Inhibition of angiogenesis by nonsteroidal anti-inflammatory drugs: insight into mechanisms and implications for cancer growth and ulcer healing.

Angiogenesis, the formation of new capillary blood vessels, is essential not only for the growth and metastasis of solid tumors, but also for wound and ulcer healing, because without the restoration of blood flow, oxygen and nutrients cannot be delivered to the healing site. Nonsteroidal anti-inflammatory drugs (NSAIDs) such as aspirin, indomethacin and ibuprofen are the most widely used drugs for pain, arthritis, cardiovascular diseases and, more recently, the prevention of colon cancer and Alzheimer disease. However, NSAIDs produce gastroduodenal ulcers in about 25% of users (often with bleeding and/or perforations) and delay ulcer healing, presumably by blocking prostaglandin synthesis from cyclooxygenase (COX)-1 and COX-2 (ref. 10). The hypothesis that the gastrointestinal side effects of NSAIDs result from inhibition of COX-1, but not COX-2 (ref. 11), prompted the development of NSAIDs that selectively inhibit only COX-2 (such as celecoxib and rofecoxib). Our study demonstrates that both selective and nonselective NSAIDs inhibit angiogenesis through direct effects on endothelial cells. We also show that this action involves inhibition of mitogen-activated protein (MAP) kinase (ERK2) activity, interference with ERK nuclear translocation, is independent of protein kinase C and has prostaglandin-dependent and prostaglandin-independent components. Finally, we show that both COX-1 and COX-2 are important for the regulation of angiogenesis. These findings challenge the premise that selective COX-2 inhibitors will not affect the gastrointestinal tract and ulcer/wound healing.

Studies of the hepatic excretory defects in essential fatty acid deficiency. Their possible relationship to the genesis of cholesterol gallstones.

Male hamsters were fed normal and essential fatty acid (EFA)-deficient diets for at least 12 wk before bile duct cannulation. With [32P]phosphate, hepatic synthesis of lecithin was similar, but biliary excretion of newly synthesized lecithin was significantly reduced in EFA-deficient compared to that in normal hamsters. Hepatic uptake of intravenously infused taurocholate (TC) and taurochenodeoxycholate (TCDC) were similar in both groups of animals. However, biliary excretion of intravenously infused TC was significantly reduced in EFA-deficient hamsters, whereas that of TCDC-was unchanged. The absolute rate of biliary cholesterol excretion was similar in both groups. Canalicular bile flow, as measured by [14C]erythritol clearance after functional nephrectomy, was significantly lower, with both the bile salt-dependent and independent fractions of this flow being diminished in EFA-deficient hamsters infused with TC. It is concluded that EFA deficiency leads to impaired biliary excretion of taurocholate, lecithin, and water, while cholesterol transport is unaffected, and thus results in supersaturation of bile with respect to cholesterol and production of lithogenic bile.

Prostaglandin protection of human isolated gastric glands against indomethacin and ethanol injury. Evidence for direct cellular action of prostaglandin.

Isolated human gastric glands from surgical specimens were preincubated in an oxygenated medium with placebo or 16,16 dimethyl prostaglandin E2 (dmPGE2) and incubated at 37 degrees C in either medium alone, medium containing 4.43 mM indomethacin or medium containing 8% ethanol. We assessed the viability of gland cells with fast green exclusion, release of lactate dehydrogenase (LDH) into the medium, and ultrastructural damage by scanning and transmission electron microscopy. Both indomethacin and ethanol significantly reduced the viability of placebo-pretreated glands, increased LDH release into the medium, and produced prominent ultrastructural damage. DmPGE2 significantly reduced both indomethacin and ethanol-induced injury, increased the number of viable cells, reduced LDH release, and diminished the extent of ultrastructural damage. These studies indicate that PG protection of gastric mucosal cells has a direct cellular action that is not limited to replacement of depleted endogenous PGs. PG protection in our experiments did not depend on PG's previously described systemic actions, such as protection of the microvessels, preservation of the mucosal blood flow, or stimulation of bicarbonate and mucus secretion.

Modeling splanchnic hemodynamics after distal splenorenal shunt: a computer simulation and sensitivity analysis.

