1. Art and Diseño

METABOLIC AND BLOOD COAGULATION CHANGES DURING
INTRAVENOUS FAT INFUSIONS
Arris and Gale Lecture delivered at the Royal College of Surgeons of England
on
12th July 1967
by
D. J. Reid, D.M., M.Ch., F.R.C.S.
Consultant Surgeon, Brighton, Lewes and Mid-Sussex Hospitals
IN SURGERY TO-DAY, emphasis has moved from speed in operative surgery
to the overall care of the patient during the whole period of treatment.
Much attention has been paid to problems such as anaesthesia, restoration
of fluid and electrolyte balance, the haemodynamics of shock and the
protection of blood flow to the kidneys. Of equal importance, however,
and often neglected, is the nutrition of the patient; and although periods
of relative starvation are well tolerated by the majority of patients undergoing routine surgical procedures, there are cases where adequate attention
to this aspect of treatment may prove crucial to the survival of the patient.
During all the stages of treatment in surgical patients there is an increase
in demand for calories, which is often far greater than realized. Three to
four times the number of calories required for a normal patient at rest
may be required for the same patient in this early post-operative phase.
Fever increases metabolic rate. Destruction of body tissues has to be
made good by protein synthesis which requires energy. Wound healing
requires protein and energy. At these times oral intake of food is often
impossible, especially in operations on the gastro-intestinal tract. The
deficiency is therefore more acute as demand increases and supplies are
withheld.
The body withstands starvation by drawing on reserves. At first
glycogen stores in muscle and liver are utilized, but these are soon exhausted. The supplies of liver glycogen, which amount to only an average
of 400 Gm., are rather less than the average daily intake of carbohydrate.
Energy requirements are then met by the catabolism of fat and body
protein.
The length of time which any individual can survive a period of total
starvation naturally depends on the extent of available body stores of fat
and protein. Women's greater resistance to starvation was observed at
Stalingrad in 1942 and in Holland in 1945. This is presumably due to the
greater body stores of fat in women.
For the well nourished with large stores of body fat, it is possible to
remain in positive nitrogen balance for a considerable time on severely
deficient calorie diets, provided a small quantity of protein is available in
the diet. For the poorly nourished, this period is greatly reduced and
322
METABOLIC AND BLOOD COAGULATION CHANGES
both the hazards of a surgical operation and the period of convalescence
will be increased. If a patient is maintained for a long period in a state of
malnutrition, hepatic dysfunction and hypoproteinaemia will follow.
Wound healing is delayed, decreased tolerance to shock producing circumstances occurs, patients may exhibit an instability of blood pressure disproportionate to any actual haemorrhage and may respond unsatisfactorily
to transfusion of blood and plasma.
The gastro-intestinal tract is undoubtedly the best route to administer
nourishment, but often this is impossible at the time when it would be most
beneficial, so the intravenous route becomes necessary as a method of
patient feeding.
To provide 3,000 Cal. intravenously in the form of 5 per cent dextrose
would require the administration of 15 litres. Fat when completely
metabolized can yield 9 Cal./Gm. Intralipid 20 per cent, a new form of
intravenous fat emulsion containing glycerol, provides 2,000 calories per
litre.
A form of intravenous fat which is safe has long eluded the researcher.
The reason for this is that, in the normal human, fat exists in the circulation
in a highly dispersed state, 1 micron or less in diameter, the droplets
exhibiting Brownian movement. For a fat emulsion to be tolerated in the
body, it must exist in this form. In order to achieve this state of dispersion,
it is necessary to lower the surface charge on the fat droplets by utilization
of an emulsifying agent.
It is the toxicity of the emulsifying agents, which appear to be toxic in
direct relationship to their value as emulsifying agents, that has limited the
development of fat emulsions. A number of emulsions are now available
which appear to be well tolerated, although it would be unwise to accept
their complete safety yet.
The material to be presented was carried out first to confirm that fat
administered intravenously has an effect on body metabolism and secondly
that its administration does not adversely affect the process of blood
coagulation. The emulsion Intralipid, manufactured by the Vitrium
Company of Sweden, has been used throughout this work. Intralipid is a
soya bean oil in water emulsion with egg yolk phosphatide 1.2 per cent as
the emulsifying agent. Glycerol 2.5 per cent is added to maintain isotonicity. 150 ml. N. sodium hydroxide is added to 500 litres of the emulsion
to counteract the effect of free fatty acids liberated from the emulsion
during storage. The particle size is approximately 1 micron in diameter.
The emulsion has a shelf life of approximately one year. When exposed
to high temperatures free fatty acids are liberated spontaneously from the
emulsion and it is therefore necessary to store the material at 4° C. There
is, however, no free fatty acid present in the emulsion when it is adequately
stored. Fractionation of the phospholipid fraction present in the egg yolk
phosphatide shows that 30 per cent is present in the form of phosphatidyl
ethanolamine and the remainder is lecithin.
