Shota Muraki 1, Tomohiro Sasaki 1, Krishnakant Dasika 2, Rebecca Baillie 2, Christina Friedrich 2, Renee Myers 2, Vincent Hurez 2, Seongryul Kim 3, Shingo Uno 3, Hiroyuki Fujiki 4, Yukinobu Takeshita 4, Yosuke Kawai 1
1 Office of Clinical Pharmacology, Department of Biometrics, Headquarters of Clinical Development, Otsuka Pharmaceutical Co., Ltd. (, ), 2 Rosa & Co, LLC (, ), 3 Department of Clinical Science 1 Group 2, Headquarters of Clinical Development, Otsuka Pharmaceutical Co., Ltd. (, ), 4 Department of Renal and Cardiovascular Research, New Drug Research Division, Otsuka Pharmaceutical Co., Ltd. (, )
Objectives:
In fluid overloaded heart failure, relieving congestion while minimizing electrolyte disturbances is critical. Tolvaptan, a vasopressin V2 receptor antagonist, increases the excretion of free water and shows a characteristic pattern in which urine output rises on the first day and then gradually decreases toward a steady state. This time dependent aquaretic pattern has not been well reproduced by simplified kidney-focused quantitative systems pharmacology (QSP) models. Our objective was to build and calibrate a more comprehensive mechanistic QSP model that describes the reported time course of urine output after tolvaptan initiation in patients with fluid overloaded heart failure.
Methods:
We expanded a kidney and circulation QSP platform [1] to better represent heart failure physiology. Regulation of sodium was represented separately in plasma and in interstitial fluid to reflect differences between vascular and interstitial spaces in edematous states. An osmotically inactive sodium storage pool was also added, following a reported modeling framework developed to account for high electrolyte free water clearance with sodium glucose cotransporter 2 (SGLT2) inhibitors by allowing nonosmotic sodium storage [2]. These structural elements were intended to explain how water loss can occur with little change in urinary sodium. In addition, we implemented an endogenous arginine vasopressin (AVP) mediated negative feedback in which small increases in plasma sodium (used here as a surrogate for osmolality) raise the apparent EC50 for V2 antagonism, gradually reducing the aquaretic effect of tolvaptan. Model calibration used published clinical time course data for urine output and serum sodium, and simulations evaluated movement of water and sodium between plasma and interstitial compartments and the influence of these processes on observed urine output trajectories.
Results:
Representing plasma and interstitial sodium regulation separately, together with the inactive sodium storage pool, led to a small increase in plasma sodium after tolvaptan administration that strengthened the gradient between plasma and interstitial compartments. Consequently, water moved from interstitial fluid to plasma and sodium moved from plasma to interstitial fluid, with excess interstitial sodium sequestered in the inactive pool. These processes allowed free water clearance of tolvaptan without substantial natriuresis or clinically meaningful hypernatremia. When this sodium regulation structure was combined with the AVP mediated feedback, the model described the day 1 peak in urine output, the subsequent attenuation, and stabilization by around day 7 after the start of tolvaptan treatment. The simulations also suggested a preferential reduction of interstitial fluid relative to blood volume following the start of tolvaptan, which is consistent with relief of congestion and edema through fluid redistribution.
Conclusions:
By incorporating separate regulation of sodium in plasma and interstitial fluid and AVP mediated feedback, the model described the characteristic time course of tolvaptan induced aquaresis in fluid overloaded heart failure. The model clarifies how substantial aquaresis can occur with minimal natriuresis and without excessive increases in plasma sodium, and it provides a physiologic basis for understanding fluid redistribution under V2 antagonism.
References:
[1] Friedrich C, Muraki S, Sasaki T, et al. Adaptation of a published kidney disease QSP model to represent autosomal-dominant polycystic kidney disease and evaluate treatment options. Poster presented at: American Conference of Pharmacometrics 2024; November 10–13, 2024; Phoenix, AZ, USA.
[2] 2) Hallow KM, Helmlinger G, Greasley PJ, McMurray JJV, Boulton DW. Why do SGLT2 inhibitors reduce heart failure hospitalization? A differential volume regulation hypothesis. Diabetes Obes Metab. 2018;20(3):479-487.
Reference: PAGE 34 (2026) Abstr 12104 [www.page-meeting.org/?abstract=12104]
Poster: Drug/Disease Modelling - Other Topics