- Email: [email protected]
Redistribution of organ blood flow after hemorrhage and resuscitation in full-term piglets
Redistribution of Organ Blood Flow After Hemorrhage and Resuscitation in Full-Term Piglets By D. Lynn Dyes,
Randall W. Powell, Albert N. Swafford, Jr, Dale C. Schmacht,
W. Scott Roberts, John J. Ferrara, and Jeffrey L. Ardell Mobile, Alabama 0 Newborn piglets (aged 1 to 2 days and 7 to 14 days) were used to study (1) the redistribution of organ blood flow after a 25% acute blood loss and (2) the response to resuscitation with shed blood (20 mL/ kg), crystalloid (normal saline [NS] or lactated Ringer’s [LR]; 60 mL/ kg), and colloid (Dextran-40, 20 mL/kg). Hemodynamic parameters showed little differences in the response to hemorrhage and resuscitation. The two age groups had no significant differences in parameters or blood flow (results combined). The animals maintained flow to the heart and central nervous system (CNS) and had significantly decreased flow to the kidneys and splanchnic organs. In the gastrointestinal tract, the small intestine was affected most severely, with a significant decrease in blood flow, especially to the mucosa. In all organ systems, Dextran 40 restored blood flow to levels significantly above the baseline. Shed blood and crystalloid restored flow to organs sustaining decreased flow, but crystalloid did not restore flow to the baseline level in the kidney and all segments of the gastrointestinal tract. Copyright Q 1994 by W.B. Saunders Company INDEX WORDS: Hypovolemia, regional blood flow, intestinal mucosa ischemia, resuscitation.
A
NUMBER of conditions occurring before, at the time of, and after birth can result in hypovolemia in the newborn infant. The response of the neonate to the hypovolemic state may be dependent on multiple factors. The stage of maturity of the autonomic nervous system may govern the distribution of organ blood flow. In addition, the resuscitation fluid (colloid or crystalloid) may impact on the neonate’s response to this stress. This study was undertaken to determine whether specific organ blood flow distributions after hypovolemia were age-dependent, and to determine if the response to resuscitation was dependent on the choice of fluid. MATERIALSAND METHODS
Studies were performed in neonatal piglets. using techniques previously established in this laboratory.’ All experimental procedures were approved by the Institutional Care and Use Committee of the University of South Alabama College of Medicine, and its guidelines were employed. Preparation of Model Fasted full-term piglets were obtained at the ages 1 to 2 days and 7 to 14 days. After intramuscular administration of ketamine (50 mgikg) and chlorpromazine (25 mgikg), tracheotomy and intubation were performed. A pressure-cycled Byrd ventilator (Byrd Corp. Palm Springs. CA) (rate. 10 to 15 breaths per minute: JournalofPediatricSurgery,
Vol29,No
8(August),1994:
pp 1097.1102
positive pressure. 15 to 25 cm HzO) was used to ventilate the piglet with room air through the tracheostomy tube. A cutdown in the groin was performed, and the femoral vein was catheterized for purposes of fluid administration and blood withdrawal. Ipsilateral cannulation of the femoral artery was performed, with the catheter tip positioned in the abdominal aorta to monitor blood pressure and heart rate. obtain blood samples, withdraw blood, and administer fluids. Through a left thoracotomy. the left atria1 appendage was exposed and cannulated for injection of microspheres. Temperature, continuously monitored by rectal probe. was maintained at 36 to 375°C with a heat lamp and/or heating pad. Experimental Protocol After stabilization of a piglet. the first of four randomly selected radiolabeled microspheres was injected to establish baseline measurements of organ blood flows. Hypovolemia by hemorrhage was induced by withdrawal of blood (20 mL/kg body weight) over 10 to 20 minutes from the femoral arterial and venous catheters. Fifteen and sixty minutes (respectively) after completing the blood withdrawal, the second and third microspheres were injected. The animal was then resuscitated with crystalloid (60 mL/kg body weight, in lactated Ringer’s or normal saline), colloid (20 mL/kg body weight; Dextran 40). or reinfusion of the shed blood. The reinfusion occurred over approximately the same period as the blood withdrawal (10 to 20 minutes). The fourth microsphere was administered 30 minutes after completing the resuscitation. Fifteen minutes after injection of the fourth microsphere. the piglet was euthanized with a bolus of saturated potassium chloride administered through the left atria1 cannula. The following organs were immediately removed. weighed, and counted for radioactivity: esophagus, stomach, duodenum. small intestine (subdivided into proximal. middle, and distal segments), colon (cecum. proximal, middle, distal. rectosigmoid), pancreas, liver, spleen, kidney, skeletal muscle, skin. heart, and brain. The mucosa from segments of the gastrointestinal tract (esophagus, stomach, small intestine, and colon) was isolated from the remaining visceral wall to assess mucosal blood flow distinctively from the underlying muscularis, Blood flow to each of the above organs was determined by the reference organ technique,? as detailed below. Control animals were subjected to the same surgical procedures, but were not subjected to hemorrhage or resuscitation. Organ blood flow was measured using lS+m radiolabeled microspheres. All microspheres (cerium 141, strontium 85, scandium 46. chromium 51) were mechanically agitated and then
From the Departments of Surge? and Physiology, Universiv of South Alabama College sf Medicine, Mobile, AL. Presented at the 1993 Annual Meeting of the Section on Surgery of the American Academy of Pediatrics. Washington, DC. October 29-31, 1993. Supported by the American Heart Association. Alabama Afiliate. Address reprint requests to Randall W. Powell, MD, Division of Pediatric Sutgey, 2451 Fillingim St, Mobile, AL 36617. Copyright 8~8 1994 by W. B. Saunders Company 1)022-336819312908-0030$03.00~0 1097
1098
DYESS ET Al
rapidly injected through the left atria1 cannula to allow for adequate mixing. Simultaneously, blood was withdrawn from the aortic cannula at a constant rate, over a 90-second interval, using a Harvard pump. The order in which the microspheres were infused was randomized. After euthanasia and organ harvesting, tissue and blood samples were placed in a LKB gamma counter, and radioactivity in each tissue sample and from each of the four aortic arterial blood withdrawals was determined and was corrected for energy overlap. Blood flow to the various organ samples (Qr) was calculated for each microsphere injection on a per-gram net-weight basis, according to the equation
2ref
QT = RT x
where Rr and Rrcr are radioactive counts in the tissue and reference blood flow samples, and Qret is the withdrawal rate of the aortic arterial blood sample, ie, the reference organ technique.
