- Email: [email protected]
Urgent transjugular intrahepatic portosystemic shunt for control of acute variceal bleeding
Vol. 93, No. I. 1998 ISSN OOO2-9270/98/$19.00 PI1 SOOO2-9270(97)00026-9
THE AMERICANJOURNALOF ‘iASTROENTEROLOCY Copyright 0 1998 by Am. COIL of Gastroenterology Published by Elsevier Science Inc.
Urgent Transjugular Intrahepatic Portosystemic Shunt for Control of Acute Variceal Bleeding Rafael Bafiares, M.D., Marta Casado, M.D., Jose Manuel Rodriguez-Laiiz, M.D., Fernando Camtiiiez, M.D., Ana Matilla, M.D., Antonio Echenagusia, M.D., Gonzalo Sim6, M.D., Belen Piqueras, M.D., Gerard0 Clemente, M.D., and Enrique Cos, M.D. Liver Unit. Department
of Gastroenterology, and Interventional Vascular Radiology Unit, Department Hospital General Universitario Gregorio MaraAdn, Madrid, Spain
patients is required before TIPS. (Am J Gastroenterol 1998;93:75-79. 0 1998 by Am. Coll. of GastroenterolopY)
Endoscopic sclerotherapy and pharmacological therapy are widely used in the treatment of acute variceal hemorrhage. However, they fail at arresting acute bleeding in 20-30% of bleeding episodes. The efficacy of transjugular intrahepatic portosystemic shunt (TIPS) in the prevention of recurrent variceal bleeding has been proved recently, but the effectiveness and safety of urgent TIPS in the treatment of acute variceal bleeding refractory to conventional therapy are still under evaluation. Methods: Over 4.5 yr, 358 variceal hemorrhage episodes were treated in our hospital. Pharmacological and endoscopic therapy failed to control hemorrhage in 93 episodes. Thirty-two patients died because of uncontrolled massive bleeding. In 56 patients, TIPS (Strecker stent) was performed after temporary control of the episode with balloon tamponade. Results: Eleven of 56 patients with urgent TIPS belonged to Child-Pugh class A, 22 to class B, and 23 to class C. The mean time between indication and insertion was 17 + 10 h (range 4-24 h). Control of bleeding was achieved in 53 patients (95%). Eight patients had recurrent bleeding at 1 month after TIPS, seven of them during the first week after the procedure. The l-month actuarial probability of rebleeding was 22%. The main complications of the procedure were massive hemoperitoneum (n = l), cardiorespiratory arrest (n = 2), cardiac failure (n = l), acute renal failure- (n = 2), and bacteremia (n = 7). Operative mortality (30 days) was 28%. The actuarial probability of survival at 30 days was significantly lower in Child-Pugh class C than in class A or B (48% vs 90%; p < 0.001). The presence of ascites, hepatic encephalopathy, and serum albumin level before TIPS were independent prognostic factors associated with the risk of operative mortality. Conclusions: Urgent TIPS is an effective alternative for the treatment of acute variceal bleeding refractory to endoscopic and pharmacological therapy, but sometimes is associated with major complications. Because of the high operative mortality rate in patients with severe liver failure, careful selection of Objective:
Received Apr.
2, 1997:
accepted Aug.
of Radiology,
INTRODUCTION Gastrointestinal bleeding due to ruptured esophageal or gastric varices is the most severe complication of portal hypertension syndrome. Despite therapeutic advances, its mortality rate is approximately 30% (l), and is higher in patients with severe liver failure (2, 3). Among the different therapeutic approaches used to control variceal hemorrhage, endoscopic sclerotherapy and pharmacological therapy with vasoactive drugs are the most widely used procedures, achieving primary hemostasis in >80% of cases (4, 5). Nevertheless, early rebleeding occurs frequently, reaching 30-50% (6, 7). Therefore, 25-30% of the variceal bleeding episodes initially treated with vasoactive drugs or sclerotherapy persist, requiring another therapeutic alternative. Emergency shunting surgery has been considered the best therapeutic approach when standard pharmacological and endoscopic treatments fail to control acute variceal bleeding because it has a high hemostatic efficacy and prevents variceal rebleeding. However, the operative mortality rate is >50% in most studies, being higher in patients with severe hepatic failure, thus reducing its application mainly in patients with Child class C cirrhosis (8, 9). Transjugular intrahepatic portosystemic shunt (TIPS) is a percutaneous procedure that allows partial decompression of the portal venous system by inserting an expandable calibrated stent between the intrahepatic branches of the portal vein and a hepatic vein. This nonoperative technique avoids the risks of a laparotomy, suggesting that TIPS could be used in patients with advanced hepatic cirrhosis. In preliminary studies, TIPS has proven effective in preventing variceal rebleeding (10-15). However, there is not enough information concerning the efficacy of emergency TIPS in acute variceal bleeding that is not controlled with pharmacological and endoscopic treatment.
26, 1997. 75
76
BARARES et al.
AJG - Vol. 93, No. I, 1998 358 VARICEAL
BLEEDING
EPISODES +
4 Primary
hemostasis
(ES** or Ph*)
Treatment
failure
ii Early rebleeding 71 Treatment
(ES or Ph)
93
265 Definitive
hemostasis
Alternative
1 treatment
(ES or Ph)#
194
(ES or Ph)#
.c
i.!
+Failure 40 Dead
Definitive
hemostasis
Definitive
30
1
4 for VB**** Emergency 11 29 I 1EMERGENCY
hemostasis
Failure 55
35
TIPS Emergeicy I
TIPS
27 + Itfailed)
TIPS 56 PATIENTS
S***Dead
p
1
6
for &**** 21
1
* Pharmacologic therapy: Somatostatin or Terlipressin **ES: Endoscopic sclerotherapy ““*PCS: Portacaval sbunt **** VB: Uncontrolled variceal bleeding ir/4 pakx~sdid not receive alternative treatment because they had diffuse hepatocellular carcinoma FIG.
