- Email: [email protected]
Transjugular intrahepatic portosystemic shunt in the treatment of ectopic varices secondary to portal hypertension
W1411
obtained by analysis o[ twemy-ibur-honr ambulatory electrocardiographic recordings using both the time and the treqnency domain methods, Results: The analysis of heart rate variabdily showed significantly lower values ot the following parameters in cirrhotic patients: InHF (patmnts: 5 2 • 03, controls: 5,8 • 06; p = 0.03); rMSSD (patients: 32 • 10, controls:47 • 10 ms; p = O 0 1 ) and pNN50% ( patients: 9 • controls: 19• p=O.02), indicating a decreased parasympathetic activaty in this group of patients Conclusions: Our results suggest impaired efferent vagal function in patients with Child's grade B nonalcoholic cirrhosis. This abnormality may contribute to the hemodynamic disturbances of cirrhosis.
Transjugular Intrahepatic Portosystemic Shunt (TIPS) Before Surgery in Cirrhotic Patients: A Retrospective Comparative Study Evelyme Vinci, Pierre Perreauh, Louis Bouchard, Denis Bernard, Ranges Wassef, Carole Richard, Gilles Pomier-Layrargues Surgery in cirrhotic patients is associated with a high morbidity and mortality related to portal hypertension and liver insuffciency. It has been show'~ that the amount of intraoperative transfusions and post-operative ascites a~ independent predictors of survival in cirrhotic patients tbllowing surgery- (Garison et al. Ann. 5urg 1984;199:648) Therefore relief of portal hypertension before surgery appears to be a logical approach to prevent these complications and hopefully improve post-operative sup~vaI. We evaluated prospectively the clinicai outcome of I8 patients (mean age: 58 years) with cirrhosis (7 alcoholics and 11 non-alcohokcs) who underwent TIPS placement before antrectomy (5), colectomy (10), small bowel resection (1), pancreatectomy (1) and nephrectomy (l). TIPS was performed 72 +/* 21 days before surgery and induced a marked decrease in portohepatic gradient from 21.4 +/- 3.9 to 8 4 +/- 3.4 mmHg Cirrhotic patients (n = ]7) who underwent elective surgery" without pre-operative TIPS placement were used as a historical control group. Both groups were matched for age, etiology of cirrhosis, indications for surgeD', type of surgery and coagulation parameters. However, the Pugh score was significantly higher in the TIPS group (72 +/- 1.3 vs 6 2 +/- 13). No signifcant difference was observed for operative blood loss, post-operative complications, duration of hospitafsation, and ] month (83% vs 88%) or 1 year cumulative survival rate (54% vs 63%). A multiple logistic regression analysis confirmed the absence of beneficial effects of pre-operattve TIPS on 1 year survival rate after adjustment for pt~-operative Pugh score difference between the two groups. Therefore, the present study shows that pre-operative TIPS placement does not improve' postoperative evolution in cirrhotic patients with mild or moderate liver f~anction impairment.
W1409 Model for End-Stage Liver Disease (MELD) Predicts Survival in Patients with Variceal Bleeding Kiran M. Bambha Mamje Vat~ Ijperen, Michaei Malinchoc, Walter K Kremers, Ray W. Kim, Russell H Wiesner, Pamck S ICamath BACKGROUND: The Chfid-Turcotte-Pugh (CTP) score has traditionally been used to stratfl}"
patients w~th variceal bleeding according to risk for mortality. Draw,backs of CTP include suhjective assessment of ascites and encephalopathy and the lack of standardization of measurement of albumin and prothrombin time. The Model tor End-Stage Liver Disease (MELD) derived from serum creatmine, bilimbin, and pmthrombm time INR has recently been found to be superior to CTP in stratifying patients at risk tor mortality with cirrhosis of the liver. AIM: To determine whether MEId9 can stratify" patients with variceaI Needing according to risk fbr mu~ality METHODS: From a prospectively kept database, we identified 389 patients between 1988-2002 v/no tulfifed Baveno criteria for acute variceal bleeding. Of these, 299 (77%) patients had the requisite data toe MELD score within 24 hours of the vanceal bleed. Since 1993, treatntent has been octreotide and endoscopic variceal Iigation (n = 205) while, prior to 1993, treatroem was vasc~pressin and either sclerotberapy or banding ( n = 9 4 ) Patient surwval was analyzed using tog rank test and Cox proportional hazards model with patients followed for a maximum of one year post Needing. RESULTS: The mo~3ality thr the cotwrt was 97% at 7 days, 25.7% at 42 days, and 39,1% at one year. 'v\~en pauents were grouped according to MELD score < i 0 , 10-20, and >20, the MELD scorn was a sginficant predictor of increased mortality (p<0.OO1)(Table). With ever,/" 10 unit increase in MELD score, the increase in mortality at one year was 66% (95% CI:4492%). ]'his change was independent of the time period (which is a surrogate for type of treatment) CONCLUSIONS: MELD can accurately rank patients at risk of mortality followmg a ~"anceal Need irrespecm'e of treatment and may be used m trials to stratify patients.
