Sclerotherapy vs. Esophageal Transection vs. Distal Splenorenal Shunt for the Clinical Management of Esophageal Varices in Patients with Child Class A and B Liver Function: A Prospective Randomized Trial SEIGO

KITANO,YASUNORI ISO, W O T O HASHIZUME, HIROHIKO YAMAGA, NOBUHIROKOYANAGI, HIROYA WADA, TETSUYA IWANAGA, MASAWKIOHTA AND mIZ0 SUGIMACHI Department of Surgery ZZ, Kyushu University, Faculty of Medicine, Fukuoka 812, Japan

Ninety-six patients with good liver function (Child class A or B) and esophageal varices were randomly assigned to one of three groups given different treatments: endoscopic injection sclerotherapy (n = 32), esophageal transection (n = 32) or distal splenorenal shunt (n = 32). Five patients (5.2%) had to be excluded from this study because severe chronic pancreatitis made separation of the distal splenic vein from the pancreatic bed difficult. Esophageal transection was performed for these patients. No deaths occurred during the 30 days of treatment. The 5-yrcumulative bleeding rates were 0%, 5.9% and 12.9% in the endoscopicinjection sclerotherapy, esophageal transection and distal splenorenal shunt groups, respectively (nostatistical significance). In no case in the three groups did death occur because of variceal bleeding. Sixteen patients died, mainly because of underlying liver disease; four were in the endoscopic injection sclerotherapy group, five were in the esophaged transection group and seven were in the distal splenorenal shunt group. No statistically significant difference in survival rate among the three groups was found. These results show that endoscopic injection sclerotherapy is a satisfactory alternative to esophageal transection or distal splenorenal shunt for the clinical management of patients with esophageal varices. (HEPATOLOGY 1992;15:63-68.)

In Japan, nonshunting procedures and selective shunts such as esophageal transection ( E T ) with extensive devascularization of the lower esophagus and the upper stomach (1)and distal splenorenal shunt (DSRS) (2) were once widely performed after portal decompression surgery, but were abandoned due to the extremely high rate of occurrence of postoperative encephalopathy. A nationwide survey, however, showed

Reeeived October 13, 1990; accepted August 28, 1991. Address reprint requests to: Seigo Kitano, M.D., Department of Surgery 11, Kyushu University, Faculty of Medicine Fukuoka 812, Japan. 31/1/33692

good results under elective and prophylactic conditions (3). The risk of variceal hemorrhage was assessed by endoscopy using the criteria of the Japanese Research Society for Portal Hypertension (4). Endoscopic injection sclerotherapy (EIS) to treat patients with esophageal varices was resurrected in the 1970s (5, 6) because of its success in halting acute variceal bleeding and preventing recurrent episodes of bleeding. After considerable experience with EIS, we proposed that removal of the esophageal mucosa might well be an endpoint for prevention of recurrence of esophageal varices after EIS in the long run (7). When compared with other emergency procedures such as ET, EIS was shown to be superior to surgery in that patients lived longer and incidence of repeat bleeding was rare (8). Other investigators (9-11) reported no advantage for EIS; hence no consensus exists as to whether EIS guarantees fewer episodes of repeat bleeding and better long-term survival rates in patients with cirrhosis and esophageal varices. We report here findings in a prospective randomized trial to assess the effects of elective EIS in comparison with ET and DSRS for treating patients with esophageal varices. PATIENTS AND METHODS From April 1985 to March 1990, 96 consecutive Child class A and B patients were included in a prospective randomized clinical trial to investigate the efficacy, safety and complications of EIS vs. ET vs. DSRS in the elective treatment of esophageal varices. Thirty-five of the 96 patients had had at least one endoscopically proven episode of variceal bleeding. Written, informed consent was obtained from each patient. For long-term follow-up, we selected patients younger than 65 yr who had values of less than 35% on the indocyanine green test (normal = 0% to 10%). All the patients were Japanese and living in Japan. During the same period, 821 patients admitted to our clinics who were considered poor candidates for surgery were not included in this trial and were treated with EIS, mainly because of poor liver function (Child class C) and concomitant, inoperable hepatoma. Patients with acute variceal bleeding or recurrent varices after surgery were also excluded from this

63

KITANO ET AL.

64

; !%!

TABLE 1. Characteristicsof the three groups at entry

u

2

100-7

Age T I

70 -

5 -

6050 -

?

