Endoscopic cyanoacrylate versus transjugular intrahepatic portosystemic shunt for gastric variceal bleeding: a single-center U.S. analysis

ORIGINAL ARTICLE: Clinical Endoscopy

Endoscopic cyanoacrylate versus transjugular intrahepatic portosystemic shunt for gastric variceal bleeding: a single-center U.S. analysis Nicholas J. Procaccini, MD, JD, MS, Abdullah M. S. Al-Osaimi, MD, Patrick Northup, MD, MS, Curtis Argo, MD, MS, Stephen H. Caldwell, MD Charlottesville, Virginia, USA

Background and Objectives: Gastric variceal hemorrhage treatment remains a difficult issue for clinicians. There is controversy regarding whether first-line treatment should be endoscopic therapy with cyanoacrylate glue or placement of a transjugular intrahepatic portosystemic shunt (TIPS). We compared these methods on the basis of rebleeding, survival, and complications. Design, Setting, Patients, and Interventions: This was a retrospective cohort analysis of cirrhotic patients with gastric variceal hemorrhage treated with endoscopic cyanoacrylate therapy or TIPS placement at a single U.S. center from 1997 to 2007. The groups were compared for rebleeding at 72 hours, 3 months, and 1 year; survival rates at 3 months and 1 year; and acute and extended complications and morbidity. Main Outcome Measurements and Results: A total of 105 patients were included. There were no significant pretreatment differences between the 2 groups in age, sex, MELD (Model for End-Stage Liver Disease) score at the time of admission, or cause of liver disease. There were no significant differences in rebleeding at 72 hours, 3 months, and 1 year; survival at 3 months and 1 year; and aggregate long-term survival or acute complications. However, the TIPS group had a higher rate of long-term morbidity requiring hospitalization (41% with a TIPS and 1.6% in the cyanoacrylate arm, P ! .0001). Limitations: Retrospective and uncontrolled samples. Conclusion: In patients with similar characteristics, cyanoacrylate therapy performed as well as a TIPS in controlling and preventing gastric variceal hemorrhage with no significant differences in survival. Patients receiving cyanoacrylate therapy experienced significantly less long-term morbidity related to therapy than patients who received a TIPS. Cyanoacrylate therapy appears to be safe and effective and compares favorably with TIPS therapy. (Gastrointest Endosc 2009;70:881-7.)

See CME section; p. 976. Copyright ª 2009 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 doi:10.1016/j.gie.2009.03.1169

bleed within 2 years after diagnosis.2 Of these, gastric variceal bleeding constitutes approximately 10% to 20% of variceal bleeding in portal hypertensive patients.3,4 Gastric varices can present unique difficulties; unlike esophageal varices, they pose particular therapeutic challenges because of their size and location. Traditional endoscopic therapies with band ligation or sclerotherapy have proven to be significantly less effective in acute control and prophylaxis of gastric varices.4,5 Often, gastric variceal bleeding is associated with a more severe course than esophageal variceal bleeding, with greater transfusion requirements, higher costs of care, and worse outcomes.4,5 The increased severity of gastric varices is likely associated with their distinct anatomy and physiology, particularly with fundal varices. Anatomically, fundal varices (Sarin classes6 GOV2 [gastroesphageal varices, gastric varices

www.giejournal.org

Volume 70, No. 5 : 2009 GASTROINTESTINAL ENDOSCOPY 881

More than 5.5 million people in the United States are estimated to have chronic liver disease.1 This number is expected to increase in the future with the increasing incidence of hepatitis C and obesity-related nonalcoholic steatohepatitis. In conjunction with this increased prevalence of cirrhosis, chronic complications such as variceal bleeding will also increase. It is estimated that approximately 20% to 30% of cirrhotic patients will have a variceal Abbreviation: TIPS, transjugular intrahepatic portosystemic shunt. DISCLOSURE: All authors disclosed no financial relationship relevant to this publication.

