0016-5107/90/3606-0567$02.00 GASTROINTESTINAL ENDOSCOPY Copyright © 1990 by the American Society for Gastrointestinal Endoscopy

Prospective randomized comparison of esophageal variceal sclerotherapy agents: sodium tetradecyl sulfate versus sodium morrhuate Stephen A. McClave, MD, Steven C. Kaiser, MD Richard A. Wright, MD, James L. Edwards, MD Kenneth R. Kranz, MD Louisville, Kentucky

We designed a prospective randomized study to evaluate differences in efficacy and complication rate between the two most commonly used sclerosing agents, sodium tetradecyl sulfate (STD) and sodium morrhuate (MOR). Of 41 patients with acute index variceal bleeding initially evaluated, 21 were randomized to receive 0.75% STD and 20 to receive 1.6% MOR diluted with 50% dextrose. Overall mortality for the STD group was 38% and that for the MOR group was 25% (NS). Control of acute bleeding was achieved in 86% of the STD patients and 90% of the MOR patients (NS). Overall, obliteration was achieved in only 33% of the STD group and 25% of the MOR group; but in those patients who remained in the study over 3 months, obliteration was achieved in 87 and 83%, respectively (NS). There was a trend toward higher rebleeding in the STD group compared with the MOR group (81% vs. 51%) (p=0.078), but there was no significant difference between the STD and MOR patients with regard to transfusion requirements (8.4 units/patient vs. 8.7 units/patient), ulceration (53% vs. 40%), or stricture formation (9.5% vs. 0.0%). The results of this study suggest that STD and MOR are clinically equivalent sclerosing solutions, and that bias favoring use of one agent over the other may be unfounded. (Gastrointest Endosc 1990; 36:567-571)

A number of factors impact on the success of esophageal variceal sclerotherapy in controlling acute bleeding, preventing rebleeding, and achieving obliteration of esophageal varices. Few of these factors, including type and concentration of sclerotherapy agents, volume and pattern of injections, and schedule of sessions, have been studied prospectively in a controlled manner. 1-8 Five agents appear in the literature as potential variceal sclerosants; polidocanol, sodium tetradecyl (STD), sodium morrhuate (MaR), ethanol, and ethanolamine 0Ieate. 9 - 13 A number of studies, both human l - 5 and animal,I4-17 have shown differences in efficacy and complication rates between these agents; but most of the human studies (particularly with MaR

Received March 28,1990. For revision June 10, 1990. Accepted July 25,1990. From the Department of Medicine, University of Louisville School of Medicine, Louisville, Kentucky. Reprint requests: Stephen A. McClave, MD, Department of Medicine, University of Louisville, School of Medicine, Louisville, Kentucky 40292. VOLUME 36, NO.6, 1990

and STD) are retrospective/8 .1 9 or only in abstract form. 20- 24 We designed a study using dilute concentrations of the two most commonly used single agents (STD and MaR) to evaluate efficacy and complication rate in order to select the better of the two sclerosing agents. METHODS

Only those patients with initial "index" esophageal variceal bleeding, admitted to Humana Hospital University of Louisville or Louisville Veterans Administration Medical Center, were included in this study. All patients entered in the study presented with a major upper gastrointestinal bleed, were endoscoped within 6 hours of admission, and were found to have esophageal varices with no other obvious source of bleeding (although they may not have been actively bleeding at the time of endoscopy). Patients were excluded for failure to obtain informed consent, if the bleeding source was found to be nonvariceal, or if the patient received multiple sclerotherapy sessions at outlying hospitals. This study was approved by the Committee for Human Investi567

