Evaluation and follow-up of patients with cirrhosis and oesophageal varices

Abstract

Bleeding from ruptured oesophago-gastric varices is the most severe complication of cirrhosis; despite the progress achieved over the past two decades [[1]Pagliaro L. D'Amico G. Pasta L. Tinè F. Aragona E. Politi F. et al.Efficacy and efficiency of treatments in portal hypertension.in: de Franchis R. Portal hypertension II. Proceedings of the second Baveno international consensus workshop on definitions, methodology and therapeutic strategies. Blackwell Science, Oxford1996: 159-179Google Scholar] in the management of this medical emergency, mortality figures are still in the range of 20–30% within 6 weeks of the bleeding episode. Therefore, prevention of variceal bleeding is an important goal both for cirrhotic patients and for the physicians dealing with them. Since oesophago-gastric varices are a consequence of portal hypertension, knowledge of the natural history of this condition may help in making important decisions about the diagnosis, follow-up and treatment of patients with cirrhosis. We know from longitudinal studies that varices eventually develop in all cirrhotic patients [[2]Christensen E. Fauerholdt L. Schlichting P. Juhl E. Poulsen H. Tygstrup N. Aspects of natural history of gastrointestinal bleeding in cirrhosis and the effect of prednisone.Gastroenterology. 1981; 81: 944-952Abstract Full Text PDF PubMed Scopus (188) Google Scholar], and that once they have developed, they tend to increase in size and to bleed [[3]D'Amico G. Luca A. Natural history. Clinical-haemodynamic correlations. Prediction of the risk of bleeding.Bailliere's Clin Gastroenterol. 1997; 11: 243-256Abstract Full Text PDF PubMed Scopus (200) Google Scholar]. We also know that the prevalence of varices is higher in decompensated than in compensated cirrhosis [[3]D'Amico G. Luca A. Natural history. Clinical-haemodynamic correlations. Prediction of the risk of bleeding.Bailliere's Clin Gastroenterol. 1997; 11: 243-256Abstract Full Text PDF PubMed Scopus (200) Google Scholar], and that large varices have a higher propensity to bleed than small ones [4D'Amico G. Pagliaro L. Bosch J. Pharmacologic treatment of portal hypertension: an evidence-based approach.Semin Liver Dis. 1999; 19: 475-505Crossref PubMed Scopus (629) Google Scholar, 5North-Italian Endoscopic Club for the Study and Treatment of Esophageal VaricesPrediction of the first variceal hemorrhage in patients with cirrhosis of the liver and esophageal varices.N Engl J Med. 1988; 319: 983-989Crossref PubMed Scopus (1060) Google Scholar]. However, at a given point in time, a variable proportion of patients will not have varices; in addition, progression of varices from small to large is variable; finally, only about one-third of patients with varices will bleed. The current policy is to treat all patients with large varices prophylactically with non-selective beta-blockers [[6]de Franchis R. Updating consensus in portal hypertension: report of the Baveno III consensus workshop on definitions, methodology and therapeutic strategies in portal hypertension.J Hepatol. 2000; 33: 846-852Abstract Full Text Full Text PDF PubMed Scopus (485) Google Scholar]. However, the most efficient strategy for identifying such patients is not clearly defined; to achieve this goal, two important questions need to be answered:(a)Can we identify the patients with varices before doing endoscopy?(b)How frequently should patients be re-endoscoped to detect varices at risk of bleeding?Answering the first question would have important economic implications. In fact, the reported prevalence of oesophageal varices is variable [[7]Pascal J.P. Calès P. Desmorat H. Natural history of esophageal varices.in: Bosch J. Rodès J. Recent advances in the pathophysiology and treatment of portal hypertension. Serono Symposia review no. 22. Serono Symposia, Rome1989: 127-142Google Scholar], ranging in different series between 24% and 80%. Thus, screening all cirrhotic patients with upper GI endoscopy to detect the presence of varices implies a number of unnecessary endoscopies. Predicting the presence of oesophageal varices by non-invasive means would restrict the performance of endoscopy to those patients with a high probability of having varices. Until recently, scanty data were available on this matter [[8]Pagliaro L. D'Amico G. Pasta L. Politi F. Vizzini G. Traina M. et al.Portal hypertension in cirrhosis: natural history.in: Bosch J. Groszmann R.J. Portal hypertension, pathophysiology and treatment. Blackwell Scientific, Oxford1994: 72-92Google Scholar], and a recent consensus workshop on portal hypertension recommended that all patients with cirrhosis should be evaluated by endoscopy to ascertain the presence of varices [[6]de Franchis R. Updating consensus in portal hypertension: report of the Baveno III consensus workshop on definitions, methodology and therapeutic strategies in portal hypertension.J Hepatol. 