:

5 =

5

40-

302010

0 c

HEPATOLOGY

C u m u l a t i v e survival r a t e s

__ . _

__-

EIS _ ET ~ DSRS

72.6%

_

Cumulative bleeding rates

- - - - f -___

J

- - - - - - - - - - - -12.9% r _ - - ~

_ _ _ _J

5.9% 0.0%

br)"

Sex (M :F)

ET

EIS (n = 32)

(n = 32)

DSRS (n = 32)

54.9 t 7.0 19: 13

49.2 2 7.6 21:11

50.4 t 9.8 24:8

Liver disease

Cirrhosis (alcoholic) IPH Previous episode of variceal bleeding Child's score" Child's class A B

28 (8) 5 12

4 12 6.0

?

0.8

study because of high risk for perioperative mortality or difficult surgical manipulation. Prophylactic treatment was given to patients with high-risk varices (those patients with blue varices of F2 or F3 grade and those patients with varices showing moderate to severe degree of one of the following: red wale markings ( + +,+ + +), cherry-red spots ( + + ,+ + +) or hematocystic spots ( + ) (4, 12). Randomization was performed when the patient was found to be a good candidate for any of three treatment modalities. Patients were assigned to one of the three therapeutic groups using the block-randomization method and the envelope method. This randomization was fully explained and each patient agreed to the assignment. The quantitative classification as proposed by Pugh et al. (13) was used. Patients assigned to undergo EIS were given injections of 5%ethanolamine oleate (by the overtube technique (14) at the initial session and with the free-hand technique at following sessions) until eradication of esophageal varices was attained. Details of the endoscopical procedures we used have been reported (7, 14). EIS was carried out at 1-wk intervals and terminated when a superficial circumferential ulcer had developed in the lower esophagus. Follow-up endoscopy was scheduled for 4 to 6 wk after the last session and then at 3-mo intervals. In patients randomly assigned to ET, transection of the esophagus was performed. The abdomen was opened through a bilateral subcostal incision. After resection of the spleen, the peritoneal reflexion around the cardia of the stomach was incised and the lower esophagus (7 cm above the esophagogastric junction) and the proximal stomach were extensively devascularized (modified Sugiura procedure) (1).The vagal nerve was also resected. The end-to-end anastomosis autosuture stapling device (EEA United States Surgical Corp., Norwalk, CT) was introduced into the esophagus through a small anterior gastrostomy opening. The esophageal wall was inverted by a silk ligature tied around the central rod between the cartridge and anvil, 2 to 3 cm above the esophagogastric junction. The anvil and cartridge were approximated and, by

6.2 i 0.9

24 8 6

Ascites ICG (retention rate 22.2 t 8.8 at 15 min) (%)" Albumin (grddl)" 3.8 ? 0.4 Total bilirubin (mgfdl)" 1.0 * 0.5 AST (IU/L)" 59.8 f 16.8 ALT (IU/L)" 43.0 t 21.0 Prothrombin 76.6 17.4 index (%)"

3 11 6.3

20 12 8 21.4

5

20 12 7 8.0

3.6 ? 0.4 0.9 2 0.3 61.2 i 21.2 46.5 ? 23.2 75.2 i 14.5

IPH = idiopathic portal hypertension; ICG test (normal = 0% to 10%). "Values expressed as mean c S.D.

i: 0.9

=

21.5 t 6.9 3.7

k

0.5

1.1 4 0.4

55.6 i: 20.2 41.0 t 18.4 70.9 5 17.3

indocyanine green

squeezing the trigger, the esophagus was transected and reanastomosed. The cartridge used was 28 mm in diameter for all patients. The excised tissue ring was examined for completeness; anastomotic leakage of dye infused into the esophagus and the stomach was nil in all cases. The gastrostomy opening was closed and pyloroplasty was performed. Oral intake of food was initiated after no anastomotic leakage was demonstrated by radiocontrast 7 to 10 days after surgery. Patients assigned to undergo DSRS underwent modified DSRS with splenopancreatic disconnection (15, 16). Technical details and long-term results of the modified DSRS were as reported elsewhere (16). Patients in whom the shunt could not be performed (for technical reasons) were considered initial failures of the procedure and were excluded in calculating survival and bleeding rates. Persistence and size of esophageal varices were assessed by follow-up endoscopy. Shunt patency was verified by angiography and ultrasonography approximately 3 wk after surgery or if any variceal bleeding was found. The principal author (S.K.) performed the critical operative procedures or was present in units at each session of EIS or at surgery. All patients were compliant and returned as scheduled for follow-up endoscopy. For statistical analysis, qualitative variables were compared using the x2 test with Yate's correction as needed. Quantitative variables were compared using unpaired Student's t test. The Kaplan-Meier method was used to plot survival and bleeding rates.

