Contrasting acute and chronic effects of tolvaptan on serum osmolality in the EVEREST trial

Aims In the EVEREST trial, tolvaptan improved symptoms and body weight during hospitalization for heart failure (HF), but did not improve post-discharge cardiovascular outcomes. We hypothesized that this disconnect between the short- and long-term effects may be related to changes in serum osmolality. We describe the longitudinal profile of osmolality and its response to tolvaptan during and after hospitalization for HF. Methods and results EVEREST enrolled 4133 patients hospitalized for HF and reduced EF. Serum osmolality data were available in 3744 (91%). We assessed the effects of tolvaptan on serum osmolality and related these effects to in-hospital changes in body weight and physician-assessed symptoms. Calculated values of osmolality, determined from serum sodium, blood urea nitrogen, and glucose, were 3-4 mOsm/kg higher than concurrently measured serum osmolality at enrolment and discharge in both treatment arms. During hospitalization in the placebo group, serum osmolality slightly increased throughout hospitalization, whereas serum sodium decreased and blood urea nitrogen increased until discharge. Tolvaptan increased osmolality by hospital day 1, but this effect diminished by post-discharge week 4-8 and disappeared by post-discharge week 56. In-hospital changes in osmolality were poorly correlated with in-hospital changes in body weight and physician-assessed dyspnoea. Conclusion Tolvaptan increased serum osmolality during hospitalization for HF, a time frame when the drug also improved signs and symptoms of HF. However, this effect on osmolality declined in the early post-discharge period, when tolvaptan failed to influence clinical outcomes. Serum osmolality, which can be estimated based on readily available laboratory parameters, may be a marker or a target of response to tolvaptan.