Search

Portal and hepatic circulation can now be safely accessed using endoscopic ultrasound (EUS). EUS-guided needle access of the portal vein is performed clinically at select tertiary centers for measurement of portal pressure gradients in patients with chronic liver disease and sampling of portal venous thrombus to diagnose malignancy. We propose that this novel clinical technique can be applied in research studies to allow blood collection from and profiling of portal and hepatic circulation. In this technical report, we present and highlight the technical aspects, feasibility, and safety of EUS: guided portal venous blood collection. As a proof of the concept and the utility of this technique in metabolic research and biomarker assessment and discovery, we present a pilot metabolite profiling study of portal venous blood in a small cohort of patients with cirrhosis and a comparison with a group without cirrhosis. Despite the very small diameter of the endoscopic needle used for the blood collection, the portal samples have the same quality as those collected from systemic circulation, and they can be used for the same downstream applications. Finally, we propose an analytical workflow to screen for promising metabolites that could qualify for further studies to determine their utility as sensitive, early candidate biomarkers of hepatic fibrosis, portal shunt, and hypertension. We hope that this report could stimulate and facilitate the widespread use of EUS-guided techniques for the profiling of portal circulation, which could potentially open a new field of scientific inquiry. NEW & NOTEWORTHY The technical aspects, feasibility, and safety of endoscopic ultrasound (EUS)-guided needle access for portal venous blood collection are presented in this technical report. Despite the very small diameter of the endoscopic needle, portal blood samples have the same quality as those collected from systemic circulation. As a proof of the concept and the utility of this technique in metabolic research and biomarker assessment and discovery, we present a pilot metabolite profiling study of portal venous blood in a small cohort of patients with cirrhosis and a comparison with a group without cirrhosis.

The influence of circulating urotensin II (UII) on liver disease and portal hypertension is unknown. We aimed to evaluate whether UII executes a pathogenetic role in the development of hepatic fibrosis and portal hypertension. UII was administered by continuous infusion over 4 wk in 20 healthy rats divided into three treatment groups, controls (saline, n = 7), low dose (UII, 1 nmol x kg(-1) x h(-1), n = 8), and high dose (UII, 3 nmol x kg(-1) x h(-1), n = 5). Hemodynamic parameters and morphometric quantification of fibrosis were assessed, and profibrotic cytokines and fibrosis markers were assayed in hepatic tissue. UII induced a significant dose-dependent increase in portal venous pressure (5.8 +/- 0.4, 6.4 +/- 0.3, and 7.6 +/- 0.7, respectively, P = 0.03). High-dose UII infusion was associated with an increase in hepatic transcript for transforming growth factor-beta (P < 0.05) and platelet-derived growth factor-beta (P = 0.06). Liver tissue hydroxyproline was elevated in the high-dose group (P < 0.05). No systemic hemodynamic alterations were noted. We concluded that UII infusion elevates portal pressure and induces hepatic fibrosis in normal rats. This response may be mediated via induction of fibrogenic cytokines. These findings have pathophysiological implications in human liver disease where increased plasma UII levels have been observed.

Portal hypertension (PHT) is a complication of liver cirrhosis and directly increases mortality and morbidity by increasing the propensity of venous hemorrhage. There are two main underlying causations for PHT, increased hepatic resistance and systemic hyperdynamic circulation. Both are related to localized aberrations in endothelial nitric oxide synthase (eNOS) function and NO biosynthesis. This study investigates the importance of eNOS and systemic hyperdynamic-associated hyperemia to better understand the pathophysiology of PHT. Wild-type and eNOS(-/-) mice were given the hepatotoxin CCl(4) for 4-12 wk. Hepatic fibrosis was determined histologically following collagen staining. Portal venous pressure, hepatic resistance, and hyperemia were determined by measuring splenic pulp pressure (SPP), hepatic portal-venous perfusion pressure (HPVPP), abdominal aortic flow (Qao), and portal venous flow (Qpv). Hepatic fibrosis developed equally in wild-type and eNOS(-/-) CCl(4)-exposed mice. SPP, Qao, and Qpv increased rapidly in wild-type CCl(4)-exposed mice, but HPVPP did not. In eNOS(-/-) CCl(4) mice, Qao was not increased, SPP was partially increased, and HPVPP and Qpv were increased nonsignificantly. We concluded that the systemic hyperemia component of hyperdynamic circulation is eNOS dependent and precedes increased changes in hepatic resistance. Alternative mechanisms, possibly involving cyclooxygenase, may contribute. eNOS maintains normal hepatic resistance following CCl(4)-induced fibrosis. Consequently, increased portal pressure following chronic CCl(4) exposure is linked to hyperdynamic circulation in wild-type mice and increased hepatic resistance in eNOS(-/-) mice.

