- Email: [email protected]
Endotoxin and transient hypoxia cause severe acidosis in the piglet
Endotoxin
and Transient
Hypoxia
Cause Severe Acidosis
By Jonathan
Columbus, l Purpose: Both hypoxia and gram-negative sepsis are thought to play a role in the development of necrotizing enterocolitis (NEC). Endotoxin, a lipopolysaccharide (LPS), is a potent mediator of gram-negative sepsis. The author investigated the effect of LPS and hypoxia on arterial and mesenteric venous blood gas values in a piglet model. Methods: 16 piglets (mean age, 9 days; mean weight, 3.2 kg) were anesthetized and mechanically ventilated. Catheters were placed in the aorta and the superior mesenteric vein (SMV). After a 30-minute stabilization period, piglets were randomly assigned to four experimental groups: normoxic ventilation (FIo~, 0.211, normoxic ventilation and LPS infusion (200 pg/kg, intravenously), hypoxic ventilation (FIo~, 0.10 for 20 minutes), or hypoxic ventilation and LPS infusion. All subjects were then monitored for an additional 30 minutes (recovery period). Multiple, paired blood gas samples were obtained from the aorta and SMV during the stabilization, experimental, and recovery periods. Results: Piglets subjected to both hypoxia and LPS experienced a much more severe acidosis in both the aorta (pH, 7.10 f 0.08) and SMV (pH, 7.03 f 0.09) than piglets subjected to either hypoxia or LPS alone (P < ,051. In addition, LPS lowered the arterial oxygen saturation in piglets exposed to acute, transient hypoxia (36 r 4% v 59 -c 12%. P < .05). Conclusion: This study suggests that the combination of transient hypoxia and gram-negative sepsis may act synergistically to produce both a severe acidosis and decreased tissue oxygenation. Copyright o 1997 by W.B. Saunders Company
INDEX WORDS: sis, necrotizing
Endotoxin, enterocolitis.
lipopolysaccharide,
hypoxia,
sep-
T
HE ETIOLOGY of necrotizing enterocolitis (NEC) is considered to be multifactorial, and both hypoxia and sepsis are thought to play a role.’ Transient hypoxic events are common in criticahy ill neonates as a result of respiratory distress syndrome, atelectasis. pneumothorax. or endotracheal tube plugging or dislodgment. Sepsis frequently occurs in this population because premature neonates lack a fully competent immune system2 and the newborn’s gut is rapidly colonized with microflora, including gram-negative bacteria.3 Many of the deleterious effects in gram-negative bacterial infections are mediated by endotoxin, a lipopolysaccharide (LPS). High doses of endotoxin and severe hypoxia independently cause damage to the intestine in animal models.” The purpose of this study is to develop a model for investigating the possible synergistic effect of LPS and hypoxia by measuring standard blood gas parameters in both the arterial and mesenteric circulation in the piglet. MATERIALS
AND
METHODS
This experimental protocol was approved by the Institutional Ammal Care and Use Committee of Children’s Hospital. Columbus. Sixteen piglets (mean age, 9 days: mean weight, 3 2 kg) were anesthetized with JournalofPediatric
Surgery,
Vol 32, No 7 (July),
1997, pp 1123-l
126
in the Piglet
I. Groner
Ohio a combmation of Tiletamine HCl, 3 mg/kg intramuscularly (IM). and Zolazepam HCl. 3 mg/kg IM, (Telazol, Fort Dodge Labs. Fort Dodge, IA); and Xylazine, 4 mg/kg IM (Rompum, Miles. Inc. Shawnee Mission, KS). Anesthesia was maintained using intermittent intravenous inJections of Pentobarbital Sodium (Nembutol, Abbott Laboratories. N Chicago. IL). A tracheotomy was performed. and the animals were ventilated with a mechamcal ventilator (Harvard Apparatus, South Nstick, MA) set to deliver 30 breaths per minute at a tidal volume of 65 mL per breath of room air. Fme bore polyethylene catheters (PE-20, Clay-Adams, Parsippany, NJ) were placed in the left internal jugular vein and the aorta via the femoral artery. A midventral laparotomy was then created, and the confluence of the mesenteric veins was identified at the base of the small bowel mesentery. The superior mesenteric vein (SMV) was cannulated by direct puncture m a retrograde fashion with a polyethylene catheter. It was secured with a cyanoacrylate ester glue (Duro Quick Gel, Loctite Corp, Rocky Hill, CT). and it did not occlude SMV blood flow. Body temperature was