& 1990 S. Karger A G , Basel ÜO12-2823/90/0457-006052.75/0
Digestion 1990:45(suppl I ):60-65
A Study of Octreotide in Oesophageal Varices Ruth McKee Surgical Unit. Royal Infirmary, Glasgow. UK
Key Words. Octreotide • Sandostatin • Oesophageal varices
Patients with portal hypertension who de velop oesophageal varices are difficult man agement problems. When bleeding from the varices occurs, rapid control of bleeding is necessary to prevent further deterioration in liver function. Although injection sclerother apy can both control the acute bleed and commence definitive treatment of the var ices. this skill may not always be imme diately available and. in any case, the patient will probably require a period of resuscita tion. An immediate means of control of va riceal bleeding is therefore necessary.
At present two principal methods of im mediate control of variceal bleeding are used. Oesophageal tamponade is effective [1] but is unpleasant for the patient and has been associated with respiratory complica tions. Vasopressin, a generalized vasocon strictor, also constricts the coronary and ce rebral vessels and serious cardiac side effects have been reported [2]. Naturally occurring somatostatin has been thought to act as a selective presplanchnic vasoconstrictor and therefore has been advocated for use in vari ceal bleeding. In two studies [3, 4] compar
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 9:27:19 PM
Abstract. A comparison of octreotide infusion (25 yig/h) and placebo in 16 stable cirrhotic patients revealed a 30% reduction in transhepatic venous gradient between 0 and 60 min in the octreotide group without an effect on systemic haemodynamics. In a separate trial, 40 patients with active varicea! bleeding were randomized to octreotide infusion (25 pg/h for 48 h) or oesophageal tamponade. The 2 treatments gave comparable control of variceal bleed ing. Tolerance of treatment was significantly better in the octreotide group. In summary, octreotide infusion is simple to administer, has few side effects, and may be of use in the immediate control of oesophageal bleeding.
ing somatostatin with vasopressin in variceal bleeding, naturally occurring somatostatin was found to be at least as effective as vaso pressin with a reduced incidence of side effects. However, naturally occurring soma tostatin has a plasma half-life of a few min utes and is unstable in solution. This means that practical difficulties may occur during its administration. Octreotide (Sandostatin, SMS 201-995: Sandoz Pharmaceuticals) has a half-life of 45-60 min in plasma and is sta ble in solution. If it has similar haemody namic effects to naturally occurring soma tostatin. this new drug may be more useful in variceal bleeding. In the University Department of Surgery, Glasgow, a 2-year study of octreotide in por tal hypertension has been performed. Ini tially the effects of octreotide on haemody namics in stable cirrhotic patients were stud ied and thereafter a controlled clinical trial examined its efficacy in active bleeding from oesophageal varices.
Effect of Octreotide on Portal and Systemic Haemodynamics
In our first study we examined the effects of octreotide infusion on both portal and sys temic haemodynamics in stable cirrhotic pa tients [5], Sixteen patients were included, of whom 9 were given octreotide infusion and 7 acted as controls. Patients and Methods All 16 patients had experienced previous, endoscopically proven bleeding from oesophageal varices. All patients had been hacmodynamically stable for at least 7 days prior to study and each patient gave writ ten. informed consent. Patients with cardiac or renal disease were excluded, as were patients who were tak
61
ing vasoactive drugs and patients who had abnormal coagulation screens. The patients underwent right heart catheteriza tion using a Swan-Ganz thermodilution catheter un der radiological control. Following a period of at least 30 min bed rest, initial measurements of cardiac index and systemic vascular resistance were made with the catheter tip in the pulmonary artery. An American-Edwards cardiac output computer was used to compute these values. Pulse rate and arterial blood pressure were also measured at this time. Fol lowing these systemic haemodynamic measurements, the catheter tip was manipulated into one of the hepatic veins under radiological control. The pres ence of a venous pressure wave was confirmed and then measurements of free and wedged hepatic ve nous pressure were made. Transhepatic venous gra dient was then calculated by subtracting the free from the wedged hepatic venous pressure. After the initial measurements had been taken. 9 patients were given an intravenous infusion of 25 pg/h of octreotide in 125 ml normal saline. The 7 patients who acted as controls were not given the drug. At 60, 120 and 180 min from the beginning of the infusion period, the above measurements were repeated. Blood samples were taken at 15-min inter vals for plasma levels of octreotide.