Gastric varices therapy

associated with esophageal varices type 2 located in the fundus] and IGV1 [isolated gastric varices, type 1, no relation to esophageal varices type 1 located in the fundus]) are associated with high-flow vessels originating in gastrorenal, gastrophrenic, or cardiophrenic shunts, which may have more aggressive bleeding potential.7 In terms of first-line therapy for gastric varices, there are 2 primary options: endoscopically administered tissue adhesives and radiologic transjugular intrahepatic portosystemic shunt (TIPS) placement. TIPS placement has been the first-line therapy in the United States, primarily because of the limited availability of cyanoacrylate tissue adhesives, which are widely used in other regions of the world. Although not approved by the U.S. Food and Drug Administration, the availability of similar polymers such as an 8-carbon octylcyanoacrylate (Dermabond; Johnson & Johnson, New Brunswick, NJ) and other formulations of the 4 carbon N-butyl-2-cyanoacrylate (Indermil, Covidien, Bermuda) has led to their increasing use for treating gastric varices in the United States, and many centers are accruing significant experience.8 However, there have been no prospective head-to-head comparisons in the literature of cyanoacrylate glue therapy and TIPS in North America. A retrospective study from the United Kingdom comparing a TIPS and N-butyl-2-cyanoacrylate (Histoacryl; B. Braun, Melsungen, Germany) found a higher early (!30 days) rebleeding rate for cyanoacrylate therapy, but also found no significant differences in survival, morbidity, and complications.9 Although the study tracked survival outcomes for only a 6-month period, their analysis found that cyanoacrylate therapy achieved significantly lower costs than a TIPS.9 Notably, the rates of early rebleeding with cyanoacrylate therapy of approximately 30% reported in this study were higher than other studies from the literature.8,10-13 We recently reported data from a pilot study of 92 patients who underwent enbucrilate (N-butyl-2-cyanoacrylate) (Histoacryl; B. Braun) therapy in a multicenter study at the University of Virginia, University of Maryland, and University of California, San Diego.14 Given the debate regarding optimal use of endoscopic cyanoacrylate therapy versus TIPS placement and the need for a prospective and randomized trial, the previous performance of this pilot study afforded us the opportunity to compare extended outcomes in patients treated over a 10-year period.14 The primary endpoints of this study were survival at 3 months and 1 year, the secondary endpoints were rebleeding at 72 hours, 3 months, and 1 year, and secondary outcome measures were complication rates of advancing liver disease and morbidity during the entire follow-up period.

Procaccini et al

Capsule Summary What is already known on this topic d

Transjugular intrahepatic portosystemic shunt (TIPS) placement has been first-line therapy for gastric variceal bleeding in the United States because of the limited availability of cyanoacrylate tissue adhesives.

What this study adds to our knowledge d

d

In a retrospective cohort analysis of 105 cirrhotic patients with gastric variceal hemorrhage, cyanoacrylate therapy was as good as TIPS placement in controlling and preventing rebleeding. Patients receiving cyanoacrylate therapy experienced significantly less long-term morbidity related to therapy than those who received a TIPS.

rylate glue therapy or TIPS placement for the treatment of gastric variceal bleeding. All patients in both groups provided informed consent before their respective procedures. The patients analyzed in the cyanoacrylate arm of the study were 61 patients with cirrhosis and gastric variceal hemorrhage treated at the University of Virginia who participated in a multicenter study conducted from 1997 to 2004. Cyanoacrylate therapy was used with additional informed consent with approval from the University of Virginia Institutional Review Board. Patients were eligible if they had actively bleeding gastric varices or evidence of recent upper GI bleeding as evidenced by hematemesis or melena and a concurrent decrease in hemoglobin requiring transfusion, with the presence of gastric varices with high-risk stigmata, including red wale spots, adherent clot to the varix, or a plug (nipple) sign on the varix confirmed by endoscopy. Exclusion criteria for the cyanoacrylate study were proven hepatocellular carcinoma; pulmonary arteriovenous malformation; allergy to cyanoacrylate, ethiodol or iodine; uncertainty of the source of upper GI bleeding; pregnancy; known cardiac septal defects; and known hepatopulmonary syndrome. Patients from the TIPS arm of the study were all patients treated at the University of Virginia for actively or recently hemorrhaging gastric varices with a TIPS from 1997 to 2007. All of these patients had previous endoscopic confirmation of gastric varices as the source of bleeding identical to the cyanoacrylate group.

Initial patient management

This study was a retrospective examination of patients with cirrhosis who underwent either endoscopic cyanoac-

The majority of patients were transferred from outside hospitals with ongoing octreotide infusion (because of the unavailability of terlipressin in the United States). Bleeding was determined by a significant decrease in hemoglobin or hematocrit requiring packed red cell transfusion or evidence of hematemesis or melena with hemodynamic insufficiency. All patients in both the cyanoacrylate and

882 GASTROINTESTINAL ENDOSCOPY Volume 70, No. 5 : 2009

www.giejournal.org

METHODS

Procaccini et al

TIPS arms had diagnostic endoscopy before the intervention to confirm the presence of gastric varices as the source of bleeding. In the immediate stabilization period, subjects from both the TIPS and cyanoacrylate arms received Sengsten-Blakemore tubes as needed.

