Radiology. Published online /radiol Radiology 2005; 237: Abbreviation: EV esophageal varices

Transcription

1 Evgeny Farber, MD Doron Fischer, MD Rami Eliakim, MD Nira Beck-Razi, MD Ahuva Engel, MD Ella Veitsman, MD Irit Chermesh, MD Kamel Yassin, MD Diana Gaitini, MD Michael Libes, MD Shai Linn, MD, PhD Soboh Soboh, MD Yaacov Baruch, MD Published online /radiol Radiology 2005; 237: Abbreviation: EV esophageal varices 1 From the Liver Unit (E.F., E.V., Y.B.), Department of Radiology (D.F., N.B., A.E., D.G.), Department of Gastroenterology (R.E., I.C., K.Y., M.L.), and Epidemiology Unit (S.L.), Rambam Medical Center, Efron St, Haifa 31096, Israel; Department of Internal Medicine A, Poriya Hospital, and Bruce Rappaport Faculty of Medicine, Technion- Israel Institute of Technology, Haifa, Israel (S.S.); and School of Public Health, Haifa University, Haifa, Israel (S.L.). Received September 22, 2004; revision requested November 26; revision received December 12; accepted January 18, Address correspondence to Y.B. ( Authors stated no financial relationship to disclose. Author contributions: Guarantors of integrity of entire study, E.F., D.F., R.E., N.B., A.E., E.V., I.C., D.G., Y.B.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, E.F., D.F., R.E., N.B., E.V., D.G., M.L., S.L., Y.B.; clinical studies, all authors; statistical analysis, E.F., D.F., R.E., S.L., Y.B.; and manuscript editing, E.F., D.F., R.E., E.V., D.G., S.L., Y.B. RSNA, 2005 Esophageal Varices: Evaluation with Esophagography with Barium versus Endoscopic Gastroduodenoscopy in Patients with Compensated Cirrhosis Blinded Prospective Study 1 PURPOSE: To prospectively evaluate the accuracy of esophagography with barium in diagnosis of esophageal varices (EV) in patients with compensated cirrhosis, with endoscopic gastroduodenoscopy as the reference standard. MATERIALS AND METHODS: In this study, which was approved by the local Helsinki Committee and in which all patients consented to participate, 61 patients with cirrhosis (34 men, 27 women; mean age, 61 years; range, years) received a diagnosis clinically or with liver biopsy. In 87% (n 53) of patients, Child-Pugh classification was A; in 13% (n 8), Child-Pugh classification was B. They were evaluated with endoscopic gastroduodenoscopy, according to Japanese general criteria. Esophagography was performed within 3 weeks of endoscopic gastroduodenoscopy, and EV were assigned grades as follows: 0, no EV were seen; 1, EV manifested as very mild irregularities of the folds; and 2, the irregularity of the folds (EV) was clearly present. They were also assigned grades for shape and size: grade F0, no EV detected; grade F1, small straight EV; grade F2, slightly enlarged tortuous EV occupying less than one-third of esophageal lumen; and grade F3, large coil-shaped EV that occupied more than one-third of esophageal lumen. The sensitivity and specificity and positive and negative predictive values of esophagography for identification of each grade of EV were calculated separately, as was the 95% confidence interval. RESULTS: All large EV (grades F2 and F3) were diagnosed at esophagography. Sensitivity declined with small EV (grade F1) to 71. The overall sensitivity of esophagography was 89% (95% confidence interval: 75.9%, 96.5%), the overall specificity was 83% (95% confidence interval: 64.5%, 94.7%), the overall positive predictive value was 89%, and the overall negative predictive value was 83% (95% confidence interval: 64.5%, 94.7%). Overall accuracy was 87%. CONCLUSION: Esophagography is highly accurate in diagnosis of EV and can be considered a viable noninvasive alternative for determination of patients who should be selected for prophylactic treatment. RSNA, 2005 Current guidelines in both Europe and the United States recommend that patients with cirrhosis undergo endoscopic gastroduodenoscopy every 1 2 years for the presence of esophageal varices (EV) (1,2). It is further recommended that patients with large EV be 535