A growing body of evidence indicates that heme degradation products may counteract the deleterious consequences of hypoxia and/or ischemia-reperfusion injury. Because heme oxygenase (HO)-1 induction after adverse circulatory conditions is known to be protective, and because females in the proestrus cycle (with high estrogen) have better hepatic function and less hepatic damage than males after trauma-hemorrhage, we hypothesized that estrogen administration in males after trauma-hemorrhage will upregulate HO activity and protect the organs against dysfunction and injury. To test this hypothesis, male Sprague-Dawley rats underwent 5-cm laparotomy and hemorrhagic shock (35-40 mmHg for 93 +/- 2 min), followed by resuscitation with four times the shed blood volume in the form of Ringer lactate. 17beta-Estradiol and/or the specific HO enzyme inhibitor chromium mesoporphyrin (CrMP) were administered at the end of resuscitation, and the animals were killed 24 h thereafter. Trauma-hemorrhage reduced cardiac output, myocardial contractility, and serum albumin levels. Portal pressure and serum alanine aminotransferase levels were markedly increased under those conditions. These parameters were significantly improved in the 17beta-estradiol-treated rats. Estradiol treatment also induced increased HO-1 mRNA expression, HO-1 protein levels, and HO enzymatic activity in cardiac and hepatic tissue compared with vehicle-treated trauma-hemorrhage rats. Administration of the HO inhibitor CrMP prevented the estradiol-induced attenuation of shock-induced organ dysfunction and damage. Thus the salutary effects of estradiol administration on organ function after trauma-hemorrhage are mediated in part via upregulation of HO-1 expression and activity.

We examined the role of thromboxane A2 (TXA2) in LPS-induced hyperresponsiveness of hepatic portal circulation to endothelins (ETs) and whether Kupffer cells are the primary source of TXA2 release in response to ET-1 in endotoxemia. After 6 h of LPS (1 mg/kg body wt ip) or saline (control), liver was isolated and perfused with recirculating Krebs-Henseleit bicarbonate buffer at a constant flow rate (100 ml.min(-1).kg body wt(-1)). ET-1 (10 pmol/min) was infused for 10 min. Portal pressure (PP) was continuously monitored during perfusion. Perfusate was sampled for enzyme immunoassay of thromboxane B2 (TXB2; the stable metabolite of TXA2) and lactate dehydrogenase (LDH) assay. ET-1 infusion resulted in a significantly greater increase of PP in the LPS group than in controls. Both TXA2 synthase inhibitor furegrelate (Fureg) and TXA2 receptor antagonist SQ-29548 (SQ) substantially blocked enhanced increase of PP in the LPS group (4.9 +/- 0.4 vs. 3.6 +/- 0.5 vs. 2.6 +/- 0.6 mmHg for LPS alone, LPS + Fureg, and LPS + SQ, respectively; P < 0.05) while having no significant effect on controls. GdCl3 for inhibition of Kupffer cells had similar effects (4.9 +/- 0.4 mmHg vs. 2.9 +/- 0.4 mmHg for LPS alone and GdCl3 + LPS, respectively; P < 0.05). In addition, the attenuated PP after ET-1 was found concomitantly with significantly decreased releases of TXB2 and LDH in LPS rats treated with Fureg, SQ, and GdCl3 (886.6 +/- 73.4 vs. 110.8 +/- 0.8 vs. 114.8 +/- 54.7 vs. 135.2 +/- 45.2 pg/ml, respectively; P < 0.05). After 6 h of LPS, Kupffer cells in isolated cell preparations released a significant amount of TXA2 in response to ET-1. These results clearly indicate that hyperresponsiveness of hepatic portal circulation to ET-1 in endotoxemia is mediated at least in part by TXA2-induced receptor activation, and Kupffer cells are likely the primary source of increased TXA2 release.

