Article Text

Treatment of variceal haemorrhage
1. A K BURROUGHS
1. Department of Liver Transplantation and
2. Hepatobiliary Medicine,
3. Royal Free Hospital School of Medicine,
4. Pond Street,
5. London NW3 2QG, UK

## Statistics from Altmetric.com

Randomised studies of emergency therapy for variceal bleeding are difficult to design, conduct and evaluate. Despite international consensus definitions covering descriptive characteristics of patients, evaluation of end points and checklists for presentation of results, many trials are difficult to interpret.1 The multicentre English study by Jenkins et al(Gut1997;41:526–33) is a case in point. In this randomised study injection sclerotherapy (n=77) was compared with octreotide 50 μg/h (n=73) over a 48 hour period, after which sclerotherapy was used in the octreotide group. By its nature this comparison is unblinded for both the clinician and the patient. Although it could be argued that end points for control of bleeding are fairly “hard end points”, this is not the case in practice. I highlight some examples. Balloon tamponade was added for 12 hours in both trial groups: in the injection group the indication was if there was persistent oozing at endoscopy and in the octreotide group, the indication was “if the bleeding persisted after starting octreotide”, presumably after the endoscopy had been completed. Although the numbers needing tamponade were similar (sclerotherapy n=20, octreotide n=13), I suspect the severity of bleeding was different as the parameters of assessment were not the same (endoscopic oozing versus clinical evidence of persistent bleeding). Neither use of balloon tamponade was counted as a failure to control bleeding.  Further need of balloon tamponade and/or haematemesis or melaena accompanied by changes in pulse or blood pressure (not specified) or decrease in haemoglobin (not specified) within 48 hours constituted failure to control bleeding (n=14, injection sclerotherapy; n=11, octreotide). This was the same in both groups. However, if transfusion requirements are evaluated to corroborate the purported similar efficacy then again the data are not presented in a standard way. Transfusion need is described, but only from hospital admission to the end of 48 hours, and not from randomisation. There was a median delay of 23 hours (sclerotherapy) and 19 hours (octreotide) before patients received treatment from time of bleeding, but the median time to admission to the trial hospital was five hours (sclerotherapy) and six hours (octreotide). Clearly, patients were being transfused in these intervals; as this is a measure of severity of bleeding this should have been evaluated.

There were more deaths over 60 days in the octreotide group—relative risk 1.91 (95% confidence interval 0.97 to 3.78, p=0.06). This may be a chance result, as mortality has not been increased in any other study of octreotide versus other treatment in variceal bleeding.  Overall, the cautious would interpret this study as suggesting that octreotide does not offer any advantage over sclerotherapy. This is because a diagnostic endoscopy must still be performed; if octreotide is used injection sclerotherapy is only delayed (by 48 hours) and there is no gain in efficacy and there could be an increased mortality rate. Why spend more money on octreotide? The way the results are presented in this paper does not justify delaying sclerotherapy by 48 hours by using octreotide, but it must be emphasised that this is not a study in which conclusions can be drawn easily in favour of octreotide or sclerotherapy.  The study by Sauer et al(Gastroenterology1997;113:1623–31) compares transjugular intrahepatic portosystemic stent shunt (TIPS) and long term sclerotherapy for the prevention of variceal rebleeding. This type of trial is easier to design and evaluate than acute trials of variceal bleeding. Six trials have now been published in full and another two are in abstract form. My personal view has always been that the place of an elective TIPS could never be as a straight replacement for endoscopic techniques. The experience of distal splenorenal shunt versus sclerotherapy, albeit in well compensated cirrhotic patients, has shown that survival was the same but rebleeding was far less in the shunted group.2 However, encephalopathy rates were higher in the shunted group. Rescue shunts when sclerotherapy failed were found to be useful. Distal spleno-renal shunts are still good operations in compensated cirrhotic patients with the appropriate indications. What could have one expected of TIPS trials versus sclerotherapy? Essentially a similar result to the surgical shunt trials, but with some important caveats: (1) more grade C patients would receive TIPS than in the surgical shunts trials; (2) an increased survival rate per procedure as surgical mortality is avoided; (3) decreased survival long term as more decompensated patients are treated with TIPS and TIPS is a total shunt thereby increasing the likelihood of shunt induced liver failure; (4) decreased efficacy in terms of rebleeding because TIPS block; (5) equivalent or worse encephalopathy rates because although TIPS block—reducing encephalopathy—they are total shunts, and the population with TIPS includes more grade C patients; (6) considerable long term expense in order to monitor shunts and keep them patent. These a priori considerations have been fulfilled in this trial. The authors conclude the same in their study of 83 patients (rebleeding: TIPS 23%, sclerotherapy 57%; survival: TIPS 69%, sclerotherapy 67%), and suggest TIPS should be confined to sclerotherapy failures. Moreover, survival was slightly worse with TIPS on meta-analysis.3 Lastly, one can argue that today, comparison of TIPS should be made with banding ligation, which has been shown to be better than sclerotherapy.4However, this would not change the message that TIPS is not a first line treatment for the prevention of rebleeding once the initial bleeding episode has been controlled.

