- Email: [email protected]
Left gastric vein hemodynamics and variceal recurrence in patients undergoing prophylactic endoscopic ligation of high-risk esophageal varices
Left gastric vein hemodynamics and variceal recurrence in patients undergoing prophylactic endoscopic ligation of high-risk esophageal varices Akio Matsumoto, MD, Norihiro Hamamoto, MD, Atsuko Ohnishi, MD, Hirofumi Miyoshi, MD, Kazunori Sugi, MD, Hiroshi Kojima, MD, Masanobu Kayazawa, MD, Hiroshi Morikawa, MD, Ichiro Hirata, MD, Ken-ichi Katsu, MD Osaka, Japan
Background: Early recurrence of esophageal varices remains problematic after endoscopic variceal ligation. To evaluate the efficacy of prophylactic endoscopic ligation for esophageal varices at high risk for bleeding, the relationship between left gastric vein hemodynamics and variceal recurrence was investigated. Methods: Thirty-five patients with cirrhosis underwent endoscopic variceal ligation. Angiography was performed in all patients before treatment and after eradication of varices to study left gastric vein hemodynamics. Results: Before treatment, 12 patients had hepatopetal flow in the left gastric vein (type I), 17 had hepatofugal flow (type II), and 6 had hepatofugal flow with an extra-esophageal shunt (type III). In type I and III patients, the direction of blood flow in the left gastric vein did not change after eradication of varices. Type II patients showed bi-directional flow in the left gastric vein after treatment. Varices recurred in all but one type II patient and in one type I patient during follow-up (mean 36.7 months). The 2-year recurrence-free rate was higher in type I patients (p = 0.0001) and type III patients (p = 0.0002) than in type II patients. Conclusions: Prophylactic ligation seems to be a safe and useful procedure, especially in patients with type I or III hemodynamics in the left gastric vein before treatment. (Gastrointest Endosc 1999;50:768-74.)
In recent years, there have been marked advances in the endoscopic treatment of esophageal varices. Sclerotherapy, the conventional method, is highly effective for controlling acute bleeding and for preventing recurrence of bleeding.1 However, sclerotherapy has to be performed carefully to prevent complications associated with the injected sclerosant.2 Endoscopic variceal ligation (EVL) was developed by Stiegmann et al.3 as an alternative to sclerotherapy. This procedure is easier to perform and has achieved wide popularity, but its impact on the portal hemodynamics and long-term efficacy have not been clearly established. Because the left gastric vein (LGV) is the predominant collateral route for the formation of esophageal varices,4 evaluation of hemodynamics in the LGV can be used to select patients for prophylactic EVL. The present study was therefore performed to define which patients will redevelop high-risk esophageal varices Received November 18, 1998. For revision April 27, 1999. Accepted July 15, 1999. From the Second Department of Internal Medicine, Osaka Medical College, Takatsuki, Osaka, Japan. Reprint requests: Akio Matsumoto, MD, Second Department of Internal Medicine, Osaka Medical College, 2-7, Daigaku-cho, Takatsuki, Osaka, Japan; fax: 0726-84-6532. Copyright © 1999 by the American Society for Gastrointestinal Endoscopy 0016-5107/99/$8.00 + 0 37/1/101477 768
GASTROINTESTINAL ENDOSCOPY
after EVL by investigating the relationship between angiographic features in the LGV and variceal recurrence. PATIENTS AND METHODS After consultation with three physicians specializing in the treatment of varices, we selected 35 consecutive patients managed at our institution from April 1994 to April 1996. The selection criteria were as follows: (1) highrisk esophageal varices with a Beppu score less than –0.38 (i.e., varices were F2 [enlarged tortuous] or F3 [large coilshaped] in form and showed at least one of the following: red wale markings [++, +++], cherry red spots [++, +++], or hematocystic spots [+]);5,6 (2) no history of upper GI bleeding; (3) no gastric varices or ectopic varices; (4) no portal thrombosis; and (5) no malignant diseases including hepatocellular carcinoma. In all 35 patients, the underlying disease was liver cirrhosis, with the diagnosis of cirrhosis confirmed by clinical, biochemical, or histologic means. Portal hypertensive gastropathy was slight or absent, with a “snake skin pattern” limited to the upper body of the stomach, as described by McCormack et al.7 The patients all underwent prophylactic EVL alone. Before treatment, written informed consent was obtained from each patient, and the study protocol was reviewed and approved by the ethics committee of our medical college. Esophageal variceal ligation An endoscope (XQ20; Olympus Optical Co., Ltd., Tokyo, Japan), an overtube (Sumitomo Bakelite, Tokyo, VOLUME 50, NO. 6, 1999
Endoscopic ligation of high-risk esophageal varices
A
A Matsumoto, N Hamamoto, A Ohnishi, et al.
B
Figure 1. LGV hemodynamics before EVL (type I). Venous phase in celiac (A) and left gastric (B) arteriograms. Esophageal varices and hepatopetal blood flow in the LGV are visible. Large black arrow, LGV; white arrow, esophageal varices; small black arrows, palisade zone. Japan), and a Varioligator ligation kit (Top, Tokyo, Japan) were used. For sedation, flunitrazepam (Rohypnol, Roche, Tokyo, Japan) was given intravenously at a dose of 0.3 to 0.4 mg immediately before overtube placement. Ligation was begun just below the gastroesophageal junction along each variceal channel. Successive bands were placed as close to the site of initial ligation as possible. Each variceal channel was ligated with 3 to 4 elastic bands to a distance of 5 to 8 cm from the gastroesophageal junction. After 1 week, any remaining varices were ligated. Portions of the mucosa between the varices near the gastroesophageal junction were also ligated. After another week, eradication of the varices and formation of a circumferential ulcer at the gastroesophageal junction were confirmed. The goals of treatment were to achieve circumferential ulcer formation near the gastroesophageal junction and the endoscopic disappearance of varices. Additional treatment was performed when needed. The interval between each treatment session was 1 week. Angiography Before initial EVL and within 2 weeks after a sufficient therapeutic effect was confirmed endoscopically, angiography was performed in all patients. After an overnight fast, the patient was given an intramuscular injection of 25 mg of hydroxyzin (Atarax P, Pfizer, Tokyo, Japan). Digital subtraction angiography was performed with a DAR1200 (Shimazu, Kyoto, Japan). The catheters used were a 5F JA 003 (Catex, Tokyo, Japan) and a 3F SP Catheter (Terumo, Tokyo, Japan). A catheter was introduced via the right femoral artery by Seldinger’s method. After celiac and VOLUME 50, NO. 6, 1999
superior mesenteric arteriography, the splenic and left gastric arteries were visualized selectively. Iopamidol (iopamiron; Schering AG, Berlin, Germany) was injected at a rate of 4 mL/sec to a total volume of 20 mL for celiac and superior mesenteric arteriography, 3 mL/sec to a total volume of 20 mL for splenic arteriography, and 2 mL/sec to a total volume of 8 mL for left gastric arteriography. Angiograms were reviewed independently by two investigators, and hemodynamics in the LGV as well as the presence of collateral vessels were evaluated from the venous phase. In addition, the maximum vessel size in the palisade zone8 was measured. Follow-up Endoscopy was performed at 3-month intervals to detect any recurrence of varices and to observe portal hypertensive gastropathy. Recurrence was defined as the redevelopment of high-risk varices similar to those present before treatment. EVL was repeated as soon as recurrence or bleeding from varices was detected. Bleeding was defined as any episode of hematemesis or melena. The trial endpoints were recurrence or bleeding from esophageal varices, death, loss to follow-up, and maintenance of a recurrence-free status for 24 months after the treatment. Statistical analysis All data are shown as the mean ± SD. The unpaired Student’s t test was used to compare age, Child-Pugh score,9 number of EVL sessions, number of ligations, and maximum vessel size in the palisade zone. Fisher’s exact test was used for other data. Cumulative recurrence-free GASTROINTESTINAL ENDOSCOPY
769
A Matsumoto, N Hamamoto, A Ohnishi, et al.
