University of Groningen Fibroblast Growth Factor 23: a ... · Fibroblast Growth Factor 23: A Bridge...

University of Groningen

Fibroblast Growth Factor 23: a Bridge Between Bone Minerals and Renal Volume HandlingHumalda, Jelmer Kor

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2016

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Humalda, J. K. (2016). Fibroblast Growth Factor 23: a Bridge Between Bone Minerals and Renal VolumeHandling. [Groningen]: Rijksuniversiteit Groningen.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 26-09-2020

https://www.rug.nl/research/portal/en/publications/fibroblast-growth-factor-23-a-bridge-between-bone-minerals-and-renal-volume-handling(36312a6e-7477-4aa3-bb7d-79e596781853).html



Fibroblast Growth Factor 23: A Bridge Between Bone Minerals and Renal Volume Handling

Jelmer K. Humalda

ColofonJelmer K. HumaldaTitle: Fibroblast Growth Factor 23: A Bridge Between Bone Minerals and Renal Volume Handling

This project was financially supported by:University Medical Center GroningenJunior Scientific Masterclass, Faculty of Medicine, Groningen.Financial support by Amgen B.V., Astellas Pharma Europe B.V., Chipsoft, the Dutch Kidney Foun-dation, Fresenius Medical Care B.V., Graduate School of Medical Sciences, Niernieuws.nl, Noord Negentig, Sanofi Nederland and Timing Uitzendbureau for the printing of this thesis is gratefully acknowledged.

Cover: Berber A. HumaldaInvitation: Berber A. HumaldaEleven Bridges: Berber A. HumaldaCopyright © Jelmer K. Humalda, Groningen, 2016.All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form without explicit prior permission of the author. ISBN: 978-90-367-9193-9 (Printed book) ISBN: 978-90-367-9190-8 (Digital)

Layout and printed by Gildeprint

Fibroblast Growth Factor 23: A Bridge Between Bone Minerals and Renal Volume Handling

Proefschrift

ter verkrijging van de graad van doctor aan deRijksuniversiteit Groningen

op gezag van derector magnifi cus prof. dr. E. Sterken

en volgens besluit van het College voor Promoti es.

De openbare verdediging zal plaatsvinden op

maandag 28 november 2016 om 14.30 uur

door

Jelmer Kor Humalda geboren op 11 mei 1988

te Rott erdam

Fibroblast Growth Factor 23: A Bridge Between Bone Minerals

and Renal Volume Handling

Proefschrift

ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen

op gezag van de rector magnificus prof. dr. E. Sterken

en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op

maandag 28 november 2016 om 14.30 uur

door

Jelmer Kor Humalda

geboren op 11 mei 1988 te Rotterdam

PromotorProf. dr. G.J. Navis

CopromotorDr. M.H. de Borst

BeoordelingscommissieProf. dr. R. Sanderman Prof. dr. P.M. ter Wee Prof. dr. D.J.A. Goldsmith

Promotor Prof. dr. G.J. Navis Copromotor Dr. M.H. de Borst Beoordelingscommissie Prof. dr. R. Sanderman Prof. dr. P.M. ter Wee Prof. dr. D.J.A. Goldsmith

ParanimfenDrs. M.R.M. San GiorgiDrs. N.F. Casteleijn

Paranimfen Drs. M.R.M. San Giorgi Drs. N.F. Casteleijn

ContentsChapter 1 Introduction and Aims of Thesis 9

Part I ‒ Minerals, Volume and Fibroblast Growth Factor 23

Chapter 2 Vitamin D Analogues to Target Residual Proteinuria: Potential Impact on Cardiorenal Outcomes

27

Chapter 3 Dietary Sodium Restriction: a Neglected Therapeutic Opportunity in Chronic Kidney Disease

41

Chapter 4 Fibroblast Growth Factor 23 and the Antiproteinuric Response to Dietary Sodium Restriction During Renin-Angiotensin-Aldosterone System Blockade

57

Chapter 5 Fibroblast Growth Factor 23 and Cardiovascular Mortality after Kidney Transplantation

77

Chapter 6 Fibroblast Growth Factor 23 Correlates with Volume Status in Haemodialysis Patients and is Not Reduced by Haemodialysis

97

Part II ‒ Dietary Interventions

Chapter 7 Response of Fibroblast Growth Factor 23 to Volume Interventions in Arterial Hypertension and Diabetic Nephropathy

119

Chapter 8 Concordance of Dietary Sodium Intake and Concomitant Phosphate Load: Implications for Sodium Interventions

135

Chapter 9 High Potassium Intake Reduces Fibroblast Growth Factor 23 to Increase Renal Phosphate Reabsorption

151

Chapter 10 The SUBLIME Approach: Efficacy and Cost-Effectiveness of a Blended Care Self-Management Approach Facilitated by E-health for Dietary Sodium Restriction in Patients with Chronic Kidney Disease

165

Chapter 11 Summary and General Discussion 187Samenvatting (met uitleg voor niet-ingewijden) 209Zusammenfassung 213

Dankwoord 217About the Author 225Supplements 229

Chapter 1Introduction and Aims of Thesis

11

Introduction and Aims of Thesis

1Introduction

Chronic Kidney Disease – When Renal Function Sinks

The prevalence of chronic kidney disease (CKD) varies among European countries between 3 and 17 per 100 persons (1). In the Netherlands, CKD prevalence has been estimated between 7.6 and 10.4 patients per 100 persons, according to the large observational cohort studies LifeLines (1) and PREVEND (2). A reduced renal function, commonly expressed as the estimated glomerular filtration rate (eGFR), is associated with a 3‒19-fold increased mortality risk and an increased risk of progression to end stage renal disease (ESRD), depending on age and disease severity (3). Blockade of the renin-angiotensin-aldosterone system (RAAS) is the mainstay of CKD treatment. It aims at reduction of rate of renal function loss and of cardiovascular complica-tions by reduction of blood pressure and proteinuria (4, 5). RAAS blockade interferes with the hemodynamic effects of increased RAAS-activity, i.e. systemic and renal vasoconstriction, as well as with its effects on volume status, and its pro-fibrotic effects. Despite proven efficacy however, risk reduction is at best modest. Treatment with angiotensin receptor blockers (ARB) delays progression to ESRD with a mere four to eight months (6, 7). RAAS blockade was introduced in the 1980’s as antihypertensive treatment. Its renoprotective properties were investigated in the 90’s, and large RCTs corroborated use of RAAS blockade for reno- and cardioprotection in CKD around 2000. Unfortunately, from that time until now in 2016 little progress has been made to further retard renal function loss or protect from cardiovascular complications. Our understanding of renal disease, and the factors driving progressive renal function loss, is far from complete, and other mechanisms than the RAAS are likely also involved. In this thesis, we address the interaction of RAAS-activity, sodium- and volume homeostasis, with another major hormonal system, that governs mineral-bone homeostasis.

In CKD the RAAS is aberrantly activated. In line, the prevalence of hypertension increases as renal function declines: for instance from 18.1 per cent when eGFR is 90–119 mL/min/1.73m2 to 82.1 per cent when eGFR has dropped below 30 mL/min/1.73m2 (8). RAAS blockade (e.g. angiotensin-converting enzyme [ACE] inhibitors or ARBs) reduces blood pressure and protein-uria, and these effects are assumed to be main factors for cardiovascular and renal protection. However, ACE inhibitors or ARBs reduce proteinuria only by some 40% on the average (9). The more ‘residual’ proteinuria remains, the higher the residual risk for subsequent cardiovascular events (10) and for progressive renal function loss (11). Combination of ACE inhibition with ARBs was initially thought to enhance RAAS blockade efficacy and improve outcomes. However, large RCTs pointed out that dual-blockade does not provide additional benefits, and even may cause more harm (12-14), as discussed more elaborately in (→ Chapter 2). Therefore, additional strategies to improve efficacy of RAAS blockade are urgently needed. The role of extracellular volume overload has long since been identified as an important mediator of RAAS blockade efficacy, based on the consistent observation that high salt intake blunts the effect of RAAS

Chapter 1

12

blockade in patients with hypertension (15) and with CKD (16). Moreover, volume overload correlates with proteinuria, and identifies patients that have greater therapeutic benefit from sodium restriction and/or diuretics on blood pressure and proteinuria (17, 18). The other way round, proteinuria in itself may induce sodium retention and thus promote volume overload (19). All in all, there is compelling evidence for an interaction of volume overload with the effects of RAAS blockade, with blunting of efficacy during volume-overload. More recent data suggest that phosphate status may interact with efficacy of RAAS blockade as well, and more-over, that phosphate status may interact with volume status. Intriguingly, nephrotic ‒and hence probably volume overloaded‒ pediatric patients have higher serum phosphate concentrations despite preserved eGFR (20). Moreover, higher serum phosphate levels are associated with impaired cardiorenal protective effect of ACE-inhibitors (21). This suggests interaction between bone-mineral and volume-hemodynamic pathways in pathophysiology, and in the therapeutic response to RAAS blockade in CKD. This placed phosphate and mineral bone homeostasis firmly in our sights.

Phosphate and Fibroblast Growth Factor 23 – Between Balancing and Capsizing

The kidney controls serum phosphate concentrations by regulating its excretion. Only in the setting of ESRD, hyperphosphatemia occurs and is strongly associated with increased mortality (22). However, in earlier stages of CKD, serum phosphate is regulated to remain within the normal range (23). Even within this normal range, higher serum phosphate concentrations are associated with a higher risk of mortality in patients with patients with CKD (24, 25) and the general population (26), suggesting that mild disturbances in the regulatory system already have clinical impact. Of note, in CKD patients with higher serum phosphate concentrations, cardiorenal protection by RAAS blockade is less effective (21). Notwithstanding the unequivocal beneficial effect of reducing serum phosphate concentrations in haemodialysis patients (27), strategies to reduce serum phosphate in earlier stages of CKD lack currently evidence for long-term cardiovascular protection.

Phosphate regulation is under control of three different hormones (28). Active, 1,25-dihydroxy-vitamin D3 (1,25D) increases serum calcium and phosphate levels, predominantly by increasing gastrointestinal and renal tubular reabsorption. Parathyroid hormone (PTH) decreases renal phosphate reabsorption –and stimulates the activation of vitamin D. In 2001, fibroblast growth factor 23 (FGF-23) was discovered as the culprit of several bone disorders that are characterized by excessive renal phosphate loss (29, 30). FGF-23 inhibits tubular phosphate reabsorption; in-hibits 1-alpha-hydroxylase, the enzyme that converts inactive 25-hydroxy vitamin D3 into 1,25D (31); and inhibits PTH secretion (32, 33). FGF-23 is able to bind to the FGF-receptor 1c only when the coreceptor α-klotho is present (34). Indeed, mice lacking αklotho or FGF-23 share the same phenotype of premature ageing and ectopic calcification, partly due to high calcium and phosphate concentrations (28, 35, 36). FGF-23 thus counterbalances the effects of vitamin

13


1D on phosphate homeostasis. An additional layer of complexity is added by the interaction of FGF-23 and vitamin D with the RAAS (37). FGF-23 lowers 1,25D, and in this fashion attenuates the inhibiting effect of vitamin D on the RAAS (38). Moreover, FGF-23 may induce left ventricular hypertrophy (39), possibly by aberrant stimulation of the FGF-receptor 4 in the cardiac myocytes (40). The actions of FGF-23 are summarized in Figure 1.

1,25D

PTH

Phosphate1,25D

NaPi PO4 excretionFGF-23

Figure 1. Schematic representation of FGF-23 physiology. FGF-23 signals via the FGF receceptor 1c (FGFR1c) and its obligate coreceptor, α-klotho (Kl) the parathyroid gland to inhibit production of parathyroid hormone (PTH); inhibits expression of sodium-phosphate (NaPi) cotransporters in the kidney, thus stimulate phosphate excretion in the urine; inhibits conversion of inactive vitamin D to active, 1,25-dihydroxy-vitamin D3 (1,25D). Dashed red ar-row signifies pathophysiologic association of FGF-23 and left ventricular hypertrophy, possibly mediated by FGFR-4 signaling or FGF-23 induced volume overload.

The principal site of FGF-23 production is the osteocyte. FGF-23 production increases in re-sponse to 1,25D (31), PTH (41) and phosphate (42), effectively closing negative feedback loops. Also higher calcium concentrations may stimulate FGF-23 production (43), as may inflammatory cytokines or iron deficiency (44). The strongest determinant of FGF-23 concentrations, however, is renal function.

In persons with preserved renal function, the normal value of FGF-23 is typically <100 RU/mL (45). FGF-23 concentrations increase early in the course of CKD (46), long before disturbances in PTH and phosphate values are apparent (23). Allegedly, this increase serves to stimulate phosphate excretion in the remaining functioning nephrons. With the progression of CKD, FGF-23 rises exponentially from values in the hundreds in CKD stage 3–4 to reach extreme values in haemodialysis patients: 10 000 to even 100 000 RU/mL, as illustrated in Figure 2.

Chapter 1

14

Figure 2. Summary of FGF-23 concentrations reported in this thesis. FGF-23 rises exponentially as renal function declines. This occurs earlier and to a more extreme extent than parathyroid hormone (PTH, left Y- axis) and serum phosphate concentrations (right Y-axis). Data points reflect patients with CKD (chapter 4, 7), haemodialysis patients (chapter 6) and renal transplant recipients (chapter 5).

Even the presence of minimal residual renal function in haemodialysis patients is associated with lower FGF-23 levels (47, 48). After renal transplantation and thus restoration of renal function, FGF-23 levels generally return to normal values within 3 months (49). Higher FGF-23 concentrations are associated with an increased mortality risk in haemodialysis patients (50, 51), predominantly cardiovascular mortality in predialysis CKD (52, 53), and even with cardiovascular and all-cause mortality in the general population (54, 55). Because the adverse consequences of high FGF-23 concentrations have repeatedly been demonstrated, FGF-23 has been hypothesized to be a target for intervention. Accordingly, known determinants of FGF-23 have been targeted relentlessly in efforts to lower FGF-23 levels.

A Barrage of Interventions Targeting Fibroblast Growth Factor-23

Efforts to reduce FGF-23 concentrations consist of strategies aimed at various elements, includ-ing dietary, immunological and hemodynamic interventions summarized in table 1.

Table 1. Strategies to reduce FGF-23 concentrations or limit its consequences. Strategies in italic are based on unpublished data.

Strategy Rationale

Dietary strategies

Phosphate restriction Reduce intake of a phosphate.

Phosphate binders Reduce intestinal absorption of phosphate

Sodium restriction May attenuate effects of FGF-23 on volume status

Potassium supplementation May lower FGF-23 and increase serum phosphate

Antibodies to FGF-23 Inhibit FGF-23 action in body

Intensified haemodialysis schedule Reduce phosphate and fluid accumulation, both determinants of FGF-23

15


1Because FGF-23 regulates phosphate homeostasis, dietary phosphate was considered first. Net phosphate reabsorption can be reduced by limiting intake by a phosphate-restricted diet, and deployment of phosphate binders, which are commonly used in ESRD. High dietary phosphate intake results in higher FGF-23 concentrations in healthy volunteers (42, 56, 57). However, in patients with CKD, dietary phosphate restriction did not achieve a reduction of FGF-23 (58). The efficacy of phosphate binders on FGF-23 varies, with absent effects on FGF-23 by calcium-based phosphate binders (59); the effect of lanthanum carbonate that varied from none (60) to strong if combined with dietary phosphate restriction (58); and marked reductions by sevelamer treatment (59). Third, antibodies to FGF-23 have been tested, however full-blockade of FGF-23 function resulted in extreme calcifications and premature death of animals (61), underscoring the physiological necessity of FGF-23 for keeping calcium-phosphate balance in check. Fourth, in haemodialysis, better phosphate control can be achieved by more frequent, or longer dialy-sis sessions (62); or longer hemodiafiltration (63): both strategies also achieve lower FGF-23 concentrations in haemodialysis (63, 64), however both are obviously not applicable to the predialysis CKD population. A fifth strategy follows from the finding that FGF-23 is a particular strong predictor of heart failure rather than ischemic cardiovascular events (53, 65). These observational data have led to the hypothesis that FGF-23 may induce volume overload. Ex-perimental studies found that FGF-23 induces expression of the sodium-chloride cotransporter (NCC) to stimulate sodium retention and induce hypertension (66). This effect could be over-come by the diuretic hydrochlorothiazide. This emphasizes that FGF-23 and the renal regulation of the elements phosphorus (P) and sodium (Na) are possibly intertwined, and interventions that target sodium (diuretics, a low sodium diet) may thus influence phosphate regulation. Consequently, both elements may be identified as twin-targets for dietary restriction. A sixth strategy involves potassium supplementation. Potassium supplementation increases serum phosphate levels (67). This suggest that potassium is another dietary component capable of influencing phosphate regulation, possibly potassium could lower FGF-23 concentrations. If so, potassium would be another target for dietary intervention. However, achievement of dietary intake targets is difficult: notwithstanding dietary recommendations (4), few patients with CKD achieve adherence to dietary sodium restriction (68). Therefore, better strategies for sustained changes in dietary habits are direly needed.

Need for a Multidisciplinary Approach for Enhancing Elements of a Healthy Lifestyle

Restriction of phosphate uptake can be achieved by pharmacological and dietary interventions. Pharmacological phosphate binder therapy is costly, not fully effective and comes with a sub-stantial pill-burden for patients (69). Compliance to a phosphate-restricted diet is difficult and challenging, as phosphate is ubiquitously present in our food products (70), although avoidance of phosphate-rich additives is feasible and facilitates control of hyperphosphatemia.

Chapter 1

16

Several dietary components affect outcome in CKD and the general population. A healthy lifestyle must thus address such dietary factors including, but not restricted to, phosphate. A higher intake of sodium puts patients with CKD at higher risk for progression to ESRD (71) and increased risk for cardiovascular complications despite treatment (72). Even in the general population, high sodium intake is associated with higher blood pressure (73) and increased car-diovascular mortality especially in overweight subjects (74), whereas higher potassium intake may be protective against development of hypertension (75) and mortality (76). Phosphate intake is not routinely analyzed, however the results are mixed between detrimental (77), absent (25), or protective effects (78) in the general, CKD and mixed population respectively. Conversely, the concentration of phosphate in the serum phosphate is a strong predictor of renal disease progression (21) and mortality (24). Sodium restriction has been assessed in several trials in our center (79-82), and elsewhere (83, 84). These studies typically consist of short-term (6 weeks) interventions, and many patients relapse in old habits after participation (68). Of course, a sustained change resulting from dietary interventions is needed to elicit true impact in clinical practice. A sustained change in health behavior however is notoriously dif-ficult to achieve. To achieve and maintain a successful change in health behavior, behavioral interventions can be deployed, for example based on the self-regulation therapy (85). Guidance in the attainment of goals throughout the phases of behavioral change in self-regulation theory however can be time-extensive ‒and hence costly. E-health technology has emerged rapidly over the last few years, and may have potential to support the process of behavioral change and self-management skills of patients with CKD, but experience so far is limited, and not up to the level of evidence-based medicine. All in all, there is an urgent need for novel approaches that can achieve long-term beneficial changes in dietary elements of a healthy lifestyle.

Outline of Thesis

RAAS blockade is the ship-of-the-line in prevention of progressive renal function loss. However, RAAS blockade only delays end stage renal disease by 4‒8 months and indiscriminate combina-tion of several RAAS blocking agents together may be harmful, so alternative approaches to enhance the protective effects of RAAS blockade are needed. These approaches could be phar-macological or dietary intervention, or their combination. This thesis addresses interactions between RAAS/volume status, and regulators of phosphate-, bone- and mineral homeostasis, to identify better treatment strategies based on combined modulation of RAAS/volume status, and of factors involved in phosphate‒bone-mineral homeostasis. We expand this introduction with a review of vitamin D as adjunct to improve antiproteinuric efficacy of RAAS blockade (→ Chapter 2). As became clear, sodium intake remains a treatment target in CKD. This deserves a more elaborate explanation, as we review in (→ Chapter 3). Because FGF-23 may interact with the renin-angiotensin-aldosterone system (37) and higher serum phosphate concentrations

17


1reduce the protective effects of RAAS blockade (21), we set out to investigate whether higher FGF-23 concentrations may impair the efficacy of intensification of RAAS blockade by a low sodium diet (→ Chapter 4). FGF-23 emerged as a robust risk factor for cardiovascular mortality in haemodialysis patients (50, 51), but after renal transplantation FGF-23 concentrations return to near-normal values (49). We hypothesized that relatively higher FGF-23 concentrations in stable renal transplant recipients still herald a high risk of cardiovascular mortality (→Chapter 5). Because FGF-23 is correlated more strongly with volume overload rather than ischemic cardiovascular events (65), we suggested that volume overload itself may be a determinant of FGF-23 levels. Therefore, we assessed in haemodialysis patients the relation between ultrafil-tration volume – the amount of volume that needs to be withdrawn– during a haemodialysis session and FGF-23 concentrations (→ Chapter 6). We also investigated whether the steep drop in serum phosphate during haemodialysis may lower FGF-23 values (→ Chapter 6). To expand the hypothesis that FGF-23 may directly induce volume overload by effects on sodium handling (66), we assessed, the other way round, whether changes in volume status may influence FGF-23 concentrations in patients with preserved renal function (→ Chapter 7). The elements sodium and phosphorus are literally connected at the molecular level in many food products as sodium-phosphate salts (86), and a diet rich of these additives increases FGF-23 levels (87). In a broader perspective, we investigated if sodium intake and phosphate intake are correlated in different patient populations (→ Chapter 8). Another viewpoint arose from the observation that potassium supplementation increases serum phosphate levels (67). We suggested that potassium exerts this effect by reducing FGF-23 concentrations (→ Chapter 9).

This thesis addresses several dietary elements that affect renal and cardiovascular outcomes. However, to realize the therapeutic potential of dietary improvements in clinical practice, a sustained change in dietary behavior must be established. Individual behavioral counseling of patients is time-extensive and costly. E-health technology could facilitate counseling and sup-port behavioral approaches based on self-regulation theory. Therefore, we investigated the fea-sibility and cost-efficacy of a multidisciplinary e-health approach aimed at reduction of dietary sodium intake in renal patients in the randomized, multicenter SUBLIME trial (→ Chapter 10). A general discussion of all these chapters in the perspective of current literature will be provided in (→ Chapter 11).

Chapter 1

18

References

1. Bruck K, Stel VS, Gambaro G, et al. CKD Prevalence Varies across the European General Population. J.Am.Soc.Nephrol. 2015; 27: 2135-47.

2. de Zeeuw D, Hillege HL, de Jong PE. The kidney, a cardiovascular risk marker, and a new target for therapy. Kidney Int.Suppl. 2005; (98): S25-9.

3. Hallan SI, Matsushita K, Sang Y, et al. Age and association of kidney measures with mortality and end-stage renal disease. JAMA 2012; 308: 2349-2360.

4. Werkgroep Richtlijn Chronische Nierschade, Rensma PL, Hagen EC, et al. Richtlijn voor de behandeling van atiënten met Chronische Nierschade (CNS). 2009.

5. Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group. KDIGO 2012 Clinical Practice Guideline for the Evaluation and Management of Chronic Kidney Disease. Kidney Int.Suppl. 2013; 3: 1-150.

6. Lewis EJ, Hunsicker LG, Clarke WR, et al. Renoprotective effect of the angiotensin-receptor antagonist irbesartan in patients with nephropathy due to type 2 diabetes. N.Engl.J.Med. 2001; 345: 851-860.

7. Brenner BM, Cooper ME, de Zeeuw D, et al. Effects of losartan on renal and cardiovascular outcomes in patients with type 2 diabetes and nephropathy. N.Engl.J.Med. 2001; 345: 861-869.

8. Inker LA, Coresh J, Levey AS, Tonelli M, Muntner P. Estimated GFR, Albuminuria, and Complications of Chronic Kidney Disease. Journal of the American Society of Nephrology 2011; 22: 2322-2331.

9. Kunz R, Friedrich C, Wolbers M, Mann JF. Meta-analysis: effect of monotherapy and combination therapy with inhibitors of the renin angiotensin system on proteinuria in renal disease. Ann.Intern.Med. 2008; 148: 30-48.

10. Holtkamp FA, de Zeeuw D, de Graeff PA, et al. Albuminuria and blood pressure, independent targets for cardioprotective therapy in patients with diabetes and nephropathy: a post hoc analysis of the combined RENAAL and IDNT trials. Eur.Heart J. 2011; 32: 1493-1499.

11. Ruggenenti P, Perna A, Remuzzi G, GISEN Group Investigators. Retarding progression of chronic renal disease: the neglected issue of residual proteinuria. Kidney Int. 2003; 63: 2254-2261.

12. Mann JF, Schmieder RE, McQueen M, et al. Renal outcomes with telmisartan, ramipril, or both, in people at high vascular risk (the ONTARGET study): a multicentre, randomised, double-blind, controlled trial. Lancet 2008; 372: 547-553.

13. Parving HH, Brenner BM, McMurray JJ, et al. Cardiorenal end points in a trial of aliskiren for type 2 diabe-tes. N.Engl.J.Med. 2012; 367: 2204-2213.

14. Fried LF, Emanuele N, Zhang JH, et al. Combined Angiotensin inhibition for the treatment of diabetic nephropathy. N.Engl.J.Med. 2013; 369: 1892-1903.

15. Navis G, de Jong PE, Donker AJ, van der Hem GK, de Zeeuw D. Moderate sodium restriction in hypertensive subjects: renal effects of ACE-inhibition. Kidney Int. 1987; 31: 815-819.

16. Heeg JE, de Jong PE, van der Hem GK, de Zeeuw D. Efficacy and variability of the antiproteinuric effect of ACE inhibition by lisinopril. Kidney Int. 1989; 36: 272-279.

17. Slagman MC, Waanders F, Vogt L, et al. Elevated N-terminal pro-brain natriuretic peptide levels predict an enhanced anti-hypertensive and anti-proteinuric benefit of dietary sodium restriction and diuretics, but not angiotensin receptor blockade, in proteinuric renal patients. Nephrol.Dial.Transplant. 2012; 27: 983-990.

18. Hung SC, Kuo KL, Peng CH, et al. Volume overload correlates with cardiovascular risk factors in patients with chronic kidney disease. Kidney Int. 2014; 85: 703-709.

19. Svenningsen P, Friis UG, Versland JB, et al. Mechanisms of renal NaCl retention in proteinuric disease. Acta Physiol.(Oxf) 2013; 207: 536-545.

19


1 20. de Seigneux S, Courbebaisse M, Rutkowski JM, et al. Proteinuria Increases Plasma Phosphate by Altering

Its Tubular Handling. Journal of the American Society of Nephrology 2015; 26: 1608-1618. 21. Zoccali C, Ruggenenti P, Perna A, et al. Phosphate may promote CKD progression and attenuate renopro-

tective effect of ACE inhibition. J.Am.Soc.Nephrol. 2011; 22: 1923-1930. 22. Floege J, Kim J, Ireland E, et al. Serum iPTH, calcium and phosphate, and the risk of mortality in a European

haemodialysis population. Nephrol.Dial.Transplant. 2011; 26: 1948-1955. 23. Isakova T, Wahl P, Vargas GS, et al. Fibroblast growth factor 23 is elevated before parathyroid hormone and

phosphate in chronic kidney disease. Kidney Int. 2011; 79: 1370-1378. 24. Kestenbaum B, Sampson JN, Rudser KD, et al. Serum phosphate levels and mortality risk among people

with chronic kidney disease. J.Am.Soc.Nephrol. 2005; 16: 520-528. 25. Selamet U, Tighiouart H, Sarnak MJ, et al. Relationship of dietary phosphate intake with risk of end-stage

renal disease and mortality in chronic kidney disease stages 3-5: The Modification of Diet in Renal Disease Study. Kidney Int. 2015; 89: 176-84.

26. Chang AR, Grams ME. Serum phosphorus and mortality in the Third National Health and Nutrition Exami-nation Survey (NHANES III): effect modification by fasting. Am.J.Kidney Dis. 2014; 64: 567-573.

27. Fernández-Martín JL, Martínez-Camblor P, Dionisi MP, et al. Improvement of mineral and bone metabo-lism markers is associated with better survival in haemodialysis patients: the COSMOS study. Nephrology Dialysis Transplantation 2015; 30: 1542-1551.

28. Haussler MR, Whitfield GK, Kaneko I, et al. The role of vitamin D in the FGF23, klotho, and phosphate bone-kidney endocrine axis. Rev.Endocr Metab.Disord. 2012; 13: 57-69.

29. Shimada T, Mizutani S, Muto T, et al. Cloning and characterization of FGF23 as a causative factor of tumor-induced osteomalacia. Proc.Natl.Acad.Sci.U.S.A. 2001; 98: 6500-6505.

30. Shimada T, Muto T, Urakawa I, et al. Mutant FGF-23 responsible for autosomal dominant hypophospha-temic rickets is resistant to proteolytic cleavage and causes hypophosphatemia in vivo. Endocrinology 2002; 143: 3179-3182.

31. Liu S, Tang W, Zhou J, et al. Fibroblast growth factor 23 is a counter-regulatory phosphaturic hormone for vitamin D. J.Am.Soc.Nephrol. 2006; 17: 1305-1315.

32. Ben-Dov IZ, Galitzer H, Lavi-Moshayoff V, et al. The parathyroid is a target organ for FGF23 in rats. J.Clin.Invest. 2007; 117: 4003-4008.

33. Galitzer H, Ben-Dov IZ, Silver J, Naveh-Many T. Parathyroid cell resistance to fibroblast growth factor 23 in secondary hyperparathyroidism of chronic kidney disease. Kidney Int. 2010; 77: 211-218.

34. Goetz R, Beenken A, Ibrahimi OA, et al. Molecular insights into the klotho-dependent, endocrine mode of action of fibroblast growth factor 19 subfamily members. Mol.Cell.Biol. 2007; 27: 3417-3428.

35. Kuro-o M, Matsumura Y, Aizawa H, et al. Mutation of the mouse klotho gene leads to a syndrome resem-bling ageing. Nature 1997; 390: 45-51.

36. Shimada T, Kakitani M, Yamazaki Y, et al. Targeted ablation of Fgf23 demonstrates an essential physiologi-cal role of FGF23 in phosphate and vitamin D metabolism. J.Clin.Invest. 2004; 113: 561-568.

37. de Borst MH, Vervloet MG, ter Wee PM, Navis G. Cross talk between the renin-angiotensin-aldosterone system and vitamin D-FGF-23-klotho in chronic kidney disease. J.Am.Soc.Nephrol. 2011; 22: 1603-1609.

38. Li YC, Kong J, Wei M, Chen ZF, Liu SQ, Cao LP. 1,25-Dihydroxyvitamin D(3) is a negative endocrine regulator of the renin-angiotensin system. J.Clin.Invest. 2002; 110: 229-238.

39. Faul C, Amaral AP, Oskouei B, et al. FGF23 induces left ventricular hypertrophy. J.Clin.Invest. 2011; 121: 4393-4408.

40. Grabner A, Amaral A, Schramm K, et al. Activation of Cardiac Fibroblast Growth Factor Receptor 4 Causes Left Ventricular Hypertrophy. Cell Metabolism. 2015;22(6):1020-1032.

Chapter 1

20

41. Lavi-Moshayoff V, Wasserman G, Meir T, Silver J, Naveh-Many T. PTH increases FGF23 gene expression and mediates the high-FGF23 levels of experimental kidney failure: a bone parathyroid feedback loop. Am.J.Physiol.Renal Physiol. 2010; 299: F882-9.

42. Ferrari SL, Bonjour JP, Rizzoli R. Fibroblast growth factor-23 relationship to dietary phosphate and renal phosphate handling in healthy young men. J.Clin.Endocrinol.Metab. 2005; 90: 1519-1524.

43. Quinn SJ, Thomsen AR, Pang JL, et al. Interactions between calcium and phosphorus in the regulation of the production of fibroblast growth factor 23 in vivo. Am.J.Physiol.Endocrinol.Metab. 2013; 304: E310-20.

44. David V, Martin A, Isakova T, et al. Inflammation and functional iron deficiency regulate fibroblast growth factor 23 production. Kidney Int. 2015; 89: 135-46.

45. Jonsson KB, Zahradnik R, Larsson T, et al. Fibroblast growth factor 23 in oncogenic osteomalacia and X-linked hypophosphatemia. N.Engl.J.Med. 2003; 348: 1656-1663.

46. Larsson T, Nisbeth U, Ljunggren O, Juppner H, Jonsson KB. Circulating concentration of FGF-23 increases as renal function declines in patients with chronic kidney disease, but does not change in response to variation in phosphate intake in healthy volunteers. Kidney Int. 2003; 64: 2272-2279.

47. Wang M, You L, Li H, et al. Association of circulating fibroblast growth factor-23 with renal phosphate excretion among hemodialysis patients with residual renal function. Clin.J.Am.Soc.Nephrol. 2013; 8: 116-125.

48. Viaene L, Bammens B, Meijers BK, Vanrenterghem Y, Vanderschueren D, Evenepoel P. Residual renal func-tion is an independent determinant of serum FGF-23 levels in dialysis patients. Nephrol.Dial.Transplant. 2012; 27: 2017-2022.

49. Wolf M, Weir MR, Kopyt N, et al. A Prospective Cohort Study of Mineral Metabolism After Kidney Trans-plantation. Transplantation 2015; 100: 184-93..

50. Jean G, Terrat J, Vanel T, et al. High levels of serum fibroblast growth factor (FGF)-23 are associated with increased mortality in long haemodialysis patients. Nephrology Dialysis Transplantation 2009; 24: 2792-2796.

51. Gutierrez OM, Mannstadt M, Isakova T, et al. Fibroblast growth factor 23 and mortality among patients undergoing hemodialysis. N.Engl.J.Med. 2008; 359: 584-592.

52. Isakova T, Xie H, Yang W, et al. Fibroblast growth factor 23 and risks of mortality and end-stage renal disease in patients with chronic kidney disease. JAMA 2011; 305: 2432-2439.

53. Seiler S, Rogacev KS, Roth HJ, et al. Associations of FGF-23 and sKlotho with cardiovascular outcomes among patients with CKD stages 2-4. Clin.J.Am.Soc.Nephrol. 2014; 9: 1049-1058.

54. Lutsey PL, Alonso A, Selvin E, et al. Fibroblast growth factor-23 and incident coronary heart disease, heart failure, and cardiovascular mortality: the Atherosclerosis Risk in Communities study. J.Am.Heart Assoc. 2014; 3: e000936.

55. Ix JH, Katz R, Kestenbaum BR, et al. Fibroblast growth factor-23 and death, heart failure, and cardiovascu-lar events in community-living individuals: CHS (Cardiovascular Health Study). J.Am.Coll.Cardiol. 2012; 60: 200-207.

56. Antoniucci DM, Yamashita T, Portale AA. Dietary phosphorus regulates serum fibroblast growth factor-23 concentrations in healthy men. J.Clin.Endocrinol.Metab. 2006; 91: 3144-3149.

57. Vervloet MG, van Ittersum FJ, Buttler RM, Heijboer AC, Blankenstein MA, ter Wee PM. Effects of dietary phosphate and calcium intake on fibroblast growth factor-23. Clin.J.Am.Soc.Nephrol. 2011; 6: 383-389.

58. Isakova T, Barchi-Chung A, Enfield G, et al. Effects of dietary phosphate restriction and phosphate binders on FGF23 levels in CKD. Clin.J.Am.Soc.Nephrol. 2013; 8: 1009-1018.

59. Yilmaz MI, Sonmez A, Saglam M, et al. Comparison of calcium acetate and sevelamer on vascular function and fibroblast growth factor 23 in CKD patients: a randomized clinical trial. Am.J.Kidney Dis. 2012; 59: 177-185.

21


1 60. Urena-Torres P, Prie D, Keddad K, et al. Changes in fibroblast growth factor 23 levels in normophospha-

temic patients with chronic kidney disease stage 3 treated with lanthanum carbonate: results of the PREFECT study, a phase 2a, double blind, randomized, placebo-controlled trial. BMC Nephrol. 2014; 15: 71-2369-15-71.

61. Shalhoub V, Shatzen EM, Ward SC, et al. FGF23 neutralization improves chronic kidney disease-associated hyperparathyroidism yet increases mortality. J.Clin.Invest. 2012; 122: 2543-2553.

62. Daugirdas JT, Chertow GM, Larive B, et al. Effects of frequent hemodialysis on measures of CKD mineral and bone disorder. J.Am.Soc.Nephrol. 2012; 23: 727-738.

63. Cornelis T, van der Sande FM, Eloot S, et al. Acute hemodynamic response and uremic toxin removal in con-ventional and extended hemodialysis and hemodiafiltration: a randomized crossover study. Am.J.Kidney Dis. 2014; 64: 247-256.

64. Zaritsky J, Rastogi A, Fischmann G, et al. Short daily hemodialysis is associated with lower plasma FGF23 levels when compared with conventional hemodialysis. Nephrol.Dial.Transplant. 2014; 29: 437-441.

65. Scialla JJ, Xie H, Rahman M, et al. Fibroblast growth factor-23 and cardiovascular events in CKD. J.Am.Soc.Nephrol. 2014; 25: 349-360.

66. Andrukhova O, Slavic S, Smorodchenko A, et al. FGF23 regulates renal sodium handling and blood pres-sure. EMBO Mol.Med. 2014; 6: 744-759.

67. Sebastian A, Hernandez RE, Portale AA, Colman J, Tatsuno J, Morris RC,Jr. Dietary potassium influences kidney maintenance of serum phosphorus concentration. Kidney Int. 1990; 37: 1341-1349.

68. de Borst MH, Navis G. Sodium intake, RAAS-blockade and progressive renal disease. Pharmacol Res Avail-able online 31 March 2016, ISSN 1043-6618, http://dx.doi.org/10.1016/j.phrs.2016.03.037.

69. Tonelli M, Pannu N, Manns B. Oral phosphate binders in patients with kidney failure. N.Engl.J.Med. 2010; 362: 1312-1324.

70. Kalantar-Zadeh K, Gutekunst L, Mehrotra R, et al. Understanding sources of dietary phosphorus in the treatment of patients with chronic kidney disease. Clin.J.Am.Soc.Nephrol. 2010; 5: 519-530.

71. Vegter S, Perna A, Postma MJ, Navis G, Remuzzi G, Ruggenenti P. Sodium intake, ACE inhibition, and progression to ESRD. J.Am.Soc.Nephrol. 2012; 23: 165-173.

72. Lambers Heerspink HJ, Holtkamp FA, Parving HH, et al. Moderation of dietary sodium potentiates the renal and cardiovascular protective effects of angiotensin receptor blockers. Kidney Int. 2012; 82: 330-337.

73. Geleijnse JM, Kok FJ, Grobbee DE. Blood pressure response to changes in sodium and potassium intake: a metaregression analysis of randomised trials. J.Hum.Hypertens. 2003; 17: 471-480.

74. Tuomilehto J, Jousilahti P, Rastenyte D, et al. Urinary sodium excretion and cardiovascular mortality in Finland: a prospective study. The Lancet 2001; 357: 848-851.

75. Kieneker LM, Gansevoort RT, Mukamal KJ, et al. Urinary potassium excretion and risk of developing hyper-tension: the prevention of renal and vascular end-stage disease study. Hypertension 2014; 64: 769-776.

76. O’Donnell M, Mente A, Rangarajan S, et al. Urinary sodium and potassium excretion, mortality, and cardiovascular events. N.Engl.J.Med. 2014; 371: 612-623.

77. Chang AR, Lazo M, Appel LJ, Gutierrez OM, Grams ME. High dietary phosphorus intake is associated with all-cause mortality: results from NHANES III. Am.J.Clin.Nutr. 2014; 99: 320-327.

78. Palomino HL, Rifkin DE, Anderson C, Criqui MH, Whooley MA, Ix JH. 24-Hour Urine Phosphorus Excretion and Mortality and Cardiovascular Events. Clinical Journal of the American Society of Nephrology 2013; 8: 1202-1210.

79. Vogt L, Waanders F, Boomsma F, de Zeeuw D, Navis G. Effects of dietary sodium and hydrochlorothiazide on the antiproteinuric efficacy of losartan. J.Am.Soc.Nephrol. 2008; 19: 999-1007.

Chapter 1

22

80. Slagman MC, Waanders F, Hemmelder MH, et al. Moderate dietary sodium restriction added to angio-tensin converting enzyme inhibition compared with dual blockade in lowering proteinuria and blood pressure: randomised controlled trial. BMJ 2011; 343: d4366.

81. Kwakernaak AJ, Krikken JA, Binnenmars SH, et al. Effects of sodium restriction and hydrochlorothiazide on RAAS blockade efficacy in diabetic nephropathy: a randomised clinical trial. Lancet Diabetes Endocrinol. 2014; 2: 385-395.

82. de Vries LV, Dobrowolski LC, van den Bosch JJ, et al. Effects of Dietary Sodium Restriction in Kidney Trans-plant Recipients Treated With Renin-Angiotensin-Aldosterone System Blockade: A Randomized Clinical Trial. Am.J.Kidney Dis. 2016; 67: 936-44.

83. McMahon EJ, Bauer JD, Hawley CM, et al. A randomized trial of dietary sodium restriction in CKD. J.Am.Soc.Nephrol. 2013; 24: 2096-2103.

84. Ekinci EI, Thomas G, Thomas D, et al. Effects of salt supplementation on the albuminuric response to telmisartan with or without hydrochlorothiazide therapy in hypertensive patients with type 2 diabetes are modulated by habitual dietary salt intake. Diabetes Care 2009; 32: 1398-1403.

85. Maes S, Karoly P. Self-Regulation Assessment and Intervention in Physical Health and Illness: A Review. Appl.Psychol. 2005; 54: 267-299.

86. Lampila LE. Applications and functions of food-grade phosphates. Ann.N.Y.Acad.Sci. 2013; 1301: 37-44. 87. Gutierrez OM, Luzuriaga-McPherson A, Lin Y, Gilbert LC, Ha SW, Beck GR,Jr. Impact of phosphorus-based

food additives on bone and mineral metabolism. J.Clin.Endocrinol.Metab. 2015; jc20152279.

Part IMinerals, Volume and

Fibroblast Growth Factor 23

Chapter 2Vitamin D Analogues to Target Residual Proteinuria: Potential Impact on Cardiorenal OutcomesJelmer K. HumaldaDavid J.A. GoldsmithRavi ThadhaniMartin H. de Borst

Nephrol Dial Transplant. 2015 Dec;30(12):1988-94

Chapter 2

28

Abstract

Residual proteinuria, the amount of proteinuria that remains during optimally dosed renin-angiotensin-aldosterone system (RAAS) blockade, is an independent risk factor for progressive renal function loss and cardiovascular complications in chronic kidney disease (CKD) patients. Dual RAAS blockade may reduce residual proteinuria but without translating into improved cardiorenal outcomes at least in diabetic nephropathy; rather, dual RAAS blockade may increase the risk of adverse events. These findings have challenged the concept of residual proteinuria as an absolute treatment target. Therefore new strategies must be explored to address whether by further reduction of residual proteinuria using interventions not primarily targeting the RAAS benefit in terms of cardiorenal risk reduction would accrue. Both clinical and experimental intervention studies have demonstrated that vitamin D can reduce residual proteinuria through both RAAS-dependent and RAAS-independent pathways. Future research should prospectively explore vitamin D treatment as an adjunct to RAAS blockade in an interventional trial exploring clinically relevant cardiorenal endpoints.

29

Vitamin D and Cardiorenal Outcomes

2

Review

Proteinuria: a Target for Therapy in Chronic Kidney Disease

The presence of proteinuria is a risk factor for adverse cardiorenal outcome across the spectrum of CKD stages. In the general population, proteinuria is associated with renal function loss in-dependently of baseline renal function (1) while in patients with established CKD, the presence of proteinuria is a risk factor for the development of end-stage renal disease (ESRD) (2). Recent large-scale analyses from the CKD prognosis consortium revealed that the relative risk for ESRD by eGFR and albuminuria was independent of the presence of diabetes (3) or hypertension (4), highlighting the importance of proteinuria per se as a predictor of clinical outcomes. Proteinuria is also strongly associated with cardiovascular mortality, independent of other cardiovascular risk factors (5).

Successful proteinuria reduction has been shown to lower the risk of reaching both renal and cardiovascular end-points in concert with, but also independent of, blood pressure reduction (6, 7), while further blood pressure reduction over and above adequate proteinuria reduction in non-diabetic CKD may have no additional benefit in terms of ESRD prevention (8). Proteinuria is thus proposed as an independent treatment target that should be resolutely addressed in order to reduce the risk of progressive renal function loss and cardiovascular complications.

Renin-Angiotensin-Aldosterone System Blockade for Proteinuria: Assets and Boundaries

Renin-angiotensin-aldosterone system (RAAS) blockade is the mainstay of treatment for pro-teinuric CKD, both in diabetic and in non-diabetic patients. RAAS blockade reduces, but rarely abrogates, proteinuria, resulting in ‘residual proteinuria’ (or ‘treatment-resistant proteinuria’). Bolstering the importance of addressing residual proteinuria, the extent of residual proteinuria is associated with the rate of renal function loss across populations (9) while it also determines the remaining cardiovascular risk (6). Under optimal conditions, each RAAS-blocking agent is able to reduce proteinuria by around 40% (ratio of means ARB versus placebo 0.66, ARB and ACEi equally effective) (10). In adherence to ‘the rule of halves’ – one drug halves the amount of proteinuria, whereas the additive antiproteinuric effect of the second drug is only a modest 25% reduction, i.e. ‘halved’ effect – addition of another RAAS-blocking agent resulted in only 25% further reduction of proteinuria (ratio of means 0.76-0.78 of combination therapy versus ARB or ACEi, respectively). Exposure to high doses of two agents targeting the same pathway may provide further reductions in blood pressure and proteinuria; however, this no longer translates into further incremental outcome benefits but rather clear safety signals emerging in the form of hyperkalemia and acute kidney injury (Table 1).(11) Combination of a RAAS-blocking compound (ACEi or ARB) with the direct renin inhibitor aliskiren even increased the cardiorenal risk, along with a higher incidence of hyperkalemia and hypotension (12).

Chapter 2

30

Table 1. Three major randomized-controlled trials aimed at reducing residual proteinuria and cardiovascular out-come by combined RAAS blockade

ONTARGET (11) ALTITUDE (12) VA NEPHRON-D (13)

Population 25,620 patients ≥55 years with diabetes and end-organ damage or with atherosclerotic vascular disease

8,561 patients ≥35 years with diabetes and microalbuminuria, macroalbuminuria or cardiovascular disease

1,448 patients with type 2 diabetes and macroalbuminuria

Renal function Mean eGFR 74 mL/min/1.73 m2

eGFR <60 mL/min/1.73 m2: n=8,034eGFFR <30 mL/min/1.73 m2: n=263

Mean eGFR 57 mL/min/1.73 m2

eGFR <60 mL/min/1.73 m2: n=5,778eGFR <30 mL/min/1.73 m2: n=210

Mean eGFR 54 mL/min/1.73 m2

eGFR <60 mL/min/1.73 m2: n=894eGFR <30 mL/min/1.73 m2: none

Intervention • Telmisartan 80 mg/d• Ramipril 10 mg/d• Telmisartan + ramipril

ACEi/ARB therapy combined with:• Aliskiren 300 mg/d• Placebo

Losartan 100 mg/d combined with:• Lisinopril 10-40 mg/d• Placebo

Median follow-up

56 months 32.9 months (study halted prematurely)

26.4 months (study halted prematurely)

Proteinuria outcome

Combination therapy reduced the increase in albuminuria compared with ramipril monotherapy (21% vs 31%, P=0.0009).

Combination therapy decreased albumin-to-creatinine ratio more than monotherapy (between group difference: 14% [95% CI 11% to 17%]).

Combination therapy decreased albumin-to-creatinine ratio more (786 to 517 mg/g) than losartan monotherapy (829 to 701 mg/g), P<0.001.

Primary outcome

Combination therapy had an increased occurrence of the composite renal endpoint (dialysis, doubling serum creatinine, death; HR 1.09 [1.01-1.18], P=0.037).

Combination therapy showed no beneficial effect on the primary composite endpoint (cardiovascular events; renal events i.e. ESRD, RRT needed but not given, death by renal cause, doubling of creatinine).

There was no benefit of combination therapy on primary endpoint (eGFR decline of ≥30mL/min/1.73m2 or 50% reduction, ESRD or death), secondary endpoint (eGFR decline or ESRD) or tertiary endpoints (cardiovascular events, eGFR slope).

Safety concern Increased occurrence of the primary composite renal endpoint with combination therapy.

Combination therapy was associated with higher incidence of hyperkalemia (11.6% vs 7.2%) and hypotension (12.1% vs 8.3%), both P<0.001.

Combination therapy increased the risk of hyperkalemia (6.3 vs 2.6 events per 100 person-years) and acute kidney injury (12.2 vs 6.7 events per 100 person-years), both P<0.001).

Abbreviations: ACEi, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker; eGFR, estimated glomerular filtration rate, ESRD, end-stage renal disease; HR, hazard ratio.

The dissociation of proteinuria reduction from improved cardiorenal outcomes during and beyond optimally-dosed RAAS blockade raises the question whether we have reached the maxi-mum beneficial effect of RAAS blockade or whether we are reaching the toxicity threshold of the pharmacological instruments we choose to deploy. Agents with a different pharmacological

31


2

and side effect profile, i.e. with limited or no effect on blood pressure and serum potassium concentrations, are urgently needed to address whether the relationship between proteinuria and cardiorenal outcomes may persist when using a different interventional treatment modal-ity to further reduce proteinuria. One important type of drugs currently under investigation to this extent is the endothelin antagonist. Atrasentan, a selective endothelin A receptor an-tagonist, was recently demonstrated to reduce albuminuria and improve blood pressure and lipid spectrum with manageable fluid overload-related adverse events in patients with type 2 diabetic nephropathy receiving RAAS inhibitors (14). The effect of atrasentan on hard endpoints in this population is investigated by the currently ongoing SONAR trial in over 4000 patients (NCT01858532).

Vitamin D in Chronic Kidney Disease

CKD is characterized by a progressive inability to generate active vitamin D (1,25(OH)2-vitamin D, calcitriol) from its precursor 25(OH)-vitamin D (calcidiol). CKD patients are also more commonly deficient in 25(OH)-vitamin D in comparison with subjects with normal renal function (15). The urinary loss of vitamin D bound to albumin and its carrier protein vitamin D-binding protein (VDBP) might predispose patients to vitamin D deficiency. In line with this concept, proteinuria reduction by ACE inhibition also reduced urinary VDBP loss (16). The impact of urinary VDBP loss on vitamin D status and the potential of antiproteinuric therapy to improve vitamin D status in patients with non-nephrotic range proteinuria however remain to be established. Genetic variation in VDBP may also influence 25(OH)-vitamin D concentrations, although this may not translate into different concentrations of bioavailable 25(OH)-vitamin D (17). Moreover, in CKD the vitamin D-activating enzyme 1-alpha hydroxylase (Cyp27b1) is suppressed by increased fi-broblast growth factor 23 (FGF-23) concentrations (18). The progressive deregulation of vitamin D metabolism with deteriorating renal function is mirrored by the fact that vitamin D deficiency itself may contribute to progressive renal function loss (19). If untreated, this vicious circle may well be one of the forces driving progressive kidney disease, bone disease and possibly also cardiovascular disease in CKD patients.

Vitamin D, RAAS Blockade and CKD Progression

Over the past decade, several animal studies demonstrated the capacity of supplementation with either endogenous vitamin D (cholecalciferol, calcidiol or calcitriol) or its analogues (e.g. paricalcitol) to reduce proteinuria along with renal inflammation, glomerulosclerosis and in-terstitial fibrosis in models of fibrotic and inflammatory chronic kidney disease, as reviewed elsewhere (20)). In humans, observational studies have documented associations between the use of vitamin D supplements (calcitriol and calcitriol or alfacalcidol, respectively) and survival in CKD (21, 22). Together, these findings set the stage for a number of prospective randomized controlled trials (RCTs) addressing the effect of vitamin D analogues on proteinuria as a surrogate endpoint. We recently performed a systematic review of all RCTs with either endogenous-active

Chapter 2

32

vitamin D (calcitriol) or its synthetic analogue paricalcitol as an antiproteinuric intervention. During follow-up, active vitamin D analogues reduced proteinuria on average by 16%, whereas proteinuria increased by 6% in patients receiving control treatment (P<0.001) (23).

Of interest, these results were obtained in the majority of cases (84% overall) against the background of pre-existing chronic RAAS blockade, underlining the capacity of vitamin D ana-logues to reduce residual proteinuria. The additive effect of RAAS blockade and vitamin D on proteinuria could point towards interactions between the RAAS and vitamin D (24) as well as towards RAAS-independent renoprotective effects of vitamin D. Of note, a recent study found that ramipril reduces FGF-23 levels in Stage 1-2 CKD patients with diabetic nephropathy, with-out clear effects on PTH, 25(OH)-vitamin D or calcium levels (25). In a very recent study, it was found that proteinuria in itself affects renal phosphate handling, increasing serum phosphate and FGF-23 levels (26). This may very well explain the observed relationship between protein-uria (reduction) and FGF-23. In turn, reduced FGF-23 levels as a consequence of ACE inhibitor therapy may very well increase the bioavailability of active 1,25(OH)2-vitamin D, contributing to proteinuria control. Thus, the capacity of ACE inhibitors to lower proteinuria may in fact be in part through an effect of FGF-23 and vitamin D. Departing from a long known association be-tween calcitriol and renin levels, elegant experimental studies have elucidated interference of the vitamin D receptor (VDR) with a transcription factor binding site in the (pro)renin promoter region, inhibiting renin expression (27). Consequently, VDR-/- mice display strongly increased renal and circulating renin concentrations as well as increased angiotensin II generation (28). In man, vitamin D deficiency is accompanied by increased circulating angiotensin II concentrations and blunted renal plasma flow responses to infused angiotensin II, indicating both systemic and intrarenal RAAS activation (29). The capacity of vitamin D analogues to reduce the compensa-tory induction of renin during RAAS blockade may at least in part explain their renoprotective effect in addition to RAAS blockade.

Despite this relatively well-defined negative regulatory effect on renin production, VDR ago-nists should not be considered equal to conventional RAAS blockers (Figure 1). From a clinical perspective, vitamin D analogues such as paricalcitol may display limited to no effects on blood pressure (30), nor have hyperkalemia as a side effect, making them more suitable for combi-nation with RAAS blockers further to downtitrate proteinuria in more advanced CKD, where hyperkalemia can be more problematic (Table 1).

33


2

Renal fibrosis

TGF-beta/SMAD

Progressive renal function loss

Angiotensin II

Angiotensin I

Angiotensinogen

Renin

ACE

AT1R

ACEi/ARB

Vitamin D

Figure 1. Schematic representation of complementary renoprotective actions of ACEi/ARB and vitamin D in chron-ic kidney disease. Vitamin D may provide renoprotection through RAAS-mediated effects, i.e. by suppression of renin gene expression. This effect is through VDR-signaling. The second renoprotective pathway of vitamin D is through reduction of TGF-beta/SMAD signaling.Negative regulation indicated by blue arrows, positive regulation by red arrows. ACEi, angiotensin converting en-zyme inhibitor; ARB, angiotensin II receptor blocker; AT1R, angiotensin II type 1 receptor; TGF-beta, transforming growth factor-beta. See text for references.

Vitamin D supplementation yields renoprotective effects beyond the RAAS. Recently published elegant studies revealed that novel vitamin D analogues may also exert antifibrotic effects by influencing the TGF-beta/Smad pathway (31). Using synthetic ligands based on the structure of the VDR–ligand complex that were specific to reduce TGF-beta/Smad signaling, Ito et al reduced renal fibrosis in an animal model, while hypercalcemia was avoided. The use of 1,25(OH)2-vitamin D-derived synthetic ligands may allow more potent blockade of pro-fibrotic pathways in chronic kidney disease, with a more limited tendency to hypercalcemia, which commonly limits vitamin D uptitration in clinical practice. A question that remains to be addressed is whether chemical compounds specifically targeting TGF-beta/Smad but without activating classical VDR-mediated pathways will take away not only hypercalcemia but also beneficial effects such as those resulting from renin suppression. Vitamin D supplementation also exerts beneficial effects on endothelial function. Vitamin D directly regulates endothelial nitric oxide synthase (eNOS) in mice, and mice lacking VDR demonstrated endothelial dysfunction and arterial stiffness (32). In line, VDR activation by paricalcitol improves endothelium-dependent vasodilatation in patients with Stage 3–4 CKD (33).

Chapter 2

34

Interestingly, also the precursor vitamin D compound cholecalciferol may lower proteinuria. In a recently published clinical trial, patients with Stage 3–4 CKD and albuminuria were randomized to receiving 666 IU/day of cholecalciferol or no treatment. After 6 months of follow-up, the albumin-to-creatinin ratio (ACR) had decreased by 53.2% (95% CI 66.0–27.0%), whereas the ACR increased by 7.1% (−25.3 to +53.3%) in the control group (P = 0.005 between groups) (34).

Potential Side Effects of Vitamin D Treatment

A finding that should be interpreted as a warning sign is the increased calcium-phosphate product observed in patients treated with vitamin D when given either as cholecalciferol (34) or as active vitamin D (analogues), as documented in some but not all trials (23). On one hand, an increased calcium load elicited by vitamin D may predispose to the development of adynamic bone disease, accompanied by a higher fracture risk but possibly also vascular calcification (35). On the other hand, a higher level of serum phosphate may not only promote vascular calcifica-tion but also (partially) blunt the renoprotective effect of RAAS blockade (36). The long-term effects of a higher serum phosphate level induced by vitamin D treatment may warrant close monitoring. Nevertheless, it is reassuring that even in the high-risk population of haemodialysis patients, serum calcium is only associated with increased mortality when its levels are higher than 2.75 mmol/L (37). Treatment decisions should be guided by individual patient character-istics, e.g. favoring paricalcitol over calcitriol in the case of a high–normal serum calcium in the CKD-Mineral Bone Disease (MBD) patient at increased risk for vascular calcification (38). Thus, the benefit–risk ratio of vitamin D added to RAAS blockade, when compared with dual RAAS blockade, is driven by different efficacy and safety profiles; whether this translates into clinically relevant benefits overwhelming its side effects remains to be established.

Vitamin D and Cardiovascular Outcome in CKD

The cardiovascular effects of vitamin D deficiency and its supplementation in CKD are not well understood. Epidemiological studies have suggested associations between vitamin D deficiency and cardiovascular morbidity and mortality in CKD patients (39), and the use of vitamin D supplements (predominantly the native form cholecalciferol) has been associated with a survival advantage (40). However, prospective intervention studies designed to establish the cardioprotective effects of vitamin D supplementation have yielded mixed results. In the Women’s Health Initiative study conducted in postmenopausal women, combined calcium and vitamin D3 supplementation failed to materially reduce the risk for myocardial infarction or death by coronary heart disease (41). In the PRIMO trial, paricalcitol treatment had no effect on left ventricular mass index (30), but did lower left atrial volume index and brain natriuretic peptide in CKD patients (42). Therefore, at this moment, it is premature to conclude that vita-min D supplementation can influence the strongly increased cardiovascular risk to which CKD patients are exposed. On the other hand, since a considerable number of studies have now documented reduction of (residual) proteinuria –a well-established cardiorenal marker in CKD–

35


2

it is very worthwhile to further clarify the position of vitamin D treatment as an adjunct to RAAS blockade. Of note, a large randomized trial addressing among others whether cholecalciferol can provide primary prevention against cardiovascular disease in 20 000 men in the USA(VITAL) is currently ongoing (ClinicalTrials.gov identifier: NCT01169259).

A secondary analysis of the other VITAL study, which was performed in patients with diabetic nephropathy (43), suggested that the capacity of paricalcitol to reduce albuminuria depends on the patient’s dietary sodium status. This is relevant given the fact that dietary sodium intake, through impacting volume status, is another modifiable determinant of the antiproteinuric response to RAAS blockade.

RAAS Blockade and Dietary Sodium

High dietary sodium intake is associated with attenuated antiproteinuric and long-term reno-protective effects of RAAS blockade in both diabetic and non-diabetic proteinuric CKD. Further, numerous trials have now demonstrated that sodium restriction further reduces residual pro-teinuria during RAAS blockade, with a possible role for volume markers to identify patients who might benefit most from sodium restriction, as reviewed in Ref. (44). In hypertensive patients, every 1 g of sodium intake per day was associated with a 14% increased risk of coronary heart disease (45). These reports underscore the necessity to reduce sodium intake below 5 g of sodium chloride per day in CKD patients, as indicated in current guidelines. A recent double-blinded, placebo-controlled trial demonstrated that proteinuria can be halved in just 2 weeks by reducing sodium excretion from 168 to 75 mmol/day in hypertensive Stage 3–4 CKD (46). Notwithstanding recent data reporting absence of an association of high sodium excretion and renal failure per se (47), high sodium intake remained associated with cardiovascular morbid-ity and mortality in a worldwide study (48). Taking into account that a reduction of dietary sodium by 33-44 mmol/day lowered cardiovascular risk by 25% in 744 prehypertensive patients after 10-15 years follow-up (49), efforts to reduce and optimize sodium intake in patients at risk according to current guidelines should be increased. The persistent effect across different populations –particularly in CKD – favors modulation of sodium intake as a strategy to intensify RAAS blockade efficacy. Alternative strategies to increase the cardiorenal protective effect of RAAS blockade include combination therapy with diuretics, targeting obesity, and moderate protein restriction, as reviewed elsewhere (50).

As both vitamin D and dietary sodium restriction have a strong potential to lower residual proteinuria during RAAS blockade, the combination of these strategies seems prudent. A multicenter, randomized controlled crossover trial addressing the combined impact of dietary sodium restriction and the vitamin D analogue paricalcitol on residual proteinuria in 50 non-diabetic patients is currently ongoing; results are expected in 2015 (ViRTUE trial, Dutch trial register NTR2898). The interaction between sodium intake and paricalcitol on albuminuria is

Chapter 2

36

currently under investigation in 112 diabetic patients (PROCEED trial, ClinicalTrials.gov identi-fier: NCT01393808).

Summary and Future Directions

Although proteinuria has been considered the key target for renoprotective therapy, recent studies using double RAAS blockade have dissociated proteinuria reduction per se from pro-gressive renal function loss. Since the boundaries of RAAS blockade-based treatment are now better defined, this paves the way to study adjunctive antiproteinuric therapies with different pharmacological and side effect profiles. Given the emerging independent observations that vitamin D analogues are able to reduce proteinuria, even when added to RAAS blockade, fu-ture studies are warranted to investigate whether further reduction of proteinuria beyond the maximally tolerated dose of RAAS blockade can further improve cardiorenal prognosis of CKD patients. Disturbed calciumphosphate metabolism is probably the most prominent safety signal that should be accounted for in such studies. How cardiorenal protective and adverse effects translate into hard patient-specific outcomes remains therefore to be established in future large-scale prospective RCTs, designed to address the effects of vitamin D supplementation added to background RAAS blockade on hard cardiorenal end points.