The distal splenorenal shunt operation was specifically designed to preserve portal flow and maintain elevated portal pressure. However, although this goal is met in the immediate postoperative period, flow decreases over time, and in as many as 75% of alcoholic patients, portal flow is lost in the first year. Various explanations have been offered for this observation, and modifications of the original operation have been proposed (splenopancreatic disconnection). Although other portacaval shunts have been successfully modeled as electrical circuits, this approach has never been described for the distal splenorenal shunt. In this study we developed a computer program that modeled the distal splenorenal shunt as an electrical circuit. We performed an analysis to determine the sensitivity of portal flow to changes in each resistance element and then performed a series of simulation experiments to critically examine the various explanations offered for the gradual changes in hepatic hemodynamics. We found that portal flow was most sensitive to resistance in the renal vein followed by resistance in the anastomosis. The simulation experiments suggested a new alternative to splenopancreatic disconnection--restricting the ability of the splenic vein, anastomosis, or renal vein to dilate. Additional clinical studies will be needed to test these predictions.

Quantitation and fractionation of nutrient hepatic blood flow in normal persons, in persons with portal hypertensive cirrhosis, and after small-diameter portacaval H grafts.

Patients maintaining portal perfusion following small-diameter portacaval H grafts have better survival and lower portasystemic encephalopathy rates than those with reversed flow. To determine why this is so, we measured nutrient hepatic blood flow with the use of 99m-Tc-diisopropyl-IDA (DISIDA) clearance pharmacokinetics fractionated into its hepatic arterial and portal venous components. Patients with cirrhosis and portal hypertension had significantly lower nutrient hepatic blood flow than normal persons; this was due almost entirely to reduced portal flow. In patients with prograde portal flow after small-diameter H grafts nutrient hepatic blood flows were nominally reduced from levels seen in patients with portal hypertensive cirrhosis. Postoperative patients with reversed portal flow had significantly less nutrient hepatic blood than those with prograde flow. There was no evidence of significant hepatic arterial compensation for lost portal flow. Of four hemodynamic variables--portal flow direction, portal flow, arterial flow, and nutrient hepatic blood flow--only nutrient hepatic blood flow showed an independent correlation with clinical outcome. Portal perfusion is a critical factor in maintenance of adequate nutrient hepatic blood flow, primarily because hepatic arterial flow does not compensate chronically for lost portal perfusion.

Serial measurement of portal hemodynamics after partial portal decompression.

In a serial analysis of splanchnic hemodynamics, we compared partial with total portal decompression in 16 alcoholic cirrhotic patients who underwent portacaval shunts for variceal hemorrhage. Partial decompression was achieved with 8 or 10 mm polytetrafluorethylene portacaval H grafts and aggressive collateral ligation. Total decompression was achieved with larger diameter H grafts (12 or 14 mm). Early and follow-up (mean interval, 18 months) postoperative studies of portal hemodynamics included: direct measurement of shunt gradients, scintigraphic quantitation of portal and mesenteric flow distribution to the liver, and a portal and splenic collateral scoring system developed from standardized splenic venography. Partial portal decompression reduced portal pressure by 43% +/- 8% compared with 81% +/- 5% after total decompression (p less than 0.01). Scintigraphy demonstrated that partial decompression provided a greater fraction of portal flow to the liver than did total decompression (57% +/- 9% versus 2% +/- 1% intrahepatic radioactivity) and mesenteric flow distribution (14.5% +/- 5.4% versus 1.2% +/- 0.7%). Only one patient with partial decompression had a significant loss of portal perfusion during the interval studies. Significantly more residual collaterals were visualized in patients with partial decompression than in those with total decompression, and interval studies showed no significant changes from early studies. We conclude that partial decompression maintains higher portal pressures, more residual collaterals, and a greater fraction of portal and mesenteric flow to the liver than does total decompression. A modest but uniform reduction of portal pressure minimizes stimulus for new collateral formation and further shunting of portal flow.

Epidermal growth factor protects portal hypertensive gastric mucosa in ischemia/reperfusion: the role of capillary endothelia and prostaglandins.