Endoscopic ligation of high-risk esophageal varices
B
A
Figure 2. LGV hemodynamics before EVL (type II). Venous phase in superior mesenteric (A) and left gastric (B) arteriograms. Hepatofugal blood flow into esophageal varices via the LGV is visible. Large black arrow, LGV; white arrow, esophageal varices; small black arrows, palisade zone. curves were drawn using the Kaplan-Meier method,10 and significance was tested using the log-rank test.11 The level of significance was set at p < 0.05.
RESULTS The therapeutic goals were achieved in all patients. The esophageal ulcers resulting from ligation were small and did not hinder subsequent ligation of the mucosa between varices. There was no bleeding from the esophageal ulcers or from other sites, and no systemic complications. Observation of hemodynamics in the LGV before treatment revealed the following 3 patterns without observer error. In type I, there was no reflux from the splenic or portal vein and all blood flow entered the portal vein (hepatopetal flow) (Fig. 1). In type II, there was reflux from the splenic or portal vein into the esophageal varices via the LGV (hepatofugal flow) (Fig. 2). In type III, blood flow from the portal vein entered the esophageal varices and paraesophageal veins via the LGV (hepatofugal flow with a large extra-esophageal shunt). After treatment, type I and III patients exhibited the same hemodynamics in the LGV as before EVL, whereas type II patients had bidirectional blood flow in the LGV. No other collateral veins were observed in any patient before or after treatment. The palisade zone persisted in all patients. The clinical characteristics of the three groups are summarized in Table 1. There were no significant differences in age, gender, etiology of cir770
GASTROINTESTINAL ENDOSCOPY
rhosis, and extent of varices among them, but liver dysfunction was more severe in type II patients than in either type I patients (p = 0.002 by Child-Pugh classification, p = 0.042 by Child-Pugh score) or type III patients (p = 0.0352 by Child-Pugh score). The maximum vessel size in the palisade zone was larger in type II patients than in either type I or type III patients (p < 0.0001), whereas the number of EVL sessions and ligations was lower in type I or type III patients than in type II patients (p < 0.0001). No patient was lost to follow-up, therefore the final number of subjects was also 35. The mean follow-up period for type I, II, and III patients was 33.6 ± 7.4 months, 36.7 ± 6.5 months, and 36.8 ± 6.8 months, respectively. There were no episodes of GI bleeding, no development of gastric varices, no aggravation of portal hypertensive gastropathy, and no development of portal thrombosis. During followup, there was also no change in the Child-Pugh classification, and no deaths. Two of the type I patients and 3 of the type II patients developed hepatocellular carcinoma. Varices recurred in 17 patients, of whom only 1 had type I and 16 had type II hemodynamics before EVL. In the type II patients, varices always recurred within 9 months after treatment, with the shortest period being 3 months (mean 6.8 months). Among these patients, 3 had a second episode of recurrence. Among the 18 patients without recurrence, 11 had type I and 6 had type III hemodynamics before EVL. The cumulative recurVOLUME 50, NO. 6, 1999
Endoscopic ligation of high-risk esophageal varices
A Matsumoto, N Hamamoto, A Ohnishi, et al.
Table 1. Profile of patients undergoing prophylactic EVL No. of patients (men) Age (yr) Etiology of cirrhosis (A/B/C) Child-Pugh classification (A/B/C) Child-Pugh score Endoscopic features of varices* (F2/F3) No of EVL sessions No of ligations Maximum size in palisade vessels (mm)
Type I
Type II
Type III
12 (10) 56.3 ± 2.7 2/2/8 10/2/0 6.1 ± 1.2 9/3 2.2 ± 0.4 24 ± 1 1.5 ± 0.2
17 (14) 58.4 ± 4.3 3/4/10 3/11/3 8.1 ± 1.9 11/6 3.3 ± 0.7 30 ± 1 2.3 ± 0.2
6 (4) 55.8 ± 5.2 1/2/3 4/2/0 6.2 ± 1.2 5/1 2.2 ± 0.4 23 ± 1 1.7 ± 0.2
EVL, Endoscopic variceal ligation; A, alcoholic liver disease; B, hepatitis B; C, hepatitis C. *According to the criteria of Japanese Research Society for Portal Hypertension.5
rence-free rate for all patients was 54.3% in the first year and 51.4% in the second year. The recurrencefree rate for type I patients was 100% in the first year and 91.7% in the second year, whereas the respective rate for type II patients was 5.9% and 5.9%, and the rate for type III patients was 100% and 100% (Fig.3). The cumulative 2-year recurrence-free rate was significantly higher in type I patients (p = 0.0001) and type III patients (p = 0.0002) than in type II patients. DISCUSSION EVL has been shown to have few complications and to be more effective for controlling bleeding esophageal varices than sclerotherapy.12,13 However, it has been suggested that use of EVL alone is associated with early variceal recurrence.14,15 Concerning the treatment procedure, standard EVL16 as well as intensive EVL17 are currently used. In standard EVL,16 only the varices observed endoscopically are ligated, whereas, in intensive EVL,17 both the varices and the surrounding mucosa in which varices may develop are treated. We previously reported18 that recurrent variceal bleeding could be significantly reduced when the aim of sclerotherapy was not only variceal eradication at endoscopy but also circumferential ulcer formation in the distal esophagus. Kitano et al.19 have also shown that recurrence can be significantly reduced by inducing fibrosis in the mucosa of the distal esophagus. We therefore performed ligation of both the varices and the intervening mucosa with the aim of creating a circumferential ulcer at the gastroesophageal junction. With this method, the extent of ligation was intermediate between standard EVL16 and intensive EVL.20 It seems likely that these different procedures would have different influences on portal hemodynamics and a different therapeutic effect. Portal hemodynamics have been studied by measurement of hepatic venous pressure gradient,21 VOLUME 50, NO. 6, 1999
transarterial portography,22 percutaneous transhepatic portography,23 Doppler US,24 magnetic resonance angiography,25 and color Doppler EUS.20 Although hepatic venous pressure gradient is considered to be an independent predictor of variceal bleeding,21 the relationship between portal hemodynamics and redevelopment of varices after treatment has not been studied in detail. Because esophageal varices are visualized in the venous phase on left gastric arteriogram,26 it is obvious that not only hepatofugal blood flow in the LGV but also blood flow in the left gastric artery are important for variceal development. Accordingly, we evaluated hemodynamics in the LGV including the findings from left gastric arteriography before and after EVL. Four distinct anatomic zones were identified around the gastroesophageal junction, and all were predominantly localized within the esophageal mucosa; the gastric zone was a 2 to 3 cm circumferential arrangement of varices in the proximal stomach organized longitudially within the lamina propria and submucosa. The palisade zone was identified as a continuation of the gastric zone beginning at the gastroesophageal junction and extending cephalad for 2 to 3 cm. Palisade veins anastomosed with four or five submucosal veins from the gastric zone. The perforationg zone was located 3 to 5 cm above the gastroesophageal junction, whereas the truncal zone was located 5 cm from the gastroesophageal junction and was 8 to 10 cm in length.8 The present angiographic study showed disappearance of blood flow to the perforating zone, but blood flow to the palisade zone was still present after EVL. Although all of the subjects still had a palisade zone after variceal eradication, only 16 of the type II patients and one of the type I patients developed variceal recurrence. This suggests that persistence of the palisade zone does not always lead to variceal recurrence. In 11 type I patients without recurrence, LGV blood flow was hepatopetal GASTROINTESTINAL ENDOSCOPY
771
Cumulative Recurrence-free Rates (%)
A Matsumoto, N Hamamoto, A Ohnishi, et al.
