original article - circulation: heart...

994

Renal dysfunction is often present in patients with heart failure and is related to a poorer prognosis.1,2 Patients

with chronic heart failure (CHF) are often treated with angio-tensin-converting enzyme inhibitors and mineralocorticoid receptor antagonists, although they often cause a further dete-rioration of renal function. Patients with acute heart failure (AHF) are often treated with loop diuretics, which might also further deteriorate renal function. To date, no drugs that are used for treating chronic or AHF have been shown to improve renal function. Serelaxin, a recombinant human relaxin-2, is a promising and novel investigational therapy for the treatment

of AHF. In the RELAXin in Acute Heart Failure (RELAX-AHF) study, early infusion of serelaxin was associated with a decrease in dyspnea and lower 180-day all-cause mortality in patients admitted for AHF.3 In addition, the study showed that treatment with serelaxin is associated with lower serum creati-nine and plasma cystatin C concentrations compared with pla-cebo.4 Other studies also indicate possible beneficial effects of serelaxin on renal function. A study investigating the renal effects of serelaxin in 11 human volunteers showed a marked increase of renal plasma flow (RPF) by 47% compared with baseline levels, but no significant change was observed in glo-merular filtration rate (GFR).5 In models of chronic kidney disease, relaxin showed protective effects on kidney structure

Original Article

© 2014 American Heart Association, Inc.

Circ Heart Fail is available at http://circheartfailure.ahajournals.org DOI: 10.1161/CIRCHEARTFAILURE.114.001536

Background—Serelaxin is a promising therapy for acute heart failure. The renal hemodynamic effects of serelaxin in patients with chronic heart failure are unknown.

Methods and Results—In this double-blind, randomized, placebo-controlled, multicenter study, patients with New York Heart Association Class II to III chronic heart failure, left ventricular ejection fraction ≤45%, and estimated glomerular filtration rate (GFR) 30 to 89 mL/min per 1.73 m2 received intravenous serelaxin 30 μg/kg per day or placebo for 24 hours. Primarily, we assessed the difference between serelaxin and placebo on renal plasma flow (para-aminohippuric acid clearance) and GFR (iothalamate clearance) over 8 to 24 hours. All 22 patients from 1 clinical site were excluded from primary analyses before unblinding because of implausible measurements. The primary analysis comprised 65 patients, mean age was 68 (±10) years, 89% were male, mean estimated GFR was 64 (±19) mL/min per 1.73 m2, and 34% had New York Heart Association Class III symptoms. Renal plasma flow increased by 29% with serelaxin and 14% with placebo (13% relative increase with serelaxin; P=0.0386), whereas GFR changes did not differ significantly during 8 to 24 hours. Filtration fraction increased by 36% with serelaxin and 62% with placebo (16% relative decrease with serelaxin; P=0.0019) during 8 to 24 hours. Changes in systolic blood pressure were largely similar, and creatinine clearance did not differ between groups. Adverse event rates were similar with serelaxin (20.5%) and placebo (25.0%).

Conclusions—In patients with chronic heart failure, serelaxin increased renal plasma flow and reduced the increase in filtration fraction compared with placebo, but did not affect GFR. These results suggest beneficial renal hemodynamic effects in patients with chronic heart failure.

Clinical Trial Registration—URL: http://www.clinicaltrials.gov. Unique identifier: NCT01546532. (Circ Heart Fail. 2014;7:994-1002.)

Key Words: clinical trial ■ heart failure ■ hemodynamics ■ kidney

Received March 12, 2014; accepted September 29, 2014.From the University of Groningen, University Medical Center Groningen, The Netherlands (A.A.V., S.M., R.H.J.A.S., G.J.N.); Novartis Pharma

AG, Basel, Switzerland (M.D., A.J., D.L.); Institute of Cardiology, Warsaw, Poland (J.S.); University of Maryland, Baltimore (S.S.G.); and Novartis Pharmaceuticals, East Hanover, NJ (Y.Z.).

The Data Supplement is available at http://circheartfailure.ahajournals.org/lookup/suppl/doi:10.1161/CIRCHEARTFAILURE.114.001536/-/DC1. Correspondence to Adriaan Voors, MD, PhD, Department of Cardiology, University Medical Center Groningen, Hanzeplein 1, PO Box 30.001, 9700 RB

Groningen, The Netherlands. E-mail [email protected]

Renal Hemodynamic Effects of Serelaxin in Patients With Chronic Heart Failure

A Randomized, Placebo-Controlled Study

Adriaan A. Voors, MD, PhD; Marion Dahlke, MD, PhD; Sven Meyer, MD; Janina Stepinska, MD, PhD; Stephen S. Gottlieb, MD; Andrew Jones, BSc, MSc;

Yiming Zhang, PhD; Didier Laurent, PhD; Riemer H.J.A. Slart, MD, PhD; Gerjan J. Navis, MD, PhD

Clinical Perspective on p 1002

by guest on July 21, 2018http://circheartfailure.ahajournals.org/

Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from

http://www.clinicaltrials.gov

http://circheartfailure.ahajournals.org/lookup/suppl/doi:10.1161/CIRCHEARTFAILURE.114.001536/-/DC1

mailto:[email protected]

http://circheartfailure.ahajournals.org/
















Voors et al Renal Hemodynamic Effects of Serelaxin in CHF 995

and function.6 However, prospective controlled studies on the renal hemodynamic effects of serelaxin in patients with heart failure, using the gold standard methods for measuring GFR and RPF, are lacking. We assessed, therefore, the renal hemo-dynamic effects of serelaxin in patients with symptomatic CHF, impaired renal function, and a baseline systolic blood pressure (SBP) ≥110 mm Hg.

Methods

Study Design, Population, and TreatmentThis study was a double-blind, randomized, parallel-group, pla-cebo-controlled, multicenter study. Inclusion criteria were CHF with standard oral therapy, including stable oral furosemide 40 to 240 mg/d or equivalent, reduced left ventricular ejection fraction (≤45% within the past 6 months), brain natriuretic peptide ≥100 pg/mL or N-terminal prohormone of brain natriuretic peptide ≥400 pg/mL, New York Heart Association Class II to III, worsening symptoms within the previous 3 months, and mild-to-moderate re-nal impairment (estimated GFR [eGFR] 30–89 mL/min per 1.73 m2). Key exclusion criteria comprised SBP <110 mm Hg, acute contrast-induced nephropathy (at the time of randomization), ad-ministration of intravenous radiographic contrast agent ≤72 hours before randomization or treatment with nonsteroidal anti-inflam-matory drugs, and current or planned treatment with any intra-venous vasoactive therapies or mechanical support. Furthermore, patients with clinically significant hepatic impairment, defined as any hepatic encephalopathy or total bilirubin >50 μmol/L or spon-taneous international normalized ratio >2.0, were not considered for the study. Interventional treatment consisted of serelaxin (30 μg/kg per day) or matching placebo, each provided as a 24-hour intravenous infusion.

A central randomization scheme was used, and patients were ran-domized using consecutive randomization numbers in blocks of 4 for each of the 2 stratification factors: patients with prescribed daily dose of furosemide <120 mg or ≥120 mg orally, or equivalent dose of other loop diuretics, respectively.

Primary End PointsThe 2 primary outcome measures were (1) para-aminohippuric acid (PAH) plasma clearance for RPF, and (2) iothalamate (IOTH) plasma clearance for GFR. The primary interval of interest was 8 to 24 hours after start of infusion. Renal blood flow (RBF) was defined as a coprimary end point in the study protocol and can be derived as RPF/(1−hematocrit). However, because RPF results are predomi-nantly presented in the literature, RPF instead of RBF was treated as the coprimary end point, as specified in the statistical analysis plan that was finalized before database closure.

Secondary outcome variables included filtration fraction (FF), de-fined as the ratio of GFR to RPF in percentage, changes in urinary flow rate (diuresis), sodium excretion (natriuresis), and changes in SBP and diastolic blood pressure (DBP).