Statistical Analysis For purposes of initial data analyses, the piglets were divided into two groups based on age (1 to 2 day, 7 to 14 day), and these groups were subdivided based on the resuscitation fluid (crystalloid, colloid, or blood). The mean response (? 1 SE) was determined for each variable (blood pressure, heart rate, cardiac output, peripheral resistance, arterial blood gases, and regional blood flows) before the induction of hypovolemia (time 0, baseline, or control data). Each of these variables was also determined at 15 and 60 minutes of hypovolemia and 30 minutes after resuscitation. Within-group comparisons between baseline, 15 minutes and 60 minutes of hypovolemia, and 30 minutes postresuscitation were analyzed by correlated analysis of variance. Between-group effects were likewise subjected to analysis of variance. Whenever the F ratio was significant, both protected I (least squared difference) and Newman-Keul’s follow-up tests were applied to identify the mean differences having statistical significance. A P value of less than .05 was considered statistically significant.
RESULTS
Fifty-eight piglets successfully underwent the experimental procedures, and data were collected. There were no significant differences between the age groups; therefore, the results of these two groups were combined. Hereafter, “full-term piglets” refers to both age groups. The number of animals and the respective body weights for each treatment group are noted in Table 1. The results of hemodynamic monitoring of the piglets are presented in Table 2. Blood pressure, heart rate, cardiac output, and total peripheral resistance remained stable throughout the duration of the experiment in time-control animals, indicating stability of the model with the operative preparation. The Table 1. Group Characteristics Resuscitation
Fluid
n
Weight
(gl
+- 240 1897 + 169
Control
9
Blood
16
Crystalloid
16
1500 + 92
mxtran
11
2386 2 329
NOTE.
Weight data are expressed as mean f SEM.
1728
BLOOD FLOW AFTER HEMORRHAGE
AND RESUSCITATION
degree of hypovolemia did not consistently alter these parameters. Ventilatory support was manipulated to maintain the arterial blood pH between 7.35 and 7.45. Minimal alterations (not significant) in Pcoz and PO? were observed during the course of the experiments, and these were attributed to manipulation of the ventilatory support. Organ blood flow distributions at time 0, 15 and 60 minutes of hypovolemia, and 30 minutes after resuscitation are shown in Figs 1 and 2. Hypovolemia did not result in significant changes in flow to the heart or central nervous system (CNS). Resuscitation with blood did not significantly impact on flow to the heart or CNS; however, resuscitation with Dextran resulted in significant increases in flow to these organs. Crystalloid resuscitation also increased heart flow. Flow to the kidney, spleen, and pancreas decreased significantly as a result of hypovolemia. Resuscitation with blood and crystalloid increased flow to these organs; however, these fluids were not able to return flows to baseline values. Resuscitation with Dextran increased flow to the kidney, spleen, and pancreas, to rates greater than baseline values. In addition, the flow to these organs resulting from Dextran resuscitation was significantly greater than the final flow resulting from the other resuscitation regimens. The changes in blood flow and vascular resistance to segments of the gastrointestinal tract are shown in
5.0
HEART
KIDNEY
Fig 3. Blood flow to all segments of the gastrointestinal tract decreased significantly with the hypovolemic insult. In addition, the vascular resistances in the viscera consistently increased significantly over baseline values. Resuscitation with blood or crystalloid increased blood flow to the gastrointestinal tract, compared with the flow during hypovolemia, but the flow remained below baseline during the observation period. In contrast, Dextran resuscitation resulted in restoration of blood flow to all segments, with the final flows significantly greater than baseline. Dextran also returned vascular resistance to baseline values. When the resuscitation fluids were compared with each other, Dextran resulted in significantly greater flow than did blood or crystalloid in all segments of the gastrointestinal tract. Evaluation of flow to the layers of the gastrointestinal wall (Fig 4) indicated that hypovolemia diminished flow to both mucosa and muscularis, and both layers responded similarly to Dextran resuscitation, with a rebound increase in flow. These effects were demonstrated in all sections of the gastrointestinal tract. The effects of crystalloid and blood resuscitation on the gastrointestinal wall layers were similar to their effects on the whole wall; neither fluid was able to completely reverse the impact of hypovolemia and restore flows to baseline levels.