Results Table I shows the clinical details of the 16 patients included in this study. It is worth noting that their modified Child's grading [6] is rather better than our usual spectrum of patients because of the exclusion of pa tients with abnormal coagulation. Figure 1 shows the principal Finding with regard to systemic haemodynamics through out the study period in the 2 groups of patients. There was no significant change in pulse rate or arterial blood pressure in either group, and no difference between the groups. A significant reduction in cardiac index be tween 0 and 60 min was seen in both groups, with no significant differences between the groups. This was associated with a signifi cant increase in systemic vascular resistance.
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 9:27:19 PM
A Study of Octreotide in Oesophageal Varices
62
McKee
Fig. 1. Change in cardiac index with time in the 2 groups of stable cirrhotic patients in the study of the haemodynamic effects of octreotide (Sandostatin. SMS 201-995). The octreotide group was given an intra venous infusion of 25 pg octreotide between 0 and 60 min. Values are mean ± standard error.
Number Age. years Pugh’s modification of Child's A B C Aetiology of liver disease Alcoholic cirrhosis Primary biliary cirrhosis Chronic active hepatitis Sclerosing cholangitis
Control group
Octreotide group
7 57 ±3.0 grade 4 2 i
9 52 ±2.0
6 1 0 0
4 3 2 7 0 1 1
Age is given as mean ± standard error.
Therefore this fall in cardiac index could not be attributed to the octreotide infusion. These changes in cardiac index have been reported by at least 1 other group [7], Pa tients with hypodynamie circulation due to congestive cardiac failure increase their car-
diac index with bed rest and it may be that the decrease in cardiac index seen in liver disease patients is due to the effect of bed rest in patients with an abnormally hyperdy namic circulation. The effect of octreotide infusion on portal haemodynamics in these patients is shown in figure 2. A 30% reduction in transhepatic venous gradient between 0 and 60 min was seen in the octreotide group, while no such reduction was seen in the control group. This difference was statistically significant (re peated analysis of variance with Greenhouse-Geisscr correction: p = 0.032). Fig ure 3 shows the plasma concentrations of octreotide in patients given the drug infu sion. consistent with a plasma half-life of 45-60 min. Conclusion This initial study therefore demonstrated that, in stable cirrhotic patients, octreotide infusion causes a significant reduction in portal pressure without effect on systemic haemodynamics.
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 9:27:19 PM
Table I. Clinical details for 16 cirrhotic patients who took part in the study of the haemodynamic effects of octreotide
A Study of Octreotide in Oesophageal Varices
63
Fig. 2. Change in transhepatic venous gradient with time in the 2 groups of stable cirrhotic patients in the study of the haemodynamic ef fects of octreotide (SMS 201-995). The octreotide group was given an intravenous infusion of 25 pg oc treotide between 0 and 60 min. Val ues are mean ± standard error.
Fig. 3. Plasma concentrations of octreotide in 8 stable cirrhotic pa tients given an intravenous infusion of 25 gg octreotide (SMS 201-995) between 0 and 60 min. in the study of the haemodynamic effects of oc treotide. Values are mean ± stan dard error.
