lack of a critical cardiac output and critical systemic oxygen delivery during low cardiac output in...

Lack of a critical cardiac output and critical systemic oxygen delivery during low cardiac output in the third trimester in the pregnant sheep

Wayne Evans, MD, Susan C. Capeile, MD, and Daniel I. Edelstone, MD

Pittsburgh, Pennsylvania

OBJECTIVE: We sought to determine whether a critical cardiac output and oxygen delivery exist in normal pregnancy. We also sought to determine the role of fetoplacental oxygen demand on maternal oxygen transport variables in response to decreased maternal cardiac output. STUDY DESIGN: We studied 10 adult female sheep, 5 nonpregnant and 5 pregnant. We placed a flow-directed thermodilution catheter in the pulmonary artery and a balloon-tipped catheter in the right atrium of the sheep. We also placed a catheter for pressure monitoring and blood sampling in the descending thoracic aorta in both the mother and fetus. We decreased maternal cardiac output by incremental inflation of the right atrial balloon. We measured maternal cardiac output by intravenous bolus thermodilution technique. We also measured maternal and fetal acid-base status and serum lactate concentrations. We calculated a variety of maternal cardiorespiratory variables, including systemic oxygen delivery, systemic oxygen consumption, and fractional whole body tissue oxygen extraction. RESULTS: The nonpregnant sheep displayed a critical cardiac output below which there was an abrupt decrease in oxygen consumption. In contrast, there was no critical level of cardiac output in the pregnant sheep. Maternal oxygen consumption was linearly dependent on cardiac output. Maximum fractional oxygen extraction was significantly lower in the pregnant sheep than in the nonpregnant sheep. CONCLUSION: States of low cardiac output in the pregnant sheep are associated with a lack of a critical cardiac output; the flow-dependent oxygen consumption observed is the result of either an impairment in tissue oxygen extraction or some degree of metabolic arrest or a combination of both. If this unique cardiac output-oxygen consumption relationship is seen in human pregnancy, it could have significant implications in the care of the critically ill obstetric patient. (Am J Obstet Gynecol 1996;175:222-8.)

Key words: Critical cardiac output, oxygen consumption, tissue oxygen extraction, metabolic arrest, pregnant sheep

Under normal conditions oxygen consumption is the indicator of an organism's ability to use oxygen at the

tissue level. A drop in oxygen consumption may denote an impairment of effective oxygen utilization. This may occur because of a decrease in oxygen delivery or blood oxygen content. During states of decreased oxygen deliv-

ery the organism maintains oxygen consumption by ex- tracting oxygen more effectively from the microcircula- tion. When the oxygen supply is satisfactory, oxygen con-

From the Division of Maternal-Fetal Medicine, the Department of Ob- stetrics, Gynecology, and Reproductive Sciences, Magee-Womens Hospi- tal/Magee-Womens Research Institute, University of Pittsburgh School of Medicine. Presented in part at the Fifteenth Annual Meeting of the Society of Perinatal Obstetricians, Atlanta, Georgia, January 23-28, 1995. Received for publication November 13, 1995; revised February 7, 1996; accepted February 12, 1996. Reprint requests: Wayne Evans, MD, University of Wisconsin-Madison Medical School, Milwaukee Clinical Campus, Sinai Samaritan Medical Cent~ Department of Obstetrics and Gynecology, 2000 W. Kilbourn Ave., P O. Box 342, Milwaukee, W1 53201-0342. Copyright © 1996 by Mosby-Year Book, Inc. 0002-93 78/96 $5. O0 + 0 6/1/72909

sumption is said to be supply or flow independent.

However, there is a point of decreased oxygen delivery or cardiac output below which oxygen consumption cannot be maintained, where oxygen extraction becomes impaired. This point is the critical cardiac output or critical oxygen delivery. Below this point oxygen consumption falls as oxygen supply is inadequate or oxygen demand

increases; oxygen consumption can only be returned to- ward normal by increasing either blood flow or blood oxygen content. 1 Critical cardiac output is a widely accepted concept and has been demonstrated in newborn, infant, and adult nonpregnant animals and humans.