CNS
SPLEEN
6.5
r
PANCREAS
r
0
3.0
-
i3
2.5 -
g
2.0
2
1.5-
-
m 1.0 -
0.0
0 156030
0 156030
0 156030
Fig 1. Regional blood flow to the heart, kidney, and central narvous system (CNS) at time 0 (control or baseline), 15 and 60 minutes of hypovolemia, and 30 minutes after resuscitation. Squares denote blood; triangles denote crystalloid; diamonds denote DexWan. V < 65 versus baseline values; +P < .05 versus 15 and 60 minutes of hypovolemia (within group comparisons); #P c -05 versus blood; ‘P c .05 versus crystalloid.
0.5
-
0.0
’
-, 0
15
60
30
0
15
60
30
Fig 2. Regional blood flow to the spleen and pancreas at time 0 (control or baseline), 15 and 60 minutes of hypovolemia. and 30 minutes after resuscitation. Squares denote blood; triangles denote crystalloid; diamonds denote Dextran. lP < .05 versus baseline values; +P c .05 versus 15 and 60 minutes of hypovolemia (within group comparisons): #P c .05 versus blood: +P c .05 versus crystalloid.
DYESS ET AL
ESO-ST0
3.0
3.
DUO-ILE
CEC-REC
T *+
r 2.5
2.5 1
2.0 aE
1.5 ~
1.5
002 ;;ilg
1.0
1.0
-
0.5
#
0.0
l
p;;
it -
tin*+
*
*
*
Oe5 0.0 250
200
1 *
T6h
w
\+
010 0
16
150
60
90
0
16 00
30
MUSCULARIS
MUCOSA
Fig 4. Blood flow to mucosa and muscularis layers of the gastrointestinal tract at baseline (0). 15 and 60 minutes of hypovolemia, and 30 minutes after resuscitation. Squares denote blood; triangles denote crystalloid; diamonds denote Daxtran. *P c .05 versus baseline values; +P < .05 versus 15 and 60 minutes of hypovolamia (within group comparisons); #P -z .05 versus blood; +P < .05 versus crystalloid.
0
16
ESO-ST0
80
90
Fig 3. Segmental blood flow and vascular resistances to the gastrointestinal tract at baseline (0). 15 and 60 minutes of hypovolemia, and 30 minutes after resuscitation. Squares denote blood; triangles denote caystalloid; diamonds denote Daxtran. *P c .05 versus baseline values; +P < .05 versus 15 and 60 minutes of hypovolemia (within group comparisons); *P c .05 versus blood; + P < .05 versus crystalloid.
DUO-ELE
CEC-REC
BLOOD FLOW AFTER HEMORRHAGE
1101
AND RESUSCITATION
DISCUSSION
Studies of hypovolemic shock were limited to mature animals until 1968, when Rowe and ArcilIa3 described the effects of acute hemorrhage in the newborn puppy (2 hours to 9 days of age). They reported a mortality rate of 59% when 35% of the blood volume was removed in this neonatal model. The initial hemodynamic response was a decrease in heart rate, but this returned to normal in the animals that survived the insult. Mean arterial pressure (MAP) and cardiac output also decreased, with very little change in central venous pressure (CVP). Ramenofsky et al4 studied the response of newborn pigs to acute hemorrhage, using radiolabeled microsphere injections to determine perfusion to different organ systems. Their studies indicated significant decreases in flow to the gastrointestinal tract, kidneys, skin, and muscle. In contrast, perfusion was sustained or increased to the heart and brain. LeGal also evaluated the response of newborn piglets to a hemorrhagic insult. He observed decreases in heart rate, MAP, and CVP. Laptook et al6 reported regional differences in brain flow in the piglet in response to hypotension resulting from acute blood loss. Buckley et al? studied the cardiovascular response to hemorrhage in newborn piglets of varying ages. Tachycardia occurred in most animals, and the MAP response to arterial hemorrhage was agedependent. Renal, femoral, and carotid flow decreased in all animals, but increases in femoral artery resistance to both arterial and venous hemorrhage were seen only in 2-week-old animals. In the present study, no significant differences were noted in regional blood flow (of the organs examined) between the two age groups. Little information is available concerning the response of regional blood flow restoration after hemorrhage in a neonatal model. In adult rats, Scannell, et al7 showed that crystalloid solutions (lactated Ringer’s) did not restore the intestinal-tract blood flow to baseline levels at two and three times the blood loss. Wang and Chaudry,# again in adult rats, showed a similar failure of crystalloid resuscitation to restore
cardiac output at two or three times the blood loss; four times the volume of blood loss restored cardiac output but only transiently in a noncontinuous-bloodloss model. In the present study, the full-term newborn piglets had little disturbance of hemodynamic parameters in response to a 25% blood volume hemorrhage, indicating a mature neural-hormonal response to acute hypovolemia. Blood flow redistribution sacrificed flow to the kidneys and splanchnic organs to maintain flow to the heart and brain. Resuscitation with shed blood usually returned decreased flows to baseline or near baseline levels; however, crystalloid replacement at three times the blood volume loss did not restore blood flow to baseline in the kidney and gastrointestinal tract, indicating a persistent ischemic insult to these organs. In the gastrointestinal tract, significant decreases in blood flow occurred at all levels, but the small intestine was affected most severely, with signihcant decreases in mucosal blood flow, which were not restored to baseline by shed blood or crystalloid solution. Mucosal ischemia has been implicated as a causative factor in the development of necrotizing enterocolitis9%ln and in bacterial translocation” and persistent mucosal ischemia, and it may potentiate the development of these problems. However, the increased perfusion seen with Dextran 40 resuscitation may magnify shock-induced ischemia-reperfusion injury by increasing the blood flow and oxygen delivery to primed ischemic tissue.” The marked increases in blood flow resulting from Dextran 40 could be considered beneficial in situations in which acute hemorrhage occurs and then is controlled. However, in cases with continuous hemorrhage, the increased flows and decreased vascular resistance induced by Dextran 40 could result in increased blood loss as has been seen in continuing hemorrhage models with hypertonic saline.‘” Dextran is not widely used in the human neonate, and other colloid solutions or mixtures of colloid and crystalloid solutions need to be evaluated to allow appropriate resuscitation of the microcirculation after acute hemorrhage in the neonate.
