from correction of the hyponatremia rather than from
delayed postanoxic encephalopathy.
In addition, it is interesting to speculate why diabetes
mellitus and diabetes insipidus have also been reported
to occur in the face of severe hyponatremia, respiratory
arrest, and cerebral swelling with grooving of the uncus
and cerebellum pressure cones [TI). Whether this clinical syndrome represents a new entity [7) or reflects
systemic manifestations in patients who became brain
dead from massive cerebral edema is very unclear. The
description of infarction of the pituitary gland in Fraser
and Arieff’s study C71 bears a striking resemblance to
previously reported pathological changes in the pituitary in brain death [lo). Others also suggested that
brain death resulting in terminal hypothalamic and pituitary destruction may cause t h s specific observation
of diabetes mellitus and diabetes insipidus 111).
In any case, at the Mayo Clinic we failed to identify
this entity of postoperative respiratory arrest associated
with hyponatremia in a study of 290,815 fernales. Certainly, patients may have had postoperative hyponatremia without seizures, respiratory arrest, or persistent metabolic encephalopathy, but the clinical
relevance in this subset is uncertain.
Our negative results do not intend to minimize the
importance of the observations of Arieff and his associates. The absence of catastrophic postoperative hyponatremia syndrome in young, healthy females in our
study, however, may help in further defining its true
prevalence. Any retrospective survey of this magnitude
has the potential of missing index patients; nevertheless, our survey in Mayo Clinic-affiliated hospitals convinces us that the entity is extremely uncommon.
References
1. Deutch S, Goldberg M, Dripps RD. Postoperative hyponatremia with the inappropriate release of antidiuretic hormone.
Anesthesiology 1966;27:2 50-2 5 6
2. Chung HM, Kluge R, Schrier RW, Anderson RT. Postoperative
hyponatremia: a prospective study. Arch Intern Med 1986;146:
133-136
3. Arieff AI. Hyponatremia, convulsions, respiratory arrest, and
permanent brain damage after elective surgery in healthy
women. N Engl J Med 198(;;314:1529-1535
4. Berl T. Treating hyponatremia. Damned if we do and damned
if we don’t. Kidney Int 1990;37:1006-1018
5. Kurland LT, Molgaard CA. The patient record in epidemiology.
Sci Am 1981;245:54-63
6. Ayus JC, Wheeler JM, Arieff AI. Postoperative hyponatremic
encephalopathy in menstruant women. Ann Intern Med 1992;
117:891-897
7. Fraser CL, Arieff AI. Fatal central diabetes mellitus and insipidus
resulting from untreated hyponatremia: a new syndrome. Ann
Intern Med 193O;112:113-1 19
8. Plum F, Posner JB, Hain RF. Delayed neurologic deterioration
after anoxia. Arch Intern Med 1962;110:18-25
9. Ginsberg MD. Delayed neurological deterioration following
hypoxia. Adv Neurol 1979;26:2 1-44

10. McCormick WF, Halmi NS. The hypophysis in patients with
coma di.pass6 (“respirator brain”). Am J Clin Pathol 1970;54:

374-383
11. Plum F, Posner JB. Encephalopathy from hyponatremia. Ann
Intern Med 1970;113:84 (Letter)

Neurological Complications
from Fat Emulsion
Therapy
Paul E. Schulz, MD,” Susan P. Weiner, MD,”t§
Leslie M. Haber, MD,$ Dawna D. Armstrong, MI)$

and Marvin A. Fishman, MD”t

Fat emulsion therapy is convenient for parenterally administering calories and essential fats. We report 2 children with neurological complications of fat emulsion
therapy that arose before any systemic findings. The
complications included focal and generalized seizures,
weakness, and altered mental status. Biopsy and autopsy
findings included cerebral endothelial and intravascular
lipid deposition. Early recognition of fat emulsion therapy complications is essential as the neurological cornplications are potentially reversible with alteration of the
parented diet.
Schulz PE, Weiner SP, Haber LM, Armstrong DD,
Fishman MA. Neurological complications from fat
emulsion therapy. Ann Neurol 1994;35:628-630

Fat emulsion therapy (FET) was developed in 1935
as a means of parenterally administering calories and
essential fatty acids. Initial oil preparations often
caused multiple complications including fever and renal failure {l},and often resulted in excessively high
levels of circulating fat 12) associated with widespread
emboli [3}. The introduction of improved emulsifiers
for soybean-based FET in the early 1960s, however,
greatly decreased the morbidity associated with FET;
nonetheless, both symptomatic [4] and asymptomatic
[ 3 ) neurological lesions have subsequently been reported. One patient was a 5-month-old infant who received a 10% soybean oil emulsion (Intralipid) for 4

From the Departments of *Neurology, t Pediatrics, and $Neuropathology, Baylor College of Medicine, Neurology Service, Texas Children’s Hospital, Houston, TX.
Received Jul 14, 1992, and in revised form Dec 8, 1993. Accepted
for publication Dec 9, 1993.
Address correspondence to Dr Schulz, Department of Neurology,
NB-302, Baylor College of Medicine, One Baylor Plaza, Houston,
TX 77030.
$Present address: 1850 Union Street, San Francisco, CA 94123.