Having established that an intravenous infusion of octreotide causes a 30% reduc tion in portal pressure in cirrhotic patients, a controlled clinical trial of its efficacy in ac tive variceal bleeding was performed. Our previous best means of immediate control of
variceal bleeding was oesophageal tampo nade [8] and therefore octreotide infusion was compared with this treatment. Patients and Methods For entry to the trial, patients were required to have endoscopicallv proven active bleeding from oe sophageal varices. The patients were then random ized to initial treatment with either octreotide infu
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 9:27:19 PM
Controlled Trial of Octreotide in Variceal Bleeding
64
Results Forty variceal bleeds were entered into this study, 20 being randomized to each treatment group. The group allocated to oesophageal tamponade contained a slightly greater number of modified Child’s C grade patients, but this did not reach statistical sig nificance (table 1). Control of bleeding over the first 4 h after entry to the trial was achieved in 19 bleeds in the tamponade group and in 18 bleeds in the octreotide group. Complete control of bleed ing over the 48 h trial period was achieved in 14 of the tamponade group and in 10 of the octreotide group. Two patients in the tam ponade group died as a result of severe hepa torenal failure before the end of the 48-hour period. There was no significant difference in the incidence of chest infection between the 2 groups. Chest complications were more closely associated with rebleeding than with the method of treatment. No symptomatic side effects were seen in the octreotide infu sion group, whereas, in 17 of the 20 bleeds treated with oesophageal tamponade, the pa tient complained of discomfort due to the tube. Fifteen patients survived the admis sion in the tamponade group compared with all the patients in the octreotide group.
Conclusion In this clinical trial octreotide infusion gave comparable control of variceal bleeding to oesophageal tamponade. Tolerance of treatment was significantly better in the oc treotide group.
References 1 Panes J. Teres J. Bosch .1. Rodes J: Efficacy of balloon tamponade in treatment of bleeding gas tric and oesophageal varices. Dig Dis Sci 1988:33: 454-459. 2 Slotnik IL, Teigland JD: Cardiac accidents follow ing vasopressin injection (‘Pitressin'). JAMA 1951:146:1126-1129. 3 Kravetz D. Bosch J. Teres J. et al: Comparison of somatostatin and vasopressin infusions in treat ment of acute variceal hemorrhage. Hepatology 1984:4:442-446. 4 Jenkins SA, Baxter JN. Corbett W. et al: A pro spective randomised controlled clinical trial com paring somatostatin and vasopressin in control ling acute variceal haemorrhage. Br Med .1 I985:i: 275-278. 5 Pringle SD. McKee RE. Garden DJ. Lorimer AR. Carter DC: The effect of SMS 201-995 on portal and systemic haemodynamics in cirrhosis. Ali ment Pharmacol Ther 1988:2:451-459. 6 Pugh RMH. Murray-Lyon IM. Dawson JL, Pie troni MC. Williams R: Transection of the oesoph agus for bleeding oesophageal varices. Br .1 Surg 1973:60:646-649. 7 Lebrec D. Hillon P. Munoz C. et al: The effect of propranolol on portal hypertension in patients with cirrhosis: a haemodynamic study. Hepato logy 1982:2:523-527. 8 Haddock G. Garden J. McKee RF. Anderson JR. Carter DC: Oesophageal tamponade in the man agement of acute variceal hemorrhage. Dig Dis Sci 1989:34:913-918.
Ruth McKee Surgical Unit Royal Infirmary Glasgow G4 QSF (UK)
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 9:27:19 PM
sion or oesophageal tamponade. Patients assigned to treatment with octreotide received an intravenous infusion of 25 pg/h of octreotide over 48 h. Patients in the oesophageal tamponade group had a 4-lumen Minnesota modification of the Sengstaken-Blakemore tube passed. Both gastric and oesophageal bal loons were left inflated for the first 24 h. but from 24 to 48 h only the gastric balloon was in Hated. In both groups fibreoptic endoscopy and injection sclerother apy were performed at 48 h. Bleeding control, compli cations. and symptomatic side effects were recorded. Evidence of chest infection was looked for particu larly carefully.
McKee
A Study of Octreotide in Oesophageal Varices
65
Discussion them with a series of pharmacologically complicated manoeuvres. Dr. Harris: The incidence of side effects from tam ponade seems low'er than that usually reported in the literature. Am I correct? Dr. McKee: The reports in the literature vary very widely. An institution where tamponade is the stan dard treatment for bleeding from oesophageal varices and is performed often, and where the patients are looked after in one ward by one set of nursing staff, can give very good results. I think our results fall into that category. The reports where oesophageal tamponade is said to be very dangerous come from units where it is used as a last-ditch measure. We do see problems in patients transferred from other hospitals. Dr Jenkins (Liverpool): How was active bleeding assessed? We have admitted perhaps 200 patients over the last 3 or 4 years, and about 30% of them were actually bleeding. My second point relates to the dose. Our dose-response study on wedged hepatic venous pressure suggests that you might have a rather low dose at 25 gg/h octreotide. Dr McKee: We used that dose in our acute study because we had shown a difference in wedged hepatic venous pressure in our stable cirrhotics with that dose. Certainly I think it might be worth giving a bolus of 25 pg octreotide before starting infusion and it may indeed be that a higher dose might be more effective. We defined active variceal bleeding as bleeding seen down the endoscope which would nor mally have caused us to put down a Sengstaken tube. Thus, we did not randomize anyone who was simply oozing a very small amount from their varices. All patients required blood transfusion: all had evidence of shock with increased pulse rate and reduced ar terial blood pressure.