Patients(n) % MortalitT"7 ~ % M o ~ 42 days % M~:JI~ 365 days
MELD<10 84 <1.0 9.5 19.9
MELD10-20 117 9,6 21,7 34,5
W1412 Transjugular Intrahepatic Portosystemic Shunt in the Treatment of Ectopic Varices Secondary to Portal Hypertension Ramasamy Saravanan, Manu Nayar, Peter Rowlands, Richard Mc Williams, Howard Smart, Martin Lombard Background Bleeding from ectopic (rectal and stomal) varices is a well recognized complication of portal hypertension It can be massive and life threateningTbere is little data in the literature on managing such problems and the optimal treatment for rectal varices has yet to be established Methods We retrospectively reviewed our institution's experience of patients w~th ectopic vanceal bleeding who underwent transjugular intrahepatic portosystemic shunting (TIPSS) for recurrent bleeding not responding to conservative management. Resuhs Over an eleven year period (1992-2002) we identified eight patients who underwent TIPSS for ectopic variceal haemorrhage: Four patients bled from rectal varices and 4 from stomal varices. TIPSS was successful in seven patmnts: the one failm~ being due to an anatomical abnormality of the portal vein. There were four males and three females with a mean age of 60.5 years (range 43-73 years) The Chflds Pugh grade of the patients was A = 3, B = 2 and C = 2. The follow up period range from 7 days to 46 months. The mean portal pressure gradient before TIPSS was 156 mm Hg (range12-22) and al'ter T1PSS dropped to 5.1 mm Hg(range 3.12).TIPSS success~hlly controlled bleeding in all patients. Rebleeding occurred in three patients two of whom died. The remaining patient had a blocked TIPSS and successfully underwent repeat scenting which re-established patency. Four patients (Childs B = 2, Childs C = 2) died within 60 days of TIPSS due to multiorgan failure. All of these four had a significantly devated bflimbin (mean value of 30 mgs/dl) and/or a raised creatiinne (mean value 3.3mgs/dl) with advanced Childs Pugh grades. All three patmnts with Childs A liver disease were alive at one year Conclusion In our experience TIPSS can be used effectively to treat ectopic vanceal bleeding secondary" to pm'tal hypertension. In this small senes, patients with Chdds grade A liver disease appear to do well with TIPSS. Those with advanced liver disease (Childs B & C) have a unitbnnly poor outcome, pamcuiarIy when assc:ciated with renal impairment, In these patients ectopic vanceal haemorrhage is likely to represent a terminal event
MELD>20 98 22A 52.5 71,5
W1410 Prevalence and Natural History of Gastric Antra[ Vascular Ectasia (GAVE) in Patients Undergoing Orthotopic Liver Transplantation (OLT) Eric M. Ward, Hugo Bonatti, Victor I. Machicao, Rolland C. Dickson, Denise M. Hamois, Barry G Rosser, Raj Satyanarayana, Jeffery L Steers, Massimo Raimondo Background: GAVE is a well-recognized cause of gastrointestinal (GI) hemorrhage associated with portal hypertension. Ahhongh 30% of patients with GAVE have liver disease, the prevalence of GAVE in pauents with cirrhosis is not known. The aims of this study are: a) to describe fl~e prevalence and natural history of GAVE in patients with end-stage liver disease (ESLD) undergoing OLT; b) to compare the prevalence of GAVE in ESLD patients to that in patients with iron deficiency- anemia (IDA) or GI bleeding (G[B). Methods: We rewewed clinical records of patients who underwent OLT at our institution from 2/1998 to 2/20(/1 in addition, the clinical records of consecutive patients referred to our GI endoscopy lab from 11/1999 to 11/2002 for the evaluation of IDA or GIB wet~ also reviewed. Demographic and clit'.ical details were recorded. Results: Three hundred patients underwent OLT Of these, 166 were evaluated pre-operatively with esophagogastroduodenoscopy (EGD) lhree (I.8%) of these patients were found to have GAVE before OLT. Two patients had cry'ptogenic cirrhosis and 1 had hepatitis C In 2 of the patients, EGD performed after OLT demonstrated reso[ution of GAVE. 1Ne third patient did not undergo