mamtained throughout the procedure using a warming pad and heat lamps. When the surgical preparation was completed, all animals underwent a 30.minute stabilization period and were then randomly divided into one of four experimental groups. Group 1 piglets (n = 4) were ventilated with room air throughout the expenment and served as controls. Group 2 piglets (n = 4) were ventilated with room air and received an mtravenous injection of 200 pg/kg LPS (from Escherichia co/r K-235, Sigma Chemical Co, St Louis, MO) at the start of the experimental period. Group 3 piglets (n = 4) were ventilated with a gas mixture containing 10% oxygen and 90% nitrogen for 20 minutes. Group 4 piglets received LPS (200 pg/kg) and were ventilated with the 10% oxygen and 90% nitrogen mixture for 20 minutes. After the 20.mmute experimental period, groups 3 and 4 were returned to room air ventilation. and all animals were monitored for an additional 30 minutes (recovery period). Paired sets of blood samples (0.4 mL per sample) were drawn from the aortic and SMV catheters during the stabilization. experimental, and recovery phases usmg heparimzed I-mL syringes. Samples were immediately analyzed for pH. PO?, and Pco~. and oxygen saturation using an automated blood gas analyzer (ABL 300. Radiometer Medical A/S. Copenhagen Denmark). For each treatment group, data from the three phases of the experiment (stabilization, experimental, recovery) were averaged. Treatment groups were then compared using one-way analysis of variance for overall significance and the Student-NewmanKeuls test for intergroup compartsons. A P value of less than .05 was considered sigmficant. RESULTS
There were no differences in pH, Pco2, PO,, and oxygen saturation levels among the four groups during
From the Drpavtmem of Surgeq ChildretI S Hospital, Columbus, OH. Preserlted at the 1996 Anmtal Meeting of the Sectron on Swgev) oj the American Academy of Pediatrics, Boston, Massachusetts, October 26-30, 1996. Address reprrnt requests to Jonathan I. Gronel: MD, 555 S 18th St, Sate 6C, Columbus, OH 43205. Copynght o 1997 by U!B. Saunders Company 0022-3468/97/3207-0040$03.00/O
1123
1124
JONATHAN
the stabilization period. During the experimental period, the pH values in both aorta (7.20 t 0.10) and SMV (7.10 2 0.17) samples from group 4 were more acidotic than group 1 (arterial, 7.35 2 0.02; SMV, 7.34 t 0.03) and group 3 (arterial, 7.40 + 0.02; SMV, 7.34 i 0.05). Group 4 animals also demonstrated a marked arterial acidosis during the recovery phase of the experiment (pH, 7.11 + 0.08) compared with group 1 (7.3 1 ? 0.02), group 2 (7.23 5 0.01) or group 3 (7.24 ? 0.01, Fig 1). The same result was observed in the mesenteric venous circulation. The recovery phase pH level in group 4 (7.03 t 0.09) was significantly lower than that in group 1 (7.29 + 0.02), group 2 (7.19 + 0.01). or group 3 (7.26 -t 0.03, Fig 2). There were no differences in PCO? levels in either the arterial or SMV samples among the groups during the experimental or recovery phases of the experiment (Figs 3 and 4). Group 4 had a significantly lower oxygen saturation (36 t 4%) during transient hypoxia in comparison with group 3 (59 rt 4%, Fig 5). DISCUSSION
NEC, which accounts for 15% of deaths in newborns less than 1,500 g, continues to be a major cause of morbidity and mortality in premature infants.s The cause of NEC is multifactorial, and numerous animal models have been developed in an attempt to duplicate the pathophysiological conditions that are thought to be necessary to initiate this disease process. Several recent models have focused on intestinal ischemia as the key inciting event for NEC. Bhatia et aI6 used mesenteric arterial occlusion for 1 hour combined with intraluminal injection of platelet activating factor (PAF) and various nutrients to produce histological injury in the rat intestine.6 Musemeche used repetitive mesenteric arterial occlusion and intraluminal PAF in rats to produce similar results.7 Unfortunately, histological ex-
m
Group
Group 2
m
Group 3
n
Group 4
5 7 20
7 IO
7.00 stablllzatmn
experImental
recovery
Fig 2. Mean pH of blood samples from SMV during each phase of the study. Error bars represent one standard deviation. Groups are explained in the text. *P < .05 “groups 1 and 3. +P < .05 vgroups 1,2, and 3.