The plasma IOTH and PAH clearances were calculated as (in-travenous infusion rate)/(plasma concentration) at each time point. We adjusted the infusion rate for the IOTH and PAH continuous infusion by individual eGFR (using the Modification of Diet in Renal Disease formula). The following formulas were used: infu-sion rate (mg/h) for IOTH=1.2×eGFR (mL/min) and infusion rate (mg/h) for PAH=5.5×eGFR (mL/min). For pooled time intervals (8–24 hours, 0–24 hours, and 24–28 hours), plasma concentra-tion in the above formula was replaced by the time-weighted aver-age plasma concentration (area under the curve divided by time). Baseline plasma clearance was derived using the 0-hour IOTH/PAH concentration measurement.

The urinary PAH and IOTH clearances over each urine collec-tion interval and pooled time intervals were calculated as (amount excreted in urine)/(plasma concentration area under the curve). For baseline (–3 to 0 hours), because plasma concentration was not

available for the –3-hour time point, the area under the curve was predicted from a 1-compartment model fitted to the first 3 data points (0, 2, and 4 hours):

Ct = (IV injection bolus dose/V)×exp(−K×t) + (IV infusionrate/(V×K))×(1−exp(−k×t))+error

Where V and K are the volume of distribution and elimination rate constant, respectively. Specifically:

AUC = (IV injection bolus dose/(V×K))×(1−exp(−3×K))+ (IV infusion rate/(V×K×K))×(−1+exp(−3×k))+3× (IV infusion rate/(V×K))

Plasma clearance of PAH is considered to be the gold standard for RPF, and plasma clearance of IOTH was chosen over urinary clear-ance as the coprimary end point because fewer than half of the pa-tients had change from baseline values available as a result of the difficulty in estimating the baseline values (the 1-compartment model often failed to converge).

AssessmentsTo assess renal hemodynamics, patients were placed in the re-cumbent position, and an intravenous bolus injection of PAH and IOTH was administered 3 hours before the start of study drug infusion, followed immediately by constant intravenous infusion (adjusted for individual eGFR) during study drug infusion, and for 4 hours after stopping study drug infusion (31 hours in total). Notably, oral loop diuretics were last to be taken by patients 24 hours before randomization. Parallel to study drug infusion, all patients were converted from their oral loop diuretic dose to a con-tinuous intravenous infusion of furosemide, outlasting study drug infusion by 4 hours. The total daily dose of continuous intrave-nous furosemide was 50% of the prior total daily oral dose of fu-rosemide. All patients were required to be on a stable (≥4 weeks) dose of furosemide 40 to 240 mg/d orally or equivalent dose of other loop diuretic at the time of study enrollment. If a patient had been taking an oral loop diuretic other than furosemide, the equivalent dose of oral furosemide was calculated and used to de-termine the furosemide dose for continuous intravenous infusion (eg, furosemide 40 mg=torsemide 20 mg=bumetanide 1 mg). All patients were treated with evidence-based medications for treat-ment of heart failure and had to be on a stable dose, particularly of loop diuretics (≥4 weeks). When domiciled, patients were to be on a standard weight-maintaining diet, ideally with small, caloric/salt equivalent portions approximately every 2 hours during the day until a light dinner at ≈7 PM. The total amount of Na/24 hours was ≈80 to 100 mmol. Patients could drink water ad libitum to ensure adequate hydration for urine collection and a target fluid intake of ≥240 mL every 4 hours during waking hours was recommended. Study patients were asked to void their bladder before/at the end of each urine collection interval. In general, meals were to be pro-vided when all other study procedures scheduled for that time had been completed.

Serial blood and urine samples were analyzed to assess safety and effects on renal function and biomarkers. Circulatory biomarkers comprised N-terminal prohormone of brain natriuretic peptide and cystatin C.

Uniquely, the effects of serelaxin on RBF and redistribution of blood flow between the cortical and medullary kidney tissue were analyzed by positron emission tomography (PET)/computed to-mography scanning with 15O-water in a subset of nine patients. Details of the PET imaging procedure are outlined in the Data Supplement. In brief, patients were scanned using a Biograph mCT PET/computed tomography system (Siemens Medical Systems, Knoxville, TN), with a PET axial field of view of 21 cm. Low-dose computed tomography was used for attenuation correction of the region of interest (abdomen) and for anatomic delineation of the kidney aorta and the abdominal aorta. Renal cortical and medullary blood perfusion of the different regions was estimated by fitting the PET measured time activity curves to a validated 1-compartment model for 15O-water.7


Dow

nloaded from


996 Circ Heart Fail November 2014

EthicsThe study was approved by all local Ethics Committees and complied with the Declaration of Helsinki guidelines. Written informed con-sent was obtained from all patients.

Power and Sample Size ConsiderationsAssuming standard deviations (SD) of 40% for RPF and 45% for GFR, and the respective true mean treatment differences of 25.5% and 28.6%, 40 patients per group were predicted to provide at least 80% probabil-ity that the estimated treatment difference in percentage change from baseline is at least 17.8% for RPF and 20% for GFR and is statisti-cally significant at the 0.05 level based on a 2-sided test. Assuming a dropout rate of 10%, 44 patients per group needed to be randomized. In a similar trial, it was shown that an adenosine A1 receptor antago-nist increased RPF and GFR by 32% and 24%, respectively, relative to placebo treatment, over the first 8 hours after the start of study drug infusion.8 An ≈20% lower effect on RPF and an ≈20% higher effect on GFR were considered to be clinically relevant and potentially achiev-able by serelaxin. The true mean treatment difference of 25.5% for RPF and 28.6% for GFR was, therefore, assumed. The respective estimated treatment difference of 17.8% and 20% was based on back calculation given a sample size of 40 patients per treatment arm.

Statistical AnalysesStatistical analyses were performed using SAS software. Relevant baseline variables were summarized and compared between the 2 treatments using methods appropriate for the underlying distribution. Specifically, continuous data were summarized using sample size, mean with SD, geometric mean with 95% confidence interval (bio-chemical variables only), and compared by a 2-sample t test, whereas categorical data were summarized using sample size, absolute and relative frequency and compared by a Fisher exact test.

ANCOVA was performed for time-weighted average (area under the curve divided by time) change from baseline for RPF, GFR, and FF in the log domain over 8 to 24, 0 to 24, and 24 to 28 hours, sepa-rately, using treatment and diuretic dose strata as classification factors and the corresponding log-transformed baseline values as covariate. Results were back-transformed to the original scale and expressed as ratio to baseline or percentage change from baseline. In addi-tion, changes from baseline in RPF, GFR, and FF at each time point were analyzed. One patient in the serelaxin group had no PAH and IOTH concentration data at 24 and 26 hours. In addition, a total of 4 measurements from each of the 2 treatment groups (4 patients in the serelaxin group, 2 patients in the placebo group) were taken between 24 and 28 hours after PAH/IOTH infusion had stopped so were set to missing. The last observation carried forward method was used to impute these missing values before any statistical summary and analysis. The least squares mean difference and associated 95% confidence interval, as well as the P value, were obtained from the treatment comparison. Summaries of RPF, GFR, and FF, and of the percentage change in RPF, GFR, and FF from baseline based on ratio of geometric means (geomeans), were presented, and the geomeans plotted over time by treatment group.

Further secondary variables included changes in urinary flow rate (diuresis), sodium excretion rate (natriuresis), and creatinine clearance. Brachial SBP and DBP were summarized using descriptive statistics.

As a post hoc analysis, an ANCOVA with treatment and diuretic dose stratum as the classification factors, and baseline as a covariate, was performed on change from baseline creatinine clearance in the log domain (measurements including baseline were log-transformed before analysis) for each urinary pooled time interval (8–24, 0–24, and 24–28 hours). For brachial SBP and DBP, a post hoc ANCOVA with treatment as the classification factor, and baseline as a covariate, was performed on change from baseline values. Post hoc analyses were also performed for plasma/serum variables and urine flow rate as well as sodium excretion rate. Safety analyses comprised all pa-tients on study drug infusion with ≥1 postbaseline safety assessment. Patients were analyzed according to treatment received. A 2-sided P<0.05 was considered statistically significant.

The statistical analyses were performed by inVentive Health Clinical, and all results were verified or reviewed by Novartis. The non-Novartis authors did not have direct access to the database. Novartis is ultimately accountable for the accuracy of the analysis results.

