Experimental Necrotizing Enterocolitis: The Role of Polymorphonuclear Neutrophils By C. Musemeche,

M. Caplan, W. Hsueh, X. Sun, and A. Kelly Chicago, Illinois

l Polymorphonuclear neutrophils (PMNs) play an important role in inflammation. Activated PMNs adhere to the vascular wall and release reactive

oxygen radicals and en-

zymes, producing vascular injury. In the present study, we investigated whether PMNs play an important role in the pathogenesis of experimental necrotizing enterocolitis (NEC). NEC was induced in rats using platelet activating factor (PAF, 1 pg/kg) and bacterial endotoxin (LPS, 1 mg/kg) intravenously. Neutropenia was accomplished by parenteral injection of Vinblastine (VB, 0.75 mg/kg) 4 days before the experiment to deplete the total white blood cell (WBC) and neutrophil counts. The animals were divided into 4 groups: (I) 1 kg/kg PAF; (2) 1 mg/kg LPS; (3) 1 pg/kg PAF + 1 mg/kg LPS; and (4) PMN depleted, 1 pg/ kg PAF + 1 mg/kg LPS. Combined administration of PAF and LPS produced prolonged hypotension (blood pressure 53.5 * 13.8 mm Hg at 2 hours), leukopenia (4,082 ? 497.4), hemoconcentration (hematocrit 44.5% & 1.1%). reduced intestinal perfusion (74% 2 13.3%). and segmental

bowel necrosis. However,

in

VB-treated animals combined PAF + LPS induced only mild hypotension (84.3 2 9.2 mm Hg at 2 hours) and no hemoconcentration. In these animals the intestinal perfusion was normal, no bowel necrosis was observed, and the intestinal myeloperoxidase activity f.0034 f .0017 U/g tissue) was significantly lower than that of the nondepleted group (.0075 2 .0012 U/g tissue). We conclude that the presence of neutrophils and/or neutrophil products play a major role in the pathogenesis of NEC. Copyright o 1991 by W.B. Saunders Company INDEX WORDS: Necrotizing factor, intestinal ischemia.

enterocolitis;

platelet activating

A

FT’ER 20 YEARS of investigation, the etiology of necrotizing enterocolitis (NEC) remains ob-

scure. It is generally believed that intestinal ischemic injury plays a role in the pathogenesis of this disease, which is frequently fatal in premature infants. The manner in which inflammatory mediators produce NEC has not been thoroughly explored. Platelet activating factor (PAF), an endogenous phospholipid mediator produced by a wide variety of cells, is produced by ischemic intestinal tissue. Previous work has shown that injection of PAF in rats causes intestinal necrosis and lesions that are morphologically indistinguishable from those found in human NEC.’ It also has been shown that bacterial endotoxin (LPS) acts synergistically with PAF to induce bowel necrosis. Although the presence of neutrophils in intestinal conditions has long been recognized, the functional significance of neutrophils and neutrophil products in intestinal inflammation is only now becoming appar-