Conflict of Interest Statement

D.J.A.G. has received speaking and consulting honoraria from Abbott, Amgen, Genzyme, Keryx Pharmaceuticals and Shire. R.T. has received a coordinating center grant from Abbott to the Massachusetts General Hospital and speaker’s fees and travel support from Abbott. R.T. is also a consultant to Fresenius Medical Care. M.H.D.B. has received non-financial support from Abbott to the University Medical Center Groningen for an ongoing clinical trial, and lecture fees from Amgen. J.K.H. has none to declare.

Acknowledgements

This work is supported by a consortium grant from the Dutch Kidney Foundation (NIGRAM con-sortium, grant no CP10.11). Dr. De Borst is supported by grants from the Dutch Kidney Founda-tion (grant no KJPB.08.07) and the Netherlands Organization for Scientific Research (Veni grant). The funding sources had no role in the preparation, review, or approval of the manuscript.

37


2

References

1. Turin TC, James M, Ravani P, et al. Proteinuria and rate of change in kidney function in a community-based population. J.Am.Soc.Nephrol. 2013; 24: 1661-1667.

2. McQuarrie EP, Traynor JP, Taylor AH, et al. Association Between Urinary Sodium, Creatinine, Albumin, and Long-Term Survival in Chronic Kidney Disease. Hypertension 2014; 64: 111-117.

3. Fox CS, Matsushita K, Woodward M, et al. Associations of kidney disease measures with mortality and end-stage renal disease in individuals with and without diabetes: a meta-analysis. Lancet 2012; 380: 1662-1673.

4. Mahmoodi BK, Matsushita K, Woodward M, et al. Associations of kidney disease measures with mortality and end-stage renal disease in individuals with and without hypertension: a meta-analysis. Lancet 2012; 380: 1649-1661.

5. Hillege HL, Fidler V, Diercks GF, et al. Urinary albumin excretion predicts cardiovascular and noncardiovas-cular mortality in general population. Circulation 2002; 106: 1777-1782.

6. Holtkamp FA, de Zeeuw D, de Graeff PA, et al. Albuminuria and blood pressure, independent targets for cardioprotective therapy in patients with diabetes and nephropathy: a post hoc analysis of the combined RENAAL and IDNT trials. Eur.Heart J. 2011; 32: 1493-1499.

7. Schmieder RE, Mann JF, Schumacher H, et al. Changes in albuminuria predict mortality and morbidity in patients with vascular disease. J.Am.Soc.Nephrol. 2011; 22: 1353-1364.

8. Ruggenenti P, Perna A, Loriga G, et al. Blood-pressure control for renoprotection in patients with non-diabetic chronic renal disease (REIN-2): multicentre, randomised controlled trial. Lancet 2005; 365: 939-946.


10. Kunz R, Friedrich C, Wolbers M, Mann JF. Meta-analysis: effect of monotherapy and combination therapy with inhibitors of the renin angiotensin system on proteinuria in renal disease. Ann.Intern.Med. 2008; 148: 30-48.

11. Mann JF, Schmieder RE, McQueen M, et al. Renal outcomes with telmisartan, ramipril, or both, in people at high vascular risk (the ONTARGET study): a multicentre, randomised, double-blind, controlled trial. Lancet 2008; 372: 547-553.



14. de Zeeuw D, Coll B, Andress D, et al. The Endothelin Antagonist Atrasentan Lowers Residual Albuminuria in Patients with Type 2 Diabetic Nephropathy. J.Am.Soc.Nephrol. 2014; 25: 1083-1093.

15. LaClair RE, Hellman RN, Karp SL, et al. Prevalence of calcidiol deficiency in CKD: a cross-sectional study across latitudes in the United States. Am.J.Kidney Dis. 2005; 45: 1026-1033.

16. Doorenbos CR, de Cuba MM, Vogt L, et al. Antiproteinuric treatment reduces urinary loss of vitamin D-binding protein but does not affect vitamin D status in patients with chronic kidney disease. J.Steroid Biochem.Mol.Biol. 2012; 128: 56-61.

17. Powe CE, Evans MK, Wenger J, et al. Vitamin D-binding protein and vitamin D status of black Americans and white Americans. N.Engl.J.Med. 2013; 369: 1991-2000.

18. Shimada T, Hasegawa H, Yamazaki Y, et al. FGF-23 is a potent regulator of vitamin D metabolism and phosphate homeostasis. J.Bone Miner.Res. 2004; 19: 429-435.

Chapter 2

38

19. Goncalves JG, de Braganca AC, Canale D, et al. Vitamin d deficiency aggravates chronic kidney disease progression after ischemic acute kidney injury. PLoS One 2014; 9: e107228.

20. Mirkovic K, van den Born J, Navis G, de Borst MH. Vitamin D in chronic kidney disease: new potential for intervention. Curr.Drug Targets 2011; 12: 42-53.

21. Naves-Diaz M, Alvarez-Hernandez D, Passlick-Deetjen J, et al. Oral active vitamin D is associated with improved survival in hemodialysis patients. Kidney Int. 2008; 74: 1070-1078.

22. Shoben AB, Rudser KD, de Boer IH, Young B, Kestenbaum B. Association of oral calcitriol with improved survival in nondialyzed CKD. J.Am.Soc.Nephrol. 2008; 19: 1613-1619.

23. de Borst MH, Hajhosseiny R, Tamez H, Wenger J, Thadhani R, Goldsmith DJ. Active vitamin D treatment for reduction of residual proteinuria: a systematic review. J.Am.Soc.Nephrol. 2013; 24: 1863-1871.


25. Yilmaz MI, Sonmez A, Saglam M, et al. Ramipril Lowers Plasma FGF-23 in Patients with Diabetic Nephropa-thy. Am.J.Nephrol. 2014; 40: 208-214.

26. de Seigneux S, Courbebaisse M, Rutkowski JM, et al. Proteinuria Increases Plasma Phosphate by Altering Its Tubular Handling. J.Am.Soc.Nephrol. 2014; doi:10.1681/ASN.2014010104.

27. Yuan W, Pan W, Kong J, et al. 1,25-dihydroxyvitamin D3 suppresses renin gene transcription by blocking the activity of the cyclic AMP response element in the renin gene promoter. J.Biol.Chem. 2007; 282: 29821-29830.

28. Li YC, Kong J, Wei M, Chen ZF, Liu SQ, Cao LP. 1,25-Dihydroxyvitamin D(3) is a negative endocrine regulator of the renin-angiotensin system. J.Clin.Invest. 2002; 110: 229-238.

29. Forman JP, Williams JS, Fisher ND. Plasma 25-hydroxyvitamin D and regulation of the renin-angiotensin system in humans. Hypertension 2010; 55: 1283-1288.

30. Thadhani R, Appelbaum E, Pritchett Y, et al. Vitamin D therapy and cardiac structure and function in patients with chronic kidney disease: the PRIMO randomized controlled trial. JAMA 2012; 307: 674-684.

31. Ito I, Waku T, Aoki M, et al. A nonclassical vitamin D receptor pathway suppresses renal fibrosis. J.Clin.Invest. 2013; 123: 4579-4594.

32. Andrukhova O, Slavic S, Zeitz U, et al. Vitamin D is a regulator of endothelial nitric oxide synthase and arterial stiffness in mice. Mol.Endocrinol. 2014; 28: 53-64.

33. Zoccali C, Curatola G, Panuccio V, et al. Paricalcitol and endothelial function in chronic kidney disease trial. Hypertension 2014; 64: 1005-1011.

34. Molina P, Gorriz JL, Molina MD, et al. The effect of cholecalciferol for lowering albuminuria in chronic kidney disease: a prospective controlled study. Nephrol.Dial.Transplant. 2014; 29: 97-109.

35. Brandenburg VM, Floege J. Adynamic bone disease—bone and beyond. NDT Plus 2008; 1: 135-147. 36. Zoccali C, Ruggenenti P, Perna A, et al. Phosphate may promote CKD progression and attenuate renopro-

tective effect of ACE inhibition. J.Am.Soc.Nephrol. 2011; 22: 1923-1930. 37. Floege J, Kim J, Ireland E, et al. Serum iPTH, calcium and phosphate, and the risk of mortality in a European

haemodialysis population. Nephrol.Dial.Transplant. 2011; 26: 1948-1955. 38. Mazzaferro S, Goldsmith D, Larsson TE, Massy ZA, Cozzolino M. Vitamin D metabolites and/or analogs:

which D for which patient? Curr.Vasc.Pharmacol. 2014; 12: 339-349. 39. Drechsler C, Pilz S, Obermayer-Pietsch B, et al. Vitamin D deficiency is associated with sudden cardiac

death, combined cardiovascular events, and mortality in haemodialysis patients. Eur.Heart J. 2010; 31: 2253-2261.

40. Autier P, Gandini S. Vitamin D supplementation and total mortality: a meta-analysis of randomized con-trolled trials. Arch.Intern.Med. 2007; 167: 1730-1737.

39


2

41. Hsia J, Heiss G, Ren H, et al. Calcium/vitamin D supplementation and cardiovascular events. Circulation 2007; 115: 846-854.

42. Tamez H, Zoccali C, Packham D, et al. Vitamin D reduces left atrial volume in patients with left ventricular hypertrophy and chronic kidney disease. Am.Heart J. 2012; 164: 902-9.e2.

43. de Zeeuw D, Agarwal R, Amdahl M, et al. Selective vitamin D receptor activation with paricalcitol for reduction of albuminuria in patients with type 2 diabetes (VITAL study): a randomised controlled trial. Lancet 2010; 376: 1543-1551.

44. Humalda JK, Navis G. Dietary sodium restriction: a neglected therapeutic opportunity in chronic kidney disease. Curr.Opin.Nephrol.Hypertens. 2014; 23: 533-540.

45. Joosten MM, Gansevoort RT, Mukamal KJ, et al. Sodium excretion and risk of developing coronary heart disease. Circulation 2014; 129: 1121-1128.

46. McMahon EJ, Bauer JD, Hawley CM, et al. A randomized trial of dietary sodium restriction in CKD. J.Am.Soc.Nephrol. 2013; 24: 2096-2103.

47. Fan L, Tighiouart H, Levey AS, Beck GJ, Sarnak MJ. Urinary sodium excretion and kidney failure in nondia-betic chronic kidney disease. Kidney Int. 2014; 86: 582-588.


49. Cook NR, Cutler JA, Obarzanek E, et al. Long term effects of dietary sodium reduction on cardiovascular disease outcomes: observational follow-up of the trials of hypertension prevention (TOHP). BMJ 2007; 334: 885-888.

50. Laverman GD, de Zeeuw D, Navis G. Between-patient differences in the renal response to renin-angioten-sin system intervention: clue to optimising renoprotective therapy? J.Renin Angiotensin Aldosterone Syst. 2002; 3: 205-213.

Chapter 3Dietary Sodium Restriction: a Neglected Therapeutic Opportunity in Chronic Kidney DiseaseJelmer K. HumaldaGerjan Navis

Curr Opin Nephrol Hypertens. 2014 Nov;23(6):533-40

Chapter 3

42

Abstract

Purpose of review

Restriction of dietary sodium is recommended at a population level as well as for groups at high cardiovascular risk, and chronic kidney disease (CKD). This review addresses recent evidence for the protective effect of dietary sodium restriction in CKD patients specifically.

Recent findings

Sodium intake in CKD populations is generally high, and often above population average. Recent data demonstrated that moderately lower sodium intake in CKD patients is associated with sub-stantially better long-term outcome of renin-angiotensin-aldosterone system (RAAS)-blockade, in diabetic and non-diabetic CKD, related to better effects of RAAS blockade on proteinuria, independent of blood pressure. This is in line with better short term efficacy of RAAS blockade during moderate sodium restriction in diabetic and nondiabetic CKD. This effect of sodium restriction is likely mediated by its effects on volume status. Sustainable sodium restriction can be achieved by approaches on the basis of behavioral sciences.

Summary

Moderate restriction of dietary sodium can substantially improve the protective effects of RAAS blockade in CKD, by specific renal effects apparent from proteinuria reduction. The latter precludes straightforward extrapolation of data from non-renal populations to CKD. Concerns regarding the adverse effects of a very low sodium intake should not distract from the protec-tive effects of moderate sodium restriction. Prospective studies should assess the efficacy and sustainability of different strategies to target high sodium intake in CKD, along with measures at population level.

Video abstract: http://links.lww.com/CONH/A14

43

Dietary Sodium Restriction

3

Review

Introduction

Restriction of dietary sodium to a maximum of 5 grams of salt (sodium chloride) daily for an adult, corresponding to ca. 2000 mg sodium, is among the top-priorities of the WHO for the combat of chronic noncommunicable diseases (1). In line, for chronic kidney disease (CKD) patients, the 2012 Kidney Disease Improving Global Outcomes guideline recommends the reduction of daily sodium intake to less than 2000 mg/90 mmol per day (2). Salt intake varies widely between dif-ferent countries, but generally exceeds the recommended amount in most communities where data are available. Excess sodium intake is associated with considerable morbidity and mortality, and, hence, substantial costs in terms of health expenditure. It has been estimated that a 3-g reduction of salt intake would reduce healthcare costs by $10–24 billion per year (3). Of note, these estimates were based on cardiovascular disease only: inclusion of the costs related to sodium-related morbidity and mortality in CKD would have increased these figures even further.

Sodium Intake in Chronic Kidney Disease Patients

Among the many studies on CKD patients in the literature, only a minority reports 24-h sodium excretion. In these studies, average 24-h sodium excretion in CKD patients is usually in the range between 160 and 240 mmol (4-7). This is in the same range as in the general population, or even higher. This is remarkable, considering the fact that these patients were under dedicated nephrology care. It could indicate either neglect of sodium status, or failure of current strate-gies in renal care to achieve sustained reduction of dietary sodium, or reflect the association between risk behavior and risk to develop CKD.

Effect of Dietary Sodium Restriction in CKD Patients

In CKD, blood pressure is usually sodium sensitive. Moreover, renal protein loss is reduced by dietary sodium restriction. This has been shown for sodium restriction as a single measure, as well as for sodium restriction as an add-on to antihypertensive treatment (8, 9). Interestingly, proteinuria reduction by sodium restriction remains significant after adjustment for the fall in blood pressure, suggesting an independent renoprotective effect of sodium restriction, both as a single measure and in combination with blockade of the renin-angiotensin-aldosterone system (RAAS) (8). A blood pressure-independent effect of dietary sodium on the kidney is substanti-ated by data in healthy volunteers, in which dietary sodium restriction reduces albuminuria within the normal range, without a detectable effect on blood pressure (10).

Interaction between Sodium Intake and Effects of RAAS Blockade

In CKD, RAAS-blockers are first-line therapy for the treatment of hypertension and proteinuria, and accordingly the vast majority of CKD patients are on maintenance treatment with either an angiotensin-converting-ezyme (ACE) inhibitor or an angiotensin-receptor blocker (ARB).

Chapter 3

44

It has been known for over 2 decades that the effects of RAAS-blockers are blunted by high sodium intake (8, 11, 12). Of note, by interfering with the buffering action of the RAAS on the hemodynamic consequences of altered sodium status, RAAS blockade renders blood pressure sodium-sensitive, thus creating better therapeutic opportunities for sodium restriction, even in previously sodium-resistant patients. Sodium restriction increases the top of the dose-response of RAAS blockade for both blood pressure and proteinuria (13). The effect of moderate sodium restriction during RAAS blockade on blood pressure and proteinuria is approximately similar to the effect of adding a diuretic, whereas the maximum effect is achieved by their combination in nondiabetic (8) as well as diabetic CKD (Figure 1) (5). The reduction in blood pressure and proteinuria is accompanied by a slight decrease in renal function.Of note, in the latter study, habitual sodium intake was very high, that is an average of 224 mmol Na+ /day. During intervention, sodium intake decreased to 148 mmol/day. Although this was still substantially above the recommended levels, nevertheless, it was associated with a significant reduction in blood pressure and proteinuria.

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

CE: A.S; MNH/230612; Total nos of Pages: 8;

MNH 230612

In a head-to-head comparison, moderate dietarysodium restriction during monotherapy ACE-inhibi-tor (ACEi) more effectively reduced blood pressureandproteinuria than adding anARB [12]. Dual RAAS-blockade has now been deemed obsolete, due toworse renal outcome in long-term studies [14–16]related to increased risk for acute renal functiondeterioration. The current data show that moderatedietary sodium restriction as add-on tomonotherapyRAAS-blockade provides a more effective alternativeto dual blockade. Of note, this study also contained adual blockade-sodium restriction arm. This regimenhad potent effects on blood pressure and proteinuria,accompanied by a substantial reduction in renalfunction. The latter was much more pronouncedthan the mild decrease in renal function duringsingle blockade at similar sodium restriction. Thedecrease in renal function was fully reversible afterchanging the regimen, demonstrating its hemody-namicnature,but illustrates thatdualRAAS-blockade

severely compromises the kidneys’ capacity tomain-tain glomerular filtration during sodium restriction,whichmayexplain the increased risk for acute kidneyinjury associated with dual blockade in long-termstudies.

INTERACTION BETWEEN SODIUM INTAKEAND LONG-TERM OUTCOME OF RENIN–ANGIOTENSIN–ALDOSTERONE SYSTEM-BLOCKADE

Recently, the effect of sodium intake on the long-term outcome of RAAS-blockade was analyzed indata sets from the large intervention trials thatprovided the empirical basis for our current RAAS-blockade-based treatment regimens in CKD. Datafrom the ACE-inhibitor arm from the Ramipril Effi-cacy In Nephropathy (REIN) trial were analyzed bytertiles of sodium intake, assessed as urinary sodiumexcretion throughout the study [6]. By design, blood

NaC

l exc

reti

on (

g/d

)

Rec

omm

ende

d in

take

14

12

10

8

6

4

NS SR NS + HCT SR + HCT

ACEi

P < 0.001 P < 0.001

P < 0.001

Pro

tein

uria

(g

/d)

2.0

1.5

1.0

0.5

0.0

NS SR NS + HCT SR + HCT

ACEi

FIGURE 1. Sodium excretion (upper panel) during four different 6-week treatment periods by a rotation schedule in patientswith diabetes and CKD on ACE-inhibitor therapy. NS intake was very high and accordingly also during SR without and withhydrochlorothiazide, sodium intake remained above recommended levels. Nevertheless, proteinuria (lower panel) wasreduced significantly, as was blood pressure (data not shown). CKD, chronic kidney disease; HCT, hydrochlorothiazide; NS,normal sodium; SR, sodium restriction. Adapted from [5&&].

Dietary sodium restriction Humalda and Navis

1062-4821 � 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins www.co-nephrolhypertens.com 3

Figure 1. Sodium excretion (upper panel) during four different 6-week treatment periods by a rotation schedule in patients with diabetes and CKD on ACE-inhibitor therapy. NS intake was very high and accordingly also during SR without and with hydrochlorothiazide, sodium intake remained above recommended levels. Nevertheless, protein-uria (lower panel) was reduced significantly, as was blood pressure (data not shown). CKD, chronic kidney disease; HCT, hydrochlorothiazide; NS, normal sodium; SR, sodium restriction. Adapted from (5**).

45


3

This study also allows inferences on the clinical relevance of the so-called ‘sodium paradox’ in patients with diabetes and CKD. The ‘sodium paradox’ refers to the rise in glomerular filtration rate and filtration fraction that has been reported during low sodium diet in uncomplicated type I diabetes and experimental diabetes. It is partly attributed to excess intrarenal RAAS activity. Concerns regarding the possible adverse renal effects of hyperfiltration may have contributed to the underrated role of sodium restriction in patients with diabetes and CKD, but the current data do not support the relevance of the sodium paradox for the clinical condition of diabetes with CKD during RAAS blockade.

In a head-to-head comparison, moderate dietary sodium restriction during monotherapy ACE-inhibitor (ACEi) more effectively reduced blood pressure and proteinuria than adding an ARB (11). Dual RAAS blockade has now been deemed obsolete, due to worse renal outcome in long-term studies (14-16) related to increased risk for acute renal function deterioration. The current data show that moderate dietary sodium restriction as add-on to monotherapy RAAS blockade provides a more effective alternative to dual blockade. Of note, this study also contained a dual blockade-sodium restriction arm. This regimen had potent effects on blood pressure and proteinuria, accompanied by a substantial reduction in renal function. The latter was much more pronounced than the mild decrease in renal function during single blockade at similar sodium restriction. The decrease in renal function was fully reversible after changing the regimen, demonstrating its hemodynamic nature, but illustrates that dual RAAS blockade severely compromises the kidneys’ capacity to maintain glomerular filtration during sodium restriction, which may explain the increased risk for acute kidney injury (AKI) associated with dual blockade in long-term studies.

Interaction between Sodium Intake and Long-Term Outcome of Renin-Angiotensin-Aldosterone System Blockade

Recently, the effect of sodium intake on long-term outcome of RAAS blockade was analyzed in data sets from the large intervention trials that provided the empirical basis for our current RAAS blockade-based treatment regimens in CKD. Data from the ACE-inhibitor arm from the Ramipril Efficacy In Nephropathy (REIN) trial were analyzed by tertiles of sodium intake, assessed as urinary sodium excretion throughout the study (6). By design, blood pressure was titrated to less than 140/90 mmHg in all patients by adding antihypertensives on top of the ACEi, with a diuretic as the first titration step. Accordingly, blood pressure was not different by sodium intake. However, proteinuria reduction was less effective in the higher tertiles of sodium intake and was associated with a substantially worse long-term renal outcome, with 60% of the patients in the upper tertile reaching the renal end point versus 20% in the lowest tertile (Figure 2).

Chapter 3

46



MNH 230612

pressure was titrated to less than 140/90 mmHg inall patients by adding antihypertensives on topof the ACEi, with a diuretic as the first titrationstep. Accordingly, blood pressure was not differentby sodium intake. However, proteinuria reductionwas less effective in the higher tertiles of sodiumintake and was associated with a substantially worselong-term renal outcome, with 60% of the patientsin the upper tertile reaching the renal end pointversus 20% in the lowest tertile (Fig. 2).

In the Reduction of Endpoints in NIDDM withthe Angiotensin II Antagonist Losartan Study(RENAAL)/Irbesartan Diabetic Nephropathy Trial(IDNT) data set the conventional treatment armwas also analyzed. In these studies, blood pressurewas also titrated by design.With conventional treat-ment, overall outcome was worse, but not differentby sodium intake. In the ARB arm, however, the riskto reach a renal or cardiovascular end point inpatients in the highest tertile was approximatelytwo-fold higher than in the lowest tertile. Accord-ingly, the treatment benefit from the ARB overconventional antihypertensives was present in thelower tertile only (Fig. 3, [4]). In line with the REINdata, the lack of protective action of RAAS-blockadein those ingesting excessive sodium was associatedwith lack of antiproteinuric effect, despite similarblood pressure.

Thus, in CKD patients on RAAS-blockade,a modestly lower dietary sodium intake, in therange recommended for the general population, is

associated with substantial benefits regarding renaland cardiovascular outcome. This difference inoutcome occurs despite adequate blood pressurecontrol in those on the highest sodium intake andis associated with persistence of proteinuria. Ofnote, in these studies, the titration schedule ledto more diuretic use in patients who ingestedexcess sodium. Apparently, this did not sufficientlyprevent the adverse effect of excessive sodiumintake on long-term outcome. This is somewhatdisconcerting, considering the similarity of theshort-term effects of sodium restriction and diu-retics. It could be that control of volume overloadwas still insufficient, as high sodium intake isassociated with diuretic resistance [17] or, theother way round, that diuretic use is associatedwith side-effects that adversely affect long-termoutcome, such as for instance potassium depletionor hyperuricemia [18].

CAN WE GO TOO LOW?

Observational data from various studies, showing aJ-curve between sodium intake and renal and car-diovascular outcome, have raised concern on thesafety of rigorous sodium restriction [19–21]. Theseobservational data should be interpreted withcaution, as a habitual salt intake below 5g daily isa rarity in the outpatient population, and may wellbe an indicator of concomitant conditions com-promising nutritional status as well as outcome.Moreover, quantification of sodium intake wasquestionable in some of the studies, by lack of24-h urine data on sodium intake. This may havecontributed to the substantial differences in thelevel of the nadir of the J-curve. Obviously, thisdebate is seriously hampered by a lack of prospectivelong-term sodium intervention data in CKD. Never-theless, the presence of a J-curve has been reportedin several independent studies (albeit not all [22

&&

])and should be given serious consideration. Inexperimental renal disease, we previously foundthat a regimen of ACEi and rigorous sodium restric-tion reduced blood pressure, proteinuria, andglomerular damage, but unexpectedly aggravatedtubulo-interstitial damage [23]. This was also foundin healthy rats, so, it is not a particularity of themodel but a generalizable adverse effect of the com-bination of RAAS-blockade or rigorous sodiumrestriction. This is disconcerting, as, moreover, theinterstitial damage was not readily apparent fromnoninvasive parameters, so that if this occurred inpatients it would go unnoticed. This once moreunderlines the need for better noninvasive markersof tubulo-interstitial damage. Considering theconsistent association of interstitial damage with

1.0

0.8

0.6

0.4

0.2

0.0

% o

f pat

ien

ts w

ith

ou

t ES

RF

0 12 24 36 48

Months of follow-up

Logrank: P < 0.001

LSD

MSD

HSD

FIGURE 2. Renal survival in patients on ACEi by tertile ofsalt intake LSD 7.1g/day, MSD 10.8g/day, and HSD14.2g/day. ESRF, end stage renal failure; HSD, highsodium diet; LSD, low sodium diet; MSD, medium sodiumdiet. Adapted from [6].

Diagnostics and techniques

4 www.co-nephrolhypertens.com Volume 23 � Number 00 � Month 2014

Figure 2. Renal survival in patients on ACEi by tertile of salt intake LSD 7.1 g/day, MSD 10.8 g/day, and HSD 14.2 g/day. ESRF, end stage renal failure; HSD, high sodium diet; LSD, low sodium diet; MSD, medium sodium diet. Adapted from (6).

In the Reduction of Endpoints in NIDDM with the Angiotensin II Antagonist Losartan (RENAAL)/Irbesartan Diabetic Nephropathy Trial (IDNT) data set the conventional treatment arm was also analyzed. In these studies, blood pressure was also titrated by design. With conventional treatment, overall outcome was worse, but not different by sodium intake. In the ARB arm, however, the risk to reach a renal or cardiovascular end point in patients in the highest tertile was approximately two-fold higher than in the lowest tertile. Accordingly, the treatment benefit from the ARB over conventional antihypertensives was present in the lower tertile only (Figure 3, (4)). In line with the REIN data, the lack of protective action of RAAS blockade in those ingest-ing excessive sodium was associated with lack of antiproteinuric effect, despite similar blood pressure.

Thus, in CKD patients on RAAS blockade, a modestly lower dietary sodium intake, in the range recommended for the general population, is associated with substantial benefits regarding renal and cardiovascular outcome. This difference in outcome occurs despite adequate blood pressure control in those on the highest sodium intake, and is associated with persistence of proteinuria. Of note, in these studies, the titration schedule led to more diuretic use in subjects that ingested excess sodium. Apparently, this did not sufficiently prevent the adverse effect of excessive sodium intake on long-term outcome. This is somewhat disconcerting, considering the similarity of the short-term effects of sodium restriction and diuretics. It could be that control of volume overload was still insufficient, as high sodium intake is associated with diuretic resistance (17) or, the other way round, that diuretic use is associated with side effects that adversely affect long-term outcome, such as for instance potassium depletion, or hyperuricemia (18).

47


3

Can We Go Too Low?

Observational data from various studies, showing a J-curve between sodium intake and renal and cardiovascular outcome, have raised concern on the safety of rigorous sodium restriction (19-21). These observational data should be interpreted with caution, as a habitual salt intake below 5 g daily is a rarity in the outpatient population, and may well be an indicator of concomi-tant conditions compromising nutritional status as well as outcome. Moreover, quantification of sodium intake was questionable in some of the studies, by lack of 24-h urine data on sodium intake. This may have contributed to the substantial differences in the level of the nadir of the J-curve. Obviously, this debate is seriously hampered by lack of prospective long-term sodium intervention data in CKD. Nevertheless, the presence of a J-curve has been reported in several independent studies (albeit not all (22)) and should be given serious consideration. In experimental renal disease we previously found that a regimen of ACEi and rigorous sodium restriction reduced blood pressure, proteinuria, and glomerular damage, but unexpectedly aggravated tubulo-interstitial damage (23). This was also found in healthy rats, so, it is not a particularity of the model but a generalizable adverse effect of the combination of RAAS block-ade and rigorous sodium restriction. This is disconcerting, as, moreover, the interstitial damage was not readily apparent from non-invasive parameters, so if this would occur in patients, it would go unnoticed. This once more underlines the need for better non-invasive markers of



MNH 230612

long-term renal outcome, this could provide apotential explanation for the association of verylow sodium intake with worse renal outcome. Itsmechanisms deserve further explanation, but couldinclude intrarenal ischemia, excess reactive renin oraldosterone activation, or other causes.

MECHANISMS UNDERLYING THEINTERACTION OF SODIUM INTAKE WITHRENIN–ANGIOTENSIN–ALDOSTERONESYSTEM-BLOCKADE EFFICACY

Several mechanisms could explain the interactionbetween sodium intake and the effects of RAAS-blockade. First, high sodium intake suppressesRAAS-activity in the circulation, and blocking asuppressed cascade cannot be expected to havemuch effect. For tissue RAAS-activity other mech-anisms may be relevant as high sodium increasestissue conversion of angiotensin (ang) I [24] andannihilates the effects of ACEi on tissue ang I con-version [25]. Moreover, low sodium potentiates theACEi-induced increase of vasodilator and antifi-brotic substances such as ang 1–7 [26] and AcSDKP[27]. Finally, it is important to realize that highsodium intake is often associated with other nutri-tional factors that can affect renal function and theresponse to RAAS-blockade by themselves, such asfor instance protein intake [10,12,28].

SODIUM INTAKE AND VOLUME STATUSAND THEIR EFFECTS ON RESPONSE TORENIN–ANGIOTENSIN–ALDOSTERONESYSTEM-BLOCKADE

Sodium intake is generally recognized as an import-ant determinant of the extracellular volume (ECV),and it is plausible that the effect of high sodiumintake on the response to RAAS-blockade is to animportant extent determined by its effect on ECV.This is supported by the similarity of the effects ofdiuretic and dietary sodium restriction on RAAS-blockade efficacy (at least on short term) [5

&&

,8,29]and by the predictive effect of volume markers onthe effect of sodium restriction and/or diuretic onthe efficacy of RAAS-blockade [30], as outlined inmore detail in the next paragraph.

In this respect, it is important to realize for agiven sodium intake that volume status can be verydifferent between patients, depending on the avid-ity of sodium retaining mechanisms. The latter canvary by renal condition, withmore avid retention inproteinuric patients [31

&&

], in diabetes, and over-weight or obesity [32]. Accordingly, volume over-load can be present even if sodium intake isadequately restricted, as occurs in severely nephroticpatients as an extreme example. Conversely, volumeexpansion can be absent despite a high sodiumintake, due to highly efficient sodium excretion,

Favors Favors P forARB Non-RAASi trend

Renal outcome

Na:Cr <121 mmol/g 40/173 75/219 0.57 (0.39–0.84)

121 < Na:Cr <153 mmol/g 54/175 72/218 1.00 (0.70–1.42) <0.001

Na:Cr ≥153 mmol/g 56/151 75/241 1.37 (0.96–1.96)

Overall 150/499 222/678 0.92 (0.75–1.14)

Cardiovascular outcome

Na:Cr <121 mmol/g 45/173 64/219 0.63 (0.43–0.92)

121 < Na:Cr <153 mmol/g 62/175 62/218 1.02 (0.73–1.43) 0.021

Na:Cr ≥153 mmol/g 59/151 72/241 1.25 (0.89–1.75)

Overall 166/499 198/678

0.5 1.0 1.5 2.0

0.93 (0.75–1.15)

Hazard ratio (95% CI)

Non-RAASiARB

No. of events/patients Hazard ratio(95% CI)

FIGURE 3. Effect of sodium intake by tertile of urinary Na/creatinine (Cr) on the treatment benefit of ARB for renal (upperpanels) and cardiovascular (lower panels) outcome in patients with type 2 diabetes and nephropathy. Urinary sodiumexcretion in the subsequent tertiles corresponded to a dietary salt intake of 8.9, 10.9, and 12.2g, respectively. ARB,angiotensin-receptor blocker; RAAS, renin–angiotensin–aldosterone system. Adapted from [4].

Dietary sodium restriction Humalda and Navis

1062-4821 � 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins www.co-nephrolhypertens.com 5

Figure 3. Effect of sodium intake by tertile of urinary Na/creatinine (Cr) on the treatment benefit of ARB for re-nal (upper panels) and cardiovascular (lower panels) outcome in patients with type 2 diabetes and nephropathy Urinary sodium excretion in the subsequent tertiles corresponded to a dietary salt intake of 8.9, 10.9 and 12.2 g, respectively. ARB, angiotensin-receptor blocker; RAAS, renin-angiotensin-aldosterone system. Adapted from (4).

Chapter 3

48

tubulo-interstitial damage. Considering the consistent association of interstitial damage with long-term renal outcome, this could provide a potential explanation for the association of very low sodium intake with worse renal outcome. Its mechanisms deserve further explanation, but could include intrarenal ischemia, excess reactive renin or aldosterone activation, or other causes.

Mechanisms Underlying the Interaction of Sodium Intake with Renin-Angiotensin-Aldosterone-Blockade Efficacy

Several mechanisms could explain the interaction between sodium intake and effects of RAAS blockade. First, high sodium intake suppresses RAAS-activity in the circulation, and blocking a suppressed cascade cannot be expected to have much effect. For tissue RAAS-activity other mechanisms may be relevant as high sodium increases tissue conversion of angiotensin (ang) I (24) and annihilates the effects of ACEi on tissue ang I conversion (25). Moreover, low sodium potentiates the ACEi-induced increase of vasodilator and antifibrotic substances such as ang1-7 (26), and AcSDKP (27). Finally, it is important to realize that high sodium intake is often as-sociated with other nutritional factors that can affect renal function and the response to RAAS blockade by themselves, such as for instance protein intake (10, 11, 28).

Sodium Intake and Volume Status and Their Effects on Response to Renin-Angiotensin-Aldosterone System Blockade

Sodium intake is generally recognized as an important determinant of the extracellular volume (ECV), and it is plausible that the effect of high sodium intake on response to RAAS blockade is to an important extent determined by its effect on ECV. This is supported by the similarity of the effects of diuretic and dietary sodium restriction on RAAS blockade efficacy (at least on short term)(5, 8, 29) and by the predictive effect of volume markers on the effect of sodium restriction and/or diuretic on the efficacy of RAAS blockade (30), as outlined in more detail in the next paragraph.

In this respect, it is important to realize for a given sodium intake that volume status can be very different between patients, depending on the avidity of sodium retaining mechanisms. The latter can vary by renal condition, with more avid retention in proteinuric patients (31), in diabetes, and overweight or obesity (32). Accordingly, volume overload can be present even if sodium intake is adequately restricted, as occurs in severely nephrotic patients as an extreme example. Conversely, volume expansion can be absent despite a high sodium intake, due to highly efficient sodium excretion, or highly effective nonosmotic storage mechanism (33). Re-cent data suggest that fibroblast growth factor 23 (FGF-23) is involved in renal sodium retention (34) and volume status (35), thus providing a mechanistic link between the adverse effects of high sodium intake and high phosphate levels (36) on therapy response to RAAS blockade. In

49


3

line with this assumption, high FGF-23 was associated with an impaired response to low sodium diet on top of ACEi (Humalda, Am J Kidney Dis, in press).

Titrating Sodium and Volume Status in Chronic Kidney Disease

Assessment of volume status may be useful to guide therapy in CKD patients on RAAS blockade. In patients with elevated N-terminal pro-brain natriutetic peptide (NT-proBNP), as a marker for volume expansion, sodium intervention by diet, diuretic, or both, reduces blood pressure and proteinuria, whereas a normal NT-proBNP predicts a minor or non-significant effect (Figure 4) (30). Apparently, in the absence of cardiac disease, mild elevation of NT-proBNP indicates subclinical volume expansion as a suitable target for intervention. Accordingly, the volume intervention reduces NT-proBNP. If it normalizes, further volume targeting does not result in a clinical response; if it still above normal, additional volume targeting (e.g., adding diuretic to the diet) results in a clinical response. These data in proteinuric patients suggest that it could be relevant to not only assess sodium intake, but also corresponding volume status, as this could potentially prevent overzealous sodium restriction or diuretic treatment, and its adverse consequences in CKD patients.



MNH 230612

or a highly effective nonosmotic storagemechanism[33]. Recent data suggest that fibroblast growthfactor 23 (FGF23) is involved in renal sodium reten-tion [34

&

] and volume status [35], thus providing amechanistic link between the adverse effects of highsodium intake and high phosphate levels [36] ontherapy response to RAAS-blockade. In line with thisassumption, high FGF23 was associated with animpaired response to low sodium diet on top ofACEi (Humalda, Am J Kidney Dis, in press).

TITRATING SODIUM AND VOLUMESTATUS IN CHRONIC KIDNEY DISEASE

Assessment of volume status may be useful to guidetherapy in CKD patients on RAAS-blockade. Inpatients with elevated N-terminal pro-brain natriu-retic peptide (NT-proBNP), as a marker for volumeexpansion, sodium intervention by diet, diuretic, orboth, reduces blood pressure and proteinuria,whereas a normal NT-proBNP predicts a minor ornonsignificant effect (Fig. 4) [30]. Apparently, inthe absence of cardiac disease, mild elevation ofNT-proBNP indicates subclinical volume expansionas a suitable target for intervention. Accordingly,the volume intervention reduces NT-proBNP. If it

normalizes, further volume targeting does not resultin a clinical response; if it is still above normal,additional volume targeting (e.g., adding diureticto the diet) results in a clinical response. These datain proteinuric patients suggest that it could berelevant to not only assess sodium intake, but alsocorresponding volume status, as this could poten-tially prevent overzealous sodium restriction or diu-retic treatment, and its adverse consequences inCKD patients.

Quantification of ECV is not part of the clinicalroutine in CKD, and validation of the availablevolume markers such as NT-proBNP will probablybe cumbersome, due to interference of cardiac func-tional status. Yet, it might be worthwhile to putmore effort in assessing volume status per se, notonly for its role in modifying the response to RAAS-blockade per se, but also as a possible independentcardiovascular risk factor in CKD patients [37

&

].

MANAGING SODIUM INTAKE IN CHRONICKIDNEY DISEASE: IMPORTANCE OFBEHAVIORAL APPROACHES

How should sodium status be targeted in the clinic?It is increasingly recognized that current strategies

BaselineChange in mean arterial pressure from baseline Change in proteinuria from baseline

NT-proBNP ≤ 125 pg/ml NT-proBNP >125 pg/ml NT-proBNP ≤ 125 pg/ml NT-proBNP >125 pg/ml

0.00

−5.00

−10.00

−15.00

−20.00

−25.00

mm

Hg

g/2

4h

0.00

−1.00

−2.00

−3.00

−4.00

P = 0.23

P = 0.001

P = 0.004

P = 0.25

P = 0.08

P = 0.03

Change from baseline: effect of ARBChange from baseline: effect of ARB + DiureticsChange from baseline: effect of ARB + Diuretics + LS

FIGURE 4. Predictive value of NT-proBNP at baseline (untreated) for the responses of blood pressure (left panel) andproteinuria (right panel) to ARB, add-on sodium restriction, and add-on sodium restriction plus diuretic. NT-proBNP above theupper level of normal (>125pg/ml) was not associated with the response to ARB, but predicted a more pronounced responseto volume intervention by sodium restriction and thiazide. A similar predictive value of NT-proBNP for clinical efficacy ofvolume intervention was found for NT-proBNP during monotherapy ARB (not shown). ARB, angiotensin-receptor blocker; LS,low sodium. Adapted from [32].

Diagnostics and techniques

6 www.co-nephrolhypertens.com Volume 23 � Number 00 � Month 2014

Figure 4. Predictive value of NT-proBNP at baseline (untreated) for the responses of blood pressure (left panel) and proteinuria (right panel) to ARB, add-on sodium restriction, and add-on sodium restriction plus diuretic. NT-proBNP above the upper level of normal (>125pg/ml) was not associated with the response to ARB, but predicted a more pronounced response to volume intervention by sodium restriction and thiazide. A similar predictive value of NT-proBNP for clinical efficacy of volume intervention was found for NT-proBNP during monotherapy ARB (not shown). ARB, angiotensin-receptor blocker; LS, low sodium. Adapted from (32).

Chapter 3

50

Quantification of ECV is not part of the clinical routine in CKD, and validation of the available vol-ume markers such as NT-proBNP will probably be cumbersome, due to interference of cardiac functional status. Yet, it might be worthwhile to put more effort in assessing volume status per se, not only for its role in modifying the response to RAAS blockade per se, but also as a possible independent cardiovascular risk factor in CKD patients (37).

Managing Sodium Intake in Chronic Kidney Disease: Importance of Behavioral Approaches.

How should sodium status be targeted in the clinic? It is increasingly recognized that current strategies to change dietary habits are ineffective, as illustrated by data from the Masterplan study, in which support by trained nurses was effective in improving compliance with pharmaco-logical guidelines, but not in improving compliance with lifestyle measures (7). There is compel-ling evidence from behavioral sciences that sustained lifestyle changes require a dedicated, behavioral approach (38-40). Such approaches are not yet part of the clinical routine in renal care, but are being tested currently (SUBLIME, ClinicalTrials.gov identifier NCT02132013).

Monitoring Sodium Intake from 24-h Urine

The gold standard for assessment of sodium intake is from well collected 24-h urine, as dietary recall and food frequency questionnaires are notoriously unreliable for assessment of sodium intake. This relates to the fact that only 15% of the sodium ingested is added during cooking or during meals, whereas the remainder is present in the food in hidden form, as additives in processed foods (41). As collection of 24-h urine is considered cumbersome by many, and moreover prone to collection errors, these data are not routinely available in many centers. However, assessment of 24-h creatinine excretion allows detection of collection errors by test-ing observed creatinine excretion versus creatinine excretion expected from anthropometric data (42). Moreover, additional nutritional factors relevant to outcome CKD can also be reliably assessed from 24-h urine, such as phosphate intake, protein intake, potassium (43), magnesium (44), sulphate (45), and finally, the absolute value of 24-h creatinine excretion, by its association with muscle mass, is a robust marker of physical fitness, and predictor of mortality risk (46). Thus, 24-h urine can provide a multidimensional nutrition and fitness profile relevant to CKD patients that renders the investment in terms of patient instruction and urine collection even more worthwhile.

Conclusion

Control of sodium and volume status is crucial in the management of diabetic and nondiabetic CKD patients for control of blood pressure and proteinuria, and eventually prevention of pro-gressive renal function loss and its complications. This is particularly so in patients on RAAS blockade, as sodium overload interferes with its therapeutic efficacy, as apparent from persis-tence of proteinuria, even when blood pressure is well controlled. In most CKD patients habitual sodium intake is too high, despite medical supervision. In both diabetic and nondiabetic CKD a

51


3

moderately lower dietary sodium, even at levels substantially above the recommended amount, is associated with a substantially better response to RAAS blockade in short-term interventions, and a substantially better renal and cardiovascular outcome in post hoc analyses of hard end point studies. Concerns have been raised on the safety of rigorous sodium restriction on the basis of a J-curve for sodium intake and outcome, with higher risk not only at higher sodium in-takes, but also the lower end. Safety concerns on rigorous sodium restriction should not distract from the considerable potential benefits of moderate sodium restriction in the vast majority of CKD patients, in whom sodium intake is high or very high. Population measures, including the action of government and industry, are important to facilitate reduction of sodium intake (3). Moreover, it is crucial to develop better strategies for lifestyle management in CKD patients. This should include monitoring of dietary sodium (as well as other relevant dietary factors) from 24-h urine, as well as integration of behavioral approaches into regular care.

Keypoints

– High sodium intake blunts the therapeutic benefit of RAAS blockade on short term and long term in CKD patients.

– The blunting of the long-term benefits of RAAS blockade in CKD relates to persistent protein-uria and cannot be overcome by better blood pressure control by adding antihypertensives.

– Moderate sodium restriction substantially improves the responses to RAAS blockade in CKD, even when the remaining sodium intake is still above recommended levels.

– Sodium intake should be assessed from 24-h urine in patients on RAAS blockade, in particu-lar when therapy response is unsatisfactory.

– Concomitant assessment of volume markers may help to preclude overzealous sodium and volume targeting.

– Behavioral approaches are required to achieve long-term changes in dietary habits.

Acknowledgements

None.

Conflicts of interest

Disclosure of funding: G.N. is advisor for Astra Zeneca. J.K.H. has nothing to disclose.

Chapter 3

52

References

Papers of particular interest, published within the annual period of review, have been high-lighted as:* of special interest** of outstanding interest (1) Beaglehole R, Bonita R, Horton R et al. Priority actions for the non-communicable disease crisis. Lancet

2011;377:1438-1447. (2) Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group. KDIGO 2012 Clinical Practice

Guideline for the Evaluation and Management of Chronic Kidney Disease. Kidney Int.Suppl. 2013;3: 1-150. (3) Bibbins-Domingo K, Chertow GM, Coxson PG et al. Projected effect of dietary salt reductions on future

cardiovascular disease. N.Engl.J.Med. 2010;362:590-599. (4) Lambers Heerspink HJ, Holtkamp FA, Parving HH et al. Moderation of dietary sodium potentiates the renal

and cardiovascular protective effects of angiotensin receptor blockers. Kidney Int. 2012;82:330-337. **(5) Kwakernaak AJ, Krikken JA, Binnenmars SH et al. Effects of sodium restriction and hydrochlorothiazide on

RAAS blockade efficacy in diabetic nephropathy: a randomised clinical trial. Lancet Diabetes Endocrinol. 2014;2:385-395.

First intervention trial showing the effects of sodium restriction, diuretic, and their combination in patients with diabetes and nephropathy on RAAS-blockade. Moderate dietary sodium restriction considerably increased the efficacy of RAAS-blockade. This article also reports on sodium excretion in the unselected recruitment population, showing that sodium intake in this outpatient population of patients with diabetes and ne-phropathy was considerably above the population average in The Netherlands.

(6) Vegter S, Perna A, Postma MJ et al. Sodium intake, ACE inhibition, and progression to ESRD. J.Am.Soc.Nephrol. 2012;23:165-173.

(7) van Zuilen AD, Bots ML, Dulger A et al. Multifactorial intervention with nurse practitioners does not change cardiovascular outcomes in patients with chronic kidney disease. Kidney Int. 2012;82:710-717.

(8) Vogt L, Waanders F, Boomsma F et al. Effects of dietary sodium and hydrochlorothiazide on the antipro-teinuric efficacy of losartan. J.Am.Soc.Nephrol. 2008;19:999-1007.

*(9) McMahon EJ, Bauer JD, Hawley CM et al. A randomized trial of dietary sodium restriction in CKD. J.Am.Soc.Nephrol. 2013;24:2096-2103.

Elegant study demonstrating salt sensitivity in CKD patients as sodium restriction lowered blood pressure, pro-teinuria, and ECV.

(10) Krikken JA, Lely AT, Bakker SJ, Navis G. The effect of a shift in sodium intake on renal hemodynamics is determined by body mass index in healthy young men. Kidney Int. 2007;71:260-265.

(11) Navis G, de Jong PE, Donker AJ et al. Moderate sodium restriction in hypertensive subjects: renal effects of ACE-inhibition. Kidney Int. 1987;31:815-819.

(12) Slagman MC, Waanders F, Hemmelder MH et al. Moderate dietary sodium restriction added to angiotensin converting enzyme inhibition compared with dual blockade in lowering proteinuria and blood pressure: randomised controlled trial. BMJ 2011;343:d4366.

(13) Wapstra FH, Van Goor H, Navis G et al. Antiproteinuric effect predicts renal protection by angiotensin-converting enzyme inhibition in rats with established adriamycin nephrosis. Clin.Sci.(Lond) 1996;90:393-401.

53


3

(14) Mann JF, Schmieder RE, McQueen M et al. Renal outcomes with telmisartan, ramipril, or both, in people at high vascular risk (the ONTARGET study): a multicentre, randomised, double-blind, controlled trial. Lancet 2008;372:547-553.

(15) Fried LF, Emanuele N, Zhang JH et al. Combined Angiotensin inhibition for the treatment of diabetic nephropathy. N.Engl.J.Med. 2013;369:1892-1903.

(16) Parving HH, Brenner BM, McMurray JJ et al. Cardiorenal end points in a trial of aliskiren for type 2 diabe-tes. N.Engl.J.Med. 2012;367:2204-2213.

(17) Wilcox CS, Mitch WE, Kelly RA et al. Response of the kidney to furosemide. I. Effects of salt intake and renal compensation. J.Lab.Clin.Med. 1983;102:450-458.

(18) Bellomo G. Uric acid and chronic kidney disease: A time to act? World J.Nephrol. 2013;2:17-25. (19) Thomas MC, Moran J, Forsblom C et al. The association between dietary sodium intake, ESRD, and all-

cause mortality in patients with type 1 diabetes. Diabetes Care 2011;34:861-866. (20) Ekinci EI, Clarke S, Thomas MC et al. Dietary salt intake and mortality in patients with type 2 diabetes.

Diabetes Care 2011;34:703-709. (21) O’Donnell MJ, Yusuf S, Mente A et al. Urinary sodium and potassium excretion and risk of cardiovascular

events. JAMA 2011;306:2229-2238. **(22) McQuarrie EP, Traynor JP, Taylor AH et al. Association Between Urinary Sodium, Creatinine, Albumin, and

Long-Term Survival in Chronic Kidney Disease. Hypertension 2014; 64:111-117.This study, with 8.5 years of follow-up, unveils that the relation between sodium intake and adverse renal outcome

is independent from, and additive to, proteinuria, moreover, they could not discern a J-curve effect. (23) Hamming I, Navis G, Kocks MJ, van Goor H. ACE inhibition has adverse renal effects during dietary sodium

restriction in proteinuric and healthy rats. J.Pathol. 2006;209:129-139. (24) Boddi M, Poggesi L, Coppo M et al. Human vascular renin-angiotensin system and its functional changes

in relation to different sodium intakes. Hypertension 1998;31:836-842. (25) Kocks MJ, Buikema H, Gschwend S et al. High dietary sodium blunts affects of angiotensin-converting

enzyme inhibition on vascular angiotensin I-to-angiotensin II conversion in rats. J.Cardiovasc.Pharmacol. 2003;42:601-606.

(26) Kocks MJ, Lely AT, Boomsma F et al. Sodium status and angiotensin-converting enzyme inhibition: effects on plasma angiotensin-(1-7) in healthy man. J.Hypertens. 2005;23:597-602.

(27) Kwakernaak AJ, Waanders F, Slagman MC et al. Sodium restriction on top of renin-angiotensin-aldosterone system blockade increases circulating levels of N-acetyl-seryl-aspartyl-lysyl-proline in chronic kidney disease patients. J.Hypertens. 2013;31:2425-2432.

(28) Bellizzi V, Di Iorio BR, De Nicola L et al. Very low protein diet supplemented with ketoanalogs improves blood pressure control in chronic kidney disease. Kidney Int. 2007;71:245-251.

(29) Ekinci EI, Thomas G, Thomas D et al. Effects of salt supplementation on the albuminuric response to telmisartan with or without hydrochlorothiazide therapy in hypertensive patients with type 2 diabetes are modulated by habitual dietary salt intake. Diabetes Care 2009;32:1398-1403.

(30) Slagman MC, Waanders F, Vogt L et al. Elevated N-terminal pro-brain natriuretic peptide levels predict an enhanced anti-hypertensive and anti-proteinuric benefit of dietary sodium restriction and diuretics, but not angiotensin receptor blockade, in proteinuric renal patients. Nephrol.Dial.Transplant. 2012;27:983-990.

**(31) Svenningsen P, Friis UG, Versland JB et al. Mechanisms of renal NaCl retention in proteinuric disease. Acta Physiol.(Oxf) 2013;207:536-545.

Excellent review on mechanisms of sodium retention, for example, activation of ENaC by urinary plasmin in ne-phrotic syndromes.

Chapter 3

54

(32) Visser FW, Krikken JA, Muntinga JH et al. Rise in extracellular fluid volume during high sodium depends on BMI in healthy men. Obesity (Silver Spring) 2009;17:1684-1688.

(33) Machnik A, Neuhofer W, Jantsch J et al. Macrophages regulate salt-dependent volume and blood pressure by a vascular endothelial growth factor-C-dependent buffering mechanism. Nat.Med. 2009;15:545-552.

*(34) Andrukhova O, Slavic S, Smorodchenko A et al. FGF23 regulates renal sodium handling and blood pres-sure. EMBO Mol.Med. 2014; 6:744-759.

First article to demonstrate involvement of bone-mineral disease key-hormone FGF23 to upregulation of NCC and thus fluid retention

(35) Baia LC, Humalda JK, Vervloet MG et al. Fibroblast growth factor 23 and cardiovascular mortality after kidney transplantation. Clin.J.Am.Soc.Nephrol. 2013;8:1968-1978.

(36) Zoccali C, Ruggenenti P, Perna A et al. Phosphate may promote CKD progression and attenuate renopro-tective effect of ACE inhibition. J.Am.Soc.Nephrol. 2011;22:1923-1930.

*(37) Hung SC, Kuo KL, Peng CH et al. Volume overload correlates with cardiovascular risk factors in patients with chronic kidney disease. Kidney Int. 2014;85:703-709.

In this article, volume overload was assessed by bioimpedance in predialysis CKD patients and found to correlate with proteinuria, blood pressure and higher antihypertensive medication use. This article demonstrates that volume status might be objectified by bedside measurements and may have clinical consequences.

(38) Cook NR, Cutler JA, Obarzanek E et al. Long term effects of dietary sodium reduction on cardiovascu-lar disease outcomes: observational follow-up of the trials of hypertension prevention (TOHP). BMJ 2007;334:885-888.

(39) Robare JF, Bayles CM, Newman AB et al. The “10 keys” to healthy aging: 24-month follow-up results from an innovative community-based prevention program. Health Educ.Behav. 2011;38:379-388.

(40) Zhang SX, Guo HW, Wan WT, Xue K. Nutrition education guided by Dietary Guidelines for Chinese Resi-dents on metabolic syndrome characteristics, adipokines and inflammatory markers. Asia Pac.J.Clin.Nutr. 2011;20:77-86.

*(41) Carrigan A, Klinger A, Choquette SS et al. Contribution of food additives to sodium and phosphorus con-tent of diets rich in processed foods. J.Ren.Nutr. 2014;24:13-9, 19e1.

Elegant study quantifying the abundance of sodium in food additives, and thus enforcing the need for structured interventions and aid for CKD patients in order to reduce sodium intake.

(42) Ix JH, Wassel CL, Stevens LA et al. Equations to estimate creatinine excretion rate: the CKD epidemiology collaboration. Clin.J.Am.Soc.Nephrol. 2011;6:184-191.

(43) Smyth A, Dunkler D, Gao P et al. The relationship between estimated sodium and potassium excretion and subsequent renal outcomes. Kidney Int. 2014; doi: 10.1038/ki.2014.214. [Epub ahead of print]

(44) Joosten MM, Gansevoort RT, Mukamal KJ et al. Urinary and plasma magnesium and risk of ischemic heart disease. Am.J.Clin.Nutr. 2013;97:1299-1306.

(45) van den Berg E, Pasch A, Westendorp WH et al. Urinary sulfur metabolites associate with a favor-able cardiovascular risk profile and survival benefit in renal transplant recipients. J.Am.Soc.Nephrol. 2014;25:1303-1312.

(46) Sinkeler SJ, Kwakernaak AJ, Bakker SJ et al. Creatinine excretion rate and mortality in type 2 diabetes and nephropathy. Diabetes Care 2013;36:1489-1494.

Chapter 4Fibroblast Growth Factor 23 and the Antiproteinuric Response to Dietary Sodium Restriction During Renin-Angiotensin-Aldosterone System BlockadeJelmer K. HumaldaHiddo J. Lambers HeerspinkArjan J. KwakernaakMaartje C.J. SlagmanFemke WaandersMarc G. VervloetPieter M. Ter WeeGerjan NavisMartin H. De Borst

on behalf of the NIGRAM Consortium Am J Kidney Dis. 2015 Feb;65(2):259-66.

Chapter 4

58

Abstract

Background: Residual proteinuria during renin-angiotensin-aldosterone system (RAAS) blockade is a major renal and cardiovascular risk factor in chronic kidney disease. Dietary sodium restric-tion potentiates the antiproteinuric effect of RAAS blockade, but residual proteinuria remains in many patients. Previous studies linked high fibroblast growth factor 23 (FGF-23) levels with volume overload; others linked higher serum phosphate levels with impaired RAAS blockade efficacy. We hypothesized that FGF-23 reduces the capacity of dietary sodium restriction to potentiate RAAS blockade, impairing the antiproteinuric effect.

Study Design: Post hoc analysis of cohort data from a randomized crossover trial with two 6-week study periods comparing proteinuria after a regular-sodium diet with proteinuria after a low-sodium diet, both during background angiotensin-converting enzyme inhibition.

Setting & participants: 47 nondiabetic patients with CKD with residual proteinuria ( median protein excretion, 1.9 [IQR, 0.8-3.1] g/d; mean age 50 ± 13 [SD] years; creatinine clearance, 69 [IQR, 50-110] mL/min).

Predictor: Plasma carboxy-terminal FGF-23 levels.

Outcomes: Difference in residual proteinuria at the end of the regular-sodium versus low-sodium study period. Residual proteinuria during low sodium adjusted for proteinuria during regular-sodium diet period.

Results: Higher baseline FGF-23 level was associated with reduced antiproteinuric response to dietary sodium restriction (standardized β = –0.46; P = 0.001; model R2 = 0.71). For every 100-RU/mL increase in FGF-23 level, the antiproteinuric response to dietary sodium restriction was reduced by 10.6%. Higher baseline FGF-23 level was a determinant of more residual proteinuria during low-sodium diet (standardized β = 0.27; P = 0.003) in linear regression analysis adjusted for baseline proteinuria (model R2 = 0.71). There was no interaction with creatinine clearance (P interaction = 0.5). Baseline FGF-23 level did not predict changes in systolic or diastolic blood pressure upon intensified antiproteinuric treatment.

Limitation: Observational study, limited sample size.

Conclusions: FGF-23 levels are associated independently with impaired antiproteinuric response to sodium restriction in addition to RAAS blockade. Future studies should address whether FGF-23–lowering strategies may further optimize proteinuria reduction by RAAS blockade combined with dietary sodium restriction.

59

FGF-23 and Proteinuria Reduction

4

Introduction

Blockade of the renin-angiotensin-aldosterone system (RAAS) retards progressive renal function loss in proteinuric chronic kidney disease (CKD) through lowering blood pressure and protein-uria (1, 2). However, residual proteinuria may remain and the renoprotective efficacy of RAAS blockade-based therapy is incomplete in many patients. The impact of residual proteinuria is reflected by less effective long-term renal and cardiovascular protection in individuals with a poor initial antiproteinuric response (3-6). Accordingly, several strategies have been developed to improve the antiproteinuric response to RAAS blockade, including control of volume overload by dietary sodium restriction and diuretics (7). However, residual proteinuria remains common, particularly in patients who are volume overloaded (8).

Recent data point toward a potential role for phosphate metabolism in resistance to RAAS block-ade. In a post hoc analysis of the REIN (Ramipril Efficacy in Nephropathy) study, higher serum phosphate level was associated independently with worse long-term kidney disease outcome in RAAS blockade of patients with proteinuric CKD (9), yet the underlying mechanism remained unclear. The phosphaturic hormone fibroblast growth factor 23 (FGF-23) plays a key role in regu-lating phosphate metabolism. FGF-23 levels are increased early in CKD, long before an increase in serum phosphorus level happens, in order to keep phosphate balance.(10) Several studies identified associations between higher FGF-23 levels and increased risk of mortality and CKD progression (11-14). Proteinuria emerged as an independent correlate of FGF-23 level in several cohorts (15-17). Interestingly, recent studies suggested that FGF-23 level also is associated with both parameters of volume overload and increased risk of congestive heart failure in patients with CKD (18-20). High FGF-23 levels thus may contribute to antiproteinuric therapy resistance by promoting volume retention.

We therefore hypothesized that patients with CKD with higher FGF-23 levels are less susceptible to dietary sodium restriction in addition to RAAS blockade, resulting in attenuated proteinuria reduction. This hypothesis was addressed in a post hoc analysis of cohort data from a clinical trial comparing residual proteinuria on a regular-sodium versus a low-sodium diet, during back-ground angiotensin-converting enzyme (ACE) inhibition. In a secondary analysis, we addressed whether FGF-23 is associated with proteinuria reduction in response to dual RAAS blockade (angiotensin receptor blockade plus ACE inhibition) compared to ACE-inhibitor monotherapy.

Chapter 4

60

Methods

Study design

We performed a post hoc analysis of cohort data from a clinical trial comparing residual proteinuria on a regular-sodium versus a low-sodium diet. The original trial was a crossover, placebo-controlled, randomized, controlled trial; the protocol has been described in detail previously (21). In short, 52 nondiabetic patients with CKD and residual proteinuria with protein excretion > 1 g/d under maximally dosed ACE inhibition (lisinopril 40 mg/d) underwent sub-sequent 6-week treatment periods with either a regular, liberal sodium diet (regular-sodium group target of 200 mmol/d of sodium [~4800 mg/d of sodium, 12 g/d of sodium chloride]) or a low-sodium diet (low-sodium group, target of 50 mmol/d of sodium [~1200 mg of sodium, or 3 g/d of sodium chloride]), respectively. All patients underwent all study periods in random order. After each 6-week treatment period, patient examination, blood sample and 24-hour urine sample collection took place. Blood and urine samples were stored at –80°C until labora-tory measurement. All patients gave written informed consent. This study was performed in accordance with the Declaration of Helsinki. In the original study, patients underwent 4 study periods in a 2 × 2 factorial design, in which a regular- or low-sodium diet was combined with either ACE-inhibitor monotherapy or ACE-inhibitor plus angiotensin receptor blockade therapy (valsartan 320 mg/d), respectively, in randomized order. We excluded patients who did not adhere to treatment, which was considered to have occurred when a patient demonstrated both higher sodium excretion and increased proteinuria under low-sodium diet compared to regular-sodium diet. One patient met both criteria and was excluded from further analysis. Re-inclusion did not materially change our main findings. No baseline FGF-23 was available for 4 patients; they were excluded from this study.

