- Email: [email protected]
Has sclerotherapy altered the management of patients with variceal bleeding?
Has Sclerotherapy Altered the Management of Patients with Variceal Bleeding? John Terblanche, ChM, FCSSA, FRCS, Cape Town,South Africa Sclerotherapy is currently the primary treatment of choice for the majority of patients who present with esophageal variceal bleeding. Although it has altered the management of these patients, unanswered questions and controversies remain. Patients with acute variceal bleeding should preferably be treated in a specialized center. The primary treatment ~should be immediate sclerotherapy, when possible~!Portosystemic shunts and esophageal transection should be reserved for the 5% to 10% of patients in whom sclerotherapy fails to control acute bleeding. There are several treatment options for long-term management after a varieeal bleeding episode. Selerotherapy is one option and has become the primary treatment in most major centers. All patients with end-stage liver disease must be considered for liver transplantation, and sclerotherapy should be the primary method of treatment in those who are selected. Pharmacologic therapy remains controversial. I propose that portosystemic shunts and devascnlarization and transection operations be reserved for those few patients in whom sclerotherapy fails to eradicate the varices and to prevent recurrent variceal bleeding. Patients in whom sclerotherapy is unsuccessful should be identified and treated early.
T
he,, short answer to the question posed m" the Utle" l _ Has Sclerotherapy Altered the Management of Patients with Variceal Bleeding?" is "yes." However, this needs to be qualified. Dean Warren, probably more than any of the "surgical shunters," has helped to answer this question and to establish the true role of sclerotherapy in management as a result of his group's seminal controlled trial, which compared sclerotherapy plus shunt salvage with the Warren shunt alone [I]. This presentation reviews certain unanswered questions, as well as the role of sclerotherapy in the management of patients with acute variceal bleeding and in the long-term management of patients after a variceal bleeding episode. Assessment of prophylaxis in patients with From the Department of Surgery and the Medical Research Council Liver Research Center, Universityof Cape Town Medical School,Cape Town, South Africa. Requests for reprints should be addressed to John Terblanche, ChM, Department of Surgery, University of Cape Town Medical School, Observatory 7925, Cape Town, South Africa.
varices that have not yet bled is outside the scope of this paper; however, my views on prophylaxis have been presented in several recent publications [2-4]. SCLEROTHERAPY: UNANSWERED QUESTIONS Technical considerations: Certain technical questions have been resolved, including the use of the flexible endoscope rather than the rigid endoscope used in the past. The superiority of the flexible endoscope has been confirmed in a controlled trial [5]. On the other hand, many of the questions posed a decade ago still remain unanswered despite innumerable published studies. Sclerotherapy may be performed by direct intravariceal injection. Here, the aim is to cause variceal thrombosis and thereby achieve both control of acute variceal bleeding and prevention of subsequent recurrent episodes of variceal bleeding (Figure 1A). Intravariceal injections are usually limited to a single injection tretment per varix, with the needle puncture sited immediately above the esophagogastric junction, although multiple injections are used for each varix by some workers. The intravariceal technique has been most widely used in the United States and the United Kingdom. The alternative paravarieeal technique is shown in Figure 1B. There are several variations of this technique, the most widely used being that proposed by Paquet [6], where multiple 0.5-mL to 1mL injections of polidocanol are placed circumferentiaUy around the lower third of the esophagus, commencing at the esophagogastric junction and moving up the esophagus in a helical fashion. The concept is that acute bleeding is controlled by compression of the bleeding varix by submucosal swellings produced by the injections, and that in the long-term, the overlying mucosa is thickened, thereby preventing further bleeding while preserving deeper variceal channels. A combination of intravariceal and paravariceal injection sclerotherapy is depicted in Figure 1C. This was first proposed by the author's group [7] and Soehendra et al [8] of Germany, with the aim of combining the advantages of the former two techniques and in the hope that this would provide a summation of these advantages. My group currently uses the combined technique in the management of acute variceal bleeding and an intravariceal technique for repeated sclerotherapy. The one trial that compared intravariceal and paravariceal sclerotherapy used a relatively unusual agent, absolute alcohol, and demonstrated that the intravariceal technique was superior [9]. Kitano et al [10] have recently proposed a novel three-part sclerotherapy management protocol: thrombosis is induced in the varices by intravariceal injections; this is followed by removal of the esophageal mucosa by repeated paravarieeal injections; and finally reepithelialization is allowed to occur. They have reported remarkable results with the use of ethanolamine
THE AMERICAN JOURNAL OF SURGERY VOLUME 160 JULY 1990
37
TERBLANCHE
A
:.*:
c
D
Figure 1. A flexible endoscope is used for intravariceal injection (A), paravariceal (submucosal) injection (B), and combined paravariceal and intravariceal injection (C), whereas a rigid esophagoscope with a slot is used for intravariceal injection (D). Reprinted from [2], with permission from the publisher.
oleate for this three-phase treatment in a group of 141 patients. There were no recurrent varices and only minimal early recurrent bleeding at a median follow-up of 14.6 months. I would, however, advocate caution in considering this approach until further controlled trials have shown it to be effective and safe in other groups of patients. The technique could be hazardous with certain sclerosing agents, such as sodium tetradecal sulfate, particularly if deep ulceration had occurred prior to thrombosis of the varices, which could result in massive uncontrolled bleeding. Timing of selerotherapy: The timing of intravariceal sclerotherapy in assessing long-term management has been subjected to two controlled trials [11,12]. Both concluded that injection treatments should be performed at least at weekly intervals until the varices are eradicated. Many groups, including our own, have failed to adhere to a weekly regimen, largely due to logistic considerations. Our 10-year analysis confirmed this to be a problem [13]. With paravariceal sclerotherapy, repeat endoscopy has been performed at weekly or more frequent intervals, although this has not been subjected to controlled trial analysis. The timing of the combined technique still has to be subjected to properly controlled trials. Sclerosant agents: The most widely used agent for intravariceal and combined techniques is 5% ethanolamine oleate. Polidocanol has been the main agent used 38
