• Conflict of interest: Nothing to report.


We are grateful to Dr. Ozdogan for his interest in our article aimed to investigate the possible influence of emergency endoscopic treatment on HVPG in patients with cirrhosis and with bleeding esophageal varices (BEV). We completely agree with him that the investigators should stop a trial if during the study period a consensus has been reached regarding the treatment modalities under evaluation. However, this does not appear to be the issue in our study. The author refers to a single topic symposium on portal hypertension and variceal bleeding reported in HEPATOLOGY in 1998. At that period there was consensus suggesting only that endoscopic measures are the first choice of treatment for the control of BEV. It is true that the metaanalyses of D'Amico et al. (1999)1 and de Franchis and Primignani (1999)2 favored the early administration of a vasoactive drug when variceal bleeding is suspected. However, their data were based only on two double blind controlled trials one of which suffered from difficulties in interpretation since the placebo group contained patients with more severe liver disease at randomization versus the treated group.2, 3 Later on, the question on the optimal role of vasoactive drugs was again brought out by an excellent review from Garcia-Tsao.4 The author recommends endoscopic therapy as the therapy of choice for the management of acute variceal bleeding since it stops bleeding in 80%-90% of patients. The goal of vasoactive drugs would be the prevention of early rebleeding. Recently, Banares et al.5 showed the superiority of combination therapy in achieving the 5-day haemostasis in BEV without, however, any improvement in mortality. Furthermore, D'Amico et al. (2003)6 concluded that emergency sclerotherapy should not be performed as the first line treatment of BEV because vasoactive drugs achieve control of bleeding in 83% of patients and therefore, endoscopic therapy might be added only in pharmacological treatment failures. However, the definitions of the end points and the selection of trials included in the above metaanalysis have been seriously criticized.7 Hence it is obvious than even in 2003 more trials were required in order to determine further potential advantages of combined therapy in the management of these patients. Therefore, our study which was conducted (1998-2001) under the guidelines of good clinical practice is absolutely documented according to the recommendations for the management of patients with BEV. As we have shown, in BEV endoscopic therapy increases HVPG which is sustained after sclerotherapy, but not band ligation, and this resulted in a higher rebleeding rate. Taking into consideration (1) the above results; (2) the available data suggesting that vasoactive drugs may reduce HVPG8; and (3) the safety, efficacy and easy of administration of these drugs compared to endotherapy,6 we consider that the early administration of vasoactive drugs is mandatory in all patients with cirrhosis and with BEV. Therefore, in the forthcoming era of future trials, we believe that withholding vasoactive drugs in patients with BEV is not justified.

Alec Avgerinos* †, Jiannis Vlachogiannakos*, Spilios Manolakopoulos*, Sotirios A. Raptis†, * 2nd Department of Gastroenterology, Evangelismos Hospital, Athens, Greece, † 2nd Department of Medicine Propaedeutic and Research Unit, Attikon University Hospital, Chaidari, Greece.