628 Copyright 0 1994 by the American Neurological Association

months, then had a grand mal seizure and became coThere was evidence of multiorgan dysfuncmatose [4].
tion. FET was discontinued and the patient was treated
symptomatically. There were no permanent neurological deficits. Another patient on FET (Intralipid), w h o
died of other causes, had a clinically silent accumulation of fat in the cerebrovasculature as well as in other
organs [ 3 ] . We report 2 patients receiving FET in
whom neurological complications occurred as t h e
prominent and presenting features. T h e y were previously presented in abstract form [ S } .

Patient Reports
Patient 1
A 9-year-old girl with aplastic anemia was admitted to the
hospital for bone marrow transplantation. Baseline laboratory
data included normal liver function tests and triglyceride
(TG) levels. As part of a transplantation protocol, FET (Intralipid) was initiated on day 2 (1-2 mgikgiday, weight =
23 kg) and continued for 30 days. Transplantation was performed on day 10. After transplantation, her hospital course
was complicated by multiple episodes of septicemia.
O n hospital day 30 she had a right focal seizure with an
associated right hemiparesis and generalized hyperreflexia.
A head computed tomographic (CT) scan showed bilateral
hypodensities in the cerebral hemispheres, more prominent
on the right, with slight enhancement. An electroencephalogram (EEG) showed a left temporal delta focus. Cerebrospinal fluid (CSF) examination was normal. Serum TG was elevated at 227 mddl. TG levels were monitored serially with
the FET being decreased intermittently because of elevated
TG levels. H e r neurological examination eventually normalized.
O n hospital day 78 she had a left focal seizure. H e r temperature was 103°F and she was somnolent, followed commands poorly, and had a mild left hemiparesis. A head C T
scan showed multiple bilateral hypodensities that had increased in size versus her original examination. An EEG
showed bilateral diffuse slowing with recurrent epileptiform
discharges in the right occipitotemporal area. The CSF was
normal except for a protein of 102 mddl. Serum TG was
447 mgldl. Her left hemiparesis resolved over 2 days. A
brain biopsy 2 days later showed perivascular edema and
neutral lipid in the pericytes of many capillaries (Fig A, filled
arrow). Bacterial, AFB, and fungal stains were negative. Because of the biopsy finding of excessive lipid, FET was discontinued.
Four months after admission, when her mental status was
normal and she had no weakness, a C T scan showed moderate atrophy with only one small left parietal lucency remaining. An EEG showed persistent right occipital slowing.
Her improvement occurred despite persistently elevated TG
levels between 400 and 1,400 mg/dl. Multiple medical complications including Candida sepsis and Pneumocysti’ carini
pneumonia eventually led to her death 7 months after admission.
Autopsy revealed Pneumocystis pneumonia, Candida endocarditis, and chronic cholecystitis. Examination of the brain

A

B
(A) A n electron micrographfrom the brain biopsy of Patient I
shows lipid (filled arrow) and lipofascin (open arrow) in brain
capillary pericytes (bar = 1.0 pni, x 44,000before 55% reduction). (B) A n autopss)
from the brain o f Patient 2,
. . sDeiimen
examined under light mirroscop~with oil red 0 staining. illurtrates intravascular deposition of lipid (bar = 10 pm, x
1.000 bejore J396 reduction).

disclosed minimal swelling and a recent subarachnoid hemorrhage over the left inferior and posterior occipitotemporal
area. Old areas of hemorrhagic necrosis measuring 1 to 1.5
cm in the right frontal, left medial frontal, and right occipital
lobes were present. Microscopic study demonstrated multiple areas of necrosis and hemorrhage. Pericytes contained
large amounts of neutral lipid.

Patient 2
A 9-year old boy with cystic fibrosis was admitted because
of fever and increased sputum production. Because of poor
oral intake, FET (Intralipid) was begun on hospital day 6 as
a 20%. solution at 40 mlihr (5.1 mgikgiday; weight = 19.5
kg). The next day he was nauseated, agitated, and had hemoptysis. That afternoon he had a right focal seizure that generalized. He was lethargic and followed simple commands
poorly, cranial nerve and motor functions were normal, and

Brief Communication: Schulz et al: Ncurological Complications from Fat Emulsion

629

his deep-tendon reflexes were decreased but present. Laboratory data showed a mild respiratory acidosis and were otherwise normal exccpt for a prothrombin time of 15 seconds, a
partial thromboplastin time 2 120 seconds, and a TG level
of 1,575 mg/dl. A CT scan of the head was normal. An EEG
performed during a seizure showed constant spike activity
emanating from the entire right hemisphere. Interictdy
there was a spike focus in the right occipital region. The
patient was begun on dilantin and phenobarbital but continued to have right focal seizures that generalized. The patient’s
pulmonary function deteriorated and he died the next
morning.
The general autopsy, described previously 161, revealed
hepatosplenomegaly and end-stage cystic fibrosis changes in
the lungs. Microscopic studies with fat stains demonstrated
the presence of fat emboli in multiple capillaries and arterioles leading to areas of necrosis in the spleen, liver, kidney,
and lymph nodes. A bone marrow aspirate showed numerous
eosinophils. Other organs showed changes consistent with
long standing cystic fibrosis. There was mild cerebral edema
(brain weight, 1,490 gm; normal, 1,275 gm), with no evidence of herniation, and an area of petechial hemorrhage
in the right occipital lobe. Microscopic examination showed
numerous intra-arteriolar and capillary neutral lipid emboli
(Fig B). Some capillaries and arterioles had aggregates of
inflammatory cells consistent with acute terminal ischemia.
Viral cultures and bacterial and fungal stains were negative.