2000; 33: 846-852Abstract Full Text Full Text PDF PubMed Scopus (485) Google Scholar]. Over the last 5 years a number of studies [9Garcia-Tsao G. Escorsell A. Zakko M. Patch D. Matloff D. Grace N. et al.Predicting the presence of significant portal hypertension and esophageal varices in compensated cirrhotic patients (abstract).Hepatology. 1997; 26: 360AGoogle Scholar, 10Schepis F. Cammà C. Niceforo D. Magnano A. Pallio S. Cinquegrani M. et al.Which patients should undergo endoscopic screening for esophageal varices detection?.Hepatology. 2001; 33: 333-338Crossref PubMed Scopus (248) Google Scholar, 11Zaman A. Becker T. Lapidus J. Benner K. Risk factors for the presence of varices in cirrhotic patients without history of variceal hemorrhage.Arch Int Med. 2001; 161: 2564-2570Crossref PubMed Scopus (134) Google Scholar, 12Madhotra R. Mulcahy H.E. Willner I. Reuben A. Prediction of esophageal varices in patients with cirrhosis.J Clin Gastroenterol. 2002; 34: 81-85Crossref PubMed Scopus (167) Google Scholar] have addressed the issue of identifying patients with varices before doing endoscopy. In general, a low platelet count has been constantly found to be related to the presence of varices. Other prognostic factors identified in different studies include a portal vein diameter on ultrasound scan of ≥13 mm [[10]Schepis F. Cammà C. Niceforo D. Magnano A. Pallio S. Cinquegrani M. et al.Which patients should undergo endoscopic screening for esophageal varices detection?.Hepatology. 2001; 33: 333-338Crossref PubMed Scopus (248) Google Scholar], an advanced Child–Pugh class [[11]Zaman A. Becker T. Lapidus J. Benner K. Risk factors for the presence of varices in cirrhotic patients without history of variceal hemorrhage.Arch Int Med. 2001; 161: 2564-2570Crossref PubMed Scopus (134) Google Scholar], hypoalbuminaemia and the presence of telangiectasias [[9]Garcia-Tsao G. Escorsell A. Zakko M. Patch D. Matloff D. Grace N. et al.Predicting the presence of significant portal hypertension and esophageal varices in compensated cirrhotic patients (abstract).Hepatology. 1997; 26: 360AGoogle Scholar], low prothrombin activity [[10]Schepis F. Cammà C. Niceforo D. Magnano A. Pallio S. Cinquegrani M. et al.Which patients should undergo endoscopic screening for esophageal varices detection?.Hepatology. 2001; 33: 333-338Crossref PubMed Scopus (248) Google Scholar] and splenomegaly [[12]Madhotra R. Mulcahy H.E. Willner I. Reuben A. Prediction of esophageal varices in patients with cirrhosis.J Clin Gastroenterol. 2002; 34: 81-85Crossref PubMed Scopus (167) Google Scholar]. A prognostic model based on a platelet count of <100,000/mm3, a portal vein diameter >13 mm and a prothrombin activity of <70% has been developed [[10]Schepis F. Cammà C. Niceforo D. Magnano A. Pallio S. Cinquegrani M. et al.Which patients should undergo endoscopic screening for esophageal varices detection?.Hepatology. 2001; 33: 333-338Crossref PubMed Scopus (248) Google Scholar]; however, when this model was tested in an independent patients series [[13]Riggio O. Angeloni S. Nicolini G. Merli M. Merkel C. Endoscopic screening for esophageal varices in cirrhotic patients (letter).Hepatology. 2002; 35: 501-502Crossref PubMed Scopus (32) Google Scholar], the results were disappointing, since 42% of the patients classified by the model as having the highest risk had no varices, and 34% of those classified as having the lowest risk had varices.Knowledge of the rate of development and growth of oesophageal varices would have important implications because it would help in optimizing the intervals for follow-up endoscopy, which aims at detecting varices at risk of bleeding before they actually bleed, in order to start prophylactic treatment. Too short intervals would unnecessarily increase the workload of endoscopy units, while too long intervals would increase the risk of patients bleeding between endoscopies. This important problem has been addressed by the paper by Merli et al. [[14]Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar] published in this issue. In this study, 206 cirrhotic patients, 113 without and 93 with small varices, were included and followed up with yearly endoscopies for a mean of about 3 years. The rate of incidence of new varices was about 9% per year. None of several parameters evaluated at baseline correlated with the incidence of new varices. The rate of progression of varices to medium-sized or large in patients