To test whether hemorrhagic shock and resuscitation (HSR) alters the vascular responsiveness of the portohepatic circulation to endothelins (ETs), we studied the macro- and microcirculatory effects of the preferential ET(A) receptor agonist ET-1 and of the selective ET(B) receptor agonist sarafotoxin 6c (S6c) after 1 h of hemorrhagic hypotension and 5 h of volume resuscitation in the isolated perfused rat liver ex vivo using portal pressure-flow relationships and epifluorescence microscopy. Although HSR did not cause major disturbances of hepatic perfusion per se, the response to ET-1 (0.5 x 10(-9) M) was enhanced, leading to greater increases in portal driving pressure, total portal resistance, and zero-flow pressures and more pronounced decreases in portal flow, sinusoidal diameters, and hepatic oxygen delivery compared with time-matched sham shock controls. In sharp contrast, the constrictive response to S6c (0.25 x 10(-9) M) remained unchanged. Thus HSR primes the portohepatic circulation for the vasoconstrictive effects of ET-1 but does not alter the effects of the ET(B) receptor agonist S6c. The enhanced sinusoidal response may contribute to the subsequent development of hepatic microcirculatory failure after secondary insults that are associated with increased generation of ET-1.

Although hepatic blood flow increases significantly during early sepsis [as produced by cecal ligation and puncture (CLP)], it is not known whether this is due to the increase in portal or hepatic arterial blood flows. To study this, rats were subjected to CLP, after which they and sham-operated rats received either 3 or 6 ml normal saline/100 g body wt subcutaneously (i.e., all rats received crystalloid therapy). Blood flow in various organs was determined by using a radioactive microsphere technique at 5 and 20 h after CLP or sham operation. Portal blood flow was calculated as the sum of blood flows to the spleen, pancreas, gastrointestinal tract, and mesentery. Total hepatic blood flow was the sum of portal blood flow and hepatic arterial blood flow. A significant increase in portal blood flow and in total hepatic blood flow was observed at 5 h after CLP (i.e., early sepsis), and this was not altered by doubling the volume of crystalloid resuscitation after the induction of sepsis. In contrast, hepatic arterial blood flow during early sepsis was found to be similar to control; however, it was significantly reduced in late sepsis (i.e., 20 h after CLP). Cardiac output was significantly higher than the control in early sepsis. However, even in late sepsis, cardiac output and total hepatic blood flow were not significantly different from controls. These results indicate that the increased total hepatic blood flow during early hyperdynamic sepsis is solely due to the increased portal blood flow.

To clarify the roles that portocollateral resistance ("backward-flow" theory) and portal flow ("forward-flow" theory) play in maintaining chronic portal hypertension, we studied, in a rat model with prehepatic portal hypertension, the hemodynamic changes that occur when portocollateral resistance is reduced and high portal venous inflow is maintained. In 30 portal-hypertensive rats the constriction around the portal vein was removed 4 days after induction of portal hypertension, 30 rats were used as portal vein-constricted controls, and 30 additional rats were subjected to a sham operation. The removal of the ligature constricting the portal vein was followed by an immediate decrease in portal pressure (from 16.3 +/- 0.8 to 9.6 +/- 0.8 mmHg, P less than 0.001). Two days after the ligature removal, hyperdynamic circulation was still evident and was characterized by a decreased splanchnic arteriolar resistance and an increased portal venous inflow. The coexistence of high portal venous inflow and normal portal pressure indicates that high portal venous inflow per se is not sufficient to produce an increase in portal pressure when it faces a low-resistance vascular bed. We conclude that portal hypertension is induced by the interaction of an abnormally high portal venous inflow and high resistance offered to the flow by the portocollateral vessels. Neither the forward-flow theory nor the backward-flow theory can be applied solely to explain the increased portal pressure.

Experimental cirrhosis was produced in dogs by the sporadic feeding of dimethylnitrosamine for the purpose of studying the temporal relationships between urinary sodium retention, plasma volume expansion, and ascites formation. Sodium retention started about 16 days following the onset of cirrhosis and preceded ascites formation by about 10 days. Plasma volume increased by 9% (P less than 0.05) within 3-4 days of sodium retention and expanded further as ascites accumulated. Splanchnic plasma volume was greater by 161 ml in 10 cirrhotic dogs with ascites than in 14 normal dogs. Nonsplanchnic volume was greater by 96 ml (P less than 0.05). Thus, the "effective" as well as the splanchnic component of the vascular space was expanded. Paracentesis did not cause the re-formation of ascites in five dogs as long as dietary salt was denied. Refeeding permitted reaccumulation of ascites and further plasma volume expansion. Renal perfusion remained constant as dogs became progressively cirrhotic. We conclude that ascites formation depends on the prior retention of urinary sodium, and occurs as an "overflow" phenomenon. A contracted effective plasma volume does not appear to be necessary for continuing sodium retention.