gation (at each of the two hospitals) in January 1986. Informed consent was obtained from all patients. Patients were randomized at the time of initial bleed to receive either STD 0.75% or MOR 1.6%, achieved by diluting the agents with 50% dextrose. Sclerotherapy was performed with the "free hand" technique, with all injections attempted to be intravariceal. Retractable 25-gauge sclerotherapy needles (Microvasive Corp., Watertown, Mass.) were used with either single channel or double channel gastroscopes (Olympus Corporation XQlO or 2TlO, Lake Success, N.Y.). Injections of 2 to 3 ml each were made circumferentially beginning in the distal esophagus near the esophago-gastric junction and proceeding cephalad. One injection was made per varix, unless additional injections were deemed necessary by the endoscopist. A total of 20 to 24 ml of sclerosant was injected per session. No over-the-scope balloons or overtubes were used. All procedures were performed with the patient placed in the left lateral decubitus position. Patients received topical anesthetic spray to the oropharynx and mild preoperative sedation with intravenous meperidine, midazolam, or diazepam. Varices were sclerosed weekly for a total of three sessions, beginning at the time of initial bleed, and then continued monthly thereafter to obliteration. Obliteration was determined subjectively by the endoscopist on the basis of size and appearance (as well as the resistance to injection) ofthe remaining varices. A number of factors were recorded and followed prospectively, including overall appearance and percent obliteration of the remaining varices, pattern of injections, complication rate (ulcers and strictures), mortality rate, and transfusion requirements. Ulcers were defined as depressions in the mucosal surface with overlying exudate. If an ulcer was noted at a sclerotherapy session, the period from the previous session (with no ulcer) to the subsequent session (with ulcer healed) was evaluated. If bleeding occurred during this period, the ulcer was considered for purposes of this study to be related to recurrent bleeding (even if active bleeding was not seen at endoscopy). Gastric varices were defined by that component of the varices which extended below the squamocolumnar mucosal junction onto the gastric fundus. Statistical evaluation was done with chi-square analysis of tabulated data, using Yates continuity correlation when evaluating only two variables.

mean of 7.97 months (median 3.0 months, range 1 to 41 months). Nineteen patients in each group were male, with the average age for the STD group 49.4 years, and for the MOR group 54.1 years. The etiology of chronic liver disease in these patients included Laennec's cirrhosis in 32, cryptogenic cirrhosis in 4, primary biliary cirrhosis in 3, and primary sclerosing cholangitis and portal vein thrombosis in 1 each. Child's classification showed that 10 patients were class A, 23 patients were class B, and 8 patients class C, with all three classes equally distributed between the STD and MOR groups (p = 0.326) (Table 1). Efficacy of these two agents, based on mortality rate, control of index bleed, and rate of obliteration, appeared to be equal. The mortality rate (over the 3.5year study period) for the STD group was 38%, and that of the MOR group was 25%, a difference which was not statistically significant (p = 0.572). Mortality was directly related to rebleeding in eight patients (five STD, three MOR). In six other patients, mortality was related to underlying liver disease and concomitant organ failure (respiratory four, renal one, cardiac one) without bleeding. Control of index bleed was achieved in 86% of the STD patients, and 90% of the MOR patients. Failure to control index bleed occurred in five patients (three STD, two MOR). All five died despite additional medical therapy in four (Pitressin, Sengstaken-Blakemoore tube, repeat sclerotherapy), and surgical shunt placement in one. Obliteration rate for the entire study group was poor, with only 7 of the STD patients (33%) and 5 ofthe MOR patients (25%) achieving obliteration. However, in those patients who remained in the study over 3 months, obliteration was achieved in 87% of the STD patients (7 of 8) and 83% of the MOR patients (5 of 6), a difference which was not statistically significant (p = 0.924) (Table 2). Twenty-eight patients remained in the study less than three months. Of those, 11 sustained early death, 6 refused further therapy, 3 relocated to distant medical centers, 1 received liver transplant, and 7 were late

RESULTS

Of 47 patients initially evaluated for participation in the study, 6 patients were excluded for errors in diagnosis and violation of the study protocol. Of the 41 patients included in the study, 21 were randomized to received STD and 20 were randomized to receive MOR. A total of 202 sessions were performed on these patients (96 sessions STD, 106 MOR), with an overall average of approximately 5 sessions per patient (4.5 sessions/patient STD, 5.3 MOR). Fourteen patients in each group received at least three or more sessions, with the STD group averaging 16.2 ml of sclerosant per session, and the MOR group receiving 18.2 ml of sclerosant per session. Patients were followed for a 568

Table 1. Patient epidemiology

Tetradecyl Male/female Average age (yr) Etiology: liver disease ETOH PBC Other Child's class

(N = 21)

Morrhuate (N = 20)

19/2

19/1

49.4

54.2

16

16

2

1 3

3

A

4

6

B C

12

11

5

3

GASTROINTESTINAL ENDOSCOPY

Table 2. Efficacy

Mortality rate Control index bleed Obliteration After 3 mo Sessions to obliteration

Tetradecyl (N = 21)

Morrhuate (N = 20)

