Randomised trial of octreotide for long term management of cirrhosis after variceal haemorrhage

Patients and methods

Patient evaluation— Entry was confined to patients with cirrhosis confirmed by biopsy after a first variceal bleed. Variceal bleeding was controlled in all patients with two sessions of injection sclerotherapy, and the patients were readmitted to hospital for assessment of their liver disease 3 weeks after their initial presentation. Liver biopsy was performed, and the degree of hepatic dysfunction graded according to the criteria of Child and Turcotte.16 Portal pressure was measured indirectly by measuring wedged hepatic venous pressure through the femoral vein and subtracting vena cava pressure to determine the hepatic venous pressure gradient. Hepatocyte function was measured by breath testing with aminopyrine labelled with carbon-14,17 and the activity of the hepatic and splenic reticuloendothelial system was measured by single photon emission computed tomography using sulphur colloid labelled with technetium-99.18

Patient recruitment—As octreotide had not previously been used long term in cirrhotic patients after variceal haemorrhage, there were no data to inform a power calculation. Thus pragmatic considerations determined that the sample size was the number of patients who would complete the trial within three years. Thirty two patients satisfying the entry requirements were randomly allocated treatment after complete assessment of their liver disease and 6 weeks after the onset of their first variceal bleed. A randomisation code with equal numbers of the alternative treatments was generated, using random number tables which were unknown to the clinicians allocating treatment to patients. Randomisation was done with sequentially numbered, sealed envelopes using the previously determined code to either the combination of octreotide (Sandoz Pharmaceuticals, Camberley) and injection sclerotherapy or injection sclerotherapy alone. The study was approved at the outset by the Royal Liverpool University Hospital Ethics Committee, and all patients gave informed consent.

Injection sclerotherapywas carried out using ethanolamine oleate and an intravariceal technique (aliquots of 2 ml, maximum of 20 ml) every 3 weeks until all oesophageal varices were obliterated.

Octreotide—Patients randomly allocated combination treatment received 50 μg octreotide subcutaneously twice daily for 6 months. Patients were taught to self administer between breakfast and lunch, with the second injection 12 hours later before retiring. The use of octreotide vials could not be checked as they had to be kept refrigerated. Placebo injections and blinding were not used as it was anticipated that these would have reduced compliance.

Recurrent bleeding—Patients who had recurrent variceal bleeds were admitted and treated by injection sclerotherapy as above. In some patients endoscopy showed haemorrhage from oesophageal ulcers, which were then treated with somatostatin 250 μg/h.

End of trial—Six months after entry into the trial all surviving patients were readmitted for reassessment as in the initial evaluation (without liver biopsy).

Statistical analysis—Non-parametric methods were used throughout the statistical analysis.19 Categorical measurements were tabulated for each group and compared using a two tailed Fisher's exact test or χ² analysis. Results of continuous measurements were summarised using medians, interquartile ranges, and 95% confidence intervals. Comparisons between the two patient groups were performed using Mann-Whitney U tests, survival was plotted using the Kaplan-Meier method, and the log rank test was used to assess differences in outcome between the two groups.

Characteristics of patients

The two groups of patients were well matched in age, sex, severity of liver dysfunction, cause of liver disease, and hepatic venous pressure gradient (Table 1).

Rebleeding

During the 6 months of the trial, variceal haemorrhage occurred in one of the 16 patients receiving combined treatment, whereas seven patients receiving only injection sclerotherapy rebled from varices, two of them on two occasions. This difference was significant (P=0.037, Fisher's exact test). Two patients rebled from oesophageal ulcers, one in each group.

Survival

During the 6 months of the trial 5 patients died, all in the group receiving sclerotherapy alone; 4 died after recurrent variceal bleeding. At the end of the trial patients receiving the combined treatment showed a significant improvement in survival compared with those receiving sclerotherapy alone (P<0.02, log rank test), and this was maintained for 12 further months, but no longer. This improvement was maintained for about 12 months after the end of treatment (1).

Kaplan-Meier survival plot of patients treated with octreotide and injection sclerotherapy or sclerotherapy alone. Numbers denote surviving patients

• Download figure
• Open in new tab
• Download powerpoint

Other measures

All measures at 6 months include only patients alive at 6 months, other than Child's grade. Child's grade is given at 6 months, except for the patients who died before the trial ended: their grade is given at entry and at the readmission when they died (Table 2).