To determine the effect of pentobarbital sodium anesthesia on the rat with portal hypertension due to portal vein stenosis, four groups of rats were studied. Cardiac output and regional blood flow were measured by radioactive microspheres in anesthetized and conscious sham-operated and portal-hypertensive rats. Anesthesia markedly decreased cardiac output in both sham-operated (109.7 +/- 4.6 vs. 77.8 +/- 1.4 ml/min, P less than 0.001) and portal-hypertensive rats (130.1 +/- 7.6 vs. 93.8 +/- 5.3 ml/min, P less than 0.01). In spite of this diminution in cardiac output, pentobarbital did not significantly change absolute blood flow values of splanchnic organs in either group. However, the fractions of cardiac output perfusing the splanchnic organs were significantly increased by pentobarbital in both groups because of the decrease in cardiac output: sham operated, anesthetized, 22.86 +/- 1.19% vs. conscious, 14.83 +/- 1.02%, P less than 0.001; and portal hypertensive, anesthetized, 26.67 +/- 0.71% vs. conscious, 19.07 +/- 1.44%, P less than 0.001. The hyperdynamic circulation of the portal vein-stenosed rat compared with the sham-operated rat continued to manifest itself with significantly increased portal pressure, cardiac output, and splanchnic blood flow, whether the animal was anesthetized or awake. We conclude that, despite marked hemodynamic changes induced by pentobarbital, the rat with portal vein stenosis remains a useful experimental model of portal hypertension.

Systemic and splanchnic hemodynamics of the chronic bile duct-ligated rat were characterized by radioactive microspheres. Conscious and pentobarbital sodium-anesthetized, bile duct-ligated and sham-operated rats had cardiac output and regional organ blood flows determined. The conscious bile duct-ligated rat compared with the sham-operated showed a hyperdynamic circulation with an increased cardiac output (153.3 +/- 9.8 vs. 112.6 +/- 6.0 ml/min, P less than 0.005) and portal tributary blood flow (21.32 +/- 1.43 vs. 12.79 +/- 1.47 ml/min, P less than 0.005). Pentobarbital sodium anesthesia induced marked hemodynamic changes in both sham-operated and bile duct-ligated rats. The latter group was especially sensitive to its effects; thus, comparison of cardiac output and portal tributary blood flow between anesthetized bile duct-ligated and sham-operated rats showed no significant differences. We conclude that the rat with cirrhosis due to chronic bile duct ligation is an excellent model for hemodynamic investigations but should be studied in the conscious state, since pentobarbital sodium anesthesia eliminates the hyperdynamic circulation.

The possible contribution of hepatocellular damage and portal-systemic shunting to hyperinsulinism in cirrhosis was studied in 23 cirrhotics, 8 of whom had a surgical portacaval shunt, and 16 controls by measuring insulin and the connecting peptide (C-peptide) concentrations in simultaneous samples of peripheral arterial and hepatic venous blood. The fractional hepatic insulin extraction (0.48 +/- 0.06, mean +/- SE) was normal in cirrhosis. The hepatic insulin elimination rate was directly related to arterial insulin levels (r = 0.91, P less than 0.001) even at very high circulating levels. Extrahepatic insulin metabolism was measured across the kidney and lower limb. There were no significant differences between cirrhotics and control subjects in relation to renal (0.25 +/- 0.05 vs. 0.23 +/- 0.04) and lower limb insulin extraction (0.14 +/- 0.07 vs. 0.19 +/- 0.04). While in the control group hepatic venous insulin (0.143 +/- 0.018 pmol/ml) markedly exceeded the peripheral insulin concentration (0.083 +/- 0.009 pmol/ml, P less than 0.01), the contrary was found in cirrhotics with end-to-side portacaval shunt in whom all the pancreatic venous effluent is shunted to the systemic circulation (hepatic venous insulin, 0.130 +/- 0.028 pmol/ml; peripheral, 0.234 +/- 0.037 pmol/ml; P less than 0.01). Portal-hypertensive cirrhotics without a surgical portacaval shunt also had hepatic venous insulin levels (0.132 +/- 0.029 pmol/ml) below peripheral arterial insulin concentrations (0.205 +/- 0.041 pmol/ml, P less than 0.01). The study suggests that hyperinsulinism in cirrhosis is not the result of an intrinsic defect of hepatic insulin metabolism but of the spontaneous shunting of portal blood to the systemic circulation.