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 9:27:19 PM
Dr. Far/liing: The absolute difference in mortality between the octreotide group and the tamponade group was only 5. Were these patients all Child’s group C? Dr. McKee: Yes. they were. Dr. Farthing: I think this raises a key issue when performing small studies, that it is essential to stratify the initial randomization for severity of disease. You probably should not overplay that statistical differ ence, because if the patients had been stratified by severity at the outset you would not have a differ ence. Dr. McKee: Yes. I think that is right. The trouble with performing this sort of study is the small number of patients available. It would have been much more difficult to organize if we have tried to stratify before randomization, particularly if you remember the practicalities of assessing patients to a Child’s grading when they have just arrived with a variceal bleed. Their liver function tests are not going to be imme diately available. But obviously it would be useful. Dr. Farthing: I would agree. It is interesting some times to run a clinical grading - that is. your initial clinical assessment - against the subsequent Child's grading. I think you would not have done badly if perhaps you had done that. Dr. Chalmers (Leeds): I think you were a little unfair about vasopressin. Many of the disadvantages you quoted can be partially or totally overcome by intravenous nitrate therapy. Have you any experience of that? Dr. McKee: We use vasopressin very little in the Glasgow Royal Infirmary for the reasons that I have outlined. Given that patients have an acute assault on their haemodynamic system, having lost a consider able amount of blood, we feel a little wary of treating
Digestion 1990:45(suppl I ):60-65
A Study of Octreotide in Oesophageal Varices Ruth McKee Surgical Unit. Royal Infirmary, Glasgow. UK
Key Words. Octreotide • Sandostatin • Oesophageal varices
Patients with portal hypertension who de velop oesophageal varices are difficult man agement problems. When bleeding from the varices occurs, rapid control of bleeding is necessary to prevent further deterioration in liver function. Although injection sclerother apy can both control the acute bleed and commence definitive treatment of the var ices. this skill may not always be imme diately available and. in any case, the patient will probably require a period of resuscita tion. An immediate means of control of va riceal bleeding is therefore necessary.
At present two principal methods of im mediate control of variceal bleeding are used. Oesophageal tamponade is effective [1] but is unpleasant for the patient and has been associated with respiratory complica tions. Vasopressin, a generalized vasocon strictor, also constricts the coronary and ce rebral vessels and serious cardiac side effects have been reported [2]. Naturally occurring somatostatin has been thought to act as a selective presplanchnic vasoconstrictor and therefore has been advocated for use in vari ceal bleeding. In two studies [3, 4] compar
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 9:27:19 PM
Abstract. A comparison of octreotide infusion (25 yig/h) and placebo in 16 stable cirrhotic patients revealed a 30% reduction in transhepatic venous gradient between 0 and 60 min in the octreotide group without an effect on systemic haemodynamics. In a separate trial, 40 patients with active varicea! bleeding were randomized to octreotide infusion (25 pg/h for 48 h) or oesophageal tamponade. The 2 treatments gave comparable control of variceal bleed ing. Tolerance of treatment was significantly better in the octreotide group. In summary, octreotide infusion is simple to administer, has few side effects, and may be of use in the immediate control of oesophageal bleeding.