Laboratory measurements

Plasma FGF-23 was determined in blood samples obtained at the end of each treatment pe-riod using a human FGF-23 enzyme-linked immunosorbent assay directed against the carboxy terminus (Immutopics Inc). Blood and urinary electrolytes, blood lipids and proteins were measured using an automated multianalyzer (Modular; Roche Diagnostics). Parathyroid hor-mone (PTH) was measured in EDTA plasma using radioimmunoassay. As described previously, 25-hydroxyvitamin D3 (25[OH]D3) and 1,25-dihydroxyvitamin D3 (1,25[OH]2D3) levels were deter-mined by isotope-dilution online solid-phase extraction liquid chromatography–tandem mass spectrometry and radioimmunoassay, respectively (22). Proteinuria was measured in 24-hour urine collections with a turbidimetric assay using benzethonium chloride (Modular). Renin and aldosterone were measured with an immunoradiometric assay (CisBio) and radioimmunoassay (Siemens), respectively. Creatinine clearance was calculated form creatinine concentrations in plasma and 24-hour urine collections. To assess RAAS activity, we calculated the aldosterone to

61


4

renin ratio (ARR) (23). Data obtained at the end of the ACE-inhibition plus regular-sodium diet period were considered baseline values.

Statistical analysis

Data are presented as mean ± standard deviation (SD) when normally distributed or median and interquartile range (IQR) when skewed. We used paired t tests to assess differences between treatment periods after natural log transformation when appropriate. Kruskal-Wallis nonpara-metric tests were used to compare data among tertiles of FGF-23 levels. Uni- and multivariable linear regression analyses were used to assess which covariates at baseline (ACE inhibition + regular-sodium diet) were associated with (natural log–transformed) proteinuria after a low-sodium diet (ACE inhibition + low-sodium diet). We chose absolute proteinuria rather than relative change values to avoid data reduction. We included baseline values for systolic and diastolic blood pressure, proteinuria, creatinine clearance, serum FGF-23, serum phosphate, 24-hour phosphate excretion, 24-hour urea excretion (a marker of protein intake), PTH, 25(OH)D3, 1,25(OH)2D3, body mass index, N-terminal probrain natriuretic peptide (NT-proBNP), renin, and aldosterone in our analyses. We constructed multivariable models with possible potential confounders, primarily using contributing variables from univariate regression. We report standardized β and unstandardized coefficients with 95% confidence intervals (CIs). Standard-ized β reflects change in dependent natural log–transformed proteinuria expressed in standard deviations for every 1-SD increase in the independent covariate (ie, FGF-23). Non-normally distributed variables were natural log–transformed when appropriate. We added dummy vari-ables for the 4 possible treatment sequences in the initial models to assess possible carryover effect. We tested for effect modifications by invoking multiplicative interaction terms in the final model. We tested for interaction by covariates that correlated univariately with residual proteinuria (NT-proBNP level, systolic blood pressure, and age) and/or might confound the relation between FGF-23 level and proteinuria (25[OH]D3 level, PTH level, creatinine clearance, and estimated glomerular filtration rate [eGFR; calculcated with the CKD-EPI (CKD Epidemiol-ogy Collaboration) equation]) because vitamin D status or reduced kidney function may affect proteinuria, and FGF-23 and PTH levels are closely related. Second, we repeated the analysis for relative change values (percent proteinuria reduction between the regular- and low-sodium diets). Similar analyses were performed to assess the association between baseline covariates and systolic or diastolic blood pressure after maximum antiproteinuric therapy. We also used linear regression analysis to assess the association between (natural log–transformed) FGF-23 and NTproBNP or ARR values, respectively. A 2-sided P < 0.05 was considered statistically significant. Data were analyzed with PASW Statistics, version 18.0, for Windows (SPSS Inc) and GraphPad Prism, version 5.01 (GraphPad Software Inc).

Chapter 4

62

Results

Study Participants and Effects of Dietary Sodium Restriction

The study cohort consisted of patients with nondiabetic CKD stages 1 to 3 (median creatinine clearance, 69 mL/min) who had a mean age of 50 ± 13 years and mean body mass index of 27.5 ± 4.3 kg/m2; 38 of 47 (81%) were men. Detailed patient characteristics have been published previously (21). For this analysis, we compared the effect of a low-sodium diet during 6 weeks with a regular-sodium diet during 6 weeks, as adjunct to background ACE inhibition. As shown in Figure 1, residual proteinuria decreased markedly during the low-sodium (P<0.001) compared to the regular-sodium diet. Creatinine clearance decreased concomitantly with proteinuria reduction (Table S1, supplementary material). Plasma FGF-23 (P = 0.7), serum phosphate (P = 0.2) levels, and 1,25(OH)2D3 (P = 0.2) levels did not change under the low-sodium versus regular-sodium conditions, whereas phosphate excretion decreased after a low-sodium diet (P = 0.02; Table S1).

FGF-23 and Therapy Response to Dietary Sodium Restriction

We first assessed univariate associations between FGF-23 levels and outcome parameters dur-ing the regular-sodium diet (Table 1). Across tertiles of FGF-23, proteinuria was similar, whereas creatinine clearance was lower (P < 0.001) and values for PTH (P = 0.004), aldosterone (P = 0.01), systolic blood pressure (P = 0.03), and age (P = 0.006) were higher in the higher FGF-23 tertiles.

Next, we investigated associations between FGF-23 at baseline (ie, during a regular-sodium diet) and outcome parameters after 6 weeks of a low-sodium diet. When the low-sodium diet study period was considered, differences in creatinine clearance and PTH persisted across tertiles of baseline FGF-23 (Table S2). Although this did not reach statistical significance (P = 0.08), there was nominally more residual proteinuria during the low-sodium diet period in patients with the highest FGF-23 levels at baseline. Patients with higher baseline FGF-23 levels displayed less pronounced proteinuria reduction (Figures 1 and 2). Patients in the highest FGF-23 tertile dur-ing the regular-sodium diet had higher NT-proBNP levels (P = 0.04), had higher plasma sodium levels (P = 0.01), and tended to have higher diastolic blood pressures (P = 0.05) during the low-sodium diet.

We subsequently used linear regression analysis to assess which parameters at baseline (ie, during the regular-sodium diet) would best predict residual proteinuria during the low-sodium diet. Univariate regression analysis identified baseline proteinuria, NT-proBNP level, FGF-23 level, 25(OH)D3 level, age, and systolic blood pressure as significant correlates (Table 2). In multivariable regression analysis, baseline FGF-23 predicted residual proteinuria independent of baseline proteinuria (Table 3). Adjustment for treatment sequence did not affect results.

63


4

Table 1. Clinical and Laboratory Parameters During ACE Inhibitions Plus Regular-Sodium Diet, by FGF-23 Tertile

Unit 1st Tertile: (80 ‒131 RU/mL)

2nd Tertile: (134 ‒183 RU/mL)

3rd Tertile: (211 ‒556 RU/mL)

Reference value

P-value

Male/female 14/2 10/6 15/1 0.6

Age y 39 (35‒53) 47 (44 ‒57) 58 (48 ‒67) 0.006

BMI kg/m2 24 (22 ‒30) 27 (24 ‒31) 28 (26 ‒33) 0.08

Systolic blood pressure mmHg 129 (118 ‒145) 118 (114 ‒127) 134 (128 ‒150) 0.03

Diastolic blood pressure mmHg 77 (71 ‒88) 75 (68 ‒86) 83 (75 ‒94) 0.3

24-h sodium excretion mmol/d 196 (167 ‒243) 160 (150 ‒197) 161 (146 ‒197) 100 ‒250 0.1

24-h proteinuria g/d 1.8 (0.7 ‒2.8) 1.9 (0.7 ‒3.4) 1.8 (0.9 ‒3.2) <0.3 0.8

Creatinine clearance mL/min 117 (75 ‒164) 67 (51 ‒100) 50 (36 ‒68) <0.001

NT-proBNP ng/L 62 (19 ‒123) 52 (20 ‒231) 159 (32 ‒552) <125 0.2

Plasma sodium mEq/L 140 (138 ‒142) 140 (139 ‒142) 142 (140 ‒143) 132 ‒144 0.05

Serum calcium mg/dL 9.4 (0.0 ‒9.8) 9.4 (9.3 ‒9.7) 9.0 (8.3 ‒10.0) 9.0 ‒11.0 0.9

Serum phosphate mg/dL 3.1 (2.8 ‒3.3) 3.3 (2.9 ‒3.9) 3.4 (3.0 ‒4.1) 2.0 ‒4.0 0.2

24-h phosphate excretion mg/d 1032 (888‒1252) 875 (630‒1056) 906 (681‒1266) <1486 0.2

25(OH)D3 ng/mL 22 (18 ‒30) 18 (12 ‒26) 24 (21 ‒31) >32 0.4

1,25(OH)2D3 pg/mL 41 (26 ‒46) 29 (23 ‒41) 33 (22 ‒39) 15 ‒77 0.3

Parathyroid hormone pg/mL 37 (26 ‒50) 41 (32 ‒53) 76 (49 ‒113) <82 0.004

Renin pg/mL 61 (15 ‒87) 120 (24 ‒971) 40 (21 ‒173) 3.5 ‒28.5 0.3

Aldosterone ng/dL 6.50 (4.33 ‒7.94) 6.14 (2.89 ‒9.39) 13.72 (7.22 ‒15.16) 2.0 ‒23.8 0.01

Aldosterone to renin ratio 2.04 (0.50 ‒2.51) 0.59 (0.13 ‒2.26) 2.63 (0.82 ‒8.46) <50 0.08

Note: Data are depicted per tertile of baseline FGF-23 level. Values for categorical variables are given as num-ber; values for continuous variables as median [interquartile range]. P values are obtained from Kruskal-Wallis test. Conversion factors for units: calcium in mg/dL to mmol/L, ×0.2495; serum phosphate in mg/dL to mmol/L, ×0.323; phosphate excretion in mg/d to mmol/d, ×0.0323; 25(OH)D3 in ng/mL to nmol/L, ×2.496; 1,25(OH)2D3 in pg/mL to pmol/L, ×2.6; aldosterone in ng/dL to nmol/L, × 0.0277. Abbreviations: 25(OH)D3, 25-hydroxyvitamin D3; 1,25(OH)2D3, 1,25-dihydroxyvitamin D3; ACE, angiotensin-converting enzyme; BMI, body mass index; BP, blood pressure; FGF-23, fibroblast growth factor 23; NT-proBNP, N-terminal probrain natriuretic peptide.

Figure 1. Proteinuria after regular- (RS) or low-sodium (LS) diet, per tertile of fibroblast growth factor 23 (FGF-23). Mean proteinuria (standard deviation) in the full cohort and depicted per tertile of FGF-23 during the RS diet.

Chapter 4

64

We found no evidence of interaction between FGF-23 level and residual proteinuria by NT-proBNP level, creatinine clearance, 25(OH)D3 level, systolic blood pressure, or PTH level (all P for interaction > 0.1). There was also no interaction when we used the eGFR instead of creatinine clearance (P for interaction = 0.9). NT-proBNP level correlated with FGF-23 level during both the regular-sodium (r = 0.323; P=0.03) and low-sodium diet (r = 0.502; P < 0.001) study periods.

We also analyzed the relative proteinuria reduction achieved by a low-sodium diet. In univariate regression analysis, only FGF-23 level (r = –0.394; P = 0.01; Figure 1) and diastolic blood pres-sure (r = 0.357; P = 0.02) were associated with the antiproteinuric response. In multivariable analysis, both FGF-23 level (standardized β = –0.464; P = 0.001) and diastolic blood pressure (standardized β = 0.432; P=0.002) contributed to this final model (R2 = 0.34). Again, no evidence of interaction with creatinine clearance was found. The B, or effect size, of FGF-23 was –0.106 (95% CI, –0.164 to –0.047). This suggests that for every 100-relative units (RU)/mL greater FGF-23 level, the antiproteinuric efficacy of a low sodium diet is decreased by 10.6%.

We performed similar analyses to assess the influence of baseline FGF-23 level on changes in blood pressure. In multivariable regression models, only baseline systolic blood pressure (standardized β = 0.879; P < 0.001), but neither baseline FGF-23 (standardized β = –0.005; P = 0.9) nor the other previously used covariates, was a significant determinant of systolic blood pressure after a low-sodium diet. Results were similar for diastolic blood pressure (FGF-23 standardized β = –0.032; P=0.8).

Secondary Analyses

The ARR showed a positive trend with borderline statistical significance among FGF-23 tertiles during regular- (P = 0.08; Table 2) and low-sodium diets (P = 0.05; Table 3). Furthermore, FGF-23 levels during the regular-sodium diet were associated with ARR during the low-sodium diet

Figure 2. Baseline fibroblast growth factor 23 (FGF-23) and proteinuria reduction after the low-sodium diet. Uni-variate linear regression fit line for the correlation of FGF23 level during regular-sodium (RS) diet and percent pro-teinuria reduction (UPR): r = –0.394, P = 0.01. Dashed lines delineate the 95% confidence interval for the regression line.

65


4

(r = 0.330; P = 0.03). Subsequent testing for effect modifications suggested interaction in the association between FGF-23 level and residual proteinuria by the ARR (P for interaction = 0.03). In another secondary analysis, we assessed whether FGF-23 also was associated with the anti-proteinuric response to angiotensin receptor blockade when added to baseline ACE inhibition. Clinical and laboratory parameters during dual RAAS blockade (on the regular-sodium diet) are presented in Table S3. The antiproteinuric response to dual blockade was less pronounced than the response to sodium restriction (Table S1). Accordingly, only baseline proteinuria was associ-ated with residual proteinuria in multivariable analysis (Table S4).

Discussion

The presence of residual proteinuria despite optimally dosed RAAS blockade contributes to long-term adverse kidney disease and cardiovascular outcomes (3-6) Therefore, it is pivotal to identify determinants of therapy resistance in patients with CKD. In the current study, the key finding is that higher plasma FGF-23 levels independently predicted impaired antiproteinuric

Table 2. Univariate Linear Regression Analysis: Determinants During Regular-Sodium, Diet of Residual Proteinuria After the Switch to a Low-Sodium Diet

Determinant Standardized β P-value

Ln( proteinuria) 0.802 <0.001

NT-proBNP (in ng/L) 0.494 0.001

FGF-23 (in RU/mL) 0.469 0.001

Systolic blood pressure (in mmHg) 0.398 0.007

Age (in y) 0.380 0.01

25(OH)D3 (in ng/mL) –0.330 0.04

BMI (in kg/m2) –0.292 0.06

Parathyroid hormone (in pg/mL) 0.298 0.07

Serum phosphate (in mg/dL) 0.297 0.07

Diastolic blood pressure (in mmHg) 0.274 0.09

Creatinine clearance (in mL/min) –0.240 0.1

24-h phosphate excretion (in mg/d) –0.196 0.2

Serum calcium (in mg/dL) –0.244 0.3

Sex 0.136 0.4

Aldosterone (in ng/dL) 0.126 0.4

Renin (in pg/mL) 0.121 0.4

1,25(OH)2D3 –0.118 0.5

Plasma sodium (in mEq/L) 0.033 0.8

Abbreviations: 25(OH)D3, 25-hydroxyvitamin D3; 1,25(OH)2D3, 1,25-dihydroxyvitamin D3; BMI, body mass index; BP, blood pressure; FGF-23, fibroblast growth factor 23; NT-proBNP, N-terminal probrain natriuretic peptide.

Chapter 4

66

response to a low-sodium diet intervention added to background ACE inhibition. In other words, higher FGF-23 level was associated with resistance to volume depletion as a strategy to potenti-ate RAAS blockade – the current standard therapy in CKD.

Zoccali et al. previously demonstrated an association between higher serum phosphate level and worse long-term kidney disease outcome of ACE inhibition in proteinuric patients with CKD in the REIN study (9), suggesting interference of phosphate metabolism with RAAS blockade efficacy. Unfortunately, in the REIN study, no biobank was available to measure FGF-23 (9). In our study, phosphate levels were in the normal range and were not associated with residual proteinuria. The normal phosphate levels in our study correspond to the less advanced stage of kidney function loss as compared to the REIN study. In earlier stages of CKD, FGF-23 is a more sensitive parameter of altered phosphate metabolism (10). Accordingly, our study is in agree-ment with the findings by Zoccali et al. (9), indicating that deregulated phosphate metabolism may affect the efficacy of RAAS blockade–based therapy.

FGF-23 level was associated with reduced antiproteinuric response to sodium restriction and performed stronger and more consistently than other correlates in our multivariable analyses.

Table 3. Multivariable Linear Regression Analysis: Determinants During Regular-Sodium Diet of Residual Protein-uria After the Switch to a Low-Sodium Diet

Baseline determinant B (95% CI) Standardized β P-value R2

Model 1

Proteinuria, per 1-log unit greater 0.787 (0.599 to 0.974) 0.729 <0.001 0.71

FGF-23, per 10-RU/mL greater 0.022 (0.008 to 0.036) 0.272 0.003

Model 2



Age, per 1-y older –0.004 (–0.018 to 0.010) –0.060 0.5

Female sex –0.156 (–0.542 to 0.230) –0.072 0.4

Model 3



Creatinine clearance, per 10-mL/min greater 0.011 (–0.028 to 0.050) 0.065 0.6

NT-proBNP, per 10-ng/L greater 0.005 (–0.001 to 0.001) 0.185 0.1

25(OH)D3 per 1-ng/mL greater 0.009 (–0.010 to 0.027) 0.102 0.3

Systolic BP, per 1-mmHg greater –0.012 (–0.023 to 0.003 –0.238 0.05

Note: In table, log refers to natural logarithm. Abbreviations and definitions: 25(OH)D3, 25-hydroxyvitamin D3; 95% CI, 95% confidence interval; B, unstandardized coefficient; β, standardized coefficient; Baseline determinants, base-line determinants of the natural log of the residual proteinuira after low-sodium diet; BP, blood pressure; FGF-23, fibroblast growth factor 23; NT-proBNP, N-terminal probrain natriuretic peptide.

67


4

The association between FGF-23 level and impaired therapy response could not be reproduced by exchanging FGF-23 level for other markers of deranged bone-mineral metabolism (PTH, phosphorus, or vitamin D level). We could not demonstrate an interaction between FGF-23 level and creatinine clearance or eGFR. These findings suggest that FGF-23 has a specific effect on therapy response to volume depletion during background RAAS blockade. The association between FGF-23 level and residual proteinuria was not observed for dual RAAS blockade during the regular-sodium diet, which could be explained by the comparatively small antiproteinuric effect of dual RAAS blockade versus sodium restriction (Table S1). Whether FGF-23 level pre-dicts the response to dual RAAS blockade should be tested in a population with a stronger antiproteinuric effect by dual RAAS blockade.

The mechanism underlying the association of FGF-23 level with residual proteinuria is of interest, but cannot be derived with certainty from our study. However, volume overload, a well-known determinant of therapy resistance (24-26), could play a role. Recent studies found associations between FGF-23 levels and parameters of volume status and with an increased risk of congestive heart failure in patients with CKD (18-20). Of note, FGF-23 may directly affect renal sodium transport to promote sodium retention and increase blood pressure (27). Thus, higher FGF-23 levels may be linked to a state of volume overload that can insufficiently be overcome by dietary sodium restriction, resulting in more residual proteinuria. Alternatively, the degree of preexisting kidney damage is a known determinant of therapy response (28). Because the association with residual proteinuria was independent of kidney function and proteinuria and FGF-23 affects tubular phosphate transport, FGF-23 level might be considered a marker of tubulointerstitial damage that could affect therapy response. In line, reduced renal expression of the FGF-23 co-receptor klotho could have contributed to higher FGF-23 levels at least in part; renal klotho loss also may contribute to deregulated sodium homeostasis through promoting hyperaldosteronism (29). Finally, interactions between FGF-23 and the (intrarenal) RAAS may influence the response to antiproteinuric therapy. FGF-23 lowers 1,25(OH)2D3 levels by increas-ing the enzyme 24-hydroxylase and suppressing 1α-hydroxylase (30). Because active vitamin D suppresses renin synthesis, high FGF-23 levels in this way might lead to increased intrarenal RAAS-activation (31). However, in our analyses, a role for 1,25(OH)2D3 was not apparent. In this clinical study, no direct assessment of intrarenal RAAS activity could be obtained. Indirect assessment of RAAS blockade efficacy nevertheless supports possible interference of FGF-23 with the RAAS. During the low-sodium diet the ARR consistently increased per FGF-23 tertile, indicating higher aldosterone levels for a given renin level and hence less effective pharmaco-logic RAAS blockade. Part of this dissociation of renin and aldosterone may be explained by ACE2 (angiotensin I–converting enzyme 2), which counteracts RAAS activation by degrading both angiotensin I and II (32). ACE2 neoexpression has been seen in the renal endothelium of patients with CKD (33), and recent experimental data suggest that FGF-23 suppresses renal ACE2 expression (34), resulting in increased angiotensin II generation, which could translate

Chapter 4

68

into a higher ARR. Our finding that the association between baseline FGF-23 level and residual proteinuria after ACE inhibition plus a low-sodium diet depends on the ARR (P for interaction = 0.03) supports our hypothesis that FGF-23 impairs the therapy response through interaction with the RAAS.

To our knowledge, this is the first clinical study to investigate the relationship between FGF-23 and therapy response. Strengths include its crossover design, which increases the power of our findings and rigorous statistical analysis. Moreover, we were able to assess the role of several potential confounders. This study also has limitations, the first being its post hoc nature prone to unmeasured confounding. It must be stated that the effect size in our analyses is modest, where addition of baseline FGF-23 increases the explained variance of residual proteinuria from 64% to 71%, with baseline proteinuria (ie, during the regular-sodium diet) contributing most strongly. Furthermore, our population was heterogeneous in terms of type of underlying disease; nevertheless, positive findings could be demonstrated consistently. The number of patients with very high FGF-23 levels (>300 RU/mL) was small; further studies are needed to confirm the association between FGF-23 and therapy response across the spectrum of CKD. Our study focused on a population with residual proteinuria during monotherapy RAAS blockade because this is the most relevant population for studies of therapy resistance. This preselection, and accordingly monotherapy ACE inhibition as baseline condition, precludes conclusions on interference of FGF-23 with RAAS blockade per se, and studies including an untreated baseline value would be needed to that purpose. Notwithstanding, our results imply that FGF-23 is a relevant factor that attenuates the efficacy of ACE inhibition and dietary sodium restriction, that is, standard treatment in accordance with current guidelines (35). Another limitation is that our sample population was entirely of Northern European ancestryt; therefore, our findings cannot be generalized to other ethnic populations. We could not account for seasonal variations (eg, vitamin D levels), and the design of our study precluded long-term follow-up and hard end points.

In conclusion, higher baseline FGF-23 levels were independently associated with resistance to intensified antiproteinuric therapy, consisting of dietary sodium restriction in addition to baseline ACE-inhibition. Future studies should prospectively investigate whether reduction of FGF-23 levels improves the therapy response to (intensified) RAAS blockade and improves renal and cardiovascular prognosis for patients with CKD.

Acknowledgements

The NIGRAM (Nier-Gerichte Research: van Arterie naar Mens) Consortium comprises Pieter M. ter Wee, Marc G. Vervloet (VU University Medical Center, Amsterdam, the Netherlands); René J.

69


4

Bindels, Joost G. Hoenderop (Radboud University Medical Center Nijmegen, the Netherlands); Gerjan Navis, Jan-Luuk Hillebrands, and Martin H. de Borst (University Medical Center Gronin-gen, the Netherlands).Part of this work was presented at the American Society of Nephrology Kidney Week, October 30-November 4, 2012, San Diego, CA (FR-OR067).

Support

This work is supported by a consortium grant from the Dutch Kidney Foundation (NIGRAM con-sortium grant CP10.11). Dr de Borst is supported by grants from the Dutch Kidney Foundation (KJPB.08.07) Drs. de Borst and Lambers Heerspink are supported by the Netherlands Organization for Scientific Research (Veni grants). The funding sources had no role in the design and conduct of the study; the collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or the decision to submit the report for publication. The original study was supported by an unrestricted grant from Novartis (CVAL489ANL08).

Financial Disclosure

The authors declare that they have no other relevant financial interests.

Chapter 4

70

Supplementary material

Table S1. Clinical and Laboratory Parameters per Study Period

Variables Unit ACEi+ regular sodium

ACEi+ low sodium

ACEi+ARB+RS Reference value

Clinical and general laboratory parameters

Systolic blood pressure mmHg 131±18 122±17* # 130±21

Diastolic blood pressure mmHg 79±13 73±12* # 77±14

Creatinine clearance mL/min 69 (50-110) 70(43-92)* #¥ 71(53-103)

Proteinuria g/d 1.9 (0.8-3.1) 0.8 (0.5-1.4)* # 1.5 (0.6-3.3)* <0.3

24-h sodium excretion mmol/d 185±52 106±48* # 179±69 100-250

Plasma sodium mEq/L 141±3 139±3* # 141±3 132-144

Body weight kg 88.3±16.9 85.4±15.9* # 87.9±16.2

Renin pg/mL 59 (23-181) 168 (44-386)* # 57 (25-230)* 3.5-28.5

Aldosterone ng/dL 7.22 (4.33-13.18) 12.27 (6.50-17.33)* # 7.22 (5.05-10.11) 2.0-23.8

Aldosterone to renin ratio pg/mL/pg/mL 1.40 (0.49-3.38) 0.69 (0.23-2.60)* 1.05 (0.23-3.09)* <50

Mineral metabolism parameters

Serum phosphate mg/dL 3.28±0.64 3.25±0.58 3.28±0.61 2.0-4.0

24-h phosphate excretion mg/d 913 (705-1158) 812 (686-1008)* 1002 (692-1223) <1486

Serum calcium$ mg/dL 9.34±0.51 9.45±0.43# 9.37±0.54 9.0-11.0

FGF-23 RU/mL 146 (119-244) 156 (112-221) 147 (106-205) <125

25(OH)-vitamin D3 ng/mL 23±9 24±9 23±11 >32

1,25(OH)2-vitamin D3 pg/mL 33±12 31±10 15-77

Parathyroid hormone pg/mL 48 (34-77) 44 (31-75) 51 (30-102) <82

NT-proBNP ng/L 64 (24-223) 50 (21-175)* 69 (14-201) <125

Data are presented as mean ± standard deviation or median (IQR) as appropriate. * = P<0.05 vs. ACEi+RS, # = P<0.05 vs. ACEi+ARB+RS;. Abbreviations: ACEi = angiotensin converting enzyme inhibitor, NT-proBNP= NT-probrain natriuretic peptide. $Calcium data were available for 21, 18, and 17 patients per treatment group, respectively. Con-version factors for units: aldosterone in ng/dL to nmol/L, × 0.0277; serum phosphate in mg/dL to mmol/L, ×0.323; phosphate excretion in mg/d to mmol/d; ×0.0323, serum calcium in mg/dL to mmol/L, × 0.25; 25(OH)-vitamin D3 in ng/mL to nmol/L, ×2.496; 1,25(OH)2-vitamin D3 in pg/mL to pmol/L, ×2.6. ¥Mean±SD creatinine clearance in LS+ACEi group vs RS+ACEi group: 82±44 vs 71±35 mL/min, respectively.

71


4

Table S2. Clinical and Laboratory Parameters during ACEi + Low Sodium Diet




Reference value

P-value

Gender (male/female) 14/2 10/6 15/1 0.6

Systolic blood pressure mmHg 121 (106-134) 113 (109-120) 129 (119-139) 0.2

Diastolic blood pressure mmHg 71 (61-75) 71 (66-79) 81 (66-84) 0.05

24-h sodium excretion mmol/d 120 (63-139) 86 (74-135) 93 (67-117) 100-250 0.5

24-h proteinuria g/d 0.5 (0.2-1.0) 0.9 (0.5-1.2) 0.9 (0.5-1.7) <0.3 0.08

Proteinuria reduction % 66 (42-73) 51 (17-63) 38 (31-54) 0.06

Creatinine clearance mL/min 92 (75-110) 68 (50-85) 44 (31-64) 0.001

NT-proBNP ng/L 26 (16-54) 48 (20-221) 99 (40-322) <125 0.03

Plasma sodium mEq/l 139 (138-141) 139 (136-140) 141 (139-143) 132-144 0.01

Serum calcium mg/dL 9.5 (9.4-10.0) 9.6 (9.3-9.8) 9.1 (8.8-9.4) 9.0-11.0 0.05

Serum phosphate mg/dL 3.4 (2.8-3.9) 3.1 (3.0-3.5) 3.6 (3.0-4.0) 2.0-4.0 0.7

24-h phosphate excretion mg/d 979 (715-1196) 793 (602-975) 764 (686-889) <1486 0.1

25(OH)-vitamin D3 ng/mL 23 (13-33) 22 (17-30) 24 (17-38) >32 0.7

1,25(OH)2-vitamin D3 pg/mL 31 (24-37) 34 (23-44) 31 (25-35) 15-77 0.7

Parathyroid hormone pg/mL 31 (22-42) 43 (28-73) 71 (53-95) <82 0.001

Renin pg/mL 314 (138-713) 126 (43-1240) 82 (39-174) 3.5-28.5 0.07

Aldosterone ng/dL 9.39 (5.78-16.61) 12.64 (6.50-19.86) 13.00 (7.76-27.80) 2.0-23.8 0.3

Aldosterone to renin ratio

pg/mL/pg/mL 0.30 (0.22-0.58) 0.71 (0.09-4.35) 1.56 (0.90-3.96) <50 0.05

Data are depicted per tertile of baseline FGF-23. Data are presented as median (IQR). P-values are obtained from Kruskal-Wallis test. Conversion factors for units: serum calcium in mg/dL to mmol/L, ×0.2495; serum phosphate in mg/dL to mmol/L, ×0.323; phosphate excretion in mg/d to mmol/d; ×0.0323, 25(OH)-vitamin D3 in ng/mL to nmol/L, ×2.496; 1,25(OH)2-vitamin D3 in pg/mL to pmol/L, ×2.6; aldosterone in ng/dL to nmol/L, × 0.0277.

Chapter 4

72

Table S3. Clinical and Laboratory Parameters of the Study Participants during ACEi + ARB + Regular Sodium Diet




Reference value

P -va l -ue

Gender (male/female) 14/2 10/6 15/1 0.6

Systolic blood pressure mmHg 126 (111-136) 116 (106-133) 140 (131-150) 0.03

Diastolic blood pressure mmHg 73 (64-84) 71 (63-90) 85 (76-88) 0.1

24-h sodium excretion mmol/d 187 (120-231) 184 (146-206) 175 (101-224) 100-250 0.9

24-h proteinuria g/d 0.9 (0.3-2.7) 2.0 (1.0-3.7) 1.6 (0.8-3.3) <0.3 0.2

Proteinuria reduction % 30 (–11-53) 23 (–9-37) 15 (–27-29) 0.2

Creatinine clearance mL/min 102 (68-139) 77 (57-95) 51 (39-72) 0.001

NT-proBNP ng/L 52 (7-122) 67 (14-171) 127 (20-376) <125 0.2

Plasma sodium mEq/L 141 (140-142) 141 (138-142) 141 (140-143) 132-144 0.5

Serum calcium mg/dL 9.2 (9.2-9.6) 9.7 (9.3-10.0) 9.4 (8.3-9.6) 9.0-11.0 0.3

Serum phosphate mg/dL 3.24 (2.83-3.49) 3.31 (2.88-3.68) 3.53 (2.73-3.97) 2.0-4.0 0.4

24-h phosphate excretion mg/d 1026 (699-1462) 856 (671-1172) 966 (636-1099) <1486 0.2

25(OH)-vitamin D3 ng/mL 22 (15-26) 18 (12-27) 26 (19-34) >32 0.4

1,25(OH)2-vitamin D3 pg/mL 43 (33-47) 35 (23-49) 31 (21-42) 15-77 0.1

Parathyroid hormone pg/mL 41 (28-57) 48 (31-69) 94 (52-126) <82 0.1

Renin pg/mL 90 (26-276) 142 (20-708) 55 (29-204) 3.5-28.5 0.6

Aldosterone ng/dL 5.05 (4.33-7.22) 6.50 (2.89-9.39) 10.11 (7.22-17.33) 2.0-23.8 0.003

Aldosterone to renin ratio pg/ml/pg/mL 0.48 (0.21-2.19) 0.71 (0.05-4.80) 2.12 (0.84-4.98) <50 0.2

Data are depicted per tertile of baseline FGF-23. Data are presented as median (IQR). P-values are obtained from Kruskal-Wallis test. Conversion factors for units: serum calcium in mg/dL to mmol/L, ×0.2495; serum phosphate in mg/dL to mmol/L, ×0.323; phosphate excretion in mg/d to mmol/d; ×0.0323, 25(OH)-vitamin D3 in ng/mL to nmol/L, ×2.496; 1,25(OH)2-vitamin D3 in pg/mL to pmol/L, ×2.6; aldosterone in ng/dL to nmol/L, × 0.0277.

Table S4. Multivariable Linear Regression Analysis: Determinants during RS of Residual Proteinuria during Dual RAAS Blockade

Baseline determinants of the Ln residual proteinuria after dual RAAS blockade

Model No. Baseline determinant B 95% CI B Standardized β P-value R2

Univariate 1 Proteinuria, per Ln (g/d) 1.113 0.912-1.314 0.863 <0.001 0.74

+FGF-23 2 Proteinuria, per Ln (g/d) 1.096 0.890-1.302 0.849 <0.001 0.75

FGF-23, per 10 RU/mL 0.006 –0.009-0.021 0.064 0.4

73


4

References

1. The GISEN Group (Gruppo Italiano di Studi Epidemiologici in Nefrologia). Randomised placebo-controlled trial of effect of ramipril on decline in glomerular filtration rate and risk of terminal renal failure in protein-uric, non-diabetic nephropathy. Lancet 1997; 349: 1857-1863.

2. Giatras I, Lau J, Levey AS. Effect of angiotensin-converting enzyme inhibitors on the progression of nondia-betic renal disease: a meta-analysis of randomized trials. Angiotensin-Converting-Enzyme Inhibition and Progressive Renal Disease Study Group. Ann.Intern.Med. 1997; 127: 337-345.

3. Schmieder RE, Mann JF, Schumacher H, et al. Changes in albuminuria predict mortality and morbidity in patients with vascular disease. J.Am.Soc.Nephrol. 2011; 22: 1353-1364.


5. Laverman GD, de Zeeuw D, Navis G. Between-patient differences in the renal response to renin-angioten-sin system intervention: clue to optimising renoprotective therapy? J.Renin Angiotensin Aldosterone Syst. 2002; 3: 205-213.

6. Lambers Heerspink HJ, de Borst MH, Bakker SJ, Navis GJ. Improving the efficacy of RAAS blockade in patients with chronic kidney disease. Nat.Rev.Nephrol. 2013; 9: 112-121.


8. Hung SC, Kuo KL, Peng CH, et al. Volume overload correlates with cardiovascular risk factors in patients with chronic kidney disease. Kidney Int. 2014;85(3):703-709.

9. Zoccali C, Ruggenenti P, Perna A, et al. Phosphate may promote CKD progression and attenuate renopro-tective effect of ACE inhibition. J.Am.Soc.Nephrol. 2011; 22: 1923-1930.

10. Isakova T, Wahl P, Vargas GS, et al. Fibroblast growth factor 23 is elevated before parathyroid hormone and phosphate in chronic kidney disease. Kidney Int. 2011; 79: 1370-1378.


12. Fliser D, Kollerits B, Neyer U, et al. Fibroblast growth factor 23 (FGF23) predicts progression of chronic kidney disease: the Mild to Moderate Kidney Disease (MMKD) Study. J.Am.Soc.Nephrol. 2007; 18: 2600-2608.


14. Wolf M, Molnar MZ, Amaral AP, et al. Elevated fibroblast growth factor 23 is a risk factor for kidney transplant loss and mortality. J.Am.Soc.Nephrol. 2011; 22: 956-966.

15. Ix JH, Shlipak MG, Wassel CL, Whooley MA. Fibroblast growth factor-23 and early decrements in kidney function: the Heart and Soul Study. Nephrol.Dial.Transplant. 2010; 25: 993-997.

16. Vervloet MG, van Zuilen AD, Heijboer AC, et al. Fibroblast growth factor 23 is associated with proteinuria and smoking in chronic kidney disease: An analysis of the MASTERPLAN cohort. BMC Nephrol. 2012; 13: 20.

17. Lundberg S, Qureshi AR, Olivecrona S, Gunnarsson I, Jacobson SH, Larsson TE. FGF23, albuminuria, and disease progression in patients with chronic IgA nephropathy. Clin.J.Am.Soc.Nephrol. 2012; 7: 727-734.

18. Baia LC, Humalda JK, Vervloet MG, et al. Fibroblast growth factor 23 and cardiovascular mortality after kidney transplantation. Clin.J.Am.Soc.Nephrol. 2013; 8: 1968-1978.


Chapter 4

74



22. Doorenbos CR, de Cuba MM, Vogt L, et al. Antiproteinuric treatment reduces urinary loss of vitamin D-binding protein but does not affect vitamin D status in patients with chronic kidney disease. J.Steroid Biochem.Mol.Biol. 2012; 128: 56-61.

23. Trenkel S, Seifarth C, Schobel H, Hahn EG, Hensen J. Ratio of serum aldosterone to plasma renin con-centration in essential hypertension and primary aldosteronism. Exp.Clin.Endocrinol.Diabetes 2002; 110: 80-85.

24. Buter H, Hemmelder MH, Navis G, de Jong PE, de Zeeuw D. The blunting of the antiproteinuric efficacy of ACE inhibition by high sodium intake can be restored by hydrochlorothiazide. Nephrol.Dial.Transplant. 1998; 13: 1682-1685.

25. Esnault VL, Ekhlas A, Delcroix C, Moutel MG, Nguyen JM. Diuretic and enhanced sodium restriction results in improved antiproteinuric response to RAS blocking agents. J.Am.Soc.Nephrol. 2005; 16: 474-481.

26. Ekinci EI, Thomas G, Thomas D, et al. Effects of salt supplementation on the albuminuric response to telmisartan with or without hydrochlorothiazide therapy in hypertensive patients with type 2 diabetes are modulated by habitual dietary salt intake. Diabetes Care 2009; 32: 1398-1403.

27. Andrukhova O, Slavic S, Zeitz U, Shalhoub V, Lanske B, Erben RG. FGF23 Regulates Renal Sodium Handling and Blood Pressure [Abstract]. J.Am.Soc.Nephrol. 2012; 23: 46A.

28. Kramer AB, Laverman GD, van Goor H, Navis G. Inter-individual differences in anti-proteinuric response to ACEi in established adriamycin nephrotic rats are predicted by pretreatment renal damage. J.Pathol. 2003; 201: 160-167.

29. Lindberg K, Amin R, Moe OW, et al. The Kidney Is the Principal Organ Mediating Klotho Effects. [published online ahead of print May 22, 2014]. J.Am.Soc.Nephrol. 2014. http://dx.doi.org/10.1681/ASN.2013111209.

30. Shimada T, Yamazaki Y, Takahashi M, et al. Vitamin D receptor-independent FGF23 actions in regulating phosphate and vitamin D metabolism. Am.J.Physiol.Renal Physiol. 2005; 289: F1088-95.


32. Hamming I, Cooper ME, Haagmans BL, et al. The emerging role of ACE2 in physiology and disease. J.Pathol. 2007; 212: 1-11.

33. Lely AT, Hamming I, van Goor H, Navis GJ. Renal ACE2 expression in human kidney disease. J.Pathol. 2004; 204: 587-593.

34. Dai B, David V, Martin A, et al. A Comparative Transcriptome Analysis Identifying FGF23 Regulated Genes in the Kidney of a Mouse CKD Model. PLoS One 2012; 7: e44161.

35. National Kidney Foundation K/DOQI Kidney Disease Quality Outcome Initiative. K/DOQI Clinical Practice Guidelines on Hypertension and Antihypertensive Agents in Chronic Kidney Disease: Guideline 9 Pharma-cological Therapy: Nondiabetic Kidney Disease. 2004.

Chapter 5Fibroblast Growth Factor 23 and Cardiovascular Mortality after Kidney TransplantationLeandro C. BaiaJelmer K. HumaldaMarc G. VervloetGerjan NavisStephan J.L. BakkerMartin H. de Borst

on behalf of the NIGRAM Consortium Clin J Am Soc Nephrol. 2013 Nov;8(11):1968-78.

Chapter 5

78

Abstract

Background and objectives: Circulating fibroblast growth factor 23 (FGF-23) is associated with adverse cardiovascular outcomes in CKD. Whether FGF-23 predicts cardiovascular mortality af-ter kidney transplantation, independent of measures of mineral metabolism and cardiovascular risk factors, is unknown.

Design, setting, participants, & measurements: The association between plasma C-terminal FGF-23 and cardiovascular mortality was analyzed in a single-center prospective cohort of 593 stable kidney transplant recipients (mean age ± SD, 52±12 years; 54% male; estimated GFR, 47±16 ml/min per 1.73 m2), at a median of 6.1 (interquartile range, 2.7–11.7) years after transplantation. Multivariable Cox regression models were built, adjusting for measures of renal function and mineral metabolism; Framingham risk factors; the left ventricular wall strain markers midregional fragment of pro–A-type natriuretic peptide (MR-proANP) and N-terminal-probrain natriuretic peptide (NT-proBNP); and copeptin, the stable C-terminal portion of the precursor of vasopressin.

Results: In multivariable linear regression analysis, MR-proANP (β=0.20, P<0.001), NT-proBNP (β=0.18, P<0.001), and copeptin (β=0.26, P<0.001) were independently associated with FGF-23. During follow-up for 7.0 (interquartile range, 6.2–7.5) years, 128 patients (22%) died, of whom 66 (11%) died due to cardiovascular disease; 54 (9%) had graft failure. FGF-23 was associated with an higher risk of cardiovascular mortality in a fully adjusted multivariable Cox regression model (hazard ratio [HR], 1.88 [95% confidence interval (CI), 1.11 to 3.19]; P=0.02). FGF-23 was also independently associated with all-cause mortality (full model HR, 1.86 [95% CI, 1.27 to 2.73]; P=0.001). Net reclassification improved for both cardiovascular mortality (HR, 0.07 [95% CI, 0.01 to 0.14]; P<0.05) and all-cause mortality (HR, 0.11 [95% CI, 0.05 to 0.18]; P<0.001).

Conclusions: Plasma FGF-23 is independently associated with cardiovascular and all-cause mortality after kidney transplantation. The association remained significant after adjusting for measures of mineral metabolism and cardiovascular risk factors.

79

FGF-23 and CV Mortality in Kidney Transplantation

5

Introduction

Kidney transplantation is the treatment of choice for patients with ESRD. As a result of improved surgical techniques, immunosuppressive therapies, and prevention of opportunistic infections, the short-term prognosis of kidney transplant recipients in particular has improved dramatically over the past decades. However, the risk of premature death due to cardiovascular disease re-mains greatly increased in kidney transplant recipients compared with the general population (1).

Deregulation of calcium/phosphate metabolism is common in CKD and in kidney transplant recipients with impaired renal function (2). In patients with CKD, circulating levels of fibroblast growth factor 23 (FGF-23), parathyroid hormone (PTH), and phosphate have been identified as independent risk factors for cardiovascular disease and all-cause mortality (3-9). Experimental studies demonstrated that FGF-23 may be directly involved in the development of left ventricu-lar hypertrophy (LVH) (10). Whether FGF-23 predicts cardiovascular mortality in renal transplant recipients is unknown.

Our primary hypothesis was that plasma FGF-23 is a risk factor for cardiovascular mortality in renal transplant recipients, independent of Framingham risk factors (recipient age and sex, systolic BP, antihypertensive treatment use, smoking status, diabetes mellitus, plasma total cholesterol, and HDL cholesterol), known determinants of FGF-23 (including estimated GFR [eGFR] and proteinuria (11), and factors related to phosphate metabolism. We also addressed whether FGF-23 is associated with the left ventricular wall strain markers midregional fragment of pro–A-type natriuretic peptide (MR-proANP) and N-terminal-probrain natriuretic peptide (NT-proBNP), and with copeptin, the stable C-terminal portion of the precursor of vasopres-sin (12), and whether these cardiac markers influence the association between FGF-23 and cardiovascular mortality.

Materials and Methods

Study Population

In this prospective cohort study, all renal transplant recipients with a graft functioning for at least 1 year who visited our outpatient clinic (University Medical Center Groningen, The Netherlands) between August 2001 and July 2003 were eligible to participate. This study was a post hoc analysis of a previously published study (13); plasma samples from 593 patients were available. All participants provided written informed consent. Patients were enrolled at median of 6.1 (interquartile range [IQR], 2.7–11.7) years after transplantation. Baseline visits were postponed if signs of infection were present. Patients diagnosed with cancer other than cured skin cancer were not eligible. A total of 606 patients signed written informed consent. The institutional

Chapter 5

80

review board approved the study protocol (METc 01/039), which was in adherence with the Declaration of Helsinki. The clinical and research activities being reported are consistent with the Principles of the Declaration of Istanbul as outlined in the Declaration of Istanbul on Organ Trafficking and Transplant Tourism.

Endpoints

The major outcome of the study was cardiovascular mortality. All-cause mortality and death-censored graft failure were secondary outcomes. Adequate collection of up-to-date data on events and mortality was ensured by our continuous surveillance system; general practitioners or referring nephrologists were contacted in case the current status of a patient was unknown. Cause of death was obtained by linking the number of the death certificate to the primary cause of death as coded by a physician, according to the International Classification of Diseases, Ninth Revision (ICD-9). Cardiovascular death was defined as ICD-9 codes 410–447 (14). Follow-up was completed until May 19, 2009. There was no loss to follow-up.

Patient Characteristics and Laboratory Measurements

Relevant transplant characteristics were taken from a database containing information on all renal transplantations performed at our center since 1968. Standard immunosuppressive treatment are published previously (13). Current medication, including active vitamin D use (al-facalcidol or calcitriol); presence of diabetes mellitus; and cardiovascular history were extracted from the medical record. Diabetes mellitus was defined according to the American Diabetes Association guidelines: fasting plasma glucose ≥126 mg/dl or use of antidiabetic medication (15). Cardiovascular history was defines as a history of myocardial infarction, percutaneous transluminal angioplasty or stenting of coronary or peripheral arteries, bypass operation of coronary or peripheral arteries, intermittent claudication, amputation for vascular reasons, transient ischemic attack, or ischemic cerebrovascular accident. BP pressure was measured as described previously (13). Smoking status was recorded with a self-report questionnaire.

Upon entry in the cohort (baseline), blood was drawn after an 8- to 12-hour overnight fasting period. EDTA plasma samples were stored at –80°C until assessment of biochemical parameters for this study. Plasma C-terminal FGF-23 levels were determined by sandwich ELISA (Immutop-ics, San Clemente, CA), with intra-assay and interassay coefficient of variation of <5% and <16%, respectively (16). Plasma creatinine concentrations were determined using a modified version of the Jaffe method (MEGA AU 510, Merck Diagnostic, Darmstadt, Germany). PTH and 1,25(OH)2 vitamin D were measured in EDTA plasma using radioimmunoassay, and 25(OH)-vitamin D levels were determined by isotope dilution–online solid phase extraction liquid chromatogra-phy–tandem mass spectrometry (17). NT-proBNP levels were measured by immunoassay on an ELECSYS2010 instrument (ELECSYS proBNP, Roche Diagnostics, Germany). MR-proANP was measured with a sandwich immunoassay (MR-proANP LIA; B.R.A.H.M.S) (18). Plasma copeptin

81


5

was measured using a sandwich immunoassay (B.R.A.H.M.S GmbH/Thermo Fisher Scientific, Hennigsdorf/Berlin, Germany) (19). Serum albumin, calcium, cholesterol, C-reactive protein, glucose, hemoglobin, and phosphate, and urinary phosphate, sodium, total protein, and urea were determined by routine laboratory measurements. we calculated eGFR using the CKD-Epidemiology Collaboration formula (20).

Statistical analyses

Variable distribution was tested with histograms and probability plots. Normally distributed variables are presented as mean ± SD, and skewed variables as median (interquartile range). Baseline characteristics upon entry into the cohort were compared between tertiles of FGF-23 using one-way ANOVA, Kruskal-Wallis test, and chi-squared tests as appropriate. Skewed data were natural log-transformed for correlation analysis, Cox regression, and backward and forward linear analysis.

The associations between (log-transformed) FGF-23 levels and MR-proANP, NT-proBNP, and copeptin were assessed by Pearson correlation analysis. Backward linear regression was used to identify correlates of plasma FGF-23 levels. The following covariates were tested: age, sex, donor type (deceased or living), history of one or more acute rejection episodes, cold ischemia time, waist circumference, cardiovascular history, Framingham risk score factors, eGFR, protein-uria, serum phosphate, 24-hour urinary phosphate excretion (representing phosphate intake), 24-hour urinary urea excretion (representing protein intake (21)), plasma 25(OH)-vitamin D, 1,25(OH)2 vitamin D, PTH, C-reactive protein, albumin, hemoglobin, MR-proANP, NT-proBNP and copeptin levels, use of angiotensin-converting enzyme inhibitor or angiotensin receptor blocker (given potential interactions between the renin-angiotensin-aldosterone system and FGF-23 (22)), use of active vitamin D, and number of antihypertensive drugs. Variables significantly associated with FGF-23 levels using backward linear regression analysis were subsequently tested in a forward linear regression model. Variables that were significant in this model were considered independent correlates of FGF-23 levels.

Risk ratios and conditional maximum likelihood estimates of the rate ratio for cardiovascular and all-cause mortality and graft failure were calculated per tertile of FGF-23. Confidence intervals were calculated using the Fisher Exact test. To further analyze the association between FGF-23 and outcomes, tertiles of FGF-23 were plotted in Kaplan Meier curves with log rank test. Subsequently, multivariable Cox regression models were built to assess the association between FGF-23 and outcomes, adjusted for potential confounders. We tested for effect modifications by invoking multiplicative interaction terms. Sensitivity analyses were performed in a predefined subgroup of patients with an eGFR between 30 and 90 mL/min per 1.73m2 (n=493).

Chapter 5

82

To determine the goodness of fit of the predictive models, the Akaike Information Criterion (AIC) was used. AIC is a statistical estimate of the trade-off between the likelihood of a model against its complexity. A lower AIC value indicates a better model fit. Furthermore, discrimination, or the ability of the risk prediction models to distinguish those with an event (i.e., cardiovascular or all-cause mortality or graft failure) from those who do not, was evaluated with the Harrell C-index. The C-index is analogous to the area under the receiver operating characteristic curve, for which larger values indicate better discrimination.

We assessed net reclassification improvement (NRI) for predefined risk categories of cardiovas-cular and all-cause mortality (<5%, 5%–10%, and >10%), and calculated integrated discrimina-tion improvement (IDI) (23). The NRI provides reclassification tables constructed separately for study participants with and without events and quantifies the correct movement between categories: upwards for events and downwards for nonevents.

Proportionality assumptions were tested using Schoenfeld tests and log-minus-log survival plots. All statistical analyses were performed using SPSS software, version 16.0 for Windows (IBM, Armonk, NY); Stata/SE software, version11.0 for Windows (College Station, TX); and OpenEpi (openepi.com). P values <0.05 were considered to represent statistically significant differences.

Results

Patient Characteristics

Patient characteristics are summarized in Table 1. Overall, FGF-23 levels were 140 [IQR, 95–219] RU/ml, and eGFR was 47 ± 16 ml/min per 1.73 m2. None of the patients used dietary phosphate binders. Patients in the highest tertile FGF-23 tertile were older; were more likely to have received a graft from a deceased donor, to have had an acute rejection episode and to be a smoker; had a longer cold ischemia time; had a higher waist circumference and systolic and diastolic BP; had worse renal function; had lower serum albumin, HDL cholesterol, 1,25(OH)2 vitamin D, and hemoglobin; had higher serum phosphate, NT-proBNP, MR-proANP, copeptin, PTH, and CRP levels; and had more proteinuria.

FGF-23 is Independently Associated with Cardiac Markers

Plasma FGF-23 levels were associated with NT-proBNP (r2=0.12; P<0.001), MR-proANP (r2=0.18; P<0.001), and copeptin (r2=0.18; P<0.001) levels in univariate analysis (Figure 1).

Using backward and forward multivariable linear regression models, we found that copeptin was strongly and independently associated with plasma FGF-23 levels (Table 2). Other independent correlates of FGF-23 levels were eGFR, serum phosphate, hemoglobin, and donor type (living or

83


5

Table 1. Clinical Characteristics of the Cohort

Plasma FGF-23 tertiles

All patientsn=593

Tertile 1n=197

Tertile 2n=198

Tertile3 n=198

P-value

Plasma C-terminal FGF-23 (RU/mL) 140 (95‒219) 83 (68‒96) 141 (124‒162) 311 (220‒567) <0.001

Age (yr) 52 ± 12 50 ± 13 52 ± 12 53 ± 11 0.04

Men (%) 54 56 51 55 0.62

Caucasian race (%) 95 94 94 95 0.41

Cause of kidney disease (%) 0.75

Primary GN 28 26 30 29

GN due to vasculitis / autoimmune disease 6 9 5 6

Tubulointerstitial nephritis / pyelonephritis 16 20 15 14

Polycystic kidney disease 18 16 18 20

Hypertension 6 7 5 5

Diabetes 3 4 3 3

Other / unknown 23 18 24 23

Dialysis duration (mo) 27 (13‒48) 24 (13‒43) 31 (14‒48) 26 (13‒52) 0.40

Deceased donor transplant (%) 86 78 92 91 <0.001

Cold ischemia time (h) 22 (15‒27) 20 (7‒27) 22 (16‒26) 22 (16‒29) 0.04

Total warm ischemia time (min) 35 (30‒45) 35 (30‒45) 35 (30‒45) 36 (30‒45) 0.92

Total number of HLA mismatches (n) 1 (0‒2) 1 (0‒2) 1 (0‒2) 1 (0‒2) 0.41

History of acute rejection (%) 45 39 44 52 0.04

Transplant vintage (yr) 6.1 (2.7‒11.7) 5.6 (2.4‒11.1) 5.8 (3.0‒11.5) 7.2 (3.5‒12.2) 0.06

Waist circumference (cm) 97 ± 14 94 ± 13 97 ± 14 100 ± 13 <0.001

Systolic BP (mmHg) 153 ± 23 148 ± 21 153 ± 22 157 ± 25 <0.001

Diastolic BP (mmHg) 90 ± 10 89 ± 10 89 ± 9 91 ± 10 <0.001

Current smoker (%) 22 18 18 30 <0.001

Current diabetes (%) 18 19 17 18 0.88

Laboratory results

Measurements in serum/plasma

Creatinine (mg/dL) 1.5 (1.3‒1.9) 1.4 (1.1‒1.6) 1.5 (1.3‒1.7) 1.9 (1.4‒2.5) <0.001

Albumin (g/dL) 4.1 ± 0.3 4.1 ± 0.3 4.1 ± 0.4 4.0 ± 0.3 <0.001

Calcium (mg/dL) 9.6 ± 0.6 9.5 ± 0.6 9.6 ± 0.7 9.6 ± 0.7 0.20

Total cholesterol (mg/dL) 218 ± 42 214 ± 35 219 ± 38 221 ± 51 0.13

HDL cholesterol (mg/dL) 42 ± 12 45 ± 13 42 ± 12 40 ± 12 <0.001

LDL cholesterol (mg/dL) 137 ± 39 137 ± 31 136 ± 39 138 ± 46 0.84

NT-proBNP (pg/mL) 299 (131‒672) 180 (92‒419) 285 (130‒567) 457 (240‒1517) <0.001

MR-proANP (pmol/L) 164 (103‒275) 126 (87‒188) 153 (100‒231) 257 (158‒390) <0.001

Copeptin (pmol/L) 9.3 (5.0‒19.5) 6.5 (3.8‒10.4) 8.6 (4.7‒17.4) 16.1 (7.7‒30.0) <0.001

C-reactive protein (mg/L) 2.1 (0.8‒4.9) 1.7 (0.6‒3.7) 2.1 (0.8‒4.5) 3.0 (1.2‒7.8) <0.001

Glucose (mg/dL) 87 ± 23 86 ± 22 87 ± 26 88 ± 20 0.59

Chapter 5

84

deceased). Results were highly similar when, instead of copeptin, MR-proANP (coefficient, 0.20 [95% CI, 0.12 to 0.28]; P<0.001) or NT-proBNP (coefficient, 0.18 [95% CI, 0.11 to 0.25]; P<0.001) were forced into the final forward regression model.

Baseline Plasma FGF-23 and Outcomes

During a median follow-up of 7.0 (IQR, 6.2–7.5) years, 128 patients (22%) died, 66 (11%) of cardiovascular disease; 54 (9%) had graft failure. Kaplan-Meier survival plots demonstrated a higher risk of cardiovascular mortality, all-cause mortality, and graft failure per increasing tertile of plasma FGF-23 (Figure 2).

Table 1. (continued)


All patientsn=593

Tertile 1n=197

Tertile 2n=198

Tertile3 n=198

P-value

Phosphate (mg/dL) 3.3 ± 0.7 3.1 ± 0.6 3.2 ± 0.6 3.6 ± 0.7 <0.001

Parathyroid hormone (pg/mL) 86 (57‒131) 72 (51‒112) 88 (61‒121) 101 (68‒193) <0.001

25-OH vitamin D (ng/mL) 21 ± 9 22 ± 10 21 ± 8 22 ± 10 0.49

1,25-OH2 vitamin D (pg/mL) 42 ± 18 47 ± 17 42 ± 19 35 ± 15 <0.001

Hemoglobin (g/dL) 13.7 ± 1.5 14.2 ± 1.3 13.8 ± 1.3 13.1 ± 1.7 <0.001

Estimated GFR (CKD-EPI) (mL/min/1.73m2) 47 ± 16 55 ± 13 48 ± 14 37 ±15 <0.001

Measurements in 24h-urine

sodium excretion (mEq/24 hr) 140 ± 68 139 ± 68 138 ± 61 143 ± 75 0.66

phosphate excretion (g/24hr) 2.3 ± 0.8 2.3 ± 0.9 2.4 ± 0.8 2.2 ± 0.8 0.18

proteinuria (g/24 hr) 0.2 (0.0-0.5) 0.2 (0.0-0.3) 0.2 (0.0-0.4) 0.3 (0.2-0.8) <0.001

urea excretion (mEq/24 hr) 375 ± 114 380 ± 110 388 ± 114 357 ± 116 0.03

Medication

Ca supplements (%) 7 5 6 10 0.07

Active vitamin D (alfacalcidol or calcitriol) (%) 9 8 5 14 0.02

ACE inhibitor or ARB (%) 34 30 34 39 0.17

Prednisone dose, mg/day (%) 10 (7.5-10) 10 (7.5-10) 10 (7.5-10) 10 (7.5-10) 0.58

Calcineurin inhibitor (%) 78 77 83 75 0.12

Azathioprine (%) 34 29 34 38 0.17

Mycophenolate mofetil (%) 40 49 41 31 <0.001

Sirolimus (%) 2 2 2 2 0.88

Values expressed with a plus/minus sign are the mean ± SD; values with parentheses are median and interquar-tile range. All variables were determined at enrollment except gender, race, cause of the original kidney disease, cold ischemia time, total warm ischemia time and total number of HLA mismatches, which were assessed at time of transplantation. FGF-23, fibroblast growth factor 23; NT-proBNP, N-terminal probrain natriuretic peptide; MR-proANP, mid-region pro‒A-type natriuretic peptide; CKD-EPI, CKD-Epidemiology Collaboration; ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker.

85


5

Figure 1. Scatter diagrams illustrating the associations between fibroblast growth factor 23 (FGF23) and (A) N-ter-minal-probrain natriuretic peptide (NT-proBNP), (B) pro–A-type natriuretic peptide (MR-proANP), and (C) copeptin.

Chapter 5

86

Risk ratios for cardiovascular mortality, all-cause mortality, and graft failure were significantly increased in the highest FGF-23 tertile (Table 3).In Cox regression models, the highest tertile of FGF-23 was consistently associated with a higher risk of cardiovascular mortality after adjustment for Framingham risk factors, history of

Table 2. Multivariable Linear Regression Analysis for Correlates of Plasma Fibroblast Growth Factor 23 Levels

Variable Coefficient for FGF-23 (95 CI) P-value

eGFR (CKD-EPI) –0.32 (–0.41 to –0.23) <0.001

Ln copeptin 0.26 (0.17 to 0.35) <0.001

Serum phosphate 0.15 (0.07 to 0.23) <0.001

Hemoglobin –0.14 (–0.22 to –0.06) <0.001

Donor type (0=living; 1=deceased) –0.10 (–0.18 to –0.03) 0.004

Excluded from the model: age, sex, history of one or more acute rejection episodes, cold ischemia time, waist circumference, cardiovascular history, systolic and diastolic BP, smoking status, diabetes mellitus, total cholesterol, proteinuria, 24-urinary phosphate and urea excretion, plasma 25(OH)-vitamin D, 1,25(OH)2 vitamin D, C-reactive protein, albumin, parathyroid hormone, use of angiotensin-converting enzyme inhibitor or angiotensin receptor blocker, use of active vitamin D, and total number of antihypertensive drugs. Coefficients are standardized β values. Model fit: R2=0.41. FGF-23, fibroblast growth factor 23; CI, confidence interval; eGFR, estimated GFR; CKD-EPI, CKD-Epidemiology Collaboration; Ln, natural log.

Figure 2. Kaplan-Meier survival plots of fibroblast growth factor 23 (FGF-23) tertiles and cardiovascular mortality. Higher FGF-23 levels are associated with an increased incidence of cardiovascular mortality.

87


5

cardiovascular disease, measures of renal function, phosphate metabolism, and cardiac wall stress (Table 4). To assess whether the association between FGF-23 and cardiovascular mortality was influenced by other factors, interaction analyses were performed. We found no evidence for interaction in the association between FGF-23 and cardiovascular mortality by eGFR (P for interaction=0.51), proteinuria (P=0.62), 24-hour urinary phosphate excretion (P=0.50), frac-tional phosphate excretion (P=0.87), or any Framingham risk factor. In addition, no significant interaction by NT-proBNP (P for interaction=0.26), MR-proANP (P=0.32), or copeptin (P=0.19) was observed. NT-proBNP and MR-ANP but not copeptin were independently associated with cardiovascular mortality in the full model (data not shown). The final models provided the best fit (lowest AICs); all models displayed good discrimination as suggested by relatively high Harrell C-indices (Table 4). As a sensitivity analysis, full multivariable Cox regression models were re-peated in a subgroup of patients with an eGFR between 30 and 90 ml/min per 1.73m2 (n=493). In this subgroup, plasma FGF-23 remained an independent risk factor for cardiovascular mortal-ity (full model hazard ratio [HR], 2.43 [95% CI, 1.19 to 4.93] per 1 SD of natural log [ln]FGF-23; P=0.01) and all-cause mortality (HR, 2.12 [95% CI, 1.29 to 3.49] per 1 SD of lnFGF-23; P=0.003).

