university of groningen renal function after solid organ transplantation … · 2016-03-08 ·...

University of Groningen

Renal function after solid organ transplantationBroekroelofs, Jan

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:2000

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Broekroelofs, J. (2000). Renal function after solid organ transplantation. s.n.

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: 30-10-2020

https://www.rug.nl/research/portal/en/publications/renal-function-after-solid-organ-transplantation(cce96ab0-b754-44f0-b4e5-7bff74b80a05).html

https://www.rug.nl/research/portal/en/publications/renal-function-after-solid-organ-transplantation(cce96ab0-b754-44f0-b4e5-7bff74b80a05).html

RENAL FUNCTION AFTER SOLID ORGAN TRANSPLANTATION



Jan Broekroelofs


Jan Broekroelofs

Broekroelofs J.

Renal function after solid organ transplantation.

Proefschrift Rijksuniversiteit Groningen. Met literatuuropgave en Nederlandse samenvatting.

ISBN 90-367-1317-X

NUGI 742

© Copyright 2000 J. Broekroelofs

All rights are reserved. No part of this publication may be reproduced, stored in a retrieval system, or

transmitted in any form or by means, mechanically, by photocopying, recording, or otherwise, without the

written permission of the author.

Printed by Hellinga b.v. Leeuwarden, The Netherlands.

RIJKSUNIVERSITEIT GRONINGEN


Proefschrift

ter verkrijging van het doctoraat in de

Medische Wetenschappen

aan de Rijksuniversiteit Groningen

op gezag van de

Rector Magnificus, dr. D.F.J. Bosscher,

in het openbaar te verdedigen op

woensdag 15 november 2000

om 14.15 uur

door

Jan Broekroelofs

Geboren op 6 november 1961

te Hardenberg

Promotores Prof. dr. P.E. de Jong

Prof. dr. G.J. Navis

Prof. dr. D. de Zeeuw

Referent Dr. C.A. Stegeman

Beoordelingscommissie Prof. dr. B.L. Kasiske

Prof. dr. G. Koëter

Prof. dr. L.C. Paul

Paranimfen Fenny Broekroelofs

Willem Dikkers

Voor Pelle, Maartje, Gijs en Joost

Contents

Prologue History and Scope 11

Chapter 1 Prevention of renal function loss after non-renal solid organ transplantation: How can nephrologists help to keep the kidneys 15out of the line of fire?Nephrol Dial Transplant 1999; 14: 1841-3

Chapter 2 Creatinine-based estimation of rate of long-term renal function loss in lung transplant recipients. Which method is preferable? 23J Heart Lung Transplant 2000; 19: 256-62

Chapter 3 Renal haemodynamics after lung transplantation: a prospective study 37Transplantation 1996; 61: 1600-5

Chapter 4 Long-term renal outcome after lung transplantation is predicted by the 1 month post-operative renal function loss 51Transplantation 2000; 69: 1624-8

Chapter 5 Early renal function loss after lung transplantation: analysis of peri-operative risk factors. 65Submitted

Chapter 6 Risk factors for long term renal survival after renal transplantation: a role for angiotensin-converting enzyme (insertion/deletion)polymorphism? 79J Am Soc Nephrol 1998; 9: 2075-81

Chapter 7 Summary and General Discussion 95

Chapter 8 Samenvatting 107

Appendix Lancet 1998; 351: 1064 119Lancet 1998; 352: 69-70

Dankwoord 125

Curriculum Vitae 127

Prologue

History and Scope

History and scope of this thesis

Solid organ transplantation already has a long history. Among them renal transplantation hasthe longest history and thus follow up after successful renal transplantation is largest. Theintroduction of cyclosporin A in 1983 has been a breakthrough. Until then acute rejectionwas the most important complication after renal transplantation and often difficult to treat.Renal survival at 1 year after transplantation improved significantly after the introduction ofcyclosporin A. However, renal graft survival from 1 year after transplantation onwardshardly changed after the introduction of cyclosporin A and still shows nearly the same trendas before. Within the Division of Nephrology in Groningen in 1966 the first successful renaltransplantation has been performed. Therefore, the experience with renal dysfunction afterrenal transplantation is large. In chapter 1 we discuss that insights, but also many questions,gathered from experience with dysfunction after renal transplantation could be helpful in theunderstanding and unravelling of renal dysfunction after lung transplantation, the ultimategoal being adequate risk assessment and prevention of renal function loss in these non-renalsolid organ transplantations. Heart-, lung-, liver- and heart-lung transplantation have a much shorter history and in thefirst years transplant- and therefore patient survival was very poor. Between 1963 and 1979about 40 lung transplantations have been performed and all but one patient died within onemonth after transplantation1,2. Until now the total number of lung transplants isapproximately 6400 and survival rate is increasing due to better immunosuppressivestrategies. The introduction of cyclosporin A in lung transplantation has been important forincreasing survival rates. At 1 and 5 years after transplantation the survival rate is currentlyapproximately 70 and 40% respectively3. As the number of recipients after lungtransplantation with prolonged survival is increasing, serious long-term side effects ofcyclosporin A such as renal dysfunction are becoming a relevant problem in this specificpopulation. From the experiences with cyclosporin nephrotoxicity in renal transplantation, it wasanticipated that renal dysfunction might occur after lung transplantation as well. Therefore,from the start of the lung transplantation program onwards, monitoring of renal function wasincluded in the routine evaluation during screening for transplantation, as well as duringfollow up after transplantation.In this thesis we studied renal function after renal as well as after lung transplantationbecause it is supposed that, apart from obvious differences, similarities will be present. Ourstudies in recipients after kidney and lung transplantation were aimed to further evaluate thecourse, aetiology and monitoring of renal function loss after solid organ transplantation.Renal function loss after solid organ transplantation is a multifactorial process. Largepopulations are needed to analyse the different factors contributing to this loss of renal

13

function and unravelling the different factors is a prerequisite to develop renoprotectivestrategies. The golden standard to monitor renal function is serial measurement of glomerular filtrationrate (GFR). In clinical practise, however, this is not always feasible due to logistic, financial,or patient related problems. In patients after solid organ transplantation great changes inbody mass and composition may occur. Serum creatinine, which is widely used as a markerfor renal function, is critically dependent on body composition. To assess the utility but alsothe limitations of serum creatinine as a marker of renal function we analysed in chapter 2different creatinine based methods to measure renal function in lung transplant recipients.Before adequate prevention of renal function loss after lung transplantation can becomereality, insights in the course and predictors of renal function loss after lung transplantationhave to be gathered. In the closely monitored cohort transplanted at our institution weanalysed this course and possible predictors for short and long-term renal function loss afterlung transplantation. In chapter 3 we measured renal haemodynamics before and at severaltime points after lung transplantation. We discuss differences in renal haemodynamicsbefore and after lung transplantation within the group of lung transplant recipients. Inchapter 4 we quantified renal function loss during more long-term follow up in patientsafter lung transplantation. Analysed end points were the slope of GFR from 6 months post-transplantation onwards and the GFR at 24 months after transplantation and we discussthe factors associated with this long-term renal prognosis. Early post-operative renalfunction loss is considerable after successful lung transplantation and therefore, in chapter 5,we analysed predictors for renal function loss at 1 month after transplantation.As proven in experimental and human studies, a major system involved in progression ofrenal disease is the renin-angiotensin-aldosterone system. A genetic polymorphism resultingin measurable distinct phenotypes within this system, the angiotensin-converting enzym(insertion/deletion) (ACE (I/D)) gene, has been described. As the population of recipientsafter lung transplantation is still too small to perform such an analysis, the influence of theACE (I/D) polymorphism on graft survival after renal transplantation was studied asdescribed in chapter 6. Since we were interested in the influence of this geneticpolymorphism on progressive renal function loss, pure graft survival (patients dying with afunctioning graft were censored, excluding the first post-transplant year) was studied.

14

PROLOGUE

References

1 Patterson GA, Cooper JD. Status of Lung Transplantation. Surgical Clinics of North America 1988;

68: 545-58

2 Derom F, Barbier F, Ringoir S, Versieck J, Rolly G, Nerzsenyi G, Vermeire P, Vrints L. Ten-month

survival after lung homotransplantation in man. J Thorac Cardiovasc Surg 1971; 61: 835-46

3 Internetsite: www.ishlt.com

15

Chapter 1

Prevention of renal function loss after non-renal solid organ transplantation:

How can nephrologists help to keep the kidneys out of the line of fire?

Broekroelofs J, Stegeman CA, Navis GJ, De Jong PENephrol Dial Transplant 1999; 14: 1841-3

After solid organ transplantation, be it renal or non-renal, renal function loss is common1,2,3,4.In renal transplantation, chronic renal function deterioration as an important cause of long term graft loss is well recognized1. In this population, studies aimed at elucidating itsmechanisms and improving long-term renal allograft prognosis are performed. Whereas inrecipients after non-renal solid organ transplantation progressive renal function loss is animportant problem as well, the knowledge on renal morbidity in these populations isrelatively limited and scattered. Improvements in non-renal solid organ transplantation haveled to improved patient and graft survival. The burden of renal morbidity in thesepopulations grows with the increasing number of recipients, and with the increasing numbersurviving long enough to develop clinically significant renal problems. Therefore, itbecomes mandatory to develop renoprotective strategies in these high-risk groups. As failingkidneys after non-renal organ transplantation share many features, such as vascularobliteration with ischemic glomerular collapse and sclerosis, tubular atrophy, and interstitialfibrosis5 with chronic renal transplant failure, insights derived from renal transplantationmay be useful in this respect1. However, subtle differences have been found at the matrixprotein level between kidneys after renal as compared to non-renal transplantation, pointingto some differences in pathogenic mechanisms6. Clearly, insights derived from renaltransplantation need to be combined with knowledge on specific risk factors for eachparticular population in non-renal transplant recipients. The other way round, insights fromrenal function loss in non-renal transplant recipients may increase our understanding of non-immunological factors in chronic renal transplant failure.

Renal function loss in non-renal as compared to renal transplant recipients

Chronic renal function loss occurs in many patients after initially successful renaltransplantation. Multiple risk factors, both related and unrelated to the allograft status of thegrafted kidney, have been identified, suggesting a complex and multifactorial pathogenesis1.After renal transplantation, renal function initially improves and later on stabilises in mostrecipients. Acute deteriorations caused by preservation trauma or acute rejection are clearlyrelated to the allograft status of the kidney. Slowly progressive renal function loss firstoccurs after months or even years. Although immunological factors are clearly involved, asacute rejection episodes are the strongest predictor for subsequent chronic renal transplantfailure1, non-immunological factors such as blood pressure7 or cyclosporin toxicity play animportant role. In contrast, in heart or lung transplant recipients renal function loss is mostprominent the first 6 months after transplantation with stabilisation or slow progressionthereafter2,5. So, despite clear differences in the early post-transplant course and the allograftstatus of the transplanted kidney, long-term renal function loss may be comparable, bothclinically and histologically, with involvement of non-immunological factors in both renal

CHAPTER 1

17

and other solid organ recipients. Identifying the importance and role of these factors in thedifferent transplant recipient populations is the necessary first step for the development ofpreventive measures.

Diversity in renal function loss after non-renal solid organ transplantation

Cyclosporin A and tacrolimus are important causes of both acute and chronic renal functionloss after non-renal solid organ transplantation5,7,8. Despite that all receive cyclosporin A ortacrolimus, the severity and course of renal dysfunction displays great diversity between andwithin different populations of non-renal transplant recipients. As compared to heart- or lungrecipients with 5-10% end-stage renal failure within 5-7 years after transplantation2, livertransplant recipients are clearly less at risk3. Also, within the groups of heart or lungtransplant recipients the individual course of post-transplant renal function ranges fromrapid deterioration to long-term stability2,4. This diversity in renal prognosis suggestsdifferences in exposure and susceptibility to nephrotoxic insults, which are related both tothe type of organ transplant and individual patient factors.Differences in immunosuppressive dosing regimens may contribute to differences in renaldamage. High cyclosporin levels and exposure early post-transplant, common after heartand lung transplantation and combined with increased susceptibility caused by peri-operative instability and the use of extra-corporeal circulation, may be a factor. A recentanalysis of factors predicting loss of renal function 1 month after lung transplantation in 83patients in our centre identified items such as duration of the operation, hypotensiveepisodes (mean arterial pressure (MAP) <70 mmHg), use of aminoglycoside antibiotics andearly post-operative diuresis as important determinants.Pre-existing renal status could also explain differences in renal function loss after solidorgan transplantation. Atherosclerotic vascular disease, present in heart transplant recipientswith ischemic heart disease, may predispose to increased renal susceptibility. Among lungtransplant recipients, patients with cystic fibrosis have a particularly poor renal prognosisdespite normal pre-transplant renal function9, which may be related to prior and currentexposure to aminoglycoside antibiotics, diabetes mellitus, or to pre-existent renal tubularfunction disorders or microcalcinosis. In liver transplantation, pre-transplant renal diseaserelated to the hepatic disease may be present. On the other hand, patients with cardiac failureor pulmonary hypertension with severely impaired renal function pre-transplantation, oftenshow renal function improvement following successful heart or lung transplantation4.Likewise, successful liver transplantation cures hepatorenal syndrome.

18

PREVENTION OF RENAL FUNCTION LOSS

Prevention of renal function loss after non-renal solid organ transplantation

Less nephrotoxic immunosuppressives may reduce renal morbidity, but until nowcyclosporin A or tacrolimus are indispensable to preserve graft function. In the early post-transplantation phase, the nephrotoxicity of these agents may be aggravated byhaemodynamic, operation related instability or concomitant use of other nephrotoxic agents.As induction therapy with newer monoclonal agents and addition of mycophenolate mofetilare available, critical reappraisal of the moment of introduction of cyclosporin or tacrolimusand the targeted levels may reduce this early toxicity. During long-term follow up, carefulmonitoring of trough cyclosporin or tacrolimus levels, withdrawal of these agents or the useof calcium-entry-blockers10 are possible tools. Early diagnosis and adequate therapy ofcyclosporin-induced hypertension is likely to be important as in native kidney diseases andrenal transplantation hypertension is associated with worse renal prognosis7. Whether thesepotential preventive measures will result in relevant renoprotection in non-renal transplantrecipients remains, however, to be investigated.

Monitoring of renal function after solid organ transplantation

Accurate monitoring is a prerequisite for the prevention of renal function loss. After non-renal solid organ transplantation serum creatinine is the usual parameter for renalfunction during follow up. Its interpretation, however, is subject to the confounding effectsof (changes in) muscle mass and tubular creatinine secretion. Following renal and non-renalorgan transplantation body composition may change considerably, with increased (greatermuscle mass by improved well-being and exercise capacity) or decreased (muscle masswasting after operation or infections; corticosteroids) creatinine generation. Measurement ofcreatinine clearance may circumvent some of these problems, but induces inaccuracies byurine collection errors. Moreover, early after transplantation the extent of nephrotoxicdamage may be masked by the favourable effects of improved circulatory status in patientswith prior heart failure or cor pulmonale, or by correction of hepatorenal syndrome.Long-term renal function monitoring by more sophisticated methods like iothalamateclearance is laborious, demanding on patient compliance and costly. These disadvantages,however, may be outweighed by its accuracy, which provides the possibility to detect smallchanges in GFR leading to early identification of high risk individuals or subgroups11. Evenwith sophisticated renal haemodynamic measurements, however, renal function mayinadequately reflect changes in the kidney. In a small follow up study after cardiactransplantation repeated renal biopsies showed clear-cut progression of glomerular andinterstitial fibrosis while the glomerular filtration rate had stabilised or even improved12. Asafter renal transplantation, where serial renal biopsies have been found to be a good

CHAPTER 1

19

predictor of long-term graft failure13, serial renal biopsies in patients after non-renal solidorgan transplantation, albeit invasive, may be relevant for our understanding of thepathophysiology and prediction of possible progression or reversibility.

Conclusion

Renal dysfunction after non-renal solid organ transplantation is a problem of growingimportance. Nephrologists should play an active role in the care and research in thesepatients as early as the pre- and peri-transplantation phase, less patients may come tonephrological attention only after end-stage renal failure has become inevitable. Expertisefrom renal transplantation, by virtue of large patient numbers and better documentation ofrenal structural damage may be turned into benefit for these populations as well. This mayhelp to keep the kidneys out of the line of fire -and the patients out of the dialysis units-, andimprove our understanding of chronic graft dysfunction after renal transplantation as well.Improvements in renal risk assessment, monitoring and prevention of renal failure areclearly needed in patients after non-renal transplantation. It will be worth the effort!

20


References

1 Paul LC. Chronic renal transplant loss. Kidney Int 1995; 47: 1491-9.

2 Pattison JM, Petersen J, Kuo P, Valantine V, Robbins C, Theodore J. The incidence of renal failure

in one hundred consecutive heart-lung transplant recipients. Am J Kidney Dis 1995; 26: 643-8.

3 Klompmaker IJ, Homan van der Heide JJ, Tegzess AM et al. Effects of cyclosporin. A withdrawal

on renal function and renal stimulation in liver transplant patients treated with triple drug

immunosuppression for over two years. Nephrol Dial Transplant 1994; 9: 1629-33.

4 Navis GJ, Broekroelofs J, Mannes GPM, Van der Bij W, Tegzess AM, De Jong PE. Renal

haemodynamics after lung transplantation: a prospective study. Transplantation 1996; 61: 1600-5.

5 Myers BD, Ross J, Newton L, Luetscher J, Perloth M. Cyclosporin-associated chronic

nephropathy. N Engl J Med 1984; 311: 699-705.

6 Abrass CK, Berfield AK, Stehman-Breen C, Alpers CE, Davis CL. Unique changes in interstitial

extracellular matrix composition are associated with rejection and cyclosporine toxicity in human

renal allograft biopsies. Am J Kidney Dis 1999; 33: 11-20.

7 Opelz G, Wujciak T, Ritz E. Association of chronic kidney graft failure with recipient blood

pressure. Kidney Int 1998; 53: 217-22.

8 Spencer CM, Goa KL, Gillis JC. Tacrolimus. An update of pharmacology and clinical efficacy in

the management of organ transplantation. Drugs 1997; 54: 925-75.

9 Broekroelofs J, Navis GJ, Stegeman CA, Van der Bij W, De Jong PE. Lung transplantation [letter].

Lancet 1998; 351: 1064.

10 Chan C, Maurer J, Cardella C, Cattran D, Pei Y. A randomised controlled trial of verapamil on

cyclosporine nephrotoxicity in heart and lung transplant recipients. Transplantation 1997; 63:

1435-40.

11 Broekroelofs J, De Haan J, Stegeman CA, Navis GJ, De Zeeuw D, De Jong PE. Validity of slopes

calculated from reciprocal serum creatinine (1000/Cr) or Cockroft clearance (ClCockr) in detecting

rate of renal function loss after lung transplantation [abstract]. J Am Soc Nephrol 1998; 9: 68.

CHAPTER 1

21

12 Myers BD, Sibley R, Newton L, Tomlanovich SJ et al. The long-term of cyclosporin-associated

chronic nephropathy. Kidney Int 1988; 33: 590-600.

13 Isoniemi H, Taskinen E, Hayry P. Histological chronic allograft damage index accurately predicts

chronic renal allograft rejection. Transplantation 1994; 58: 1195-8.

22


Chapter 2

Creatinine-based estimation of rate of long-term renal function lossin lung transplant recipients. Which method is preferable?

Broekroelofs J, Stegeman CA, Navis GJ, De Haan J, Van der Bij W, De Boer WJ, De Zeeuw D, De Jong PEJ Heart Lung Transplant 2000; 19: 256-62

Abstract

Progressive renal function loss during long-term follow up is common after lungtransplantation and close monitoring is warranted. Since changes in creatinine generationand excretion may occur after lung transplantation, the reliability of creatinine-basedmethods of renal function assessment to serial measurements of glomerular filtration rate(GFR) were compared in this population.Renal function with serial measurements of GFR by iothalamate clearance every 6 monthsafter transplantation was studied in a cohort of 40 lung transplant recipients with at least 24months of follow up, transplanted between November 1990 and October 1995 in this centre.The correlation between the rate of renal function loss calculated from the slope of GFR andthe following creatinine-based indices: the reciprocal of serum creatinine (1/Screatinine),Cockcroft clearance and Levey estimation were analysed. The absolute difference betweenGFR and Cockcroft clearance and Levey estimation pre- and at several points post-transplantation was also studied. The slopes of 1/Screatinine (r=0.85), Cockcroft clearance (r=0.86), and the Levey estimation(r=0.84) correlated significantly with the slope of GFR as measured by iothalamateclearance. However, all creatinine-based slopes underestimate the rate of GFR loss.Cockcroft clearance and the reciprocal value of Screatinine do not detect small GFR losses.During long-term follow up a time-dependent discrepancy between Cockcroft clearance,Levey estimation and GFR was observed which may partially explain the observations forthis population.Creatinine-based slopes correlate with GFR slopes after lung transplantation, butconsistently underestimate the rate of GFR decline. The Levey estimation is the mostsensitive method to detect small GFR losses and may be preferable when no GFRmeasurement is available. In special conditions when an accurate renal function assessmentis needed true GFR may be necessary.

CHAPTER 2

25

Introduction

As progressive renal function loss is common after heart- and lung transplantation1,2 closemonitoring of renal function is warranted. The standard used to monitor chronic renalfunction loss is serial measurement of glomerular filtration rate (GFR)3. In clinical practice,however, this is not always feasible. It would be convenient to rely on renal functionestimates based on Screatinine only. Screatinine depends not only on GFR but also on muscle mass,and renal tubular creatinine secretion induces additional bias. To account for the resultingerror several correction formulas such as the Cockcroft-Gault formula4 and the recentestimation developed by Levey5, are available. These formulas were empirically derived,and validated mostly in cross-sectional studies in renal patients.Whether the assumptions on body mass composition and tubular creatinine secretion thatunderlie these estimations similarly apply to lung transplant recipients is unknown. Specificalterations of tubular creatinine secretion, due to drug use6 or the underlying disease7, arelikely to occur in this population. Their reliability for longitudinal renal follow up,moreover, depends on the assumptions that body mass composition and creatinine secretiondo not change over time8,9,10.The reliability of Screatinine-based estimates of long-term renal loss by comparing theseestimates with serial GFR measurements by iothalamate clearance was investigated in lungtransplant recipients. For long-term renal monitoring, calculation of the slope of renalfunction loss from repeated measurements has the advantage of circumvening the bias ofshort-term fluctuations and providing an estimate of renal prognosis. Thus, the slopes ofmeasured GFR over time was compared with the slopes derived from 1/Screatinine, theCockcroft-Gault formula and the Levey formula, respectively, in a closely monitored cohortof lung transplant recipients with at least 24 months of follow up. For all three Screatinine-basedmethods the agreement as to the rate of renal function loss was assessed, as well as thesensitivity for detection of small renal function loss. In addition, the Cockcroft-Gaultclearance, the GFR estimated from the Levey equation and the GFR measured byiothalamate clearance were compared to see if the relations remained constant over time.

Patients and Methods

Patients

Consecutive patients, receiving a bi- or unilateral lung transplant between November 1990and October 1995, with at least 24 months of follow up and at least 4 GFR measurementsfrom 6 months onwards were included in this study. The latter criterion required for anaccurate calculation of the slopes of GFR, 1/Screatinine, Cockcroft-Gault estimated creatinine

26

CREATININE-BASED ESTIMATION

clearance (Cockcroft clearance) and GFR estimated according to the Levey equation (Leveyestimation) over time. Of seventy patients transplanted in this period, 40 patients fulfilledthese criteria. Thirty patients were excluded due to patient death (n=20) or re-transplantation(n=2) within 24 months after transplantation or for lacking GFR measurements (n=8).

Methods

Immunosuppressive treatment is described in detail elsewhere1. Briefly, induction therapywas given for 7 days with rabbit anti-human thymocyte globulin (RIVM, Bilthoven, TheNetherlands). Maintenance immunosuppression consisted of a triple-drug regimen withcyclosporin A (CyA) (Sandimmune or Neoral), azathioprine, and corticosteroids, startingduring transplantation. The target CyA serum trough level was initially 400 µg/l tapering to150 µg/l, which was considered the lowest acceptable level. All patients received post-operative antibiotic prophylaxis with ceftazidime. Antibiotics were changed, if necessary, inaccordance with sputum or other cultures. Aminoglycosides were used when necessary,under close monitoring of serum levels. Prophylaxis for herpes infections consisted of oralaciclovir (4x200 mg) during the first 6 months, and prophylaxis for Pneumocystis cariniiconsisted of oral co-trimoxazole 800/160 mg every other day for the lifetime of the patient.Renal function studies were performed in all patients during the pre-transplant workup andwere repeated every 6 months after transplantation. Screatinine was measured (SMA(C)autoanalyser; Technicon Instruments, Tarrytown, NY, USA) on the same day as GFR. Fromthese values 1/Screatinine was calculated and creatinine clearance was estimated according to theformula of Cockcroft and Gault (Cockcroft clearance)4:

estimated creatinine clearance (ml/min) = (140 – age) x bodyweight (kg)Screatinine (mg/dl) x 72

(for women the outcome is multiplied by 0.85)

Screatinine was also used to predict GFR with the MDRD (Modification of Diet in RenalDisease) study equation 7 (Levey estimation)5:

estimated GFR (ml/min per 1.73 m2 = 170 x SCr(mg/dl)-0.999 x age-0.176 x SU(mg/dl)

-0.170 x SAlb(g/dl)0.318

(for women the outcome is multiplied by 0.762 and for black persons with 1.180)

GFR was measured as the urinary clearance of constantly infused 125I-iothalamate.Simultaneous measurement of 131I-hippuran clearance allowed to correct for errors inducedby incomplete bladder emptying as described previously. GFR measurements performed in

CHAPTER 2

27

this way have a variation coefficient of only 2.2 %, which allows accurate follow up of renalfunction3. The measured GFR and the estimated Cockcroft clearance were corrected to 1.73m2 of body surface.

Statistical analysis

Data are presented as mean with standard deviation and where indicated with total range,unless stated otherwise. Differences in continuous variables between and within groupswere tested with the unpaired and paired Student’s t-test, respectively. The rate of long-termrenal function loss was defined as the individual slope over time of the change in GFR, ofl/Screatinine, of the Cockcroft clearance or of the Levey estimation, and calculated by leastsquares linear regression. Slopes were calculated, using all available data points, as of 6months post-transplant because a biphasic course of renal function in patients after lungtransplantation was found previously1. Correlation between the individual slopes over timeof GFR and l/Screatinine, of the Cockcroft formula and of the Levey estimation were tested withleast squares linear regression analysis. Regression coefficients are given with the 95%confidence interval (CI)11. To assess whether the relation between GFR, Cockcroft clearanceand the Levey estimation is constant over time, the absolute difference between GFR andCockcroft clearance and Levey estimation, respectively, was also studied pre- and at severaltime points post-transplantation.