1. Clinical course of patients with variceal bleeding.
The aim of this study was to asses the applicability, effectiveness, and safety of urgent TIPS in the treatment of acute variceal bleeding refractory to conventional therapy. MATERIALS
AND METHODS
From January 1992 to May 1996. 358 consecutive episodes of acute variceal bleeding in 277 patients were treated in the Gastrointestinal Bleeding Unit at the Hospital General Universitario “Gregorio Marafi6n.” The diagnosis of variceal hemorrhage was established by an emergency endoscopy if active bleeding was identified from a varix. signs of recent hemorrhage were found in a varix. or there was fresh blood in the stomach and esophageal or gastric varices as the only potential sources of upper _rastrointestinal bleeding (16). Control of hemorrhage. cessation of hemorrhage. rebleeding, and early rebleeding were defined according to the guidelines set at a consensus meeting held in Baveno, Italy, in 1990 (16). An episode of variceal bleeding was defined as refractory if primary hemostasis could not be obtained with endoscopic and pharmacological therapy or if uncontrolled early rebleeding occurred. Characteristics of the patients and treatments received are shown in Figure 1. Pharmacological treatment with somatostatin (250 pg/h after a bolus of 250 kg) or terlipressin (2 mg/4 h) or endoscopic sclerotherapy (ethanolamine oleate 3-5 ml injected in each varix column) achieved primary hemostasis in 74% of cases. When the initial therapy failed at arresting the acute bleeding episode, the patients received alternative treatment with endoscopic sclerotherapy or vasoactive drugs
(sclerotherapy when pharmacological treatment was the first-line treatment, and vice versa). This strategy controlled an additional 10% of cases. If these measures failed, a balloon tamponade was installed to achieve temporary control of bleeding until TIPS could be performed. The indication for TIPS was established when temporary hemodynamic stability was obtained. Twenty-one patients died because of massive bleeding despite the treatment used, including eight patients with diffuse hepatocellular carcinoma. In these patients, we could not achieve temporary control of the bleeding with balloon tamponade because of massive bleeding, so the TIPS indication could not be established. Portacaval shunt was performed in six patients. Early recurrent variceal bleeding (26% of the patients with primary hemostasis) was treated with the previously described approach, achieving hemostasis in 57% of the episodes. Eleven of 38 patients, including four with diffuse hepatocellular carcinoma, in whom initial treatment failed to stop bleeding, died because of massive rebleeding. TIPS was performed in the remaining patients. All patients who received TIPS were duly informed of the potential risks and benefits of the procedure and gave written consent. TIPS was performed as described previously (17). After TIPS, pressure measurements and ultrasonographic, endoscopic, and angiographic studies were scheduled at 1, 3, and 6 months and when clinically indicated. During the first 30 days after the procedure, angiography and pressure measurements were performed whenever the patients had recurrent bleeding. Shunt dysfunction was defined by the
A JG - January I998
TIPS
IN ACUTE
VARICEAL
BLEEDING
77
TABLE 1 Characteristics
Child-Pugh
of the Parients Having Emergency TIPS
A & B
90%
N*
Characteristic
Age (yr) Sex (male/female) Alcoholic liver disease Chjld class A B C Source of bleeding Esophageal Fundal Encephalopathy before TIPS Ascites before TIPS Transfusion requirements (packed red cell units)
512 11 39 (70%) 17 (30%) 33 (59%) 11 (25%) 22 (34%) 23 (41%) 37 (66%) 19 (34%) 12 (21%) 32 (57%) 10.5 2 5
* Total N = 56.
presence of a portacaval pressure gradient (PPG) > 12 mm Hg in at least one pressure measurement during the follow-up examinations. Qualitative variables were compared using Fisher’s exact test, and quantitative variables were compared with the unpaired Student’s t test. Results are expressed as mean + SD, and p < 0.05 was considered statistically significant. Survival curves were calculated by the Kaplan-Meier method and compared using the log-rank test. Multivariate analysis (stepwise logistic regression) was used to identify the independent prognostic factors associated with operative mortality (30 days) after TIPS. Statistical procedures were performed with a statistical analysis program package (RSIGMA, Horus Hardware, Madrid, Spain). RESULTS Fifty-six patients (15% of the variceal bleeding episodes) underwent urgent TIPS. All patients had portal hypertension due to chronic liver disease: alcoholic cirrhosis in 33 patients, cirrhosis due to hepatitis B or C in 21 patients, and primary biliary cirrhosis in two patients. Eleven patients (19%) belonged to Child-Pugh class A, 22 (39%) to class B, and 23 to class C (41%). The need for TIPS was significantly more frequent in patients whose source of bleeding was fundal varices than in patients with an esophageal source (25% vs 9%; p < 0.01). Characteristics of the patients are shown in Table 1. In all but one patient in whom TIPS was indicated, the procedure was completed. The mean time between indication and insertion of TIPS was 17 2 10 h (range 4-24 h). PPG significantly decreased after TIPS from 20 t 5 mm Hg to 8 _C4 mm Hg @ < 0.001). In all except two cases, PPG was < 12 mm Hg after the procedure. TIPS achieved hemostasis in 53 patients (95%). In the remaining three patients, PPG was < 12 mm Hg after the procedure. One patient had portal vein thrombosis before
I
Child-Pugh
I
C 48%
0
0
10
20
30 Days
FIG. 2. Actuarial probability of survival for emergency TIPS patients with severe liver disease (Child-Pugh class C) or mild to moderate liver disease (Child-Pugh class A or B).
TIPS placement, and acute stent thrombosis occurred thereafter. Another patient died of diffuse bleeding secondary to coagulopathy, and the other .died because of massive hemorrhage. Variceal rebleeding within 1 month after TIPS placement occurred in eight patients (14%), seven of them during the first week after the procedure. All patients who rebled had shunt dysfunction, requiring reintervention with balloon dilatation or the insertion of an additional stent. Four patients with shunt dysfunction received a new coaxial stent, leading to complete patency, and the remaining three died because of uncontrolled hemorrhage associated with severe liver failure and/or infection despite reintervention. The l-month actuarial probability of rebleeding was 22%. The rate of rebleeding was similar in patients with esophageal vat-ices and in patients with gastric varices. Twenty patients (35%) had complications associated with the proce.dure. Accidental intrahepatic biliary duct puncture or hepatic capsule puncture was noted in eight patients, resulting in clinically significant intraperitoneal hemorrhage in two. Cardiac arrest occurred immediately after the procedure in two patients. After resuscitation, one patient had severe anoxic brain damage and died 12 days after. The other patient fully recovered after resuscitation. The remaining complications were cardiac failure in one patient, acute renal failure in two, and bacteremia in seven. Severe complications related to the procedure (including intraperitoneal hemorrhage, cardiac arrest, cardiac failure, acute renal failure, and clinically relevant bacteremia) were significantly more frequent in patients with urgent TIPS than in those having it electively (64 patients in our series) (17% vs 4%; p < 0.05). Operative (30 days) mortality after TIPS was 28%. Four patients died because of variceal hemorrhage. The actuarial probability of survival at 30 days was significantly lower in Child-Pugh class C patients than in class A or B (48% vs 90%; p < 0.001) (Fig. 2). All but one patient with a Child-Pugh score > 11 points died during the first 30 days after the procedure, versus 11% of patients with a score 5 11 (p < 0.001).