Months
Figure 3. Cumulative variceal recurrence-free rate for each group.
after treatment, indicating that blood from the left gastric artery did not ascend beyond the palisade zone but instead entered the LGV. Hashizume et al.22 found that when patients with high-risk esophageal varices had no cephalad collateral vessels on transarterial portography, varices were easily eradicated. Vianna et al.8 reported that blood flow in the palisade zone is “to and fro.” Koyanagi et al.27 measured the blood pressure in several parts of the portal system simultaneously at laparotomy and found that the LGV pressure was higher than the portal pressure in about one third of the subjects with esophageal varices. They suggested that the presence of a local hyperhemodynamic state28 in the upper gastric body is important for the formation of esophageal varices. Although their data were obtained under the nonphysiologic conditions of anesthesia and laparotomy, the results suggested that there are situations where the LGV has hepatopetal blood flow. In patients with portal hypertension, arteriovenous anastomoses in the submucosa at the gastric cardiac region may become dilated,28 leading to local elevation of the venous pressure. EVL blocks the collateral circulation and causes increased congestion in the gastric mucosa,29 which dilates the arteriovenous anastomoses. This is thought to lead to progression of the hyperhemodynamic state and elevation of the local venous pressure, with hepatopetal blood flow from the LGV to the portal vein being increased as a result. Additionally, all type III patients (who had hepatofugal blood flow in the LGV with a large extra-esophageal shunt before ligation) also remained recurrence-free. Kimura et al.30 reported that esophageal varices are less prominent when blood flow from the varices enters the azygos vein at the distal esophagus. According to Lo et al.,31 the portal pressure is reduced after ligation when there is marked shunt formation. Therefore, it was suggested that blood flow via the paraesophageal veins 772
GASTROINTESTINAL ENDOSCOPY
Endoscopic ligation of high-risk esophageal varices
entered the azygos vein to reduce the variceal pressure in these patients. Conversely, in type II patients with recurrence after a relatively short period, hepatofugal flow in the LGV increased as a result of the permanently elevated portal pressure, which seemed to promote variceal recanalization and recurrence via the remaining palisade zone without development of other collateral veins including paraesophageal veins. Our type II patients exhibited the same angiographic pattern as those studied by Hashizume et al.22 who had varices that were difficult to eradicate. Lo et al.31 reported that an elevated portal venous pressure after ligation is an important factor in the recurrence of bleeding. Arakawa et al.32 stated that the main body of the palisade vessels moves from the lamina propria into the submucosa with the development of varices. Hashizume et al.33 reported that the vascular pattern in the lower esophagus could be classified into two types on portography, the palisading type (<2 mm) and the bar type (≥2 mm). Our type II patients had more bar type vessels in the palisade zone than was found in the other groups. Hashizume et al.34 demonstrated that in patients whose palisade zone vessels were the bar type, the superficial and deep submucosal veins were markedly dilated and connected with the gastric varices. Nagamine et al.20 reported that patients who responded poorly to intensive EVL had highly developed gastric varices. In intensive EVL, an average of 41 ligations (range 18 to 76) and an average of 3.2 sessions per patient are required.20 In our method, an average of 21 ligations (range 23 to 31) and an average of 2.5 sessions per patient were required. The reason for fewer ligations and treatment sessions with our method compared with intensive EVL was probably that the proximal gastric mucosa just below the gastroesophageal junction was also ligated. Consequently, the beginning of the palisade zone was at least partly affected. In our patients, recurrence almost always occurred about 6 months after variceal eradication. In patients undergoing standard EVL16 or in those who respond poorly to intensive EVL, 20 recurrence usually appears about 6 months after treatment. The 2-year recurrence rate in our study was 48% (17 of 35) for all subjects; it has been reported to be about 45% after intensive EVL,20 similar to the rate in our study. Therefore it seems that the extent of ligation does not influence variceal recurrence, and that adequate treatment of the palisade zone in the deep submucosa is difficult by EVL alone.35 Controversy exists as to the efficacy of prophylactic sclerotherapy for esophageal varices.36 Because EVL is a safer method, prophylactic treatment can be justified. Lay et al.37 performed a prospective VOLUME 50, NO. 6, 1999
Endoscopic ligation of high-risk esophageal varices
randomized study in patients with viral induced cirrhosis with a similar proportion to those in the present study (76% vs. 83%) and found that preventive ligation improved the nonbleeding rate and the survival of Child-Pugh classification A patients. In the present study, 63% of our subjects were in ChildPugh classification A, and the cumulative recurrence-free rate was significantly higher in type I and III patients, which included more subjects with relatively good liver function compared with type II patients. Matsutani et al.24 performed a study using Doppler US and found that blood flow in the LGV changed from hepatopetal to hepatofugal with increasing portal pressure and that blood flow was hepatofugal in all patients with large coil-shaped varices. This may explain why there were more recurrence-free type I and III patients: although all of the subjects had a risk of variceal bleeding of greater than 90%, patients who required fewer ligations and treatment sessions to eradicate varices may have had less variceal blood flow than the type II patients who needed more treatment sessions. It is difficult to predict the clinical outcome of EVL based on pretreatment endoscopic features, but pretreatment angiography is useful in predicting whether esophageal varices may recur. Because angiography is relatively invasive, further study by noninvasive methods such as Doppler US,24 magnetic resonance angiography,25 or measurement of variceal pressure38 should be performed to accurately evaluate LGV hemodynamics. When prophylactic treatment is completed safely and rapidly, with initial variceal bleeding being prevented for a long period, the prognosis of patients with cirrhosis will be improved. From these results, we conclude that prophylactic EVL can be expected to achieve intermediate-term control of initial variceal bleeding. Because varices are likely to recur after a relatively short period in type II patients, it is important to promote fibrosis of the esophageal mucosa at potential sites of recurrence. Hence, more intensive treatment is recommended for the palisade zone in type II patients, for example, concurrent use of a sclerosant.39 A prospective randomized study is required to make a comparison between prophylactic β-blocker treatment40 and EVL using survival as the end point. REFERENCES 1. Copenhagen Esophageal Varices Sclerotherapy Project. Sclerotherapy after first variceal haemorrhage in cirrhosis: a randomized multicenter trial. N Engl J Med 1984;311:1594600. 2. Schuman BM, Beckman JW, Tedesco FJ, Griffin JW, Assad RT. Complications of endoscopic injection sclerotherapy: a review. Am J Gastroenterol 1987;82:823-30. VOLUME 50, NO. 6, 1999
A Matsumoto, N Hamamoto, A Ohnishi, et al.
