HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2311030967v1 231/1/231    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Ferral, H. Articles by McMahan, C. A. Search for Related Content PubMed PubMed Citation Articles by Ferral, H. Articles by McMahan, C. A.

Survival after Elective Transjugular Intrahepatic Portosystemic Shunt Creation: Prediction with Model for End-Stage Liver Disease Score¹

¹ From the Dept of Radiology, Div of Cardiovascular and Special Interventions (H.F., D.W.P., C.R.Y.), Dept of Medicine, Div of Gastroenterology and Nutrition (K.V.S.), and Dept of Pathology (C.A.M.), University of Texas Health Science Center at San Antonio; Dept of Radiology, Div of Interventional Radiology, Ohio State University, Columbus (P.G.); and Dept of Radiology, Vascular and Interventional Radiology, East Carolina University, Greenville, NC (V.S.A.). Received June 23, 2003; revision requested August 29; revision received September 22; accepted October 21. Address correspondence to H.F., Affiliated Radiologists, SC, 1725 W Harrison St, Suite 456, Chicago, IL 60612 (e-mail: hferral@rushradiology.org).

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: To evaluate the ability of a model of end-stage liver disease (MELD) score to predict survival in a diverse group of patients who underwent elective transjugular intrahepatic portosystemic shunt (TIPS) creation in two tertiary care institutions.

MATERIALS AND METHODS: Patients who underwent elective TIPS creation in two institutions between May 1, 1999, and June 1, 2002, were selected. Patients who underwent emergency TIPS creation were excluded. One hundred sixty-six patients met the inclusion criteria. The MELD score was computed and compared with the survival rate. Survival curves were estimated with Kaplan-Meier product limit estimates and were compared with the log-rank test. Accuracy of the model was evaluated with the c statistic.

RESULTS: The survival rate for all patients was 88.4% at 30 days, 78.1% at 3 months, and 71.8% at 6 months. Significantly lower survival rates were found in patients with MELD scores of 18 or more in comparison to those with MELD scores of 17 or less (P = .001). The c statistic for prediction of 3-month survival on the basis of the MELD score was 0.76. The early (30-day) death rate for this series was 11.4%. There was a significant difference in the 30-day mortality rate between patients with MELD scores of 17 or less and those with scores of 18 or more (P = .001). Patients who underwent TIPS creation for the management of refractory ascites had a significantly lower survival rate in comparison to that for the management of variceal bleeding (P = .001).

CONCLUSION: Results confirm that after elective TIPS creation, patients with a MELD score of 18 or more have a significantly lower 3-month survival rate than do those with a MELD score of 17 or less.

© RSNA, 2004

Index terms: Hypertension, portal, 957.723 • Liver, interventional procedures, 957.1268 • Shunts, portosystemic, 957.1268

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Factors that affect patient survival after transjugular intrahepatic portosystemic shunt (TIPS) creation have been extensively studied (1–6). Elevated serum bilirubin and creatinine levels before TIPS creation have been proposed as important predictors of poor patient prognosis (3,5,7). A model of end-stage liver disease (MELD) score has been recently described (8). The MELD score is calculated on the basis of three objective indicators: serum bilirubin level, serum creatinine level, and international normalized ratio. The MELD score used in the current study is a modified version of a model described by Malinchoc et al (9) that was developed to predict the 3-month survival rate of patients with cirrhosis who underwent elective TIPS creation. If the calculated risk of death before TIPS creation was too high (80%), the patients were advised not to undergo the procedure (9). The model of Malinchoc et al was evaluated previously (10,11), and the investigators confirmed that patients with a risk of 1.8 or more (MELD score of 18 or more) had a significantly lower 3-month survival rate than did those with a risk score lower than 1.8 (MELD score of less than 18). The investigators emphasized the importance of evaluation of larger and more diverse patient populations to further validate the applicability of this prognostic model (10,11). The purpose of the present study was to evaluate the ability of the MELD score to predict survival in a diverse group of patients who underwent elective TIPS creation in two tertiary care institutions.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Patients
This retrospective study was approved by the review board of each institution. Informed consent was not required for the study, although written informed consent was obtained for the TIPS creation procedures. The medical records of all patients who underwent TIPS creation in two institutions (referred to as institution 1 and institution 2) between May 1, 1999, and June 1, 2002, were reviewed. The data from institution 1 included procedures performed in two hospitals, while the data from institution 2 were from one hospital. The data collected from the medical records included the following: patient age, sex, ethnicity, and mental status; cause of liver cirrhosis; degree of ascites; liver function test results; and portosystemic gradients before and after TIPS creation. The Child-Pugh scores and MELD scores were calculated on the basis of data obtained within 7 days before TIPS creation. If multiple sets of laboratory test results existed, the latest set obtained immediately before the TIPS creation was used. The date of the last clinic visit or the date of liver transplantation defined the follow-up period for all surviving patients. If there were no clinic visits recorded in the medical records, the patients or their families were contacted by telephone by one of the investigators (H.F., P.G., C.R.Y.), and a brief interview was conducted to determine the patient’s clinical course after TIPS creation. The mean follow-up time for the patients that were alive at last contact was 11 months 3 days (range, 4 days to 42 months 24 days).