Higher FGF-23 levels were also associated with a higher risk of all-cause mortality, both when analyzed per tertile and as continuous variable (Table 4). Higher tertiles of FGF-23 were also as-sociated with a higher risk of death-censored graft failure in crude analyses, but significance was lost after adjustment for eGFR and proteinuria (Table 4). There was no significant interaction between the risk of graft failure and cardiovascular mortality (P=0.38).

Table 3. Events and Risk Ratios for Cardiovascular and All-cause Mortality and Graft Failure per Fibroblast Growth Factor 23 Tertile


All patients Tertile 1 Tertile 2 Tertile 3 P-value

Cardiovascular mortality

Events / patient-years 66 / 3722 11 / 1343 17 / 1260 38 / 1119

Incidence rate/1000 person-yrs 17.7 8.2 13.5 34.0

Risk ratio (95 CI) Reference 1.65 (0.73-3.89) 4.15 (2.08-8.99) <0.001

All-cause mortality

Events / patient years 128 / 3722 24 / 1343 33 / 1260 71 / 1119



Graft failure

Events / patient years 54 / 3722 8 / 1343 8 / 1260 38 / 1119



Chapter 5

88

Comparative Analyses

To compare the performance of FGF-23 with other measures of mineral metabolism (fractional phosphate excretion, PTH, phosphate, 25-OH vitamin D or 1,25[OH]2 vitamin D) as individual predictors of cardiovascular mortality, separate Cox regression analyses were performed for each of the six exposures, adjusted for age, gender, cardiovascular history, eGFR, proteinuria, and Framingham risk factors. As indicated in Figure 3, FGF-23 (highest tertile HR, 3.45 [95% CI, 1.57 to 7.58]; P=0.001) but not the other exposures were strongly associated with cardiovascu-lar mortality.

Table 4. Cox Regression Analyses for Cardiovascular and All-cause Mortality and Graft failure per Fibroblast Growth Factor 23 Tertile or per 1 SD of Natural Log Fibroblast Growth Factor 23

Plasma FGF-23 tertiles Natural Log FGF-23

Variable Tertile 1 Tertile 2 Tertile 3 P-valuea per 1 SD P-value AIC Harrell C

Cardiovascular mortality, HR (95CI)

Model 1 Reference 1.66 (0.78-3.54) 4.24 (2.17-8.29) <0.001 1.97 (1.52-2.55) <0.001 645 0.706

Model 2 Reference 1.62 (0.75-3.50) 4.02 (1.90-8.50) <0.001 1.98 (1.44-2.74) <0.001 648 0.706

Model 3 Reference 1.39 (0.55-3.50) 4.20 (1.90-9.81) <0.001 1.93 (1.33-2.80) 0.001 800 0.801

Model 4 Reference 1.59 (0.58-4.14) 4.06 (1.55-9.98) 0.003 2.27 (1.37-3.75) 0.001 479 0.810

Model 5 Reference 1.86 (0.67-5.18) 2.96 (1.06-8.26) 0.04 1.88 (1.11-3.19) 0.02 467 0.843

All-cause mortality, HR (95 CI)

Model 1 Reference 1.47 (0.87-2.48) 3.64 (2.29-5.78) <0.001 1.99 (1.65-2.39) <0.001 1245 0.694

Model 2 Reference 1.36 (0.80-2.32) 2.94 (1.73-4.98) <0.001 1.86 (1.46-2.36) <0.001 1237 0.693

Model 3 Reference 1.03 (0.55-1.93) 3.15 (1.69-5.85) <0.001 1.93 (1.45-2.56) <0.001 1542 0.768

Model 4 Reference 1.06 (0.53-2.11) 2.63 (1.32-5.24) 0.004 2.18 (1.50-3.18) <0.001 869 0.778

Model 5 Reference 1.17 (0.58-2.36) 2.23 (1.11-4.49) 0.02 1.86 (1.27-2.73) 0.001 858 0.799

Graft failure, HR 95 CI)

Model 1 Reference 1.06 (0.40-2.83) 6.29 (2.92-13.44) <0.001 3.37 (2.56-4.42) <0.001 515 0.762

Model 2 Reference 0.54 (0.20-1.46) 1.15 (0.47-2.84) 0.41 1.49 (0.99-2.24) 0.06 463 0.861

Model 3 Reference 0.53 (0.18-1.57) 0.80 (0.29-2.24) 0.99 1.15 (0.73-1.82) 0.55 540 0.885

Model 4 Reference 0.35 (0.10-1.22) 0.39 (0.12-1.30) 0.25 1.00 (0.59-1.71) 0.99 327 0.907

Model 5 Reference 0.32 (0.09-1.21) 0.37 (0.11-1.32) 0.25 1.03 (0.59-1.78) 0.93 331 0.909

Model 1: crude modelModel 2: adjusted for estimated GFR (CKD-Epidemiology Collaboration formula) and proteinuriaModel 3: Model 2 + adjusted for cardiovascular disease history, Framingham risk factors (recipient age and sex, systolic BP, antihypertensive treatment use, smoking status, diabetes mellitus, plasma total cholesterol, and HDL cholesterol), and transplant vintageModel 4: Model 3 + adjusted for serum phosphate, 1,25(OH)2-vitamin D, parathyroid hormone levels and active vitamin D useModel 5: Model 4 + adjusted for N-terminal-probrain natriuretic peptide, pro–A-type natriuretic peptide, and co-peptinHR, hazard ratio; CI, confidence interval; AIC, Aikaike Information Criterion.aP for linear trend.

89


5NRI and IDI

The NRI was calculated to assess the putative clinical utility of FGF-23 as a marker of (cardio-vascular) mortality in addition to Framingham risk factors. Reclassification improved mainly in the patients who survived, where adding FGF-23 to the model resulted in reclassification into a lower-risk category (Table 5). Overall, net reclassification improved for both cardiovascular mor-tality (0.07 [95% CI, 0.01 to 0.14]; P<0.05) and all-cause mortality (0.11 [95% CI, 0.05 to 0.18]; P<0.001). Accordingly, IDI, which does not rely on prespecified risk categories but represents a continuous measure, was 0.03 (95% CI, 0.01 to 0.05; P=0.004) for cardiovascular mortality and 0.05 (95% CI, 0.03 to 0.07; P<0.001) for all-cause mortality.

Discussion

This study identified high plasma FGF-23 levels as an independent risk factor for cardiovascular mortality in kidney transplant recipients. The association was consistent in regression models adjusted for renal function, measures of mineral metabolism, and cardiovascular risk factors. The significantly improved NRI suggests that FGF-23 levels may have an additional value to Framingham risk factors to predict the risk of cardiovascular mortality in renal transplant re-cipients.

Our findings are in line with previous data linking high FGF-23 levels with incident cardiovas-cular disease and mortality in the CKD population (3,24,25), and with a higher risk of all-cause mortality after kidney transplantation (26). Furthermore, FGF-23 levels have been associated with established cardiovascular risk factors and with a higher risk of cardiovascular events in the general population (27,28). Although patients in the highest FGF-23 tertile were character-ized by an overall high cardiovascular risk profile, the associations between plasma FGF-23 and

Figure 3. Comparative analysis of fibroblast growth factor 23 (FGF-23), fractional phosphate excretion (FE PO4), parathyroid hormone, phosphate (P), 25(OH)-vitamin D, or 1,25(OH)2 vitamin D as independent risk factors for car-diovascular (CV) mortality. The model was adjusted for age, sex, cardiovascular history, estimated GFR, proteinuria, and Framingham risk factors. For each exposure, the lowest tertile served as the reference group (R).

Chapter 5

90

(cardiovascular and all-cause) mortality remained significant after adjusting for several major cardiovascular risk markers, including Framingham risk factors, suggesting that FGF-23 is a strong and independent risk marker. The potential clinical relevance of FGF-23 in addition to established risk markers is further supported by improved reclassification (NRI and IDI). Fur-thermore, FGF-23, but not serum phosphate, PTH or vitamin D, was independently associated

Table 5. Reclassification Based on Fibroblast Growth Factor 23 LevelsEstimated risk of cardiovascular mortality:

Without FGF-23 With FGF-23 (n)

<5 5-10 >10 Total (n)

Patients with cardiovascular mortality

<5 2 1 3

5-10 6 6

>10 2 55 57

Total 2 9 55 66

Patients without cardiovascular mortality

<5 186 12 1 199

5-10 46 75 15 136

>10 27 15 178

Total 232 114 167 513

Net reclassification improvement was 0.07 (95% confidence interval [CI], 0.01 to 0.14; P<0.05).

Estimated risk of all-cause mortality:

Without FGF-23 With FGF-23 (n)

<5 5-10 >10 Total (n)

Patients with cardiovascular mortality

<5 1 1 1 3

5-10 3 3 2 8

>10 3 114 117

Total 4 7 117 128

Patients without cardiovascular mortality

<5 92 6 98

5-10 24 41 9 74

>10 5 44 230 279

Total 121 91 239 451

Net reclassification improvement was 0.11 (95% CI, 0.05 to 0.18; P<0.001).Multivariable model adjusted for Framingham risk factors (recipient age and gender, systolic BP, antihypertensive treatment use, smoking status, diabetes mellitus, plasma total cholesterol and HDL cholesterol) with or without plasma C-terminal fibroblast growth factor 23 level (FGF-23).

91


5

with cardiovascular mortality (Figure 3), suggesting a specific role for FGF-23. The observed association between elevated FGF-23 levels and a higher cardiovascular risk initially positioned FGF-23 as a biomarker of phosphate toxicity. More recently, animal studies demonstrated that FGF-23 has ‘off-target’ effects, directly contributing to the development of LVH (10). We found that FGF-23 was independently associated with MR-proANP and NT-proBNP, both markers of left ventricular wall strain, and with copeptin, a stable peptide derived from the vasopressin precursor (29). All three cardiac markers are used as markers of heart failure (30). Although these associations are in line with a potential role for FGF-23 in LVH, we could not demonstrate modulation of the association between FGF-23 and cardiovascular mortality by these cardiac markers. Furthermore, only a small amount of FGF-23 was explained by copeptin, MR-proANP and NT-proBNP (Figure 1), suggesting that FGF-23 is independently related to cardiovascular and all-cause mortality.

In a previous study, the association between FGF-23 and cardiovascular events was modified by sex (31). We could not confirm this effect modulation, which could be explained by differences in renal function between the Heart and Soul study (eGFR, 71 ± 23 mL/min per 1.73 m2) and our cohort (eGFR, 47±16 mL/min per 1.73 m2).. Data from 24-hour urine collections allowed us to include phosphate intake (assessed by 24-hour urinary phosphate excretion) (32) and excretion (fractional phosphate excretion) in our analyses. Neither measure interacted with the association between FGF-23 levels and cardiovascular mortality.

FGF-23 was also associated with an increased risk of graft failure in univariate analysis, in line with a previous study (26). However, the association was no longer significant after adjustment for eGFR and proteinuria in addition to known risk factors for graft failure (Table 3). We hypoth-esize that renal function after transplantation has a stronger effect on the risk of graft failure than on the risk of mortality (33,34).

Possible limitations of our study include its post hoc nature and the use of ICD-9 codes rather than an independent adjudication committee to define the endpoint cardiovascular mortality. Although we adjusted for several confounders, including measures of renal function, which appeared to be the most important determinant of FGF-23 levels, the possibility of residual con-founding cannot be fully excluded because not all determinants of FGF-23 are currently known. The sensitivity analyses in patients with an eGFR of 30–90 mL/min per 1.73 m2 reduces the risk of confounding by renal function The fact that diabetic and hypertensive nephropathy were under-represented in our cohort, and that our cohort consisted almost entirely of Caucasian patients, limits the generalizability of our study. Because FGF-23 was measured only at a single time point in this cohort, we could not perform a repeated-measures analysis. Prevalent cohorts such as this may suffer from survivorship biases; these generally lead to an underestimation of the actual hazard ratio (35). Using a relatively large number of covariates in the final Cox

Chapter 5

92

regression model might result in overfitting the model; however, the stepwise Cox models and the low AIC in the final model suggest that overfitting was no issue in our analyses. Finally, we did not have extensive structural or functional cardiac data (e.g., from echocardiography) available to further explore the mechanisms behind the observed association between FGF-23 and cardiovascular mortality. Strong points of this study, on the other hand, are the use of 24-hour urine collections, which enabled us to adjust for proteinuria and phosphate excretion, the availability of multiple markers of cardiovascular risk, and the complete follow-up.

In conclusion, to our knowledge this is the first study to identify FGF-23 as an independent risk factor for cardiovascular mortality after kidney transplantation. Although it may be relevant to consider therapies reducing FGF-23 levels (36) to improve cardiovascular outcomes, it should be kept in mind that high FGF-23 levels may serve an important physiological goal, namely to keep phosphate balance. A recent study demonstrated that specific FGF-23 blockade with a neutralizing antibody did reduce secondary hyperparathyroidism but increased serum phos-phate, aortic calcification and mortality (37). Whether specific reduction of FGF-23 or rather reducing phosphate load may improve cardiovascular prognosis after kidney transplantation should be addressed in future prospective studies.

Disclosures

None.

Acknowledgements

The authors would like to thank Wendy Dam for excellent technical assistance. M.H.D.B is supported by personal development grants from the Dutch Kidney foundation (KJPB.08.07) and the University Medical Center Groningen (Mandema stipend). This work is supported by a consortium grant from the Dutch Kidney Foundation (NIGRAM consortium, grant no. CP10.11). The NIGRAM consortium consists of the following principal investigators: Piet ter Wee and Marc Vervloet (VU University Medical Center, Amsterdam, the Netherlands); René Bindels and Joost Hoenderop (Radboud University Medical Center Nijmegen, the Netherlands); and Gerjan Navis, Jan-Luuk Hillebrands, and Martin de Borst (University Medical Center Groningen, the Netherlands).

93


5

References

1. Jardine AG, Gaston RS, Fellstrom BC, Holdaas H. Prevention of cardiovascular disease in adult recipients of kidney transplants. Lancet 378: 1419-1427, 2011

2. Larsson T, Nisbeth U, Ljunggren O, Juppner H, Jonsson KB. Circulating concentration of FGF-23 increases as renal function declines in patients with chronic kidney disease, but does not change in response to variation in phosphate intake in healthy volunteers. Kidney Int 64: 2272-2279, 2003

3. Arnlov J, Carlsson AC, Sundstrom J, Ingelsson E, Larsson A, Lind L, Larsson TE. Higher fibroblast growth factor-23 increases the risk of all-cause and cardiovascular mortality in the community. Kidney Int 83: 160-166, 2013

4. Kestenbaum B, Sampson JN, Rudser KD, Patterson DJ, Seliger SL, Young B, Sherrard DJ, Andress DL. Serum phosphate levels and mortality risk among people with chronic kidney disease. J Am Soc Nephrol 16: 520-528, 2005

5. Fliser D, Kollerits B, Neyer U, Ankerst DP, Lhotta K, Lingenhel A, Ritz E, Kronenberg F, MMKD Study Group, Kuen E, Konig P, Kraatz G, Mann JF, Muller GA, Kohler H, Riegler P. Fibroblast growth factor 23 (FGF23) predicts progression of chronic kidney disease: The mild to moderate kidney disease (MMKD) study. J Am Soc Nephrol 18: 2600-2608, 2007

6. Gutierrez OM, Januzzi JL, Isakova T, Laliberte K, Smith K, Collerone G, Sarwar A, Hoffmann U, Coglianese E, Christenson R, Wang TJ, deFilippi C, Wolf M. Fibroblast growth factor 23 and left ventricular hypertrophy in chronic kidney disease. Circulation 119: 2545-2552, 2009

7. Gutierrez OM, Mannstadt M, Isakova T, Rauh-Hain JA, Tamez H, Shah A, Smith K, Lee H, Thadhani R, Juppner H, Wolf M. Fibroblast growth factor 23 and mortality among patients undergoing hemodialysis. N Engl J Med 359: 584-592, 2008

8. Kovesdy CP, Ahmadzadeh S, Anderson JE, Kalantar-Zadeh K. Secondary hyperparathyroidism is associated with higher mortality in men with moderate to severe chronic kidney disease. Kidney Int 73: 1296-1302, 2008

9. Tonelli M, Curhan G, Pfeffer M, Sacks F, Thadhani R, Melamed ML, Wiebe N, Muntner P. Relation between alkaline phosphatase, serum phosphate, and all-cause or cardiovascular mortality. Circulation 120: 1784-1792, 2009

10. Faul C, Amaral AP, Oskouei B, Hu MC, Sloan A, Isakova T, Gutierrez OM, Aguillon-Prada R, Lincoln J, Hare JM, Mundel P, Morales A, Scialla J, Fischer M, Soliman EZ, Chen J, Go AS, Rosas SE, Nessel L, Townsend RR, Feldman HI, St John Sutton M, Ojo A, Gadegbeku C, Di Marco GS, Reuter S, Kentrup D, Tiemann K, Brand M, Hill JA, Moe OW, Kuro-O M, Kusek JW, Keane MG, Wolf M. FGF23 induces left ventricular hypertrophy. J Clin Invest 121: 4393-4408, 2011

11. Vervloet MG, van Zuilen AD, Heijboer AC, Ter Wee PM, Bots ML, Blankestijn PJ, Wetzels JF, MASTERPLAN group study. Fibroblast growth factor 23 is associated with proteinuria and smoking in chronic kidney disease: An analysis of the MASTERPLAN cohort. BMC Nephrol 13: 20, 2012

12. Struck J, Morgenthaler NG, Bergmann A. Copeptin, a stable peptide derived from the vasopressin precur-sor, is elevated in serum of sepsis patients. Peptides 26: 2500-2504, 2005

13. de Vries AP, Bakker SJ, van Son WJ, van der Heide JJ, Ploeg RJ, The HT, de Jong PE, Gans RO. Metabolic syndrome is associated with impaired long-term renal allograft function; not all component criteria con-tribute equally. Am J Transplant 4: 1675-1683, 2004

14. Zelle DM, Corpeleijn E, van Ree RM, Stolk RP, van der Veer E, Gans RO, Homan van der Heide JJ, Navis G, Bakker SJ. Markers of the hepatic component of the metabolic syndrome as predictors of mortality in renal transplant recipients. Am J Transplant 10: 106-114, 2010

Chapter 5

94

15. American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care 31 Suppl 1: S55-60, 2008

16. Heijboer AC, Levitus M, Vervloet MG, Lips P, ter Wee PM, Dijstelbloem HM, Blankenstein MA. Determina-tion of fibroblast growth factor 23. Ann Clin Biochem 46: 338-340, 2009

17. Doorenbos CR, de Cuba MM, Vogt L, Kema IP, van den Born J, Gans RO, Navis G, de Borst MH. Antiprotein-uric treatment reduces urinary loss of vitamin D-binding protein but does not affect vitamin D status in patients with chronic kidney disease.J Steroid Biochem Mol Biol 128: 56-61, 2012

18. van Hateren KJ, Alkhalaf A, Kleefstra N, Groenier KH, de Jong PE, de Zeeuw D, Gans RO, Struck J, Bilo HJ, Gansevoort RT, Bakker SJ. Comparison of midregional pro-A-type natriuretic peptide and the N-terminal pro-B-type natriuretic peptide for predicting mortality and cardiovascular events. Clin Chem 58: 293-297, 2012

19. Morgenthaler NG, Struck J, Alonso C, Bergmann A. Assay for the measurement of copeptin, a stable peptide derived from the precursor of vasopressin. Clin Chem 52: 112-119, 2006

20. Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF,3rd, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, Coresh J, CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration). A new equation to estimate glomerular filtration rate. Ann Intern Med 150: 604-612, 2009

21. Mitch WE. Beneficial responses to modified diets in treating patients with chronic kidney disease. Kidney Int Suppl (94): S133-5, 2005

22. de Borst MH, Vervloet MG, ter Wee PM, Navis G. Cross talk between the renin-angiotensin-aldosterone system and vitamin D-FGF-23-klotho in chronic kidney disease. J Am Soc Nephrol 22: 1603-1609, 2011

23. Pencina MJ, D’Agostino RB S, D’Agostino RB,Jr, Vasan RS. Evaluating the added predictive ability of a new marker: From area under the ROC curve to reclassification and beyond. Stat Med 27: 157-72; discussion 207-12, 2008

24. Seiler S, Reichart B, Roth D, Seibert E, Fliser D, Heine GH. FGF-23 and future cardiovascular events in patients with chronic kidney disease before initiation of dialysis treatment. Nephrol Dial Transplant 25: 3983-3989, 2010

25. Kendrick J, Cheung AK, Kaufman JS, Greene T, Roberts WL, Smits G, Chonchol M, HOST Investigators. FGF-23 associates with death, cardiovascular events, and initiation of chronic dialysis. J Am Soc Nephrol 22: 1913-1922, 2011

26. Wolf M, Molnar MZ, Amaral AP, Czira ME, Rudas A, Ujszaszi A, Kiss I, Rosivall L, Kosa J, Lakatos P, Kovesdy CP, Mucsi I. Elevated fibroblast growth factor 23 is a risk factor for kidney transplant loss and mortality. J Am Soc Nephrol 22: 956-966, 2011

27. Gutierrez OM, Wolf M, Taylor EN. Fibroblast growth factor 23, cardiovascular disease risk factors, and phosphorus intake in the health professionals follow-up study. Clin J Am Soc Nephrol 6: 2871-2878, 2011

28. Arnlov J, Carlsson AC, Sundstrom J, Ingelsson E, Larsson A, Lind L, Larsson TE. Serum FGF23 and risk of cardiovascular events in relation to mineral metabolism and cardiovascular pathology. Clin J Am Soc Nephrol 2013 Jan 18 Epub ahead of print

29. Abbasi A, Corpeleijn E, Meijer E, Postmus D, Gansevoort RT, Gans RO, Struck J, Hillege HL, Stolk RP, Navis G, Bakker SJ. Sex differences in the association between plasma copeptin and incident type 2 diabetes: The prevention of renal and vascular endstage disease (PREVEND) study. Diabetologia 55: 1963-1970, 2012

30. Sabatine MS, Morrow DA, de Lemos JA, Omland T, Sloan S, Jarolim P, Solomon SD, Pfeffer MA, Braunwald E. Evaluation of multiple biomarkers of cardiovascular stress for risk prediction and guiding medical therapy in patients with stable coronary disease. Circulation 125: 233-240, 2012

31. Ix JH, Chonchol M, Laughlin GA, Shlipak MG, Whooley MA. Relation of sex and estrogen therapy to serum fibroblast growth factor 23, serum phosphorus, and urine phosphorus: The heart and soul study. Am J Kidney Dis 58: 737-745, 2011

95


5

32. Vervloet MG, van Ittersum FJ, Buttler RM, Heijboer AC, Blankenstein MA, ter Wee PM. Effects of dietary phosphate and calcium intake on fibroblast growth factor-23. Clin J Am Soc Nephrol 6: 383-389, 2011

33. Schnitzler MA, Lentine KL, Gheorghian A, Axelrod D, Trivedi D, L’Italien G. Renal function following living, standard criteria deceased and expanded criteria deceased donor kidney transplantation: Impact on graft failure and death. Transpl Int 25: 179-191, 2012

34. Yarlagadda SG, Coca SG, Formica RN,Jr, Poggio ED, Parikh CR. Association between delayed graft function and allograft and patient survival: A systematic review and meta-analysis. Nephrol Dial Transplant 24: 1039-1047, 2009

35. Arrighi HM, & Hertz-Picciotto I. The evolving concept of the healthy worker survivor effect. Epidemiology 5: 189-196, 1994

36. Wolf M. Update on fibroblast growth factor 23 in chronic kidney disease. Kidney Int 82: 737-747, 2012 37. Shalhoub V, Shatzen EM, Ward SC, Davis J, Stevens J, Bi V, Renshaw L, Hawkins N, Wang W, Chen C, Tsai

MM, Cattley RC, Wronski TJ, Xia X, Li X, Henley C, Eschenberg M, Richards WG. FGF23 neutralization improves chronic kidney disease-associated hyperparathyroidism yet increases mortality. J Clin Invest 122: 2543-2553, 2012

Chapter 6Fibroblast Growth Factor 23 Correlates with Volume Status in Haemodialysis Patients and is not Reduced by HaemodialysisJelmer K. HumaldaIneke J. RiphagenSolmaz AssaYoran M. HummelRalf WesterhuisMarc G. VervloetAdriaan A. VoorsGerjan NavisCasper F.M. FranssenMartin H. De Borst

on behalf of the NIGRAM Consortium Nephrol Dial Transplant. 2016 Sep;31(9):1494-501.

Chapter 6

98

Abstract

Background: Recent data suggest a role for fibroblast growth factor 23 (FGF-23) in volume regulation. In haemodialysis patients, a large ultrafiltration volume (UFV) reflects poor volume control, and both FGF-23 and a large UFV are risk factors for mortality in this population. We studied the association between FGF-23 and markers of volume status including UFV, as well as the intradialytic course of FGF-23, in a cohort of haemodialysis patients.

Methods: Observational, post hoc analysis of 109 prevalent haemodialysis patients who un-derwent a standardized, low-flux, haemodialysis session with constant ultrafiltration rate. We measured UFV, plasma copeptin and echocardiographic parameters including cardiac output, end-diastolic volume and left ventricular mass index at the onset of the haemodialysis session. We measured the intradialytic course of plasma C-terminal FGF-23 (corrected for haemocon-centration) and serum phosphate levels at 0, 1, 3 and 4 hours after onset of haemodialysis and analysed changes with linear mixed effect model.

Results: Median age was 66 [interquartile range 51-75] years, 65% was male with a weekly Kt/V 4.3 ± 0.7 and dialysis vintage of 25.4 [8.5-52.5] months. In univariable analysis, predialysis plasma FGF-23 was associated with UFV, end-diastolic volume, cardiac output, early diastolic velocity e′ and plasma copeptin. In multivariable regression analysis UFV correlated with FGF-23 (standardized β 0.373, P<0.001, model R2: 57%), independent of serum calcium and phosphate. The association between FGF-23 and echocardiographic volume markers was lost for all but cardiac output upon adjustment for UFV. Overall, FGF-23 levels did not change during dialysis (7627 [3300-13514] to 7503 [3109-14433] RU/mL; P=0.98), whereas phosphate decreased (1.71 ± 0.50 to 0.88 ± 0.26 mmol/L; P<0.001).

Conclusions: FGF-23 was associated with volume status in haemodialysis patients. The strong association with UFV suggests that optimization of volume status, for example by more intensive haemodialysis regimens, may also benefit mineral homeostasis. A single dialysis session did not lower FGF-23 levels.

99

FGF-23 Correlates with Volume Status in HD

6

Introduction

End-stage renal disease (ESRD) is characterized by profound abnormalities in mineral metabo-lism parameters including phosphate, calcium and the phosphate-regulating hormone fibroblast growth factor 23 (FGF-23). FGF-23 levels increase exponentially during the course of chronic kidney disease (CKD) to reach extremely high levels in haemodialysis (1). Higher circulating levels of FGF-23 have been linked to an increased risk of mortality in both CKD (2, 3) and hae-modialysis patients (4, 5). Furthermore, high FGF-23 levels are associated with an increased risk of cardiovascular events, and particularly with cardiovascular complications related to volume overload as congestive heart failure (6). FGF-23 emerged as such a robust predictor for adverse outcome, suggesting that FGF-23 is not merely a biomarker but also exerts direct pathophysi-ological effects. Indeed, preclinical studies demonstrated that FGF-23 induces left ventricular hypertrophy in mice (7), although in rats neither high FGF-23 levels nor treatment with FGF-23 antibodies changed the minimal incidence of left ventricular hypertrophy compared with controls (8). FGF-23 also did not correlate with ventricular hypertrophy in the participants of the Heart and Soul Study (9). Another putative pathophysiological effect of FGF-23 is that it may directly stimulate tubular sodium transport to promote sodium retention (10). Thus, it is of utmost importance to identify determinants of FGF-23 in haemodialysis patients and explore possible modes of intervention.

Renal function is a major determinant of FGF-23. Previous studies in haemodialysis patients re-vealed that the presence of residual renal function (RRF) is associated with FGF-23 independent of serum phosphate and calcium levels, and that anuric patients display higher FGF-23 levels than patients with residual diuresis (11, 12). Haemodialysis patients requiring a higher ultrafil-tration volume (UFV) have an increased mortality risk (13, 14). FGF-23 is also associated with parameters of volume overload such as NT-proBNP and the vasopressin precursor copeptin, independent of renal function (15), in line with a potential link between FGF-23 and volume status. Therefore, in the current study, we aimed to assess whether UFV is an independent determinant of FGF-23, among other parameters of volume homeostasis.

Despite the accumulating evidence supporting the association between a high FGF-23 level and adverse cardiorenal outcomes in all stages of CKD, little solid data exist on the effects of low-flux haemodialysis treatment itself on FGF-23 levels. Haemodialysis lowers serum phosphate levels markedly. As serum phosphate and FGF-23 levels are strongly correlated, an intradialytic effect on FGF-23 levels might be expected, and was suggested by a preliminary study (16), even when FGF-23 is probably not dialysed due to its size. Therefore, we also investigated the intradialytic change in FGF-23 levels.

Chapter 6

100


Study Design and Measurements

We performed an observational study in a haemodialysis cohort recruited at the Dialysis Center Groningen and the University Medical Center Groningen. The protocol has been described pre-viously (17). In short, patients who underwent thrice-weekly haemodialysis for >3 months were eligible. Patients with severe heart failure (NYHA class IV) were excluded from participation. All measurements took place at the end of the longest interdialytic interval. Patients under-went a bicarbonate low-flux haemodialysis session with polysulfone hollow-fibre dialyser (F8; Fresenius Car, Bad Hamburg, Germany). Blood and dialysate flows were 250-350 mL/min and 500 mL/min, respectively, with constant ultrafiltration rate. The dialysate bath consisted of 139 mmol/L sodium, 1.0 or 2.0 mmol/L potassium, calcium (1.5 mmol/L), magnesium (0.5 mmol/L), bicarbonate (34 mmol/L), acetate (3.0 mmol/L) and glucose (1.0 g/L). Blood samples were obtained at the start of dialysis, 1 h, t3 h hours and at the end of a regular 4-h haemodialysis session. Dialysate was obtained 30 minutes after the onset of the dialysis session. The study was performed in accordance to the Declaration of Helsinki and approved by the Medical Ethical Committee of the University Medical Center Groningen. The study took place between March 2009 and March 2010.

Clinical and Laboratory Measurements

Haematocrit and electrolytes (sodium, calcium, phosphate, total calcium and ionized calcium) were measured by routine laboratory procedures. We measured plasma FGF-23 with human FGF-23 (C-terminal) enzyme-linked immunosorbent assay (ELISA; Immutopics, Inc., San Clem-ente, CA, USA) in stored plasma samples. The concentration of plasma FGF-23 was corrected for haemoconcentration or -dilution according to Schneditz et al. (18). In a subset of patients, FGF-23 was measured in the dialysate. Plasma copeptin was measured using a sandwich immu-noassay (B.R.A.H.M.S GmbH/Thermo Fisher Scientific, Hennigsdorf/Berlin, Germany). ProANP was measured using an automated sandwich immunoassay (MR-ProANP KRYPTOR, B.R.A.H.M.S. GmbH, Hennigsdorf/Berlin, Germany). Blood pressure was measured at all four time points. All patients underwent echocardiography 30 min before start of the haemodialysis session aimed at assessing cardiac function in general and development of ventricular systolic dysfunction in particular. Three experienced technicians performed the echocardiography, and one expe-rienced technician performed all analyses offline (17). Data on RRF (≥ 1 mL/min) and residual diuresis (defined as ≥200 mL/day) were extracted from patient records. RRF was calculated from a 24-hour collection together with blood sampling using urinary and serum creatinine values: RRF= [(urinary creatinine excretion/ serum creatinine) / 1440].

101


6

Statistical Analysis

Analyses were performed using SPSS version 22.0 for Windows (IBM Corporation, Chicago, IL, USA) and GraphPad Prism, version 5.01 (GraphPad Software Inc). Data are reported as mean ± standard deviation for normally distributed data and median (interquartile range) for non-normally distributed data. Nominal data are presented as the total number of patients with percentage [n (%))]. A two-sided P-value <0.05 was considered to be statistically significant.

We used uni- and multivariable linear regression analyses to identify determinants of natural logarithmic transformed (Ln) plasma FGF-23 levels at the start of the haemodialysis session. In univariable regression analysis, we first studied known determinants of FGF-23 [serum phos-phate, serum calcium, (serum) calcium–phosphate product (only univariable analysis) age and RRF –defined as creatinine clearance ≥1 mL/min–]; putative determinants as Kt/V for dialysis efficacy; UFV, plasma sodium, copeptin, systolic and diastolic blood pressure as proxies for vol-ume status; ionized calcium as it may perform better than serum calcium; and gender and body mass index as possible confounders. We constructed a multivariable model with UFV as inde-pendent covariate and other significant contributing determinants of the univariable analysis, thus adjusting for known determinants of FGF-23 (calcium and phosphate levels). We aimed to construct the optimal model with using as few variables as possible. Non-normally distributed variables were Ln-transformed when necessary to meet the assumptions of linear regression. We report standardized regression coefficients. We tested for effect modifications by invoking multiplicative interaction terms. As volume homeostasis may influence cardiac structural and functional parameters, we further investigated the correlation of FGF-23 with these parameters in a novel model. We investigated echocardiography parameters [cardiac output, stroke volume, heart rate, end-diastolic volume, end-systolic volume, left ventricular mass index (LVMI), ejec-tion fraction (biplane Simpson’s method), tricuspid annular plane systolic excursion (TAPSE) and tissue Doppler early diastolic velocity (mean e′ )]) and proANP in addition to UFV and copeptin in uni- and multivariable linear regression analysis. Changes in blood pressure and concentrations of FGF-23, copeptin, total and ionized calcium, phosphate and sodium before and after dialysis were analysed with paired t tests, after Ln transformation when appropriate. The course of FGF-23 and phosphate during the HD session was analysed with a linear mixed effect model. We calculated the change in relative blood volume (RBV) using haematocrit (Ht) values: RBV(%) = (Ht before/ Ht after dialysis session) – 1 × 100%. The correlation of changes in FGF-23 and phosphate or RBV was tested with Pearson’s correlation test.

Chapter 6

102

Results

Study Population

The 109 haemodialysis patients in our cohort were 66 (51-75) years old, had a dialysis vintage of 25.4 (8.5-52.5) months and weekly Kt/V of 4.3 ± 0.7. The mean UFV during the haemodialysis session was 2553 ± 777 mL. As shown in Table 1, before the onset of the haemodialysis session, patients had a mildly elevated blood pressure (systolic 140 ± 23 mmHg and diastolic 79 ± 15 mmHg), had high plasma copeptin levels, were hyperphosphatemic and displayed very high FGF-23 levels, as expected. Most patients used antihypertensive and calcium/phosphate-related treatments and had a mildly reduced systolic cardiac function as reflected by cardiac output and ejection fraction, and signs of diastolic dysfunction as reflected by early diastolic velocity, e′.

Determinants of FGF-23 at the Onset of the Haemodialysis Session

Univariable regression analyses identified serum phosphate, UFV, age and Ln copeptin as strong correlates of Ln FGF-23 levels at the onset of the haemodialysis session (Table 2). Total calcium and Kt/V were also positively associated with high FGF-23 levels. Notably, the calcium × phos-phate product (Ca × Pi) correlated strongest with Ln FGF-23 (standardized β = 0.655, P<0.001). In multivariable regression analysis, total calcium, phosphate and UFV were independent predictors of FGF-23 levels (Table 3, Model 1). In this model, each 100 mL increase of UFV resulted in an in-crease of Ln-transformed FGF-23 with effect size B = 0.053 (95% confidence interval: 0.033-0.073). Thus, for every extra litre of UFV, FGF-23 levels increased by e0.53= 170%, e.g. from 8103 to 13 766 RU/mL. Age, copeptin and Kt/V did not contribute to this model (Table 3, Model 2), nor did RRF (Model 3). We found no evidence for interaction by serum phosphate for the association between UFV and FGF-23, after introduction of the multiplicative interaction serum phosphate × UFV to a multivariable regression model with serum phosphate and UFV and Ln FGF-23 as dependent variables (Figure 1, P-interaction=0.3). Use of ionized calcium instead of serum calcium did not materially change our findings, and ionized calcium did not correlate with FGF-23 in univariable analysis. Therefore, the correlation of (ionized) Ca×Pi product with FGF-23 appears to be driven by phosphate. FGF-23 did not differ between patients who did or did not use calcium-containing phosphate binders (P=0.8) or use vitamin D or phosphate binders (P=0.9).

FGF-23 Correlates with Echocardiographic Volume Parameters

We further investigated the correlation of FGF-23 with cardiac structural and functional param-eters using echocardiography and the volume marker copeptin. Regarding the echocardiogra-phy parameters, FGF-23 correlated strongly with the systolic indices cardiac output, apparently driven by stroke volume in univariable analysis. FGF-23 also correlated with the diastolic indices e′ mean and end-diastolic volume (Table 4). Notably, FGF-23 was not associated with left ven-tricular mass index (LVMI), LV ejection fraction, TAPSE or proANP levels. Copeptin correlated significantly with FGF-23. In multivariable analysis, only cardiac output contributed further

103


6

Table 1. Baseline Characteristics

Unit Value Reference value

Gender (male/female) 71/38

Age years 66 (51‒75)

Dialysis vintage months 25.4 (8.5‒52.5)

BMI kg/m2 25.9 ± 4.5 <25

Systolic blood pressure mmHg 140 ± 23 <140

Diastolic blood pressure mmHg 79 ± 15 <90

FGF-23 RU/mL 7627 (3300‒13514) <125

Plasma sodium mmol/L 138 ± 3 132‒144

Serum calcium mmol/L 2.31 ± 0.16 9.0‒11.0

Serum phosphate mmol/L 1.71 ± 0.50 2.0‒4.0

Copeptin pmol/L 142 (91‒245) <14

proANP pmol/L 754 (510‒1080)

RRF ≥1mL/min n (%) 35 (32.1)

Residual diuresis ≥ 200 mL/d n (%) 39 (35.8)

UFV mL 2553 ± 777

Kt/V 4.3 ± 0.7 >3.6

Weight above target weight kg 2.4 ± 1.5

Medication use

Antihypertensivesa n (%) 75 (70.1)

RAAS blockade n (%) 23 (21.5)

Diuretics n (%) 8 (7.5)

Calcium or phosphate treatment n (%) 85 (79.4)

Calcium-containing phosphate binders n (%) 45 (41.3)

Echocardiographic parameters

Cardiac output L/min 3.6 ± 1.2

Stroke volume mL 52 ± 16

Heart rate beats/min 74 ± 14 60‒100

End-diastolic volume mL 106 ± 32

End-systolic volume mL 54 ± 22

Tricuspid annular plane systolic excursion mm 23 ± 5 >17

e′ tissue Doppler early diastolic velocity cm/s 6.6 ± 2.1 >7.8

LV mass index, males g/m2 99 ± 25 49-115

LV mass index, females g/m2 84 ± 25 43‒95

LV ejection fraction (Simpson) % 50 ± 10 >52%

BMI, body mass index; FGF-23, fibroblast growth factor 23; RU/mL, relative units per mililitre; RAAS blockade, blockade of renin-angiotensin-aldosterone system, either by angiotensin-converting enzyme inhibitors or angio-tensin receptor blockers; proANP, proatrial natriuretic peptide; LV, left ventricular. a, defined as use of RAAS block-ade, calcium channel blockers, beta blockers and/or diuretics.

Chapter 6

104

Table 2. Correlates of FGF-23 at Onset of Haemodialysis Session.

Determinant Standardized β P-value

Ca×Pi product 0.655 <0.001

Serum phosphate 0.629 <0.001

UFV 0.579 <0.001

Age –0.406 <0.001

Ln copeptin 0.269 0.005

Kt/V –0.224 0.023

Serum calcium 0.201 0.037

Diastolic blood pressure 0.148 0.134

Gender –0.141 0.145

RRF ≥1mL/min –0.100 0.335

BMI 0.089 0.376

Residual diuresis ≥ 200 mL/day –0.091 0.376

Systolic blood pressure –0.052 0.602

Ionized calcium 0.032 0.760

Dialysis vintage –0.026 0.789

Plasma sodium –0.001 0.995

Univariable regression analysis of natural log-transformed FGF-23. Abbreviations: BMI= Body Mass Index; Ca × Pi, calcium–phosphate product; UFV, ultrafiltration volume; RRF, residual renal function; FGF-23, fibroblast growth factor 23.

Table 3. Multivariable Linear Regression Analysis: Determinants of FGF-23 Levels at Onset of Dialysis Session

Determinant Standardized β P-value R2

1 Phosphate 0.484 <0.001 0.57

Calcium 0.247 <0.001

UFV (per 100mL) 0.373 <0.001

2 Phosphate 0.480 <0.001 0.58

Calcium 0.234 <0.001

UFV (per 100mL) 0.352 <0.001

Age –0.063 0.4

Copeptin 0.076 0.3

Kt/V 0.043 0.5

3 Phosphate 0.520 <0.001 0.62

Calcium 0.253 <0.001

UFV (per 100mL) 0.343 <0.001

Residual renal function ≥1 mL/min –0.091 0.2

Model 1 is the optimal regression model, that used least determinants to predict most of the variations in FGF-23 levels. Model 2 assessed the effect of all significant correlates from the univariable analysis. Model 3 assessed whether RRF may attenuate the results. FGF-23, natural log-transformed fibroblast growth factor 23; Standardized β, standardized coefficient; UFV, ultrafiltration volume; RRF, residual renal function

105


6

Figure 1. Both in patients with phosphate levels below and above median, UFV correlated with Ln-transformed FGF-23. Regression lines for FGF-23 and UFV are depicted per levels of serum phosphate above (continuous) and below (dashed) its median. FGF-23, fibroblast growth factor 23; PO4, phosphate.

Table 4. Univariable Correlates of FGF-23 at Onset of Dialysis Session and Echocardiographic and Biochemical Markers of Volume Status

Univariable Multivariable

Determinant Standardized β P-value Standardized β P-value

UFV 0.579 <0.001 0.462 <0.001

Cardiac output 0.389 <0.001 0.279 0.02

Stroke volume 0.345 0.001

Heart rate 0.097 0.3

proAVP (copeptin) 0.269 0.005 0.105 0.3

End-diastolic volume 0.286 0.007

e′ mean 0.228 0.02 0.004 0.9

End-systolic volume 0.148 0.2

LVMI 0.079 0.5 –0.112 0.4

Ejection fraction (Simpson) 0.047 0.7

TAPSE 0.027 0.8 –0.111 0.3

Ln proANP –0.007 0.9 0.042 0.7

ProANP and FGF-23 were natural log (Ln) transformed. As cardiac output is the product of stroke volume and heart rate, these factors were not included in multivariable analysis. Similarly, ejection fraction, end-diastolic and end-systolic volume were excluded due to their strong relation with cardiac output. FGF-23, fibroblast growth factor 23; UFV, ultrafiltration volume; proAVP, provasopressin; LVMI, left ventricular mass index; TAPSE, tricuspid annular plane systolic excursion; proANP; pro-A-type natriuretic peptide.

Chapter 6

106

when UFV was included in the model. However, when we forced cardiac output into our final regression model (table 3, model 1), it did not contribute (standardized β = 0.085, P=0.3).

Figure 2. The mean change of FGF-23, serum phosphate and serum calcium (with standard deviation) at the differ-ent time points during one a haemodialysis session. P-value reflects linear mixed effect model for the course of the variables. FGF-23, fibroblast growth factor 23.

The Effect of Haemodialysis on FGF-23

During the haemodialysis session, serum phosphate levels declined significantly (Figure 2, mixed linear model, P < 0.001), whereas haemodialysis treatment had no effect on FGF-23 levels (P=0.98) and a small effect on calcium levels. Changes in serum phosphate were not correlated with changes in FGF-23 at any of the time points (correlation of changes after 60 minutes R= –0.076, P=0.5; after 180 minutes R= –0.025, P=0.8; after end of dialysis R= –0.096, P=0.4). We could not detect FGF-23 in the dialysate in any patient when analysing a subgroup of 10 hae-modialysis patients, supporting the observation that FGF-23 was not cleared by haemodialysis.

Table 5. Parameters during Haemodialysis

Unit At start haemodialysis

At end haemodialysis

P-value

Systolic blood pressure mmHg 140 ± 23 132 ± 26 <0.001

Diastolic blood pressure mmHg 79 ± 15 73 ± 14 <0.001

FGF-23 RU/mL 7627 (3300‒13514) 7503 (3109‒14433) 0.2

Plasma sodium mmol/L 138 ± 3 139 ± 2 0.03

Serum calcium mmol/L 2.31 ± 0.16 2.49 ± 0.13 <0.001

Ionized calcium mmol/L 1.21 ± 0.09 1.25 ± 0.05 <0.001

Serum phosphate mmol/L 1.71 ± 0.50 0.88 ± 0.26 <0.001

Copeptin pmol/L 142 (91‒245) 164 (98‒292) <0.001

The course of clinical and biochemical parameters during haemodialysis. FGF-23 and copeptin were Ln-transformed before t-test. Abbreviations: FGF-23, fibroblast growth factor 23; RU/mL, relative units per millilitre.

107


6

Changes in serum or ionized calcium, both of which increased during haemodialysis (Table 5), did not correlate with changes in FGF-23. In order to more closely delineate the effects of indi-vidual volume changes during the haemodialysis session on FGF-23 levels, we did find that the change in FGF-23 correlated positively with the change in RBV (figure 3): patients with a more pronounced reduction in blood volume had also a more pronounced reduction of FGF-23 levels, whereas patients whose blood volume increased demonstrated an increase in FGF-23 levels.

Figure 3. The relative change of blood volume during one haemodialysis session correlated with the change in FGF-23 levels. Pearson’s correlation coefficient R and its P-value are reported. One outlying observation was ex-cluded (change RBV –6%, FGF-23 +138%). FGF-23, fibroblast growth factor 23.

Discussion

FGF-23 has recently been identified as a major non-classical cardiovascular risk factor in the haemodialysis population. The main finding of our study is that FGF-23 correlated strongly both with UFVand with echocardiographic and biochemical markers of volume overload. UFV thus emerged as a novel, strong and independent determinant of FGF-23 levels in haemodialysis patients. This adds to emerging data connecting FGF-23 with volume status. Secondly, we found that haemodialysis treatment, although it strongly reduced serum phosphate, did not achieve a uniform reduction of FGF-23 levels: a reduction in relative blood volume led to a decrease, whereas an increment in relative blood volume increased FGF-23 levels.

The high FGF-23 levels we observed are in line with other reports on FGF-23 in haemodialysis (4, 5) and peritoneal dialysis (19). The interpretation of the observed independent association between FGF-23 and UFV is complex. Deranged fluid homeostasis is particularly common in the haemodialysis setting, and chronic fluid overload is strongly associated with an increased

Chapter 6

108

risk of cardiovascular events and mortality (20, 21). In addition, a higher UFV is associated with an increased risk of sudden death (14). Both UFV (14) and FGF-23 (4) are determinants of mortality in large haemodialysis populations. A high UFV reflects by definition a state of volume overload, and indeed overhydration as measured by body composition monitoring is associ-ated with mortality as well (22). A mechanistic explanation linking FGF-23 to volume overload may lie in the sympathetic nervous system. Volume overload and the consequent large volume shifts during dialysis may lead to sympathetic overactivity, which have been shown to directly induce FGF-23 production in bone in experimental studies (23). Although heart rate was not independently associated with FGF-23 in our models, this does not exclude a potential role for the sympathetic nerve system. Indeed, when 11 patients were converted from three to six times weekly haemodialysis, this lead to a reduction of UFV from 2.4 to 1.5 L and a concomitant reduction of sympathetic activity, as assessed by muscle sympathetic nerve activity (24). In addition, sympathetic activity as measured by norepinephrine levels correlated strongly with mortality and concentric left ventricular hypertrophy ––the consequence of prolonged volume overload–– in 197 haemodialysis patients (25). FGF-23 and sympathetic overactivity may thus be intertwined heralds of adverse outcome in haemodialysis patients. We could not demon-strate an association between RRF and FGF-23, as reported by others (11, 12, 26), possibly because of our small sample size and a relatively large proportion of anuric patients. The inverse correlation of age and FGF-23 might seem counterintuitive, but has been reported before (27), possibly suggesting less compliance to phosphate-restriction in the younger patients. The strong association between FGF-23 and calcium–phosphate product has been reported earlier in an animal model (28). It is surprising that ionized calcium does not seem to correlate with FGF-23, whereas serum calcium shows only a weak correlation. Both calcium parameters when multiplied by phosphate correlate strongly with FGF-23, and calcium is an independent variable in multivariable regression analysis, apparently bolstering the role of phosphorus as primary determinant of FGF-23 levels in haemodialysis.

FGF-23 correlated strongly with functional markers of volume overload, i.e. cardiac output, stroke volume, diastolic volume and copeptin, but not with the structural marker LVMI. This is at variance with earlier observations in predialysis CKD (29, 30) and HD patients (31), although Wald et al. (32). also did not observe an association between LVMI and FGF-23 using MRI in HD patients. FGF-23 did not correlate with ejection fraction. This contrasts with the observa-tion that FGF-23 levels were elevated in patients with an ejection fraction <40% in a cohort of 885 patients with generally preserved renal function who underwent coronary angiography, (33). This may be explained by our exclusion of NYHA class IV patients or alternatively by their exclusion of patients on dialysis. Sharma et al. (34) found no difference in ejection fraction on baseline across tertiles of FGF-23 in 110 haemodialysis patients, however, higher FGF-23 levels did correlate with a more pronounced loss of ejection fraction after 1.9 years follow-up. Our findings may be reconciled with this report in that a prolonged higher stroke volume and higher

109


6

end-diastolic volume in a high FGF-23 setting may eventually lead to decompensation and thus a lower stroke volume and ejection fraction. Alternatively, our relatively small cohort or differ-ences in inclusion criteria may have influenced these associations. FGF-23 also correlated with copeptin, a hormone that is increased in haemodialysis patients and correlated with volume stimuli (35). Taken together, we interpret these findings as that FGF-23 correlates more closely with volume parameters rather than cardiac function and dimensions. This is supported by the fact that UFV remained the strongest determinant of FGF-23 levels, independent of other echocardiographic covariates that pointed to a hyperdynamic volume status.

In contrast with a previous study demonstrating a 19% reduction in intact FGF-23(16), we could not detect a relevant change in C-terminal FGF-23 during a haemodialysis session. We used another method to correct for haemodilution effects (18); however, this cannot explain the dis-crepancy as this correction did not affect FGF-23 levels by >10%. A possible explanation might be the use of another ELISA assay. We measured C-terminal FGF-23, because this assay is most suitable for repeated measures, least affected by stability issues and certainly in ESRD closely reflects the amount of intact FGF-23 (4, 36, 37). This is supported by the absence of FGF-23 in the dialysate, although the FGF-23 ELISA has not been validated for use in other biomaterials than plasma. FGF-23 clearance might not have been expected due to the size of the intact form (32kDa) precluding its passage through the dialyser membrane; even smaller fragments of 9 kDa are not likely filtered by the low-flux dialysation procedure (38). FGF-23 could not be detected at least in significant quantities by our assay detecting both intact FGF-23 and its C-terminal fragments.

Our findings may seem in contrast with the fact that an intensified haemodialysis schedule does lower FGF-23 levels (39); yet better long-term phosphate control achieved by intensified haemodialysis (40) is likely to be accompanied by lower FGF-23 levels on the longer term. In that light, we could discern an univariable association of Kt/V and FGF-23, an indicator that long-term dialysis efficacy may contribute to lower FGF-23 levels. An intensified haemodialysis schedule leads to a lower UFV per dialysis session and thus less pronounced volume shifts, this too may lower FGF-23. Indeed, a decrease in RBV in one haemodialysis session already correlated with a small decrease in FGF-23 levels in our patients. Notably, recent studies on alternative dialy-sis modalities reported that a single 8-hour online haemodiafiltration (HDF) session reduced FGF-23 levels by 48.6% (41) and that a reduction after 3-4 hours of online HDF (55.7%) is also greater than after 3-5 hours of high-flux haemodialysis (36.2%) (42). FGF-23 levels remained stable after a year of high-flux haemodialysis, whereas FGF-23 increased further in the low-flux haemodialysis group (43). In line with our observations, FGF-23 did not change significantly after a regular 4-h haemodialysis session and FGF-23 could not be detected in both regular and HDF dialysate, despite the larger pores in HDF(41). Possibly, this effect of HDF consists partly of the entrapment of FGF-23 molecules in the membrane in adjunct to the first discernible

Chapter 6

110

effects of prolonged phosphate and volume status control. It would be interesting to investigate whether strategies aimed at the reduction of the interdialytic weight gain, consistently reduce FGF-23 in a prospective study.

Strengths of our study include the fact that measurements took place during a standardized heamodialysis session, wherein we could assess intra-individual changes in FGF-23 levels at multiple time points. Further, the well-documented and extensive study protocol allowed us to assess different known and new determinants of FGF-23 at the same time. Moreover, the availability of extensive echocardiographic parameters enabled us to investigate further the putative relation of FGF-23 with volume status in our patients. Limitations of our study include its observational design, precluding definite conclusions on a cause-effect relationship between volume status and FGF-23 levels, the risk of residual confounding and the availability of a serum creatinine measurement after but not before 24-hour urine collection for the determination of RRF.

In conclusion, we found that a larger UFV was strongly and independently correlated with higher FGF-23 levels. Volume overload may thus contribute to the excessive FGF-23 levels in haemodialysis patients. This would suggests a change in the direction of effects, as according to previous hypotheses formed in pre-dialysis CKD, FGF-23 was considered to induce volume overload (3, 6). FGF-23 levels do not change during a single haemodialysis session, whereas serum phosphate was reduced by haemodialysis. Our findings support the current paradigm that intensified haemodialysis may improve both volume status and mineral metabolism on the long term as means to lower FGF-23 and hence reduce the high risk of adverse events in haemodialysis patients.

Acknowledgements

Statement of Financial Support

This work is supported by a consortium grant from the Dutch Kidney Foundation (NIGRAM con-sortium, grant no. CP10.11). The NIGRAM consortium consists of the following principal investi-gators: Pieter M. ter Wee, M.G.V. (VU University Medical Center, Amsterdam, the Netherlands), René J. Bindels and Joost G. Hoenderop (Radboud University Medical Center Nijmegen, the Netherlands), G.N., Jan-Luuk Hillebrands and M.H.d.B (University Medical Center Groningen, the Netherlands). M.H.d.B is supported by a grant from the Netherlands Organization for Scientific Research (Veni grant 016.146.014). I.J.R. is supported by TI Food and Nutrition, a public–private partnership on pre-competitive research in food and nutrition. The Dutch Kidney Foundation financially supported the original study (Grant C08.2279). A.A.V. (University of Groningen, the Netherlands) is supported by research grants from Alere, Singulex and Sphingotec. The funding

111


6

sources had no role in the design and conduct of the study, the collection, management, analy-sis and interpretation of the data, preparation, review or approval of the manuscript, or the decision to submit the report for publication. Parts of this work were presented at the European Renal Association-European Dialysis and Transplantation Association (ERA-EDTA) conference 2014, Amsterdam, the Netherlands (MP507) and at the American Society of Nephrology Kidney Week 2014, Philadelphia, PA (TH-PO789).

Disclosures

None.

Chapter 6

112

References

1. Isakova T, Wahl P, Vargas GS, et al. Fibroblast growth factor 23 is elevated before parathyroid hormone and phosphate in chronic kidney disease. Kidney Int. 2011; 79: 1370-1378.







8. Shalhoub V, Shatzen EM, Ward SC, et al. FGF23 neutralization improves chronic kidney disease-associated hyperparathyroidism yet increases mortality. J.Clin.Invest. 2012; 122: 2543-2553.

9. Agarwal I, Ide N, Ix JH, et al. Fibroblast Growth Factor–23 and Cardiac Structure and Function. Journal of the American Heart Association 2014; 3000584 doi: 10.1161/JAHA.113.00584.



12. Iwasawa H, Nakao T, Matsumoto H, Okada T, Nagaoka Y, Wada T. Phosphate handling by end-stage kidneys and benefits of residual renal function on phosphate removal in patients on haemodialysis. Nephrology (Carlton) 2013; 18: 285-291.

13. Flythe JE, Kimmel SE, Brunelli SM. Rapid fluid removal during dialysis is associated with cardiovascular morbidity and mortality. Kidney Int. 2011; 79: 250-257.

14. Jadoul M, Thumma J, Fuller DS, et al. Modifiable practices associated with sudden death among hemo-dialysis patients in the Dialysis Outcomes and Practice Patterns Study. Clin.J.Am.Soc.Nephrol. 2012; 7: 765-774.

15. Baia LC, Humalda JK, Vervloet MG, et al. Fibroblast growth factor 23 and cardiovascular mortality after kidney transplantation. Clin.J.Am.Soc.Nephrol. 2013; 8: 1968-1978.

16. Yamamoto T, Nascimento MM, Hayashi SY, et al. Changes in circulating biomarkers during a single hemo-dialysis session. Hemodial Int. 2013; 17: 59-66.

17. Assa S, Hummel YM, Voors AA, et al. Hemodialysis-induced regional left ventricular systolic dysfunction: prevalence, patient and dialysis treatment-related factors, and prognostic significance. Clin.J.Am.Soc.Nephrol. 2012; 7: 1615-1623.

18. Schneditz D, Putz-Bankuti C, Ribitsch W, Schilcher G. Correction of plasma concentrations for effects of hemoconcentration or hemodilution. ASAIO J. 2012; 58: 160-162.

19. Isakova T, Xie H, Barchi-Chung A, et al. Fibroblast growth factor 23 in patients undergoing peritoneal dialysis. Clin.J.Am.Soc.Nephrol. 2011; 6: 2688-2695.

113


6

20. Ferrario M, Moissl U, Garzotto F, et al. Effects of fluid overload on heart rate variability in chronic kidney disease patients on hemodialysis. BMC Nephrol. 2014; 15: 26-2369-15-26.

21. Zoccali C, Torino C, Tripepi R, et al. Pulmonary congestion predicts cardiac events and mortality in ESRD. J.Am.Soc.Nephrol. 2013; 24: 639-646.

22. Wizemann V, Wabel P, Chamney P, et al. The mortality risk of overhydration in haemodialysis patients. Nephrology Dialysis Transplantation 2009; 24: 1574-1579.

23. Kawai M, Kinoshita S, Shimba S, Ozono K, Michigami T. Sympathetic activation induces skeletal Fgf23 expression in a circadian rhythm-dependent manner. J.Biol.Chem. 2014; 289: 1457-1466.

24. Zilch O, Vos PF, Oey PL, et al. Sympathetic hyperactivity in haemodialysis patients is reduced by short daily haemodialysis. J.Hypertens. 2007; 25: 1285-1289.

25. Zoccali C, Mallamaci F, Tripepi G, et al. Norepinephrine and concentric hypertrophy in patients with end-stage renal disease. Hypertension 2002; 40: 41-46.

26. Wang M, You L, Li H, et al. Association of circulating fibroblast growth factor-23 with renal phosphate excretion among hemodialysis patients with residual renal function. Clin.J.Am.Soc.Nephrol. 2013; 8: 116-125.

27. Mizuiri S, Nishizawa Y, Yamashita K, et al. Lower serum fibroblast growth factor-23 levels may suggest malnutrition in maintenance haemodialysis patients. Nephrology (Carlton) 2014; 19: 568-573.


29. Smith K, deFilippi C, Isakova T, et al. Fibroblast growth factor 23, high-sensitivity cardiac troponin, and left ventricular hypertrophy in CKD. Am.J.Kidney Dis. 2013; 61: 67-73.

30. Gutierrez OM, Januzzi JL, Isakova T, et al. Fibroblast growth factor 23 and left ventricular hypertrophy in chronic kidney disease. Circulation 2009; 119: 2545-2552.

31. Kovacs A, Tapolyai M, Celeng C, et al. Impact of hemodialysis, left ventricular mass and FGF-23 on myo-cardial mechanics in end-stage renal disease: a three-dimensional speckle tracking study. Int.J.Cardiovasc.Imaging 2014; 30: 1331-1337.

32. Wald R, Goldstein MB, Wald RM, et al. Correlates of left ventricular mass in chronic hemodialysis recipi-ents. Int.J.Cardiovasc.Imaging 2014; 30: 349-356.

33. Seiler S, Cremers B, Rebling NM, et al. The phosphatonin fibroblast growth factor 23 links calcium-phosphate metabolism with left-ventricular dysfunction and atrial fibrillation. Eur.Heart J. 2011; 32: 2688-2696.

34. Sharma S, Joseph J, Chonchol M, et al. Higher fibroblast growth factor-23 concentrations associate with left ventricular systolic dysfunction in dialysis patients. Clin.Nephrol. 2013; 80: 313-321.

35. Ettema EM, Kuipers J, Assa S, et al. Changes in Plasma Copeptin Levels during Hemodialysis: Are the Physiological Stimuli Active in Hemodialysis Patients? PLoS One 2015; 10: e0127116.

36. Smith ER. The use of fibroblast growth factor 23 testing in patients with kidney disease. Clin.J.Am.Soc.Nephrol. 2014; 9: 1283-1303.

37. Smith ER, Cai MM, McMahon LP, Holt SG. Biological Variability of Plasma Intact and C-Terminal FGF23 Measurements. J.Clin.Endocrinol.Metab. 2012; 97: 3357‒3365.

38. Ronco C C, Bowry S, Ronco C. Nanoscale modulation of the pore dimensions, size distribution and struc-ture of a new polysulfone-based high-flux dialysis membrane. Int.J.Artif.Organs 2001-10; 24: 726-35.

39. Zaritsky J, Rastogi A, Fischmann G, et al. Short daily hemodialysis is associated with lower plasma FGF23 levels when compared with conventional hemodialysis. Nephrol.Dial.Transplant. 2014; 29: 437-441.

40. Daugirdas JT, Chertow GM, Larive B, et al. Effects of frequent hemodialysis on measures of CKD mineral and bone disorder. J.Am.Soc.Nephrol. 2012; 23: 727-738.

Chapter 6

114

41. Cornelis T, van der Sande FM, Eloot S, et al. Acute hemodynamic response and uremic toxin removal in con-ventional and extended hemodialysis and hemodiafiltration: a randomized crossover study. Am.J.Kidney Dis. 2014; 64: 247-256.