3. Stiegmann GV, Cambre T, Sun JH. A new endoscopic elastic band ligating device. Gastrointest Endosc 1986;32:230-3. 4. Watanabe K, Kimura K, Matutani S, Ohto M, Okuda K. Portal hemodynamics in patients with gastric varices: a study in 230 patients with esophageal and/or gastric varices using portal vein catheterization. Gastroenterology 1988;95:434-40. 5. Idezuki Y, Japanese Research Society for Portal Hypertension. General rules for recording endoscopic findings of esophagogastric varices (1991). World J Surg 1995;19:420-3. 6. Beppu K, Inokuchi K, Koyanagi N, Nakayama S, Sakata H, Kitano S, et al. Prediction of variceal hemorrhage by esophageal endoscopy. Gastrointest Endosc 1981;27:213-8. 7. McCormack TT, Sims J, Eyre-Brook I, Kennedy H, Goepel J, Johnson AG, et al. Gastric lesions in portal hypertension: inflammatory gastritis or congestive gastropathy? Gut 1985;26: 1226-32. 8. Vianna A, Hayes PC, Moscoso G, Driver M, Portmann B, Westaby D, et al. Normal venous circulation of the gastroesophageal junction: a route to understanding varices. Gastroenterology 1987;93:876-89. 9. Pugh RNH, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60:646-9. 10. Kaplan EL, Meier P. Nonparametric estimation from incomplete observation. J Am Stat Assoc 1958;54:457-81. 11. Peto R, Pike MC, Armitage P. Design and analysis of randomized clinical trials requiring prolonged observation of each patient. Br J Cancer 1977;35:1-39. 12. Gimson AES, Ramage JK, Panos MZ, Hayllar K, Harrison PM, Williams R, et al. Randomised trial of variceal banding ligation versus injection sclerotherapy for bleeding oesophageal varices. Lancet 1993;342:391-4. 13. Laine L, Cook D. Endoscopic ligation compared with sclerotherapy for treatment of esophageal variceal bleeding: a meta-analysis. Ann Intern Med 1995;123:280-7. 14. Hashizume M, Ohta M, Kawanaka H, Kishihara F, Sugimachi K. Recurrence rate of oesophageal varices with endoscopic banding ligation followed by injection sclerotherapy [letter]. Lancet 1994;344:1643. 15. Stiegmann GV, Goff JS, Michaletz-Onody PA, Korula J, Lieberman D, Saeed ZA, et al. Endoscopic sclerotherapy as compared with endoscopic ligation for bleeding esophageal varices. N Engl J Med 1992;326:1527-32. 16. Masumoto H, Toyonaga A, Oho K, Iwao T, Tanikawa K. Ligation plus low-volume sclerotherapy for high-risk esophageal varices: comparisons with ligation therapy or sclerotherapy alone. J Gastroenterol 1998;33:1-5. 17. Nagamine N, Ido K, Ueno N, Kimura K, Kawamata T, Kawada H, et al. The usefulness of ultrasonic microprobe imaging for endoscopic variceal ligation. Am J Gastroenterol 1996;91:523-9. 18. Miyoshi H, Matsumoto A, Umegaki E, Oka M, Hirata I, Ohshiba S. Endoscopic evaluation of the therapeutic effect of sclerotherapy for esophageal varices. Gastrointest Endosc 1993;39:37-42. 19. Kitano S, Koyanagi N, Iso Y, Higashi H, Sugimachi K. Prevention of recurrence of esophageal varices after endoscopic injection sclerotherapy with ethanolamine oleate. Hepatology 1987;7:810-5. 20. Nagamine N, Ueno N, Tomiyama T, Aizawa T, Tano S, Wada S, et al. A pilot study on modified endoscopic variceal ligation using endoscopic ultrasonography with color doppler function. Am J Gastroenterol 1998;93:150-5. 21. Garcia-Tsao G, Groszmann RJ, Fisher RL, Conn HO, Atterbury CE, Glickman M. Portal pressure, presence of gastroesophageal varices and variceal bleeding. Hepatology 1985;5:419-24. GASTROINTESTINAL ENDOSCOPY
773
A Matsumoto, N Hamamoto, A Ohnishi, et al.
22. Hashizume M, Kitano S, Tanoue K, Ohta M, Tomikawa M, Kawanaka H, et al. Sclerotherapy-resistant esophageal varices with enormously enlarged cephalad collateral vessels predictable using portography. Hepatogastroenterol 42;1995:551-6. 23. Lunderquist A, Vang J. Transhepatic catheterization and obliteration of the coronary vein in patients with portal hypertension and esophageal varices. N Engl J Med. 1974; 291:646-9. 24. Matsutani S, Furuse J, Ishii H, Mizumoto H, Kimura K, Ohto M. Hemodynamics of the left gastric vein in portal hypertension. Gastroenterology 1993;105:513-8. 25. Edelman RR, Zhao B, Liu C, Wentz KU, Mattle HP, Finn JP, et al. MR angiography and dynamic flow evaluation of the portal venous system. AJR 1989;153:755-60. 26. Reuter SR, Atkin TW. High-dose left gastric angiography for demonstration of esophageal varices. Radiology 1972;105: 573-8. 27. Koyanagi N, Inokuchi K, Beppu K, Nagamine K, Hashizume M, Sugimachi K. Hyperdynamic circulation in the left gastric venous area in patients with portal hypertension: angiological assessment. Jpn J Surg 1985;15:134-9. 28. Koyanagi N, Inokuchi K, Nakayama S, Sakata H, Beppu K. Decreased arteriovenous flow resistance in the left gastric venous area in cirrhotic patients. Eur J Clin Invest 1981;11: 355-9. 29. Yoshikawa I, Murata I, Nakano S, Otsuki M. Effects of endoscopic variceal ligation on portal hypertensive gastropathy and gastric mucosal blood flow. Am J Gastroenterol 1998;93:71-4. 30. Kimura T, Moriyasu F, Kawasaki T, Someda H, Tamada T, Yamashita Y, et al. Relationship between esophageal varices and azygos vein evaluated by cineportography. Hepatology 1991;13:858-64. 31. Lo GH, Liang HL, Lai KH, Chang CF, Hwu JH, Chen SM, et al. The impact of endoscopic variceal ligation on the pressure of the portal venous system. J Hepatol 1996;24:74-80.
Endoscopic ligation of high-risk esophageal varices
32. Arakawa M, Kage M. The anatomy and pathomorphology of esophageal varices. In: Okuda K, Benhamou JP, editors. Portal hypertension. 1st ed. Tokyo: Springer-Verlag; 1991. p. 415-28. 33. Hashizume M, Kitano S, Yamaga H, Higashi H, Sugimachi K. Angioarchitectural classification of esophageal varices and paraesophageal veins in selective left gastric venography. Arch Surg 1989;124:961-6. 34. Hashizume M, Kitano S, Sugimachi K, Sueishi K. Threedimensional view of the vascular structure of the lower esophagus in clinical portal hypertension. Hepatology 1988; 8:1482-7. 35. Marks RD, Arnold MD, Baron TH. Gross and microscopic findings in the human esophagus after esophageal variceal band ligation: a postmortem analysis. Am J Gastroenterol 1993;88:272-4. 36. Fardy JM, Laupacis A. A meta-analysis of prophylactic endoscopic sclerotherapy for esophageal varices. Am J Gastroenterol 1994;89:1938-48. 37. Lay CS, Tsai YT, Teg CY, Shyu WS, Guo WS, Wu KL, et al. Endoscopic variceal ligation in prophylaxis of first variceal bleeding in cirrhotic patients with high-risk esophageal varices. Hepatology 1997;25:1346-50. 38. Nevens F, Bustami R, Scheys I, Lesaffre E, Fevery J. Variceal pressure is a factor predicting the risk of a first variceal bleeding: a prospective cohort study in cirrhotic patients. Hepatology 1998;27:15-9. 39. Bhargava DK, Pokharna R. Endoscopic variceal ligation versus endoscopic variceal ligation and endoscopic sclerotherapy: a prospective randomized study. Am J Gastroenterol 1997;92:950-3. 40. Lebrec D. Pharmacological treatment of portal hypertension: present and future. J Hepatol 1998;28:896-907.
Bound volumes available to subscribers Bound volumes of Gastrointestinal Endoscopy are available for the 1998 issues from the Publisher at a cost of $95.00 ($111.00 international, $118.77 Canada) for Vol. 49 (January-June) and Vol. 50 (July-December). Shipping charges are included. Each bound volume contains a subject and author index and all advertising is removed. Copies are shipped within 60 days after publication of the last issue of the volume. The binding is durable buckram with the journal name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact Mosby, Inc., Subscription Services Department, 11830 Westline Industrial Drive, St. Louis, MO 63146-3318, USA; phone 800-453-4351 or 314-453-4351.