323
D. J. REID
The metabolic action of Intralipid has been studied in two ways. First,
nitrogen balance studies have been carried out and, secondly, changes in
oxygen consumption and respiratory quotient have been measured in
patients receiving the emulsion.
In interpreting the results of nitrogen balance studies there are various
points which must be borne in mind.
1. A period of adjustment is required before a balance experiment
can be started, to ensure that the subject is in a stable metabolic state and
has come into equilibrium on a stabilized diet.
2. Estimations of intake calculated on the weight of various items constituting the diet are not satisfactory. A stable oral intake cannot be
relied upon as the protein content of various foods may vary considerably.
Intake must be estimated by the technique of sampling and measured
directly for its nitrogen content.
3. There may be spontaneous variations in nitrogen output which
make measurements over a short period unreliable.
4. Adequate controls are necessary for interpretation of results. This
makes experiments during the operative phase extremely difficult to plan
and assess as there is no way of comparing two identical patients in the
same situation. A surgical operation in the middle of a balance period
introduces so many unknown quantities, such as increased energy
demand and degrees of surgical trauma, that it is difficult to interpret results
which include this phase.
5. Experiments are often carried out on the assumption that faecal
excretion of nitrogen remains constant. This cannot be accepted where
finer differences in nitrogen balance are expected.
6. Frequently in experimental work published on this subject, it is
stated that nitrogen requirements were supplied intravenously in the form
of plasma, albumen or protein hydrolysates. These substances may
remain in the body tissue fluids for some time before being excreted or
metabolized. They may therefore give a false impression of positive
nitrogen balance over short periods, especially when compared with groups
who were not supplied in this way.
In a recent critical analysis of metabolic balance studies by Isaksson and
Sjogren (1964), attention was drawn to nitrogen excretion in the sweat
as a cause of false balance studies. Loss of nitrogen in the sweat might
amount to anything between 20 and 130 mg. per 100 ml. sweat. Mitchell
and Hamilton (1949) observed that, with a daily intake of 98 Gm. protein,
the nitrogen excretion in sweat of men profusely sweating averaged 152 mg.
per hour or 22.5 per cent of the total daily output. During conditions of
minimal or normal sweating 15 mg. per hour occurred or 2.7 per cent of
the total daily output. The changes are important in assessing where
equilibrium lies and investigations must be carried out in air-conditioned
surroundings at a constant temperature and humidity.
324
METABOLIC AND BLOOD COAGULATION CHANGES
With these points in mind the following investigation has been carried
out.
Six patients understanding the experimental nature of these investigations were admitted to the Metabolic Ward of St. Thomas's Hospital.
It was necessary to choose patients first who would benefit from intravenous feeding and secondly who were in stable states of metabolism and
not suffering from any form of gastro-intestinal disturbance that might
produce diarrhoea or vomiting.
TABLE I
THE COMPOSITION OF THE ORAL DIET THROUGHOUT THE EXPERIMENTAL PERIOD IN THE
SIX PATIENTS STUDIED
Case
No.
Oral Diet
throughout
Experiment
(g. per day)
Protein
Fat
Carbohydrate
Protein
Total Calorie
Intake during
Calorie
Periods of
Value
Intravenous Feeding
(calories per day) (calories per day)
1983
983
39
47
106
2604
2
57
1604
Fat
63
Carbohydrate 189
39
1901
3 Protein
901
Fat
10
Carbohydrate 163
2700
4 Protein
46
1200
37
Fat
Carbohydrate 165
37
1495
2495
5 Protein
70
Fat
Carbohydrate 167
39
1905
6 Protein
905
11
Fat
Carbohydrate 162
During periods 7 and 8 in Case 6 the otal intake was increased to: protein 40 g.
fat 76 g., and carbohydrate 245 g., yielding 1,824 calories per day.
1
Each patient was stabilized for five days on the diet of known composition before any measurements were made. Each patient was studied in
five-day periods, over a total period of 30 days in Cases 1-5, and 40 days in
Case 6. The patients received oral diet as shown in Table I. The first
two five-day periods were control periods during which the patients
received the oral diet only. Water was allowed freely throughout the
periods of study. During the second two five-day periods the oral diet
continued unchanged but the patients received additional calories in the
form of intravenous fat bringing the total calories administered to the
levels indicated in the right-hand column. In the sixth case there were an
additional two five-day periods of study during which the additional
calories supplied intravenously in periods 3 and 4 were supplied orally in
periods 7 and 8. Daily urine collections were made from 10 a.m. to
325
D. J. REID
10 a.m. The patients started each 24 hour urine save at 10 a.m. with an
empty bladder. At the end of each 24-hour period, the volume of urine
passed in that period was measured and a sample set aside for nitrogen
estimation. The Kjeldahl technique was used for nitrogen estimation.
