Initial eGFR Changes Predict Response to Tolvaptan in ADPKD

Visual Abstract

Abstract

Key Points

• This post hoc analysis of the Tolvaptan Efficacy and Safety in Management of Autosomal Dominant Polycystic Kidney Disease and Its Outcomes 3:4 study investigated the long-term predictive potential of initial changes in eGFR.

• Initial eGFR change from baseline to week 3 proved to be a significant and independent indicator of the long-term effects of tolvaptan.

• No correlation was found between the initial change in eGFR and the annual rate of percent growth in total kidney volume.

Background

Tolvaptan, the only pharmaceutical treatment available for autosomal dominant polycystic kidney disease (ADPKD), reduced the rates of total kidney volume (TKV) increase and kidney function decline in patients with ADPKD in the global phase 3 Tolvaptan Efficacy and Safety in Management of Autosomal Dominant Polycystic Kidney Disease and Its Outcomes (TEMPO) 3:4 study. Since tolvaptan initiation is associated with an initial decline in the eGFR, this post hoc analysis of the TEMPO 3:4 study investigated whether initial changes in eGFR from baseline to week 3 after tolvaptan administration can predict its longer-term effects on eGFR and TKV in patients with ADPKD.

Methods

eGFR was estimated using the CKD Epidemiology Collaboration equation at baseline and up to month 36. TKV was estimated using standardized kidney magnetic resonance imaging at baseline and after 12, 24, and 36 months of tolvaptan treatment. The effect of tolvaptan on kidney function and kidney volume was evaluated by measuring changes in eGFR from week 3 and TKV from baseline up to 36 months. All 961 patients randomized to receive tolvaptan in TEMPO 3:4 were included in this analysis.

Results

Initial change in eGFR from baseline to week 3 was a significant and independent predictor of the mean rate of change in eGFR per year. By contrast, there was no association between initial change in eGFR and the rate of percent growth in TKV per year.

Conclusions

Changes in eGFR after 3 weeks of treatment are likely due to the pharmacologic effect of tolvaptan, and these initial changes are predictive of the long-term effects of tolvaptan treatment.

Introduction

Autosomal dominant polycystic kidney disease (ADPKD) is an inherited condition characterized by the development of numerous fluid-filled cysts that cause progressive enlargement of the kidney.¹ ADPKD is associated with a progressive decline in kidney function and leads to ESKD in many individuals.² Intracellular cAMP levels are elevated in ADPKD, which play a vital role in cystogenesis.³ These elevated levels of cAMP are mainly due to mutations in the PKD1 or PKD2 genes, which encode integral membrane proteins, and whose mutations alter intracellular calcium homeostasis.⁴ Vasopressin, via its action on the V2 receptors, is a powerful agonist for cAMP generation.^4,5 V2 receptors may also be involved in the development of the progressive tubulointerstitial fibrosis that accompanies the expansion of cysts in patients with ADPKD, via interaction with yes-associated protein and connective tissue growth factor.⁶

Tolvaptan, a vasopressin V2 receptor antagonist,⁴ is the only pharmaceutical treatment available for ADPKD and is widely prescribed globally.⁷ The global phase 3 Tolvaptan Efficacy and Safety in Management of Autosomal Dominant Polycystic Kidney Disease and Its Outcomes (TEMPO) 3:4 study demonstrated the efficacy of tolvaptan in patients with ADPKD, with an inhibitory effect on kidney volume increase (49% reduction in rate of change of total kidney volume [TKV] at the end of 3 years [P < 0.001]) and a suppressive effect on the decrease in kidney function (26% reduction in rate of kidney function decline [P < 0.001]) observed with treatment.⁸ However, in TEMPO 3:4, tolvaptan was associated with a high incidence of adverse events, including events related to aquaresis and hepatic adverse events not related to ADPKD.⁸ The balance of efficacy and tolerability must be considered for each patient with ADPKD in the early stage of tolvaptan treatment, but there is limited information to allow clinicians to predict which patients are most likely to benefit from tolvaptan.^9–11

