Boezaart_CH05.

qxd 4/26/06 12:37 PM Page 51

CHAPTER 5
Fat Embolism
ANDRE R. COETZEE

 INTRODUCTION emboli.7 Long bone fractures, of the lower extremities in

particular, seem to cause fat emboli.8,9
Fat embolism (FE) is the partial obstruction of multiple However, the presence of fat emboli is not synony-
blood vessels by fat globules.1 This is usually a tempo- mous with FES and, because of under reporting, differ-
rary phenomenon. The fat embolism syndrome (FES) is ences in diagnostic criteria, and studies in different popu-
the clinical manifestation of FE. lations, the reported incidence of FES varies from 0.25 to
29 percent.10–12 Perioperative FES, after immediate fixation
of fractures, has a reported incidence of 0 to 5 percent.9,13
 HISTORY A report on more than 900 cases of surgery for lower
extremity fracture gave an incidence of 0.2 percent.3
In initial experiments, milk injected into the circulation
of dogs produced FE. This was followed by oil-injection
studies in the nineteenth century.2,3 However, the  PATHOLOGY
pathology and clinical picture were appreciated only in
1862, when fat was detected in the lungs of trauma The etiology of FE and FES is classified as traumatic and
patients.2,3 Zenker postulated that the fat in the lung was nontraumatic.
aspirated, but Wagner showed that fractures of bones
released fat into the circulation.2,3 Traumatic
The clinical manifestation of FE, that is, FES, was The essential pathology is the entrance of marrow fat
described in 1873 by Bergman,3 and 350 cases were into veins. This process is enhanced by
described in 1913.2 In a review, Lehman and Moore in
19274 postulated a metabolic mechanism to explain the Multiple fractures
FES, while Sevitt drew a distinction between systemic Unstable fractures
and pulmonary fat emboli.1 Peltier concluded that the Increased pressure in the medullary cavity
morbidity associated with FE was caused by fatty acids Shock, hypoxia, and stress, although this will not
and associated inflammation.5 cause FES by itself 14
The attachment of veins to bone prevents their col-
lapse; even during episodes of low venous pressure,
 INCIDENCE such as circulatory shock.15 Fractures of bones with a
higher density of vessels, such as the isthmus of the
Fat emboli in the lungs of patients with skeletal trauma proximal femoral shaft, are more likely to cause the
are common.6 Analysis of blood taken from the pul- release of fat into the circulation.9
monary artery of patients with long bone or pelvic frac- Mechanics dictate that if the medullary pressure
tures demonstrated that 7 percent of patients had fat is raised when veins are torn, more fat will enter the
Boezaart_CH05.qxd 4/26/06 12:37 PM Page 52

52 PART I GENERAL PRINCIPLES

circulation. Hence closed fractures result in more fat fractures.27 Hypoxia was less common in patients with
emboli.9,16 Reaming of medullary spaces during surgery decompressed fractures.9
and nailing of prostheses (which raises the medullary Neutral fat (not free fatty acids) can obstruct the pul-
pressure to 500 mmHg) tend to cause more fat monary arterial system, causing acute pulmonary artery
emboli.17,18 Venting of the medullary cavity before insert- hypertension and acute right ventricular failure.28–31 The
ing instruments therefore decreases the incidence of fat resultant decrease in cardiac output together with
emboli.19 Young men are more prone than others to inflammatory changes in the lung will worsen hypoxia.
develop FES.20 Obstruction of capillaries is not restricted to the
Burns and soft tissue injury can also result in FES,21 but lung but occurs in the myocardium, brain, kidney, and
it is less common, even though fat emboli are common. other organs when fat enters the systemic circulation.
The mechanisms by which fat enters the systemic circu-
Nontraumatic lation are either through a patent foramen ovale, which
opens once the right atrial pressure is raised (because of
These are rare causes of FES and they are classified as
acute pulmonary hypertension associated with acute
follows3:
lung injury or mechanical obstruction), or through the
pulmonary capillaries. Patent foramina ovale were, how-
Procedure-Related
ever, not found in many patients with FES, and this sup-
This group includes intraosseous fluid administration,
ports the capillary transit theory.32 Another way in which
intraoperative autotransfusion, and lipid-soluble radio-
fat may enter the systemic circulation is via pulmonary-
contrast.
bronchial anastomoses if the obstruction of the pul-
monary bed results in reverse flow.
Disease-Related
Chylomicrons are small (1 µm in diameter) but, under
Sickle cell hemoglobin C, pancreatitis, fat necrosis of the
the influence of mediators, coalesce into 10- to 40-µm fat
omentum, and immunosuppressive disease are typical
globules,4 which are thought to obstruct the capillaries in
examples.
the lung and the systemic capillary bed.33 One study
demonstrated a significant increase in cholesterol in fat
Drug-Related
obtained from the pulmonary capillaries of patients with
Intravenous hyperalimentation, intraarterial cisplatinum,
FES,34 but this observation has not been confirmed.
and long-term steroid administration are included here.