42. Patrier L, Dupuy AM, Granger Vallee A, et al. FGF-23 removal is improved by on-line high-efficiency hemo-diafiltration compared to conventional high flux hemodialysis. J.Nephrol. 2013; 26: 342-349.

43. Schneider A, Schneider MP, Krieter DH, et al. Effect of High-Flux Dialysis on Circulating FGF-23 Levels in End-Stage Renal Disease Patients: Results from a Randomized Trial. PLoS One 2015; 10: e0128079.

Part IIDietary Interventions

Chapter 7Response of Fibroblast Growth Factor 23 to Volume Interventions in Arterial Hypertension and Diabetic NephropathyJelmer K. Humalda†

Sarah Seiler-Mußler†

Arjan J. KwakernaakMarc G. VervloetGerjan NavisDanilo FliserGunnar H. HeineMartin H. de Borst

† Contributed equally.

Medicine, In Press.

Chapter 7

120

Abstract

Background: Fibroblast growth factor 23 (FGF-23) rises progressively in chronic kidney disease and is associated with adverse cardiovascular outcomes. FGF-23 putatively induces volume retention by up-regulating the sodium-chloride cotransporter (NCC). We studied whether, conversely, interventions in volume status affect FGF-23 concentrations.

Methods: Post hoc analysis of 1) a prospective saline infusion study with 12 patients with arterial hypertension who received 2L of isotonic saline over 4 hours, and 2) a randomized controlled trial with 45 diabetic nephropathy (DN) patients on background ACE-inhibition (ACEi), who underwent four 6-week treatment periods with add-on hydrochlorothiazide (HCT) or placebo, combined with regular (RS) or low sodium (LS) diet in a cross-over design. Plasma C-terminal FGF-23 was measured by ELISA (Immutopics) after each treatment period in DN and before and after saline infusion in hypertensives.

Results: The patients with arterial hypertension were 45 ± 13 (mean ± SD) years old with an eGFR of 101 ± 18 mL/min/1.73 m2. Isotonic saline infusion did not affect FGF-23 (before infu-sion: 68 median [1st‒3rd quartile: 58‒97] RU/mL, after infusion: 67 [57‒77] RU/mL, P=0.37). DN patients were 65 ± 9 years old. During ACEi+RS treatment eGFR was 65 ± 25 mL/min/1.73 m2 and albuminuria 649 mg/d [230‒2008 mg/d]. FGF-23 level was 94 [73‒141] RU/mL during ACE-inhibition therapy. FGF-23 did not change significantly by add-on HCT (99 [74‒148] RU/mL), low sodium diet (99 [75‒135 RU/mL]), or their combination (111 [81‒160 RU/mL]), P=0.15.

Conclusions: Acute and chronic changes in volume status did not materially change FGF-23 in hypertensive patients and DN, respectively. Our data do not support a direct feedback loop between volume status and FGF-23 in hypertension or diabetic nephropathy.

121

FGF-23 and Volume Interventions

7

Introduction

The phosphaturic hormone fibroblast growth factor 23 (FGF-23) is a central regulator of calcium-phosphate metabolism. In moderate to severe chronic kidney disease (CKD), FGF-23 concentrations progressively increase in an attempt to keep phosphate balance (1). More re-cently, a higher concentration of FGF-23 in CKD patients has been linked with an increased risk of cardiovascular events, particularly with decompensated heart failure, which in CKD is often caused by hypervolaemia (as reviewed in (2)). Similar associations between plasma FGF-23 con-centrations and the incidence of decompensated heart failure were found among individuals without overt CKD (3-5). In line, animal experiments suggest that FGF-23 may directly induce left-ventricular hypertrophy via the calcineurin-NFAT pathway (6, 7). Yet, the role of FGF-23 in cardiovascular disease in patients without advanced CKD is subject of debate (8, 9), and other pathophysiological pathways may be involved (10). Notably, in addition to the proposed direct effects of FGF-23 on the myocardium, volume overload could contribute to the association between increased FGF-23 and decompensated heart failure observed in cohort studies. This hypothesis is fueled by recent experimental data, which suggest that FGF-23 may activate the sodium-chloride co-transporter (NCC) in the distal tubule, inducing volume expansion, hyper-tension and subsequently left-ventricular hypertrophy (11). The diuretic hydrochlorothiazide, which inhibits the NCC in the distal tubule, prevented these presumed off-target effects. These observations add to previously documented independent associations between FGF-23 and markers of volume status (12), and could at least partly explain the previously observed as-sociation between a higher FGF-23 level and an impaired anti-proteinuric response to volume depletion (13). Conversely, recent data suggest that changes in volume status may modulate FGF-23 concentrations at least in haemodialysis (14). This suggests the presence of a feedback loop, where FGF-23 increases volume load, and an increase in volume may reduce FGF-23. Vice versa, a reduction in volume load may in turn increase FGF-23. In the current study, we investigated the effect of interventions in volume status on FGF-23 in two independent settings. We firstly analyzed the acute effects of intravenous volume loading on plasma FGF-23 in twelve hypertensive individuals without overt CKD. Secondly, we investigated the chronic impact of dietary sodium restriction and hydrochlorothiazide therapy on plasma FGF-23 in patients with diabetic CKD during standardized angiotensin-converting enzyme (ACE) inhibition. To expand knowledge on the hypothesized bi-directional relationship between FGF-23 and volume regula-tion, we analyzed among the same patients the extent by which FGF-23 plasma concentrations predict the antiproteinuric response to dietary sodium restriction and hydrochlorothiazide therapy.

Chapter 7

122


Study design and measurements

To investigate the effects of acute volume load on FGF-23 plasma concentrations, we analyzed 12 outpatients with arterial hypertension but without overt CKD (defined as an estimated GFR < 60 ml/min/1.73 m2) who received an infusion of 2 liters isotonic saline in 4 hours. ETDA-plasma samples to measure FGF-23 concentrations were obtained immediately before and after admin-istration of the infusion. The acute volume expansion study was approved by the local Medical Ethical Committee in Saarland, Germany.

Furthermore, to investigate the long-term effects of modifications in volume status on FGF-23, we performed a post hoc analysis on a randomized, placebo-controlled, double-blinded cross-over intervention trial addressing the effects of sodium restriction and thiazide diuretic treatment during background ACE inhibition. The original study protocol has been described previously (Dutch trial registry number 2366, (15)). Briefly, 45 patients with type 2 diabetes and diabetic nephropathy were included. Diabetic nephropathy was defined as albuminuria >30 mg/day, urinary albumin excretion >20 mg/L or urinary albumin creatinine ratio >2.5 mg/mmol for men and >3.5 mg/mmol for women. Patients had a creatinine clearance >30 mL/min and did not have signs of another primary renal disease. Main exclusion criteria were: pres-ence of type 1 diabetes, renovascular disease, a cardio- or cerebrovascular event <3 months ago, overt hyperkalemia (> 6.0 mmol/L) or nephrotic syndrome, renal transplant recipients, use of immunosuppressants, blood pressure >180/100 mmHg and contraindications to the use of lisinopril or hydrochlorothiazide. All patients were titrated to maximum dose ACE-inhibition (lisinopril 40 mg per day), with discontinuation of other RAAS-blockers/diuretics and stable dose of other antihypertensives. Patients underwent four subsequent 6-week treatments periods in random order: hydrochlorothiazide (50 mg per day) or placebo, combined with either a sodium restricted diet (targeting 50 mmol a day or ~3 grams NaCl /day) versus a liberal sodium diet. Patients had one or two dietary counseling sessions with a dietitian and received a list with the sodium content of general used food products in the Netherlands. After each treatment period measurements were performed, blood samples were taken by venipuncture and 24-hour urine was collected. The study was conducted in accordance with the Declaration of Helsinki and ap-proved by the Medical Ethical Committee of the University Medical Center Groningen (protocol number 2010/288).

Laboratory measurements

For these analyses we determined FGF-23 by human FGF-23 enzyme-linked immunosorbent assay directed against the carboxy-terminus (Immutopics Inc, San Clemente, CA: low cut-off 1.5 RU/ml; high cut-off 1500 RU/ml, , intra-assay and inter-assay coefficients of variation of <5% and <16%, respectively (16)) in EDTA plasma samples obtained at baseline and after each treatment.

123


7

Blood samples were stored at -80°C and analyzed in batch. For 37 patients EDTA-plasma was available for FGF-23 measurements at all four treatment periods. Blood and urinary electrolytes were measured by Roche Modular multi-analyser (Roche Diagnostics, Mannheim, Germany). Albuminuria was measured in 24-hour urine samples in single-batch by benzethonium chloride-based turbidimetric assay. Renin and aldosterone were measured with a chemiluminescence immunoassay (LIASON Aldosterone and LIASON Direct Renin, DiaSorin Deutschland GmbH, Dietzenbach, Germany). We calculated estimated glomerular filtration rate (eGFR) with the creatinine-based Chronic Kidney Disease Epidemiology Collaboration group equation (CKD-EPI) (17).


Data management and statistical analysis were performed with SPSS Statistics 22. Data are re-ported as mean ± standard deviation or median (1st-3rd quartile), as appropriate, after assessing normality by plotting the data. Statistical significance of changes in FGF-23 plasma, aldosterone and renin concentrations before and after sodium-chloride infusion were assessed by Wilcoxon Signed Rank test. In the DN cohort, data were natural log (ln)-transformed, as appropriate, and were compared by the Friedman-test for dependent variables in case of skewed distribution. Normally distributed dependent variables were compared by one-way ANOVA with repeated measures. Multivariable regression analysis was performed to investigate the association of baseline FGF-23 with ln-transformed residual proteinuria after each of the four individual treat-ment periods in the diabetic nephropathy cohort. We first assessed the relationship between FGF-23 as independent and residual proteinuria as dependent variable in univariate regression analysis as described earlier(16). Subsequently, we studied the relationship between FGF-23 and residual proteinuria at the end of each treatment period in a model adjusted for “base-line” proteinuria, i.e. proteinuria during regular sodium diet and placebo. Finally, we further adjusted for creatinine clearance, a potential confounder of the relation between FGF-23 and antiproteinuric response, and repeated these analyses with estimated GFR. We constructed multiplicative interaction terms for FGF-23 and proteinuria, creatinine clearance and estimated GFR respectively.

Results

Volume Loading and FGF-23 Concentrations in Hypertension

We first studied the effect of intravenous sodium loading on plasma FGF-23 in 12 hypertensive individuals without CKD stage 3 or higher, i.e. with eGFR > 60 mL/min/1.73m2. These patients were 45 ± 13 years old and had normal renal function; further characteristics are presented in Table 1. Median FGF-23 plasma concentrations at baseline were 68 [58-97] RU/mL. The infusion of 2 liters isotonic saline in 4 hours did not change FGF-23 concentrations (P = 0.37, Figure

Chapter 7

124

1). Plasma renin concentration did not significantly change (from 4.5 [1.3-14.4] pg/mL to 1.8 [0.8-9.6] pg/mL, P = 0.24), whereas aldosterone decreased significantly from 86 [70-140] pg/mL to 58 [0-64] pg/mL as expected (P = 0.003).

Table 1. Baseline Characteristics

HT patients DN patients

n = 12 n = 45

Age, years 45 ± 13 65 ± 9

Male gender, n (%) 4 (33) 38 (84)

BMI, kg/m2* 28 ± 4 32 ± 5

eGFR, mL/min/1.73 m2 101 ± 18 65 ± 25

Proteinuria, g/d 0.1 [0.1-0.2] 1.1 [0.5-3.2]

Albuminuria, mg/d 11 [7-14] 649 [230-2008]

Systolic blood pressure, mmHg 178 ± 33 147 ± 16

Diastolic blood pressure, mmHg 104 ± 18 82 ± 10

Creatinine clearance, mL/min 113 ± 27 101 ± 47

Sodium excretion, mmol/24h 183 ± 61 224 ± 76

HbA1c, % N/A 7.1 ± 0.8

Diabetes duration, years N/A 11.8 ± 7.6

Macrovascular disease, n (%) 1 (8) 21 (47)

Coronary artery disease, n (%) 1 14 (31)

Stroke (CVA, TIA) 0 6 (13)

Peripheral artery disease, n (%) 0 7 (16)

Antihypertensives

ACE-inhibitor, n (%) 1 (8) 45 (100)

Beta blocker, n (%) 0 27 (60)

Alpha blocker, n (%) 4 (33) 4 (9)

Calcium-channel blocker, n (%) 5 (42) 27 (60)

Diuretic therapy 0 0

Vitamin D treatment, n (%) 0 3 (7)

Phosphate binder treatment, n 0 0

Abbreviations: HT, arterial hypertensive; DN, diabetic nephropathy; BMI, body mass index; eGFR, estimated glo-merular filtration rate; HbA1c, glycated hemoglobin; CVA, cerebrovascular accident; TIA, transient ischemic attack; ACE-inhibitor, angiotensin-converting enzyme inhibitor.

Volume Reduction and FGF-23 Concentrations in Diabetic Nephropathy

Baseline characteristics of the study population are depicted in Table 1.The DN patients were 65 ± 9 years old with a mean eGFR of 65 ± 25 mL/min/1.73 m2 and proteinuria of 1.1 g/d (0.5-3.2 g/d). During ACEi monotherapy and regular sodium diet, plasma FGF-23 concentration was 94 [73-141] RU/mL. Six weeks of treatment with add-on hydrochlorothiazide did not significantly

125


7

change the FGF-23 plasma concentration (post-treatment FGF-23 levels are presented in Table 2). Similarly, 6 weeks of add-on low sodium diet did not affect FGF-23 plasma concentration. Combination therapy of both low sodium diet and hydrochlorothiazide, in addition to ACEi treatment, resulted in a non-significant increase in FGF-23 to 111 [81-160] RU/mL (P=0.15). Treatment with only hydrochlorothiazide or a low sodium diet lowered proteinuria, as reported before (15), with a stronger proteinuric reduction after combination therapy. Similarly, both individual and combined interventions influenced volume status, as reflected by a reduction in body weight (Table 2). Only combination therapy resulted in a decrease in creatinine clear-ance, whereas hydrochlorothiazide treatment in itself also reduced eGFR (15). Serum calcium and phosphate and urinary phosphate excretion did not differ across treatment periods (Table 2). Urinary calcium excretion dropped from 1.6 (0.9‒3.3) mmol/d during regular sodium diet and background ACEi therapy to 1.0 (0.4‒2.6) mmol/d with add-on hydrochlorothiazide, 1.3 (0.6‒2.7) mmol/d with add-on low sodium diet and to 0.7 (0.4‒1.7) mmol/d with combination therapy, respectively (P < 0.001).

FGF-23 and Antiproteinuric Response to Volume Interventions in DN

The residuals of proteinuria were normally distributed after Ln-transformation. In univariable regression analysis, FGF-23 was significantly associated with residual proteinuria during regular sodium diet and add-on hydrochlorothiazide. The association was not significant during low sodium diet, but was again significant for the combination therapy of hydrochlorothiazide with low sodium diet (Table 3, model 1). In multivariable regression analysis, as expected, baseline proteinuria during regular sodium diet outperformed FGF-23 as a correlate of proteinuria after add-on hydrochlorothiazide (Table 3, model 2). Also during low sodium diet, FGF-23 was not sig-nificantly associated with the antiproteinuric response when adjusted for baseline proteinuria (Table 3). FGF-23 plasma concentrations were also not significantly correlated with the anti-proteinuric response in the combination therapy group. When we used estimated GFR instead of creatinine clearance, this did not materially change the results. Invocation of multiplicative

Figure 1. Effect of 2 liters of saline infusion on FGF-23 concentrations after 4 hours.P-value reflects Wilcoxon Signed Rank test. Abbrevia-tions: FGF-23, fibroblast growth factor 23; RU, relative units.

Chapter 7

126 127


7

Tabl

e 2.

Clin

ical

Par

amet

ers d

urin

g Di

ffere

nt T

reat

men

t Per

iods

Varia

bles

Uni

tAC

Ei+

RSAC

Ei+H

CT+R

SAC

Ei+

LSAC

Ei+H

CT+L

SRe

fere

nce

valu

eP-

valu

e

Clin

ical

and

gen

eral

labo

rato

ry p

aram

eter

s

Prot

einu

riag/

d1.

1 (0

.5‒3

.2)

0.7

(0.5

‒1.7

)0.

7 (0

.4‒2

.2)

0.5

(0.3

‒1.3

)<0

.3<0

.001

Prot

einu

ria re

ducti

on%

-27

(38)

28 (2

6)46

(29)

<0.0

01

Prot

einu

ria/c

reati

nine

g/m

mol

0.07

(0.0

4‒0.

22)

0.04

(0.0

3‒0.

14)

0.05

(0.0

3‒0.

18)

0.03

(0.0

2‒0.

10)

<0.0

01

Albu

min

uria

mg/

d64

9 (2

30‒2

008)

342

(215

‒120

5)30

7 (1

47‒1

072)

261

(125

‒878

)<3

0<0

.001

Syst

olic

blo

od p

ress

ure

mm

Hg14

7 ±

1613

5 ±

1614

1 ±

1612

9 ±

14<1

40<0

.001

SBP

redu

ction

%-

7.6

± 8.

03.

4 ±

8.6

11.2

± 8

.3<0

.001

Body

wei

ght

kg10

1.8

± 18

.210

0.1

± 17

.599

.5 ±

18.

198

.1 ±

17.

8<0

.001

Hem

oglo

bin

mm

ol/L

8.6

± 0.

88.

5 ±

0.7

8.5

± 0.

98.

5 ±

0.9

7.5‒

10.6

0.59

eGFR

mL/

min

/1.7

3 m

265

± 2

560

± 2

765

± 2

759

± 2

5<0

.001

Crea

tinin

e cl

eara

nce

mL/

min

101

± 47

97 ±

47

99 ±

48

88 ±

42

0.00

4

Body

wei

ght

kg10

2 ±

1810

0 ±

1899

± 1

898

± 1

8<0

.001

24h

sodi

um e

xcre

tion

mm

ol/d

224

± 76

224

± 88

148

± 65

164

± 71

100‒

250

<0.0

01

Plas

ma

sodi

umm

mol

/L14

0.3

± 3.

113

9.6

± 3.

213

9.7

± 3.

313

8.0

± 3.

713

5‒14

5<0

.001

Min

eral

met

abol

ism

par

amet

ers

FGF-

23RU

/mL

94 (7

3‒14

1)99

(74‒

148)

99 (7

5‒13

5)11

1 (8

1‒16

0)<1

250.

15

Seru

m p

hosp

hate

mm

ol/L

1.00

± 0

.15

1.01

± 0

.14

1.01

± 0

.14

1.04

± 0

.18

0.70

‒1.5

00.

31

24h

phos

phat

e ex

creti

onm

mol

/d25

.8 ±

11.

425

.6 ±

10.

422

.9 ±

9.0

23.2

± 9

.7<4

8.0

0.13

Seru

m c

alci

umm

mol

/L2.

33 ±

0.1

12.

34 ±

0.1

02.

32 ±

0.1

02.

36 ±

0.1

12.

20‒2

.60

0.17

24h

calc

ium

exc

retio

nm

mol

/d1.

6 (0

.9‒3

.3)

1.0

(0.4

‒2.6

)1.

3 (0

.6‒2

.7)

0.7

(0.4

‒1.7

)2.

5‒7.

5<0

.001

Abbr

evia

tions

: ACE

i, an

giot

ensin

-con

verti

ng e

nzym

e in

hibi

tor;

RS, r

egul

ar s

odiu

m d

iet;

HCT,

hyd

roch

loro

thia

zide;

SBP

, sys

tolic

blo

od p

ress

ure;

eGF

R, e

stim

ated

glo

mer

ular

fil

trati

on ra

te; F

GF-2

3, fi

brob

last

gro

wth

fact

or 2

3.

Chapter 7

126 127


7

Tabl

e 3.

Mul

tivar

iabl

e Re

gres

sion

Anal

ysis

for L

n Re

sidua

l Pro

tein

uria

afte

r Diff

eren

t Tre

atm

ents

Mod

elDe

term

inan

tB

95%

CI B

Stan

dard

ized

β β

P-va

lue

R2

ACEi

+HCT

1FG

F-23

0.5

82 0

.133

-1.0

30 0

.412

0.01

0.17

2FG

F-23

0.2

77 0

.016

-0.5

38 0

.196

0.04

0.75

Prot

einu

ria 0

.663

0.5

08-0

.818

0.7

90<0

.001

3FG

F-23

0.2

33–0

.046

-0.5

13 0

.165

0.10

0.75

Prot

einu

ria 0

.649

0.4

91-0

.808

0.7

75<0

.001

CrCl

–0.0

02–0

.005

-0.0

02–0

.088

0.37

ACEi

+LS

1FG

F-23

0.4

56–0

.124

-1.0

35 0

.264

0.12

0.07

2FG

F-23

0.0

29–0

.233

-0.2

91 0

.017

0.82

0.83

Prot

einu

ria 0

.928

0.7

73-1

.083

0.9

06<0

.001

3FG

F-23

0.0

86–0

.192

-0.3

64 0

.050

0.53

0.84

Prot

einu

ria 0

.928

0.7

88-1

.103

0.9

23<0

.001

CrCl

0.0

02–0

.002

-0.0

06 0

.094

0.24

ACEi

+HCT

+LS

1FG

F-23

0.5

28 0

.024

-1.0

31 0

.343

0.04

0.12

2FG

F-23

0.2

20–0

.132

-0.5

72 0

.143

0.21

0.61

Prot

einu

ria 0

.669

0.4

60-0

.878

0.7

33<0

.001

3FG

F-23

0.2

70–0

.109

-0.6

48 0

.175

0.16

0.62

Prot

einu

ria 0

.685

0.4

70-0

.899

0.7

49<0

.001

CrCl

0.0

02–0

.003

-0.0

07 0

.093

0.45

FGF-

23 a

nd p

rote

inur

ia w

ere

Ln-t

rans

form

ed. A

ll de

term

inan

t va

riabl

es a

re u

nder

ACE

i+re

gula

r so

dium

die

t co

nditi

ons.

Abb

revi

ation

s an

d de

finiti

ons:

Ln,

nat

ural

log-

tran

sfor

med

; B, u

nsta

ndar

dize

d co

effici

ent;

95%

CI,

95%

con

fiden

ce in

terv

al; S

tand

ardi

zed

β, st

anda

rdize

d co

effici

ent;

ACEi

, ang

iote

nsin

con

verti

ng e

nzym

e in

hibi

tor;

HCT,

hy

droc

hlor

othi

azid

e; L

S, lo

w so

dium

die

t; FG

F-23

, fibr

obla

st g

row

th fa

ctor

23;

CrC

l, cr

eatin

ine

clea

ranc

e.

Chapter 7

128

interaction terms did not suggestdemonstrate interaction between FGF23 and proteinuria, creatinine clearance or estimated GFR, respectively (all P-interaction > 0.1).

Discussion

In the current study we tested the hypothesis that volume intervention would impact FGF-23 concentrations in two independent settings, namely in patients with hypertension with pre-served renal function and in DN patients. Such finding would support the existence of a negative feedback loop, where volume expansion could suppress FGF-23, while volume depletion could increase FGF-23 as counterpart regulatory response to FGF-23-induced sodium retention. However, neither acute volume expansion nor chronic volume depletion changed FGF-23 con-centrations.

Cardiovascular disease is highly prevalent in patients with CKD and the main cause of mortality in patients with CKD. Increased FGF-23 plasma concentrations are known to be independent predictors of adverse cardiovascular outcome in patients with CKD and in individuals with normal renal function (3, 18-21). In these observational studies FGF-23 was more compel-lingly associated with acute heart failure than with atherosclerotic events. Given the consistent associations between FGF-23 and markers of volume status in previous studies (22-24), and the implicated role for FGF-23 in volume homeostasis (11, 13, 22), we sought to investigate whether, conversely, an acute increase in volume status influences FGF-23 concentrations. We found that acute expansion of extracellular volume by sodium-chloride infusion did not reduce FGF-23 concentrations in patients with arterial hypertension.

This negative result may be explained by the short interval between the volume interven-tion and the FGF-23 measurement. In comparison, the increase of FGF-23 following dietary phosphate intake takes multiple hours to develop (25). On the other hand, acute changes in volume status such as cardiogenic shock are known to suddenly increase FGF-23 to far higher concentrations within a day and even on admission, respectively (26). Secondly, we also as-sessed the effects of chronic interventions, after homeostatic readjustment could have taken place. The diabetic nephropathy patients had FGF-23 concentrations that are typically observed in patients with mildly impaired renal function. The volume-depleting interventions of dietary sodium restriction and hydrochlorothiazide did not significantly increase FGF-23 concentrations in these patients. Intensification of treatment with hydrochlorothiazide or a low sodium diet did not change FGF-23 concentrations. This is in line with our earlier observations in non-diabetic proteinuric CKD patients, where low sodium diet or add-on angiotensin-receptor blockade did not increase FGF-23 concentrations (13). However, combination of low sodium diet with HCT did show a small but non-significant increase in FGF-23 concentrations. This small increase is

129


7

probably caused by the concomitant drop in renal function, as FGF-23 starts to increase mark-edly when renal function drops below ~60 mL/min/1.73 m2 (1). In that perspective, a greater increase in FGF-23 concentrations may have been expected. Of note, the observed decrease in renal function is considered an indicator of therapeutic efficacy, since such effect has been associated with long-term preserved renal function (27). Our results suggest that the relation-ship between FGF-23 and volume status is not bidirectional in patients with mild to moderate chronic kidney disease and in patients with normal renal function. The effects of FGF-23 on volume status as proposed by others may be considered an ‘off-target effect’ of FGF-23, rather than that volume status triggers a FGF-23 response in CKD patients with mildly to moderately impaired renal function. In the setting of haemodialysis, on the other hand, the more extreme changes in volume status may result in a stronger correlation with FGF-23 concentrations (14).

Since both dietary sodium restriction and diuretic therapy reduced body weight and residual proteinuria, and volume depletion is known to potentiate the anti-proteinuric response (28), we subsequently analyzed the relationship between FGF-23 and residual proteinuria. However, we could not demonstrate a strong relation of FGF-23 with the response to anti-proteinuric response. This finding seems at variance with our earlier report, where a high FGF-23 concen-tration was correlated with an impaired antiproteinuric response to low sodium diet (13). In the current study, the correlation of FGF-23 with residual proteinuria was lost when we adjusted for renal function. An explanation may be that in the current study there were more patients with lower proteinuria levels (13 of 45 had proteinuria <0.5 g/d, whereas in our previous study only 5 of 47 had proteinuria <0.5 g/d). Patients with higher proteinuria reabsorb sodium more avidly (29), which makes sodium restriction a particular helpful strategy in these patients (as reviewed elsewhere (30)). In addition, in the current study, the lower proteinuria levels before therapy intensification may have precluded the identification of determinants of proteinuria after therapy intensification in our regression analyses. Also, the smaller sample size likely led to less statistical power to detect an effect. Although non-significant, our current findings in patients with diabetic nephropathy point toward a similar direction of effects as in our earlier report.

Strengths of our study include the use of a randomized controlled cross-over trial where we could assess three treatment-combinations in the same subjects targeting volume status. Further, we performed a prospective experiment in patients who received intravenous sodium-chloride. FGF-23 plasma concentrations were determined using the same assay, enabling a comparison of the effects. Limitations of the study that deserve to be mentioned are the post hoc nature of the analysis in diabetic nephropathy study and the small number of patients increasing the chance of a type II error (false negative finding), particularly in the infusion experiment. Therefore, larger experiments are needed to confirm our findings. Since phosphate intake was not controlled during the studies, changes in dietary phosphate might have influenced

Chapter 7

130

the results, although we did not observe any differences in urinary phosphate excretion or serum phosphate in response to volume interventions. Additionally, in the infusion experiment we investigated patients with arterial hypertension, who could have suffered from subclinical volume retention, which could have prevented a significant response of FGF-23 levels to acute expansion of extracellular volume.

Further, the interaction between FGF-23 and volume status appears to be stronger in protein-uric CKD, and proteinuria is a strong correlate of FGF-23 concentration (31), suggesting that the effects of fluid administration should be assessed in arterial hypertensive patients with proteinuria. The diabetic nephropathy patients were all on background ACE-inhibitor therapy. This may have altered the FGF-23‒klotho‒vitamin D axis, uncoupling the effect of increased renin expression by angiotensin 2 on FGF-23 concentrations, so that an additional effect of sodium restriction or hydrochlorothiazide on FGF-23 plasma concentrations might have been precluded (32).

In conclusion, we could not demonstrate any effect of acute or chronic volume interventions on plasma FGF-23 concentrations in patients with diabetic nephropathy or hypertension and normal renal function, respectively. Our data do not support a direct feedback mechanism of volume status on FGF-23. Future studies may address whether lowering of FGF-23 by other means in patients prone to volume retention improves outcomes.

Acknowledgements

This work is supported by a consortium grant from the Dutch Kidney Foundation (NIGRAM consortium, grant no CP10.11). The NIGRAM consortium consists of the following principal investigators: Pieter M. ter Wee, Marc G. Vervloet (VU University Medical Center, Amsterdam, the Netherlands), René J. Bindels, Joost G. Hoenderop (Radboud University Medical Center Ni-jmegen, the Netherlands), Gerjan Navis, Jan-Luuk Hillebrands and Martin H. de Borst (University Medical Center Groningen, the Netherlands). Dr. De Borst is supported by a grant from the Netherlands Organization for Scientific Research (Veni grant 016.146.014). The funding sources had no role in the design and conduct of the study, the collection, management, analysis, and interpretation of the data, preparation, review, or approval of the manuscript, or the decision to submit the report for publication.

131


7

Conflict of Interest Statement

None declared. Parts of this work were presented at the 3rd Workshop of the CKD-MBD working group of the European Renal Association-European Dialysis and Transplantation Association (ERA-EDTA), 5 December 2015, Milan, Italy.

Chapter 7

132

References

1. Isakova T, Wahl P, Vargas GS, et al. Fibroblast growth factor 23 is elevated before parathyroid hormone and phosphate in chronic kidney disease. Kidney Int. 2011;79(12):1370-1378.

2. Scialla JJ, Wolf M. Roles of phosphate and fibroblast growth factor 23 in cardiovascular disease. Nat Rev. 2014;10(5):268-278.

3. Ix JH, Katz R, Kestenbaum BR, et al. Fibroblast growth factor-23 and death, heart failure, and cardio-vascular events in community-living individuals: CHS (Cardiovascular Health Study). J Am Coll Cardiol. 2012;60(3):200-207.

4. Lutsey PL, Alonso A, Selvin E, et al. Fibroblast growth factor-23 and incident coronary heart disease, heart failure, and cardiovascular mortality: the Atherosclerosis Risk in Communities study. J Am Heart Assoc. 2014;3(3):e000936.

5. Kestenbaum B, Sachs MC, Hoofnagle AN, et al. Fibroblast growth factor-23 and cardiovascular disease in the general population: the Multi-Ethnic Study of Atherosclerosis. Circ Fail. 2014;7(3):409-417.

6. Faul C, Amaral AP, Oskouei B, et al. FGF23 induces left ventricular hypertrophy. J Clin Invest. 2011;121(11):4393-4408.

7. Grabner A, Amaral AP, Schramm K, et al. Activation of Cardiac Fibroblast Growth Factor Receptor 4 Causes Left Ventricular Hypertrophy. Cell Metab. 2015;22(6):1020-1032.

8. Agarwal I, Ide N, Ix JH, et al. Fibroblast growth factor-23 and cardiac structure and function. J Am Heart Assoc. 2014;3(1):e000584.

9. Xie J, Yoon J, An S-W, Kuro-o M, Huang C-L. Soluble Klotho Protects against Uremic Cardiomyopathy Independently of Fibroblast Growth Factor 23 and Phosphate. J Am Soc Nephrol. 2015;26(5):1150-1160.

10. Shalhoub V, Shatzen EM, Ward SC, et al. FGF23 neutralization improves chronic kidney disease-associated hyperparathyroidism yet increases mortality. J Clin Invest. 2012;122(7):2543-2553.

11. Andrukhova O, Slavic S, Smorodchenko A, et al. FGF23 regulates renal sodium handling and blood pres-sure. EMBO Mol Med. 2014;6(6):744-759.

12. Baia LC, Humalda JK, Vervloet MG, et al. Fibroblast growth factor 23 and cardiovascular mortality after kidney transplantation. Clin J Am Soc Nephrol. 2013;8(11):1968-1978.

13. Humalda JK, Heerspink HJL, Kwakernaak AJ, et al. Fibroblast growth factor 23 and the antiproteinuric re-sponse to dietary sodium restriction during renin-angiotensin-aldosterone system blockade. Am J Kidney Dis. 2015;65(2):259-266.

14. Humalda JK, Riphagen IJ, Assa S, et al. Fibroblast growth factor 23 correlates with volume status in hae-modialysis patients and is not reduced by haemodialysis. Nephrol Dial Transplant. November 2015. Epub ahead of print. doi:10.1093/ndt/gfv393.

15. Kwakernaak AJ, Krikken JA, Binnenmars SH, et al. Effects of sodium restriction and hydrochlorothiazide on RAAS blockade efficacy in diabetic nephropathy: a randomised clinical trial. lancetDiabetes Endocrinol. 2014;2(5):385-395.

16. Heijboer AC, Levitus M, Vervloet MG, et al. Determination of fibroblast growth factor 23. Ann Clin Bio-chem. 2009;46(Pt 4):338-340.

17. Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):604-612.

18. Scialla JJ, Xie H, Rahman M, et al. Fibroblast growth factor-23 and cardiovascular events in CKD. J Am Soc Nephrol. 2014;25(2):349-360.

19. Seiler S, Rogacev KS, Roth HJ, et al. Associations of FGF-23 and sKlotho with cardiovascular outcomes among patients with CKD stages 2-4. Clin J Am Soc Nephrol. 2014;9(6):1049-1058.

20. Parker BD, Schurgers LJ, Brandenburg VM, et al. The associations of fibroblast growth factor 23 and un-carboxylated matrix Gla protein with mortality in coronary artery disease: the Heart and Soul Study. Ann Intern Med. 2010;152(10):640-648.

21. Udell JA, Morrow DA, Jarolim P, et al. Fibroblast growth factor-23, cardiovascular prognosis, and ben-efit of angiotensin-converting enzyme inhibition in stable ischemic heart disease. J Am Coll Cardiol. 2014;63(22):2421-2428.

22. Baia LC, Heilberg IP, Navis G, de Borst MH, investigators N. Phosphate and FGF-23 homeostasis after kidney transplantation. Nat Rev. 2015;11(11):656-666.

23. Gruson D, Lepoutre T, Ketelslegers J-M, Cumps J, Ahn SA, Rousseau MF. C-terminal FGF23 is a strong predictor of survival in systolic heart failure. Peptides. 2012;37(2):258-262.

24. Gruson D, Ferracin B, Ahn SA, Rousseau MF. Comparison of fibroblast growth factor 23, soluble ST2 and Galectin-3 for prognostication of cardiovascular death in heart failure patients. Int J Cardiol. 2015;189:185-187.

25. Vervloet MG, van Ittersum FJ, Buttler RM, Heijboer AC, Blankenstein MA, ter Wee PM. Effects of dietary phosphate and calcium intake on fibroblast growth factor-23. Clin J Am Soc Nephrol. 2011;6(2):383-389.

26. Fuernau G, Pöss J, Denks D, et al. Fibroblast growth factor 23 in acute myocardial infarction complicated by cardiogenic shock: a biomarker substudy of the Intraaortic Balloon Pump in Cardiogenic Shock II (IABP-SHOCK II) trial. Crit Care. 2014;18(6):713.

27. Apperloo AJ, de Zeeuw D, de Jong PE. A short-term antihypertensive treatment-induced fall in glomerular filtration rate predicts long-term stability of renal function. Kidney Int. 1997;51(3):793-797.

28. Vogt L, Waanders F, Boomsma F, de Zeeuw D, Navis G. Effects of dietary sodium and hydrochlorothiazide on the antiproteinuric efficacy of losartan. J Am Soc Nephrol. 2008;19(5):999-1007.

29. Svenningsen P, Friis UG, Versland JB, et al. Mechanisms of renal NaCl retention in proteinuric disease. Acta Physiol (Oxf). 2013;207(3):536-545. doi:10.1111/apha.12047.

30. Humalda JK, Navis G. Dietary sodium restriction: a neglected therapeutic opportunity in chronic kidney disease. Curr Opin Nephrol Hypertens. 2014;23(6):533-540.

31. Vervloet MG, van Zuilen AD, Heijboer AC, et al. Fibroblast growth factor 23 is associated with proteinuria and smoking in chronic kidney disease: An analysis of the MASTERPLAN cohort. BMC Nephrol. 2012;13:20.

32. de Borst MH, Vervloet MG, ter Wee PM, Navis G. Cross talk between the renin-angiotensin-aldosterone system and vitamin D-FGF-23-klotho in chronic kidney disease. J Am Soc Nephrol. 2011;22(9):1603-1609.

Chapter 8Concordance of Dietary Sodium Intake and Concomitant Phosphate Load: Implications for Sodium InterventionsJelmer K. Humalda†

Charlotte A. Keyzer†

S. Heleen BinnenmarsArjan J. KwakernaakMaartje C.J. SlagmanGoos D. LavermanStephan J. L. BakkerMartin H. de BorstGerjan Navis

†contributed equallyNutr Metab Cardiovasc Dis. 2016 Aug;26(8):689-96.

Chapter 8

136

Abstract

Background and aims: Both a high dietary sodium and high phosphate load are associated with an increased cardiovascular risk in patients with chronic kidney disease (CKD), and possibly also in non-CKD populations. Sodium and phosphate are abundantly present in processed food. We hypothesized that (modulation of) dietary sodium is accompanied by changes in phosphate load across populations with normal and impaired renal function.

Methods and Results: We first investigated the association between sodium and phosphate load in 24-h urine samples from healthy controls (n = 252), patients with type 2 diabetes mellitus (DM, n = 255) and renal transplant recipients (RTR, n = 705). Secondly, we assessed the effect of sodium restriction on phosphate excretion in a nondiabetic CKD cohort (ND-CKD: n = 43) and a diabetic CKD cohort (D-CKD: n = 39). Sodium excretion correlated with phosphate excretion in healthy controls (R = 0.386, P < 0.001), DM (R = 0.490, P < 0.001), and RTR (R = 0.519, P < 0.001). This correlation was also present during regular sodium intake in the intervention studies (ND-CKD: R = 0.491, P < 0.001; D-CKD: R = 0.729, P < 0.001). In multivariable regression analysis, sodium excretion remained significantly correlated with phosphate excretion after adjustment for age, gender, BMI, and eGFR in all observational cohorts. In ND-CKD and D-CKD moderate sodium restriction reduced phosphate excretion (31 ± 10 to 28 ± 10 mmol/d; P = 0.04 and 26 ± 11 to 23 ± 9 mmol/d; P = 0.02 respectively).

Conclusions: Dietary exposure to sodium and phosphate are correlated across the spectrum of renal function impairment. The concomitant reduction in phosphate intake accompanying sodium restriction underlines the off-target effects on other nutritional components, which may contribute to the beneficial cardiovascular effects of sodium restriction.

137

Concordance of Dietary Sodium Intake and Concomitant Phosphate Load

8

Introduction

Dietary interventions form an essential component of the treatment of chronic kidney disease (CKD). Sodium restriction is beneficial for patients in all stages of CKD, reviewed in Ref. (1), and a restriction to <5 g of salt [<2000 mg of sodium] daily is advised in CKD guidelines (2). Notwithstanding these recommendations, most CKD patients consume almost twice as much salt: about 9 g a day, which reflects the high sodium intake in the Western general population (3, 4). This directly hampers the efficacy of renin-angiotensin-aldosterone system (RAAS) blockade, the standard therapy for patients with chronic kidney disease (5).

Phosphate restriction is nowadays only advised in the setting of end stage renal disease (ESRD), but has been proposed as treatment target earlier in predialysis CKD (6, 7). This recommen-dation is based on evidence that higher serum phosphate concentrations are associated with increased mortality in patients with moderately impaired renal function (8) and even in the healthy population (9). High-normal serum phosphate concentrations also correlate with an impaired response to RAAS blockade in CKD patients (10, 11).

Dietary interventions typically address one single nutrient, i.e. ‘avoid phosphate-rich products’. This reductionist nutrient approach is one of the reasons why preventive nutrition did not suc-ceed in the prevention of diet-related chronic diseases over the last decades (12). Assessing food as whole products or dietary patterns may be a more fruitful strategy.

Reducing dietary phosphate intake is a challenge, as phosphate is present ubiquitously in food products (13). Additive-rich, processed products can easily contain 66% more phosphate than its non-phosphate based preservative equivalent (14). Moreover, the bioavailability of additive-derived inorganic phosphate is almost 100%, whereas phosphate from animal or vegetable sources is far less avidly absorbed (60% and 40%, respectively (15)). As many additives contain both sodium and phosphate (e.g. disodiumdiphosphate), it is not surprising that a recent RCT found that an additive-enriched diet increases sodium and phosphate intake concomitantly by 60% (16). These data suggest that intake of sodium and phosphate are concordant in subjects on a Western diet. If so, dietary sodium restriction can also be anticipated to modulate phos-phate intake, as an off-target effect.

To test these assumptions we first analyzed the association between sodium and phosphate excretion in 24-hour urine collections obtained from prospective cohort studies in CKD and non-CKD populations. Secondly, we studied the effect of a dietary sodium intervention on both sodium and phosphate excretion, in a post-hoc analysis of two clinical trials in CKD patients.

Chapter 8

138

Methods

Study Population

Observational cohortsWe studied three independent observational cohorts recruited in two different centers in the Netherlands.First, we recruited a cohort of healthy controls (HC), consisting of participants in a kidney donor screening program at the University Medical Center Groningen, The Netherlands. Participants had no history of CKD, cardiovascular disease or diabetes, nor did they receive dietary counsel-ing on sodium restriction. Mild hypertension (below 140/90 mmHg with 1-2 antihypertensive drugs) was allowed. More details regarding the healthy controls have been published previously (17).Second, a cohort of diabetics (DM) without overt renal dysfunction was recruited in the ZGT Hospital in Almelo, The Netherlands (METc2008/240), and served as reference diabetes patients as reported earlier (18).Third, a cohort was recruited consisting of renal transplant recipients (RTR) who visited our outpatient clinic between 2008 and 2010 with a functioning graft > 1 year (METc2008/186). Detailed information about this cohort has been published previously (18).For all cohorts, patients with missing 24-hour urine values on sodium or phosphate were ex-cluded for this analysis.

Intervention studiesThe intervention study in nondiabetic CKD patients (ND-CKD) was performed in patients with CKD with blood pressure >125/75 mmHg, creatinine clearance 30 mL/min with no upper limit, and >1.0 g per day proteinuric kidney disease (Dutch Trial Register NTR675), in four Dutch cen-ters (Medical Center Leeuwarden, University Medical Center Groningen, ZGT Hospital Almelo, Martini Hospital Groningen). Main exclusion criteria were diabetes mellitus, blood pressure >180/110 or renal function loss > 6 mL/min/year. The original study investigated the antiprotein-uric efficacy of combination of angiotensin receptor blockade (ARB) with angiotensin-converting enzyme inhibitors (ACEi) –also known as dual blockade– and compared this to the effect of a low sodium diet. All patients underwent 4 six-week treatment periods in a randomized, cross-over design: use of ACEi monotherapy with placebo versus ACEi combined with ARB, in the setting of a low sodium diet or regular sodium diet. For the current study we focus on the six week sodium restriction period targeting a 50 mmol/d Na intake compared to a six week regular sodium in-take period, both during background ACEi (lisinopril 40 mg daily) therapy. Patients received 2-4 counseling sessions with a dietitian, a list with the sodium content of common food products in the Netherlands, were asked to refrain from adding salt to food and to replace sodium-rich with sodium-poor products. The dietitian did not receive a script or training other than the instruction to target 50 mmol/d and 200 mmol/d sodium per day for the low and regular sodium

139


8

intake treatment arms, while keeping other dietary factors, including protein intake, as stable as possible. Dietary compliance was assessed halfway during treatment period by 24-hour urine collection. During regular sodium diet patients were asked to maintain nutritional habits. Data collection was performed at the end of each treatment period. For extensive details we refer to the protocol documented elsewhere (5).

In another study with a similar design, 45 diabetic CKD patients (D-CKD) underwent a six week treatment period with regular sodium intake (maintaining dietary habits) and sodium restriction targeting 50 mmol/day (NTR2366) (18), in three medical centers (ZGT Hospital Almelo, Medi-cal Center Leeuwarden, University Medical Center Groningen). Data collection was performed at the end of each treatment period. Here, patients received 1‒2 counseling sessions with a dietitian and further similar advises as mentioned above. Patients without 24-hour urine values on sodium or phosphate were excluded for this analysis.

Measurements

Creatinine and elektrolytes were measured with routine laboratory methods. Sodium intake and phosphate intake were estimated from 24-hour urinary excretion in all cohorts. In the observational cohorts, 24-hour urine was collected in containers with 5 mL oil and 50 mL chlorhexidine. The intervention trials did not use preservatives for 24-hour urine collections. As there are concerns that phosphate may precipitate when urine pH > 7.0, we performed a sensitivity analysis excluding individuals with urine pH > 7.0. Estimated glomerular filtration rate (eGFR) was calculated with the CKD-EPI equation (19). Clinical measurements were performed at the time of the outpatient clinic visit in all patients.

Statistics

We report mean and standard deviations or median (1st‒3rd quartile) as appropriate. Differences in means for continuous variables were assessed by ANOVA, Kruskal‒Wallis or χ2 as appropriate. As urea excretion was not available in the DM cohort, means between healthy controls and RTR were compared by t-test. The correlation between phosphate and sodium excretion was assessed by Pearson’s correlation test. We used linear multivariable regression analysis with sodium excretion as dependent and phosphate excretion as independent covariate in a first model. Then we constructed the second model together with covariates that may confound the relation: age, gender and BMI to adjust for overt differences in body composition, and eGFR to adjust for differences in solute clearance capacity. In the third model, we introduced urea excre-tion to reflect differences in dietary intake of protein. In the fourth model, calcium excretion was added to account for intestinal calcium absorption as a proxy for calcium intake. Interactions were assessed by invoking multiplicative interaction terms.

Chapter 8

140

In the intervention trials we assessed the effect of dietary sodium restriction on phosphate excretion in patients that had complete 24-h urine collections by paired t-tests per study or Wilcoxon Signed Rank test as appropriate, and analyzed the associations between the percent change in sodium and phosphate excretion with Pearson’s correlation test. Relative changes in excretion between treatment periods were calculated as follows: relative change = (excretion at regular sodium – excretion at low sodium) /excretion at regular sodium × 100%.

Results

Study Populations

We investigated three independent observational cohorts recruited in two different centers in the Netherlands (Table 1). The first consisted of 252 healthy controls (HC), aged 53 ± 10.6 years with an eGFR of 91.1 ± 14.0 mL/min/1.73m2, the second of 255 patients with diabetes that were 63.2 ± 8.9 years old with an eGFR of 72.3 ± 24.4 mL/min/1.73m2, and the third of 705 renal transplant recipients (RTR) aged 53.0 ± 12.8 years with an eGFR of 52.2 ± 20.1 mL/min/1.73m2 on median 5.4 (interquartile range 1.9‒12.2) years after transplantation (17). We also included cross-sectional analyses of two patient cohorts derived from randomized controlled trials, both during regular sodium intake and during low sodium intake (Table 2). The ND-CKD patients were 51.3 ± 13.9 years old and had an eGFR of 59.3 ± 29.1 mL/min/1.73m2 during regular sodium intake. The D-CKD patients were 64.0 ± 8.6 years old, had an eGFR of 66.5±25.2 mL/min/1.73m2 during regular sodium intake, and had a HbA1c of 7.1 ± 0.8%.

Sodium and Phosphate Excretion

Sodium excretion was similar among patients and healthy controls (Table 1). Mean 24-h phos-phate excretion was between 25 and 31 mmol per day. The 24-h phosphate and sodium excretion correlated strongly in all groups: R = 0.386, P < 0.001 in healthy controls, R = 0.490, P < 0.001 in diabetic patients and R = 0.519, P < 0.001 in RTR (Figure 1). In multivariable regression analysis sodium excretion remained significantly correlated with phosphate excretion after adjustment for age, gender, BMI and eGFR in healthy controls (Standardized beta [St. β] = 0.252, P < 0.001, R2 = 0.30), DM (St. β = 0.386, P < 0.001, R2 = 0.35) and RTR (St. β = 0.391, P < 0.001, R2 = 0.38, table 3, model 2). Additional adjustment for urea excretion –reflecting protein intake–did not in-fluence the association between sodium and phosphate excretion (Table 3, model 3). In healthy controls however, significance for sodium excretion was lost after addition of urea excretion. This may suggest an interaction between sodium excretion and urea excretion, i.e. concomitant intake food high in sodium and protein, that explains the variability in phosphate excretion. Indeed, the standardized regression coefficient of sodium excretion also decreased in RTR from 0.391 to 0.11 after introduction of urea in model 3. We found no significant interaction between sodium excretion and urea excretion in its relation to phosphate excretion (P-interaction = 0.7 in

141


8

Table 1: Clinical and Biochemical Parameters of the Observational Cohorts.

HC n= 252 DM n= 255 RTR n= 705

Age, years 53.3 ± 10.6 63.2 ± 8.9 53.0 ± 12.8

Male, n(%) 116 (46) 137 (54) 401 (57)

Weight, kg 79.4 ± 13.8 96.7 ± 18.9 80.4 ± 16.6

BMI 26.0 ± 3.4 33.1 ± 6.0 26.7 ± 4.8

Vitamin D use, n (%) 0 8 (3) 174 (25)

eGFR, ml/min 91.1 ± 14.0 72.3 ± 24.4 52.2 ± 20.1

Systolic blood pressure 125 ± 14 141 ± 16 136 ± 18

Diastolic blood pressure 76 ± 9 76 ± 10 83 ± 11

Serum sodium, mmol/L 142 ± 1.9 138 ± 3.0 141 ± 3

serum phosphate, mmol/L 1.07 ± 0.18 0.99 ± 0.18 0.96 ± 0.21

Urinary sodium, mmol/day 194.2 ± 71.6 189.6 ± 79.4 157.1 ± 62.0

Urinary phosphate, mmol/day 28.1 ± 9.6 26.4 ± 10.9 25.0 ± 8.9

Urinary calcium, mmol/day 5.0 (3.4‒6.8) 3.2 (1.5‒5.2) 2.4 (1.1‒3.9)

Proteinuria, g/day 0.0 (0.0‒0.2) 0.2 (0.1‒0.4) 0.2 (0.0‒0.4)

Urea excretion, mmol/day 404 ± 119 N/A 388 ± 114

Creatinine excretion, mmol/day 13.2 ± 4.2 13.3 ± 4.3 11.6 ± 3.5

Abbreviations: HC, healthy controls, DM, diabetes mellitus patients; RTR, renal transplant recipients; BMI, Body Mass Index; eGFR, estimated Glomerular Filtration Rate; N/A, not available.

Table 2: Clinical and Biochemical Parameters of the Intervention Studies after Regular Sodium Treatment Period

ND-CKD, n = 43 D-CKD, n = 39

Regular Sodium

Low Sodium P-value Regular sodium

Low Sodium P-value

Age, years 51.3 ± 13.9 - 64.0 ± 8.6 -

Male, n(%) 36 (84) - 33 (85) -

BMI 27.5 ± 4.2 - 32.4 ± 5.1 -

Vitamin D use, n(%) 4 (9) 2 (5)

Weight, kg 88.9 ± 17.1 86.3 ± 16.3 <0.001 102.3 ± 18.6 100.7 ± 18.7 <0.001

eGFR, ml/min 59.3 ± 29.1 54.6 ± 26.7 0.05 66.5 ± 25.2 66.7 ± 26 0.6

Systolic blood pressure, mmHg 135 ± 20 125 ± 18 <0.001 146 ± 16 140 ± 16 0.008

Diastolic blood pressure, mmHg 81 ± 14 73 ± 12 <0.001 82 ± 10 78 ± 10 0.007

Serum sodium, mmol/L 141 ± 3 139 ± 3 0.003 140 ± 3 140 ± 3 0.06

Serum phosphate, mmol/L 1.06 ± 0.21 1.11 ± 0.18 0.1 0.99 ± 0.15 1.01 ± 0.14 0.4

Urinary sodium, mmol/day 188.7 ± 58.8 104.4 ± 40.9 <0.001 232.5 ± 72.2 150 ± 69 <0.001

Urinary phosphate, mmol/day 30.7 ± 9.9 28.3 ± 10.1 0.04 26.5 ± 11.5 23.4 ± 9.0 0.02

Urinary urea, mmol/day 386 ± 119 353 ± 109 0.06 422 ± 137 449 ± 197 0.5

Urinary potassium, mmol/day 76 ± 23 75 ± 24 0.3 78 ± 26 83 ± 34 0.3

Urinary creatinine, mmol/day 13.8 ± 4.1 13.5 ± 4.1 0.2 14.3 ± 4.2 13.8 ± 4.0 0.3

Proteinuria, g/day 2.0 (0.9-3.5) 0.9 (0.5-1.7) <0.001 1.1 (0.5-3.2) 0.6 (0.4-2.1) <0.001

Abbreviations: CKD, Chronic Kidney Disease patients without diabetes; D-CKD, CKD patients with diabetes; BMI, Body Mass Index; BSA, Body Surface Area; eGFR, estimated Glomerular Filtration Rate.

Chapter 8

142 143


8

healthy controls and P-interaction = 0.3 in RTR). Introduction of calcium excretion improved all models but did not influence the association between sodium and phosphate excretion (Table 3, model 4). One healthy control (pH 7.16) and six RTRs had urine pH > 7.0 (maximum pH = 7.68). Exclusion of these individuals did not alter conclusions of our analysis. Vitamin D use was only common in the RTR cohort (Tables 1 and 2) and did not materially influence our results. Sodium excretion correlated with phosphate excretion in the vitamin D users (n = 174, St. β = 0.485, P < 0.001) and non-vitamin D users (n = 531, (St. β = 0.528, P < 0.001). Vitamin D use did not attenuate our regression models, e.g. when introduced to model 1 of table 3 (R2 increased from 0.27 to 0.29; coefficient for vitamin D use, St. β = ‒0.150, P < 0.001; coefficient for sodium excretion, St. β = 0.508, P < 0.001).

Figure 1. Correlation of 24-hour Sodium Excretion and Phosphate Excretion in the three Observational Cohorts.

Chapter 8

142 143


8

Tabl

e 3:

Mul

tivar

iabl

e Li

near

Reg

ress

ion

Anal

ysis

of D

eter

min

ants

of P

hosp

hate

Exc

retio

n in

Obs

erva

tiona

l Coh

orts

.

HCDM

*RT

R

Mod

elVa

riabl

eSt

. Bet

aP-

valu

eR2

St. B

eta

P-va

lue

R2St

. Bet

aP-

valu

eR2

1So

dium

exc

retio

n0.

386

<0.0

010.

150.

490

<0.0

010.

240.

519

<0.0

010.

27

2So

dium

exc

retio

n0.

252

<0.0

010.

300.

389

<0.0

010.

350.

391

<0.0

010.

38

BMI

0.20

3<0

.001

–0.0

650.

20.

163

<0.0

01

Gend

er–0

.316

<0.0

01–0

.285

<0.0

01–0

.245

<0.0

01

Age

–0.1

200.

10–0

.250

<0.0

01–0

.028

0.4

eGFR

(CKD

-EPI

)0.

010

0.9

–0.0

020.

90.

200

<0.0

01

3So

dium

exc

retio

n0.

099

0.07

0.49

0.38

9<0

.001

0.35

0.11

1<0

.001

0.58

BMI

0.08

20.

10–0

.065

0.2

0.12

5<0

.001

Gend

er–0

.194

<0.0

01–0

.285

<0.0

01–0

.154

<0.0

01

Age

–0.0

950.

13–0

.250

<0.0

01–0

.061

0.02

eGFR

(CKD

-EPI

)–0

.026

0.7

–0.0

020.

90.

156

0.00

1

Ure

a ex

creti

on0.

518

<0.0

01N

/AN

/A0.

554

<0.0

01

4So

dium

exc

retio

n0.

054

0.29

50.

540.

337

<0.0

010.

400.

097

0.00

20.

60

BMI

0.09

10.

056

–0.0

470.

40.

108

<0.0

01

Gend

er–0

.217

<0.0

01–0

.274

<0.0

01–0

.180

<0.0

01

Age

–0.1

020.

084

–0.2

24<0

.001

–0.0

79<0

.001

eGFR

(CKD

-EPI

)–0

.039

0.5

–0.1

210.

050.

085

0.00

1

Ure

a ex

creti

on0.

454

<0.0

01N

/AN

/A0.

509

<0.0

01

Calc

ium

exc

retio

n0.

242

<0.0

010.

265

<0.0

010.

188

<0.0

01

Abbr

evia

tions

: HC,

hea

lthy

cont

rols,

DM

, dia

bete

s m

ellit

us p

atien

ts; R

TR, r

enal

tran

spla

nt re

cipi

ents

; St.

Beta

, sta

ndar

dize

d be

ta; B

MI,

Body

Mas

s In

dex;

eGF

R, e

stim

ated

Gl

omer

ular

Filt

ratio

n Ra

te; *

Ure

a ex

creti

on m

easu

rem

ents

wer

e no

t ava

ilabl

e (N

/A) f

or th

e DM

coh

ort.

Chapter 8

144

Intervention Studies

We subsequently studied the effect of an intervention in sodium intake, namely moderate so-dium restriction, on phosphate intake as reflected by urinary phosphate excretion. In ND-CKD, sodium restriction from 189 ± 56 to 106 ± 48 mmol/d was accompanied by a reduction in phos-phate excretion from 31 ± 10 to 28 ± 10 mmol/d (P = 0.04). In D-CKD, even a moderate sodium restriction from 224 ± 76 to 148 ± 65 mmol/d led to a concomitant reduction of phosphate excretion from 26 ± 11 to 23 ± 9 mmol/d (P = 0.02, Figure 2). Urinary phosphate and sodium excretion during regular sodium intake correlated strongly in ND-CKD (R = 0.491) and in D-CKD (R = 0.729, both P < 0.001). The relative reduction in urinary sodium excretion and phosphate excretion correlated poorly (ND-CKD: R = 0.248, P = 0.11, and D-CKD: R = 0.065, P = 0.7).

To investigate whether the change in phosphate excretion in response to dietary sodium restric-tion was driven by changes in protein intake, we subsequently adjusted our analyses for the change in 24-hour urinary urea excretion. This further weakened the association between the change in sodium and phosphate excretion (ND-CKD: St. β = –0.047, P = 0.7, D-CKD St. β = 0.107,

Figure 2. Concomitant Effects of a Low Sodium Diet on Phosphate and Urea Excretion.

145


8

P = 0.7). Although sodium restriction did not lower urea excretion significantly (Figure 2), the percent change in urea excretion correlated in itself strongly with percent phosphate reduction in ND-CKD (St. β = 0.634, P < 0.001) and correlated borderline-significantly in D-CKD (St. β = 0.439, P = 0.08).

24-hour excretion of phosphate (left Y-axis), and urea and sodium (right Y-axis) under regular and low sodium diet in ND-CKD (upper panel) and D-CKD (lower panel). P-value reflects paired t-test. ND-CKD, nondiabetic chronic kidney disease; D-CKD, diabetic chronic kidney disease; NS, not-significant.

Discussion

In this analysis we confirm that sodium and phosphate intake are strongly correlated across different stages of chronic kidney disease and in healthy controls. Moreover, a dietary interven-tion aimed solely at sodium restriction achieved a mild but significant concomitant, off-target reduction in phosphate load.

The sodium intake of 10‒12 g of sodium chloride a day in this study equals or is even higher than the already superfluous sodium intake of the general Dutch population of 8.5 g a day (4). This is far more than the maximum of 5 g per day as recommended by chronic kidney disease guideline (20). Also for the general population, the WHO recommends to reduce worldwide sodium in-take to less than 5 g per day for every person (21). The phosphate intake can be estimated from the 24-hour phosphate excretion. The phosphate excretion of our patients was 25‒30 mmol per day [~800‒1000 mg/day], which is comparable with the mean excretion of 1008 mg/day in 481 patients with normal renal function in the PREMIER study (22). This corresponds with an estimated intake by dietary recall of around 43 mmol/day [~1400 mg/day] (23), assuming that 70% of all phosphorus intake is absorbed in the intestine. As of yet, there is no target value for phosphate intake for the healthy population. A phosphate-restricted diet in the setting of ESRD would target a phosphate intake of 700 mg per day, i.e. roughly half of ‘normal’ dietary intake.

The coincidence of high sodium load with a high phosphate load is in line with our hypoth-esis. Food additives contribute substantially to both sodium and phosphate intake (16). Many phosphate-based food additives also contain sodium. For example the mono-, di- and trisodiumphosphates that are used ubiquitously in baking products, beverages, processed cheeses and the sodiumtripolyphosphates used for conservation and stabilizing of meat and fish products (13). Although sodium content is routinely expressed on labels on food products, its phosphate content is not quantified nor clearly mentioned. The high concurrent sodium and phosphate load in our Western patients is at striking variance with the low sodium and phos-

Chapter 8

146

phate excretion rates in individuals of African ancestry living in Africa (24), quantifying the effect of the superfluous, additive-rich Western diet. Furthermore, the correlation between sodium excretion and phosphate excretion was independent from urea excretion in RTR, but not in the healthy controls. This may suggest that RTR are particularly susceptible for the contribution of phosphate-rich additives to their sodium/phosphate load, whereas the correlation of sodium and phosphate in healthy controls appears to be mainly protein-driven. As the correlation was attenuated in RTR, of course protein intake played a large role in the RTR population too. Alternatively, the sodium-phosphate excretion association in healthy controls may have become insignificant because of the smaller size of this cohort.

We report that an intervention targeting solely sodium intake, also achieves a reduction in phosphate excretion. The 10% reduction of 3 mmol/day [~ 92 mg/day] is subtle, however, in perspective of the 5.6 mmol/day [173 mg/day, 23%] reduction achieved by a trial that actively targeted phosphate intake it should not be discarded as trivial (23). Also in ten healthy controls, the change from one week on a low-additive diet to one week on an additive-enhanced diet in-creased phosphate excretion by 4.0 mmol/day [124 mg/day, 20%] (25). Most sodium restriction trials tend to not report urinary phosphate excretion, and vice versa. Thus, it is not surprising yet often overlooked that an intervention aimed at sodium restriction may also exerts effects on other nutrients. It is well-known that dietary sodium restriction leads to a lower protein intake determined by urea excretion (5). This was not significant in our diabetic CKD interven-tion study, maybe because this population had a different dietary pattern (e.g. a bit more meat, and far more added salt or salty snacks), as reflected by higher urea excretion compared with the nondiabetic CKD intervention study. Consequently, the D-CKD patients may strongly reduce sodium intake by reducing added salt, without changing protein intake. Because food frequency questionnaires were not available, we could not identify differences in dietary patterns. Alter-natively, the effect may have been absent due to a too small sample size. In line, in 22 RTR sodium restriction did not significantly reduce urea excretion (26). Nevertheless, the relative reduction of urea correlated with the reduction of phosphate excretion levels and obliterated the contribution of the reduction in sodium excretion in the intervention studies. This suggests that although sodium restriction may partly reduce protein-associated phosphate, the main effect may be reduction of non-organic phosphate intake, i.e. additives.