Results

Patient characteristics are shown in table 1 (pag. 28) . Most patients were transplanted foremphysema, while a minority had pulmonary hypertension, cystic fibrosis or lung fibrosis asthe cause of respiratory failure. Renal function loss is most rapid in the first 6 months post-transplant as can be seen in figure 1 (pag. 29). After 6 months a more gradual fall is evidentlater on in most patients. The mean individual slope of GFR from 6 months aftertransplantation onwards was -7.3 ± 6.6 ml.min-1.yr-1 per 1.73 m2 (range, –22.8 to +6.1).Before transplantation Screatinine was 0.079 ± 0.018 mmol/l (range 0.050 to 0.133). Screatinine

increased to 0.120 ± 0.028 mmol/l (range 0.056 to 0.187) at 6 months, 0.137 ± 0.030 mmol/l(range 0.067 to 0.202) at 12 months, 0.148 ± 0.040 mmol/l (range 0.092 to 0.235) at 24months, 0.167 ± 0.061 mmol/l (range 0.086 to 0.296) at 36 months, and 0.175 ± 0.061mmol/l (range 0.097 to 0.309) at 48 months post-transplant.The mean slope of 1/Screatinine from 6 months after transplantation onwards was -0.53 ± 0.79 l.mmol-1.yr-1 (range, –2.58 to +1.32). Individual slopes of 1/Screatinine correlatedsignificantly with the individual slopes of GFR (r=0.86). The intercept with the horizontalaxis was –2.2 ml.min–1.yr–1. 1.73 m-2 (95% CI, –5.1 to –0.2) (figure 2A, pag. 30).

28


The mean slope of the Cockcroft clearance from 6 months after transplantation onwards was-4.1 ± 5.7 ml.min-1.yr-1.1.73 m-2 (range, –18.6 to +11.3). A significant correlation was presentbetween the individual slopes of the Cockcroft clearance and GFR (r=0.87). The interceptwith the horizontal axis was –1.8 ml.min–1.yr –1 per 1.73 m2 (95% CI, –4.5 to 0.3). Theregression line is not parallel to the line of identity, with a regression coefficient of 0.75(95% CI, 0.61 to 0.89), indicating a mean underestimation of 25% of the rate of renalfunction loss as measured by GFR slope over the whole range (figure 2B, pag. 30).The mean slope of the Levey estimation from 6 months after transplantation onwards was–4.3 ± 5.1 ml.min-1.yr-1.1.73 m-2 (range, –16.1 to +10.0) with a significant correlationbetween the individual slopes of the Levey estimation and GFR as well (r=0.84). Theintercept with the horizontal axis was –0.6 ml.min–1.yr–1.1.73 m-2 (95% CI, –3.5 to +1.2).

CHAPTER 2

29

Table 1 Patient characteristics in the group of 40 lung transplant recipients.

* p < 0.001 as compared to GFR** mean ± standard deviation (total range)

Number (n) 40Mean age (yr) 43 ± 10Female/male 17/23Pulmonary diagnosis• Pulmonary hypertension 6• Emphysema 25• Cystic fibrosis 6• Lung fibrosis 3Pre-transplantation*

• Serum creatinine (mmol.l-1) 0.078 ± 0.018(0.050 to 0.133)

• 1/Screatinine (l.mmol-1) 13.1 ± 3.0(7.5 to 20)

• GFR (ml.min-1.1.73m-2) 100 ± 22(65 to 171)

• Cockcroft clearance (ml.min-1.1.73m-2) 91 ± 21 *(64 to 152)

• Levey formula (ml.min-1.1.73m-2) 93 ± 22 *(60 to 156)

Follow up (months) ** 47 ± 15(24 to 78)

The regression line is not parallel to the line of identity, with a regression coefficient of 0.65(95% CI, 0.51 to 0.79), indicating a mean underestimation of 35% of the rate of renalfunction loss as measured by GFR slope over the whole range (figure 2C, pag. 30).The intercept with the horizontal axis of the regression lines of both the slopes of 1/Screatinine (figure 2A, pag. 30) and Cockcroft clearance (figure 2B, pag. 30) indicates thatthese methods have a higher threshold for detecting GFR decline than iothalamate clearance.The regression line of the slopes of the Levey estimation intercepts the horizontal axis moreclosely to the origin (figure 2C, pag. 30), indicating that this method is somewhat moresensitive than the other creatinine-based estimates in the detection of small losses in GFR.Both the correlations between 3 creatinine-based slopes and the slopes of measured GFR

30


Figure 1 Mean glomerular filtration rate GFR (ml/min/1.73m2) (A), reciprocal value of Screatinine (l/mmol)

(B), creatinine clearance estimated according to the Cockcroft-Gault formula (ml/min/1.73m2)

(C), and GFR estimated according to the Levey equation (ml/min/1.73m2) (D) pre- and post-

transplantation in 40 lung transplant recipients. The error bars indicate ± 1 standard deviation.

The numbers above the horizontal axis indicate the number of patients with GFR measurements

at the time point.

CHAPTER 2

31

Figure 2 Correlation between slopes over time of measured glomerular filtration rate (GFR), reciprocal

value of Screatinine (A), slopes of Cockcroft-Gault estimated creatinine clearance (B), and slopes of

the GFR estimated according to the Levey equation (C) in 40 lung transplant recipients with at

least 4 measurements during follow up. The solid line indicates the calculated regression line,

the dotted line indicates the line of identity (B and C).

Figure 3 Mean difference (ml/min/1.73m2) between Cockcroft-Gault estimated creatinine clearance (A),

estimated GFR according to the Levey equation (B) and measured glomerular filtration rate

(ml/min/1.73m2) in 40 lung transplant recipients. Error bars indicate the standard error of the

mean differences.

and the intercepts with the horizontal axis were not different for male and female lungtransplant recipients (data not shown).The mean of the differences between Cockcroft clearance, GFR estimated from the Leveyequation, and GFR measured by iothalamate clearance before and at several time pointsafter transplantation is shown in figure 3. Remarkably, pre-transplantation and during thefirst year after lung transplantation Cockcroft clearance underestimates GFR, but duringfurther follow up it overestimates GFR (figure 3A). The GFR estimated from the Leveyequation underestimates true GFR both before and after transplantation. The magnitude ofthe underestimation tends to decrease over time (figure 3B). These results are independentfrom pulmonary diagnosis and gender (data not shown).

Discussion

The population of patients receiving a lung transplant is relatively small. In order to be ableto properly analyse the course of renal function in this population, renal function ismonitored by frequent 125I-iothalamate clearance measurements in this centre. Prior analysishas demonstrated that this allows accurate calculation of the rate of renal function loss inindividual patients, and alleviates the need for large populations to obtain sufficientstatistical power to detect differences3. Serial measurements of GFR showed a considerable and progressive renal function loss aftertransplantation in the patients. Overall, the slopes of the reciprocal value of Screatinine, ofCockcroft clearance and of the Levey estimation correlate reasonably well with the renalfunction slopes measured by serial iothalamate clearances. However, the creatinine-basedslopes underestimate the rate of GFR decline. First, the slopes of the reciprocal of Screatinine andof the Cockcroft clearance have a higher threshold to detect renal function decline than GFRmeasured by iothalamate clearance. A zero slope as calculated by these methods suggests astable renal function, but in fact corresponds to a GFR loss of approximately – 2ml.min–1.yr–1.1.73 m-2. Of the creatinine-based methods, the Levey method appears to havethe lowest threshold to detect GFR decline. Second, and more important from a clinicalperspective, the creatinine-based methods also underestimate the rate of GFR decline whenrenal function loss is more pronounced, i.e. by some 25% with the Cockcroft formula and by35% with the Levey equation. Furthermore, the large interindividual variability, asillustrated by the data in figure 2, leads to a wide confidence interval in the prediction of theGFR slope based on creatinine-based methods. Therefore, available creatinine-basedmethods to monitor renal function decline in this population suffer from considerable, andclinically relevant flaws.We found that the relation of iothalamate clearance with the Cockcroft clearance and theLevey-estimate of GFR was not constant over time. An overestimation of GFR by the

32


Cockcroft clearance occurred gradually with longer follow up. Likewise, theunderestimation of GFR by the Levey equation gradually diminished over time. Thesechanges over time could well explain the significantly lower slopes as compared to serialiothalamate clearances of renal function decline calculated by these creatinine-basedmethods. Decreased creatinine generation, increased tubular creatinine secretion, or bothover time could both cause the observed changes during long-term follow up10,12. Decreasedcreatinine generation over time could be due to loss of muscle mass as a result of long-termuse of corticosteroids. As we have no measurements of lean body mass, or of creatinineexcretion in our patients, however, we cannot substantiate this assumption. Increasedsecretion of creatinine has been reported with specific mutations (DF508) underlying cysticfibrosis7, but in our population the discrepancy was present irrespective of diagnosis. The use of cyclosporin A and co-trimoxazole in our population would be expected todecrease rather than increase tubular creatinine secretion. These medications wereunchanged during follow up and thus do not explain our observation. The relativecontribution of tubular secretion in renal creatinine clearance is known to increase withdeteriorating renal function10, which could explain the discrepancy over time. If so, onewould expect similar observations in other populations as well, but no systematical time-dependent difference between creatinine-based estimates of GFR and iothalamateclearance has been found until now12,13. Finally, the accuracy of the estimation of creatinineclearance or GFR from Screatinine depends on the assay used to measure Screatinine

14, which mayexplain differences between studies but not within one single centre.The Levey- and Cockcroft-Gault estimations of renal function use anthropometric and otherdata to estimate 24-hour creatinine generation without collecting 24-hour urine, in order toimprove the relation between individual Screatinine and renal function. Even without measured24-hour urine creatinine excretion the Levey estimation is claimed to provide a moreaccurate estimation of GFR than does measured creatinine clearance5. Both the Cockcroft-Gault formula and the Levey equation have been derived from cross-sectional data inpopulations that are grossly comparable to our lung transplant recipients with respect to age,gender and renal function. The Cockcroft-Gault formula modelled measured creatinineclearance from Screatinine and 24-hour urine creatinine excretion in a large data set obtained inmales and females with different levels of renal function and body composition as a functionof anthropometric characteristics and Screatinine only. The Levey estimation modelled GFR asmeasured by iothalamate clearance to a combination of anthropometric and laboratory datain male and female patients with renal disease and moderate degrees of renal functionimpairment. Both equations include a multiplication factor for females to correct the resultsfor the gender dependent muscle mass differences. Although our groups are small, the datado not suggest that the discrepancies found between measured and estimated GFR arecaused by a difference in only one of the sexes. For a proper interpretation of our findings it

CHAPTER 2

33

is important to realise that the population of lung transplant recipients is not only clearlydifferent from the populations studied by Cockcroft, Gault and Levey, but that the latterestimations were calculated from cross-sectional data only. Consequently, the modelsderived from these populations are not subject to bias due to changes in anthropometriccharacteristics and creatinine generation. In our population, however, a major change inclinical condition is elicited by the lung transplantation. Apparently, this results in an alteredrelationship between true and estimated GFR, before and after transplantation. Moreover,we show that during long-term follow up, with the patients in a relatively stable clinicalcondition, the relationship between estimated and true GFR is not constant over time. Themaximum follow up in our study was 48 months and it is unclear whether this timedependent effect on the relation between GFR, Cockcroft clearance and Levey formula hasalready levelled off at that time.What would be the implications of our findings? Clearly, monitoring of renal function byScreatinine only elicits a systematical underestimation of the rate of renal function loss in lungtransplant recipients. Measurement of Screatinine is a simple method for estimating renalfunction, but our findings demonstrate its limited reliability as a tool to monitor long-termrenal function loss in this population. From our data we cannot establish whether calculationof creatinine clearance from 24-hour urine might provide a better estimate, as this wouldonly correct for bias by decreased creatinine generation but not for bias by increased tubularcreatinine secretion. In clinical practice, accurate collection of 24-hour urine requires patientcompliance and introduces additional sources of error. In general, Screatinine-based methods tomonitor renal function during long-term follow up can only be reliable if the assumptions onthe relation between Screatinine and GFR are valid and do not change over time in the populationstudied. Our study shows that these criteria are not met in the population of lung transplantrecipients. Despite its costs serial GFR measurement is therefore preferable in observationaland intervention studies on renal function in these populations. Even in clinical care, underspecific circumstances GFR measurement may be necessary for correct evaluation of renalfunction. For instance when renal function is a determinant in acceptance for lungtransplantation or has major impact in the treatment during follow up.In conclusion, although slopes calculated from Screatinine-based estimates of renal functioncorrelate with slopes of measured GFR decline after lung transplantation, they consistentlyunderestimate the rate of GFR decline. This inaccuracy will have to be accounted for in thisspecific population when using creatinine-based monitoring.

34


References


haemodynamics after lung transplantation: a prospective study. Transplantation1996; 61: 1600-5.

2 Goldstein DJ, Zuech N, Sehgal V, Weinberg AD, Drusin R, Cohen D. Cyclosporin-associated

end-stage nephropathy after cardiac transplantation: incidence and prognosis. Transplantation

1997; 63: 664-8.

3 Apperloo AJ, De Zeeuw D, Donker AJM, De Jong PE. Precision of glomerular filtration rate

determinations for long-term slope calculation is improved by simultaneous infusion of 125I-iothalamate and 131I-hippuran. J Am Soc Nephrol 1989; 7: 567-71.

4 Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron

1976; 16: 31-41.

5 Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate method to estimate

glomerular filtration rate from serum creatinine: a new prediction equation. Ann Intern Med 1999;

130: 461-70.

6 Stegeman CA, Cohen Tervaert JW, De Jong PE, Kallenberg CGM. Trimethoprim-

sulfamethoxazole (co-trimoxazole) for the prevention of relapses of Wegener’s granulomatosis.

N Engl J Med 1996; 335: 16-20.

7 Windstetter D, Schaefer F, Schärer K et al. Renal function and renotropic effects of secretin in

cystic fibrosis. Eur J Med Res 1997; 2: 43-6.

8 Walser M. Assessing renal function from creatinine measurements in adults with chronic renal

failure. Am J Kidney Dis 1998; 32: 23-31.

9 Levey AS, Berg RL, Gassman JJ, Hall PM, Walker WG. Creatinine filtration, secretion and

excretion during progressive renal disease. Kidney Int 1989; 27: S73-S0.

10 Shemesh O, Golbetz H, Kriss JP, Myers BD. Limitations of creatinine as a filtration marker in

glomerulopathic patients. Kidney Int 1985; 28: 830-8.

11 Altman DG, Gardner MJ. Calculating confidence intervals for regression and correlation. Br Med

J 1988; 296: 1238-40.

CHAPTER 2

35

12 Walser M, Drew HH, LaFrance ND. Creatinine measurements often yield false estimates of

progression in chronic renal failure. Kidney Int 1988; 34: 412-8.

13 Bedros FV, Kasiske BL. Estimating glomerular filtration rate from serum creatinine in renal

transplant recipients. J Am Soc Nephrol 1998; 9: 666A.

14 Kemperman FA, Silberbusch J, Slaats EH et al. Glomerular filtration rate estimation from plasma

creatinine after inhibition of tubular secretion: relevance of the creatinine assay. Nephrol Dial

Transplant 1999; 14: 1247-51.

36


Chapter 3

Renal haemodynamics after lung transplantation:a prospective study

Navis GJ, Broekroelofs J, Mannes GPM, Van der Bij W, De Boer WJ, Tegzess AM, De Jong PETransplantation 1996; 61: 1600-5

Abstract

Renal function impairment is common after solid organ transplantation, due to thenephrotoxicity of cyclosporin. Moreover, in patients with severe respiratory failure, renalfunction is often impaired. This renal function impairment may predispose patients forfurther renal function impairment after lung transplantation. Therefore, renalhaemodynamics were measured in 44 patients prior to lung transplantation and 1, 6, 12, 18,24 and 30 months after transplantation. After transplantation, a decline in renal functionoccurred, with a progressive fall in glomerular filtration rate (GFR) of 33±4 % at 12 monthsand 46±9 % at 30 months. Effective renal blood flow (ERBF) fell by 22±5% at 12 monthsand remained stable thereafter. Changes in effective renal plasma flow (ERPF) were lesspronounced than those of ERBF, due to a fall in haematocrit after transplantation. Bloodpressure and renal vascular resistance increased significantly, consistent with the effects ofcyclosporin A. Prior to transplantation, renal function impairment with intense renalvasoconstriction had been found in a subset of the patients. Remarkably, the decrease inrenal function after transplantation was less pronounced in patients with renal functionimpairment prior to transplantation, as indicated by significant negative correlationsbetween pre-transplantation GFR and the percentage change in GFR after transplantation,and pre-transplantation ERPF and the percentage change in ERPF after transplantation. Thissuggests that the net course of renal haemodynamics after lung transplantation is the resultof the opposed effects of cyclosporin A nephrotoxicity and the favourable effects of thenormalisation of respiratory status. In conclusion, after lung transplantation a decline inrenal function occurs that is less pronounced in patients with renal function impairment andintense renal vasoconstriction prior to transplantation. Such a renal function impairment,therefore, should not be considered a contra-indication to lung transplantation.

CHAPTER 3

39

Introduction

Renal function impairment is common after solid organ transplantation. A considerable partof this morbidity can be attributed to the nephrotoxicity of the immunosuppressive agentcyclosporin A (CyA), as has been demonstrated both in renal transplant recipients and inpatients with normal kidneys, such as heart and heart-lung transplant recipients and CyA-treated patients with multiple sclerosis1,2,3,4,5. Risk factors predisposing for renalfunction impairment after transplantation are elevated CyA trough levels, the use ofconcomitant nephrotoxic drugs, and, in renal transplantation, the number of rejectionepisodes, renal hypoperfusion and primary poor renal function6,7. In lung transplant recipients, a rise in serum creatinine has been reported aftertransplantation8. No detailed data on renal function, however, have been reported yet. This isall the more important because, in patients with respiratory failure, renal perfusion isfrequently considerably impaired, even before transplantation9,10,11,12. Such a renalhypoperfusion could predispose for renal function loss after lung transplantation. Thus far,no data are available to test this assumption. Therefore, we included measurements of renalhaemodynamics in the pre-operative workup and in the follow up of patients who underwenta lung transplantation. In this article, we report on the course of renal haemodynamics afterlung transplantation in 44 patients with a follow up of 1-30 months.

Patients and methods

Patients

Between November 1991 and September 1994 49 lung transplantations were performed inour centre, one of which was a retransplantation. Patients were eligible for transplantation ifthey had severe pulmonary disease with substantial limitation in activities of daily living, inwhich medical therapy was ineffective or unavailable. Life expectancy had to be less than12-18 months.No pretransplantation renal function measurements were available in two patients; thereforethese patients were not included in this study. Two patients died due to primary graft failure;because of the lack of renal follow up these patients were not included either. Of theremaining 44 patients, six patients died from pulmonary or cardiac causes after 6, 11, 12, 13,14 and 15 months. Of the 44 patients evaluated 27 were male and 17 female. Mean age was41±2 years. Diagnoses were primary pulmonary hypertension (n=6), cystic fibrosis (n=12),emphysema (n=19, of which 12 cases were due to alpha-1-antitrypsin deficiency); fibrosis(n=4), and sarcoidosis, bronchiectasias and histiocytosis X (each n=1). With the exception ofone patient who had an asymptomatic proteinuria of 2 g/24h, all patients had a normal

40

RENAL HAEMODYNAMICS AFTER LUNG TRANSPLANTATION

urinalysis and none of the patients had any signs of overt renal disease prior totransplantation. Of the patients with cystic fibrosis, three had insulin-dependent diabetesmellitus and four had impaired glucose tolerance not requiring therapy. Albuminuria was notpresent in any of these patients. Ten patients received a single lung transplant; the others hada bilateral lung transplant.

Methods

Induction therapy was given during the first 7 days after transplantation with rabbit anti-human thymocyte globulin (Anti-Humaan Thymocyten Globuline, RIVM, Bilthoven, The Netherlands). Immunosuppression consisted of a triple-drug protocol of CyA(Sandimmune), azathioprine and corticosteroids, starting day 1. During the first three weekstarget CyA serum level was 400 ug/l; thereafter, 150 ug/l was considered the lowestacceptable level. CyA trough levels were measured by high-performance liquidchromatography daily in the postoperative period and at each outpatient visit during followup. Episodes of rejection were treated with intravenous bolus therapy methylprednisolone(Solumedrol, Upjohn, Kalamazoo; 500-1000 mg, depending on body weight) daily for threedays. All patients received antibiotic prophylaxis, normally ceftazidime, during the first 48hours after transplantation. Antibiotics could be changed in accord with sputum cultures:aminoglycosides were added only when necessary, under close monitoring of serum druglevels. Prophylaxis for herpes infections consisted of aciclovir during the first 6 months andprophylaxis for Pneumocystis carinii consisted of 960 mg co-trimoxazole every other day,lifelong. Renal function studies were performed during the workup for transplantation. Post-transplant renal function studies were done 1 month after surgery if the clinicalcondition was stable. If not, the measurement was postponed until a stable condition hadbeen reached. Follow up measurements were done 6 months after transplantation andsubsequently at 6-month intervals. Renal function was measured as the urinary clearance ofcontinuously infused 131I-Hippuran and 125I-iothalamate, reflecting effective renal plasmaflow (ERPF) and glomerular filtration rate (GFR), respectively13. Filtration fraction (FF) wascalculated as GFR/ERPF, and effective renal blood flow (ERBF) was calculated as ERPF/(1-[haematocrit/100]). Renal vascular resistance (RVR) was calculated as (mean arterialpressure/ERBF)*80 and expressed in dyne.s.cm-5. Data were normalised for 1.73 m2 bodysurface area. Normal values for these parameters in our centre, derived from measurementsin healthy volunteers, range from 320 to 680 ml/min/1.73m2 with a median value of 443ml/min/1.73m2, for ERPF and from 90 to 145 ml/min/1.73m2, with a median value of 109ml/min/1.73m2 for GFR. Median value for FF in healthy volunteers was 0.25 (range of 0.22-0.28). For RVR it was 10.2 dyne.s.cm-5 (range of 7.2-13.3 dyne.s.cm-5)14.

CHAPTER 3

41

Data handling

Data are presented as means ± SEM. Statistical evaluation was carried using the two-tailedWilcoxon test for paired data and by linear regression analysis. For comparison amongdifferent patient groups, analysis of variance was used, followed by Tukey-Kramer’smultiple comparison test.

Results

Data on the course of blood pressure, haematocrit and renal haemodynamics aftertransplantation are given in table 1 (pag. 43). GFR and FF both fell significantly. Thedecrease in GFR was already apparent 1 month after transplantation with a gradual furtherdecline during follow up. After 12-30 months, GFR was 30-40 % lower than beforetransplantation. Mean ERPF remained relatively unaffected; only at 12 months aftertransplantation a significant decrease was found. As haematocrit had fallen significantlyafter transplantation, we also calculated ERBF. ERBF had already significantly fallen at 1

42


Figure 1 Individual pretransplantation values for glomerular filtration rate (GFR), effective renal plasma

flow (ERPF), filtration fraction (FF) and renal vascular resistance (RVR). The bars indicate

range and median of these parameters in healthy volunteers. For each renal haemodynamic

parameter, patients are grouped according to pulmonary diagnosis, i.e. (from left to right): cystic

fibrosis (cf); miscellaneous (misc) and primary pulmonary hypertension (ph). Each renal

parameter was significantly different for the different diagnosis groups (p<0.001) (analysis of

variance), in the post-hoc comparison test, only FF in cystic fibrosis versus miscellaneous was

not significantly different.

month after transplantation, with a gradual further decline during follow up. Blood pressurerose after transplantation, with a mean increase of 14±4 % after 12 months. By then,antihypertensive treatment was needed in 9 out of 20 patients, whereas none of thesepatients had been hypertensive before transplantation. Consequent to the fall in ERBF andthe rise in blood pressure RVR increased significantly. Urinalysis remained normal in allpatients; in the patient with asymptomatic proteinuria, urinary protein loss disappeared aftertransplantation.To assess whether pre-existent renal function impairment predisposes for renal function lossafter transplantation, individual data were considered. Individual pretransplantation valuesfor ERPF, GFR, FF, and RVR are given in figure 1. GFR was below normal in 12 patients. Infive patients, GFR was elevated: three of these patients had insulin-dependent diabetesmellitus and two had impaired glucose tolerance, all due to cystic fibrosis. ERPF was belowthe normal range in 13 out of 44 patients. A FF above the normal range was found in 16patients, consistent with a state of renal vasoconstriction, as also evident from the RVR,which was elevated above the normal range in these patients. It is readily apparent fromfigure 1 (pag. 41) that renal functional parameters were not evenly distributed acrosspatients with different pulmonary disorders. Particularly in patients with pulmonaryhypertension, GFR tended to be impaired and there was a distinctly more severe impairmentin ERPF, and consequently an extremely elevated FF and RVR. In patients with cysticfibrosis, on the other hand, both GFR and ERPF were normal or supranormal, with a normalFF and RVR. The course of GFR for individual patients is given in figure 2 (left panel).Considerable interindividual differences are apparent in the course of GFR early aftertransplantation. One month after transplantation pronounced decreases in GFR were foundin patients with normal or elevated GFR prior to transplantation. In patients with renal

CHAPTER 3

Figure 2 The course of glomerular filtration rate (left panel) and of effective renal plasma flow (right

panel) in individual patients.

43

44


Tab

le 1

.M

ean

±SE

M o

f bl

ood

pres

sure

, hae

mat

ocri

t, r

enal

hae

mod

ynam

ics,

and

CsA

tro

ugh

leve

ls b

efor

etr

ansp

lant

atio

n an

d du

ring

fol

low

up

aM

AP,

mea

n ar

teri

al b

lood

pre

ssur

e; H

ct, h

aem

atoc

rit.

bP<

0.05

; pai

red

Wilc

oxon

test

ver

sus

pret

rans

plan

tatio

n va

lue.

c P<

0.01

; pai

red

Wilc

oxon

test

ver

sus

pret

rans

plan

tatio

n va

lue.

nM

APa

Hct

aG

FRE

RPF

ER

BF

FFR

VR

CsA

trou

ghle

vel

Pre-

Tx

4496

±1

42±1

105±

438

7±19

662±

300.

