging’ of blood.12 Coagulation abnormalities areencountered frequently in FES. The combination of activation ofplatelets, vascular endothelial effects and microaggregate collec-tion might be catalysed by the presence of fat.

2.4. Systemic and paradoxical embolisation

Given that fat is released into the venous return to the rightheart, it is a curiosity that it causes systemic effects, sometimeswithout pulmonary effects.13,14 It has been suggested that thiscould be via a patent foramen ovale (PFO) which is not uncommonin the general population, with an incidence of about 25%.15

Although closure of PFO does reduce emboli, whether a PFO waspresent or not makes no difference to the incidence of FES and it isclear that emboli are seen in patients without PFO.3,16e18 Trans-pulmonary systemic fat embolisation does occur and has beenshown in dogs without a patent foramen ovale.19 The size anddeformability of the fat micelles may allow this pulmonary transit.

There is no clear individual mechanism that can readily explainall the abnormalities seen in the syndrome. As shall be discussedlater the distribution of lesions on cerebral MRI might provide cluestowards multiple mechanisms but to date it is conjecture.

3. Clinical features of the syndrome

3.1. Associations

There are several reported causes of fat embolism syndrome.These are listed below and may be considered as preconditionswhere fat embolism syndrome (FES) might be seen. Trauma withlong bone fractures and major elective orthopaedic procedures arethe most common causes. Other causes are rarer. Liposuction hasthe potential to force fat into the circulation and there are manycase reports of it causing Fat Embolism Syndrome (FES). Medicalconditions such as hepatic necrosis or the fatty liver may predis-pose. Drugs such as intralipid,, parenteral nutrition or propofolhave the potential to cause the syndrome although themechanismsmay differ in that the fat emulsions may coalesce and causemechanical obstruction of the vascular tree and hence localdamage. FES is implicated in sickle cell disease. In an acute sicklecrisis the pulmonary macrophages often stain positive for fat and

Bronchoalveolar Lavage may be a useful adjunct to diagnosis.20

Bone marrow necrosis may release fat and certainly non-specificmarkers such as lipases are raised. In a recent study the incidenceappeared very high (33%).20e22

Main causesTrauma e long bone fracturesJoint replacements e prosthetic placementScoliosis surgery

Mechanical disruption of adipocytesSoft tissue injury. (Crush and Blast injuries)LiposuctionLiver failure e fatty liver.

Mechanical disruption of the bone marrowBone marrow harvestBone marrow transplant

Exogenous fatParenteral nutritionPropofol infusionLymphograophy

Non-specificBurnsExtra corporeal circulationAcute sickle crisisAcute pancreatitisDecompression sicknessAltitude sickness

3.2. Diagnosis

Fat embolism syndrome is a collection of symptoms and signs.As with any syndrome it is a clinical pattern rather than an easilydefined entity. It is helpful to consider the condition in terms ofpreconditions that need to exist, symptoms that may occur andphysical signs that may or may not be present and to a lesser orgreater degree. It may even then be a diagnosis of exclusion as itoverlaps with other conditions.

While many of the signs may be suggestive of the condition,none are absolutely pathognomonic.

It is also important to emphasise the wide range of clinicalpresentation from almost imperceptible sub-clinical signs to themore common gradual onset or the fulminating crisis withpulmonary and systemic embolisation of fat, right ventricularfailure and cardiovascular collapse.

FES can present in a fulminant manner after the trauma orduring the operation. The more classical onset is gradual over12e36 h following injury or surgery, with increasing hypoxaemiaoften associated with neurological symptoms such as confusion,drowsiness, or coma. It may be accompanied by fever and a char-acteristic petechial rash, often in the axillae but may be on the faceor in the conjunctivae.

To diagnose this syndrome requires the right preconditions toexist (see associations above). Gurd’s classification is helpful andcan be seen in Table 1.