Biochemical Hypothesis
 PATHOPHYSIOLOGY
This hypothesis proposes that the fat globules in the
systemic and pulmonary circulation originate from fat
Three main hypotheses have been put forward to
normally present in the blood, presumably altered by
explain FES:
physical and more probably biochemical factors, result-
Mechanical hypothesis ing in either toxic or obstructive pathology.33
Biochemical hypothesis The toxic hypothesis suggests that fat is broken
A combination of mechanical and biochemical down to free fatty acids (FFA), which cause injury to cap-
mechanisms illaries and pneumocytes. The resultant clinical picture is
similar to that of acute lung injury with hemorrhagic
lung injury to the pulmonary interstitium, edema, and
Mechanical Hypothesis
pneumonitis.5,35,36 The effects of FFA on the pulmonary
This hypothesis proposes that the embolized fat causes tissue have been confirmed and there is a correlation
mechanical obstruction of the pulmonary and systemic between serum FFA levels and hypoxia.37,38 However,
capillaries. Fat has been demonstrated in the circulation this association does not necessarily imply causation. In
soon after injury,22 and fat in the lungs of animals has addition, there is uncertainty about the origin of the FFA,
been shown to originate from the bone marrow.22–24 with one group postulating that it is formed in the lung
Experiments in rabbits have demonstrated that raising by the activity of pulmonary lipase on fat, while others
the intramedullary pressure at a fracture site increased postulate that it originates from the body fat.5,35,36
the extravasation of marrow fat.23 In humans, significant Trauma and catecholamines do result in FFA release,
amounts of fat have been found in the ipsilateral femoral but whether this is important in the development of FES
vein of a fractured femur during surgery.25,26 is uncertain.39,40 If one considers the number of surgical
The lung capillaries are prominently affected in FE, procedures performed and/or trauma cases managed
and there appears to be a relationship between the annually, the relatively low incidence of FES argues
decrease in arterial PaO2 in FES and compressed closed against this theory as a sole explanation for FES. Some
Boezaart_CH05.qxd 4/26/06 12:37 PM Page 53

CHAPTER 5 FAT EMBOLISM 53

studies have suggested that the serum albumin neutral- temperature (>37.8oC; 100.4oF) and tachycardia (>100
izes the effect of FFA on the systemic circulation.41–43 beats per minute) occur in the majority of patients with
There is also a suggestion that FFA is raised more in the subacute and mild form of FES.50 Approximately 50
patients with decreased liver function. This hypothesis to 90 percent of patients with lower extremity fractures
has some credibility, since shock, hypoxia, and anesthe- suffer from some degree of hypoxia (defined as a PaO2
sia decrease liver blood flow. Under these circumstances, <80 mmHg; 10.6 kPa),50–52 and FES is viewed as the most
together with raised catecholamines, FE may well become likely explanation for the recorded desaturation of the
a clinical entity. arterial blood. Increased alveolar ventilation, resulting in
various degrees of hypocapnia, also occurs in patients
Combined Mechanical and with subclinical FES.48 The latter study demonstrated a
Biochemical Hypothesis 48 percent incidence of thrombocytopenia (platelet
count less than 200,000/µL), which has been regarded as
Because mechanical and biochemical hypotheses sepa-
an important indicator for FES48 and suggested to be
rately do not appear to completely explain the FES, a
linked to the lung injury in FES. However, some of the
combination of the two has been proposed.5,34 The fat
conclusions in this regard rest on indirect evidence
embolization of the capillaries by marrow is the initiating
(AaDo2) and are uncertain, since the AaDo2 is affected
event and thereafter biochemical factors come into play.
by numerous factors and does not correlate well with the
calculated shunt (Qs/Qt).53
Other Theories
An echocardiographic study showed marrow particles
Neutral fat has a thromboplastic effect, which activates in the right atrium in 60 percent of patients in whom
the clotting cascade and could lead to disseminated bones had been manipulated during surgery.54
intravascular coagulation (DIC) and eventually depletion
of clotting factors. The end result is spontaneous bleed-
ing.44 This tendency, coupled with the suppression of
Nonfulminant Form (Subacute
the fibrinolytic system by trauma, results in the accumu-
Form of FES)
lation of red blood cells, platelets, leukocytes, and fibrin This is the more classic and well-known mode of
in the microvascular bed, including the lung.45–47 The presentation.
aggregates and clots will be filtered in part by the lung, Patients often have petechial rashes on the anterior
and this contributes to the clinical picture of pulmonary upper portion of the chest, especially in the areas adja-
parenchymal dysfunction often associated with FES. An cent to the axillae. Other areas commonly affected are
inverse relationship has been demonstrated between the shoulders, mucous membranes (mouth), and con-
platelet count and the degree of pulmonary dysfunction junctiva.35,41,55 These rashes usually appear 12 to 96 h
in these patients.48 after the injury and may last from hours to a week. It is
DIC is not always associated with FES; in the important that an active search be made for these telltale
author’s practice, however, it does seem to occur regu- signs of FES. It has been suggested that the upper areas
larly, albeit more often as subtle DIC patterns. of the body are more affected because fat is lighter than
Another hypothesis is that trauma depresses the blood, but this is at best an uncertain explanation.
immune system, but the evidence for this as a cause of Histology of the skin lesions shows occlusion and dis-
the FES is scanty.49 tention of the capillaries with fat globules and increased
permeability of the capillaries.56
Neurologic signs are present in up to 86 percent of
patients with FES.55,57 These vary in severity and include
 CLINICAL PRESENTATION
confusion, stupor, coma, and even decerebrate rigidity.
Confusion is the more common presenting symptom and
FES usually presents in one of three degrees of severity3:
convulsions have been reported. Signs are nonlateraliz-
Subclinical ing and are usually progressive, starting with confusion
Nonfulminant subacute and progressing to more severe encephalopathy in many
Fulminant patients. Localizing signs such as hemiplegia, apraxia,
visual scotomata, and conjugate eye deviation occur but
are rare.57 The pathophysiology of neurologic dysfunc-
Subclinical Form
tion is thought to be the direct result not of cerebral
The clinical signs are nonspecific and often confused with hypoxia but rather of embolization. Cerebral edema may
postoperative symptoms ascribed to pain, discomfort, or be induced by the FFA causing vascular disruption.10,58
the postoperative inflammatory process. Tachypnea The respiratory system is often also involved in FES,
(defined as a breathing rate in excess of 25/min), elevated and this syndrome is among the causes of acute lung
Boezaart_CH05.qxd 4/26/06 12:37 PM Page 54