From a scientific point-of-view this non-specificity of sodium restriction, i.e. off-target effects on phosphate intake, may be bothersome. On the other hand, this reflects the real-life situa-tion and simply emphasizes that sodium, protein and phosphate are overly-represented in the Western diet. Whilst this technically confounds dietary sodium intervention studies, this may offer at the same time an additional clinical benefit: a double-edged sword. One explanation may be that improved adherence to sodium restriction (i.e. avoiding processed foods, additives) concomitantly reduces phosphate load, although this did not translate to a marked correlation

147


8

between relative change in sodium excretion and phosphate excretion in our study. Also, recent concerns about adverse effects of an overzealous sodium restriction may be influenced by ef-fects on other particular nutrients or malnutrition in general. This serves as an example of the effect of sodium restriction on other nutrients.

The strength of this study is that we could combine data from observational studies with the effects of sodium-based interventions in randomized clinical trials. Moreover, our populations cover a broad spectrum of the nephrology outpatient clinic, allowing for generalization of our data. For this study, we could rely on 24-hour urinary excretions as an estimate for sodium and phosphate intake in a stable outpatient setting. No food frequency questionnaires were available in all cohorts. Also in the trial conditions of the dietary intervention studies, due to the relative long intervention period, there are no detailed data on the actual intake. This reflects real-life outpatient conditions, but may be considered a limitation. Our observations thus rely on the premise that 24-hour urinary excretion reflects intake. Taking into account that there are also non-osmotic buffering capacities for sodium (27) and changes in bone-metabolism for phosphate were not assessed, one cannot state that every mmol of sodium eaten is eventu-ally excreted in the steady state. Notwithstanding, a 24-hour urine collection remains the gold standard for dietary intake of the electrolytes sodium and phosphate. Indeed, dietary recall consistently underestimates sodium intake (28), and aforementioned mechanisms would only serve to attenuate the found association rather than confound it.

In conclusion, we found that across different patient populations sodium and phosphate intake are closely related, and that intervention aimed at reduction of sodium also reduces phos-phate. Future studies should explore the interaction between sodium and phosphate handling thoroughly. In the meantime, moderate reduction of sodium intake appears to have beneficial effects on phosphate load. This “off-target” effect supports dietary prescriptions aimed at avoid-ance of processed foods, which should be enforced by dietitians and physicians.

Acknowledgements

C.A.K. and M.H.d.B are partly supported by a consortium grant from the Dutch Kidney Founda-tion (NIGRAM consortium, grant no CP10.11). The NIGRAM consortium consists of the following principal investigators: Pieter M. ter Wee, Marc G. Vervloet (VU University Medical Center, Am-sterdam, the Netherlands), René J. Bindels, Joost G. Hoenderop (Radboud University Medical Center Nijmegen, the Netherlands), Gerjan Navis, Jan-Luuk Hillebrands and Martin H. de Borst (University Medical Center Groningen, the Netherlands). M.H.d.B. is supported by a grant from the Netherlands Organization for Scientific Research (Veni grant no 016.146.014).

Chapter 8

148

References

1. Humalda JK, Navis G. Dietary sodium restriction: a neglected therapeutic opportunity in chronic kidney disease. Curr.Opin.Nephrol.Hypertens. 2014; 23: 533‒540.

2. Kidney Disease: Improving Global Outcomes (KDIGO) Blood Pressure Work Group. KDIGO Clinical Practice Guideline for the Management of Blood Pressure in Chronic Kidney Disease. Kidney Int.Suppl. 2012; 2: 337‒414.

3. van Zuilen AD, Bots ML, Dulger A, et al. Multifactorial intervention with nurse practitioners does not change cardiovascular outcomes in patients with chronic kidney disease. Kidney Int. 2012; 82: 710‒7.

4. Hendriksen MA, van Raaij JM, Geleijnse JM, Wilson-van den Hooven C, Ocke MC, van der ADL. Monitoring salt and iodine intakes in Dutch adults between 2006 and 2010 using 24 h urinary sodium and iodine excretions. Public Health Nutr. 2013; 1‒8.


6. Cozzolino M, Urena-Torres P, Vervloet MG, et al. Is chronic kidney disease-mineral bone disorder (CKD-MBD) really a syndrome? Nephrol.Dial.Transplant. 2014; 29: 1815‒1820.

7. Scialla JJ, Wolf M. Roles of phosphate and fibroblast growth factor 23 in cardiovascular disease. Nat.Rev.Nephrol. 2014; 10: 268‒278.

8. Kestenbaum B, Sampson JN, Rudser KD, et al. Serum phosphate levels and mortality risk among people with chronic kidney disease. J.Am.Soc.Nephrol. 2005; 16: 520‒528.

9. Chang AR, Grams ME. Serum phosphorus and mortality in the Third National Health and Nutrition Exami-nation Survey (NHANES III): effect modification by fasting. Am.J.Kidney Dis. 2014; 64: 567‒573.

10. Zoccali C, Ruggenenti P, Perna A, et al. Phosphate may promote CKD progression and attenuate renopro-tective effect of ACE inhibition. J.Am.Soc.Nephrol. 2011; 22: 1923‒1930.

11. Di Iorio BR, Bellizzi V, Bellasi A, et al. Phosphate attenuates the anti-proteinuric effect of very low-protein diet in CKD patients. Nephrol.Dial.Transplant. 2013; 28: 632‒640.

12. Fardet A, Rock E. The healthy core metabolism: A new paradigm for primary preventive nutrition. J.Nutr.Health Aging 2015; 1‒9.

13. Kalantar-Zadeh K, Gutekunst L, Mehrotra R, et al. Understanding sources of dietary phosphorus in the treatment of patients with chronic kidney disease. Clin.J.Am.Soc.Nephrol. 2010; 5: 519‒530.

14. Cupisti A, Benini O, Ferretti V, Gianfaldoni D, Kalantar-Zadeh K. Novel differential measurement of natural and added phosphorus in cooked ham with or without preservatives. J.Ren.Nutr. 2012; 22: 533‒540.

15. Gonzalez-Parra E, Gracia-Iguacel C, Egido J, Ortiz A. Phosphorus and nutrition in chronic kidney disease. Int.J.Nephrol. 2012; 597605.

16. Carrigan A, Klinger A, Choquette SS, et al. Contribution of food additives to sodium and phosphorus content of diets rich in processed foods. J.Ren.Nutr. 2014; 24: 13‒9, 19e1.

17. van den Berg E, Pasch A, Westendorp WH, et al. Urinary sulfur metabolites associate with a favorable cardiovascular risk profile and survival benefit in renal transplant recipients. J.Am.Soc.Nephrol. 2014; 25: 1303‒1312.

18. Kwakernaak AJ, Krikken JA, Binnenmars SH, et al. Effects of sodium restriction and hydrochlorothiazide on RAAS blockade efficacy in diabetic nephropathy: a randomised clinical trial. Lancet Diabetes Endocrinol. 2014; 2: 385‒395.

19. Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann.Intern.Med. 2009; 150: 604‒612.

149


8

20. Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group. KDIGO 2012 Clinical Practice Guideline for the Evaluation and Management of Chronic Kidney Disease. Kidney Int.Suppl. 2013; 3: 1‒150.

21. Beaglehole R, Bonita R, Horton R, et al. Priority actions for the non-communicable disease crisis. Lancet 2011; 377: 1438‒1447.

22. Chang A, Batch BC, McGuire HL, et al. Association of a reduction in central obesity and phosphorus intake with changes in urinary albumin excretion: the PREMIER study. Am.J.Kidney Dis. 2013; 62: 900‒907.

23. Williams PS, Stevens ME, Fass G, Irons L, Bone JM. Failure of dietary protein and phosphate restriction to retard the rate of progression of chronic renal failure: a prospective, randomized, controlled trial. Q.J.Med. 1991; 81: 837‒855.

24. Eckberg K, Kramer H, Wolf M, et al. Impact of westernization on fibroblast growth factor 23 levels among individuals of African ancestry. Nephrol.Dial.Transplant. 2015 Apr;30:630‒635.

25. Gutierrez OM, Luzuriaga-McPherson A, Lin Y, Gilbert LC, Ha SW, Beck GR,Jr. Impact of phosphorus-based food additives on bone and mineral metabolism. J.Clin.Endocrinol.Metab. 2015;100:4264‒4271. jc20152279.

26. de Vries LV, Dobrowolski LC, van den Bosch JJ, et al. Effects of Dietary Sodium Restriction in Kidney Trans-plant Recipients Treated With Renin-Angiotensin-Aldosterone System Blockade: A Randomized Clinical Trial. Am.J.Kidney Dis. 2016 ; 67: 936-44.

27. Machnik A, Neuhofer W, Jantsch J, et al. Macrophages regulate salt-dependent volume and blood pressure by a vascular endothelial growth factor-C-dependent buffering mechanism. Nat.Med. 2009; 15: 545‒552.

28. De Keyzer W, Dofková M, Lillegaard ITL, et al. Reporting accuracy of population dietary sodium intake using duplicate 24 h dietary recalls and a salt questionnaire. Br.J.Nutr. 2015; 113: 488‒497.

Chapter 9High Potassium Intake Reduces Fibroblast Growth Factor 23 to Increase Renal Phosphate ReabsorptionJelmer K. HumaldaLieke GijsbersJ. Marianne GeleijnseIneke J. RiphagenGerjan NavisStephan J. L. BakkerMartin H. De Borst

In Preparation (Embargo).

Chapter 9

152

Abstract

Background: High plasma concentrations of the phosphate-regulating hormone fibroblast growth factor 23 (FGF-23) are robustly associated with cardiovascular morbidity and mortality. The determinants of FGF-23 concentration are not completely understood, hampering develop-ment of strategies to reduce FGF-23. Since dietary potassium increases serum phosphate, we hypothesized that potassium supplementation would reduce FGF-23 levels.

Methods: Post hoc analysis of a randomized, blinded, placebo-controlled cross-over trial in which 36 untreated pre-hypertensives underwent three study periods of four weeks on a fully controlled, on-site diet combined with capsules containing 3 grams of potassium/day, 3 grams of sodium/day or placebo, respectively, in random order. Dietary potassium intake was kept constant at 2.3 grams/d. Blood and 24h-urine were collected after each study period. Plasma C-terminal FGF-23 was measured by ELISA, parathyroid hormone (PTH) by radioimmunoassay. Outcomes at the end of each supplementation period were compared with placebo using linear mixed models.

Results: Participants (67% male) were 66.0 ± 9.3 (mean ± SD) years old and had an eGFR (cystatin C-based CKD-EPI) of 84 ± 13 mL/min/1.73m2, that was not influenced by potassium supplementation (P=0.6). Potassium supplementation increased urinary potassium excretion from 55.3 ± 16.7 to 118.1 ± 32.2 mmol/d (P<0.001). Serum phosphate increased from 1.10 ± 0.19 to 1.15 ± 0.18 mmol/L (P=0.004), in line with an increase in TmP/GFR from 0.93 ± 0.21 to 1.01 ± 0.20 mmol/L (P<0.001). Plasma FGF-23 decreased from 114.3 RU/mL (geometric mean; 95% CI, 96.2 to 135.8) to 108.5 RU/mL (95% CI, 93.0 to 126.6; P=0.01), without changes in PTH and 25(OH)-vitamin D3. Excretion of sodium, phosphate and urea did not materially change.

Conclusions: Dietary potassium supplementation induced a reduction in FGF-23, an increased TmP/GFR, the tubular set point for phosphate reabsorption, and a higher serum phosphate level. This effect was independent of PTH or 25(OH)-vitamin D3. Sodium supplementation lowered FGF-23 and increased serum phosphate. Potassium supplementation may be a novel strategy to reduce FGF-23 levels and subsequently improve adverse outcomes in chronic kidney disease and heart failure.

153

FGF-23 and Potassium Supplementation

9

Introduction

Fibroblast growth factor 23 (FGF-23) has been recognized as a key regulator of phosphate ho-meostasis. FGF-23 stimulates renal phosphate excretion, and inhibits the production of parathy-roid hormone (PTH) and active, 1,25-dihydroxy vitamin D (1). A higher concentration of FGF-23 is associated with higher mortality risk, particularly in high-risk populations as haemodialysis patients (2, 3) and predialysis chronic kidney disease (4, 5), but also in the general population (6, 7). However, our understanding of determinants and regulators of FGF-23 is incomplete, and this hampers efforts to reduce FGF-23. Serum phosphate concentrations and renal function are tightly correlated with FGF-23 concentrations, but strategies that reduce serum phosphate lev-els are at best modestly effective in lowering FGF-23 (8, 9). Possibly, dietary factors contribute to FGF-23 physiology.

A recent study demonstrated that the Western diet is associated with higher circulating FGF-23 levels (10). Besides being rich in phosphate, the Western diet is also relatively poor in potassium content. Lower potassium intakes are associated with higher risk of cardiovascular events (11), end stage renal disease (12), and mortality (13). Interestingly, the observational study found an inverse association between potassium intake and FGF-23 levels (10). Studies from the 1980s have demonstrated that potassium supplementation increases serum phosphate concentra-tions in healthy volunteers independent of PTH (14). In preclinical studies, potassium increased tubular reabsorption of phosphate (15). The mechanism of this effect of potassium on phos-phate is unknown. We hypothesize an inhibiting effect of potassium on FGF-23 concentrations as a candidate mechanism. If so, potassium supplementation will lower FGF-23 as a mechanism that increases phosphate concentration. To test this hypothesis we measured phosphate and FGF-23 in a post hoc analysis of a randomized, full-dietary, placebo-controlled trial of potassium supplementation.

Methods

Participants

We analyzed a double-blinded, randomized, placebo-controlled, crossover study that assessed the effects of potassium and sodium supplementation on blood pressure. The protocol has extensively been described (16). Eligible participants were 40‒80 years old, with a fasting of-fice systolic blood pressure of 130‒159 mmHg. Exclusion criteria were diabetes mellitus, renal diseases including chronic kidney disease, gastrointestinal and liver diseases. Participants were also ineligible for participation if they were current smokers; had a BMI > 40 kg/m2; used medi-cation that affects cardiovascular system; nutritional supplements; were on a energy-restricted

Chapter 9

154 155


9

or medically prescribed diet; had unstable weight or excessive alcohol use. Participants were recruited from December 2011 to April 2012.

Study Design

Eligible participants were on a fully-controlled diet according to individual energy needs. The research facility supplied 90% of the daily energy needs, the remaining 10% were chosen by the participants from a list of products that were low in sodium and potassium. A 2500 kcal-diet provided 2.3 grams of potassium, 2.4 grams of sodium and 82.2 grams of protein. After a run-in period of 1 week on the diet, participants underwent three 4-week treatment periods. The treatment periods consisted of the daily use of 8 capsules (Microz, Geleen, the Netherlands), that contained in total 3 grams of potassium, 3 grams of sodium (7.5 grams of salt) or placebo, in randomized order.

Measurements

Participants underwent venous blood sampling after each treatment period at fixed time points of the day throughout the study, and handed in a 24-h urinary collection. Serum, EDTA-plasma and urine samples were stored at –80°C, and electrolytes were assessed by routine labora-tory procedures (Modular P, Roche Diagnostics, Mannheim, Germany). C-terminal FGF-23 was determined in EDTA-plasma by enzyme-linked immunosorbent assay (ELISA, Immutopics, San Clemente, CA, USA). Parathyroid hormone (PTH) was measured by radioimmunoassay, and 25-hydroxy vitamin D3 (25[OH]-vitamin D3) with isotope dilution–online solid phase extraction liquid chromatography–tandem mass spectrometry, both in EDTA-plasma. The tubular maxi-mum reabsorption / GFR (TmP/GFR) was calculated as a measure of the phosphate threshold using the next formula (17):

TmP/GFR = serum phosphate (mmol/L) – (urinary phosphate [mmol/L] × serum creatinine [mmol/L]) / urinary creatinine [mmol/L]

Statistical Analysis

Normally-distributed data are presented as means ± standard deviation, whereas data that did not follow the normal distribution are presented as geometric mean and 95% confidence interval. For each outcome measure, we used a mixed-effects model with covariance structure compound symmetry to estimate the effect of active treatment compared with placebo. Fixed effects were ‘treatment’ and ‘period’, random effect was participant number. Variables were natural log (Ln) transformed when appropriate, as assessed with histograms and Q-Q plots. Analysis were performed in SAS 9.3 (SAS Institute, Cary, North Carolina, USA).

Chapter 9

154 155


9

Tabl

e 1.

Effe

cts o

f Pot

assiu

m o

r Sod

ium

Sup

plem

enta

tion

on M

iner

al B

one

Met

abol

ism P

aram

eter

s

Mea

n ±

SDM

ean

diffe

renc

e (9

5%)

Seru

mPo

tass

ium

Plac

ebo

Sodi

umPo

tass

ium

vs P

lace

boP-

valu

eSo

dium

vs P

lace

boP-

valu

e

Seru

m p

hosp

hate

, mm

ol/L

1.15

± 0

.19

1.10

± 0

.19

1.06

± 0

.21

0.05

(0.0

2 to

0.0

9)0.

004

‒0.0

4 (‒

0.08

to 0

.00)

0.03

Seru

m c

alci

um, m

mol

/L2.

34 ±

0.0

82.

34 ±

0.0

62.

33 ±

0.0

8‒0

.01

(‒0.

03 to

0.0

2)0.

6‒0

.01

(‒0.

04 to

0.0

1)0.

2

FGF-

23, R

U/m

La10

8.5

(93.

0 to

126

.6)

114.

3 (9

6.2

to 1

35.8

)10

8.7

(91.

7 to

128

.8)

‒0.0

5 (‒

0.09

to ‒

0.01

)0.

01‒0

.05

(‒0.

09 to

‒0.

01)

0.02

PTH,

pm

ol/L

a4.

36 (3

.84

to 4

.94)

4.37

(3.8

9 to

4.9

0)4.

37 (3

.93

to 4

.85)

0.00

(‒0.

07 to

0.0

6)0.

90.

00 (‒

0.06

to 0

.07)

0.9

25(O

H)-v

itam

in D

3b59

.0 ±

19.

059

.0 ±

16.

958

.3 ±

18.

10.

9 (‒

1.6

to 3

.3)

0.5

‒0.8

(‒3.

3 to

1.7

)0.

5

Urin

e

Sodi

um e

xcre

tion,

mm

ol/2

4h96

.5 ±

39.

010

5.1

± 39

.720

2.9

± 54

.8‒8

.9 (‒

25.4

to 7

.6)

0.3

97.6

(81.

0 to

114

.1)

<0.0

01

Pota

ssiu

m e

xcre

tion,

mm

ol/2

4h11

8.1

± 32

.255

.3 ±

16.

753

.2 ±

16.

662

.9 (5

4.9

to 7

0.8)

<0.0

01‒2

.2 (‒

10.2

to 5

.7)

0.6

Phos

phat

e ex

creti

on, m

mol

/24h

24.3

6 ±

9.58

24.3

9 ±

8.58

24.4

5 ±

7.27

‒0.0

2 (‒

2.47

to 2

.43)

0.99

0.05

(‒2.

40 to

2.5

0)0.

98

Calc

ium

exc

retio

n, m

mol

/24h

4.05

± 2

.15

4.28

± 1

.91

5.45

± 2

.51

‒0.2

4 (‒

0.69

to 0

.21)

0.3

1.16

(0.7

0 to

1.6

1)<0

.001

Ure

a ex

creti

on,m

mol

/24h

372

± 10

335

5 ±

9336

3 ±

120

16 (‒

11 to

43)

0.2

7 (‒

20 to

34)

0.6

TmP/

GFR

1.01

± 0

.20

0.93

± 0

.21

0.91

± 0

.22

0.07

(0.0

3 to

0.1

1)<0

.001

‒0.0

2 (‒

0.06

to 0

.01)

0.2

Frac

tiona

l exc

retio

n of

pho

spha

te, %

13.3

± 4

.215

.8 ±

5.8

14.7

± 4

.5‒2

.50

(‒3.

75 to

‒1.

25)

<0.0

01‒1

.01

(‒2.

26 to

0.2

4)0.

1

a) V

alue

s ar

e ge

omet

ric m

ean

and

95%

CI f

or F

GF-2

3 an

d PT

H, d

iffer

ence

s ar

e ch

ange

in L

n-tr

ansf

orm

ed v

aria

ble.

b) D

ue to

thre

e m

issin

g sa

mpl

es, a

naly

sis p

erfo

rmed

for

pota

ssiu

m o

n n=

36 (n

o m

issin

g da

ta),

plac

ebo

n=35

, sod

ium

n=3

4, re

spec

tivel

y. A

bbre

viati

ons:

SD,

stan

dard

dev

iatio

n; F

GF-2

3, fi

brob

last

gro

wth

fact

or 2

3; P

TH, p

arat

hyro

id

horm

one;

Tm

P/GF

R, tu

bula

r max

imum

reab

sorp

tion

of p

hosp

hate

per

glo

mer

ular

filtr

ation

rate

.

Chapter 9

156

Results

Population Characteristics

The 36 participants were 65.8 years old (range 47‒80), predominantly male (67%) with a body mass index of 27.2 ± 4.6 kg/m2. Participants had slightly elevated blood pressure at screening (average 145/81 mmHg). Baseline characteristics have been reported previously (16, 18).

Effect of Potassium Supplementation

Potassium supplementation increased potassium excretion by 62.9 mmol (Table 1). This in-crease of 62.8 mmol (2459 mg) is in accordance with ~80% gastrointestinal uptake of the 3 grams supplemented potassium. Serum phosphate concentration increased from 1.10 ± 0.19 to 1.15 ± 0.19 mmol/L (P= 0.004), whereas serum calcium concentration was unchanged. The increase in serum phosphorus was paralleled by an increase in phosphate tubular reabsorp-tion, as assessed by TmP/GFR, from 0.93 ± 0.21 to 1.01 ± 0.20. Urinary phosphate excretion did not change (Figure 1, panels A‒C). We analyzed three hormones that are involved in renal phosphate handling. Potassium supplementation reduced FGF-23 concentrations from 114.3 RU/mL (geometric mean; 95% CI, 96.2 to 135.8) to 108.5 RU/mL (95% CI, 93.0 to 126.6; P=0.01). Potassium supplementation did not influence 25(OH)-vitamin D3 or PTH concentrations (Figure 1, panels D‒F). Potassium supplementation did not change sodium or urea excretion.

Effects of Sodium Supplementation

Sodium supplementation increased 24-h urinary sodium excretion by 97.6 mmol/day, whereas potassium excretion did not change (P=0.6). Serum phosphate levels were a bit lower, and surprisingly, FGF-23 concentrations were lower as well (Table 2). Here, TmP/GFR did not change (P=0.1), neither did 25(OH)-vitamin D3, PTH nor 24-hour phosphate excretion.

Discussion

In this randomized, placebo- and fully dietary controlled trial potassium supplementation reduced FGF-23 concentrations, with a concomitant rise of renal phosphate reabsorption and serum phosphate levels, without an effect on PTH or 25(OH)-vitamin D3. These data suggest that potassium supplementation may be a novel strategy to reduce FGF-23 levels in prehyper-tensives.

A stimulatory effect of potassium on phosphate reabsorption has been described in rats already in 1983 (15). At variance with the animal study, PTH did not change in our study and is there-fore unlikely to explain the observed effect on serum phosphate. Our results are in line with a previous study in healthy human volunteers, where potassium supplementation increased

157


9

serum phosphate levels (14). Although our patients were older and prehypertensive, they were otherwise healthy and the mean increase of 0.05 mmol/L in serum phosphate is comparable with the 0.07 mmol/L increase Sebastian et al. reported (14). The same study underscored the specificity of potassium, because the effect was present regardless whether potassium bicar-

A

C

E

B

D

F

Figure 1. Effect of potassium supplementation on serum phosphate (A), phosphate excretion (B) and TmP/GFR (C). The rise of phosphate levels was paralleled by a decrease in FGF-23 (D), without effect on PTH (E) or vitamin D (F). Depicted are unadjusted means and standard errors, or geometric means and 95% confidence intervals for FGF-23 and PTH. Abbreviations: TmP/GFR, tubular maximum reabsorption of phosphate per glomerular filtration rate; FGF-23, fibroblast growth factor 23; PTH, parathyroid hormone

Chapter 9

158

bonate or potassium chloride was used. In line, epidemiological studies report that a population living in Ghana had a diet more rich in vegetables (and thus potassium), accompanied by higher serum phosphate, lower fractional excretion of phosphate, and lower FGF-23 concentrations compared with a population of African ancestry that lived in the United States (19). This mirrors our results of potassium treatment, although potassium intake was not reported. To summarize our findings, we propose that potassium supplementation lowers FGF-23 to increase TmP/GFR, and thus renal phosphate reabsorption. This temporarily reduces phosphate excretion, until a new steady state with higher serum phosphate concentration has been reached (Figure 2).

Figure 2. Schematic representation of our hypothesis. Potassium supplementation starts at day 0. This results in a drop of FGF-23, and thus increases the renal set point for phosphate reabsorption, TmP/GFR. Phosphate excre-tion temporarily falls, and reaches a new steady state at a higher serum phosphate concentration. Question marks depict uncertainty about the time course of changes in FGF-23 and PTH, as measurements took place only after 30 days.

The increase in serum phosphate may be disconcerting, as frank hyperphosphataemia (>1.78 mmol/L) is an established risk factor for mortality in haemodialysis patients (20). Even more, serum phosphate concentrations >1.13 mmol/L were associated with increased mortality risk in patients with CKD (21). However, whether a potassium-induced rise of serum phosphate levels is also detrimental cannot be derived from these data.

We also evaluated the effect of sodium supplementation on FGF-23. Although sodium lowered FGF-23 concentrations, this effect was not accompanied by changes in TmP/GFR, and occurred together with a counterintuitive reduction of serum phosphate concentrations. This may be explained by sodium-induced calcium excretion, which was inversely correlated with FGF-23 concentrations in previous studies (22), and low serum calcium levels correlate positively with low FGF-23 levels (23). In line, the calcium × phosphate product is a stronger correlate of FGF-23 concentrations than when phosphate or calcium are considered individually (24). Although se-rum calcium levels did not change in our study, it is possible that increased calciuria contributed to a tendency for a lower calcium × phosphate product, which may have elicited the reduction in FGF-23 concentrations.

159


9

The mechanism how a potassium load reduces FGF-23 is unknown. A possible explanation for this interaction may lie in the effects of potassium load on the renal tubuli. A potassium load inhibits the sodium (Na+) chloride cotransporter (NCC) by rapid dephosphorylation in mice (25), regardless whether potassium is administrated orally (25) or intravenously (26). This NCC dephosphorylation also occurs when potassium is supplemented to a low sodium diet (27), this mirrors our study design. Putatively, this dephosphorylation interacts with upstream WNK/SPAK signaling, which has been implied to lower FGF-23 concentrations (28). Alternatively, potassium may directly reduce the formation of FGF-23 in the osteocytes in the bone. The extracellular matrix in the bone has fivefold higher potassium concentration compared with extracellular fluid, a gradient that is maintained by active transport mechanisms (29). We postulate that the bone may serve as a buffer for potassium load. Possibly, the osteocytes are sensitive to increases in potassium load and respond by reducing the production of FGF-23.

Part of the beneficial effects of high potassium intake may thus be mediated by potassium-induced reduction of FGF-23 concentrations, and the ensuing putative beneficial effects on vol-ume status (18) may overcome the possible adverse effects of a higher serum phosphate level. This explains why a diet rich in potassium and low FGF-23 concentrations are both robustly associated with better cardiovascular outcomes. For potassium this is traditionally explained by its blood pressure-lowering effects. A meta-analysis found that for a mean increase of 51 mmol potassium per day, blood pressure was reduced by 3.3/2.1 mmHg (30). Potassium supplementa-tion also lowered blood pressure in our cohort (16). This may be due to a facilitated sodium excretion (25-27, 31). Interestingly, an interaction between volume status and FGF-23 has been demonstrated in experimental studies where FGF-23 increases sodium retention and blood pressure by upregulation of NCC (32), further corroborating that potassium and FGF-23 may be related by their effects on volume status. So, the drop of FGF-23 reduces the putative stimu-latory effect of FGF-23 on NCC (32), thus facilitating sodium-potassium exchange in order to maintain potassium balance, at the expense of an increase in serum phosphate concentrations.

Reduction of FGF-23 is of paramount importance, given its strong association with increased mortality across several populations (2-7). Several lines of evidence suggest that FGF-23 itself is involved in increased cardiovascular morbidity and mortality as an effector and not merely a correlate. Two different hypotheses exist to explain the detrimental effect of FGF-23 on the cardiovascular system. First, FGF-23 may directly promote left ventricular hypertrophy (33), pos-sibly by stimulation of the FGF4 receptor (34). The more recent hypothesis suggests that FGF-23 may induce volume retention (32). In line, FGF-23 may increase risk of heart failure (5, 35). Our results indicate a novel and readily targetable pathway that can reduce FGF-23 concentrations.

Strengths of this study include the fully-controlled diet and the use of potassium, sodium and placebo capsules. The crossover design increased power, as every participant served as his own

Chapter 9

160

control. Weaknesses of this study are the limited sample size and that measurements were only performed at the end of each diet period. The exact time at which reduction of FGF-23 occurred is therefore unknown, and conclusions whether the reduction persists over time cannot be drawn. This is not of trivial importance, as phosphate supplementation temporarily increased FGF-23 levels after four weeks, only to return to baseline values after eight weeks of supplementation (36) ‒without affecting plasma phosphate levels. Obviously, such studies do not exist for potassium at this time. Finally, our study does not allow to identify the mechanism of the effect of potassium supplementation on FGF-23 levels.

In conclusion, we demonstrated in a fully dietary and placebo-controlled study that potassium supplementation increases phosphate retention and reduces FGF-23. Our study suggests that FGF-23 connects potassium and phosphate homeostasis; this may explain why a lifestyle rich in potassium may yield beneficial cardiovascular and renal outcomes.

Acknowledgements.

The authors acknowledge W.A. Dam and B.M. Aarts for technical assistance. The original re-search was supported by research grant CH001 from TI Food and Nutrition, a public-private partnership on precompetitive research in food and nutrition. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

161


9

References

1. Haussler MR, Whitfield GK, Kaneko I, et al. The role of vitamin D in the FGF23, klotho, and phosphate bone-kidney endocrine axis. Rev.Endocr Metab.Disord. 2012; 13: 57-69.







8. Isakova T, Barchi-Chung A, Enfield G, et al. Effects of dietary phosphate restriction and phosphate binders on FGF23 levels in CKD. Clin.J.Am.Soc.Nephrol. 2013; 8: 1009-1018.

9. Yilmaz MI, Sonmez A, Saglam M, et al. Comparison of calcium acetate and sevelamer on vascular function and fibroblast growth factor 23 in CKD patients: a randomized clinical trial. Am.J.Kidney Dis. 2012; 59: 177-185.

10. Eckberg K, Kramer H, Wolf M, et al. Impact of westernization on fibroblast growth factor 23 levels among individuals of African ancestry. Nephrol.Dial.Transplant. 2015; 30: 630-635.

11. O’Donnell MJ, Yusuf S, Mente A, et al. Urinary sodium and potassium excretion and risk of cardiovascular events. JAMA 2011; 306: 2229-2238.

12. Smyth A, Dunkler D, Gao P, et al. The relationship between estimated sodium and potassium excretion and subsequent renal outcomes. Kidney Int. 2014; 86(6): 1205-12.



15. Jaeger P, Bonjour JP, Karlmark B, et al. Influence of acute potassium loading on renal phosphate transport in the rat kidney. Am.J.Physiol. 1983; 245: F601-5.

16. Gijsbers L, Dower JI, Mensink M, Siebelink E, Bakker SJ, Geleijnse JM. Effects of sodium and potassium supplementation on blood pressure and arterial stiffness: a fully controlled dietary intervention study. J.Hum.Hypertens. 2015; 29: 592-598.

17. Parfitt AM. Misconceptions IV--the hypophosphatemia of primary hyperparathyroidism is the result of renal phosphate wasting. Bone 2004; 35: 345-347.

18. Riphagen IJ, Gijsbers L, van Gastel MD, et al. Effects of potassium supplementation on markers of osmo-regulation and volume regulation: results of a fully controlled dietary intervention study. J.Hypertens. 2016; 34(2):215-20.

19. Yuen SN, Kramer H, Luke A, et al. Fibroblast Growth Factor-23 (FGF-23) Levels Differ Across Populations by Degree of Industrialization. The Journal of Clinical Endocrinology & Metabolism 2016; jc.2015-3558.

Chapter 9

162

20. Floege J, Kim J, Ireland E, et al. Serum iPTH, calcium and phosphate, and the risk of mortality in a European haemodialysis population. Nephrol.Dial.Transplant. 2011; 26: 1948-1955.

21. Kestenbaum B, Sampson JN, Rudser KD, et al. Serum phosphate levels and mortality risk among people with chronic kidney disease. J.Am.Soc.Nephrol. 2005; 16: 520-528.

22. Sigrist M, Tang M, Beaulieu M, et al. Responsiveness of FGF-23 and mineral metabolism to altered dietary phosphate intake in chronic kidney disease (CKD): results of a randomized trial. Nephrol.Dial.Transplant. 2013; 28: 161-169.

23. Rodriguez-Ortiz ME, Lopez I, Muñoz-Castañeda JR, et al. Calcium Deficiency Reduces Circulating Levels of FGF23. Journal of the American Society of Nephrology 2012; 23: 1190-1197.


25. Sorensen MV, Grossmann S, Roesinger M, et al. Rapid dephosphorylation of the renal sodium chloride cotransporter in response to oral potassium intake in mice. Kidney Int. 2013; 83: 811-824.

26. Rengarajan S, Lee DH, Oh YT, Delpire E, Youn JH, McDonough AA. Increasing plasma [K+] by intravenous potassium infusion reduces NCC phosphorylation and drives kaliuresis and natriuresis. Am.J.Physiol.Renal Physiol. 2014; 306: F1059-68.

27. van der Lubbe N, Moes AD, Rosenbaek LL, et al. K+-induced natriuresis is preserved during Na+ deple-tion and accompanied by inhibition of the Na+-Cl- cotransporter. Am.J.Physiol.Renal Physiol. 2013; 305: F1177-88.

28. Pathare G, Foller M, Michael D, et al. Enhanced FGF23 serum concentrations and phosphaturia in gene targeted mice expressing WNK-resistant SPAK. Kidney Blood Press.Res. 2012; 36: 355-364.

29. Rubinacci A, Benelli FD, Borgo E, Villa I. Bone as an ion exchange system: evidence for a pump-leak mecha-nism devoted to the maintenance of high bone K(+). Am.J.Physiol.Endocrinol.Metab. 2000; 278: E15-24.


31. Penton D, Czogalla J, Loffing J. Dietary potassium and the renal control of salt balance and blood pressure. Pflugers Arch. 2015; 467: 513-530.



34. Grabner A, Amaral A, Schramm K, et al. Activation of Cardiac Fibroblast Growth Factor Receptor 4 Causes Left Ventricular Hypertrophy. Cell Metabolism. 2015; 1;22(6):1020-32


36. Trautvetter U, Jahreis G, Kiehntopf M, Glei M. Consequences of a high phosphorus intake on mineral metabolism and bone remodeling in dependence of calcium intake in healthy subjects - a randomized placebo-controlled human intervention study. Nutr.J. 2016; 15: 7-016-0125-5.

Chapter 10The SUBLIME Approach: Efficacy and Cost-Effectiveness of a Blended Care Self-Management Approach facilitated by E-health for Dietary Sodium Restriction in Patients with Chronic Kidney DiseaseJelmer K. HumaldaGerald KlaassenHanne de VriesYvette MeulemanLara C. VerschuurLilian (E.) J.M. StraathofTrijntje KokGoos D. LavermanWillem Jan W. J. BosPaul J.M. van der BoogKarin M. VermeulenOlivier A. Blanson HenkemansWilma OttenMartin H. De BorstSandra van DijkGerjan Navis

for the SUBLIME Investigators. In Preparation (Embargo).

Chapter 10

166

Abstract

Objective: Assess a multidisciplinary blended care self-management approach aimed at sodium restriction for efficacy, cost-effectiveness and implementation analysis

Design: Randomised, unblinded trial in 100 participants. Intervention was designed in co-cre-ation with and evaluated by patients, consisting of e-health embedded in clinical care. Control group received care as usual.

Setting: Nephrology outpatient clinics in four Dutch hospitals (two teaching hospitals).

Participants: Adults with chronic kidney disease or renal transplant recipients with eGFR >25 mL/min/1.73m2, hypertension and sodium intake >130 mmol/day. 89/100 patients completed the maintenance phase.

Intervention: Dietary sodium restriction, supported by a web-based self-management system with individual e-coaching and two group meetings in the intervention phase of 3 months, fol-lowed by 6 months maintenance phase with continued access to the website but without group meetings.

Main outcome measure: 24-hour urinary sodium excretion, office blood pressure, cost-effec-tiveness after 3 months (intervention phase) and 9 months (maintenance phase).

Results: Patients (44% renal transplant recipients) were on average 56.6 ± 12.4 (SD) years old with eGFR 55.0 ± 22.0 mL/min/1.73m2 at baseline. During the intervention phase, the interven-tion reduced 24h sodium excretion from 188 ± 63 to 147 ± 55 mmol/day (linear mixed effects model effect ‒24.8 mmol/day (95% CI, ‒49.6 to ‒0.1; P<0.05) compared with control). The intervention reduced systolic blood pressure (SBP) from estimated marginal mean 140 (SE, 3) mmHg to 132 (3) mmHg (P<0.001; effect ‒4.7 mmHg (95% CI, ‒10.7 to 1.3; P=0.12) compared with control). After the maintenance phase, sodium excretion remained lower than baseline in the intervention (159 (8) mmol/day; P=0.001), but also decreased in the control group (154 (9) mmol/day; P<0.001; P=0.6 between groups). SBP remained lower in the intervention group compared with baseline: 131 (3) mmHg (P<0.001), ‒4.3 mmHg (95% CI, ‒10.2 to 1.7; P=0.16) compared with control. The incremental cost-effectiveness ratio was €0.39 [95%CI: – €10 to €16] per 1 mmol Na and €4 [– €88 to €106] per 1 mmHg reduction in the intervention phase for the intervention versus control group, as evidenced by bootstrap analyses. We found no effect on quality of life. Participants (focus groups) appreciated the program.

167

SUBLIME: a Blended Care Approach for Sodium Restriction

10

Conclusion: The SUBLIME intervention reduced sodium intake and blood pressure, moreover this reduction was maintained. Data in the control group suggest a trial-effect during the main-tenance phase.

Trial registration: ClinicalTrials.gov identifier NCT02132013

Funding body: The Netherlands Organization for Health Research and Development (ZonMw) project 837001005, Doelmatigheidsonderzoek 2013-2015. Dutch Kidney Foundation.

Chapter 10

168

Introduction

The worldwide average salt intake by far exceeds 5 grams (or ~2000 milligrams/ 90 mmol of sodium) per day, the recommended value by the World Health Organization (1). Patients with chronic kidney disease (CKD) are particularly sensitive to excess sodium as reviewed in (2), and are strongly advised to limit sodium intake in guidelines (3-5). Observational studies revealed the potential of moderately reduced sodium intake, suggesting that every 1 gram less sodium intake is associated with a 15% lower risk of cardiovascular complications, a 15% lower risk for end stage renal disease (ESRD) in diabetic CKD (6), and a 10% lower risk for ESRD in nondiabetic CKD (7). A cost-effectiveness simulation stated that 3 gram reduction of salt intake would yield $10‒24 billion worldwide annually (8); this figure would be enhanced when taking ESRD and the accompanying excessive costs for renal replacement therapy (RRT) into account. In the Neth-erlands, CKD prevalence is estimated between 7.6 and 10.4 patients per 100 persons (10, 11), about 750 000 suffer from CKD Stage 3 or higher (4) and in 2015 16 316 patients were on RRT, of whom 6461 on dialysis according to the Dutch registry RENINE (www.renine.nl). Healthcare expenditure for RRT totals 750 million euros in the Netherlands. This emphasizes that sodium restriction may be beneficial from both a medical and macroeconomical perspective.

Current approaches to sustainably reduce sodium intake are largely unsuccessful: a recent re-view of observational and intervention studies performed in over 10,000 CKD patients revealed that the average sodium intake in CKD patients is 164 mmol/day, even in the dedicated setting of the nephrology outpatient clinic (12). Behavioral approaches, starting from the behavioral roots of dietary habits, may provide more fruitful strategies. Participants that received dietary and intensive behavioral counseling in the Trials of Hypertension Prevention I and II had a 25% lower risk of cardiovascular events after 10‒15 years follow-up, and this intervention group reported a higher incidence of dislike of salty foods, use of low sodium products, and monitoring of daily sodium intake in the follow-up questionnaire (13). A theoretical framework such as the self-regulation theory (SRT) can provide a sound theoretical basis and practical guidelines for development of a lifestyle intervention (14). In this SRT framework, a change in dietary intake is approached as a behavioral change. SRT distinguishes three phases of behavioral change: goal-selection or motivational phase, to develop the intention to change one’s behavior; active goal pursuit or action phase, where behavior is actually changed; and a maintenance phase where the new behavior is implemented (15). A qualitative focus group study in 25 patients with CKD identified facilitators and barriers for reduction of sodium intake in the perspective of SRT, yielded the following recommendations for successful interventions: provide information on sodium content, strengthen intrinsic motivation, provide self-monitoring, enforce refusal skills regarding adverse environmental triggers, stimulate social support by involving partners, and provide coaching and evaluation of goals in the maintenance phase (15). In support of a SRT-based approach, in the ESMO intervention study we achieved effective sodium restric-

169


10

tion in CKD patients, although the effect did not persist after the intervention (Meuleman et al., Am J Kidney Dis, in press). Taken together, behavioral intervention could be a successful approach to achieve sodium restriction. However, sustained efficacy requires specific efforts aimed at maintenance. Moreover, coaching approaches involving one-to-one counseling are costly and time-expensive. To be effectively implemented in healthcare, an intervention should be cost-effective. E-health may facilitate a self-regulation‒based approach in a cost-effective way. To achieve these aims, we designed the Sodium Burden lowered by Lifestyle Intervention: self-Management and E-health technology (SUBLIME) study, that builds on SRT-experiences from ESMO. The intervention included group counseling and a web-based self-management system instead of intensive individual coaching, and was followed by a maintenance period to achieve sustained efficacy. Here we evaluate the SUBLIME intervention on efficacy and cost-effectiveness, and explore possibilities for implementation in clinical practice.

Methods

Participants

SUBLIME was a randomised, unblinded controlled trial that aimed to assess the efficacy, cost-effectiveness and barriers and facilitators for implementation of a novel self-management pro-gram for reduction of dietary sodium intake. We screened patients that visited the outpatient clinics of the participating centers. Main inclusion criteria were age ≥ 18 years; chronic kidney disease or kidney transplantation, with an eGFR ≥ 25 mL/min/1.73m2; urinary sodium excretion at the last two visits both > 130 mmol/day or > 150 mmol/day at the last visit; systolic blood pressure > 135 mmHg or diastolic blood pressure > 85 mmHg or well-controlled blood pres-sure with antihypertensive therapy; sufficient command of the Dutch language; access to and ability to use the internet; and provision of written informed consent. Main exclusion criteria were rapidly and persistently progressive renal function loss, not from acute, intermittent ori-gin; blood pressures > 170 mmHg systolic, or > 95 mmHg diastolic, or < 95 mmHg systolic not responding to withdrawal of antihypertensive medications; a history of cardiovascular event (myocardial infarction, cerebrovascular incident) < 6 months ago; renal transplantation < 1 year ago; medical conditions likely to interfere with the completion of the study at discretion of the treating nephrologist; previous participation in the ESMO trial (Dutch trial registry NTR2917), as this trial relied on a similar behavioral approach (self-management enhanced by motivational interviewing, self-monitoring, strengthening self-efficacy) and may thus confound the current study. Participants were recruited from June 2014 to March 2015 at the nephrology outpatient clinics in the four participating centers: Leiden University Medical Center, Leiden; Sint Anto-nius Hospital, Nieuwegein; University Medical Center Groningen, Groningen; and ZGT Hospital, Almelo. The medical ethics board approved the study protocol (METc2014/075). The study has

Chapter 10

170

been registered at ClinicalTrials.gov with identifier NCT02132013, and was performed in ac-cordance to the Declaration of Helsinki.

Intervention

The SUBLIME intervention aimed at reducing dietary sodium intake. A program was developed based on self-regulation theory. Participants randomised to the SUBLIME intervention were allocated to a coach and received a home blood pressure monitoring device (Microlife Watch BP Home). The coaches were dietitians, lifestyle professionals or research nurses, all trained by certified lifestyle professionals in the use of the intervention. The intervention consisted of a 3-month intervention phase followed by a 6-month maintenance phase (Figure 1).

Figure 1. Design of the SUBLIME trial. Arrows indicate study visits at baseline, after intervention (3 months) and maintenance phase (9 months), respectively.

At the start of the intervention phase, the coach gave the participants access to a web-based self-management system dedicated to sodium restriction. The coach also offered a detailed instruc-tion on how to use the web-based self-management system. The web-based self-management system consisted of different modules that addressed components of self-regulation theory (→ Supplement A). These components included exercises to strengthen intrinsic motivation and self-monitoring with a diary coupled with a food databank where individual food products could be entered. Next, the program would display in bar graphs sodium intake in total and per prod-uct. The algorithm translating food intake to sodium intake was based on the NEVO tables (16). Options for change could be entered and the effect for e.g. reducing the amount of or replacing a product for a low sodium alternative was displayed in bar graphs; exercises to increase self-efficacy, goal-setting, social support, and dealing with relapse; and a summary page that sum-marized a ‘change plan’. Exercises addressed subcomponents, e.g. for self-efficacy the perceived barriers, solutions and current self-assessment of self-efficacy. Coaches could view the change plan and use this as input for e-coaching, coaches to support the patient in setting tailored goals, rather than a more directive coaching approach. Participants were invited to attend two 2-h group meetings led by the coach with 5‒8 participants and their partners. Group meetings took place in a nonclinical part of the hospital and addressed self-regulation skills regarding reducing salt intake, i.e. goal setting; self-monitoring; refusal skills; and relapse prevention, and

171


10

knowledge about hidden salt in food. Participants received individual e-coaching by telephone or email in the intervention phase of 3 months with a minimum of two e-coaching sessions.

In the maintenance phase of 6 months e-coaching was limited to 1–2 moments without group meetings or other counseling.

The control group received care-as-usual, which in the Netherlands consists of annually to up to three-monthly outpatient clinic visits. All participants visited the outpatient clinics at baseline, 3 months and 9 months for anthropomorphic and blood pressure measurements; blood sam-pling by venipuncture; 24-hour urine collection; assessment of medication use; and they also filled out a questionnaire at each time point. The baseline questionnaire was distributed after randomization.

Objectives and Outcomes

The main objective of this study was to assess the efficacy and cost-effectiveness of a web-based self-management approach on reducing dietary sodium intake, defined as urinary sodium excre-tion. The primary outcome sodium excretion was measured by a 24-hour urine collection. Blood and urinary electrolytes were measured with routine laboratory procedures (Roche Modular). The secondary outcomes were blood pressure; proteinuria; cost-effectiveness; physical and mental health-related quality of life, and self-management skills; and evaluation of barriers and facilitators for implementation of the SUBLIME approach. eGFR was calculated with the CKD Epidemiology Collaboration formula (17). Blood pressure was measured in a standardized fashion at the outpatient clinic, in upright sitting position, after 5 minutes of rest with an auto-mated oscilometric device (WatchBP Home, Microlife), three times with a 1-minute interval in accordance with European Society of Hypertension guidelines (18), we took the mean of the second and third reading for analysis. Proteinuria was measured in 24-hour urinary collection. Changes in medication were explicitly asked at each control visit.

The questionnaires included sociodemographic factors, Short Form-12 (19), EuroQol-5D (20), and Partners In Health scale (21). Physical and mental health-related quality of life was mea-sured using the Short Form (SF)-12. The scoring of SF-12 ranged from 0 to 100. Physical health summary scores (PHS) and mental health summary scores (MHS) were calculated for each item by converting the Likert scale used in the questionnaire to a score from 0 to 100, where higher scores indicate a better health-related quality of life. The SF-12 demonstrated good reliability with Cronbach alpha values of 0.86 for mental health and 0.89 for physical health, respectively (22). Self-management skills were assessed using the Partners In Health (PIH) scale, which showed a good reliability (Cronbach alpha = 0.82) (21).

Chapter 10

172

Healthcare expenditure costs were calculated based on health care consumption data and pro-ductivity losses measured on a patient level with a CRF and patient questionnaire, and valued according to the Dutch guideline for costs studies (23). Incremental Cost-Effectiveness Ratios were computed for each of the major outcome measures separately. Displaying the additional cost per unit of improvement of health with the SUBLIME approach, compared to standard care. In addition, Cost Utility ratios were computed, based on EQ-5D defined utilities, displaying the additional costs of SUBLIME per Quality Adjusted Life Year (QALY) compared to standard care.

For future implementation, the identification of facilitators and barriers started with meetings with end users including representatives of the Dutch Kidney Patients Association to discuss design of the study and the prototype of the web-based self-management system. Thus, users could indicate points for improvement which led to including the interactive submodule for visualizing the effect of changing food products for low sodium alternatives on sodium intake. After completion of the study we organized focus groups to evaluate the intervention and identify barriers and facilitators. Every center organized a focus group for participants whom completed the intervention, aiming at 6 participants per group. Each focus group followed an interview protocol covering the three aspects of the intervention: web-based self-management system, group meetings, and individual coaching. The focus group was led by a representative of the Dutch Kidney Patients Association who had experience with focus groups, and the session was observed by two note-takers from TNO. The sessions were also audiotaped after permission was granted by the participants. One note-taker attended all four focus groups (WO), and was checked by the second observer and the discussion leader. These registrations were qualitatively analyzed to identify barriers and facilitators for implementation.

Sample Size

The sample size calculation was based on a target difference of 2 grams of salt (correspond-ing to 34 mmol sodium/day) between the control and intervention group. Based on data from previous studies (24-26) we assumed a standard deviation of 40 mmol/day. To obtain a power of 80% to demonstrate a 34 mmol difference, 42 patients per group are needed. Accounting for possible drop-out, and 10% loss of data due to inaccurate urine collections, we targeted a size of 49 patients per group.

Randomisation

A randomisation list was generated automatically, stratified per participating center. The randomisation list was concealed from (a) local coordinators until all interventions had been allocated; (b) personnel involved in selection, recruitment and outpatient management of participants; and (c) from the principal investigators throughout the whole study. Upon receipt of a signed informed consent form, the local coordinator allocated a study number, contacted

173


10

the list coordinator and received the randomisation result. Due to the design of the intervention with active behavioral counseling, blinding was not feasible.


Data are reported as mean ± standard deviation (SD) for continuous variables that follow the normal distribution, or median (first to third quartile) for skewed continuous variables. Categori-cal variables are reported as frequency (percentage). We performed an intention-to-treat analy-sis on the primary and secondary outcomes using a linear mixed-effects model with restricted maximum likelihood approach and scaled identity (ID) covariance structure for the outcomes sodium excretion, systolic and diastolic blood pressure. Fixed effects were treatment group, time and time × treatment group, random effect was participant number. We report estimated marginal means and standard errors for continuous variables in our linear mixed-effects models. PIH and SF-12 data were non-normally distributed. Therefore, within-group differences over time were tested using Wilcoxon Signed Rank test. PIH and SF-12 delta scores were normally distributed, hence the between-groups differences at the three time points were tested using in-dependent samples t-test. Occurrence of antihypertensive dose reduction between control and intervention was compared with Fisher’s exact test, similarly for dose increase. We conducted two sensitivity analyses. First, we performed a sensitivity analysis for completion of 24-hour urine collection, by calculating the estimated creatinine excretion rate (eCER) according to Ix’ Equation D (27). We repeated the mixed model analysis for sodium excretion in patients whose 24-hour creatinine excretion fell within 30% and 35% eCER, respectively. Second, although 24-hour urine collection is the gold standard, it may vary between outpatient clinic visits. Because one inclusion criterion was sodium excretion >130 mmol/d, this study is vulnerable for regres-sion to the mean (inclusion bias). We therefore conducted a subgroup analysis in 34 patients to compare baseline sodium excretion with average sodium excretion at the outpatient clinic during two years before baseline. P-values smaller than 0.05 were considered to represent statistical significance. Analyses were performed with PASW Statistics, version 22.0 (SPSS Inc.); STATA Statistical Software: Release 13 (StataCorp.); and Graphpad Prism, version 5.01 (Graph-Pad Software Inc.). For the cost effectiveness analysis, composite outcomes were calculated by dividing the difference in mean costs between the two treatment alternatives by the difference in effect. In addition, bootstrapping (5000 replications) was performed to estimate 95% confi-dence intervals. Furthermore, a cost effectiveness plane was made. The south east quadrant shows the replications in which the new intervention is less costly and more effective, the south west quadrant shows less costly and less effective replications, north west is less effective and more costly, and the north east quadrant shows the more costly and more effective replications.

Chapter 10

174

Results

Parti cipants Flow

We randomised 99 pati ents: 52 pati ents to the interventi on group and 47 pati ents to the control group. Five pati ents did not att end the baseline visit for various reasons, as depicted in Figure 2. The group size was 50 and 44 for interventi on and control group, respecti vely. 89 pati ents completed the study, there were 5 drop-outs in the interventi on group: three parti cipants had no ti me for further parti cipati on, one pati ent deceased, and one was lost to follow-up.

Enro

llmen

t:

Part

icip

ants

A

lloca

tion

: Pa

rtic

ipan

ts

Allo

cati

on:

Care

Pro

vide

rs

Follo

w-u

p:

Part

icip

ants

Figure 2. Flow diagram of the SUBLIME trial.

Care providers: 4 teams in 4 centers observed controls. Each center observed median 11 (min 4, max 18) participants

Care providers: 4 teams in 4 centers performed the intervention. Each center treated median 13 (min 6, max 19) participants

Received informed consent, n=100

Informed consent not-identifiable (only signature, no name.)

Randomised, n=99

Received Control, n=44 Drop-out: 0

Received Intervention, n=45 Drop-out: 5 - 3 too busy - 1 lost to follow-up - 1 deceased

Allocated to intervention, n=52 Received allocated intervention, n=50 Did not receive, n=2 - 2 withdrawal informed consent

Allocated to control, n=47 Received allocated control, n=44 Did not receive, n=3 - 2 no shows - 1 violated inclusion criteria

Figure 2. Flow diagram of the SUBLIME trial.

175


10

Baseline Characteristics

Participants were 56.6 ± 12.4 (mean ± SD) years old, 15/94 were female, and patients had a mean eGFR of 55.0 ± 22.0 mL/min/1.73m2. Forty-four percent of the participants were renal transplant recipients. Baseline characteristics were similar between control and intervention groups (Table 1). At baseline, participants had a mean sodium excretion of 188 ± 63 mmol/day and a systolic blood pressure of 139 ± 17 mmHg.

Table 1. Baseline Characteristics.

Total Control Intervention

n 94 44 50

Age, years 56.6 ± 12.4 58.2 ± 13.2 55.1 ± 11.5

Female gender, n 15 (16%) 8 (18%) 7 (14%)

eGFR (CKD-EPI), mL/min/1.73m2 55.0 ± 22.0 54.3 ± 21.6 55.6 ± 22.6

History of DM, none 65 (69%) 30 (68%) 35 (70%)

DM I 7 (7%) 3 (7%) 4 (8%)

DM II 22 (23%) 11 (25%) 11 (22%)

History of Dialysis, n 27 (29%) 12 (27%) 15 (30%)

Renal Transplant Recipient, n 41 (44%) 19 (43%) 22 (44%)

Antihypertensive drug use, n 90 (96%) 41 (93%) 49 (98%)

Number of classes 2.0±1.0 2.0 ± 1.1 2.1 ± 1.0

RAAS blockade, n 70 (74%) 32 (73%) 38 (76%)

Beta-blocker, n 41 (44%) 15 (34%) 26 (52%)

Calcium channel antagonist, n 33 (35%) 16 (36%) 17 (34%)

Diuretic, n 40 (43%) 22 (50%) 18 (36%)

Calcineurin-inhibitor use 27 (29%) 12 (27%) 15 (30%)

Possession HBPM 64 (68%) 35 (80%) 29 (58%)

Uses never 11 (17%) 6 (17%) 5 (17%)

Uses daily 5 (8%) 2 (6%) 3 (10%)

Uses weekly 19 (30%) 10 (9%) 9 (31%)

Uses monthly 29 (45%) 17 (49%) 12 (41%)

Body mass index, kg/m2 28.6 ± 5.3 28.4 ± 5.0 28.7 ± 5.6

Caucasian, n 89 (95%) 40 (91) 49 (98)

Higher educated, n 39 (41%) 19 (43) 20 (40)

Abbreviations: eGFR, estimated glomerular filtration rater; CKD-EPI, Chronic Kidney Disease Epidemiology Collabo-ration formula; DM, diabetes mellitus; RAAS, renin‒angiotensin‒aldosterone system; HBPM, home blood pressure monitor.

Primary and Secondary Clinical Outcomes after Intervention Phase

In the intervention group sodium excretion was reduced from 188 ± 63 mmol/day to 147 ± 55 mmol/day (Figure 3). Compared with the control group, this reflected an additional effect of the

Chapter 10

176

intervention of ‒24.8 (95% confidence interval, ‒49.6 to ‒0.1; P<0.05) mmol/day. The control group demonstrated a nominally but not significant reduction in sodium excretion (Table 2). There was a concomitant drop in systolic blood pressure from 140 ± 16 to 131 ± 14 mmHg; and in diastolic blood pressure from 84 ± 9 to 80 ± 9 mmHg in the intervention group (Figure 4). Linear mixed-effects model analysis confirmed that this was a significant reduction compared to baseline with estimated marginal mean, 140 (standard error, 3) to 132 (3) mmHg, P=0.001 (table 2). In the control group systolic blood pressure did not significantly change 139 (3) to 136 (3) P=0.08,. Only 14 patients had proteinuria ≥ 1.0 g/d (8 intervention, 6 control patients). Antihypertensive drug use in the control group was reduced in 1 and increased in 3 patients whereas it was reduced in 5 and increased in 3 patients in the intervention group (Fisher’s exact test, P=0.2 for dose reduction, P=0.99 for dose increase). The intervention group received 2.8 ± 1.2 times e-coaching, 4 patients had only one e-coaching moment and 8 did not receive e-coaching (2 due to drop-out) according to logs of the coaches.

Table 2. Baseline Values and Effects on Sodium Excretion and Blood Pressure after Intervention and Maintenance Phase

Meana (SE) Intervention Meana (SE) Control Effect Intervention (95% CI)b

Months 0 3 9 0 3 9 ∆ Intervention ∆ Maintenance

Na, mmol/24h 187.6(7.9)

147.5††(8.2)

159.3†(8.4)

188.8(8.5)

173.5(8.8)

153.6††(8.6)

‒24.8*(‒49.6 to ‒0.06)

6.9(‒17.8 to 31.6)

Systolic BP, mmHg 139.6(2.5)

131.8††(2.5)

131.5††(2.5)

139.2(2.6)

136.1(2.7)

135.3(2.6)

‒4.7(‒10.7 to 1.3)

‒4.3(‒10.2 to 1.7)

Diastolic BP, mmHg 83.9(1.4)

80.5†(1.4)

79.2††(1.5)

83.3(1.5)

81.7(1.5)

80.1†(1.5)

‒1.8(‒5.5 to 2.0)

‒1.5(‒5.2 to 2.3)

Linear mixed-effects model for change after intervention phase (3 months) and maintenance phase (9 months) compared with baseline (0). a Estimated marginal means and standard error (SE). b Effect of interaction term time × treatment with 95% confidence interval (CI). * P < 0.05 versus control group; † P < 0.05 versus baseline within group (intervention or control); †† P < 0.001 versus baseline within group (intervention or control). Abbreviations: SE, standard error; CI, confidence interval; BP, blood pressure.

Figure 3. Sodium excretion as assessed by 24-h urine collection at baseline, after intervention (3 months) and maintenance phase (9 months). * denotes P < 0.05 versus control group. P-values represent change within group compared with baseline.

177


10

Primary and Secondary Clinical Outcomes after Maintenance Phase

In the intervention group the effect on sodium excretion persisted throughout the maintenance phase at 157 ± 64 mmol/day (Table 2 and Figure 3).In the control group a reduction of sodium excretion was found after the maintenance phase to 154 ± 40 mmol per day). Accordingly, there was no significant difference in sodium excretion between intervention and control group after the maintenance phase. Also the intervention effect on blood pressure was maintained, without however, significant differences from control. Antihypertensive drug use was reduced during the maintenance phase in 5 patients and increased in 3 patients, in both control and intervention group (Fisher’s exact test, both P=0.99). 11 patients had proteinuria ≥ 1.0 g/d (5 intervention, 6 control patients). In the intervention group 33 patients received 2.1 ± 0.6 times e-coaching, 17 patients did not receive e-coaching (5 due to drop-out) according to logs of the coaches.

Sensitivity Analyses

To account for possible confounding by urine collection errors, we repeated the analysis for patients with a creatinine excretion in all three 24-hour urinary collections within 30% of the estimate creatinine excretion rate. This yielded similar results: a ‒26.1 mmol/day (95%CI, ‒58.6 to 6.4; P=0.11) reduction in sodium excretion in 21 control vs. 23 intervention patients. Within 35% limit, the effect was ‒22.7 mmol/day (95%CI, ‒51.3 to 6.0; P=0.12) for 30 intervention patients compared with 29 control patients. Under-collection thus did not materially influenced our results.

To account for possible regression to the mean, we tested in a subset of patients (n=34) whether sodium excretion before the study was different from baseline. We compared mean sodium excretion during all subsequent visits from 2 years before baseline to baseline with baseline values. This was not significantly different (182.7 ± 47 mmol/day before; 192.8 ± 55.3 mmol/day

Figure 4. Office blood pressure, after intervention (3 months) and maintenance phase (9 months). No difference with P < 0.05 versus control group. P-values represent change within group compared with baseline.