38%

25% 90% 25% 83% (5/6) 9.6

86%

33% 87% (7/8) 9.2

p

0.572 0.953 0.924

entries (still receiving treatment at the close of the study). The complication rates between these two agents likewise appeared to be equal. Although there was a trend toward a higher rebleeding rate in the STD group compared with the MOR group (81% vs. 50%), this difference did not reach statistical significance (p = 0.078). Over 70% ofthe rebleedingepisodes occurred prior to or including the third session, and only 8 of the 41 study patients rebled after the fourth sclerotherapy session. Seventy-five percent ofthe rebleeding episodes (30 of 40) were felt to be due to recurrent variceal hemorrhage. In 10 of the 40 episodes, the presence of an esophageal post-sclerotherapy ulcer, duodenal ulcer, severe gastritis, or Mallory-Weiss tear mayor may not have accounted for the rebleeding. The total number of packed red blood cell units transfused was exactly equal, at 175 total units transfused for each group. This averaged 8.4 units per patient in the STD group, and 8.7 units per patient in the MOR group. Post-sclerotherapy ulcers occurred in 53 % of the STD group (13 separate ulcer episodes in 11 patients), and 40% of the MOR group (10 separate ulcer episodes in 8 patients), a difference which was not statistically significant (p = 0.630). Overall, the ulcers tended to be shallow and superficial, and were not clinically significant as evidenced by the fact that only a small percentage appeared to be related to rebleeding episodes. Of 23 separate ulcer episodes, 7 were considered to possibly be related to recurrent bleeding. Of those seven, however, only two ulcers were documented to be bleeding at the time of endoscopy. In the other five episodes, ulcers were present at the time of rebleeding, but Mallory-Weiss tears or clots on varices suggested a bleeding site other than the ulcers. Sixteen of the ulcer episodes were clearly not related to recurrent bleeding, as the ulcers appeared and then healed without clinical consequence. In 21 of these 23 episodes, further sclerotherapy was performed. No bleeding occurred from the time of sclerotherapy to ulcer healing in all 21. Post-sclerotherapy strictures occurred in only two of the STD patients (9.5%) and none of the MOR group, again, a difference that was not statistically significant (p = 0.490) (Table 3). Both patients were treated easily with Maloney dilation. VOLUME 36, NO.6, 1990

Various parameters, including gastric varices, red color markings, and Child's classification, were evaluated for their ability to distinguish those patients in the study most likely to rebleed or die. The presence of gastric varices (or the persistence of gastric varices despite sclerosis and obliteration of esophageal varices), failed to predict either rebleeding or mortality. Sixty-seven percent (14 of 21) of those patients with positive gastric varices rebled, compared with 65% (13 of 20) of those patients without gastric varices. Mortality rate was 28 and 35%, respectively, in those patients with and without gastric varices. Red color marks (cherry red spots or red wale marks) surprisingly did not predict rebleeding in our study. Although 75% of those patients with red color marks rebled (12 of 16), this was not significantly different from the 60% of patients without red color marks who rebled (15 of 25) (p = 0.515). As would be expected, red color markings did not predict mortality, as 31 % of those patients with positive markings and 32% of those without markings died in our study. Although Child's classification showed a strong clinical trend toward predicting mortality, the difference between classes did not react statistical significance. Mortality rates were 10% in class A, 30% in class B, and 63% in class C (p = 0.057). Rebleeding occurred in 50% of the Child's class A patients, 60% in class B, and 75% in class C, differences again which were not significant (p = 0.459). DISCUSSION

Although a number of "recipes" for esophageal variceal sclerosant solutions have been reported in the literature, no single agent or combination of agents has emerged as the clear-cut leader in terms of safety, efficacy, and low complication rates. There is little data to support that a combination of agents is superior to any single agent, and the role of certain additives is questionable. Thrombin, a frequent additive, has been shown to be inactivated when mixed with dextrose, morrhuate, or tetradecyl.ll Cephalothin, another additive, is probably more valuable for its irritating properties than for its antibiotic effect. ll Dextrose at 50% and normal saline have been shown to Table 3. Complications

Rebleeding :53 sessions 2:4 sessions PRBC units transfused Average units/patient Ulcers Stricture

Tetradecyl (N = 21)

Morrhuate (N = 20)