amination of clinical NEC specimens seldom demonstrates evidence of mesenteric arterial occlusion, such as large vessel thrombi.2 Other investigators have shown that asphyxia produces intestinal necrosis in newborn rats.s.9 However, case-controlled studies have failed to document an increased incidence of NEC in infants who have perinatal asphyxia.lO,” Most often, NEC develops without the catastrophic circulatory events these models use to produce ischemic intestinal damage.i2 A unique neonatal piglet model of NEC has been reported by Di Lorenzo et a1.13In this study, ischemic lesions were produced when segments of intestine (isolated by ligatures) are injected intraluminally with a mixture of bovine casein, calcium gluconate, and propionic acid. They postulate that the organic acid damages the intestinal mucosa and casein promotes growth of gram-negative bacteria. The proliferation of gramnegative bacteria in the presence of impaired gut mucosal Group
T
experimental
@%%
7 30
m
stablkailon
1
I. GRONER
1
i%%$$
Group
2
m
Group
3
m
Group
T
T
1
recovery
Fig 1. Mean pH of blood samples from abdominal aorta during each phase of the study. Error bars represent one standard deviation. Groups are explained in the text. *P < .05 “groups 1 and 3. +P < .05 v groups 1,2, and 3.
expenmental
Fig 3. Mean Pcoz from aorta during bars represent one standard deviation.
each phase
recovery of the study.
Error
4
LPS AND
m
HYPOXIA
Group
CAUSE
1
&@#
ACIDOSIS
Group
IN PIGLETS
2
m
Group
1125
3
0
Group
m
4
Group3
0
Group4
8 I
40 2 g P
30 20 10 L
0 stablhzation
experimental
recovery
stab
exP
ret
stab
aorta Fig 4. Mean Pco, from SMV during bars represent one standard deviation.
each phase
of the study.
ret
exP
-SMV
Error Fig 5. Mean “group 3.
barrier function promotes the translocation of bacteria and endotoxin, an important pathogenic mechanism for NEC.14 The combined effect of endotoxin and hypoxia has been investigated in a rat model by Caplan et a1.15The combination of a 2 mg/kg intravenous infusion of LPS and ventilation with 5% oxygen (for 90 minutes) was found to have synergistic effects on hypotension, metabolic acidosis, and intestinal injury. By contrast, the study described herein involves a tenfold lower dose of LPS and a milder and much briefer hypoxic stress than the model cited above. Nevertheless, this protocol resulted in a profound metabolic acidosis both in the arterial and mesenteric venous circulation without the “catastrophic”
oxygen
saturation
during
transient
hypoxia.
*PC
.05
circulatory aberrations used in many NEC models. The acidosis was caused by the synergism between transient hypoxia and LPS and did not occur with either agent alone. In addition, this model demonstrates that endotoxin causes a significant decrement in arterial oxygen saturation during hypoxia. Interestingly, NEC-like lesions develop spontaneously in “runted” (small for gestational age) piglets and, in some cases, full thickness bowel wall necrosis occurs.16 Future studies will determine if histological changes resembling NEC develop in this piglet model, and, if the aberrations caused by the combination of transient hypoxia and LPS are dose dependent or influenced by the age of the experimental subject.