54 PART I GENERAL PRINCIPLES

injury (ALI), the severe form of which is called acute res- The lungs are affected in various ways: lung edema
piratory distress syndrome (ARDS). The clinical picture (even in the upper lung lobes) as well as reduced com-
includes tachypnea, cyanosis or some degree of arterial pliance and poor gas transfer resulting in hypoxia.
hemoglobin desaturation, and a respiratory alkalosis. Involvement of the upper lobe is explained by the
The pathophysiology of the acute lung injury is that hypothesis that fat is lighter than blood and hence will
of poor pulmonary compliance resulting in tachypnea, affect the upper zones. What argues against this idea is
low ventilation/perfusion ratios, either Qva/Qt or Qs/Qt that the toxic effects of FFA damage the endothelium of
causing hypoxia, and an increased Vd/Vt. This is accom- the vasculature; hence areas with higher perfusion, such
panied by various degrees of pulmonary hypertension, as the dependent parts of the lung, will be affected more.
which jeopardize right ventricular function and cardiac Another explanation for the upper lobe edema is the
output, resulting in mixed venous blood desaturation. The acute pulmonary hypertension associated with acute
latter will be aggravated by the raised temperature that lung injury. This will increase perfusion of the upper
commonly occurs in patients with FES. The combination lobes and may highlight the edema seen in those areas
of mixed venous blood desaturation and pulmonary shunt of the lung that are normally less well perfused (because
will further depress the arterial oxygenation. of gravity). The picture can progress to a full-blown
The pulmonary effects of FES usually have a rela- ARDS with all the clinical signs and symptoms of severe
tively benign course, and by the third day—barring new acute lung injury requiring respiratory support.
fat emboli or infection—the lung parenchyma starts to The right heart is severely affected by acute increases
improve rapidly. in the afterload. Acute lung injury is virtually always
There are reports of patients who had FES without associated with ARDS, and data show that by the time
pulmonary involvement.59–61 The retina is affected in the patient is admitted to the intensive care unit with a
approximately 50 percent of patients.62 In 4 percent of diagnosis of ARDS, acute pulmonary hypertension is
patients with long bone fractures, retinopathy presents already present.66 The patient will have tachycardia,
subclinically.63 Eye lesions, consisting of cotton-wool raised jugular venous pressure, and raised right ventric-
spots and flame-like hemorrhages, are indistinguishable ular end-diastolic pressure. Acute cor pulmonale results
from the Purtscher retinopathy described in extraocular in systemic hypotension, intraventricular septal shift, and
trauma.63,64 The cotton-wool areas are microinfarcts in low cardiac output. The arterial pressure will swing with
the nerve fiber layer of the eye63 and the pathophysiology the ventilator (pulsus paradoxus) because of the low left
is the same as in other organs—i.e., capillary engorge- ventricular end-diastolic volume; but because of the sep-
ment and loss of endothelial integrity. This results in tal shift, the end-diastolic pressure in the left ventricle is
hemorrhage, edema, and microinfarction. The lesions raised, and this will aggravate the tendency to lung
disappear in weeks but permanent central scotomata edema. Left ventricular ischemia occurs because of the
have been described. low arterial pressure, and the risk of ischemic right and
left ventricular failure is significant.67 The resultant low
cardiac output, especially in the presence of fever,
Fulminant FES
results in low mixed venous hemoglobin oxygen satura-
This form of FES contains all the elements of nonfulmi- tion. This, together with the shunt or low V/Q lung units,
nant FES but is more severe. It has been described after aggravates arterial hypoxia, which cannot be corrected
the release of limb tourniquets, reaming of the femur, with mechanical ventilation or increased inspiratory frac-
insertion of intramedullary prostheses, closed reductions tion of oxygen. Angina has been reported, which is pre-
of fractures, and manipulation of a deformed hip. sumed to result from emboli in the coronary arteries or
However, it can also occur in the absence of surgery and cor pulmonale.68
several hours after surgery.28,58,65 In the anesthetized Acute renal failure occurs relatively seldom and it is
patient, it may present as sudden hypotension, wheezing, thought that less neutral fat emboli enter the kidneys
tachycardia or bradycardia, poor lung compliance, pul- than other organs. In acute circulatory failure, as in cor
monary edema, cyanosis, or unexplained bleeding. pulmonale, the neurohumoral response of the body sig-
Failure to regain consciousness after orthopaedic surgery nificantly decreases the renal blood flow, decreasing the
and anesthesia should raise the possibility of cerebral FES. number of renal emboli.69
In the unanesthetized patient, the signs and symp- We have observed acute and severe diffuse
toms of the nonfulminant type occur, including more intravascular coagulation and clotting disorder (DIC)
severe cerebral symptoms such as agitation, choreoa- during surgery on long bones. This occurred without
thetosis, seizures, psychosis, or coma. These may pre- other prodromata of the FES. The patient started to bleed
cede fever, acute respiratory distress, significant hemo- spontaneously, and special examinations showed raised
dynamic instability, oliguria, renal failure, and clotting D-dimers, low fibrinogen and platelets, raised interna-
disorders.3 tional normalized ratio (INR) and partial thromboplastin
Boezaart_CH05.qxd 4/26/06 12:37 PM Page 55