At our institutions, patients are referred to the interventional radiology service for elective TIPS creation if they have recurrent variceal bleeding after at least three sessions of sclerotherapy or banding, if they have persistent tense ascites despite adequate medical management with diuretics, if they require repeat paracentesis at least once every other week to control tense ascites, or if they have an elevated serum creatinine or electrolyte imbalance secondary to the use of diuretics to control tense ascites. Emergency TIPS creation is defined as the procedure performed in a patient who is actively bleeding or who is hemodynamically unstable after emergency sclerotherapy or banding. Two hundred consecutive patients underwent TIPS creation in this period. One hundred sixty-six patients underwent elective TIPS creation, and their records were selected for analysis. Thirty-four patients were excluded on the basis of emergency (33 patients) or unsuccessful (one patient) TIPS creation. Characteristics of the patients included in the current study are described on the basis of institution and on an overall basis in Table 1. The 89 patients from institution 1 comprised 72% (n = 64) Latinos and 28% (n = 25) non-Latinos, whereas the 77 patients from institution 2 comprised 88% (n = 68) non-Latinos and 12% (n = 9) African Americans.

Mortality and early mortality (defined as death that occurred within 30 days after TIPS creation) were calculated and compared on the basis of these same MELD score categories. Additional data analysis included comparison of patient mortality on the basis of the combined MELD score and presence or absence of ascites and on the basis of the combined MELD score and Child-Pugh score. The c statistic was calculated and compared for the MELD score alone, ascites alone, and a combination of both, as well as for the Child-Pugh score alone, the MELD score alone, and a combination of both.

MELD Score Calculation
The MELD score was calculated with the United Network for Organ Sharing modification of the original formula: MELD score = 9.6 x loge(creatinine mg/dL) + 3.8 x loge (bilirubin mg/dL) + 11.2 x loge (INR) + 6.4, where INR is international normalized ratio. The MELD score can be easily calculated on the Web site that we used for our calculations: depts.washington.edu/uwhep/calculations/meldscore.htm (accessed starting June 1, 2002).

Statistical Analysis
The differences between groups were evaluated by means of an independent-sample t test, the ² test, or the Fisher exact test. The c statistic, which is equivalent to the area under the receiver operating characteristic curve, was used as a global measurement of discrimination. A c statistic between 0.8 and 0.9 indicates excellent diagnostic accuracy, and a c statistic between 0.7 and 0.8 is considered acceptable (8). Survival curves were computed by means of Kaplan-Meier product-limit estimates and were compared by means of the log-rank test (SAS; SAS Institute, Cary, NC). A P value of less than .05 was considered to indicate a statistically significant difference.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Patient Characteristics and Technical Outcomes
The characteristics of the patients on the basis of institution and on an overall basis are given in Table 1. One hundred (60.2%) of the 166 patients underwent TIPS creation to treat ascites, and 66 (39.8%) underwent TIPS creation to treat recurrent variceal bleeding. Seven patients who underwent TIPS creation to treat cirrhotic hydrothorax were included in the group of patients with ascites. Other than the differences in patient ethnicity, few significant differences were found between the patients in the two institutions (Table 1). Figure 1 shows the survival curves for the study group on the basis of institution. No significant differences in patient survival were found between institutions 1 and 2 (P = .293). Subsequent results are based on the combined data from both institutions. The survival rate for all patients was 88.4% at 30 days, 78.1% at 3 months, and 71.8% at 6 months.

View larger version (26K): [in this window] [in a new window] [Download PPT slide]   Figure 1. Kaplan-Meier survival curves based on two institutions show no differences.

