Portal systemic collateral development: Is it a trophic adaptation mechanism to hepatic deprivation?

See article in J. Gastroenterol. Hepatol. 2005; 20: 1867–1872. Portal hypertension is usually secondary to intra or extrahepatic obstruction of portal flow, although it usually develops in patients with chronic liver diseases, which are one of the most common causes of death in adults worldwide.1 It is also the main complication of cirrhosis.2 Portal hypertension is characterized by a pathological increase in portal pressure, associated with splenomegaly, and by the development of portal-systemic collateral circulation, which diverts portal flow to the systemic circulation bypassing the liver.3–5 The most serious clinical consequences are closely related to the development of portal-systemic collateral vessels.2 The importance of these collateral vessels has encouraged extensive research into the mechanisms underlying their formation, as well as the pathophysiological and pharmacological properties of the collateral venous bed in portal hypertensive states. This research includes an interesting and elegant study by Chan et al. published in this journal, which uses the in situ perfusion model of collateral circulation to investigate the vascular response of norepinephrine (NE) and acetylcholine (ACh) in portal-systemic collaterals, in an experimental model of cirrhosis by common bile duct ligation.6 This study shows that NE has a vasoconstrictor effect on the portal-systemic collateral vessels of rat with biliary cirrhosis, and that ACh produces vasodilatation of the NE-preconstricted collaterals.6 The authors also demonstrate that these vascular effects are, in turn, modulated by nitric oxide and prostaglandins.6 These results agree with those obtained by Mosca et al.7 in portal hypertensive-rats, in which both NE and 5-hydroxytryptamine (5-HT) increased the perfusion pressure through the portal-systemic collaterals. It has been shown that ACh, in turn, dilated the NE-preconstricted collaterals.7 Also, the competitive inhibitor of nitric oxide synthase, N-omega-nitro-L-arginine (L-NNA), has been demonstrated to increase collateral resistance and prevent ACh-induced dilatation of the collaterals.7 The study developed by Chan et al. continues on from lengthy research in this field. These authors have previously demonstrated the vasoconstrictor effect of vasopressin, endothelin-1 (ET-1) and somatostatin on the collateral vessels in partial portal vein ligated-rats.8–12 They also reported the vasoconstrictor action of ET-1 on the collaterals of bile duct-ligated cirrhotic rats.13 The common triggering factor for both cirrhotic and non-cirrhotic portal hypertension is an increased resistance to portal venous flow. As a result of high portal pressure, portal-systemic collateral circulation is thought to be a natural response to provide new outlets for congested blood in the splanchnic venous system and to increase the splanchnic vascular bed.14 There are many factors involved in the development and dilation of the portal-systemic collaterals, as well as the regulation of their blood flow.15 Development of the portal-systemic collateral circulation characteristic of portal hypertension occurs by a combination of nitric oxide (NO)-dependent dilation of preexisting vessels and growth factor-mediated angiogenesis and neovascularization.16,17 Portal hypertension has been shown to be a very potent angiogenic stimulus,14,18 although other factors may also play a role in the development of portal-systemic collateral vessels.15,18–21 NO, carbon monoxide (CO) and tumor necrosis factor (TNF)-α levels, which increase early on in portal hypertension22–24 and are all mediators of angiogenesis,25–27 could be involved in the development of collateral circulation. Therefore, NO, CO and TNF-α induce the expression of growth factors, such as basic fibroblastic growth factor (bFGF) and vascular endothelial growth factor (VEGF), which are the principal mediator molecules of angiogenesis.27–29 It has been recently demonstrated that the development of portal-systemic collateral vessels in portal hypertension is a VEGF-dependent angiogenic process. This was shown by a significant and marked inhibition in their formation in both partial portal vein ligated-rats and mice treated with anti-VEGF receptor-2 monoclonal antibodies.1 It has been hypothesized that portal-systemic collaterals may mimic an arteriovenous fistula, in which the high pressure portal blood connects with the lower pressure systemic venous circulation.14 Although these collaterals decompress the portal circulation, it is especially interesting that the portal pressure remains elevated, despite the presence of high-grade portal-systemic shunting.30 Hyperdynamic circulation with splanchnic hyperemia and portal hypertension, rather than increasing perfusion of the liver, results in an increased shunting of portal blood into the systemic circulation via portal-systemic collateral anastomoses.14,31 Therefore, progressive collateral shunting exacerbates the reduction in portal blood flow to the