by the majority of advocates of the paravariceal technique and by some for the combined technique. It has been difficult to prove which agent is better since the characteristics of agents required for successful intravariceal or paravariceal techniques clearly vary. Several controlled trials have been reported. Ethanolamine oleate was shown to be superior to polidocanol [14] and sodium tetradecal sulfate [15] using one technique, whereas polidocanol was superior to absolute alcohol using a different injection technique [16]. Combination therapy: Although sclerotherapy plays a major role in patients with variceal bleeding, there is increasing evidence that it should not be the only treatment considered and used. Dean Warren's excellently executed controlled trial demonstrated that the combination of sclerotherapy with shunt salvage, using the Warren shunt in cases of sclerotherapy failure (31%), was superior to the use of the Warren shunt alone in longterm management after a variceal bleeding episode [1]. Although two other controlled trials have failed to confirm this finding, trial design and technical differences probably account for this discrepancy [17,18]. An important problem with sclerotherapy management is that patients continue to have major variceal bleeding until their varices are eradicated. The use of pharmacologic therapy as an adjunct during this phase might prove useful because both sclerotherapy and propranolol have been shown to be equally effective when compared in controlled trials of long-term therapy [19]. In one controlled study, early rebleeding was significantly reduced when propranolol was administered for 14 days, commencing 24 hours after the initial bleeding varices had been successfully treated [20]; however, further combination therapy trials are required. Anatomic and physiologic factors: Recent anatomic and physiologic studies have provided new data and understanding. Has this altered our management or our concept of why sclerotherapy works? The answer is a qualified yes. Varices almost invariably bleed from the lower 5 cm of the esophagus. The reason for this is not yet clearly understood, although it is better explained by recent studies. Bleeding probably results from internal disruption or explosion of the varices as a result of increased pressure from within the varix or because of thinning of the overlying vascular supporting structures, and is not caused by external erosion from the lumen of the esophagus by acid-pepsin digestion or by the trauma of swallowing solids. In an analysis of recent anatomic studies, our group has postulated why sclerotherapy is successful [21]. Prediction of variceal bleeding: This has proved to be one of the most controversial areas. Most groups have found it exceedingly difficult to predict which patient will bleed, either for the first time or after previous variceal bleeding. It has also proved difficult to define which patients with acute variceal bleeding will stop active bleeding with simple measures, which patients will continue to bleed, and which patients will have an early recurrence after bleeding has been controlled. The most widely proclaimed predictive index has been the complex endoscopic classification of varices proposed by Beppu et al [22] of
THE AMERICANJOURNALOF SURGERY VOLUME160 JULY1990
SCLEROTHERAPY FOR VARICEAL BLEEDING
Japan. However, in a controlled trial, they were only able to predict a first episode of variceal bleeding in 20% of their control patients [23]. Therefore, in the treated group in this trial, 80% of the patients were, by implication, subjected to unnecessary major surgery. The simple endoscopic and coagulation signs used by Paquet [24] to predict variceal hemorrhage in German patients proved more accurate, with a 66% predictive value of a first episode of variceal bleeding. This is probably partly the result of different types of patients being included in the studies. These discrepancies have led various groups to search for combined clinical and endoscopic findings that would be more accurate. The North Italian Endoscopic Club [25] has proposed such a predictive index, but their index requires further evaluation in controlled clinical trials for both its value in prophylactic therapy and in assessment as a predictor of further variceal bleeding in patients being considered for therapy after a bleeding episode. Etiologic and geographic variations: There are major variations in the cause of the underlying liver disease and in the severity of liver disease in different geographic areas, and therefore, by implication, variations in the chance of a patient dying. This has caused major problems in comparing studies from different areas and in making recommendations on the basis of such studies. I believe that meta-analysis of several randomized trials from disparate areas, like that published by Infante-Rivard et al [26], is difficult to justify because of these geographic and etiologic variations. Cause of death: Patients with portal hypertension and variceal bleeding due to underlying liver disease die either as a result of continued active variceal bleeding or as a result of deteriorating liver function and liver failure. Whereas bleeding frequently aggravates the liver failure, the converse is also true; this leads to a vicious cycle, which ends in death. It is unusual for patients to die of active variceal bleeding today because sclerotherapy has markedly improved management [27], but the underlying liver disease, nevertheless, determines the prognosis. Few patients with severe underlying liver disease survive 5 to 10 years, despite the use of repeated sclerotherapy
[13]. SCLEROTHERAPY FOR ACUTE VARICEAL BLEEDING All patients with suspected acute variceal bleeding require immediate hospital admission and expert management, preferably in an intensive care unit. Resuscitation includes crystalloids and blood component therapy, particularly for massive hemorrhage. An unanswered question is whether such patients should be treated in specialized portal hypertension units or whether they can be as successfully treated by primary-care physicians, gastroenterologists, or surgeons in local or district hospitals. I believe that such patients should be transferred to a specialized center and placed under the care of a combined team with particular expertise in portal hypertension management whenever possible, because they may require more than one form of complex therapy. The diagnosis of variceal bleeding must be confirmed
by emergency endoscopy, because at least 30% of patients in whom it is suspected do not have Varices, and up to one third of patients with varices will be bleeding from another source [28,29]. The role of pharmacologic therapy remains controversial. My views have been expressed in recent publications [2,3,30]. My co-workers and myself advocate intravenous infusion of vasopressin (0.4 units/min) plus sublingual nitroglycerine (one tablet each hour for 6 hours). Balloon tamponade is also controversial but should not be. If active variceal bleeding is not controlled pharmacologically or by immediate sclerotherapy, the patient must have a balloon tube inserted. Correct insertion will invariably control the bleeding; thereby allowing time to resuscitate the patient and to summon expert help for subsequent definitive emergency m a n a g e m e n t [2,29]. Definitive therapy should occur within 6 to 12 hours, at the time that the balloon tube is removed. Sclerotherapy is currently the mainstay of treatment for acute variceal bleeding; however, the role of immediate or urgent shunting or esophageal transection is being reevaluated. A policy of initial measures to control hemorrhage, followed by urgent sclerotherapy, has been the most widely practiced regimen in the past decade. Three controlled trials [31-33] and innumerable uncontrolled studies [2] have confirmed that sclerotherapy controls acute variceal bleeding in 90% to 95% of patients. Sclerotherapy has also been shown to be more effective than either balloon tamponade or vasopressin used alone. A single injection treatment will control variceal bleeding in more than 70% of patients; therefore, only 30% of patients require further sclerotherapy or other treatment [29]. A second injection will increase the control rate to 90% or 95%. Unfortunately, various published trials are difficult to compare because of differences in design. There has been a shift in emphasis from delayed emergency sclerotherapy to the use of sclerotherapy during the first emergency endoscopy. This view is supported by two controlled trials [34,35] and is my recommended therapy. Immediate sclerotherapy requires a high degree of skill when attempted in patients with active bleeding, but it is possible if the correct technique is used. Immediate sclerotherapy Should also be undertaken in patients with varices that have stopped bleeding, as some of these patients would otherwise have further major life-threatening bleeding before subsequent sclerotherapy could be performed. We have defined failure of emergency sclerotherapy as an acute episode ofvariceal bleeding that occurs within hours or days of emergency injection sclerotherapy and that requires blood transfusion and subsequent emergency management [2]. We and others have demonstrated that emergency sclerotherapy, using one or two injection treatments, is successful in controlling acute variceal bleeding in 90% to 95% of patients. The 5% to 10% of patients in whom sclerotherapy fails must be considered for other forms of management. Analysis of our first prospective evaluation of tamponade and sclerotherapy showed that the mortality approached 90% in those few patients who had further episodes of acute variceal bleeding after two sclerotherapy treatments during a single