REFERENCES 1. Dyess DL. Peeples CL, Ardell JL, induced blood flow distribution in premature J Pediatr Surg 28:1396-1400, 1993
et al: Indomethacinand full-term piglets.
2. Buckley BJ. Gootman N. Nagelberg JS, et al: Cardiovascular responses to arterial and venous hemorrhage in neonatal swine. Am J Physiol 247:R626-R666, 1984 3. Rowe MI. Arcilla newborn to hemorrhage. 4. Ramenofsky
RA: Hemodynamic adaptation J Pediatr Surg 3:278-285,1968
ML, Connolly
RJ, Keough
of the
EM, et al: Differen-
tial organ perfusion in the hypovolemic neonate: animal study. J Pediatr Surg 16:955-959, 1981
A neonatal
5. LeGal YM: Effects of acute hemorrhage on some physiological parameters of the cardiovascular system in newborn pigs. Biol Neonate 44:210-281, 1983 6. Laptook A, Stonestreet BS, Oh W: Autoregulation of brain blood flow in the newborn piglet: Regional differences in flow reduction during hypotension. Early Hum Dev 6:99-107. 1982 7. Scannell
G, Clark
L, Waxman
K: Regional
flow
during
1102
DYESS ET AL
experimental hemorrhage and crystalloid resuscitation: Persistence of low flow to the splanchnic organs. Resuscitation 23:217225,1992 8. Wang P, Chaudry IH: Crystalloid resuscitation restores but does not maintain cardiac output following severe hemorrhage. .I Surg Res 50:163-169,1991 9. Rowe Ml: Necrotizing enterocolitis, in Welch KJ, Randolph JG, Ravitch MM, et al (eds): Pediatric Surgery. Chicago, IL, Year Book Medical 1986, pp 944-958 10. Touloukian RJ, Posch JN, Spencer RP: The pathogenesis of ischemic gastroenterocolitis of the neonate: Selective gut mucosal
ischemia in asphyxiated neonatal piglets. J Pediatr Surg 7:194-205, 1972 11. Baker JW, Deitch EA, Li M, et al: Hemorrhagic shock induces bacterial translocation from the gut. J Trauma 28:896-906, 1988 12. Behrman SW, Fabian TC, Kudsk KA, et al: Microcirculatory flow changes after initial resuscitation of hemorrhagic shock with 7.5% hypertonic saline/6% Dextran 70. J Trauma 31:589-600,199l 13. Gross D, Landau EH, Lkin B, et al: Quantitative measurement of bleeding following hypertonic saline therapy in “uncontrolled” hemorrhagic shock. J Trauma 29:79-83, 1989
Discussion D.R. King (Columbus, OH): I have a couple of methodological questions. The rules are a little different for this kind of study than for some, and I wonder if you did any determinations of blood volume to demonstrate that your animals at the end of an hour were in fact hypovolemic? I also wonder if you had any differences that you could demonstrate between the piglets that were 1 to 2 days old and those that were considerably older, by most of our estimations at a week to 2 weeks. D.L. Dyess (response): We did not determine blood volume after the hypovolemia. We did initially separate the two groups of animals with respect to age, and with each Auid regimen, and there were no statistically significant differences in the responses that we measured, Therefore, we grouped our data. M. W. Harrison (Portland, OR): The brain blood flow increased substantially in the postresuscitation phase on one of your graphs. I’m curious, do you think this might represent a detriment to our trauma resuscitation patients? Many of them have central nervous system injuries. D.L. Dyess (response): I think that could be a consideration. In our studies and in some previous works, the perfusion to the heart and brain is maintained as opposed to the other organs, and this may truly be detrimental. D. Cohn (Dallas, TX): I really admire anyone who undertakes any hemorrhagic shock model in neonatal animals of any sort because it’s very labor-intensive. You didn’t state the number of animals you had in your experiment, or at least I didn’t hear it. There are a couple of problems I had with your
study. I don’t think the degree of hypovolemia that you induced in these animals is sufficient to cause serious physiological or metabolic problems, and I think any intravascular replacement will give you a good response. The other thing is, in your crystalloid group, you have mixed Ringer’s lactage and normal saline. The volume of normal saline that you gave these animals is probably not sufficient to produce a dilutional acidosis, but the added chloride load may contribute to some acidosis and the decreased blood flow to the organs. Secondly, the volume status of the two animal groups appears to be different prestudy, because all the organ flows are higher in the Dextran group than they are in the crystalloid group. The last question is, did you monitor the temperature of these animals? If they’re the least bit hypothermic, their organ flow is reduced significantly. D. L. Dyess (response): The total number of animals in this particular set of experiments was 58, with anywhere from nine to 16 in each group. With respect to the crystalloid resuscitation, we used either lactated Ringer’s or normal saline. We performed statistical analyses between these different infusions to see if we could determine any difference in the resuscitation, and we could not. With respect to the volume of fluid that we administered, we did replace it 3 to 1 with crystalloid. We monitored temperature throughout the experiment. With just a heating pad and a warmed environment, we’ve not had a problem with temperature.