Discussion
The complex of multiorgan failure in association with
FET has been termed the “fat overload syndrome”
(FOS). We report neurological complications of FET
as the initial clinical manifestations of FOS. The complications included focal and generalized seizures,
weakness, and encephalopathy. The head CT scan indicated multifocal, asymmetrical abnormalities in 1 patient but was normal in the other. The EEG showed
focal abnormalities in both patients.
Several lines of evidence suggest that lipid was involved in the etiology of the symptoms observed. In
Patient 2, the autopsy revealed intravascular lipid deposition leading to areas of necrosis in multiple organs,
as noted previously 171, and included the brain. In Patient l, the brain biopsy demonstrated large amounts
of lipid in pericytes and endothelial cells in association
with high serum TG levels. In addition, there were
systemic complications similar to those previously reported in association with FET. While intravascular
lipid was not noted, pathological studies were performed after the symptoms had resolved. Cyclosporin
can cause encephalopathy and seizures and result in
diffuse 181 or focal 193 white matter changes on CT
scan; however, it does not appear to have been responsible for the symptoms in our patients since it was not
used in Patient 2, and Patient 1 had additional symptoms and findings that were more consistent with FOS
than cyclosporin toxicity.

630 Annals of‘ Neurology Vol 35 No 5 May 1994

The mechanism by which lipid was deposited is unclear [lo-123; however, at least two factors may have
contributed to its being deposited. First, a rapid rise of
TG levels may have contributed as suggested by Patient 2 whose T G levels rose rapidly to 1,575 mg/dl
and were associated with the onset of symptoms. Second, multiple infections, medications, and liver dysfunction may have contributed to elevated TG levels
in Patient 1 as suggested by finding that a rise in liver
function tests around day 75, which might be associated with a decrease in TG clearance, was followed by
a rise in TG levels shortly thereafter.
The use of FET is very common in neurological,
medical, and pediatric settings. Our patients illustrate
that neurological complications can occur and can be
the presenting and principal signs of fat overload. The
cardinal neurological symptoms of fat overload appear
to be multifocal deficits and focal seizures. Patient 1
and the patient reported by Belin and colleagues 143
illustrate the importance of early recognition of FOS
as early intervention may reverse the neurological
deficits.
References
1. Taylor RF, Buckner CD. Fat overload from 10 percent soybean
oil emulsion in a marrow transplant recipient. West J Med 1982;
136:345-349
2. Heyman MB, Storch S, Ament ME. The fiat overload syndrome,
report of a case and literature review. AmJ Dis Child 1981;135:
628-630
3. Hessov I, Melson F, Haug A. Postmortem findings in three
patients treated with intravenous fat emulsions. Arch Surg 1979;
114166-68
4. Belin RP, Bivins BA, JonaJZ, Young VL. Fat overload with a
1096 soybean emulsion. Arch Surg 1976;11:1391-1393
5. Weiner SP, Schulz PE, Haber LA, et al. Neurologic complications of Intralipid therapy: the fat overload syndrome. Ann Neurol 1987;22:452
6. Haber LM,Hawkins EP, Seilheimer DK, Saleen A. Fat overload
syndrome: an autopsy study with evaluation of the coagulopathy.
Am J Clin Pathol 1988;90:223-227
7. Freund U, Krausz Y, Levij IS, Eliakim M. Iatrogenic lipidosis
following prolonged intravenous hyperalimentation. Am J Clin
Nutr 1975:28:1156-1160
8. DeGroen PC, Aksamit AJ, Rakela J, et al. Central nervous system toxicity afcer liver transplanration: the role of cyclosporin
and cholesterol. N Engl J Med 1987;317:861-866
9. RL Hughes. Cyclosporin-related central nervous system toxicity
in cardiac transplantation. N Engl J Med 1990;323:420-421
10. Campbell AN, Freedman MH, Pencherz DB, Zlotkin SH.
Bleeding disorder from the “fat overload” syndrome. J Parenter
Enter Nutr 1984;8:447-449
11. Hullman G, Pearson HJ, Fraser I, Bell PRF. Agglutination of
Intralipid by sera of acutely ill patients. Lancet 1982;8313:
1426-1427
12. Burnham WR, Cockbill SR, Hepinstall S, Harrison S. Blood
platelet behavior during infusion of an Intrahpid-based intravenous feeding mixture. Postgrad Med J 1982;58:152-153