In some pathologic disorders the normal relationship of oxygen delivery or cardiac output to oxygen consumption is lost. These disorders do not display the normal maintenance of oxygen consumption with decreasing oxygen delivery. Either oxygen consumption decreases at all levels of decreased oxygen delivery or the critical level of oxygen delivery is considerably higher than in the normal nonpathologic state. ~' ~ Examples of pathologic disorders that display this oxygen delivery (cardiac out-

222

Volume 175, Number 1 Evans, Capelle, and Edelstone 223 Am J Obstet Oynecol

put) to oxygen consumption alteration include the adult

respiratory distress syndrom e , sepsis, septic shock, mul- tiple trauma, severe pancreatitis, and extensive burns. This oxygen delivery-oxygen consumption alteration has also been shown to occur in preeclampsia, a pathologic

disorder of pregnancy? However, the relevance of a lack of a critical cardiac output in severe preeclampsia is un- known because the normal pregnancy state has not been

studied. Therefore the response of oxygen consumption to decreasing cardiac output in normal pregnancy is not known. We know that some of the physiologic features of pregnancy resemble some of the pathophysiologic features of adult respiratory distress syndrome. However,

adult respiratory distress syndrome is a pathologic disorder, whereas pregnancy is not. In this study we sought to determine the level of critical cardiac output and oxygen

delivery in pregnancy. We hypothesized that, because pregnancy is a normal condition, it should display tile

same oxygen delivery-oxygen consumption relationship noted in normal nonpathologic conditions.

Material and methods

The study protocol was approved by the Institutional Animal Care and Use Committee of the Magee-Womens Hospital and Research Institute. We studied 10 adult female sheep of mixed breed. We studied two series of

animals: (1) nonpregnant (n= 5) and (2) late-gestation

pregnant (n = 5). Gestational age was based on the number of days from date of conception. For this study we

used the method to gradually reduce maternal cardiac output developed by Fahey and Lister 5' 6 in experiments

on neonatal and infant lambs. Animal preparation. We anesthetized each animal with

2.5% thiopental and a mixture of 0.5% to 0.7% halothane (Abbott Laboratories, Abbott Park, Ill.) with 2.5 L /min oxygen. We catheterized the right atrium of each sheep

with a Bardex 20F Foley catheter with a 75 ml balloon .(C.R. Bard, Urologic Division, Covington, Ga.) through a right jugular venous cutdown. We also catheterized the

puhnonary artery by a percutaneous cannulation of the teft jugular vein with an 8.5F introducer sheath (Arrow,

Reading, Pa.) and a 7.5F Swan-Ganz thermodilution pulmonary artery catheter (Baxter-Edwards Critical Care Di- vision, Irvine, Calif.). We confirmed proper placement of each catheter by noting characteristic pressure wave- forms when the catheters were connected to a multichannel monitor and recorder (Beckman R611, Beckman In- struments, Shaumberg, Ill.) by way- of sampling stopcocks and transducers. We placed a polyvinyl catheter (Tygon, Akron, Ohio) in the descending thoracic aorta by a cut-

down of the left hind limb artery. We connected the aortic catheter to the multichannel monitor and ob-

served the aortic pressure tracing. We confirmed placement and function of the right atrial Foley catheter by inflating the balloon and noting a decrease in blood

pressure, indicating a decrease in cardiac output. We then deflated the balloon and noted a return of the aortic

pressure to baseline. All channels of the puhnonary artery catheter were flushed with heparin and saline solution and capped.

In the pregnant sheep we performed a laparotomy and hysterotomy after the maternal catheterizations were

completed. We placed a Tygon catheter in a fetal hind-

limb artery. Each sheep was given procaine penicillin and streptomycin before surgery and daily until the study was

terminated.

Study design. We studied each animal after a postoper- ative period of 1 or 2 days. We transferred the sheep to the study area by use of a holding pen. We connected the

aortic catheter, pulmonary artery catheter, and fetal arterial catheter to the multichannel recorder by way of sam-

pling stopcocks and transducers. The thermodilution

probe of the pulmonary artery catheter was connected to a cardiac output computer. The sheep was allowed to acclimate to its surroundings for approximately 1 hour.