28±0

.01

12.9

±0.8

NA

1 m

onth

4410

0 ±

234

±185

±337

2±17

556±

240.

23±0

.01

15.7

±0.8

222±

11%

chan

ge6

±2b

-19±

2c-1

6±4c

2±5

-11±

4c-1

6±3c

30±6

c

6 m

onth

s30

103±

336

±177

±431

3±25

534±

280.

23±0

.01

17.1

±1.1

211±

20%

chan

ge8±

3b-1

5±3c

-26±

3c-5

±5-1

5±4c

-20±

3c41

±9c

12 m

onth

s20

107±

336

±166

±430

0±15

459±

240.

22±0

.01

20.1

±1.4

217±

11%

chan

ge14

±4c

-16±

3c-3

3±4c

-10±

5c-2

2±5c

-24±

4c60

±12c

18 m

onth

s15

97±3

36±1

66±5

294±

1445

8±33

0.22

±0.0

118

.4±1

.618

4±17

% c

hang

e3±

5-1

8±3c

-32±

5c-6

±8-1

5±8b

-27±

4c44

±19c

24 m

onth

s12

100±

437

±165

±529

0±16

466±

220.

23±0

.02

17.7

±1.2

241±

37%

cha

nge

3±4

-15±

3c-3

2±6c

-5±9

-11±

9-2

7±5c

27±1

1c

30 m

onth

s6

90±4

34±2

52±7

244±

2037

3±28

0.21

±0.0

320

.0±1

.826

6±46

%ch

ange

-4±5

-19±

5b-4

2±9b

-8±1

4-1

5±17

-39±

4c41

±2.4

c

function impairment prior to transplantation, on the other hand, the fall in GFR tends to bemodest or even absent at that time. From 6 months onwards, the course of GFR wass moreuniform, that is, a gradual decline. A statistically significant negative correlation was foundbetween pretransplantation GFR and the percentage change in GFR after transplantation atall measurements. Thus, the decline in GFR was less pronounced in patients with a low GFRprior to transplantation. This is illustrated in figure 3 for the data obtained 1 month (r=-0.53;p<0.001) and 24 months after transplantation (r=-0.58, p<0.05), respectively.The course of ERPF in individual patients is also shown in figure 2 (pag. 42) (right panel).Again, interindividual differences were most prominent early after transplantation. At thattime, increases in ERPF were found mainly in patients with a low ERPF prior totransplantation. From 6 months onward, the course of ERPF was more uniform in that ERPFtended to stabilise. A statistically significant negative correlation existed betweenpretransplantation ERPF with the percentage change in ERPF after transplantation at all

CHAPTER 3

45

Figure 3 Correlations between pretransplantation glomerular filtration rate and the percentage change in

glomerular filtration rate (left panels) and between pre transplantation effective renal plasma

flow (ERPF) and the percentage change in effective renal plasma flow (right panels) for the data

obtained 1 month and 24 months after transplantation.

measurements. Figure 3 illustrates this for the data after one month (r = -0.57; p<0.001) and24 months (r = -0.92; p<0.001). Apparently, the fall in ERPF was less pronounced or evenabsent in patients with a low ERPF prior to transplantation.In view of the pretransplantation differences in renal function between patients withdifferent pulmonary conditions, and the correlation between pretransplantation renalfunction and the course following transplantation, we also analysed our data to see whetherpulmonary diagnoses were related to the course of renal function after transplantation. Onemonth after transplantation the change in GFR was -28±7 % in cystic fibrosis patients(n=12), -17±4 % in patients with miscellaneous disorders (n=26), and 16±9 % in patientswith pulmonary hypertension. These changes were significantly different between thediagnosis groups (p<0.01). During further follow up, this trend persisted, but due to thesmall number of patients in the cystic fibrosis and the pulmonary hypertension groups, nomeaningful statistical comparison could be made.

Discussion

Following lung transplantation, we noted a decline in renal function, with a mean fall inGFR of 33±4 % below baseline after 12 months, and 42±9 percent after 30 months. Thisimpairment of GFR is in the same order of magnitude as has been reported for hearttransplant recipients and is somewhat larger than the fall in renal function in liver transplantrecipients in our centre15. Thus far, no patients experienced end-stage renal failure. Thecourse of GFR in individual patients as shown in figure 2 (pag. 42), however, gives reasonfor concern, as the decline in GFR tends to be progressive at least in some patients. Mean ERPF was relatively unaffected after transplantation. Only at 12 months aftertransplantation did the decrease reach statistical significance. At first glance, this relativelack of effect on ERPF may seem at variance with the well-known renal vasoconstrictiveaction of CyA. Calculation of ERBF and RVR, however, reveals a significant fall in ERBFand a considerable, statistically significant rise in RVR. The relative lack of effect on ERPF,therefore, should not be attributed to the absence of renal vasoconstrictive effects aftertransplantation, but to the fall in haematocrit, allowing for preservation of plasma flow inspite of a decrease in blood flow. The fall in haematocrit could well be explained by theimprovement in oxygenation after transplantation, but iatrogenic factors such as theazathioprine treatment and repeated blood sampling for diagnostic purposes may havecontributed as well. The overall long-term decline in renal function, with an increase in RVR and a rise in bloodpressure, is consistent with the well-known nephrotoxic effects of CyA. In this particularpopulation other potentially nephrotoxic drugs, such as co-trimoxazole and aciclovir in allpatients, and aminoglycosides and amphotericine-B in some, may also have played a role.

46


One of the purposes of our study was to assess whether renal function impairment prior tolung transplantation predisposes for increased renal function loss after transplantation.Remarkably, however, renal function loss after lung transplantation appeared to be lesspronounced in patients with renal function impairment prior to transplantation. To explain this seeming discrepancy, we considered the nature of the renal functionimpairment prior to transplantation. It was characterised by a disproportionate impairmentof renal perfusion, as indicated by a low ERPF, an elevated FF, and an elevated RVR. Suchrenal function impairment with intense renal vasoconstriction has been found in severalstudies in patients with respiratory failure9,10,11,12. Mechanisms thought to contribute to theserenal abnormalities are a low cardiac output, hypoxemia, acid-base disturbances,polycythemia, or a combination of those10,11,12, 16. Early studies suggest that the renalvasoconstriction reflects the degree of respiratory insufficiency9,12, becauss renal functionimpairment is more severe in patients with more severe respiratory failure and, moreover,renal function improves when respiratory status improves. It is remarkable, therefore, thatrenal function was relatively normal in most of our patients, despite the advanced state oftheir respiratory disorder. It should be noted, however, that our population differs from thosein other studies. Thus far, only patients with chronic obstructive pulmonary disease havebeen described, whereas a variety of pulmonary disorders is represented in our population.Our data on pretreatment renal function clearly demonstrate the heterogeneity of renalhaemodynamics across the different diagnosis groups. Moreover, our data suggest that thisheterogeneity may be relevant to the course of renal function after transplantation. Of our patients suffering from cystic fibrosis, three had diabetes mellitus and four hadimpaired glucose tolerance. In five of these subjects, GFR was elevated, consistent withdiabetic hyperfiltration. The more severe fall in GFR early after transplantation in thesepatients could be due to the vasoconstrictive effects of CyA unmasking diabetichyperfiltration, but it may also be related to the often more complicated postoperative coursein these patients and the more frequent requirement of nephrotoxic antibiotics. Our data asyet do not allow to differentiate between these factors. Long-term follow up, as well as anevaluation of nephrotoxic assaults, is needed for this particular group of patients. Adiagnosis of pulmonary hypertension, on the other hand, was associated with more severerenal function impairment before transplantation and a more favourable course aftertransplantation. In these patients generally right-sided heart failure is present, and theresulting forward failure can be expected to impair renal perfusion and elicit the state ofintense renal vasoconstriction found in these patients. Improvement of circulatory statusafter transplantation can be expected to have a favourable effect on renal function. In our patient population in general, presumably the net effect on renal haemodynamics inour patients is the result of the opposed effects of the nephrotoxicity of CyA (and othernephrotoxic drugs) and the favourable effects of the normalisation of respiratory status on

CHAPTER 3

47

renal function. The latter effect could be due to improved oxygenation, normalisation ofacid-base status and cardiac performance, the fall in haematocrit and to a combination ofthose factors 11,12. Such a beneficial effect of transplantation itself on renal function isstrongly suggested by the time course of the changes in renal haemodynamics aftertransplantation particularly by the fact that the relatively favourable course of renal functionin patients with impaired pretransplantation function is mainly due to the changes in GFRand ERPF early after transplantation, as evident from figures 2 (pag. 42). In interpreting our renal blood flow measurements, it should be kept in mind that CyA canlower renal extraction ratio of PAH-derived tracers3. Prior to transplantation, renal tracerextraction was presumably normal, as respiratory failure does not affect tracer extraction17.We did not measure renal extraction of 131I-Hippuran, however, because renal veincatheterisation is an invasive procedure. Our data, therefore, may to a certain degreeunderestimate true renal blood flow after transplantation, and overestimate RVR due to aCyA-induced decrease in renal extraction ratio of 131I-Hippuran.In conclusion, in recipients of a lung transplant, clinically relevant impairment of renalfunction associated with the development of hypertension occurs, consistent with thenephrotoxic effects of CyA. Further follow up is needed to delineate the long-term outcomein renal function in this population and its particular sub-populations. In patients withpretransplantation renal function impairment characterised by intense renalvasoconstriction, renal function loss following transplantation is less pronounced. Thus,such a renal function impairment should not be considered a contraindication fortransplantation.

Acknowledgements

Part of these data has been published in abstract form in JASN 5; p 401, 1994.We gratefully acknowledge the technical and secretarial assistance of Mrs M van Kammen, Mrs A Drent-Bremer and Mrs A Formsma.

48


References

1 Lewis RM, Van Buren CT, Radovancevic B, Frazier OH, Janney RP, Powers PL, Golden DL,

Giannakis JG, Macris MP, Kerman RH, Kahan BD. Impact of long-term cyclosporin

immunosuppressive therapy on native kidneys versus renal allografts: serial renal function in heart

and kidney transplant recipients. J Heart Lung Transplant 1991; 10: 63-70.

2 Modry DL, Oyer PE, Jamieson SW, Stinson EB, Baldwin JC, Reitz BA, Dawkins KD, McGregor

CGA, Hunt SA, Moram M, Myers B, Shumway NM. Cyclosporin in heart and heart-lung

transplantation. Can J Surg 1985; 28: 274-82.

3 Myers BD, Sibley R, Newton L, Tomlanovich SJ, Boshkos C, Stinson E, Luetscher JA, Whitney

DJ, Krasny D, Coplon NS, Perlroth MG. The long-term course of cyclosporin-associated chronic

nephropathy. Kidney Int 1988; 33: 590-600.

4 Bertani T, Ferrazi P, Schieppati A, Ruggenenti P, Gamba A, Parenzan L, Mecca G, Perico N,

Imberti O, Remuzzi A, Remuzzi G. Nature and extent of glomerular injury induced by cyclosporin

in heart transplant recipients. Kidney Int 1991; 40: 243-50.

5 Tegzess AM, Doorenbos BM, Minderhoud JM, Donker AJM. Prospective serial renal function studies

in patients with non-renal disease treated with cyclosporin A. Transplant Proc 1988; 20: 530-3.

6 Mihatsch MJ, Steiner K, Abeywickrama KH, Landmann J, Thiel G. Risk factors for the

development of chronic cyclosporin-nephrotoxicity. Clin Nephrol 1988; 29: 165-75.

7 Sabbatini M, De Nicola L, Sansone G, Conte G. Renal hypoperfusion as the primary cause of

cyclosporin-induced nephropathy. Nephrol Dial Transplant 1993; 8: 794-9.

8 Zaltzman JS, Pei Y, Maurer J, Patterson A, Cattran DC. Cyclosporin nephrotoxicity in lung

transplant recipients. Transplantation 1992; 54: 875-8.

9 Kilburn KH, Dowell AR. Renal function in respiratory failure. Arch Int Med 1971; 127: 754-62.

10 Farber MO, Roberts LR, Weinberger MH, Robertson GL, Fineberg NS, Manfredi F. Abnormalities

of sodium and H2O handling in chronic obstructive lung disease. Arch Int Med 1982; 142: 1326-30.

11 Wilcox CS, Payne J, Harrison BDW. Renal function in patients with chronic hyperaemia and cor

pulmonale following reversal of polycythemia. Nephron 1982; 30: 173-7.

CHAPTER 3

49

12 Aber GM, Bishop JM. Serial changes in renal function, arterial gas tension and the acid-base status

in patients with chronic bronchitis and oedema. Clin Sci 1965; 28: 511-25.

13 Donker AJM, Van der Hem GK, Sluiter WJ, Beekhuis H. A radio-isotope method for simultaneous

determination of the glomerular filtration rate and the effective renal plasma flow. Neth J Med

1977; 20: 97-103.

14 Ter Wee PM, Smit AJ, Rosman JB, Sluiter WJ, Donker AJM. Effect of intravenous infusion of

low-dose dopamine on renal function in normal individuals and in patients with renal disease. Am

J Nephrol 1986; 6: 42-6.

15 Klompmaker IJ, Homan van der Heide JJ, Tegzess AM, Meijer S, Haagsma EB, Verwer R, Slooff

MJH. Effects of cyclosporin A withdrawal on renal function and renal stimulation in liver

transplant patients treated with triple-drug immunosuppression for over 2 years. Nephrol Dial

Transplant 1994; 9: 1629-33.

16 Malizzia E. Renal function and haemodynamics in primary and secondary polycythemia. Acta Med

Scand 1956; 154: 399-406.

17 Brodwall EK, Storstein O, Skåland K, Vinje O. Renal oxygen consumption and medullary oxygen

tension in chronic hypoxia in man. Acta Med Scand 1971;190: 541-4.

50


Chapter 4

Long-term renal outcome after lung transplantation is predicted bythe 1 month post-operative renal function loss

Broekroelofs J, Navis GJ, Stegeman CA, Van der Bij W, De Boer WJ, De Zeeuw D, De Jong PETransplantation 2000; 69: 1624-8

Abstract

Progressive renal function loss is common after lung transplantation. To designrenoprotective strategies, identification of early predictors for long-term renal function losswould be useful.We prospectively analysed renal function (glomerular filtration rate(GFR); 125I-iothalamateclearance) in a closely monitored cohort of 57 lung transplant recipients with at least 24months of follow up transplanted between November 1990 and September 1996 in ourcentre. Analysed end points were the slope of GFR from 6 months post-transplant onwardsand the GFR at 24 months after transplantation.Before transplantation GFR was 100 ml/min (median, range 59 to 163). It decreased to 67ml/min (29 to 123) at six months, 53 ml/min (17 to 116) at 24 months and 51 ml/min (20 to87) at 36 months after transplantation. The magnitude of the loss of GFR 1 month posttransplantation was the only factor significantly correlated with absolute GFR at 24 monthsafter transplantation. Pulmonary diagnosis was significantly associated with long-term rateof renal function loss. Median loss of GFR was greatest in patients with cystic fibrosis (-10ml/min/yr, range –14 to –6 ml/min/yr), preserved in pulmonary hypertension (-1 ml/min/yr,range –6 to +7 ml/min/yr) and in between in emphysema (-6 ml/min/yr, range –27 to +12ml/min/yr). No other factors could be identified.In lung transplant recipients the 1 month post-operative loss of GFR is an early marker forlong-term renal prognosis. Pulmonary diagnosis appears to be a relevant predictor as well.These factors may guide further research and the development of preventive strategies.

CHAPTER 4

53

Introduction

Progressive renal function loss is an important complication of solid organ transplantation. Theseverity of long-term renal function loss is particularly prominent in heart and heartlungtransplant recipients, with end stage renal failure in up to 10 percent of the patients that survivefor five years or more after transplantation1. In liver transplant recipients renal function lossappears to be less prominent2. We previously reported considerable long-term renal functionloss in lung transplant recipients3. The interindividual difference in renal function loss waslarge, ranging from a slightly impaired but stable renal function in a minority of patients tosevere and rapid renal function loss in others. For the design of strategies for the prevention of long-term renal function loss it would beimportant to identify patients at high risk for long-term renal function loss at an early stage. Wepreviously found that interindividual differences in renal function loss were particularlymarked early after transplantation3.In the present study, therefore, we questioned whether early renal function loss could serve toidentify patients at risk for long-term renal function loss. To this purpose we analysed renalfunction data in a prospectively monitored cohort of lung transplant recipients transplantedbetween November 1990 and September 1996 in our centre. We previously found that pre-transplant renal function was influenced by pulmonary diagnosis3. To account for diagnosisrelated confounders, therefore, in the present study we analysed the course of renal function fordiagnosis groups separately.

Patients and Methods

Patients

Consecutive patients, receiving a bi- or unilateral lungtransplant, between November 1990 andSeptember 1996 with at least 24 months of follow up and at least 4 GFR measurements from 6months post-transplant onwards were included in this study. The latter criterion was taken toallow an accurate calculation of the GFR slope over time. Of ninety-two patients transplantedin this period, fifty-seven patients fulfilled these criteria. Twenty-nine died within 24 monthsafter transplantation and in six patients serial GFR measurements were not available due topatient refusal, inability to perform the GFR measurement or follow up at another hospital.

Methods

Induction therapy was given the first 7 days after transplantation with rabbit anti-humanthymocyte globulin (RIVM, Bilthoven, The Netherlands). Immunosuppression consisted of a

54

LONG TERM RENAL OUTCOME

triple-drug protocol of oral cyclosporin A (CyA), azathioprine, and corticosteroids, startingat day 1. During the first three weeks, the target CyA serum level was 400 ug/l; thereafter,150 ug/l was considered the lowest acceptable level. Initial CyA dosing schedules werebased on individual assessment of CyA kinetics that had been performed pre-transplantation.Based on these kinetic data, CyA was administered twice a day, but in all patients with cysticfibrosis three times a day. CyA trough levels were measured by high-performance liquidchromatography daily in the postoperative period and at each outpatient visit during followup. The corn-oil based soft gel cap formulation (Sandimmune, Novartis Pharma B.V.,Arnhem, The Netherlands) was used until September 1995. After September 1995 allpatients were started on or switched to the microemulsion formulation (Neoral, NovartisPharma B.V., Arnhem, The Netherlands). Rejection episodes were treated with intravenousbolus therapy methylprednisolone (500-1000 mg, depending on body weight; Solumedrol,Upjohn, Kalamazoo, MI) daily for 3 days. All patients received antibiotic prophylaxis withceftazidime. Antibiotics were changed, if necessary, in accordance with sputum or othercultures: aminoglycosides were added when necessary, under close monitoring of serumlevels. Prophylaxis for herpes infections consisted of oral aciclovir (4x200 mg) during thefirst 6 months, and prophylaxis for Pneumocystis carinii consisted of oral co-trimoxazole800/160 mg every other day (lifelong).Renal function studies were performed in all patients during the pre-transplant workup.Posttransplant renal function was measured 1 month after surgery if the clinical conditionwas stable. If not, the measurement was postponed until a stable condition had been reached.Follow up measurements were done every 6 months after transplantation. Glomerular filtration rate (GFR) was measured as the urinary clearance of constantlyinfused 125I-iothalamate. Simultaneous measurement of 131I-Hippuran clearance allows tocorrect for errors induced by incomplete bladder emptying as described previously4. GFRmeasurement performed in this way has a variation coefficient of only 2.2 %, which allowsaccurate follow up of renal function loss4.

Data analysis

Data are presented as median with ranges. Not only data for the cohort as a whole arepresented, but also data grouped according to diagnostic category; i.e. for pulmonaryhypertension, emphysema and cystic fibrosis separately. To identify patients at high risk forlong-term renal function loss we analysed for long-term renal function loss in a dual fashion,i.e. the rate of long-term GFR decline over time, and GFR at 24 months after transplantation,respectively. The rate of GFR decline was defined as the individual slope of GFR over time(ml/min/yr) as of 6 months after transplantation and calculated by least squares linearregression. By taking this starting point, the obtained data can be considered to reflect the

CHAPTER 4

55

process of long-term renal function loss devoid of bias by fluctuations by early, potentiallyreversible events. Early renal function loss was defined as the difference between the GFR atone month after transplantation and the pre-transplant GFR.During the follow up of this cohort, CyA treatment was switched from the corn-oil basedformulation to the microemulsion formulation. To analyse for possible effects of differencesin bioavailability, we analysed trough levels and oral dosage before and after conversion tothe microemulsion formulation of CyA in patients converted later than 6 months aftertransplantation. Furthermore, the rate of long-term GFR decline until conversion in patientson the corn oil based soft gel cap formulation for at least 24 months, was compared with therate of GFR decline in patients treated with the microemulsion formulation only.


Differences between or within groups were analysed by unpaired and paired Wilcoxon ranksum test, respectively. Three group comparisons were tested by non-parametric Kruskal-Wallis ANOVA followed by Dunn’s multiple comparison test. Correlations between GFR at24 months and GFR slope with possible predictive factors were tested by non-parametricSpearman’s rank correlation. All reported p-values are two-tailed. A two-tailed p-value < 0.05 was considered to indicate statistical significance.

Results

Patient characteristics are shown in table 1 (pag. 56), grouped according to pulmonarydiagnosis. Patients with cystic fibrosis were significantly younger than the others. Mostpatients received a bilateral transplantation, except for those with pulmonary hypertension inwhom all but two patients received a unilateral transplant. Cystic fibrosis patients had asignificantly higher pre-transplantation GFR than patients with pulmonary hypertension oremphysema. Prior to transplantation diabetes was present in 3 patients with cystic fibrosis.None of the patients had proteinuria. Median follow up was 48 months (range 24 to 84) andnot different between the diagnosis groups. Before transplantation GFR was 100 ml/min (59 to 163) in the group as a whole decreasingto 67 ml/min (29 to 123) at six months, 53 ml/min (17 to 116) at 24 months and 51 ml/min(20 to 87) at 36 months after transplantation. Figure 1 (pag. 56) shows the individual serialGFR measurements grouped according to pulmonary diagnosis. Renal function loss is mostrapid in the first six months after transplantation and there is large interindividual variabilityin the course of renal function. The absolute early renal function loss (at 1 month after transplantation) as well as the rate oflong-term renal function loss (ml/min/yr) from 6 months post-transplant onwards are shown

56


CHAPTER 4

57

Figure 1 Individual serial glomerular filtration rate (GFR) measurements prior to and after lung

transplantation in 57 patients grouped according to pulmonary diagnosis.

Table 1 Patient characteristics according to pulmonary diagnosis in 57 patients (median (range))

PH=pulmonary hypertension, E=emphysema, CF=cystic fibrosis# pre-transplantation * p<0.001 CF vs E and PH ** p<0.01 CF vs E and PH

PH E CFN 8 42 7Age 41 (29-53) 47 (26-64) 26 (20-31) *Female 5 15 3Unilateral transplant 6 4 0GFR in ml/min # 84 (72-124) 100 (59-143) 124 (96-163) **MAP in mmHg # 93 (61-100) 96 (77-117) 88 (87-98)Follow up in months 60 (36-84) 42 (24-84) 54 (36-78)

in figure 2, grouped according to pulmonary diagnosis. In cystic fibrosis patients, both theearly renal function decline and the long-term rate of renal function loss were morepronounced than in patients with emphysema or pulmonary hypertension. In patients withcystic fibrosis the early change was –50 ml/min (range, -22 to –61 ml/min) as compared to–29 ml/min (range, -78 to +23 ml/min) in emphysema and –10 ml/min (range, -27 to +31ml/min) in pulmonary hypertension. The rate of long-term GFR loss was greatest in patientswith cystic fibrosis (-10 ml/min/yr, range –14 to –6 ml/min/yr), and significantly different(p<0.01) from patients with pulmonary hypertension (-1 ml/min/yr, range –6 to +7ml/min/yr). In the emphysema patients the rate of GFR loss was inbetween (-6 ml/min/yr,range –27 to +12 ml/min/yr) and significantly different (p<0.05) from patients withpulmonary hypertension. Calculation of the GFR slopes was based on 7 (median, range 4 to14) serial GFR measurements; there was no difference between the diagnosis groups in thenumber of GFR measurements. Other than pulmonary diagnosis no factors were found to besignificantly associated with the rate of long-term GFR loss.Mean arterial blood pressure increased from 93 mmHg (range 61 to 100) pre-transplant to 98 mmHg (range 73 tot 130) and 101 mmHg (range 73 to 120) at 6 and 24 months post-transplant, respectively. Blood pressure was not different between the 3 pulmonarydiagnosis groups at any of these time points (data not shown).Trough CyA levels were not different for the three different diagnosis groups at 1, 6 and

58


Figure 2 Individual early glomerular filtration rate (GFR) change in ml/min (GFR change at 1 month

compared to pre-transplant GFR) (left) and GFR slope in ml/min/year calculated from 6 months

post-transplant onwards (right) in 57 patients according to pulmonary diagnosis.