774
GASTROINTESTINAL ENDOSCOPY
VOLUME 50, NO. 6, 1999
Background: Early recurrence of esophageal varices remains problematic after endoscopic variceal ligation. To evaluate the efficacy of prophylactic endoscopic ligation for esophageal varices at high risk for bleeding, the relationship between left gastric vein hemodynamics and variceal recurrence was investigated. Methods: Thirty-five patients with cirrhosis underwent endoscopic variceal ligation. Angiography was performed in all patients before treatment and after eradication of varices to study left gastric vein hemodynamics. Results: Before treatment, 12 patients had hepatopetal flow in the left gastric vein (type I), 17 had hepatofugal flow (type II), and 6 had hepatofugal flow with an extra-esophageal shunt (type III). In type I and III patients, the direction of blood flow in the left gastric vein did not change after eradication of varices. Type II patients showed bi-directional flow in the left gastric vein after treatment. Varices recurred in all but one type II patient and in one type I patient during follow-up (mean 36.7 months). The 2-year recurrence-free rate was higher in type I patients (p = 0.0001) and type III patients (p = 0.0002) than in type II patients. Conclusions: Prophylactic ligation seems to be a safe and useful procedure, especially in patients with type I or III hemodynamics in the left gastric vein before treatment. (Gastrointest Endosc 1999;50:768-74.)
In recent years, there have been marked advances in the endoscopic treatment of esophageal varices. Sclerotherapy, the conventional method, is highly effective for controlling acute bleeding and for preventing recurrence of bleeding.1 However, sclerotherapy has to be performed carefully to prevent complications associated with the injected sclerosant.2 Endoscopic variceal ligation (EVL) was developed by Stiegmann et al.3 as an alternative to sclerotherapy. This procedure is easier to perform and has achieved wide popularity, but its impact on the portal hemodynamics and long-term efficacy have not been clearly established. Because the left gastric vein (LGV) is the predominant collateral route for the formation of esophageal varices,4 evaluation of hemodynamics in the LGV can be used to select patients for prophylactic EVL. The present study was therefore performed to define which patients will redevelop high-risk esophageal varices Received November 18, 1998. For revision April 27, 1999. Accepted July 15, 1999. From the Second Department of Internal Medicine, Osaka Medical College, Takatsuki, Osaka, Japan. Reprint requests: Akio Matsumoto, MD, Second Department of Internal Medicine, Osaka Medical College, 2-7, Daigaku-cho, Takatsuki, Osaka, Japan; fax: 0726-84-6532. Copyright © 1999 by the American Society for Gastrointestinal Endoscopy 0016-5107/99/$8.00 + 0 37/1/101477 768
GASTROINTESTINAL ENDOSCOPY
after EVL by investigating the relationship between angiographic features in the LGV and variceal recurrence. PATIENTS AND METHODS After consultation with three physicians specializing in the treatment of varices, we selected 35 consecutive patients managed at our institution from April 1994 to April 1996. The selection criteria were as follows: (1) highrisk esophageal varices with a Beppu score less than –0.38 (i.e., varices were F2 [enlarged tortuous] or F3 [large coilshaped] in form and showed at least one of the following: red wale markings [++, +++], cherry red spots [++, +++], or hematocystic spots [+]);5,6 (2) no history of upper GI bleeding; (3) no gastric varices or ectopic varices; (4) no portal thrombosis; and (5) no malignant diseases including hepatocellular carcinoma. In all 35 patients, the underlying disease was liver cirrhosis, with the diagnosis of cirrhosis confirmed by clinical, biochemical, or histologic means. Portal hypertensive gastropathy was slight or absent, with a “snake skin pattern” limited to the upper body of the stomach, as described by McCormack et al.7 The patients all underwent prophylactic EVL alone. Before treatment, written informed consent was obtained from each patient, and the study protocol was reviewed and approved by the ethics committee of our medical college. Esophageal variceal ligation An endoscope (XQ20; Olympus Optical Co., Ltd., Tokyo, Japan), an overtube (Sumitomo Bakelite, Tokyo, VOLUME 50, NO. 6, 1999
Endoscopic ligation of high-risk esophageal varices
A
A Matsumoto, N Hamamoto, A Ohnishi, et al.
B
Figure 1. LGV hemodynamics before EVL (type I). Venous phase in celiac (A) and left gastric (B) arteriograms. Esophageal varices and hepatopetal blood flow in the LGV are visible. Large black arrow, LGV; white arrow, esophageal varices; small black arrows, palisade zone. Japan), and a Varioligator ligation kit (Top, Tokyo, Japan) were used. For sedation, flunitrazepam (Rohypnol, Roche, Tokyo, Japan) was given intravenously at a dose of 0.3 to 0.4 mg immediately before overtube placement. Ligation was begun just below the gastroesophageal junction along each variceal channel. Successive bands were placed as close to the site of initial ligation as possible. Each variceal channel was ligated with 3 to 4 elastic bands to a distance of 5 to 8 cm from the gastroesophageal junction. After 1 week, any remaining varices were ligated. Portions of the mucosa between the varices near the gastroesophageal junction were also ligated. After another week, eradication of the varices and formation of a circumferential ulcer at the gastroesophageal junction were confirmed. The goals of treatment were to achieve circumferential ulcer formation near the gastroesophageal junction and the endoscopic disappearance of varices. Additional treatment was performed when needed. The interval between each treatment session was 1 week. Angiography Before initial EVL and within 2 weeks after a sufficient therapeutic effect was confirmed endoscopically, angiography was performed in all patients. After an overnight fast, the patient was given an intramuscular injection of 25 mg of hydroxyzin (Atarax P, Pfizer, Tokyo, Japan). Digital subtraction angiography was performed with a DAR1200 (Shimazu, Kyoto, Japan). The catheters used were a 5F JA 003 (Catex, Tokyo, Japan) and a 3F SP Catheter (Terumo, Tokyo, Japan). A catheter was introduced via the right femoral artery by Seldinger’s method. After celiac and VOLUME 50, NO. 6, 1999
superior mesenteric arteriography, the splenic and left gastric arteries were visualized selectively. Iopamidol (iopamiron; Schering AG, Berlin, Germany) was injected at a rate of 4 mL/sec to a total volume of 20 mL for celiac and superior mesenteric arteriography, 3 mL/sec to a total volume of 20 mL for splenic arteriography, and 2 mL/sec to a total volume of 8 mL for left gastric arteriography. Angiograms were reviewed independently by two investigators, and hemodynamics in the LGV as well as the presence of collateral vessels were evaluated from the venous phase. In addition, the maximum vessel size in the palisade zone8 was measured. Follow-up Endoscopy was performed at 3-month intervals to detect any recurrence of varices and to observe portal hypertensive gastropathy. Recurrence was defined as the redevelopment of high-risk varices similar to those present before treatment. EVL was repeated as soon as recurrence or bleeding from varices was detected. Bleeding was defined as any episode of hematemesis or melena. The trial endpoints were recurrence or bleeding from esophageal varices, death, loss to follow-up, and maintenance of a recurrence-free status for 24 months after the treatment. Statistical analysis All data are shown as the mean ± SD. The unpaired Student’s t test was used to compare age, Child-Pugh score,9 number of EVL sessions, number of ligations, and maximum vessel size in the palisade zone. Fisher’s exact test was used for other data. Cumulative recurrence-free GASTROINTESTINAL ENDOSCOPY
769
A Matsumoto, N Hamamoto, A Ohnishi, et al.
