We thank Dr. Krag and colleagues for their comments on our article on hyponatremia in patients treated with terlipressin for severe gastrointestinal bleeding due to portal hypertension.1 We believe that the most important pathogenic factor in the development of a decrease in serum sodium concentration in patients with gastrointestinal bleeding treated with terlipressin is a reduction in solute-free water clearance due to the effect of terlipressin on V2 receptors located in the collecting duct cells, as nicely demonstrated by the same authors in a recent study.2 Nevertheless, an increased natriuresis can also participate.
With respect to the duration of treatment with terlipressin, the 2010 Baveno Consensus Conference on Portal Hypertension indicates that pharmacological treatment of acute variceal bleeding should be maintained for 5 days.3 However, in response to the comment of Krag et al., suggesting a shorter duration of therapy (of only 2 days), we have reanalyzed our data. This new analysis shows that serum sodium concentration decreased from 134.9 ± 6.6 mEq/L before the start of therapy to 130 ± 7.1 mEq/L after 2 days of treatment (P < 0.0001). Most importantly, 58% of patients developed a decrease in serum sodium concentration of more than 5 mEq/L (in these patients serum sodium concentration decreased from 136.4 ± 6.2 mEq/L to 127.7 ± 6.9 mEq/L; P < 0.0001), with 16% of patients developing a decrease of more than 10 mEq/L (Fig. 1). Therefore, our results suggest that even with duration of terlipressin treatment of only 2 days, serum sodium concentration should be monitored to allow an early detection of hyponatremia. We also would like to emphasize that hypotonic fluids should be avoided in patients with gastrointestinal bleeding treated with terlipressin to reduce the occurrence of hypovolemic hyponatremia, particularly in patients with low MELD scores that are at greatest risk to develop this complication.