THE AMERICANJOURNAL OF SURGERY VOLUME160 JULY 1990 39
TERBLANCHE
hospital admission when low-risk Child's class A patients were excluded [36]. Therefore, bleeding in such patients must be controlled temporarily with balloon tamponade followed by either a portosystemic shunt or an esophageal transection [2,3,36]. Unfortunately, during the initial evaluation, we have been unable to predict which patients will not respond to sclerotherapy. If such a prediction was possible, selected patients could undergo immediate surgery. This leads to the important question, "Is there a place for immediate or early shunting or esophageal transection?" Unfortunately, the answer is not yet known. The role of emergency shunt procedures has received increasing support recently. Orloff et al [37] have championed emergency portacaval shunts over the years. In an important study, Cello et al [38] demonstrated that portacaval shunts were more effective than sclerotherapy in preventing early and late recurrences of bleeding in a group of high-risk patients, but there was no difference in survival between the shunted and control groups, There are also advocates of the early use of surgical transection. On the basis of their controlled trial, Burroughs et al [39], from the Royal Free Hospital in London, have concluded that emergency stapel transection is as safe as sclerotherapy and, in their hands, is more effective than a single sclerotherapy procedure. However, their recommendation is that a transection should be offered after two injection treatments have failed. This is the identical conclusion reached earlier by our group [36]. I predict that either emergency portacaval shunting or emergency esophageal transection will be performed more frequently in the future for management of acute variceal bleeding. S C L E R O T H E R A P Y FOR LONG-TERM MANAGEMENT Sclerotherapy is one of the available options in longterm management after an episode of variceal bleeding, but is it the best primary option for the majority of patients? This is a difficult question, but Warren's trial suggested that sclerotherapy with shunt salvage was better than the use of the Warren selective distal splenorenal shunt alone [1]. Note that, with the exception of hepatic transplantation, no other therapy actually treats the underlying liver disease. Furthermore, some forms of therapy, such as shunting, may worsen the liver disease. The best management may hinge on the underlying causes of portal hypertension, which differ in various parts of the world. In Western countries, alcoholic cirrhosis and certain types of nonalcoholic cirrhosis are common. Emphasis will be placed on the management of these patients. In addition to repeated injection sclerotherapy and liver transplantation, the other options are conservative medical management with close follow-up to detect an additional bleeding episode and adminstration of specific therapy should it recur; portosystemic shunts; transection and devascularization; and long-term pharmacoligic therapy [2,3]. The latter alternative remains highly controversial. The problems of lifelong therapy and compliance compound the difficulty of assessing rec-
40
ommendations in the literature. I believe that the use of drugs should currently be restricted to major institutions with a special interest in portal hypertension. All patients who have had variceal bleeding should be considered for hepatic transplantation. This is the only therapy that cures the underlying liver disease as well as the portal hypertension. If patients are potential transplant candidates, the therapy used should not interfere with a subsequent liver transplant. We therefore recommend sclerotherapy as the primary treatment for such patients. If sclerotherapy fails, we prefer simple esophageal transection to shunting. The five completed controlled clinical trials, and innumerable uncontrolled studies, have demonstrated that repeated sclerotherapy effectively eradicates esophageal varices in most patients [40-44]. After eradication, the incidence of recurrent variceal bleeding is markedly reduced, although it still occurs frequently until variceal eradication is achieved. The ability of repeated sclerotherapy treatments to increase survival has yet to be determined. Only one of the five controlled trials [42] showed a clear improvement in survival with sclerotherapy. There are several important problems related to repeated sclerotherapy in long-term management. Varices are difficult to eradicate in some patients, whereas others continue to have life-threatening bleeding during the period of variceal eradication. Long-term sclerotherapy should not be continued when varices cannot be eradicated or bleeding episodes continue to recur. Instead, surgery should be performed at an early stage [3]. This combination of sclerotherapy with surgical salvage if treatment fails, as recommended by Warren [1], is probably the best long-term strategy for the majority of patients. Other problems with long-term sclerotherapy include the need for lifelong follow-up with repeated injections because varices recur in time. Surviving patients place an increasing burden on hospital resources even when sclerotherapy can be performed on an outpatient basis. Although the complications of sclerotherapy do not usually pose a major problem, they are cumulative with time [45]. The complications in the 10-year Cape Town experience have been detailed in recent publications [13,27,45]. It is important to attempt to identify those patients in whom sclerotherapy is likely to ultimately fail at an early stage of management and use surgical treatment as an alternative. An important, but seldom discussed, subset of patients is the high-risk group with end-stage liver disease, particularly alcoholic cirrhosis, in whom an early definitive decision must be made not to treat the patient. This is difficult, but should be seriously entertained; otherwise, hospital resources will be fruitlessly utilized on patients whose condition is not salvageable. In conclusion, I believe that sclerotherapy has altered the management of patients with portal hypertension and has led to more rational therapy. However, it has probably been overused in the past decade to the detriment of patient care in some centers, including our own. Sclero-
THE AMERICAN JOURNAL OFSURGERY VOLUME160 JULY 1990
SCLEROTHERAPY FOR VARICEAL BLEEDING