ing somatostatin with vasopressin in variceal bleeding, naturally occurring somatostatin was found to be at least as effective as vaso pressin with a reduced incidence of side effects. However, naturally occurring soma tostatin has a plasma half-life of a few min utes and is unstable in solution. This means that practical difficulties may occur during its administration. Octreotide (Sandostatin, SMS 201-995: Sandoz Pharmaceuticals) has a half-life of 45-60 min in plasma and is sta ble in solution. If it has similar haemody namic effects to naturally occurring soma tostatin. this new drug may be more useful in variceal bleeding. In the University Department of Surgery, Glasgow, a 2-year study of octreotide in por tal hypertension has been performed. Ini tially the effects of octreotide on haemody namics in stable cirrhotic patients were stud ied and thereafter a controlled clinical trial examined its efficacy in active bleeding from oesophageal varices.
Effect of Octreotide on Portal and Systemic Haemodynamics
In our first study we examined the effects of octreotide infusion on both portal and sys temic haemodynamics in stable cirrhotic pa tients [5], Sixteen patients were included, of whom 9 were given octreotide infusion and 7 acted as controls. Patients and Methods All 16 patients had experienced previous, endoscopically proven bleeding from oesophageal varices. All patients had been hacmodynamically stable for at least 7 days prior to study and each patient gave writ ten. informed consent. Patients with cardiac or renal disease were excluded, as were patients who were tak
61
ing vasoactive drugs and patients who had abnormal coagulation screens. The patients underwent right heart catheteriza tion using a Swan-Ganz thermodilution catheter un der radiological control. Following a period of at least 30 min bed rest, initial measurements of cardiac index and systemic vascular resistance were made with the catheter tip in the pulmonary artery. An American-Edwards cardiac output computer was used to compute these values. Pulse rate and arterial blood pressure were also measured at this time. Fol lowing these systemic haemodynamic measurements, the catheter tip was manipulated into one of the hepatic veins under radiological control. The pres ence of a venous pressure wave was confirmed and then measurements of free and wedged hepatic ve nous pressure were made. Transhepatic venous gra dient was then calculated by subtracting the free from the wedged hepatic venous pressure. After the initial measurements had been taken. 9 patients were given an intravenous infusion of 25 pg/h of octreotide in 125 ml normal saline. The 7 patients who acted as controls were not given the drug. At 60, 120 and 180 min from the beginning of the infusion period, the above measurements were repeated. Blood samples were taken at 15-min inter vals for plasma levels of octreotide.
Results Table I shows the clinical details of the 16 patients included in this study. It is worth noting that their modified Child's grading [6] is rather better than our usual spectrum of patients because of the exclusion of pa tients with abnormal coagulation. Figure 1 shows the principal Finding with regard to systemic haemodynamics through out the study period in the 2 groups of patients. There was no significant change in pulse rate or arterial blood pressure in either group, and no difference between the groups. A significant reduction in cardiac index be tween 0 and 60 min was seen in both groups, with no significant differences between the groups. This was associated with a signifi cant increase in systemic vascular resistance.
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 9:27:19 PM
A Study of Octreotide in Oesophageal Varices
62
McKee
Fig. 1. Change in cardiac index with time in the 2 groups of stable cirrhotic patients in the study of the haemodynamic effects of octreotide (Sandostatin. SMS 201-995). The octreotide group was given an intra venous infusion of 25 pg octreotide between 0 and 60 min. Values are mean ± standard error.
Number Age. years Pugh’s modification of Child's A B C Aetiology of liver disease Alcoholic cirrhosis Primary biliary cirrhosis Chronic active hepatitis Sclerosing cholangitis
Control group
Octreotide group
7 57 ±3.0 grade 4 2 i
9 52 ±2.0
6 1 0 0
4 3 2 7 0 1 1
Age is given as mean ± standard error.
Therefore this fall in cardiac index could not be attributed to the octreotide infusion. These changes in cardiac index have been reported by at least 1 other group [7], Pa tients with hypodynamie circulation due to congestive cardiac failure increase their car-
diac index with bed rest and it may be that the decrease in cardiac index seen in liver disease patients is due to the effect of bed rest in patients with an abnormally hyperdy namic circulation. The effect of octreotide infusion on portal haemodynamics in these patients is shown in figure 2. A 30% reduction in transhepatic venous gradient between 0 and 60 min was seen in the octreotide group, while no such reduction was seen in the control group. This difference was statistically significant (re peated analysis of variance with Greenhouse-Geisscr correction: p = 0.032). Fig ure 3 shows the plasma concentrations of octreotide in patients given the drug infu sion. consistent with a plasma half-life of 45-60 min. Conclusion This initial study therefore demonstrated that, in stable cirrhotic patients, octreotide infusion causes a significant reduction in portal pressure without effect on systemic haemodynamics.