(PH=pulmonary hypertension, E=emphysema, CF=cystic fibrosis)

24 months post-transplant, although oral CyA dose was higher at all time points of follow upfor patients with cystic fibrosis as compared to patients with emphysema and pulmonaryhypertension (table 2). As the oral CyA formulation was changed during the study period,the possible impact of this change on renal function was analysed. Twenty-six patients wereconverted after more than 6 months therapy on the corn oil based soft gel cap to themicroemulsion formulation. No difference was found between trough CyA levels before(176 ug/l; range 127 to 375) and after (170 ug/l; range 124 to 214) conversion tomicroemulsion formulation. The oral CyA dose was marginally but significantly lower(p=0.001) after conversion to the microemulsion formulation of CyA (4.0 mg/kg/day (2.2 to 10.9) before, and 3.8 mg/kg/day (2.0 to 8.4) after conversion), consistent with ahigher bioavailability of the microemulsion formulation. Eighteen patients were on the cornoil based soft gel cap formulation for at least 24 months. The GFR slope over the period oftime until conversion was –5.4 ml/min/yr (-17.2 to –2.0). In twenty five patients, treated bythe microemulsion form only, the GFR slope was similar, i.e -5.9 ml/min/yr (-26.8 to 11.6).There was no difference in early post-operative renal function loss between the two groupseither. Trough CyA levels and oral dose were also similar (data not shown).At 24 months after transplantation GFR was 52 ml/min (range; 26 to 67) in patients with

CHAPTER 4

59

Table 2 Oral cyclosporin A dose (mg/kg/day at 1, 6, and 24 months post-transplant)and trough cyclosporin A levels (median and range from the preceedingperiod in ug/l) according to pulmonary diagnosis at 1, 6, and 24 months afterlung transplantation in 57 patients. PH=pulmonary hypertension,E=emphysema, CF=cystic fibrosis

# median trough level in the preceding 5 months## median trough level in the preceding 6 monthsa p<0.001 compared to PH and E

1 month 6 months 24 monthsOral dose Level Oral dose Level Oral dose Level(mg/kg/d) (µg/l) (mg/kg/d) (µg/l)# (mg/kg/d) (µg/l)##

PH 4.2 273 3.3 181 3.4 168(3.3-6.8) (238-326) (2.6-5.8) (147-213) (1.4-7.4) (148-205)

E 4.5 288 4.1 174 3.7 160(0.9-9.5) (225-344) (0.9-7.1) (150-226) (0.6-7.0) (125-367)

CF 8.9a 314 9.8a 191 8.2a 179(4.8-21.8) (230-425) (5.4-11.1) (151-291) (4.5-10.9) (168-252)

cystic fibrosis, 51 ml/min (range; 17 to 116) in those with emphysema and 62 ml/min (range;50 to 84) in those with pulmonary hypertension and not significantly different between these3 diagnosis groups. Univariate analysis revealed that the magnitude of the early loss of GFR(from pre-transplant to 1 month after transplantation) was significantly correlated with GFRat 24 months after transplantation. Patients with the largest early post-operative fall in GFRhad the lowest absolute GFR at 24 months after transplantation (figure 3). No other factorspredictive of the GFR at 24 months were identified. Pre-transplant GFR was weaklyassociated with the GFR at 6 months post-transplantation (r=0.287, p=0.034), but not withGFR at 24 months post-transplantation (r=0.054, NS).

Discussion

The population of patients receiving a lung transplant is relatively small. Renal functionmonitoring by frequent 125I-iothalamate clearance measurements affords accurate calculationof the rate of renal function loss in individual patients with a low variation coefficient, thusallowing proper analysis despite the relatively small number of patients4. As expected,considerable renal function loss occurred after lung transplantation with a great between-patient variability and a clear biphasic course. Importantly, by univariate analysis long-termrenal prognosis could be predicted from the loss of GFR at one month post-transplantation.

60


Figure 3 Correlation between the early glomerular filtration rate (GFR) change in ml/min (GFR change at

1 month compared to pretransplant GFR) and the absolute GFR in ml/min at 24 months after

transplantation.

In addition, renal prognosis was different for patients with different underlying pulmonaryconditions.In lung transplant recipients, many factors may contribute to renal function loss. Pre-transplantation characteristics (pulmonary diagnosis and pre-transplant GFR) and short-termperi-operative factors (haemodynamic instability, type of surgery, use of nephrotoxicantibiotics, infectious complications and the use of CyA), as well as long-term factors (useof CyA and blood pressure) may be involved. Many factors are related to either the treatmentregimen (and thus can be modified if identified as deleterious), to medical complications ora combination of these. For such multifactorial conditions, a multivariate analysis to assessthe relative importance of different independent risk factors would be helpful to analyselong-term renal function loss after lung transplantation. The small number of recipients afterlung transplantation in our study, however, precluded such an analysis. By univariate analysis we found that the early GFR loss – but not pre-transplant GFR – is apredictor of renal function 2 years after transplantation. This finding not only allows earlyidentification of patients at high risk for renal function loss, but also suggests that early peri-operative renal damage bears lasting impact on long-term renal prognosis. In the firstmonth after transplantation the trough CyA levels were deliberately higher than afterwards.These high levels could contribute to the early renal function loss and suggest a criticalreappraisal of the necessity of these high target levels in the first month after transplantation.It is unknown however, whether an alternative CyA dosing regimen is feasible withoutlosing immunosuppressive potency.During the follow up of this patient cohort the formulation of CyA was altered. Theconversion to the microemulsion formulation of CyA might have impact on renal functionby improving gastrointestinal absorption and bioavailability of CyA5,6, by altering peak andor trough levels. Our data, however, do not allow to support an effect of conversion to themicroemulsion formulation on long-term GFR loss.We previously found that pre-transplant renal function was influenced by pulmonarydiagnosis3. To account for diagnosis related confounders, therefore, in the present study weanalysed the course of renal function for diagnosis groups separately. Interestingly, early aswell as long-term renal function loss were related to pulmonary diagnosis. Renal functionloss was largest in cystic fibrosis, smallest in pulmonary hypertension and in-between inemphysema, implicating that a global assessment of renal risk can already be made prior totransplantation. Whether these differences in renal prognosis are related to differences innephrotoxic insults between the groups or to other diagnosis-associated factors, cannot beascertained from our data. CyA trough levels were similar for the diagnosis groups.However, cystic fibrosis patients used higher oral CyA doses, presumably reflecting reducedbioavailability5,6. Moreover, cystic fibrosis patients used 3 in stead of 2 doses daily. Themore frequent and higher dosing required to obtain target trough serum levels may therefore

CHAPTER 4

61

have contributed to the worse renal prognosis in these patients7. More frequent exposure tonephrotoxic antibiotics, or specific renal vulnerability in cystic fibrosis may also have beeninvolved8,9,10,11. In pulmonary hypertension, on the other hand, renal prognosis appears to befavourable both on long-term and on short-term, as apparent from figure 1 (pag. 56). This isremarkable, considering the presence of considerable renal function impairment with intenserenal vasoconstriction prior to transplantation in these patients3. The favourable short-termoutcome in pulmonary hypertension might be explained by improvement of renal perfusionby the normalisation of cardiac output after transplantation. The reason for the long-termresistance against the combined nephrotoxic insults associated with lung transplantation,however, can not be inferred from our data. Notwithstanding the impact of diagnosis, in all three diagnosis groups the individualdifferences in the course of renal function was considerable, emphasizing the importance ofnon-diagnosis related factors.What are the implications of our findings? The first rough step to identify patients at risk forrenal function loss seems possible. Given the limitations of our study, a study with largernumbers of recipients will be needed to identify more specific predictors of renal functionloss after lung transplantation. Still, pending the results of such studies, it would beworthwhile to study renoprotective measures in high risk patients – such as rigourous bloodpressure control, avoidance of nephrotoxic antibiotics, more intensive monitoring and, ifpossible, tapering, of CyA levels – in the population after lung transplant recipients12,13. Theimpact of early renal function loss on long-term renal prognosis implicates thatidentification and possible modification of peri-operative nephrotoxic insults may have thepotential to improve long-term renal function. In conclusion, in lung transplant recipients the early post-operative GFR loss is an earlypredictor of a high risk for long-term renal function loss. Diagnosis appears to be related tolong-term outcome as well. These factors might guide further research in to mechanisms ofrenal damage and thus for the development of preventive strategies in high risk patients.

62


References

1 Goldstein DJ, Zuech N, Sehgal V, Weinberg AD, Drusin R, Cohen D. Cyclosporin-associated

end-stage nephropathy after cardiac transplantation: incidence and prognosis. Transplantation

1997; 63: 664-8.

2 Klompmaker IJ, Homan van der Heide JJ, Tegzess AM, Slooff A. Effects of cyclosporin A

withdrawal on renal function and renal stimulation in liver transplant patients treated with triple

drug immunosuppression for over two years. Nephrol Dial Transpl 1994; 9: 1629-33.



4 Apperloo AJ, De Zeeuw D, Donker AJM, De Jong PE. Precision of glomerular filtration

rate determinations for long-term slope calculation is improved by simultaneous infusion of

125I-iothalamate and 131I-hippuran. J Am Soc Nephrol 1996; 7: 567-71.

5 Cooney CF, Fiel SB, Shaw LM, Cavarocchi NC. Cyclosporine bioavailability in heart-lung

transplant candidates with cystic fibrosis. Transplantation 1990; 49: 821-3.

6 Tan KKC, Hue KL, Strickland S, Trull AK, Smyth RL, Scott JP, Kelman AW, Whiting B,

Higenbottam TW, Wallwork J. Altered pharmokinetics of cyclosporin in heart-lung transplant

recipients with cystic fibrosis. Ther Drug Monit 1990; 12: 520-4.

7 Fisher NC, Nightingale PG, Gunson BK, Lipkin GW, Neuberger JM. Chronic renal failure

following liver transplantation: a retrospective analysis. Transplantation 1998; 66(1): 59-66.

8 Katz SM, Krueger LJ, Falkner B. Microscopic nephrocalcinosis in cystic fibrosis. N Engl J Med

1988; 319: 263-6.

9 Davis CA, Abramowsky CR, Swinehart G. Circulating immune complexes and the nephropathy of

cystic fibrosis. Hum Pathol 1984; 15: 244-7.

10 Abramowsky CR, Swinehart GL. The nephropathy of cystic fibrosis: a human model of chronic

nephrotoxicity. Hum Pathol 1982; 13: 934-9.

11 Strandvik B, Berg U, Kallner A, Kusoffsky B. Effect on renal function of essential fatty acid

supplementation in cystic fibrosis. J Pediatr 1989; 115: 242-50.

CHAPTER 4

63

12 Chan C, Maurer J, Cardella C, Cattran D, Pei Y. A randomized controlled trial of verapamil

on cyclosporine nephrotoxicity in heart and lung transplantation recipients. Transplantation 1997;

63: 1435-40.

13 Navis GJ, de Jong PE, De Zeeuw D. ACE-inhibitors: panacea for progressive renal disease? Lancet

1997; 349: 1852-3.

64


Chapter 5

Early renal function loss after lung transplantation: analysis of peri-operative risk factors

Broekroelofs J, Loef BG, Stegeman CA, Navis GJ, Epema AH, Van der Bij W, De Boer WJ,De Zeeuw D, De Jong PESubmitted

Abstract

Early renal function loss after lung transplantation is considerable and exerts a negativelasting impact on long-term renal prognosis. Therefore, identification of risk factors for thisearly renal function loss would be useful.We prospectively analysed the change in GFR (125I-iothalamate clearance) in 83 lungtransplant recipients with at least 1 month follow up. The change in GFR at 1 month wascorrelated with baseline patient-, transplant-,operation-, and early post-operativecharacteristics.Before transplantation GFR was 102 ± 23 ml/min in the group as a whole, decreasing to 79± 24 ml/min at 1 month post-transplantation. The GFR change in the first month was –23 ± 26 ml/min. Effective renal plasma flow (ERPF) and filtration fraction (FF) alsoshowed a decrease at 1 month after transplantation. A large variability between pulmonarydiagnosis groups and individual patients was observed. GFR loss was greatest in patientswith cystic fibrosis (-37 ± 26 ml/min) and smallest in patients with pulmonary hypertension(-3 ± 22 ml/min). With multivariate analysis a higher pre-transplant GFR, a bilateraltransplantation, post-transplant revision for bleeding, a diminished urinary production and amean arterial pressure < 70 mmHg during the first 24-hours post-transplantation wereassociated with the loss of GFR at 1 month after transplantation, while pulmonary diagnosiswas not.In lung transplant recipients risk factors for early renal function loss are related toperioperative haemodynamic parameters. These factors tend to cluster with certainpulmonary diagnoses.

CHAPTER 5

67

Introduction

Progressive renal function loss is common after lung transplantation1 and after other solidorgan transplantations2,3, causes considerable morbidity and may lead to renal failurerequiring renal replacement therapy. To develop strategies for the prevention of renalfunction loss, it is important that risk factors for renal function loss are identified4. Werecently reported that the magnitude of renal function loss at one month after transplantationis a major predictor of long-term renal outcome, thus early renal function loss exerts a largeand lasting negative influence on renal prognosis5. Therefore improvement of long-termrenal outcome may be possible by prevention of renal damage in the peri-operative and earlypost-operative phase.Early renal function loss differs considerably between individual patients. This variabilitymay be related to differences in pre-transplant renal function as well as in peri-operativefactors such as haemodynamics, cyclosporin A levels, and the use of other nephrotoxicagents. The purpose of the present study was to identify risk factors that explain thesedifferences in early renal function loss, in order to identify targets for future preventivemeasures. We prospectively analysed predictors for renal function loss at 1 month aftertransplantation in a closely monitored cohort of lung transplant recipients transplantedbetween November 1991 and September 1997 at our center.


Patients

Consecutive patients, accepted for transplantation after evaluation according to a stepwiseselection procedure6, receiving a bi- or unilateral lungtransplant between November 1991and September 1997 were included. Glomerular filtration rate measurements were routinelyperformed before and 1 month after lung transplantation. Of one-hundred and five patientstransplanted in this period, eighty-three patients were evaluated. Nine patients had diedwithin the first month after transplantation and in thirteen patients the pre- or 1-month posttransplantation GFR measurement had not been performed due to logistic or technicalinability. Patient records for the lung transplant recipients are kept and maintained accordingto a set protocol for the anaestesia period and for the ICU stay. Thus, a detailed database onthe peri-operative phase and the ICU stay is available. For the present study, variablespotentially relevant to renal function changes (table 1, pag. 68) were extracted from thisdatabase.

68

EARLY RENAL FUNCTION LOSS AFTER LUNG TRANSPLANTATION

Immunosuppression and infection prophylaxis

Induction therapy was given the first 7 days after transplantation with rabbit anti-humanthymocyte globulin (RIVM, Bilthoven, The Netherlands) to a maximum of 3 doses in thefirst postoperative week. Maintenance immunosuppression consisted of a triple-drugprotocol of cyclosporin A (CyA), azathioprine, and corticosteroids, starting at day 1. During

CHAPTER 5

69

Table 1 Factors analysed for their possible association with the change in glomerularfiltration rate (GFR) 1 month after lung transplantation as compared to pre-transplant GFR in 83 patients.

* present or absent† pulmonary hypertension, emphysema, cystic fibrosis or other‡ mean of all values during days 2 to 28

Patient characteristicsage (y)genderbodyweight (kg)ody mass index (kg/m2)pulmonary diagnosis†pre-transplant GFR (ml/min)pre-transplant ERPF (ml/min)

Operation characteristicsuni- or bilateral transplantoperation duration (min)extracorporeal oxygenation*perfusion duration (min)MAP < 50 mmHg (min)MAP < 70 mmHg (min)iuresis (ml/kg/h)PaO2 at end of operationFiO2 at end of operationPEEP at end of operationnumber of transfusions

Early post-operative (24 h)MAP < 70 mmHg (min)vasopressor support*number of transfusionsPaO2 at end of 24 h (kPa)FiO2 at end of 24hPEEP at end of 24h (cm H2O)diuresis (ml/kg/h)cardiac arrhytmias*

Follow up (day 2 to 28)length of ICU stay (d)days on mechanical ventilationotal dose of ATG (mg/kg)rejection episodes*CMV infection*aminoglycoside antibiotics*cyclosporin trough level (µg/L)re-thoracotomy*

the first three weeks, the target 12 hour trough CyA serum level was 400 ug/L; thereafter,150 ug/L was considered the lowest acceptable level. CyA was administered twice a day, butin all patients with cystic fibrosis three times a day. CyA trough levels were measured byhigh-performance liquid chromatography daily in the postoperative period and doses wereadjusted as needed. Rejection episodes were treated with intravenous bolus therapymethylprednisolone (500-1000 mg, depending on body weight) daily for 3 days. All patientsreceived antibiotic prophylaxis with ceftazidime. Antibiotics were changed, if necessary, inaccordance with sputum and other cultures: aminoglycosides were added when necessary,under close monitoring of serum peak and trough levels. Prophylaxis for herpes infectionsconsisted of oral aciclovir (4x200 mg), and prophylaxis for Pneumocystis carinii consistedof oral co-trimoxazole 800/160 mg every other day.

Renal function measurement

Renal function studies were routinely performed in all patients during the pre-transplantworkup from September 1989 onwards. Post transplant renal function was measured 1month after surgery if the clinical condition was stable. If not, the measurement waspostponed until a stable condition had been reached.Glomerular filtration rate (GFR) was measured as the urinary clearance of constantlyinfused 125I-iothalamate during 4 hours. Effective renal plasma flow (ERPF) wassimultaneously measured by 131I-hippuran clearance allowing to correct for errors induced byincomplete bladder emptying as described previously7. GFR measurement performed in thisway have a variation coefficient of 2.2 %. Filtration fraction (FF) was calculated as the ratioGFR/ERPF.

Surgical procedure and Intensive Care management

The surgical procedure was performed as previously described8,9. When two lungs wereimplanted, the bilateral sequential method was used. All patients with pulmonaryhypertension were transplanted using cardiopulmonary bypass. In the other patientscardiopulmonary bypass was initiated only in case of haemodynamic or respiratoryinstability. Anesthesia was performed according to a fixed protocol. After insertion of peripheralvenous and radial arterial cannulae under local analgesia, anesthesia was induced withsufentanil and midazolam. Tracheal intubation was performed with pancuronium. Aflowdirected pulmonary artery catheter was inserted into the right internal jugular vein.Anesthesia was maintained with sufentanil, midazolam and pancuronium.Postoperative the patients were treated according to a standardized protocol. All patients

70


received a low dose of dopamine as long as they were ventilated mechanically. Cardiac outputwas measured with a thermodilution pulmonary artery catheter. A cardiac index of >2,2 l/m2

and a mean arterial pressure > 70 mm Hg was attained with fluid resuscitation and inotropicsupport. Patients were sedated until haemodynamically and respiratory stable.


Data are presented as means ± standard deviation and where indicated with ranges. As we havepreviously found that both pre-transplant renal function and the course of renal function aftertransplantation are different for patients with different underlying pulmonary conditions, dataare presented for the whole cohort, but also for the different diagnostic categories, i.e.pulmonary hypertension, cystic fibrosis, emphysema, and a restgroup. The restgroup containspatients with lung fibrosis, sarcoidosis, re-transplantation due to bronchiolitis obliterans,histiocytosis or destroyed lungs after chemotherapy. Short term renal function loss was definedas the absolute change in glomerular filtration at 1 month post-transplantation as compared topre-transplantation GFR. Possible predictors for short term renal function loss were identifiedby univariate analysis. The factors analysed included baseline patient characteristics, operationcharacteristics, characteristics from the first 24-hours post-transplantation and during longerfollow up from day 2 to 28 (table 1, pag 68). Differences in continuous variables between fourgroups were tested with the Student-t test or the non-parametric Wilcoxon rank sum test, whereappropriate. Differences in continuous variables between more than two groups were tested byANOVA or the non parametric Kruskall Wallis test. Post tests were performed with theBonferroni method. Correlations between loss of GFR at 1 month post-transplantation withcontinuous variables were tested by non-parametric Spearman’s rank correlation. To assess therelative importance and independence of the different factors multivariate regression analysiswith loss of GFR at 1 month post-transplantation as the dependent factor was performed withthe variables with a p-value < 0.10 by univariate analysis. A stepwise backward procedure wasfollowed. A twotailed p-value < 0.05 was considered to indicate statistical significance.

Results

Pre-transplant and operation characteristics are shown in table 2, grouped according topulmonary diagnosis. Patients with cystic fibrosis were significantly younger compared to theemphysema patients and the patients in the restgroup and had a significantly higherpretransplant GFR and ERPF as compared to patients with pulmonary hypertension andemphysema. Patients with pulmonary hypertension had a significantly lower ERPF, and asignificantly higher FF compared to the other diagnosis groups. In all but one of the patientswith pulmonary hypertension a unilateral transplantation was performed while most other

CHAPTER 5

71

patients received a bilateral transplantation. Operation duration was significantly longer incystic fibrosis, as compared to pulmonary hypertension and emphysema patients. The time oncardiopulmonary bypass was also longer in cystic fibrosis patients as compared to emphysemapatients.

72


Table 2 Patient and operation characteristics according to pulmonary diagnosis in 83patients receiving lung transplantation

Pulmonary Emphysema Cystic fibrosis Restgrouphypertension (E) (CF) (R)(PH)

N 11 45 14 13Age (yr) 36 ± 10 49 ± 6 29 ± 7 ‡ 45 ± 12

(22-53) (35-62) (20-45) (23-64)Male/female 5/6 23/22 8/6 9/4Uni/bilateral 10/1 3/42 0/14 4/9GFR pre-transplant 81 ± 21 102 ± 17 121± 25 † 102 ±25(ml/min) (43-124) (66-150) (96-163) (59-143)ERPF pre-transplant 232 ± 56 ‡‡ 352 ± 70 480 ± 110 380 ± 108(ml/min) (155-330) (207-578) (328-695) (235-573)FF pre-transplant 0.35 ± 0.07 ‡‡ 0.29 ± 0.03 0.25 ± 0.02 0.27 ± 0.01

(0.24-0.46) (0.21-0.38) (0.22-0.29) (0.20-0.35)Time on perfusion (min) 166 ± 55 108 ± 94 215 ± 82 * 129 ± 106

(85-223) (0-339) (105-339) (0-255)Operation time (min) 324 ± 90 †† 322 ± 93 434 ± 98 ** 345 ± 82

(209-483) (160-592) (341-684) (220-502)Revision for bleeding 2 8 4 2Diuresis first 24 h (ml) 3735 ± 1582 3058 ± 1106 2813 ± 1468 3215 ±1343

(1830-7100) (1170-6415) (560-5060) (900-5545)MAP < 70 mmHg first 24 h 66 ± 58 59 ± 102 77 ± 146 47 ± 51(minutes) (0-195) (0-420) (0-540) (0-180)Cyclosporin level in 1st mo 286 ± 36 295 ± 34 315 ± 70 295 ± 33(µg/l) (238-353) (225-366) (230-425) (245-361)

* CF vs E** CF vs E† CF vs PH/E

p<0.01p<0.001p<0.05

†† PH vs CF‡ CF vs E/R‡‡ PH vs E/CF/R

p<0.05p<0.001p<0.001

Before transplantation GFR was 102 ± 23 ml/min (range; 43 to 163) in the group as a whole,decreasing to 79 ± 24 ml/min (range; 37 to 146) at 1 month post-transplantation. The GFRchange in the first month after transplantation was –23 ± 26 ml/min (range; -82 to +38), witha large variability between pulmonary diagnosis groups and between individual patientswithin the diagnosis groups (figure 1A). The loss of GFR at one month after transplantationis significantly greater in patients with cystic fibrosis (-37 ± 26 ml/min) as compared to theother diagnosis groups. At the other end of the spectrum patients with pulmonaryhypertension show almost no loss of GFR at one month after transplantation (-3 ± 22ml/min). Patients with emphysema and patients in the restgroup had losses of GFR at 1 month intermediate to the CF and PH patient. At 1 month after transplantation renalfunction was no longer different between the diagnosis groups (data not shown).ERPF was 362 ± 107 ml/min (range; 155 to 695) before transplantation in the group as awhole and decreased to 344 ± 95 ml/min (range; 180 to 677) at 1 month after transplantation.

CHAPTER 5

73

figure 1. Change in renal function and haemodynamic parameters at 1 month after lung transplantation in 83

lung transplant recipients grouped according to pulmonary diagnosis. The horizontal lines indicate

the group means.

In contrast to the other 3 groups, ERPF increased after transplantation in patients withpulmonary hypertension (figure 1B).The relative changes in GFR were greater than the relative changes in ERPF in allpulmonary diagnosis groups. As a result, FF decreased from 0.29 ± 0.05 (range; 0.20 to0.47) before transplantation to 0.23 ± 0.04 (range; 0.14 to 0.26) at 1 month aftertransplantation. The decrease in FF after transplantation was significant in all pulmonarydiagnosis groups, but the FF remained significantly higher in patients with pulmonaryhypertension compared to the other groups (figure 1C).To identify risk factors for the change in GFR at 1 month post-transplantation we firstperformed univariate analysis. The variables associated with the change in GFR at 1 monthpost-transplantation compared to pre-transplantation are listed in table 3 (p-value < 0.10). Ahigher pre-transplant GFR, a diagnosis of cystic fibrosis, a longer operation or ICU-stay anda diminished urinary production in the first 24-hours post-transplantation were significantlyassociated with a greater decline in renal function, as was a bilateral lungtransplant,

74


Table 3 Variables associated with the change in glomerular filtration rate (GFR) at 1 month posttransplantation as compared to pre-transplant GFR with a p-value < 0.10 by univariate analysis in 83 patients.

* data presented as mean ± SD (total range) for the group of 83 patients

Variable value p-valuepulmonary diagnosis pulmonary hypertension, 0.0067

emphysema,cystic fibrosis, otherpre-transplant GFR (ml/min)* 102 ± 23 (43 - 163) < 0.0001uni-/bilateral lungtransplant 17 / 66 0.0009operation duration (min)* 345 ± 99 ( 160 - 684) 0.006diuresis during operation (ml/kg/h)* 3.3 ± 2.1 (0.1 - 13.3) 0.049re-thoracotomy (yes/no) 16 / 67 0.007diuresis 24h post operative (ml/kg/h)* 2.1 ± 0.9 (0.5 - 5.2) 0.0011mean arterial pressure <70 mmHg 61± 99 (0 – 540) 0.10during 24h post operative (min)duration of ICU admission* 9 (5 - 54) 0.052rejection episode(s) (yes/no) 36 / 47 0.085infection episode(s) 24 / 59 0.069aminoglycoside antibiotics 18 / 65 0.016number of transfusions (units)* 8 ± 8 (0 - 55) 0.026

postoperative revision for bleeding, the use of aminoglycosides and a greater number oftransfusions. More episodes with rejection and infectious complications and a mean arterialpressure < 70 mmHg were also, although not significantly, associated with a greater declineof GFR at 1 month. Remarkably cyclosporin through levels in the first month aftertransplantation were not associated with GFR loss at 1 month after transplantation.To assess the independent risk factors for the loss of GFR at 1 month post-transplantation,multivariate regression analysis was performed. A higher pre-transplant GFR (p<0.001), a bilateral transplant (p=0.007), revision for bleeding (p=0.017), a diminished urinaryproduction (p=0.019) and a mean arterial pressure < 70 mmHg (p=0.059) during the 24-hours post-transplantation were identified as covariates with a p-value < 0.10. The modelincluding these covariates had an adjusted r2 of 0.39 (table 4).