2 TABLE 1 Demographic Characteristics of Patients Characteristic Patients* Total no. of patients 61 Sex M 34 (56) F 27 (44) Mean age (y) 61 (36 76) Cause of liver disease Hepatitis C virus 29 (48) Hepatitis B virus 9 (15) Alcoholic liver disease 9 (15) Nonalcohol steatohepatitis 9 (15) Autoimmune hepatitis 3 (5) Primary biliary cirrhosis 2 (3) * Numbers in parentheses are percentages except where specified otherwise. Numbers in parentheses are the range. Percentages do not add up to 100% because of rounding. TABLE 2 Clinical and Biochemical Characteristics of 61 Patients Characteristic Value patients treated with beta-blockers* 26 (43) Child-Pugh classification* A 53 (87) B 8 (13) Mean score in all patients (5 8) Ascites* Mild 3 (5) Moderate 1 (2) Encephalopathy* 2 (3) Splenomegaly* 35 (57) Total bilirubin level (mg/dl) ( ) Albumin level (g/l) ( ) International normalized ratio ( ) Aspartate aminotransferase level (IU/mL) (10 255) Alanine aminotransferase level (IU/mL) (9 224) Platelet count ( 10 3 / L) (39 336) * Data are numbers of patients. Numbers in parentheses are percentages. Data are the mean standard deviation. Numbers in parentheses are the range. To convert to micromoles per liter, multiply by treated with beta-blockers and that those with small EV be rechecked. This approach was established on the basis of the high risk of bleeding from EV, which are associated with severity of liver disease and the size of EV; both of these factors are the most important predictors of bleeding (3,4). During their lifetime, 60% 80% of patients with cirrhosis develop EV at a rate of 5% per year, and the progression from small to large EV occurs in 5% 10% of the patients after the 1st year (5 8). The risk of mortality after the first occurrence of bleeding is substantial, and, therefore, prophylactic measures are necessary (9 11). Primary prevention with beta-blockers leads to a major risk reduction in the first occurrence of bleeding from EV, with a tendency to improve survival, especially in patients with large EV (9,12,13). Thus, early diagnosis and follow-up for EV in cirrhotic patients are necessary. Endoscopy of the upper gastrointestinal tract, which is the most common and accurate procedure for evaluation of EV, is unpleasant and relatively costly (14,15). It also carries a risk, albeit small, of complications, such as perforation of the esophagus, aspiration of gastric content, and bacteremia (16,17). In addition, sedation with benzodiazepines for this procedure can substantially intensify hepatic encephalopathy (18). To overcome the need for sedation, intranasal endoscopy has been suggested for detection of EV (19). In the past 15 years, a number of noninvasive methods have been developed for the prediction of the presence of EV. With some of these methods, a combination of clinical parameters, such as platelet count, spleen size, and abdominal ultrasonographic (US) findings (20), is used, but the overall sensitivity of the methods does not exceed 70% (21 25). A close relationship between size of EV and some important Doppler US parameters, such as portal flow velocity, congestion index, and waves from hepatic veins, has been reported (26,27). Recently, the need to reduce the cost of diagnostic endoscopy versus the need for prophylactic treatment led some investigators to suggest the use of prophylactic treatment for all patients with cirrhosis (28). This approach has not gained much support because of the relatively high rate of side effects of propranolol and the lack of study results that confirm the effectiveness of such treatment (29,30). in early reports suggested that endoscopy is superior to radiologic examination (31), yet the possible use of a noninvasive technique, such as esophagography with barium, has been neglected. Esophagography is a safe, simple, and noninvasive means to visualize EV. Still, the diagnostic value of this method is limited (32). Efforts to increase the sensitivity of this method by adding administration of anticholinergic drugs were limited because of side effects (33). With better coating of the esophageal mucosa with high-density barium sulfate for detection of mucosal irregularities, overall sensitivity and specificity of 85% and 87%, respectively, were achieved (34). The interest among clinicians and radiologists in implementation of this technique has not been great, probably because of the complicated method of assigning scores. Therefore, the purpose of our study was to prospectively evaluate the accuracy of esophagography with barium for the diagnosis of EV in patients with compensated cirrhosis and portal hypertension,withendoscopicgastroduodenoscopy as the reference standard. MATERIALS AND METHODS Patients From November 2002 to May 2003, 61 consecutive ambulatory patients (34 men, 27 women; mean age, 61 years; range, years) with cirrhosis were offered participation in our study, which was approved by the local Helsinki Committee. All patients consented to participate in the study. All patients underwent endoscopic gastroduodenoscopy first, and within 1 month (mean interval, 23 days 6.3 [standard deviation]; range, days) after, they underwent esophagography. At entry into the study, each patient underwent a physical examination and blood tests, which included a complete blood cell count, coagulation profile, and biochemical parameters (albumin, aspartate aminotransferase, alanine aminotransferase, and creatinine levels). In all 61 patients, cirrhosis was diagnosed with liver biopsy and clinical parameters: 38 (62%) had postnecrotic disease; nine (14%) had nonalcoholic steatohepatitis; nine (14%) had alcoholic liver disease; and five (8%) had autoim- 536 Radiology November 2005 Farber et al

3 Figure 1. Left posterior oblique barium esophagrams. (a) Patient without EV (grade 0). (b) Patient with small EV (grade 1) (arrows). Note the irregularity of the mucosa at the lower esophagus (arrows). (c) Patient with large grade 2 EV. TABLE 3 at Endoscopic Gastroduodenoscopy Diagnosed with Esophagography with Barium Classification at Esophagography* Grade F0 mune hepatitis and primary biliary cirrhosis. Patients were assigned scores according to the Child-Pugh classification (35); in 53 (87%) of patients, the Child- Pugh classification was A, and in eight (13%) of them, the Child-Pugh classification was B. In most of them, the Child- Pugh scores were low the mean score was without ascites (compensated cirrhosis) (Tables 1,2). Endoscopic Gastroduodenoscopy of Upper Gastrointestinal Tract Gastroduodenoscopy endoscopy of the upper gastrointestinal tract was performed by three experienced gastroenterologists (R.E., with 20 years of experience; I.C., with 10 years of experience; K.Y., with 15 years of experience) and a 1st-year gastroenterologist (E.F.), always in teams of two, by using an endoscope (Q145; Olympus, Tokyo, Japan) after intravenous premedication with 5 mg of Endoscopic Grade F1 midazolam (Taro Pharmaceutical International, Yakum, Israel). EV were described and assigned grades, according to previously published criteria (36). The EV were assigned grades according to their shape and size: grade F0, no EV detected; grade F1, small straight EV; grade F2, slightly enlarged tortuous EV occupying less than one-third of the esophageal lumen; and grade F3, large coil-shaped EV that occupied more than one-third of the esophageal lumen. For evaluation of the stomach, signs and grades of portal hypertensive gastropathy were described, according to classification of the Northern Italian Endoscopic Club for the Study of Esophageal Varices index (37). No complications were noted after endoscopic gastroduodenoscopy. Esophagography Grade F2 Grade F3 Grade 0 (n 24 [39]) Grade 1 (n 18 [29]) Grade 2 (n 19 [31]) * Numbers in brackets are percentages. A team of two experienced radiologists (D.F., with 10 years of experience, and N.B., with 5 years of experience) performed esophagography. Fasting was not needed. The patient was asked to lie down in a supine left lateral oblique position, and imaging was focused on the middle and lower areas