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 9:27:19 PM
Table I. Clinical details for 16 cirrhotic patients who took part in the study of the haemodynamic effects of octreotide
A Study of Octreotide in Oesophageal Varices
63
Fig. 2. Change in transhepatic venous gradient with time in the 2 groups of stable cirrhotic patients in the study of the haemodynamic ef fects of octreotide (SMS 201-995). The octreotide group was given an intravenous infusion of 25 pg oc treotide between 0 and 60 min. Val ues are mean ± standard error.
Fig. 3. Plasma concentrations of octreotide in 8 stable cirrhotic pa tients given an intravenous infusion of 25 gg octreotide (SMS 201-995) between 0 and 60 min. in the study of the haemodynamic effects of oc treotide. Values are mean ± stan dard error.
Having established that an intravenous infusion of octreotide causes a 30% reduc tion in portal pressure in cirrhotic patients, a controlled clinical trial of its efficacy in ac tive variceal bleeding was performed. Our previous best means of immediate control of
variceal bleeding was oesophageal tampo nade [8] and therefore octreotide infusion was compared with this treatment. Patients and Methods For entry to the trial, patients were required to have endoscopicallv proven active bleeding from oe sophageal varices. The patients were then random ized to initial treatment with either octreotide infu
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 9:27:19 PM
Controlled Trial of Octreotide in Variceal Bleeding
64
Results Forty variceal bleeds were entered into this study, 20 being randomized to each treatment group. The group allocated to oesophageal tamponade contained a slightly greater number of modified Child’s C grade patients, but this did not reach statistical sig nificance (table 1). Control of bleeding over the first 4 h after entry to the trial was achieved in 19 bleeds in the tamponade group and in 18 bleeds in the octreotide group. Complete control of bleed ing over the 48 h trial period was achieved in 14 of the tamponade group and in 10 of the octreotide group. Two patients in the tam ponade group died as a result of severe hepa torenal failure before the end of the 48-hour period. There was no significant difference in the incidence of chest infection between the 2 groups. Chest complications were more closely associated with rebleeding than with the method of treatment. No symptomatic side effects were seen in the octreotide infu sion group, whereas, in 17 of the 20 bleeds treated with oesophageal tamponade, the pa tient complained of discomfort due to the tube. Fifteen patients survived the admis sion in the tamponade group compared with all the patients in the octreotide group.
Conclusion In this clinical trial octreotide infusion gave comparable control of variceal bleeding to oesophageal tamponade. Tolerance of treatment was significantly better in the oc treotide group.
References 1 Panes J. Teres J. Bosch .1. Rodes J: Efficacy of balloon tamponade in treatment of bleeding gas tric and oesophageal varices. Dig Dis Sci 1988:33: 454-459. 2 Slotnik IL, Teigland JD: Cardiac accidents follow ing vasopressin injection (‘Pitressin'). JAMA 1951:146:1126-1129. 3 Kravetz D. Bosch J. Teres J. et al: Comparison of somatostatin and vasopressin infusions in treat ment of acute variceal hemorrhage. Hepatology 1984:4:442-446. 4 Jenkins SA, Baxter JN. Corbett W. et al: A pro spective randomised controlled clinical trial com paring somatostatin and vasopressin in control ling acute variceal haemorrhage. Br Med .1 I985:i: 275-278. 5 Pringle SD. McKee RE. Garden DJ. Lorimer AR. Carter DC: The effect of SMS 201-995 on portal and systemic haemodynamics in cirrhosis. Ali ment Pharmacol Ther 1988:2:451-459. 6 Pugh RMH. Murray-Lyon IM. Dawson JL, Pie troni MC. Williams R: Transection of the oesoph agus for bleeding oesophageal varices. Br .1 Surg 1973:60:646-649. 7 Lebrec D. Hillon P. Munoz C. et al: The effect of propranolol on portal hypertension in patients with cirrhosis: a haemodynamic study. Hepato logy 1982:2:523-527. 8 Haddock G. Garden J. McKee RF. Anderson JR. Carter DC: Oesophageal tamponade in the man agement of acute variceal hemorrhage. Dig Dis Sci 1989:34:913-918.