Randall W. Powell, Albert N. Swafford, Jr, Dale C. Schmacht,
W. Scott Roberts, John J. Ferrara, and Jeffrey L. Ardell Mobile, Alabama 0 Newborn piglets (aged 1 to 2 days and 7 to 14 days) were used to study (1) the redistribution of organ blood flow after a 25% acute blood loss and (2) the response to resuscitation with shed blood (20 mL/ kg), crystalloid (normal saline [NS] or lactated Ringer’s [LR]; 60 mL/ kg), and colloid (Dextran-40, 20 mL/kg). Hemodynamic parameters showed little differences in the response to hemorrhage and resuscitation. The two age groups had no significant differences in parameters or blood flow (results combined). The animals maintained flow to the heart and central nervous system (CNS) and had significantly decreased flow to the kidneys and splanchnic organs. In the gastrointestinal tract, the small intestine was affected most severely, with a significant decrease in blood flow, especially to the mucosa. In all organ systems, Dextran 40 restored blood flow to levels significantly above the baseline. Shed blood and crystalloid restored flow to organs sustaining decreased flow, but crystalloid did not restore flow to the baseline level in the kidney and all segments of the gastrointestinal tract. Copyright Q 1994 by W.B. Saunders Company INDEX WORDS: Hypovolemia, regional blood flow, intestinal mucosa ischemia, resuscitation.
A
NUMBER of conditions occurring before, at the time of, and after birth can result in hypovolemia in the newborn infant. The response of the neonate to the hypovolemic state may be dependent on multiple factors. The stage of maturity of the autonomic nervous system may govern the distribution of organ blood flow. In addition, the resuscitation fluid (colloid or crystalloid) may impact on the neonate’s response to this stress. This study was undertaken to determine whether specific organ blood flow distributions after hypovolemia were age-dependent, and to determine if the response to resuscitation was dependent on the choice of fluid. MATERIALSAND METHODS
Studies were performed in neonatal piglets. using techniques previously established in this laboratory.’ All experimental procedures were approved by the Institutional Care and Use Committee of the University of South Alabama College of Medicine, and its guidelines were employed. Preparation of Model Fasted full-term piglets were obtained at the ages 1 to 2 days and 7 to 14 days. After intramuscular administration of ketamine (50 mgikg) and chlorpromazine (25 mgikg), tracheotomy and intubation were performed. A pressure-cycled Byrd ventilator (Byrd Corp. Palm Springs. CA) (rate. 10 to 15 breaths per minute: JournalofPediatricSurgery,
Vol29,No
8(August),1994:
pp 1097.1102
positive pressure. 15 to 25 cm HzO) was used to ventilate the piglet with room air through the tracheostomy tube. A cutdown in the groin was performed, and the femoral vein was catheterized for purposes of fluid administration and blood withdrawal. Ipsilateral cannulation of the femoral artery was performed, with the catheter tip positioned in the abdominal aorta to monitor blood pressure and heart rate. obtain blood samples, withdraw blood, and administer fluids. Through a left thoracotomy. the left atria1 appendage was exposed and cannulated for injection of microspheres. Temperature, continuously monitored by rectal probe. was maintained at 36 to 375°C with a heat lamp and/or heating pad. Experimental Protocol After stabilization of a piglet. the first of four randomly selected radiolabeled microspheres was injected to establish baseline measurements of organ blood flows. Hypovolemia by hemorrhage was induced by withdrawal of blood (20 mL/kg body weight) over 10 to 20 minutes from the femoral arterial and venous catheters. Fifteen and sixty minutes (respectively) after completing the blood withdrawal, the second and third microspheres were injected. The animal was then resuscitated with crystalloid (60 mL/kg body weight, in lactated Ringer’s or normal saline), colloid (20 mL/kg body weight; Dextran 40). or reinfusion of the shed blood. The reinfusion occurred over approximately the same period as the blood withdrawal (10 to 20 minutes). The fourth microsphere was administered 30 minutes after completing the resuscitation. Fifteen minutes after injection of the fourth microsphere. the piglet was euthanized with a bolus of saturated potassium chloride administered through the left atria1 cannula. The following organs were immediately removed. weighed, and counted for radioactivity: esophagus, stomach, duodenum. small intestine (subdivided into proximal. middle, and distal segments), colon (cecum. proximal, middle, distal. rectosigmoid), pancreas, liver, spleen, kidney, skeletal muscle, skin. heart, and brain. The mucosa from segments of the gastrointestinal tract (esophagus, stomach, small intestine, and colon) was isolated from the remaining visceral wall to assess mucosal blood flow distinctively from the underlying muscularis, Blood flow to each of the above organs was determined by the reference organ technique,? as detailed below. Control animals were subjected to the same surgical procedures, but were not subjected to hemorrhage or resuscitation. Organ blood flow was measured using lS+m radiolabeled microspheres. All microspheres (cerium 141, strontium 85, scandium 46. chromium 51) were mechanically agitated and then
From the Departments of Surge? and Physiology, Universiv of South Alabama College sf Medicine, Mobile, AL. Presented at the 1993 Annual Meeting of the Section on Surgery of the American Academy of Pediatrics. Washington, DC. October 29-31, 1993. Supported by the American Heart Association. Alabama Afiliate. Address reprint requests to Randall W. Powell, MD, Division of Pediatric Sutgey, 2451 Fillingim St, Mobile, AL 36617. Copyright 8~8 1994 by W. B. Saunders Company 1)022-336819312908-0030$03.00~0 1097
1098
DYESS ET Al
rapidly injected through the left atria1 cannula to allow for adequate mixing. Simultaneously, blood was withdrawn from the aortic cannula at a constant rate, over a 90-second interval, using a Harvard pump. The order in which the microspheres were infused was randomized. After euthanasia and organ harvesting, tissue and blood samples were placed in a LKB gamma counter, and radioactivity in each tissue sample and from each of the four aortic arterial blood withdrawals was determined and was corrected for energy overlap. Blood flow to the various organ samples (Qr) was calculated for each microsphere injection on a per-gram net-weight basis, according to the equation
2ref
QT = RT x
where Rr and Rrcr are radioactive counts in the tissue and reference blood flow samples, and Qret is the withdrawal rate of the aortic arterial blood sample, ie, the reference organ technique.