View larger version (31K): [in this window] [in a new window] [Download PPT slide]   Figure 2. Kaplan-Meier survival curves based on MELD scores show a significant decrease in patient survival with increasing MELD score.

Predictors of Survival in Addition to MELD Score
Patient survival was analyzed on the basis of cause of cirrhosis; there was no significant difference in survival rates between groups A and B (P = .736). There were also no differences in survival rates between groups A and B for patients with MELD scores of 17 or less (P = .978) and for those with scores of 18 or more (P = .201). The mortality rates at 30 days and 3 and 6 months are shown in Table 3.

View larger version (25K): [in this window] [in a new window] [Download PPT slide]   Figure 3. Kaplan-Meier survival curves based on indication for TIPS creation show significantly higher mortality in patients who underwent TIPS creation to treat refractory ascites.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

In the current study, significant differences were found between institutions according to patient sex, race, ethnicity, origin of cirrhosis, albumin levels, and portosystemic gradients before TIPS creation. Despite these differences, survival rates were similar between the two institutions. These findings suggest that race, ethnicity, procedure technique, and patient treatment strategies in different institutions have little effect on patient survival after elective TIPS creation. Survival rates were similar when the patients were separated by cause of cirrhosis. This finding supports the deletion of the cause of cirrhosis in the modification of the MELD score (8).

Another interesting finding in the current study was the significantly lower survival rate in patients who underwent elective TIPS creation to treat refractory ascites. These findings correlate with those of other authors (10,12). In the group of patients with MELD scores of 17 or less, the survival rates for patients with refractory ascites were lower than those for patients who underwent TIPS creation to treat variceal bleeding. The negative effect of the presence of ascites was lost in the subgroup of patients with MELD scores of 18 or more, probably because they are sicker overall and the presence or absence of ascites is less important. Development of ascites is the most common complication in patients with cirrhosis and is considered to be a poor prognostic sign (13,14). The medical management of refractory ascites is problematic, and the two best current options are large-volume paracentesis with albumin infusion and TIPS creation (13,14).

Findings in two recent studies demonstrate that TIPS creation is superior to large-volume paracentesis and albumin infusion in the management of refractory ascites (14,15). Despite a lower incidence of recurrent ascites in patients who undergo TIPS creation, however, the survival rate and quality of life are not improved after TIPS creation (14,15). These two studies were prospective randomized trials with strict patient inclusion criteria to provide a highly selected group of patients with potential for a better outcome. The patients in the current study represent a nonselected population treated in two tertiary care centers. Although ascites in some of our patients did not meet the strict definition of refractory ascites, all collections were massive and difficult to control. Despite these differences in patient population, results in the current study still support the concept that TIPS creation should probably not be performed to treat refractory ascites in patients with advanced liver failure. We agree that only patients with relatively better liver function should be selected as candidates to undergo TIPS creation to treat refractory ascites (14,16). Despite the fact that the presence of ascites had a negative effect on survival in patients with MELD scores of 17 or less, the addition of ascites to the MELD score did not greatly improve the prognostic ability of the model, because the c statistic only increased to 0.79 for 3-month mortality and to 0.80 for 6-month mortality. These increases in the c statistic were not significant.

Decreased survival was also found in patients with Child-Pugh scores of 10 or more. Addition of the Child-Pugh score to the MELD score increased the discrimination for 3- and 6-month mortality. We think that the MELD score is still preferable to the Child-Pugh score because of its improved ability to delineate deranging liver failure (17); however, it would be interesting to evaluate discrimination with the combination of these two scores in a larger series of patients.

Results in the current study showed relatively lower c statistics for the MELD score than were shown by other authors (8,10). Salerno et al (10) reported a c statistic of 0.84 for 3-month survival, while Kamath et al (8) reported a c statistic of 0.87 in hospitalized patients with cirrhosis who did not undergo TIPS creation and 0.80 in ambulatory patients with noncholestatic cirrhosis. It is not entirely clear why the c statistics in our study were lower. We note some potentially important differences in the patient population. Patients in the study of Salerno et al were mainly treated for variceal bleeding, whereas our patient population was mainly treated for ascites that was difficult to control. In addition, 82% of the former patients had Child-Pugh class A and B disease, whereas 62% of the patients in our study had such disease. On the other hand, mean MELD scores for the two patient groups described by Kamath et al were 9 and 10, whereas the mean MELD score for our patients was 15. Comparatively, our patients were expected to have a worse prognosis. On the other hand, Angermayr et al (12) recently evaluated a large number of patients who underwent elective TIPS creation. The c statistic for 3-month survival in their series was 0.72, which is similar to that in our study (0.76). It is possible that the predictive power of the MELD score is better for patients with relatively better liver function.