of the esophagus. For best mucosal coating, we used two barium sulfate formulations as follows: In the initial radiographic series, usually six or seven views, we used ml of high-density barium sulfate, 60% wt/vol (E-Z HD; E-Z-Em, Westbury, NY). After the initial esophagographic study, we obtained another six or seven views with 5 10 ml of barium sulfate paste, 98% wt/vol (E-Z-Paste; E-Z-Em). All views were acquired during esophageal relaxation with mucosal coating with a single swallowing of contrast agent, and they were obtained in 5 8 minutes altogether. The EV manifested as irregularities of the esophageal folds. We arbitrarily divided all radiologic findings into three grades: grade 0, no EV were seen; grade 1, EV manifested as very mild irregularities of the folds; and grade 2, the irregularity of the folds (EV) was clearly present (Fig 1). The interpretation of the esophagrams was performed independently by the two radiologists, without any knowledge of the results of endoscopic gastroduodenoscopy. Telephone Survey To include the patients point of view in our study, we conducted a telephone interview (E.F.) in regard to the preferred method for evaluation of EV. The tele- Volume 237 Number 2 Esophageal Varices 537

4 TABLE 4 Ability of Esophagraphy with Barium to Depict EV according to Grade Assigned at Endoscopic Gastroduodenoscopy Radiology Grade of EV phone interview was conducted 65 days after the last patient completed participation in the study. Statistical Analysis The statistical analysis was performed (S.L.) by using software (SPSS, version 11.0, ; SPSS, Chicago, Ill) and the SCRN procedure of Computer Programs for Epidemiologists (PEPI, version 4, 2001; Sagebrush, Salt Lake City, Utah). Results of endoscopic gastroduodenoscopy were considered the reference standard. The sensitivity (true-positive findings/ [true-positive findings false-negative findings]) and specificity (true-negative findings/[true-negative findings falsepositive findings]) of esophagography to identify EV of all grades were calculated, and a 95% confidence interval was estimated for these parameters. The positive and negative predictive values were also calculated, since we wished to describe the sensitivity and specificity of the diagnostic method rather than to detect statistically significant causal associations. Thus, all patients eligible to be included in the study during the study period were recruited. Furthermore, power analyses were not performed a priori. The unit of analysis was the occurrence and detectability of any EV (whether single or multiple) considered together in a patient. Thus, there was no need for correction for multiple EV (data clustering). RESULTS True-Positive True-Negative In 24 (39%) patients, no EV (grade F0) were detected at endoscopic gastroduodenoscopy. In 14 (23%) patients, grade F1 EV were detected; in 17 (28%) patients, grade F2 EV were detected; and in six (10%) patients, grade F3 EV were detected. Portal hypertensive gastropathy was found in 28 (46%) patients, and in these 28 patients, it was assigned a grade of mild in 22 (79%) and of severe in six (21%). False-Positive False-Negative Sensitivity (%) Esophagography did not disclose EV in 24 (39%) patients (grade 0). In four of these patients, endoscopic gastroduodenoscopy disclosed grade F1 EV; that is, there were four false-negative results with esophagography. Eighteen (30%) patients had grade 1 EV (mild esophageal fold irregularity). In four patients from this group, endoscopic gastroduodenoscopy did not depict EV (grade F0); thus, there were four false-positive results. Grade F2 EV were identified at endoscopic gastroduodenoscopy in four patients of this latter group. Grade 2 EV were found at esophagography in 19 (31%) patients. All patients in this group had large EV, as depicted at endoscopic gastroduodenoscopy (grades F2 and F3) (Table 3). The overall sensitivity and specificity (the capability of esophagography to help in the identification of EV of all grades) was 89% and 