therapy is likely to remain the initial therapy of choice for most patients, but treatment failure must be identified early and the patient treated by a portosystemic shunt or a transection and devascularization procedure. REFERENCES 1. Warren WD, Henderson JM, Millikan W J, et al. Distal splenorenal shunt versus endoscopic sclerotherapy for long-term management of variceal bleeding: preliminary report of a prospective, randomized trial. Ann Surg 1986; 203: 454-62. 2. Terblanche J, Burroughs AK, Hobbs KEF. Controversies in the management of bleeding esophageal varices. N Engl J Med 1989; 320i 1393-8 and 1469-75. 3. Terblanche J. The surgeon's role in the management of portal hypertension. Ann Surg 1989; 209: 381-95. 4. Anonymous. Prophylactic sclerotherapy of oesophageal varices; is it justified? Lancet 1988; 1: 1369-70. 5. Bornman PC, Kahn D, Terblanche J, Worthley C, Spence RA, Krige JJ. Rigid versus fiberoptic endoscopic injection sclerotherapy: a prospective randomized controlled trial in patients with bleeding esophageal varices. Ann Surg 1988; 208: 175-8. 6. Paquet K-J. Endoscopic paravariceal injection sclerotherapy of the esophagus--indications, technique, complications: results of a period of 14 years. Gastrointest Endosc 1983; 29: 310-5. 7. Terblanche J, Bornman PC, Jonker MAT, Kirsch RE, Saunders SJ. Injection sclerotherapy of esophageal varices. Sem Liver Dis 1982; 2: 233-41. 8. Soehendra N, de Heer K, Kempeneers I, Runge M. Sclerotherapy of esophageal varices: acute arrest of gastrointestinal hemorrhage or long-term therapy? Endoscopy 1983; 15: 136-40z 9. Sarin SK, Nanda R, Sachdev G, Chari S, Anand BS, Broor SL. Intravariceal versus paravariceal sclerotherapy: a prospective, controlled, randomised trial. Gut 1987; 28: 657-62. 10. Kitano S, Koyanagi N, Iso Y, Higashi H, Sugimachi K. Prevention of recurrence of esophageal varices after endoscopic injection sclerotherapy with ethanolamine oleate. Hepatotogy ;1987; 7: 810-5. 11. Westaby D, Melia WM, Macdougall BR, Hegarty JE, Williams R. Injection sclerotherapy for oesophageal varices: a prospective randomised trial of different treatment schedules. Gut 1984; 25: 129-32. 12. Sarin SK, Sachdev G, Nanda R, Batra SK, Anand BS. Comparison of the two time schedules for endoscopic sclerotherapy: a prospective randomised controlled study. Gut 1986; 27: 710-3. 13. Terblanche J, Kahn D, Bornman PC. Long-term injection sclerotherapy treatment for esophageal varices: a 10-year prospective evaluation. Ann Surg 1989; 210: 725-31. 14,. Kitano S, Iso Y, Koyanagi N, Higashi H, Sugimachi K. Ethanolamine oleate is superior to polidoeanal (aethoxysklerol) for endoscopic injection sclerotherapy of esophageal varices: a prospective randomized trial. Hepatogastroenterology t987; 34: i9-23. 15. Kitano S, Iso Y, Yamaga H, Hashizume N, Higashi H, Sugimachi K. Trial of sclerosing agents in patients with oesophageal varices. Br J Surg 1988; 75: 751-3. 16. Atamkuri SP, Bhargava DK, Sharma MP. Endoscopic sclerotherapy for esophageal varices: a prospective, randomised trial of absolute alcohol versus polidocanol. Indian J Gastroenterol 1988; 7: 87-9. 17. Rikkers LF, Burnett DA, Volentine GD, Buchi KN, Cormier RA. Shunt surgery versus endoscopic sclerotherapy for long-term treatment of variceal bleeding: early results of a randomized trial. Ann Surg 1987; 206: 261-71. 18. Teres J, Bordas JM, Bravo D, et al. Sclerotherapy vs. distal splenorenal shunt in the elective treatment of variceal hemorrhage: a randomized controlled trial. Hepatology 1987; 7: 430-6. 19. Fleig WE, Stange EF, Hunecke R, et aL Prevention of recur-
rent bleeding in cirrhotics with recent variceal hemorrhage: prospective, randomized comparison of propranolol and sclerotherapy. Hepatology 1987; 7: 355-61. 20. Marbet UA, Straumann A, Gyr KE, et al. Reduction in early recurrence of variceal bleeding by propranolol. Seand J Gastroenterol 1988; 23: 369-74. 21. Spence RAJ, Terblanche J. Venous anatomy of the lower oesophagus: a new perspective on varices. Br J Surg 1987; 74: 65960. 2 2 . Beppu K, Inokuchi K, Koyanagi N, et al. Prediction of variceal hemorrhage by esophageal endoscopy. Gastrointest Endosc 1981; 27: 213-8. 23. Inokuchi K. Prophylactic portal nondecompression surgery in patients with esophageal varices: an interim report. Ann Surg 1984; 200: 61-5. 24. Paquet K-J. Prophylactic endoscopic sclerosing treatment of the esophageal wall in varices: a prospective controlled trial. Endoscopy 1982; 14: 4-5. 25. The North Italian Endoscopic Club for the Study and Treatment of Esophageal Varices. Prediction of the first variceal hemorrhage in patients with cirrhosis of the liver and esophageal varices: a prospective multicenter study. N Engl J Med 1988; 319: 983-9. 26. Infante-Rivard C, Esnaola S, Villeneuve JP. Role of endoscopic variceal sclerotherapy in the long-term management of variceal bleeding: a meta-analysis. Gastroenterology 1989; 96: 1087-92. 27. Kahn D, Bornman PC, Terblanche J. A 10-year prospective evaluation of balloon tube tamponade and emergency injection sclerotherapy for actively bleeding oesophageal varices. HPB Surg 1989; 1: 207-19. 28. Novis BH, Duys P, Barbezat GO, Clain J, Bank S, Terblanche J. Fiberoptic endoscopy and the use of the Sengstaken tube in acute gastrointestinal haemorrhage in patients with portal hypertension and varices. Gut 1976; 17: 258-63. 29. Terblanche J, Yakoob HI, Bornman PC, et al. Acute bleeding varices: a five-year prospective evaluation of tamponade and sclerotherapy. Ann Surg 1981; 194: 521-30. 30. Anonymous. Management of acute variceal bleeding. Lancet 1988; 2: 999-1000. 31. Barsoum MS, Bolous FI, E1-Rooby AA, Rizk-Allah MA, Ibrahim AS. Tamponade and injection sclerotherapy in the management of bleeding oesophageal varices. Br J Surg 1982; 69: 76-8: 32. Paquet K-J, Feussner H. Endoscopic sclerosis and esophageal balloon tamponade in acute hemorrhage from esophagogastric varices: a prospective controlled randomized trial. Hepatology 1985; 5: 580-3. 33. Larson AW, Cohen H, Zweiban B, et al. Acute esophageal variceal sclerotherapy: results of a prospective randomized controlled trial. JAMA 1986; 255: 497-500. 34. Westaby D, Hayes PC, Gimson AES, Poison R J, Williams R. Controlled clinicaltrial of injection sclerotherapy for active variceal bleeding. Hepatology 1989; 9: 274-7. 35. Moreto M, Zaballa M, Berne A, Ibanez S, Ojembarrena E, Rodriguez A. A randomized trial of tamponade or sclerotherapy as immediate treatment for bleeding esophageal varices. Surg Gynecol Obstet 1988; 167: 331-4. 36. Bornman PC, Terblanche J, Kahn D, Jonker MA, Kirsch RE. Limitations of multiple injection sclerotherapy sessions for acute variceal bleeding. S Afr Med J 1986; 70: 34-6. 37. Orloff M J; Bell RH Jr, Hyde PV, Skivolocki WP. Long-term results of emergency portacaval shunt for bleeding esophageal varices in unselected patients with alcoholic cirrhosis. Ann Surg 1980; 192: 325-40. 38. Cello JP, Grendell JH, Crass RA, Weber TE, Trunkey DD. Endoscopic sclerotherapy versus portacaval shunt in patients with severe cirrhosis and acute variceal hemorrhage: long-term followup. N Engt J Med 1987; 316: 11-5. 39. Burroughs AK, Hamilton G, Phillips A, Mezzanotte G, Mclntyre N, Hobbs KEF. A comparison of sclerotberapy with staple
THE AMERICAN JOURNAL OFSURGERY
VOLUME 160 JULY 1990
41
TERBLANCHE
transection of the esophagus for the emergency control of bleeding from esophageal varices. N Engl J Med 1989; 321: 857-62. 40. Terblanche J, Bornman PC, Kahn D, et al. Failure of repeated injection sclerotherapy to improve long-term survival after oesophageal variceal bleeding: a five-year prospective controlled clinical trial. Lancet 1983; 2: 1328-32. 41. The Copenhagen Esophageal Varices and Sclerotherapy Project. Sclerotherapy after first variceal hemorrhage in cirrhosis: a randomized multi-center trial. N Engl J Med 1984; 311: 1594-600. 42. Westaby D, Macdougall BR, Williams R. Improved survival following injection sclerotherapy for esophageal varices: final anal-
42
ysis of a controlled trial. Hepatology 1985; 5: 827-30. 43. Korula J, Balart LA, Radvan G, et al. A prospective, randomized controlled trial of chronic esophageal variceal sclerotherapy. Hepatology 1985; 5: 584-9. 44. Soderlund C, Ihre T. Endoscopic sclerotherapy v. conservative management of bleeding oesophageal varices: a 5-year prospective controlled trial of emergency and long-term treatment. Acta Chir Scand 1985; 151: 449-56. 45. Kahn D, Jones B, Bornman PC, Terblanche J. Incidence and management of complications after injection sclerotherapy: a 10year prospective evaluation. Surgery 1989; 105: 160-5.