Chapter 10

178

at baseline, P=0.3) and markedly higher than sodium excretion at the end of the maintenance phase. There was no difference between patients allocated to intervention or control group (intervention 185.9 ± 54.5 versus control 178.7 ± 37.2 mmol/day before, P=0.7; 198.8 ± 52.8 versus 185.2 ± 59.2 mmol/day at baseline, P=0.5).

Effects on Quality of Life and Self-Management Skills

Quality of life and self-management skills were not significantly different between the groups at baseline. No within-group differences over time were observed in self-management skills and quality of life (Table 3). After the maintenance phase however, a significantly higher change in PHS score was observed in the intervention group compared to the change in the control group.

Table 3. Baseline Values and Effects on Self-Management Skills and Quality of Life after Intervention and Mainte-nance Phase.

Median [IQR] Intervention Median [IQR] Control Mean difference intervention group

Mean difference control group

Months 0 3 9 0 3 9 ∆ 3 ∆ 9 ∆ 3 ∆ 9

PIH-score 86[72-102]

93[79-101]

91[76-104]

97[82-105]

96[84-104]

96[80-106]

4.46 2.33 -2.09 0.68

SF-12, physical health summary score

79[59-922]

89[65-92]

91[58-92]

83[54-92]

54[33-92]

58[38-92]

-1.94 4.07* -9.44 -5.63

SF-12, mental health summary score

83[72-90]

83[69-93]

86[75-93]

83[73-93]

80[47-87]

80[64-86]

-0.87 0.08 -4.62 -4.47

Mean differences for change after intervention phase (3 months) and maintenance phase (9 months) compared with baseline (0). Data are non-normally distributed and shown as median [IQR]. Differences within-groups over time were tested with Wilcoxon sign test. Differences over time between groups were tested with Independent Samples T-test. *denotes a significant change over time in intervention group versus control group (P < 0.05).Abbreviations: PIH, Partners in Health scale; SF, Short Form.

Cost-Effectiveness

The intervention group gained 0.02 QALYs, the control group 0.05 QALYs as assessed by EQ-5D. Quality of life and self-management skills were not significantly different between groups at baseline (Table 2). The incremental cost-effectiveness ratio (ICER) was €0.39 [95%CI: –10 to 16€] per 1 mmol Na+ reduction and €4 [–88 to 106€] per 1 mmHg systolic blood pressure reduction in the intervention phase (intervention versus control). This means that, compared to the control group, for every 4 euro invested, blood pressure is reduced by 1 mmHg. The cost-effectiveness plane in Figure 5 demonstrates this effect and the strong efficacy of the intervention on blood pressure at acceptable costs. The results were similar for blood pressure after 9 months with €7 [–98 to 123€] per mmHg, as evidenced by bootstrap analyses. For sodium restriction, the

179


10

ICER became negative with €‒7 [–68 to 85€] per 1 mmol Na+ reduction. This follows from the

nominally lower sodium excretion in the control group compared to the intervention group. The cost-utility analysis for QALY at 9 months is similarly inconclusive with a negative ICER of ‒ €1422 (95% CI, ‒ €22.685 to €25.055).

Figure 5. Cost-effectiveness plane for systolic office blood pressure. Data points reflect bootstrap analyses (5000 replications). Incremental effect in mmHg and incremental costs in Euros are depicted compared to control group. The upper-right quadrant reflects non-dominance of the intervention: more effective but at higher costs. The low-er-right quadrant reflects dominance of the intervention: more effective and less costly.

Barriers and Facilitators for Implementation: Focus Groups and Logging Data

Thirty-six out of fifty intervention participants were invited to attend a focus group, because at that moment in time they had completed the intervention. In total, 21 patients participated (response rate 56 %) and in addition 5 partners and 1 daughter also took part in a focus group. The focus group consisted of 5‒6 participants per center. Three participants indicated that they did not use the web-based self-management system (one insisted in reducing sodium intake at once without further help, two experienced difficulties in accessing the module). The 18 participants who used the system thought the exercises clearly formulated and easy to use, but questioned the necessity to complete exercises that addressed intrinsic motivation “because we are already motivated to reduce our salt-intake, otherwise we would not have participated.”. However, the user-friendliness of the intake monitoring module was the largest perceived bar-rier to implementation: filling out the diary was time-intensive; not all products were available in the database, or not retrievable when searched for by a synonym; and it is difficult to esti-mate salt content from food prepared in restaurants or combined products. Patients generally valued the ‘options for change’ menu. They used the ‘change options’ to reduce salt intake and to plan in a week how to compensate for excess salt intake. Most participants used the

Chapter 10

180

system the first 2‒4 months. According to logging data, 44/50 participants used the e-Health module. 1647 recordings of dietary intake were registered during the 9-month study period. Participants registered 4256 (55.4%) main meals (breakfast, lunch, dinner) and 3428 (44.6%) snacks. Participants registered on average 37.4 days. 6 participants registered less than 5 days, 21 participants registered between 5 and 31 days, 12 registered between 32 and 93 days, and 5 participants registered for more than 94 days. If they registered, they usually registered every other day. When asked to which extent the modules provided insight in actual salt consumption, participants strongly valued the monitoring module (average 8.3/10): “You think you know what you eat, until you write it down.”. Most participants valued the possibility for e-coaching. Some patients expressed the wish for a reminder after a longer period of non-use. Most participants would have appreciated “an unannounced reminder contact” in the maintenance phase, lest they would be triggered to adhere to the program. The group meetings were valued for practical advices, increased awareness, exchange of experiences and contact with fellow renal patients. An exercise where participants could write themselves a postcard was less appreci-ated. Participants stressed that support of the partner/family was important. Two partners of participants reported that their own antihypertensive medication was reduced, accordingly due to reduced sodium consumption and another partner suggested partners should have been allowed to hand-in a 24-hour urine collection as well to quantify their sodium reduction. Food intake and salt reduction impacts not only patients, but also their family. Most participants did not feel their relation with their treating nephrologist had changed due to SUBLIME. Overall, participants valued participation in SUBLIME with 7.8/10, and would recommend use of this web-based self-management system to other patients with renal disease, notwithstanding cur-rent sub-optimal user-friendliness.

Discussion

The multidisciplinary, self-management‒ and web-based SUBLIME intervention reduced sodium intake in patients with chronic kidney disease after the intervention phase, with a concomitant reduction in blood pressure compared to baseline. After a maintenance phase of 6 months the reduction in sodium excretion and blood pressure was maintained, whereas in the control group sodium excretion was also reduced compared to baseline. The intervention was cost-effective for sodium restriction and blood pressure reduction.

The SUBLIME study has several strengths. The intervention was designed from start in a mul-tidisciplinary setting, with psychologists, nephrologists, a lifestyle professional and dietitians in co-creation with patients with CKD, represented by the Dutch Kidney Patients Association. We measured sodium intake with the gold standard method 24-hour urine excretion, checked for effects by collections errors and confirmed the representativeness of baseline sodium excre-

181


10

tion in a subgroup. Of note, patients valued the feedback by hard biomedical data. Our study population consists of a mixture of different CKD stages and renal transplant recipients, improv-ing generalization of the results. Further, the intervention was thoroughly evaluated by focus groups and logging data, to identify barriers and possibilities for eventual wide implementation.

A potential weakness of this study is selection bias. By design the study population reflects a more motivated portion of the outpatient population, because patients were willing and able to co-create, and participate in a RCT. Further, our intervention relied heavily on access to in-ternet and literacy. In the Netherlands internet penetration is 95.5% (as of November 15, 2015) (28), which is markedly higher than European average of 74.5%, which impairs generalization. Further, the intervention is not suitable for (digital) illiterates and thus results cannot be gener-alized to the whole patient population. It must also be mentioned that the feedback on sodium intake by the web-based system relies on availability of validated data on sodium content of food products, as provided for the current study by the NEVO tables (16), that cover most food products consumed in The Netherlands. As dietary habits and food products are substantially different in different countries, application in countries other than The Netherlands will require specific adaptation to local food habits.

There was a clear difference in sodium excretion between intervention and control after the intervention phase, and moreover, the effects were maintained during a six month maintenance phase. However, in the control group a reduction in dietary sodium intake occurred during the maintenance phase as well. This suggests a trial-effect, supported by the observation that baseline values did not differ from pre-study outpatient clinic values. There are several possible explanations for this trial-effect. First, regression to the mean may have occurred, although this is not likely a problem according to our subgroup analysis on pre-study sodium excretion values. Second, patients that participate in a RCT are generally more motivated than those who decline participation. This was also literally expressed during the focus groups in intervention patients. Also, patients allocated to control group may be more willing to reduce sodium intake a priori and could have become more vigilant to their sodium intake, because they had to collect 24-hour urine collections. Three-monthly 24-hour urine collection may not fully reflect care as usual for some patients, and could not be performed in a blinded fashion. Some participants indeed expressed disappointment about being allocated to the control group at the baseline visit. This so-called contamination effect is a major bias of RCTs in behavioral interventions that by design cannot be performed in a blinded fashion (29). Our patients scored relatively high on physical and mental health scales as assessed by SF-12 compared to studies performed in simi-lar populations (30, 31). The observed difference in change in SF-12 between intervention and control group suggests that patients feel better about their health after the maintenance phase. Notably, our participants had relatively high self-management skills, indicated by PIH-scores, at baseline which might explain why their score did not further improve during the study. This also

Chapter 10

182

bolsters the notion that our population may be selected. Indeed, the majority of the patients already possessed a home blood pressure monitor at the start of the study.

The effect on sodium restriction observed here is comparable to other behavioral interventions that addressed sodium intake. The 18-month PREMIER study in untreated (pre)hypertensives consisted of biweekly behavioral counseling in the first half year that aimed at weight reduc-tion alone, or combined with adherence to the DASH diet (32), or advise-only (33). Sodium excretion was reduced with 31.6, 32.6 and 20.6 mmol/day respectively, which is comparable to the 41 mmol/day change achieved in our intervention phase, and also in line with 44 and 33 mmol/day reductions achieved in the TOHP trials (13). There are very few studies that in-vestigated behavioral interventions in CKD for sodium restriction. The MASTERPLAN study was performed in a setting similar to SUBLIME, namely a multicenter outpatient clinical study in the Netherlands. It investigated in 788 Dutch CKD patients a nurse-led intervention that targeted eleven treatment targets, including adherence to dietary sodium intake < 2000 mg (90 mmol) per day (34). MASTERPLAN did not address all components of self-regulation theory (34) and had a long intervention phase of two years, with on average 7.2 outpatient clinic visits yearly (35). The MASTERPLAN intervention had no effect on sodium excretion (150 vs 148 mmol/day). The authors contribute this to contamination bias and to the observation that multiple-target approaches tend to be ineffective. Of note, this study was effective in increasing compliance to several pharmacological interventions, but with none of the lifestyle factors. This reinforces the rationale of a specific behavioral approach for lifestyle related factors. In line with this as-sumption, the ESMO intervention in CKD based on SRT showed effective reduction of sodium excretion and blood pressure (Meuleman et al., Am J Kidney Dis, in press).

Cost-effectiveness of a sodium intervention program in CKD was previously assessed in the CanPREVENT study. The study addressed 8 surrogate targets in patients with CKD, and did not achieve better control of most risk factors e.g. blood pressure or cholesterol, notably sodium intake was not targeted (36). The CanPREVENT relied on nurse-led self-management based intervention. Notwithstanding, the cost-effectiveness analysis of CanPREVENT was deemed favorable in that the intervention reduced costs without reducing quality of life for patients (37). The reported, not significant difference of 0.046 QALYs gained in the intervention group (n= 238) reflect a similar effect size of the 0.02 and 0.05 QALYs gained in our analysis. We feel that these results are not significant nor representative due to the strong influence of indi-vidual events in such small populations and the limited sensitivity of the EQ-5D questionnaire for effects of the intervention on quality of life. Higher sodium intake correlates with higher antihypertensive drug use in 141 patients with CKD Stage 4 and 5 (38). In SUBLIME there was more dose reduction in the intervention group after the intervention phase, while after the maintenance phase both groups displayed a similar incidence of dose reduction, in line with the effects on sodium intake.

183


10

What are the implications of our study? It presents an effective, multidisciplinary integrated strategy of blended care for sodium management in CKD, developed in co-creation, and appre-ciated by patients, thus providing a basis for affordable programs for management of sodium-intake in CKD. The evaluation by patients provided several options for further improvements, and suggestions for improved user-friendliness of the interface, that may prove useful during further implementation. By its affordability it fulfills a prerequisite for large scale application, be it in clinical practice, or in the context of prospective intervention studies.

Acknowledgements

For the SUBLIME Investigators: project leaders P.J.M. van der Boog, S. van Dijk, G.J. Navis; proj-ect coordination: J.K. Humalda; the investigators: O.A. Blanson Henkemans, W.J.W. Bos, M.H. de Borst, G. Klaassen, T. Kok, G.D. Laverman, Y. Meuleman, W. Otten, H. Piels, E.J.M. Straathof, K.M. Vermeulen, L.C. Verschuur, H. de Vries, G.A. Welker.

We acknowledge the contribution of: E. Buskens, E. Corpeleijn, C. van Daelen, M.W. Dijk-Schaap, B. Gabel, J.D. de Groot, E. van Houdt, C.G.J. Indemans, I.N. Kunnekes, H. Piels, T. Rövekamp, I. Schultze, A. Spijker, M. Storm, I.M. van Weverwijk, D. Wiefferink.

The SUBLIME study was funded by a grant from the Netherlands Organization for Health Research and Development (ZonMw, project number 837001005 of the ‘Doelmatigheidsonder-zoek 2013-2015’ program) and by the Dutch Kidney Foundation (number PV48). The e-health module was developed by Bonstato.

Chapter 10

184

References

1. Beaglehole R, Bonita R, Horton R, et al. Priority actions for the non-communicable disease crisis. Lancet 2011; 377: 1438-1447.

2. Humalda JK, Navis G. Dietary sodium restriction: a neglected therapeutic opportunity in chronic kidney disease. Curr.Opin.Nephrol.Hypertens. 2014; 23: 533-540.

3. National Kidney Foundation. K/DOQI Clinical Practice Guidelines on Hypertension and Antihypertensive Agents in Chronic Kidney Disease. Am J Kidney Dis 2004; 43:S1-S290 (suppl 1)

4. Werkgroep Richtlijn Chronische Nierschade, Rensma PL, Hagen EC, et al. Richtlijn voor de behandeling van atiënten met Chronische Nierschade (CNS). 2009.


6. Lambers Heerspink HJ, Holtkamp FA, Parving HH, et al. Moderation of dietary sodium potentiates the renal and cardiovascular protective effects of angiotensin receptor blockers. Kidney Int. 2012; 82: 330-337.


8. Bibbins-Domingo K, Chertow GM, Coxson PG, et al. Projected effect of dietary salt reductions on future cardiovascular disease. N.Engl.J.Med. 2010; 362: 590-599.

9. de Borst MH, Navis G. Dietary salt reductions and cardiovascular disease. N.Engl.J.Med. 2010; 362: 2224-5; author reply 2225-6.

10. Bruck K, Stel VS, Gambaro G, et al. CKD Prevalence Varies across the European General Population. J.Am.Soc.Nephrol. 2016; 27(7):2135-47.

11. de Zeeuw D, Hillege HL, de Jong PE. The kidney, a cardiovascular risk marker, and a new target for therapy. Kidney Int.Suppl. 2005; (98): S25-9.


13. Cook NR, Cutler JA, Obarzanek E, et al. Long term effects of dietary sodium reduction on cardiovascular disease outcomes: observational follow-up of the trials of hypertension prevention (TOHP). BMJ 2007; 334: 885-888.

14. Maes S, Karoly P. Self-Regulation Assessment and Intervention in Physical Health and Illness: A Review. Appl.Psychol. 2005; 54: 267-299.

15. Meuleman Y, Ten Brinke L, Kwakernaak AJ, et al. Perceived Barriers and Support Strategies for Reducing Sodium Intake in Patients with Chronic Kidney Disease: a Qualitative Study. Int.J.Behav.Med. 2015; 22: 530-539.

16. NEVO-online versie 2013/4.0, RIVM, Bilthoven. 17. Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann.Intern.

Med. 2009; 150: 604-612. 18. Mancia G, Fagard R, Narkiewicz K, et al. 2013 ESH/ESC Practice Guidelines for the Management of Arterial

Hypertension. Blood Press. 2014; 23: 3-16. 19. Korevaar JC, Merkus MP, Jansen MA, et al. Validation of the KDQOL-SF: a dialysis-targeted health measure.

Qual.Life Res. 2002; 11: 437-447. 20. EuroQol--a new facility for the measurement of health-related quality of life. The EuroQol Group. Health

Policy 1990; 16: 199-208.

185


10

21. Petkov J, Harvey P, Battersby M. The internal consistency and construct validity of the partners in health scale: validation of a patient rated chronic condition self-management measure. Qual.Life Res. 2010; 19: 1079-1085.

22. Ware JE, Kosinski M., Turner-Bowker D., Gandek B. How to score version 2 of the SF-12 health survey (with a supplement documenting version 1). QualityMetric Incorporated, Lincoln, R.I.: 2002.

23. Hakkaart- van Roijen L, Tan SS, Bouwmans CAM. Handleiding voor kostenonderzoek. Methoden en stan-daard kostprijzen voor economische evaluaties in de gezondheidszorg. CVZ 2010.




27. Ix JH, Wassel CL, Stevens LA, et al. Equations to estimate creatinine excretion rate: the CKD epidemiology collaboration. Clin.J.Am.Soc.Nephrol. 2011; 6: 184-191.

28. Miniwatts Marketing Group. Internet World Stats. 2015. http://internetworldstats.com/stats.htm. 29. de Bruin M, McCambridge J, Prins JM. Reducing the risk of bias in health behaviour change trials: improv-

ing trial design, reporting or bias assessment criteria? A review and case study. Psychol.Health 2015; 30: 8-34.

30. Deng N, Allison JJ, Fang HJ, Ash AS, Ware JE,Jr. Using the bootstrap to establish statistical significance for relative validity comparisons among patient-reported outcome measures. Health.Qual.Life.Outcomes 2013; 11: 89-7525-11-89.

31. McClellan WM, Abramson J, Newsome B, et al. Physical and psychological burden of chronic kidney disease among older adults. Am.J.Nephrol. 2010; 31: 309-317.

32. Appel LJ, Champagne CM, Harsha DW, et al. Effects of comprehensive lifestyle modification on blood pressure control: main results of the PREMIER clinical trial. JAMA 2003; 289: 2083-2093.

33. Chang A, Batch BC, McGuire HL, et al. Association of a reduction in central obesity and phosphorus intake with changes in urinary albumin excretion: the PREMIER study. Am.J.Kidney Dis. 2013; 62: 900-907.

34. Van Zuilen AD, Wetzels JF, Bots ML, Van Blankestijn PJ, MASTERPLAN Study Group. MASTERPLAN: study of the role of nurse practitioners in a multifactorial intervention to reduce cardiovascular risk in chronic kidney disease patients. J.Nephrol. 2008; 21: 261-267.

35. van Zuilen AD, Bots ML, Dulger A, et al. Multifactorial intervention with nurse practitioners does not change cardiovascular outcomes in patients with chronic kidney disease. Kidney Int. 2012; 82: 710-717.

36. Barrett BJ, Garg AX, Goeree R, et al. A Nurse-coordinated Model of Care versus Usual Care for Stage 3/4 Chronic Kidney Disease in the Community: A Randomized Controlled Trial. Clinical Journal of the American Society of Nephrology 2011; 6: 1241-1247.

37. Hopkins RB, Garg AX, Levin A, et al. Cost-effectiveness analysis of a randomized trial comparing care models for chronic kidney disease. Clin.J.Am.Soc.Nephrol. 2011; 6: 1248-1257.

38. Boudville N, Ward S, Benaroia M, House AA. Increased Sodium Intake Correlates With Greater Use of Antihypertensive Agents by Subjects With Chronic Kidney Disease. American Journal of Hypertension 2005; 18: 1300-1305.

Chapter 11Summary and General Discussion

189

Summary and General Discussion

11

Summary.

The work in this thesis addressed the interaction between mineral metabolism, with a particular focus on fibroblast growth factor 23 (FGF-23), and volume homeostasis, mostly in the context of renin-angiotensin-aldosterone system (RAAS) blockade –the mainstay of chronic kidney disease treatment. In the first part of the thesis, we assessed the role of FGF-23 in association with several read-outs of volume status. In the second part of the thesis, we investigated the effect of several dietary interventions on phosphate and FGF-23 and conclude with a prospective randomized controlled trial that deploys a multidisciplinary approach may be used to achieve a sustainable change in dietary elements of a healthy lifestyle.

RAAS blockade lowers blood pressure and proteinuria and thus protects the kidney from pro-gressive fibrosis and further renal function loss. However, RAAS blockade does not completely abrogate proteinuria. The more ‘residual’ proteinuria remains, the higher the risk for cardiorenal complications. Interestingly, emerging studies indicate that several dietary factors (i.e. vitamin D, sodium and phosphate) may influence the efficacy of RAAS blockade to reduce proteinuria.

In (→ Chapter 2), we reviewed the role of vitamin D as adjunct to RAAS blockade for residual proteinuria reduction. We describe safety concerns of indiscriminate combinations of RAAS blockade. We suggested that vitamin D by inhibitory effects on RAAS on the one hand, and RAAS-independent inhibition of fibrosis on the other hand, may preserve renal function. Vitamin D supplementation does not seem to abolish the increased cardiovascular risk in patients with CKD, although results of prospective studies are yet to follow. The review confirmed residual proteinuria as a treatment target, identified an interaction between mineral bone homeostasis with RAAS, identified a possible interplay of sodium intake with vitamin D, but also attenuated expectations of vitamin D supplementation alone for augmentation of RAAS blockade-based renoprotective effects.

In (→ Chapter 3) we reviewed the role of sodium intake in patients with CKD. Sodium intake is high in both the general population and patients with CKD. The latter is problematic, as high sodium intake blunts the efficacy of RAAS blockade. We summarized that even a modest reduc-tion of sodium intake has beneficial effects on blood pressure and proteinuria, emphasized that patients with CKD are particularly prone to sodium retention, and we suggested that parameters of volume overload may identify patients who benefit most from a moderate sodium restriction.

Besides high sodium intake, increased concentrations of serum phosphate are also associated with an impaired antiproteinuric response to RAAS blockade. We found in (→ Chapter 4) that higher concentrations of the phosphate-regulating hormone FGF-23 were independently asso-ciated with higher residual proteinuria after sodium restriction combined with RAAS blockade.

Chapter 11

190

Moreover, FGF-23 was correlated with N-terminal probrain natriuretic peptide (NT-proBNP), a marker for volume overload, and increased aldosterone to renin ratio (ARR), a marker for less successful RAAS blockade. Thus, FGF-23 correlated with a more volume-overloaded phenotype and with an impaired efficacy of main pharmaceutical and dietary intervention options.

In (→ Chapter 5) we expanded these findings to renal transplant recipients (RTR). Although FGF-23 levels may be reduced after transplantation, RTR with persistently higher FGF-23 concentra-tions remain at increased risk for cardiovascular mortality. FGF-23 independently contributed to established cardiovascular risk factors. Moreover, FGF-23 was correlated with NT-proBNP, mid-regional fragment of pro‒A-type natriuretic peptide (MR-proANP), and copeptin, a surrogate marker for vasopressin. These three hormones could be considered as markers for increased volume strain on the heart.

Patients with end stage renal disease are fully dependent on haemodialysis to reduce volume overload. FGF-23 concentrations reach extreme values in haemodialysis patients. We identified in (→ Chapter 6) that haemodialysis patients who had a greater interdialytic gain of volume and thus a necessity for higher ultrafiltration volumes, were the ones with the highest FGF-23 concentrations. We thus identified ultrafiltration volume as a novel determinant of FGF-23 in haemodialysis patients, independent from known determinants phosphate and calcium. Here too, we found that echocardiographic and biochemical markers of volume status correlated with FGF-23, namely cardiac output, copeptin and end-diastolic volume. Also, patients with a greater reduction of relative blood volume had a greater reduction of FGF-23 levels during one dialysis session. Taken together, the abovementioned findings imply a strong interplay between mineral-bone parameters including FGF-23, volume status and RAAS blockade efficacy.

Next, we investigated whether, vice versa, volume loading or volume reduction by sodium restriction could change FGF-23 concentrations in (→ Chapter 7). We found that acute saline infusion did not change FGF-23 in arterial hypertensive patients, nor did 6 weeks of sodium restriction change FGF-23 in patients with diabetic nephropathy. As sodium interventions ap-pear not to influence FGF-23 –whereas phosphate interventions do–, it is key to explore the relation between these dietary components.

We reported that sodium intake and phosphate intake were strongly correlated in different populations in (→ Chapter 8): namely healthy controls, diabetes mellitus patients and RTR. Moreover, a modest sodium restriction in patients with CKD was accompanied by a reduction in phosphate intake. We hypothesized that this is partly due to phosphate-rich additives in processed foods, which may explain why a phosphate-additive rich diet has been found to in-crease FGF-23 concentrations. A healthy diet thus would be low in both sodium and phosphate. Conversely, healthy food choices are rich in potassium, and high potassium intake is associated

191


11

with lower blood pressure. Potassium supplementation was found to increase phosphate reab-sorption. This suggests that potassium may influence phosphate regulation.

In (→ Chapter 9), we thus hypothesized a role for potassium in FGF-23 regulation. Potassium supplementation lowered FGF-23 concentrations in prehypertensive participants, accompanied by a rise in tubular phosphate reabsorption and serum phosphate concentrations, without ef-fects on other mineral-bone hormones as parathyroid hormone or vitamin D. This may warrant follow-up research on the effect of a prolonged change in potassium intake on FGF-23 and other mineral-bone parameters.

However, sustainable change in behavior is notoriously difficult to achieve. To address this problem, (→ Chapter 10) reported the first results of the randomized controlled, multicenter trial SUBLIME. SUBLIME is based on the behavioral change theory self-regulation, and aimed to facilitate a moderate sodium restriction. To this end, SUBLIME deployed an E-health module that addressed the several stages of behavioral change, group meetings, and individual e-coaching. SUBLIME reduced sodium excretion and blood pressure in the intervention group after the 3-month intervention phase, however after 6-month maintenance phase also the control group reduced sodium excretion. We contribute this to increased awareness which warrants further investigation. The SUBLIME intervention was valued by the participants, and their feedback offers guidance for future development of E-health interventions that can facilitate behavioral change in dietary lifestyle.

General Discussion.

Chronic Kidney Disease‒ When Deranged FGF-23 and Volume Status Meet.

The work in this thesis encompasses several potential causes and effects of deranged fibroblast growth factor 23 (FGF-23) regulation across the spectrum of chronic kidney disease (CKD). We particularly focused on the potential interaction between FGF-23 and volume status. Progres-sive derangement of volume homeostasis is apparent throughout the course of renal disease, with high incidences of hypertension (1) and heart failure (2, 3) in patients with CKD. Particularly in haemodialysis patients, more pronounced fluid accumulation is associated with a higher risk of mortality (4-7). Potassium has recently been re-discovered as a modulator of blood pres-sure, where higher potassium intake has been consistently linked with lower blood pressure and reduced cardiovascular risk (8, 9). Interestingly, there seems to be interaction between sodium and potassium intake, concertedly influencing volume status (10). We thus expanded known bone (calcium, phosphate) and classical renal (glomerular filtration rate) determinants of FGF-23 with novel determinants: a high ultrafiltration volume in haemodialysis patients and potassium intake in patients at risk for development of hypertension, respectively; to conclude

Chapter 11

192

that potassium, phosphate and sodium are connected on a far more complex level. We also observed that high FGF-23 concentrations predispose to an impaired efficacy of antiproteinuric therapies that target volume status. In this thesis, we thus explored the interaction between FGF-23, a main player in the phosphate-regulating hormonal system, and measures of deranged sodium and volume homeostasis in different clinical settings.

FGF-23 and Volume Overload‒ Correlation, Cause and Effect.

FGF-23 and Markers of Volume OverloadFGF-23 correlates with measures of volume overload in several populations. In 1851 elderly Italian individuals from the general population, FGF-23 concentrations were associated with higher levels of N-terminal probrain natriuretic peptide (NT-proBNP), a marker of left ventricular wall strain and generally used as marker for volume overload in the setting of heart failure (11). FGF-23 correlates with BNP in heart failure patients (12) and with NT-proBNP levels in patients that underwent elective coronary angiography, also in patients with preserved cardiac and renal function (13). We confirmed the association in patients with chronic kidney disease in (→ Chapter 4) and renal transplant recipients in (→ Chapter 5). These findings are robust, as FGF-23 also correlated with copeptin, a surrogate marker for the vasopressin which is the hormone that facilitates water retention, and mid-regional type atrial natriuretic peptide, another marker for ventricular wall strain (→ Chapter 5). In haemodialysis, patients with more fluid overload require more fluid withdrawal per dialysis session. We found that this so-called ultrafiltration volume and FGF-23 are strongly correlated, which supports our notion that differ-ences in required ultrafiltration volume may explain the wide variance in FGF-23 concentrations in haemodialysis patients (→ Chapter 6). Of note, this is independent from the earlier described determinants phosphate, calcium and residual renal function (14-16), and thus suggests that FGF-23 may connect mineral metabolism with volume status, both of which are progressively deregulated in CKD.

FGF-23 and Clinical Outcomes in Settings of Volume OverloadFGF-23 has been linked to new-onset heart failure, as well as adverse outcomes in volume-overloaded patients, both with and without (primary) renal disease. First, higher FGF-23 concentrations are associated with a higher risk for development of heart failure in predialysis CKD patients (2, 3), but also in the general population (17-19). In the Atherosclerosis Risk in Communities study, the 1592 of 11638 participants in the highest FGF-23 category had a signifi-cantly increased risk of new-onset heart failure, even after adjustment for renal function and cardiovascular risk factors (20). In line, in the Cardiovascular Health Study, 3107 individuals from the general population were followed for a median of 10.5 years. During the study, 1730 deaths and 697 heart failure events were reported. Participants in the highest FGF-23 quartile had a 2.47-fold increased risk of developing heart failure, and these associations were stronger in

193


11

patients with CKD (17). In the same population, FGF-23 was not associated with sudden cardiac death (18), which suggests that FGF-23 is not associated with fatal arrhythmic cardiovascular events (e.g., ventricular fibrillation) but rather with progressive heart failure (18). In fact, the only arrhythmic cardiovascular disorder that does correlate with FGF-23 is rarely lethal but strongly related to volume overload: atrial fibrillation. This has been demonstrated in the gen-eral population (21), patients who underwent coronary angiography (13), and haemodialysis patients (22).

Second, in patients with established heart failure, an elevated FGF-23 is a marker of increased disease severity and poor prognosis, both in adults (23) and in children (24). Even in the extreme and acute setting of cardiogenic shock, higher FGF-23 concentrations herald poor outcomes (25, 26). The sudden increase of FGF-23 concentrations in acute distress and the adverse prognosis of higher concentrations, has also been observed in end stage liver disease (27), acute kidney injury after cardiac surgery (28), and sepsis in patient with CKD (29). Whether FGF-23 is also increased in response of the formation of an acellular membrane upon intra-ocular lens implant (30), remains to be elucidated. Of note, some of these conditions may correlate with loss of klotho, the co-receptor of FGF-23, for example in dehydration (31). CKD may be perceived as a state of klotho loss, leading to high FGF-23 concentrations. Contrary to FGF-23, klotho does not correlate with cardiovascular or renal outcomes (3, 32), a discrepancy that could be attributed to analytical difficulties in available klotho assays (33). As the potential role for klotho in the clinical setting remains elusive, we chose to focus this thesis on FGF-23.

A long-term effect of volume overload is left ventricular hypertrophy (LVH). High FGF-23 con-centrations are associated with an increased incidence of LVH (34) and higher left ventricular mass index (LVMI) in many reports (35, 36). However, when the heart was assessed with MRI instead of echography, the association between FGF-23 and LVMI was not present (37). In line with this report, we could not demonstrate a correlation between FGF-23 and LVMI in haemodialysis patiens (→ Chapter 6) either, suggesting that the correlation may not solely rely on volume overload, but partly relies on other factors that affect dimensions of the heart (36, 38). In line, the association between FGF-23 and cardiovascular mortality in renal transplant recipients remained significant after adjustment for markers of volume overload (→ Chapter 5 ), also suggesting that the adverse effects of FGF-23 on the heart are not only exerted via volume overload.

Nevertheless, the relationship between FGF-23, parameters of volume homeostasis, and clinical outcomes related to volume overload are in line with the concept that FGF-23 connects deregu-lated mineral metabolism and volume homeostasis, an observation not restricted to the CKD population.

Chapter 11

194

FGF-23 and Volume-targeting TherapiesThe consistent association between FGF-23 and volume status raises the question whether the relationship between FGF-23 and volume status is simply unidirectional, namely that FGF-23 induces volume overload. The other way round, volume status may also be a valid target to reduce high FGF-23 concentrations. There are, however, several issues undermining this hy-pothesis. For example, in the aforementioned pediatric patient population (24), diuretic use was associated with FGF-23, and FGF-23 correlated strongly with NT-proBNP. However, the association diuretic‒FGF-23 could result from several distinct cause‒effect relations. Diuretics may increase FGF-23 levels through effects on renal function; or diuretics are prescribed to patients that were already volume-overloaded (confounding by indication) which induced FGF-23 production, analogous to our findings in haemodialysis patients in (→ Chapter 6); or FGF-23 may merely identify patients with volume overload without a causal relation.

In non-diabetic CKD patients (→ Chapter 4), we report that patients with CKD in the highest ter-tile of FGF-23 also demonstrated a higher blood pressure, aldosterone levels and nominally but not significantly higher NT-proBNP levels, with similar proteinuria compared with lower tertiles. When these patients adhered to a low sodium diet, FGF-23 concentrations did not materially change whereas blood pressure and proteinuria were reduced markedly (39). Also in patients with CKD and diabetes (→ Chapter 7), low-sodium diet or use of the diuretic hydrochlorothia-zide did not change FGF-23 levels, despite profound effects on blood pressure and albuminuria (40). Baseline FGF-23 concentrations were associated with an impaired antiproteinuric re-sponse to the low sodium diet in nondiabetic patients with proteinuric CKD (→ Chapter 4), and this association was discernible yet less obvious in diabetic CKD (→ Chapter 7). This suggests that although FGF-23 and volume status are correlated and can predict therapeutic response, alteration of volume status does not necessarily induce a change in FGF-23 concentrations in patients already on RAAS blockade. Start of RAAS blockade on the other hand, reduced blood pressure, proteinuria and FGF-23 levels in 67 diabetic patients with CKD stage 1 (41), in line with reports in experimental diabetic CKD where ACE inhibition reduced renal expression of FGF-23, with a very modest effect on serum concentrations (42). The latter corroborates the bone as the principal site of FGF-23 production (43), with the possibility of ectopic production of FGF-23 in injured tissue as has been suggested for kidney (42) and heart (44). However, in the setting of heart failure myocardial FGF-23 production was not induced, conversely circulating FGF-23 concentrations did rise (45). We interpret these findings as follows: in volume overload, the contribution of ectopically expressed FGF-23 to circulating FGF-23 concentrations is negligible compared with the contribution of bone-produced FGF-23. This is in accordance with the hy-pothesis where the bone is at the center of a systemic interaction between mineral metabolism and volume homeostasis (46).

195


11

FGF-23 and Minerals‒ From Single Elements to Complex Systems in Renal Health

FGF-23 and Sodium: New Kid and Usual Suspect Intertwined.First, indirect evidence suggesting an effect of FGF-23 on tubular sodium handling, and thus volume homeostasis, emerged in 2012 (47). When Ste-20-related kinase (SPAK) is modified to resist the inhibiting actions of the fourth With-no-lysin (K) kinase, WNK4, FGF-23 is upregulated. The WNK-SPAK kinase system is crucial for regulation of the expression of several sodium chan-nels in the renal tubuli, e.g. the sodium(Na+)-chloride cotransporter (NCC) and the epithelial Na+ channel (ENaC), and because its details are beyond the periscope of this manuscript, we refer to excellent reviews (48, 49). In short, WNK4 inhibits SPAK, SPAK phosphorylates and so stimulates NCC action, promoting sodium retention. Aberrant SPAK activation led to higher FGF-23 concentrations (47). More conclusive evidence followed in 2014 when Andrukhova et al. demonstrated that administration of recombinant FGF-23 (rFGF-23) to mice stimulated WNK4 production and increased NCC expression by 40% (50). Indeed, rFGF-23-treated animals displayed reduced sodium excretion and higher blood pressure, and hydrochlorothiazide ab-rogated rFGF-23 induced effects (50). Conversely, in this study WNK4 functioned as a positive regulator of NCC, which is at variance with its general assumed role as negative regulator of NCC activity (48, 49). This ambivalence has been described earlier, and apparently the role of WNK4 varies in specific conditions (51). Nevertheless, rFGF-23 actually reduced aldosterone levels in mice fed a low-sodium diet (50). This suggests that FGF-23 may have such profound effects on sodium retention and blood pressure, that it partly substitutes the classical effect of aldosterone. This could be part of the explanation why FGF-23 and sequelae from volume overload as heart failure and LVH steam jointly and relentlessly to adverse outcomes in our patients. Other way round, it implies that patients with high FGF-23 levels, i.e. renal patients, may be more susceptible to the effects of a given amount of sodium intake.

Sodium in Health and Chronic Kidney Disease: Salt in the Balance.In the general population, every 6 gram increase in salt intake correlated with a 45% increased risk of cardiovascular disease 26% higher risk of mortality (52). This is in line with the findings of O’Donnell et al. where a salt intake > 6 grams per day was associated with increased risk of car-diovascular events and mortality (8). In this study, the adverse effect of sodium intake was more pronounced in hypertensive patients. Sodium intake in renal patients is on average 164 mmol/day (53), which is comparable to the estimated global sodium consumption of 172 mmol/d (10 grams of salt), and glaringly deviant from guidelines that recommend a salt intake of less than 85 mmol (5 grams) per day (54-56). This was not different in the cohorts studied in this thesis (→ Chapter 4, 7, 8). The concept that particular population subgroups are more susceptible to the deleterious effects of high sodium intake is also reflected by findings from the PREVEND cohort, showing an association between high sodium intake and adverse cardiovascular effects only in patients with an elevated NT-proBNP at baseline (57). As reviewed in (→ Chapter 2), pa-

Chapter 11

196

tients with CKD are particularly salt-sensitive, i.e. their blood pressure increases and decreases in par with sodium intake. High sodium intake in renal patients was associated with a higher risk for ESRD (58), although a post hoc analysis of the Modification of Diet in Renal Disease (MDRD) found no association between baseline high sodium intake (>130 mmol/day) and ESRD (59). This discrepancy may follow from higher sodium intake and proteinuria in the Ramipril Efficacy In Nephropathy population (REIN) (58), although it must be stated that the MDRD study consisted of a three year dietary intervention, and changes in lifestyle/dietary intake in this period were not accounted for. This is less likely a problem in REIN, as patients were only recom-mended to limit sodium intake but did not receive extensive dietary counseling as in MDRD. Moreover, RAAS blockade was only used by 36% of the MDRD participants (59), this also may explain this discrepancy, as the combination of RAAS blockade and volume targeting by sodium restriction is synergetic (60). Reduction of sodium intake potently reduces blood pressure in both normotensive and hypertensive populations without other diseases (61-63). For example, in a meta-analysis of 6970 persons a linear dose-response relationship was observed: every 100 mmol reduction of sodium intake reduced blood pressure with 3.8 mmHg (63), in line with an earlier systematic review (64). A cost-effectiveness analysis found that even a modest reduction of sodium intake to less than 130 mmol/day (7.6 grams of salt per day), would annually reduce the incidence of coronary heart disease, stroke and myocardial infarction; reduce deaths-from-any-cause with 44.000 per year; and save 10‒24 billion dollars of health care expenditure, every year (65)! In this analysis, costs of kidney disease were not taken into account. We expect that this would have increased the benefits of sodium restriction even further, given our findings in the SUBLIME trial (→ Chapter 10).

Twenty-four-hour urine collection is still the gold standard for assessment of dietary sodium intake. The abovementioned studies –and studies conducted by ourselves‒ rely on the premise that in the steady state, sodium excretion equals sodium intake. This deserves more nuance, following the discovery of nonosmotic sodium buffering mechanisms in the skin (66) and macrophages (67), and rhythmic variability in sodium excretion (68). However, 24-hour urinary sodium excretion remains the best estimate for sodium intake (68). In our studies, participants had 4‒9 weeks time to reach steady state, which would minimize the effect of nonosmotic buff-ering. However, the daily variance may explain the relative modest results, and partly explain anecdotal discrepancies between measured sodium excretion and the participant’s perception of sodium intake. One solution lies in urinary collections of three or even nine days in a row (68), although this may face resistance for reasons of feasibility. Alternatively, we could embrace this variability as guidance in joint clinical decision-making and thus step back from overzealous reliance on isolated figures: instead, interpret a urinary excretion rate in context of previous measurements and in combination with other clinical information, e.g. blood pressure and other nutrients that can readily be measured in 24-hour urine collection. This sets the stage for personalized dietary management.

197


11

Phosphate and Sodium: A Deleterious Dietary Duo?The identification of multiple dietary factors, or dietary profiles, by means of 24-hour urine collections offers more information than focus on a single nutrient. We applied this to study the concordance of sodium and phosphate in our diet. In (→ Chapter 8), we describe that sodium and phosphate excretion are tightly correlated in several patient populations. We considered this of relevance, as high serum phosphate levels are associated with a higher risk of mortality in CKD (69). In contrast, conflicting data have emerged regarding the relationship between phosphate intake and serum phosphate (70). Moreover, the association between high phosphate intake and long-term outcomes is not unequivocal. Although high phosphate intake has been associated with increased mortality (71) and left ventricular mass (72), conversely in the Heart and Soul study it was associated with fewer cardiovascular events (73), and a link between phosphorus intake and mortality could not be discerned in NHANES III (74) nor in the MDRD study (70).

In addition, assessment of dietary intake often relies on food frequency questionnaires or dietary recall and this was found to correlate weakly at best to fasting serum phosphate levels (75), in line with the earlier notion that dietary intake is not the primary determinant of serum phosphate levels. On the one hand, fasting serum phosphate levels may underestimate the ef-fect of phosphate loading due to diurnal variation, while on the other hand dietary assessment underestimates the contribution of phosphate-rich additives (76). Diurnal variation may also explain the absence of an effect for phosphate : creatinine ratio on mortality in fasting, morning urinary samples (77).

Part of the explanation of the conflicting evidence for the relation of phosphate load with adverse outcomes may thus lie in our inability to adequately determine phosphate intake. Oral intake is difficult to estimate, and gut phosphate absorption can vary due to regulation by 1,25(OH)2 vitamin D and putatively other regulatory mechanisms (78), and also due to variations in bioavailability depending on phosphate source (79). Twenty-four-hour urine collections are therefore the parameter of choice to assess effects of phosphate restriction or binder therapy on intestinal uptake of phosphate (80). Although the relation between lowering phosphate intake and outcome is well-established in ESRD, it is unknown what role phosphate intake plays in the setting of the general population.

Lowering dietary phosphate intake is a challenge in itself, as phosphate is present ubiquitously in food products (81, 82). A major part of the phosphate is ‘hidden’ as food-additives, indeed additive-rich products can easily contain 66% more phosphate than its non-phosphate based equivalent preservative (83). Moreover, the bioavailability of additive-derived inorganic phos-phate is almost 100%, whereas phosphate from animal or vegetable sources is far less avidly absorbed (60% and 40%, respectively (79)). This may explain why a adherence to a vegetarian

Chapter 11

198

diet for one week lowered FGF-23 levels in healthy volunteers (84), and the Western diet was associated with higher phosphate excretion and FGF-23 concentrations in participants from African ancestry living in the U.S. compared with participants living in rural Nigeria. Mere avoid-ance of food additives achieved a reduction in serum phosphate levels in haemodialysis patients (85). As many additives contain both sodium and phosphate (e.g. disodiumdiphosphate), it is not surprising that a recent RCT found that an additive-enriched diet increases sodium and phosphate intake concomitantly 60% (86), in line with our findings in (→ Chapter 8). These re-sults corroborate a more holistic approach in nutritional sciences, assessing shared components of dietary patterns rather than focusing on a single element.

FGF-23 Reduction by Dietary Phosphate: a Complex SystemAnother RCT confirmed that one week of additive-enriched diet increased phosphate and sodium excretion, and also increased FGF-23 concentrations (87). An increase in FGF-23 could be reconciled with the potentially adverse effects of high phosphate intake. However, this hy-pothesis is torpedoed by compensatory metabolic changes observed over time after sustained changes in phosphate intake. Trautvetter et al. described that high phosphorus intake achieved by inorganic sodium phosphate supplements increased FGF-23 only in the first four weeks in 62 healthy volunteers. After eight weeks of phosphate supplementation, FGF-23 concentrations returned to their normal values while phosphate excretion remained elevated (88). So, high di-etary phosphate intake increases FGF-23 on the short-term (two to five days) (89-91), increases phosphate excretion without influence on fasting serum phosphate levels (88-91) –although a circadian rhythm of serum phosphate concentrations is observable (91)‒, but the effect on FGF-23 does not persist. We hypothesize that the physiological role of FGF-23 is to induce a change in the set point for tubular maximum phosphate reabsorption (TmP, when expressed per nephron TmP/GFR), and is not necessary to maintain a given set point. For example, a change in FGF-23 may result in decrease or increase of sodium phosphate cotransporter expression in the tubuli, these transporters may remain in place long after the change in FGF-23 levels. FGF-23 could be considered the push to the bridge engine order telegraph (“half ahead to full ahead”), the transporters as increased engine activity, acceleration as phosphate excretion and serum phosphate concentration as vessel speed, which will eventually reach a new steady state: for phosphate excretion at the same level as baseline/intake, for serum phosphate at a higher/lower concentration. This role of FGF-23 as set point regulator of phosphate balance has been postulated earlier (92). In summary, although our understanding of FGF-23 has evolved, the factors that actually drive (the change in) circulating FGF-23 and phosphate concentrations remain to be elucidated.

FGF-23 and Potassium: a New Horizon.Our findings in (→ Chapter 9) suggest a role for potassium in FGF-23 regulation. The first ob-servations that potassium supplementation stimulates phosphate reabsorption and increases

199


11

serum phosphate concentrations originate from the 1980’s (93, 94). Conversely, potassium deficiency downregulated NaPi transporters to induce phosphaturia (95). We extended these findings in a post hoc analysis of the KaNa trial in healthy prehypertensives, where we found that 4 weeks of potassium supplementation reduced FGF-23 concentrations, without an effect on the other phosphate-regulating hormones vitamin D and PTH. This reduction came along an increment in the set point for phosphate reabsorption (tubular maximum reabsorption per GFR, TmP/GFR), and consequently higher serum phosphate concentrations (Figure 1).

Figure 1. Schematic representation of the effects of potassium intake on regulation of phosphate homeostasis in the KaNa trial, Chapter 9. Potassium supplementation starts at day 0. This results in a drop of FGF-23, and thus increases the renal set point for phosphate reabsorption, TmP/GFR. Phosphate excretion temporarily falls, and reaches a new steady state at a higher serum phosphate concentration. Question marks depict uncertainty about the time course of changes in FGF-23 and PTH, as measurements took place only after 30 days.

Phosphate excretion was unchanged after four weeks, which means that phosphate intake did not change and a new steady state must have been reached within the study period (92). The ramifications of this finding are potentially immense, as there are very few strategies that achieve a reduction of FGF-23 levels (→ Chapter 1). These findings in a post hoc analysis are in line with large epidemiological studies showing that high potassium intake protects against renal function loss (96) and cardiovascular events (8), putatively due to blood pressure lowering effects (97). Our findings may add a new paradigm. Since osteocytes are the main source of circulating FGF-23, it seems that potassium regulates FGF-23 in bone, either directly or via an unknown intermediate. We hypothesize the existence of an elegant auxiliary mechanism for maintaining potassium homeostasis. Based upon the observation that potassium concentra-tions in bone extracellular fluid are five-fold that of serum (98), we postulate that potassium loading may be sensed by osteocytes through 1) potassium‒induced effects on osteocyte membrane polarization, 2) through potassium‒induced NCC phosphorylation (99, 100) as NCC is present in osteocytes (101), or 3) potassium‒mediated effects on the WNK4/SPAK signal-ing pathway (102) which may also be present in osteocytes given the association of a WNK4 variant with osteoporosis (103). WNK/SPAK signaling has been has been implicated in FGF-23

Chapter 11

200

regulation (47). Reduction of FGF-23 may in turn lower renal NCC expression (104), in order to facilitate potassium excretion by ENaC. Circumstantial evidence consists of the sharp rise of FGF-23 in the setting of cardiogenic shock (25, 26), end-stage liver disease (27), acute kidney injury (28), sepsis (29), heart failure (45) and chronic hypoxia (105), i.e. a state of multi-organ hypoperfusion and/or hypoxia, also affecting the bone, which seems to elicit FGF-23 release. The hypoperfusion of the bone syncytium may result in a loss of the potassium gradient, and the FGF-23 response may be auxiliary to cope with imminent potassium and phosphate leak from bone-in-distress. This would also explain the strong linear relation of large volume shifts in a haemodialysis session with an extreme FGF-23 level in (→ Chapter 6).

Nephrologists may be reluctant to advocate a liberal potassium diet because of concerns for hyperkalemia ‒ the dreaded complication of overzealous RAAS blockade. This resulted in premature termination of the VA-NEPHRON study, which investigated ACE inhibitor and ARB combination (106); as well downed the ALTITUDE trial (107), the successor of the ominously named AVOID study (108) that examined the direct renin inhibitor aliskiren. Again, we ought to separate here intake, excretion and the serum concentration: the mechanisms that are at play in coping with a potassium-rich diet may be beneficial. Spironolactone increases serum potassium more than losartan, even when renal excretion is constant. Either, spironolactone changes the set point for blood potassium concentrations, or potassium regulation outside the kidney occurs (109). High potassium excretion protected from renal function loss and ESRD incidence in a post hoc analysis of ONTARGET and TRANSCEND trials (96). Ironically, ONTARGET was the first study that reported adverse renal outcomes after dual RAAS blockade (110). High potassium excre-tion estimated from morning urine samples was associated with less cardiovascular events and mortality (8) and lower blood pressure (9) in the general population. In line, the lowest tertile of 24-hour urine excretion of potassium had an increased risk of developing hypertension (111). In a follow-up study of two Trials On Hypertension Prevention (TOHP I and TOHP II), a higher sodium to potassium ratio was associated with a greater incidence of cardiovascular disease (112). Potassium may be a proxy for low intake of fruits and vegetables, cereals, or vitamin K deficiency. To put in in other words, the model that integrated both sodium and potassium excretion was most informative. In a broader perspective, potassium is likely a missing element in our understanding of sodium and phosphate homeostasis, bolstering the idea of complex adaptive system where bone, heart and kidney health are connected on multiple levels. Fur-thermore, these findings again support the concept of focusing on dietary profiles, rather than individual components, in clinical practice.

Toward an Integrated Approach‒ Plotting a New Course for a Healthy Lifestyle.

The work in this thesis emphasizes the necessity of a broad perspective for understanding the derangements in bone‒mineral and other electrolytes in renal disease. Multiple organs inter-act with each other by several cross-talk mechanisms in a dynamic, adaptive or detrimental,

201


11

interactive way. These interactions connect kidney, bone and cardiovascular health in a complex web. Although the reductionalist approach ‒ dissecting a problem to the smallest entity (e.g., a chemical element as sodium) to be the cause of a single effect‒ aided in developing new hypotheses and initial tremendous progress, it may hamper further progress in preventive medicine, according to Fardet et al. (113). Despite identification of several harmful or beneficial components in our diet, prevention of welfare diseases has made minimal progress. Fardet et al. advocate to focus more on the healthy state and define contributors to the healthy core metabolism on the level of diet (instead of nutrient) (113). Indeed, a healthy diet encompasses avoidance of processed foods and a liberal intake of fruits, vegetables and nuts (114). Although a reductionist would attribute the beneficial effect solely to low sodium, low phosphate, and high potassium content of this diet, by definition unknown contributors and interactions are missed. Medication adherence in hypertension patients with CKD displays a similar need for improvement as in patients without CKD: three out of ten forgets their medication, 5% reports being careless about medication use (115), and if we use sodium excretion as proxy for compli-ance, adherence to a healthy diet is close to none (→ Chapter 2). A multidisciplinary approach involving pharmacists improved medication adherence and consequently blood pressure (116), and a Canadian program involving nurses and dietitians with therapy aimed at primary and secondary prevention in CKD was cost-effective (117). In the SUBLIME study (→ Chapter 10), ‒ a multidisciplinary approach developed in co-creation with patients, which consisted of behav-ioral approaches, dietary counseling, e-health technology and self-management‒ was highly valued by patients and achieved a sustained reduction in sodium excretion and blood pressure. SUBLIME focused on sodium both because of its health impact in CKD and because its intake can be objectively quantified. However, the intervention could easily be expanded to new food components, and dietary patterns. At the same time, registration of food intake could offer information on separate nutrients and their interactions. Unraveling the specific roles of FGF-23 helped us realize that in general, human health is the sum of many interactions. Personalized medicine should set goals and align several homeostatic set points in a manner that is most beneficial and appropriate for the individual patient. With our growing understanding of bone-mineral pathophysiology, we hope to contribute to the ultimate goal of healthy ageing for all.

J.K.H.

Chapter 11

202

References

1. Inker LA, Coresh J, Levey AS, Tonelli M, Muntner P. Estimated GFR, Albuminuria, and Complications of Chronic Kidney Disease. Journal of the American Society of Nephrology 2011; 22: 2322-2331.



4. Zoccali C, Torino C, Tripepi R, et al. Pulmonary congestion predicts cardiac events and mortality in ESRD. J.Am.Soc.Nephrol. 2013; 24: 639-646.

5. Agarwal R. Hypervolemia Is Associated With Increased Mortality Among Hemodialysis Patients. Hyperten-sion 2010; 56: 512-517.

6. Flythe JE, Kimmel SE, Brunelli SM. Rapid fluid removal during dialysis is associated with cardiovascular morbidity and mortality. Kidney Int. 2011; 79: 250-257.

7. Wizemann V, Wabel P, Chamney P, et al. The mortality risk of overhydration in haemodialysis patients. Nephrology Dialysis Transplantation 2009; 24: 1574-1579.


9. Mente A, O’Donnell MJ, Rangarajan S, et al. Association of urinary sodium and potassium excretion with blood pressure. N.Engl.J.Med. 2014; 371: 601-611.

10. Adrogue HJ, Madias NE. Sodium and potassium in the pathogenesis of hypertension. N.Engl.J.Med. 2007; 356: 1966-1978.

11. Sabatine MS, Morrow DA, de Lemos JA, et al. Evaluation of multiple biomarkers of cardiovascular stress for risk prediction and guiding medical therapy in patients with stable coronary disease. Circulation 2012; 125: 233-240.

12. Gruson D, Lepoutre T, Ketelslegers JM, Cumps J, Ahn SA, Rousseau MF. C-terminal FGF23 is a strong predic-tor of survival in systolic heart failure. Peptides 2012; 37: 258-262.

13. Seiler S, Cremers B, Rebling NM, et al. The phosphatonin fibroblast growth factor 23 links calcium-phosphate metabolism with left-ventricular dysfunction and atrial fibrillation. Eur.Heart J. 2011; 32: 2688-2696.



16. Iwasawa H, Nakao T, Matsumoto H, Okada T, Nagaoka Y, Wada T. Phosphate handling by end-stage kidneys and benefits of residual renal function on phosphate removal in patients on haemodialysis. Nephrology (Carlton) 2013; 18: 285-291.


18. Deo R, Katz R, de Boer IH, et al. Fibroblast Growth Factor 23 and Sudden Versus Non-sudden Cardiac Death: The Cardiovascular Health Study. American Journal of Kidney Diseases.

19. Arnlov J, Carlsson AC, Sundstrom J, et al. Serum FGF23 and risk of cardiovascular events in relation to mineral metabolism and cardiovascular pathology. Clin.J.Am.Soc.Nephrol. 2013; 8: 781-786.

203


11


21. Mathew JS, Sachs MC, Katz R, et al. Fibroblast growth factor-23 and incident atrial fibrillation: the Multi-Ethnic Study of Atherosclerosis (MESA) and the Cardiovascular Health Study (CHS). Circulation 2014; 130: 298-307.

22. Nowak A, Friedrich B, Artunc F, et al. Prognostic value and link to atrial fibrillation of soluble Klotho and FGF23 in hemodialysis patients. PLoS One 2014; 9: e100688.

23. Poelzl G, Trenkler C, Kliebhan J, et al. FGF23 is associated with disease severity and prognosis in chronic heart failure. Eur.J.Clin.Invest. 2014; 44: 1150-1158.

24. Isakova T, Houston J, Santacruz L, et al. Associations between fibroblast growth factor 23 and cardiac characteristics in pediatric heart failure. Pediatr.Nephrol. 2013; 28: 2035-2042.

25. Fuernau G, Poss J, Denks D, et al. Fibroblast growth factor 23 in acute myocardial infarction complicated by cardiogenic shock: a biomarker substudy of the Intraaortic Balloon Pump in Cardiogenic Shock II (IABP-SHOCK II) trial. Crit.Care 2014; 18: 713-014-0713-8.

26. Poss J, Mahfoud F, Seiler S, et al. FGF-23 is associated with increased disease severity and early mortality in cardiogenic shock. Eur.Heart J.Acute Cardiovasc.Care. 2013; 2: 211-218.

27. Prie D, Forand A, Francoz C, et al. Plasma fibroblast growth factor 23 concentration is increased and predicts mortality in patients on the liver-transplant waiting list. PLoS One 2013; 8: e66182.

28. Christov M, Waikar SS, Pereira RC, et al. Plasma FGF23 levels increase rapidly after acute kidney injury. Kidney Int. 2013; 84: 776-785.

29. Dounousi E, Torino C, Pizzini P, et al. Effect of inflammation by acute sepsis on intact fibroblast growth factor 23 (iFGF23) and asymmetric dimethyl arginine (ADMA) in CKD patients. Nutr.Metab.Cardiovasc.Dis. 2016; 26: 80-83.

30. Humalda D, Blanksma LJ, Jongebloed WL, Worst JG. Membrane formation and endothelial cell growth on an intra-ocular lens in the human eye, a SEM study. Doc.Ophthalmol. 1986; 61: 241-246.

31. Tang C, Pathare G, Michael D, Fajol A, Eichenmuller M, Lang F. Downregulation of Klotho expression by dehydration. Am.J.Physiol.Renal Physiol. 2011; 301: F745-50.

32. Seiler S, Wen M, Roth HJ, et al. Plasma Klotho is not related to kidney function and does not predict adverse outcome in patients with chronic kidney disease. Kidney Int. 2013; 83: 121-128.

33. Heijboer AC, Blankenstein MA, Hoenderop J, de Borst MH, Vervloet MG, NIGRAM consortium. Laboratory aspects of circulating alpha-Klotho. Nephrol.Dial.Transplant. 2013; 28: 2283-2287.

34. Gutierrez OM, Januzzi JL, Isakova T, et al. Fibroblast growth factor 23 and left ventricular hypertrophy in chronic kidney disease. Circulation 2009; 119: 2545-2552.

35. Negishi K, Kobayashi M, Ochiai I, et al. Association between fibroblast growth factor 23 and left ventricular hypertrophy in maintenance hemodialysis patients. Comparison with B-type natriuretic peptide and car-diac troponin T. Circ.J. 2010; 74: 2734-2740.


37. Wald R, Goldstein MB, Wald RM, et al. Correlates of left ventricular mass in chronic hemodialysis recipi-ents. Int.J.Cardiovasc.Imaging 2014; 30: 349-356.

38. Grabner A, Amaral A, Schramm K, et al. Activation of Cardiac Fibroblast Growth Factor Receptor 4 Causes Left Ventricular Hypertrophy. Cell Metabolism. 2015;22(6):1020-1032.

39. Slagman MC, Waanders F, Vogt L, et al. Elevated N-terminal pro-brain natriuretic peptide levels predict an enhanced anti-hypertensive and anti-proteinuric benefit of dietary sodium restriction and diuretics,

Chapter 11

204

but not angiotensin receptor blockade, in proteinuric renal patients. Nephrol.Dial.Transplant. 2012; 27: 983-990.


41. Yilmaz MI, Sonmez A, Saglam M, et al. Ramipril Lowers Plasma FGF-23 in Patients with Diabetic Nephropa-thy. Am.J.Nephrol. 2014; 40: 208-214.

42. Zanchi C, Locatelli M, Benigni A, et al. Renal expression of FGF23 in progressive renal disease of diabetes and the effect of ACE inhibitor. PLoS One 2013; 8: e70775.

43. Mirams M, Robinson BG, Mason RS, Nelson AE. Bone as a source of FGF23: regulation by phosphate? Bone 2004; 35: 1192-1199.

44. Andrukhova O, Slavic S, Odorfer KI, Erben RG. Experimental Myocardial Infarction Upregulates Circulating Fibroblast Growth Factor-23. J.Bone Miner.Res. 2015; 30: 1831-1839.

45. Andersen IA, Huntley BK, Sandberg SS, Heublein DM, Burnett JC. Elevation of circulating but not myocar-dial FGF23 in human acute decompensated heart failure. Nephrology Dialysis Transplantation 2015;31(5): 767-72.

46. Vervloet MG, Massy ZA, Brandenburg VM, et al. Bone: a new endocrine organ at the heart of chronic kidney disease and mineral and bone disorders. Lancet Diabetes Endocrinol. 2014; 2: 427-436.

47. Pathare G, Foller M, Michael D, et al. Enhanced FGF23 serum concentrations and phosphaturia in gene targeted mice expressing WNK-resistant SPAK. Kidney Blood Press.Res. 2012; 36: 355-364.

48. Hoorn EJ, Nelson JH, McCormick JA, Ellison DH. The WNK kinase network regulating sodium, potassium, and blood pressure. J.Am.Soc.Nephrol. 2011; 22: 605-614.

49. Moes AD, van der Lubbe N, Zietse R, Loffing J, Hoorn EJ. The sodium chloride cotransporter SLC12A3: new roles in sodium, potassium, and blood pressure regulation. Pflugers Arch. 2014; 466: 107-118.


51. McCormick JA, Ellison DH. The WNKs: atypical protein kinases with pleiotropic actions. Physiol.Rev. 2011; 91: 177-219.

52. Tuomilehto J, Jousilahti P, Rastenyte D, et al. Urinary sodium excretion and cardiovascular mortality in Finland: a prospective study. The Lancet 2001; 357: 848-851.