83% accordingly. The 95% confidence intervals were 75.9 to 96.5 and 64.5 to 94.7, respectively (Table 4). Overall positive and negative predictive values were similar to sensitivity and specificity values (as a result of an equal number of false-positive and false negative results). Positive predictive value is the capability of esophagography to predict a diagnosis. Overall accuracy was 87% (53 of 61). Within the subgroups, the sensitivity and specificity (the ability of esophagography to depict EV according to each grade) were 71% and 83%, respectively, for small EV (grade F1) (95% confidence interval: 44.5%, 90% for sensitivity and 64.5%, 94.5% for specificity). All large EV (grades F2 and F3) were detected by using esophagography. As a result of this finding, both sensitivity and specificity of esophagography (ability of esophagography to depict large EV) were 100% (95% confidence interval: 60%, 100%) (Table 4). The results of the telephone survey showed that 39 patients reported that they preferred esophagography and that Specificity (%) Figure 2. Diagram shows suggested algorithm for use of esophagography with barium (BaEs) and endoscopic gastroduodenoscopy (referred to as endoscopy ). According to our data and the medical literature, we believe that the classification of the incidence of EV will be approximately 40% for grade 0, 30% for grade 1 (mild irregularity of the folds), and 30% for grade 2 (irregularity of the folds was clearly present). Repeat studies are recommended every 2 years for patients with cirrhosis and no EV. Other tests refers to other diagnostic imaging procedures for determination of spleen size and portal hypertension. only six patients preferred endoscopic gastroduodenoscopy. Four patients had no preference. DISCUSSION Positive Predictive Value (%) Negative Predictive Value (%) Overall (75.9, 96.5) 83 (64.5, 94.7) 89 (75.9, 96.5) 83 (64.5, 94.7) F (44.5, 90) 83 (64.5, 94.5) 71 (44.5, 90) 83 (64.5, 94.5) F (83.9, 100) 100 (86, 100) 100 (83.9, 100) 100 (86, 100) F (60, 100) 100 (86, 100) 100 (60, 100) 100 (60, 100) Note. Numbers in parentheses represent the 95% confidence interval expressed as percentages. On the basis of the results of our prospective study, we suggest that esophagography may have a role in the planning of a program for primary prevention of bleeding from EV in patients with cirrhosis. Esophagography used to be the main diagnostic procedure for EV before endoscopic gastroduodenoscopy was introduced. The capability of endoscopic gastroduodenoscopy to be used for both diagnosis and treatment of upper gastrointestinal tract complications, and its high availability, allowed it to become the primary diagnostic tool (31). Over the past 20 years, researchers in only one study compared esophagography with endoscopic gastroduodenoscopy, and the results of that study showed that esophagography is very sensitive for the diagnosis of EV (34). 538 Radiology November 2005 Farber et al

5 That study did not receive widespread attention, and the method was not implemented by clinicians or radiologists, possibly because of the rather complex grading of EV with esophagrams. A debate about the preferred approach for evaluation and preventive treatment of bleeding from EV has arisen (38,39), and suggestions to use noninvasive techniques are increasing (15,21,22,26,40). The alternative to using barium esophagography was not discussed in this debate, probably because gastroduodenoscopy is not considered to be a purely noninvasive technique (30). It is not surprising that only half of the patients with cirrhosis actually undergo diagnostic endoscopic gastroduodenoscopy prior to transplantation (41). The low clinical awareness of and high cost of endoscopic gastroduodenoscopy are probably the major factors for the decreased use of this modality. Because underuse of EV evaluation may contribute to mortality and morbidity of patients with cirrhosis, some investigators have suggested the use of universal treatment with betablockers, thus avoiding the use