вход по аккаунту


Brain microemboli during cardiac surgery or aortography.

код для вставкиСкачать
Brain Microemboli During
Cardiac Surgery or Aortography
D. M. Moody, MD," M. A. Bell, DPhi1,"i V. R. Challa, MD,? W. E. Johnston, MD,§ and D. S . Prough, MD§
We have observed many focal dilatations or very small aneurysms in terminal arterioles and capillaries of 4 of 5 patients
and 6 dogs who had recently undergone cardiopulmonary bypass. A smaller number of sausagelike dilatations distended medium-sized arterioles. Two other patients had a small number of the same microvascular changes following
proximal aortography. Thirty-four patients and 6 dogs not undergoing cardiopulmonary bypass had none. (A 35th
patient who had not undergone cardiopulmonary bypass or aortography showed a small number of dilatdtions;
mediastinal air was a suggested source.) Some of the dilatations exhibited various forms of birefringence. Because most
of the dilatations appear empty, we speculate that they are the sites of gas bubbles or fat emboli that have been
removed by the solvents used in processing. These microvascular events, occurring only in conjunction with major
arterial interventions, may be the anatomical correlate of the neurological deficits or moderate to severe intellectual
dysfunction seen in at least 24% of patients after cardiac surgical procedures assisted by cardiopulmonary bypass.
Moody DM, Bell MA, Challa VR, Johnston WE, Prough DS. Brain microemboli during
cardiac surgery or aortography. Ann Neurol 1990;28:477-486
The incidence of severe central nervous system sequelae after cardiopulmonary bypass (CPB) has declined as modern CPB techniques have developed. In
a retrospective review of medical records, only 3.496
of patients had altered mental state and only 1% had
stroke [l]. Newer evidence establishes that when
patients undergoing cardiac operations are enrolled
prospectively in an investigation of motor and higher
integrative cerebral function after CPB, persistent neurological deficits and moderate to severe intellectual
dysfunction can be found in at least 24% {2-61. Thus,
of the more than 300,000 patients in the United States
every year who undergo operations assisted by CPB,
72,000 may have neurological or neuropsychological
impairment after othewise szlccessfzll surgical procedures.
Using the alkaline phosphatase histochemical staining technique for thick celloidin sections, we studied
the brain microvasculature of 5 patients and 6 dogs
after recent CPB. In all, there were dramatic and ubiquitous alterations (Figs 1-7) consisting of focal small
capillary and arteriolar dilatations (SCADs), or microaneurysms. These changes were not seen in 40 brains
from both human beings and dogs that had not undergone CPB exposure or major proximal arterial manipulation [7).
Human brains received at autopsy were entered in our microvascular study protocol and prepared according to our
previously reported method { 8 , 91.The 4 3 patients ranged in
age from 29 to 96 years. Table 1 gives clinical information
and perfusion data on the 5 patients who underwent CPB, as
well as 2 additional patients who underwent invasive procedures involving the left-sided circulation; these data represent the patients with SCADs. (Clinical data are not included
in Table 1 for 1 patient who had not undergone CPB and
who showed a small number of SCADS after dying with
mediastinal air from a ruptured esophagus, or for another
patient who had aortography but no SCADs; both these
patients, however, appear in Table 3.) The others in the
series died of a wide variety of causes. Of the 8 patients who
underwent cardiac surgical procedures or aortography, 5
were hypertensive and all were oider than 50 years. Of the
35 patients who did not have cardiac surgical procedures or
aortography (including the patient with mediastinal air described above), 24 were hypertensive and 24 were older than
50 years. Many of these brains were included in our reports
on the anatomical arrangement of the cerebral microvasculature [9}and on changes in the cerebral vasculature associated
with hypertension and aging [lo}.
Large, thick blocks of tissue (up to 5 x 5 x I cm) were
cut from 10 cerebral and brainstem areas according to a
protocol. An average of 7.5% by weight of the human brain
was processed and inspected, more if pathological changes
From the Departments of *Radiology, ?Pathology, +Anatomy, and
§Anesthesiology, Bowman Gray School of Medicine, Wake Forest
University, Winston-Salem, NC.
Address correspondence to Dr Moody, Department of Radiology,
Bowman Gray School of Medicine, Wake Forest University, Winston-Salem, NC 27 103.
Received Feb 27, 1990, and in revised form May 7. Accepted for
publication May 7, 1990.