REFERENCES 1. Kosloske AM: A unifying hypothesis for pathogenesis and prevention of necrotizing enterocolitis. J Pediatr 117:S68-S74, 1990 2. Kliegman RM: Neonatal necrotizing enterocolitis, in Wyllie R, Hyams JS (eds): Pediatric Gastrointestinal Disease, Philadelphia, PA, Saunders, 1993, pp 788-798, Chap 61 3. Motil KJ: Development of the gastrointestinal tract, in Wyllie R, Hyams JS (eds): Pediatric Gastrointestinal Disease, Philadelphia, PA, Saunders, 1993, pp 3-16, Chap 61 4. Hsueh W, Caplan MS, Sun X, et al: Platlet-activating factor, tumor necrosis factor, hypoxia and necrotizing enterocolitis. Acta Paediatr Suppl396:11-17, 1994 5. Necrotizing enterocolitis, in Rowe MI, O’Neill JA, Grosfeld J, et al (eds): Essentials of Pediatric Surgery, St. Louis, MO. Mosby, 1995. pp 526-535, Chap 59 6. Bhatia AM, Fedderson RM, Musemeche CA: The role of luminal nutrients in intestinal injury from mesenteric reperfusion and plateletactivating factor in the developing rat. J Surg Res 63:152-156, 1996 7. Musemeche CA, Baker JL. Feddersen RM: A model of intestinal ischemia in the neonatal rat utilizing superior mesenteric artery occlusion and intraluminal platelet activating factor. J Surg Res 58:124-727,1995 8. Caplan MS,
Hedlund E, Adler L, et al: Role of asphyxia feeding in a neonatal rat model of necrotizing enterocohtis. Pediatr 14: 1017-1028, 1994
and Path
9. Okur H, Kucukaydm M, Kose K, et al: Hypoxia-induced necrotizing enterocolitis in the immature rat: The role of lipid peroxidation and management by vitamin E. J Pediatr Surg 30:1416-1419, 1995 10. Stoll BJ. Kanto WP, Glass RI, et al: Epidemiology of necrotizing enterocolitis: A case control study. J Pediatr 96:447-451, 1980 11. Decurtis M; Paone M, Vetrano G. et al: a case control study of necrotizing enterocohtis occurring over 8 years in a neonatal intensive care unit. Eur J Pediatr 1146:398, 1987 12. Nowicki PT, Nankervis CA: The role of the circulation in the pathogenesis of necrotizing enterocolitis. Clm Perinatol 21:219-234. 1992 13. Di Lorenzo M, Bass J, Krantis A: An intraluminal model of necrotizing enterocolitis in the developing neonatal piglet. J Pediatr Surg 39:1138-1142, 1995 14. Deitch EA: Role of bacterial translocation in necrotizing enterocolitis. Acta Paediatr Supp 396:33-36, 1994 15. Caplan MS, Kelly A, Hsueh W: Endotoxin and hypoxia-Induced necrosis in rats: The role of platelet activatmg factor. Pediatr Res 31:428-434, 1992 16. Thornbury JC, Sibbons PD, van Velzen D, et al: Histological investigations mto the relationship between low birth weight and spontaneous bowel damage m the neonatal piglet. Pediatr Path01 13:59-59, 1993
1126
JONATHAN
I. GRONER
Discussion A.G. Coran (Ann Arboi; MI): When you subject animals to low FIO* of lo%, such as you did here, they usually compensate by hyperventilating and dropping their C02, which will have an effect on the ultimate pH value. Why didn’t that happen in this group of piglets? We have tried a similar experiment in rabbits and they all get low CO2 levels as a result of hyperventilation. J.Z. Groner (response): Because of the mechanical ventilation and the general anesthesia these animals were under, controlled ventilation was maintained throughout the experiment. Thus, their Pcoz levels were controlled. That is the reason that there was no hyperventilation during the experiment. J. Gosche (New Haven, CT): Both hypoxia and endotoxin can affect systemic hemodynamic parameters as well as cardiac output. Do you have any measures of hemodynamic parameters during these studies? Also, because you are looking at whether these are mediators of NEC, did you look at the intestine, and are
there any histological changes that occur after these experiments? J.Z. Groner (response): We only followed arterial blood pressure. Most animals experience significant hypotension during the LPS infusion but then recover. In terms of looking at the intestine, we do have intestine from each of these animals sitting in paraffin blocks awaiting analysis. It is a very interesting question, because piglets that are spontaneously runted and killed will have intestinal lesions that are similar to NEC with ischemic changes, which suggests this would be a very good model for investigating NEC. D. Hackney (Toronto, Ontario): Could you please comment on the degree of resuscitation that each group was given? Was there any mortality in group IV? J.Z. Groner (response): The only resuscitation was replacing the blood volume that was drawn out from the blood gas samplings. There was a piglet in group IV that did not complete the experiment and died, and that pig was excluded from the analysis.