of endoscopic gastroduodenoscopy and decreasing the burden of cost from this modality (28). Interobserver variation that often occurs with endoscopic gastroduodenoscopy can be a problem in the interpretation of radiologic findings. Absence of clear-cut criteria for interpretation also may have hindered the adoption of esophagography. To make interpretation easier and clearer, we simplified the grades of EV depicted with esophagography to only three, hoping that this simplification will facilitate use of esophagography with barium and the grading system and render it more likely to be adopted. Furthermore, we outlined an algorithm offering an approach for the use of esophagography or endoscopic gastroduodenoscopy in the evaluation of EV in patients with cirrhosis (Fig 2). Finally, patients surveyed indicated a clear preference for esophagography over endoscopic gastroduodenoscopy. The limitation of this study was the lower capacity of esophagography to aid in the diagnosis of gastropathy and gastric varices. Therefore, it can be used as a tool only for initial diagnosis and follow-up of patients with EV. In this study, after ascertaining that esophagography with barium is a highly accurate procedure for diagnosis of EV in patients with compensated liver disease, it clearly is an alternative option for endoscopic gastroduodenoscopy. Use of this method for both diagnosis and subsequent follow-up will help the clinician to select patients for prophylactic treatment at a lower cost and with increased convenience to patients. References 1. Grace ND, Groszmann RJ, Garcia-Tsao G, et al. Portal hypertension and variceal bleeding: an AASLD single topic symposium. Hepatology 1998;28(3): Jalan R, Hayes PC. UK guidelines on the management of variceal haemorrhage in cirrhotic patients. British Society of Gastroenterology. Gut 2000;46(suppl 3 4): III1 III D Amico G, Pagliaro L, Bosch J. Pharmacological treatment of portal hypertension: an evidence-based approach. Semin Liver Dis 1999;19: The Northern Italian Endoscopic Club for the Study of Esophageal Varices: a prospective multicenter study. N Engl J Med 1988; 319: Christensen E, Fauerholtdt L, Schlichting P, et al. Aspects of the natural history of gastrointestinal bleeding in cirrhosis and the effect of prednisone. Gastroenterology 1981;81: Cales P, Desmorat H, Vinel JP, et al. Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding. Gut 1990;31: Merli M, Nicolini G, Angeloni S, et al. Incidence and natural history of small esophageal varices in cirrhotic patients. J Hepatol 2003;38(3): Zoli M, Merkel C, Magalotti D, et al. Natural history of cirrhotic patients with small esophageal varices: a prospective study. Am J Gastroenterol 2000;95(2): Hayes PC, Davis JM, Lewis JA, et al. Metaanalysis of value of propranolol in prevention of variceal haemorrhage. Lancet 1990; 336: D Amico G, De Franchis R; Cooperative Study Group. Upper digestive bleeding in cirrhosis. Post-therapeutic outcome and prognostic indicators. Hepatology 2003; 38(3): Nidegger D, Ragot S, Berthelemy P, et al. Cirrhosis and bleeding: the need for very early management. J Hepatol 2003;39(4): Poynard T, Cales P, Pasta L, et al. Betaadrenergic-antagonist drugs in the prevention of gastrointestinal bleeding in patients with cirrhosis and esophageal varices: an analysis of data and prognostic factors in 589 patients from four randomized clinical trials. Franco-Italian Multicenter Study Group. N Engl J Med 1991; 324: Lebrec D, Poynard T, Hillon P, Benhamou JP. Propranolol for prevention of recurrent gastrointestinal bleeding in patients with cirrhosis: a controlled study. N Engl J Med 1981;305(23): Spiegel BM, Targownik L, Dulai GS, et al. Endoscopic screening for esophageal varices in cirrhosis: is it ever cost effective? Hepatology 2003;37(2): Schepis F, Camma C, Niceforo D, et al. Which patients with cirrhosis should undergo endoscopic screening for esophageal varices detection? Hepatology 2001;33: Nashef SA, Pagliero KM. Instrumental perforation of the esophagus