Journal of Pediatric Surgery, Vol 26, No 9 (September), 1991: pp 1047-1050

ent. Stimulated neutrophils produce superoxide and other highly reactive oxygen products capable of inducing cellular injury.’ The purpose of this study was to evaluate the effects of vinblastine-induced neutropenia on the production of NEC in an established model. MATERIALS AND METHODS Young male Sprague-Dawley rats weighing 70 to 80 g were anesthetized with intraperitoneal pentobarbital. A tracheotomy was performed to facilitate breathing. The carotid artery and jugular vein were catheterized for continuous recording of blood pressure, drug administration and blood sampling. PAF (1-O-akyl2-0-acetyl-sn-glycero-3-phosphocholine; Calbiochem-Behring, La Jolla, CA) was diluted from a stock solution in ethanol and stored at -70°C. Working solutions were prepared in 2.5 mg/dL of bovine serum albumin saline solution. Lipopolysaccharide-B (LPS) from Sulmoneffu typhosa was purchased from Difco (Detroit. MI). Fresh saline solution was prepared before each use. Vinblastine sulphate (VB; Cetus, Emeryville, CA ) was dissolved in physiological saline and injected intravenously (0.75 mg/kg) 4 days prior to the acute experiment to achieve polymorphonuclear neutrophil (PMN) depletion.3 To assess the effect of this treatment on peripheral leukocyte counts, venous blood samples were obtained before injection and at various times after injection by pricking hindpaw veins. Total white blood cell (WBC) and differential counts were determined using unopettes and Wright stained smears of peripheral blood, respectively (Table 1). In the first part of the experiment the animals were divided into 4 groups: (1) 1 )*g/kg PAF (n = 4); (2) 1 mglkg LPS (n = 4); (3) I @g/kg PAF + 1 mg/kg LPS (n = 5); and (4) PMN depleted, 1 kg/kg PAF + 1 mgikg LPS (n = 5). Blood samples were taken periodically to determine hematocrit and WBC counts. The extent of intestinal perfusion was determined at 2 hours with 2 mL of 5% Evans Blue. The length of nonstained bowel was recorded and sections were taken for histological examination to confirm the presence of necrosis. In the second part of the experiment the entire small bowel was homogenized and intestinal myeloperoxidase (MPO) activity has determined in PAF + LPS-treated animals with and without PMN depletion (n = 10) after the method of Krawisz et al.4 MPO is an enzyme produced by neutrophils and its activity in tissue correlates with the number of PMNs present. This assay was performed to

From the Children> Memorial Hospital, Evanston Hospital, and Northwestern University Chicago, IL. Presented at the 42nd Annual Meeting ofthe Surgical Section of the American Academy of Pediatrics, Boston, Massachusetts, October 6-7, 1990. Address reprint requests to Catherine A. Musemeche, MD, Division of Pediatric Surgev, University of Texas Medical School at Houston, 6431 Fannin, Suite 6.264, Houston, TX 77030. Copyright o 1991 by WB. Saunders Company 0022-3468191/2609-0008$03.0010 1047

MUSEMECHE

1048

ET AL

Table 1. Effect of Vioblastine on Systemic Leukocyte and

50

Neutrophil Counts Group

Total WBC

No pretreatment Vinblastine treatment

Neutrophils

15,500 + 450.0

630 + 69

2,700 t 312.9

55 ? 29

40 a**

C &

30

confirm that PMN depletion was achieved in the organ of interest, the intestine, as well as systemically.

Several major systemic changes were observed in animals treated with PAF + LPS (Figs 1 to 3). These rats developed marked hypotension (blood pressure 53.5 + 13.8 mm Hg) (Fig l), moderate leukopenia (4,062 + 497.4), and hemoconcentration (hematocrit 44.5% + 1.1%). In contrast, PMN-depleted animals treated with combined PAF + LPS developed only mild hypotension (blood pressure 84.3 + 9.2 mm Hg at 2 hours) and no hemoconcentration. The difference in hemoconcentration (Fig 2) between nondepleted (44.5 f 1.1) and PMN-depleted (28.6 + 2.3) animals was significant (P < -001, t test). Systemic leukopenia (Fig 3), which develops after PAF + LPS administration during the experiment, was significantly more severe in nondepleted animals (P < .OOl, t test). Hypotension and leukopenia are known systemic effects of PAF administrati0n.l.j Low-dose PAF (1 kg/kg) or 1 mg/kg LPS alone did not cause any microscopic or gross bowel lesions. In contrast, all 4 animals treated with the combination of PAF and LPS developed intestinal hypoperfusion and focal necrosis. The gross involvement varied from 10% to 60% of the small intestine and microscopically from mild to moderate degrees of necrosis. In all 5 PMN-depleted animals there was normal intestinal perfusion after combined PAF + LPS treatment.