and Transient
Hypoxia
Cause Severe Acidosis
By Jonathan
Columbus, l Purpose: Both hypoxia and gram-negative sepsis are thought to play a role in the development of necrotizing enterocolitis (NEC). Endotoxin, a lipopolysaccharide (LPS), is a potent mediator of gram-negative sepsis. The author investigated the effect of LPS and hypoxia on arterial and mesenteric venous blood gas values in a piglet model. Methods: 16 piglets (mean age, 9 days; mean weight, 3.2 kg) were anesthetized and mechanically ventilated. Catheters were placed in the aorta and the superior mesenteric vein (SMV). After a 30-minute stabilization period, piglets were randomly assigned to four experimental groups: normoxic ventilation (FIo~, 0.211, normoxic ventilation and LPS infusion (200 pg/kg, intravenously), hypoxic ventilation (FIo~, 0.10 for 20 minutes), or hypoxic ventilation and LPS infusion. All subjects were then monitored for an additional 30 minutes (recovery period). Multiple, paired blood gas samples were obtained from the aorta and SMV during the stabilization, experimental, and recovery periods. Results: Piglets subjected to both hypoxia and LPS experienced a much more severe acidosis in both the aorta (pH, 7.10 f 0.08) and SMV (pH, 7.03 f 0.09) than piglets subjected to either hypoxia or LPS alone (P < ,051. In addition, LPS lowered the arterial oxygen saturation in piglets exposed to acute, transient hypoxia (36 r 4% v 59 -c 12%. P < .05). Conclusion: This study suggests that the combination of transient hypoxia and gram-negative sepsis may act synergistically to produce both a severe acidosis and decreased tissue oxygenation. Copyright o 1997 by W.B. Saunders Company
INDEX WORDS: sis, necrotizing
Endotoxin, enterocolitis.
lipopolysaccharide,
hypoxia,
sep-
T
HE ETIOLOGY of necrotizing enterocolitis (NEC) is considered to be multifactorial, and both hypoxia and sepsis are thought to play a role.’ Transient hypoxic events are common in criticahy ill neonates as a result of respiratory distress syndrome, atelectasis. pneumothorax. or endotracheal tube plugging or dislodgment. Sepsis frequently occurs in this population because premature neonates lack a fully competent immune system2 and the newborn’s gut is rapidly colonized with microflora, including gram-negative bacteria.3 Many of the deleterious effects in gram-negative bacterial infections are mediated by endotoxin, a lipopolysaccharide (LPS). High doses of endotoxin and severe hypoxia independently cause damage to the intestine in animal models.” The purpose of this study is to develop a model for investigating the possible synergistic effect of LPS and hypoxia by measuring standard blood gas parameters in both the arterial and mesenteric circulation in the piglet. MATERIALS
AND
METHODS
This experimental protocol was approved by the Institutional Ammal Care and Use Committee of Children’s Hospital. Columbus. Sixteen piglets (mean age, 9 days: mean weight, 3 2 kg) were anesthetized with JournalofPediatric
Surgery,
Vol 32, No 7 (July),
1997, pp 1123-l
126
in the Piglet
I. Groner
Ohio a combmation of Tiletamine HCl, 3 mg/kg intramuscularly (IM). and Zolazepam HCl. 3 mg/kg IM, (Telazol, Fort Dodge Labs. Fort Dodge, IA); and Xylazine, 4 mg/kg IM (Rompum, Miles. Inc. Shawnee Mission, KS). Anesthesia was maintained using intermittent intravenous inJections of Pentobarbital Sodium (Nembutol, Abbott Laboratories. N Chicago. IL). A tracheotomy was performed. and the animals were ventilated with a mechamcal ventilator (Harvard Apparatus, South Nstick, MA) set to deliver 30 breaths per minute at a tidal volume of 65 mL per breath of room air. Fme bore polyethylene catheters (PE-20, Clay-Adams, Parsippany, NJ) were placed in the left internal jugular vein and the aorta via the femoral artery. A midventral laparotomy was then created, and the confluence of the mesenteric veins was identified at the base of the small bowel mesentery. The superior mesenteric vein (SMV) was cannulated by direct puncture m a retrograde fashion with a polyethylene catheter. It was secured with a cyanoacrylate ester glue (Duro Quick Gel, Loctite Corp, Rocky Hill, CT). and it did not occlude SMV blood flow. Body temperature was mamtained throughout the procedure using a warming