Two dissimilar hemodynamic hypotheses, the "backward flow" theory and the "forward flow" theory, have been advanced to define splanchnic hemodynamics in portal hypertension. An animal model with portal hypertension and high-grade portal-systemic shunting, the portal vein-stenotic rat, was studied to determine whether a hemodynamic picture compatible with either theory would develop. Splanchnic and systemic hemodynamics and portal-systemic shunting were measured by radioactive microsphere techniques. The portal-hypertensive rats (portal pressure, 12.8 +/- 0.5 vs. 8.3 +/- 0.4 mmHg) with greater than 95% portal-systemic shunting had a 60% increase in portal venous inflow (23.46 +/- 2.54 vs. 14.97 +/- 1.61 ml/min; P less than 0.01) with a concomitant 50% decrease in splanchnic arteriolar resistance (3.86 +/- 0.43 vs. 7.60 +/- 0.80 dyn . s . cm-5 . 10(5); P less than 0.001) compared with control rats. Cardiac index (391 +/- 17 vs. 250 +/- 20 ml . min-1 . kg-1) was elevated 50% (P less than 0.001), and total peripheral resistance (7.1 +/- 0.4 vs. 11.7 +/- 0.8 dyn . s . cm-5 . 10(4)) was decreased 60% (P less than 0.001). The resistance to portal blood flow in portal vein-stenotic rats (4.77 +/- 0.57 dyn . s . cm-5 . 10(4)) was similar to the resistance to portal blood flow in control rats (4.82 +/- 0.43 dyn . s . cm-5 . 10(4)), indicating that the hyperdynamic portal venous inflow, not resistance, provided the main impetus for maintaining the elevated portal venous pressure. The splanchnic hemodynamic observations directly support the forward flow theory of portal hypertension. The relation between splanchnic arteriolar resistance and total peripheral resistance (r = 0.67; P less than 0.01) indicated that the systemic hemodynamic parameters were secondarily altered by the splanchnic hemodynamic changes. This animal model of chronic portal hypertension gave evidence for a generalized splanchnic arteriolar vasodilation occurring in the presence of high-grade portal-systemic shunting.

In rats with portal hypertension induced by partial ligation of the portal vein, we have recently demonstrated an increased portal venous inflow that becomes an important factor in the maintenance of portal hypertension. The sequence of events that leads into this circulatory disarray is unknown. We evaluated chronologically the chain of hemodynamic changes that occurred after portal hypertension was induced by partial ligation of the portal vein. In this model it is possible to follow, from the initiation of the portal-hypertensive state, the interaction between blood flow and resistance in the portal system as well as the relation between the development of portal-systemic shunting and the elevated portal venous inflow. The study was performed in 45 portal-hypertensive rats and in 29 sham-operated rats. Blood flow and portal-systemic shunting were measured by radioactive microsphere techniques. The constriction of the portal vein was immediately followed by a resistance-induced portal hypertension characterized by increased portal resistance (9.78 +/- 0.89 vs. 4.18 +/- 0.71 dyn X s X cm-5 X 10(4), mean +/- SE, P less than 0.01), increased portal pressure (17.7 +/- 0.9 vs. 9.5 +/- 0.6 mmHg, P less than 0.001), and decreased portal venous inflow (3.93 +/- 0.26 vs. 6.82 +/- 0.49 ml X min-1 X 100 g body wt-1, P less than 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)

High serum lactate may not reflect the severity of endotoxin shock: the lactate load could even be formed immediately after the endotoxin challenge. During the first 30 min after endotoxin injection (Escherichia coli; 1.5 mg/kg iv) into anesthetized dogs (4 mg.kg-1.h-1 etomidate, n = 19) we studied arterial lactate concentration; contributions of portal and splanchnic (n = 6), renal and pulmonary (n = 7), and femoral (n = 6) vascular beds to the early lactate rise; and regional O2 extraction and blood flow (microspheres). In control dogs (n = 5, no endotoxin), we found no significant hemodynamic and biochemical changes. Endotoxin caused an immediate decrease in blood pressure, cardiac output, and organ perfusion, followed by recovery after approximately 5 min to approximately 75% of preshock values at t = 30 min (except for renal blood flow, which remained low). Arterial lactate concentration started to increase almost immediately after endotoxin and increased rapidly until t = 15 min (to 300%) and then leveled off, but in spite of the hemodynamic recovery it remained elevated. A major part of the early increase in lactate concentration can be explained by splanchnic lactate production. The total splanchnic bed released more lactate than the portal bed, indicating that the liver produces lactate. We conclude that the lactate concentration later in canine endotoxin shock depends on events that occur during early shock in which the liver may play a crucial role.