78 Variables
BANARES
et al.
AJG - Vol. 93, No. 1. 1998
TABLE 2 Wiflz SigniJican? Predictive Value for Operative Mortality Univariate
Variable Ascites before TIPS Hepatic encephalopathy &fore TIPS Serum albumin C2.7 g/L before TIPS Serum bilirubin >2 mg/dl before TIPS Prothrombin activity GO% before TIPS Infection after TIPS
in
Analysis
Deaths
Survivors
D Value
16 (94%)
17 (42%)
8 (50%)
4 (10%)
-Co.01
13 (81%)
17 (42%)
CO.05
13 (81%)
19 (47%)
CO.05
13 (81%)
14 (35%)
-Co.01
11 (69%)
15 (37%)
CO.1
On univariate analysis, only the presence of hepatic encephalopathy and ascites; the values of serum albumin, serum bilirubin, and prothrombin activity before TIPS; and the presence of infection after TIPS were statistically significant as prognostic factors of operative mortality (Table 2). On multivariate analysis, hepatic encephalopathy before TIPS 0, = 0.04), ascites before TIPS @ = O.Ol), and serum albumin <2.7 g/L 0, = 0.02) were the only prognostic factors independently associated with the risk of operative mortality. DISCUSSION Acute variceal bleeding refractory to endoscopic and pharmacological treatment is a life-threatening complication of portal hypertension syndrome that produces a high mortality rate. Despite recent advances in endoscopic or pharmacological treatment, a large number of episodes of bleeding are refractory to conventional therapy. In these patients, emergency shunt surgery is associated with high mortality, mainly in those patients with poor liver function. Theoretically, TIPS has the advantages of shunt surgery in terms of portal decompression without the risks associated with a laparotomy in cirrhotic patients. Recently, the efficacy of TIPS was suggested in the prevention of rebleeding due to ruptured esophageal varices when pharmacological and endoscopic therapy had failed. For these reasons, TIPS has been considered a promising therapy for the treatment of acute refractory variceal bleeding (10-15). However, the role of TIPS in this particular setting has not been completely established, even though the majority of series mention this important aspect. The results of our study show the efficacy, safety, and applicability of TIPS performed on an urgent basis in 56 patients with acute bleeding when other treatments had failed. It is important to underline that this was not a controlled study aimed to compare TIPS with other therapeutic alternatives, so conclusions must be reached carefully. In this series, a relatively low number of patients had refractory bleeding (95 of 358), and TIPS could be performed in the majority of cases. TIPS was completed in all
but one patient in whom the indication was established. It is important to note that TIPS was applied in the majority of patients who achieved the indication. The TIPS procedure was associated with a significant hemodynamic effect, consisting of a marked decrease in PPG, which was associated with a high hemostatic rate, similar to the rate obtained with shunt surgery (18 -22). However, early rebleeding was relatively frequent and in all cases was associated with a PPG >12 mm Hg. This fact may be associated with improper TIPS placement; however, the technique was similar to that reported previously. Indeed, it is well known that the first week after an acute variceal bleeding episode is a period of high risk of rebleeding as a result of several factors, such as the presence of blood in the gut and the increase of portal pressure after blood transfusion. This suggests that urgent TIPS should be submitted to a more intensive early follow-up protocol. Despite this finding, the efficacy of TIPS in refractory variceal bleeding seems to be very high, mainly when compared with the rate of early variceal rebleeding after pharmacological or endoscopic treatment (4, 5). Onemonth rebleeding rates after emergency shunt surgery seem to be lower than those after TIPS in some series, but in other reports, the rebleeding rate after surgery was similar to that reported in our study (18, 19). Other series of urgent TIPS have shown similar values of rebleeding: Barange et al. (23) reported 23%, Jalan et al. (24) 15.6%, McCormick et al. (25) 30%, and Sanyal et al. (26) 18% when evaluating 40, 19,20, and 32 patients, respectively; these differences could be related to a different time of follow-up. It is important to note that patients with fundal vat-ices needed TIPS more frequently than patients with esophageal varices, suggesting that the hemostatic efficacy of endoscopic and pharmacological therapy is lower in the former group. Urgent TIPS could be used as a relevant therapeutic approach in patients with gastric variceal hemorrhage. Urgent TIPS occasionally is associated with a relatively high number of complications. It should be noted that most of them did not have important clinical consequences. However, on three occasions they played a major role in the death of the patients. It is important to underline that major complications were related to the urgency of the procedure and not to the degree of liver dysfunction. When considering mortality, it is important to note that the operative mortality rate was significantly higher in patients belonging to grade C of the Child-Pugh classification, reaching almost 100% in patients with a score >11 points. These data have been reported in other series of TIPS (23-26). Similarly, the operative mortality rate after surgical shunt is higher in patients with poor liver function (20-22). This fact is not related to a higher rate of rebleeding, suggesting that an impairment of hepatic function due to both the hemorrhagic episode and the decrease of portal flow induced by TIPS could be a life-threatening situation in patients with poor liver function. Supporting this finding, some investigators have suggested recently that TIPS induces a worsening in liver function tests (27, 28). These data
A JG - Januar?, 1998
TIPS IN ACUTE
suggest the need for a careful assessment of the indications for emergent TIPS in patients with poor liver function. Furthermore, all the variables with independent prognostic value for operative mortality were related to the degree of liver dysfunction. In summary, this study shows that urgent TIPS can be an effective alternative in the treatment of acute variceal bleeding refractory to. endoscopic and pharmacological therapy. The main limitations of this procedure are the requirement of temporary control of bleeding and the existence of important complications in some instances. The elevated operative mortality rate in patients with marked liver impairment necessitates careful selection before the procedure. Randomized clinical trials should be performed to compare the TIPS procedure with other alternatives such as shunting surgery to confirm these results. ACKNOWLEDGMENTS This study was supported by grants from Fondo de Investigaciones Sanitarias (FIS 0240/94). Reprint requests and correspondence: Rafael Baiiares, M.D., Seccidn de Hepatologfa. Servicio Aparato Digestivo 6300. Hospital General Universitario Gregorio Marat%%. Dr Esquerdo 46, 28007, Madrid, Spain.