Copyright 0 1990 by the American Neurological Association 477
Table 1. Clinical and Perfusion Characteristics of Patients with SCADs
70, F
64, F
65, F
75, F
52, M
57, F
69, M
Time from
to Death
Biatrial myxoma removal
Repeat CABG
Cardiac transplant
(1) Coronary angiography
(2) Coronary angiography
(1) Axillobifemoral bypass graft
(2) Aortography
of CPB
of Cross-clamp
Duration of
“The anesthetic agent used was fentanyl in each instance.
‘This patient had blood-filled capillary aneurysms, but not SCADs as strictly defined. See Discussion and Figure 4
‘Did not have CPB.
SCADs = small capillary and arteriolar dilatations; MVR = mitral valve replacement; CABG = coronary artery bypass graft; AVR = aortic
valve replacement; VAD = ventricular assist device; CPB = cardiopulmonary bypass; B = bubble oxygenator; M = membrane oxygenator;
NA = not applicable.
were identified at the time of gross cutting. The tissue blocks
were fixed in cold, weak formalin followed by progressively
higher concentrations of alcohol before celloidin embedding
and sectioning at 100, 500, and 1,000 pm on a base sledge
microtome. The sections were stained histochemically, using
the activity of the native nonspecific alkaline phosphatase
enzyme present in the endothelium of capillaries, arterioles,
and the smallest arteries. The final reaction product is brownblack lead sulfide, which makes the microvascular bed of the
100-pm sections suitable for examination by light microscopy after being counterstained with cresyl violet acetate and
light green, Congo red, or a myelin stain, and then being
mounted and coverslipped. The presence of the lead precipitate also makes the endothelium relatively opaque to lowenergy X rays. High-resolution contact microradiographs of
the 1,000- and 500-pm-thick sections were made at 7 to 10
kVp with the copper anode tube of an ISRR-60 microradiographic unit (Softex Co., Tokyo, Japan) using Kodak SO343 high-resolution single-emulsion film (Eastman Kodak
Co., Rochester, NY). The microradiographs were covered
with mounting medium and coverslips and then examined
under a hght microscope.
Tissue blocks for routine paraffin sections and hematoxylin-eosin (H & E) stain were also obtained from the opposite
side of the brain and, where appropriate, from areas adjacent
to the phosphatase-stained sections. The H & E-stained sections were studied for vessel wall changes and overall correlation with the findings in the vascular phosphatase preparations.
SCADs can be reliably detected by scanning a 100-pmthick section with the 10 x objective lens of the light microscope. The unequivocal sighting of even one SCAD placed a
subject in the “with SCADs” category. Detailed counts and
descriptive analyses of SCADs are extremely time-con-
Annals of Neurology
Vol 28
4 October 1990
suming and have been carried out on only a few sections;
numbers detected in more casual scans of the different subjects with positive findings ranged from two SCADs per
large section to a few hundred, For a subject to be categorized as “without SCADs,” a minimum of two 4 x 4-cm
sections (from the basal ganglia area) had to be thoroughly
scanned at 10 x ,with every vessel being observed, without a
single SCAD being found. Usually, several more sections
from different areas of the same brain were also examined
less rigorously.
For 3 of the patients who underwent CPB, a bubble oxygenator had been used; for the remaining 2, a membrane
type. Both types had been combined with a 25-km poresized arterial line filter (see Table 1). In all instances, fieldaspirated blood had been passed through a filter (rated at
755%efficient for 35-pm particles) incorporated into the cardiotomy reservoir.
Twelve brains of mongrel dogs were donated to us at the
termination of experiments in other laboratories. Six of the
dogs were euthanized immediately after experiments performed by one of us (W. E. J.) and a colleague in studies of
cerebral or coronary blood flow during CPB. During the
experiments, microspheres were injected into the inflow
line. For all 6 dogs, a bubble oxygenator and a 25-pm poresized arterial line filter were used; further perfusion data for
these dogs is found in Table 2. None of the post-CPB dogs
were hypertensive. One dog (Dog 7) died after institution of
anesthesia before CPB could be started; this dog was used as
a control. The 5 remaining control dogs were involved in
hypertension experiments (3 were hypertensive, 2 were not),
but nothing was injected into these proximal aortas. Whole
coronal dog brain slices 1 cm thick were cut at the levels of
the basal ganglia and pons without prior cooling, and were
immediately put into either cold fixative or cold 70% al-
Table 2. PerJusion Characteristics of Dogs UnLrgoing Cardiopulmonavy"
of CPB
Dog Number
of Cross-clamp
Duration of
"The anesthetic agents used in each instance were diazepam, thiamylal, and fenranyl.
bMicrospheres only, without CPB. This animal had no SCADs.