Injection technique All personnel were outfitted with protective eye gear, and the subjects’ eyes were covered before working with the cyanoacrylate compound. Refrigerated (5 C) enbucrilate was mixed with ethiodol in a 1:1 ratio and gently stirred and drawn into a 23-gauge Markon-Haber cannula (WilsonCook Medical, Winston-Salem, NC). The Markon-Haber cannula was flushed with sterile ethiodol and then loaded with the enbucrilate-ethiodol mixture. Primarily, flexible sigmoidoscopes were used during the therapeutic portion of the procedure. The shorter 60- to 70-cm endoscopes allowed increased tip deflection, which facilitated visualization of the fundus. The endoscope suction channel was flushed with lubricant (olive oil) before insertion of the catheter. Simethicone (3-4 mL) was added to the olive oil to decrease adherence of the cyanoacrylate polymer to the working channel during aspiration after injection. The preloaded sclerotherapy catheter was advanced through the endoscope into the stomach, and the needle was inserted directly in the gastric varix. Typically, a small drop of polymer mixture was expressed just before injection of the varix to ensure liquidity of the cyanoacrylate mixture. Sterile water (0.8 mL) was flushed through the catheter to deliver the glue mixture into the varix, and the needle was then withdrawn while glue was still flowing from the cannula to decrease the risk of needle embedment into the varix. Typically, 2 mL of the 1:1 mixture of lipodol and cyanoacrylate was administered per injection. A new injection needle was used for each injection to ensure that there was no premature polymerization within the cannula. Injections were repeated until the gastric varices appeared to be occluded, as judged by probing with a blunt probe. The median number of injections per session was 3.

Gastric varices therapy

hematocrit requiring packed red cell transfusion or evidence of hematemesis or melena with hemodynamic insufficiency. The bleed was determined to be from gastric varices if it met the same endoscopic criteria as described in the initial evaluation. Patients in the TIPS group generally had follow-up scheduled in hepatology clinic at the University of Virginia, but did not have follow-up endoscopy arranged per protocol as in the cyanoacrylate group. Rebleeding in the TIPS group was defined by identical criteria as for the cyanoacrylate group.

Data acquisition and statistical evaluation Complete data for patients in the study were acquired through the electronic medical records system at the University of Virginia. When there were gaps in the data, outside medical records were requested for clarification. The data review and requests for medical records were approved by the Institutional Review Board at the University of Virginia. Data for survival and rebleeding were determined by clinic and hospital follow-up where possible, and when no clinical information was available, by the Social Security Death Index for survival data.16 Cumulative survival was compared for the 2 groups by using a log-rank test. A Kaplan-Meier survival curve was used to illustrate all-cause survival for the 2 groups with censoring at death and to account for variable follow-up time between the 2 groups. Morbidity events were determined by subsequent hospitalization based on the medical records. Results of the study were analyzed by using the Student t test (2 tailed) for continuous data and Fisher’s exact test (2 tailed) for categorical outcomes. Statistical computation was calculated by using SAS version 9.1.3 (SAS Institute, Cary, NC) and Minitab version 15 (Minitab, State College, Pa).

RESULTS

Patients in the cyanoacrylate group were followed with systematic endoscopic surveillance. One to 4 weeks after their initial gluing procedure, patients underwent endoscopy with repeat gluing, if indicated, and then every 3 to 6 months or with any episode of rebleeding. Rebleeding was defined by a significant decrease in hemoglobin or

A total of 105 patients were analyzed for this study, 61 patients in the cyanoacrylate arm and 44 patients in the TIPS arm. The majority of patients who had gastric variceal bleeding from 1997 to 2004 underwent cyanoacrylate therapy. After the study ended in 2004, all patients underwent TIPS placement for gastric varices. Of the 44 patients who underwent TIPS placement, 29 (65.9%) received the stent with GORE-TEX as opposed to 15 (34.1%) who received the uncovered metal stent. One subject who received cyanoacrylate therapy was excluded from the analysis because he underwent elective splenorenal shunt surgery before meeting any clinical endpoints for analysis. At presentation, there were no statistically significant differences between the 2 cohorts for the parameters of age, sex, severity of liver disease (based on the MELD [Model for EndStage Liver Disease] score measured on admission before treatment with cyanoacrylate or TIPS placement), or cause of liver disease (Table 1). Also, there were no significant differences in preprocedure hemoglobin readings between the 2 groups.

