Renoprotection by Long-Term Low-Dose Tolvaptan in Patients with Heart Failure and Hyponatremia

Figures

• 

  Figure 1. Flow diagram of the study.

• 

  Figure 2. (A) Between‐group comparison of the daily dose‐to‐baseline dose ratio for loop diuretics to month 9. The analysis was performed in the propensity score‐matched cohort. Daily dose‐to‐baseline dose ratios for loop diuretics at 3, 6, and 9 months are shown in each treatment group. The analysis was restricted to subjects receiving any loop diuretics at baseline and those followed up for ≥9 months. (B) Incidence rate ratio of a loop diuretic dose reduction in TLV users compared with never‐users stratified by baseline serum Na of 135 mEq/L. Error bars represent 95% confidence intervals. ‘T’ and ‘N’ represent TLV users and never‐users, respectively. The analysis was performed in the propensity score‐matched cohort. The loop diuretic dose‐sparing effect of TLV therapy was pronounced in patients with hyponatremia.

• 

  Figure 3. Between‐group comparison of eGFR over time in the propensity score‐matched cohort. Error bars represent 95% confidence intervals. Estimated GFR (eGFR) over time was compared between the treatment groups by a mixed effects model with time‐dependent eGFR as a dependent variable. (A) TLV users had a significantly higher eGFR trajectory than never‐users. (B) However, this difference was extinguished after additional adjustment for the number of loop diuretic dose reductions during the follow‐up period.

• 

  Figure 4. Estimated GFR over time stratified by baseline serum Na (cut‐off at 135 mEq/L) in the propensity score‐matched cohort. Error bars represent 95% confidence intervals. The renoprotective effect of TLV therapy in terms of eGFR trajectories was pronounced in patients with hyponatremia.

• 

  Figure 5. Renoprotection of TLV therapy in terms of annualized eGFR slope was dependent on the baseline serum Na levels. Shaded area represents 95% confidence intervals. The y‐axis represents the difference in annualized eGFR slopes between the treatment groups (∆eGFR slope = [annualized eGFR slope in TLV users] − [annualized eGFR slope in never‐users]). The x‐axis represents baseline serum Na levels. (A) A multivariable fractional polynomial interaction analysis was performed in the propensity score‐matched participants. (B) The analysis was restricted to participants who did not undergo a loop diuretic dose reduction during the follow‐up period.