CPB = cardiopulmonary bypass; B
bubble oxygenator; M
membrane oxygenator; NA = not applicable.
cohol. Sections from chese slices were subsequently stained
for alkaline phosphatase and analyzed for SCADs in the
same manner as those from the human brains.
All brain material from patients and experimental animals exposed to CPB had SCADS (or a comparable
alteration). We observed a few SCADs in 2 patients
(Patients 6 and 7) who had not undergone cardiac operations, but who had undergone proximal aortography. Of the 35 patients and 6 dogs not undergoing
CPB o r major left-sided arterial manipulation, only 1
had a few scattered SCADs. This patient died with a
perforated esophagus after air was insufflated in an
attempt at dilation. The remaining patients and dogs
(40 subjects) had no SCADs.
The SCADs observed in these specimens occurred
in medium-sized arterioles, terminal arterioles, and
capillaries. The dilatations ranged in size from 10 to 40
pm in diameter, and had a predilection for bifurcation
points. In the terminal arterioles and capillaries, the
SCADs were elliptical and measured 10 to 15 pm in
diameter (see Fig 1). In larger arterioles, they were
frequently as large as 40 p m in diameter (see Fig 2)
and were commonly elongated or sausage-shaped.
Their walls appeared normal, albeit stretched and thin.
SCADs appeared to be more consistent with a dilated
and empty lumen than with a swollen endothelial cell;
any material originally present in the lumina was not
evident following tissue preparation. No SCADs were
seen in the postcapillary or larger venules, which had
little or no alkaline phosphatase but could be seen with
the counterstain. Some of these SCADs exhibited the
property of birefringence (see Fig 1). There was no
favored location in the brain or upper spinal cord for
these SCADs, but they were found in proportion to
the density of the capillary bed, that is, more were
found in cortex and deep gray nuclei than in white
Fig 1. Aneurysmal dilatati0n.r with difluse birefringence in a
brah capillary from a patient following cardiopulmonary bypas.r. The capillaty arites from the arteriole that is out of focus to
the left. The lower panel is the same field seen with the aid of
crossed polarizing filters. Dgfiise birefringence is in the wall
of the dilated segment. not the bmen. The wall is thin and
stretched. Approximately half of these SCADs exhibit birefringence. Bar = 25 ,um. (100-pm-thick celloidin section .stained
for alkaline phosphatase.1
Moody et al: Brain Emboli after CPB 479
F i g 2. Sausagelike dilatations in a medium-sized arteriole from
white matter of a patient who had recently undergone cardiopulmonary bypass. These putative emboli are 40 pa in diameter
and are probably more dangerous than the smaller ones (iee text).
Notice that blood elements are displaced out of the lumen at the
site of the clear swellings (see also Pig 7).Another of these emboli
is seen in a smaller arteriole (arrow). Bar = 100 pm. (100pm-thick celloidin section stained for a l k u h e phosphatase.)
Fig 3. Large cylindrical embolus with more distal (daughter?)
emboli in same vascular system from cortical gray matter in a
patient after cardiopulmonaty bypass. Surface of brain is toward
top. Thi.c is a typical 20-pm cortical arteriole dividing into multiple branches. One limb has a long clear embohs in it. Distal
branches have characteristic candelabra configuration, and one
segment has multiple small emboli (arrows). Some of the latter
are slight4 out OffDcusin this 100-p-thick section. Identifving small daughter emboli in the same .tJaJcularsystem distal to a
larger embolus would be extremely difficult with ordinary 5 - t o
10-papreparations. Bar = 100 p n . (100-pm-thick celloidin
section, stained for alkaline phosphatase.)
matter. They were often found in groupings (see Figs
1-3, 6, 7). In 1 patient (Patient 2), we counted 1,740
SCADs in a large 100-km-thick coronal section that
included basal ganglia, frontal cortex, and intervening
white matter.
SCADs in the thicker (500 to 1,000 Km) microradiographed sections were often obscured by superimposed vessels, but a few convincing examples
were recognized (see Fig 5).
A slightly different kind of capillary microaneurysm
was found in 1 patient (Patient 4 ) at the margins of
several intracerebral hemorrhages (see Fig 4). In this
patient, CPB was performed 27 days before death. No
SCADs were found apart from the hemorrhages.
All the dogs in Table 2 had microspheres injected
into the systemic circulation. The first 3 dogs each had
five doses of 15-pm microspheres during CPB for
blood flow determinations. From 1 of these dogs, a
single coronal whole-brain section at the level of the
basal ganglia, 100 pm thick, had 297 microspheres;
399 SCADs were counted in this same section, and
it is possible that some of the lucent SCADs were
missed while opaque microspheres were easy to identify (see Fig 6).