CHAPTER 5 FAT EMBOLISM 55

time (PTT), and excessive levels of fibrinogen break- A progressive fall in hemoglobin has been reported,76
down products. Some degree of clotting disorder is, in with a 30 percent decrease over 48 h in 75 percent of
the author’s experience, very common in patients with patients with FES. This is thought to be the result of
the fulminant form of FES. intraalveolar hemorrhage.5 However, this hypothesis is
uncertain, since a similar aggressive decrease in hemoglo-
bin does not regularly occur in patients with other causes
 SPECIAL EXAMINATIONS of ARDS.
Cytologic examination of urine and sputum could be
The FES is a syndrome and the diagnosis is therefore pri- useful in confirming the clinical diagnosis. Sudan red
marily a clinical one. Special examinations can help to staining will detect free fat globules in the macrophages.
confirm it and indicate the degree of organ dysfunction. This procedure takes about 2 to 3 h, but its value is ques-
The arterial blood gases, more often than not, show tionable, since it has given negative results in a group of
a low PaO2, and this is often described as a diagnostic patients with a positive diagnosis of FES.77 This highlights
criterion. However, there are many other reasons for a the point that FES is primarily diagnosed clinically, and
low PaO2, and the absence of a reduction in arterial the value of confirmatory tests is subject to their specificity
oxygen tension does not exclude the diagnosis.13,60,61,70 and sensitivity.
Hypoxia is usually accompanied by a respiratory alkalosis, The cryostat test can be used to identify fat globules
but once mechanical respiratory failure occurs, the PaCO2 in blood. Peripheral blood is rapidly frozen and the clot
increases because of alveolar hypoventilation, which sectioned for microscopic examination.50,78 This test has
results from respiratory muscle fatigue. a limitation in that most of the fat will be found in the
X-rays of the lungs are nonspecific and lag behind veins draining the source area and not in the peripheral
the clinical syndrome. In the presence of FES, the typical blood. Because the fat globules are trapped in the pul-
x-ray findings are useful and help to confirm the diagnosis. monary capillaries, a blood sample from a pulmonary
The typical pattern is that of a “snowstorm.”71–73 Bilateral artery catheter in a wedged position would increase pos-
patchy infiltrates occur mainly in the perihilar and basilar itive results for the test.49,78 Bronchoalveolar lavage
areas, usually sparing the apices. This classic picture can be (BAL) may contain cells with fat droplets.79 This was seen
modified by pulmonary edema, which can occur because in 31 percent of patients with FES but in only 2 percent of
of the septal shift associated with acute pulmonary those without FES.
hypertension and the resultant raised left atrial pressure Blood lipids do not correlate with the severity of
or primary (usually ischemic) left ventricular failure. FES.10 Activated complement (C5a) increases in FES, but
Dilatation of the right heart (acute cor pulmonale) this increase is not specific for this syndrome.77
can sometimes be seen.
Computed tomography (CT) scans of the lungs show
the subsegmental perfusion defects and ventilation/  DIAGNOSIS
perfusion scans confirm the ventilation/perfusion mis-
match. Indeed, CT and V/Q scan findings can be diagnostic The only pathognomic sign for fat embolism is the
when radiologic findings are negative.74 petechiae. Other clinical signs and symptoms and most
A CT scan of the brain may show the absence of of the readily available special examinations are nonspe-
cortical sulci and compressed lateral ventricles due to cific. However, FES is a syndrome, and the diagnosis
brain edema. Hemorrhagic brain infarcts have been rests on a constellation of findings. If all the signs and
recorded.58 symptoms are present, the diagnosis is usually not diffi-
Acute cor pulmonale patterns and/or myocardial cult; but if some of them are lacking, the diagnosis may
ischemia may be revealed by an electrocardiogram (ECG), be uncertain.
depending on the severity of the FES and its effects on the For patients with lower body skeletal trauma, Gurd
pulmonary and coronary circulation. and Wilson10 have compiled the following diagnostic
Coagulation disorders, although not specific for criteria:
FES, require quantification for effective treatment, since
One of the major manifestations; i.e., petechiae and
the INR, PTT, fibrinogen, D-dimer, fibrinogen breakdown
pulmonary or cerebral involvement
products, and platelet count will identify either acceler-
Four of the five minor manifestations; i.e., pyrexia,
ated clotting and/or clot lysis.75 Fibrinogen, an acute-
tachycardia, jaundice, renal or retinal changes,
phase protein that should be increased because of the
and fat macroglobulinemia
trauma and stress, is often decreased or may be normal
because of the activation of the coagulation system. As The approach of Gurd and Wilson is limited by the
the patient improves, the fibrinogen usually increases absence of arterial blood gas findings and other lung
(within 3 to 5 days). pathology descriptors. Lindeque70 therefore expanded
Boezaart_CH05.qxd 4/26/06 12:37 PM Page 56