Discussion

In accordance with our previous findings in a smaller number of patients, we foundconsiderable renal function loss in the first month after lung transplantation with a greatbetween-patient variability1,5. In the peri- and post-operative period in lung transplantrecipients many factors may contribute to renal function loss, including haemodynamicinstability, the use of cardiopulmonary bypass, high cyclosporin A levels and the use of othernephrotoxic medication. By univariate analysis we found several factors associated with theloss of GFR at 1 month after transplantation. Most of the factors identified point to

CHAPTER 5

75

Table 4 Variables associated with the change in glomerular filtration rate (GFR) at 1 month post-transplantation as compared to pre-transplant GFR with a p-value < 0.10 by multivariate regression analysis in 83 patients.*

* Adjusted R square for the model including these variables 0.391. Constant 43.7 (95%CI 12.8 to 74.7).

Variable (unit) p-value Coefficient (95% CI)pre-transplant GFR (ml/min) < 0.001 -0.39 (-0.60 to -0.18)bi-/unilateral lung transplant 0.007 -15.9 (-27.4 to -4.5)re-thoracotomy (yes/no) 0.017 -14.4 (-26.1 to -2.7)diuresis 24h post operative (ml/kg/h) 0.019 6.0 (1.0 to 11.0)mean arterial pressure <70 mmHg 0.059 -0.04 (-0.09 to 0.00)ring 24h post operative (min)

haemodynamic instability in the peri-operative phase –presumably associated with reducedrenal perfusion- as the common denominator. Using multivariate analysis pre-transplantGFR, a bilateral transplantation, revision for bleeding and a lower diuresis and mean arterialpressure in the first 24-hours were found to be independently associated with GFR loss at 1 month. Four of these 5 factors again are indicative for peri-operative haemodynamicinstability. So far, risk factors appear to be highly comparable to those in other populationsin the peri-operative setting. However, some remarkable discrepancies with otherpopulations undergoing major surgery are also apparent.In other populations it is usually prior renal function impairment, and not an elevated GFRthat predisposes to renal function loss after for example coronary artery bypass surgery10. Incontrast with patients undergoing coronary artery bypass surgery, in lung transplantrecipients a higher pre-transplant GFR is not a protection against post-transplant renalfunction loss. The heterogeneity in renal function characteristics in the lung transplantpopulation may be relevant in this respect. Differences in pre-transplant GFR in lungtransplant recipients may not merely reflect quantitative differences in prior renal damage.Cor pulmonale, which is evident especially in patients with pulmonary hypertension, isassociated with a serious compromised cardiac output pre-transplantation which elicitsintense renal vasoconstriction1. In this setting, lung transplantation clearly improves cardiacoutput and consequently renal perfusion as demonstrated by the increase in ERPF followinglung transplantation in patients with pulmonary hypertension. Therefore, renal dysfunctionprior to transplantation characterised by intense renal vasoconstriction should not beconsidered a contraindication for lung transplantation1. On the other hand the higher GFRpre-transplantation in cystic fibrosis patients may reflect hyperfiltration making thesepatients especially vulnerable for nephrotoxic insults.In accordance with our earlier studies, a significant difference was apparent betweendiagnosis groups in both pre-transplant renal function as well as post-operative renalfunction loss. The impact of diagnosis is clearly relevant from the point of view ofpretransplant assessment of renal risk. Moreover, it raises the question whether pulmonarydiagnosis as such is a risk factor for renal function loss, or, alternatively, whether diagnosisis a marker for a favourable or poor risk factor profile. Our present analysis finds a relationbetween pulmonary diagnosis and renal function loss on univariate-, but not on multivariateanalysis, supporting the assumption that pulmonary diagnosis is associated with clusteringof renal risk factors. We identified 5 independent risk factors; these were apparently notevenly distributed over the diagnosis groups. A bilateral transplant is independentlyassociated with loss of GFR at 1 month after transplantation. The majority of patients withcystic fibrosis and emphysema received a bilateral and all but one patient with pulmonaryhypertension a unilateral transplant. A bilateral transplant may be a more difficult procedurewhich could result in a longer operation duration and a longer time on perfusion, implicating

76


a greater risk for haemodynamic instability in these patients. A higher pre-transplant GFRwas mainly found in cystic fibrosis in whom it may reflect hyperfiltration. Although notsignificant, in patients with cystic fibrosis the duration of mean arterial pressure below 70mmHg in the first 24-hours post-operative was longer and in a greater percentage of thesepatients, as compared to the others, post-operative revision for bleeding was necessary.Probably as a consequence of the lower mean arterial pressure in the first 24-hourspostoperative diuresis was in cystic fibrosis also lower as well. These findings suggest thatdiagnosis as such is not a risk factor for GFR loss at 1 month after transplantation but thatclustering of diagnosis related risk factors is responsible for the relation of renal functionloss at 1 month after transplantation and pulmonary diagnosis. Remarkably, cyclosporin Atrough levels were not found to be associated with early renal function loss. The target levelin the first month after transplantation is 400 mg/l and the mean level in our population wasapproximately 300 mg/l irrespective diagnosis or the amount of early renal function loss.These high levels are certainly strongly responsible for severe renal function loss. However,the mean target level is also 300 mg/l in those patients who hardly show early renal functionloss or even improve after lung transplantation (data not shown). The latter situation mainlyoccurs in patients with pulmonary hypertension with a compromised renal function prior totransplantation and improvement of cardiac situation after a successful lung transplantation.This suggests that the net course of renal function impairment after lung transplantation isthe result of the opposed effects of cyclosporin nephrotoxicity and the favourable effects ofthe normalization of respiratory or cardiopulmonary status. Other nephrotoxic insults werealso not found to be independently associated with early renal function loss because they areprobably overruled by other more powerful haemodynamic parameters. On long-term follow up, as opposed to other non-transplant populations the peri-operativeloss of renal function seems irreversible on long-term evaluation5. This suggests that afterlung transplantation recovery of renal function is blocked and that the initial loss of functionis intensified. In this specific population nephrotoxic medication, i.c. cyclosporin, has to becontinued to preserve transplant function. Cyclosporin, therefore, may play an importantrole in this lack of improvement of renal function after the peri-operative phase.What are the implications? The factors associated with the loss of GFR at 1 month aftertransplantation are indicative of haemodynamic instability and as a consequence difficult tobe influenced. Trying to maintain the mean arterial pressure at a level of at least 70 mmHg inthe peri-operative phase could be a target in this specific population and for future research.However, development of pulmonary oedema is a constant source of concern in recipientsafter lung transplantation making it difficult to reach this goal. Haemodynamic instabilitymay cause a state of renal hypoperfusion. Until now the use of cyclosporin, underminingrenal perfusion, is inevitable. Postponing the introduction of cyclosporine until stabilisationof haemodynamics has been reached or tapering cyclosporine levels could be tools to

CHAPTER 5

77

78


prevent further reduction of renal perfusion. Whether these preventive measurements willpreserve renal function but also transplant function has still to be investigated. For thefuture, development of less nephrotoxic but also strong immunosuppressive drugs could betools to preserve renal function.In conclusion, risk factors for early renal function loss following lung transplantation arerelated to peri-operative haemodynamic parameters. These risk factors tend to cluster withcertain pulmonary diagnoses. Pre- and early post-transplant identification of these riskfactors and thereby patients at high risk for renal function loss seems possible with adequatemonitoring and may give opportunities for prevention of renal function loss associated withlung transplantation.

References




following liver transplantation: a retrospective analysis. Transplantation 1998; 15: 59-66.

3 Goldstein DJ, Zuech N, Seghal V, Weinberg AD, Drusin R, Cohen D. Cyclosporin-associated

endstage nephropathy after cardiac transplantation. Transplantation 1997; 63: 664-8.

4 Broekroelofs J, Stegeman CA, Navis GJ, De Jong PE. Prevention of renal function loss after

non-renal solid organ transplantation. How can nephrologists keep the kidneys out of the line of

fire? Nephr Dial Transplant 1999; 14: 1841-3.

5 Broekroelofs J, Navis GJ, Stegeman CA, Van der Bij W, De Boer W, De Zeeuw D, De Jong PE. A

poor long-term renal outcome after lung transplantation is predicted by the 1 month post-operative

renal function loss. Transplantation 2000; 69: 1624-8.

6 Mannes GP, De Boer WJ, Van der Bij W, Grevink RG, Koeter GH. Three hundred patients referred

for lung transplantation. Experiences of the Dutch Lung Transplantation Program. Chest 1996;

109: 408-13.


determinations for long-term slope calculation is improved by simultaneous infusion of 125I-iothalamate and 131I-hippuran. J Am Soc Nephrol 1996; 7: 567-72.

8 Pasque MK, Cooper JD, Kaiser LR. An improved technique for bilateral lung transplantaion. Ann

Thorac Surg 1990; 49: 785-91.

9 Cooper JD. The evolution of techniques and indications for lung transplantation. Ann Surg 1990;

212: 249-55.

10 Anderson RJ, O’Brien M, Mawwhinney S, Villanueava CB, Moritz TE, Sethi GK, Henderson WG,

Hammermeister KE, Grover FI, Shroyer L. Renal failure predisposes patients to adverse outcome

after coronary artery bypass surgery. Kidney International 1999; 55: 1057-62.

CHAPTER 5

79

Chapter 6

Risk factors for long-term renal survival after renal transplantation:a role for angiotensin-converting enzyme (insertion/deletion)

polymorphism?

Broekroelofs J, Stegeman CA, Navis GJ, Tegzess AM, De Zeeuw D, De Jong PEJ Amer Soc Nephrol 1998; 9: 2075-81

Abstract

Chronic progressive renal function loss is a main cause of long-term graft loss after initiallysuccessful renal transplantation. Transplanted kidneys share some risk factors for renalfunction loss, such as hypertension or proteinuria, with diseased native kidneys. Recently, ithas been shown that renal function loss is influenced by the angiotensin-converting enzyme(ACE)(insertion/detection [I/D]) genotype in renal disease in diseased native kidneys. Thisstudy examines whether donor or recipient ACE (I/D) genotype is a risk factor for graft lossafter renal transplantation.To avoid bias by acute events, graft survival was studied, with patients dying with afunctioning graft censored, starting at 12 months after transplantation in a cohort of 367patients transplanted between 1987 and 1994 with at least two years of follow up. Meanfollow up was 58 months. ACE (I/D) genotype was determined by PCR on stored donor andrecipient lymphocytes.Neither donor nor recipient ACE (I/D) genotype was associated with graft survival.However, Cox proportional hazards analysis identified recipient, but not donor, ACE (I/D)genotype D-allele to be independently associated with a shorter time to graft loss insubgroups of patients at high risk for graft loss defined by a creatinine clearance < 50 ml/min(n=108, p=0.017) or proteinuria ≥ 0.5 g/24h at 12 months (n=97, p=0.0051) aftertransplantation.In conclusion, recipient ACE (I/D) genotype was associated with time to graft loss in aspecific high risk subgroup of our population. This suggests that the effect of ACE (I/D)genotype on graft survival only becomes apparent when other risk factors aresimultaneously present.

CHAPTER 6

81

Introduction

Chronic renal function loss occurs in a substantial number of patients with an initiallysuccessful renal transplantation and is a main cause of graft loss during long-term follow up.Multiple risk factors have been identified, suggesting that the pathogenesis of thisprogressive renal function loss is multifactorial1,2. Some of these factors, such as HLA-mismatching and previous rejection episodes are specific for renal transplants, whereasothers, such as high blood pressure and proteinuria are risk factors common to both nativeand transplanted kidneys. In chronic renal disease in native kidneys, the insertion/deletion(I/D) polymorphism of the angiotensin-converting enzyme (ACE) gene has recently beenidentified as a risk factor for progressive renal function loss and decreased renalsurvival3,4,5,6,7. The association of the D-allele with an increased risk for renal failure in aspectrum of renal disorders of different origin suggests that the D-allele is a renal risk factorregardless of the primary cause of renal damage8.In the present study, therefore, the primary objective was to identify whether the ACE (I/D)genotype contributes to graft loss after renal transplantation. The study of thispolymorphism in renal transplantation provides the unique opportunity to investigatewhether the recipient or donor ACE (I/D) genotype is associated with renal risk. This mightallow us to distinguish between the influence of tissue and systemic ACE genotype and thuscould provide a clue to mechanism of action of ACE activity in renal disease progression9,10.To avoid bias induced by acute pathology, such as technical failure and therapy-resistantepisodes of acute rejection, the study involved a cohort of patients with a functioning graft12 months after transplantation. To account for the multifactorial nature of graft loss afterrenal transplantation, the influence of recipient or donor ACE (I/D) genotype on graftsurvival was assessed both by univariate and multivariate survival analysis including otheridentified risk factors for graft loss.


Patients

We retrospectively studied data from all patients transplanted with a cadaveric renal graftbetween April 1987 and December 1994 who had at least 12 months of follow up with afunctioning graft. Patients were included if stored lymphocytes of both donor and recipientwere available for ACE I/D genotype determination.Immunosuppressive treatment consisted of 3 mg/kg cyclosporin A intravenously for 72hours, followed by oral cyclosporin A (10 mg/kg) in 2 divided doses, with dose adjustmentsto obtain trough levels of 200 to 250 ng/ml for the first 3 months and 150 to 200 ng/ml

82

RISK FACTORS FOR LONG-TERM RENAL SURVIVAL AFTER RENAL TRANSPLANTATION

thereafter, and 20 mg/day prednisolone, tapered over 8 weeks to a maintenance dose of 10mg/day. Patients with anti-HLA antibodies (maximum panel reactivity > 60%) receivedinduction with murine monoclonal anti-CD3 antibody (OKT3, 5mg/day) for 12 daysfollowed by cyclosporin A started at day 10, azathioprine (1.5 mg/kg/day) and prednisolone(1 mg/kg/day initially and tapered to 10 mg/day after 8 weeks). Acute rejection episodesdiagnosed on clinical findings or by renal biopsy were treated with maximal two courses1000 mg of methylprednisolone intravenously for 3 consecutive days. Steroid-resistant oracute vascular rejection was treated with rabbit anti-thymocyte globulin (4 mg/kg onalternating days for 10 days). ACE (I/D) genotype was determined by PCR on storedlymphocytes from donor and recipient11. Mistyping was checked by intron specific primersas described by Shanmugan et al12. Serum and urinary creatinine concentrations and 24-hour urinary protein excretion weredetermined by standard laboratory techniques.

Data analysis

Survival curves until graft failure were calculated by the Kaplan-Meier method. Graft failurewas defined as the need to restart renal replacement therapy or death of the patient due torenal failure. Patients dying with a functioning graft were censored at the moment of death.Thus, renal survival is analysed as pure graft survival. Data are given as means with standarddeviation (SD). Differences in categorical variables between groups were tested with Chi-square test. Differences in continuous variables between two groups were tested withthe Student-t test or the non-parametric Wilcoxon rank sum test, where appropriate.Differences in continuous variables between more than two groups were tested by ANOVAor the non parametric Kruskall Wallis test. Post tests were performed with the Bonferronimethod. Differences in survival between groups were tested by logrank test. Coxproportional hazards analysis with time to graft failure as the dependent variable wasperformed; independent vaiables tested were recipient and donor ACE I/D genotype, age,gender, number of HLA class I and II mismatches, ischaemia times, number of acuterejection episodes, and creatinine clearance, blood pressure, the use of antihypertensivemedication and proteinuria at 12 months after transplantation. Hazard ratios with 95%confidence interval calculated from the exponential in the regression model, including allcovariates with a p-value < 0.1, are reported as relative risks. All p-values are two-tailed.

Results

Four hundred and ninety-seven patients received a cadaveric renal transplant at our centrefrom April 1987 to December 1994. In 62 patients, either donor or recipient material for

CHAPTER 6

83

ACE (I/D) genotype determination was unavailable. Of the remaining 435 patients, 68(15.6%) lost their graft within 12 months after transplantation due to technical failure (21 patients, 4.8%), death (15 patients, 3.4%), therapy resistant rejection (25 patients, 5.7%),or other causes (7 patients, 2.0%). Genotype distribution in patients with graft failure within12 months after transplantation was similar to that in patients with a functioning graft at 12months (II: 21, ID: 30, and DD: 17 for recipient ACE (I/D) genotype and II: 18, ID: 30, andDD: 20 for donor ACE (I/D) genotype, respectively).The characteristics of the 367 patients with a functioning graft at 12 months aftertransplantation are shown in table 1, grouped according to recipient ACE (I/D) genotype.The ACE (I/D) genotype distribution is in accordance with the Hardy-Weinbergequilibrium13. No statistically significant differences in baseline characteristics were foundbetween the three recipient ACE (I/D) genotype groups. At 12 months after transplantation,approximately 70% of the recipients used antihypertensive medication without significantdifferences between the three recipient ACE (I/D) genotype groups. Less than 10% of the

84


Table 1 Baseline characteristics at the moment of cadaveric transplantation andrenal function and blood pressure at 12 months after transplantation in 367patients with functioning grafts at 12 months grouped according to recipientACE (I/D) genotype.

# Results are given as mean ± standard deviation* at 12 months after transplantation

Recipient ACE (I/D) genotype II ID DDNumber 91 187 89Male/female 56/35 102/85 55/34Recipient age (year)# 44 ± 12 45 ± 13 43 ± 13Donor age (year)# 36 ± 16 37 ± 15 38 ± 17Total number of HLA mismatches# 1.2 ± 0.9 1.2 ± 0.9 1.3 ± 1.0Number of HLA DR mismatches# 0.3 ± 0.6 0.2 ± 0.4 0.1 ± 0.3Acute rejection episodes in first year (%) 36 41 44Serum creatinine level (mmol/l)#,* 170 ± 85 158 ± 59 171 ± 87Creatinine clearance (ml/min)#,* 61 ± 22 61 ± 21 65 ± 25Proteinuria ≥ 0.5 g/24h (%)* 31 24 28Mean arterial blood pressure (mmHg)#,* 112 ± 13 113 ± 12 113 ± 13Antihypertensive medication (%)* 70 68 74Follow up (months)# 59 ± 24 58 ± 24 55 ± 25

CHAPTER 6

85

Figure 2. Death-censored graft survival curves for patients with a functioning graft at 12 months after

renal transplantation (n=367) grouped according to donor ACE (I/D) genotype. The number of

patients for each donor ACE (I/D) genotype still at risk during follow up is indicated at the

bottom of the figure. Graft survival is not statistically significantly different between the 3

groups (log-rank test).

time (months)

Figure 1. Death-censored graft survival curves for patients with a functioning graft at 12 months after

renal transplantation (n=367) grouped according to recipient ACE (I/D) genotype. The number

of patients for each recipient ACE (I/D) genotype still at risk during follow up is indicated at the

bottom of the figure. Graft survival is not statistically significantly different between the 3

groups (log-rank test).

time (months)

patients in all groups were treated with ACE inhibitors. Also, the number of antihypertensiveagents used was not different between the different groups. Grouped according to donorACE (I/D) genotype (109, 163 and 95 with II, ID and DD genotype, respectively) nodifferences in any of the above parameters were found either (data not shown). Graft survival was not significantly different between the three recipient ACE (I/D) genotypegroups (figure 1); during follow up 6, 14 and 7 patients lost their graft in the recipient II, IDand DD genotype group, respectively. Patient survival was not different between thesegroups; 8, 23, and 11 patients died with a functioning graft in the recipient II, ID and DDgenotype group, respectively. Grouped according to donor ACE (I/D) genotype nosignificant difference in graft survival was found between these groups either (figure 2); 7, 9and 11 patients lost their graft during follow up in the donor II, ID and DD genotype group,respectively. Patient survival also was not different between these three groups; 11, 17 and14 died with a functioning graft in the donor II, ID and DD genotype groups, respectively. Toidentify risk factors for graft loss, univariate analysis of graft survival was performed by log-rank test. Creatinine clearance below 50 ml/min at 12 months (p<0.0001; relative risk (RR)4.82; 95% confidence interval (CI), 2.21 to 10.52), proteinuria ≥ 0.5 g/24h at 12 months(p<0.0001; RR 8.86; 95% CI, 3.74 to 20.96), and an acute rejection episode in the first yearafter transplantation (p=0.046; RR 2.12; 95% CI, 1.00 to 4.52) were identified as variablesassociated with graft loss. The presence of ≥ 1 class I HLA mismatches (p=0.074; RR 2.74;95% CI, 0.89 to 8.85) just failed to reach statistical significance.To assess the relative importance and interaction of the different risk factors for graft loss,Cox proportional hazards analysis was performed. Creatinine clearance (p<0.0001) and aproteinuria of ≥ 0.5 g/24h (p<0.0001) at 12 months, recipient age (p=0.013), ≥ HLA class Imismatches (p=0.011), and recipient ACE (I/D) genotype (p=0.053) were identified ascovariates with a p-value < 0.10. The model including these covariates had a r2 value of 0.33.A lower creatinine clearance at 12 months was significantly associated with an increased riskfor graft loss during long-term follow up, as was a lower recipient age. Proteinuria ≥ 0.5 g/24h at 12 months and the presence of ≥ 1 class I HLA mismatches were alsosignificantly associated with an increased risk for graft loss. The presence of one or two Dalleles in the recipient ACE (I/D) genotype is associated with time to graft loss at borderlinestatistical significance (table 2, pag. 86).To identify a possible interaction between different risk factors, subgroups of patients with apoor renal prognosis, i.e. those with creatinine clearance < 50 ml/min at 12 months aftertransplantation (n=108; 29% of all patients) or with proteinuria ≥ 0.5 g/24h (n=97; 26% ofall patients), were analysed separately.In the patients with a creatinine clearance < 50 ml/min at 12 months after transplantation 17of 27 graft losses (63%) occurred. In these patients acute rejection episodes during the first year (54 of 108 (50%) vs 94 of 259 (36%), respectively; p=0.033) and proteinuria

86


≥ 0.5 g/24h at 12 months (40 of 108 (37%) vs 57 of 259 (22%), respectively; p=0.004) weremore frequently found compared to patients with a creatinine clearance ≥ 50 ml/min at 12months (n=259). Also, donor age was significantly higher in this group (45 ± 25.5 vs 34 ±15.2 year; p<0.0001). Otherwise the characteristics at transplantation and 12 months aftertransplantation were similar. In particular, no significant differences in blood pressure or the

CHAPTER 6

87

Table 2 Relative risk with 95% confidence interval (CI) for renal graft loss calculatedfrom the Cox proportional hazards analysis model#.

# Given are results from 367 patients with functioning grafts at 12 months with time until graft failure as dependent variable including all covariates with a p-value < 0.10.

* At 12 months after transplantation

Variable (unit) Relative risk (95% CI) p-valueCreatinine clearance (ml/min)* 0.93 (0.91 to 0.95) < 0.0001Proteinuria ≥ 0.5 g/24h* 7.48 (2.93 to 19.1) < 0.0001≥ 1 HLA class I mismatch 5.47 (1.47 to 20.3) 0.011Recipient age (year) 0.96 (0.93 to 0.99) 0.013Recipient ACE (I/D) genotype (per D allele) 1.73 (0.99 to 3.01) 0.053

Figure 3. Death-censored graft survival curves for patients with a creatinine < 50 ml/min at 12 months

after transplantation (n=108) grouped according to recipient ACE (I/D) genotype. The number

of patients still at risk during follow up is indicated at the bottom of the figure. Graft survival is

not statistically significantly different between the 3 groups (log-rank test).

time (months)

use of antihypertensives agents at 12 months after transplantation were found. Figure 3shows graft survival according to recipient ACE I/D genotype in the group with a creatinineclearance < 50 ml/min. Log-rank did not identify significant differences in graft survivalbetween patients with different recipient or donor (data not shown) ACE (I/D) genotype inthis subgroup. Cox proportional hazards analysis identified creatinine clearance, proteinuria≥ 0.5 g/24h at 12 months, the presence of ≥ 1 HLA class I mismatch, recipient age, andrecipient ACE (I/D) genotype as variables independently associated with time to graft loss(table 3).In the patients with proteinuria ≥ 0.5 g/24h at 12 months 20 of 27 graft losses (74%)occurred. These patients had an increased frequency of acute rejection episodes during thefirst year (53 of 97 (55%) vs 95 of 270 (35%), respectively; p=0.0004), a lower creatinineclearance at 12 months (56 ± 26 vs 64 ± 20 ml/min, respectively; p=0.0086), and higherdonor age at time of transplantation (40 ± 17 vs 36 ± 16 year, respectively; p=0.039)compared to the patients with proteinuria < 0.5 g/24h at 12 months (n=270). No significantdifferences in blood pressure or the use of antihypertensive agents at 12 months aftertransplantation were found. Figure 4 (pag. 88) shows graft survival according to recipientACE (I/D) genotype in this subgroup. By log-rank test, no significant differences in graftsurvival between the 3 recipient ACE (I/D) genotype groups were found. In this subgroup,Cox proportional hazards analysis identified creatinine clearance at 12 months, recipientACE (I/D) genotype, and recipient age as variables independently associated with time tograft loss (table 4, pag. 88).

88



# Given are the results from a subgroup of 108 patients with a creatinine clearance < 50 ml/min at 12 months with time until graft failure as dependent variable includingall covariates with a p-value < 0.10 (r2 of the model including these variables: 0.48).


Variable (unit) Relative risk (95% CI) p-valueCreatinine clearance (ml/min)* 0.89 (0.85 to 0.94) < 0.0001Proteinuria ≥ 0.5 g/24h* 34.7 (3.97 to 303) 0.0013≥ 1 HLA class I mismatch 9.74 (1.70 to 55.7) 0.011Recipient age (year) 0.94 (0.90 to 0.99) 0.012Recipient ACE (I/D) genotype (per D allele) 2.54 (1.18 to 5.47) 0.017

CHAPTER 6

89

Figure 4. Death-censored graft survival curves for patients with proteinuria > 0.5 g/24h at 12 moths after

transplantation (n=97) grouped according to recipient ACE (I/D) genotype. The number of

patients at risk during follow up is indicated at the bottom of the figure. Graft survival is not

statistically significantly different between the 3 groups (log-rank test).

time (months)


# Given are the results from 97 patients with proteinuria ≥ 0.5 g/24h at 12 months with time until graft failure as dependent variable including all covariates with a p-value < 0.10# (r2 of the model including these variables: 0.49).