Table 3 relates the occurrence of SCADs found in
certain interventional procedures. The sensitivity and
specificity of the finding (SCADs) were calculated
from a 2 x 2 contingency table (Table 4).
480 Annals of Neurology
Vol 28
No 4 October 1990
F i g 4. Microradiograph showing 3 of the marzj capillaqi and
terminal arteriolar aneuvysms found at the margin of an intracerebral hemorrhage (ldt).The patient had coronary artery bypass
grafting with iardiopulmonavy bypass-assisted uentilation 2 7
&ys before death. No SCADs were seen apart from the hemorrhages. These aneurysms are slight(y difjfirent from other
SCADs described in the present report, because the lumina are
jilled with blood products, that is, they are not clear (arrows).
Compare this microradiograph with F i g 5 . Bar = 100 pm.
(500-pm-thick celloidin section stained for alkaline phosphatase.)
Alkaline Phosphatase Vasczllar Stain
The arteries of the pial-arachnoidal plexus and their
larger perforating branches to the cerebrum are negative for alkaline phosphatase. The enzyme, which is
active in the membranes of endothelial cells, first appears in a patchy or streaky fashion in penetrating and
intraparenchymal vessels with diameters of approximately 200 down to 50 pm, that is, in the smallest
arteries and largest arterioles. In this size range, exchange of nutrients begins to occur C11, 121. The
smaller arterioles and the capillary bed are strongly
positive for alkaline phosphatase; venules do not usually stain for alkaline phosphatase {S, 11, 131.
Fig 5 . Microradiograph shouiing embolus in arteriole following
cardiopulmonaq bypass. Arrow is o n a dilatation with an
empty lumen. Bur = 100 pm. (SOO-pm-thick celloidin section
stained far alkaline
The phosphatase histochemical technique produces
an endothelial map of the afferent cerebral microvasculature, and is particularfy suited for analysis of multiple emboli in the microcirculation when compared
with older injection techniques because no injection (a
source of artifactual air bubbles) of either fixative or
dye into the vascular tree is required. Although there
were some slight differences in the preparation of
these tissues in our study (cooling brain versus immediate cutting, alcohol versus formalin fixation),
SCADs were seen in each instance depending only on
whether CPB or proximal circulatory manipulation
had been performed.
Importance of SCADs
It has been established that cerebral blood flow decreases dramatically during CPB for reasons that are
not entirely clear C14-18). Microemboli 16, 15, 19,
201 and factors modifying metabolic demand t 2 l l have
been implicated. The microvascular alterations we report are small empty capillary and arteriolar dilatations,
Moody et al: Brain Emboli after CPB
Fig 6. Microspheres andSCADs in white matter of a dog following cardiopulmonay bypass. Direction of blood flow is from
top to bottom of illustration. A black microsphere is lodged in the
oujice of a capillary. SCADs are seen distal to the microsphere
in an area that is at a d i f f e n t plane of focus (arrow) and also
in an adjacent capillaty. Microsphere = 15 pm. (100-pi-thtck
celloidin section stained for alkaline phosphatase.,
Fig 7. A f n u SCADs were observed in 5-prn hematoxylin-eosin
paraffin sections. Previous analysis of thick alkaline pho.rphatase-stained sections had determined that this patient had “millions” of SCAD.i after cardiopulmonary hj9aA.r. Note that the
blood elements are displaced away from the clear spaces in the
lamina, and that the vascular walls are thin and stretched.
Bars = 25 pm.
or SCADs, which presumably result from emboli during CPB. They occur in sufficient numbers and sizes
that mental alterations might be expected.