REFERENCES JC, Calvet X, TerCs J. et al. Factors which influence the I. Garcia-Pagln prognosis of the first episode of variceal haemorrhage in patients with cirrhosis. In: Bosch J. Rodis J. eds. Recent advances in the pathophysiology and therapy of portal hypertension. Bologna, Italy: Serono Symposia Review, 1989:287-302. 2. Pugh RNH. Murray-Lyon MDIM. Dawson JL. et al. Transection of the oesophagus for bleeding oesophageal vances. Br J Surg 1973:60:64664. PA. Natural history and prognosis of 3. Burroughs AK. McCormick variceal bleeding. Baillieres Clin Gastroenterol 1992:6:437-50. 4. Planas R, Quer JC. Boix J. et al. A prospective randomized trial comparing somatostatin and sclerotherapy in the treatment of acute variceal bleeding. Hepatology 1994:20:370-5. 5. Shields R. Jenkins SA. Baxter JN. et al. A prospective randomized controlled trial comparing the efficacy of homatostatin with injection sclerotherapy in the control of bleeding oe\ophageal varices. J Hepatol 1992;16:128-37. De Dombal FT. Clarke IR. Clays SE. et al. Prognostic factors in upper gastrointestinal bleeding. Endoscopy 19X6: I8:6-IO. Cardin F. Gori G, McCormick PA. et al. A predictive model for very early rebleeding from varices. Gut l990:3 I :A 1204 (abstract 1. Orloff MJ. Dugnay RJ, Koster LD. Criteria for selection of patients for emergency portocaval shunt. Am J Surg 1977: 134: 146-5 I. Edmonson HT. Jackson FC. Juler GK. et al. Clinical investigation of the portocaval shunt: A report of early survival from emergency operation. Ann Surg 1971:173:372-80. 10. Sanyal AJ. Freedman AM, Purdum PP. et al. Transjugular intrahepatic
II
12.
13.
14.
IS.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
‘7.
28.
VARICEAL
BLEEDING
79
portosystemic shunt versus sclerotherapy for prevention of recurrent variceal hemorrhage: A randomized prospective trial. Gastroenterology 1994:106:A975 (abstract). LaBerge JM. Somberg KA, Lake JR, et al. Two-year outcome following transjugular intrahepatic portosystemic shunt for variceal bleeding: Results in 90 patients. Gastroenterology 1995;108: 1143-51. Riissle M, Haag K, Ochs A, et al. The transjugular intrahepatic portosystemic stent-shunt procedure for variceal bleeding. N Engl J Med 1994;330:165-71. Cabrera J, Maynar M, Granados R. et al. Transjugular intrahepatic portosystemic shunt vs sclerotherapy in the elective treatment of variceal hemorrhage. Gastroenterology 1996;110:832-9. Merli M. Riggio Oia I, Capocaccia L, et al. Transjugular intrahepatic portosystemic shunt versus endoscopic sclerotherapy in preventing variceal rebleeding: Preliminary results of a randomized controlled trial. Hepatology 1994:20:A43 (abstract). Shiffman ML, Jeffers L. Hoofnagle JH, et al. The role of transjugular intrahepatic portosystemic shunt for treatment of portal hypertension and its complications: A conference sponsored by the National Digestive Diseases Advisory Board. Hepatology 1996:22:1591-7. De Francis R. Pascal JP. Ancona E. et al. Definitions, methodology and therapeutic strategies in portal hypertension. J Hepatol 1992; l5:25661. Echenagusia AJ, Camuiiez F. Simo G. et al. Variceal hemorrhage: Efficacy of transjugular intrahepatic portasystemic shunts created with Strecker stents. Radiology 1994:192:235-40. Cello P, Grendell JH. Crass RA. et al. Endoscoqic sclerotherapy vs portacaval shunt in patients with severe cirrhosis and acute variceal hemorrhage. N Engl J Med 1984:311:1589-94. Orloff MJ. Bell RH Jr. Orloff MS. et al. Prospective randomized trial of emergency portacaval shunt and emergency medical therapy in unselected cirrhotic patients with bleeding varices. Hepatology 1994; 20:863-72. Johansen K. Prospective comparison of partial versus total portal decompression for bleeding esophageal varices. Sug Gynecol Obstet 1992;175:528-34. Chabert M, Page Y. Cadi F. et al. Calibrated side-to-side portacaval anastomosis in the treatment of bleeding from ruptured esophageal varices. Results in 38 cirrhotic patients. Gastroenterol Clin Biol 1990; I4:698-704. Malt RA. Abbott W. Warshaw A. et al. Randomized trial of emergency mesocaval shunt and portacaval shunts for bleeding esophageal varices. Am J Surg 1978;135:584-9. Barange K. Rousseau H. Vine1 JP. et al. TIPS as an emergency procedure in actively bleeding patients with advanced cirrhosis. Hepatology 1994:20:108A (abstract). Jalan R. John TG, Redhead DN. et al. A comparative study of emergency TIPSS and esophageal transection in the management of uncontrolled variceal hemorrhage. Am J Gastroenterol 1995;9011932-7. McCormick PA, Dick R. Panagou EB, et al. Emergency transjugular intrahepatic portasystemic stent shunting as salvage treatment for uncontrolled variceal bleeding. Br J Surg I994:81: 1324-27. Sanyal AJ. Freedman AM. Luketic VA. et al. Transjugular intrahepatic portosystemic shunts for patients with active variceal hemorrhage unresponsive to sclerotherapy. Gastroenterology 1996;l I 1:138-46. Haag K. Lu S. Rijsle M. et al. The influence of TIPS on portal hemodynamics and liver function with respect to hepatic encephalopathy and survival. Hepatology 1995:22:296A (abstract). Rodriguez-Laiz JM. Baiiares R. Echenagusia A, et al. Effects of transjugular intrahepatic portosystemic shunt (TIPS) on splanchnic and systemic hemodynamics. and hepatic function in patients with portal hypertension: Preliminary results. Dig Dis Sci 1995;40:2221-7.