with small varices at entry was 12%, 25% and 31% at 1, 2 and 3 years, respectively. On the other hand, the occurrence of medium-sized or large varices was much lower in patients with no varices at baseline endoscopy, reaching 9% at 3 years. Independent predictors of the enlargement of varices were the alcoholic aetiology of cirrhosis, Child–Pugh class B or C and the presence of red wale markings on varices at baseline endoscopy. The 2 year bleeding rate was significantly higher in patients with small varices than in those with no varices at baseline endoscopy (12% vs. 2%; P<0.01). The presence of red wale markings on varices at baseline was the only prognostic factor for bleeding. Interestingly, only one patient without varices and three with small varices at entry bled before large varices were detected at follow-up endoscopy.How does this paper compare with previous studies? As far as the incidence of new varices is concerned, the figure reported by Merli et al. [[14]Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar] is similar to those observed by others [2Christensen E. Fauerholdt L. Schlichting P. Juhl E. Poulsen H. Tygstrup N. Aspects of natural history of gastrointestinal bleeding in cirrhosis and the effect of prednisone.Gastroenterology. 1981; 81: 944-952Abstract Full Text PDF PubMed Scopus (188) Google Scholar, 8Pagliaro L. D'Amico G. Pasta L. Politi F. Vizzini G. Traina M. et al.Portal hypertension in cirrhosis: natural history.in: Bosch J. Groszmann R.J. Portal hypertension, pathophysiology and treatment. Blackwell Scientific, Oxford1994: 72-92Google Scholar, 15Gentilini P. Laffi G. La Villa G. Romanelli R.G. Buzzelli G. Casini-Raggi V. et al.Long course and prognostic factors of virus-induced cirrhosis of the liver.Am J Gastroenterol. 1997; 92: 66-72PubMed Google Scholar, 16Primignani M. Albè R. Preatoni P. Carnevale P. Bianchi M.B. Parravicini E. et al.‘De novo’ development of esophageal varices in patients with a recent histologic diagnosis of liver cirrhosis (abstract).Gastroenterology. 1998; 114: 1324AGoogle Scholar]. One major exception is the study by Calès et al. [[17]Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar], who, in 41 patients of cirrhotics of alcoholic aetiology with a relatively large proportion of Child B and C patients, observed 1 and 2 year incidences of varices of 23% and 50%, respectively. A possible explanation of this discrepancy is that in the other studies the proportions of alcoholic cirrhosis and of patients with advanced liver disease were much lower than in the Calès study. The fact that Merli et al. [[14]Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar] were unable to identify any predictor of the ‘de novo’ appearance of varices, while disparate prognostic factors were identified in the other studies, suggests that no strong predictor exists. Concerning the progression of varices from small to large, the rates reported by other authors vary greatly: 25% over 6 years by Pagliaro et al. [[8]Pagliaro L. D'Amico G. Pasta L. Politi F. Vizzini G. Traina M. et al.Portal hypertension in cirrhosis: natural history.in: Bosch J. Groszmann R.J. Portal hypertension, pathophysiology and treatment. Blackwell Scientific, Oxford1994: 72-92Google Scholar], 45% at 3 years by Zoli et al. [[18]Zoli M. Merkel C. Magalotti D. Gueli C. Grimaldi M. Gatta A. et al.Natural history of cirrhotic patients with small esophageal varices: a prospective study.Am J Gastroenterol. 2000; 95: 503-508Crossref PubMed Google Scholar], 70% at 2 years by Calès et al. [[17]Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar], 30% at 2 years in another study by the same author [[19]Calès P. Oberti F. Payen J.L. Naveau S. Guyader D. Blanc P. et al.Lack of effect of propranolol in the prevention of large esophageal varices in patients with cirrhosis.Eur J Gastroenterol Hepatol. 1999; 11: 741-745Crossref PubMed Scopus (120) Google Scholar], and 15% at 16 months in the paper by Merkel et al. [[20]Merkel C. Angeli P. Marin R. Zanella P. Felders M. Bernardinello E. et al.Beta-blockers in the prevention of aggravation of esophageal varices in patients with cirrhosis and small esophageal varices: interim analysis of a controlled clinical trial (abstract).Hepatology. 1998; 28: 453AGoogle Scholar]. This large variability of results is probably related to differences in the populations studied and to the different endoscopic follow-up intervals used in the studies. As far as predictors of variceal progression are concerned, the only factors identified in more than one study are the Child–Pugh class at baseline [14Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar, 17Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar], or its worsening during follow-up [17Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar, 18Zoli M. Merkel C. Magalotti D. Gueli C. Grimaldi M. Gatta A. et al.Natural history of cirrhotic patients with small esophageal varices: a prospective study.Am J Gastroenterol. 