We measured and recorded maternal aortic and pulmonary artery pressures continuously. In the pregnant subjects we also measured fetal heart rate and fetal arte-

rial blood pressure continuously. We measured maternal heart rate by either continuous cardiotachometry with the multichannel recorder, or we counted the number of

aortic pressure waves within a given period noted on the chart paper.

To determine whether the order in which we reduced

cardiac output mattered experimentally, we performed three experiments with sequential-ordered reduction in

cardiac output and threeexperiments with random-ordered reduction of cardiac output. The responses of all oxygen transport variables did not differ in both the

sheep who underwent random-ordered reduction and the sheep who underwent sequential-ordered" reduction.

Therefore we performed all subsequent experiments with sequential reduction of cardiac output because this shortened the experimental protocol considerably.

We collected data at baseline (right atrial balloon vol-

ume of zero) and with each right atrial balloon inflation.

We collected all samples and recorded all measurements when the sheep was quiet and physiologically in the steady state. We defined steady state by a stable blood pressure, heart rate, cardiac output, and mixed venous oxygen saturation for >3 minutes. I fa stable steady state was noted during this 3-minute period, we then waited an additional 30 minutes before collecting data. Maternal blood pressure and heart rate were measured from the aortic pressure tracing. Cardiac output was measured by an intravenous bolus thermodilution technique with normal saline solution at room temperature used as the indicator. Cardiac output was calculated by a cardiac out-

put computer (American Edwards Laboratory-COM-1, Baxter-Edwards Critical Care Division) connected to the

224 Evans, Capelle, and Edelstone July 1996 AmJ Obstet Gynecol

Table I. Baseline hemodynamic and cardiorespiratory variables

Nonpregnant Pregnant

No. Weight (kg) Cardiac output (ml/min/kg) Oxygen delivery (ml oxygen/min/kg) Oxygen consumption (ml oxygen/min/kg) Fractional oxygen extraction (%) Lactate concentration (mmol/L) Mean arterial pressure (ram Hg) Systemic vascular resistance (dyne/sec/cm -5) Heart rate (beats/rain) Hemoglobin concentration (gm/dl) Hematocrit (%)

5 5 56.40 _+ 4.41 55.42 _+ 4.1

126.00 _+ 16.73 146.34 -+ 12.51 16.02 + 1.80 15.78 -+ 1.71 4.44 _+ 0.65 6.74 ± 0.50* 0.27 ± 0.02 0.43 _+ 0.02* 1.60 _+ 0.20 2.06 + 0.23*

96.27 ± 4.71 84.73 + 3.50* 1058.24 ± 64.40 837.90 ± 82.00*

96.00 ± 7.60 106.60 ± 11.(t3" 10.68 ± 0.48 9.11 + 0.46* 33.00 ± 0.62 27.00 ± 0.30*

*p < 0.05 t test.

pulmonary artery thermistor. We obtained five cardiac

output measurements. We eliminated the highest and lowest measurements and averaged the remaining three values to determine cardiac output. Core body tempera-

ture was continuously monitored with the cardiac output computer. Blood samples were drawn from aortic and pulmonary artery catheters for gas tensions (Corning

Blood Gas Analyzer, International Laboratories, Lexing-

ton, Mass.), oxygen saturation and hemoglobin concentration (Radiometer OSM2 Hemoximeter, Radiometer-

Copenhagen), hematocrit, and lactate concentration. The experiment was terminated when either the arterial

blood pressure, cardiac output, or mixed venous oxygen saturation failed to reach a stable steady state at a new

balloon volume. Data analysis. We used standard formulas to calculate

the derived hemodynamic and cardiorespiratory vari-

ables, including mean arterial pressure and systemic vascular resistance. 7 Arterial oxygen content (Ca02) and

mixed venous oxygen content (Cv02) were also calcu-

lated by standard formulas. Cao 2 and Cvo 2 were used to calculate oxygen delivery, oxygen consumption, and frac-

tional tissue oxygen extraction. Cardiac output was measured by thermodilution. All values except mean arterial blood pressure and systemic vascular resistance were in-

dexed to body weight in kilograms. We compared mean baseline hemod}mamic and car-

diorespiratory variables for both the nonpregnant and pregnant groups. These data are expressed as means _+ SD and were analyzed statistically by paired t test,

p < 0.05 is considered statistically significant. We then divided the data into three arbitrary catego-

ries on the basis of the level of cardiac output: (1) baseline cardiac output, (2) <50% reduction in cardiac output, and (3) >50% reduction in cardiac output. We analyzed these data by two-way analysis of variance with p < 0.05 considered as statistically significant. If there was a significant difference by analysis of variance, we then analyzed the data with Newman-Keuls post hoc testing.