Several post hoc analyses of the TEMPO 3:4 study revealed that the greatest response to tolvaptan occurred in patients with ADPKD who had greater suppression of urine osmolality (uOsm) (i.e., patients who had better eGFR at baseline)¹⁰ and higher percentage increase in copeptin from baseline to 3 weeks.¹¹

Recent data from small, single-center studies in Japan have shown that the total tolvaptan dose significantly affects the change in eGFR¹² and that initial eGFR changes correlate with certain patient outcomes.⁹ A greater initial eGFR decline with tolvaptan treatment was associated with annual eGFR change and better kidney prognosis; in addition, the fractional excretion of urea nitrogen and free water both increased significantly after tolvaptan treatment.⁹

The current post hoc analysis of the TEMPO 3:4 study investigated whether initial changes in eGFR observed between baseline and week 3 after tolvaptan administration can predict the longer-term effects of tolvaptan on eGFR and TKV in patients with ADPKD.

Methods

Study Design

This was a post hoc exploratory analysis of tolvaptan recipients from the TEMPO 3:4 study (NCT00428948), the details of which have been published previously.⁸ In brief, TEMPO 3:4 was a phase 3, multicenter, double-blind, placebo-controlled 3-year study, conducted to assess the safety and efficacy of tolvaptan in patients with ADPKD. Patients aged 18–50 years with a diagnosis of ADPKD were included and were required to have a TKV of ≥750 ml and creatinine clearance of ≥60 ml/min. Patients were randomized (2:1) to tolvaptan or matching placebo. Tolvaptan was administered twice a day for 36 months (at the highest tolerated dose), following a 3-week titration period. During the first week of the titration period, patients in the active arm received tolvaptan 45 mg in the morning and 15 mg in the afternoon. The dose was then increased to 60/30 mg and 90/30 mg each week as tolerated.

The study was conducted in accordance with the principles of the Declaration of Helsinki and the International Conference of Harmonisation Good Clinical Practice Guidelines, with an institutional review board or ethics committee at each site approving the study protocol before initiation. All participants provided written informed consent.

Assessments

GFR was estimated using the creatinine-based CKD Epidemiology Collaboration equation^13,14 and determined at baseline, at randomization (week 3), weekly during the dose escalation phase, and every 4 months during treatment up to month 36 after tolvaptan initiation. Reciprocal creatinine (which is independent of spurious changes in body weight) was more likely to represent an accurate assessment of change in kidney function. The formula used to calculate reciprocal serum creatinine was 1/Pcr, where Pcr=serum creatinine concentration (mg/dl). TKV was assessed using standardized kidney magnetic resonance imaging at baseline and at months 12, 24, and 36 after the initiation of tolvaptan.¹⁵ BUN and fasting uOsm were assessed at baseline.

Outcomes

The effects of tolvaptan on kidney function and kidney volume were evaluated using the eGFR mean rate of change calculated from week 3 of tolvaptan administration up to 36 months and the TKV (left and right kidneys) rate of percent growth calculated from baseline up to 36 months.

Statistical Methods

Descriptive statistics were used in this analysis, with means and SD used for continuous variables and number of patients and percentages used for categorical variables. To determine correlation between changes in TKV and eGFR and initial change in eGFR, both the Spearman and Pearson correlation coefficients, with 95% confidence intervals (CIs), were calculated.

Univariate regression models were used to determine the association between sex, age, baseline eGFR, systolic BP (SBP), body mass index (BMI), TKV, BUN, and fasting uOsm and initial changes in eGFR from baseline to week 3 after tolvaptan administration (independent/explanatory variables) and the eGFR mean rate of change per year and the rate of percent growth in TKV per year (dependent/objective variables), respectively. Multivariate regression models were used to determine the association between initial changes in eGFR after tolvaptan administration and the two dependent variables (mean rate of change per year in eGFR and rate of percent growth in TKV) after adjusting for other independent variables. Univariate and multivariate regression models for predicting eGFR at week 3 were also performed using the abovementioned independent variables. Owing to its distribution, the rate of percent growth in TKV per year was natural logarithm-transformed in analyses.