54. Beaglehole R, Bonita R, Horton R, et al. Priority actions for the non-communicable disease crisis. Lancet 2011; 377: 1438-1447.


56. Werkgroep Richtlijn Chronische Nierschade, Rensma PL, Hagen EC, et al. Richtlijn voor de behandeling van atiënten met Chronische Nierschade (CNS). 2009

57. Joosten MM, Gansevoort RT, Mukamal KJ, et al. Sodium excretion and risk of developing coronary heart disease. Circulation 2014; 129: 1121-1128.


59. Fan L, Tighiouart H, Levey AS, Beck GJ, Sarnak MJ. Urinary sodium excretion and kidney failure in nondia-betic chronic kidney disease. Kidney Int. 2014; 86: 582-588.

205


11


61. Sacks FM, Svetkey LP, Vollmer WM, et al. Effects on Blood Pressure of Reduced Dietary Sodium and the Dietary Approaches to Stop Hypertension (DASH) Diet. N.Engl.J.Med. 2001; 344: 3-10.


63. Mozaffarian D, Fahimi S, Singh GM, et al. Global sodium consumption and death from cardiovascular causes. N.Engl.J.Med. 2014; 371: 624-634.

64. Aburto NJ, Ziolkovska A, Hooper L, Elliott P, Cappuccio FP, Meerpohl JJ. Effect of lower sodium intake on health: systematic review and meta-analyses. BMJ 2013; 346: f1326.

65. Bibbins-Domingo K, Chertow GM, Coxson PG, et al. Projected effect of dietary salt reductions on future cardiovascular disease. N.Engl.J.Med. 2010; 362: 590-599.

66. Titze J, Lang R, Ilies C, et al. Osmotically inactive skin Na+ storage in rats. Am.J.Physiol.Renal Physiol. 2003; 285: F1108-17.

67. Machnik A, Neuhofer W, Jantsch J, et al. Macrophages regulate salt-dependent volume and blood pressure by a vascular endothelial growth factor-C-dependent buffering mechanism. Nat.Med. 2009; 15: 545-552.

68. Lerchl K, Rakova N, Dahlmann A, et al. Agreement Between 24-Hour Salt Ingestion and Sodium Excretion in a Controlled Environment. Hypertension 2015; 66: 850-857.

69. Kestenbaum B, Sampson JN, Rudser KD, et al. Serum phosphate levels and mortality risk among people with chronic kidney disease. J.Am.Soc.Nephrol. 2005; 16: 520-528.

70. Selamet U, Tighiouart H, Sarnak MJ, et al. Relationship of dietary phosphate intake with risk of end-stage renal disease and mortality in chronic kidney disease stages 3-5: The Modification of Diet in Renal Disease Study. Kidney Int. 2015; 89: 176-84.

71. Chang AR, Lazo M, Appel LJ, Gutierrez OM, Grams ME. High dietary phosphorus intake is associated with all-cause mortality: results from NHANES III. Am.J.Clin.Nutr. 2014; 99: 320-327.

72. Yamamoto KT, Robinson-Cohen C, de Oliveira MC, et al. Dietary phosphorus is associated with greater left ventricular mass. Kidney Int. 2013; 83: 707-714.

73. Palomino HL, Rifkin DE, Anderson C, Criqui MH, Whooley MA, Ix JH. 24-Hour Urine Phosphorus Excretion and Mortality and Cardiovascular Events. Clinical Journal of the American Society of Nephrology 2013; 8: 1202-1210.

74. Murtaugh MA, Filipowicz R, Baird BC, Wei G, Greene T, Beddhu S. Dietary phosphorus intake and mortality in moderate chronic kidney disease: NHANES III. Nephrology Dialysis Transplantation 2012; 27: 990-996.

75. de Boer IH, Rue TC, Kestenbaum B. Serum Phosphorus Concentrations in the Third National Health and Nutrition Examination Survey (NHANES III). American Journal of Kidney Diseases 2009; 53: 399-407.

76. Gutierrez OM. The connection between dietary phosphorus, cardiovascular disease, and mortality: where we stand and what we need to know. Adv.Nutr. 2013; 4: 723-729.

77. Dominguez JR, Kestenbaum B, Chonchol M, et al. Relationships Between Serum and Urine Phosphorus With All-Cause and Cardiovascular Mortality: The Osteoporotic Fractures in Men (MrOS) Study. American Journal of Kidney Diseases 2013; 61: 555-563.

78. Berndt T, Thomas LF, Craig TA, et al. Evidence for a signaling axis by which intestinal phosphate rapidly modulates renal phosphate reabsorption. Proc.Natl.Acad.Sci.U.S.A. 2007; 104: 11085-11090.

79. Gonzalez-Parra E, Gracia-Iguacel C, Egido J, Ortiz A. Phosphorus and nutrition in chronic kidney disease. Int.J.Nephrol. 2012; 2012: 597605.

80. Sprague SM, Abboud H, Qiu P, Dauphin M, Zhang P, Finn W. Lanthanum Carbonate Reduces Phosphorus Burden in Patients with CKD Stages 3 and 4: A Randomized Trial. Clinical Journal of the American Society of Nephrology 2009; 4: 178-185.

Chapter 11

206

81. Kalantar-Zadeh K, Gutekunst L, Mehrotra R, et al. Understanding sources of dietary phosphorus in the treatment of patients with chronic kidney disease. Clin.J.Am.Soc.Nephrol. 2010; 5: 519-530.

82. Lampila LE. Applications and functions of food-grade phosphates. Ann.N.Y.Acad.Sci. 2013; 1301: 37-44. 83. Cupisti A, Benini O, Ferretti V, Gianfaldoni D, Kalantar-Zadeh K. Novel differential measurement of natural

and added phosphorus in cooked ham with or without preservatives. J.Ren.Nutr. 2012; 22: 533-540. 84. Moe SM, Zidehsarai MP, Chambers MA, et al. Vegetarian compared with meat dietary protein source and

phosphorus homeostasis in chronic kidney disease. Clin.J.Am.Soc.Nephrol. 2011; 6: 257-264. 85. Sullivan C, Sayre SS, Leon JB, et al. Effect of food additives on hyperphosphatemia among patients with

end-stage renal disease: a randomized controlled trial. JAMA 2009; 301: 629-635. 86. Carrigan A, Klinger A, Choquette SS, et al. Contribution of food additives to sodium and phosphorus

content of diets rich in processed foods. J.Ren.Nutr. 2014; 24: 13-9, 19e1. 87. Gutierrez OM, Luzuriaga-McPherson A, Lin Y, Gilbert LC, Ha SW, Beck GR,Jr. Impact of phosphorus-based

food additives on bone and mineral metabolism. J.Clin.Endocrinol.Metab. 2015; jc20152279. 88. Trautvetter U, Jahreis G, Kiehntopf M, Glei M. Consequences of a high phosphorus intake on mineral

metabolism and bone remodeling in dependence of calcium intake in healthy subjects - a randomized placebo-controlled human intervention study. Nutr.J. 2016; 15: 7-016-0125-5.

89. Ferrari SL, Bonjour JP, Rizzoli R. Fibroblast growth factor-23 relationship to dietary phosphate and renal phosphate handling in healthy young men. J.Clin.Endocrinol.Metab. 2005; 90: 1519-1524.

90. Burnett SM, Gunawardene SC, Bringhurst FR, Juppner H, Lee H, Finkelstein JS. Regulation of C-terminal and intact FGF-23 by dietary phosphate in men and women. J.Bone Miner.Res. 2006; 21: 1187-1196.

91. Vervloet MG, van Ittersum FJ, Buttler RM, Heijboer AC, Blankenstein MA, ter Wee PM. Effects of dietary phosphate and calcium intake on fibroblast growth factor-23. Clin.J.Am.Soc.Nephrol. 2011; 6: 383-389.

92. Parfitt AM. Misconceptions IV--the hypophosphatemia of primary hyperparathyroidism is the result of renal phosphate wasting. Bone 2004; 35: 345-347.

93. Jaeger P, Bonjour JP, Karlmark B, et al. Influence of acute potassium loading on renal phosphate transport in the rat kidney. Am.J.Physiol. 1983; 245: F601-5.


95. Breusegem SY, Takahashi H, Giral-Arnal H, et al. Differential regulation of the renal sodium-phosphate cotransporters NaPi-IIa, NaPi-IIc, and PiT-2 in dietary potassium deficiency. Am.J.Physiol.Renal Physiol. 2009; 297: F350-61.

96. Smyth A, Dunkler D, Gao P, et al. The relationship between estimated sodium and potassium excretion and subsequent renal outcomes. Kidney Int. 2014; 86: 1205-12.

97. Aburto NJ, Hanson S, Gutierrez H, Hooper L, Elliott P, Cappuccio FP. Effect of increased potassium intake on cardiovascular risk factors and disease: systematic review and meta-analyses. BMJ 2013; 346: f1378.

98. Rubinacci A, Benelli FD, Borgo E, Villa I. Bone as an ion exchange system: evidence for a pump-leak mecha-nism devoted to the maintenance of high bone K(+). Am.J.Physiol.Endocrinol.Metab. 2000; 278: E15-24.

99. Sorensen MV, Grossmann S, Roesinger M, et al. Rapid dephosphorylation of the renal sodium chloride cotransporter in response to oral potassium intake in mice. Kidney Int. 2013; 83: 811-824.

100. Rengarajan S, Lee DH, Oh YT, Delpire E, Youn JH, McDonough AA. Increasing plasma [K+] by intravenous potassium infusion reduces NCC phosphorylation and drives kaliuresis and natriuresis. Am.J.Physiol.Renal Physiol. 2014; 306: F1059-68.

101. Dvorak MM, De Joussineau C, Carter DH, et al. Thiazide Diuretics Directly Induce Osteoblast Differentia-tion and Mineralized Nodule Formation by Interacting with a Sodium Chloride Co-Transporter in Bone. Journal of the American Society of Nephrology 2007; 18: 2509-2516.

207


11

102. van der Lubbe N, Moes AD, Rosenbaek LL, et al. K+-induced natriuresis is preserved during Na+ deple-tion and accompanied by inhibition of the Na+-Cl- cotransporter. Am.J.Physiol.Renal Physiol. 2013; 305: F1177-88.

103. Mendes AI, Mascarenhas MR, Matos S, et al. A WNK4 gene variant relates to osteoporosis and not to hypertension in the Portuguese population. Mol.Genet.Metab. 2011; 102: 465-469.

104. Andrukhova O, Slavic S, Zeitz U, Shalhoub V, Lanske B, Erben RG. FGF23 Regulates Renal Sodium Handling and Blood Pressure [Abstract]. J.Am.Soc.Nephrol. 2012; 23: 46A.

105. Hanudel MR, Wesseling-Perry K, Gales B, et al. Effects of acute kidney injury and chronic hypoxemia on fibroblast growth factor 23 levels in pediatric cardiac surgery patients. Pediatr.Nephrol. 2015; 31: 661-9.



108. Parving HH, Persson F, Lewis JB, Lewis EJ, Hollenberg NK, AVOID Study Investigators. Aliskiren combined with losartan in type 2 diabetes and nephropathy. N.Engl.J.Med. 2008; 358: 2433-2446.

109. Van Buren PN, Adams-Huet B, Nguyen M, Molina C, Toto RD. Potassium handling with dual renin-angio-tensin system inhibition in diabetic nephropathy. Clin.J.Am.Soc.Nephrol. 2014; 9: 295-301.

110. Mann JF, Schmieder RE, Dyal L, et al. Effect of telmisartan on renal outcomes: a randomized trial. Ann.Intern.Med. 2009; 151: 1-10, W1-2.

111. Kieneker LM, Gansevoort RT, Mukamal KJ, et al. Urinary potassium excretion and risk of developing hyper-tension: the prevention of renal and vascular end-stage disease study. Hypertension 2014; 64: 769-776.

112. Cook NR, Obarzanek E, Cutler JA, et al. Joint effects of sodium and potassium intake on subsequent car-diovascular disease: the Trials of Hypertension Prevention follow-up study. Arch.Intern.Med. 2009; 169: 32-40.

113. Fardet A, Rock E. The healthy core metabolism: A new paradigm for primary preventive nutrition. J.Nutr.Health Aging 2015; 1-9.

114. Gezondheidsraad. Richtlijnen goede voeding 2015. 2015; publicatienr. 2015/24. 115. Muntner P, Judd SE, Krousel-Wood M, McClellan WM, Safford MM. Low Medication Adherence and Hy-

pertension Control Among Adults With CKD: Data From the REGARDS (Reasons for Geographic and Racial Differences in Stroke) Study. American Journal of Kidney Diseases 2010; 56: 447-457.

116. Santschi V, Chiolero A, Colosimo AL, et al. Improving Blood Pressure Control Through Pharmacist Interven-tions: A Meta-Analysis of Randomized Controlled Trials. Journal of the American Heart Association 2014; 3:: e000718.

117. Hopkins RB, Garg AX, Levin A, et al. Cost-effectiveness analysis of a randomized trial comparing care models for chronic kidney disease. Clin.J.Am.Soc.Nephrol. 2011; 6: 1248-1257.

118. Di Lullo L, Gorini A, Bellasi A, et al. Fibroblast growth factor 23 and parathyroid hormone predict extent of aortic valve calcifications in patients with mild to moderate chronic kidney disease. Clin.Kidney J. 2015; 8: 732-736.

119. van Binsbergen JJ, Delaney BC, van Weel C. Nutrition in primary care: scope and relevance of output from the Cochrane Collaboration. The American Journal of Clinical Nutrition 2003; 77: 1083S-1088S.

120. Kerstens MN, Kobold ACM, Volmer M, Koerts J, Sluiter WJ, Dullaart RPF. Reference Values for Aldosterone–Renin Ratios in Normotensive Individuals and Effect of Changes in Dietary Sodium Consumption. Clinical Chemistry 2011; 57: 1607-1611.

121. Kamphuis CBM, Turrell G, Giskes K, Mackenbach JP, van Lenthe FJ. Life course socioeconomic conditions, adulthood risk factors and cardiovascular mortality among men and women: A 17-year follow up of the GLOBE study. Int.J.Cardiol. 2013; 168: 2207-2213.

209

Samenvatting

11

Samenvatting. Met uitleg voor niet-ingewijden

Dit proefschrift behandelt de wisselwerking tussen enerzijds het hormoon fibroblast groeifactor 23 (FGF-23) en de andere elementen uit de botmineraalhuishouding, en anderzijds zouten vo-lumehuishouding. De nieren hebben een centrale rol in deze regelsystemen. In het eerste deel van dit proefschrift beschrijven we de rol van FGF-23 in relatie tot verschillende uitkomstmaten die wijzen op volumebelasting (het hebben van teveel vocht in het lichaam, overvulling). In het tweede deel onderzoeken we het effect van dieetmaatregelen op fosfaat en FGF-23, om te besluiten met een gerandomiseerd, discipline-overstijgend onderzoek dat gebruikt kan worden om een duurzame dieetverandering gericht op elementen van gezonde voeding te kunnen bewerkstelligen.

De nieren zijn van cruciaal belang voor het menselijk lichaam. Zij zuiveren het bloed van afval-producten, produceren het hormoon erytropoïetine (EPO) dat bloedaanmaak stimuleert, en activeren vitamine D dat nodig is voor het behouden van calcium en fosfaat. Elke nier bestaat uit vele kleine filtersystemen, ‘nefronen’. Het nefron begint met een zeef, de glomerulus, waarlangs het bloed stroomt onder druk. Het filtraat dat zo uit het bloed wordt geperst is de voorurine, de grotere stoffen als eiwitten en bloedcellen blijven in het bloed. Deze voorurine, het ultrafiltraat, stroomt door een buis, de tubulus. In de wand van deze buis zitten allerlei pompen die zouten en mineralen uit de voorurine weer terughalen, zodat het lichaam deze weer kan gebruiken. Andere pompen geven juist afvalstoffen af aan de voorurine. Op deze manier bewaken de nieren de balans in de hoeveelheid water, zouten en andere mineralen, en zuren in het men-selijk lichaam. In patiënten met chronische nierschade functioneren de nefronen niet goed: de zeven (glomeruli) kunnen beschadigd zijn, de pompen kunnen niet voldoende werken of hele nefronen zijn verloren gegaan. Beschadigde glomeruli lekken eiwit, dat dus in urine terecht komt: proteïnurie. Door verlittekening zitten andere glomeruli juist helemaal dicht. Hierdoor vermindert de glomerulaire filtratie snelheid (GFR) van de nieren: de GFR is dan ook een maat voor nierfunctie. Als er langer dan drie maanden sprake is van eiwitverlies in de urine en/of verlies van filtratiefunctie, spreekt men van chronische nierschade.

De hoeveelheid zout in het lichaam, en daarmee de hoeveelheid vocht en dus bloeddruk, wordt onder andere geregeld door het renine-angiotensine-aldosteron systeem (RAAS). De hormonen uit het RAAS stimuleren de opname van zout in de nier vanuit de voorurine en zorgen voor het vernauwen van de bloedvaten. Hierdoor neemt zowel de bloeddruk toe in het lichaam als de filtratiedruk in de nier. Bij nierziekten is het RAAS overmatig geactiveerd, wat kan leiden tot nog meer nierschade. RAAS-remmende medicijnen verlagen de bloeddruk en het eiwitverlies in de urine, en beschermen zo de nieren tegen verdere beschadiging. De RAAS-remmers verlagen echter zelden het eiwitverlies tot nul. Hoe meer eiwitverlies er overblijft, ‘residuale proteïnurie’

Chapter 11

210

of restproteïnurie, hoe groter het risico op hart- en vaatziekten. Recente onderzoeken tonen aan dat bepaalde onderdelen uit het dieet zoals vitamine D, natrium (natrium vormt samen met chloride [keuken]zout) en fosfaat de effectiviteit van RAAS-remmers beïnvloeden.

In hoofdstuk 2 beschrijven we de toegevoegde waarde van vitamine D aan RAAS-remmers voor wat betreft het verminderen van eiwitverlies. We benoemen dat het combineren van verschil-lende RAAS-remmers mogelijk gevaarlijk is, omdat het overmatig beperken van de filtratiedruk en natriumopname uit de urine een potentieel gevaarlijke stijging van het kalium tot gevolg kan hebben. Verder beschrijven we dat vitamine D aan de ene kant door invloed op het RAAS en aan de andere kant ook onafhankelijk van het RAAS de nier mogelijk kan beschermen. Het toedie-nen van extra vitamine D lijkt het risico op hart- en vaatziekten voor patiënten met chronische nierschade niet te verminderen, maar prospectief onderzoek zal hierover uitsluitsel moeten bieden. Dit alles overwegende, hebben we minder hoge verwachtingen van vitamine D alleen als aanvulling op RAAS-remming voor langdurige bescherming van de nieren. Er moet dus verder onderzoek gedaan worden naar het optimaliseren van nierbescherming met RAAS-remming.

In hoofdstuk 3 bieden we een overzicht van de rol van zoutinname in patiënten met chroni-sche nierschade. In de algemene bevolking en in patiënten met chronische nierschade is de zoutinname aan de hoge kant. Dat laatste is een probleem, omdat een hoge zoutinname de effecten van RAAS-remmers teniet doet. Zelfs een bescheiden vermindering van de zoutinname verlaagt de bloeddruk en het eiwitverlies. Patiënten met chronische nierschade houden zout en daarmee vocht extra makkelijk vast. Wij stellen dat de mate van vocht vasthouden kan helpen om in te schatten welke patiënten het meest gebaat zijn bij een zoutbeperking. Naast een hoge zoutinname is ook een hoge fosfaatwaarde in het bloed geassocieerd met een verminderd ef-fect van RAAS-remmers op eiwitverlies.

In hoofdstuk 4 beschrijven we dat hoe hoger de concentratie van het fosfaatregulerende hor-moon FGF-23 is, hoe hoger de restproteïnurie onder het combineren van RAAS-remmers met een zoutbeperking. FGF-23 correleerde daarnaast ook sterk met het NT-proBNP. Dit hormoon is een maat voor volumebelasting. Daarnaast ging een hoger FGF-23 ook gepaard met een hogere aldosteron/renine-waarde in het bloed. Dit gaf aan dat de RAAS-remming minder succesvol was. FGF-23 gaat dus gepaard met volumebelasting en met een verminderde effectiviteit van de belangrijkste behandelopties: namelijk RAAS-remmers en een zoutbeperking.

In hoofdstuk 5 onderzoeken we of ook in niertransplantiepatiënten hoge FGF-23 waarden gere-lateerd zijn aan verminderde bescherming. FGF-23 concentraties nemen explosief toe naarmate de nierfunctie verslechtert. Patiënten die afhankelijk zijn van dialyse (nierspoeling), hebben FGF-23 waarden die tot 2000 maal hoger zijn dan in gezonde personen. Niertransplantatie verlaagt de FGF-23 waarden tot bijna normale waarden. Maar niertransplantatiepatiënten die

211

Samenvatting

11

een relatief hoger FGF-23 hebben, blijken een groter risico te hebben op hart- en vaatziekten. FGF-23 verklaart dit hogere risico, ook wanneer we rekening houden met andere, al langer bekende risicofactoren voor hart- en vaatziekten zoals bloeddruk, roken en cholesterol. Verder blijkt FGF-23 sterk gerelateerd aan de hormonen NT-proBNP, MR-proANP en copeptin. Je kan deze drie hormonen zien als een maat voor hoeveel (bloed)volume het hart moet rondpompen.

Patiënten met eindstadium nierfalen zijn volledig afhankelijk van dialyse om van het overschot aan vocht af te komen. In hoofdstuk 6 is vastgesteld dat hemodialysepatiënten die tussen de dialyse sessies meer vocht vasthouden, en aan wie dus tijdens de dialyse meer vocht moet worden onttrokken, diegene zijn met de meest extreme FGF-23 concentraties. Deze hoeveel-heid vocht, het ultrafiltratie volume, is daarmee een nieuwe voorspeller van FGF-23 waardes in hemodialysepatiënten. Dit verband was onafhankelijk van bekende voorspellers van FGF-23 zoals fosfaat en calcium. Het ultrafiltratie volume verklaart dus deels de grote verschillen tussen de patiënten onderling in extreme FGF-23 waarden. Ook hier vinden we dat bloed- en echogra-fische waarden die iets zeggen over volumestatus ook gerelateerd zijn met FGF-23: namelijk de copeptin bloedwaarde, het hartminuutvolume (hoeveel bloed het hart per minuut rondpompt), en het eind-diastolisch volume (hoeveel het hart gevuld is vóór elke hartslag). Patiënten bij wie het bloedvolume het meest verlaagd wordt, hebben ook een verlaging van FGF-23 waarden. Er is dus sprake van een sterke wisselwerking tussen botmineraalfactoren als FGF-23, de volume status en het RAAS. Vervolgens onderzoeken we in hoofdstuk 7 of het toedienen of beperken van vocht FGF-23 concentraties kan beïnvloeden. Dit is niet het geval: het toedienen van vocht aan patiënten met een hoge bloeddruk heeft geen effect op FGF-23. Ook een zoutbeperking (en daarmee het verminderen van het vocht in het lichaam) heeft geen effect op FGF-23 in patiënten met nierschade door suikerziekte. Interventies op zout beïnvloeden FGF-23 kennelijk dus niet, terwijl interventies op fosfaat wél FGF-23 beïnvloeden. Het is daarom nuttig om de relatie tussen zout en fosfaat in de voeding te onderzoeken.

In hoofdstuk 8 blijkt dat zouten fosfaatinname sterk aan elkaar gerelateerd zijn in zowel gezonde proefpersonen, in patiënten met suikerziekte en in niertransplantatiepatiënten. We zien ook dat een zoutbeperking in patiënten met nierschade ook een kleine verlaging van de fosfaatinname tot gevolg heeft. Wij denken dat dit komt doordat producten die veel zout bevat-ten, ook veel fosfaat bevatten. Kant-en-klare producten bevatten naast zout ook verschillende smaak- en conserveringsstoffen die gebaseerd zijn op fosfaat. Dit verklaart ook waarom een dieet dat veel van deze toevoegingen bevat, het FGF-23 verhoogt.

Een gezond dieet bevat weinig zout en fosfaat maar vaak véél kalium. Mensen met een hogere kaliuminname hebben een lagere bloeddruk. Uit experimenten met het geven van extra kalium blijkt dat dit het vasthouden van fosfaat stimuleert. Dit impliceert dat kalium dus de fosfaat-huishouding beïnvloedt. In hoofdstuk 9 veronderstellen we een rol voor kalium in de regulatie

Chapter 11

212

van FGF-23. Kaliumsupplementen verlagen FGF-23 concentraties in proefpersonen met een hoog-normale bloeddruk. Dit gaat gepaard met een verhoogde neiging tot het vasthouden van fosfaat, en daarmee een lichte verhoging van het fosfaat in het bloed. Andere hormonen die betrokken zijn bij de fosfaathuishouding zoals vitamine D en het bijschildklierhormoon PTH veranderen niet.

Voor veel patiënten is het niet eenvoudig om hun dieet te veranderen. Gedragsverandering is immers lastig. In hoofdstuk 10 testten we een nieuwe strategie hiervoor in het SUBLIEM-onderzoek. Dit onderzoek is een gerandomiseerd onderzoek, waarin patiënten loten voor interventie of controlegroep, uitgevoerd in verschillende centra. De SUBLIEM-interventie is gebaseerd op de zelfregulatie theorie, een theorie over gedragsverandering. SUBLIEM heeft als doel de zoutinname te verminderen. Hiertoe maakt SUBLIEM gebruik van een interactieve website, die de verschillende stadia van gedragsverandering doorloopt, groepsbijeenkomsten en individuele begeleiding via telefoon of email. In de SUBLIEM-interventiegroep is na 3 maanden de zoutinname en bloeddruk verlaagd. Maar na de volhoudfase van 6 maanden laat ook de controlegroep een verlaging van zoutinname zien. We denken dat dit komt doordat de deelnemers in de controlegroep ook meer bewust zijn van hun zoutinname, maar dit moet nog verder uitgezocht worden. In elk geval werd de SUBLIEM-studie erg gewaardeerd door de deelnemers. De tips en adviezen die wij van hen hebben gekregen, zullen leidend zijn voor de ontwikkeling van nieuwe E-health toepassingen die met gedragsverandering proberen een gezonde voeding te stimuleren.

213

Zusammenfassung

11

Zusammenfassung.

Im Rahmen dieser Thesis wird die Interaktion zwischen dem Knochen-Mineral Metabolismus, insbesondere mit Bezug auf Fibroblast Wachsfaktor 23 (FGF-23), und dem Volumenhaushalt behandelt, mit Fokussierung auf eine medikamentöse Hemmung des Renin-Angiotensin-Aldosteron-Systems (RAAS), welche die Grundlage der Behandlung von chronischem Nieren-versagen darstellt. Im 1. Teil der Thesis erforschten wir die Rolle von FGF-23 in Beziehung zu Parametern des Volumenhaushaltes. Im 2. Teil der Thesis haben wir den Effekt verschiedener Nahrungsinterventionen auf Phosphat und FGF-23 analysiert. Zum Schluss beschreiben wir eine prospektive, randomisierte kontrollierte Studie unter Nutzung eines multidisziplinären Ansatzes, damit langfristige Änderungen in Elementen einer gesunden Nahrung erreicht werden.

Eine RAAS Hemmung reduziert den Blutdruck als auch Proteinurie, so dass die Nieren vor einer Fibrose und konsekutiv vor dem Verlust der Nierenfunktion geschützt werden. Trotzdem erreicht eine RAAS Hemmung keine völlige Reduktion der Proteinurie. Je höher die Proteinurie, desto höher ist das Risiko für ein Herz- und Nierenversagen. Neue Ergebnisse identifizieren manche Ernährungfaktoren (z.B. Vitamin D, Natrium und Phosphat), die die Effektivität der RAAS Hemmung auf Proteinurie beeinflussen.

In Kapitel 2 geben wir eine Übersicht über die Möglichkeiten von Vitamin D, um eine Proteinu-rie weiter zu reduzieren. Wir benennen Sicherheitsbedenken für unbeschränkte Kombinationen von RAAS Hemmern, und dass Vitamin D durch direkte Hemmung des RAASs als auch durch eine RAAS-unabhängige Reduktion der Fibrose die Nierenfunktion schützt. Vitamin D Supplemente verringern nicht das erhöhten Risiko für Herzversagen für Patienten mit Nierenversagen, obwohl die letzten Ergebnisse von prospektiven Studies noch nicht da sind. Diese Übersichtsarbeit be-stätigt, dass eine Reduktion einer Proteinurie immer noch ein Behandlungsziel ist, identifiziert eine Wechselwirkung zwischen Knochen-Mineralhaushalt und RAAS sowie eine mögliche Wech-selwirkung zwischen Natriumaufnahme und Vitamin D, aber senkt auch die Erwartungen einer Vitamin D Supplementation zur Verbesserung der RAAS-Hemmung-basierten Renoprotektion.

Kapitel 3 gibt eine Übersicht über den Stellenwert der Natriumaufnahme bei Patienten mit chronischer Nierenerkrankung. Die Natriumeinnahme ist in der Allgemeinbevölkerung und bei chronisch Nierenkranken erhöht. Dies ist insofern problematisch, dass eine höhere Natrium-einnahme die Effektivität von RAAS Hemmern beeinträchtigt. Wir postulieren, dass schon eine mäßige Reduktion der Natriumaufnahme eine positive Wirkung auf den Blutdruck und eine Proteinurie hat. Wir betonen zudem, dass Patienten mit chronischer Nierenerkrankung für eine Natriumretention äußerlich empfindlich sind, und wir schlagen vor, dass Parameter des Volu-menhaushaltes Patienten identifizieren, die am meisten von einer bereits mäßigen Reduktion der Natriumaufnahme profitieren könnten. Neben einer höhen Natriumeinnahme, sind auch

Chapter 11

214

erhöhte Serum Phosphatwerte mit einer verringerten Effektivität der RAAS Hemmung auf die Proteinurie assoziiert.

Wir beschreiben in Kapitel 4, dass erhöhte Konzentrationen des Phosphatregulierenden Hor-mons FGF-23 unabhängig mit erhöhten Proteinurie assoziiert sind, trotz Natriumbeschränkung kombiniert mit RAAS Hemmung. FGF-23 korrelierte darüberhinaus mit N-terminalem pronatri-uretischem Peptid B (NT-proBNP), als Korrelat für eine Überwässerung. FGF-23 korrelierte auch mit einem erhöhten Aldosteron-Renin-Quotient, ein Maß für eine weniger erfolgreiche RAAS Hemmung. FGF-23 korrelierte somit mit einem mehr Volumen-belasteten Phänotypus und mit einer Einschränkung der Effektivität von medikamentösen und diätetischen Behandlungsmög-lichkeiten.

In Kapitel 5 erweitern wir die Ergebnisse durch Untersuchung von Patienten nach Nieren-transplantation. Obwohl die FGF-23 Konzentration nach Nierentransplantation sinkt, haben Patienten mit einer relativ erhöhten FGF-23 Konzentration ein erhöhtes Risiko auf eine erhöhte kardiovaskuläre Mortalität, so dass FGF-23 einen von klassischen Risikofaktoren unabhängigen kardiovaskulären Risikofaktor darstellt. FGF-23 korrelierte mit NT-proBNP, mid-regional pronat-riuretische Peptid A (MR-proANP), und dem Prohormon von Vasopressin, Copeptin. Diese drei Hormone kann man als Merkmale für eine erhöhte Volumenbelastung des Herzens betrachten.

Patienten mit endgültigem Verlust der Nierenfunktion benötigen eine Nierenersatztherapie, um die Volumenbelastung zu reduzieren. FGF-23 erreicht extreme Werte insbesondere bei Hämo-dialysepatienten. Wir stellen in Kapitel 6 dar, dass die Patienten, die höheren Volumenentzug brauchen, auch die Patienten mit den höchsten FGF-23 Werten sind. Anscheinend stellt das Ultrafiltrationsvolumen einen Determinant von FGF-23 dar, unabhängig von den bekannten Determinanten wie beispielsweise Calcium und Phosphat. Wir fanden auch, dass echografische als auch biochemische Merkmale des Volumenhaushaltes mit FGF-23 korrelierten: nämlich Herzminutenvolumen, Copeptin und das enddiastolische Volumen. Je höher die Reduktion des relativen Blutvolumens während der Dialysebehandlung, desto höher ist auch die Reduktion von FGF-23. Die vorgenannten Befunde unterstützen eine starke Wechselwirkung zwischen Knochenmineralmetabolismus, Volumenstatus und Effektivität der RAAS Hemmung.

Demzufolge untersuchten wir in Kapitel 7, ob Volumenbelastung oder Volumenbeschränkung zu Änderungen der FGF-23 Werte führen könnten. Es stellte sich heraus, dass Kochsalzinfusio-nen die FGF-23 Werte nicht bei Patienten mit arterieller Hypertonie nicht veränderten. Darüber hinaus änderte sechs Wochen diätetischer Natriumbeschränkung auch nicht die Höhe der FGF-23 Werte bei Patienten mit diabetischer Nephropathie.

215

Zusammenfassung

11

Da Modifikationen der Natriumzufuhr FGF-23 nicht beeinflussen, im Gegensatz zu Änderungen der Phosphataufnahme, ist es wichtig die Relation zwischen diesen zwei Komponenten zu explo-rieren. Wir berichteten, dass die Natrium- Phosphat-Aufnahme in verschiedenen Populationen stark miteinander korrelieren (gesunde Probanden, Patienten mit Diabetes mellitus und Patien-ten mit Z. n. Nierentransplantation) (Kapitel 8). Passend trat bei bereits geringer Reduzierung der Natriumeinnahme auch eine Reduzierung der Phosphataeinnahme auf. Teilweise führen wir diese Beobachtungen auf phosphatreiche Nahrungszusätzen zurück, das ebenso erklärt, warum eine phosphatenreiche Diät FGF-23 Spiegel erhöht. Somit sollte eine gesunde Diät wenig Natri-um und Phosphat enthalten. Daneben sollte sie kaliumreich sein, da eine hohe Kaliumaufnahme mit niedrigem Blutdruck assoziiert ist. Kaliumsupplemente erhöhen die Phosphatreabsorption in den Nieren, so dass Kalium anscheinend den Phosphathaushalt beeinflusst.

In Kapitel 9 darstellen wir die Hypothese, dass Kalium in die Regulation von FGF-23 eingreift. Kaliumsupplemente reduzierten den FGF-23 Spiegel in prähypertensiven Patienten, in Be-gleitung einer Zunahme der tubulären Phosphatreabsorption und der serumphosphatwerte, ohne Effekte auf andere Knochenstoffwechselh Hormone wie z.B. Parathormon und Vitamin D. Dieser Befund bietet die Grundlage zukünftiger Forschungsprojekte, die den Einfluss der Kali-umaufnahme auf FGF-23 und andere Parameter des Knochenmineralhaushaltes untersuchen. Trotzdem ist eine langfristige Änderung des Ernährungsverhaltens sehr schwer beizubehalten.

Als Versuch dieses Problem zu lösen, beschreiben wir in Kapitel 10 die ersten Ergebnissen der randomisierten, kontrollierten, multizentrischen Studie SUBLIME. SUBLIME basiert auf der behavioristischen Selbstregulationstheorie, und hat als Ziel die diätetische Natriumaufnahme zu reduzieren. Dazu setzte SUBLIME ein E-healthmodul ein, welches die verschiedenen Phasen der Verhaltungsänderungen adressierte, unter anderem mit Gruppentreffen und individuel-lem E-coaching. SUBLIME reduzierte die Natriumausscheidung und den Blutdruck nach drei Monaten Interventionsphase in der Interventionsgruppe, aber nach dem sechs Monaten Auf-rechterhaltungsstadium reduzierte sich die Natriumausscheidung auch in der Kontrollegruppe. Wir schreiben das, der erhöhten Aufmerksamkeit der Studienteilnehmer zu, die Gegenstand zukünftiger Untersuchungen sein soll. Die SUBLIME Intervention wurde von den Studienteilneh-mern sehr geschätzt, deren Feedback für die weitere Entwicklung von E-health Interventionen grundlegend ist, und die Verhaltungsänderungen im diätetischen Lifestyle zum Ziel haben.

Dankwoord

219

D

DankwoordPromoveren doe je nooit alleen. Met het einde in zicht is het moment daar om iedereen te bedanken die heeft bijgedragen aan de totstandkoming van dit proefschrift. Het is een uitdaging om hierin volledig te zijn. Ik wil bij deze dan ook allen bedanken die deel uitmaken van het complexe, dynamische systeem waaruit het leven als arts-onderzoeker bestaat.

Allereerst dank ik alle deelnemers van de diverse onderzoeken. Zonder jullie deelname aan deze dikwijls zeer intensieve studies is wetenschappelijk onderzoek onmogelijk.

Prof. dr. G.J. Navis, beste Gerjan, inmiddels werken we bijna vijf jaar samen. Van begin af aan heb je mij gestimuleerd mijzelf te ontwikkelen en te verbreden, zowel binnen het onderzoek als daarbuiten. Het geleerde bleek nodig: voor de coördinatie en voltooiing van de multicenter studie SUBLIEM moesten we alle zeilen bijzetten. Bedankt dat ik altijd op je kon rekenen wan-neer een storm van deadlines zich aandiende. Daarnaast benadrukte je altijd het belang van het uitdragen van wetenschappelijke kennis naar een breed publiek, bijvoorbeeld op Noorderzon. Maar ook de les ‘dat het betere de vijand is van het goede’ blijkt zeer vaak te gelden. Bedankt voor alle kansen die je mij geboden hebt!

Dr. M.H. de Borst, beste Martin, wat was het een mooi traject! Jij verenigt alle eigenschappen van een uitstekende copromotor: betrouwbaar, benaderbaar, een heldere onderzoekslijn, snelle feedback, gevoel voor humor en steun wanneer nodig. Inspirerend hoe jij als arts, onderzoeker en familieman het leven weet te combineren. Onze brainstormsessies hebben mij van begin af aan enthousiast gemaakt voor onderzoek. Het is fijn dat we die sessies gedurende dit traject behouden hebben. Deze palavers hebben inmiddels een vloot aan ideeën en acroniemen voor de komende jaren opgeleverd. Hopelijk kunnen we samen een koers uitzetten naar nieuwe gebieden.

De leden van mijn beoordelingscommissie wil ik bedanken voor het beoordelen van dit proefschrift:

Prof. dr. R. Sanderman. Dank dat u de tijd heeft willen nemen om mijn proefschrift te beoorde-len, hopelijk versterkt dit de verbinding tussen de medische en gedragswetenschappen.

Prof. dr. P.M. Ter Wee, beste Piet. Ik heb je leren kennen als de gedreven voorzitter van het NIGRAM-consortium. Deels tijdens de inspirerende bijeenkomsten in de verschillen gewesten van het land, maar ook tijdens diverse buitenlandse congressen. Hartelijk dank voor het beoor-delen van dit proefschrift.

Prof. dr. D.J.A. Goldsmith, dear David. It was inspiring to have been given the opportunity to collaborate with such an esteemed leader in the field. Your vast knowledge combined with

220

Dankwoord

your friendliness and your willingness to cooperate with and inspire the young is what science ought to be all about. Our collaboration enriched my vocabulary tremendously. I can only hope that one does not become wamblecropt from the ubiquitously present puns in this thesis. I am grateful and honored that you took place in the Reading Committee.

Drs. M.R.M. San Giorgi, beste Mies. We kennen elkaar sinds het begin van de studie en zijn in de loop van jaren naar elkaar toegegroeid. Je staat altijd voor mij klaar, zowel voor het vieren van successen als het omgaan met tegenslag. Op de racefiets of hardlopend hielden we elkaar scherp tijdens onze promotietrajecten. Promoveren is immers te vergelijken met een duurloop: je moet op tijd wat drinken en energie over houden om uiteindelijk te kunnen sprinten. De gastvrijheid van jou en Ilse is een grote steun voor mij geweest, fijn dat je zo op mij gepast hebt. Dank dat je mijn paranimf wilt zijn, het is een eer om jou aan mijn zijde te hebben.

Drs. N.F. Casteleijn, beste Niek. Wat hebben we veel vlieguren gemaakt. Jouw arbeidsethos is indrukwekkend, met een even indrukwekkend resultaat. Het ontspannen voor of na congres konden we dan ook goed gebruiken. Hierdoor hebben we een paar van de mooiste plaatsen ter wereld mogen zien. En de koudste. Ik mis je nu al als vanzelfsprekende co-piloot en ka-mergenoot. Knap hoe jij alles in balans weet te houden. Voor jouw steun als paranimf ben ik buitengewoon dankbaar.

Onderzoek begint traditiegetrouw in het Triade-gebouw. Collega-triarii, Charlotte en Laura de Vries, wat een mooie start was dat! Deeluitmakend van de laatste linie hoorden we elke dag de geheimzinnige generatoren uitschakelen. Wij moesten door: abstract deadlines, cupjes stickeren, poli’s screenen. Inmiddels weten we de balans beter te vinden en op onze reserves te letten, ‘winter is coming’. Maar met humor, een drupje cynisme en een snufje taalpurisme slaan we ons overal doorheen!

Na mijn bevordering tot de Brug kwam ik bij Janneke op de kamer. Goodmorning Moneypenny! Jouw onverwoestbare optimisme ten aanzien van gebruikersvijandelijke elektronische registra-tieformulieren en protocollen werkt aanstekelijk. Dank voor zoveel vrolijkheid op de vroege morgen. Harmke, het was een bijzondere tijd waarin we veel hebben meegemaakt: met de slag om het open raam als terugkerend verzetje. Solmaz, mijn promotietraject is ook dankzij jouw werk(plek) goed verlopen! Dan mijn kamergenoten-in-opleiding-tot-nefroloog: Femke, Folkert, Susan, Janke, Heleen en Mark. Dankzij jullie hield ik ook contact met de kliniek, dit bleek extra plezierig toen ik begon met het jullie welbekende HYP1/2 hypertensiespreekuur.

Captains on the Bridge, collega’s van de naburige kamer. Maarten, altijd scherp op FGF-23 sym-posia, het gedenkwaardige bezoek aan Milaan zal mij altijd bijblijven…Prosit! Over wijnlanden gesproken, dank aan Michele en Antonio voor zuidelijke invloeden. Gerald, mooi team vormden we voor SUBLIEM. Ineke en Nicole, jullie gezelligheid en excellente stata-voering wordt gemist.

221

Dankwoord

D

Arjan, dank voor je hulp bij het afronden van diverse artikelen. Tsjitske en Annemarijn, het was een mooie cursus dalen-met-mountainbike in San Diego. Joline, jouw financiële bijstand is voor elke onderzoeker onmisbaar.

Nefrologie in de Triade: Lianne en Lyanne, wat een mooie reizen hebben we gemaakt over (en in) de oceaan. Fijn dat het bijna weer vrijdag is. Irina, wij komen elkaar altijd weer tegen, grote bewondering voor jouw onafhankelijkheid en reislust! Brigit, dank voor jouw hulp met FGF-23 bepalingen. Onderzoek bij de Nefrologie is zo leuk door veel meer collega’s: Arno, Bjorn, Coby, Debbie, Dineke, Dorien, Edwin, Elise, Esmée, Ilse, Isidor, Janine, Jacob, Laura Harskamp, Marieke, Marco, Maryse, Michel, Rik, Sara, Wouter, en alle anderen oud en nieuw. Dank voor de wetenschappelijke discussies, lunches, borrels, sinterklaasavonden, een dagje skiën en congres-bezoek: Join Nephrology and See the World.

Verder ben ik dankbaar voor de samenwerking en samenscholing met nefrologisch Nederland. Op de ASN, ERA-EDTA, Nefrologiedagen, Winterschool en PLAN-dagen: intensieve dagen met mooie avonden. Alle projectleden van het SUBLIEM onderzoek wil ik bedanken voor de goede samenwerking onder hoge tijdsdruk, met name projectleiders Paul, Sandra en Gerjan en uit-voerders te velde Anke, Hanne, Irma, Inger, Trijntje en alle anderen. Yvette, bedankt voor het teamwork in ESMO en SUBLIEM. Dank aan de leden van NIGRAM consortium voor de goede samenwerking in de afgelopen jaren. Marc Vervloet, inspirerend om met zo’n enthousiaste nefroloog samen te mogen werken aan diverse artikelen. Melissa, het was net werken in Milaan, jammer dat we dit jaar geen kerstinkopen kunnen doen.

Lara en Lilian, Team Worldcup, jullie waren onmisbaar voor de uitvoering van de SUBLIEM studie. Meer dan tienduizend stickers en cupjes, honderden bloed- en urinebuisjes, en span-nende ritjes door het land voor studielogistiek: geweldig! Wendy, bedankt voor jouw hulp bij het verrichten van de vele FGF-23 bepalingen. De collega’s van de nierfunctiekamer: Marian, Roelie en Dirkina, bedankt voor de gastvrijheid. Winnie, hart en nieren van de afdeling, super hoe je elke promovendus op weg helpt in de logistiek in het UMCG. Erna, bedankt voor je hulp en geduld wanneer poli’s verplaatst of samengevoegd konden/moesten worden.

Laura Meems, ik ben benieuwd in welk gremium we elkaar gaan tegenkomen. Na mooie jaren in medisch onderwijs ben jij nu cutting-edge cardiologisch onderzoek aan het doen in de VS. Daar moet een brug naar te slaan zijn. Jouw onuitputtelijkheid in werk en goede wijn is legendarisch. Bij jouw eigen promotie maakte je een diepe indruk op jouw paranimfen Rosa en ikzelf. Ik hoop dat ik voor mijn verdediging maar half zo ontspannen ben als jij was. Rosa, dank voor de mooie gesprekken en vergaderingen in Barrel! Anne Koning, heerlijk om nieuwe hypothesen te gene-reren, we moesten die eens testen.

222

Dankwoord

Ilse, Sjoukje, Elmire; zeker aan het begin van onze promotietrajecten veel gehad aan het wielrennen, hardlopen en spelletjesavonden. Dat we maar zo blessurevrij mogelijk door onze opleidingen mogen komen!

Blijven leren is belangrijk. Erik, Patrick: over uitdagingen gesproken. Als welkome afwisseling hebben we een mooi stukje werk mogen afleveren voor het nieuwe curriculum. Dank aan Ta-lentweb Groningen dat ik ook na kantooruren mocht blijven leren. Zo mogelijk nog intensiever is de Nederlandse Vereniging voor PAD-Alumni. Bestuursgenoten Alette, Anke, Jirsi, Maarten, Sandra; ik kijk altijd uit naar onze vergaderingen en leerzame reizen met de NVPA. Het is ge-nieten om met zo’n bevlogen, enthousiaste en ambitieuze groep mensen op pad te gaan. Rob, bedankt voor jouw kritische blik! Cees, Paul, Marritt; buitengewoon gezellig én uitdagend om samen het eerste Lustrum te organiseren.

Heren van Sneek: Dirk-Jan, Geert, Hidde en Joost. Vriendschappen die —ondanks Risk— al twin-tig jaar bestaan. Dank voor de hilarische gesprekken, briljante vakanties en de zeilweekendjes-waarvoor-je-ook-een-week-had-weggekund. Fijn dat we zo gewoon zijn gebleven. Ik ben benieuwd wie van ons wanneer zijn eerste jacht koopt.

Ik prijs mijzelf gelukkig deel uit te maken van hechte families. Humalda, Roorda en Schrale. Het is altijd fijn jullie neven, nichten, ooms en tantes, weer te zien!

Mijn pake, Kor Humalda. Van jongs af aan hebben Afke en u ons gestimuleerd: bijvoorbeeld met boeken als verjaardagsgeschenken waaronder mijn eerste encyclopedie. Later heeft u laten zien hoe ongelofelijk sterk en zorgzaam u bent. We moeten helaas Beppe missen. U bent een ware pater familias. Ik bin hiel grutsk op dy.

Pake en Beppe Roorda, jullie blijven in gedachten bij ons.

Nynke en Berber, zusjes! Ik ben superblij met jullie. De familiediners zijn legendarisch, on-getwijfeld dankzij het subtiele gevoel voor humor dat wij in ons leven ontwikkeld en verfijnd hebben. Hierin zijn we ondanks onze verschillen één. Berber, supergaaf dat je de voorkant van mijn boekje hebt ontworpen. Nynke, sorry van de bomen. Zonder gekheid, jouw inzet voor een betere wereld is inspirerend, we moeten we daarvan leren. Kusjehoi!

Lieve pa en ma, wat is het fijn om zulke lieve ouders te hebben. Met onvoorwaardelijke steun hebben jullie mij altijd alle kansen geboden. Het is fijn dat wij altijd op de familie kunnen terug-vallen, en bij spanning de rust van Sneek kunnen opzoeken. Mam, zorgzaam en supersportief. Pap, altijd rustig, hardwerkend en vriendelijk voor eenieder. Jullie zijn mijn voorbeelden. Ik hoop dat we samen nog eens de (gehele) Elfstedentocht mogen rijden. Knuffel!

Jelmer

Splice the mainbrace.

About the Author

227

A

About the AuthorJelmer Kor Humalda was born in Rotterdam, 11 May 1988, and raised in Sneek. After graduating from gymnasium RSG Magister Alvinus, he chose to study Medicine at the University of Gronin-gen in 2006. He graduated both Bachelor and Master cum laude in January, 2013. He conducted his research elective at Nephrology with supervisors Martin de Borst and Gerjan Navis, which led to admission to the MD/PhD programme in 2013. April 2013, he was assigned the coordinat-ing role of the multidisciplinary, multicenter trial SUBLIME that was successfully completed in December 2015. The work of this thesis led to several oral and poster presentations at inter-national conferences, notably oral presentations at the American Society of Nephrology Kidney Week in 2012 and oral presentations at European Renal Association – European Dialysis and Transplantation Association congresses in 2013 and 2015. In 2014 he was elected to participate in the European Society of Hypertension Summer School in Varna, Bulgaria. The acquisition of international contacts at these events proved invaluable for several parts of this thesis.

During his Medicine studies, in keeping with family tradition in education, he was chair of ProMed student representation from 2007‒2009; held position in the Decentral Admission Committee and Project Group G2010+, Curricular Development and Update; and was tutor for G2010 Bachelor students in small group education. He graduated from University of Groningen Master Honours ‘Leadership’ Programme in 2012. During his PhD-track, he achieved the Basic Qualification Education for students in 2013 and became member of TalentWeb Groningen in 2014. He was involved in development of tutor assignments for the first academic year of the new medical curriculum G2020 2014‒2015. Underscoring his affinity for Dutch-German cultural interactions, he joined the board of the Nederlandse Vereniging voor PAD-Alumni in 2014. In 2015 he was selected to participate in the UMCG Talent Development Program ‘UMCG NEXT-III’, and completed assignments related to Healthy at Work, Education and Research programmes. Upon completion of the thesis, Jelmer started formal training in Internal Medicine at Isala Klinieken, Zwolle, June 2016.

SupplementsPreface .

Supplement A: E-health Website ‘SaltModule’ of the SUBLIME Study Page 231.

Supplement B: ‘Salt Down, Spice Up!’‒ An Example of Societal Impact Page 235.

230

Supplements

Preface.The clinical studies in this thesis offer a panoply of means for societal interactions. We be-lieve that societal interaction ought to be the logical consequence of a clinical trial. In these Supplements, we share our experiences wherein we aimed to bring academic knowledge to the general public in a fun and easy accessible way. In (→ Supplement A), we describe the e-health intervention deployed in the SUBLIME study in more detail. The e-health intervention ‘SaltMod-ule’ is partly based and partly a more-developed version of the publicly available website www.mijnnierinzicht.nl, customized for the SUBLIME study. However, after implementation of the feedback of our participants, this may well form the basis of a new publicly accessible version. In (→ Supplement B) we describe a HealthHack we performed at the arts, theatre and music festival ‘Noorderzon’. HealthHacks are smart, easy ways to ‘hack’ your daily routine in order to implement a more healthier routine. Our HealthHack aimed to raise awareness of excess salt consumption: “Salt Down, Spice Up!’.

231

Supplements

S

Supplement A.

E-Health Website ‘SaltModule’ of the SUBLIME Study

Jelmer Humalda

The development processThe SUBLIME study (→ Chapter 10) deployed an e-health application based on self-regulation theory, to facilitate behavioral change in sodium intake in patients with chronic kidney disease (CKD). The e-health intervention consisted of a website created by Bonstato BV and third-party IT-developers. Its infrastructure was based on www.mijnnierinzicht.nl, a publicly accessible website of Bonstato BV and owned by its director Hannie Piels. This website consists of a food diary function that has been linked with the NEVO-tables, and thus all available information on nutrients. This diary function is suitable for monitoring of goals, i.e. sodium intake. Self-monitoring is only a part of the self-regulation therapy. The medical psychologists (Sandra van Dijk, Yvette Meuleman) within the project group SUBLIME designed several exercises to address other components of self-regulation theory as outlined below. The e-health module (SaltMod-ule, Dutch: Zoutmodule) underwent pilot testing with volunteers and patients with CKD to improve its interface. Nierpatiënten Vereniging Nederland (NVN, Dutch Kidney Patients Associa-tion) was also involved in the development of SaltModule. They strongly advised to maintain the option to visualize changes by moving switches (Figure 2). This was the most-expensive part of SaltModule to design and it was designated to be scrubbed, however after consultation with NVN representatives and the positive results from pilot tests, the project group allocated additional budget to implement this part into the SaltModule.

Implementing e-health: SaltModule.The SaltModule is a text-based website that was only available for SUBLIME participants who were randomised to the intervention group. The SaltModule consists of two lines (Let’s Start and Evaluation) that consist of different submodules: Introduction, Risk, Motivation, Monitor-ing, Self-efficacy, Goals, Support, and Options for Change. We briefly describe the components of the Let’s Start module below.

Further, the Monitoring menu reports the sodium content and offers options for change (Figure 2).

Self-Efficacy In this exercise patients identify barriers and possible solution to achieve sodium restricition. They state their own strengths and weaknesses regarding sodium restriction in a

232

Supplements

multiple choice menu. Next, they are asked to what extent the barriers are a problem, given the self-reported solutions. Finally, patients score their own confidence in their ability, or self-efficacy, to maintain a sodium restriction.

Goals After a brief introduction on how to set achievable goals (SMART, specific, measurable, achievable, relevant and time-bound. Patients review their earlier answers on barriers and op-tions for change and are then triggered to set their own goal

Support In this exercise patients are asked to identify social support: who could help them to achieve their sodium goal, how could they help (multiple options e.g. “by helping reading contents on food products”. Next, they are asked how they will reward themselves when a goal has been reached.

Introduction General information about the study

Risks Brief overview about the relation between salt, blood pressure with cardiovascular and renal events.

Motivation Patients are asked what they consider important in their life by selecting a picture. Next, they are asked how a healthy lifestyle fits within these important things in life. Finally, they are asked for reasons why they do or do not pay attention to salt, and to score on a 1‒10 scale how important they think a sodium restriction is in their own situation.

Monitoring Here patients can fill-out the diary and select food products that constitute their meals. SaltModule remembers earlier choices, so filling out the food diary will take less time (Figure 1.).

Figure 1. Monitoring/Diary Diary of the Monitoring Module. Upper left are preferred products (‘voorkeuren’), that have been filled out earlier at this time of day (breakfast). Patients can specify units and quantities of a given food product.

233

Supplements

S

Change plan a summary of the previous answers, that is used by the coach as guidance for e-coaching.

The Evaluation module consists of Introduction, Monitoring, Experiences, Relapse, Self-Efficacy, Goals, Support, and Change Plan.

Introduction Here is explained that behavioral change is a continuous process, and that the purpose of this module is to evaluate the excecution of the Change Plan so far.

Monitoring Similar as above, patients can register their food intake.

Experiences Patients are asked what went well and what could be improved in the last period.

Figure 2. Monitoring/Options for Change. Ranks food products in a given time period for their sodium content. The switches (red circle) can be moved to assess the effect of reducing portions. Alternatively, patients can replace a product by clicking the change symbol (blue circle), and select for example ‘low-sodium bread’. The effect of reduction/change is displayed in blue bars (upper panel), the current intake in gray bars.

234

Supplements

Relapse Patients are reassured that goals are not always met (and that this is not a bad thing). They are invited to describe a situation where they failed to achieve a sodium restriction goal and to possible challenges that may emerge. Next, they are asked to identify why it was difficult and change their value of the Barriers from the Let’s Start program. Finally, patients could score their self-efficacy.

Goals, Support, and Change Plan are similar as above. This way, patients update their Change Plan, that forms the basis for consecutive e-coaching moments.

Although the abovementioned order is recommended, patients can access all components at their own discretion. To conclude, SaltModule is an e-health application that offers several tools to enhance the self-regulation skills of patients to support a sustainable sodium restriction.

Figure 3. The Change plan summaries why health is important, important reasons to restrict sodium, reiterates the set goals and steps from the Options for Change menu. Finally, two persons are asked for support in finding recipes and reinforce good behavior.

235

Supplements

S

Supplement B.

HealthHack ‘Salt Down, Spice Up!’ – An Example of Societal Impact

Jelmer Humalda

A summer evening at the Groningen modern arts festival Noorderzon. Amidst starting artists, modern theater and several food corners, citizens of Groningen gather around a small table to experience a salient experiment. Cups are filled, swirled and tasted and the brave participants are asked to score the solution on its saltiness: “1 for nothing, to 10 for seawater!”

The Experiment: ‘Salt Down, Spice Up!’The experiment is straight-forward. Four to ten participants taste 4 solutions in succession, and are asked to score them for their ‘saltiness’. Solution A consists of a 17 g/L saline solution, whereas B and D contain 10 g/L. C is plain water. The scores are kept on a flip-over board. Frequently, participants tend to overestimate the D solution as the most salt one, and typically rank the solution D firmly above the equal saline B solution. At the disclosure, the presenter outlines what happened: you got used to the salt in A, and thus did not recognize the salt in B, that was just as salty as D! This equals what’s going on in our society, we do not recognize salt in our food anymore, because we are used to it…our taste has been wrecked by salt!

Partipants are then rewarded with a small sample of spices –without salt. This “Northern Spice” mix has been developed by our dietitian Trijntje Kok, and came with a recipe to demonstrate that tasty and salty are not synonyms. Next, participants could move to the next table. They could try to order food products in order of salt content, guided by a dietitian. Included were eye-openers: canned vegetables are not necessarily salt-rich, whereas ready pasta sauce and other processed foods contains far more salt than the equivalent of fresh vegetables. Strikinlgy, a pizza contains more than the recommended daily allowance of salt intake.

236

Supplements

Figure 1. “Rien ne va plus, hoeveel zout proeft u?” Jelmer Humalda pours the next round of saline, Trijntje Kok (right) prepares for the educational part.

BackgroundThis experiment was developed in the setting of HealthHack050, an initiative led by Ritzo ten Cate in collaboration with Jan Bouwhuis from the UMCG. Young employees of the UMCG, part train-ees of the UMCG Next Talent Development Program, were asked to think of easy-to-implement exercises or daily routines to increase health of the general population. These are called ‘health hacks’ ‒smart tricks to improve or ‘hack’ your health‒. We were asked to test these initiatives with the local population on the eleven-day Noorderzon festival. In the brainstorm sessions healthy nutrition was coined, which triggered us to consider salt and led us to invite dietitians and experts to join the group. Other initiatives included several physical exercises and games, social exercises and brainstorm sessions, and volunteers executed the exercises on several days of the Festival. ‘Salt down, Spice Up!’ was performed on five consecutive evenings by dietitians, trainees, PhD students and professors. ‘Salt Down, Spice Up!’ is therefore a joint product of interactions between JongUMCG, TalentWeb Groningen, UMCG NEXT Trainees, dietitians, PhD students and medical staff. It was a lot of fun to do, and proved to be an easy, energetic, and cheap way of raising awareness of sodium content in food products in the general population.

DisseminationIn the months that followed the experiment was performed in different settings: for 2nd year medical students as adjunct to a lecture, at the Dutch Nephrology Days and the Dutch Dietitians Days. The Dutch Kidney Foundation kindly requested the protocol, and implemented it for the World Kidney Day Event where the HealthHack was one of the experiments performed

237

Supplements

S

at primary schools by doctors and researchers. The group that contributed to the HealthHack is dynamic and expanding, and more demonstrations and follow-ups on the HealthHack are planned.

Acknowledgements

HealthHack050 ‘Salt Down, Spice Up!’ was powered by: Jan Bouwhuis, Ritzo Ten Cate, Nicole Dijk, Wil Heikamp, Heleen Hoogeveen, Jelmer Humalda, Maarten de Jong, Trijntje Kok, Gerjan Navis, Juliette Schuurmans, Laura de Vries, Marianne Zwolsman. Due to the non-formal organi-zation, this list is ever-changing and likely not complete.

Reference

Press coverage: http://www.rug.nl/research/behavioural-cognitive-neurosciences/news/bcn-newsletter100.pdf. Page 26. H Hoogeveen. Last accessed 10.04.2016.

238

Supplements

Recipe by Trijntje Kok

Het Noorderkruid

Een kruidenmengsel passend bij Groningen, authentiek en met sterk karakter. Er zit geen zout in verwerkt waardoor het mengsel zeer gezond is. Door de unieke combinatie van kruiden lek-ker voor jong en oud en te combineren in vele gerechten.

De ingrediënten:

1 x cayennepeper, 2 x gemberpoeder, 1 x kaneel, ½ x kardemon, 2 x komijn, 2 x koriander1 x kruidnagelpoeder, ½ x nootmuskaat, 3 x zwarte peperHieronder een voorbeeld van een recept:

Zoet-hartige couscous met perzik (4 personen)

300 g kipfilet, in grove stukken gesnedensap van 1 citroen4 el olijfolie2 el Noorderkruid300 g couscous100 g rozijnen400 ml water4 wilde perziken (of 8 halve perziken uit blik)4 el honing60 g ongezouten cashewnoten, grof gehakt15 g munt, grof gesneden1. Schep in een kom de kip om met 3 el citroensap, 2 el olijfolie en 1 ½ el kruidenmix. Verhit

een grill- of koekenpan en bak de kip in 12-15 minuten gaar.2. Doe de couscous met de rozijnen en ½ el kruidenmix in een kom en schenk er 400 ml kokend

water bij. Dek de kom af en laat 10 minuten staan.3. Halveer de perziken en verwijder de pit. Snijd het vruchtvlees in plakjes.4. Roer de couscous los met een vork en voeg de honing, de rest van het citroensap en de rest

van de olie toe.5. Schep de couscous in een grote schaal en verdeel de kip en de perziken erover.6. Garneer met de noten en de munt.

University of Groningen Fibroblast Growth Factor 23: a ... · Fibroblast Growth Factor 23: A Bridge...

Documents

Transcript of University of Groningen Fibroblast Growth Factor 23: a ... · Fibroblast Growth Factor 23: A Bridge...