We then used a third method of analysis to determine

the points of critical cardiac output for each sheep and for each series of sheep (pooled data). To determine critical cardiac outputs, we analyzed the data continu-

ously and then plotted the data on an x-y graph with cardiac output on the x axis and oxygen consumption on the y axis. We used a technique described by Samsel and Schumacker, ~ which has been used and verified in other animal studies involving critical cardiac output) ' 6 This

technique consists of constructing pairs of regression

lines using all points once and only once to define the

response of oxygen consumption to decreased cardiac output. We chose the grouping of points resulting in minimum overall residual sum of squares as the best pair of regression lines. The point of intersection of these

regression lines defined the critical cardiac output and the oxygen consumption at these points.

Results

We studied two series of adult female sheep: nonpreg-

nant (n = 5) and late-gestation pregnant (n = 5). The gestational ages of the pregnant sheep ranged from 122 to

135 days' gestation from the date of conception. Baseline

hemodynamic and cardiorespiratory data are shown in Table I. When the groups of animals were compared, we found no statistically significant difference in baseline weight. As expected, baseline cardiac output was noted to be higher in the pregnant animal but was not statistically significant (p = 0.11). We found no significant difference

in baseline oxygen delivery between the nonpregnant and pregnant sheep. Oxygen consumption was significantly higher in the pregnant sheep (p = 0.02). Fractional

tissue oxygen extraction and serum lactate concentration were higher in the pregnant group. Systemic vascular resistance and hemoglobin concentration were lower in

the pregnant animals. Cardiac output and systemic oxygen delivery were di-

rectly correlated (Fig. 1). Therefore we analyzed all data in relationship to cardiac output. We analyzed oxygen consumption in relation to cardiac output (Fig. 2). We observed that the nonpregnant animals displayed a bi-

Volmne 175, Number 1 Evans, Capelle, and Edelstone 225 AmJ Obstet Oynecol

a ,

L.

). v

~ E

0

2 5 -

2 0 "

1 5 -

1 0 .

5 -

. / / ! | |

5 0 1 0 0 1 ~ 0

Cardiac Output ml/min/kg

I Nonpregnant

|

2 0 0

b.

L.

). v

r~ E

N 0

2 5 -

2 0 -

1 5 -

1 0 -

5 .

IPregnant I

# / / ! i i |

50 100 150 200


Fig. 1. Relationship of systemic oxygen delivery to cardiac output. Systemic oxygen delivery correlates directly with cardiac output in both nonpregnant (a) and pregnant (h) sheep (~ = 0.961 and 0.975, respectively).

phasic response of cardiac ou tpu t to oxygen consump-

tion (Fig. 2, a). The response of decreasing cardiac out-

pu t to oxygen consumpt ion in the p regnan t sheep was

l inear (Fig. 2, b). In o ther words, there was no critical

cardiac ou tpu t in the p regnan t group.

To more fully characterize these differences between

n o n p r e g n a n t and p regnan t sheep, we compared the

mean values of each variable in n o n p r e g n a n t and preg-

nan t animals in relat ion to cardiac ou tpu t levels at base-

line, _<50% reduc t ion and >50% reduc t ion (Table II). We

observed that oxygen delivery corre la ted directly with

changes in cardiac ou tpu t in both groups. The decrease

in oxygen consumpt ion was significantly greater in preg-

nan t than in n o n p r e g n a n t sheep. The n o n p r e g n a n t

sheep demons t ra ted a greater tissue oxygen extract ion

capacity than the p regnan t sheep did. This is fur ther

ev idenced in Fig. 3. Here we see the differences in tissue

oxygen extract ion in each series of animals. Mean maxi-

mal oxygen extract ion in the n o n p r e g n a n t animals was

a .