Variable (unit) Relative risk (95% CI) p-valueCreatinine clearance (ml/min)* 0.91 (0.88 to 0.94) < 0.0001≥ 1 HLA class I mismatch - not significantRecipient age (year) 0.96 (0.93 to 0.99) 0.0084Recipient ACE (I/D) genotype (per D allele) 3.02 (1.39 to 6.54) 0.0051

Discussion

The presence of one or two D-alleles in the recipient, but not donor, ACE (I/D) genotype wasfound to be associated with a shorter time to graft loss in subgroups at high risk for graft lossfrom a retrospective cohort of 367 patients with a functioning graft 12 months aftercadaveric renal transplantation. However, the relation between ACE (I/D) genotype and graftsurvival was not found by univariate analysis and could only be demonstrated aftercontrolling for other risk factors for graft loss by multivariate analysis. For the whole cohortof 367 patients, this relation was only of borderline statistical significance.Multiple risk factors have previously been identified for long-term renal graft loss. Theinability to demonstrate an association between graft survival and either recipient or donorACE (I/D) genotype for the whole population studied could mean that such an association isabsent, or that its influence is obscured by selection bias or by interaction with other riskfactors. Because our study was performed retrospectively, it may have been subject toselection bias, for instance due to effects of ACE (I/D) genotype on patient or kidneysurvival during the first year after transplantation. Genotype distribution in the 68 patientswith graft loss or death within one year after transplantation was not different from thepatients included in the study (i.e., those with a functioning graft at 12 months).Furthermore, in the group of patients studied, the distribution of ACE (I/D) genotypes was inaccordance with the Hardy-Weinberg equilibrium and the D-allele frequency was similar tothat in the normal population in the Netherlands14. Although this does not exclude selectionwith certainty, it renders such selection effects less likely. A recently published cohort studyin 269 renal transplant recipients also failed to show an association between recipient ordonor ACE (I/D) genotype and graft survival15. In contrast to our study, the latter studyanalysed graft survival from the moment of transplantation, and follow up was restricted to30 months after transplantation. The same authors recently reported a large case-controlstudy in which no differences were found in the recipient and donor ACE (I/D) genotypedistribution in patients with graft survival less than 3 years compared to patients with graftsurvival of at least 3 years16. The mean graft survival in the patients with graft survival lessthan 3 year, however, was only 5.2 months. The differences in time frame and, therefore,patient selection should be taken into account when comparing the results of these twoanalyses.Progressive renal function loss after renal transplantation is a multifactorial process. Toassess the relative importance of other risk factors for graft loss and the possible interactionof these risk factors with recipient or donor ACE (I/D) genotype, we applied differentapproaches. First, Cox proportional hazards analysis was used to account for the influence ofother risk factors. This analysis showed the recipient, but not donor, ACE (I/D) genotype tobe an independent determinant of graft loss in the whole population, albeit of borderline

90


statistical significance, with a relative risk of 1.73 per recipient D-allele present. The otherrisk factors identified in this analysis, a low creatinine clearance, a greater urinary proteinloss, more HLA class I mismatches and a lower recipient age, are in accordance withprevious findings, with the sole exception of recipient blood pressure. In this small study,blood pressure was not identified as an independent risk factor associated with graft survival,which is in contrast with larger studies17. This indicates that the risk factor profile in ourpopulation is largely in line with reports from literature1,2. This suggests that in renaltransplant recipients, recipient ACE (I/D) genotype may influence time to graft loss, but thatits effect is not prominent, or only operative in the presence of other risk factors. Anotherexplanation could be that our study group was of inadequate size and the number of eventstoo small to detect differences in graft survival associated with ACE (I/D) genotype alone.In addition, we analysed the effect of ACE (I/D) genotype in two subgroups of our patientcohort with an increased renal risk identified by a low creatinine clearance (< 50 ml/min)or proteinuria (≥ 0.5 g/24h) at 12 months after transplantation, respectively. Interestingly,in these two subgroups, the presence of the recipient D-allele was associated with time tograft loss. This more clearcut effect of the D-allele in these subpopulations compared withthe overall population might reflect the greater statistical power in subpopulations with agreater proportion of events. On the other hand, it could also implicate that the D-alleleexerts an effect on graft survival only when other risk factors are simultaneously present.Current evidence on the nature of the renal risk associated with the D-allele, as alsoapparent from a large meta-analysis that addressed cardiovascular risk18, indicates that theD-allele acts as a course modifying gene rather than as a disease inducing gene8. Our dataare consistent with this view because we did not find a difference in long-term graftsurvival, as demonstrated by the convergence of the survival curves after 5 years for thethree recipient ACE (I/D) genotype groups, but did find a difference in time to graft loss inthose destined to loose their graft. Thus, presence of a D-allele does not appear to enhancerenal risk in itself, but once a sequence of events leading to progressive renal function lossis initiated by whatever cause, its course is more rapid in presence of the D-allele.We did not find any association between graft loss and donor ACE (I/D) genotype. Acomparison between the risk associated with donor versus recipient ACE (I/D) genotypecould provide a clue as to the mechanism of risk modulation by ACE (I/D) genotype. The D-allele is associated with increased levels of circulating as well as tissue ACE and,possibly, although not uniformly, with enhanced conversion of angiotensin I in angiotensinII19,20. It should be noted that it is still unknown whether the D-allele is just a marker or amediator of increased renal risk. Nevertheless, the above findings, taken together with thekey role of angiotensin II in the pathophysiology of progressive renal function loss, fuelled the hypothesis that increased circulating or tissue ACE activity is a mediator of theincreased renal risk associated with the D-allele. An alternative hypothesis could be that

CHAPTER 6

91

graft-infiltrating mononuclear cells, which in vitro have been found to express ACEactivity influenced by ACE (I/D) genotype polymorphism, are a major source of renaltissue ACE activity after renal transplantation21,22. The present results, however, do notsupport a role for donor derived renal tissue ACE as a mechanism of risk modulation byACE (I/D) genotype in renal transplant recipients. Finally, like many studies on theinfluence of genetic polymorphisms like ACE (I/D) genotype, our study was aretrospective one. Clearly, to allow for definite conclusions, large prospective studies aremandatory. In conclusion, in patients with a high risk for graft loss, we found an association betweenthe recipient, but not donor, ACE gene D-allele and time to graft loss. This suggests that anadverse effect of the D-allele on renal prognosis is present, but only becomes manifestwhen other risk factors for graft loss are simultaneously present.

Acknowledgement

Part of this study has been presented at the Annual Meeting of The American Society ofNephrology, November 3 - 6, 1996, New Orleans (Abstract: J Am Soc Nephrol. 7:1903A,1996).

92


References

1 Hostetter TH, Cohen JJ, Harrington JT, Madias NE, Zusman CJ. Chronic transplant rejection.

Kidney Int 1994; 46: 266-79.

2 Pirsch JD, Ploeg RJ, Gange S, D’Alessandro AM, Knechtle SJ, Sollinger HW, Belzer FO.

Determinants of graft survival after renal transplantation. Transplantation 1996; 61: 1581-6.

3 Van Essen GG, Rensma PL, De Zeeuw D, Sluiter WJ, Scheffer H, Apperloo AJ, De Jong PE.

Association between angiotensin-converting enzyme gene polymorphism and failure of

renoprotective therapy. Lancet 1996; 347: 94-5.

4 McLaughlin KJM, Harden PN, Ueda S, Boulton-Jones JM, Connell JMC, Jardine AG. The role of

genetic polymorphism of angiotensin converting enzyme in the progression of renal disease.

Hypertension 1996; 28: 912-5.

5 Parving HH, Jacobsen P, Tarnow L, Rossing PP, Lecerf L, Poirier O, Cambien F. Effect of deletion

polymorphism of angiotensin converting enzyme gene on progression of diabetic nephropathy

during inhibition of angiotensin converting enzyme: observational follow up study. BMJ 1996;

313: 591-4.

6 Yoshida H, Kuriyama S, Atsumi Y, Tomonari H, Mitarai T, Hamaguchi A, Kubo H, Kawaguchi Y,

Kon V, Matsuoka K, Ichikawa I, Sakai O. Role of the deletion polymorphism of the angiotensin

converting enzyme gene in the progression and therapeutic responsiveness of IgA nephropathy.

J Clin Invest 1995; 96: 2162-9.

7 Harden PN, Geddes C, Rowe PA, McIlroy JH, Boulton-Jones M, Rodger RSC, Junor BJR, Briggs

JD, Connell JMC, Jardine AG. Polymorphism in angiotensin-converting-enzyme gene and

progression of IgA nephropathy. Lancet 1995; 345: 1540-2.

8 Navis GJ, De Jong PE, De Zeeuw D. Insertion/deletion polymorphism of the angiotensin

converting enzyme gene: a clue to heterogenity of renal prognosis and renal response to therapy?

Nephrol Dial Transplant 1997; 12: 1097-100.

9 Rigat B, Hubert C, Alhenc-Gelas F, Cambien F, Corvol P, Soubrier F. An Insertion/Deletion

polymorphism in the angiotensin-I-converting enzyme gene accounting for half the variance of

serum enzyme levels. J Clin Invest 1990; 86: 1343-6.

CHAPTER 6

93

10 Danser JAH, Schalekamp MADH, Bax WA, Van den Brink AM, Saxena PR, Riegger GA,

Schunkert H. Angiotensin-converting-enzyme in the human heart: effect of the insertion/deletion

polymorphism. Circulation 1995; 92: 1387-8.

11 Rigat B, Hubert C, Corvol P, Soubrier F. PCR detection of the insertion/deletion polymorphism of

the human angiotensin-converting enzyme gene (DCP1). Nucleic Acids Res 1992; 20: 1433.

12 Shanmugan V, Sell KW, Saha BK. Mistyping ACE Heterozygotes. PCR Methods Appl 1993; 3:

120-1.

13 Elston RC, Forthofer R. Testing for Hardy-Weinberg equilibrium in small samples. Biometrics

1977; 33: 536-42.

14 Schmidt S, Van Hooft IMS, Grobbee DE, Ganten D, Ritz E. Polymorphism of the angiotensin I

converting enzyme gene and its predisposition to high bloodpressure. Hypertension 1993; 21: 455-

60.

15 Beige J, Scherer S, Weber A, Engeli S, Offermann G, Opelz G, Distler A, Sharma AM.

Angiotensin-converting enzyme genotype and renal allograft survival. J Am Soc Nephrol 1997; 8:

1319-23.

16 Beige J, Offermann G, Distler A, Sharma AM. Angiotensin-converting enzyme insertion/deletion

genotype and long-term renal allograft survival. Nephrol Dial Transplant 1998; 13: 735-8.

17 Opelz G, Wujciak T, Ritz E. Association of chronic kidney graft failure with recipient blood

pressure. Kidney Int 1998; 53: 217-22.

18 Staessen JA, Wang JG, Ginocchio G, Saavedra AP, Soubrier F, Vlietinck R, Fagard R. The

insertion/deletion polymorphism of the angiotensin converting enzyme gene and cardiovascular

risk: a meta-analysis (Abstract). J Hypertens 1997; 15: s86.

19 Buikema H, Pinto YM, Rooks G, Grandjean JG, Schunkert H, Van Gilst WH. The deletion

polymorphism of the angiotensin-converting enzyme gene is related to phenotypic differences in

human arteries. Eur Heart J 1996; 17: 787-94.

20 Kitamura H, Moriyama T, Izumi M, Yokoyama K, Ueda N, Kamada T, Imai E. Angiotensin

I-converting enzyme insertion/deletion polymorphism: potential significance in nephrology.

Kidney Int Suppl 1996; 55: s101-3.

94


21 Friedland J, Setton C, Silverstein E. Induction of angiotensin converting enzyme in human

circulating monocytes in culture. Biophys Res Commun 1978; 83: 843-9.

22 Costerousse O, Allegrini J, Lopez M, Alhenc-Gelas F. Angiotensin I-converting enzyme in human

circulating mononuclear cells: genetic polymorphism of expression in T-lymphocytes. Biochem J

1993; 290: 33-40.

CHAPTER 6

95

Chapter 7

Summary and General Discussion

Summary

The studies described in this thesis focus on the problem of renal chronic function lossfollowing solid organ transplantation from a nephrologist point of view. Nephrologists havebeen and are still confronted with renal function loss in native kidney diseases. The last 3decades chronic renal function loss has become an important item after renal transplantation.In renal transplantation, due to the greatly improved short-term results, a clear shift inattention and research from the peri- and early post-transplantation phase to the long-termfollow up is made. However, intriguingly one does not only observe chronic renal functionloss after renal transplantation. Renal function also deteriorates shortly after transplantationof other organs such as lung, liver and heart. In non-renal solid organ transplantation, withit’s generally shorter history, progressive renal function loss poses an important andclinically relevant problem as well, but has so far not been extensively studied. Given theexperience obtained in renal transplant recipients nephrologists, as described in chapter 1,may have an important role in the necessary efforts to identify patients at risk and developrenoprotective strategies in the non-renal solid organ transplant recipients.Monitoring of renal function is a prerequisite to study renal function loss or to evaluate theimpact of intervention on renal function. Serum creatinine is commonly used as a measureof renal function. However, after solid organ transplantation great changes in creatininegeneration and excretion may occur, which may compromise the accuracy of creatinine-based renal function assessment. In order to assess the reliability of thecreatinine-based methods to detect renal function loss in this population in chapter 2 wecompared these methods with the golden standard of measurement of glomerular filtrationrate every 6 months after transplantation in a cohort of 40 lung transplant recipients followedfor at least 24 months. The rate of renal function loss calculated from the slope of measuredglomerular filtration rate and the commonly used creatinine-based indices 1/SCreatinine,Cockcroft clearance, and the Levey formula correlated significantly. However, allcreatinine-based slopes underestimated the rate of renal function loss by 20 to 30% and areless accurate to detect small losses in glomerular filtration rate. The discrepancy betweenmeasured glomerular filtration rate and creatinine-based measures of renal function provedto be dependent on the time of follow up since transplantation. This may explain theconsequent underestimation and proves that creatinine generation and excretion is not stablefollowing lung transplantation. Therefore, these methods are not well suited to study renalfunction loss in this population. The next studies describe serial measurements of glomerular filtration rate following lungtransplantation. These studies were performed to elucidate the course of renal function lossafter lung transplantation and to determine possible predictors of this renal function loss. Inchapter 3 we report renal haemodynamics in 44 patients before and at several points after

CHAPTER 7

97

lung transplantation. A substantial decline in renal function occurred, with a progressive fallin GFR of 33% at 12 months and 42% at 30 months after transplantation. Effective renalplasma flow fell by 22% at 12 months and remained stable thereafter. After transplantationblood pressure and renal vascular resistance increased significantly, consistent with theeffect of cyclosporin A. Prior to transplantation, renal function impairment with signs ofintense renal vasoconstriction was found in a subset of the patients. As especially lungtransplant recipients with pulmonary hypertension showed this pre-transplant renalhaemodynamic profile, the suggestion is that this profile may be secondary to right sidedcardiac failure. Remarkably, the decrease in renal function after transplantation was lesspronounced in patients with renal function impairment prior to transplantation: a significantnegative correlation between pre-transplantation GFR and ERPF and the percentage changein GFR and ERPF after transplantation was found. These findings suggest that renalfunction impairment in combination with intense renal vasoconstriction prior to lungtransplantation may be functionally and not structurally, and that the net course of renalfunction impairment after lung transplantation is the result of the opposed effects ofcyclosporin nephrotoxicity on the one hand and the favorable effects of the normalization ofrespiratory or cardiopulmonary status on renal perfusion on the other hand. In chapter 4 we extended the data on renal function before, at 1 month, and every 6 monthsfollowing transplantation in 57 lung transplant recipients with at least 24 months of followup. A clear biphasic pattern of renal function loss was found. The glomerular filtration ratedecreased profoundly from 100 ml/min before transplantation to 67 ml/min at six monthsafter transplantation with most of the loss already present at 1 month. From 6 months aftertransplantation onwards it was followed by a more gradual decline with a GFR of 53 ml/minat 24 months and 51 ml/min at 36 months after transplantation. The magnitude of the loss ofGFR at 1 month after transplantation was the only factor significantly correlated withabsolute GFR at 24 months after transplantation, which was one of the endpoints of thestudy. Another marker of renal progression, the slope of GFR as of 6 months aftertransplantation, was found to be significantly associated with pulmonary diagnosis. Medianloss was greatest in patients with cystic fibrosis (-10 ml/min/yr), preserved in pulmonaryhypertension (-1 ml/min/yr) and in-between in patients with emphysema (-6 ml/min/yr). Noother factors could be identified. Given the large and lasting negative impact of early renalfunction loss on renal prognosis, the obvious next step was to analyze determinants for renalfunction loss at 1 month after transplantation. To this purpose we analysed in chapter 5 acohort of 83 recipients after lung transplantation. Pre-transplant characteristics,haemodynamical peri-operative parameters, complications, and treatment modalitiesoccurring during the first month were correlated to the change in GFR between pre- and 1month post-transplantation. The mean decline in GFR in the first month was –23 ml/min.Multivariate regression analysis identified a higher pre-transplant GFR, a bilateral

98

SUMMARY AND GENERAL DISCUSSION

transplant, post-operative surgical revision for bleeding and a diminished diuresis and aprolonged period of mean arterial pressure < 70 mmHg in the first 24 hours aftertransplantation as factors independently associated with a greater decline in GFR at 1month post-transplantation. Four of these five factors are indicative for peri-operativehaemodynamic instability and these risk factors appear to be highly comparable with otherpopulations in the peri-operative phase. On the other hand, a higher pre-transplant GFR asa risk factor for early renal function loss and the fact that renal function loss is irreversibleare remarkable findings that appear to be specific for this population. As a rule renalfunction impairment is reported to be a risk factor for post-operative renal failure. Theexplanation of this discrepancy may be the diversity in course of renal function for patientswith different pulmonary diagnosis. A significant difference was apparent betweendiagnosis groups in both pre-transplant as well as post-operative renal function loss. In thispopulation a compromised renal function may be due to intense renal vasoconstriction, asin patients with pulmonary hypertension, and improvement of cardiopulmonary status mayrestore renal perfusion. A higher GFR, as in cystic fibrosis, may reflect glomerularhyperfiltration making them especially vulnerable for nephrotoxic insults. The impact ofdiagnosis is clearly of clinical relevance but it raises the question whether pulmonarydiagnosis is a risk factor or merely a marker for a favourable or poor risk factor profile. Ourresults show that the risk factors associated with early renal function loss are not evenlydistributed over the diagnosis groups supporting the assumption that pulmonary diagnosisis associated with clustering of renal risk factors. An important result from this analysis is that patients at risk for renal function loss afterlung transplantation can be identified as early as one month after or even prior totransplantation. Attention can and should, therefore, be focussed on renoprotectivemeasures in high risk patients. Whether it will be possible to improve short and long-termrenal prognosis in this particular population by additional renoprotective interventionsremains to be investigated. Chronic renal function loss occurs in a substantial proportion ofpatients with an initially successful renal transplantation as well. Until now multiple riskfactors, immunological and non-immunological, have been identified, suggesting that itspathogenesis is multifactorial. Some of these factors, such as HLA-mismatching andprevious rejection episodes are specific for renal transplants, whereas others, such as highblood pressure, a compromised renal function and proteinuria are risk factors common toboth native and transplanted kidneys. Genetic factors may also play a role in long-termrenal prognosis after renal transplantation. The renin-angiotensin-aldosterone system is, asproven in many experimental and human studies, involved in progression of renal disease.The angiotensin-converting enzyme (ACE) gene may influence this system and has beenidentified as a risk factor for progressive renal function loss and decreased renal survival innative kidneys.

CHAPTER 7

99

Whether this ACE (I/D) gene and therefore the RAAS-system plays a role in graft survivalafter renal transplantation justifies further research in this population. In contrast with thegroup of recipients after lung transplantation the group of patients after renaltransplantation is large enough to analyze the influence of ACE genotype on renal survival.To this purpose we studied in chapter 6 a group of 435 patients who received a cadavericrenal transplant. We conducted our analysis in 367 patients with a functioning graft at 12months after transplantation to avoid bias by acute pathology, such as technical failure andtherapy resistant episodes of acute rejection. Patients dying with a functioning graft werecensored and donor and recipient ACE (I/D) genotype had to be available. The latter mightallow to distinguish between the influence of tissue and systemic ACE genotype and thuscould provide a clue to the mechanism of action of the impact of ACE-genotype in renaldisease. We found that in the group as a whole graft loss in DD genotype was worse,however, no statistical significance could be found. In patients with a high risk for graft losswe found an association between recipient, but not donor, ACE gene D-allele and time tograft loss. This relation could only be demonstrated after controlling for other risk factorsfor graft loss by multivariate analysis. This suggests that an adverse effect of the D-allele onrenal prognosis is present but only becomes manifest when other risk factors for graft lossare simultaneously present. Recipient ACE gene D-allele plays an independent role in long-term graft survival after renal transplantation. This D allele is reported to be associated withhigher systemic ACE levels and may therefore modify the renin-angiotensin-aldosterone-system. Whether genetic factors such as the ACE (I/D) genotype also influence renalsurvival in lung transplant recipients has still to be investigated. In our series, the group ofrecipients after lung transplantation is still too small. However, the number of patientsliving long enough after lung transplantation is increasing.

General discussion

Almost a decade ago, the lung transplantation program started in our center. At the time,renal transplantation already had a history of more than 25 years. From the 10-yearexperience with the use of cyclosporin A in this renal population, the awareness of itsnephrotoxicity was well-established. However, the relative contribution of cyclosporinnephrotoxicity in chronic long-term function loss in renal allografts had – and still has – notbeen delineated. Not only in renal allografts1 but also in subjects with non-renal disorders2,3

treatment with cyclosporin A had been shown to exert nephrotoxic effects.The experience with solid organ transplantation other than the kidney – although in smallerpopulations and with shorter duration of follow up – had already pointed towards obviousrenal function loss after transplantation. However, it was also clear that the severity of renalfunction loss in native kidneys was very different for different transplant populations. It

100


was relatively mild – and hardly leading to clinical symptoms – in liver transplantrecipients4, but in heart- and lung-transplant recipients renal function loss had beenobserved with progression towards end-stage renal failure in approximately 10 percent ofthe patients five year after transplantation5.From this state of knowledge, at the start of the lung transplantation program it could beanticipated that renal function loss would occur in this particular population. As to itsseverity and clinical impact however, no valid prediction could be made. Triggered by therenal abnormalities in the first patients screened for transplantation – all with pulmonaryhypertension – and faced with the prospect of a population at a certain, but undefined, renalrisk, it was decided to include renal monitoring in the work up and follow up of thisparticular population6. The purpose of this renal program was to identify, within the shortestpossible time-frame, the nature and magnitude of short- and long-term renal function loss inthis particular population in order to be able to develop renoprotective strategies if needed.To this purpose, a close renal monitoring strategy was followed, based on serial renalhaemodynamic measurements. Renal haemodynamics were measured during the work upfor transplantation and at one month after transplantation – i.e. the patients being free frompost-operative instability and complications – and subsequently at six-month intervals. Thisparticular monitoring schedule – renal haemodynamics at six month intervals – wasadopted because prior studies by Apperloo et al. on long-term renal function loss in chronicrenal patients had shown that his approach allows to define the rate of renal function losswith great accuracy, within a relatively short time-frame7. Moreover, the unreliability ofcreatinine measurements during cyclosporin treatment, as well as the anticipated specificrenal haemodynamic effects of cyclosporin were reasons not to rely on creatinine-basedrenal function parameters in the renal monitoring of this population, but to use iothalamateclearances. The renal monitoring schedule was implemented immediately after the start ofthe first lung transplantations in our center. Consequently, we now avail of a unique seriesof observations, including pre-transplant renal haemodynamics, in virtually all transplantedpatients.The purposes of the renal monitoring program in the lung transplant population can besummarized as follows. First, we pursued the descriptive characterization of renal functionloss in lung transplant recipients to define the severity of renal risk in this specificpopulation. The next purpose, once it had become apparent that renal morbidity wasconsiderable, was to attempt to identify the risk factors for renal function loss in thispopulation, in order to enable to design preventive strategies. In this respect we sought forpatient-related risk factors – i.e. factors that can be identified already before transplantation,thus allowing pre-operative risk assessment – and treatment-related risk factors, that canpotentially be modified by adapting treatment protocols. In addition, we evaluated themonitoring strategy itself, in order to develop a future monitoring strategy that is both

CHAPTER 7

101

simple and effective, i.e. validated for correct identification of renal function loss, andsuitable to identify high risk patients at an early stage.

Descriptive characterization

Renal function loss after lung transplantation is considerable and a clinically relevantproblem and the course of renal function loss in recipients after lung transplantation shows abiphasic pattern8,9. The greatest amount of renal function is lost in the first 6 months andthereafter it is stabilizing in a subset of patients but in the majority of the patients renalfunction is declining with a great between-patient variability. The rate of renal function lossas of 6 months after transplantation was analysed by linear regression9. Whether the declineof renal function in this specific population on long-term remains a linear process isuncertain. Indeed, until now our long-term observation suggests a tri-phasic pattern withstabilization of renal function after initial loss of renal function. This observation has clearlyimpact on clinical as well as analytical evaluation of this population. Further follow up andseparate long-term analyses may have to be performed to answer specific questions. Within the group of lung transplant recipients as a whole a heterogeneity in renal functioncharacteristics is present and differences in pre-transplant GFR may not merely reflectquantitative differences in prior renal damage. Recipients with a compromised renalfunction pre-transplantation characterized by intense renal vasoconstriction, as in patientswith pulmonary hypertension, appeared to be protected against severe renal function lossafter transplantation whereas a higher pre-transplant GFR, as in cystic fibrosis, appeared tobe a risk factor for a worse long- and short-term renal prognosis8,9.

Identification of risk factors

Cyclosporin A plays without any doubt an important role in the process of renal functionloss after solid organ transplantation, but the diversity in the course of renal function in thepopulation after renal and lung transplantation makes clear that cyclosporin is not the onlyfactor responsible for this loss, unless cyclosporin toxicity varies widely between differentindividuals. In renal transplant recipients the course of graft survival is diverse as well. As inother renal populations10 the ACE (I/D) genotype also affected renal survival in renaltransplant recipients. Those recipients with the DD genotype and a worse renal functionand/or proteinuria at 1 year after transplantation are at risk for early graft loss11. Our analysis in lung transplant recipients until now has shown that the first blow seems to behalf the battle. This means that the amount of early renal function loss has major impact onlong-term renal prognosis. Pulmonary diagnosis appeared to be a relevant factor as well9.Risk factors for early renal function loss tend to cluster with pulmonary diagnosis:

102


peri-operative haemodynamic compromise affecting renal perfusion and pre-transplant GFRplay an important role in short-term renal function loss and therefore on long-term renalprognosis12. Further and longer follow up in a greater number of patients will be needed toanswer questions about other risk factors for renal function loss in this specific population.