THE AMERICANJOURNALOF ‘iASTROENTEROLOCY Copyright 0 1998 by Am. COIL of Gastroenterology Published by Elsevier Science Inc.
Urgent Transjugular Intrahepatic Portosystemic Shunt for Control of Acute Variceal Bleeding Rafael Bafiares, M.D., Marta Casado, M.D., Jose Manuel Rodriguez-Laiiz, M.D., Fernando Camtiiiez, M.D., Ana Matilla, M.D., Antonio Echenagusia, M.D., Gonzalo Sim6, M.D., Belen Piqueras, M.D., Gerard0 Clemente, M.D., and Enrique Cos, M.D. Liver Unit. Department
of Gastroenterology, and Interventional Vascular Radiology Unit, Department Hospital General Universitario Gregorio MaraAdn, Madrid, Spain
patients is required before TIPS. (Am J Gastroenterol 1998;93:75-79. 0 1998 by Am. Coll. of GastroenterolopY)
Endoscopic sclerotherapy and pharmacological therapy are widely used in the treatment of acute variceal hemorrhage. However, they fail at arresting acute bleeding in 20-30% of bleeding episodes. The efficacy of transjugular intrahepatic portosystemic shunt (TIPS) in the prevention of recurrent variceal bleeding has been proved recently, but the effectiveness and safety of urgent TIPS in the treatment of acute variceal bleeding refractory to conventional therapy are still under evaluation. Methods: Over 4.5 yr, 358 variceal hemorrhage episodes were treated in our hospital. Pharmacological and endoscopic therapy failed to control hemorrhage in 93 episodes. Thirty-two patients died because of uncontrolled massive bleeding. In 56 patients, TIPS (Strecker stent) was performed after temporary control of the episode with balloon tamponade. Results: Eleven of 56 patients with urgent TIPS belonged to Child-Pugh class A, 22 to class B, and 23 to class C. The mean time between indication and insertion was 17 + 10 h (range 4-24 h). Control of bleeding was achieved in 53 patients (95%). Eight patients had recurrent bleeding at 1 month after TIPS, seven of them during the first week after the procedure. The l-month actuarial probability of rebleeding was 22%. The main complications of the procedure were massive hemoperitoneum (n = l), cardiorespiratory arrest (n = 2), cardiac failure (n = l), acute renal failure- (n = 2), and bacteremia (n = 7). Operative mortality (30 days) was 28%. The actuarial probability of survival at 30 days was significantly lower in Child-Pugh class C than in class A or B (48% vs 90%; p < 0.001). The presence of ascites, hepatic encephalopathy, and serum albumin level before TIPS were independent prognostic factors associated with the risk of operative mortality. Conclusions: Urgent TIPS is an effective alternative for the treatment of acute variceal bleeding refractory to endoscopic and pharmacological therapy, but sometimes is associated with major complications. Because of the high operative mortality rate in patients with severe liver failure, careful selection of Objective:
Received Apr.
2, 1997:
accepted Aug.
of Radiology,
INTRODUCTION Gastrointestinal bleeding due to ruptured esophageal or gastric varices is the most severe complication of portal hypertension syndrome. Despite therapeutic advances, its mortality rate is approximately 30% (l), and is higher in patients with severe liver failure (2, 3). Among the different therapeutic approaches used to control variceal hemorrhage, endoscopic sclerotherapy and pharmacological therapy with vasoactive drugs are the most widely used procedures, achieving primary hemostasis in >80% of cases (4, 5). Nevertheless, early rebleeding occurs frequently, reaching 30-50% (6, 7). Therefore, 25-30% of the variceal bleeding episodes initially treated with vasoactive drugs or sclerotherapy persist, requiring another therapeutic alternative. Emergency shunting surgery has been considered the best therapeutic approach when standard pharmacological and endoscopic treatments fail to control acute variceal bleeding because it has a high hemostatic efficacy and prevents variceal rebleeding. However, the operative mortality rate is >50% in most studies, being higher in patients with severe hepatic failure, thus reducing its application mainly in patients with Child class C cirrhosis (8, 9). Transjugular intrahepatic portosystemic shunt (TIPS) is a percutaneous procedure that allows partial decompression of the portal venous system by inserting an expandable calibrated stent between the intrahepatic branches of the portal vein and a hepatic vein. This nonoperative technique avoids the risks of a laparotomy, suggesting that TIPS could be used in patients with advanced hepatic cirrhosis. In preliminary studies, TIPS has proven effective in preventing variceal rebleeding (10-15). However, there is not enough information concerning the efficacy of emergency TIPS in acute variceal bleeding that is not controlled with pharmacological and endoscopic treatment.
26, 1997. 75
76
BARARES et al.
AJG - Vol. 93, No. I, 1998 358 VARICEAL
BLEEDING
EPISODES +
4 Primary
hemostasis
(ES** or Ph*)
Treatment
failure
ii Early rebleeding 71 Treatment
(ES or Ph)
93
265 Definitive
hemostasis
Alternative
1 treatment
(ES or Ph)#
194
(ES or Ph)#
.c
i.!
+Failure 40 Dead
Definitive
hemostasis
Definitive
30
1
4 for VB**** Emergency 11 29 I 1EMERGENCY
hemostasis
Failure 55
35
TIPS Emergeicy I
TIPS
27 + Itfailed)
TIPS 56 PATIENTS
S***Dead
p
1
6
for &**** 21
1
* Pharmacologic therapy: Somatostatin or Terlipressin **ES: Endoscopic sclerotherapy ““*PCS: Portacaval sbunt **** VB: Uncontrolled variceal bleeding ir/4 pakx~sdid not receive alternative treatment because they had diffuse hepatocellular carcinoma FIG.