o

E--. ,="

~ E

x 0

10 -

7 .5 -

5=

2 .5 -

0

@

0 500 i i "

0 IO0 150 200

INonpregnant ]

o

=E ~0 t- O L~

>,. N

b.

e-

E

10 -

7 . 5 -

5 -

2 . 5 -


/ I 0 ' ! I J

0 1 0 0 1 $ 0 2 0 0

Cardiac O u t p u t ml/min/kg

Fig. 2. Response of oxygen consumption to decreasing cardiac output. Critical cardiac output in nonpregnant sheep (a) is 41 ml/min/kg. In pregnant sheep (b) no critical cardiac output level was demonstrated.

0.90 (range 0.52 to 0.99), whereas in the p regnan t ani-

mals the mean maximal oxygen extract ion was 0.74

(range 0.61 to 0.84).

We compared critical values of cardiac ou tput as a

funct ion of oxygen consumpt ion with the critical value of

cardiac ou tpu t as a funct ion of serum lactate concentra-

tion. In Fig. 4 we super imposed the cardiac ou tput versus

oxygen consumpt ion curves, as seen in Fig. 2, on the

graphs depict ing the response of serum lactate concen-

tration to decreasing cardiac ou tpu t in both the nonpreg-

nant (Fig. 4, a) and p regnan t (Fig. 4, b) sheep. In both

groups as cardiac ou tpu t was r educed serum lactate con-

centrat ion remained stable unti l a certain level o f cardiac

ou tput was reached. However, two differences were seen

between the n o n p r e g n a n t and p regnan t animals. The

first difference is that the level of cardiac ou tpu t that

corresponds to an abrupt rise in lactate concent ra t ion is

significantly h igher in the p regnan t animals compared

with the n o n p r e g n a n t animals (70 and 43 m t / m i n / k g ,


Table II. Compar ison of mean cardiorespiratory variables with varying degrees of decreased cardiac ou tpu t

I Baseline cardiac <50% Decreased cardiac >50% Decreased cardiac output output output

Cardiac output Nonpregnam 126.00 _+ 16.73 107.91 _+ 18.03" 30.43 _+ 8.41" Pregnant 146.34 _+ 12.51 104.66 _+ 17.51" 31.68 _+ 10.20"

Oxygen delivery Nonpregnant 16.02 -+ 1.80 13.81 _+ 2.09* 4.74 _+ 1.20" Pregnant 15.78_ 1.71 12.04 _+ 1.85",]- 4.63_ 1.28"

Oxygen consumption Nonpregnant 4.44 _+ 0.65 4.20 + 0.91 4.02 _+ 0.75 Pregnant 6.74 + 0.50]- 5.23 _+ 0.73",]- 2.81 --. 0.64',t

Oxygen extraction Nonpregnant 0.27 + 0.02 0.39 _+ 0.05* 0.80 _+ 0.07* Pregnant 0.43 -+ 0.02]- 0.45 _+ 0.04 0.64 _+ 0.05*,l-

Lactate concentration Nonpregnant 1.60 -+ 0.20 1.82 + 0.30* 6.97 + 2.41" Pregnant 2.06 _+ 0.23t 2.52 + 0.27 9.03 _+ 2.04*,]-

Analysis of variance, Neuman-Keuls, was used to determine significance. *p < 0.05, baseline versus <50% reduction in cardiac output versus >50% reduction in cardiac output in either pregnant or

nonpregnant. ]-p < 0.05, nonpregnant versus pregnant for each level of cardiac output.

respectively). Second, there is a dif ference in the re-

sponse of oxygen consumpt ion at cardiac outputs above

the break points no ted in the cardiac output - lac ta te con-

centrat ion curves. In the n o n p r e g n a n t sheep ne i the r lac-

tate nor oxygen consumpt ion was affected by reduced

cardiac ou tpu t at cardiac outputs above the break point,

whereas in the p regnan t sheep oxygen consumpt ion con-

tinually fell wi thout an associated rise in lactate concen-

tration.