2000; 95: 503-508Crossref PubMed Google Scholar]. The observation by Merli et al. [[14]Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar] that the rate of bleeding was significantly higher in patients with small varices at baseline than in those without is in keeping with previous data [[17]Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar].How can we use the above information to decide the strategies to prevent variceal bleeding?Concerning the possibility of restricting screening endoscopy to the patients at the highest risk of having varices, the available data are disappointing, since no prognostic model to predict the presence of varices has been proven precise enough to select patients. Therefore, the recommendation [[6]de Franchis R. Updating consensus in portal hypertension: report of the Baveno III consensus workshop on definitions, methodology and therapeutic strategies in portal hypertension.J Hepatol. 2000; 33: 846-852Abstract Full Text Full Text PDF PubMed Scopus (485) Google Scholar] that all cirrhotic patients should be screened for the presence of varices at the time of the initial diagnosis of cirrhosis still stands.As far as deciding the timing of follow-up endoscopy is concerned, we must first decide which level of risk (i.e. what proportion of patients bleeding before starting prophylactic treatment) we are willing to accept. If we set this level at 10%, then patients with no varices at baseline can be re-endoscoped at 3 year intervals. In fact, the rate of development of large varices and the rate of bleeding in these patients are both below the 10% threshold [8Pagliaro L. D'Amico G. Pasta L. Politi F. Vizzini G. Traina M. et al.Portal hypertension in cirrhosis: natural history.in: Bosch J. Groszmann R.J. Portal hypertension, pathophysiology and treatment. Blackwell Scientific, Oxford1994: 72-92Google Scholar, 14Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar, 17Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar]. For patients with small varices at baseline, the picture is much more complicated, since the reported incidences range between 6% [[8]Pagliaro L. D'Amico G. Pasta L. Politi F. Vizzini G. Traina M. et al.Portal hypertension in cirrhosis: natural history.in: Bosch J. Groszmann R.J. Portal hypertension, pathophysiology and treatment. Blackwell Scientific, Oxford1994: 72-92Google Scholar] and 70% [[17]Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar] at 2 years for large varices and between 4% at 16 months [[20]Merkel C. Angeli P. Marin R. Zanella P. Felders M. Bernardinello E. et al.Beta-blockers in the prevention of aggravation of esophageal varices in patients with cirrhosis and small esophageal varices: interim analysis of a controlled clinical trial (abstract).Hepatology. 1998; 28: 453AGoogle Scholar] and 24% at 2 years [[17]Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar] for bleeding. Clearly, more information is needed on this point. For the time being, it appears reasonable to maintain the recommended interval of 1–2 years [[6]de Franchis R. Updating consensus in portal hypertension: report of the Baveno III consensus workshop on definitions, methodology and therapeutic strategies in portal hypertension.J Hepatol. 2000; 33: 846-852Abstract Full Text Full Text PDF PubMed Scopus (485) Google Scholar], adopting the shorter interval for patients with alcoholic cirrhosis [14Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar, 17Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar], with more severe impairment of liver function [8Pagliaro L. D'Amico G. Pasta L. Politi F. Vizzini G. Traina M. et al.Portal hypertension in cirrhosis: natural history.in: Bosch J. Groszmann R.J. Portal hypertension, pathophysiology and treatment. Blackwell Scientific, Oxford1994: 72-92Google Scholar, 14Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar, 17Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar] and with endoscopic risk signs [5North-Italian Endoscopic Club for the Study and Treatment of Esophageal VaricesPrediction of the first variceal hemorrhage in patients with cirrhosis of the liver and esophageal varices.N Engl J Med. 1988; 319: 983-989Crossref PubMed Scopus (1060) Google Scholar, 14Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar]. Bleeding from ruptured oesophago-gastric varices is the most severe complication of cirrhosis; despite the progress achieved over the past two decades [[1]Pagliaro L. D'Amico G. Pasta L. Tinè F. Aragona E. Politi F. et al.Efficacy and efficiency of treatments in portal hypertension.in: de Franchis R. Portal hypertension II. Proceedings of the second Baveno international consensus workshop on definitions, methodology and therapeutic strategies. Blackwell Science, Oxford1996: 159-179Google Scholar] in the management of this medical emergency, mortality figures are still in the range of 20–30% within 6 weeks of the bleeding episode. Therefore, prevention of variceal bleeding is an important goal both for cirrhotic patients and for the physicians dealing with them. Since oesophago-gastric varices are a consequence of portal hypertension, knowledge of the natural history of this condition may help in making important decisions about the diagnosis, follow-up and treatment of patients with cirrhosis. We know from longitudinal studies that varices eventually develop in all cirrhotic patients [[2]Christensen E. Fauerholdt L. Schlichting P. Juhl E. Poulsen H. Tygstrup N. Aspects of natural history of gastrointestinal bleeding in cirrhosis and the effect of prednisone.Gastroenterology. 1981; 81: 944-952Abstract Full Text PDF PubMed Scopus (188) Google Scholar], and that once they have developed, they tend to increase in size and to bleed [[3]D'Amico G. Luca A. Natural history. Clinical-haemodynamic correlations. Prediction of the risk of bleeding.Bailliere's Clin Gastroenterol. 1997; 11: 243-256Abstract Full Text PDF PubMed Scopus (200) Google Scholar]. We also know that the prevalence of varices is higher in decompensated than in compensated cirrhosis [[3]D'Amico G. Luca A. Natural history. Clinical-haemodynamic correlations. Prediction of the risk of bleeding.Bailliere's Clin Gastroenterol. 1997; 11: 243-256Abstract Full Text PDF PubMed Scopus (200) Google Scholar], and that large varices have a higher propensity to bleed than small ones [4D'Amico G. Pagliaro L. Bosch J. Pharmacologic treatment of portal hypertension: an evidence-based approach.Semin Liver Dis. 1999; 19: 475-505Crossref PubMed Scopus (629) Google Scholar, 5North-Italian Endoscopic Club for the Study and Treatment of Esophageal VaricesPrediction of the first variceal hemorrhage in patients with cirrhosis of the liver and esophageal varices.N Engl J Med. 1988; 319: 983-989Crossref PubMed Scopus (1060) Google Scholar]. However, at a given point in time, a variable proportion of patients will not have varices; in addition, progression of varices from small to large is variable; finally, only about one-third of patients with varices will bleed. The current policy is to treat all patients with large varices prophylactically with non-selective beta-blockers [[6]de Franchis R. Updating consensus in portal hypertension: report of the Baveno III consensus workshop on definitions, methodology and therapeutic strategies in portal hypertension.J Hepatol. 2000; 33: 846-852Abstract Full Text Full Text PDF PubMed Scopus (485) Google Scholar]. However, the most efficient strategy for identifying such patients is not clearly defined; to achieve this goal, two important questions need to be answered:(a)Can we identify the patients with varices before doing endoscopy?(b)How frequently should patients be re-endoscoped to detect varices at risk of bleeding? Answering the first question would have important economic implications. In fact, the reported prevalence of oesophageal varices is variable [[7]Pascal J.P. Calès P. Desmorat H. Natural history of esophageal varices.in: Bosch J. Rodès J. Recent advances in the pathophysiology and treatment of portal hypertension. Serono Symposia review no. 22. Serono Symposia, Rome1989: 127-142Google Scholar], ranging in different series between 24% and 80%. Thus, screening all cirrhotic patients with upper GI endoscopy to detect the presence of varices implies a number of unnecessary endoscopies. Predicting the presence of oesophageal varices by non-invasive means would restrict the performance of endoscopy to those patients with a high probability of having varices. Until recently, scanty data were available on this matter [[8]Pagliaro L. D'Amico G. Pasta L. Politi F. Vizzini G. Traina M. et al.Portal hypertension in cirrhosis: natural history.in: Bosch J. Groszmann R.J. Portal hypertension, pathophysiology and treatment. Blackwell Scientific, Oxford1994: 72-92Google Scholar], and a recent consensus workshop on portal hypertension recommended that all patients with cirrhosis should be evaluated by endoscopy to ascertain the presence of varices [[6]de Franchis R. Updating consensus in portal hypertension: report of the Baveno III consensus workshop on definitions, methodology and therapeutic strategies in portal hypertension.J Hepatol. 