A parallel that supports this contention exists in the
microsphere method of measuring blood flow in experimental animals: microspheres of a size that are
trapped in the microvasculature during the first pass
through the brain are injected into the left-sided circulation. For measuring cerebral blood flow, 15-pm
spheres are optimal [22), as they lodge in the smallest
arterioles. In animal experiments, a sufficient number
of 15-pm microspheres are administered in one dose
to provide approximately 400 microspheres per gram
of brain tissue (for the dog, this has been determined
to be 1 x 106spheres). This dose will obstruct 0.01%
of brain capillaries [23}. Dogs given as many as 25 x
106 (15-pm) microspheres, that is, 25 doses yielding
10,000 spheres/@, appear grossly intact, but much
smaller numbers of larger (50-p.m) microspheres pro-
duce both distortion in flow patterns and neurological
injury c23). The larger the embolus, the more proximally it will lodge and the larger the affected distal
arteriolar and capillary territories will be. In the present series, 1 dog (Dog 3) that had received approximately 5 x lo6 (15-pmj microspheres while on CPB
had 1.3 times as many SCADs as microspheres (399
SCADs to 297 microspheres). This count was made
from a single 100-pm-thick coronal whole-brain section at the level of the basal ganglia. Many of the
SCADs were considerably larger than 15 pm. In this
specimen, extrapolation would suggest a total brain
load of 2,000 microspheres per gram and 2,600
SCADs per gram. This total density of 10- to 15-pm
SCADs still might not cause neurological injury in human beings, but there may be a sufficient number of
larger and more dangerous emboli to cause deterioration of complex psychomotor function. It is difficult to
compare postoperative sequelae in human beings and
Annals of Neurology Vol 28
No 4 October 1990
Table 3. Pwvalence of SCADJ
No. of Brains
with SCADs
Type of Specimen
No. of Brains
without SCADs
Human," post CPB
Canine, post CPB
Total human and canine, post CPB only
Human" with recent proximal arteriography
Total human and canine with recent CPB or proximal arterio$wPhy
Total No. of
Brains Examined
l b
Human" without recent arteriography or CPB
Canine without CPB (or arteriography)
Total human and canine without recent CPB or proximal areriography
"Human brains examined comprise totals of lines 1, 4 , and 6 = 43.
bPatient 4 died 27 days after CPB. Had blood-filled capillary aneurysms, but not SCADs as strictly defined. Was counted as a negative. See
Discussion and Fig. 4.
'Died after rupture of esophagus. Established pathway for air to enter systemic circulation existed. (See Discussion.)
small capillary and arteriolar dilatations; CPB
carcliopulmonary bypass.
l a & 4. Relationship of SCADs to Proximal Circulatoy
Procedure i n Human Beings and dog^
"Sensitivity of the finding (SCADS), 12/13 = 92%. If the finding
(SCADs) is present, there is a 92%, chance that proximal vascular
manipulation has occurred.
bSpecificity of the finding (SCADs), 40/42 = 95%. If the finding
(SCADs) is absent, there is a 95% chance that proximal vascular
manipulation has not occurred.
small capillary and arteriolar dilatations
dogs. Most patients appear grossly neurologically intact after cardiac surgical procedures, and detailed testing is required to elicit deficits; comparable tests of
complex psychomotor function cannot be performed
in dogs.
In 1 patient (Patient 2) whose SCADs we have
counted in detail, 1,740 SCADs were found in a coronal hemisection at the level of the basal ganglia, incorporating deep and cortical gray matter as well as white
matter. This represented a tissue volume of 148 mm3
(corrected for 30% shrinkage), yielding a SCAD density of 11.76/mm3 or 11,760/cm3 (4.5 times that estimated for the dog just described). As 1 cm3 of brain
weighs approximately 1 gm, and this patient's brain
weighed 1,305 gm, we calculate the patient's total
brain load to have been 11,760 x 1,305, that is, 15.3
X lo6 SCADs.
Prevukncr ofSCADs
The appearance of SCADs in the microvessels of only
those brains that had been exposed to CPB or an interventional left-sided arteriographic procedure is a coincidence too striking to be discounted. We have given
details of 13 such cases (7 patients and 6 dogs, see
Tables 1, 2), and contrasted these findings with those
for 40 unaffected subjects (34 patients and 6 dogs)
with no manipulation of the proximal left-sided circulation.
Three patients exhibiting SCADs deserve special
comment. Patient 4 had CPB 27 days before death,
with focal neurological findings 7 days before death
attributable to several large intraparenchymal hemorrhages confirmed by computed tomography. No
SCADs were found in nine large blocks representing
parts of the brain away from the hemorrhages, suggesting that SCADs (which presumably were present during CPB) are transient and eventually disappear. There
were, however, numerous arteriolar and capillary microaneurysms adjacent to the hemorrhages (see Fig 4).
These were similar to the SCADs seen in other specimens except that their lumina were filled with blood
products [24]. We don't know if SCADs were the
forerunner of these aneurysms or were related to the
hemorrhage. All other subjects with SCADs and CPB
died within 10 days of operation.
Patient 6 had complete coronary and left ventricular
arteriography at an outside institution. This procedure
was repeated at our hospital 15 days later because of
the patient's worsening cardiac status. She died before
operation could be performed. SCADs found in the
brain of this patient were in no way different from the
Moody et
Brain Emboli after CPB 483
ones seen in post-CPB patients. This finding raises
the distinct possibility that these putative emboli can
also be liberated during arteriography proximal to the
brain. Of the patients without SCADs, only 1 had had
recent proximal arteriography. All of the patients with
SCADs and CPB had arteriography prior to operation.