Comment

Our study shows that critical cardiac ou tpu t and critical

systemic oxygen delivery do no t exist in the third trimes-

ter of the p regnan t sheep. This was unexpec t ed because

pregnancy is no t a pathologic disorder. We assumed that

the materna l cardiorespiratory response to a gradual re-

duct ion in cardiac ou tpu t would resemble that of the

n o n p r e g n a n t organism. In the normal n o n p r e g n a n t state

the cardiac ou tpu t -oxygen consumpt ion relat ionship is

biphasic, with a plateau or stabilization of oxygen con-

sumpt ion in response to decreasing cardiac ou tpu t until a

critical cardiac ou tpu t is r eached and oxygen consump-

t ion falls. Ia We did no t see this biphasic cardiac o u t p u t -

oxygen consumpt ion in the p regnan t sheep. Instead we

saw a l inear cardiac ou tpu t -oxygen consumpt ion rela-

t ionship. The re are two possible explanat ions for this

un ique cardiac ou tpu t -oxygen consumpt ion relat ionship

we observed in response to decreasing cardiac ou tpu t in

the p regnan t sheep.

The first explanat ion is impa i rmen t of tissue oxygen

extraction. U n d e r condi t ions of decreased cardiac out-

put, the p regnan t animal 's ability to extract oxygen f rom

the microcircula t ion is impaired. We know that in te rm

pregnancy baseline oxygen extract ion may be increased

as compensa t ion for the r ightward shift in the oxyhemo-

globin dissociation curve associated with normal preg-

nancy. 9 Also, tissue oxygen extract ion normal ly increases

as a response to the decreased hemoglob in concent ra t ion

and arterial oxygen con ten t seen in pregnancy. 1°-13 We did

no t know the ability of the p r e g n a n t sheep to increase

tissue oxygen extract ion in response to decreasing car-

diac output. If impa i rmen t of tissue oxygen extract ion is

the mechan i sm for this cardiac ou tpu t -oxygen consump-

tion de rangement , we would expect to see a difference in

oxygen extract ion in response to reduced cardiac ou tput

between the n o n p r e g n a n t and the p regnan t sheep. Fig. 3

demonstra tes the dif ference in tissue oxygen extract ion

between p regnan t and n o n p r e g n a n t sheep. We observed

that the n o n p r e g n a n t sheep had a h igher mean maximal

oxygen extract ion at 0.90 than the p regnan t sheep, whose

mean maximal oxygen extract ion was 0.74 (p< 0.05).

These data suggest that impa i rmen t of tissue oxygen ex-

traction is at least partly responsible for the lack o f a

critical cardiac ou tpu t in the p regnan t sheep.

The second possible explanat ion for this lack of a criti-

cal cardiac ou tpu t in normal pregnancy is a p h e n o m e n o n

called "metabol ic arrest. ''14 Dur ing metabol ic arrest tis-

sues stop using oxygen but do no t resort to anaerobic

metabolism. Oxygen consumpt ion falls, yet there is no

tissue hypoxia or metabol ic acidosis. This is demon-

strated in Fig. 4, which depicts the response of serum

lactate concent ra t ion to decreasing cardiac output. The

data in the n o n p r e g n a n t sheep show that, as cardiac

ou tput is reduced, the serum lactate concent ra t ion does

no t change unti l a certain cardiac ou tput level is reached,

below which lactate concent ra t ion abruptly rises. Not ice

that this break po in t in lactate concent ra t ion coincides

with the critical cardiac ou tpu t in relat ion to oxygen

consumpt ion. In the p regnan t sheep, as cardiac ou tpu t

was reduced, lactate concent ra t ion r ema ined stable unti l

Volume 175, Number 1 Evans, Capelle, and Edelstone 227 AmJ Obstet Gynecol

a . e- I - o ,,,m

{o [,- 0 . 7 5 -

N

c - O . 5 -

0 . 2 5 -

I-- o

,.>... INonpregnant I

510 i i i 100 150 200

c-

O

U

L.