tbllow-up EGD, but no ciinically apparent G1B occurred during 23 months of follow-up. No patients received endoscopic treatment or portosystemic shunting. By way of comparison, GAVE was found in 28 of 956 patients (3%) undergoing EGD for the evaluation of IDA or GIB. Fourteen (1.5%) of chase had co-existing GAVE and established liver disease at the time of endoscopy. rbe remainder had GAVE but no established liver disease. Conclusions: GAVE was present in 1.8% of patients who underwent EGD bdbre OLT over a 3-year period. This finding is likely" an accurate estimation of the prevalence of GAVE in patients undergomg OLT, as nearly all patients with GAVE would have been referred for endoscopy for the evaluation of anemia or GIB As liver di~ase has been reported to be present in a high percentage of patients with GAVE, it is possfble that patients wiCn ESLD and GAVE generally are too fll or do not survive long enough to undergo OLT. GAVE definitely resolved in two of these patients after OLT, and the third did not experience clinically significant GIB ali.er surgery. These findings are consistent with a previous report docmnenting resolution of GAVE after n I T The prevalence of GAiT in patients without liver disease reid'reed for IDA is similar to the prevalem:e of GAVF in patients undergoing OLT
W1413 Transjugular lntrahepatic Portosystemic Shunting (TIPSS) - Eleven Year Experience in a Regional Referral Centre Manu Nayar, Saravanan Ramaswan~y, Peter Rowlands, Richard iMcWilliams, Howard Smart, Martin Lombard BACKGROUND Since the introduction of TIPSS in 1988 it has been v~idely promoted as a treatment for uncontmlled "v'aricealbleeding and refractory ascites. There are tew long term follow up studies. AIMS AND METHODS The aim of this retrospective analysis was to study the efficacy of TIPSS and to look at five year survival, We assessed the efffct of Childs Pugh grade, previous management, imervals of referral, complications and mortality. We studied the case notes of all patiems who underwent TIPSS between 1992 and 2002 RESULTS 124 patients had 148 procedures per[omaed during the study period, Of these the procedure failed in eight patients because of technical difficulty ie. angulated or blocked portal vein or difficult access. 73 (63%) patients were referred from 13 hospitals in our region. Mean age of the group was 51.5 years (range 18-87 ),ears) and 59% were male. Portal hypertension was caused by alcoholic liver disease m 81% with vh-al hepatitis B or C being an additional etiology in 11 (12%) of these patients. Other causes included crypeogenic cirrhosis (8%), primary bifary cirrhosis (6%) and others. 92% of patients presented with variceal Needing (esophageal, gastric, rectal or stomal) while the rest were patients vdth refractory ascites. The patients with variceal bleeding had a mean of 1,9 bleeds (range 1-4) prior to TIPSS, The mean time from the decision to undertake TIPSS to performing the procedure was 32 days (range 1-56). The Childs Pugh score of the patients was A = 13%, B = 24% and C = 63%. The mean pre TIPSS pressure gradient was 2 1 5 mmHg (range 10-48 mmHg) and the mean post TIPSS pressure gradient was 7,7 mmHg (range 0-16 mmHg).38 patients (33%) had 46 episodes of rebleeding at anman interval of 134 days after the procedure (range 1-780
A-665
AGA Abstracts
obtained by analysis o[ twemy-ibur-honr ambulatory electrocardiographic recordings using both the time and the treqnency domain methods, Results: The analysis of heart rate variabdily showed significantly lower values ot the following parameters in cirrhotic patients: InHF (patmnts: 5 2 • 03, controls: 5,8 • 06; p = 0.03); rMSSD (patients: 32 • 10, controls:47 • 10 ms; p = O 0 1 ) and pNN50% ( patients: 9 • controls: 19• p=O.02), indicating a decreased parasympathetic activaty in this group of patients Conclusions: Our results suggest impaired efferent vagal function in patients with Child's grade B nonalcoholic cirrhosis. This abnormality may contribute to the hemodynamic disturbances of cirrhosis.