The arteriography itself was not the principal factor in
producing SCADS in our specimens: none of 6 dogs
with CPB and SCADs had arteriography (although microspheres were injected in each instance).
The third patient (not listed in Table 1) had a few
scatrered SCADs that were presumably due to a ruptured esophagus, which gives access to an established
route for pulmonary venous air embolism that passes
ultimately into the left-sided circulation.
Distribution of SCADs
SCADs appeared in multiples in the same vessel, or in
clusters near each other, more often than would be
expected to occur randomly. Similarly, two or more
microspheres frequently occurred near each other in
the same vascular system in dogs; they were often seen
in combination with SCADs (see Fig 6). In the case of
microspheres, this phenomenon might reflect some
physical attraction between the particles. (This is a
minor flaw in the microsphere method of measuring
blood flow; steps are routinely taken to minimize
clumping, but as we have seen, it is not entirely eliminated.) We believe there is a better explanation for the
grouping of SCADS-daughter emboli breaking away
from the larger parent to migrate downstream. Thus,
smaller SCADs, apparently not large enough to cause
neural dysfunction, might have arisen from larger,
more proximal SCADs.
It is likely that SCADs represent iatrogenic emboli.
They are distributed in the brain with a frequency that
corresponds to the expected volume of blood flow,
that is, more are found in the cortical and deep nuclear
gray matter than in white matter. When a sausagelike
dilatation is found in a medium-sized arteriole, smaller,
elliptical SCADs can often be identified in terminal
arterioles and capillaries in the same vascular system
downstream (see Fig 3). This point again emphasizes
the advantage of the alkaline phosphatase stain and
thick celloidin sections; such an analysis would be impossible with ordinary paraffin 5- to 10-pm preparations.
SCADs Are Not Cbarcot-BoucbardAneuysms
Charcot-Bouchard intraparenchymal aneurysms are a
result of hypertension. They occur in small arteries and
are visible to the naked eye, being 200 to 1,000 pm in
diameter 125, 261. These persistent aneurysms may be
the precursor of some (but not all) hypertensive ganglionic hemorrhages and, in recent times, are rare (we
have never seen one in our material). A Charcot484 Annals of Neurology Vol 28 No 4 October 1990
Bouchard aneurysm does not have an empty lumen,
and its vascular wall is diseased. SCADs are smaller
(10 to 40 pm), and occur in much smaller vessels; they
have an empty lumen and a normal wall (although thin
and stretched), and are apparently transient.
Pathological Correlations, Birefringence,
and Clinical Sign$cance
Supplementary blocks of brain tissue from all our human subjects were embedded in paraffin and sectioned
before staining with the usual neuropathological stains
such as H & E. Often these blocks came from slices
facing those taken for alkahne phosphatase histochemical processing. The H & E-stained sections were cut T
bm thick, that is, 20 times thinner than the celloidinphosphatase sections for light microscopy and 100
to 200 times thinner than the microradiographed
sections. In the H & E-stained sections, only small
segments of the microvasculature are present, and
SCADs are difficult to identify; in a patient who had
millions of them (estimated from our alkaline phosphatase preparations) only a few candidates for SCADs
could be found (see Fig 7).
The SCADs appeared in H & E-stained preparations as distended vessel segments in which the lumina
were clear and free of blood products, but the blood
cells were packed up against both proximal and distal
contours of each SCAD in the lumen of the vessel (see
Fig 7). Birefringent SCADs can occasionally be recognized in H & E-stained paraffin sections with the aid of
crossed polarizing lenses. Particulate emboli during
CPB from agents such as glove powder, silicone antifoam products, fat, blood products, and tubing particles have been discussed previously 115, 27-30]. All
but fat have a different appearance from SCADs and
have been largely eliminated by modern surgical techniques and CPB technology using arterial line and cardiotomy reservoir filters.
We have seen a few particulate emboli in celloidin
and paraffin preparations. In one human brain that we
have analyzed in detail (we counted and categorized
every SCAD in a 4 x 4-cm basal ganglia section), less
than 1% of microaneurysms contained particulate matter and exhibited a Maltese-cross form of birefringence
in the bmen. We think that these particles represent
dust, glove powder, o r other exogenous material that
will never be totally excludable from this sort of procedure; in the procedure for a dog, in which precautions
may have been less rigorous, 20% of the SCADs
showed luminal Maltese-cross birefringence. Most
SCADs are different in that they are free of particulate emboli and are empty; in human beings, 42%
of SCADs have diffuse birefringence in their walls.