Exclusion of Results From One Study Site Before Database ClosureA blinded preliminary data review (before database closure) revealed clinically implausible outlier values for plasma and urine PAH and IOTH data considering the constant rate of administration of PAH and IOTH in this study. These outlier data largely originated from 1 study site (Figures I and II in the Data Supplement). Indeed, 20 of 22 patients from this site were considered outliers for PAH values. The reasons for the implausible PAH/IOTH values could not be identified, and a possi-ble administration/sampling error could not be excluded. Consequently, all data from this study site (n=22 patients) were excluded from the primary pharmacodynamic analyses, and the results presented in the main body of the article refer to the subset of all patients from the other sites. The results of the total per-protocol study cohort, including the data from the site with questionable sampling, are provided in the Data Supplement (Tables I–III in the Data Supplement).

ResultsBaseline CharacteristicsOverall, 87 patients with CHF were randomized in 4 coun-tries at 13 sites. The primary analysis was performed with data from 65 patients (serelaxin n=28; placebo n=37) (Figure III in the Data Supplement). Details of baseline patient characteris-tics are shown in Table 1.

Primary Outcome Variables

Renal Plasma FlowSerelaxin increased RPF from baseline in the 8 to 24 hours of infusion by 29% whereas a 14% increase was reported in the placebo-treated patients (relative increase of 13% with sere-laxin; P=0.0386; Figure 1A). In addition, RPF treatment dif-ferences were statistically significantly in favor of serelaxin for RPF least square (LS) geometric mean ratios to baseline over 0 to 24 hours and 24 to 28 hours, with treatment differences of 16% for both (P=0.0042 and 0.0115, respectively; Table 2). The maximum RPF change from baseline was 56% at 4 hours in the serelaxin group and 20% at 4 and 24 hours in the placebo group (data shown in Table I in the Data Supplement).

Glomerular Filtration RateThere was no statistically significant treatment difference for the coprimary end point of GFR change from baseline over 8 to 24 hours, determined by plasma IOTH clearance (Fig-ure 1B). Treatment differences in GFR change from baseline over 0 to 24 hours and 24 to 28 hours were also nonsignifi-cant (Table 2). However, steady state of IOTH did not seem to have been achieved on start of the study drug infusion, which resulted in an underestimation of GFR during the first hours of infusion for both treatments.

To verify that the cause of this rising GFR over time was a methodological issue and not a true effect, we also determined GFR by urinary IOTH clearance and creatinine clearance. The results support the primary analysis showing no treatment differ-ence comparing serelaxin with placebo, but do not show a rise in GFR (Figure 2) during study drug infusion in either treatment group.


Dow

nloaded from



Secondary Outcome VariablesFiltration FractionFF increased from baseline over 8 to 24 hours, 0 to 24 hours, and 24 to 28 hours in both the serelaxin and placebo groups.

However, the FF changes from baseline over these time points were lower in serelaxin-treated patients compared with pla-cebo-treated patients. Over 8 to 24 hours, serelaxin caused a relative decrease of FF by 16% compared with placebo

Table 1. Baseline Characteristics (Evaluable Population)

Serelaxin Placebo Total

(n=28) (n=37) P Value (N=65)

Male, n (%) 25 (89.3) 33 (89.2) 1.000 58 (89.2)

White, n (%) 27 (96.4) 36 (97.3) 1.000 63 (96.9)

Age, y 68.9 (10.1) 67.1 (9.8) 0.4535 67.9 (9.9)

Body mass index, kg/m2 29.5 (5.0) 28.5 (3.8) 0.3515 29.0 (4.4)

Systolic blood pressure, mm Hg 133.0 (18.8) 125.3 (14.5) 0.0646 128.6 (16.8)

Diastolic blood pressure, mm Hg 80.9 (10.4) 78.1 (9.1) 0.2566 79.3 (9.7)

NYHA classification, n (%)

Class II 17 (60.7) 26 (70.3) 0.4410 43 (66.2)

Class III 11 (39.3) 11 (29.7) 22 (33.8)

Prior HF hospitalization, n (%) 19 (67.9) 27 (73.0) 0.7843 46 (70.8)

Left ventricular ejection fraction, %* 32.1 (8.4) 28.9 (7.7) 0.1237 30.3 (8.1)

Ischemic pathogenesis, n (%) 19 (67.9) 25 (67.6) 1.000 44 (67.7)

AICD, % 6 (21.4) 10 (27.0) 0.7727 16 (24.6)

Pacemaker, % 4 (14.3) 4 (10.8) 0.7171 8 (12.3)

Cardiac resynchronization therapy, % 4 (14.3) 5 (13.5) 1.000 9 (13.8)

Concomitant HF drugs at baseline, %

ACE inhibitor 20 (71.4) 30 (81.1) 0.3885 50 (76.9)

ARB 7 (25.0) 2 (5.4) 0.0324 9 (13.8)

β-Blocker 21 (75.0) 24 (64.9) 0.4272 45 (69.2)

Digoxin 3 (10.7) 7 (18.9) 0.4948 10 (15.4)

MRA 19 (67.9) 27 (73.0) 0.7843 46 (70.8)

†Glomerular filtration rate, mL/min per 1.73 m2

63.9 (18.9) 65.0 (20.1) 0.8243 64.5 (19.5)

Renal plasma flow, mL/min 425.7 (246.2) 344.1 (124.6) 0.0859 379.2 (189.6)

Glomerular filtration rate, mL/min 55.5 (27.5) 52.2 (25.4) 0.6201 53.6 (26.2)

Filtration fraction 14.0 (5.4) 16.3 (8.5) 0.2150 15.3 (7.3)

Plasma/serum variables

Sodium, mmol/L 141.1 (2.3) 141.6 (3.1) 0.4426 141.4 (2.8)

Potassium, mmol/L 4.5 (0.4) 4.4 (0.4) 0.2253 4.4 (0.4)

Creatinine, μmol/L 105.0 (27.5) 105.7 (30.5) 0.9268 105.4 (29.0)

Urea, mmol/L 10.2 (3.8) 9.5 (3.7) 0.4968 9.8 (3.7)

Urine variables

Creatinine clearance, mL/min 90.4 (51.7) 103.2 (40.1) 0.2936 97.6 (45.6)

Urine flow rate, mL/h 114.6 (97.5) 97.9 (83.3) 0.4800 105.1 (89.3)

Sodium excretion rate, mmol/h 26.5 (21.8) 21.3 (18.3) 0.3234 23.5 (19.8)

Circulatory biomarkers

NT-proBNP, ng/L‡ 1425.0 (975.0, 2082.8) 1310.6 (1013.4, 1695.1) 0.7013 1360.1 (1097.8, 1685.0)

Cystatin C, mg/L§ 1.29 (1.15, 1.44) 1.22 (1.11, 1.35) 0.4697 1.25 (1.16, 1.34)

Values are mean (SD) for clinical characteristics and geometric mean (95% confidence interval) for biochemical parameters, unless otherwise stated. ACE indicates angiotensin-converting enzyme; AICD, automatic implantable cardioverter-defibrillators; ARB, angiotensin receptor blocker; HF, heart failure; MRA, mineralocorticoid receptor antagonists; NT-proBNP, N-terminal prohormone of brain natriuretic peptide; and NYHA, New York Heart Association.

*Serelaxin n=27, placebo n=37, total N=64.†Estimated glomerular filtration rate, derived using Modification of Diet in Renal Disease formula from serum creatinine at baseline,

measured by the central laboratory.‡Serelaxin n=27, placebo n=34, total N=61.§Serelaxin n=27, placebo n=35, total N=62.


Dow

nloaded from



(P=0.0019). Similar relative decreases in FF were found over 0 to 24 hours and 24 to 28 hours by 16% (P=0.0004) and 22% (P<0.0001), respectively (Figure 3).

Blood PressureFrom mean baselines of 133.0±18.8 mm Hg in the serelaxin group and 125.3±14.5 mm Hg in the placebo group, changes in brachial SBP during the infusion period were largely simi-lar between patients treated with either serelaxin or placebo (Figure 4A). Mean SBP remained below baseline after the end of the infusion. Baseline brachial DBP was also simi-lar between the serelaxin group and the placebo group

(Figure 4B). Only at the 8-hour time point were changes from baseline in brachial SBP and DBP significantly greater in serelaxin-treated patients compared with placebo. It is of note that changes in SBP and DBP were adjusted for baseline blood pressure differences.