Endoscopic ligation of high-risk esophageal varices
B
A
Figure 2. LGV hemodynamics before EVL (type II). Venous phase in superior mesenteric (A) and left gastric (B) arteriograms. Hepatofugal blood flow into esophageal varices via the LGV is visible. Large black arrow, LGV; white arrow, esophageal varices; small black arrows, palisade zone. curves were drawn using the Kaplan-Meier method,10 and significance was tested using the log-rank test.11 The level of significance was set at p < 0.05.
RESULTS The therapeutic goals were achieved in all patients. The esophageal ulcers resulting from ligation were small and did not hinder subsequent ligation of the mucosa between varices. There was no bleeding from the esophageal ulcers or from other sites, and no systemic complications. Observation of hemodynamics in the LGV before treatment revealed the following 3 patterns without observer error. In type I, there was no reflux from the splenic or portal vein and all blood flow entered the portal vein (hepatopetal flow) (Fig. 1). In type II, there was reflux from the splenic or portal vein into the esophageal varices via the LGV (hepatofugal flow) (Fig. 2). In type III, blood flow from the portal vein entered the esophageal varices and paraesophageal veins via the LGV (hepatofugal flow with a large extra-esophageal shunt). After treatment, type I and III patients exhibited the same hemodynamics in the LGV as before EVL, whereas type II patients had bidirectional blood flow in the LGV. No other collateral veins were observed in any patient before or after treatment. The palisade zone persisted in all patients. The clinical characteristics of the three groups are summarized in Table 1. There were no significant differences in age, gender, etiology of cir770
GASTROINTESTINAL ENDOSCOPY
rhosis, and extent of varices among them, but liver dysfunction was more severe in type II patients than in either type I patients (p = 0.002 by Child-Pugh classification, p = 0.042 by Child-Pugh score) or type III patients (p = 0.0352 by Child-Pugh score). The maximum vessel size in the palisade zone was larger in type II patients than in either type I or type III patients (p < 0.0001), whereas the number of EVL sessions and ligations was lower in type I or type III patients than in type II patients (p < 0.0001). No patient was lost to follow-up, therefore the final number of subjects was also 35. The mean follow-up period for type I, II, and III patients was 33.6 ± 7.4 months, 36.7 ± 6.5 months, and 36.8 ± 6.8 months, respectively. There were no episodes of GI bleeding, no development of gastric varices, no aggravation of portal hypertensive gastropathy, and no development of portal thrombosis. During followup, there was also no change in the Child-Pugh classification, and no deaths. Two of the type I patients and 3 of the type II patients developed hepatocellular carcinoma. Varices recurred in 17 patients, of whom only 1 had type I and 16 had type II hemodynamics before EVL. In the type II patients, varices always recurred within 9 months after treatment, with the shortest period being 3 months (mean 6.8 months). Among these patients, 3 had a second episode of recurrence. Among the 18 patients without recurrence, 11 had type I and 6 had type III hemodynamics before EVL. The cumulative recurVOLUME 50, NO. 6, 1999
Endoscopic ligation of high-risk esophageal varices
A Matsumoto, N Hamamoto, A Ohnishi, et al.
Table 1. Profile of patients undergoing prophylactic EVL No. of patients (men) Age (yr) Etiology of cirrhosis (A/B/C) Child-Pugh classification (A/B/C) Child-Pugh score Endoscopic features of varices* (F2/F3) No of EVL sessions No of ligations Maximum size in palisade vessels (mm)
Type I
Type II
Type III
12 (10) 56.3 ± 2.7 2/2/8 10/2/0 6.1 ± 1.2 9/3 2.2 ± 0.4 24 ± 1 1.5 ± 0.2
17 (14) 58.4 ± 4.3 3/4/10 3/11/3 8.1 ± 1.9 11/6 3.3 ± 0.7 30 ± 1 2.3 ± 0.2
6 (4) 55.8 ± 5.2 1/2/3 4/2/0 6.2 ± 1.2 5/1 2.2 ± 0.4 23 ± 1 1.7 ± 0.2
EVL, Endoscopic variceal ligation; A, alcoholic liver disease; B, hepatitis B; C, hepatitis C. *According to the criteria of Japanese Research Society for Portal Hypertension.5
rence-free rate for all patients was 54.3% in the first year and 51.4% in the second year. The recurrencefree rate for type I patients was 100% in the first year and 91.7% in the second year, whereas the respective rate for type II patients was 5.9% and 5.9%, and the rate for type III patients was 100% and 100% (Fig.3). The cumulative 2-year recurrence-free rate was significantly higher in type I patients (p = 0.0001) and type III patients (p = 0.0002) than in type II patients. DISCUSSION EVL has been shown to have few complications and to be more effective for controlling bleeding esophageal varices than sclerotherapy.12,13 However, it has been suggested that use of EVL alone is associated with early variceal recurrence.14,15 Concerning the treatment procedure, standard EVL16 as well as intensive EVL17 are currently used. In standard EVL,16 only the varices observed endoscopically are ligated, whereas, in intensive EVL,17 both the varices and the surrounding mucosa in which varices may develop are treated. We previously reported18 that recurrent variceal bleeding could be significantly reduced when the aim of sclerotherapy was not only variceal eradication at endoscopy but also circumferential ulcer formation in the distal esophagus. Kitano et al.19 have also shown that recurrence can be significantly reduced by inducing fibrosis in the mucosa of the distal esophagus. We therefore performed ligation of both the varices and the intervening mucosa with the aim of creating a circumferential ulcer at the gastroesophageal junction. With this method, the extent of ligation was intermediate between standard EVL16 and intensive EVL.20 It seems likely that these different procedures would have different influences on portal hemodynamics and a different therapeutic effect. Portal hemodynamics have been studied by measurement of hepatic venous pressure gradient,21 VOLUME 50, NO. 6, 1999
transarterial portography,22 percutaneous transhepatic portography,23 Doppler US,24 magnetic resonance angiography,25 and color Doppler EUS.20 Although hepatic venous pressure gradient is considered to be an independent predictor of variceal bleeding,21 the relationship between portal hemodynamics and redevelopment of varices after treatment has not been studied in detail. Because esophageal varices are visualized in the venous phase on left gastric arteriogram,26 it is obvious that not only hepatofugal blood flow in the LGV but also blood flow in the left gastric artery are important for variceal development. Accordingly, we evaluated hemodynamics in the LGV including the findings from left gastric arteriography before and after EVL. Four distinct anatomic zones were identified around the gastroesophageal junction, and all were predominantly localized within the esophageal mucosa; the gastric zone was a 2 to 3 cm circumferential arrangement of varices in the proximal stomach organized longitudially within the lamina propria and submucosa. The palisade zone was identified as a continuation of the gastric zone beginning at the gastroesophageal junction and extending cephalad for 2 to 3 cm. Palisade veins anastomosed with four or five submucosal veins from the gastric zone. The perforationg zone was located 3 to 5 cm above the gastroesophageal junction, whereas the truncal zone was located 5 cm from the gastroesophageal junction and was 8 to 10 cm in length.8 The present angiographic study showed disappearance of blood flow to the perforating zone, but blood flow to the palisade zone was still present after EVL. Although all of the subjects still had a palisade zone after variceal eradication, only 16 of the type II patients and one of the type I patients developed variceal recurrence. This suggests that persistence of the palisade zone does not always lead to variceal recurrence. In 11 type I patients without recurrence, LGV blood flow was hepatopetal GASTROINTESTINAL ENDOSCOPY
771
Cumulative Recurrence-free Rates (%)
A Matsumoto, N Hamamoto, A Ohnishi, et al.