Statistical Analysis For purposes of initial data analyses, the piglets were divided into two groups based on age (1 to 2 day, 7 to 14 day), and these groups were subdivided based on the resuscitation fluid (crystalloid, colloid, or blood). The mean response (? 1 SE) was determined for each variable (blood pressure, heart rate, cardiac output, peripheral resistance, arterial blood gases, and regional blood flows) before the induction of hypovolemia (time 0, baseline, or control data). Each of these variables was also determined at 15 and 60 minutes of hypovolemia and 30 minutes after resuscitation. Within-group comparisons between baseline, 15 minutes and 60 minutes of hypovolemia, and 30 minutes postresuscitation were analyzed by correlated analysis of variance. Between-group effects were likewise subjected to analysis of variance. Whenever the F ratio was significant, both protected I (least squared difference) and Newman-Keul’s follow-up tests were applied to identify the mean differences having statistical significance. A P value of less than .05 was considered statistically significant.
RESULTS
Fifty-eight piglets successfully underwent the experimental procedures, and data were collected. There were no significant differences between the age groups; therefore, the results of these two groups were combined. Hereafter, “full-term piglets” refers to both age groups. The number of animals and the respective body weights for each treatment group are noted in Table 1. The results of hemodynamic monitoring of the piglets are presented in Table 2. Blood pressure, heart rate, cardiac output, and total peripheral resistance remained stable throughout the duration of the experiment in time-control animals, indicating stability of the model with the operative preparation. The Table 1. Group Characteristics Resuscitation
Fluid
n
Weight
(gl
+- 240 1897 + 169
Control
9
Blood
16
Crystalloid
16
1500 + 92
mxtran
11
2386 2 329
NOTE.
Weight data are expressed as mean f SEM.
1728
BLOOD FLOW AFTER HEMORRHAGE
AND RESUSCITATION
degree of hypovolemia did not consistently alter these parameters. Ventilatory support was manipulated to maintain the arterial blood pH between 7.35 and 7.45. Minimal alterations (not significant) in Pcoz and PO? were observed during the course of the experiments, and these were attributed to manipulation of the ventilatory support. Organ blood flow distributions at time 0, 15 and 60 minutes of hypovolemia, and 30 minutes after resuscitation are shown in Figs 1 and 2. Hypovolemia did not result in significant changes in flow to the heart or central nervous system (CNS). Resuscitation with blood did not significantly impact on flow to the heart or CNS; however, resuscitation with Dextran resulted in significant increases in flow to these organs. Crystalloid resuscitation also increased heart flow. Flow to the kidney, spleen, and pancreas decreased significantly as a result of hypovolemia. Resuscitation with blood and crystalloid increased flow to these organs; however, these fluids were not able to return flows to baseline values. Resuscitation with Dextran increased flow to the kidney, spleen, and pancreas, to rates greater than baseline values. In addition, the flow to these organs resulting from Dextran resuscitation was significantly greater than the final flow resulting from the other resuscitation regimens. The changes in blood flow and vascular resistance to segments of the gastrointestinal tract are shown in
5.0
HEART
KIDNEY
Fig 3. Blood flow to all segments of the gastrointestinal tract decreased significantly with the hypovolemic insult. In addition, the vascular resistances in the viscera consistently increased significantly over baseline values. Resuscitation with blood or crystalloid increased blood flow to the gastrointestinal tract, compared with the flow during hypovolemia, but the flow remained below baseline during the observation period. In contrast, Dextran resuscitation resulted in restoration of blood flow to all segments, with the final flows significantly greater than baseline. Dextran also returned vascular resistance to baseline values. When the resuscitation fluids were compared with each other, Dextran resulted in significantly greater flow than did blood or crystalloid in all segments of the gastrointestinal tract. Evaluation of flow to the layers of the gastrointestinal wall (Fig 4) indicated that hypovolemia diminished flow to both mucosa and muscularis, and both layers responded similarly to Dextran resuscitation, with a rebound increase in flow. These effects were demonstrated in all sections of the gastrointestinal tract. The effects of crystalloid and blood resuscitation on the gastrointestinal wall layers were similar to their effects on the whole wall; neither fluid was able to completely reverse the impact of hypovolemia and restore flows to baseline levels.
CNS
SPLEEN
6.5
r
PANCREAS
r
0
3.0
-
i3
2.5 -
g
2.0
2
1.5-
-
m 1.0 -
0.0
0 156030
0 156030
0 156030
Fig 1. Regional blood flow to the heart, kidney, and central narvous system (CNS) at time 0 (control or baseline), 15 and 60 minutes of hypovolemia, and 30 minutes after resuscitation. Squares denote blood; triangles denote crystalloid; diamonds denote DexWan. V < 65 versus baseline values; +P < .05 versus 15 and 60 minutes of hypovolemia (within group comparisons); #P c -05 versus blood; ‘P c .05 versus crystalloid.