The early death rate in the current study is comparable to the death rates reported in previous series (1,2,7,11,18). We found a significantly higher early mortality rate in patients with a MELD score of 18 or more compared with that in patients with a MELD score of 17 or less. Patients with a MELD score of 25 or more had a 42% early mortality rate, which was much higher that that in any other subgroup, but we had expected the rate to be even higher. End-stage liver failure is definitely associated with a poorer prognosis, but there are other factors that participate in early mortality rate after elective TIPS creation that are still elusive. We think that the changes in portal and hepatic arterial flows induced by the nonselective portosystemic shunting process after TIPS creation play an important role in the precipitation of early mortality. Patients at risk for hemodynamic imbalance after TIPS creation should be identified before the procedure, and they should be further evaluated (18–20).

We think that the application of mathematic models to predict patient survival after invasive procedures is subject to several limitations. We understand that the final outcome for a given patient is determined by a multitude of factors, and it is difficult to apply rigid guidelines based on these models to decide patient treatment strategies. We do not intend to deny any patient the chance to undergo a potentially life-saving procedure by strictly applying the results of a mathematic model. Our goal is to apply the model in an attempt to identify those patients in whom the invasive procedure may be potentially harmful. With this information, we should be able to discuss risks and benefits of the procedure with referring physicians, patients, and family members with a more objective approach.

Results in the current study confirm the usefulness of the MELD score as a disease-severity index to predict survival in patients who undergo elective TIPS creation. Our results strongly suggest that patients with MELD scores of 25 or more should not undergo elective TIPS creation, and alternative treatment strategies should be considered. Patients with refractory ascites and a MELD score of 25 or more probably should not undergo TIPS creation but should be treated with repeat paracentesis as the treatment of choice. Discussions with the referring physicians, patient, and family members are important in the decision about treatment strategies for patients with refractory ascites. The most important application of these results is in decisions about the use of available resources. We think that this important information can help patients make decisions about treatment and can enable physicians to explain the therapeutic options to patients, their families, and referring colleagues, with an objective perspective about the expected outcomes.

	FOOTNOTES

 
Abbreviations: MELD = model of end-stage liver disease, TIPS = transjugular intrahepatic portosystemic shunt

Author contributions: Guarantors of integrity of entire study, H.F., P.G., D.W.P., C.A.M.; study concepts, H.F., C.A.M.; literature research, H.F., C.A.M.; clinical studies, K.V.S., H.F., P.G., V.S.A., D.W.P.; data acquisition, C.R.Y., P.G., V.S.A., K.V.S., H.F., D.W.P.; data analysis/interpretation, P.G., V.S.A., C.A.M., H.F.; statistical analysis, H.F., C.A.M.; manuscript preparation and editing, H.F., C.A.M.; manuscript definition of intellectual content, revision/review, and final version approval, all authors