RESULTS Patients Characteristics. Thirty-two patients were ran-

domly assigned to the EIS group, to the ET group and to the DSRS group. Data obtained from the past history and on admission are summarized in Table 1. The three groups were comparable at randomization with regard to age, sex, Child class and origin of liver disease. Twelve patients in both the

Vol. 15, No. 1,1992

65

SCLEROTHERAF'Y, ESOPHAGEAL TRANSECTION OR DISTAL SPLENORENAL SHUNT

TABLE 2. Alternative treatments and outcomes of patients in the DSRS group excluded from the trial after randomization

Patient

Age (yr)/sex

Alternative treatment

Blood transfusion (ml)

30 33

ET ET

3,000

25 27 10

ET ET ET

Duration of alcohol use (yr)

Reason for exclusion ~

1 2

5 1M 55m

3

45m 47iM 32iM

4 5

Marked adhesion of splenic vein Marked fibrosis and small caliber (7 mm) of splenic vein Marked adhesion of splenic vein Marked adhesion of splenic vein Splenic artery thrombosis due to preoperative angiography

TABLE 3. Perioperative status of patients in the three groups EIS Variables

(n = 32)

Duration of proce0.6 ? 0.2 dure (hr)" Blood transfusions required Blood transfusions per patient ( X 1000 ml)" Hospitalization (daysla 25.8 2 4.1 Hospital mortality -

Outcome (mo after procedure)

~

ET (n = 32)

3.5 -t 0.6'

-

EIS

DSRS (n = 27)

4.5

1.2'

?

1I d

0.06 -c 0.27

0.4 t 0.5'

-

Dead (22) Alive (34) Alive (26)

1,200

TABLE 4. Clinical results

2

26.9 2 6.2

Alive (54) Alive (43)

-

26.3

&

5.2

Parameters

Mean follow-up, in months (range) Late mortality Late survival (%) Bleeding Varices Duodenal ulcer Gastritis

(n = 32)

ET (n = 32)

40.8 (9-72) 4 28 87.5

38.7 (6-72) 5 27 84.4

-

1 1 3

1

DSRS

(n = 27) 41.3 I111-72)

7 20 74.1 4 1 -

-

"Data expressed as mean t S.D. bp < 0.001 compared with EIS group. "p < 0.01 compared with ET group. dp < 0.005 compared with EIS and ET groups. 'p < 0.01 compared with ET group; p < 0.001 compared with EIS group.

EIS and ET groups and 11patients in the DSRS group had at least one episode of endoscopically proven variceal bleeding. The remaining patients were given prophylactic treatment because signs of high risk were present on endoscopy. At randomization, no patient had a concomitant hepatoma. thclusions. Five (15.6%)of the 32 patients assigned to the DSRS group had to be excluded from the trial after randomization and during actual surgery because of chronic pancreatitis so severe that separation of the splenic vein from the pancreatic bed was not feasible. The alternative treatment used and the final outcome are summarized in Table 2. In two of the five patients we saw recurrence of the varices (2 and 3 yr, respectively, after surgery; EIS was performed to eradicate the varices. No patients bled during follow-up. Short-term Outcome. Eleven (40.7%)of the 27 patients in the DSRS group who were subjected to shunting required blood transfusion; no patients in the EIS group required transfusion, and two patients in the ET group had to be transfused. Between the DSRS and the other two groups was a statistically significant difference (p < 0.005) in the number of patients requiring blood transfusion during surgery (Table 3). The patients in the DSRS group required significantly more transfusions than did those in the ET (p < 0.01) and EIS groups (p < 0.001). The 32 EIS patients underwent 4.5 & 0.3 (mean S.D.) sessions per patient (range = 3 to 6 sessions). Patients received 57.6 t 8.3 ml5% ethanolamine oleate (range = 38 to 66 ml); 24.8 h 4.2 ml, 15.8 5 3.8 ml, 11.0 h 3.2 ml and 5.8 ? 3.3 ml were administered at the first, second, third and