Other Secondary Outcome VariablesNo treatment differences were observed in sodium, potassium, creatinine, urea levels, and creatinine clearance (Table 3). N-terminal prohormone of brain natriuretic peptide baseline values were similar in the 2 treatment groups, and no signifi-cant change from baseline was seen in either treatment group at any time point. Cystatin C baseline values were also similar in the 2 treatment groups and increased with time in the pla-cebo group, but not in the serelaxin group. Both urinary flow rate and sodium excretion rate were significantly higher in the placebo-treated patients (Table 3). The mean equivalent oral furosemide dose during the 24-hour infusion period was 60.1 (±6.85) mg in serelaxin-treated patients and 75.2 (±5.96) mg in placebo-treated patients (Table 3).

PET/Computed Tomography DataOn day 1 (baseline), tissue blood perfusion as estimated from the K1 rate constant appeared to be ≈20% greater in the renal medulla than in the cortex of serelaxin-treated patients, whereas no difference was observed in the placebo group (Figure 5). On day 2, K1 values remained stable in placebo patients, however increased by ≈35% with sere-laxin, in both cortex and medulla. Accordingly, pre- and post-treatment values of the corresponding medulla/cortex ratios remained ≈1.2 and ≈1.0 with serelaxin and placebo, respectively. For both groups, the volume of distribution also appeared to be well below 1.0 and stable over the 24-hour treatment period (ie, ≈0.6 and ≈0.8 mL plasma/mL tissue for the cortex and the medulla, respectively), most likely indicative of a relatively strong rediffusion of H

215O

from the renal tissue into the circulation (ie, k2 rate con-stant, Figure IV in the Data Supplement). In line with this,

Figure 1. Hemodynamic results. Primary outcome variables renal plasma flow (A) and glomerular filtration rate (B) over time.

Table 2. Time-Weighted Average Ratio to Baseline of Renal Plasma Flow and Glomerular Filtration Rate (Coprimary End points) and Filtration Fraction (Secondary End Point)

Time-Weighted Average Ratio to Baseline, h Serelaxin (n=28) Placebo (n=37)

Treatment Difference* (95% Confidence Interval) P Value

Renal plasma flow

0–24 1.31 (1.05) 1.13 (1.04) 1.16 (1.05–1.28) 0.0042

8–24 1.29 (1.05) 1.14 (1.05) 1.13 (1.01–1.27) 0.0386

24–28 1.35 (1.05) 1.16 (1.05) 1.16 (1.03–1.30) 0.0115

Glomerular filtration rate

0–24 1.60 (1.04) 1.66 (1.04) 0.96 (0.88–1.06) 0.4336

8–24 1.78 (1.06) 1.87 (1.05) 0.95 (0.84–1.07) 0.3932

24–28 1.92 (1.06) 2.13 (1.05) 0.90 (0.79–1.03) 0.1371

Filtration fraction

0–24 1.20 (1.05) 1.44 (1.04) 0.84 (0.76–0.92) 0.0004

8–24 1.36 (1.05) 1.62 (1.05) 0.84 (0.75–0.94) 0.0019

24–28 1.41 (1.06) 1.81 (1.05) 0.78 (0.69–0.88) <0.0001

Values are LS geomean ratio to baseline (SE), unless otherwise stated.*Serelaxin to placebo.


Dow

nloaded from



serelaxin induced an ≈30% increase in the reverse transport of H

215O both in the cortex and the medulla (Figure IV in

the Data Supplement) whereas no change in k2 was detected in placebo-treated patients. Finally, V

blood data (Figure V in

the Data Supplement) showed that the blood volume con-tributed to only ≈20% of the PET signal (ie, the PET signal emanates from the tissue), was similar in both cortex, and medulla and was unaffected by treatment.

Safety ParametersDetails on safety data refer to the safety analysis set of all 87 randomized patients. Eight serelaxin-treated patients (20.5%) and 12 placebo-treated patients (25.0%) reported ≥1 adverse event. The majority of adverse events were mild in nature (serelaxin: seven patients [17.9%]; placebo: nine patients [18.8%]) or moderate (serelaxin: 1 patient [2.6%]; placebo:

2 patients [4.2%]) in severity. Overall, five serious adverse events were reported in 4 patients, and 2 had suspected rela-tionship to study drug. One patient treated with serelaxin (2.1%) reported flushing, possibly caused by inadvertent administration of IOTH/PAH bolus, and 3 patients treated with placebo (6.3%) reported nausea or hypotension, hydro-thorax, or an ischemic stroke. In 2 of the patients (1 patient in each group) who met the blood pressure safety criterion (decrease in SBP to <90 mm Hg), the study drug was stopped prematurely and permanently. In 1 serelaxin-treated patient, the infusion rate was halved based on the predefined criteria. In safety laboratory assessments, no changes in hematology or clinical chemistry measurements were notable or considered to be clinically relevant.

DiscussionIn this double-blind study, using the gold standard methods for measuring renal hemodynamics, serelaxin increased RPF and reduced the increase in FF compared with placebo, but did not affect GFR.

The increase in RPF in serelaxin-treated patients is of interest and may be of clinical relevance. Central hemo-dynamic studies of serelaxin in patients with AHF showed that serelaxin decreased the mean pulmonary artery pres-sure and pulmonary capillary wedge pressure, indicating cardiac unloading.9,10 However, in these studies, serelaxin did not increase cardiac index. An increase in cardiac index as an explanation for the increase in RPF with serelaxin in this study is, therefore, unlikely. Another explanation for the increase in RBF in this study might be related to a possible reduction in venous congestion with serelaxin, which has been

Figure 3. Filtration fraction (A), sodium excretion (B), urine flow rate (C), and creatinine clearance (D) over time. IQR indicates interquartile range.

Figure 2. Glomerular filtration rate (GFR; median±interquartile range) assessed by urinary iothalamate clearance.


Dow

nloaded from



related to renal perfusion in several studies.11–14 However, as we did not measure venous pressures, we cannot support this potential explanation. Third, the increase in RPF in the absence of an increase in GFR might be explained by a direct renal vasorelaxant effect of serelaxin, probably by both renal afferent and efferent vasorelaxation, leading to unloading of the glomerulus.

Several studies have shown that RBF (and, in turn, RPF) is the main driver of preservation of renal function, both in AHF and CHF.11–14 It could, therefore, be expected that an increase in RPF results in an improvement in GFR. Correspondingly, a recent biomarker analysis of the RELAX-AHF trial showed that serelaxin-treated patients had lower serum creatinine and cystatin C values compared with placebo-treated patients.4 However, GFR remained unchanged in the present study, and results may, therefore, seem contradictory. Because we could confirm the absence of a treatment effect of serelaxin on urinary IOTH and creatinine clearances, the treatment comparison for plasma IOTH clearance is considered valid, despite the methodological limitation with regards to steady state attainment of IOTH at baseline and during the early hours of study drug infusion for both serelaxin and placebo treatments.

Most importantly, it needs to be emphasized that this study included patients with CHF whereas RELAX-AHF included patients with AHF.3 Patients with CHF were chosen for the current study as large volume shifts occurring in patients with AHF may mask treatment effects in a relatively small study like this. However, it is reasonable to assume that renal function is more seriously compromised during acute hemodynamic dete-rioration in patients with acute decompensated heart failure.1 Interestingly, both Ljungman et al11 and Smilde et al14 showed that in patients with CHF, GFR remains stable even when renal blood pressure dropped. As a result, FF increases, probably by efferent vasoconstriction, to maintain glomerular filtration. However, this increase in FF is associated with an increase in intraglomerular pressures, potentially leading to renal damage in the long term. In the current study, serelaxin resulted in a

relative reduction of FF, thereby unloading the glomerulus, potentially leading to preservation of renal function over time. The absence of improvement of GFR in our study is supported by the lack of marked differences in serum creatinine, blood urea nitrogen, and cystatin C values between the serelaxin- and placebo-treated CHF patients.