56 PART I GENERAL PRINCIPLES

on the Gurd and Wilson approach by including the be gauged from the central venous pressure response to
following: a small fluid load, abnormal a or c waves, and the
appearance of acute tricuspid incompetence. In severe
PaO2 < 60 mmHg (8 kPa) on a FiO2 of 0.21
FES, a pulmonary artery catheter is necessary to deter-
PaCO2 > 55 mmHg (7.3 kPa) or a pH < 7.3
mine right ventricular failure. The latter is diagnosed by
Respiratory rate >35 breaths per minute (even after
critically evaluating the right ventricular stroke work
sedation)
(RVSW) against the pulmonary artery pressure. If RVSW
Increased work of breathing as manifest by dyspnea,
remains normal or does not increase as the pulmonary
use of accessory muscles, and a tachycardia
artery pressure increases, right ventricular failure is
As can be seen from the above, the criteria for diag- occurring. This implies that the function of the right ven-
nosis are fairly nonspecific except for petechiae. Even tricle changes from maximal efficiency to maximum
the inclusion of the lung descriptors does not make criteria stroke work. This is indicated by lower than expected
more specific, since the latter could signify lung injury RVSW in the presence of a raised pulmonary artery pres-
without FES.3 sure. It is mandatory to support right ventricular function
Based on individual organ dysfunction, which can with inotropes and to maintain the systemic pressure so
occur in FES, there is an extensive list of differential as to prevent hypotension and right ventricular ischemic
diagnoses that should be considered.3 However, the key failure. The choice of inotrope depends on the systemic
issues in diagnosing the syndrome remain a high degree blood pressure: if the perfusion pressure is still normal,
of suspicion, petechiae and other organ involvement. dobutamine is useful. If the perfusion pressure is low, an
The timing of the clinical signs may also assist in the alpha agonist is required and epinephrine or norepi-
diagnosis. For example, the development of petechiae nephrine is a more appropriate choice. Pulmonary
soon after bone manipulation or injury would very likely vasodilators are singularly unsuccessful, and since drugs
suggest FES; on the other hand, if it occurs after 3 days like isoproterenol will cause systemic hypotension and
and after stabilization of the fractures, the diagnosis is increase the pulmonary shunt, they should not be used.
less likely (although not excluded). The same can be said for all of the other generally avail-
The clinical picture may differ in patients under able nonspecific vasodilators. The role of inhaled nitric
general anesthesia as well as muscle relaxation and oxide has not been established, and in view of the
mechanical ventilation, since the telltale cerebral and mechanical theory of obstruction, it is doubtful that it
pulmonary signs can obviously not be observed in them. would be helpful.80 If there is significant pulmonary
vasoconstriction in addition to the capillary obstruction,
then there would be reason to expect that nitric oxide
 TREATMENT could help in unloading the right ventricle (and improving
arterial oxygenation).81
There is no specific treatment for FES; the only option is Studies indicate that albumin should be used for
to continue life support until the patient recovers. volume restoration because, in addition to this, it binds the
The management of shock is very important in the FFA and thereby reduces the potential for lung injury.82,83
outcome of FES. Data indicate that mortality was higher Excessive fluid will, as in all cases of acute lung injury (ALI),
in patients with systolic blood pressure <100 mmHg and increase the amount of interstitial fluid in the lungs and
a heart rate in excess of 120 beats per minute.53 Because could worsen the venous admixture or pulmonary shunt.
this was a retrospective study and it is not ethical to do a Ventilatory support is often mandatory to restore
trial in which hypovolemia and incipient shock is not arterial oxygenation. The decision to support oxygenation
treated, it is not known whether the increased mortality and ventilation rests on known criteria. These maneuvers
was due to FES and incipient shock or hypovolemia and include additional inspired oxygen, noninvasive and inva-
shock alone. It was most likely the latter. Hemodynamic sive ventilation. Given that the basic pulmonary pathology
shock alone can cause acute lung injury, and this could is that of low ventilation/perfusion units resulting from a
well be aggravated by FES. In addition, hypovolemia in reduction in the functional residual capacity, the judi-
the presence of acute pulmonary hypertension could cious application of positive end-expiratory pressure
have a significant effect on the cardiac output. with or without lung recruitment is valuable. The man-
Excessive fluid loading is detrimental to the thin- agement of mechanical ventilation and alveolar pressure
walled right ventricle in the presence of an acute increase is guided by the current knowledge of volume trauma,
in afterload. Intravascular fluid restoration should there- and protective ventilatory strategies should be employed
fore be done cautiously and monitored closely. A central when dealing with these patients.84,85
venous catheter is valuable, since it will give indirect Of all the causes for acute lung injury, fat embolism is
guidance as to the volume status and also indicate one of the more benign etiologies. The acute ventilatory
whether the right ventricle is under strain. The latter can difficulty usually lasts for approximately 3 days, after
Boezaart_CH05.qxd 4/26/06 12:37 PM Page 57