Preventive strategies

Identifying patients at risk for renal function loss in the population of recipients after solidorgan transplantation gives us the opportunity to start preventive strategies in recipients atrisk. In native kidneys the renin-angiotensin-aldosterone system plays an important role inprogression of renal disease13 and the use of ACE-inhibitors is a current intervention topreserve renal function14. We found that the DD-genotype of this system, associated withhigher ACE-levels, is a risk factor for graft loss after renal transplantation in high riskgroups11. Early introduction of an ACE-inhibitor could therefore be a possible intervention topreserve graft function in high risk groups with a compromised renal function and/orproteinuria at 1 year after transplantation. Starting-point could also be to strive for anoptimal renal function at 1 year without proteinuria. Better donor selection, living donortransplantation or reducing the number of episodes with rejection by further improvement ofimmunosuppressive medication could also be tools to improve long-term graft survival.Identification of lung transplant recipients at risk for renal function loss seems possible aswell. Renal perfusion plays, as we have found12, a crucial role in the first month aftertransplantation. Therefore, trying to achieve an optimal volume status in the peri-operativephase could be a target for future research. However, the perceived risk of pulmonary edemaor respiratory distress syndrome often precludes liberal fluid resuscitation in this population.The use of cyclosporin, undermining renal perfusion, is still inevitable. Tapering cyclosporinlevels, preventing cyclosporin toxicity or postponing the introduction of cyclosporin untilrecovery of renal perfusion could be tools to prevent further compromise of renal perfusion.The development of less nephrotoxic but also strong immunosuppressive drugs should belooked for. Whether, as in native kidneys15,16, good blood pressure control will positivelyaffect long-term renal prognosis in this population has still to be investigated. The choice ofdrug, calcium-antagonists or ACE-inhibitors, is also a question that still has to be answered.

Monitoring renal function

In transplant recipients great changes in creatinine generation and excretion may occurduring follow up and therefore the reliability of the creatinine-based methods of renalfunction assessment was questioned. Underestimation of the rate of renal function loss andthe inability to detect small GFR losses17 made clear that in specific clinical situations and in

CHAPTER 7

103

research programs these methods are not the proper tools to measure renal function.However, the Levey method had the lowest treshold and the time dependent discrepancybetween this method and GFR is diminishing over time and we do not know whether it hasalready levelled off. Longer follow up will learn us more about the accuracy of this methodin measuring renal function and if it can replace iothalamate clearances.In the population after renal transplantation graft loss is used as a well known end-point inmost of the studies in this field. For preventive strategies in patients at risk for graft loss therate of renal function loss will probably give us more information about the effect of themeasures taken and strategies can be changed if the desired effect will not be reached. In thepopulation after renal transplantation changes in creatinine generation and excretion may, asin lung transplant recipients, occur as well. Testing the reliability of the creatinine-basedmethods to measure renal function in this population has to be performed if we want to usethe rate of renal function loss along with graft loss.Graft survival or the rate of renal function loss, with the pitfalls mentioned earlier, are notthe only or proper ways to reflect changes in the kidney and other tools may be used toidentify patients at risk for renal function loss early after transplantation. Assessment ofrenal damage by performing renal biopsy or measuring renal reserve capacity can learn usmore about the seriousness of the renal damage at a certain point and may be importantprognostic signs to identify patients at risk for end-stage renal disease.

Conclusion

Renal function loss after solid organ transplantation is a clinically relevant problem and aswe have shown a multifactorial process. Pre-transplantation-, peri-operative- and patientcharacteristics play a role and a multidisciplinary approach seems necessary. The GroningenLung Transplantation program proves that along with the surgeon, anaesthesist and intensivecare doctor the nephrologist plays an important role in attending recipients after solid organtransplantation.

104


References

1 Kon SP, Templar J, Dodd SM, Rudge CJ, Raftery MJ. Diagnostic contribution of renal allograft

biopsies at various intervals after transplantation. Transplantation 1997; 63: 547-50.

2 Lowe NJ, Wieder JM, Rosenbach A, Johnson K, Kunkel R, Bainbridge C. Long-term low-dose

cyclosporine therapy for severe psoriasis: effects on renal function and structure. J Am Acad

Dermatol 1996; 35: 710-9.

3 Vitale AT, Rodriguez A, Foster CS. Low-dose cyclosporin A therapy on treating chronic,

noninfectious uveitis. Ophthalmology 1996; 103: 365-73.

4 Klompmaker IJ, Homan van der Heide JJ, Tegzess AM et al. Effects of cyclosporin A withdrawal


immunosuppression for over two years. Nephrol Dial Transplant 1994; 9: 1629-33.

5 Pattison JM, Petersen J, Kuo P, Valentine V, Robbins C, Theodore J. The incidence of renal failure

in one hundred consecutive heart-lung transplant recipients. Am J Kidney Dis 1995; 26: 643-8.

6 Mannes GP, De Boer WJ, Van der Bij W, Grevink RG, Koeter GH. Three hundred patients referred

for lung transplantation. Experiences of the Dutch Lung Transplantation Program. Chest 1996;

109: 408-13.


determinations for long-term slope calculation is improved by simultaneous infucion of 125I-iothalamate and 131I-hippuran. J Am Soc Nephrol 1996; 7: 567-72.



9 Broekroelofs J, Stegeman CA, Navis GJ, van der Bij W, De Zeeuw D, De Jong. Long-term renal

outcome after lung transplantation is predicted by the 1 month post-operative renal function loss.

Transplantation 2000; 69: 1624-8.

10 Van Essen GG, Rensma PL, De Zeeuw D, Sluiter WJ, Scheffer H, Apperloo AJ, De Jong PE.

Association between angiotensin-converting enzyme gene polymorphism and failure of

renoprotective therapy. Lancet 1996; 347: 94-5.

CHAPTER 7

105

11 Broekroelofs J, Stegeman CA, Navis GJ, Scheffer H, Tegzess AM, De Zeeuw D, De Jong PE. Risk

factors for long-term renal survival after renal transplantation: a role for ACE (I/D) polymorphism?

J Am Soc Nephrol 1998; 9: 2075-81

12 Broekroelofs J, Loef BG, Stegeman CA, Navis GJ, Epema AH, Van der Bij W, De Boer WJ, De

Zeeuw D, De Jong PE. Early renal function loss after lung transplantatin: analysis of peri-opertive

risk factors. (submitted)

13 Mclaughlin KJM, Harden PN, Ueda S, Boulton-Jones JM, Connell JMC, Jardine AG. The role of

genetic polymorphism of angiotensin converting enzyme in the progression of renal disease.

Hypertension 1996; 28: 912-5.

14 Gansevoort RT, De Zeeuw D, De Jong PE. Long-term benefits of the antiproteinuric effect of ACE

inhibition in non-diabetic renal disease. Am J Kid Dis 1993; 2: 202-6.

15 Brazy PC, Stead WW, Fitzwilliam JF. Progression to renal insufficiency: Role of blood pressure.

Kidney Int 1989; 35: 670-4.

16 Parving HH, Andersen AR, Smidt UM, Hommel E, Mathiesen ER, Svendsen PA. Effect of

antihypertensive treatment on kidney function in diabetic nephropathy. Br Med J 1987; 294: 1443-7.

17 Broekroelofs J, Stegeman CA, Navis GJ, De Haan J, Van der Bij W, De Boer WJ, De Zeeuw D, De

Jong PE. Creatinine-based estimation of renal function during follow up in lung tansplant

recipients. Which method is preferable? J Heart Lung Transplant 2000; 19: 256-62.

106


Chapter 8

Samenvatting

Samenvatting

Dit proefschrift beschrijft het probleem van chronisch nierfunctieverlies na orgaan-transplantatie. Chronisch nierfunctieverlies is een van de belangrijkste aandoeningen in denefrologie. Bij niertransplantaties heeft, door de sterk verbeterde resultaten op korte termijn,een duidelijke verschuiving van aandacht en onderzoek van de vroege fase na transplantatienaar de follow up op lange termijn plaatsgevonden. In de afgelopen 30 jaar is nierfunctie-verlies na niertransplantatie dan ook eveneens een belangrijk probleem geworden. Intrige-rend is echter dat er niet alleen na niertransplantaties nierfunctieverlies wordt gezien. Denierfunctie verslechtert ook kort na transplantatie van andere organen als de longen, de leveren het hart. Bij deze transplantaties levert nierfunctieverlies eveneens een belangrijk enklinisch relevant probleem op, wat tot nu toe echter nog niet uitgebreid onderzocht is.Expertise van de nefroloog is hier onmisbaar bij de identificatie en begeleiding vanrisicopatiënten, zoal beschreven in hoofdstuk 1.Meting van de nierfunctie is een eerste vereiste bij onderzoek naar nierfunctieverlies of bijde evaluatie van het effect van een ingestelde behandeling op de nierfunctie. Gewoonlijkwordt het serumcreatinine, afbraakprodukt van spieren, gebruikt als maat voor de nier-functie. Na transplantatie van organen kunnen er echter grote veranderingen in de aanmaaken uitscheiding van het creatinine optreden, wat afbreuk kan doen aan de nauwkeurigheidwaarmee de nierfunctie op grond van creatininebepalingen wordt beoordeeld. Om debetrouwbaarheid van op creatininebepaling gebaseerde methoden ter detectie van nier-functieverlies in deze populatie te beoordelen hebben we in een cohort van 40 long-transplantatie patiënten die gedurende tenminste 24 maanden na transplantatie gevolgd zijn,elke 6 maanden deze methoden vergeleken met de meest nauwkeurige methode voor debepaling van de glomerulaire filtratiesnelheid (GFR), de gouden standaard met iothalamaat(hoofdstuk 2). De over de tijd gemeten verandering in glomerulaire filtratiesnelheidvergeleken met de op creatinine gebaseerde methoden correleerde significant. Met alle opcreatinine gebaseerde methoden werd de mate van nierfunctieverlies over de tijd echter 20tot 30% onderschat. Verder bleken deze methoden minder nauwkeurig voor het detecterenvan een langzame achteruitgang van de glomerulaire filtratiesnelheid. De discrepantietussen de gemeten glomerulaire filtratiesnelheid en de op creatinine gebaseerde bepalingenvan de nierfunctie bleek afhankelijk te zijn van de follow up tijd na transplantatie. Dit kan deonderschatting van nierfunctieverlies verklaren en vormt het bewijs dat aanmaak enuitscheiding van creatinine na een longtransplantatie niet stabiel genoeg zijn om met dezemethoden in deze populatie nierfunctie adequaat te onderzoeken.De volgende hoofdstukken beschrijven seriële bepalingen van de glomerulaire filtratie-snelheid na longtransplantatie. Deze onderzoeken werden uitgevoerd om helderheid teverkrijgen over het verloop van de nierfunctie na longtransplantatie en over mogelijke

CHAPTER 8

109

factoren die met dit nierfunctieverlies samenhangen. In hoofdstuk 3 wordt verslag gedaanvan de renale hemodynamiek van 44 patiënten voorafgaand aan en op verschillendemomenten na longtransplantatie. Er trad een substantiële verslechtering van de nierfunctieop met een progressieve daling van de GFR van 33% op 12 maanden en 42% op 30 maandenna transplantatie. De effectieve renale plasma-flow (ERPF) was na 12 maanden met 22%gedaald en bleef daarna stabiel. De bloeddruk en de renale vaatweerstand namen natransplantatie significant toe, consistent met het effect van ciclosporine A. Voorafgaand aantransplantatie was opvallend dat een subgroep van patiënten een nierfunctiestoornis mettekenen van sterke renale vasoconstrictie had. Deze groep bestond uit patiënten metpulmonale hypertensie. Dit nierfunctieprofiel is secundair aan rechtszijdig hartfalen. Deachteruitgang van de nierfunctie na transplantatie was juist bij deze patiënten minder sterk.Er werd een significante negatieve correlatie tussen de GFR en de ERPF voorafgaand aantransplantatie en de procentuele verandering in de GFR en de ERPF na transplantatiegevonden. Deze bevindingen suggereren dat bij sterke renale vasoconstrictie nierfunctie-verlies voorafgaand aan longtransplantatie functioneel zou kunnen zijn in plaats vanstructureel. Het netto beloop van de nierfunctie na longtransplantatie is de resultante van detegengestelde effecten van de nefrotoxiciteit van ciclosporine enerzijds en de gunstigeeffecten van de normalisatie van de cardiopulmonale situatie op de nierperfusie anderzijds.In hoofdstuk 4 hebben we van 57 longtransplantatie patiënten, gedurende een follow up vantenminste 24 maanden, de gegevens over de nierfunctie voorafgaand aan, op 1 maand eniedere 6 maanden na transplantatie geanalyseerd. Er werd een duidelijk bi-fasisch patroonvan nierfunctieverlies gevonden. De glomerulaire filtratiesnelheid nam sterk af van 100ml/min voorafgaand aan transplantatie tot 67 ml/min op 6 maanden na transplantatie waarbijhet grootste deel van die afname al na 1 maand aanwezig was. Vanaf 6 maanden natransplantatie ging dit over in een geleidelijker afname met een GFR van 53 ml/min op 24maanden en 51 ml/min op 36 maanden na transplantatie. Het verlies van GFR op 1 maand natransplantatie was de enige significante voorspeller van de GFR op 24 maanden natransplantatie. Van een andere marker van nierfunctieachteruitgang, de verandering vanGFR over de tijd gemeten vanaf 6 maanden na transplantatie, werd gevonden dat dezesamenhing met de pulmonale diagnose. Het verlies was het grootst bij patiënten metcystische fibrose (mediaan -10 ml/min/jr), vrijwel afwezig bij pulmonale hypertensie(mediaan -1 ml) en daartussenin bij emfyseempatiënten (mediaan -6 ml/min/jr). Binnen deverschillende groepen werd er een grote variabiliteit gezien. Er konden echter geen andereverklarende factoren worden vastgesteld. Gezien de grote en ook op langere termijn nogaantoonbare slechte invloed van vroeg nierfunctieverlies op de nierprognose, was analysevan determinanten voor nierfunctieverlies (1 maand na transplantatie) een logische volgendestap. Hiertoe werd in hoofdstuk 5 een cohort van 83 longtransplantatie patiënten geana-lyseerd. De pre-transplantatie-karakteristieken, het type transplantatie, de hemodynamische

110

SAMENVATTING

peri-operatieve parameters, de complicaties en de verschillende vormen van therapie tijdensde eerste maand werden gecorreleerd aan de verandering in nierfunctie pre-transplantatie en1 maand na transplantatie. De mediane afname in GFR in de eerste maand bedroeg 23ml/min waarbij ook hier een grote variabiliteit werd gezien. Met multivariate regressie-analyse werden een hoge GFR voorafgaand aan transplantatie, een dubbelzijdige long-transplantatie, post-operatief chirurgisch ingrijpen wegens bloeding, een verminderdeurineproduktie en een langdurige periode met een gemiddelde bloedruk < 70 mmHg in deeerste 24 uur na transplantatie als factoren geïdentificeerd die onafhankelijk geassocieerdwaren met een grotere afname van de GFR op 1 maand na transplantatie. Vier van deze vijffactoren wijzen op peri-operatieve hemodynamische instabiliteit; deze risicofactoren lijkenin hoge mate vergelijkbaar te zijn met andere niet-transplantatie populaties in de peri-operatieve fase. Opmerkelijk genoeg is een hoge GFR voor transplantatie een risicofactorvoor vroeg nierfunctieverlies na transplantatie hetgeen constrasteert met ander populatiesmet post-operatief nierfunctieverlies. Ook de irreversibiliteit van het vroege nierfunctie-verlies wijkt af van andere populaties met peri-operatief nierfunctieverlies en lijkt specifiekvoor deze populatie. Normaalgesproken is juist een gestoorde nierfunctie een risicofactorvoor post-operatieve nierinsufficiëntie bij andere populaties. De diversiteit in het verloopvan de nierfunctie van patiënten met een verschillende pulmonale diagnose kan deverklaring voor deze discrepantie zijn. Zowel voorafgaand aan transplantatie als post-operatief bestond er wat betreft nierfunctieverlies een significant verschil tussen de diagnosegroepen. Bij deze patiënten na longtransplantatie zou een gestoorde pre-operatieve nier-functie het gevolg kunnen zijn van sterke renale vasoconstrictie, zoals bij patiënten met eenpulmonale hypertensie wordt gezien, en verbetering van de cardiopulmonale situatie zou denierperfusie kunnen herstellen. Een hoge GFR, zoals we in deze populatie vooral vinden bijpatiënten met cystische fibrose, kan de weerslag van glomerulaire hyperfiltratie zijn, watdeze patiënten in het bijzonder gevoelig voor nefrotoxische momenten maakt. De invloedvan de diagnose is duidelijk klinisch relevant, maar werpt de vraag op of de pulmonalediagnose zelf een risicofactor is, of alleen maar een marker voor een gunstig of eenongunstig profiel van risicofactoren. Uit onze resultaten blijkt dat de risicofactoren voorvroeg nierfunctieverlies die samenhangen met peri-operatieve instabiliteit niet gelijkelijkverdeeld zijn over de verschillende groepen. Dit ondersteunt de aanname dat pulmonalediagnose samenhangt met clustering van risicofactoren.Een belangrijk resultaat van de analyse is dat patiënten met een verhoogd risico opnierfunctieverlies na longtransplantatie al één maand na transplantatie of zelfs al vooraf-gaand aan transplantatie kunnen worden geïdentificeerd. De aandacht kan en moet daaromop deze patiënten met een verhoogd risico worden gericht. Of het bij deze specifiekepopulatie mogelijk zal zijn de prognose van de nierfunctie op korte en lange termijn dooraanvullende maatregelen te verbeteren moet nog worden onderzocht.

CHAPTER 8

111

Chronisch nierfunctieverlies komt eveneens bij een belangrijk aantal patiënten met eenaanvankelijk succesvolle niertransplantatie voor. Tot op heden zijn meerdere risicofactoren,immunologische en niet-immunologische, geïdentificeerd, wat suggereert dat de patho-genese meerdere factoren omvat. Een aantal van deze factoren, zoals HLA mismatches enepisodes met afstoting, zijn specifiek voor niertransplantaties, terwijl andere, zoals hogebloeddruk, een gestoorde nierfunctie en eiwituitscheiding in de urine, risicofactoren zijn diezowel bij natieve als bij getransplanteerde nieren vaak voorkomen. Genetische factorenkunnen eveneens een rol spelen bij de prognose van de nierfunctie op lange termijn naniertransplantatie. Het renine-angiotensine-aldosteronsysteem (RAAS) is, zoals in veleexperimentele en humane onderzoeken aangetoond is, betrokken bij de progressie vannieraandoeningen. Een insertie/deletie polymorfisme van het angiotensin-converting-enzym (ACE) gen kan dit systeem beïnvloeden en is geïdentificeerd als risicofactor voorprogressief nierfunctieverlies en een kortere renale overlevingsduur bij ziekten van natievenieren. Dit D-allel is geassocieerd met hogere systemische ACE-spiegels en zou daardoorhet renine-angiotensine-aldosteronsysteem kunnen beïnvloeden. Of dit ACE-polymorfismeen daarom het RAAS-systeem een rol speelt bij de overleving van het transplantaat na eenniertransplantatie is een vraag die nader onderzoek binnen deze populatie rechtvaardigt. Integenstelling tot de groep ontvangers van een longtransplantaat is de groep patiënten met eenniertransplantaat groot genoeg om de invloed van het ACE-genotype op renale overlevingbinnen de context van een single center studie te onderzoeken. Hiertoe hebben we inhoofdstuk 6 een groep van 435 patiënten met een postmortale niertransplantatie bestudeerd.We hebben onze analyse uitgevoerd bij 375 patiënten met een functionerend transplantaat op12 maanden na transplantatie, om invloed door acute pathologie, zoals technisch falen enniet te behandelen episoden van acute afstoting, te vermijden. Patiënten die overleden meteen functionerend transplantaat werden gevolgd tot overlijden maar werden niet alstransplantaatfalen beschouwd. Het ACE(I/D) genotype van donor en ontvanger moestbeschikbaar zijn. Met dit laatste zou het mogelijk kunnen zijn onderscheid te maken tussende invloed van weefsel- en systemisch ACE-genotype dat daardoor een aanwijzing op zoukunnen leveren voor het werkingsmechanisme van het effect van het ACE-genotype opnieraandoeningen. We vonden dat in de groep als geheel transplantaatverlies bij hetontvanger DD-genotype vaker voorkwam, er kon echter geen statistische significantieworden aangetoond met univariate analyse. De rol van het D-allel was het meest uit-gesproken bij patiënten met ook een anderszins verhoogd risico op transplantaatverlies zoalsaangegeven door een gestoorde nierfunctie dan wel eiwituitscheiding in de urine. Dit wijsterop dat er een negatieve invloed van het D-allel op de renale prognose bestaat maar dat dieinvloed met name manifest wordt wanneer er gelijktijdig sprake is van andere risico-factoren voor transplantaatverlies. Het D-allel van het ACE-polymorfisme van de ontvangerspeelt dus een onafhankelijke rol bij de transplantaatoverleving op lange termijn na

112

SAMENVATTING

niertransplantatie. Of genetische factoren als het ACE(I/D) genotype ook invloed hebben opde renale overleving bij patiënten na longtransplantatie moet nog worden onderzocht. Inonze serie is de groep longtransplantatie patiënten nog te klein.

Algemene beschouwing

Bijna tien jaar geleden werd er in het Academisch Ziekenhuis te Groningen een long-transplantatieprogramma gestart. In die tijd had de niertransplantatie al een geschiedenis vanmeer dan 25 jaar. In de tien jaar ervaring met ciclosporine A bij deze niertransplantatie-populatie is het besef van de nefrotoxiciteit van ciclosporine duidelijk doorgedrongen. Derelatieve bijdrage van ciclosporine aan chronisch functieverlies na niertransplantatie was –en is – echter nog niet in kaart gebracht. Niet alleen na niertransplantatie, maar ook bijpatiënten met niet-renale aandoeningen was van behandeling met ciclosporine A aangetoonddat die schadelijke effecten op de nier tot gevolg had.De ervaring met transplantatie van solide organen anders dan de nieren – hoewel bij eenkleinere populatie en met een kortere follow up – wees al in de richting van een duidelijknierfunctieverlies na transplantatie. Het was echter ook duidelijk dat de ernst vannierfunctieverlies bij natieve nieren in de diverse transplantatiepopulaties in hoge mateverschilde. Dit verlies was bij ontvangers van een levertransplantaat betrekkelijk gering –en leidde nauwelijks tot klinische symptomen – maar bij hart-longtransplantaties wasprogressief nierfunctieverlies waargenomen dat bij ongeveer 10% van de patiënten 5 jaarna transplantatie in een terminale nierinsufficiëntie resulteerde.Met deze kennis kon aan het begin van het longtransplantatieprogramma worden verwachtdat bij deze specifieke populatie nierfunctieverlies zou optreden. Van de ernst en deklinische gevolgen daarvan kon echter geen goede voorspelling worden gemaakt. Aangezetdoor de nierafwijkingen bij de eerste voor transplantatie gescreende patiënten – allen haddenpulmonale hypertensie – en geconfronteerd met het vooruitzicht van een populatie met eenzeker, hoewel ongedefinieerd, nier-risico werd besloten nauwkeurige controle van denierfunctie in de screeningsfase en tijdens de follow up na transplantatie te verrichten. Hetdoel van dit nierfunctieprogramma was om in een zo kort mogelijk tijdsbestek de aard en deomvang van het nierfunctieverlies op korte en lange termijn in deze specifieke populatie vastte stellen om zo nodig een renaal protectiebeleid te kunnen ontwikkelen.Voor dit doel werd een strikt beleid ter controle van de nierfunctie uitgevoerd, gebaseerd opseriële renale hemodynamische bepalingen. De renale hemodynamiek werd tijdens descreening voor transplantatie en één maand na transplantatie bepaald – dat wil zeggen bij depatiënten zonder post-operatieve instabiliteit en complicaties – en vervolgens om de zesmaanden. Voor dit specifieke controleschema werd gekozen omdat uit eerder onderzoekdoor Apperloo et al. naar nierfunctieverlies op lange termijn bij chronische nierpatiënten

CHAPTER 8

113

was gebleken dat met deze benadering de snelheid van nierfunctieverlies binnen eenbetrekkelijk kort tijdsbestek met grote nauwkeurigheid kan worden bepaald. Verder warenzowel de onbetrouwbaarheid van creatininebepalingen tijdens behandeling met ciclo-sporine, de specifieke negatieve renale effecten van ciclosporine redenen om bij de controlevan de nierfunctie van deze populatie niet op creatinine gebaseerde methoden alleen tevertrouwen, maar iothalamaat klaringen te gebruiken. Het schema voor de nierfunctie-controle werd onmiddellijk na het begin van de eerste longtransplantaties in ons centrumgestart. Daardoor hebben we nu de beschikking over een unieke reeks waarnemingen,inclusief renale hemodynamische bepalingen voorafgaand aan transplantatie, van vrijwelalle transplantatiepatiënten.De doelstellingen van dit programma binnen de longtransplantatiepopulatie kunnen als volgtworden samengevat. Ten eerste streefden we naar de beschrijvende karakterisering vannierfunctieverlies bij patiënten na longtransplantatie om de ernst van het nierfunctie-verlies in deze specifieke populatie te definiëren. Nadat duidelijk was geworden dat er eenaanzienlijke renale morbiditeit bestond, was de volgende doelstelling een poging totidentificatie van de risicofactoren voor nierfunctieverlies in deze populatie, om ontwikkelingvan preven-tieve maatregelen mogelijk te maken. In dit verband zochten we naar patiënt-gerelateerde risicofactoren – dat wil zeggen factoren die al voorafgaand aan transplantatiekunnen worden geïdentificeerd zodat pre-operatieve risicobeoordeling mogelijk wordt – entherapiegerelateerde risicofactoren, die in potentie kunnen worden gewijzigd door de behan-delprotocollen aan te passen. Vervolgens hebben we het controlebeleid zelf geëvalueerd omvoor de toekomst een controlebeleid te ontwikkelen dat zowel eenvoudig als effectief is, datwil zeggen gevalideerd voor het correcte vaststellen van nierfunctieverlies, en dat geschikt isom patiënten met een verhoogd risico in een vroeg stadium te identificeren.