In order to diagnose FES, 1 major and 4 minor criteria arerequired. Although various reports describe the presence of fatglobules in the blood or sputum this requirement poses severalproblems. While the original guidelines require daily testing for fatglobules it has been contentious. Fat globules can be found in bothtrauma patients andmoreworryingly in healthy volunteers with no

Table 1Modified from Gurd.23

Major criteria Petechial rashRespiratory symptoms e tachypnoea, dyspnoea,bilateral inspiratory crepitations, haemoptysis,bilateral diffuse patchy shadowing on chest X-rayNeurological e confusion, drowsiness, coma

Minor criteria Tachycardia> 120 beatminr1

Pyrexia> 39.4Retinal changes e fat or petechiaeRenal changes e anuria or oliguriaJaundice

Laboratory (minor) Thrombocytopaenia> 50% decrease on admission valueSudden decrease in haemoglobin level> 20%of admission valueHigh erythrocyte sedimentation rate> 71 mmhr1

Fat macroglobulaemiaFat in sputum

P. Sinha et al. / Current Anaesthesia & Critical Care 21 (2010) 277e281 279

suggestion of FES. It is clearly relatively non-specific. Pragmaticallyit is also difficult to get the laboratory to look for fat. Hence themodified Gurd’s classification may represent a more clinicallyrelevant system.

Lindeque suggested a more specific approach to the lungcomponent of FES which measures functional aspects of the lunginjury. These are very similar to those used for any other form oflung injury.

Pulmonary criteria for FES e Lindeque.24

1 A sustained Pao2 of less than 8 kPa (Fio2 0.21).2 A sustained Paco2 of more than 7.3 kPa or pH of less than 7.3.3 A sustained respiratory rate of greater than 35 breaths.minr1

even after adequate sedation.4 Increased work of breathing judged by dyspnoea, use ofaccessory muscles, tachycardia and anxiety.

The incidence of the condition is difficult to identify. Theretrospective reported incidence tends to be low with many lessthan 1% but with a fewas high as 4%. Prospective studies report a farhigher incidence of 11e19%. This is not as high as the incidence offat embolisation at post mortem where evidence of fat is verycommon but not necessarily associated with the syndrome itself.

4. Presentation

Most patients, more than 70%, have a significant respiratorycomponent25 but this can be very variable, ‘not all respiratoryfailure following orthopaedic surgery is fat embolism, and not all fatembolism results in respiratory failure’. Certainly transient post-operative hypoxaemia is common and there are many possiblecauses.

When diagnosed a high proportion of patients need ventilation,44% in one series.25 In those with X-ray changes and impairedrespiratory function requiring ventilation resolution often occursrapidly, on average 3e7 days of invasive support is all that isrequired.25,26 There is a continuumof lung injury related to FES froma groupwith sub-clinical pulmonary dysfunction at one end to thosewho remain ventilator dependent for many days, may be difficult tooxygenate and are left with long-term pulmonary sequelae.

The rash is the feature closest to being pathognomonic. It is seenin 33% of cases25 and is frequently axillary but can also be found onthe forehead, sub-mucosa and conjunctiva. The classic distributionof the rash has been attributed to accumulation of fat droplets inthe aortic arch prior to distribution through carotid and subclavianvessels to non-dependent areas. Its development may be related tostasis, endothelial damage by FFA’s or peripheral embolisation butthe exact mechanism is yet to be fully elucidated. Many of these

patients will develop a coagulopathy, also a cause of petechiae,however the rash often occurs with platelet counts above 50,000 atwhich level petechiae related to thrombocytopaenia would beunusual. It is worth remembering that in diagnosing the syndromeit is the relative platelet count rather than an absolute value that isimportant. Coagulopathy, infectious diseases and some congenitalconditions can all result in a similar rash; context is crucial.