2000; 33: 846-852Abstract Full Text Full Text PDF PubMed Scopus (485) Google Scholar]. Over the last 5 years a number of studies [9Garcia-Tsao G. Escorsell A. Zakko M. Patch D. Matloff D. Grace N. et al.Predicting the presence of significant portal hypertension and esophageal varices in compensated cirrhotic patients (abstract).Hepatology. 1997; 26: 360AGoogle Scholar, 10Schepis F. Cammà C. Niceforo D. Magnano A. Pallio S. Cinquegrani M. et al.Which patients should undergo endoscopic screening for esophageal varices detection?.Hepatology. 2001; 33: 333-338Crossref PubMed Scopus (248) Google Scholar, 11Zaman A. Becker T. Lapidus J. Benner K. Risk factors for the presence of varices in cirrhotic patients without history of variceal hemorrhage.Arch Int Med. 2001; 161: 2564-2570Crossref PubMed Scopus (134) Google Scholar, 12Madhotra R. Mulcahy H.E. Willner I. Reuben A. Prediction of esophageal varices in patients with cirrhosis.J Clin Gastroenterol. 2002; 34: 81-85Crossref PubMed Scopus (167) Google Scholar] have addressed the issue of identifying patients with varices before doing endoscopy. In general, a low platelet count has been constantly found to be related to the presence of varices. Other prognostic factors identified in different studies include a portal vein diameter on ultrasound scan of ≥13 mm [[10]Schepis F. Cammà C. Niceforo D. Magnano A. Pallio S. Cinquegrani M. et al.Which patients should undergo endoscopic screening for esophageal varices detection?.Hepatology. 2001; 33: 333-338Crossref PubMed Scopus (248) Google Scholar], an advanced Child–Pugh class [[11]Zaman A. Becker T. Lapidus J. Benner K. Risk factors for the presence of varices in cirrhotic patients without history of variceal hemorrhage.Arch Int Med. 2001; 161: 2564-2570Crossref PubMed Scopus (134) Google Scholar], hypoalbuminaemia and the presence of telangiectasias [[9]Garcia-Tsao G. Escorsell A. Zakko M. Patch D. Matloff D. Grace N. et al.Predicting the presence of significant portal hypertension and esophageal varices in compensated cirrhotic patients (abstract).Hepatology. 1997; 26: 360AGoogle Scholar], low prothrombin activity [[10]Schepis F. Cammà C. Niceforo D. Magnano A. Pallio S. Cinquegrani M. et al.Which patients should undergo endoscopic screening for esophageal varices detection?.Hepatology. 2001; 33: 333-338Crossref PubMed Scopus (248) Google Scholar] and splenomegaly [[12]Madhotra R. Mulcahy H.E. Willner I. Reuben A. Prediction of esophageal varices in patients with cirrhosis.J Clin Gastroenterol. 2002; 34: 81-85Crossref PubMed Scopus (167) Google Scholar]. A prognostic model based on a platelet count of <100,000/mm3, a portal vein diameter >13 mm and a prothrombin activity of <70% has been developed [[10]Schepis F. Cammà C. Niceforo D. Magnano A. Pallio S. Cinquegrani M. et al.Which patients should undergo endoscopic screening for esophageal varices detection?.Hepatology. 2001; 33: 333-338Crossref PubMed Scopus (248) Google Scholar]; however, when this model was tested in an independent patients series [[13]Riggio O. Angeloni S. Nicolini G. Merli M. Merkel C. Endoscopic screening for esophageal varices in cirrhotic patients (letter).Hepatology. 2002; 35: 501-502Crossref PubMed Scopus (32) Google Scholar], the results were disappointing, since 42% of the patients classified by the model as having the highest risk had no varices, and 34% of those classified as having the lowest risk had varices. Knowledge of the rate of development and growth of oesophageal varices would have important implications because it would help in optimizing the intervals for follow-up endoscopy, which aims at detecting varices at risk of bleeding before they actually bleed, in order to start prophylactic treatment. Too short intervals would unnecessarily increase the workload of endoscopy units, while too long intervals would increase the risk of patients bleeding between endoscopies. This important problem has been addressed by the paper by Merli et al. [[14]Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar] published in this issue. In this study, 206 cirrhotic patients, 113 without and 93 with small varices, were included and followed up with yearly endoscopies for a mean of about 3 years. The rate of incidence of new varices was about 9% per year. None of several parameters evaluated at baseline correlated with the incidence of new varices. The rate of progression of varices to medium-sized or large in patients with small varices at entry was 12%, 25% and 31% at 1, 2 and 3 years, respectively. On the other hand, the occurrence of medium-sized or large varices was much lower in patients with no varices at baseline endoscopy, reaching 9% at 3 