fourth sessions, respectively. All EIS patients underwent removal of the mucosa of the lower esophagus (our proposed endpoint) (7). The procedure required significantly less time (p < 0.001) in the EIS group (total time for all sessions) than in the ET and the DSRS groups, and significantly less time (p < 0.01) in the ET group than in the DSRS group. No significant difference was seen in the duration of hospitalization among the three groups. One patient in the DSRS group had an occluded shunt, and repeat bleeding occurred on the 12th postoperative day. Emergency EIS successfully halted bleeding from the esophageal varices, and eradication of varices was attained with three sessions of EIS. Another patient who underwent DSRS bled from a duodenal ulcer on the 24th postoperative day and was treated conservatively. All 96 patients were discharged from hospital. Long-term Outcome. The 96 patients were followed for 6 to 72 mo after treatment (mean = 40.2 mo). No patient was lost to follow-up. At this writing, the survival rate is 87.5% in the EIS group, 84.4% in the ET group and 74.1% in the DSRS group (Table 4). Four, five and seven patients died in the EIS, ET and DSRS groups, respectively, during follow-up. The mortality rate was 12.5%in the EIS group, 15.6%in the ET group and 25.9%in the DSRS group at 5 yr after surgery. The variceal bleeding rate was 0% in the EIS group, 3.1%(1of 32) in the ET group and 14.6%(4 of 27) in the DSRS group, with no statistical significance. Cumulative bleeding and survival rates are shown in Figure 1. The 5-yr cumulative bleeding rates were 0%, 5.9% and 12.9%in the EIS, ET and DSRS groups, respectively. No statistically significant difference was seen in cumulative bleeding or survival rates among the three groups. The causes of death and times of follow-up are shown in Table 5. One patient in the EIS group died of unknown causes but with no occurrence of gastrointestinal bleeding; two died of hepatoma and another died of heart failure. In the ET group,

66

KITANO ET AL. TABLE 5. Cause of death Age (yr)/sex

Group

1

47lF

EIS

2 3 4 5 6 7 8 9

45 69 64lF

EIS EIS EIS ET ET ET ET ET

10

53lF

DSRS

11 12 13 14 15 16

60

DSRS DSRS DSRS DSRS DSRS DSRS

Patient

53lF 54 52 52 57

64

53 54 63 64

Cause of death

Unknown (no gastrointestinal bleeding) Hepatoma Hepatoma Heart failure Liver failure Meningitis Liver failure Hepatoma Duodenal ulcer bleeding Liver failure, portal thrombosis Liver failure Hepatoma Hepatoma Hepatoma Liver failure Liver failure

%e(mo after procedure)

12 63 45 9 31 33 63 54 6

18 70 44 39 31 15 25

one patient died of massive gastrointestinal bleeding 1mo after follow-up endoscopy revealed the disappearance of esophagogastric varices and the presence of a duodenal ulcer; two patients died of liver failure and another died of meningitis. Seven patients died in the DSRS group: one died of liver failure after portal-vein thrombosis due to poor circulation 18 mo after DSRS. Three other patients died of liver failure and three more died of hepatoma. The bleeding sites, times of onset of bleeding and outcomes are shown in Table 6. One patient had massive hematemesis and died 6 mo after ET, probably of bleeding from an endoscopically evident duodenal ulcer. Five patients (18.5%)in the DSRS group had episodes of hematemesis. One patient survived an episode of bleeding from a duodenal ulcer detected during the initial hospitalization. The remaining four patients had variceal bleeding; one patient experienced bleeding during the hospital stay (2 wk after shunting) and the other three patients experienced bleeding during follow-up times of 20,21 and 68 mo after surgery. In these four patients, the shunt was occluded. EIS successfully halted the acute bleeding. At this writing, no patient has died of variceal bleeding.