Neurologically, patients can present with a wide range ofsymptoms and signs from mild confusion and disorientation toconvulsions and coma but neurological dysfunction is common. Inone series of 14 patients, 5 were unconscious, 4 demonstrateddecerebrate posturing and one had a clonic seizure.27

These symptoms may not be present on admission so thedifferential diagnosis of unexpected neurological deterioration intrauma or post-operative orthopaedic patients should includeFES.28e30 This may include a patient with no obvious neurologicalproblems preeprocedure but a significant deficit afterwards.Clearly in the trauma patient there are many other potential causesof delayed neurological deterioration.

It is important to appreciate that some patients have neuro-logical findings without respiratory dysfunction.14,31 This isparticularly likely if the neurology is focal.32

The outcome from cerebral fat embolism is often, but notalways, good. This is frequently stated in the literature but withlittle evidence to support this statement.33 There is no obvious wayof determining prognosis although recently it has been suggestedthat MRI could distinguish between reversible vasogenic oedemaand signs of more severe damage, the latter a poor prognosticindicator.34,35 The more severe cerebral injuries may be associatedwith a patent foramen ovale and it has been suggested that pre-operative closure may prevent embolisation.36

5. Investigations

As a syndrome unsurprisingly there is no single definitiveinvestigation although MRI may be getting close.

Laboratory tests are often abnormal but there are no specifictests for FES. Thrombocytopaenia (platelet count <150�109/L) andanaemia are common (37% and 67%, respectively)25 but are alsofeatures of trauma per se. Biochemistry is unhelpful, hypocalcaemiaand hypoalbminaemia commonly occur in Critical Care patients.Serum lipase and phospholipase A2 (PLA2) both tend to rise in lunginjury but these changes are not specific to FES. Fat globules may beidentified in both blood and urine but again are non-specific andmay occasionally be seen in normal volunteers.

Bronchoalveolar lavage (BAL) has been used to sample macro-phages which might be expected to contain fat in FES. Despiteattempts to calibrate the amount of fat in these cells, a lack ofspecificity for FES remains a problem.37

Previously a pulmonary artery catheter had been advocated forthe diagnosis of fat embolism by detecting a rise in mean pulmo-nary arterial blood pressure. However a recent study suggests thisis of no benefit and the majority of patients with confirmed FEShave few if any cardiovascular changes.38 The value of finding fat inpulmonary arterial blood samples is unknown.39

The chest X-ray usually shows bilateral patchy oedema likeinfiltrates. The descriptive term ‘snow storm appearance’ has beenused in more severe cases.

Computer tomography (CT) is generally unhelpful beyondshowing non-specific markers of injury and also for eliminatingother causes of neurological deterioration. One author suggeststhat the failure of a neurological picture to correlate with the CTscan is a reason to consider fat embolism.40 In severe injury it mayshow generalised cerebral oedema or high density spots butusually adds little.

P. Sinha et al. / Current Anaesthesia & Critical Care 21 (2010) 277e281280

Magnetic resonance imaging (MRI) is showing greater promise.The term ‘starfield’ describes the pattern seen due to innumerablepunctate areas of restricted diffusion.33 In recent experimentalstudies at least two types of changes have beendetected in bothDWI(DiffusionWeighted Image) and T2weighted imageswithin thefirst30 min from the insult. Firstly, multiple small non-confluent hyper-intense lesions are seen in the white and grey matter, which arethought to be microinfarcts. Secondly, there is a subtle increase inthe signal intensity with isointensity on ADC mapping (apparentdiffusion coefficient) and obvious enhancement with gadolinium.This represents areas of vasogenic oedema and should be reversible.Theymaybedue to the presence of free fatty acids and thusmayhinttowards a mechanism for some of the pathology. Furthermore theuse of DWI sequences appears useful as both types of lesions wereseen at 7 days but by 21e30 days the vasogenic lesions had dis-appeared.41 This time line appears to correspond with clinicalimprovement.34,42 It is very early days but the MRI changes arespecific enough to support a diagnosis of FES even if they are not yetpathognomonic. It would seem sensible to perform an MRI in anypatient with orthopaedic injuries who manifests with an acutealteration in mental status in the presence of a normal CT scan.42

Certainly with the right preconditions and the presence ofadequate criteria, MRI is likely to be very helpful and may be able totentatively provide some guarded prognostic information.