81% 14 3 175 8.3 53% 9.5%

50% 5 5 175 8.7 40% 0.0%

p

0.078

0.630 0.490

569

be weak sclerosants,17 and are probably used more for their dilutional effect when added to other agents. In general, the fat soluble sclerosing agents (ethanol, morrhuate, ethanolamine) have a greater incidence of systemic side effects such as fever, pleural effusion, chest pain, and ARDS, than the more water soluble agent (tetradecyl).l0 A number of studies have shown that diluting the concentration of the primary sclerosant may reduce the risk of ulceration and stricture formation. 21 ,22, 25 Unfortunately, there seems to be a correlation between sclerosing potential and rate of complications, and efforts to reduce the complication rate through dilution may result in diminished efficacy. Polidocanol has been used widely in Europe but is not currently available in the United State. Although excellent results have been achieved with paravariceal injection,26 intravariceal injection of this agent may result in significant systemic side effects. 11 , 12 In two prospective randomized trials, polidocanol was found to be less effective than ethanolamine in controlling acute bleeding and preventing rebleeding from esophageal varices. I, 3 Absolute ethanol has been used widely in India. Although once heralded as the "ideal sclerosant,"13 ethanol has had limited use in the United States due to local toxicity and excessive ulcerogenicity.9,15 Use of ethanol in combination with other agents has succeeded in reducing some of its side effects. 15 ,23 In a prospective randomized controlled trial, absolute alcohol was shown to be superior to 5% ethanolamine oleate with regard to control of acute variceal bleeding, cost, and earlier obliteration of varices (requiring fewer sessions and less volume of sclerosant). There was no difference between agents with regard to mortality, rebleeding, ulceration, stricture formation, or incidence of systemic side effects. 4 Diluting absolute ethanol to 50% succeeded in reducing pain and fever (with no difference in mortality, rebleeding, ulceration, or stricture formation) but resulted in significantly poorer control of acute bleeding and greater time to obliteration.5 Ethanolamine oleate has been used widely in Great Britain, South Africa, and Japan. 2 Although not used widely in the United States up to this time, ethanolamine oleate became the first agent in 1989 to be approved by the FDA for use in sclerosis of esophageal varices. Early animal studies suggested that 5% ethanolamine was a less effective sclerosant than 1%/3% STD and 5% MOR (although the ethanolamine was less ulcerogenic). 14, 16,17 Ethanolamine was noted to be more viscous and difficult to inject than either STD or MOR,14 Likewise, early human studies suggested that ethanolamine oleate was a less effective sclerosant than MOR, and at best, was equal to or inferior to STD. 1S,19 A more recent study supported these findings, showing that a greater number of sessions 570

were required to achieve obliteration with 5% ethanolamine compared with 3% STD (5.6 vs. 3.9 sessions, respectively).20 However, a recent prospective randomized study from Japan found 5% ethanolamine to be superior to 2% STD with regard to post-injection bleeding, late rebleeding, ulceration rate, disappearance of red color signs, and greater mucosal (not variceal) obliteration. There was no difference between agents in mortality or control of acute variceal bleeding. This latter study has been criticized though for the fact that 55% of the patients received esophageal variceal sclerosis prophylactically, having never bled from their esophageal varices. 2 A number of animal and human studies have compared MOR with STD. In early animal studies, 5% MOR was found to be an equally effective9 to slightly inferior sclerosant to 1.5% STDY Whereas MOR was found to be less ulcerogenic than STD/ 4,17 there was no difference found in systemic side effects. 15 Human studies, however, tended to contradict the findings from the animal studies. In comparison with 5% MOR, 3% STD was shown to have lower mortality (64% vs. 31 % ),23 and less systemic complications offever, chest pain, abdominal pain, and pleuritic reaction. 19,24 Whereas one study suggested greater ulcerogenicity with 5% MOR,24 three other studies found equal ulceration rates between the two agents. 21 ,22, 23 Overall, 5% MOR and 3% STD were found to be equivalent in efficacy,19,21 and rebleeding rates. 23 Modification of the primary sclerosant solution by dilution has been shown in some studies to reduce complications with only minimal loss of efficacy. Diluting to 1% STD with a final concentration of 33% ethanol in one study, significantly reduced ulceration and rebleeding rate. 23 Diluting STD from 3% to 1% with only sterile water or dextrose, has been shown to reduce the rate of ulceration 16,23 without significantly altering efficacy15,16 or rebleeding rate. 23 By lowering the concentration even further from 1% to 0.75% STD, one study succeeded in reducing the stricture rate from 66% to 15% without significant change in efficacy (obliteration) or rebleeding rate. 25 For purposes of this study, the concentrations of 0.75% STD and 1.6% MOR were chosen, indicated by the literature to have the lowest rate of ulceration and stricture (but still adequate efficacy).21,22,23,25 At these concentrations, the stricture rate was negligible with only two strictures occurring in the STD group. Although almost one-half of the patients in this study developed post-sclerotherapy ulcers (53% of the STD and 40% ofthe MOR group), the ulcers were clinically silent in that they were superficial and healed rapidly. The presence of the post-sclerotherapy ulcers correlated poorly with rebleeding and did not preclude further scleortherapy. However, despite adequate overall control of acute variceal bleeding, the concentrations used in this study were associated with a high GASTROINTESTINAL ENDOSCOPY