Transjugular Intrahepatic Portosystemic Shunt (TIPS) Before Surgery in Cirrhotic Patients: A Retrospective Comparative Study Evelyme Vinci, Pierre Perreauh, Louis Bouchard, Denis Bernard, Ranges Wassef, Carole Richard, Gilles Pomier-Layrargues Surgery in cirrhotic patients is associated with a high morbidity and mortality related to portal hypertension and liver insuffciency. It has been show'~ that the amount of intraoperative transfusions and post-operative ascites a~ independent predictors of survival in cirrhotic patients tbllowing surgery- (Garison et al. Ann. 5urg 1984;199:648) Therefore relief of portal hypertension before surgery appears to be a logical approach to prevent these complications and hopefully improve post-operative sup~vaI. We evaluated prospectively the clinicai outcome of I8 patients (mean age: 58 years) with cirrhosis (7 alcoholics and 11 non-alcohokcs) who underwent TIPS placement before antrectomy (5), colectomy (10), small bowel resection (1), pancreatectomy (1) and nephrectomy (l). TIPS was performed 72 +/* 21 days before surgery and induced a marked decrease in portohepatic gradient from 21.4 +/- 3.9 to 8 4 +/- 3.4 mmHg Cirrhotic patients (n = ]7) who underwent elective surgery" without pre-operative TIPS placement were used as a historical control group. Both groups were matched for age, etiology of cirrhosis, indications for surgeD', type of surgery and coagulation parameters. However, the Pugh score was significantly higher in the TIPS group (72 +/- 1.3 vs 6 2 +/- 13). No signifcant difference was observed for operative blood loss, post-operative complications, duration of hospitafsation, and ] month (83% vs 88%) or 1 year cumulative survival rate (54% vs 63%). A multiple logistic regression analysis confirmed the absence of beneficial effects of pre-operattve TIPS on 1 year survival rate after adjustment for pt~-operative Pugh score difference between the two groups. Therefore, the present study shows that pre-operative TIPS placement does not improve' postoperative evolution in cirrhotic patients with mild or moderate liver f~anction impairment.
W1409 Model for End-Stage Liver Disease (MELD) Predicts Survival in Patients with Variceal Bleeding Kiran M. Bambha Mamje Vat~ Ijperen, Michaei Malinchoc, Walter K Kremers, Ray W. Kim, Russell H Wiesner, Pamck S ICamath BACKGROUND: The Chfid-Turcotte-Pugh (CTP) score has traditionally been used to stratfl}"
patients w~th variceal bleeding according to risk for mortality. Draw,backs of CTP include suhjective assessment of ascites and encephalopathy and the lack of standardization of measurement of albumin and prothrombin time. The Model tor End-Stage Liver Disease (MELD) derived from serum creatmine, bilimbin, and pmthrombm time INR has recently been found to be superior to CTP in stratifying patients at risk tor mortality with cirrhosis of the liver. AIM: To determine whether MEId9 can stratify" patients with variceaI Needing according to risk fbr mu~ality METHODS: From a prospectively kept database, we identified 389 patients between 1988-2002 v/no tulfifed Baveno criteria for acute variceal bleeding. Of these, 299 (77%) patients had the requisite data toe MELD score within 24 hours of the vanceal bleed. Since 1993, treatntent has been octreotide and endoscopic variceal Iigation (n = 205) while, prior to 1993, treatroem was vasc~pressin and either sclerotberapy or banding ( n = 9 4 ) Patient surwval was analyzed using tog rank test and Cox proportional hazards model with patients followed for a maximum of one year post Needing. RESULTS: The mo~3ality thr the cotwrt was 97% at 7 days, 25.7% at 42 days, and 39,1% at one year. 'v\~en pauents were grouped according to MELD score < i 0 , 10-20, and >20, the MELD scorn was a sginficant predictor of increased mortality (p<0.OO1)(Table). With ever,/" 10 unit increase in MELD score, the increase in mortality at one year was 66% (95% CI:4492%). ]'his change was independent of the time period (which is a surrogate for type of treatment) CONCLUSIONS: MELD can accurately rank patients at risk of mortality followmg a ~"anceal Need irrespecm'e of treatment and may be used m trials to stratify patients.