6. Prevention of fat embolism

Patients presenting with trauma or for an elective prosthesishave the highest risk of developing FES but they require differentstrategies to try and mitigate that risk.

Primary prevention of trauma is an important aim but by defi-nition when these patients are admitted the damage has alreadybeen done.We cannot prevent the fat embolus caused by the injuryitself but the management of that injury provides a preventiveopportunity. The most important intervention is early fixation andideally this should happen in the first 24 h.43e47 Surgical manipu-lations in the medullary cavity, in either emergency or electivework, creates high pressures and predisposes to fat embolism.48

External fixation minimizes interference with the marrow.49 Ifthe medulla must be invaded then unreamed nails are associatedwith a lower incidence of FES than reamed nails,50 sharp reamersproduce less pressure than blunt reamers and hollow nails producelower pressure than solid nails.51 Ultrasonic reamers havea propensity to generate multiple embolic showers.52 There are,however, some simple surgical measures that can reduce the risk ofFES when performing intra-medullary surgery. This includesventing medullary cavities to reduce pressure and lavaging thecavity before inserting any prosthesis.

Corticosteriods have been used as prevention but there areobvious contraindications in patients following major trauma anda likely period of critical illness. This is clearly contentious but in theless critically ill patient’s such as an isolated long bone fracture orelective joint replacement they may have a role. A recent meta-analysis in patients with isolated long bone fractures has comedown in favour of using steroids for prevention with a NNT of 8.53

However most of the studies in the meta-analysis have significantheterogeneity and are over 20 years old, it is debatable whether,with improvements in surgical technique, the conclusions are stillvalid today.

7. Treatment

Good supportive treatment is essential in these patients whomay develop amultitude of problems. This is particularly importantin the management of the respiratory component of the syndrome.

Early diagnosis will not result in any specific interventionalthough it may be very useful for informing relatives and intempering prognosis. As there is no specific treatment that is ofbenefit it is important not to put the patient in jeopardy in order toclinch the diagnosis. In particular, careful consideration of the risksof intra-hospital transfer and the support of critically ill patientsundergoingMRI should beweighed up. If necessary the scan shouldbe delayed until the patient is in a suitable clinical state by whichtime they may have recovered what function they are going toregain, making the MRI less useful.

Steroids pose a problem. The possible benefits must be balancedagainst the infective complications especially in patients who arecritically ill. There are case reports of their use54 but both timing anddosage are ill defined with dosages of 6e90mg/kg and very variabletreatment durations. In the recent Canadian meta-analysis thenumber needed to treat was 8 but it was small (approx. 300 patients)and it focused only on prevention53 Without further evidence it isdifficult to fully endorse their use. Other drugs such as dextran,heparin and aspirin are of no proven benefit and have their ownintrinsic problems. Early pre-operative echocardiography and theclosure of any demonstrable PFOmay prevent emboli. Unfortunatelythere isnomagic solutionandasusual in ICU it isgoodsupportive careand avoidance of complications that are themainstay of therapeutics.

8. Outcome

A syndrome encompasses a wide clinical spectrum. Somepatients do not survive the acute event. Some patients with verysevere lung disease may have residual deficit and may even requirelong-term ventilation whilst others may not need any respiratorysupport. Similarly the neurological outcome is unpredictable. It iscommonly reported in the literature as a disease process witha favourable prognosis but this statement is rarely supported byrobust outcome data. MRI appears to move towards definitivediagnosis, which may be an advance on what was availablea decade ago and may even give us a better idea of prognosis.Although we may be better at diagnosing the syndrome the onlymanagement continues to be supportive care, this is unlikely tochange until we understand more about its pathophysiology.

Conflict of interest statement

The authors have no conflicts of interest to declare.