0.5
-
0.0
’
-, 0
15
60
30
0
15
60
30
Fig 2. Regional blood flow to the spleen and pancreas at time 0 (control or baseline), 15 and 60 minutes of hypovolemia. and 30 minutes after resuscitation. Squares denote blood; triangles denote crystalloid; diamonds denote Dextran. lP < .05 versus baseline values; +P c .05 versus 15 and 60 minutes of hypovolemia (within group comparisons): #P c .05 versus blood: +P c .05 versus crystalloid.
DYESS ET AL
ESO-ST0
3.0
3.
DUO-ILE
CEC-REC
T *+
r 2.5
2.5 1
2.0 aE
1.5 ~
1.5
002 ;;ilg
1.0
1.0
-
0.5
#
0.0
l
p;;
it -
tin*+
*
*
*
Oe5 0.0 250
200
1 *
T6h
w
\+
010 0
16
150
60
90
0
16 00
30
MUSCULARIS
MUCOSA
Fig 4. Blood flow to mucosa and muscularis layers of the gastrointestinal tract at baseline (0). 15 and 60 minutes of hypovolemia, and 30 minutes after resuscitation. Squares denote blood; triangles denote crystalloid; diamonds denote Daxtran. *P c .05 versus baseline values; +P < .05 versus 15 and 60 minutes of hypovolamia (within group comparisons); #P -z .05 versus blood; +P < .05 versus crystalloid.
0
16
ESO-ST0
80
90
Fig 3. Segmental blood flow and vascular resistances to the gastrointestinal tract at baseline (0). 15 and 60 minutes of hypovolemia, and 30 minutes after resuscitation. Squares denote blood; triangles denote caystalloid; diamonds denote Daxtran. *P c .05 versus baseline values; +P < .05 versus 15 and 60 minutes of hypovolemia (within group comparisons); *P c .05 versus blood; + P < .05 versus crystalloid.
DUO-ELE
CEC-REC
BLOOD FLOW AFTER HEMORRHAGE
1101
AND RESUSCITATION
DISCUSSION
Studies of hypovolemic shock were limited to mature animals until 1968, when Rowe and ArcilIa3 described the effects of acute hemorrhage in the newborn puppy (2 hours to 9 days of age). They reported a mortality rate of 59% when 35% of the blood volume was removed in this neonatal model. The initial hemodynamic response was a decrease in heart rate, but this returned to normal in the animals that survived the insult. Mean arterial pressure (MAP) and cardiac output also decreased, with very little change in central venous pressure (CVP). Ramenofsky et al4 studied the response of newborn pigs to acute hemorrhage, using radiolabeled microsphere injections to determine perfusion to different organ systems. Their studies indicated significant decreases in flow to the gastrointestinal tract, kidneys, skin, and muscle. In contrast, perfusion was sustained or increased to the heart and brain. LeGal also evaluated the response of newborn piglets to a hemorrhagic insult. He observed decreases in heart rate, MAP, and CVP. Laptook et al6 reported regional differences in brain flow in the piglet in response to hypotension resulting from acute blood loss. Buckley et al? studied the cardiovascular response to hemorrhage in newborn piglets of varying ages. Tachycardia occurred in most animals, and the MAP response to arterial hemorrhage was agedependent. Renal, femoral, and carotid flow decreased in all animals, but increases in femoral artery resistance to both arterial and venous hemorrhage were seen only in 2-week-old animals. In the present study, no significant differences were noted in regional blood flow (of the organs examined) between the two age groups. Little information is available concerning the response of regional blood flow restoration after hemorrhage in a neonatal model. In adult rats, Scannell, et al7 showed that crystalloid solutions (lactated Ringer’s) did not restore the intestinal-tract blood flow to baseline levels at two and three times the blood loss. Wang and Chaudry,# again in adult rats, showed a similar failure of crystalloid resuscitation to restore
cardiac output at two or three times the blood loss; four times the volume of blood loss restored cardiac output but only transiently in a noncontinuous-bloodloss model. In the present study, the full-term newborn piglets had little disturbance of hemodynamic parameters in response to a 25% blood volume hemorrhage, indicating a mature neural-hormonal response to acute hypovolemia. Blood flow redistribution sacrificed flow to the kidneys and splanchnic organs to maintain flow to the heart and brain. Resuscitation with shed blood usually returned decreased flows to baseline or near baseline levels; however, crystalloid replacement at three times the blood volume loss did not restore blood flow to baseline in the kidney and gastrointestinal tract, indicating a persistent ischemic insult to these organs. In the gastrointestinal tract, significant decreases in blood flow occurred at all levels, but the small intestine was affected most severely, with signihcant decreases in mucosal blood flow, which were not restored to baseline by shed blood or crystalloid solution. Mucosal ischemia has been implicated as a causative factor in the development of necrotizing enterocolitis9%ln and in bacterial translocation” and persistent mucosal ischemia, and it may potentiate the development of these problems. However, the increased perfusion seen with Dextran 40 resuscitation may magnify shock-induced ischemia-reperfusion injury by increasing the blood flow and oxygen delivery to primed ischemic tissue.” The marked increases in blood flow resulting from Dextran 40 could be considered beneficial in situations in which acute hemorrhage occurs and then is controlled. However, in cases with continuous hemorrhage, the increased flows and decreased vascular resistance induced by Dextran 40 could result in increased blood loss as has been seen in continuing hemorrhage models with hypertonic saline.‘” Dextran is not widely used in the human neonate, and other colloid solutions or mixtures of colloid and crystalloid solutions need to be evaluated to allow appropriate resuscitation of the microcirculation after acute hemorrhage in the neonate.