1. Clinical course of patients with variceal bleeding.
The aim of this study was to asses the applicability, effectiveness, and safety of urgent TIPS in the treatment of acute variceal bleeding refractory to conventional therapy. MATERIALS
AND METHODS
From January 1992 to May 1996. 358 consecutive episodes of acute variceal bleeding in 277 patients were treated in the Gastrointestinal Bleeding Unit at the Hospital General Universitario “Gregorio Marafi6n.” The diagnosis of variceal hemorrhage was established by an emergency endoscopy if active bleeding was identified from a varix. signs of recent hemorrhage were found in a varix. or there was fresh blood in the stomach and esophageal or gastric varices as the only potential sources of upper _rastrointestinal bleeding (16). Control of hemorrhage. cessation of hemorrhage. rebleeding, and early rebleeding were defined according to the guidelines set at a consensus meeting held in Baveno, Italy, in 1990 (16). An episode of variceal bleeding was defined as refractory if primary hemostasis could not be obtained with endoscopic and pharmacological therapy or if uncontrolled early rebleeding occurred. Characteristics of the patients and treatments received are shown in Figure 1. Pharmacological treatment with somatostatin (250 pg/h after a bolus of 250 kg) or terlipressin (2 mg/4 h) or endoscopic sclerotherapy (ethanolamine oleate 3-5 ml injected in each varix column) achieved primary hemostasis in 74% of cases. When the initial therapy failed at arresting the acute bleeding episode, the patients received alternative treatment with endoscopic sclerotherapy or vasoactive drugs
(sclerotherapy when pharmacological treatment was the first-line treatment, and vice versa). This strategy controlled an additional 10% of cases. If these measures failed, a balloon tamponade was installed to achieve temporary control of bleeding until TIPS could be performed. The indication for TIPS was established when temporary hemodynamic stability was obtained. Twenty-one patients died because of massive bleeding despite the treatment used, including eight patients with diffuse hepatocellular carcinoma. In these patients, we could not achieve temporary control of the bleeding with balloon tamponade because of massive bleeding, so the TIPS indication could not be established. Portacaval shunt was performed in six patients. Early recurrent variceal bleeding (26% of the patients with primary hemostasis) was treated with the previously described approach, achieving hemostasis in 57% of the episodes. Eleven of 38 patients, including four with diffuse hepatocellular carcinoma, in whom initial treatment failed to stop bleeding, died because of massive rebleeding. TIPS was performed in the remaining patients. All patients who received TIPS were duly informed of the potential risks and benefits of the procedure and gave written consent. TIPS was performed as described previously (17). After TIPS, pressure measurements and ultrasonographic, endoscopic, and angiographic studies were scheduled at 1, 3, and 6 months and when clinically indicated. During the first 30 days after the procedure, angiography and pressure measurements were performed whenever the patients had recurrent bleeding. Shunt dysfunction was defined by the
A JG - January I998
TIPS
IN ACUTE
VARICEAL
BLEEDING
77
TABLE 1 Characteristics
Child-Pugh
of the Parients Having Emergency TIPS
A & B
90%
N*
Characteristic
Age (yr) Sex (male/female) Alcoholic liver disease Chjld class A B C Source of bleeding Esophageal Fundal Encephalopathy before TIPS Ascites before TIPS Transfusion requirements (packed red cell units)
512 11 39 (70%) 17 (30%) 33 (59%) 11 (25%) 22 (34%) 23 (41%) 37 (66%) 19 (34%) 12 (21%) 32 (57%) 10.5 2 5
* Total N = 56.
presence of a portacaval pressure gradient (PPG) > 12 mm Hg in at least one pressure measurement during the follow-up examinations. Qualitative variables were compared using Fisher’s exact test, and quantitative variables were compared with the unpaired Student’s t test. Results are expressed as mean + SD, and p < 0.05 was considered statistically significant. Survival curves were calculated by the Kaplan-Meier method and compared using the log-rank test. Multivariate analysis (stepwise logistic regression) was used to identify the independent prognostic factors associated with operative mortality (30 days) after TIPS. Statistical procedures were performed with a statistical analysis program package (RSIGMA, Horus Hardware, Madrid, Spain). RESULTS Fifty-six patients (15% of the variceal bleeding episodes) underwent urgent TIPS. All patients had portal hypertension due to chronic liver disease: alcoholic cirrhosis in 33 patients, cirrhosis due to hepatitis B or C in 21 patients, and primary biliary cirrhosis in two patients. Eleven patients (19%) belonged to Child-Pugh class A, 22 (39%) to class B, and 23 to class C (41%). The need for TIPS was significantly more frequent in patients whose source of bleeding was fundal varices than in patients with an esophageal source (25% vs 9%; p < 0.01). Characteristics of the patients are shown in Table 1. In all but one patient in whom TIPS was indicated, the procedure was completed. The mean time between indication and insertion of TIPS was 17 2 10 h (range 4-24 h). PPG significantly decreased after TIPS from 20 t 5 mm Hg to 8 _C4 mm Hg @ < 0.001). In all except two cases, PPG was < 12 mm Hg after the procedure. TIPS achieved hemostasis in 53 patients (95%). In the remaining three patients, PPG was < 12 mm Hg after the procedure. One patient had portal vein thrombosis before
I
Child-Pugh
I
C 48%
0
0
10
20
30 Days
FIG. 2. Actuarial probability of survival for emergency TIPS patients with severe liver disease (Child-Pugh class C) or mild to moderate liver disease (Child-Pugh class A or B).
TIPS placement, and acute stent thrombosis occurred thereafter. Another patient died of diffuse bleeding secondary to coagulopathy, and the other .died because of massive hemorrhage. Variceal rebleeding within 1 month after TIPS placement occurred in eight patients (14%), seven of them during the first week after the procedure. All patients who rebled had shunt dysfunction, requiring reintervention with balloon dilatation or the insertion of an additional stent. Four patients with shunt dysfunction received a new coaxial stent, leading to complete patency, and the remaining three died because of uncontrolled hemorrhage associated with severe liver failure and/or infection despite reintervention. The l-month actuarial probability of rebleeding was 22%. The rate of rebleeding was similar in patients with esophageal vat-ices and in patients with gastric varices. Twenty patients (35%) had complications associated with the proce.dure. Accidental intrahepatic biliary duct puncture or hepatic capsule puncture was noted in eight patients, resulting in clinically significant intraperitoneal hemorrhage in two. Cardiac arrest occurred immediately after the procedure in two patients. After resuscitation, one patient had severe anoxic brain damage and died 12 days after. The other patient fully recovered after resuscitation. The remaining complications were cardiac failure in one patient, acute renal failure in two, and bacteremia in seven. Severe complications related to the procedure (including intraperitoneal hemorrhage, cardiac arrest, cardiac failure, acute renal failure, and clinically relevant bacteremia) were significantly more frequent in patients with urgent TIPS than in those having it electively (64 patients in our series) (17% vs 4%; p < 0.05). Operative (30 days) mortality after TIPS was 28%. Four patients died because of variceal hemorrhage. The actuarial probability of survival at 30 days was significantly lower in Child-Pugh class C patients than in class A or B (48% vs 90%; p < 0.001) (Fig. 2). All but one patient with a Child-Pugh score > 11 points died during the first 30 days after the procedure, versus 11% of patients with a score 5 11 (p < 0.001).