References

1. Kropfl A, Berger U, Neureiter H, Hertz H, Schlag G. Intramedullary pressure andbone marrow fat intravasation in unreamed femoral nailing. J Trauma 1999;42(5):946e54.

2. Barak M, Kabha M, Norman D, Soudry M, Kats Y, Milo S. Cerebral microemboliduring hip fracture fixation: a prospective study. Anesth Analg 2008 Jul;107(1):221e5.

3. Forteza AM, Koch S, Romano JG, Zych G, Bustillo IC, Duncan RC, et al. Trans-cranial doppler detection of fat emboli. Stroke 1999 Dec;30(12):2687e91.

4. Barre J, Lepouse C, Segal P. Embolism and intramedullary femoral surgery. RevChir Orthop Reparatrice Appar Mot 1997;83(1):9e21.

5. Kim YH. Incidence of fat embolism syndrome after cemented or cementlessbilateral simultaneous and unilateral total knee arthroplasty. J Arthroplasty2001 Sep;16(6):730e9.

6. Meek RN, Woodruff B, Allardyce DB. Source of fat macroglobules in fractures ofthe lower extremity. J Trauma 1972 May;12(5):432e4.

7. Liu DD, Hsieh NK, Chen HI. Histopathological and biochemical changes followingfat embolismwith administration of corn oilmicelles: a newanimalmodel for fatembolism syndrome. J Bone Joint Surg Br. 2008 Nov;90(11):1517e21.

8. Glover P, Worthley LI. Fat embolism. Crit Care Resusc 1999 Sep;1(3):276e84.9. Schnaid E, Lamprey JM, Viljoen MJ, Joffe BI, Seftel HC. The early biochemical and

hormonal profile of patients with long bone fractures at risk of fat embolismsyndrome. J Trauma 1987 Mar;27(3):309e11.

10. Jones JG, Minty BD, Beeley JM, Royston D, Crow J, Grossman RF. Pulmonaryepithelial permeability is immediately increased after embolisation with oleicacid but not with neutral fat. Thorax 1982 Mar;37(3):169e74.

P. Sinha et al. / Current Anaesthesia & Critical Care 21 (2010) 277e281 281

11. Nixon JR, Brock-Utne JG. Free fatty acid and arterial oxygen changes followingmajor injury: a correlation between hypoxemia and increased free fatty acidlevels. J Trauma 1978 Jan;18(1):23e6.

12. Hofmann S, Huemer G, Kratochwill C, Koller-Strametz J, Hopf R, Schlag G, et al.Pathophysiology of fat embolisms in orthopedics and traumatology. Orthopade1995 Apr;24(2):84e93.

13. Citerio G, Bianchini E, Beretta L. Magnetic resonance imaging of cerebral fatembolism: a case report. Intensive Care Med 1995;21(8):679e81.

14. Scopa M, Magatti M, Rossitto P. Neurologic symptoms in fat embolismsyndrome: case report. J Trauma 1994;36(6):906e8.

15. Johansson MC, Eriksson P, Dellborg M. The significance of patent foramenovale: a current review of associated conditions and treatment. Int J Cardiol2009 May 1;134(1):17e24.

16. Patel R, Stygall J, Harrington J, Newman S, Haddad F. Intra-operative cerebralmicroembolisation during primary hybrid total hip arthroplasty comparedwith primary hip resurfacing. Acta Orthop Belg 2009 Oct;75(5):671e7.

17. Thienpont E, Kaddar S, Morrison S. Paradoxical fat embolism after unce-mented total hip arthroplasty: a case report. Acta Orthop Belg 2007 Jun;73(3):418e20.

18. Forteza AM, Rabinstein A, Koch S, Zych G, Chandar J, Romano JG, et al. Endo-vascular closure of a patent foramen ovale in the fat embolism syndrome:changes in the embolic patterns as detected by transcranial Doppler. ArchNeurol 2002 Mar;59(3):455e9.