Months
Figure 3. Cumulative variceal recurrence-free rate for each group.
after treatment, indicating that blood from the left gastric artery did not ascend beyond the palisade zone but instead entered the LGV. Hashizume et al.22 found that when patients with high-risk esophageal varices had no cephalad collateral vessels on transarterial portography, varices were easily eradicated. Vianna et al.8 reported that blood flow in the palisade zone is “to and fro.” Koyanagi et al.27 measured the blood pressure in several parts of the portal system simultaneously at laparotomy and found that the LGV pressure was higher than the portal pressure in about one third of the subjects with esophageal varices. They suggested that the presence of a local hyperhemodynamic state28 in the upper gastric body is important for the formation of esophageal varices. Although their data were obtained under the nonphysiologic conditions of anesthesia and laparotomy, the results suggested that there are situations where the LGV has hepatopetal blood flow. In patients with portal hypertension, arteriovenous anastomoses in the submucosa at the gastric cardiac region may become dilated,28 leading to local elevation of the venous pressure. EVL blocks the collateral circulation and causes increased congestion in the gastric mucosa,29 which dilates the arteriovenous anastomoses. This is thought to lead to progression of the hyperhemodynamic state and elevation of the local venous pressure, with hepatopetal blood flow from the LGV to the portal vein being increased as a result. Additionally, all type III patients (who had hepatofugal blood flow in the LGV with a large extra-esophageal shunt before ligation) also remained recurrence-free. Kimura et al.30 reported that esophageal varices are less prominent when blood flow from the varices enters the azygos vein at the distal esophagus. According to Lo et al.,31 the portal pressure is reduced after ligation when there is marked shunt formation. Therefore, it was suggested that blood flow via the paraesophageal veins 772
GASTROINTESTINAL ENDOSCOPY
Endoscopic ligation of high-risk esophageal varices
entered the azygos vein to reduce the variceal pressure in these patients. Conversely, in type II patients with recurrence after a relatively short period, hepatofugal flow in the LGV increased as a result of the permanently elevated portal pressure, which seemed to promote variceal recanalization and recurrence via the remaining palisade zone without development of other collateral veins including paraesophageal veins. Our type II patients exhibited the same angiographic pattern as those studied by Hashizume et al.22 who had varices that were difficult to eradicate. Lo et al.31 reported that an elevated portal venous pressure after ligation is an important factor in the recurrence of bleeding. Arakawa et al.32 stated that the main body of the palisade vessels moves from the lamina propria into the submucosa with the development of varices. Hashizume et al.33 reported that the vascular pattern in the lower esophagus could be classified into two types on portography, the palisading type (<2 mm) and the bar type (≥2 mm). Our type II patients had more bar type vessels in the palisade zone than was found in the other groups. Hashizume et al.34 demonstrated that in patients whose palisade zone vessels were the bar type, the superficial and deep submucosal veins were markedly dilated and connected with the gastric varices. Nagamine et al.20 reported that patients who responded poorly to intensive EVL had highly developed gastric varices. In intensive EVL, an average of 41 ligations (range 18 to 76) and an average of 3.2 sessions per patient are required.20 In our method, an average of 21 ligations (range 23 to 31) and an average of 2.5 sessions per patient were required. The reason for fewer ligations and treatment sessions with our method compared with intensive EVL was probably that the proximal gastric mucosa just below the gastroesophageal junction was also ligated. Consequently, the beginning of the palisade zone was at least partly affected. In our patients, recurrence almost always occurred about 6 months after variceal eradication. In patients undergoing standard EVL16 or in those who respond poorly to intensive EVL, 20 recurrence usually appears about 6 months after treatment. The 2-year recurrence rate in our study was 48% (17 of 35) for all subjects; it has been reported to be about 45% after intensive EVL,20 similar to the rate in our study. Therefore it seems that the extent of ligation does not influence variceal recurrence, and that adequate treatment of the palisade zone in the deep submucosa is difficult by EVL alone.35 Controversy exists as to the efficacy of prophylactic sclerotherapy for esophageal varices.36 Because EVL is a safer method, prophylactic treatment can be justified. Lay et al.37 performed a prospective VOLUME 50, NO. 6, 1999
Endoscopic ligation of high-risk esophageal varices
randomized study in patients with viral induced cirrhosis with a similar proportion to those in the present study (76% vs. 83%) and found that preventive ligation improved the nonbleeding rate and the survival of Child-Pugh classification A patients. In the present study, 63% of our subjects were in ChildPugh classification A, and the cumulative recurrence-free rate was significantly higher in type I and III patients, which included more subjects with relatively good liver function compared with type II patients. Matsutani et al.24 performed a study using Doppler US and found that blood flow in the LGV changed from hepatopetal to hepatofugal with increasing portal pressure and that blood flow was hepatofugal in all patients with large coil-shaped varices. This may explain why there were more recurrence-free type I and III patients: although all of the subjects had a risk of variceal bleeding of greater than 90%, patients who required fewer ligations and treatment sessions to eradicate varices may have had less variceal blood flow than the type II patients who needed more treatment sessions. It is difficult to predict the clinical outcome of EVL based on pretreatment endoscopic features, but pretreatment angiography is useful in predicting whether esophageal varices may recur. Because angiography is relatively invasive, further study by noninvasive methods such as Doppler US,24 magnetic resonance angiography,25 or measurement of variceal pressure38 should be performed to accurately evaluate LGV hemodynamics. When prophylactic treatment is completed safely and rapidly, with initial variceal bleeding being prevented for a long period, the prognosis of patients with cirrhosis will be improved. From these results, we conclude that prophylactic EVL can be expected to achieve intermediate-term control of initial variceal bleeding. Because varices are likely to recur after a relatively short period in type II patients, it is important to promote fibrosis of the esophageal mucosa at potential sites of recurrence. Hence, more intensive treatment is recommended for the palisade zone in type II patients, for example, concurrent use of a sclerosant.39 A prospective randomized study is required to make a comparison between prophylactic β-blocker treatment40 and EVL using survival as the end point. REFERENCES 1. Copenhagen Esophageal Varices Sclerotherapy Project. Sclerotherapy after first variceal haemorrhage in cirrhosis: a randomized multicenter trial. N Engl J Med 1984;311:1594600. 2. Schuman BM, Beckman JW, Tedesco FJ, Griffin JW, Assad RT. Complications of endoscopic injection sclerotherapy: a review. Am J Gastroenterol 1987;82:823-30. VOLUME 50, NO. 6, 1999
A Matsumoto, N Hamamoto, A Ohnishi, et al.
