• Potential conflict of interest: Nothing to report.


Thank you for responding to our study,1 which is in line with all other trials of primary prophylaxis for the prevention of variceal bleeding in selection of patients with grade II or larger varices.2 Indeed, we recommend caution in performing band ligation in patients with smaller varices, as highlighted by a recent study.3 This demonstrated a high incidence of banding-related bleeding when patients with smaller varices were included. In our study, none of the patients had red signs at endoscopy, but it is not clear whether patients without red signs should be excluded from band ligation. A recent study comparing propranolol with variceal band ligation for primary prevention reported fatalities due to banding-related ulceration despite all patients having red signs.4 Banding-related bleeding occurred at the second banding session in all patients in our trial, and in these patients there was no difference in the banding schedule compared with other patients. We believe variceal band ligation remains an option for primary prophylaxis, but careful selection of patients with high-risk varices is essential.

We observed a low eradication rate, and we would respond by saying that our trial reflects what is likely to be the real-life situation in a patient with predominantly alcoholic liver disease. Compliance was an issue in our trial, in particular with patients in the banding arm. We performed a separate per-protocol analysis where the eradication rate was much higher at 73%, and did not demonstrate differences in the outcomes. We highlight the importance of eradication of varices and adhering to banding schedules. Drug therapy clearly has advantages here, because the protective effect is immediate and carvedilol therapy is particularly attractive because there is no need to titrate with the resting pulse, dosing is once daily, and the side effect profile appears to be better than with propranolol. Comparing carvedilol only with patients who have had varices eradicated would seem ideal, but in practice in the vast majority of compliant patients eradication of varices is possible.

We observed secondary gastric varices in 19% of patients in the banding arm, and none of these patients bled from gastric varices.1 Patients in the carvedilol arm did not undergo routine endoscopies after randomization. Clearly, there is a risk of the development of gastric varices with variceal band ligation, but at least in our follow-up period there did not seem to be clinical complications. A longer follow-up and a larger number of patients would be required to study this group of patients.

Dhiraj Tripathi M.D., M.R.C.P.*, Peter C. Hayes M.D., Ph.D., F.R.C.P.†, * Liver Unit, Queen Elizabeth Hospital, Edgbaston, Birmingham, UK, † Department of Hepatology, Royal Infirmary of Edinburgh, Edinburgh, UK.