BACKGROUND: Epidermal growth factor (EGF) protects gastric mucosa against a variety of injurious agents, but the mechanism is unclear. Because the abnormal microvasculature of portal hypertensive (PHT) gastric mucosa is a major target of ischemia/reperfusion (I/R) injury, we used this model to assess EGF's protective role at the microvascular level. METHODS: Rats with PHT (staged portal vein ligation) received either EGF, 20 micrograms/kg, or saline solution intravenously, with or without indomethacin pretreatment (20 mg/kg subcutaneously). I/R was produced by withdrawing blood to systemic pressures of 30 mm Hg for 20 minutes and reinfusing it. Stomachs were excised 20 minutes later and evaluated for gross and histologic necrosis, microvascular permeability, mucosal ultrastructure and vimentin, and cyclooxygenase immunofluorescence. RESULTS: In saline-treated rats, gross and histologic damage involved 46% +/- 3% of glandular mucosa and 23% +/- 3% of mucosal sections, respectively. Microvascular permeability was increased 43-fold over that of normal control rats. Vimentin immunofluorescence intensity was reduced to 36% +/- 4% that of normal control rats. EGF pretreatment reduced histologic necrosis to 2% +/- 1% (p less than 0.01). Microvascular permeability and vimentin intensity were almost normalized. Indomethacin partially reversed the mucosal protection induced by EGF. CONCLUSIONS: EGF significantly reduces I/R injury to PHT gastric mucosa. Microvascular endothelia of PHT gastric mucosa are the major target of I/R injury and the site of EGF's protective action. Prostaglandins in part mediate EGF's protective action.

The portal hypertensive gastric mucosa: histologic, ultrastructural, and functional analysis after aspirin-induced damage.

We assessed macroscopic, histologic, ultrastructural, and functional features of aspirin-induced gastric mucosal injury in portal hypertensive and sham-operated rats. Portal hypertension was produced by staged portal vein ligation. Four hours after intragastric acidified aspirin administration, intraluminal pH in portal hypertensive rats was 6.6 +/- 0.2 and 4.3 +/- 0.5 in sham-operated controls (p less than 0.01). Gross mucosal damage was significantly greater in portal hypertensive rats compared with controls (18 +/- 2 versus 7 +/- 1% of total mucosal area). Histologic deep necrosis involved 22 +/- 2% of mucosal section lengths in portal hypertensive rats compared with 7 +/- 1% in sham-operated rats (p less than 0.01). In portal hypertensive rats, histologic and ultrastructural evaluation demonstrated capillary endothelial abnormalities, arterialization of submucosal veins, and markedly greater severity of microvascular damage than in sham-operated controls. Neutralized aspirin (pH, 7.0) did not produce any significant damage detectable grossly, histologically, or by transmission electron microscopy in portal hypertensive rats. We conclude that acid-dependent aspirin-induced gastric mucosal damage is significantly increased in portal hypertension.

Reduced expression of basic fibroblast growth factor and its receptor mRNAs and proteins in portal hypertensive esophageal mucosa: a mechanism responsible for muscularis mucosae thinning and variceal rupture.

BACKGROUND: Basic fibroblast growth factor (bFGF) enhances cell migration, proliferation, and tissue integrity. This especially pertinent to the smooth muscle cells in which it stimulates cell proliferation and promotes their growth. The aim of this study was to determine whether expression of bFGF and its receptors (FGFR-1 and -2) is altered in portal hypertensive esophageal mucosa, especially in the muscularis mucosal layer, which constitutes a physical barrier to variceal rupture. METHODS: Portal hypertension (PHT) was produced by staged portal vein ligation. In 30 PHT rats and 30 sham-operated controls 2 weeks after operation, specimens of lower esophagus were obtained for (1) quantitative histologic assessment including thickness of epithelium and muscularis mucosae; (2) immunostaining with specific antibodies against bFGF and its receptors 1 and 2 (intensity of bFGF, FGFR-1 and FGFR-2 immunostaining in esophageal structures was measured with a video image system); and (3) expression of bFGF and FGFR-1 and -2 mRNAs was assessed with reverse transcription-polymerase chain reaction. RESULTS: The esophageal muscularis mucosae and epithelium overlaying large submucosal veins in PHT rats significatly thinner than those in controls (muscularis mucosae, 28.3 +/- 1.4 versus 52.2 +/- 8.0 micrometer, respectively, p<0.05); epithelium, 39.0+/- 7.1 versus 49.3 +/- 1.9 micrometer, respectively, p<0.05). The immunostaining intensity of bFGF and FGFR-2 was significantly reduced in PHT rats (42.1 +/- 2.3 and 71.3 +/- 6.5 units, respectively) versus controls (49.5 +/- 5.6 and 78.6 +/- 5.7 units, respectively, p< 0.05). Expressions of bFGF and FGFR-2 mRNAs in PHT esophageal mucosa were significantly reduced versus controls by 30.8% and 30.3%, respectively (p < 0.01, p 0.05). CONCLUSIONS: (1) Esophageal mucosa of PHT rats has thinner muscularis mucosae and reduced bFGF and FGFR-2 mRNAs and proteins. (2) Because bFGF stimulates smooth muscle cell proliferation and their growth, our findings can explain thinning of esophageal muscularis mucosae in PHT rats, thus indicating a possible mechanism for rupture of varices in the esophagus.