Beschrijvende karakterisering

Nierfunctieverlies na longtransplantatie is aanzienlijk en vormt een klinisch relevantprobleem. Nierfunctieverlies bij ontvangers van een longtransplantaat verloopt volgens eenbi-fasisch patroon. Het grootste verlies aan nierfunctie treedt in de eerste zes maanden opwaarna dit bij een subgroep van de patiënten stabiliseert. Bij het merendeel van de patiëntenneemt de nierfunctie echter verder af, waarbij tussen de patiënten onderling grote verschillenkunnen bestaan. De snelheid van nierfunctieverlies vanaf zes maanden na transplantatiewerd met behulp van lineaire regressie geanalyseerd. Of de achteruitgang van nierfunctie indeze specifieke populatie op lange termijn een lineair proces blijft, is onzeker. In feitesuggereert onze waarneming op lange termijn tot nu toe een tri-fasisch patroon waarbij denierfunctie in veel patiënten zich lijkt te stabiliseren na aanvankelijk achteruit te zijn gegaan.Helaas treedt deze stabilisatie niet in alle patiënten op en is inmiddels bij 5 patiënten

114

SAMENVATTING

nierfunctievervangende therapie noodzakelijk geworden. Op basis van deze eerderewaarnemingen zullen we met behulp van langdurige follow up en specifieke lange-termijnanalyses beter inzicht moeten gaan verkrijgen in de oorzaak van het specifiekongunstige beloop in een deel van de patiënten. De groep patiënten na longtransplantatie als geheel is heterogeen wat betreft nier-functiekarakteristieken, en de verschillen in GFR voorafgaand aan transplantatie zijn nietalleen maar een weergave van kwantitatieve verschillen van eerdere nierbeschadigingen.Ontvangers met voorafgaand aan transplantatie een gestoorde nierfunctie gekenmerkt doorsterke renale vasoconstrictie (als bij patiënten met pulmonale hypertensie) blekenbeschermd te zijn tegen ernstig nierfunctieverlies na transplantatie, terwijl een hoge GFR(als bij cystische fibrose) voorafgaand aan transplantatie een risicofactor bleek te zijn vooreen slechte renale prognose op korte en lange termijn.

Identificatie van risicofactoren

Ciclosporine A speelt zonder enige twijfel een belangrijke rol in het proces van nier-functieverlies na transplantatie van een solide orgaan, maar de diversiteit in verloop vannierfunctie in de populatie na longtransplantatie maakt duidelijk dat ciclosporine niet deenige voor dit verlies verantwoordelijke factor is. Bij ontvangers van een niertransplantaat ishet verloop van transplantaatoverleving eveneens divers. Net als bij andere nierpopulatieswerd ook bij ontvangers van een niertransplantaat de nieroverleving mede door hetACE(I/D) genotype beïnvloed. Die ontvangers met het DD-genotype en een slechte nier-functie en/of eiwituitscheiding in de urine op 1 jaar na transplantatie lopen het risico hettransplantaat in een vroeg stadium te verliezen.Uit onze analyse van ontvangers van een longtransplantaat blijkt tot nu toe dat de eerste klapeen daalder waard is. Dit betekent dat de ernst van vroeg nierfunctieverlies een belangrijkeinvloed heeft op de renale prognose op lange termijn. De pulmonale diagnose bleekeveneens een relevante factor te zijn. Risicofactoren voor vroeg nierfunctieverlies en depulmonale diagnose vertonen een neiging tot clustervorming: peri-operatieve hemo-dynamische instabiliteit die de nierperfusie negatief kan beïnvloeden en de GFR vooraf-gaand aan transplantatie spelen een belangrijke rol bij nierfunctieverlies op korte termijn, endus bij de renale prognose op lange termijn. Verdere en langduriger follow up bij een groteraantal patiënten is noodzakelijk om vragen over andere risicofactoren voor nierfunctie-verlies in deze specifieke populatie te beantwoorden.

Preventief beleid

Identificatie van patiënten met een verhoogd risico op nierfunctieverlies in de populatie na

CHAPTER 8

115

transplantatie van een solide orgaan moet ons in de gelegenheid stellen bij ontvangers meteen verhoogd risico preventieve maatregelen te starten. Bij natieve nieren speelt het renine-angiotensine-aldosteronsysteem een belangrijke rol bij de progressie van nieraandoeningenen het gebruik van ACE-remmers is een gebruikelijke interventie tot behoud van nierfunctie.We vonden dat het DD-genotype van dit systeem, geassocieerd met een hoog ACE-gehalte,bij groepen met een verhoogd risico voor transplantaatverlies na niertransplantatie eenrisicofactor is. Vroege introductie van een ACE-remmer kan daarom een mogelijkeinterventie zijn tot behoud van de functie van het transplantaat bij groepen met eengestoorde nierfunctie en/of eiwituitscheiding in de urine op 1 jaar na transplantatie. Strevennaar een optimale nierfunctie zonder eiwituitscheiding in de urine op 1 jaar na transplantatiekan ook een uitgangspunt zijn. Een betere donorselectie, levende donoren of verderereductie van het aantal episodes met afstoting door verdere verbetering van immuno-suppressieve medicatie kunnen eveneens opties zijn ter verbetering van de transplantaat-overleving op lange termijn.Identificatie van ontvangers van een longtransplantaat met een verhoogd risico vannierfunctieverlies lijkt eveneens mogelijk. Nierperfusie speelt, zoals we vonden, een crucialerol in de eerste maand na transplantatie. Daarom kunnen pogingen tijdens de peri-operatievefase een optimale vulling te verkrijgen een doelstelling voor toekomstig onderzoek zijn. Hette voorziene risico van longoedeem of van het respiratory distress syndrome vormt bij dezepopulatie echter een belangrijk bezwaar tegen liberale vochttoediening. Het gebruik vanciclosporine, wat de nierperfusie eveneens ondermijnt, is nog steeds noodzakelijk. Verlagingvan ciclosporine spiegels, preventie van ciclosporine-toxiciteit of uitstel van het starten metciclosporine totdat de nierperfusie hersteld is kunnen eveneens preventieve hulpmiddelenzijn. Studies naar minder nefrotoxische doch krachtige immunosuppressiva zullen in detoekomst uitkomst moeten bieden. Of een goede bloeddrukcontrole net als bij anderenieraandoeningen een gunstige invloed op de renale prognose op lange termijn heeft, moetnog worden onderzocht. In hoeverre een specifieke klasse van antihypertensievemedicamenten, calciumantagonisten of ACE-remmers, hierbij de voorkeur heeft is evenminduidelijk.

Controle van de nierfunctie

Bij ontvangers van transplantaten kunnen tijdens de follow up grote veranderingen in deaanmaak en uitscheiding van creatinine optreden, reden waarom vraagtekens bij debetrouwbaarheid van op creatinine gebaseerde methoden ter beoordeling van de nierfunctiewerden geplaatst. Onderschatting van de snelheid van nierfunctieverlies en het onvermogenkleine nierfunctie-verliezen bij patiënten na longtransplantie te detecteren maakten duide-lijk dat deze methoden in specifieke klinische situaties en in onderzoeksprogramma’s niet

116

SAMENVATTING

geschikt zijn voor het bepalen van de nierfunctie. De Levey-methode had echter de laagstedrempel en de tijdsafhankelijke discrepantie tussen deze methode en de GFR neemt naverloop van tijd af. We weten niet of deze trend zich verder zal voortzetten. Naarmate defollow up langer duurt, komen we meer te weten over de nauwkeurigheid van deze methodevoor het bepalen van de nierfunctie.Binnen de niertransplantatiepopulatie wordt transplantaatverlies als algemeen bekendeindpunt bij de meeste onderzoeken op dit terrein gebruikt. Wat betreft preventief beleid bijpatiënten met een verhoogd risico op transplantaatverlies zal de snelheid van nierfunctie-verlies ons waarschijnlijk meer informatie kunnen verschaffen over het effect van degenomen maatregelen en kan het beleid worden bijgesteld als het gewenste effect niet wordtbereikt. Binnen de niertransplantatiepopulatie kunnen, net als bij patiënten na longtransplan-tatie, veranderingen in de aanmaak en uitscheiding van creatinine optreden. De betrouw-baarheid van de op creatinine gebaseerde methoden om in deze populatie de nierfunctie tebepalen zal eveneens moeten worden geanalyseerd als we de snelheid van nierfunctieverliessamen met transplantaatverlies willen gebruiken.Transplantaatoverleving of de snelheid van nierfunctieverlies, met de eerder genoemdevalkuilen, zijn niet de enige methoden om veranderingen in de nieren weer te geven; ookandere middelen kunnen worden toegepast om patiënten met een verhoogd risico voornierfunctieverlies vroeg na transplantatie te identificeren. Het verrichten van een nierbiopsieof het bepalen van de renale reservecapaciteit kunnen ons mogelijk meer vertellen over deernst van de nierschade op een bepaald moment en op die grond belangrijke prognostischeaanwijzingen zijn voor de identificatie van patiënten met een verhoogd risico op terminalenierinsufficiëntie.

Slot

Nierfunctieverlies na transplantatie van een solide orgaan is een klinisch relevant probleemen, zoals we hebben laten zien, een multifactorieel proces. Pretransplantatie-, periopera-tieve- en patiënt-karakteristieken spelen een rol en een multidisciplinaire benadering lijktnoodzakelijk. Het Groningse Longtransplantatieprogramma laat zien dat de nefroloogsamen met de chirurg, de anesthesist en de intensivist een belangrijke rol speelt bij de zorgvoor patiënten na solide orgaantransplantatie.

CHAPTER 8

117

Verklarende woordenlijst

Nefroloog specialist die zich bezig houdt met nierpatiënten

Chronisch nierfunctieverlies langzame achteruitgang van nierfunctie

Natieve nieren eigen nieren

Serumcreatinine creatininespiegel in het bloed

Cohort groep

Renale hemodynamiek bloeddruk locaal in de nier

Glomerulaire filtratiesnelheid (GFR) nierfunctie op de meest nauwkeurige manier

gemeten

Effectieve renale plasma-flow (ERPF) bloedstroom in de nier

Renaal van de nier

Vaatweerstand spanning in bloedvatwand

Vasoconstrictie samengeknepen bloedvat

Ciclosporine medicijn ter voorkoming van afstoting

Nefrotoxiciteit schadelijk voor de nier

Pulmonaly hypertensie ziektebeelden met hoge bloeddruk in de longen

Cardiopulmonale situatie de toestand aangaande longen en hart

Nierperfusie nierdoorbloeding

Cystische fibrose taai slijm ziekte

Emfyseem rek uit de longen

Peri-operatieve hemodynamische instabiliteit periodes met te lage bloeddruk rondom de operatie

Renale protectiemaatregelen maatregelen ter bescherming van de nier

Immunologisch betrekking hebbende op het afweermechanisme

HLA mis-matches verschillen in het erfelijke materiaal

Renine-angiotensine-aldosteronsysteem (RAAS) het systeem in de nier wat mede de bloeddruk en

de zouthuishouding regelt

Angiotensin-converting-enzym (ACE) gen het stukje erfelijk materiaal wat het RAAS-systeem

beïnvloedt

Allel onderdeel van een gen

Terminale nierinsufficientie dusdanig ernstige nierfunctiestoornis dat

kunstnierbehandeling nodig is

Iothalamaat het stofje dat nodig is om de GFR te bepalen

Morbiditeit ziekte

ACE-remmers medicijnen die de bloeddruk reguleren via het

RAAS-systeem

Renale reservecapaciteit extra nierfunctie die kan worden behaald na

stimulatie

118

SAMENVATTING

APPENDIX

Broekroelofs J, Navis G, Stegeman CA, Van der Bij W, De Jong PELancet 1998; 351: 1064Lancet 1998, 352: 69-70

Lung transplantation

Sir, Jeffrey Hosenpud and co-workers (Jan 3, p 24)1 note the importance of pulmonary diagnosisfor survival benefit after lung transplantation. In the accompanying commentary, John Dark2

describes the lack of survival benefit in emphysema patients and argues that diagnosis is afactor to consider in allocation schemes. We wish to draw attention to pulmonary diagnosisas a determinant of post-transplantation loss of renal function.

In our centre, glomerular filtration rate (GFR; clearance of iothalamate) is measured beforelung transplantation and every 6 months after transplantation, which allows accurateevaluation of renal function. Long-term loss of renal function occurs in most patients3. Weanalysed renal function loss according to diagnosis in 39 patients who underwenttransplantation between January, 1992, and December, 1995, with at least 24 months offollow up. Seven patients with cystic fibrosis had a pretransplantation GFR of 124 (range 96to 163) ml per min per 1·73 m2 with a decrease of 60% (46 to 79) at 24 months. In 28 patientswith emphysema, pretransplantation GFR was 95 (83 to 124) with a decrease of 44% (17 to70). In four patients with pulmonary hypertension, pretransplantation GFR was 95 (83 to124) with a decrease of 36% (24 to 46). Overall, loss of renal function differed between thethree groups (p<0·05). This difference was not explained by perioperative differences, sincethe rate of long-term renal function loss (6-24 months after transplantation) also differedbetween the groups (p<0·05). Loss of renal function was most rapid in patients with cysticfibrosis (-15 [-20 to -10] ml per min every year) attenuated in pulmonary hypertension (-1 [-3 to 2] ml per min every year), and in-between in patients with emphysema (-6 [-17 to7] ml per min every year).Thus, diagnosis is relevant to renal morbidity after lung transplantation. Although ourpopulation is small, the accurate renal measurements revealed differences in loss of renalfunction between the groups within a short observation time. The main concern about renalfunction is the development of end-stage renal failure, which depends not only on the rate ofrenal function loss but also on survival time. We found that survival was better in patientswith emphysema (85% at 24 months) and pulmonary hypertension (89%) than in those withcystic fibrosis (58%). After 24 months, there were too few patients for a valid comparison ofrenal outcome. Data thus far suggest that diagnosis is relevant to renal function beyond 2years. In all four pulmonary hypertension patients, renal function was stable up to 4 years,whereas two patients with emphysema entered dialysis about 5 years after transplantation.

Our findings underline the importance of a pulmonary diagnosis when assessing the benefitof transplantation for lung transplantation candidates, in terms of survival as well as quality

121

of life. In rapidly progressive pulmonary disorders, such as cystic fibrosis, survival benefitmay outweigh a poor renal prognosis. When the pulmonary condition is only slowlyprogressive, as in many patients with primary emphysema, survival benefit is less likely. Forthese patients, the negative effect of long-term renal morbidity on initially gained quality oflife could be an important factor to consider.

Jan Broekroelofs, Gerjan Navis, Coen A Stegeman, Wim van der Bij, Paul E de Jong. Department of Medicine, Divisions of Nephrology and Pulmonology, State UniversityHospital, 9713 G2 Groningen, Netherlands

References

1 Hosenpud JD, Bennett LE, Keck BM, Edwards EB, Navick RJ . Effect of diagnosis on survival

benefit of lung transplantation for end-stage lung disease. Lancet 1998; 351: 24-27.

2 Dark JH. Priorities for lung transplantation. Lancet 1998; 351; 4-5.

3 Navis GJ, Broekroelofs J, Mannes GPM et al. Renal haemodynamics after lung transplantation:

a prospective study. Transplantation 1996; 61: 1600-05.

Renal failure after lung transplantation

Sir,Jan Broekroelofs and colleagues (April 4, p 1064)4 comment that pretransplant pulmonarydiagnosis influenced the risk of subsequent renal toxicity in their lung-transplant population.Their report did not include any data relating to doses or blood concentrations ofimmunosuppressives used, but I presume that they used either cyclosporin or tacrolimus,which have very similar renal toxicity.

The faster rate of renal decline they describe in patients transplanted for cystic fibrosis isprobably largely a reflection of higher pretransplant glomerular filtration rate in this group;a similar occurrence has been shown in liver-transplant recipients and is probably a featureof cyclosporin nephrotoxicity5 rather than preoperative diagnosis6. Further differences inlong-term renal function loss were probably a reflection of variation in metabolism ofcyclosporin or tacrolimus in the different patient groups; since these drugs are metabolisedin the liver and excreted almost exclusively in bile, any impairment of liver function (which

122

APPENDIX

commonly occurs in cystic fibrosis and -1-antitrypsin deficiency) may reduce their excretionand thus predispose to renal toxicity. The use of trough blood concentrations of cyclosporinor tacrolimus to guide dosage may to some extent reduce the risk of renal toxicity. However,we and others have found that outpatient monitoring of cyclosporin concentrations is a poorguide to risk of renal failure7,8; this is partly because of variation in individual susceptibilityto nephrotoxicity, but also perhaps difficulty in obtaining valid trough concentrations in theoutpatient setting. In fact, in the long-term we found the dosage (adjusted for weight), ratherthan cyclosporin concentration was a more reliable guide to risk of nephrotoxicity. The development of end-stage renal failure in solid organ (other than kidney) transplantrecipients carries a high mortality even with renal replacement therapy. Avoidance of this devastating complication in patients who develop renal impairment in the presence of satisfactory graft function requires early and judicious reduction in dose of cyclosporin or tacrolimus, or in some cases complete substitution with non-nephrotoxicimmunosuppressives if renal impairment has become irreversible8.

Neil C Fisher Department of Gastroenterology, New Cross Hospital, Wolverhampton WV10 0QP, UK

References

4 Broekroelofs J, Navis G, Stegeman CA, Van der Bij W, De Jong PE. Lung transplantation. Lancet

1998; 351: 1064.

5 Platz KP, Mueller AR, Blumhardt G et al. Nephrotoxicity following orthotopic liver

transplantation: a comparison between cyclosporine and FK506. Transplantation 1994; 58: 170-8.

6 Ayres R, Ismail T, Angrsani L, Buckels J, Elias E, McMaster P et al. Long-term renal function in

liver transplantation. Transpl Proc 1991; 23: 1469-70.

7 Feutren G, Miller C. Low predictive value of cyclosporine level for efficacy or renal dysfunction in

psoriasis and idiopathic nephrotic syndrome. Transpl Proc 1990; 22: 1299-302.


following liver transplantation: a retrospective analysis. Transplantation 1998; 15: 59-66.

123

Authors’ reply

Sir,We agree with Neil Fisher that cyclosporin is presumably a main factor in renal function lossin transplant recipients. We emphasise, however, that the severity of cyclosporinnephrotoxicity varies widely between different populations of recipients, suggestinginteraction with other renal risk factors. This variation is not only apparent within thepopulation of lung-transplant recipients, but also between recipients of different organtransplants. In cardiac and heart-lung9, as well as lung transplant recipients renal morbidity10

is reported to be higher than that in liver-transplant recipients11. Many factors mightcontribute to differences in renal damage among the various recipients, such as differencesin cyclosporin dosing regimen or handling, comorbidity and cotreatment, such as antibiotics.To prevent renal function loss, identification of the specific renal risk factors in the differentpopulations is essential. To identify determinants of renal function loss in solid-organ recipients, it is important todistinguish between perioperative changes and long-term loss of renal function.Perioperative renal function changes can be hard to interpret in terms of nephrotoxicity,because a successful heart, lung, or liver transplant can improve renal perfusion bynormalising cardiac output in patients with heart failure or cor pulmonale, and by curinghepatorenal syndrome, respectively. We have seen postoperative renal functionimprovement - despite cyclosporin - in patients with intense renal vasoconstriction beforelung transplantation10. Thus, early changes in net renal function result from the opposedeffects of improved renal perfusion and cyclosporin nephrotoxicity. Such haemodynamicfactors may explain why patients with a compromised renal function pretransplantationseem to do better after transplantation than those with normal pretransplantation function, assuggested by Fisher. Thus, their net change in renal function may not be a good indication ofrenal parenchymal damage. Considering that we avoided such bias by analysing short-termand long-term renal function loss separately, the difference in the long-term rate of renalfunction loss between the diagnosis groups is all the more striking.

Cyclosporin is presumably involved in renal function loss in our patients. All patients wereon cyclosporin (three of whom were converted to tacrolimus after 6, 24, and 36 months). Toaccount for differences in cyclosporin handling, our dosing regimen is based onpretransplantation assessment of individual cyclosporin kinetics. Trough cyclosporinconcentrations were similar for the diagnosis groups throughout follow up. Remarkably,cystic fibrosis patients require higher cyclosporin doses to obtain similar bloodconcentrations. This dosage regimen may well have an impact on renal function. Whethercyclosporin dose can be safely reduced, however, is still unknown.

124

APPENDIX

Introduction of less nephrotoxic immunosuppressives such as mycophenolate may reducerenal morbidity in the future. For the moment, however, cyclosporin is still indispensable.Whereas close monitoring of dosage regimen is important to reduce renal morbidity, it isalso crucial to recognise that the severity of cyclosporin nephrotoxicity is apparentlymodified by other factors that may be specific for a particular patient population.Identification of such factors may provide targets for renoprotective strategies for patients athigh risk.

Jan Broekroelofs, Coen A Stegeman, Wim van der Bij, Paul E de Jong, Gerjan Navis Department of Medicine, Division of Nephrology, University Hospital, 9713 GZGroningen, Netherlands

References

9 Pattison JM, Petersen JP, Kuo P, Valentine V, Robbins RC, Theodore J. The incidence of renal

failure in one hundred consecutive heart-lung transplant recipients. Am J Kidney Dis 1995; 26:

643-8.

10 Navis GJ, Broekroelofs J, Mannes GPM et al. Renal haemodynamics after lung transplantation:

a prospective study. Transplantation 1996; 61: 1600-5.

11 Klompmaker IJ, Homan van der Heide JJ, Tegzess AM et al. Effects of cyclosporin A withdrawal


immunosuppression for over two years. Nephrol Dial Transpl 1994; 9: 1629-33.

125

Dankwoord

Bij deze wil ik een aantal mensen bedanken die het voltooien van dit proefschrift mede tot eengoed einde hebben gebracht. Zoals al velen voor mij hebben gemeld is het schrijven van eenproefschrift een teamsport en zonder de medespelers en coaches niet te volbrengen. Debelangrijkste coaches waren ongetwijfeld Prof. Dr. Gerjan Navis en Dr. Coen Stegeman diemet hun onuitputtelijke geduld en enthousiasme keer op keer mijn onderzoeksresultaten enartikelen hebben bestudeerd en gecorrigeerd. Tevens waren beiden een voortdurend klankborden hebben we samen uitgebreide discussies gevoerd over de resultaten en de interpretatiedaarvan. Ik kan rustig stellen dat het zonder hun hulp niet was gelukt. Uiteraard hebben mijn andere promotoren Prof. Dr. Paul de Jong en Prof. Dr. Dick de Zeeuw ook een heelbelangrijke rol hierin gespeeld zodat uiteindelijk de puntjes op de ”i” werden gezet. Tevenskregen we kritische adviezen en opmerkingen van Drs. Adam Tegzess, Dr. Wim van der Bij,Drs. Wim de Boer en Drs. Bert Loef.Voor het verzamelen van de klinische gegevens heb ik de hulp gehad van een aantal medischstudenten: Alward, Tim, Laila, Maureen en Jelmer hartelijk dank hiervoor. Mijn dank gaat ookuit naar Marja van Kammen en Aly Drent-Bremer van de nierfunctiekamer die alle GFRmetingen hebben begeleid en natuurlijk ook naar de patiënten die bereid waren om heelregelmatig een deel van hun kostbare tijd op de nierfunctiekamer door te brengen. Ook mijnhuidige collegae in Leeuwarden wil ik hartelijk danken voor het feit dat ze me tot nu toe in degelegenheid hebben gesteld om wekelijks naar Groningen af te reizen voor overleg met mijnbegeleiders. Tevens dank ik de leden van de beoordelingscommissie, Prof. dr. B.L. Kasiske,Prof. dr. G. Koëter en Prof.dr. L.C. Paul voor hun beoordeling van het proefschrift.Naast het schrijven van dit proefschrift ben ik natuurlijk ook nog opgeleid tot nefrolooghetgeen op een zeer degelijke en plezierige manier plaatsvond en waarvan ik nu inLeeuwarden veel profijt heb. Bovendien weet ik de weg nog steeds te vinden als ik inLeeuwarden met ingewikkelde problematiek word geconfronteerd.De lay-out van het proefschrift werd verzorgd door Mw. Elsa Alingh Prins-van Rhijn en deillustratie op de voorkant door Mw. Margriet Schadee-Dikkers. Al met al hebben zeer veel mensen bijgedragen aan het tot stand komen van dit proefschrift enik ben vast en zeker mensen vergeten te bedanken, zodat ik bij deze al deze vergeten personenook wil bedanken.Nu hoop ik dan ook weer meer aandacht te kunnen besteden aan mijn gezin die de laatste jarenveel hebben moeten missen. Ik zat namelijk veel boven op zolder achter de computer tewerken omdat de finish en deadline van de volgende etappe bijna weer waren bereikt. Op 15 november 2000 hoop ik de finish succesvol te passeren.

Jan Broekroelofs

126

Curriculum vitae

De schrijver van dit proefschrift werd op 6 november 1961 geboren te Hardenberg. Na demiddelbare school (C.S.G. “Jan van Arkel” te Hardenberg) studeerde hij vanaf 1980geneeskunde aan de Rijksuniversiteit te Groningen. Hij behaalde zijn arts-examen inseptember 1987. Hierna vervulde hij zijn dienstplicht als arts-assistent interne geneeskundein het Militair Hospitaal “Dr. A. Mathijssen” te Utrecht (hoofd: Dr. M. van Zoeren). In maart1990 begon hij met zijn opleiding tot internist in het Sophia Ziekenhuis te Zwolle (opleider:Dr. T. Tjabbes) welke hij voortzette in oktober 1992 in het Medisch Spectrum Twente teEnschede (opleider: Dr. J.G.M. Jordans). Op 1 september 1995 registreerde hij zich alsinternist en begon in het Academisch Ziekenhuis te Groningen met zijn opleiding totnefroloog (opleider: Prof. Dr. P.E. de Jong). Tijdens deze laatste periode werd gestart met hetonderzoek wat uiteindelijk resulteerde in dit proefschrift. De nefrologie aantekening werdbehaald in februari 1998 en sindsdien is hij werkzaam als internist-nefroloog, aanvankelijkals waarnemer, in het Medisch Centrum Leeuwarden. Hij is getrouwd met Pelle Dikkers enhun kinderen heten Maartje, Gijs en Joost.

128



Illustratie omslag:Margriet Schadee-Dikkers

university of groningen renal function after solid organ transplantation … · 2016-03-08 ·...

Documents