www.giejournal.org

Volume 70, No. 5 : 2009 GASTROINTESTINAL ENDOSCOPY 883

TIPS technique TIPS procedures, using a standard technique described previously,15 were performed with the patient under general anesthesia. Notably, in 2003, the stent used for the TIPS was changed from the traditional metallic stent to a metallic stent with GORE-TEX (Viatorr; W.L. Gore & Associates, Flagstaff, Ariz) fabric lining. Overall, 29 (65.9%) received the stent with GORE-TEX and 15 (34.1%) received the uncovered metal stent.

Subsequent patient monitoring

Gastric varices therapy

Procaccini et al

TABLE 1. Patient characteristics Characteristic

Cyanoacrylate

TIPS

61

44

54.5 (14.1)

52.0 (14.1)

.37

Male

43 (70.5)

26 (59.1)

.30

Female

18 (29.5)

18 (40.9)

No. patients Age, y, mean (SD)

P value

Sex, no. (%)

MELD, start, mean (SD)

15.4 (6.3)

15.7 (8.3)

.85

Cause of liver disease, no. (%) Viral

16 (26.2)

14 (31.8)

.28

Alcohol

24 (39.3)

17 (38.6)

.32

Autoimmune

3 (4.9)

4 (9.1)

.43

NASH

6 (9.8)

5 (11.3)

.48

11 (18.0)

4 (9.1)

.20

Cryptogenic Primary biliary cirrhosis

1 (1.6)

Figure. 1. Kaplan-Meier survival curve for TIPS and cyanoacrylate groups.

0

TIPS, Transjugular intrahepatic portosystemic shunt; SD, standard deviation; MELD, Model for End-Stage Liver Disease; NASH, nonalcoholic steatohepatitis.

For the primary endpoint of patient survival at 3 months and 1 year, there were no significant differences between cyanoacrylate therapy and TIPS placement. Patients in the cyanoacrylate group had survival rates of 83.6% and 72.1% at 3 months and 1 year, respectively, compared with 79.6% and 66.7% in the TIPS group (P Z .61 and P Z .66, respectively) (Fig. 1). The average number of months that the cyanoacrylate group was followed was 73.9 months, and 47.8 months in the TIPS group (Table 2). For long-term survival, patients in the cyanoacrylate group survived for an average of 41 months and patients in the TIPS group survived for an average of 40 months (P Z .96 by log-rank test). The Kaplan-Meier survival curve for the 2 groups is illustrated as Figure 1. For the secondary endpoints of rebleeding, there were also no significant differences found in rebleeding rates of gastric varices for the 2 groups at 72 hours, 3 months, and 1 year. If patients died, they were removed from analysis for the following time period, so it was possible for therapies to have lower rates of bleeding at 1 year than 3 months. Data on rebleeding for 4 patients in the cyanoacrylate group at 3 months, 4 patients in cyanoacrylate group at 1 year, 6 patients in the TIPS group at 3 months, and 4 patients in the TIPS group at 1 year were unavailable, which accounts for the differences in the patients analyzed from the survival and rebleeding groups from the corresponding time periods illustrated in Table 2. For patients who received cyanoacrylate therapy, rebleeding rates were 6.9%, 10.6%, and 10.0% at 72 hours, 3 months,

and 1 year, respectively, whereas for those with a TIPS, the rebleeding rates were 9.5%, 20.7%, and 25.0% (P Z .71, P Z .25, and P Z .16, respectively). There was no significant difference found in the 2 groups for nongastric variceal bleeding events such as esophageal variceal bleeding and gastric antral vascular ectasia (Table 2). Both groups required frequent postprocedural followup after their initial intervention for gastric variceal hemorrhage. Patients in the cyanoacrylate group required a median of 2 procedures for obliteration of their gastric varices. This does not include regular endoscopic variceal surveillance, which was recommended on at least a yearly basis after obliteration of gastric varices. Patients in the TIPS group underwent a mean of 1.6 TIPS procedures or revisions. This does not include radiologic surveillance of the TIPS or any endoscopic surveillance. We found no significant differences in immediate procedural complications between the 2 groups, although there were 2 episodes of a possible embolus in the

884 GASTROINTESTINAL ENDOSCOPY Volume 70, No. 5 : 2009

www.giejournal.org

Figure. 2. Acute complications and extended morbidity.