THE AMERICAN JOURNAL OF SURGERY VOLUME 160 JULY 1990
T
he,, short answer to the question posed m" the Utle" l _ Has Sclerotherapy Altered the Management of Patients with Variceal Bleeding?" is "yes." However, this needs to be qualified. Dean Warren, probably more than any of the "surgical shunters," has helped to answer this question and to establish the true role of sclerotherapy in management as a result of his group's seminal controlled trial, which compared sclerotherapy plus shunt salvage with the Warren shunt alone [I]. This presentation reviews certain unanswered questions, as well as the role of sclerotherapy in the management of patients with acute variceal bleeding and in the long-term management of patients after a variceal bleeding episode. Assessment of prophylaxis in patients with From the Department of Surgery and the Medical Research Council Liver Research Center, Universityof Cape Town Medical School,Cape Town, South Africa. Requests for reprints should be addressed to John Terblanche, ChM, Department of Surgery, University of Cape Town Medical School, Observatory 7925, Cape Town, South Africa.
varices that have not yet bled is outside the scope of this paper; however, my views on prophylaxis have been presented in several recent publications [2-4]. SCLEROTHERAPY: UNANSWERED QUESTIONS Technical considerations: Certain technical questions have been resolved, including the use of the flexible endoscope rather than the rigid endoscope used in the past. The superiority of the flexible endoscope has been confirmed in a controlled trial [5]. On the other hand, many of the questions posed a decade ago still remain unanswered despite innumerable published studies. Sclerotherapy may be performed by direct intravariceal injection. Here, the aim is to cause variceal thrombosis and thereby achieve both control of acute variceal bleeding and prevention of subsequent recurrent episodes of variceal bleeding (Figure 1A). Intravariceal injections are usually limited to a single injection tretment per varix, with the needle puncture sited immediately above the esophagogastric junction, although multiple injections are used for each varix by some workers. The intravariceal technique has been most widely used in the United States and the United Kingdom. The alternative paravarieeal technique is shown in Figure 1B. There are several variations of this technique, the most widely used being that proposed by Paquet [6], where multiple 0.5-mL to 1mL injections of polidocanol are placed circumferentiaUy around the lower third of the esophagus, commencing at the esophagogastric junction and moving up the esophagus in a helical fashion. The concept is that acute bleeding is controlled by compression of the bleeding varix by submucosal swellings produced by the injections, and that in the long-term, the overlying mucosa is thickened, thereby preventing further bleeding while preserving deeper variceal channels. A combination of intravariceal and paravariceal injection sclerotherapy is depicted in Figure 1C. This was first proposed by the author's group [7] and Soehendra et al [8] of Germany, with the aim of combining the advantages of the former two techniques and in the hope that this would provide a summation of these advantages. My group currently uses the combined technique in the management of acute variceal bleeding and an intravariceal technique for repeated sclerotherapy. The one trial that compared intravariceal and paravariceal sclerotherapy used a relatively unusual agent, absolute alcohol, and demonstrated that the intravariceal technique was superior [9]. Kitano et al [10] have recently proposed a novel three-part sclerotherapy management protocol: thrombosis is induced in the varices by intravariceal injections; this is followed by removal of the esophageal mucosa by repeated paravarieeal injections; and finally reepithelialization is allowed to occur. They have reported remarkable results with the use of ethanolamine
THE AMERICAN JOURNAL OF SURGERY VOLUME 160 JULY 1990
37
TERBLANCHE
A
:.*:
c
D
Figure 1. A flexible endoscope is used for intravariceal injection (A), paravariceal (submucosal) injection (B), and combined paravariceal and intravariceal injection (C), whereas a rigid esophagoscope with a slot is used for intravariceal injection (D). Reprinted from [2], with permission from the publisher.
oleate for this three-phase treatment in a group of 141 patients. There were no recurrent varices and only minimal early recurrent bleeding at a median follow-up of 14.6 months. I would, however, advocate caution in considering this approach until further controlled trials have shown it to be effective and safe in other groups of patients. The technique could be hazardous with certain sclerosing agents, such as sodium tetradecal sulfate, particularly if deep ulceration had occurred prior to thrombosis of the varices, which could result in massive uncontrolled bleeding. Timing of selerotherapy: The timing of intravariceal sclerotherapy in assessing long-term management has been subjected to two controlled trials [11,12]. Both concluded that injection treatments should be performed at least at weekly intervals until the varices are eradicated. Many groups, including our own, have failed to adhere to a weekly regimen, largely due to logistic considerations. Our 10-year analysis confirmed this to be a problem [13]. With paravariceal sclerotherapy, repeat endoscopy has been performed at weekly or more frequent intervals, although this has not been subjected to controlled trial analysis. The timing of the combined technique still has to be subjected to properly controlled trials. Sclerosant agents: The most widely used agent for intravariceal and combined techniques is 5% ethanolamine oleate. Polidocanol has been the main agent used 38
by the majority of advocates of the paravariceal technique and by some for the combined technique. It has been difficult to prove which agent is better since the characteristics of agents required for successful intravariceal or paravariceal techniques clearly vary. Several controlled trials have been reported. Ethanolamine oleate was shown to be superior to polidocanol [14] and sodium tetradecal sulfate [15] using one technique, whereas polidocanol was superior to absolute alcohol using a different injection technique [16]. Combination therapy: Although sclerotherapy plays a major role in patients with variceal bleeding, there is increasing evidence that it should not be the only treatment considered and used. Dean Warren's excellently executed controlled trial demonstrated that the combination of sclerotherapy with shunt salvage, using the Warren shunt in cases of sclerotherapy failure (31%), was superior to the use of the Warren shunt alone in longterm management after a variceal bleeding episode [1]. Although two other controlled trials have failed to confirm this finding, trial design and technical differences probably account for this discrepancy [17,18]. An important problem with sclerotherapy management is that patients continue to have major variceal bleeding until their varices are eradicated. The use of pharmacologic therapy as an adjunct during this phase might prove useful because both sclerotherapy and propranolol have been shown to be equally effective when compared in controlled trials of long-term therapy [19]. In one controlled study, early rebleeding was significantly reduced when propranolol was administered for 14 days, commencing 24 hours after the initial bleeding varices had been successfully treated [20]; however, further combination therapy trials are required. Anatomic and physiologic factors: Recent anatomic