Patients(n) % MortalitT"7 ~ % M o ~ 42 days % M~:JI~ 365 days
MELD<10 84 <1.0 9.5 19.9
MELD10-20 117 9,6 21,7 34,5
W1412 Transjugular Intrahepatic Portosystemic Shunt in the Treatment of Ectopic Varices Secondary to Portal Hypertension Ramasamy Saravanan, Manu Nayar, Peter Rowlands, Richard Mc Williams, Howard Smart, Martin Lombard Background Bleeding from ectopic (rectal and stomal) varices is a well recognized complication of portal hypertension It can be massive and life threateningTbere is little data in the literature on managing such problems and the optimal treatment for rectal varices has yet to be established Methods We retrospectively reviewed our institution's experience of patients w~th ectopic vanceal bleeding who underwent transjugular intrahepatic portosystemic shunting (TIPSS) for recurrent bleeding not responding to conservative management. Resuhs Over an eleven year period (1992-2002) we identified eight patients who underwent TIPSS for ectopic variceal haemorrhage: Four patients bled from rectal varices and 4 from stomal varices. TIPSS was successful in seven patmnts: the one failm~ being due to an anatomical abnormality of the portal vein. There were four males and three females with a mean age of 60.5 years (range 43-73 years) The Chflds Pugh grade of the patients was A = 3, B = 2 and C = 2. The follow up period range from 7 days to 46 months. The mean portal pressure gradient before TIPSS was 156 mm Hg (range12-22) and al'ter T1PSS dropped to 5.1 mm Hg(range 3.12).TIPSS success~hlly controlled bleeding in all patients. Rebleeding occurred in three patients two of whom died. The remaining patient had a blocked TIPSS and successfully underwent repeat scenting which re-established patency. Four patients (Childs B = 2, Childs C = 2) died within 60 days of TIPSS due to multiorgan failure. All of these four had a significantly devated bflimbin (mean value of 30 mgs/dl) and/or a raised creatiinne (mean value 3.3mgs/dl) with advanced Childs Pugh grades. All three patmnts with Childs A liver disease were alive at one year Conclusion In our experience TIPSS can be used effectively to treat ectopic vanceal bleeding secondary" to pm'tal hypertension. In this small senes, patients with Chdds grade A liver disease appear to do well with TIPSS. Those with advanced liver disease (Childs B & C) have a unitbnnly poor outcome, pamcuiarIy when assc:ciated with renal impairment, In these patients ectopic vanceal haemorrhage is likely to represent a terminal event
MELD>20 98 22A 52.5 71,5
W1410 Prevalence and Natural History of Gastric Antra[ Vascular Ectasia (GAVE) in Patients Undergoing Orthotopic Liver Transplantation (OLT) Eric M. Ward, Hugo Bonatti, Victor I. Machicao, Rolland C. Dickson, Denise M. Hamois, Barry G Rosser, Raj Satyanarayana, Jeffery L Steers, Massimo Raimondo Background: GAVE is a well-recognized cause of gastrointestinal (GI) hemorrhage associated with portal hypertension. Ahhongh 30% of patients with GAVE have liver disease, the prevalence of GAVE in pauents with cirrhosis is not known. The aims of this study are: a) to describe fl~e prevalence and natural history of GAVE in patients with end-stage liver disease (ESLD) undergoing OLT; b) to compare the prevalence of GAVE in ESLD patients to that in patients with iron deficiency- anemia (IDA) or GI bleeding (G[B). Methods: We rewewed clinical records of patients who underwent OLT at our institution from 2/1998 to 2/20(/1 in addition, the clinical records of consecutive patients referred to our GI endoscopy lab from 11/1999 to 11/2002 for the evaluation of IDA or GIB wet~ also reviewed. Demographic and clit'.ical details were recorded. Results: Three