Faecal saves were collected in five-day periods, using a carmine-dye
marker to indicate the end of one period and the beginning of another.
CAR 1. A.c OaOr DUCENAL
2mwth w
CAn oT
'ADA *
N
Ca
wAm
CAK
a fwS
d~so
am
?C .'M~ ~ ~ ~ ~ ?f
C^#4TOCOCOFSMI4CHSl NJO
CCi_
bE*.C
IICE
(b)
(a)
G%e ~
016
16
(d)
~
(c)
~
~~.
9
"F
( b)
II
=l
(c)
Fig. 1. Charts I to 6 demonstrate urinary nitrogen excretion (Gm./day), nitrogen
balance and fat balance in six patients. Intravenous fat was administered between
days 10 and 20 in each patient. In the sixth patient, additional oral calories were
administered between days 30 and 40.
The charts in Figure I demonstrate the results obtained. The top row
of results show the daily urinary nitrogen excretion measured in Gm.
nitrogen per day. From the 10th to the 20th day (periods 3 and 4) daily
intravenous infusions of fat were administered as shown on the charts.
The emulsion was administered over a period of approximately 7 hours.
The middle row of results show the overall nitrogen balance in five-day
periods. Nitrogen intake is measured from the zero line downwards and
nitrogen excretion is measured from the intake line upwards. Hence the
326
MM
METABOLIC AND BLOOD COAGULATION CHANGES
patient is in positive balance when the total nitrogen excretion is less than
the total intake. In this case the nitrogen excretion line will fall below the
zero line. When the patient is in negative nitrogen balance the nitrogen
excretion line will fall above the zero line.
The lowest row of results demonstrates fat balance. Intake is measured
from the zero line upwards and output from the zero line downwards.
This part of the investigation was carried out to investigate whether faecal
fat excretion changed significantly during the periods of intravenous fat
feeding.
In Figure 1, Chart 1 demonstrates a patient in borderline balance but
slightly in negative balance in the two control periods. In the periods
during which fat was administered intravenously definite positive balance
has been achieved. In the final two control periods the patient is again in
borderline balance. No significant change in faecal fat excretion occurs
during periods 3 and 4.
Chart 2 demonstrates a patient in positive balance throughout the study,
though in borderline balance in periods 1 and 6. During periods 3 and 4
there is a slight improvement in the degree of positive balance. Again no
change occurs in faecal fat excretion during the periods of intravenous fat
feeding.
In Chart 3 the patient is in strong negative balance in the control periods
before and after intravenous fat feeding but is in positive balance during
the period of intravenous feeding. No change occurs in faecal fat excretion during periods 3 and 4.
The patient shown in Chart 4 is in strong positive balance throughout
the study. No real change has occurred during the period of intravenous
feeding and again no change in faecal fat excretion occurred in periods
3 and 4. This patient received 150 Gm. fat daily as against 100 Gm. daily
in the other patients.
Chart 5 demonstrates another patient who is in positive nitrogen balance
throughout and in whom no real change occurs either in nitrogen balance
or in faecal fat excretion during the period of intravenous fat feeding.
In the sixth patient it is seen that a strong negative balance exists in the
first control period while the patient is receiving 900 Cal. daily by mouth.
When the patient receives additional calories intravenously to bring the
total intake to 1,900 Cal. he comes into strong positive balance but returns
to a negative balance in the second control period. In periods 7 and 8
the same additional calories as were administered intravenously in periods
3 and 4 are administered orally and a similar positive nitrogen balance
is achieved.
These studies show that the patients who are in strong negative balance
in the control periods (3 and 6) come into strong positive balance when
given intravenous fat. Cases 1 and 2 are on the borderline between
positive and negative balance in the control periods and some slight in327
D. J. REID
crease in nitrogen retention occurs during the periods of intravenous
feeding. Cases 4 and 5 were in positive balance throughout the study
and no further improvement in nitrogen excretion occurred during the
periods of intravenous feeding.
No fever was recorded in any patient throughout the studies and no
toxic symptoms were observed.
TABLE II
Table II shows the Respiratory Quotient (R.Q.), Oxygen Consumption (O.C.) in
ml./min., and Metabolic Rate (M.R.) expressed in percentage change from the preinfusion level. The pre-infusion results are the mean of three separate measurements.
Patients 1 to 5 received 1.5 Gm. fat per kg. of body weight in one hour. Patients 6 to 8
received 2 Gm. fat per kg. of body weight in one hour, and patients 9 and 10 received
3 Gm. fat per kg. body weight in one hour.
Case
1
2
3
4
5
6
7
8
9
10
Pre-infusion
R.Q.