years. Independent predictors of the enlargement of varices were the alcoholic aetiology of cirrhosis, Child–Pugh class B or C and the presence of red wale markings on varices at baseline endoscopy. The 2 year bleeding rate was significantly higher in patients with small varices than in those with no varices at baseline endoscopy (12% vs. 2%; P<0.01). The presence of red wale markings on varices at baseline was the only prognostic factor for bleeding. Interestingly, only one patient without varices and three with small varices at entry bled before large varices were detected at follow-up endoscopy. How does this paper compare with previous studies? As far as the incidence of new varices is concerned, the figure reported by Merli et al. [[14]Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar] is similar to those observed by others [2Christensen E. Fauerholdt L. Schlichting P. Juhl E. Poulsen H. Tygstrup N. Aspects of natural history of gastrointestinal bleeding in cirrhosis and the effect of prednisone.Gastroenterology. 1981; 81: 944-952Abstract Full Text PDF PubMed Scopus (188) Google Scholar, 8Pagliaro L. D'Amico G. Pasta L. Politi F. Vizzini G. Traina M. et al.Portal hypertension in cirrhosis: natural history.in: Bosch J. Groszmann R.J. Portal hypertension, pathophysiology and treatment. Blackwell Scientific, Oxford1994: 72-92Google Scholar, 15Gentilini P. Laffi G. La Villa G. Romanelli R.G. Buzzelli G. Casini-Raggi V. et al.Long course and prognostic factors of virus-induced cirrhosis of the liver.Am J Gastroenterol. 1997; 92: 66-72PubMed Google Scholar, 16Primignani M. Albè R. Preatoni P. Carnevale P. Bianchi M.B. Parravicini E. et al.‘De novo’ development of esophageal varices in patients with a recent histologic diagnosis of liver cirrhosis (abstract).Gastroenterology. 1998; 114: 1324AGoogle Scholar]. One major exception is the study by Calès et al. [[17]Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar], who, in 41 patients of cirrhotics of alcoholic aetiology with a relatively large proportion of Child B and C patients, observed 1 and 2 year incidences of varices of 23% and 50%, respectively. A possible explanation of this discrepancy is that in the other studies the proportions of alcoholic cirrhosis and of patients with advanced liver disease were much lower than in the Calès study. The fact that Merli et al. [[14]Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar] were unable to identify any predictor of the ‘de novo’ appearance of varices, while disparate prognostic factors were identified in the other studies, suggests that no strong predictor exists. Concerning the progression of varices from small to large, the rates reported by other authors vary greatly: 25% over 6 years by Pagliaro et al. [[8]Pagliaro L. D'Amico G. Pasta L. Politi F. Vizzini G. Traina M. et al.Portal hypertension in cirrhosis: natural history.in: Bosch J. Groszmann R.J. Portal hypertension, pathophysiology and treatment. Blackwell Scientific, Oxford1994: 72-92Google Scholar], 45% at 3 years by Zoli et al. [[18]Zoli M. Merkel C. Magalotti D. Gueli C. Grimaldi M. Gatta A. et al.Natural history of cirrhotic patients with small esophageal varices: a prospective study.Am J Gastroenterol. 2000; 95: 503-508Crossref PubMed Google Scholar], 70% at 2 years by Calès et al. [[17]Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar], 30% at 2 years in another study by the same author [[19]Calès P. Oberti F. Payen J.L. Naveau S. Guyader D. Blanc P. et al.Lack of effect of propranolol in the prevention of large esophageal varices in patients with cirrhosis.Eur J Gastroenterol Hepatol. 1999; 11: 741-745Crossref PubMed Scopus (120) Google Scholar], and 15% at 16 months in the paper by Merkel et al. [[20]Merkel C. Angeli P. Marin R. Zanella P. Felders M. Bernardinello E. et al.Beta-blockers in the prevention of aggravation of esophageal varices in patients with cirrhosis and small esophageal varices: interim analysis of a controlled clinical trial (abstract).Hepatology. 1998; 28: 453AGoogle Scholar]. This large variability of results is probably related to differences in the populations studied and to the different endoscopic follow-up intervals used in the studies. As far as predictors of variceal progression are concerned, the only factors identified in more than one study are the Child–Pugh class at baseline [14Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar, 17Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar], or its worsening during follow-up [17Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar, 18Zoli M. Merkel C. Magalotti D. Gueli C. Grimaldi M. Gatta A. et al.Natural history of cirrhotic patients with small esophageal varices: a prospective study.Am J Gastroenterol. 