The hepatic venous pressure gradient in patients receiving combined treatment was reduced at 6 months compared with pretreatment values (median change −6.0, interquartile range −10 to −4.75 mm Hg). In contrast, hepatic venous pressure gradient in patients treated by sclerotherapy alone increased slightly at 6 months (median change 1.5, interquartile range −0.25 to 3.25 mm Hg). The difference between treatments was highly significant (P=0.0002) and in favour of a reduction in the hepatic venous pressure gradient after combined treatment in the range −13 to −6 mm Hg (95% confidence interval of the difference between treatments).

Child's grade at 6 months in patients receiving combined treatment improved significantly when compared with the grade in patients receiving sclerotherapy alone (P=0.0002, Table 2. Similarly, combined treatment was associated with a significant improvement in liver function assessed by plasma concentrations of bilirubin, albumin, and alanine aminotransferase, as well as by hepatocyte metabolism of ¹⁴C-aminopyrine (Table 3). In contrast, the reduction in alkaline phosphatase concentrations was less with combined treatment (Table 3). There was no significant difference in the median change of γ-glutamyl transferase before and after treatment between the two groups of patients (Table 3).

The activity of the hepatic and splenic reticuloendothelial system, as assessed by the uptake of ⁹⁹Tc-sulphur colloid, was similar in the two groups before treatment. At 6 months there was a significant improvement in activity with combined treatment compared with sclerotherapy alone (Table 4).

Treatment complications

The complications of octreotide treatment were minor and consisted of transient diarrhoea at the start of treatment (3 patients) and pain at the site of injection (5 patients). One patient with a history of cholelithiasis developed biliary colic while taking octreotide. Two patients bled from oesophageal ulcers after injection sclerotherapy, one from each group.

Long term octreotide and liver function

All patients surviving the 6 months of the trial showed an improvement in liver function. However, the improvements in Child's grade and plasma concentrations of albumin, bilirubin, and alanine aminotransferase were significantly greater with combined treatment. Moreover, hepatocyte function was significantly improved with this treatment but not with sclerotherapy alone. In contrast, the circulating concentrations of alkaline phosphatase were lower with sclerotherapy alone, presumably reflecting the cholestatic effect of octreotide. Nevertheless, the results of this study suggest that octreotide improves overall hepatic function in patients with cirrhotic portal hypertension. The mechanism whereby octreotide stimulates liver function remains unclear but could be due to direct or indirect effects on hepatocytes.22

In this trial octreotide stimulated hepatic reticuloendothelial activity, confirming previous observations in experimental animals.15 The phagocytic activity of the reticuloendothelial system is impaired in cirrhosis, mainly due to depression of the hepatic system, which contributes up to 80% of the total activity of the reticuloendothelial system.23 Patients with cirrhosis who have a measurably impaired hepatic reticuloendothelial system have a shorter survival than those who have normal activity, in spite of similar clinical and biochemical abnormalities.23 The reduced survival may be because clearance of endotoxin is reduced, which could contribute to liver necrosis. The clinically significant improvement in the activity of the hepatic reticuloendothelial system with combined treatment may have contributed to the improved liver function discussed.

Potential of long term somatostatin analogues

At the outset of this study there were few data on the use of somatostatin analogues in the long term management of cirrhotic portal hypertension. We therefore incorporated a detailed evaluation of the nature and severity of liver disease at the beginning and end of the trial. Furthermore, as the focus of the study was the long term prevention of variceal rebleeding, we sought patients whose condition was initially stable and who would comply with trial treatment. Nevertheless, given the positive results of this trial and the appeal of permanent pharmacological modification of portal hypertension, the administration of long term adjuvant treatment immediately after the end of the acute phase of variceal haemorrhage might be appropriate. In addition, the use of long term adjuvant treatment could be widened to include patients who have had recurrent variceal haemorrhage.

Our trial was undertaken before data became available on the relative benefits of variceal ligation over sclerotherapy.4 5 However, variceal ligation is similar to sclerotherapy in that the underlying portal hypertension remains, so long term somatostatin analogues may be similarly beneficial after variceal ligation.

The results of this controlled trial suggest that long term octreotide, used as an adjuvant to endoscopic sclerotherapy, reduces the incidence of recurrent variceal bleeding and improves survival in cirrhotic patients. Further randomised trials are needed to confirm these effects and to compare somatostatin analogues with other drugs in the long term management of portal hypertension. The recent development of long acting slow release preparations of somatostatin analogues24 make comparison with ß blockers particularly desirable.