REFERENCES 1. Dyess DL. Peeples CL, Ardell JL, induced blood flow distribution in premature J Pediatr Surg 28:1396-1400, 1993
et al: Indomethacinand full-term piglets.
2. Buckley BJ. Gootman N. Nagelberg JS, et al: Cardiovascular responses to arterial and venous hemorrhage in neonatal swine. Am J Physiol 247:R626-R666, 1984 3. Rowe MI. Arcilla newborn to hemorrhage. 4. Ramenofsky
RA: Hemodynamic adaptation J Pediatr Surg 3:278-285,1968
ML, Connolly
RJ, Keough
of the
EM, et al: Differen-
tial organ perfusion in the hypovolemic neonate: animal study. J Pediatr Surg 16:955-959, 1981
A neonatal
5. LeGal YM: Effects of acute hemorrhage on some physiological parameters of the cardiovascular system in newborn pigs. Biol Neonate 44:210-281, 1983 6. Laptook A, Stonestreet BS, Oh W: Autoregulation of brain blood flow in the newborn piglet: Regional differences in flow reduction during hypotension. Early Hum Dev 6:99-107. 1982 7. Scannell
G, Clark
L, Waxman
K: Regional
flow
during
1102
DYESS ET AL
experimental hemorrhage and crystalloid resuscitation: Persistence of low flow to the splanchnic organs. Resuscitation 23:217225,1992 8. Wang P, Chaudry IH: Crystalloid resuscitation restores but does not maintain cardiac output following severe hemorrhage. .I Surg Res 50:163-169,1991 9. Rowe Ml: Necrotizing enterocolitis, in Welch KJ, Randolph JG, Ravitch MM, et al (eds): Pediatric Surgery. Chicago, IL, Year Book Medical 1986, pp 944-958 10. Touloukian RJ, Posch JN, Spencer RP: The pathogenesis of ischemic gastroenterocolitis of the neonate: Selective gut mucosal
ischemia in asphyxiated neonatal piglets. J Pediatr Surg 7:194-205, 1972 11. Baker JW, Deitch EA, Li M, et al: Hemorrhagic shock induces bacterial translocation from the gut. J Trauma 28:896-906, 1988 12. Behrman SW, Fabian TC, Kudsk KA, et al: Microcirculatory flow changes after initial resuscitation of hemorrhagic shock with 7.5% hypertonic saline/6% Dextran 70. J Trauma 31:589-600,199l 13. Gross D, Landau EH, Lkin B, et al: Quantitative measurement of bleeding following hypertonic saline therapy in “uncontrolled” hemorrhagic shock. J Trauma 29:79-83, 1989
Discussion D.R. King (Columbus, OH): I have a couple of methodological questions. The rules are a little different for this kind of study than for some, and I wonder if you did any determinations of blood volume to demonstrate that your animals at the end of an hour were in fact hypovolemic? I also wonder if you had any differences that you could demonstrate between the piglets that were 1 to 2 days old and those that were considerably older, by most of our estimations at a week to 2 weeks. D.L. Dyess (response): We did not determine blood volume after the hypovolemia. We did initially separate the two groups of animals with respect to age, and with each Auid regimen, and there were no statistically significant differences in the responses that we measured, Therefore, we grouped our data. M. W. Harrison (Portland, OR): The brain blood flow increased substantially in the postresuscitation phase on one of your graphs. I’m curious, do you think this might represent a detriment to our trauma resuscitation patients? Many of them have central nervous system injuries. D.L. Dyess (response): I think that could be a consideration. In our studies and in some previous works, the perfusion to the heart and brain is maintained as opposed to the other organs, and this may truly be detrimental. D. Cohn (Dallas, TX): I really admire anyone who undertakes any hemorrhagic shock model in neonatal animals of any sort because it’s very labor-intensive. You didn’t state the number of animals you had in your experiment, or at least I didn’t hear it. There are a couple of problems I had with your
study. I don’t think the degree of hypovolemia that you induced in these animals is sufficient to cause serious physiological or metabolic problems, and I think any intravascular replacement will give you a good response. The other thing is, in your crystalloid group, you have mixed Ringer’s lactage and normal saline. The volume of normal saline that you gave these animals is probably not sufficient to produce a dilutional acidosis, but the added chloride load may contribute to some acidosis and the decreased blood flow to the organs. Secondly, the volume status of the two animal groups appears to be different prestudy, because all the organ flows are higher in the Dextran group than they are in the crystalloid group. The last question is, did you monitor the temperature of these animals? If they’re the least bit hypothermic, their organ flow is reduced significantly. D. L. Dyess (response): The total number of animals in this particular set of experiments was 58, with anywhere from nine to 16 in each group. With respect to the crystalloid resuscitation, we used either lactated Ringer’s or normal saline. We performed statistical analyses between these different infusions to see if we could determine any difference in the resuscitation, and we could not. With respect to the volume of fluid that we administered, we did replace it 3 to 1 with crystalloid. We monitored temperature throughout the experiment. With just a heating pad and a warmed environment, we’ve not had a problem with temperature.