78 Variables
BANARES
et al.
AJG - Vol. 93, No. 1. 1998
TABLE 2 Wiflz SigniJican? Predictive Value for Operative Mortality Univariate
Variable Ascites before TIPS Hepatic encephalopathy &fore TIPS Serum albumin C2.7 g/L before TIPS Serum bilirubin >2 mg/dl before TIPS Prothrombin activity GO% before TIPS Infection after TIPS
in
Analysis
Deaths
Survivors
D Value
16 (94%)
17 (42%)
8 (50%)
4 (10%)
-Co.01
13 (81%)
17 (42%)
CO.05
13 (81%)
19 (47%)
CO.05
13 (81%)
14 (35%)
-Co.01
11 (69%)
15 (37%)
CO.1
On univariate analysis, only the presence of hepatic encephalopathy and ascites; the values of serum albumin, serum bilirubin, and prothrombin activity before TIPS; and the presence of infection after TIPS were statistically significant as prognostic factors of operative mortality (Table 2). On multivariate analysis, hepatic encephalopathy before TIPS 0, = 0.04), ascites before TIPS @ = O.Ol), and serum albumin <2.7 g/L 0, = 0.02) were the only prognostic factors independently associated with the risk of operative mortality. DISCUSSION Acute variceal bleeding refractory to endoscopic and pharmacological treatment is a life-threatening complication of portal hypertension syndrome that produces a high mortality rate. Despite recent advances in endoscopic or pharmacological treatment, a large number of episodes of bleeding are refractory to conventional therapy. In these patients, emergency shunt surgery is associated with high mortality, mainly in those patients with poor liver function. Theoretically, TIPS has the advantages of shunt surgery in terms of portal decompression without the risks associated with a laparotomy in cirrhotic patients. Recently, the efficacy of TIPS was suggested in the prevention of rebleeding due to ruptured esophageal varices when pharmacological and endoscopic therapy had failed. For these reasons, TIPS has been considered a promising therapy for the treatment of acute refractory variceal bleeding (10-15). However, the role of TIPS in this particular setting has not been completely established, even though the majority of series mention this important aspect. The results of our study show the efficacy, safety, and applicability of TIPS performed on an urgent basis in 56 patients with acute bleeding when other treatments had failed. It is important to underline that this was not a controlled study aimed to compare TIPS with other therapeutic alternatives, so conclusions must be reached carefully. In this series, a relatively low number of patients had refractory bleeding (95 of 358), and TIPS could be performed in the majority of cases. TIPS was completed in all
but one patient in whom the indication was established. It is important to note that TIPS was applied in the majority of patients who achieved the indication. The TIPS procedure was associated with a significant hemodynamic effect, consisting of a marked decrease in PPG, which was associated with a high hemostatic rate, similar to the rate obtained with shunt surgery (18 -22). However, early rebleeding was relatively frequent and in all cases was associated with a PPG >12 mm Hg. This fact may be associated with improper TIPS placement; however, the technique was similar to that reported previously. Indeed, it is well known that the first week after an acute variceal bleeding episode is a period of high risk of rebleeding as a result of several factors, such as the presence of blood in the gut and the increase of portal pressure after blood transfusion. This suggests that urgent TIPS should be submitted to a more intensive early follow-up protocol. Despite this finding, the efficacy of TIPS in refractory variceal bleeding seems to be very high, mainly when compared with the rate of early variceal rebleeding after pharmacological or endoscopic treatment (4, 5). Onemonth rebleeding rates after emergency shunt surgery seem to be lower than those after TIPS in some series, but in other reports, the rebleeding rate after surgery was similar to that reported in our study (18, 19). Other series of urgent TIPS have shown similar values of rebleeding: Barange et al. (23) reported 23%, Jalan et al. (24) 15.6%, McCormick et al. (25) 30%, and Sanyal et al. (26) 18% when evaluating 40, 19,20, and 32 patients, respectively; these differences could be related to a different time of follow-up. It is important to note that patients with fundal vat-ices needed TIPS more frequently than patients with esophageal varices, suggesting that the hemostatic efficacy of endoscopic and pharmacological therapy is lower in the former group. Urgent TIPS could be used as a relevant therapeutic approach in patients with gastric variceal hemorrhage. Urgent TIPS occasionally is associated with a relatively high number of complications. It should be noted that most of them did not have important clinical consequences. However, on three occasions they played a major role in the death of the patients. It is important to underline that major complications were related to the urgency of the procedure and not to the degree of liver dysfunction. When considering mortality, it is important to note that the operative mortality rate was significantly higher in patients belonging to grade C of the Child-Pugh classification, reaching almost 100% in patients with a score >11 points. These data have been reported in other series of TIPS (23-26). Similarly, the operative mortality rate after surgical shunt is higher in patients with poor liver function (20-22). This fact is not related to a higher rate of rebleeding, suggesting that an impairment of hepatic function due to both the hemorrhagic episode and the decrease of portal flow induced by TIPS could be a life-threatening situation in patients with poor liver function. Supporting this finding, some investigators have suggested recently that TIPS induces a worsening in liver function tests (27, 28). These data
A JG - Januar?, 1998
TIPS IN ACUTE
suggest the need for a careful assessment of the indications for emergent TIPS in patients with poor liver function. Furthermore, all the variables with independent prognostic value for operative mortality were related to the degree of liver dysfunction. In summary, this study shows that urgent TIPS can be an effective alternative in the treatment of acute variceal bleeding refractory to. endoscopic and pharmacological therapy. The main limitations of this procedure are the requirement of temporary control of bleeding and the existence of important complications in some instances. The elevated operative mortality rate in patients with marked liver impairment necessitates careful selection before the procedure. Randomized clinical trials should be performed to compare the TIPS procedure with other alternatives such as shunting surgery to confirm these results. ACKNOWLEDGMENTS This study was supported by grants from Fondo de Investigaciones Sanitarias (FIS 0240/94). Reprint requests and correspondence: Rafael Baiiares, M.D., Seccidn de Hepatologfa. Servicio Aparato Digestivo 6300. Hospital General Universitario Gregorio Marat%%. Dr Esquerdo 46, 28007, Madrid, Spain.