In the paper by Rosemurgy et al(Surgery1997;122:794–800) a randomised comparison was made between TIPS and an 8 mm calibrated H graft portacaval shunt at an American centre. One of the advantages of TIPS, apart from avoiding a laparotomy, is the fact that the narrow diameter (less than 12 mm, effectively often only 10 mm or less) reduces the frequency of encephalopathy. The advent of fashioning surgical shunts with prosthetic material with a fixed diameter has allowed a better comparison of “like with like”. The expected advantage of a calibrated surgical shunt was better patency rates and thus reduced long term costs. In this study 40 patients were randomised to each group, two thirds were Child’s class A or B, but there were more urgent shunts in the TIPS group (28 v 8%, p=0.04). Nevertheless mortality was similar: 30% TIPS, 23% surgical. Variceal rebleeding was higher in the TIPS group (8v 0) and shunt revision higher in the TIPS group also (6 v 4). Encephalopathy rates were higher in the surgical group (3 v 7). Although the initial cost was less with TIPS, the follow up costs were higher, thus the total costs were similar long term: c. $69 000 (£43 000) for TIPS and c.$66,000 (£41 000) for surgery. The message from this paper is that a surgical shunt, at least in well compensated patients, is better than TIPS. However, at present a surgical shunt should still be reserved for failure of endoscopic methods. It remains to be seen whether a calibrated H graft portocaval shunt is better than a distal splenorenal shunt, as the calibrated shunt suffers from a higher thrombosis rate than distal the splenorenal shunt. Surgical practice in terms of shunting does not need to be changed on the basis of this paper. It reinforces the finding that with current technology TIPS is not a good long term shunt.

# Octreotide followed by sclerotherapy was not more effective than emergency injection sclerotherapy for acute variceal haemorrhage

## Objective

To compare the efficacy of a 48 hour infusion of octreotide, followed by injection sclerotherapy, with emergency injection sclerotherapy alone in the treatment of acute variceal haemorrhage.

## Design

Randomised controlled trial with 60 days of follow up.

## Setting

4 hospitals in the UK.

## Patients

150 patients with an upper gastrointestinal bleed and an endoscopically confirmed variceal bleed that resulted in a systemic disturbance or required the transfusion of ⩾2 units of blood to restore haemoglobin concentration. Exclusion criteria were oesophageal varices that were not the source of bleeding, treatment with vasoactive drugs or injection sclerotherapy within the previous 7 days, or transfer from other hospitals with an oesophageal balloon in situ. Follow up was 100%.

## Intervention

77 patients received emergency injection sclerotherapy using a fibreoptic endoscope to inject each varix with 2 to 3 ml of ethanolamine oleate. 73 patients received intravenous octreotide, 50 μg per hour, for 48 hours from the time of the first endoscopy, after which injection sclerotherapy was given.

## Main outcome measures

Number of recurrences of bleeding within 48 hours, mortality rate over 60 days, and number of complications.

## Main results

No differences in the rate of rebleeding, deaths, or complications existed between patients who received emergency injection sclerotherapy and those who received octreotide followed by sclerotherapy (table). The study had >90% power to detect a difference of 20% between the 2 treatments.

## Conclusion

Octreotide followed by sclerotherapy and immediate injection sclerotherapy alone did not differ in efficacy when used to treat acute variceal haemorrhage.

Octreotide followed by injection sclerotherapy

# Transjugular intrahepatic portosystemic stent shunt reduced the rate of variceal rebleeding in patients with cirrhosis

## Objective

To compare the efficacy of transjugular intrahepatic portosystemic stent shunt (TIPS) and endoscopic sclerotherapy plus propranolol in the prevention of variceal rebleeding in patients with cirrhosis of the liver.

## Design

Randomised controlled trial with 2 years of follow up.

Germany.

## Patients

83 patients (mean age 57 y, 58% men) who had cirrhosis and an episode of acute oesophageal variceal bleeding. Exclusion criteria included gastric varices, previous endoscopic or surgical treatment of varices, terminal illness, and septicaemia.