The property of birefringence can help in identifying
SCADs in both alkaline phosphatase-prepared specimens and H & E-stained sections.
We are not certain as to the etiology of SCADs, but
we speculate that air or fat emboli are the best possibilities. Both the appearance [3 1) and importance
132) of cerebral air emboli are well known, although
it is believed that their generation in CPB is greatly
curtailed with the use of membrane oxygenators and
filters 115, 20).
It is unlikely that SCADs represent swollen endothelial cells because of the pattern of distribution
and because some SCADs are much larger than a single endothelial cell (see Figs 2, 3). Furthermore, the
alkaline phosphatase stain, normally present in both
the luminal and abluminal plasma membranes of the
endothelial cell, should be seen splitting around a
SCAD consisting of swollen cytoplasm; even at high
magnifications, this phenomenon was not observed.
Are the SCADs clinically significant or are they an
epiphenomenon? That is, are they present but do not
cause symptoms? It would be useful to demonstrate
dysfunctional astrocytes or neurons adjacent to the
SCADs, but it must be remembered that SCADs may
be moving distally in the vascular tree as long as (pulsatile?) flow is present or while gas is undergoing reabsorption while the subject is still alive. In addition, it is
possible that the diffuse microvascular obstruction is
sufficient to impair neural function without producing
cell death.
If the well-documented neurological changes seen
following CPB are caused by the phenomenon described here, the accepted methods of cerebral blood
flow monitoring (radiolabeled xenon washout, cold
xenon computed tomography, single photon emission
computed tomography, positron emission tomographic
scanning, and sensory evoked potentials) cannot be
expected to detect their presence or extent until a
massive number are present.
This work was supported by a Jacob K Javits Neuroscience Investigator Award, National Institutes of Health grant NS 20618.
We wish to thank Chris Johnston for her expert technical assistance
and Donna McCain for her careful preparation of this manuscript.
We are grateful to Drs J. Vinten-Johansen (Bowman Gray School of
Medicine) and A. Gorman (East Carolina University School of
Medicine) for contributing research material.
1. Coffee CE, Massey EW', Roberts KB, et al. Natural history of
cerebral complications of coronary artery bypass graft surgery.
Neurology 1983;33:1416-1421
2. Slogoff S, Girgis KZ, Keats AS. Etiologic factors in neuropsychiatric complication associated with cardiopulmonary bypass.
Anesth Analg 1982;61:903-911
3. Shaw PJ, Bates D, Cardidge NEF, et al. Early intellectual dysfunction following coronary bypass surgery. Q J Med 1986;58:
4. Sotaniemi KA, Mononen H, Hokkanen TE. Long-term cerebral
outcome after open-heart surgery: a five-year neuropsychological follow-up study. Stroke 1986;17:410-416
5. Smith PLC, Treasure T, Newman SP, et al. Cerebral consequences of cardiopulmonary bypass. Lancet 1986;1:823-825
6. Aris A, Solanes H, Camari ML, et al. Arterial line filtration
during cardiopulmonary bypass. J Thorac Cardiovasc Surg 1986;
7. Moody DM, Bell MA, Challa VR. Focal encephalic microvascular alterations following cardiopulmonary bypass pump. Soc
Neurosci Abstr 1989;15:27
8. Bell MA, Scarrow WG. Staining for microvascular alkaline
phosphatase in thick celloidin sections of nervous tissue: morphometric and pathological applications. Microvasc Res 1984;
27 189-203
9. Moody DM, Bell MA, Challa VR. Features of the cerebral vascular pattern that predicts vulnerability to perfusion or oxygenation deficiency: an anatomic study. AJNR 1990;11:431-439
10. Moody DM, Santamore WP, Bell MA. Does tortuosity in cerebral arterioles impair down autoregulation in hypertensives and
elderly normotensives? A hypothesis and computer model. Clin
Neurosurg 1990 (in press)
11. Saunders RLdeCH, Bell MA. X-ray microscopy and histochemistry of the human cerebral blood vessels. J Neurosurg
12. Cervbs-Navarro J, Iglesias Rozas J. The arteriole as a site of
metabolic exchange. In: CervBs-Navarro J, Betz E, Ebhardt G,
et al., eds. Pathology of cerebrospinal microcirculation. Adv
Neurol 1978;20: 17-24
3. Bell MA, Moody DM, Angelo JN, et al. Microvascular alkaline
phosphatase panerns in deep cerebral and brainstem scruccures
of the post mortem human brain. J Neuroparhol Exp Neurol
4. Tinker JH, Roberts SL. Management of cardiopulmonary bypass. In: Kaplan JA, ed. Cardiac anesthesia, 2nd ed. Orlando,
FL:Grune & Stratton, 1987:895-926
15. Edmunds LH, Williams W. Microemboli and the use of filters
during cardiopulmonary bypass. In: Utley JR, Betleski R, eds.