pad and heat lamps. When the surgical preparation was completed, all animals underwent a 30.minute stabilization period and were then randomly divided into one of four experimental groups. Group 1 piglets (n = 4) were ventilated with room air throughout the expenment and served as controls. Group 2 piglets (n = 4) were ventilated with room air and received an mtravenous injection of 200 pg/kg LPS (from Escherichia co/r K-235, Sigma Chemical Co, St Louis, MO) at the start of the experimental period. Group 3 piglets (n = 4) were ventilated with a gas mixture containing 10% oxygen and 90% nitrogen for 20 minutes. Group 4 piglets received LPS (200 pg/kg) and were ventilated with the 10% oxygen and 90% nitrogen mixture for 20 minutes. After the 20.mmute experimental period, groups 3 and 4 were returned to room air ventilation. and all animals were monitored for an additional 30 minutes (recovery period). Paired sets of blood samples (0.4 mL per sample) were drawn from the aortic and SMV catheters during the stabilization. experimental, and recovery phases usmg heparimzed I-mL syringes. Samples were immediately analyzed for pH. PO?, and Pco~. and oxygen saturation using an automated blood gas analyzer (ABL 300. Radiometer Medical A/S. Copenhagen Denmark). For each treatment group, data from the three phases of the experiment (stabilization, experimental, recovery) were averaged. Treatment groups were then compared using one-way analysis of variance for overall significance and the Student-NewmanKeuls test for intergroup compartsons. A P value of less than .05 was considered sigmficant. RESULTS
There were no differences in pH, Pco2, PO,, and oxygen saturation levels among the four groups during
From the Drpavtmem of Surgeq ChildretI S Hospital, Columbus, OH. Preserlted at the 1996 Anmtal Meeting of the Sectron on Swgev) oj the American Academy of Pediatrics, Boston, Massachusetts, October 26-30, 1996. Address reprrnt requests to Jonathan I. Gronel: MD, 555 S 18th St, Sate 6C, Columbus, OH 43205. Copynght o 1997 by U!B. Saunders Company 0022-3468/97/3207-0040$03.00/O
1123
1124
JONATHAN
the stabilization period. During the experimental period, the pH values in both aorta (7.20 t 0.10) and SMV (7.10 2 0.17) samples from group 4 were more acidotic than group 1 (arterial, 7.35 2 0.02; SMV, 7.34 t 0.03) and group 3 (arterial, 7.40 + 0.02; SMV, 7.34 i 0.05). Group 4 animals also demonstrated a marked arterial acidosis during the recovery phase of the experiment (pH, 7.11 + 0.08) compared with group 1 (7.3 1 ? 0.02), group 2 (7.23 5 0.01) or group 3 (7.24 ? 0.01, Fig 1). The same result was observed in the mesenteric venous circulation. The recovery phase pH level in group 4 (7.03 t 0.09) was significantly lower than that in group 1 (7.29 + 0.02), group 2 (7.19 + 0.01). or group 3 (7.26 -t 0.03, Fig 2). There were no differences in PCO? levels in either the arterial or SMV samples among the groups during the experimental or recovery phases of the experiment (Figs 3 and 4). Group 4 had a significantly lower oxygen saturation (36 t 4%) during transient hypoxia in comparison with group 3 (59 rt 4%, Fig 5). DISCUSSION
NEC, which accounts for 15% of deaths in newborns less than 1,500 g, continues to be a major cause of morbidity and mortality in premature infants.s The cause of NEC is multifactorial, and numerous animal models have been developed in an attempt to duplicate the pathophysiological conditions that are thought to be necessary to initiate this disease process. Several recent models have focused on intestinal ischemia as the key inciting event for NEC. Bhatia et aI6 used mesenteric arterial occlusion for 1 hour combined with intraluminal injection of platelet activating factor (PAF) and various nutrients to produce histological injury in the rat intestine.6 Musemeche used repetitive mesenteric arterial occlusion and intraluminal PAF in rats to produce similar results.7 Unfortunately, histological ex-
m
Group
Group 2
m
Group 3
n
Group 4
5 7 20
7 IO
7.00 stablllzatmn
experImental
recovery
Fig 2. Mean pH of blood samples from SMV during each phase of the study. Error bars represent one standard deviation. Groups are explained in the text. *P < .05 “groups 1 and 3. +P < .05 vgroups 1,2, and 3.