O.C.
M.R.
R.Q.
O.C.
M.R.
R.Q.
O.C.
M.R.
R.Q.
O.C.
M.R.
R.Q.
O.C.
M.R.
R.Q.
O.C.
M.R.
R.Q.
O.C.
M.R.
R.Q.
O.C.
M.R.
R.Q.
O.C.
M.R.
R.Q.
O.C.
M.R.
0.88
220
0
1.11
157
0
0.85
166
0
0.81
206
0
0.92
225
0
0.72
156
0
0.80
206
0
0.79
201
0
0.82
242
0
0.75
171
0
1
0.86
226
+2.5
1.12
178
+16.7
0.83
200
+20.3
0.74
212
+2.4
0.79
256
+13.4
0.70
184
+18.1
0.74
232
+12.8
0.7.8
208
+3.1
0.75
260
+8.7
0.73
186
+7.6
Hours after Start of Infusion
2
3
4
5
0.95
0.80
0.81
0.88
262
263
268
232
+21.3
+18.0
+19.2
+5.6
0.94
1.14
0.97
1.02
188
178
183
173
+23.1
+16.7
+20.6
+14.1
0.88
0.82
0.82
0.88
208
168
183
187
+26.1
+1.2
+10.7
+13.2
0.82
0.94
0.78
0.97
208
246
219
198
+1.2
+20.1
-3.5
+7.1
0.83
0.78
0.85
0.82
281
271
292
268
+23.2
+18.3
+28.1
+19.6
0.93
0.74
0.76
0.84
184
194
177
193
+16.6
+25.6
+21.5
+12.8
0.80
0.89
0.77
0.84
209
245
230
230
+1.5
+18.1
+11.6
+11.6
0.84
0.82
0.92
0.84
207
226
201
217
+3.1
12.4
0
+7.3
0.78
0.77
0.77
0.74
287
276
285
263
+19.5
+9.8
+12.0
+18.4
0.74
0.75
0.75
0.76
189
195
191
191
+8.7
+11.5
+11.5
+13.0
6
-
-
0.90
221
+7.1
0.96
265
+15.8
-
-
0.73
267
+28.4
-
-
0.78
262
+7.6
-
-
Further proof of utilization of Intralipid comes from a study of oxygen
consumption and respiratory quotient in patients receiving the fat
emulsion.
Ten patients who received the emulsion were studied throughout the
period of infusion and for several hours afterwards (Table IL). Patients
1-5 received 1.5 Gm. fat/Kg. body weight over 1 hour. Patients 6-8
received 2 Gm. fat/Kg. body weight and patients 9 and 10 received
3 Gm. fat/Kg. body weight over a period of 1 hour. The pre-infusion
values shown are the mean of three separate estimations. Patients have
been studied at hourly intervals after start of the infusions. A fall in
respiratory quotient towards 0.7 would indicate the metabolism of fat.
The respiratory quotient results all show some fall during or after the
infusion compared with the average figure before the start of the infusion.
328
METABOLIC AND BLOOD COAGULATION CHANGES
Some patients had a low resting respiratory quotient. This was noted
particularly in debilitated patients who were suffering from carcinomatosis
(cases 6 and 10). A low respiratory quotient in these patients might be
expected as they are probably burning endogenous fat. In these patients
a further drop in respiratory quotient has occurred although the fall is
small (0.72 to 0.70 and 0.75 to 0.73 respectively). The fall in respiratory
quotient does not persist throughout the experiments and, although the
oxygen consumption usually remains elevated throughout the period of
study, the respiratory quotient often rises later in the experiment, in some
cases above the resting level. In one patient a respiratory quotient above 1
was recorded which must be due to persistent overbreathing.
TABLE III
RESPIRATORY QUOTIENT, OXYGEN CONSUMPTION, AND PERCENTAGE CHANGE IN
METABOLIC RATE OF PATIENT GIVEN INTRAVENOUS SALINE AND INTRALIPID ON
Two SEPARATE DAYS
Saline Infusion (510 ml. over 90 min.)
Respiratory quotient ..
Oxygen consumption
(ml. per min.)
Percentage change in
metabolic rate ..
Pre-infusion
(average of 3
Hours after Start of Infusion
1
2
3
4
readings)
0.75
0.79
0.76
0.77
0.75
268
245
243
264
272
0
-8
-8
-2
+2
5
0.78
261
-3
Intralipid Infusion (1.5 g. per kg. over 90 min.)
Pre-infusion
(average of 3
Hours after Start of Infugion
1
2
3
5
4
readings)
0.77
0.74
0.72
0.72
0.72
0.74
Respiratory quotienit ..