Ruth McKee Surgical Unit Royal Infirmary Glasgow G4 QSF (UK)
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 9:27:19 PM
sion or oesophageal tamponade. Patients assigned to treatment with octreotide received an intravenous infusion of 25 pg/h of octreotide over 48 h. Patients in the oesophageal tamponade group had a 4-lumen Minnesota modification of the Sengstaken-Blakemore tube passed. Both gastric and oesophageal bal loons were left inflated for the first 24 h. but from 24 to 48 h only the gastric balloon was in Hated. In both groups fibreoptic endoscopy and injection sclerother apy were performed at 48 h. Bleeding control, compli cations. and symptomatic side effects were recorded. Evidence of chest infection was looked for particu larly carefully.
McKee
A Study of Octreotide in Oesophageal Varices
65
Discussion them with a series of pharmacologically complicated manoeuvres. Dr. Harris: The incidence of side effects from tam ponade seems low'er than that usually reported in the literature. Am I correct? Dr. McKee: The reports in the literature vary very widely. An institution where tamponade is the stan dard treatment for bleeding from oesophageal varices and is performed often, and where the patients are looked after in one ward by one set of nursing staff, can give very good results. I think our results fall into that category. The reports where oesophageal tamponade is said to be very dangerous come from units where it is used as a last-ditch measure. We do see problems in patients transferred from other hospitals. Dr Jenkins (Liverpool): How was active bleeding assessed? We have admitted perhaps 200 patients over the last 3 or 4 years, and about 30% of them were actually bleeding. My second point relates to the dose. Our dose-response study on wedged hepatic venous pressure suggests that you might have a rather low dose at 25 gg/h octreotide. Dr McKee: We used that dose in our acute study because we had shown a difference in wedged hepatic venous pressure in our stable cirrhotics with that dose. Certainly I think it might be worth giving a bolus of 25 pg octreotide before starting infusion and it may indeed be that a higher dose might be more effective. We defined active variceal bleeding as bleeding seen down the endoscope which would nor mally have caused us to put down a Sengstaken tube. Thus, we did not randomize anyone who was simply oozing a very small amount from their varices. All patients required blood transfusion: all had evidence of shock with increased pulse rate and reduced ar terial blood pressure.
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 9:27:19 PM
Dr. Far/liing: The absolute difference in mortality between the octreotide group and the tamponade group was only 5. Were these patients all Child’s group C? Dr. McKee: Yes. they were. Dr. Farthing: I think this raises a key issue when performing small studies, that it is essential to stratify the initial randomization for severity of disease. You probably should not overplay that statistical differ ence, because if the patients had been stratified by severity at the outset you would not have a differ ence. Dr. McKee: Yes. I think that is right. The trouble with performing this sort of study is the small number of patients available. It would have been much more difficult to organize if we have tried to stratify before randomization, particularly if you remember the practicalities of assessing patients to a Child’s grading when they have just arrived with a variceal bleed. Their liver function tests are not going to be imme diately available. But obviously it would be useful. Dr. Farthing: I would agree. It is interesting some times to run a clinical grading - that is. your initial clinical assessment - against the subsequent Child's grading. I think you would not have done badly if perhaps you had done that. Dr. Chalmers (Leeds): I think you were a little unfair about vasopressin. Many of the disadvantages you quoted can be partially or totally overcome by intravenous nitrate therapy. Have you any experience of that? Dr. McKee: We use vasopressin very little in the Glasgow Royal Infirmary for the reasons that I have outlined. Given that patients have an acute assault on their haemodynamic system, having lost a consider able amount of blood, we feel a little wary of treating