in benign disease. Ann Thorac Surg 1987;44(4): Norfleet R, Mitchell D, Philo J. Does bacteremia follow upper gastrointestinal endoscopy? Am J Gastroenterol 1981;76(5): Assy N, Rosser BG, Grahame GR, Minuk GY. Risk of sedation for upper GI endoscopy exacerbating subclinical hepatic encephalopathy in patients with cirrhosis. Gastrointest Endosc 1999;49: Saeian K, Staff D, Knox J, et al. Unsedated transnasal endoscopy: a new technique for accurately detecting and grading esophageal varices in cirrhotic patients. Am J Gastroenterol 2002;97: Cottone M, D Amico G, Maringhini A, et al. Predictive value of ultrasonography in the screening of non-ascitic cirrhotic patients with large varices. J Ultrasound Med 1986;5(4): Pilette C, Oberti F, Aubé C, et al. Noninvasive diagnosis of esophageal varices in chronic liver diseases. J Hepatol 1999;31: Zaman A, Hapke R, Flora K, Rosen HR, Benner K. Factors predicting the presence of esophageal or gastric varices in patients with advanced liver disease. Am J Gastroenterol 1999;94: Madhotra R, Mulcahy HE, Willner I, et al. Prediction of esophageal varices in patients with cirrhosis. J Clin Gastroenterol 2002;34(1): Giannini E, Botta F, Borro P, et al. Platelet count/spleen diameter ratio: proposal and validation of a non-invasive parameter to predict the presence of oesophageal varices in patients with liver cirrhosis. Gut 2003;52(8): Thomopoulos KC, Labropoulou-Karatza C, Mimidis KP, et al. Non-invasive predictors of the presence of large oesophageal varices in patients with cirrhosis. Dig Liver Dis 2003;35(7): Siringo S, Bolondi L, Gaiani S, et al. Timing of the first variceal hemorrhage in cirrhotic patients: prospective evaluation of Doppler flowmetry, endoscopy and clinical parameters. Hepatology 1994;20: Gorka W, al Mulla A, al Sebayel M et al. Qualitative hepatic venous Doppler sonography versus portal flowmetry in predicting the severity of esophageal varices in hepatic C cirrhosis. AJR Am J Roentgenol 1997;169: Saab S, DeRosa V, Nieto J, Durazo F, Han S, Roth B. Costs and clinical outcomes of primary prophylaxis of variceal bleeding in patients with hepatic cirrhosis: a decision analytic model. Am J Gastroenterol 2003; 98(4): Opie LH, Yusuf S. Beta-blocking agents. In: Opie LH, ed, Drugs for the heart. 5th ed. Philadelphia, Pa: Saunders, 2001; D Amico G, Morabito A. Noninvasive markers of esophageal varices: another round, not the last. Hepatology 2004;39(1): Conn HO, Mitchell JR, Brodoff MG. A comparison of the radiologic and esophagoscopic diagnosis of esophageal varices. N Engl J Med 1961;265: Volume 237 Number 2 Esophageal Varices 539

6 32. Kirsh IE, Blackwell CC, Bennett HD. Roentgen diagnosis of esophageal varices: comparison of roentgen and esophageal findings in 502 cases. Am J Roentgenol Radium Ther Nucl Med 1955;74: Dalinka MK, Smith EH, Wolfe RD. Pharmacologically enhanced visualization of esophageal varices by Pro-Bantine. Radiology 1972;102: Ginai AZ, van Buuren HR, Hop WC, et al. Oesophageal varices: how reliable is a barium swallow? Br J Radiol 1993;66(784): Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60(8): The general rules for recording endoscopic findings on esophageal varices. Japanese Research Society for Portal Hypertension. Jpn J Surg 1980;10(1): Spina GP, Arcidiacono R, Bosh J, et al. Gastric endoscopic features in portal hypertension: final report of a consensus conference, Milan. J Hepatol 1994;21: Fleig WE. To scope or not to scope: still a question. Hepatology 2001;33(2): Chalasani N, Imperiale TF. Screening for varices in patients with cirrhosis: where do we stand? Am J Gastroenterol 2001;96(3): Chalasani N, Imperiale TF, Ismail A, et al. Predictors of large esophageal varices in patients with cirrhosis. Am J Gastroenterol 1999;94(11): Arguedas MR, McGuire BM, Fallon MB, et al. The use of screening and preventive therapies for gastroesophageal varices in patients referred for evaluation of orthotopic liver transplantation. Am J Gastroenterol 2001;96(3): Radiology November 2005 Farber et al