277
257
245
275
275
235
Oxygen consumpticin ..
(ml. per min.)
Percentage changeIin
0
+8
+17
+18
+17
metabolic rate ..
+4
Oxygen consumption increases in all patients studied. Metabolic rate
increases by an average of 21 per cent in the 10 patients, indicating a specific
dynamic action of fat. The emulsion is acting as a metabolic stimulus.
When a single patient was studied with a saline infusion and an Intralipid infusion it was found that metabolic rate was increased only in the
patient receiving the fat emulsion (Table III).
It has been shown that the oral ingestion of fat causes changes in certain
blood coagulation tests in the direction of increased coagulability.
Chylomicrons, blood platelets, certain phospholipids and some fatty
acids all increase the coagulability of blood. Certain phospholipids,
notably phosphatidyl serine and phosphatidyl ethanolamine, are highly
thromboplastic. Platelets contain these phospholipids and one of them,
phosphatidyl ethanolamine, is present in Intralipid.
There is a theory that the process of fibrin formation and fibrinolysis is
not only a process which occurs following damage to a blood vessel wall,
but that it is a continual process throughout the circulation.
329
D. J. REID
This theory is not supported, however, by any real evidence. There is
conflicting. evidence about the effect of lipoproteins on the process of
fibrinolysis. Cholesterol appears to be inhibitory in fibrinolysis whereas
neutral fat is not. The effect of Intralipid on various blood coagulation
tests, including the process of fibrinolysis, has been studied. When the
test is carried out on blood drawn from a patient who has received the
emulsion intravenously the test is referred to as an in vivo test; when the
emulsion is added to blood or plasma in a test tube the test is referred to as
an in vitro test.
The tests used in this study are the Partial Thromboplastin Time test
(Langdell et al., 1953; Matchett and Ingram, 1965), the Thrombin Generation test (Macfarlane and Biggs, 1953), the Dilute Plasma Clot Lysis test
(Mitchell and Briers, 1959) and a test of clot retraction measured by the
wire spiral method (Macfarlane, 1939).
In the Partial Thromboplastin Time (P.T.T.) test the coagulation mixture consists of platelet poor plasma (P.P.P.), Inosithin (1 mg./ml.), a
buffer solution amino tris hydroxymethyl methane and calcium chloride
solution. The effect of Intralipid on this coagulation test has been studied
TABLE IV
INTRALIPID AS A SOURCE OF PHOSPHOLIPID IN THE PARTIAL THROMBOPLASTIN TIME TEST
Clotting Times
Source of Phospholipid
(sec.)
(reagent concentrations)
62
..
..
..
..
None (buffer substituted) ..
..
..
.. ..
32
..
Inosithin 1 mg./ml ..
..
..
..
58-61
Intralipid 2 per cent-20 per cent ..
28-29
Inosithin and Intralipid together in above concentrations
by varying the phospholipid present in this mixture and recording the
clotting time in seconds. These studies represent the mean values of 10
separate tests. When no phospholipid is present in the mixture the mean
coagulation time is 62 sec. When Inosithin is present it is 32 sec. When
Intralipid in concentration of 2 per cent to 20 per cent is added it does not
significantly shorten the coagulation time compared to the times recorded
in the tubes containing no phospholipid. When both Inosithin and
Intralipid are present, the clotting time is similar to that when Inosithin
alone is present (Table IV). It would appear, therefore, that Intralipid,
even in concentrations exceeding those that would be obtained in patients
receiving the emulsion, does not shorten the coagulation time in the Partial
Thromboplastin Time test.
When the two-stage Thrombin Generation test is employed a change is
seen when Intralipid is introduced into the mixture.
In the first stage of this test 1 ml. citrated whole blood or 0.5 ml. Platelet
Poor Plasma or 0.5 ml. Platelet Rich Plasma are mixed with 0.5 ml. CaCl2
solut on (0.025 M.). At 1 minute intervals after this mixture is made
0.1 ml. are subsampled into tubes containing 0.4 ml. fibrinogen solution
330
METABOLIC AND BLOOD COAGULATION CHANGES
(0.2 G./lOO ml.) and the clotting time is recorded. The results of thrombin
600
generation tests are plotted as
against the sampling
clotting time in sec.
time in minutes. The reciprocal of the clotting time is more or less proportional to thrombin concentration.
It has been shown previously (Amris et al., 1964) that both the rate and
quantity of thrombin generation is increased by the addition of Intralipid
to whole blood, platelet rich or platelet poor plasma and these findings are
confirmed. The fat emulsion appears to replace the action of platelets.