x LaJ =

×

m ,p,. b -

b° 1 -

0 . 7 5 -

0 . 5 -

0 . 2 5 -


IPregnant I

0 510 , t i 0 100 150 200

Cardiac Output m l / m i n / k g

Fig. 3. Response of whole body tissue oxygen extraction to decreasing cardiac output. Nonpregnant sheep (a) had lower baseline tissue oxygen extraction than pregnant sheep (h) (0.27 and 0.43, respectively, p < 0.05) but a higher mean maximal tissue oxygen extraction (0.90 and 0.74, respectively), thus demonstrat- ing that pregnant animals had decreased ability to extract oxygen in face of decreased cardiac output.

a certain level o f cardiac ou tpu t was reached. There is,

however, one impor tan t di f ference be tween the p regnan t

and n o n p r e g n a n t sheep, namely, the dif ference in the

response of oxygen consumpt ion at cardiac outputs

above the break points. In the n o n p r e g n a n t sheep nei-

ther lactate nor oxygen consumpt ion is affected by re-

duced cardiac ou tpu t at cardiac outputs above the break

point, whereas in the p regnan t sheep oxygen consump-

tion is fall ing even in the absence of a rise in lactate

concentra t ion. In the p regnan t animal this lack of a rise

in lactate concent ra t ion , indicat ing an absence o f tissue

hypoxia and metabol ic acidosis, in spite of a decrease in

both cardiac ou tput and oxygen consumpt ion is the prin-

cipal characterist ic of metabol ic arrest. This is similar to a

p h e n o m e n o n that is seen in diving seals, la Seals dive to

.i~ ,%

t a E

a , 15-

1 0 -

5 -

¢,

t 4.

510 I ! i 100 150 200


b.

E 2E

10 -

5 -

O-

...::,k . .

gO I 0 0 150

Pregnant ]

I 200


Fig. 4. Response of serum lactate concentration to decreasing cardiac output. Both nonpregnant (a) and pregnant (b) sheep display biphasic response of serum lactate concentration; however, break point in lactate curve was noted at higher cardiac output level than in nonpregnant sheep (70 and 43 ml/min/kg, respectively). Break point in lactate concentration curve corresponds to critical cardiac output noted in oxygen consumption curve in nonpregnant sheep noted in Fig. 1 (dashed line). In pregnant sheep no such relationship exists between oxygen consumption curve (dashed line) and serum lactate concentration curve.

great depths, and dur ing this dive p ro longed hypoxia

may develop. In spite o f this p ro longed hypoxia, the

animal does no t have metabol ic (lactic) acidosis.

The p h e n o m e n o n of metabol ic arrest may be a teleo-

logic adaptat ion o f the p regnan t animal to maintain an

aerobic env i ronment to protect the fetus dur ing states of

low cardiac output. Three possible responses may occur

as a result o f a decrease in materna l cardiac ou tpu t and

oxygen delivery. The first response is that p ro found hyp-

oxia develops and the fetus eventually dies. The second

response is that severe hypoxia and metabol ic acidosis


develop in the mother , which eventually leads to major

mult iple organ dysfunction and death. We would expec t

to see these responses in situations where the decreased

cardiac ou tpu t was acute and severe. In the presence of

subacute, gradual reduct ions in materna l cardiac ou tpu t

a third response is feasible, that is, even in the face of

decreasing oxygen supply to the m o t h e r the m o t h e r

maintains an aerobic envi ronment . In o ther words, she

stops using oxygen to spare the fetus of the consequences

of a decreased oxygen supply. At the same t ime materna l

heal th is also spared.

Our data show that the cardiorespiratory profile we

demons t ra ted in p regnan t sheep in relat ion to decreased

cardiac ou tpu t is similar to the cardiorespiratory profile

that is seen in certain critical illnesses such as adult respi-

ratory distress syndrome, sepsis, septic shock, severe pan-

creatitis, extensive burns, and mult iple trauma. Preg-

nancy has physiologic features similar to these illnesses.

O u r data also suggest that the lack of a critical cardiac

ou tpu t is possibly conf ined to late gestation (late third

trimester). We are current ly pe r fo rming exper iments on

midt r imes ter sheep (85 to 90 days) to de t e rmine whe ther

this p h e n o m e n o n is re la ted to gestational age.