CHAPTER 5 FAT EMBOLISM 57

which the patient often improves and can be weaned The only drug that seems to have some (still debat-
from ventilatory support. Unfortunately associated able) value is corticosteroid. There is great variation in the
problems—such as acute renal failure, right ventricular suggested dose, from 9 mg/kg over 2 days to 90 mg/kg
failure, or pulmonary infection—can intervene and will over 4 days.77,89
obviously modify the course of the disease.
Although cardiac and pulmonary fat embolectomy
has been performed, it is of little value, since the emboli  PROGNOSIS
are in the capillaries.86
High doses of corticosteroids have not been shown The general mortality rate associated with FES is 10 to
to be of value in ALI.87 Because FES is often an isolated 20 percent.6,69,98 However, it appears that early fixation
problem and generally has a less malignant course than, of fractures reduces the mortality to less than 10 percent
for instance, ALI caused by sepsis, it has nevertheless (see Chap. 16).9
been suggested that steroids, given early in the course of There are, however, publications suggesting higher
the disease, could be beneficial.76,88–91 and lower mortalities. Whereas Burgher99 reported a
Aprotinin has been administered because it mortality of 33 percent, Guenter 100 reported no deaths in
decreases platelet aggregation and the release of sero- 54 patients.
tonin. A retrospective study demonstrated an increased Death is more often than not the result of respiratory
recovery and decreased mortality when aprotinin was insufficiency, but it is difficult to collect accurate data,
used in FES.10 Another study confirmed that aprotinin since there may be many intervening problems that can
will lessen the reduction in platelet count but does not be fatal.
prevent the pulmonary effects of FES.51 Permanent cerebral injury can occur, especially in
For the prevention and treatment of diffuse intravas- young patients with the fulminant variant of the disease.
cular clotting disorder, heparin should be used. For
prevention 300 U of unfractionated heparin per hour
administered intravenously was successful in preventing
REFERENCES
thrombin-induced DIC in baboons.92 For treatment of
1. Sevitt S. The significance and pathology of fat embolism.
established DIC, it is mandatory to know whether the
Ann Clin Res 9:173, 1977.
DIC is in the accelerated clotting phase, clot-lysis phase, 2. Warthin AS. Traumatic lipemia and fatty embolism. Int
or both. This can be achieved by means of a complete Clin 4:171, 1913.
coagulation scan. Heparin dosage can then be titrated 3. Capan LM, Miller SM, Patel KP. Fat embolism. Anesthesiol
against the pathology, and once the accelerated clotting Clin North Am 11:25, 1993.
is controlled, clotting factors can be given to replace 4. Lehman EP, Moore RM. Fat embolism: Including experi-
those used in the process. mental production without trauma. Arch Surg 14:621, 1927.
There is a theoretical argument that heparin may be 5. Peltier LF. Fat embolism III. The toxic properties of neu-
detrimental because it raises the FFA significantly. tral fat and free fatty acids. Surgery 40:665, 1956.
However, the need to prevent or treat DIC in these patients, 6. Peltier LF. Fat embolism syndrome: A current concept.
outweighs this possible negative effect, and the author Clin Orthop 241:241, 1969.
7. Lozman J, Deno C, Feustel PJ, et al. Pulmonary and cardio-
elects to use heparin as described.
vascular consequences of immediate fixation or conserva-
tive management of long bone fractures. Arch Surg 121:992,
1986.
 PREVENTION 8. Duis HJT, Nijsten MW, Klasen JH, et al. Fat embolism in
patients with an isolated fracture of the femoral shaft. J
Early stabilization of fractures is important in an effort to Trauma 28:383, 1988.
prevent the FES; this will reduce its incidence and 9. Fabian TC, Hoots AV, Stanford DS, et al. Fat embolism syn-
decrease the incidence of septic and pulmonary compli- drome: Prospective evaluation in 92 fracture patients. Crit
cations (see Chap. 16).7,13,93–97 Care Med 18:42, 1990.
Intraoperative drilling of a venting hole before insert- 10. Gurd AR, Wilson RI. The fat embolism syndrome. J Bone
ing a prosthesis reduces the intramedullary pressure but Joint Surg 56(B):408, 1974.
11. De Vries J, Oosterhuis JW, Oldhoff J. Bone marrow
does not prevent the FES in every patient.
embolism following cryosurgery of bone. An experimen-
The use of a tourniquet has been proposed as a tal study. J Surg Res 46:200, 1989.
method of preventing the fat from reaching the systemic 12. Moore P, James O, Saltos N. Fat embolism syndrome.
circulation. However, once the tourniquet is released, Incidence, significance and early features. Austr NZ J Surg
the fat will still reach the systemic circulation. In addi- 51:546, 1981.
tion, the accumulated fat could cause severe FES; this 13. Riska EB, Myllynen P. Fat embolism in patients with mul-
method is therefore not recommended.28,35 tiple injuries. J Trauma 22:891, 1982.
Boezaart_CH05.qxd 4/26/06 12:37 PM Page 58