Selective impairment of nutrient absorption from intestines with chronic venous hypertension.

Malnutrition is frequently associated with advanced cirrhosis. To investigate the role of portal hypertension in nutritional impairment, we developed an animal model to isolate and characterize the effects of chronic intestinal venous hypertension on intestinal nutrient absorption. We performed mesenteric arteriovenous anastomosis combined with portal vein banding in rats. Hepatic architecture and excretory function (bile flow and bile salt output) were unaltered, while severe and persistent intestinal venous hypertension was produced. We then measured in vivo absorption rates of three test nutrients (vitamin D3, valine, and tryptophan) and water. Vitamin D3 absorption was significantly impaired by intestinal congestion, while amino acid absorption was unaffected. Splanchnic hypertensive rats absorbed less water than controls. We conclude that chronic intestinal venous hypertension alone selectively impairs nutrient absorption.

Overexpression of endothelin-1 mRNA and protein in portal hypertensive gastric mucosa of rats: a key to increased susceptibility to damage?

BACKGROUND: Portal hypertension predisposes gastric mucosa to increased injury by various noxious factors. Because endothelin-1 (ET-1) is a potent vasoconstrictor that enhances gastric mucosal injury, we examined ET-1 expression in the portal hypertensive (PHT) gastric mucosa and its possible role in increased mucosal susceptibility to damage. METHODS: In gastric specimens of PHT or sham-operated rats, ET-1 mRNA expression was studied by S1-nuclease protection assay and ET-1 protein by enzyme immunoassay and immunostaining. We also determined the extent of ethanol-induced gastric mucosal necrosis in PHT and sham-operated rats after administering either a placebo or FR 139317, a selective ETA receptor antagonist. RESULTS: In PHT stomachs ET-1 mRNA expression and protein concentration were significantly increased compared with sham-operated controls: mRNA expression (ET-1/glyceraldehyde-3-phosphate-dehydrogenase ratio), 0.54 +/- 0.18 versus 0.30 +/- 0.08; protein concentration, 7.36 +/- 2.21 pg/mg versus 3.93 +/- 0.40 pg/mg, respectively; both p < 0.01. Immunofluorescence signal of ET-1 protein was predominantly localized to endothelia of gastric mucosal and submucosal vessels. In PHT stomachs FR 139317 significantly reduced mucosal necrosis (percentage of necrotic area, from 24.9 +/- 5.9% to 10.8 +/- 4.0%; p < 0.01), although it had no effect on sham-operated controls. CONCLUSIONS: Portal hypertension activates the ET-1 gene with overexpression of ET-1 protein in the gastric mucosa. Protection of PHT gastric mucosa by ETA receptor antagonist against damage indicates that overexpression of ET-1 plays an important role in increased susceptibility of PHT gastric mucosa to injury.

Clinical implications of portal hemodynamics after small-diameter portacaval H graft.

To assess the role of portal hemodynamics in the development of postshunt encephalopathy, we studied 19 patients after small-diameter portacaval H grafting (SD-PCHG). We used contrast studies as well as technetium-labeled macroaggregated albumin injected into the portal vein to assess direction of portal flow. We then quantitated the mesenteric fraction of flow perfusing the liver by injecting macroaggregated albumin into a peripheral mesenteric vein tributary. We found that none of seven patients with prograde flow by both scintigraphy and angiography developed postoperative encephalopathy, but the incidence was 58% in the remaining patients (p = 0.02). The fraction of mesenteric flow perfusing the liver after SD-PCHG was 12% +/- 4%, but this did not significantly correlate with encephalopathy rates. We conclude that after SD-PCHG, prograde portal flow minimizes encephalopathy rates. Although encephalopathy occurs in patients with predominantly reversed flow, a subgroup of patients with reversed flow remain without symptoms. The absolute fraction of mesenteric flow perfusing the liver has less influence on encephalopathy rates than has direction of portal flow. This study identifies a complex relationship between portal hemodynamics and encephalopathy.

