- Top of page
- MATERIALS AND METHODS
Portal hypertension is the main complication of liver cirrhosis and chronic hepatitis. This syndrome is responsible for the frequently fatal complication of massive gastrointestinal bleeding from ruptured oesophageal varices. Portal pressure may increase because of an increase in either portal blood flow, or intrahepatic vascular resistance, or a combination of both. It is well-established that, although the primary factor leading to portal hypertension is an increased resistance to portal blood flow, an increased portal venous inflow contributes equally and becomes especially important in advanced stages.1
Many pharmacological attempts to treat portal hypertension have aimed to correct the increased portal blood inflow with splanchnic vasoconstrictors. Somatostatin and its cyclic octapeptide analogue, octreotide, have been shown to produce an immediate reduction in portal pressure and collateral blood flow.2–7 Their immediate effect, together with the relatively high safety of both agents has provided the rational for their use in the treatment of acute variceal bleeding. The beneficial effects of somatostatin and octreotide are largely attributed to splanchnic vasoconstriction.8–10 However, the exact mechanism of action remains unknown. Although a direct effect on splanchnic vasculature cannot be excluded, indirect methods of action are under investigation. A reduction in circulating vasodilators such as glucagon has been suggested as a possible method of action of these drugs.9, 10 Recent studies have focused on somatostatin-induced changes in the equilibrium of paracrine vasoactive factors such as NO and endothelin.11–13
The present study was designed to compare the efficacy of both drugs in reducing sinusoidal pressure as measured by the wedged hepatic venous pressure, as well as to clarify whether the induced effects are accompanied by changes in the local concentration of NO.
- Top of page
- MATERIALS AND METHODS
Somatostatin has been proposed as a treatment for variceal or severe portal gastropathy bleeding because of its ability to decrease portal pressure without significant adverse systemic effects.2, 13, 16 The effect of somatostatin is possibly mediated through its ability to cause a selective splanchnic vasoconstriction and reduction of splanchnic blood flow.2, 5, 17 Although initial reports were conflicting, further haemodynamic studies proved the efficacy of somatostatin in modifying portal pressure and supported the rational for the use of the drug in acute variceal bleeding.2, 4, 5 In a double-blind, placebo-controlled haemodynamic study, Cirera et al. demonstrated that a single bolus injection of 250 μg somatostatin produced a rapid and intense fall of 52% and 45% in portal pressure gradient and azygos blood flow, respectively.4 In our study, a similar treatment caused a maximum of 34 ± 7% reduction in sinusoidal pressure in the sub-group of cirrhotic patients. This is consistent with the findings of Cirera et al., where the recorded reduction in hepatic venous pressure gradient was attributed to a significant decline in wedged hepatic venous pressure accompanied by a marked increase in free hepatic vein pressure.4
The rapid effect of bolus injection of somatostatin on hepatic circulation, confirmed by the present study, is the most pronounced effect ever observed with a pharmacological agent used for the treatment of portal hypertension and may be of great clinical relevance. According to the European ABOVE study conducted among 205 patients with cirrhosis and upper gastrointestinal bleeding, the early administration of repeated boluses of 250 μg somatostatin, followed by a continuous infusion, resulted in a reduction of cases with active variceal bleeding at the time of emergency endoscopy. As a consequence, sclerotheraphy was easier to perform, the amount of blood transfused was lower and death or rescue therapy was less frequent.18 Part of the beneficial effect of the above treatment, particularly the decrease in frequency of active bleeding during emergency endoscopy, may be attributed to the acute reduction in portal pressure gradient and collateral flow observed with the bolus administration of somatostatin.