Fat might be producing this platelet replacement action by activating the
contact mechanism of blood coagulation. In order to test this hypothesis
the Thrombin Generation Test was carried out using blood samples drawn
before and after an infusion of Intralipid. Citrated blood samples were
taken in one case using glass equipment and in the other using siliconized
oo
THROMBIN GENERATION TEST
BEFORE AND AFTER INTRALIPID INFUSION
20
20
,ao
/
I0
1
2
3
4
S
l7
S
9
o
It.
12
TIME IN MINUTES
Fig. 2. Lines 1 and 2 represent the thrombin generation test carried out on blood
withdrawn with normal needles and glassware before and after infusion of Intralipid respectively. Lines 3 and 4 represent the test carried out on blood withdrawn
with siliconized equipment before and after Intralipid infusion respectively.
needles, syringes, pipettes and glassware. Hence in the samples taken into
ordinary glass, the contact system was activated, and in those taken in
siliconized equipment the contact system was not activated (Fig. 2).
Lines 1 and 2 represent the thrombin generation curves using blood
taken with ordinary glassware before and after Intralipid infusion respectively. Lines 3 and 4 represent the thrombin generation curves before
and after the addition of Intralipid in blood withdrawn using siliconized
equipment. If Intralipid was producing a thromboplastic action by
activating the contact system, line 4 would approach line 1 on the graph.
In fact very little change is noted in the curves with the addition of fat and
the curves which result from non-contact blood (3 and 4) are closely comparable to those which result from contacted blood (1 and 2). Intralipid
does not produce an increased rate of thrombin generation, therefore, by
activating the contact system of blood coagulation.
331
D. J. REID
THROMBIN GENERATION TEST
PLATELET POOR PLASMA WITH INCREASING CONCENTRATION
OF FAT EMULSION INTRALIPID
TIME IN MINUTES
Fig. 3. The thrombin generation test carried out on platelet-poor plasma. Lines
1 to 6 represent concentrations of Intralipid 0 per cent, 2 per cent, 4 per cent,18 per
cent, 12 per cent and 16 per cent respectively.
Using platelet poor plasma in the Thrombin Generation Test it is seen
that the effect of Intralipid on thrombin generation increases as the
concentration of Intralipid in the mixture increases (Fig. 3). Lines 1 to 6
represent concentrations of Intralipid of 0, 2, 4, 8, 12 and 16 per cent.
There is little further effect, however, after 12 per cent.
When the effect of Intralipid is compared to that of a number of other
intravenous solutions commonly in use it is seen that Intralipid exerts a'far
ALL TUBES CONTAIN
80
70
ML.PLATELET POOR PLASMA
I.ML.CALCIUM OILORIDE M/40
THROMBIN GENERATION
TEST
60
600/C.T.
50
40
30
20432
10
1
2
3
4
5
6
7
8
9
10
TIME IN MINUTES
Fig. 4. Thrombin generation test carried out on platelet-poor plasma with the
addition of: (1) aminosol fructose ethanol; (2) 12 per cent dextrose; (3) 10 per cent
aminosol; (4) normal saline; (5) 5 per cent dextrose; (6) tetracycline solution
(250 mg. in 500 ml. normal saline); (7) 20 per cent Intralipid.
332
METABOLIC AND BLOOD COAGULATION CHANGES
greater effect in increasing thrombin generation than any of the others
(Fig. 4). Line 1 represents the effect of aminosol fructose ethanol, 2 represents 12 per cent dextrose, 3 represents 10 per cent aminosol, 4 represents
normal saline, 5 represents 5 per cent dextrose, 6 represents tetracycline
solution 250 mg. in 500 ml. saline solution and 7 represents the effect of
20 per cent Intralipid.
Intralipid appears therefore to exert a thromboplastic action, but its
action is not great enough to be detected by the one-stage partial thromboplastin time. It can only be detected by the two-stage thrombin generation test. Intralipid therefore has an effect in increasing the rate and
quantity of thrombin generation but the effect is slight.
TABLE V
TEST FOR ANTIHEPARIN EFFECT OF INTRALIPID BY COMPARISON WITH PLATELETS
Buffeer
Plasma
Intralipid
181 sec.
153 sec.
Platelet rich plasma
> 10 min.
> 10 min.
Platelet poor plasma
The four combinations of platelet-rich plasma or platelet-poor plasma, vs Intralipid
or buffer, were incubated with heparin 3 units/ml. After 8 hr. thrombin and fibrinogen
were added and the above clotting times recorded.
It is interesting to pursue the action of Intralipid in replacing the action
of platelets in the thrombin generation test. Blood platelets are known
to have several actions, one of which is reversal of the action of heparin.