58 PART I GENERAL PRINCIPLES

14. Cotev S, Rosenmann E, Eyal Z. The role of hypovolemic 36. Peltier LF: The diagnosis of fat embolism. Surg Gynecol
stress in the production of fat embolism in rabbits. Obstet 121:371, 1965.
Morphologic alterations in the lung. Chest 69:523, 1976. 37. Parker FB, Wax SD, Kutsajima K, et al. Hemodynamic and
15. Orlowski JP, Julius CJ, Petras RE, et al. The safety of pathological findings in experimental fat embolism. Arch
intraosseous infusions. Risk of fat and bone marrow Surg 108:70, 1974.
emboli to the lungs. Ann Emerg Med 18:1062, 1989. 38. Riseborough EJ, Herndon JH. Alterations in pulmonary
16. Manning JB, Bach AW, Herman CM, et al. Fat release after function, coalgulation and fat embolism in patients with
femur nailing in the dog. J Trauma 23:322, 1983. fractures of the lower limbs. Clin Orthop 115:248, 1976.
17. Kallos T, Enis JE, Gollan F, et al. Intramedullary pressure 39. Warner WA. Release of fatty acids following trauma. J
and pulmonary embolism of femoral medullary content Trauma 9:692, 1969.
in dogs during insertion of bone cement and prosthesis. 40. MacNamara JJ, Molot M, Dunn R, et al. Lipid metabolism
J Bone Joint Surg 56(A):1363, 1974. after trauma. Role in the pathogenesis of fat embolism.
18. Byrick RJ, Forbes D, Waddel JP. A monitored cardiovascu- J Thorac Cardiovasc Surg 63:968, 1972.
lar collapse during cemented total knee replacement. 41. Bergentz SE. Fat embolism. Prog Surg 175:128, 1968.
Anesthesiology 65:213–216, 1986. 42. Moylan JA, Birnbaum M, Katz A, et al. Fat emboli syn-
19. Cailloutte JT, Anzel SH. Fat embolism syndrome following drome. J Trauma 16:341, 1976.
the intramedullary alignment guide in total knee arthro- 43. Moylan JA, Evenson MA. Diagnosis and treatment of fat
plasty. Clin Orthop 251:199, 1990. embolism. Annu Rev Med 28:85, 1977.
20. Mokkhavesa S, Shim SS, Patterson FP. Fat embolism: 44. Saldeen T. Intravascular coalgulation in the lungs in
Clinical and experimental studies with emphasis on ther- experimental fat embolism. Acta Chir Scand 135:653,
apeutic aspects. J Trauma 9:39, 1969. 1969.
21. Emson HE. Fat embolism in 100 patients dying after 45. Keith RG, Mahoney LJ, Garvey MB. Disseminated
injury. J Clin Pathol 11:28, 1958. intravascular coalgulation: An important feature in the fat
22. Morton P, Kendall MJ. Fat embolism: Its production and embolism syndrome. Can Med Assoc J 105:634, 1971.
source of fate. Can J Surg 8:214, 1965. 46. Rammer L, Saldeen T. Inhibition of fibrinolysis in post-
23. Hausberger FX, Whitenack SH. Effect of pressure on traumatic death. Thromb Hemost 24:68, 1970.
intravasation of fate from the bone marrow cavity. Surg 47. Rennie AM, Ogston D, Cooke RJ, et al. The fibrinolytic
Gynecol Obstet 134:931, 1972. system after trauma and in patients with fat embolism.
24. Sherr S, Gertner SB. Production and recovery of pul- J Bone Joint Surg 56(B):421, 1974.
monary fat emboli in dogs. Exp Mol Pathol 21:63, 1974. 48. McCarthey B, Mammen E, LeBlanc LP, et al. Subclinical fat
25. Allardyce DB, Meek RN, Woodruff B, et al. Increasing embolism: A prospective study of 50 patients with
our knowledge of the pathogenesis of the fat extremity fractures. J Trauma 13:9, 1973.
embolism: A prospective study of 43 patients with frac- 49. Hutchins PM, Macnicol MF. Pulmonary insufficiency after
tured femoral shafts. J Trauma 14:955, 1974. long bone fractures: Absence of circulating fat or signifi-
26. Herndon JH, Bechtol CO, Crickenberger DP. Fat cant immunodepression. J Bone Joint Surg 67(B):835,
embolism during total hip replacement: A prospective 1985.
study. J Bone Joint Surg 56(7):1350, 1974. 50. Chan KM, Tham KT, Chow YN, et al. Post-traumatic fat
27. Collins JA, Gordon WC, Hudson TL, et al. Inapparent embolism: Its clinical and subclinical presentation. J
hypoxemia in casualties with wounded limbs: Pulmonary Trauma 24:45, 1984.
fat embolism? Ann Surg 167:511, 1968. 51. Sari A, Micauchi Y, Yamashita S, et al. The magnitude of
28. Hagley SR. The fulminant fat embolism syndrome. hypoxia in elderly patients with fractures of the femur
Anaesth Intens Care 11:167, 1983. neck. Anesth Analg 65:892, 1986.
29. Hagley SR. The fulminant fat embolism syndrome. 52. Tachakra SS, Sevitt S. Hypoxia after fractures. J Bone Joint
Anaesth Intens Care 11:162, 1983. Surg 57(B):197, 1975.
30. Weber KT, Janicki JS, Shroff SG, et al. The right ventricle: 53. Coetzee A, Swanevelder J, Vd Spuy G, et al. Gas exhange
Physiology and pathophysiologic considerations. Crit indices, how valid are they? S Afr Med J 85(Crit Care
Care Med 11:323, 1983. Suppl):1227, 1995.
31. Fourie P, Coetzee A, Bolliger CT. Pulmonary artery compli- 54. Heinreich H, Kremer P, Winter H, et al. Transesphageal
ance: Its role in right ventricular-arterial coupling. two-dimensional echocardiography during total hip
Cardiovasc Res 26:839, 1992. replacement. Anaesthetist 34:118, 1985.
32. Nijlsten MW, Hamer JP, Ten Huis HJ, et al. Fat embolism in 55. Pazell JA, Peltier LF. Experience with sixty three patients
patients with patent foramen ovale. Lancet 1:1271, 1989. with fat embolism. Surg Gynecol Obstet 135:77, 1972.
33. Levy D. The fat embolism syndrome. Clin Orthop 261:281, 56. Kaplan RP, Grant JN, Kaufman AJ. Dermatologic features
1990. of the fat embolism syndrome. Cutis 38:52, 1986.
34. Lequire VS, Sharpiro JL, Lequire CB, et al. A study of the 57. Jacobson DM, Terrance CF, Reinmuth OM. The neurolog-
pathogenesis of the fat embolism based on human necropsy ical manisfestations of fat embolism. Neurology 36:847,
material and animal experiments. Am J Pathol 35:999, 1986.
1959. 58. Meek RI, Fitzpatrick GJ, Phelan DM. Cerebral edema
35. Peltier LF. Fat embolism: The prophylactic value of a and the fat embolism syndrome. Int Care Med 13:291,
tourniquet. J Bone Joint Surg 38(A):835, 1956. 1987.
Boezaart_CH05.qxd 4/26/06 12:37 PM Page 59