DISCUSSION

As one of three types of treatment for esophageal varices, we performed modified DSRS (namely, total pancreatic disconnection [15, 161) to prevent postoperative portal malcirculation, a situation which may lead to an encephalopathy. The frequency of hepatic encephalopathy during follow-upin patients of the DSRS group (4%)was relatively less than reported in several series of patients treated by a classical Warren’s shunt (range = 8% to 38% [17-201). In this study, no patient had this complication in either the EIS or ET group, suggesting that EIS and ET are safer than DSRS in preventing portal-systemic encephalopathy. However, total pancreatic disconnection had to be performed. Such an approach made DSRS more difficult; the main limitation in surgical

HEPATOLOGY

procedures is the difficulty in dissecting the distal splenic vein from the pancreatic bed. The rate of completed DSRS was 84.4% (27 patients) in the 32 patients who underwent DSRS and 55.6%(five patients) in the nine patients with chronic pancreatitis due to alcohol abuse. Early mortality (within the first 30 days after therapy) was nil in the three groups. In patients treated with DSRS, surgical mortality (nil) was lower than that reported by most authors who used the classical or modified Warren shunt (2, 17-20).The lack of mortality after DSRS in this study is easily explained because the main cause of death was liver failure in patients with poor liver function. In our series, no statistically significant difference was found in late mortality during the 5-yr follow-up in the EIS group compared with the ET and DSRS groups, although this trial is too small to rule out the possibility of a type I1 error. Death may not occur unless the patient’s liver function deteriorates in patients with Child class A or B liver function. Surgical interventions with the need for less than 1,000 ml of blood transfusion does not seem to be the factor influencing the prognosis of patients undergoing surgery in the mean follow-up period of 5 yr. Since none of the 64 patients assigned to the EIS and ET groups had to be excluded (compared with 5 of the 32 patients assigned to the DSRS group), EIS or ET may be preferable to DSRS. The main cause of failure in applying DSRS is the marked adhesion of the splenic vein to the pancreas. This trend increases in patients with histories of alcohol abuse and chronic pancreatitis. The full procedures of EIS and ET were accomplished to eradicate esophageal varices in all patients assigned to the EIS or ET groups. We reported a relatively high rate of recurrence of varices after ET in the mean follow-up of 25 mo (21). In this study, only one patient has had recurrent variceal bleeding at this writing. Nevertheless, these patients must be followed closely for a longer period of time. Four (14.8%)of the 27 patients belonging to the DSRS group had variceal bleeding during follow-up. This rate of variceal bleeding after DSRS is similar to that seen after Warren’s shunt (range = 0% to 28% [17-20, 221). A comparison between the prevalences of variceal bleeding observed in the three therapeutic groups showed that EIS and ET may be more effective than DSRS in preventing variceal bleeding, though the differences are not statistically significant. Despite the occurrence of variceal bleeding in the DSRS group, the survival rate was not significantly different from those of the EIS or ET groups. The long-term survival of these patients did not seem to be influenced by variceal bleeding but rather by other variables such as liver function. The program and technique of EIS, including the sclerosing agents, affects the rate of control of acute variceal bleeding and prevents repeat bleeding. We found 5% ethanolamine oleate to be superior to 1% polidocanol(23) or 3% sodium tetradecyl sulfate (24) in

Vol. 15, No. 1,1992

SCLEROTHERAPY, ESOPHAGEAL TRANSECTION OR DISTAL SPLENORENAL SHUNT

67

TABLE6. Patients with hematemesis during follow-up Patient

1 2 3 4

5 6 7 8 9 10 11

Group

EIS ET ET ET ET ET DSRS DSRS DSRS DSRS DSRS

Bleeding sites

Gastritis Gastrointestinal bleeding Duodenal ulcer Gastritis Esophageal varices Gastritis Gastric varices Gastric v a r i e s Esophageal varices Esophageal varices Duodenal ulcer

sclerotic therapy of esophageal varices. We have proposed that the endpoint in EIS might be the removal of mucosa of the lower esophagus, where varices can recur during the long-term follow-up, to prevent recurrence of varices (7). Three reports of clinical trials comparing DSRS and EIS have appeared. Warren et al. (25) concluded that EIS has a higher repeat bleeding rate than DSRS but a significantly improved 2-yr survival rate, whereas the other two studies (11, 26) found no significant differences in long-term survival between the DSRS and EIS groups. We treated patients with varices who were at high risk of bleeding before the first variceal bleeding and as predicted endoscopically. The early prospective randomized trials (27,28) with sclerotherapy revealed that prophylactic sclerotherapy could prevent esophageal variceal bleeding and prolonged the survival rate of patients with esophagealvarices to a considerable extent in comparison with conservative treatment. In this trial, it is difficult to say whether the three treatment modalities are benifitial for patients who have never bled from varices. Final results in a prospective randomized trial (29) on the prophylactic treatment of esophageal varices showed that the bleeding rate in the control group was 46% at 5 yr, vs. 7% in the treatment group. Sauerbruch et al. (30)found that a subgroup of patients with esophageal varices and moderately decompressed alcoholic cirrhosis may benefit from prophylactic sclerotherapy. On the contrary, a recently reported, randomized, single-blind multicenter clinical trial (31) suggested that sclerothrapy should not be performed until after an initial episode of bleeding from esophageal varices has occurred because of increased mortality among men with moderate-to-severe alcoholic liver disease. This conclusion may be related to the lower bleeding rate in the control group and the higher bleeding rate in the sclerotherpy group compared with data in the report from Japan (29). Considering the high rates of bleeding and mortality at the initial episode of bleeding, prophylactic treatment could be performed in patients with high risk-varices. EIS in particular might well be recommended to prophylactically treat patients with esophageal varices, since EIS allows for significant