hundred patients underwent OLT Of these, 166 were evaluated pre-operatively with esophagogastroduodenoscopy (EGD) lhree (I.8%) of these patients were found to have GAVE before OLT. Two patients had cry'ptogenic cirrhosis and 1 had hepatitis C In 2 of the patients, EGD performed after OLT demonstrated reso[ution of GAVE. 1Ne third patient did not undergo tbllow-up EGD, but no ciinically apparent G1B occurred during 23 months of follow-up. No patients received endoscopic treatment or portosystemic shunting. By way of comparison, GAVE was found in 28 of 956 patients (3%) undergoing EGD for the evaluation of IDA or GIB. Fourteen (1.5%) of chase had co-existing GAVE and established liver disease at the time of endoscopy. rbe remainder had GAVE but no established liver disease. Conclusions: GAVE was present in 1.8% of patients who underwent EGD bdbre OLT over a 3-year period. This finding is likely" an accurate estimation of the prevalence of GAVE in patients undergomg OLT, as nearly all patients with GAVE would have been referred for endoscopy for the evaluation of anemia or GIB As liver di~ase has been reported to be present in a high percentage of patients with GAVE, it is possfble that patients wiCn ESLD and GAVE generally are too fll or do not survive long enough to undergo OLT. GAVE definitely resolved in two of these patients after OLT, and the third did not experience clinically significant GIB ali.er surgery. These findings are consistent with a previous report docmnenting resolution of GAVE after n I T The prevalence of GAiT in patients without liver disease reid'reed for IDA is similar to the prevalem:e of GAVF in patients undergoing OLT
W1413 Transjugular lntrahepatic Portosystemic Shunting (TIPSS) - Eleven Year Experience in a Regional Referral Centre Manu Nayar, Saravanan Ramaswan~y, Peter Rowlands, Richard iMcWilliams, Howard Smart, Martin Lombard BACKGROUND Since the introduction of TIPSS in 1988 it has been v~idely promoted as a treatment for uncontmlled "v'aricealbleeding and refractory ascites. There are tew long term follow up studies. AIMS AND METHODS The aim of this retrospective analysis was to study the efficacy of TIPSS and to look at five year survival, We assessed the efffct of Childs Pugh grade, previous management, imervals of referral, complications and mortality. We studied the case notes of all patiems who underwent TIPSS between 1992 and 2002 RESULTS 124 patients had 148 procedures per[omaed during the study period, Of these the procedure failed in eight patients because of technical difficulty ie. angulated or blocked portal vein or difficult access. 73 (63%) patients were referred from 13 hospitals in our region. Mean age of the group was 51.5 years (range 18-87 ),ears) and 59% were male. Portal hypertension was caused by alcoholic liver disease m 81% with vh-al hepatitis B or C being an additional etiology in 11 (12%) of these patients. Other causes included crypeogenic cirrhosis (8%), primary bifary cirrhosis (6%) and others. 92% of patients presented with variceal Needing (esophageal, gastric, rectal or stomal) while the rest were patients vdth refractory ascites. The patients with variceal bleeding had a mean of 1,9 bleeds (range 1-4) prior to TIPSS, The mean time from the decision to undertake TIPSS to performing the procedure was 32 days (range 1-56). The Childs Pugh score of the patients was A = 13%, B = 24% and C = 63%. The mean pre TIPSS pressure gradient was 2 1 5 mmHg (range 10-48 mmHg) and the mean post TIPSS pressure gradient was 7,7 mmHg (range 0-16 mmHg).38 patients (33%) had 46 episodes of rebleeding at anman interval of 134 days after the procedure (range 1-780
A-665
AGA Abstracts