The increase in RPF in this study is further supported by an independent assessment, namely a PET study in a subset of patients. In the serelaxin-treated patients, serelaxin substan-tially increased RBF, both in the cortex and medulla, whereas no changes were found in the placebo-treated patients. These findings indicate that serelaxin increases renal perfusion and support the hypothesis that the effects of serelaxin on renal function can be explained by a direct intrarenal effect.

How could the observed change in renal hemodynam-ics support long-term preservation of renal function? The increase in RBF without change in GFR leads to a reduction in FF. In animal experiments in remnant kidney models, an elevated FF, indicating elevated glomerular pressure, con-sistently predicts a more pronounced future renal function decline, whereas reduction of FF, by renin–angiotensin– aldosterone system-blockade, which induces a rise in RBF with unchanged GFR, results in protection against renal function loss, assumed to be mediated predominantly by the decrease in filtration pressure.15 This renal hemodynamic pat-tern, which is elicited by afferent and predominatly efferent vasodilation, displays an attractive similarity to the current findings with serelaxin. However, no such data are avail-able in heart failure, and substantial differences in renal hemodynamics between heart failure, where overall RBF is decreased, and remnant kidney, where nephron blood flow is increased, preclude straightforward extrapolation. In humans, moreover, longitudinal data on the prognostic effects of renal hemodynamics on outcome are virtually absent, because of the demanding nature of renal hemodynamic studies, which

Figure 4. Brachial systolic blood pressure (SBP; A) and diastolic blood pressure (DBP; B) over time.

Table 3. Secondary Outcome Variables

LS Mean Change From Baseline at 24 Hours (SE)

P ValueSerelaxin (n=28) Placebo (n=37)

Plasma/serum variables

Sodium, mmol/L –0.94 (0.42) –0.15 (0.35) 0.1496

Potassium, mmol/L –0.20 (0.06) –0.19 (0.05) 0.9052

Creatinine, μmol/L –3.74 (1.81) 0.68 (1.55) 0.0633

Urea, mmol/L –1.20 (0.24) –1.22 (0.20) 0.9483

Urine variables

Creatinine clearance, mL/min*

0.99 (1.08) 1.03 (1.07) 0.7131

Urine flow rate, mL/h* 1.52 (1.08) 1.91 (1.07) 0.0086

Sodium excretion rate, mmol/h*

1.80 (0.13) 2.56 (0.11) 0.0052

Diuretics

Equivalent oral furosemide dose, mg†

60.1 (6.85) 75.2 (5.96) 0.1009

*Values are LS geomean ratio to baseline (SE).†Values are LS mean (SE) during the 24 h study drug infusion period.


Dow

nloaded from



conflicts with the sample size needed for hard end point stud-ies. Nevertheless, we found previously that a higher FF (low RBF relative to GFR) is an independent predictor of future renal function loss in human transplant recipients.16 To the best of our knowledge, this is the only human study that pro-vides a direct analysis of the predictive effect of renal hemo-dynamics on renal and overall outcome. Interestingly, in this study in renal transplant recipients, a higher FF was also a predictor of mortality, and approximately half of the mortality was cardiovascular. A possible mechanism linking higher FF to worse overall outcome is the effect on renal sodium han-dling: elevated FF blunts renal sodium excretion by its effects on the peritubular Starling force. Of note, such a renal hemo-dynamic profile is not assumed to lead to an abrupt rise in sodium excretion, but rather to a gradual loss of sodium and water that evolves over several days.

In spite of the consistent data from the renal field, however, it is a limitation to the current study that the prognostic value of a particular renal hemodynamic pattern in heart failure patients has not been substantiated.

Serelaxin did not seem to have natriuretic and diuretic effects in this study because urinary flow rate and sodium excretion rate were even higher in patients treated with placebo. Given the observation that GFR was unchanged, and RPF increased, a reduction in sodium excretion and urinary flow rate is difficult to explain. Finally, similar to previous studies with serelaxin in AHF,3,17 the drug was well tolerated in this CHF population with a safety profile comparable to placebo.

This study has some limitations. First, the implausible out-lier values of IOTH and PAH occurred in patients treated at 1 study site. Meticulous efforts were performed to find out the

origin of these outliers, unfortunately without success. Before unblinding, a decision was taken, therefore, to exclude all 22 patients from this single site for the primary analysis, hence seriously compromising the statistical power of the study. Second, steady state of IOTH had not been reached by the start of study drug infusion, which may in part be explained by the physiological circadian rhythm in GFR.18 However, for the primary end point, the between-treatment difference was of interest and change between 8 and 24 hours was prespeci-fied in this study. We do not feel that this lack of steady state altered the findings related to the changes between serelaxin and placebo groups, which were further corroborated by uri-nary IOTH and creatinine clearance results. Finally, serelaxin is currently under investigation in patients with AHF, whereas this study was performed in patients with CHF.

In conclusion, in patients with CHF, serelaxin increased RPF and reduced glomerular FF relative to placebo, but did not increase GFR. These data indicate a direct renal vasorelax-ant effect of serelaxin.

AcknowledgmentsEditorial and artwork assistance was provided by Jonathan Brennan (MediTech Media) and was funded by Novartis Pharma AG, Basel, Switzerland. We also thank Antoon Willemsen (Department of Nuclear Medicine and Molecular Imaging, University of Groningen, Groningen, the Netherlands) for his support on the positron emission tomographic analyses. We are indebted to the multiple clinical sites involved in this study. In the Netherlands, University Medical Center, Department of Cardiology, Groningen; Antonius Hospital, Cardiology Research, Sneek; Deventer Hospital, Cardiology Research, Deventer. In Germany, Charité Universitätsmedizin Berlin, Medical Department Cardiology and Angiology, Berlin; Universitätsklinikum, Nephrology and Hypertension, Erlangen; Universitäres Herzzentrum, Hamburg GmbH, Clinic for General and Interventional Cardiology, Hamburg. In Poland, Institute of Cardiology, Warsaw; Independent Public Central Clinical Hospital 7, Silesian Medical Academy, Katowice; John Paul II Specialist Hospital, Clinical Department of Cardiac and Vascular Diseases, Krakow; Frist Clinic of Cardiology, Jagiellonian University, University Hospital, Krakow; John Paul II Western Hospital, Cardiology Department, Grodzisk Mazowiecki; Wolski Hospital, Cardiology Department, Warsaw. In the USA, University of Maryland, Division of Cardiology, Baltimore.

Sources of FundingThis study was supported by Novartis Pharma AG.

DisclosuresDr Voors has received consultancy fees and research grants from Alere, AstraZeneca, Bayer, Boehringer Ingelheim, Cardio3Biosciences, Celladon, Johnson and Johnson, Merck/MSD, Novartis, Servier, Torrent, Trevena, and Vifor. Dr Stepinska has received research grants from AstraZeneca, Cothere/Novartis, Trevena, and Servier and consulting and speaker fees from AstraZeneca, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Eli Lilly, Pfizer, and Servier. Dr Navis has received consultancy fees from AstraZeneca. Drs Dahlke, Jones, Zhang, and Laurent are employees of Novartis Pharma AG or Novartis Pharmaceuticals. The other authors report no conflicts.

References 1. Metra M, Cotter G, Gheorghiade M, Dei Cas L, Voors AA. The role of

the kidney in heart failure. Eur Heart J. 2012;33:2135–2142. 2. Damman K, Valente MA, Voors AA, O’Connor CM, van Veldhuisen DJ,

Hillege HL. Renal impairment, worsening renal function, and outcome

Figure 5. Positron emission tomography–derived renal blood perfusion as estimated from K1, the rate constant describing the transport of H2

15O from plasma to tissue, in medulla (A) and cortex (B).


Dow

nloaded from



in patients with heart failure: an updated meta-analysis. Eur Heart J. 2014;35:455–469.

3. Teerlink JR, Cotter G, Davison BA, Felker GM, Filippatos G, Greenberg BH, Ponikowski P, Unemori E, Voors AA, Adams KF Jr, Dorobantu MI, Grinfeld LR, Jondeau G, Marmor A, Masip J, Pang PS, Werdan K, Teichman SL, Trapani A, Bush CA, Saini R, Schumacher C, Severin TM, Metra M; RELAXin in Acute Heart Failure (RELAX-AHF) Investigators. Serelaxin, recombinant human relaxin-2, for treatment of acute heart failure (RELAX-AHF): a randomised, placebo-controlled trial. Lancet. 2013;381:29–39.