liver, creating a true ‘stealing’ phenomenon.31 The portal-systemic collaterals provide a route to decompress the hypertensive portal system. Despite this, the vascular resistance of the collateral bed is still greater than the resistance of the liver,32 and portal pressure does not decrease. The development of portal-systemic collaterals is precisely the central pathophysiological event that leads to complications responsible for increased morbidity and mortality in patients with portal hypertension, particularly variceal hemorrhage.32,33 Increased resistance in hepatic and portal-systemic collateral circulation can, at least partially, be modified by pharmacological agents. Changes in the flow and resistance of portal-systemic collaterals by vasoactive agents result in changes in portal pressure. Therefore, normalization of elevated portal pressure can be achieved by correcting the elevated portal resistance in the portal collateral circulation (the ‘backward’ hypothesis).4,30 This is the basis of several experimental studies on the effects of different vasoactive substances and drugs on blood flow, the resistance to portal-systemic collaterals,6–13,34–37 and their applications in the medical treatment of patients with cirrhosis.32,38,39 If establishing the significance of alterations associated with portal hypertension could help to elucidate the mechanisms involved in its production and maintenance, we could, then, speculate on the hypothetical purpose of the splanchnic response in portal hypertension. We have therefore proposed an inflammatory etiopathogenic hypothesis to explain the complications of portal hypertension. The increased pressure in the portal venous system could represent the initial stimulus that triggers the inflammatory response.40 In this scenario, the evolution of portal hypertension as an inflammatory response would be comprised of three phenotypes with a trophic significance, as proposed previously for the post-traumatic inflammatory response.41 In this response, the nervous phenotype of ischemia-reperfusion causes interstitial and cellular edema, which is an ancestral mechanism to feed the cells by diffusion. During the immune phenotype, the tissues are infiltrated by inflammatory cells and bacteria, which could become mediators in providing nutrition to the injured tissues. Finally, in the endocrine phenotype, nutrition mediated by the blood capillaries is established.41 Similarly, the three aforementioned phenotypes are expressed in prehepatic portal hypertension by partial portal vein ligation. The nervous phenotype is characterized by sympathetic hyperactivity42 and splanchnic arterial vasodilatation, associated with systemic and splanchnic hyperdynamic circulation that causes hypoxia of the mucosa.21,43–45 The second immune phenotype is characterized by an infiltration by mast cells of the intestinal mucosa and submucosa, and of the superior mesenteric lymphatic complex.46,47 Also, inflammatory mediators are released in the gut–liver axis,48 and splenomegaly occurs, with phagocytosis in germinal centers of the splenic pulp.49 Finally, the endocrine phenotype would be predominated by angiogenesis with persistence of portal-hepatic and portal-systemic collateral circulation,50 and long-term splanchnic remodeling.46,50–52 If the successive phases that constitute the post-traumatic inflammatory response have trophic properties,53 the main goal of the inflammation would not be to defend but, instead, to nourish the injured tissue.53 A similar trophic meaning could be attributed to development of the portal-systemic circulation in portal hypertension. In this condition, particularly in cirrhosis, the organism suffers a deprivation of liver function, the trophic organ in the body par excellence. This could be one of the factors that triggers the development of portal-systemic vessels. This could constitute a return to a circulation characteristic of fetal life, in which the portal blood bypasses the liver.54 The trophic significance of the portal-systemic vessels lies in their ability to systemically distribute nutritive factors of intestinal source, bypassing the liver. In this case, these nutrients, which have not been ‘precooked’ in the liver, could induce new trophic mechanisms in peripheral organs or tissues. The development of portal-systemic collaterals could therefore have a trophic meaning and represent an adaptation mechanism of the body in response to liver deprivation. This could explain the chronic persistence of these collateral vessels and why, although these provide a way to decompress the splanchnic circulation, the portal pressure does not decrease. Curiously, preservation of the portal pressure in terminolateral portocaval shunted-rats has beneficial metabolic and nutritional effects, although the underlying mechanisms for these are not explained by the authors.55 In summary, the study by Chan et al.6 represents a very important contribution to knowledge of the pathophysiology of portal hypertension, and sheds light on the possible significance of this frequent and severe condition.