19. Byrick RJ, Mullen JB, Mazer CD, Guest CB. Transpulmonary systemic fatembolism. Studies in mongrel dogs after cemented arthroplasty. Am J RespirCrit Care Med 1994;150(5 Pt 1):1416e22.

20. Lechapt E, Habibi A, Bachir D, Galacteros F, Schaeffer A, Desvaux D, et al.Induced sputum versus bronchoalveolar lavage during acute chest syndromein sickle cell disease. Am J Respir Crit Care Med 2003 Dec 1;168(11):1373e7.

21. Graham JK, Mosunjac M, Hanzlick RL. Sickle cell lung disease and suddendeath: a retrospective/prospective study of 21 autopsy cases and literaturereview. Am J Forensic Med Pathol 2007 Jun;28(2):168e72.

22. Dang NC, Johnson C, Eslami-Farsani M, Haywood LJ. Bone marrow embolism insickle cell disease: a review. Am J Hematol 2005 May;79(1):61e7.

23. Gurd AR. Fat embolism: an aid to diagnosis. J Bone Joint Surg Br. 1970 Nov;52(4):732e7.

24. Lindeque BG, Schoeman HS, Dommisse GF, Boeyens MC, Vlok AL. Fat embolismand the fat embolism syndrome. A double-blind therapeutic study. J Bone JointSurg Br. 1987 Jan;69(1):128e31.

25. Bulger EM, Smith DG, Maier RV, Jurkovich GJ. Fat embolism syndrome. A 10-year review. Arch-Surg 1997;132(4):435e9.

26. ten Duis HJ. The fat embolism syndrome. Injury 1997;28(2):77e85.27. Arthurs MH, Morgan OS, Sivapragasam S. Fat embolism syndrome following

long bone fractures. West-Indian-Med J 1993;42(3):115e7.28. Metting Z, Rodiger LA, Regtien JG, van der Naalt J. Delayed coma in head injury:

consider cerebral fat embolism. Clin Neurol Neurosurg 2009 Sep;111(7):597e600.

29. Walshe CM, Cooper JD, Kossmann T, Hayes I, Iles L. Cerebral fat embolismsyndrome causing brain death after long-bone fractures and acetazolamidetherapy. Crit Care Resusc 2007 Jun;9(2):184e6.

30. Ott MC, Meschia JF, Mackey DC, Brodersen MP, Burger C, Echols JD, et al.Cerebral embolization presenting as delayed, severe obtundation in the post-anesthesia care unit after total hip arthroplasty. Mayo Clin Proc. 2000 Nov;75(11):1209e13.

31. Jacobs S, Al Thagafi MY, Biary N, Hasan HA, Sofi MA, Zuleika M. Neurologicalfailure in a patient with fat embolism demonstrating no lung dysfunction[letter]. Intensive Care Med 1996;22(12):1461.

32. Jacobson DM, Terrence CF, Reinmuth OM. The neurologic manifestations of fatembolism. Neurology 1986 Jun;36(6):847e51.

33. Aravapalli A, Fox J, Lazaridis C. Cerebral fat embolism and the “starfield”pattern: a case report. Cases J 2009;2:212.

34. Eguia P, Medina A, Garcia-Monco JC, Martin V, Monton FI. The value of diffu-sion-weighted MRI in the diagnosis of cerebral fat embolism. J Neuroimaging2007 Jan;17(1):78e80.

35. Lee J. Gradient-echo MRI in defining the severity of cerebral fat embolism. J ClinNeurol 2008 Dec;4(4):164e6.

36. Ruiz-Gimeno JI, Ferre MA, Napal MT, Pelegrin F. Prolonged coma due to fatembolism syndrome after fracture of the femur. Rev Esp Anestesiol Reanim 2006Mar;53(3):187e90.

37. Gratadour P, Vedrinne JM, Guillaume C, Gagnieu MC, Motin J. Intra-macro-phage lipid particles collected by bronchoalveolar lavage: incidence anddiagnostic value. Ann Fr Anesth Reanim 1993;12(5):462e8.