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Rossle M, Siegerstetter V, Huber M, Ochs A. The first decade of the transjugular intrahepatic portosystemic shunt (TIPS): state of the art. Liver 1998; 18:73-89.[Medline]
2. Encarnacion CE, Palmaz JC, Rivera FJ, et al. Transjugular intrahepatic portosystemic shunt placement for variceal bleeding: predictors of mortality. J Vasc Interv Radiol 1995; 6:687-694.[Medline]
3. Rajan DK, Haskal ZJ, Clark TWI. Serum bilirubin and early mortality after transjugular intrahepatic portosystemic shunts: results of a multivariate analysis. J Vasc Interv Radiol 2002; 13:155-161.[Medline]
4. Rubin RA, Haskal ZJ, O’Brien C, Cope C, Brass C. Transjugular intrahepatic portosystemic shunting: decreased survival for patients with high APACHE II scores. Am J Gastroenterol 1995; 90:556-563.[Medline]
5. Chalasani N, Clark WS, Martin LG, et al. Determinants of mortality in patients with advanced cirrhosis after transjugular intrahepatic portosystemic shunting. Gastroenterology 2000; 118:138-144.[CrossRef][Medline]
6. Patch D, Nikolopoulou V, McCormick A, et al. Factors related to early mortality after transjugular intrahepatic portosystemic shunt for failed endoscopic therapy in acute variceal bleeding. J Hepatol 1998; 28:454-460.[CrossRef][Medline]
7. Coldwell DM, Ring EJ, Rees CR, et al. Multicenter investigation of the role of transjugular intrahepatic portosystemic shunt in management of portal hypertension. Radiology 1995; 196:335-340.[Abstract/Free Full Text]
8. Kamath PS, Wiesner RH, Malinchoc M, et al. A model to predict survival in patients with end-stage liver disease. Hepatology 2001; 33:464-470.[CrossRef][Medline]
9. Malinchoc M, Kamath PS, Gordon FD, Peine CJ, Rank J, Ter Borg PC. A model to predict poor survival in patients undergoing transjugular intrahepatic portosystemic shunts. Hepatology 2000; 31:864-871.[CrossRef][Medline]
10. Salerno F, Merli M, Cazzaniga M, et al. MELD score is better than Child-Pugh score in predicting 3-month survival of patients undergoing transjugular intrahepatic portosystemic shunt. J Hepatol 2002; 36:494-500.[CrossRef][Medline]
11. Ferral H, Vasan R, Speeg KV, et al. Evaluation of a model to predict poor survival in patients undergoing elective TIPS procedures. J Vasc Interv Radiol 2002; 13:1103-1108.[Medline]
12. Angermayr B, Cejna M, Karnel F, et al. Child-Pugh versus MELD score in predicting survival in patients undergoing transjugular intrahepatic portosystemic shunt. Gut 2003; 52:879-885.[Abstract/Free Full Text]
13. Runyon BA. Care of patients with ascites. N Engl J Med 1994; 330:337-342.[Free Full Text]
14. Sanyal AJ, Genning C, Reddy KR, et al. The North American Study for the Treatment of Refractory Ascites. Gastroenterology 2003; 124:634-641.[CrossRef][Medline]
15. Gines P, Uriz J, Calahorra B, et al. Transjugular intrahepatic portosystemic shunting versus paracentesis plus albumin for refractory ascites in cirrhosis. Gastroenterology 2002; 123:1839-1847.[CrossRef][Medline]
16. Ferral H, Bjarnason H, Wegryn SA, et al. Refractory ascites: early experience in treatment with transjugular intrahepatic portosystemic shunt. Radiology 1993; 189:795-801.[Abstract/Free Full Text]
17. Forman LM, Lucey MR. Predicting the prognosis of chronic liver disease: an evolution from Child to MELD. Hepatology 2001; 33:473-475.[CrossRef][Medline]
18. Walser EM, DeLa Pena R, Villanueva-Meyer J, Ozkan O, Soloway R. Hepatic perfusion before and after the transjugular intrahepatic portosystemic shunt procedure: impact on survival. J Vasc Interv Radiol 2000; 11:913-918.[Medline]
19. Haskal ZJ, Martin L, Cardella JF, et al. Quality improvement guidelines for transjugular intrahepatic portosystemic shunts. J Vasc Interv Radiol 2001; 12:131-136.[Medline]
20. Zipprich A, Steudel N, Behrmann C, et al. Functional significance of hepatic arterial flow reserve in patients with cirrhosis. Hepatology 2003; 37:385-392.[CrossRef][Medline]

This article has been cited by other articles:

				C. J. Yoon, J. W. Chung, and J. H. Park Transjugular Intrahepatic Portosystemic Shunt for Acute Variceal Bleeding in Patients with Viral Liver Cirrhosis: Predictors of Early Mortality Am. J. Roentgenol., October 1, 2005; 185(4): 885 - 889. [Abstract] [Full Text] [PDF]

				C. Vignali, I. Bargellini, M. Grosso, G. Passalacqua, F. Maglione, F. Pedrazzini, P. Filauri, R. Niola, R. Cioni, and P. Petruzzi TIPS with Expanded Polytetrafluoroethylene-Covered Stent: Results of an Italian Multicenter Study Am. J. Roentgenol., August 1, 2005; 185(2): 472 - 480. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2311030967v1 231/1/231    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Ferral, H. Articles by McMahan, C. A. Search for Related Content PubMed PubMed Citation Articles by Ferral, H. Articles by McMahan, C. A.