All statistical analyses were performed using SAS version 9.4. A P value of <0.05 was considered statistically significant.

Results

Patients

In TEMPO 3:4, patients were enrolled at 129 sites worldwide between January 1, 2007, and January 1, 2009. Overall, 1445 patients were randomized, 961 to tolvaptan and 484 to matching placebo.

The demographic characteristics of the TEMPO 3:4 ADPKD patient population have been reported previously.⁸ In brief, the mean age was 38.56 years and 51.51% of the patients were male (Table 1). The mean BMI was 26.25 kg/m², and the mean SBP was 128.64 mm Hg. At baseline, the mean eGFR was 81.35 ml/min per 1.73 m², with an eGFR of ≥90 ml/min per 1.73 m² in 34.45% of patients, and 60 to <90 ml/min per 1.73 m² in 48.54% of patients. The mean TKV was 1704.82 ml, with a volume of <1500 ml in 51.93% of patients and ≥1500 ml in 48.07% of patients.

Table 1.

Demographics and baseline characteristics of patients receiving tolvaptan in the Tolvaptan Efficacy and Safety in Management of Autosomal Dominant Polycystic Kidney Disease and Its Outcomes 3:4 study

Characteristic	Tolvaptan, N=961
Sex, n (%)
Male	495 (51.51)
Female	466 (48.49)
Age, yr, mean±SD	38.56±7.10
Age category, yr, n (%)
<35	256 (26.64)
≥35	705 (73.36)
eGFR, ml/min per 1.73 m 2 , mean±SD; N=958	81.35±21.02
eGFR category, ml/min per 1.73 m2, n (%)
30 to <45	28 (2.92)
45 to <60	135 (14.09)
60 to <90	465 (48.54)
≥90	330 (34.45)
SBP, mm Hg, mean±SD	128.64±13.52
SBP category, mm Hg, n (%)
<100	11 (1.14)
100 to <140	744 (77.42)
≥140	206 (21.44)
BMI, kg/m 2 , mean±SD; N=960	26.25±5.08
BMI category, kg/m2, n (%)
<25	442 (46.04)
25 to <30	336 (35.00)
≥30	182 (18.96)
TKV, ml, mean±SD	1704.82±921.27
TKV category, ml, n (%)
<1500	499 (51.93)
≥1500	462 (48.07)

BMI, body mass index; SBP, systolic BP; TKV, total kidney volume.

Predictive Value of Initial Change in eGFR

The mean±SD eGFR at 3 weeks after starting tolvaptan treatment was 76.58±21.13 ml/min per 1.73 m², with a mean decrease from baseline of 4.42±8.81 ml/min per 1.73 m² (Figure 1). This initial change in eGFR, at 3 weeks' post starting tolvaptan treatment, was significantly associated with the mean rate of change per year in eGFR (Figure 2); the larger the initial change in eGFR, the smaller the mean rate of change per year (Pearson correlation coefficient=−0.16 [95% CI, −0.23 to −0.10]; Spearman correlation coefficient=−0.18 [95% CI, −0.24 to −0.11]). On univariate regression analysis, sex, age, and baseline values of eGFR, SBP, BMI, TKV, and BUN were not significantly associated with mean rate of change in eGFR per year, but initial change in eGFR from baseline to week 3 (P < 0.0001) and higher baseline fasting uOsm (P = 0.0146) were significantly associated with it (Table 2). On multivariate regression analysis, initial change in eGFR from baseline to week 3 and fasting uOsm at baseline maintained a significant association with the mean rate of change in eGFR per year (P = 0.0002 and 0.0135, respectively); the association between BUN and the mean rate of change in eGFR per year became significant after adjusting for other factors (P = 0.0399) (Table 2).