4. Metra M, Cotter G, Davison BA, Felker GM, Filippatos G, Greenberg BH, Ponikowski P, Unemori E, Voors AA, Adams KF Jr, Dorobantu MI, Grinfeld L, Jondeau G, Marmor A, Masip J, Pang PS, Werdan K, Prescott MF, Edwards C, Teichman SL, Trapani A, Bush CA, Saini R, Schumacher C, Severin T, Teerlink JR; RELAX-AHF Investigators. Effect of serelaxin on cardiac, renal, and hepatic biomarkers in the Relaxin in Acute Heart Failure (RELAX-AHF) development program: correlation with outcomes. J Am Coll Cardiol. 2013;61:196–206.

5. Smith MC, Danielson LA, Conrad KP, Davison JM. Influence of recom-binant human relaxin on renal hemodynamics in healthy volunteers. J Am Soc Nephrol. 2006;17:3192–3197.

6. Garber SL, Mirochnik Y, Brecklin C, Slobodskoy L, Arruda JA, Dunea G. Effect of relaxin in two models of renal mass reduction. Am J Nephrol. 2003;23:8–12.

7. Alpert NM, Rabito CA, Correia DJ, Babich JW, Littman BH, Tompkins RG, Rubin NT, Rubin RH, Fischman AJ. Mapping of local renal blood flow with PET and H(2)(15)O. J Nucl Med. 2002;43:470–475.

8. Dittrich HC, Gupta DK, Hack TC, Dowling T, Callahan J, Thomson S. The effect of KW-3902, an adenosine A1 receptor antagonist, on renal function and renal plasma flow in ambulatory patients with heart failure and renal impairment. J Card Fail. 2007;13:609–617.

9. Dschietzig T, Teichman S, Unemori E, Wood S, Boehmer J, Richter C, Baumann G, Stangl K. Intravenous recombinant human relaxin in com-pensated heart failure: a safety, tolerability, and pharmacodynamic trial. J Card Fail. 2009;15:182–190.

10. Ponikowski P, Mitrovic V, Ruda M, Fernandez A, Voors AA, Vishnevsky A, Cotter G, Milo O, Laessing U, Zhang Y, Dahlke M, Zymlinski R, Metra M. A randomized, double-blind, placebo-controlled, multicentre study to assess haemodynamic effects of serelaxin in patients with acute heart failure. Eur Heart J. 2014;35:431–441.

11. Ljungman S, Laragh JH, Cody RJ. Role of the kidney in congestive heart failure. Relationship of cardiac index to kidney function. Drugs. 1990;39(suppl 4):10–21, discussion 22–24.

12. Damman K, Voors AA, Hillege HL, Navis G, Lechat P, van Veldhuisen DJ, Dargie HJ; CIBIS-2 Investigators and Committees. Congestion in chronic systolic heart failure is related to renal dysfunction and increased mortality. Eur J Heart Fail. 2010;12:974–982.

13. Damman K, Ng Kam Chuen MJ, MacFadyen RJ, Lip GY, Gaze D, Collinson PO, Hillege HL, van Oeveren W, Voors AA, van Veldhuisen DJ. Volume status and diuretic therapy in systolic heart failure and the detection of early abnormalities in renal and tubular function. J Am Coll Cardiol. 2011;57:2233–2241.

14. Smilde TD, Damman K, van der Harst P, Navis G, Westenbrink BD, Voors AA, Boomsma F, van Veldhuisen DJ, Hillege HL. Differential as-sociations between renal function and “modifiable” risk factors in pa-tients with chronic heart failure. Clin Res Cardiol. 2009;98:121–129.

15. Anderson S, Brenner BM. The role of intraglomerular pressure in the initiation and progression of renal disease. J Hypertens Suppl. 1986;4:S236–S238.

16. Bosma RJ, Kwakernaak AJ, van der Heide JJ, de Jong PE, Navis GJ. Body mass index and glomerular hyperfiltration in renal transplant re-cipients: cross-sectional analysis and long-term impact. Am J Transplant. 2007;7:645–652.

17. Teerlink JR, Metra M, Felker GM, Ponikowski P, Voors AA, Weatherley BD, Marmor A, Katz A, Grzybowski J, Unemori E, Teichman SL, Cotter G. Relaxin for the treatment of patients with acute heart failure (Pre-RELAX-AHF): a multicentre, randomised, placebo-controlled, par-allel-group, dose-finding phase IIb study. Lancet. 2009;373:1429–1439.

18. Stow LR, Gumz ML. The circadian clock in the kidney. J Am Soc Nephrol. 2011;22:598–604.

CLINICAL PERSPECTIVEWorsening renal function complicates the treatment of acute and chronic heart failure and might lead to less aggressive treatment with life-saving therapies and poorer clinical outcomes. There is, therefore, a need for heart failure treatments that improve renal function as well. Serelaxin, recombinant human relaxin-2, is a promising and novel investigational therapy for the treatment of acute heart failure and has been associated with an improvement in clinical outcomes in randomized trials to date. In addition, treatment with serelaxin has been associated with lower serum creatinine and plasma cystatin C concen-trations compared with placebo. However, prospective controlled studies on the renal hemodynamic effects of serelaxin in patients with heart failure, under controlled conditions using the gold standard methods for measuring glomerular filtration rate and renal plasma flow, are lacking. In this double-blind, randomized, placebo-controlled multicenter study in 65 systolic chronic heart failure patients, intravenous serelaxin 30 μg/kg per day for 24 hours increased renal plasma flow by 29%, com-pared with 14% on placebo (13% relative increase with serelaxin; P=0.04), whereas glomerular filtration rate did not differ significantly during 8 to 24 hours. Filtration fraction increased by 36% with serelaxin and 62% with placebo (16% relative decrease with serelaxin; P=0.002). Changes in systolic blood pressure were largely similar, and creatinine clearance did not differ between groups. Adverse event rates were similar with serelaxin (20.5%) and placebo (25.0%). The combination of an increased renal blood flow, unchanged glomerular filtration rate, and decreased filtration fraction suggests that serelaxin may have favorable effects on parameters of kidney function in patients with heart failure.


Dow

nloaded from


Jones, Yiming Zhang, Didier Laurent, Riemer H.J.A. Slart and Gerjan J. NavisAdriaan A. Voors, Marion Dahlke, Sven Meyer, Janina Stepinska, Stephen S. Gottlieb, Andrew

Randomized, Placebo-Controlled StudyRenal Hemodynamic Effects of Serelaxin in Patients With Chronic Heart Failure: A

Print ISSN: 1941-3289. Online ISSN: 1941-3297 Copyright © 2014 American Heart Association, Inc. All rights reserved.

75231is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TXCirculation: Heart Failure

doi: 10.1161/CIRCHEARTFAILURE.114.0015362014;7:994-1002; originally published online October 6, 2014;Circ Heart Fail.

http://circheartfailure.ahajournals.org/content/7/6/994World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://circheartfailure.ahajournals.org/content/suppl/2014/10/06/CIRCHEARTFAILURE.114.001536.DC1Data Supplement (unedited) at:

http://circheartfailure.ahajournals.org//subscriptions/

is online at: Circulation: Heart Failure Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further information isthe Editorial Office. Once the online version of the published article for which permission is being requested

can be obtained via RightsLink, a service of the Copyright Clearance Center, notCirculation: Heart Failurein Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:


Dow

nloaded from

http://circheartfailure.ahajournals.org/content/7/6/994

http://circheartfailure.ahajournals.org/content/suppl/2014/10/06/CIRCHEARTFAILURE.114.001536.DC1

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circheartfailure.ahajournals.org//subscriptions/


Renal hemodynamic effects of serelaxin in patients with chronic heart failure: A

randomized, placebo-controlled study

Voors. Renal hemodynamic effects of serelaxin in CHF

Adriaan A. Voors, MD, PhD;1 Marion Dahlke, MD, PhD;2 Sven Meyer, MD;1 Janina

Stepinska, MD, PhD;3 Stephen S. Gottlieb, MD;4 Andrew Jones, MD, BSc, MSc;2 Yiming

Zhang, PhD;5 Didier Laurent, PhD;2 Riemer H.J.A. Slart, MD, PhD;1 Gerjan J. Navis, MD,

PhD.1

1 University of Groningen, University Medical Center Groningen, the Netherlands

2 Novartis Pharma AG, Basel, Switzerland

3 Institute of Cardiology, Warsaw, Poland

4 University of Maryland, Baltimore, MD, USA

5 Novartis Pharmaceuticals, East Hanover, NJ, USA

SUPPLEMENTAL MATERIALS

Supplementary Methods

Positron Emission Tomography/Computed Tomography

Positron Emission Tomography Tracer and Image Acquisition

Subjects were all scanned on a Biograph mCT-64 positron emission tomography/computed

tomography (PET/CT) system (Siemens Medical Systems, TN, USA) with a PET axial field

of view of 21 cm. A low-dose CT (LDCT) was used to correct for the PET signal attenuation,

which is typically greater in deep areas or areas that are surrounded by relatively dense

structure. This LDCT scan was also used for anatomical delineation of the kidneys and the

abdominal aorta. Settings for the LDCT scan are based on the weight of the subject with

default settings of 100 kV and 30 mAs.