Time

Treatment

3 mo 17 mo 6 mo 23 mo 34 mo 60 mo 68 mo 20 mo 12 days 21 mo 24 days

-

Outcome (mo after procedure)

EIS

-

EIS EIS EIS EIS Medical

Alive (18) Dead (33) Dead (6) Alive (28) Alive (54) Alive (60) Dead (70) Alive (61) Alive (60) Dead (39) Alive (51)

improvement of liver function and better maintenance of liver function than does DSRS (25). In summary, from the results of this study, we tentatively conclude that repeated EIS, using our technique and aiming at our proposed endpoint (denudation of the mucosa of the lower esophagus) is a satisfactory alternative to ET and DSRS for elective treatment of subjects with esophageal varices (the objective being to prevent recurrence of the varices and recurrent bleeding from esophageal varices in cirrhotic patients belonging to classes A and B of the modified Child’s classification). DSRS can effectively prevent recurrent bleeding in nonalcoholic patients after a successful shunt is constructed, if no technical difficulty is encountered in disconnecting the splenic vein from the pancreatic bed. ET can also be performed by general surgeons for patients with good liver function and a small amount of bleeding during surgery, with little insult to the liver. We propose that EIS should be the first treatment of choice to prevent bleeding from esophageal varices under elective conditions in patients with Child class A and B liver function.

Acknowledgment: We thank M. Ohara for helping prepare this report. REFERENCES 1. Sugiura M, Futagawa S. A new technique for treating esophageal varices. J Thorac Cardiovasc Surg 1973;66:677-685. 2. Warren WD, Zeppa R, Fomon JJ. Selective trans-splenic decompression of gastroesophageal varices by distal splenorenal shunt. Ann Surg 1967;166:437-455. 3. Inokuchi K. Japanese Research Society for Portal Hypertension. Present status of surgical treatment of esophageal varices in Japan: a nationwide survey of 3,588 patients. World J Surg 1985;9:171-180. 4. Japanese Research Society for Portal Hypertension. The general rules for recording endoscopic findings on esophageal varies. Jpn J Surg 1980;10:84-87. 5. Johnston GW, Rogers HW.A review of 15years’ experience in the use of sclerotherapy in the control of acute haemorrhage from oesophageal varices. Br J Surg 1973;60:797-800. 6. Terblanche J , Northover JMA, Bornman PC, Kahn D, Barbezat GO, Sellars SL, Saunders SJ. A prospective evaluation of injection sclerotherapy in the treatment of acute bleeding from esophageal varices. Sureerv - * 1979:85:239-245.

68

KITANO ET AL.