and physiologic studies have provided new data and understanding. Has this altered our management or our concept of why sclerotherapy works? The answer is a qualified yes. Varices almost invariably bleed from the lower 5 cm of the esophagus. The reason for this is not yet clearly understood, although it is better explained by recent studies. Bleeding probably results from internal disruption or explosion of the varices as a result of increased pressure from within the varix or because of thinning of the overlying vascular supporting structures, and is not caused by external erosion from the lumen of the esophagus by acid-pepsin digestion or by the trauma of swallowing solids. In an analysis of recent anatomic studies, our group has postulated why sclerotherapy is successful [21]. Prediction of variceal bleeding: This has proved to be one of the most controversial areas. Most groups have found it exceedingly difficult to predict which patient will bleed, either for the first time or after previous variceal bleeding. It has also proved difficult to define which patients with acute variceal bleeding will stop active bleeding with simple measures, which patients will continue to bleed, and which patients will have an early recurrence after bleeding has been controlled. The most widely proclaimed predictive index has been the complex endoscopic classification of varices proposed by Beppu et al [22] of
THE AMERICANJOURNALOF SURGERY VOLUME160 JULY1990
SCLEROTHERAPY FOR VARICEAL BLEEDING
Japan. However, in a controlled trial, they were only able to predict a first episode of variceal bleeding in 20% of their control patients [23]. Therefore, in the treated group in this trial, 80% of the patients were, by implication, subjected to unnecessary major surgery. The simple endoscopic and coagulation signs used by Paquet [24] to predict variceal hemorrhage in German patients proved more accurate, with a 66% predictive value of a first episode of variceal bleeding. This is probably partly the result of different types of patients being included in the studies. These discrepancies have led various groups to search for combined clinical and endoscopic findings that would be more accurate. The North Italian Endoscopic Club [25] has proposed such a predictive index, but their index requires further evaluation in controlled clinical trials for both its value in prophylactic therapy and in assessment as a predictor of further variceal bleeding in patients being considered for therapy after a bleeding episode. Etiologic and geographic variations: There are major variations in the cause of the underlying liver disease and in the severity of liver disease in different geographic areas, and therefore, by implication, variations in the chance of a patient dying. This has caused major problems in comparing studies from different areas and in making recommendations on the basis of such studies. I believe that meta-analysis of several randomized trials from disparate areas, like that published by Infante-Rivard et al [26], is difficult to justify because of these geographic and etiologic variations. Cause of death: Patients with portal hypertension and variceal bleeding due to underlying liver disease die either as a result of continued active variceal bleeding or as a result of deteriorating liver function and liver failure. Whereas bleeding frequently aggravates the liver failure, the converse is also true; this leads to a vicious cycle, which ends in death. It is unusual for patients to die of active variceal bleeding today because sclerotherapy has markedly improved management [27], but the underlying liver disease, nevertheless, determines the prognosis. Few patients with severe underlying liver disease survive 5 to 10 years, despite the use of repeated sclerotherapy
[13]. SCLEROTHERAPY FOR ACUTE VARICEAL BLEEDING All patients with suspected acute variceal bleeding require immediate hospital admission and expert management, preferably in an intensive care unit. Resuscitation includes crystalloids and blood component therapy, particularly for massive hemorrhage. An unanswered question is whether such patients should be treated in specialized portal hypertension units or whether they can be as successfully treated by primary-care physicians, gastroenterologists, or surgeons in local or district hospitals. I believe that such patients should be transferred to a specialized center and placed under the care of a combined team with particular expertise in portal hypertension management whenever possible, because they may require more than one form of complex therapy. The diagnosis of variceal bleeding must be confirmed
by emergency endoscopy, because at least 30% of patients in whom it is suspected do not have Varices, and up to one third of patients with varices will be bleeding from another source [28,29]. The role of pharmacologic therapy remains controversial. My views have been expressed in recent publications [2,3,30]. My co-workers and myself advocate intravenous infusion of vasopressin (0.4 units/min) plus sublingual nitroglycerine (one tablet each hour for 6 hours). Balloon tamponade is also controversial but should not be. If active variceal bleeding is not controlled pharmacologically or by immediate sclerotherapy, the patient must have a balloon tube inserted. Correct insertion will invariably control the bleeding; thereby allowing time to resuscitate the patient and to summon expert help for subsequent definitive emergency m a n a g e m e n t [2,29]. Definitive therapy should occur within 6 to 12 hours, at the time that the balloon tube is removed. Sclerotherapy is currently the mainstay of treatment for acute variceal bleeding; however, the role of immediate or urgent shunting or esophageal transection is being reevaluated. A policy of initial measures to control hemorrhage, followed by urgent sclerotherapy, has been the most widely practiced regimen in the past decade. Three controlled trials [31-33] and innumerable uncontrolled studies [2] have confirmed that sclerotherapy controls acute variceal bleeding in 90% to 95% of patients. Sclerotherapy has also been shown to be more effective than either balloon tamponade or vasopressin used alone. A single injection treatment will control variceal bleeding in more than 70% of patients; therefore, only 30% of patients require further sclerotherapy or other treatment [29]. A second injection will increase the control rate to 90% or 95%. Unfortunately, various published trials are difficult to compare because of differences in design. There has been a shift in emphasis from delayed emergency sclerotherapy to the use of sclerotherapy during the first emergency endoscopy. This view is supported by two controlled trials [34,35] and is my recommended therapy. Immediate sclerotherapy requires a high degree of skill when attempted in patients with active bleeding, but it is possible if the correct technique is used. Immediate sclerotherapy Should also be undertaken in patients with varices that have stopped bleeding, as some of these patients would otherwise have further major life-threatening bleeding before subsequent sclerotherapy could be performed. We have defined failure of emergency sclerotherapy as an acute episode ofvariceal bleeding that occurs within hours or days of emergency injection sclerotherapy and that requires blood transfusion and subsequent emergency management [2]. We and others have demonstrated that emergency sclerotherapy, using one or two injection treatments, is successful in controlling acute variceal bleeding in 90% to 95% of patients. The 5% to 10% of patients in whom sclerotherapy fails must be considered for other forms of management. Analysis of our first prospective evaluation of tamponade and sclerotherapy showed that the mortality approached 90% in those few patients who had further episodes of acute variceal bleeding after two sclerotherapy treatments during a single