120

E 2 i

100

;

60

20 ’

Nondepleted

l

PMN Depleted

10

RESULTS

B

E

I

l

0

0 I

Oi 0

30

60

90

120

Time (mid Fig 2. Effects of PAF + LPS on hematocrit PMN-depleted animals.

in nondepleted

and

The MPO activity (Fig 4) in PMN-depleted animals treated with combined PAF and LPS was .0034 + .0017 U/g tissue and was significantly lower (P < .012, t test) than that of the nondepleted animals (.0075 + .0012 U/g tissue). DISCUSSION

The concept that PMNs may partially mediate ischemic injury was proposed by Romson et al in 1983 with respect to the postischemic myocardium.6 They observed that neutrophil depletion reduced infarct size by as much as 50% in experimental models. Since then the significance of neutrophil-mediated injury in the postischemic intestine has been investigated. In 1986, Grisham et al noted a dramatic increase in neutrophil infiltration after reperfusion of ischemic intestine.7 The proposed mechanism of injury is thought to be due to the release by stimulated neutrophils of reactive oxygen metabolites (superox-

o Nondepleted 0 **

p q.001

80

v ;

PMN Depleted

0

Nondepleted

.

PMN Depleted

20-

r 0

’ 0

I 30

I 60

I 90

I 120

o0

30

arterial blood pressure in

90

120

Time (mid

Time (mid Fig 1. Effects of PAF + LPS on-mean nondepleted and PMN-depleted animals.

60

Fig 3. Systemic leukopenia nondepleted end PMN-depleted

after PAF + LPS administration animals.

in

EXPERIMENTAL

NECROTIZING

ENTEROCOLITIS

1049

Depleted Fig 4. Myeloperoxidase activity (shaded bar) and systemic neutrophi1 counts (black bar) after PAF + LPS treatment in nondepleted and PMN-depleted animals.

ide, hydrogen peroxide, and the hydroxide radical). The gastrointestinal tract is a particularly good source of reactive oxygen metabolites due to an abundance of oxidative enzymes found in the mucosa and lamina propria and from resident phagocytic leukocytes. In previous work it has been shown that PAF alone

is sufficient to provoke intestinal necrosis only at doses that approach the lethal dose for this potent mediator. A combination of PAF and LPS has been found most effective in inducing intestinal necrosis and the data suggest that a mutually enhancing effect develops when these two compounds are administered together.’ The intestinal injury produced by this model has been shown to be mediated in part by oxygen radicals, complement, and leukotrienes.8,9 Because some of these mediators are produced by and interact with PMNs, we became interested in whether PMNs play an important role in the pathogenesis of NEC. We conclude that PMN depletion protected animals from some of the systemic and local affects of PAF + LPS treatment in this experimental model. We do not know how this occurred, although we can hypothesize that these affects are related to previously described mediators of PMN injury.7-9The exact mechanisms remain the subject for future experiments. The understanding of these complex PMNderived mediators and their manipulation may provide new insights into the treatment and prevention of NEC.

REFERENCES 1. Gonzalez-Crussi F, Hsueh W: Experimental model of ischemic bowel necrosis: The role of platelet activating factor and endotoxin. Am J Path01 112:127-135,1983 2. Grisham MB, Granger DN: Neutrophil-mediated mucosal injury: Role of reactive oxygen metabolites. Dig Dis Sci 33:6S-15S, 1988 (suppl) 3. Lemanske RF, Guthman BA, Oertel H, et al: The biologic activity of mast cell granules. VI. The effect of vinblastine-induced neutropenia on rat cutaneous late phase reactions. J Immunol 130:2837-2842,1983 4. Krawisz JE, Sharon P, Stenson WF: Quantitative assay for acute intestinal inflammation based on myeloperoxidase activity: Assessment of inflammation in rat and hamster models. Gastroenterology 87:1344-1350, 1984 5. Sanchez-Crespo M, Alonso F, Inarrea P, et al: Vascular actions of synthetic PAF-acether (a synthetic platelet-activating

factor) in the rat: Evidence for a platelet independent Immunopharmacology 4:173-1851982

mechanism.

6. Romson JL, Hook BG, Kunkel SL, et al: Reduction of the extent of ischemic myocardial injury by neutrophil depletion in the dog. Circulation 67:1016-1023, 1983 7. Grisham MB, Hernandez LA, Granger DN: Xanthine oxidase and neutrophil infiltration in intestinal ischemia. Am J Physiol 251:G567-G574,1986 8. Cueva JP, Hsueh W: Role of oxygen derived free radicals in platelet activating factor induced bowel necrosis. Gut 29:12071212,1988 9. Hsueh W, Gonzalez-Crussi F, Arroyave JL: Platelet-activating factor-induced ischemic bowel necrosis: An investigation of secondary mediators in its patbogenesis. Am J Path01 122:231-239, 1986