7. Kitano S, Koyanagi N, Is0 Y, Higashi H, Sugimachi K. Prevention of recurrence of esophageal varices after endoscopic injection 1987;7: sclerotherapy with ethanolamine oleate. HEPATOLOGY 810-815. 8. Kitano S, Hashizume M, Yamaga H, Higashi H, Sugimachi K. Sclerotherapy is superior to surgery for longer survival and less rebleeding in patients with acute variceal bleeding. Int Surg 1989;74:162-166. 9. Cello JP, Grendeli JH, Crass RA, Weber TE, Trunkey DD. Endoscopic sclerotherapy versus portacaval shunt in patients with severe cirrhosis and acute variceal hemorrhage. N Engl J Med 1987;361:11-15. 10. Huizinga WKJ, Angorn IB, Baker LW. Esophageal transection versus injection sclerotherapy in the management of bleeding esophageal varices in patients at high risk. Surg Gynecol Obstet 1985;160:539-546. 11. Teres J, Bordas JM, Bravo D, Visa J, Grande L, Garcia-Valdecasas JC, Pera C, et al. Sclerotherapy vs. distal splenorenal shunt in the elective treatment of variceal hemorrhage: a randomized con1987;7:430-436. trolled trial. HEPATOLOGY 12. Beppu K, Inokuchi K, Koyanagi N, Nakayama S, Sakata H, Kitano S, Kobayashi M. Prediction of variceal hemorrhage by esophageal endoscopy. Gastrointest Endosc 1981;27:213-218. 13. Pugh RNH, Murray-Lyon IM, Dawson JL, Pietroni HC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60:646-649. 14. Kitano S, Koyanagi N, Is0 Y , Iwanaga T, Higashi H, Sugimachi K. Prospective randomized trial comparing two injection techniques for sclerosing oesophageal varices: over-tube and free-hand. Br J Surg 1987;74:603-606. 15. Warren WD, Millikan WJ, Henderson JM, Rasheed ME, Salam AA. Selective variceal decompression after splenectomy of splenic vein thrombosis. Ann Surg 1984;199:694-702. 16. Inokuchi K, Beppu K, Koyanagi N, Nagamine K, Hashizume M, Sugimachi K. Exclusion of nonisolated splenic vein in distal splenorenal shunt for prevention of portal malcirculation. Ann Surg 1984;200:711-717. 17. Bustill RW, Bein B, Tompkins RK. Matched control study of distal splenorenal and portacaval shunts in the treatment of bleeding esophageal varices. Am J Surg 1979;138:62-67. 18. Conn HO, Resnich RH, Grace ND, Atterbury CE, Horst D, Groszmann RJ, Gazmuri P, et al. Distal splenorenal shunt vs. portal-systemic shunt: current status of a controlled trial. HEPATOLOGY 1981;1:151-160.

HEPATOLOGY

19. Rikkers LF, Rudman D, Galambos JT, Fulenwider JT, Millikan WJ, Kunter M, Smith RB, et al. A randomized, controlled trial of the distal splenorenal shunt. Ann Surg 1978;188:271-282. 20. Ridell AG, Bloor K, Hobbs KEF, Jacquet N. Elective splenorenal anastomosis. Br Med J 1972;1:731-732. 21. Koyanagi N, Is0 Y, Higashi H, Kitano S, Sugimachi K. Recurrence of varices after oesophageal transection: intraoperative and postoperative assessment by endoscopy. Br J Surg 1988;75:9-11. 22. Vang J , Simert JA, Hansson A, Bengmark S. Results of a modified distal splenorenal shunt for portal hypertension. Ann Surg 1977;185:224-228. 23. Kitano S, Is0 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. Hepato-gastroenterol 1987;34: 19-23. 24. Kitano S, Is0 Y, Yamaga H, Hashizume M, Higashi H, Sugimachi K. Trial of sclerosing agents in patients with oesophageal varices. Br J Surg 1988;75:751-753. 25. Warren WD, Henderson JM, Millikan WJ, Galambos JT, Brooks WS, Riepe SP, Salam AA, et al. Distal splenorenal shunt versus endoscopic sclerotherapy in uncontrolled variceal bleeding. Ann Surg 1986;203:454-462. 26. 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;206:261-271. 27. Paquet K-J. Prophylactic endoscopic sclerosing treatment of the esophageal wall in varices: A prospective controlled trial. Endoscopi 1982;14:4-5. 28. Witzel L. Wolberes E. Merki H. Pronhvlactic endoscoDic sclerotherapy of oesophageal varices: a prospective controlled study. Lancet 1985;1:773-775. 29. Inokuchi K. Cooperative study group of portal hypertension of Japan. Improved survival after prophylactic portal nondecompression surgery for esophageal varices: a randomized clinical 1990;12:1-6. trial. HEPATOLOGY 30. Sauerbruch T, Wotzka R, Kopcke W, Htirlin M, Heldwein W, Bayerdorffer E, Sander R, et al. Prophylactic sclerotherapy before the first episode of variceal hemorrhage in patients with cirrhosis. N Engl J Med 1988;319:8-15. 31. The Veterans Affairs cooperative variceal sclerotherapy group. Prophylactic sclerotherapy for esophageal varices in men with alcoholic liver disease. N Engl J Med 1991;324:1779-1784. "

I

."