Figure 1.

Box plot for the initial change in eGFR from baseline to week 3. The box marks the range between the upper and lower quartiles (interquartile range; Q1–Q3), the central black horizontal line indicates the median value, the triangle indicates the mean value, and the upper and lower whiskers represent the minimum/maximum values. The line between the two triangles indicates the decrease in eGFR from baseline to week 3.

Figure 2.

Association between mean rate of change in eGFR per year and initial change in eGFR from baseline to week 3. The 95% CIs for Spearman and Pearson correlation coefficients are given. CI, confidence interval; ρ_p, Pearson correlation coefficient; ρ_s, Spearman correlation coefficient.

Table 2.

Univariate and multivariate regression models for predicting eGFR mean rate of change per year^a

Regression Models and Variables	Estimate	SEM	P Value	Partial Correlation Coefficient
Univariate analysis
eGFR change from baseline at week 3, ml/min per 1.73 m2	−0.1117	0.0233	<0.0001
Sex (reference=female)	0.1430	0.4118	0.7285
Age, yr	0.0027	0.0295	0.9280
eGFR at baseline, ml/min per 1.73 m2	0.0181	0.0099	0.0675
SBP at baseline, mm Hg	0.0002	0.0152	0.9874
BMI at baseline, kg/m2	−0.0133	0.0416	0.7485
TKV at baseline, ml	−0.0003	0.0002	0.1276
BUN at baseline, mg/dl	−0.0436	0.0381	0.2530
Fasting uOsm at baseline, mOsm/kg	0.0030	0.0012	0.0146
Multivariate analysis
eGFR change from baseline at week 3, ml/min per 1.73 m2	−0.0976	0.0259	0.0002	−0.1513
Sex (reference=female)	0.4129	0.4756	0.3857	0.0352
Age, yr	0.0188	0.0363	0.6058	0.0210
eGFR at baseline, ml/min per 1.73 m2	−0.0164	0.0154	0.2878	−0.0431
SBP at baseline, mm Hg	−0.0076	0.0172	0.6588	−0.0179
BMI at baseline, kg/m2	−0.0079	0.0460	0.8639	−0.0070
TKV at baseline, ml	−0.0002	0.0003	0.4169	−0.0330
BUN at baseline, mg/dl	−0.1064	0.0517	0.0399	−0.0833
Fasting uOsm at baseline, mOsm/kg	0.0035	0.0014	0.0135	0.1000

BMI, body mass index; SBP, systolic BP; TKV, total kidney volume; uOsm, urine osmolality.

^aeGFR mean rate of change calculated for each subject using eGFR values measured from week 3 up to month 36.

By contrast, there was no association between initial change in eGFR and the rate of percent growth in TKV per year (Pearson correlation coefficient=0.01 [95% CI, −0.06 to 0.07]; Spearman correlation coefficient=0.07 [95% CI, −0.00 to 0.14]; Figure 3). On univariate regression analysis, male sex, lower age, higher baseline BMI, and higher baseline TKV all showed a significant association with greater percent growth in TKV per year (Table 3). All these factors were significantly associated with TKV rate of change on multivariate analysis (Table 3).

Figure 3.

Association between percentage change in TKV from baseline to year 3 and initial change in eGFR from baseline to week 3. The 95% CIs for Spearman and Pearson correlation coefficients are given. TKV, total kidney volume.

Table 3.