At the imaging session day, approximately 800 MBq H215O was produced on-line using a

Scanditronix MC17 cyclotron and was administrated as an intravenous bolus of

approximately 8.3 mL. Each subject had two tracer injections, the first one at baseline and the

second one 24 hours later (placebo or serelaxin). A 10-minute dynamic PET was initiated

immediately before the injection of H215O and acquired in list-mode. Blood pressure and

heart rate were measured during PET acquisition.

Image Reconstruction

For the kinetic analysis, dynamic PET images were reconstructed into 33 frames per scan

with durations of 12x5s, 12x10s, 4x30s, and 5x60s, using an iterative Siemens uHD

reconstruction approach with three iterations, 21 subsets, 5 mm Gaussian post-smoothing,

and zoom 1 to a matrix of 400*400 with all corrections. For PET co-registration purposes, the

LDCT scan (64 slices) with a transaxial scan field of view of 78 cm was reconstructed with

an axial field of view of 21 cm, matrix size of 512x512 (pixel size 0.98 mm), and a slice

thickness of 2 mm using filtered back projection and noise filtered with a standard CT filter.

Upon completion of the reconstruction, all images were reviewed for artifacts and motion.

Image Processing and Analysis

Before image processing, all scans were inspected to check alignment between the PET and

CT images. The LDCT was fused with the H215O PET scan using Syngo workstation

software (version WinNT 5.2 Service Pack 2; Siemens Medical Systems) at baseline and after

treatment administration separately. If appropriate, coordinates of fusion were registered to

overcome misregistration during further analysis. The H215O PET scan obtained after

treatment administration was also fused with the baseline H215O PET scan to ensure exactly

aligned positions of the kidneys at the different time points.

Regions of interest were drawn on the fused H215O PET/CT scans at baseline using Inveon

Research Workplace (version 3.0; Siemens Medical Systems). Time activity curves for the

cortex and medulla were provided by kidney (left, right, both) and anatomical location

(upper, middle, and lower region, including all region analysis, all in 3D axis view). The

same regions were copied into the H215O PET/CT scans obtained after treatment

administration to ensure the same regions and volumes will be analyzed.

For kinetic modeling, the arterial tracer input function was determined from dynamic PET

measurements of the abdominal aortic activity. This method has been shown to be a suitable

alternative to arterial blood sampling.1 A small region of interest (5 voxels) was defined in

the center of the aorta on the summed PET image, where all of the dynamic image frames are

overlaid to generate one static image. Drawing of the region of interest was performed on the

transaxial slices. On the caudal side, the region of interest is bounded one slice above the

aortic bifurcation. On the cranial side only the most upper slice was excluded. The time-

activity curve of this region of interest was used as the arterial input function.

Calculation of Renal Perfusion

The selected regions of the kidneys were further analyzed in the modeling program PMOD

(version 3.1, PMOD technologies Ltd, Zurich, Switzerland). Renal cortical and medullary

blood perfusion (mL·min-1·g-1) of the different regions were estimated by fitting the PET

measured time-activity curves to a validated one-compartment model for H215O 2 with the

tracer concentration in the blood of the abdominal aorta as arterial input. The fitting program

assumes a very high diffusion speed of H215O (making the tissue net uptake of H2

15O limited

by perfusion) and yields the kinetic parameter K1, which is indicative of tracer clearance

from blood to tissue and in the case of H215O equals tissue perfusion (or here renal blood

flow). Other parameters generated from this model are k2, the rate constant describing the

reverse transport of H215O, and Vblood the apparent blood volume. The blood volume gives the

fraction of the measured volume, which has the same time-activity curve as plasma. In most

tissues this is the capillary fraction. For the kidneys this may include (part of) the nephrons.

Finally the distribution volume, calculated as VT = K1/k2, gives the ratio of tissue to plasma

concentration when tissue and plasma concentration are in equilibrium.

References

1. Alpert NM, Rabito CA, Correia DJ, Babich JW, Littman BH, Tompkins RG, Rubin NT,

Rubin RH, Fischman AJ. Mapping of local renal blood flow with PET and H215O. J Nucl

Med. 2002;43:470–475.

2. Chen BC, Germano G, Huang SC, Hawkins RA, Hansen HW, Robert MJ, Buxton DB,

Schelbert HR, Kurtz I, Phelps ME. A new noninvasive quantification of renal blood flow

with N-13 ammonia, dynamic positron emission tomography, and a two-compartment

model. J Am Soc Nephrol. 1992;3:1295–1306.

Supplementary Results

Table S1. Renal Plasma Flow (mL/min) Change from Baseline – All Sites and Excluding Site 3004

All sites Excluding site 3004

Time point/ interval (hours)

Serelaxin LS-geomean

ratio to baseline (SE)

Placebo LS-geomean


Ratio of LS-geomean

ratios# (95% CI)

p-value Serelaxin LS-geomean


Placebo LS-geomean


Ratio of LS-geomean

ratios* (95% CI)

p-value

2 1.56 (1.19) 0.99 (1.16) 1.57 (1.00, 2.46) 0.0489 1.50 (1.05) 1.16 (1.05) 1.30 (1.15, 1.46) <0.0001

4 1.76 (1.16) 0.94 (1.13) 1.88 (1.29, 2.73) 0.0010 1.56 (1.05) 1.20 (1.05) 1.30 (1.16, 1.46) <0.0001

6 1.67 (1.16) 0.87 (1.13) 1.93 (1.33, 2.81) 0.0005 1.40 (1.06) 1.15 (1.05) 1.22 (1.07, 1.38) 0.0022

8 1.51 (1.15) 0.91 (1.13) 1.66 (1.14, 2.41) 0.0076 1.37 (1.06) 1.15 (1.05) 1.19 (1.04, 1.35) 0.0103

20 1.11 (1.16) 1.19 (1.14) 0.93 (0.63, 1.37) 0.7182 1.23 (1.09) 1.17 (1.08) 1.05 (0.86, 1.29) 0.6102

22 1.11 (1.18) 1.20 (1.15) 0.92 (0.60, 1.42) 0.7043 1.32 (1.06) 1.16 (1.05) 1.14 (1.00, 1.30) 0.0569

24 1.53 (1.15) 1.20 (1.13) 1.27 (0.88, 1.84) 0.2001 1.37 (1.06) 1.20 (1.05) 1.14 (1.00, 1.30) 0.0466

26 1.57 (1.19) 1.11 (1.16) 1.42 (0.91, 2.22) 0.1261 1.36 (1.06) 1.15 (1.05) 1.18 (1.04, 1.33) 0.0093

28 1.62 (1.22) 1.16 (1.19) 1.39 (0.82, 2.34) 0.2179 1.31 (1.06) 1.18 (1.05) 1.11 (0.99, 1.26) 0.0793

0–24 1.06 (1.14) 0.97 (1.12) 1.10 (0.79, 1.54) 0.5681 1.31 (1.05) 1.13 (1.04) 1.16 (1.05, 1.28) 0.0042

8–24 1.03 (1.14) 1.01 (1.12) 1.02 (0.73, 1.43) 0.8969 1.29 (1.05) 1.14 (1.05) 1.13 (1.01, 1.27) 0.0386

24–28 1.50 (1.16) 1.09 (1.14) 1.38 (0.93, 2.04) 0.1074 1.35 (1.05) 1.16 (1.05) 1.16 (1.03, 1.30) 0.0115 All sites: N=39 (serelaxin); N=48 (placebo); Excluding site 3004: N=28 (serelaxin); N=37 (placebo).

