Urine Osmolality Is a Potential Marker of Longer-Term Efficacy of Tolvaptan in Autosomal Dominant Polycystic Kidney Disease: A Post Hoc Analysis

Abstract

Key Points:

1. Post hoc analyses of the TEMPO 3:4 trial suggest that short-term reductions in urine osmolality with tolvaptan predict effects on total kidney volume and eGFR.

2. Change in urine osmolality has potential as a biomarker of treatment response and may facilitate trial design and clinical decision making.

Background: Total kidney volume (TKV) and eGFR are measures of progression and treatment response in autosomal dominant polycystic kidney disease, but utility is limited by the long follow-up required for change assessment. In an analysis of data from the 3-year TEMPO 3:4 trial, we evaluated relationships among a short-term indicator of drug activity (change in urine osmolality [Uosm]) and longer-term outcomes to evaluate Uosm as a potential marker of efficacy.

Methods: Linear regression modeling and single-point analyses assessed relationships among change in Uosm to week 3, change in TKV to month 12, and change in eGFR to month 36 in participants treated with tolvaptan (n=961) or placebo (n=483). Multivariate models evaluated the proportion of the tolvaptan treatment effect on eGFR attributable to change in Uosm.

Results: Change in TKV to month 12 and Uosm to week 3 each correlated with change in eGFR to month 36, regardless of treatment assignment. A greater decrease in Uosm from baseline to week 3 was indicative of a slower decrease in eGFR to month 36 (slope estimate of −0.01, P < 0.00001). The effect of tolvaptan on Uosm accounted for 68.8% of the treatment effect on change in eGFR to month 36. Simulations of TEMPO 3:4 under the null hypothesis (i.e., replacement of all values for change in Uosm from baseline to week 3 with values from the placebo arm only) yielded a type 1 error rate indicating an acceptable risk of falsely concluding treatment efficacy on the basis of change in Uosm as a trial end point.

Conclusions: Change in Uosm is a potential biomarker for long-term treatment outcome with tolvaptan and might expedite clinical trials and treatment decision making for drugs with similar mechanisms of action.

Trial registration: ClinicalTrials.gov NCT00428948.

Graphical abstract

Figure 1

Relationship between percent change in TKV from baseline to month 12 and percent change in eGFR from on-treatment baseline (week 3) to month 36 in TEMPO 3:4. The quartiles are overlaid to represent trends in the raw data. The Pearson correlation for percent change in TKV to month 12 and percent change in eGFR to month 36 is −0.20. TKV, total kidney volume.

Figure 2

Relationship between change in spot Uosm from baseline to week 3 and percent change in eGFR from on-treatment baseline (week 3) to month 36 in TEMPO 3:4. The quartiles are overlaid to represent trends in the raw data. The Pearson correlation for change in Uosm to week 3 and percent change in eGFR to month 36 is 0.46. EOT, end of titration; Uosm, urine osmolality.