We speculate that the physiologic adaptat ions of preg-

nancy lead to this lack of a critical cardiac ou tpu t in the

third trimester. We conclude that pregnancy, unde r con-

dit ions of low cardiac output, is associated with oxygen

consumpt ion and tissue oxygen extract ion abnormali t ies.

We also conc lude that the fetoplacental uni t is a major

stressor raising oxygen demand , which impairs oxygen

consumpt ion in low cardiac ou tpu t states. If this cardiac

ou tpu t -oxygen consumpt ion al terat ion occurs in h u m a n

pregnancy in relat ion to states of low cardiac output , then

this may have significant implicat ions in the m a n a g e m e n t

of critically ill p regnan t patients. Fur ther studies are war-

ran ted to conf i rm this f inding and to explain why preg-

nancy is associated with a f low-dependen t oxygen con-

sumpt ion and tissue oxygen extract ion impa i rmen t in low

cardiac ou tpu t states.

We thank Mr. Anthony Battelli, vivarium supervisor of the Magee-Wornens Research Institute, for his essential and invaluable contr ibut ion to this study.

REFERENCES

1. Sclmmacker PT, Cain SM. The concept of a critical oxygen delivery. Intensive Care Med 1987;13:223-9.

2. Shoemaker WC. Shock states: pathophysiology, monitoring, outcome prediction and therapy. In: Shoemaker WC, Ayres S, Grenvik A, Holbrook PR, Thompson WL, editors. Text- book of critical care. Philadelphia: WB Saunders, 1989:977- 92.

3. Shoemaker WC, Appel PL, Ka-am HB. The role of oxygen transport in the pathophysiology, prediction of outcome and therapy of shock. In: Bryan-Brown C, Ayres SM, editors. New horizons: oxygen transport and utilization. Fullerton, CA: Society of Critical Care Medicine, 1987:65-92.

4. Belfort MA, Anthony J, Saade GR, Wasserstrom N. The oxygen consumption/oxygen delivery curve in severe preeclanapsia: evidence for a fixed oxygen extraction state. Am J Obstet Gynecol 1993;169:1448-55.

5. Fahey JT, Lister G. A simple method for reducing cardiac output in the conscious lamb. AmJ Physiol 1985;249:H188- 92.

6. FaheyJT, Lister G. Postnatal changes in the critical cardiac output and oxygen transport in conscious Iambs. Am J Physiol 1987;352:H100-6.

7. Nelson LD, Rutherford EJ. Principles of hemodynamic monitoring. In: Pinsky MR, Dhainaut J-FA, editors. Patho- physiologic foundations of critical care. Baltimore: Williams & Wilkins, 1993:3-22.

8. Samsel RW, Schumacker PT. Determination of critical oxygen delivery from experimental data: sensitivity to error. J Appl Physiol 1988;64:2074-82.

9. I~ambam JR, Handte RE, Brown WU, Smith BE. Effect of normal and preeclamptic pregnancies on the oxyhemoglo- bin dissociation curve. Anesthesiology 1986;65:426-7.

10. Pivarnik JM, Lee W, Clark SL, Cotton DB, Spillman HT, Miller JE Cardiac output responses of primigravid women during exercise determined by the direct Fick technique. Obstet Gynecol 1990;75:954-9.

11. Pivarnik JM, Lee W, Spillman HT, Clark SL, Cotton DB, Miller JF. Maternal respiration and blood gases during aerobic exercise performed at moderate altitude. Med Sci Sports Exerc 1992;24:868-72.

12. Bader RA, Bader ME, Rose DJ, Braunwald E. Hemodynamics at rest and during exercise in normal pregnancy as studied by cardiac catheterization. J Clin Invest 1955;34:1524-6.

13. Rose DJ, Bader ME, Bader RA, Braunwald E. Catheterization studies of cardiac hemodynamics in normal pregnant women with reference to left ventricular work. Am J Obstet Gynecol 1956;72:233-44.

14. Hochachka PW. Metabolic arrest. Intensive Care Med 1986; 12:127-33.

15. Hochachka PW. Defense strategies against hypoxia and hy- pothermia. Science 1986;231:234-41.

lack of a critical cardiac output and critical systemic oxygen delivery during low cardiac output in...

Documents