The effect of platelets and Intralipid are compared in this function. The
plan of the experiment is shown in Table V. The experiment involves
incubation of platelet-rich and platelet-poor plasma with heparin, with and
without Intralipid. After incubation, thrombin is added and from this
mixture samples are added to fibrinogen solution. It is seen that those
200
THE EFFECT OF PLATELETS
AND FAT EMULSION ON
CLOT RETRACTION
150
PERCENTAGE
CHANGE
100
IN
CLOT VOLUME
20 40 6b
80 100
PERCENTAGE PLATELET RICH PLASMA
PERCENTAGE CONCENTRATION OF
Fig. 5. The upper line represents the effect of increasing concentrations of fat
emulsion, the lower line the effect of increasing the percentage of platelet-rich
plasma.
333
D. J. REID
tubes which contained platelet-rich plasma have clotted whereas those
which contain platelet-poor plasma, whether they contain Intralipid or
not, have failed to clot. Hence fat emulsion does not replace the action
of platelets in their capacity to reverse the action of heparin.
Platelets also have an action in the process of clot retraction. Clot
retraction was therefore measured using the wire spiral method measuring
the volume of residual serum in a clotting mixture as a measure of the
degree of retraction of the clot. It is seen that as the concentration of
platelets increased in the mixture the clot volume decreases. As the concentration of fat emulsion increases clot retraction does not increase but
initially it decreases (Fig. 5). Therefore the fat does not replace the action
of blood platelets in the process of clot retraction.
DILUTE PLASMA CLOT LYSIS TEST
WITH VARYING cONaNTRATIONS OF FAT
ENJLSION
CONC.FAT MGC;%
I.
0
soo
200
2.
3.
1000
5. 2000
S.
4.
CwT 4.
LYSIS
TIME 3s.
3s
HUS2.
1/60 Vho '/o 1/120
DILUTIONS OF PLASMA
Fig. 6.
The action of Intralipid on the process of fibrinolysis was investigated
(Fig. 6). 9 ml. venous blood was citrated with 1 ml. 3.2 per cent sodium
citrate and immediately cooled at 4° C. At 40 C. the blood specimen was
centrifuged 2,500 r.p.m. for 10 min. At 25 min. after venepuncture
dilutions of the plasma 1/60, 1/80, 1/100 and 1/120 were made up. The
dilutions were made with tris buffer alone or with the buffer containing fat
emulsion in varying concentrations 200, 500, 1,000, 2,000 mg. per cent.
4 ml. samples of dilute plasma were incubated at 370 C. in a water bath
following the addition of 0.2 ml. thrombin. The time taken for complete
lysis of the clot to occur was estimated and Figure 6 shows the results
obtained. It is seen that increasing concentrations of fat emulsion cause
334
METABOLIC AND BLOOD COAGULATION CHANGES
TABLE VI
MEASUREMENT ON ARTERIAL BLOOD
A.L. d 45 CA. Colon
J.H. d 45 Perf. G.U.
3.0 gm./Kg. B. W. 70 miii.
3.0 gm./Kg. B. W. 70 min.
Art.
Art.
Time
Art.
Art.
Time
pH
pH
pCO2
min.
pCO2
min.
0
36.4
7.47
0
32.5
7.435
70
35.4
31.5
7.44
70
7.425
110
7.43
35.0
7.428
30.0
100
165
7.438
34.5
140
7.433
30.2
7.44
34.0
30.0
225
7.420
190
285
7.43
34.2
20.2
7.420
220
7.44
31.0
405
7.405
30.0
260
30.1
300
7.420
Changes in arterial pH and pCO2 in two patients receiving intralipid 3.0 gm./Kg.
body weight over 70 minutes. (Time is expressed in minutes after the start of the
infusion.)
only a slight change in clot lysis time in the direction of increased lysis and
would appear therefore not significantly to affect the process of fibrinolysis.
In the course of this work and other work not presented over 150
infusions of Intralipid have been given and no case of venous thrombosis
has been encountered.
Prolonged protein/fat feeding on a carbohydrate-deficient diet has been
shown to produce ketonuria. It is important to know whether an intravenous fat infusion alters the pH or pCO2 of the blood during infusion.
TABLE VII
pCO2
Standard
bicarbonate
35
22.0
Baseline 7.390
1
35
7.385
21.7
2
7.389
35
21.6
38
3
7.390
22.7
40
4
7.380
22.7
39
5
7.370
22.0
39
8
7.379
22.5
9
7.398
35
22.3
10
38
22.2
7.382
36
22.2
Baseline 7.395
1
38
7.389
22.7
37
2
7.381
22.0
38
3
7.388
22.6
4
7.385
39
22.7
36
5
7.395
22.4
8
7.410
40
24.5
9
7.389
38
22.6
10
34
7.410
22.0
Baseline 7.455
35.0
25.0
1
7.455
33.5
24.9
2
7.470
33.0
25.6
3
7.455
35.0
25.5
4
25.3
7.455
35.0
5
7.455
36.0
25.8
6
7.455
34.5
25.2
*Changes in pH, pCO2 and standard bicarbonate in three patients receiving 500 ml.