Pathophysiology and techniques of cardiopulmonary bypass.
Baltimore: Williams & Wilkins, 1983:lOl-114
16. Murkin JM, Farrar JK. The influence of pulsatile vs nonpulsatile
cardiopulmonary bypass on cerebral blood flow and cerebral
metabolism. Anesthesiology 1989;7 1:A41 (Abstract)
17. Prough DS, Stump DA, Troost BT. PaCOLmanagement during
cardiopulmonary bypass: intriguing physiologic rationale, convincing clinical data, evolving hypothesis? Anesthesiology 1990;
72:3-6 (Editorial)
18. Johnston WE, Vinten-Johmsen J , DeWitt DS, et al. Effect of
PCOz on cerebral blood flow during canine cardiopulmonary
bypass. Anesthesiology 1980;71:A553 (Abstract)
19. Blauth CI, Arnold JV, Schulenberg WE, et al. Cetebral microembolism during cardiopulmonary bypass. J Thorac Cardiovasc Surg 1988;95:668-676
20. Padayachee TS, Parsons S, Theobold R, et al. The detection of
microemboli in the middle cerebral artery during cardiopulmom r y bypass: a transcranial Doppler ultrasound investigation
using membrane and bubble oxygenators. Ann Thorac Surg
21. Steen PA, Newberg L, Milde JH, Michenfelder JD. Hypothermia and barbiturates: individual and combined effects on canine
cerebral oxygen consumption. Anesthesiology 1983;58:527532
22. Marcus ML, Heisrad DD, Ehrhardt JC, Abboud FM. Total and
regional cerebral blood flow measurement with 7-lo-, 15-, 25-,
and 50-pm microspheres. J Appl Physiol 1976;40:501-507
23. Heistad DD, Marcus ML, Busija DW. Measurement of cerebral
blood flow in experimental animals with microspheres: applications of the method. In: Passonneau JV, Hawkins RA, Lust
Moody et al: Brain Emboli after CPB 485
WD, Welsh FA, eds. Cerebral metabolism and neural function.
Baltimore: Williams & Wilkins, 1980:202-2 11
Moody DM, Bell MA, Challa VR. Capillary aneurysms and
other microvascular alterations surrounding brain hemorrhage. J
Neuropathol Exp Neurol 1989;48:3 19 (abstract)
Challa VR, Bell MA, Moody DM. A combined H & E, alkaline
phosphatase and high resolution microradiographic study of
lacunes. Clin Neuropathol 1990; in press
Challa VR, Bell MA, Moody DM. Prevalence and significance
of Charcot-Bouchard aneurysms. A combined H & E, alkaline
phosphatase and microradiographic study. Proceedings of the
14th International Congress of Neurology, New Delhi, 1989:
210 (abstract)
Ghatak NR. Pathology of cerebral embolization caused by
nonthrombotic agents. Hum Pathol 1975;6:599-610
Penrp JK, Cordell AR, Johnston FR, Netsky MG. Cerebral
486 Annals of Neurology
Vol 28
No 4 October 1990
embolism by Antifoam A in a bubble oxygenator system: an
experimental and clinical study. Surgery 1960;47:784-794
Orenstein JM, Saro N, Aaron B, et al. Microembolism observed
in deaths following cardiopulmonary bypas surgery: silicone
antifoam agents and polyvinyl chloride tubing as sources of emboli. Hum Pathol 1982;13:1082-1090
Mder JA, Fonkalsrud EW, Latta HL, Maloney JV. Fat embolism associated with extracorporeal circulation and blood transfusion. Surgery 1962;51:448-451
Hekmatpanah J. Cerebral microvascular alterations in arterial air
embolism. In: Cerv6s-Navarro J, Berz B, Ebhardt G, et al, eds.
Pathology of cerebrospinal microcirculation. Adv Neurol
Helps SC, Parsons DW, Reilly PL, Gorman DF. The effect of
gas emboli on rabbit cerebral blood flow. Stroke 1990;2194-
Без категории
Размер файла
1 120 Кб
surgery, brain, aortography, cardiaca, microemboli
Пожаловаться на содержимое документа