Portal vein thrombosis associated with cirrhosis: clinical importance.

A group of 86 cirrhotics undergoing therapeutic variceal decompressive procedures were studied. Patients with portal vein thrombosis (PVT) comprised 21% of the group and more frequently had uncontrollable hemorrhage at an earlier stage of liver disease. Mortality in the 18 cirrhotics with PVT was higher (56%), mostly as a result of rebleeding. Despite adjustments for stage of liver disease and type of operation, in no cirrhotic with PVT did postshunt encephalopathy develop, compared with 32% incidence in patients without PVT (P less than .05). It is concluded that patients with cirrhosis and PVT represent a different subpopulation of cirrhotics. Once adequate variceal decompression has been achieved, their prognosis should be superior to cirrhotics without PVT because their hepatic hemodynamics are unaffected by total shunting, hence precluding further impairment of liver function as a result of acute reduction of hepatic blood flow.

Surgical treatment of enteric 'bud' fistulas in contaminated wounds. A riskless extraperitoneal method using split-thickness skin grafts.

We describe methods and results of a local extraperitoneal method of repairing enterocutaneous "bud" fistulas in abdominal-wall defects. The method is performed with local anesthesia and involves an extraperitoneal closure with skin-graft coverage. Of the nine fistulas so treated, five healed. No patient's postoperative course was set back by the repairs that failed since the method precludes intraperitoneal entrance. Two of three high-output fistulas were successfully repaired with the extraperitoneal method, reversing an otherwise stormy clinical course. We conclude that for epithelialized enterocutaneous fistulas, little is lost if our method of repair fails and much is gained if it is successful in these critically ill patients.

Salt and water balance before and after peritoneojugular and portacaval shunts.

Renal salt and water handling were studied in a patient who had refractory ascites that were due to alcoholic liver disease. The patient underwent a peritoneojugular shunt that was followed eight weeks later by a portacaval shunt. Ingestion of a 10-mEq sodium diet and albumin infusion had no effect on sodium balance but increased free water clearance before and after the operative procedures. After peritoneojugular shunt, free water clearance and sodium excretion increased. After portacaval shunt, free water clearance decreased and sodium retention occurred.

Partial portacaval shunt for variceal hemorrhage: longitudinal analysis of effectiveness.

OBJECTIVE: To determine rates of survival, long-term patency, and recurrent variceal hemorrhage among patients with alcoholic cirrhosis treated by partial portacaval shunt. DESIGN: Single-institution cohort follow-up study of 72 consecutive patients who underwent small-diameter portacaval H-graft shunt with collateral ablation during a 10-year period (1981 through 1990). Subjects were enrolled and followed up for up to 15 years. Shunt patency was assessed by portography and/or ultrasonography. We performed 7-year Kaplan-Meier analyses of survival (in 65 patients in Child classes A and B), shunt patency, and absence of variceal bleeding. SETTING: Tertiary academic referral center of the US Department of Veterans Affairs. PATIENTS: Patients with alcoholic cirrhosis were considered for operation after at least 1 proven episode of variceal hemorrhage. Patients with portal vein thrombosis were excluded; patients in Child class C underwent operation only for compelling indications. Of the 72 who underwent partial shunting, 38 were in Child class A, 27 were in class B, and 7 were in class C. INTERVENTIONS: Partial portacaval shunt (6-, 8- or 10-mm polytetrafluoroethylene H-graft with collateral ablation) and serial follow-up. MAIN OUTCOME MEASURES: Study end points were death, recurrent variceal hemorrhage, and unavailability for follow-up. Other measures included graft patency and nonvariceal rebleeding. RESULTS: Cumulative probability of 7-year patency for grafts at risk was 95%. The 7-year probability for absence of variceal bleeding in patients at risk was 92%. In 65 patients in Child classes A and B, operative mortality was 7.7% and the cumulative probability of 7-year survival was 54%. CONCLUSION: For variceal bleeding associated with alcoholic cirrhosis, the small-diameter polytetrafluoroethylene portacaval H-graft with collateral ablation affords durable patency and protection against variceal rebleeding.

Computer-derived equations for predicting survival postoperatively. Their usefulness and limitations.