rebleeding rate and a questionably low obliteration rate compared with the literature. 27 Therefore, the concentrations used in this study may not be optimal for these agents. Further studies need to be performed to determine whether more concentrated mixtures (such as 1.5% STD or 2.5% MOR) would be more beneficial. As would be expected, Child's classification in this study seemed to correlate clinically to increased mortality (but not to rebleeding). Surprisingly, red color markings 28 did not predict rebleeding (or mortality). Although the role of gastric varices in the morbidity/ mortality of variceal bleeding is uncertain, documented bleeding from gastric varices is uncommon, occurring in less than 4 to 16% of cases. 29- 31 In this study, the presence of gastric varices showed no correlation to rebleeding or mortality. The results of this project indicate that STD and MOR used as esophageal variceal sclerosants, are clinically equivalent at the concentrations studied. In terms of efficacy, there was no significant difference between agents with regard to control of index bleed (86% STD vs. 90% MOR), rate of obliteration (33% STD vs. 25% MOR), or mortality (38% STD vs. 25% MOR). Rate of complications between 0.75% STD and 1.6% MOR likewise appeared to be equal in terms ofrebleeding (81% vs. 50%), ulceration (53% vs. 40%), stricture formation (9.5% vs. 0.0%), and transfusion requirements (8.4 units/patient vs. 8.7 units/patient), respectively. This study further alludes to the clinical trade-off that occurs with diluting concentrations, that the rate of ulceration and stricture formation is reduced at the expense of decreased sclerosant potential (for obliteration and control of bleeding). The major issue in choice of a sclerosing solution then may not be selection of one agent over the other, but finding the optimum concentration of either agent which reduces complications but maintains adequate efficacy.

REFERENCES 1. Balanzo J, Sainz S, Espinos JC, et al. Efficacy of ethanolamine and polidocanol in the eradication of esophageal varices. A prospective randomized trial. Endoscopy 1989;21:251-3. 2. Kitano S, Iso Y, Yamaga H, Hashizume M, Higashi H, Sugimachi K. Trial of sclerosing agents in patients with esophageal varices. Br J Surg 1988;75:751-3. 3. Kitano S, Iso Y, Koyanagi N, Higashi H, Sugimachi K. Ethanolamine oleate is superior to polidocanol (aethoxysklerol) for endoscopic injection sclerotherapy of esophageal varices: a prospective randomized trial. Hepatogastroenterology 1987;34:1923. 4. Sarin SK, Mishra SP, Sachdev GK, Thorat V, Dalal L, Broor SL. Ethanolamine oleate versus absolute alcohol as a variceal sclerosant: a prospective randomized, controlled trial. Am J Gastroenterol 1988;83:526-30. 5. Sarin SK, Nanda R, Sachdev G. Relative efficacy and safety of absolute alcohol and 50% alcohol as variceal sclerosants. Gastrointest Endosc 1987;33:362-5. 6. Kitano S, Koyanagi N, Iso Y. Prospective randomized trial

VOLUME 36, NO.6, 1990

7. 8.

9. 10. 11.

12. 13. 14.

15. 16. 17. 18. 19. 20.

21. 22. 23. 24.