REFERENCES JC, Calvet X, TerCs J. et al. Factors which influence the I. Garcia-Pagln prognosis of the first episode of variceal haemorrhage in patients with cirrhosis. In: Bosch J. Rodis J. eds. Recent advances in the pathophysiology and therapy of portal hypertension. Bologna, Italy: Serono Symposia Review, 1989:287-302. 2. Pugh RNH. Murray-Lyon MDIM. Dawson JL. et al. Transection of the oesophagus for bleeding oesophageal vances. Br J Surg 1973:60:64664. PA. Natural history and prognosis of 3. Burroughs AK. McCormick variceal bleeding. Baillieres Clin Gastroenterol 1992:6:437-50. 4. Planas R, Quer JC. Boix J. et al. A prospective randomized trial comparing somatostatin and sclerotherapy in the treatment of acute variceal bleeding. Hepatology 1994:20:370-5. 5. Shields R. Jenkins SA. Baxter JN. et al. A prospective randomized controlled trial comparing the efficacy of homatostatin with injection sclerotherapy in the control of bleeding oe\ophageal varices. J Hepatol 1992;16:128-37. De Dombal FT. Clarke IR. Clays SE. et al. Prognostic factors in upper gastrointestinal bleeding. Endoscopy 19X6: I8:6-IO. Cardin F. Gori G, McCormick PA. et al. A predictive model for very early rebleeding from varices. Gut l990:3 I :A 1204 (abstract 1. Orloff MJ. Dugnay RJ, Koster LD. Criteria for selection of patients for emergency portocaval shunt. Am J Surg 1977: 134: 146-5 I. Edmonson HT. Jackson FC. Juler GK. et al. Clinical investigation of the portocaval shunt: A report of early survival from emergency operation. Ann Surg 1971:173:372-80. 10. Sanyal AJ. Freedman AM, Purdum PP. et al. Transjugular intrahepatic
II
12.
13.
14.
IS.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
‘7.
28.
VARICEAL
BLEEDING
79
portosystemic shunt versus sclerotherapy for prevention of recurrent variceal hemorrhage: A randomized prospective trial. Gastroenterology 1994:106:A975 (abstract). LaBerge JM. Somberg KA, Lake JR, et al. Two-year outcome following transjugular intrahepatic portosystemic shunt for variceal bleeding: Results in 90 patients. Gastroenterology 1995;108: 1143-51. Riissle M, Haag K, Ochs A, et al. The transjugular intrahepatic portosystemic stent-shunt procedure for variceal bleeding. N Engl J Med 1994;330:165-71. Cabrera J, Maynar M, Granados R. et al. Transjugular intrahepatic portosystemic shunt vs sclerotherapy in the elective treatment of variceal hemorrhage. Gastroenterology 1996;110:832-9. Merli M. Riggio Oia I, Capocaccia L, et al. Transjugular intrahepatic portosystemic shunt versus endoscopic sclerotherapy in preventing variceal rebleeding: Preliminary results of a randomized controlled trial. Hepatology 1994:20:A43 (abstract). Shiffman ML, Jeffers L. Hoofnagle JH, et al. The role of transjugular intrahepatic portosystemic shunt for treatment of portal hypertension and its complications: A conference sponsored by the National Digestive Diseases Advisory Board. Hepatology 1996:22:1591-7. De Francis R. Pascal JP. Ancona E. et al. Definitions, methodology and therapeutic strategies in portal hypertension. J Hepatol 1992; l5:25661. Echenagusia AJ, Camuiiez F. Simo G. et al. Variceal hemorrhage: Efficacy of transjugular intrahepatic portasystemic shunts created with Strecker stents. Radiology 1994:192:235-40. Cello P, Grendell JH. Crass RA. et al. Endoscoqic sclerotherapy vs portacaval shunt in patients with severe cirrhosis and acute variceal hemorrhage. N Engl J Med 1984:311:1589-94. Orloff MJ. Bell RH Jr. Orloff MS. et al. Prospective randomized trial of emergency portacaval shunt and emergency medical therapy in unselected cirrhotic patients with bleeding varices. Hepatology 1994; 20:863-72. Johansen K. Prospective comparison of partial versus total portal decompression for bleeding esophageal varices. Sug Gynecol Obstet 1992;175:528-34. Chabert M, Page Y. Cadi F. et al. Calibrated side-to-side portacaval anastomosis in the treatment of bleeding from ruptured esophageal varices. Results in 38 cirrhotic patients. Gastroenterol Clin Biol 1990; I4:698-704. Malt RA. Abbott W. Warshaw A. et al. Randomized trial of emergency mesocaval shunt and portacaval shunts for bleeding esophageal varices. Am J Surg 1978;135:584-9. Barange K. Rousseau H. Vine1 JP. et al. TIPS as an emergency procedure in actively bleeding patients with advanced cirrhosis. Hepatology 1994:20:108A (abstract). Jalan R. John TG, Redhead DN. et al. A comparative study of emergency TIPSS and esophageal transection in the management of uncontrolled variceal hemorrhage. Am J Gastroenterol 1995;9011932-7. McCormick PA, Dick R. Panagou EB, et al. Emergency transjugular intrahepatic portasystemic stent shunting as salvage treatment for uncontrolled variceal bleeding. Br J Surg I994:81: 1324-27. Sanyal AJ. Freedman AM. Luketic VA. et al. Transjugular intrahepatic portosystemic shunts for patients with active variceal hemorrhage unresponsive to sclerotherapy. Gastroenterology 1996;l I 1:138-46. Haag K. Lu S. Rijsle M. et al. The influence of TIPS on portal hemodynamics and liver function with respect to hepatic encephalopathy and survival. Hepatology 1995:22:296A (abstract). Rodriguez-Laiz JM. Baiiares R. Echenagusia A, et al. Effects of transjugular intrahepatic portosystemic shunt (TIPS) on splanchnic and systemic hemodynamics. and hepatic function in patients with portal hypertension: Preliminary results. Dig Dis Sci 1995;40:2221-7.