20 per cent Intralipid daily. The infusions were started each day at 9 a.m. and given
daily at a constant rate. Estimations were carried out daily at noon.
Day
pH
335
D. J. REID
Repeated tests on the urine of patients receiving this emulsion have failed
to show any evidence of ketonuria. In cases of metabolic acidosis, as
occur in uraemia, fat emulsions play an important supporting role in
management. If, however, the emulsion were to produce further acidosis,
its value would be reduced. An investigation was therefore undertaken to
measure changes in pH and pCO2 during intravenous fat infusions using
Intralipid. The measurements made using arterial blood are shown in
two patients each receiving 3.0 Gm./Kg. B.W. of fat over 70 min.
(Table VI). It is seen that there is virtually no change in pH or pCO2
during or after the infusions. Three further patients were studied daily
at the end of the intravenous infusion over a period of 10 days. It is again
seen that no real change in pH or pCO2 occur at any time throughout
the 10-day periods (Table VII).
CONCLUSIONS
In summary, therefore, the following conclusions are drawn.
1. Intralipid exerts a nitrogen-sparing action on those patients who are
in negative nitrogen balance. When patients in positive nitrogen balance
receive intravenous fat, no significant change in balance occurs.
2. Isocaloric replacement of intravenous calories with extra ora
calories produce the same sparing effect on nitrogen balance.
3. No significant change occurs in faecal fat excretion during the
intravenous administration of fat.
4. Intralipid causes a shift in the Thrombin Generation Test in the
direction of increased thrombin generation when fat is added to blood,
in vivo and in vitro. The coagulant action of Intralipid cannot be detected
by the one-stage coagulation test, the Partial Thromboplastin Time. The
coagulant action is therefore slight.
5. Intralipid mimics the action of blood platelets in the Thrombin
Generation test. It does not replace the action of platelets in the clot
retraction test. It is suggested that the changes demonstrated in the
Thrombin Generation test are due to phosphatidyl ethanolamine present
in the emulsifying agent of the emulsion.
6. No inhibitory action of Intralipid was noted in the process of
fibrinolysis using the Dilute Plasma Clot Lysis test.
7. Intralipid acts as a stimulus to metabolism causing increased oxygen
consumption, a fall in respiratory quotient and increased metabolic rate.
Apart from the nutritional value of the fat emulsion the increase in
metabolic rate would produce a sense of well-being in debilitated patients.
8. Ketosis has not been detected during Intralipid infusions and no
evidence of acidosis has been demonstrated.
These observations add to the evidence that this form of intravenous fat
is metabolized. It would appear to have a favourable effect on patients in
negative nitrogen balance. The emulsion exerts a mild coagulant action
but this is so slight that it can be ignored.
336
METABOLIC AND BLOOD COAGULATION CHANGES
ACKNOWLEDGEMENTS
I am grateful to Miss Gillian Wooding, Sister-in-Charge of the Metabolic
Unit, St. Thomas's Hospital, who cared for the patients undergoing metabolic study; to Professor F. T. G. Prunty, Dr. T. Pilkington, Dr. G. I. C.
Ingram, Dr. W. J. Griffiths and Dr. C. D. Marsden I am grateful for
much help.
REFERENCES
AMRIS, C. J., BROCKNER, J., and VAGN, L. (1964) Acta chir. 5cand., Suppl. 325, 70.
ISAKSSON, B., and SJOGREN, B. (1964) Excerpta med. 11, 1041.
LANGDELL, R. D., WAGNER, R. H., and BRINKHOUS, K. M. (1953) J. Lab. clin. Med. 41,
637.
MACFARLANE, R. G. (1939) Lancet, 1, 1199.
and BIGGS, R. (1953) J. clin. Path. 6, 3.
MATCHETT, M. D., and INGRAM, G. I. C. (1965) J. clin. Path. 18, 465.
MITCHELL, H. H., and HAMILTON, J. S. (1949) J. biol. Chem. 178, 345.
MITCHELL, J. R. A., and BRIERS S. M. (1959) Lancet, 2, 435.
3In e&moriam
CECILIA FLORA WEBB-JOHNSON, Baroness,
Dame of the Order of St. John of Jerusalem
Member of the Court of Patrons
LADY WEBB-JOHNSON died suddenly on 15th March during a holiday in
Teneriffe. She was the only child of Douglas Gordon MacRae and his
Lady Webb-Johnson
wife, Caroline Cecilia. In 1911 she married Alfred Webb-Johnson soon
after his appointment to the surgical staff of the Middlesex Hospital and
337