CHAPTER 5 FAT EMBOLISM 59

59. Findlay JE, DeMajo W. Cerebral fat embolism. Can Med 81. Frostell CG, Blomqvist H, Hedenstierna G, et al. Inhaled
Assoc J 131:755, 1984. nitric oxide selectively reverses human hypoxic pul-
60. Bolesta MJ. Fat embolism syndrome without adult respi- monary vasoconstriction without causing systemic vasodi-
ratory distress syndrome. NC Med J 47:257, 1986. lation. Anesthesiology 78:427, 1993.
61. Font MO, Nadal P, Betran A. Fat embolism syndrome with 82. Nixon JR, Brock-Utne JG. Free fatty acids and arterial oxy-
no evidence of pulmonary involvement [letter]. Crit Care gen changes following major injury: A correlation between
Med 17:108, 1989. hypoxia and increase in free fatty acid levels. J Trauma
62. Thomas JE, Ayyar DR. Systemic fat embolism: A diagnos- 18:23, 1978.
tic profile of 24 patients. Arch Neurol 26:517, 1972. 83. Liljedahl SO, Westermark L. Aetiology and treatment of fat
63. Chuang EL, Miller FS, Kalina RE. Retinal lesions following embolism: Report of five cases. Acta Anaesthesiol Scand
long bone fractures. Ophthalmology 92:370, 1985. 11:177, 1967.
64. Roden D, Fitzpatrick G, O’Donoghue H, et al. Purtscher’s 84. Amato MB, Barbas CS, Medeiros D, et al. Effect of a
retinopathy and fat embolism. Br J Ophthalmol 73:677, protective ventilation strategy on mortality in the acute
1989. respiratory distress syndrome. N Engl J Med 338:447,
65. Van Miert M, Thornington RE, Van Velzen D. Cardiac 1998.
arrest after massive acute fat embolism. Br J Med 303:396, 85. Acute Respiratory Distress Syndrome Network: Ventilation
1991. with lower tidal volumes as compared with traditional
66. Coetzee A, Rousseau H, Lahner D. Acute pulmonary tidal volumes for acute lung injury and the acute respira-
hypertension and right ventricular failure in adult respira- tory distress syndrome. N Engl J Med 342:1301, 2000.
tory distress syndrome. S Afr Med J 86(Cardiovasc 86. Nelson CS: Cardiac and pulmonary fat embolectomy.
Suppl):C148, 1996. Thorax 29:134, 1974.
67. Vlahakes GJ, Turley K, Hoffman JE. The pathophysiology of 87. Bernard GR, Luce JM, Sprung C, et al. High dose cortico-
failure in acute right ventricular hypertension: Hemodynamic steroids in patients with the adult respiratory distress
and clinical observations. Circulation 63:87, 1981. syndrome. N Engl J Med 317:1565, 1987.
68. Schwartz DA, Finkelstein SD, Lumb GD. Fat embolism 88. Fischer JE, Turner RH, Herndon JH, et al. Massive steroid
and the cardiac conduction system associated with sud- therapy in severe fat embolism. Surg Gynecol Obstet
den death. Hum Pathol 19:116, 1998. 132:667, 1971.
69. Roger JA, Stirt JA. Fat embolism syndrome. Semin Anesth 89. Kallenbach J, Lewis M, Zaltzman M, et al. Low dose corti-
6:176, 1987. costeroid prophylaxis against fat embolism. J Trauma
70. Lindeque BG, Schoeman HS, Dommisse GF, et al. Fat 27:1173, 1987.
embolism and the fat embolism syndrome: A double blind 90. Kreis WR, Lindenauer SM, Dent TL. Corticosteroids in
therapeutic study. J Bone Joint Surg 69(B):128, 1987. experimental fat embolism. J Surg Res 14:238, 1973.
71. Horne RH, Horne JH. Fat embolism prophylaxis. Arch 91. Shier MR, Wilson RF, James RE, et al. Fat embolism pro-
Intern Med 133:288, 1974. phylaxis. A study of four treatment modalities. J Trauma
72. Greenberg HB. Roentgenographic signs of post-traumatic 17:621, 1977.
fat embolism (editorial). JAMA 204:540, 1968. 92. Du Toit H, Coetzee A, Charlton D. Heparin treatment in
73. Maruyama Y, Little JB. Roentgen manifestations of trau- thrombin induced disseminated intravascular coagulation
matic pulmonary fat embolism. Radiology 79:945, 1962. in the baboon. Crit Care Med 19:1195, 1991.
74. Park HM, Ducret RP, Brindley DC. Pulmonary imaging in 93. Svenningsen S, Nesse O, Finsen V, et al. Prevention of fat
fat embolism syndrome. Clin Nucl Med 11:521, 1986. embolism syndrome in patients with femoral fractures.
75. Gossling HR, Pelligrini VD. Fat embolism syndrome: Immediate or delayed operative fixation? Ann Chir
Review of the pathophysiology and physiological basis of Gynecol 76:163, 1987.
treatment. Clin Orthop 165:68, 1982. 94. Talucci RC, Manning J, Lampard S, et al. Early
76. Rokkanen P, Lahndensuu M, Kataja J, et al. The syndrome intramedullary nailing of femoral shaft fractures: A cause
of fat embolism: Analysis of thirty consecutive cases com- of fat embolism syndrome. Am J Surg 146:107, 1983.
pared to trauma patients with similar injuries. J Trauma 95. Seibel R, LaDuca J, Hassett JM, et al. Blunt multiple trauma
10:299, 1970. (ISs36), femur traction and pulmonary failure: Septic state.
77. Schonfeld SA, Ploysongsang Y, Dilisio R, et al. Fat Ann Surg 202:283, 1985.
embolism prophylaxis with corticosteroids: A prospective 96. LaDuca JN, Bone LL, Seibel RW, et al. Primary open reduc-
study in high risk patients. Ann Intern Med 99:438, 1983. tion and fixation of open fractures. J Trauma 20:580,
78. Masson RG, Ruggieri J. Pulmonary microvascular cytol- 1980.
ogy: A new diagnostic application of the pulmonary artery 97. Johnson KD, Cadambi CF, Reinmuth OM. The neurologic
catheter. Chest 88:908, 1985. manifestations of fat embolism. Neurology 25:375, 1985.
79. Chastre J, Fagon J, Soler P, et al. Bronchoalveolar lavage for 98. Malik AB. Pulmonary microembolism. Physiol Rev
rapid diagnosis of the fat embolism syndrome in trauma 63:1114, 1983.
patients. Ann Intern Med 113:583, 1990. 99. Burgher LW, Dines DE, Linscheid RL, et al. Fat embolism
80. Luyt S, Coetzee A, Lahner D, et al. Nitric oxide has little and the adult respiratory distress syndrome. Anesth Analg
effect on the acute pulmonary hypertension and right 53:664, 1974.
ventricular function during acute respiratory distress syn- 100. Guenter CA, Braun TE. Fat embolism syndrome:
drome. S Afr Med J 87:639, 1997. Changing prognosis. Chest 79:143, 1981.
Boezaart_CH05.qxd 4/26/06 12:37 PM Page 60