38. Jules-Elysee KM, Yadeau JT, Urban MK. Pulmonary artery versus central venouscatheter monitoring in the outcome of patients undergoing bilateral total kneereplacement. HSS J 2009 Feb;5(1):27e30.

39. Van den Brande FG, Hellemans S, De Schepper A, De Paep R, Op De Beeck B, DeRaeve HR, et al. Post-traumatic severe fat embolism syndrome with uncommonCT findings. Anaesth Intensive Care 2006 Feb;34(1):102e6.

40. Cheatham ML, Block EF, Nelson LD. Evaluation of acute mental status change inthe nonhead injured trauma patient. Am Surg 1998 Sep;64(9):900e5.

41. Butteriss DJ, Mahad D, Soh C, Walls T, Weir D, Birchall D. Reversible cytotoxiccerebral edema in cerebral fat embolism. AJNR Am J Neuroradiol 2006 Mar;27(3):620e3.

42. Chen JJ, Ha JC, Mirvis SE. MR imaging of the brain in fat embolism syndrome.Emerg Radiol 2008 May;15(3):187e92.

43. Pape HC, Giannoudis P, Krettek C. The timing of fracture treatment in poly-trauma patients: relevance of damage control orthopedic surgery. Am J Surg2002 Jun;183(6):622e9.

44. Behrman SW, Fabian TC, Kudsk KA, Taylor JC. Improved outcome with femurfractures: early vs. delayed fixation. J Trauma 1990 Jul;30(7):792e7. discussion7e8.

45. Shaikh N, Parchani A, Bhat V, Kattren MA. Fat embolism syndrome: clinical andimaging considerations: case report and review of literature. Indian J Crit CareMed 2008 Jan;12(1):32e6.

46. Habashi NM, Andrews PL, Scalea TM. Therapeutic aspects of fat embolismsyndrome. Injury 2006 Oct;37(Suppl. 4):S68eS73.

47. Schemitsch EH, Jain R, Turchin DC, Mullen JB, Byrick RJ, Anderson GI, et al.Pulmonary effects of fixation of a fracture with a plate compared with intra-medullary nailing. A canine model of fat embolism and fracture fixation. J BoneJoint Surg Am. 1997;79(7):984e96.

48. Neudeck F, Wozasek GE, Obertacke U, Thurnher M, Schlag G. Nailing versusplating in thoracic trauma: an experimental study in sheep. J Trauma 1996;40(6):980e4.

49. Schemitsch EH, Turchin DC, Anderson GI, Byrick RJ, Mullen JB, Richards RR.Pulmonary and systemic fat embolization after medullary canal pressurization:a hemodynamic and histologic investigation in the dog. J Trauma 1998 Oct;45(4):738e42.

50. Muller C, McIff T, Rahn BA, Pfister U, Weller S. Intramedullary pressure, strainon the diaphysis and increase in cortical temperature when reaming thefemoral medullary cavityea comparison of blunt and sharp reamers. Injury1993;24(3):S22eS30.

51. Gleitz M, Hopf T, Hess T. Experimental studies on the role of intramedullaryalignment rods in the etiology of fat embolisms in knee endoprostheses. ZOrthop Ihre Grenzgeb 1996;134(3):254e9.

52. Woo R, Minster GJ, Fitzgerald Jr RH, Mason LD, Lucas DR, Smith FE. The FrankStinchfield Award. Pulmonary fat embolism in revision hip arthroplasty. ClinOrthop 1995;319:41e53.

53. Bederman SS, Bhandari M, McKee MD, Schemitsch EH. Do corticosteroidsreduce the risk of fat embolism syndrome in patients with long-bone frac-tures? A meta-analysis. Can J Surg 2009 Oct;52(5):386e93.

54. Gossling HR, Ellison LH, Degraff Jr AC. Fat embolism. the role of respiratoryfailure and its treatment. J Bone Joint Surg Am. 1974 Oct;56(7):1327e37.