LS-geomean’ geometric least-squares mean.

* Serelaxin to placebo.

Table S2. Glomerular Filtration Rate (mL/min) Change from Baseline – AllSsites and Excluding Site 3004





Placebo LS-geomean


Ratio of LS-geomean

ratios# (95% CI)



Placebo LS-geomean


Ratio of LS-geomean

ratios* (95% CI)

p-value

2 1.34 (1.09) 1.27 (1.08) 1.05 (0.84, 1.32) 0.6537 1.24 (1.04) 1.26 (1.03) 0.98 (0.91, 1.06) 0.6651

4 1.65 (1.06) 1.31 (1.05) 1.26 (1.07, 1.47) 0.0048 1.42 (1.04) 1.44 (1.03) 0.99 (0.91, 1.07) 0.7978

6 1.73 (1.08) 1.39 (1.07) 1.24 (1.02, 1.50) 0.0285 1.51 (1.04) 1.58 (1.04) 0.95 (0.87, 1.05) 0.3348

8 1.80 (1.08) 1.59 (1.07) 1.13 (0.92, 1.40) 0.2306 1.63 (1.05) 1.71 (1.04) 0.95 (0.86, 1.06) 0.3916

20 2.24 (1.08) 2.06 (1.07) 1.09 (0.89, 1.32) 0.4053 1.90 (1.06) 1.98 (1.05) 0.96 (0.84, 1.10) 0.5377

22 2.31 (1.12) 1.92 (1.10) 1.20 (0.90, 1.60) 0.2063 1.87 (1.06) 2.08 (1.06) 0.90 (0.78, 1.03) 0.1274

24 2.27 (1.09) 1.75 (1.07) 1.30 (1.05, 1.60) 0.0144 1.94 (1.07) 2.09 (1.06) 0.93 (0.81, 1.07) 0.3148

26 2.37 (1.10) 2.05 (1.08) 1.16 (0.92, 1.47) 0.2184 1.92 (1.06) 2.17 (1.06) 0.88 (0.77, 1.02) 0.0818

28 2.42 (1.12) 2.13 (1.10) 1.14 (0.86, 1.51) 0.3736 1.91 (1.06) 2.11 (1.06) 0.90 (0.79, 1.04) 0.1501

0–24 1.76 (1.07) 1.54 (1.06) 1.14 (0.96, 1.35) 0.1247 1.60 (1.04) 1.66 (1.04) 0.96 (0.88, 1.06) 0.4336

8–24 2.01 (1.08) 1.74 (1.06) 1.16 (0.96, 1.40) 0.1329 1.78 (1.06) 1.87 (1.05) 0.95 (0.84, 1.07) 0.3932

24–28 2.35 (1.08) 1.83 (1.07) 1.28 (1.05, 1.57) 0.0151 1.92 (1.06) 2.13 (1.05) 0.90 (0.79, 1.03) 0.1371 All sites: N=39 (serelaxin); N=48 (placebo); Excluding site 3004: N=28 (serelaxin); N=37 (placebo).

LS-geomean, geometric least-squares mean.


Table S3. Filtration Fraction (%) Change From Baseline – All Sites and Excluding Site 3004





Placebo LS-geomean


Ratio of LS-geomean

ratios# (95% CI)



Placebo LS-geomean


Ratio of LS-geomean

ratios* (95% CI)

p-value

2 0.85 (1.12) 1.23 (1.10) 0.69 (0.51, 0.93) 0.0148 0.81 (1.06) 1.06 (1.05) 0.76 (0.67, 0.87) <0.0001

4 0.92 (1.11) 1.37 (1.09) 0.67 (0.51, 0.88) 0.0044 0.91 (1.06) 1.17 (1.05) 0.78 (0.69, 0.88) 0.0001

6 1.00 (1.10) 1.58 (1.08) 0.63 (0.50, 0.81) 0.0002 1.06 (1.06) 1.34 (1.05) 0.79 (0.70, 0.89) 0.0001

8 1.18 (1.12) 1.68 (1.10) 0.70 (0.53, 0.93) 0.0147 1.18 (1.06) 1.45 (1.05) 0.82 (0.72, 0.92) 0.0015

20 2.06 (1.14) 1.73 (1.11) 1.19 (0.85, 1.65) 0.3121 1.51 (1.09) 1.70 (1.08) 0.88 (0.72, 1.08) 0.2188

22 2.10 (1.18) 1.64 (1.15) 1.28 (0.84, 1.96) 0.2428 1.39 (1.06) 1.78 (1.05) 0.78 (0.69, 0.88) 0.0001

24 1.47 (1.13) 1.45 (1.10) 1.01 (0.74, 1.37) 0.9535 1.40 (1.06) 1.73 (1.05) 0.81 (0.72, 0.92) 0.0008

26 1.51 (1.16) 1.83 (1.14) 0.82 (0.56, 1.22) 0.3381 1.40 (1.06) 1.86 (1.05) 0.75 (0.66, 0.86) <0.0001

28 1.50 (1.17) 1.79 (1.14) 0.84 (0.56, 1.26) 0.3958 1.44 (1.06) 1.79 (1.05) 0.80 (0.71, 0.91) 0.0006

0–24 1.64 (1.10) 1.61 (1.08) 1.02 (0.80, 1.30) 0.8820 1.20 (1.05) 1.44 (1.04) 0.84 (0.76, 0.92) 0.0004

8–24 1.95 (1.11) 1.75 (1.09) 1.11 (0.85, 1.45) 0.4370 1.36 (1.05) 1.62 (1.05) 0.84 (0.75, 0.94) 0.0019

24–28 1.55 (1.13) 1.70 (1.11) 0.91 (0.66, 1.25) 0.5693 1.41 (1.06) 1.81 (1.05) 0.78 (0.69, 0.88) <0.0001 All sites: N=39 (serelaxin); N=48 (placebo); Excluding site 3004: N=28 (serelaxin); N=37 (placebo).

LS-geomean, geometric least-squares mean.


Table S4. Creatinine Clearance – Excluding Site 3004

Exluding site 3004

Time

(hours)

Serelaxin geomean ratio to

baseline (95% CI)

Placebo geomean ratio to baseline

(95% CI)

2 1.05 (0.73, 1.51) 1.05 (0.84, 1.30)

4 1.06 (0.82, 1.37) 0.88 (0.70, 1.11)

6 0.87 (0.64, 1.18) 0.82 (0.69, 0.97)

8 1.06 (0.84, 1.34) 0.96 (0.83, 1.11)

20 0.86 (0.68, 1.10) 0.77 (0.67, 0.88)

22 1.30 (0.88, 1.92) 1.13 (0.89, 1.44)

24 1.12 (0.88, 1.42) 0.98 (0.81, 1.19)

26 0.93 (0.65, 1.32) 1.02 (0.86, 1.22)

28 1.40 (1.08, 1.80) 1.03 (0.84, 1.26)

0–24 1.04 (0.86, 1.26) 0.94 (0.81, 1.09)

8–24 1.01 (0.82, 1.23) 0.88 (0.76, 1.01)

24–28 1.14 (0.94, 1.39) 0.99 (0.81, 1.22)

Figure S1. Overlaying individual patient plasma para-aminohippuric acid profiles, all sites

(A), and excluding site 3004 (B).

Figure S2. Overlaying individual patient plasma iothalamate profiles, all sites (A), and

excluding site 3004 (B).

Figure S3. CONSORT flow of patient disposition.

IOTH, iothalamate; PAH, para-aminohippuric acid; PD, pharmacodynamic; PK,

pharmacokinetic.

Figure S4. Reverse transport of H215O from blood to renal tissue (k2 in L/min), in medulla

(A) and, cortex (B).

Figure S5. Cortical and medullary blood volume (Vblood in L) , in medulla (A) and, cortex

(B).

original article - circulation: heart...

Documents