Univariate and multivariate regression models for predicting rate of percent growth in total kidney volume per year^a

Regression Models and Variables	Estimate	SEM	P Value	Partial Correlation Coefficient
Univariate analysis
eGFR change from baseline to week 3, ml/min per 1.73 m2	0.0000	0.0002	0.8767
Sex (reference=female)	0.0314	0.0038	<0.0001
Age, yr	−0.0016	0.0003	<0.0001
eGFR at baseline, ml/min per 1.73 m2	−0.0001	0.0001	0.4393
SBP at baseline, mm Hg	0.0003	0.0001	0.0739
BMI at baseline, kg/m2	0.0024	0.0004	<0.0001
TKV at baseline, ml	0.0000	0.0000	<0.0001
BUN at baseline, mg/dl	0.0005	0.0004	0.1875
Fasting uOsm at baseline, mOsm/kg	−0.0000	0.0000	0.3745
Multivariate analysis
eGFR change from baseline at week 3, ml/min per 1.73 m2	−0.0002	0.0003	0.3784	−0.0366
Sex (male=1, female=2)	0.0203	0.0047	<0.0001	0.1756
Age, yr	−0.0023	0.0004	<0.0001	−0.2554
eGFR at baseline, ml/min per 1.73 m2	−0.0002	0.0002	0.1683	−0.0572
SBP at baseline, mm Hg	−0.0000	0.0002	0.9546	−0.0024
BMI at baseline, kg/m2	0.0020	0.0005	<0.0001	0.1787
TKV at baseline, ml	0.0000	0.0000	0.0003	0.1503
BUN at baseline, mg/dl	−0.0006	0.0005	0.2075	−0.0523
Fasting uOsm at baseline, mOsm/kg	−0.0000	0.0000	0.2952	−0.0435

BMI, body mass index; SBP, systolic BP; TKV, total kidney volume; uOsm, urine osmolality.

^aTotal kidney volume rate of percent growth calculated for each subject using total kidney volume values measured from baseline up to month 36.

Univariate and multivariate regression models for predicting eGFR at week 3 identified the baseline characteristics of age, eGFR, BMI, TKV, BUN, and fasting uOsm (univariate analysis) and age, eGFR, TKV, and BUN (multivariate analysis) as significant factors (Supplemental Table 1).

Discussion

In this post hoc analysis of the TEMPO 3:4 study, the initial decrease in eGFR from baseline to week 3 after tolvaptan administration was found to predict kidney prognosis in ADPKD. A greater initial decline in eGFR suggests a more pronounced response to tolvaptan treatment, which subsequently protects the kidney from a further decline in eGFR.⁹ This initial decline in eGFR may be due to reduction of glomerular hyperfiltration by tolvaptan through the suppression of urea recycling in the collecting tubules and/or reduction of kidney plasma flow by potential volume depletion.⁹ A similar initial decline in eGFR has been observed with renin–angiotensin–aldosterone system inhibitors¹⁶ and sodium–glucose cotransporter 2 inhibitors,¹⁷ resulting in reduced urinary protein and protecting kidney function. While renin–angiotensin–aldosterone system inhibitors expand the efferent arterioles to reduce the intraglomerular pressure and suppress glomerular hyperfiltration,¹⁶ sodium–glucose cotransporter 2 inhibitors do so via afferent arteriole vasoconstriction.¹⁷

While previous studies have demonstrated that changes in uOsm¹⁰ and in copeptin at 1 month after tolvaptan administration,¹¹ as well as the initial decrease in urinary aquaporin 2¹⁸ are also predictors of the efficacy of tolvaptan in ADPKD, these markers are not easily used in clinical practice. Predictors of tolvaptan response are needed that can be easily measured in routine clinical practice.

Both GFR and TKV are considered markers of progression in ADPKD.^19,20 There is evidence from previous studies of ADPKD that changes in TKV and in eGFR do not occur at the same rate, with changes in TKV occurring before and predictive of changes in eGFR.^19,20 However, a previous post hoc analysis of data from the TEMPO 3:4 study did not find any association between the effects of tolvaptan in slowing kidney function decline, as measured by eGFR, and changes in TKV growth rate.²¹ This is consistent with the findings of this study, in which initial changes in eGFR were predictive of improved kidney function but not of changes in TKV. Therefore, our findings suggest that tolvaptan can prevent the decline in kidney function that occurs in ADPKD, even in patients for whom tolvaptan does not reduce the rate of TKV increase. There was no significant difference in TKV at 36 months, possibly because the reduction in TKV is greater during the first year than during the second or third years.⁸