We used multivariate analysis to determine whether survival following perforations of the gastrointestinal tract could be accurately predicted from preoperative data. Of 12 variables tested, four were found to have predictive value. These were age, pulmonary disease, preoperative shock, and the attending surgeon. When these four variables were employed in a logistic regression equation on 42 patients, it correctly predicted which 21 patients died before leaving the hospital. To produce an equation useful for other hospitals, we recalculated it without the attending surgeon variable. Again, the equation was used to predict survival. The correlation of predicted vs observed outcome remained high, and, using a 2 x 2 chi 2 test, the correlation was significant. We then cross validated the three-variable model on data from a second hospital. The model accurately predicted the new data equally well. We believe that predictive models can identify risk factors in a variety of patient populations and can determine who is likely to benefit from specific treatment modalities.

Description of a continent jejunal gastrostomy.

A new technique of permanent gastrostomy interposes a vascularized jejunal conduit with an intussuscepted valve between the stomach and abdominal wall, creating a "continent jejunal gastrostomy". In a series of dogs undergoing gastrostomy, the continent jejunal gastrostomy functioned well, with no leakage, and was superior to gastric tube gastrostomy in this regard. Furthermore, there was no evidence of marginal ulceration. A patient who underwent continent jejunal gastrostomy has an excellent one-month result.

Antacid talcid activates in gastric mucosa genes encoding for EGF and its receptor. The molecular basis for its ulcer healing action.

In previous studies [Gut 35 (1994) 896-904], we demonstrated that antacid talcid (TAL) accelerates gastric ulcer healing and provides better quality of mucosal restoration within the scar than the omeprazole (OME). However, the mechanisms of TAL-induced ulcer healing are not clear. Since growth factors promote cell proliferation, re-epithelization, angiogenesis and ulcer healing, we studied whether TAL and/or OME affect expression of epidermal growth factor (EGF) and its receptors (EGF-R) in both normal and ulcerated gastric mucosae. Rats with or without acetic acid-induced gastric ulcers (n = 64) received i.g. twice daily 1 mL of either: A) placebo (PLA); B) TAL 100 mg; or C) OME 50 mg x kg(-1) for 14 d. Studies of gastric specimens: 1) ulcer size; 2) quantitative histology; 3) expression of EGF mRNAs was determined by RT/PCR; 4) gastric sections were immunostained with antibodies against EGF and its receptors. In non-ulcerated gastric mucosa of placebo or omeprazole treated group, EGF expression was minimal, while EGF-R was localized to few cells in the mucosal proliferative zone. Gastric ulceration triggered overexpression of EGF and its receptor in epithelial cells of the ulcer margin and scar. In ulcerated gastric mucosa TAL treatment significantly enhanced (versus PLA and omeprazole) expression of EGF and EGF-R. OME treatment reduced expression of EGF in ulcerated mucosa by 55 +/- 2% (P < 0.01). It is concluded that: 1) treatment with TAL activates genes for EGF and its receptor in normal and ulcerated gastric mucosae; 2) since EGF promotes growth of epithelial cells and their proliferation and migration, the above actions of TAL provide the mechanism for its ulcer healing action and improved (versus OME) quality of mucosal restoration.

Role of protein kinase a pathway in epidermal growth factor-induced liver cell repair.

To gain a better understanding of the mechanisms that control the repair process in the injured liver, the actions of epidermal growth factor (EGF) and protein kinase A (PKA) were studied. Normal rat liver cells (clone 9) were grown to confluence. Standardized excisional wounds were made with a razor blade. The extent of hepatocyte migration into the wound was measured and determined at specific time intervals using a computerized digital analyzing system. Immunostaining of F-actin was performed with a fluorescein-labeled phalloidin. EGF significantly stimulated liver cell migration, whereas specific EGF-neutralizing antibody inhibited the EGF-induced migration. Agents that activate PKA at different stages of the PKA activation pathway, including 3-isobutyl-1-methylxanthine (IBMX), forskolin, and cholera toxin, inhibited EGF-induced migration. EGF triggered formation of actin stress fibers. PKA-activating agents inhibited actin stress fiber formation and stretching of cells at the wound margin. The following conclusions were drawn: (1) In excisional wounds of hepatocyte monolayers, both EGF and PKA exert action on actin microfilaments, which are stretched by EGF and inhibited by PKA; (2) the enhanced repair of wounded hepatocyte monolayers by EGF is blocked by activation of the PKA pathway at various levels; and (3) these actions of EGF and PKA indicate their important regulatory roles in controlling the rate of hepatocyte migration and restitution following the creation of excisional wounds.