Procaccini et al

Gastric varices therapy

TABLE 2. Patient outcomes

Outcome

TABLE 3. Extended morbidity by type

Cyanoacrylate

TIPS

P value

Gastric variceal rebleeding (for surviving patients), rebleed/total (%)

Procedure TIPS Encephalopathy

72 h

4/58 (6.9)

4/42 (9.5)

.71

3 mo

5/47 (10.6)

6/29 (20.7)

.25

1y

4/40 (10.0)

6/24 (25.0)

.16

Extended morbidity

11

TIPS stenosis

6

ARF

1

Cyanoacrylate Encephalopathy after procedure

1

TIPS, Transjugular intrahepatic portosystemic shunt; ARF, acute renal failure.

Survival (%) 3 mo

51/61 (83.6) 35/44 (79.55) .61

1y

44/61 (72.1) 28/42 (66.7)* .66

Average survival, mo

41

Immediate procedural complications, no. (%)z

4 (6.6)

Extended morbidity, no. (%)x

1 (1.6)

40 4 (9.8)

.96y .72

18 (41.0) !.0001

Crossover to TIPS, no.

7

d

Crossover to cyanoacrylate, no.

d

1

Injection procedures, median

2

d

Non–gastric variceal bleeding (%)

10/61 (16.4)

7/44 (15.9)

.42

Progression to liver transplant, no.

2

5

.18

TIPS, Transjugular intrahepatic portosystemic shunt. *Two patients in this group did not have follow-up that lasted for 1 year. yBy log-rank test. zMajor procedural complications: those directly related to procedure requiring hospitalization or extended hospitalization or fatalities. xExtended procedural complications include complications as a consequence of procedure.

There was a significantly higher rate of long-term morbidity in the TIPS group than in the cyanoacrylate group. Morbidity was defined as known adverse events related to the procedural intervention that prolonged or caused a patient to be hospitalized. In the TIPS groups, 18 of 41 (41.0%) patients were subsequently hospitalized for nonbleeding-related problems (hepatic encephalopathy, TIPS stenosis, or acute renal failure caused by adjustment of diuretics after TIPS placement) compared with 1 patient (1.6%) (hospital admission for encephalopathy after follow-up endoscopy procedure) (P ! .0001, Fisher’s exact test) from the cyanoacrylate group (Table 3, Fig. 2). Ultimately, 5 patients from the TIPS group (9.1%) and 2 patients (3.3%) from the cyanoacrylate group went on to receive a liver transplant (P Z .18).

DISCUSSION

cyanoacrylate group: in 1 patient in whom fatal electromechanical dissociation developed immediately after injection and another patient who had early rebleeding requiring rescue TIPS placement owing to an existing portal vein thrombus. The patient experienced cardiac arrest several hours after TIPS placement and had a cerebral stroke without radiographic evidence of central nervous system embolic material, which should have been evident if this were glue related because of the radiopacity of ethiodol. In the TIPS group, there was 1 death involving a patient who was found to have portal vein thrombosis during TIPS placement and had a rebleeding episode after TIPS placement and a pulseless electrical activity code within 24 hours of the procedure.

In this retrospective cohort study, we compared the efficacy of cyanoacrylate therapy with that of TIPS for the treatment of gastric variceal hemorrhage. Until now, there have been no published studies comparing these 2 modalities in a single center in the United States. Despite the retrospective nature of the study, there were no significant differences in age, sex, cause of disease, or severity of liver disease (as measured by MELD score) between those patients treated with a TIPS and those treated with cyanoacrylate. No significant differences were found in rebleeding (at 72 hours, 3 months, and 1 year) or survival at 3 months and 1 year or longer term survival between the 2 groups, although the raw percentages of rebleeding and survival were more favorable in the cyanoacrylate group (Table 2). However, the group receiving a TIPS experienced significantly more problems that required hospitalization during follow-up, because of a high incidence of encephalopathy with a TIPS. The incidence of encephalopathy encountered in our study seems to be consistent with previous long-term retrospective analysis in patients who received a TIPS.17