2000; 95: 503-508Crossref PubMed Google Scholar]. The observation by Merli et al. [[14]Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar] that the rate of bleeding was significantly higher in patients with small varices at baseline than in those without is in keeping with previous data [[17]Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar]. How can we use the above information to decide the strategies to prevent variceal bleeding? Concerning the possibility of restricting screening endoscopy to the patients at the highest risk of having varices, the available data are disappointing, since no prognostic model to predict the presence of varices has been proven precise enough to select patients. Therefore, the recommendation [[6]de Franchis R. Updating consensus in portal hypertension: report of the Baveno III consensus workshop on definitions, methodology and therapeutic strategies in portal hypertension.J Hepatol. 2000; 33: 846-852Abstract Full Text Full Text PDF PubMed Scopus (485) Google Scholar] that all cirrhotic patients should be screened for the presence of varices at the time of the initial diagnosis of cirrhosis still stands. As far as deciding the timing of follow-up endoscopy is concerned, we must first decide which level of risk (i.e. what proportion of patients bleeding before starting prophylactic treatment) we are willing to accept. If we set this level at 10%, then patients with no varices at baseline can be re-endoscoped at 3 year intervals. In fact, the rate of development of large varices and the rate of bleeding in these patients are both below the 10% threshold [8Pagliaro L. D'Amico G. Pasta L. Politi F. Vizzini G. Traina M. et al.Portal hypertension in cirrhosis: natural history.in: Bosch J. Groszmann R.J. Portal hypertension, pathophysiology and treatment. Blackwell Scientific, Oxford1994: 72-92Google Scholar, 14Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar, 17Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar]. For patients with small varices at baseline, the picture is much more complicated, since the reported incidences range between 6% [[8]Pagliaro L. D'Amico G. Pasta L. Politi F. Vizzini G. Traina M. et al.Portal hypertension in cirrhosis: natural history.in: Bosch J. Groszmann R.J. Portal hypertension, pathophysiology and treatment. Blackwell Scientific, Oxford1994: 72-92Google Scholar] and 70% [[17]Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar] at 2 years for large varices and between 4% at 16 months [[20]Merkel C. Angeli P. Marin R. Zanella P. Felders M. Bernardinello E. et al.Beta-blockers in the prevention of aggravation of esophageal varices in patients with cirrhosis and small esophageal varices: interim analysis of a controlled clinical trial (abstract).Hepatology. 1998; 28: 453AGoogle Scholar] and 24% at 2 years [[17]Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar] for bleeding. Clearly, more information is needed on this point. For the time being, it appears reasonable to maintain the recommended interval of 1–2 years [[6]de Franchis R. Updating consensus in portal hypertension: report of the Baveno III consensus workshop on definitions, methodology and therapeutic strategies in portal hypertension.J Hepatol. 2000; 33: 846-852Abstract Full Text Full Text PDF PubMed Scopus (485) Google Scholar], adopting the shorter interval for patients with alcoholic cirrhosis [14Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar, 17Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar], with more severe impairment of liver function [8Pagliaro L. D'Amico G. Pasta L. Politi F. Vizzini G. Traina M. et al.Portal hypertension in cirrhosis: natural history.in: Bosch J. Groszmann R.J. Portal hypertension, pathophysiology and treatment. Blackwell Scientific, Oxford1994: 72-92Google Scholar, 14Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar, 17Calès P. Desmorat H. Vinel J.P. Caucanas J.P. Ravaud A. Gerin P. et al.Incidence of large oesophageal varices in patients with cirrhosis: application to prophylaxis of first bleeding.Gut. 1990; 31: 1298-1302Crossref PubMed Scopus (173) Google Scholar] and with endoscopic risk signs [5North-Italian Endoscopic Club for the Study and Treatment of Esophageal VaricesPrediction of the first variceal hemorrhage in patients with cirrhosis of the liver and esophageal varices.N Engl J Med. 1988; 319: 983-989Crossref PubMed Scopus (1060) Google Scholar, 14Merli M. Nicolini G. Angeloni S. Rinaldi V. De Santis A. Merkel C. et al.Incidence and natural history of small esophageal varices in cirrhotic patients.J Hepatol. 2003; 38: 266-272Abstract Full Text Full Text PDF PubMed Scopus (377) Google Scholar].