## Intervention

42 patients received sufficient Palmaz stents to cover a shunt tract of 4 cm to 6 cm. The stents were expanded from 8 mm to 12 mm in diameter to reduce portosystemic gradients to 10 mm Hg to 15 mm Hg. 41 patients received endoscopic injections of 5% ethanolamine oleate, 2 to 3 ml, and propranolol twice daily, 40 mg per day. The propranolol dose was increased by 40 mg per day to reach a target dose at which the resting heart rate was reduced by 25%.

## Main outcome measures

Recurrence of variceal bleeding, hepatic encephalopathy, and death.

## Main results

Analysis was by intention to treat. Patients who received TIPS were less likely to rebleed than those who received sclerotherapy plus propranolol (p=0.0004) (table). Patients who received TIPS had more hepatic encephalopathy than those who received sclerotherapy plus propranolol (p=0.005) (table). There was no difference in the number of deaths between patients who received TIPS and those who received sclerotherapy plus propranolol (cumulative survival rate 69% v 67%, p=0.62).

## Conclusions

Compared with endoscopic sclerotherapy plus propranolol, transjugular intrahepatic portosystemic stent shunt reduced the rate of rebleeding but increased the rate of hepatic encephalopathy in patients with cirrhosis of the liver. Survival rates did not differ between treatments.

Transjugular intrahepatic portosystemic stent shunt (TIPS) v sclerotherapy plus propranolol*

# Transjugular intrahepatic portosystemic shunt was not more cost effective than H graft portacaval shunt for bleeding varices

## Objective

To determine the cost effectiveness of transjugular intrahepatic portosystemic shunt (TIPS) compared with H graft portacaval shunt in the treatment of patients who had cirrhosis and bleeding varices due to portal hypertension.

## Design

Randomised controlled trial with >3 years of follow up.

USA.

## Patients

80 adults (mean age 54 y, 65% men) who had cirrhosis and bleeding due to portal hypertension.

## Intervention

40 patients received an H graft portal shunt using {8 mm externally reinforced polytetrafluorethylene. The graft was 3 cm from toe to toe and 1.5 cm from heel to heel.}* The pressure gradient between the portal vein and the inferior vena cava was <10 mm Hg. 40 patients received TIPS. The pressure gradient between the portal vein and the inferior vena cava was <12 mm Hg.

## Main cost and outcome measures

The main outcome measure was death. Direct costs of patient care for index admissions, subsequent admissions for complications, and follow up were determined using hospital, clinic, and physician charges in 1996 US dollars.

## Main results

The mortality rate did not differ between patients whoreceived TIPS and those who received H graft portacaval shunts (30% v 23%, p=0.45). Comparing the mean costs for TIPS and H graft portacaval shunts, there were no statistically significant differences in the cost of index admissions (mean [SD] $48 188 [$43 355] v $61 522 [$47 615], p=not significant [NS]). With follow up, the mean cost of care for each shunt was more for patients who received TIPS than for patients who received H graft portacaval shunts ($20 720 v$4512, p=0.02). There was no difference in the total cost of care between TIPS and H graft portacaval shunts (mean [SD] $69 276 [$52 712]v $66 034 [$49 118], p=NS). More patients who received TIPS had readmissions than patients receiving H graft portacaval shunts for variceal rebleeding (15%v 0%, {p=0.026}†) and shunt revisions due to occlusion or thrombosis (18% v 0%, {p=0.0012}†). No differences existed among treatments in the number of patients readmitted for cellulitis and abdominal pain, encephalopathy, ascites, shunt evaluation, and perforated gastric ulcers.

## Conclusions

Transjugular intrahepatic portosystemic shunt was not more cost effective than H graft portacaval shunt in treating patients who had cirrhosis and bleeding varices due to portal hypertension.

View Abstract

## Footnotes

• Source of funding: Sandoz Pharmaceuticals Ltd, Camberley, UK.

• For article reprint: Mr R Sutton, Clinical and Cancer Trials Unit, Department of Surgery, University of Liverpool, PO Box 147, Liverpool L69 3BX, UK. Fax +44 (0)151 706 5826.

## Footnotes

• Source of funding: not stated.

• For article reprint: Dr. P. Sauer, Medizinische Universitätsklinik, Innere Medizin IV, Bergheimer Straβe 58, D-69115, Heidelberg, Germany. Fax +49 6221 56 4922.

## Footnotes

• 3-150 Rosemurgy AS. Ann Surg1996;224:378–86.

• 3-151 p values calculated from data in article, Fisher’s exact test

• Source of funding: not stated.

• For article reprint: Dr A S Rosemurgy II, PO Box 1289, Room H-152, Tampa, FL 33601, USA. Fax +1 813 251 7396.

## Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.