The short half life of somatostatin has led to the development of synthetic octapeptide analogues with a longer biological half life, such as octreotide.5 Initial reports have shown that this agent is capable of reducing portal pressure in animals as well as in cirrhotic patients.5, 19 Subsequent studies, however, produced controversial results.6, 7 A placebo-controlled study conducted by Nevens et al., comparing the effect of octreotide with that of terlipressin on the variceal pressure, concluded that the bolus injection of 50 μg octreotide produced no change in variceal pressure.20 Furthermore, Moller et al., in a study with a small number of patients (n=13), demonstrated that a bolus injection of 100 μg octreotide followed by a continuous infusion had no effect on either the wedged hepatic and free hepatic venous pressures or on the hepatic venous pressure gradient.21
In the present study, the bolus injection of 125 μg octreotide showed, in contrast to somatostatin, a statistically significant maximum decrease of 23 ± 9% of the wedged hepatic venous pressure only at 10 min in the total group of patients. However, when cirrhotic and chronic hepatitis patients were separately analysed, no change in sinusoidal pressure could be found. Moller et al. also observed a transient and insignificant decrease in hepatic venous pressure gradient at 10 min post-injection, suggesting that a rapid desensitization to the effects of octreotide occurs in patients with cirrhosis.21 We conclude that bolus injection of octreotide is not as effective as somatostatin in acutely reducing wedged hepatic venous pressure as the observed effects are transient and insignificant. The clinical relevance of this finding remains to be established. The international trial coordinated by Burroughs et al., comparing the early administration of continuous infusion of octreotide in 193 patients with bleeding oesophageal varices before sclerotheraphy failed to show any benefit compared to placebo.22 An initial bolus was not administered in their study. It seems, therefore, that octreotide is less effective than somatostatin in acutely modifying hepatic haemodynamics.
It has been reported that both somatostatin and octreotide mediate their effects by splanchnic vasoconstriction.23 However, the exact mechanism of action remains unknown. The speed of onset of the haemodynamic effects may suggest either a direct effect on the vasculature or a modification of paracrine vasoactive factors such as nitric oxide.17, 24 Nitric oxide (NO) seems to play an important role in splanchnic vasodilation in cirrhotic patients.25 The finding of almost complete normalization of splanchnic haemodynamics, with a reduction in portal venous inflow and an increase in splanchnic vascular resistance in portal hypertensive rats by the acute administration of a non-specific NO synthesis inhibitor clearly demonstrates the importance of NO.26, 27 There are several reports indicating that somatostatin can mediate NO production both in vitro and in vivo.28, 29In situ studies on isolated saphenous arteries and veins from different species demonstrated that somatostatin may directly cause arterial vasodilation and venous vasoconstriction by NO-dependent mechanisms.12, 30 Considering the above regional differences in NO-dependent vascular responses to somatostatin, we investigated whether a bolus infusion of somatostatin modifies the local concentration of NO.
Previous investigations have demonstrated increased levels of nitric oxide in the hepatic vein of patients with cirrhosis compared to normal controls.31 We found no difference in baseline local concentrations of NO between patients with chronic hepatitis and cirrhosis. No normal controls were included in our study. An additional reason for this discrepancy may be that the majority of the patients included in our study had a low Child–Pugh score, while NO overproduction is probably a finding related to later stages of cirrhosis.15 Furthermore, we failed to demonstrate any significant changes in local levels of NO after the administration of somatostatin or octreotide, nor did we find any correlation between NO and wedged hepatic venous pressure values, either before or after injection of both drugs. Our findings are consistent with previously published data.
In an in vitro study in portal hypertensive rats, Sieber et al. observed that octreotide did not modify the vascular effects of N_-nitro-L-arginine, an antagonist of NO synthase. The authors suggested that octreotide does not exert a direct effect on NO formation.13 Sabat et al. also failed to demonstrate changes in serum levels of NO2–/NO3– after the subcutaneous administration of octreotide in cirrhotic patients.9 Furthermore, Chatila et al., using venous occlusion plethysmography, observed that an intra-arterial dose of octreotide produced an immediate vasoconstriction to the infused arm. This direct effect was not accompanied by changes in local concentration of NO products. Moreover, octreotide had no inhibitory effect on NO-induced vasodilation by metacholine.8 Finally, venoconstriction of the human saphenous vein by somatostatin has been shown to occur by an NO-independent mechanism because it was present in the absence of a functional endothelium. These reports and the findings of the present study suggest that the acute vasoactive effect of somatostatin in the sinusoidal pressure is independent of NO.
To conclude, we have shown that a single bolus injection of somatostatin can cause a rapid, significant and relatively protracted decrease in wedged hepatic venous pressure in cirrhotic patients. Octreotide is not as effective as somatostatin in acutely reducing wedged hepatic venous pressure. This somatostatin effect is probably mediated by an NO-independent mechanism.