THE AMERICANJOURNAL OF SURGERY VOLUME160 JULY 1990 39
TERBLANCHE
hospital admission when low-risk Child's class A patients were excluded [36]. Therefore, bleeding in such patients must be controlled temporarily with balloon tamponade followed by either a portosystemic shunt or an esophageal transection [2,3,36]. Unfortunately, during the initial evaluation, we have been unable to predict which patients will not respond to sclerotherapy. If such a prediction was possible, selected patients could undergo immediate surgery. This leads to the important question, "Is there a place for immediate or early shunting or esophageal transection?" Unfortunately, the answer is not yet known. The role of emergency shunt procedures has received increasing support recently. Orloff et al [37] have championed emergency portacaval shunts over the years. In an important study, Cello et al [38] demonstrated that portacaval shunts were more effective than sclerotherapy in preventing early and late recurrences of bleeding in a group of high-risk patients, but there was no difference in survival between the shunted and control groups, There are also advocates of the early use of surgical transection. On the basis of their controlled trial, Burroughs et al [39], from the Royal Free Hospital in London, have concluded that emergency stapel transection is as safe as sclerotherapy and, in their hands, is more effective than a single sclerotherapy procedure. However, their recommendation is that a transection should be offered after two injection treatments have failed. This is the identical conclusion reached earlier by our group [36]. I predict that either emergency portacaval shunting or emergency esophageal transection will be performed more frequently in the future for management of acute variceal bleeding. S C L E R O T H E R A P Y FOR LONG-TERM MANAGEMENT Sclerotherapy is one of the available options in longterm management after an episode of variceal bleeding, but is it the best primary option for the majority of patients? This is a difficult question, but Warren's trial suggested that sclerotherapy with shunt salvage was better than the use of the Warren selective distal splenorenal shunt alone [1]. Note that, with the exception of hepatic transplantation, no other therapy actually treats the underlying liver disease. Furthermore, some forms of therapy, such as shunting, may worsen the liver disease. The best management may hinge on the underlying causes of portal hypertension, which differ in various parts of the world. In Western countries, alcoholic cirrhosis and certain types of nonalcoholic cirrhosis are common. Emphasis will be placed on the management of these patients. In addition to repeated injection sclerotherapy and liver transplantation, the other options are conservative medical management with close follow-up to detect an additional bleeding episode and adminstration of specific therapy should it recur; portosystemic shunts; transection and devascularization; and long-term pharmacoligic therapy [2,3]. The latter alternative remains highly controversial. The problems of lifelong therapy and compliance compound the difficulty of assessing rec-
40
ommendations in the literature. I believe that the use of drugs should currently be restricted to major institutions with a special interest in portal hypertension. All patients who have had variceal bleeding should be considered for hepatic transplantation. This is the only therapy that cures the underlying liver disease as well as the portal hypertension. If patients are potential transplant candidates, the therapy used should not interfere with a subsequent liver transplant. We therefore recommend sclerotherapy as the primary treatment for such patients. If sclerotherapy fails, we prefer simple esophageal transection to shunting. The five completed controlled clinical trials, and innumerable uncontrolled studies, have demonstrated that repeated sclerotherapy effectively eradicates esophageal varices in most patients [40-44]. After eradication, the incidence of recurrent variceal bleeding is markedly reduced, although it still occurs frequently until variceal eradication is achieved. The ability of repeated sclerotherapy treatments to increase survival has yet to be determined. Only one of the five controlled trials [42] showed a clear improvement in survival with sclerotherapy. There are several important problems related to repeated sclerotherapy in long-term management. Varices are difficult to eradicate in some patients, whereas others continue to have life-threatening bleeding during the period of variceal eradication. Long-term sclerotherapy should not be continued when varices cannot be eradicated or bleeding episodes continue to recur. Instead, surgery should be performed at an early stage [3]. This combination of sclerotherapy with surgical salvage if treatment fails, as recommended by Warren [1], is probably the best long-term strategy for the majority of patients. Other problems with long-term sclerotherapy include the need for lifelong follow-up with repeated injections because varices recur in time. Surviving patients place an increasing burden on hospital resources even when sclerotherapy can be performed on an outpatient basis. Although the complications of sclerotherapy do not usually pose a major problem, they are cumulative with time [45]. The complications in the 10-year Cape Town experience have been detailed in recent publications [13,27,45]. It is important to attempt to identify those patients in whom sclerotherapy is likely to ultimately fail at an early stage of management and use surgical treatment as an alternative. An important, but seldom discussed, subset of patients is the high-risk group with end-stage liver disease, particularly alcoholic cirrhosis, in whom an early definitive decision must be made not to treat the patient. This is difficult, but should be seriously entertained; otherwise, hospital resources will be fruitlessly utilized on patients whose condition is not salvageable. In conclusion, I believe that sclerotherapy has altered the management of patients with portal hypertension and has led to more rational therapy. However, it has probably been overused in the past decade to the detriment of patient care in some centers, including our own. Sclero-
THE AMERICAN JOURNAL OFSURGERY VOLUME160 JULY 1990
SCLEROTHERAPY FOR VARICEAL BLEEDING