amination of clinical NEC specimens seldom demonstrates evidence of mesenteric arterial occlusion, such as large vessel thrombi.2 Other investigators have shown that asphyxia produces intestinal necrosis in newborn rats.s.9 However, case-controlled studies have failed to document an increased incidence of NEC in infants who have perinatal asphyxia.lO,” Most often, NEC develops without the catastrophic circulatory events these models use to produce ischemic intestinal damage.i2 A unique neonatal piglet model of NEC has been reported by Di Lorenzo et a1.13In this study, ischemic lesions were produced when segments of intestine (isolated by ligatures) are injected intraluminally with a mixture of bovine casein, calcium gluconate, and propionic acid. They postulate that the organic acid damages the intestinal mucosa and casein promotes growth of gram-negative bacteria. The proliferation of gramnegative bacteria in the presence of impaired gut mucosal Group
T
experimental
@%%
7 30
m
stablkailon
1
I. GRONER
1
i%%$$
Group
2
m
Group
3
m
Group
T
T
1
recovery
Fig 1. Mean pH of blood samples from abdominal aorta during each phase of the study. Error bars represent one standard deviation. Groups are explained in the text. *P < .05 “groups 1 and 3. +P < .05 v groups 1,2, and 3.
expenmental
Fig 3. Mean Pcoz from aorta during bars represent one standard deviation.
each phase
recovery of the study.
Error
4
LPS AND
m
HYPOXIA
Group
CAUSE
1
&@#
ACIDOSIS
Group
IN PIGLETS
2
m
Group
1125
3
0
Group
m
4
Group3
0
Group4
8 I
40 2 g P
30 20 10 L
0 stablhzation
experimental
recovery
stab
exP
ret
stab
aorta Fig 4. Mean Pco, from SMV during bars represent one standard deviation.
each phase
of the study.
ret
exP
-SMV
Error Fig 5. Mean “group 3.
barrier function promotes the translocation of bacteria and endotoxin, an important pathogenic mechanism for NEC.14 The combined effect of endotoxin and hypoxia has been investigated in a rat model by Caplan et a1.15The combination of a 2 mg/kg intravenous infusion of LPS and ventilation with 5% oxygen (for 90 minutes) was found to have synergistic effects on hypotension, metabolic acidosis, and intestinal injury. By contrast, the study described herein involves a tenfold lower dose of LPS and a milder and much briefer hypoxic stress than the model cited above. Nevertheless, this protocol resulted in a profound metabolic acidosis both in the arterial and mesenteric venous circulation without the “catastrophic”
oxygen
saturation
during
transient
hypoxia.
*PC
.05
circulatory aberrations used in many NEC models. The acidosis was caused by the synergism between transient hypoxia and LPS and did not occur with either agent alone. In addition, this model demonstrates that endotoxin causes a significant decrement in arterial oxygen saturation during hypoxia. Interestingly, NEC-like lesions develop spontaneously in “runted” (small for gestational age) piglets and, in some cases, full thickness bowel wall necrosis occurs.16 Future studies will determine if histological changes resembling NEC develop in this piglet model, and, if the aberrations caused by the combination of transient hypoxia and LPS are dose dependent or influenced by the age of the experimental subject.