www.giejournal.org

Volume 70, No. 5 : 2009 GASTROINTESTINAL ENDOSCOPY 885

Gastric varices therapy

Procaccini et al

Taken together, the data suggest equivalent performance in acute treatment and secondary prophylaxis of gastric variceal hemorrhage, and survival with TIPS placement and cyanoacrylate therapy with fewer long-term complications with cyanoacrylate therapy. The performance of cyanoacrylate therapy seen in this study is consistent with results of previous work18 and with the Baveeno IV Consensus Workshop that recommended cyanoacrylate injection as a first-line therapy in bleeding gastric varices.19 However, more definitive recommendations await the performance of randomized, controlled trials comparing the available interventions. The limitations of our analysis warrant discussion. The study is retrospective, which increases the likelihood of systematic bias, although the two cohorts were well matched. The cyanoacrylate group had more postprocedural surveillance, with planned endoscopies and followup clinic visits. In addition, the use of b-blockers for variceal prophylaxis was based on the discretion of the treating physicians in both groups, and there is no available information on whether one group was more likely to be on these medications. Some patients in the TIPS group underwent procedures before the switch to covered GORE-TEX stents, which may have lower rates of stenosis and potentially of rebleeding, although there were no statistically significant differences in a subgroup analysis, and the majority of the patients in the study received the Viatorr stents. Finally, it is unknown whether newer cyanoacrylates have major safety or efficacy differences from the 4-carbon enbucrilate-ethiodol mixture. There are other factors that may point to favorability of one procedure over another for specific patients. For patients with ongoing difficulties with refractory ascites, a TIPS would be attractive for alleviating both problems. Patients with previous episodes of esophageal variceal bleeding that have not been well controlled with medical prophylaxis or endoscopic techniques such as band ligation may also be better candidates for a TIPS. Overall, 7 patients in our study who had undergone cyanoacrylate therapy later went on to receive a TIPS, most for reasons of refractory bleeding, suggesting that there are cases in which a TIPS will play a role if endoscopic therapy fails. In contrast, patients with portal vein thrombosis, hepatocellular cancer that may preclude TIPS placement, or a history of recurrent encephalopathy are likely better served with cyanoacrylate therapy. We foresee increasing numbers of patients with the difficult problem of gastric varices in the future in conjunction with the explosive increase of cirrhosis from hepatitis C and nonalcoholic steatohepatitis. In the setting of increasing medical costs and diminishing resources because of demographic population shifts in the United States, it is incumbent on clinicians to use limited resources in the most cost-effective manner possible. Data from 2 previous studies, 1 by our group at the University of Virginia and another from the United Kingdom, showed significant cost benefits with cyanoacrylate therapy.9,20

Although we did not do a formal cost analysis as part of this study, the cost of TIPS placement, not including anesthesia costs, ranges from $25,000 to more than $50,000.21 This does not include previous radiologic imaging (often with magnetic resonance imaging/magnetic resonance angiography) or echocardiography, which is usually required before TIPS placement. Additionally, virtually all patients undergoing TIPS placement will have undergone at least 1 endoscopy procedure previously to confirm gastric varices as the source of bleeding. The cost of upper endoscopy with injection therapy is approximately $750 (based on 2007 Medicare reimbursement for CPT code 43243), with an approximate $250 to $400 extra for glue administration. The median number of therapeutic procedures for patients in our glue cohort was 2. With the increasing availability of tissue adhesives and the likely increased prevalence of their use in the United States, comparisons between the efficacy and cost of endoscopic therapy and a TIPS are very relevant. Our analysis indicates that endoscopic cyanoacrylate therapy is safe and as clinically effective as a TIPS with a significantly lower rate of long-term morbidity. Based on previous publications including data from our initial experience with enbucrilate, it is evident that cyanoacrylate therapy is also more attractive from a cost perspective than a TIPS. A prospective, controlled trial comparing these 2 interventions is warranted to determine optimal management.