Patient characteristics, such as sex, age, and baseline values of BMI and TKV, were predictive of the effect of tolvaptan on changes in TKV, and these results were consistent with previous reports.^8,21,22 In the TEMPO 3:4 study, being overweight or being obese were strongly associated with increased kidney growth, but not in eGFR decline; however, tolvaptan was found to be effective regardless of baseline BMI.²² This is also consistent with the findings of this study.

Our findings confirm those of smaller studies in Japan, which also demonstrated that eGFR changes during the early phase of tolvaptan administration were able to predict the longer-term effect of tolvaptan.⁹ In addition, the prognosis was better when uOsm was higher and BUN was lower at baseline. This is most likely because high uOsm indicates that the concentrating ability of the kidneys is maintained or water intake is insufficient, and low BUN suggests the preservation of renal function.

This post hoc analysis has some limitations. First, the TEMPO 3:4 study only included a few patients with more advanced kidney dysfunction (eGFR of 30–45 ml/min per 1.73 m²), with most patients having a creatinine clearance of at least 60 ml/min (eGFR of >45 ml/min per 1.73 m²); thus, these findings may not hold true in patients with more advanced kidney dysfunction. In addition, for this analysis, the CKD Epidemiology Collaboration equation was used instead of the non–race-based eGFR equation, which could limit interpretation of the results. Furthermore, we recommend the use of the non–race-based eGFR equation for future studies.

In conclusion, this post hoc analysis of the TEMPO 3:4 study demonstrated that initial changes in eGFR 3 weeks after initiation of tolvaptan treatment, which can be easily measured in clinical practice, are predictive of the long-term effects of tolvaptan.

Supplementary Material

kidney360-5-522-s001.pdf ^{(125.6KB, pdf)}

Disclosures

H. Jiang reports the following: Employer: Otsuka Pharmaceutical Development & Commercialization, Inc. M. Matsukawa reports the following: Employer: Otsuka Pharmaceutical Co., Ltd. T. Mochizuki reports the following: Consultancy: Otsuka Pharmaceutical Co., Ltd.; Research Funding: Chugai Pharmaceutical Co., JMS Co., Kyowa Kirin Co., and Otsuka Pharmaceutical Co., Ltd.; and Honoraria: Otsuka Pharmaceutical Co., Ltd. T. Tanaka reports the following: Employer: Otsuka Pharmaceutical Co., Ltd.

Funding

The TEMPO 3:4 study, this post-hoc analysis, development of the manuscript and the open access publishing fee was supported by Otsuka Pharmaceutical Co., Ltd., Japan.

Author Contributions

Conceptualization: Huan Jiang, Miyuki Matsukawa, Toshio Mochizuki, Toshiki Tanaka.

Formal analysis: Huan Jiang, Miyuki Matsukawa, Toshio Mochizuki, Toshiki Tanaka.

Methodology: Huan Jiang, Miyuki Matsukawa, Toshio Mochizuki, Toshiki Tanaka.

Writing – original draft: Huan Jiang, Miyuki Matsukawa, Toshio Mochizuki, Toshiki Tanaka.

Writing – review & editing: Huan Jiang, Miyuki Matsukawa, Toshio Mochizuki, Toshiki Tanaka.

Data Sharing Statement

Qualified researchers can submit inquiries related to Otsuka clinical research or request access to individual participant data associated with any Otsuka clinical trial at https://clinical-trials.otsuka.com/. For all approved individual participant data access re-quests, Otsuka will share anonymized individual participant data on a remotely accessible data sharing platform.

Supplemental Material

This article contains the following supplemental material online at http://links.lww.com/KN9/A450.

Supplemental Table 1. Univariate and multivariate regression models for predicting eGFR at week 3.