25. 26. 27. 28. 29. 30. 31.

comparing two injection techniques for sclerosing oesophageal varices: over-tube and free-hand. Br J Surg 1987;74:603. Sarin SK, Nanda R, Sachdev G, et al. Intravariceal versus paravariceal sclerotherapy: a prospective, controlled, randomized trial. Gut 1987;28:1657. Westaby D, Melia WB, MacDougall BRD, Hegarty JE, Williams R. Injection sclerotherapy for oesophageal varices: a prospective randomized trial of different treatment schedules. Gut 1984;25:129-32. Jensen DM. Sclerosants for injection sclerosis of esophageal varices. Gastrointest Endosc 1983;29:315-7. Sanowski RA, Waring JP. Endoscopic techniques and complications in variceal sclerotherapy. J Clin Gastroenterol 1987;9:504-13. Brooks WS. Variceal sclerosing agents. Am J Gastroenterol 1984;79:424-8. Sanowski RA. In search of a safe sclerosing agent. Endocr Rev 1985;2:31-5. Conn HO, Grace ND. Portal hypertension and sclerotherapy of esophageal varices: a point of view. Endocr Rev 1985;2:39-53. Jensen DM, Silpa ML, Tapia JI, Beilin DB, Machicado GA. Comparison of different methods for endoscopic hemostasis of bleeding canine esophageal varices. Gastroenterology 1983;84:1455-61. Shepherd MM, Lee RG, Bowers JH. Local toxicity of sclerosing agents in canine esophagus. Gastrointest Endosc 1983;29:AI88. Blenkinsopp WK. Comparison of tetradecyl sulfate of sodium with other sclerosants in rats. Br J Exp Pathol 1968;49:197201. Silpa ML, Jensen DM, Machicado GA, Tapia JI, Beilin DB. Efficacy and safety of agents for variceal sclerotherapy. Gastrointest Endosc 1982;28:A152-3. Harris OD, Dickey JD, Stephenson PM. Simple endoscopic injection sclerotherapy of esophageal varices. Aust N Z J Med 1982;12:131-5. Fearon B, Sass-Kortsak A. Management of esophageal varices in children by injection sclerosing agents. Ann Otol Rhinol Laryngol 1959;68:909-15. Chung SCS, Leung JWC, Lo KK, et al. Randomized controlled comparison of sodium tetradecyl versus ethanolamine for injection sclerotherapy of esophageal varices. Endoscopy 1988;20:A36. Brooks WS, Galambos JT. Endoscopic sclerosis of esophageal varices: evaluation of sclerosing agents. A J Gastroenterol 1982;77:A681. Brooks WS, Galambos JT. Sclerotherapy of esophageal varices: post-injection ulceration. Gastrointest Endosc 1983;29:AI91. Trudeau W, Prindiville T, Gibbert V, Young W, Siepler J. An update on the safety of sclerosing agents used in endoscopic injection sclerosis. Gastrointest Endosc 1983;29:Al68-9. Gibbert V, Feinstat T, Burns M, Trudeau W. A comparison of the sclerosing agents sodium tetradecyl sulfate and sodium morrhuate in endoscopic injection sclerosis of esophageal varices. Gastrointest Endosc 1982;28:A147-8. Boyd WP, Goldschmid S, Brady PG, Silver D, Mamel J, Nord HJ. Chronic endoscopic variceal sclerotherapy. Am J Gastroenterol 1985;80:A851. Paquet KJ. Endoscopic paravariceal injection sclerotherapy of the esophagus-indications, technique, complications: results of a period of 14 years. Gastrointest Endosc 1983;29:310-5. McClave SA. Esophageal injection sclerosis. Int J Dermatol 1987;26:244-9. Beppu K, Inokchi K, Koyanagi N, et al. Prediction of variceal hemorrhage by esophageal endoscopy. Gastrointest Endosc 1981;27:213-8. Terblanche J, Kahn D, Campbell JA, et al. Failure of repeated injection sclerotherapy to improve long term survival after esophageal variceal bleeding. Lancet 1983;2:1328-32. MacDougall BR, Westaby D, Theodossi A, et al. Increased long term survival in variceal hemorrhage using injection sclerotherapy. Lancet 1982;1:124-7. Clark AW, Westaby D, Silk DB, et al. Prospective control study of injection sclerotherapy in patients with cirrhosis and recent variceal hemorrhage. Lancet 1980;2:552-4.

571

Prospective randomized comparison of esophageal variceal sclerotherapy agents: sodium tetradecyl sulfate versus sodium morrhuate.

We designed a prospective randomized study to evaluate differences in efficacy and complication rate between the two most commonly used sclerosing age...
5MB Sizes 0 Downloads 0 Views