REFERENCES 1. Kosloske AM: A unifying hypothesis for pathogenesis and prevention of necrotizing enterocolitis. J Pediatr 117:S68-S74, 1990 2. Kliegman RM: Neonatal necrotizing enterocolitis, in Wyllie R, Hyams JS (eds): Pediatric Gastrointestinal Disease, Philadelphia, PA, Saunders, 1993, pp 788-798, Chap 61 3. Motil KJ: Development of the gastrointestinal tract, in Wyllie R, Hyams JS (eds): Pediatric Gastrointestinal Disease, Philadelphia, PA, Saunders, 1993, pp 3-16, Chap 61 4. Hsueh W, Caplan MS, Sun X, et al: Platlet-activating factor, tumor necrosis factor, hypoxia and necrotizing enterocolitis. Acta Paediatr Suppl396:11-17, 1994 5. Necrotizing enterocolitis, in Rowe MI, O’Neill JA, Grosfeld J, et al (eds): Essentials of Pediatric Surgery, St. Louis, MO. Mosby, 1995. pp 526-535, Chap 59 6. Bhatia AM, Fedderson RM, Musemeche CA: The role of luminal nutrients in intestinal injury from mesenteric reperfusion and plateletactivating factor in the developing rat. J Surg Res 63:152-156, 1996 7. Musemeche CA, Baker JL. Feddersen RM: A model of intestinal ischemia in the neonatal rat utilizing superior mesenteric artery occlusion and intraluminal platelet activating factor. J Surg Res 58:124-727,1995 8. Caplan MS,
Hedlund E, Adler L, et al: Role of asphyxia feeding in a neonatal rat model of necrotizing enterocohtis. Pediatr 14: 1017-1028, 1994
and Path
9. Okur H, Kucukaydm M, Kose K, et al: Hypoxia-induced necrotizing enterocolitis in the immature rat: The role of lipid peroxidation and management by vitamin E. J Pediatr Surg 30:1416-1419, 1995 10. Stoll BJ. Kanto WP, Glass RI, et al: Epidemiology of necrotizing enterocolitis: A case control study. J Pediatr 96:447-451, 1980 11. Decurtis M; Paone M, Vetrano G. et al: a case control study of necrotizing enterocohtis occurring over 8 years in a neonatal intensive care unit. Eur J Pediatr 1146:398, 1987 12. Nowicki PT, Nankervis CA: The role of the circulation in the pathogenesis of necrotizing enterocolitis. Clm Perinatol 21:219-234. 1992 13. Di Lorenzo M, Bass J, Krantis A: An intraluminal model of necrotizing enterocolitis in the developing neonatal piglet. J Pediatr Surg 39:1138-1142, 1995 14. Deitch EA: Role of bacterial translocation in necrotizing enterocolitis. Acta Paediatr Supp 396:33-36, 1994 15. Caplan MS, Kelly A, Hsueh W: Endotoxin and hypoxia-Induced necrosis in rats: The role of platelet activatmg factor. Pediatr Res 31:428-434, 1992 16. Thornbury JC, Sibbons PD, van Velzen D, et al: Histological investigations mto the relationship between low birth weight and spontaneous bowel damage m the neonatal piglet. Pediatr Path01 13:59-59, 1993
1126
JONATHAN
I. GRONER
Discussion A.G. Coran (Ann Arboi; MI): When you subject animals to low FIO* of lo%, such as you did here, they usually compensate by hyperventilating and dropping their C02, which will have an effect on the ultimate pH value. Why didn’t that happen in this group of piglets? We have tried a similar experiment in rabbits and they all get low CO2 levels as a result of hyperventilation. J.Z. Groner (response): Because of the mechanical ventilation and the general anesthesia these animals were under, controlled ventilation was maintained throughout the experiment. Thus, their Pcoz levels were controlled. That is the reason that there was no hyperventilation during the experiment. J. Gosche (New Haven, CT): Both hypoxia and endotoxin can affect systemic hemodynamic parameters as well as cardiac output. Do you have any measures of hemodynamic parameters during these studies? Also, because you are looking at whether these are mediators of NEC, did you look at the intestine, and are
there any histological changes that occur after these experiments? J.Z. Groner (response): We only followed arterial blood pressure. Most animals experience significant hypotension during the LPS infusion but then recover. In terms of looking at the intestine, we do have intestine from each of these animals sitting in paraffin blocks awaiting analysis. It is a very interesting question, because piglets that are spontaneously runted and killed will have intestinal lesions that are similar to NEC with ischemic changes, which suggests this would be a very good model for investigating NEC. D. Hackney (Toronto, Ontario): Could you please comment on the degree of resuscitation that each group was given? Was there any mortality in group IV? J.Z. Groner (response): The only resuscitation was replacing the blood volume that was drawn out from the blood gas samplings. There was a piglet in group IV that did not complete the experiment and died, and that pig was excluded from the analysis.