886 GASTROINTESTINAL ENDOSCOPY Volume 70, No. 5 : 2009

www.giejournal.org

ACKNOWLEDGMENT We thank Elizabeth Hespenheide for her contributions to the original cyanoacrylate studies. REFERENCES 1. The Burden of Gastrointestinal Diseases. Bethesda (Md): American Gastroenterological Association; 2001. p. 41-2. 2. de Francis R, Primignani M. Natural history of portal hypertension in patients with cirrhosis. Clin Liver Dis 2001;5:645-63. 3. Ryan BM, Stockbrugger RW, Ryan JM. A pathophysiologic, gastroenterologic, and radiologic approach to the management of gastric varices. Gastroenterology 2004;126:1175-89. 4. Trudeau W, Prindiville T. Endoscopic injection sclerosis in bleeding gastric varices. Gastrointest Endosc 1986;32:264-8. 5. Chey WD, Elta GH. Natural history of gastric varices. Gastroenterology 1993;105:599-602. 6. Sarin SK, Lahoti D, Saxena SP, et al. Prevalence, classification and natural history of gastric varices: a long-term follow-up study in 568 portal hypertensive patients. Hepatology 1992;16:1343-9. 7. Matsumoto A, Takimoto K. Gastric fundal varices: new aspects of non-surgical treatment in Japan. Nat Clin Pract Gastroenterol Hepatol 2006;3:4-5. 8. Rengstorff DS, Binmoeller KF. A pilot study of 2-octyl cyanoacrylate injection for treatment of gastric fundal varices in humans. Gastrointest Endosc 2004;59:553-8. 9. Mahadeva S, Bellamy MC, Kessel D, et al. Cost effectiveness of n-butyl-2-cyanoacrylate (Histoacryl) glue injections versus transjugular intrahepatic portosystemic shunt in the management of acute gastric variceal bleeding. Am J Gastroenterol 2003;98:2688-93.

Procaccini et al

Gastric varices therapy

10. D’Imperio N, Piemontese A, Baroncini D, et al. Evaluation of undiluted n-butyl-2-cyanoacrylate in the endoscopic treatment of upper gastrointestinal tract varices. Endoscopy 1996;28:239-43. 11. Miyazaki S, Yoshida T, Harada T, et al. Injection sclerotherapy for gastric varices using n-butyl-2-cyanoacrylate and ethanolamine oleate. Hepatogastroenterology 1998;45:1155-8. 12. Ogawa K, Ishikawa S, Naritaka Y, et al. Clinical evaluation of endoscopic injection sclerotherapy using n-butyl-2-cyanoacrylate for gastric variceal bleeding. J Gastroenterol Hepatol 1999;14:245-50. 13. Kind R, Gugliemi A, Rodella L, et al. Bucrylate treatment of bleeding gastric varices: 12 years’ experience. Endoscopy 2000;32:512-9. 14. Caldwell SH, Hespenheide EE, Greenwald BD, et al. Enbucrilate for gastric varices: extended experience in 92 patients. Aliment Pharmacol Ther 2007;26:49-59. 15. Rossle M, Haag K, Ochs A, et al. Transjugular intrahepatic portosystemic stent-shunt procedure for variceal bleeding. N Engl J Med 1994;330:165-71. 16. Social Security Death Index. Available at: http://ssdi.rootsweb.ancestry.com/. Accessed January 10, 2008. 17. ter Borg PCG, Hollemans M, van Buuren HR, et al. Transjugular intrahepatic portosystemic shunts: long-term patency and clinical results for a cohort observed for 3-9 years. Radiology 2004;237:537-45.

18. Tripathi D, Ferguson JW, Therapondos G, et al. Review article: recent advances in the management of bleeding gastric varices. Aliment Pharmacol Ther 2006;24:1-17. 19. De Franchis R. Evolving consensus in portal hypertension. Report of the Baveeno IV Consensus Workshop on methodology of diagnosis and therapy in portal hypertension. J Hepatol 2005;43:167-76. 20. Greenwald BD, Caldwell SH, Hespenheide EE, et al. N-2-butyl cyanoacrylate for bleeding gastric varices, a United States pilot study and cost analysis. Am J Gastroenterol 2003;98:1982-8. 21. Lake JR. The role of transjugular portosystemic shunting in patients with ascites. N Engl J Med 2000;342:1745-7.

www.giejournal.org

Volume 70, No. 5 : 2009 GASTROINTESTINAL ENDOSCOPY 887

Received September 30, 2008. Accepted March 20, 2009. Current affiliations: Division of Gastroenterology and Hepatology, Digestive Health Center of Excellence. University of Virginia Health System, Charlottesville, Virginia, USA. Reprint requests: Nicholas Procaccini, MD, Division of Gastroenterology and Hepatology, Department of Internal Medicine, University of Virginia Health System, PO Box 800708, Charlottesville, VA 22908-0708.