Discussion K Lally (Sun Antonio, TX): Despite intensive work for a number of years, we have yet to achieve a true model of NEC. Some animal models have examined at whole organ ischemia, but epidemiologic data make this a very unlikely inciting cause of human NEC. The model described by Dr Hsueh and her colleagues at Northwestern does not involve vascular occlusion. However, it does involve direct PAF injection, which is still not an ideal model. Do .you think

that whole-organ ischemia plays a major role in the pathogenesis of NEC, or is the etiology primarily microvascular hypoperfusion induced by inflammatory mediators such as PAF? Have you or your colleagues looked at blocking PAF therapeutically in this animal model? Your data imply that reactive oxygen metabolites, presumably generated by the PMN are the cause of tissue injury. One current theory of intestinal injury is that the initial insult is

1050

caused by xanthine oxidase-induced free radical generation that causes PMN influx and injury from PMN-generated radicals. However, xanthine oxidase levels in young rats as well as young piglets have been shown to be much lower than those present in adult animals. What do you think is the stimulus for PMN influx in your model? Why is the effect of systemically injected LPS/PAF seem primarily in the bowel? Did you look at other organs for evidence of necrosis? T. Lobe (Memphis, TN): From an experimental point of view, what happens if you give Vinblastine alone to these animals? R, Powell (Birmingham, AL): I have some of the same problems that Dr Lally did about the model itself and the title of the paper as experimental NEC. Would this be better termed “ischemic necrosis from septic shock” because of the model itself? Your dose of 1 pg/kg should result in significant derangements in the animal because PAF in a dose of about 250 ng/kg causes about a 50% decrease in pressure. Is the effect of the PAF due to its overall effect on blood pressure and a secondary effect on the gut? In the group using only the PAF you did not get much in the way of intestinal necrosis; could you comment on that? Also, how reliable is the Evans blue dye for evaluating intestinal blood flow? For example, how does it compare with fluorescein, which is probably the gold standard at least for the visual evaluation of intestinal blood flow? I think your study does reemphasize the role of the interplay between mediators and the white cell, and I think it does confirm other reports. C. Musemeche (response): With respect to the question about whole organ ischemia versus inflammation, I think many of the earlier animal models for NEC focussed on whole organ ischemia and a direct insult to the mesenteric blood supply. However, there

MUSEMECHE

ET AL

are several problems with that type of model. First, as you mentioned, that does not exactly fit the epidemiologic data we have in terms of the babies that do develop NEC. The other problem with that type of model is that it is very difficult to quantitate that type of insult. I think that recent data have shown that much of NEC is mediated in an inflammatory fashion with such mediators as PAF and tumor necrosis factor. With respect to the therapeutic manipulations, most studies involving animal models have used prophylactic treatment or pretreatment, but I do know of one experiment that Dr Hsueh conducted in Chicago in which a leukotriene antagonist was continuously infused after PAF injection and this did result in improved perfusion pressure of the intestine. With respect to what is the stimulus to PMN influx, it has been shown that reactive oxygen metabolites may have something to do with PMN influx, but the actual chemoattractants are more from the complement components that may be stimulated by the reactive oxygen components as well as leukotrienes. These can be released by oxygen radicals, but they can also be released by direct injury to the mucosa. As to what other organs are affected by this particular dose of PAF, the intestine is the only organ that is affected; however, in higher doses you can see partial necrosis of the liver. Dr Powell, you are right in saying that a more appropriate title would perhaps have to do with intestinal necrosis and not NEC per se. We are calling it experimental NEC, because it is hard to be exact about what we are producing. We do know that in newborn babies with NEC, the levels of PAF and TNF are elevated. We have used Evans blue dye in our laboratory for a number of years and have consistent results. Dr Lobe, if you give Vinblastine alone, of course you do have decreased white counts but there is no effect on the intestine.

Experimental necrotizing enterocolitis: the role of polymorphonuclear neutrophils.

Polymorphonuclear neutrophils (PMNs) play an important role in inflammation. Activated PMNs adhere to the vascular wall and release reactive oxygen ra...
485KB Sizes 0 Downloads 0 Views