3. Stiegmann GV, Cambre T, Sun JH. A new endoscopic elastic band ligating device. Gastrointest Endosc 1986;32:230-3. 4. Watanabe K, Kimura K, Matutani S, Ohto M, Okuda K. Portal hemodynamics in patients with gastric varices: a study in 230 patients with esophageal and/or gastric varices using portal vein catheterization. Gastroenterology 1988;95:434-40. 5. Idezuki Y, Japanese Research Society for Portal Hypertension. General rules for recording endoscopic findings of esophagogastric varices (1991). World J Surg 1995;19:420-3. 6. Beppu K, Inokuchi K, Koyanagi N, Nakayama S, Sakata H, Kitano S, et al. Prediction of variceal hemorrhage by esophageal endoscopy. Gastrointest Endosc 1981;27:213-8. 7. McCormack TT, Sims J, Eyre-Brook I, Kennedy H, Goepel J, Johnson AG, et al. Gastric lesions in portal hypertension: inflammatory gastritis or congestive gastropathy? Gut 1985;26: 1226-32. 8. Vianna A, Hayes PC, Moscoso G, Driver M, Portmann B, Westaby D, et al. Normal venous circulation of the gastroesophageal junction: a route to understanding varices. Gastroenterology 1987;93:876-89. 9. Pugh RNH, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60:646-9. 10. Kaplan EL, Meier P. Nonparametric estimation from incomplete observation. J Am Stat Assoc 1958;54:457-81. 11. Peto R, Pike MC, Armitage P. Design and analysis of randomized clinical trials requiring prolonged observation of each patient. Br J Cancer 1977;35:1-39. 12. Gimson AES, Ramage JK, Panos MZ, Hayllar K, Harrison PM, Williams R, et al. Randomised trial of variceal banding ligation versus injection sclerotherapy for bleeding oesophageal varices. Lancet 1993;342:391-4. 13. Laine L, Cook D. Endoscopic ligation compared with sclerotherapy for treatment of esophageal variceal bleeding: a meta-analysis. Ann Intern Med 1995;123:280-7. 14. Hashizume M, Ohta M, Kawanaka H, Kishihara F, Sugimachi K. Recurrence rate of oesophageal varices with endoscopic banding ligation followed by injection sclerotherapy [letter]. Lancet 1994;344:1643. 15. Stiegmann GV, Goff JS, Michaletz-Onody PA, Korula J, Lieberman D, Saeed ZA, et al. Endoscopic sclerotherapy as compared with endoscopic ligation for bleeding esophageal varices. N Engl J Med 1992;326:1527-32. 16. Masumoto H, Toyonaga A, Oho K, Iwao T, Tanikawa K. Ligation plus low-volume sclerotherapy for high-risk esophageal varices: comparisons with ligation therapy or sclerotherapy alone. J Gastroenterol 1998;33:1-5. 17. Nagamine N, Ido K, Ueno N, Kimura K, Kawamata T, Kawada H, et al. The usefulness of ultrasonic microprobe imaging for endoscopic variceal ligation. Am J Gastroenterol 1996;91:523-9. 18. Miyoshi H, Matsumoto A, Umegaki E, Oka M, Hirata I, Ohshiba S. Endoscopic evaluation of the therapeutic effect of sclerotherapy for esophageal varices. Gastrointest Endosc 1993;39:37-42. 19. Kitano S, Koyanagi N, Iso Y, Higashi H, Sugimachi K. Prevention of recurrence of esophageal varices after endoscopic injection sclerotherapy with ethanolamine oleate. Hepatology 1987;7:810-5. 20. Nagamine N, Ueno N, Tomiyama T, Aizawa T, Tano S, Wada S, et al. A pilot study on modified endoscopic variceal ligation using endoscopic ultrasonography with color doppler function. Am J Gastroenterol 1998;93:150-5. 21. Garcia-Tsao G, Groszmann RJ, Fisher RL, Conn HO, Atterbury CE, Glickman M. Portal pressure, presence of gastroesophageal varices and variceal bleeding. Hepatology 1985;5:419-24. GASTROINTESTINAL ENDOSCOPY
773
A Matsumoto, N Hamamoto, A Ohnishi, et al.
22. Hashizume M, Kitano S, Tanoue K, Ohta M, Tomikawa M, Kawanaka H, et al. Sclerotherapy-resistant esophageal varices with enormously enlarged cephalad collateral vessels predictable using portography. Hepatogastroenterol 42;1995:551-6. 23. Lunderquist A, Vang J. Transhepatic catheterization and obliteration of the coronary vein in patients with portal hypertension and esophageal varices. N Engl J Med. 1974; 291:646-9. 24. Matsutani S, Furuse J, Ishii H, Mizumoto H, Kimura K, Ohto M. Hemodynamics of the left gastric vein in portal hypertension. Gastroenterology 1993;105:513-8. 25. Edelman RR, Zhao B, Liu C, Wentz KU, Mattle HP, Finn JP, et al. MR angiography and dynamic flow evaluation of the portal venous system. AJR 1989;153:755-60. 26. Reuter SR, Atkin TW. High-dose left gastric angiography for demonstration of esophageal varices. Radiology 1972;105: 573-8. 27. Koyanagi N, Inokuchi K, Beppu K, Nagamine K, Hashizume M, Sugimachi K. Hyperdynamic circulation in the left gastric venous area in patients with portal hypertension: angiological assessment. Jpn J Surg 1985;15:134-9. 28. Koyanagi N, Inokuchi K, Nakayama S, Sakata H, Beppu K. Decreased arteriovenous flow resistance in the left gastric venous area in cirrhotic patients. Eur J Clin Invest 1981;11: 355-9. 29. Yoshikawa I, Murata I, Nakano S, Otsuki M. Effects of endoscopic variceal ligation on portal hypertensive gastropathy and gastric mucosal blood flow. Am J Gastroenterol 1998;93:71-4. 30. Kimura T, Moriyasu F, Kawasaki T, Someda H, Tamada T, Yamashita Y, et al. Relationship between esophageal varices and azygos vein evaluated by cineportography. Hepatology 1991;13:858-64. 31. Lo GH, Liang HL, Lai KH, Chang CF, Hwu JH, Chen SM, et al. The impact of endoscopic variceal ligation on the pressure of the portal venous system. J Hepatol 1996;24:74-80.
Endoscopic ligation of high-risk esophageal varices
32. Arakawa M, Kage M. The anatomy and pathomorphology of esophageal varices. In: Okuda K, Benhamou JP, editors. Portal hypertension. 1st ed. Tokyo: Springer-Verlag; 1991. p. 415-28. 33. Hashizume M, Kitano S, Yamaga H, Higashi H, Sugimachi K. Angioarchitectural classification of esophageal varices and paraesophageal veins in selective left gastric venography. Arch Surg 1989;124:961-6. 34. Hashizume M, Kitano S, Sugimachi K, Sueishi K. Threedimensional view of the vascular structure of the lower esophagus in clinical portal hypertension. Hepatology 1988; 8:1482-7. 35. Marks RD, Arnold MD, Baron TH. Gross and microscopic findings in the human esophagus after esophageal variceal band ligation: a postmortem analysis. Am J Gastroenterol 1993;88:272-4. 36. Fardy JM, Laupacis A. A meta-analysis of prophylactic endoscopic sclerotherapy for esophageal varices. Am J Gastroenterol 1994;89:1938-48. 37. Lay CS, Tsai YT, Teg CY, Shyu WS, Guo WS, Wu KL, et al. Endoscopic variceal ligation in prophylaxis of first variceal bleeding in cirrhotic patients with high-risk esophageal varices. Hepatology 1997;25:1346-50. 38. Nevens F, Bustami R, Scheys I, Lesaffre E, Fevery J. Variceal pressure is a factor predicting the risk of a first variceal bleeding: a prospective cohort study in cirrhotic patients. Hepatology 1998;27:15-9. 39. Bhargava DK, Pokharna R. Endoscopic variceal ligation versus endoscopic variceal ligation and endoscopic sclerotherapy: a prospective randomized study. Am J Gastroenterol 1997;92:950-3. 40. Lebrec D. Pharmacological treatment of portal hypertension: present and future. J Hepatol 1998;28:896-907.
Bound volumes available to subscribers Bound volumes of Gastrointestinal Endoscopy are available for the 1998 issues from the Publisher at a cost of $95.00 ($111.00 international, $118.77 Canada) for Vol. 49 (January-June) and Vol. 50 (July-December). Shipping charges are included. Each bound volume contains a subject and author index and all advertising is removed. Copies are shipped within 60 days after publication of the last issue of the volume. The binding is durable buckram with the journal name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact Mosby, Inc., Subscription Services Department, 11830 Westline Industrial Drive, St. Louis, MO 63146-3318, USA; phone 800-453-4351 or 314-453-4351.
774
GASTROINTESTINAL ENDOSCOPY
VOLUME 50, NO. 6, 1999