therapy is likely to remain the initial therapy of choice for most patients, but treatment failure must be identified early and the patient treated by a portosystemic shunt or a transection and devascularization procedure. REFERENCES 1. Warren WD, Henderson JM, Millikan W J, et al. Distal splenorenal shunt versus endoscopic sclerotherapy for long-term management of variceal bleeding: preliminary report of a prospective, randomized trial. Ann Surg 1986; 203: 454-62. 2. Terblanche J, Burroughs AK, Hobbs KEF. Controversies in the management of bleeding esophageal varices. N Engl J Med 1989; 320i 1393-8 and 1469-75. 3. Terblanche J. The surgeon's role in the management of portal hypertension. Ann Surg 1989; 209: 381-95. 4. Anonymous. Prophylactic sclerotherapy of oesophageal varices; is it justified? Lancet 1988; 1: 1369-70. 5. Bornman PC, Kahn D, Terblanche J, Worthley C, Spence RA, Krige JJ. Rigid versus fiberoptic endoscopic injection sclerotherapy: a prospective randomized controlled trial in patients with bleeding esophageal varices. Ann Surg 1988; 208: 175-8. 6. Paquet K-J. Endoscopic paravariceal injection sclerotherapy of the esophagus--indications, technique, complications: results of a period of 14 years. Gastrointest Endosc 1983; 29: 310-5. 7. Terblanche J, Bornman PC, Jonker MAT, Kirsch RE, Saunders SJ. Injection sclerotherapy of esophageal varices. Sem Liver Dis 1982; 2: 233-41. 8. Soehendra N, de Heer K, Kempeneers I, Runge M. Sclerotherapy of esophageal varices: acute arrest of gastrointestinal hemorrhage or long-term therapy? Endoscopy 1983; 15: 136-40z 9. Sarin SK, Nanda R, Sachdev G, Chari S, Anand BS, Broor SL. Intravariceal versus paravariceal sclerotherapy: a prospective, controlled, randomised trial. Gut 1987; 28: 657-62. 10. Kitano S, Koyanagi N, Iso Y, Higashi H, Sugimachi K. Prevention of recurrence of esophageal varices after endoscopic injection sclerotherapy with ethanolamine oleate. Hepatotogy ;1987; 7: 810-5. 11. Westaby D, Melia WM, Macdougall BR, Hegarty JE, Williams R. Injection sclerotherapy for oesophageal varices: a prospective randomised trial of different treatment schedules. Gut 1984; 25: 129-32. 12. Sarin SK, Sachdev G, Nanda R, Batra SK, Anand BS. Comparison of the two time schedules for endoscopic sclerotherapy: a prospective randomised controlled study. Gut 1986; 27: 710-3. 13. Terblanche J, Kahn D, Bornman PC. Long-term injection sclerotherapy treatment for esophageal varices: a 10-year prospective evaluation. Ann Surg 1989; 210: 725-31. 14,. Kitano S, Iso Y, Koyanagi N, Higashi H, Sugimachi K. Ethanolamine oleate is superior to polidoeanal (aethoxysklerol) for endoscopic injection sclerotherapy of esophageal varices: a prospective randomized trial. Hepatogastroenterology t987; 34: i9-23. 15. Kitano S, Iso Y, Yamaga H, Hashizume N, Higashi H, Sugimachi K. Trial of sclerosing agents in patients with oesophageal varices. Br J Surg 1988; 75: 751-3. 16. Atamkuri SP, Bhargava DK, Sharma MP. Endoscopic sclerotherapy for esophageal varices: a prospective, randomised trial of absolute alcohol versus polidocanol. Indian J Gastroenterol 1988; 7: 87-9. 17. Rikkers LF, Burnett DA, Volentine GD, Buchi KN, Cormier RA. Shunt surgery versus endoscopic sclerotherapy for long-term treatment of variceal bleeding: early results of a randomized trial. Ann Surg 1987; 206: 261-71. 18. Teres J, Bordas JM, Bravo D, et al. Sclerotherapy vs. distal splenorenal shunt in the elective treatment of variceal hemorrhage: a randomized controlled trial. Hepatology 1987; 7: 430-6. 19. Fleig WE, Stange EF, Hunecke R, et aL Prevention of recur-
rent bleeding in cirrhotics with recent variceal hemorrhage: prospective, randomized comparison of propranolol and sclerotherapy. Hepatology 1987; 7: 355-61. 20. Marbet UA, Straumann A, Gyr KE, et al. Reduction in early recurrence of variceal bleeding by propranolol. Seand J Gastroenterol 1988; 23: 369-74. 21. Spence RAJ, Terblanche J. Venous anatomy of the lower oesophagus: a new perspective on varices. Br J Surg 1987; 74: 65960. 2 2 . Beppu K, Inokuchi K, Koyanagi N, et al. Prediction of variceal hemorrhage by esophageal endoscopy. Gastrointest Endosc 1981; 27: 213-8. 23. Inokuchi K. Prophylactic portal nondecompression surgery in patients with esophageal varices: an interim report. Ann Surg 1984; 200: 61-5. 24. Paquet K-J. Prophylactic endoscopic sclerosing treatment of the esophageal wall in varices: a prospective controlled trial. Endoscopy 1982; 14: 4-5. 25. The North Italian Endoscopic Club for the Study and Treatment of Esophageal Varices. Prediction of the first variceal hemorrhage in patients with cirrhosis of the liver and esophageal varices: a prospective multicenter study. N Engl J Med 1988; 319: 983-9. 26. Infante-Rivard C, Esnaola S, Villeneuve JP. Role of endoscopic variceal sclerotherapy in the long-term management of variceal bleeding: a meta-analysis. Gastroenterology 1989; 96: 1087-92. 27. Kahn D, Bornman PC, Terblanche J. A 10-year prospective evaluation of balloon tube tamponade and emergency injection sclerotherapy for actively bleeding oesophageal varices. HPB Surg 1989; 1: 207-19. 28. Novis BH, Duys P, Barbezat GO, Clain J, Bank S, Terblanche J. Fiberoptic endoscopy and the use of the Sengstaken tube in acute gastrointestinal haemorrhage in patients with portal hypertension and varices. Gut 1976; 17: 258-63. 29. Terblanche J, Yakoob HI, Bornman PC, et al. Acute bleeding varices: a five-year prospective evaluation of tamponade and sclerotherapy. Ann Surg 1981; 194: 521-30. 30. Anonymous. Management of acute variceal bleeding. Lancet 1988; 2: 999-1000. 31. Barsoum MS, Bolous FI, E1-Rooby AA, Rizk-Allah MA, Ibrahim AS. Tamponade and injection sclerotherapy in the management of bleeding oesophageal varices. Br J Surg 1982; 69: 76-8: 32. Paquet K-J, Feussner H. Endoscopic sclerosis and esophageal balloon tamponade in acute hemorrhage from esophagogastric varices: a prospective controlled randomized trial. Hepatology 1985; 5: 580-3. 33. Larson AW, Cohen H, Zweiban B, et al. Acute esophageal variceal sclerotherapy: results of a prospective randomized controlled trial. JAMA 1986; 255: 497-500. 34. Westaby D, Hayes PC, Gimson AES, Poison R J, Williams R. Controlled clinicaltrial of injection sclerotherapy for active variceal bleeding. Hepatology 1989; 9: 274-7. 35. Moreto M, Zaballa M, Berne A, Ibanez S, Ojembarrena E, Rodriguez A. A randomized trial of tamponade or sclerotherapy as immediate treatment for bleeding esophageal varices. Surg Gynecol Obstet 1988; 167: 331-4. 36. Bornman PC, Terblanche J, Kahn D, Jonker MA, Kirsch RE. Limitations of multiple injection sclerotherapy sessions for acute variceal bleeding. S Afr Med J 1986; 70: 34-6. 37. Orloff M J; Bell RH Jr, Hyde PV, Skivolocki WP. Long-term results of emergency portacaval shunt for bleeding esophageal varices in unselected patients with alcoholic cirrhosis. Ann Surg 1980; 192: 325-40. 38. Cello JP, Grendell JH, Crass RA, Weber TE, Trunkey DD. Endoscopic sclerotherapy versus portacaval shunt in patients with severe cirrhosis and acute variceal hemorrhage: long-term followup. N Engt J Med 1987; 316: 11-5. 39. Burroughs AK, Hamilton G, Phillips A, Mezzanotte G, Mclntyre N, Hobbs KEF. A comparison of sclerotberapy with staple
THE AMERICAN JOURNAL OFSURGERY
VOLUME 160 JULY 1990
41
TERBLANCHE
transection of the esophagus for the emergency control of bleeding from esophageal varices. N Engl J Med 1989; 321: 857-62. 40. Terblanche J, Bornman PC, Kahn D, et al. Failure of repeated injection sclerotherapy to improve long-term survival after oesophageal variceal bleeding: a five-year prospective controlled clinical trial. Lancet 1983; 2: 1328-32. 41. The Copenhagen Esophageal Varices and Sclerotherapy Project. Sclerotherapy after first variceal hemorrhage in cirrhosis: a randomized multi-center trial. N Engl J Med 1984; 311: 1594-600. 42. Westaby D, Macdougall BR, Williams R. Improved survival following injection sclerotherapy for esophageal varices: final anal-
42
ysis of a controlled trial. Hepatology 1985; 5: 827-30. 43. Korula J, Balart LA, Radvan G, et al. A prospective, randomized controlled trial of chronic esophageal variceal sclerotherapy. Hepatology 1985; 5: 584-9. 44. Soderlund C, Ihre T. Endoscopic sclerotherapy v. conservative management of bleeding oesophageal varices: a 5-year prospective controlled trial of emergency and long-term treatment. Acta Chir Scand 1985; 151: 449-56. 45. Kahn D, Jones B, Bornman PC, Terblanche J. Incidence and management of complications after injection sclerotherapy: a 10year prospective evaluation. Surgery 1989; 105: 160-5.
THE AMERICAN JOURNAL OF SURGERY VOLUME 160 JULY 1990