A Longitudinal Study of Inflammation, CKD-Mineral Bone Disorder, and Carotid Atherosclerosis after Renal Transplantation

Article

A Longitudinal Study of Inflammation, CKD-MineralBone Disorder, and Carotid Atherosclerosis after RenalTransplantation

Mahmut Ilker Yilmaz,* Alper Sonmez,† Mutlu Saglam,‡ Tuncer Cayci,§ Selim Kilic,| Hilmi Umut Unal,*Murat Karaman,* Hakki Cetinkaya,* Tayfun Eyileten,* Mahmut Gok,* Yusuf Oguz,* Abdulgaffar Vural,*Francesca Mallamaci,¶** and Carmine Zoccali**

AbstractBackground and objectives The role of reversibility of nontraditional risk factors, like inflammation and CKD-mineral bone disorder, in the reduction of cardiovascular risk after renal transplantation is still scarcely defined.

Design, setting, participants, and measurements The longitudinal relationship between C-reactive protein, CKD-mineral bone disorder biomarkers, and intima media thickness was investigated in a series of 178 patients(age=32610 years) with stage 5 CKDmaintained on chronic dialysis who underwent echo-color Doppler studiesof the carotid arteries before and after renal transplantation. Smokers and patients with diabetes were excludedfrom the study. In all patients, immunosuppression was performed by a standard regimen on the basis ofcalcineurin inhibitors. Healthy controls were specifically selected to match the age and sex distribution of thepatients. Biochemical and intima media thickness assessments were repeated 6 months after transplantation.

ResultsBefore transplantation, intimamedia thickness inpatientswith stage 5CKDondialysis (average=0.960.2mm)was higher (P,0.001) than in well matched healthy controls (0.660.1 mm) and reduced substantially(222%; 95% confidence interval, 224% to 220%) after transplantation (P=0.001). GFR (multivariable-adjustedb=0.23; P,0.001), C-reactive protein (b=0.15; P,0.001), and fibroblast growth factor 23 (b=0.28; P,0.001) werethe strongest independent correlates of intima media thickness before transplantation. Similarly, longitudinalchanges in the same biomarkers were the sole independent correlates of simultaneous changes in intima mediathickness (C-reactive protein: b=0.25; fibroblast growth factor 23: b=0.26; P,0.001 for both) after renal trans-plantation. The evolution of intima media thickness after transplantation was largely independent of classic riskfactors, including BP, LDL cholesterol, and insulin resistance, as measured by homeostatic model assessment.

Conclusions Intima media thickness improves after renal transplantation. Such an improvement associates withparallel changes in serum C-reactive protein and fibroblast growth factor 23. These observations are in keepingwith the hypothesis that the decline in cardiovascular risk after transplantation, in part, depends on partialresolution of nontraditional cardiovascular risk factors, like inflammation and CKD-mineral bone disorder.

Clin J Am Soc Nephrol 10: ccc–ccc, 2014. doi: 10.2215/CJN.07860814

IntroductionThe risk of death and cardiovascular complications isabout three to four times lower in patients with renaltransplants than patients with coeval stage 5 CKD ondialysis awaiting renal transplantation (1). The declinein the risk for such events after transplantation occursin the face of an increase in the prevalence of classic riskfactors (2,3), and it is, therefore, mainly attributed tobetter control of CKD-related risk factors (4). Systemicinflammation as measured by C-reactive protein (CRP)is a powerful predictor of death (5) and progression ofatherosclerosis in patients on dialysis (6). Biomarkers ofCKD-mineral bone disorder (CKD-MBD)—includinghigh serum phosphate and parathyroid hormone(PTH), low vitamin D, and high fibroblast growth fac-tor 23 (FGF23)—are among the most indicted risk

factors for the high rate of cardiovascular events in pa-tients on dialysis (7,8). Both CRP (9) and CKD-MBD(10) improve after successful renal transplantation,and such an improvement may favorably affect vascu-lar disease in these patients.Current Kidney Disease Improving Global Outcomes

guidelines (11) remark that there is a paucity of studieson the evolution of vascular disease shown by imagingtechniques in patients on dialysis after renal transplan-tation. Carotid intima media thickness (IMT) is an in-termediate phenotype of atherosclerosis and one of thestrongest predictors of cardiovascular events, includingmyocardial infarction and stroke (12). Reduction inIMT over time entails a parallel decrease in the riskfor cardiovascular events, and this parameter is consid-ered as a valid surrogate end point by the Food and

Departments of*Nephrology,†Endocrinology,‡Radiology,§Biochemistry, and|Epidemiology,Gulhane School ofMedicine, Etlik-Ankara, Turkey; and¶Nephrology,Hypertension andRenal Transplantationand **NationalResearch Council-Institute of ClinicalPhysiology ClinicalEpidemiology of RenalDiseases andHypertension,Ospedali Riuniti,Reggio Calabria, Italy

Correspondence:Dr. Mahmut IlkerYilmaz, Department ofNephrology, GulhaneSchool of Medicine,06018 Etlik-Ankara,Turkey. Email:[email protected]

www.cjasn.org Vol 10 March, 2014 Copyright © 2014 by the American Society of Nephrology 1

. Published on December 26, 2014 as doi: 10.2215/CJN.07860814CJASN ePress

Drug Administration (13) and the European MedicinesAgency (14). The validity of IMT as a cardiovascular riskbiomarker in the CKD population is supported by studiesin patients with predialysis CKD (15) and patients on dialysis(6,16). In the sole longitudinal study focusing on IMT, thisparameter regressed in parallel with PTH levels after renaltransplantation (17). Until now, there was no sufficientlylarge study investigating the evolution of IMT after trans-plantation face to face with ongoing changes in establishedbiomarkers of inflammation and CKD-MBD and no studiesproviding information on FGF23, which is now held as themain vasculotoxic factor among CKD-MBD biomarkers (18).Probing the evolution of arterial disease and parallel

changes in pertinent risk factors after renal transplantationis of relevance in that such studies may help the interpre-tation of the substantial cardiovascular risk reductionbrought about by restored renal function (1,19).With thisbackground in mind, we studied the longitudinal associa-tion of changes in IMT in carotid arteries over time (beforeand after renal transplantation) with ongoing changes inclassic risk factors and CRP and CKD-MBD biomarkersin a large incident series of patients who underwent suc-cessful renal transplantation.

Materials and MethodsPatients and ControlsThe study protocol was approved by the ethical committee

of the Gulhane School of Medicine in Ankara, Turkey. Allconsecutive adult patients (n=272) who underwent a kidneytransplant at our center between January of 2003 and Januaryof 2012 were considered for this analysis. Some of these pa-tients are part of an ongoing prospective study encompass-ing measurements of vascular function, which has beendescribed in a separate publication (20). Because the scopeof our study was that of investigating the reversibility ofatherosclerosis (as measured by IMT) after restoration of

renal function by renal transplantation and the functionalrelationship between the longitudinal changes in IMT andCKD-MBD biomarkers, by protocol, we excluded patientswith risk factors that may distort the interpretation of carotidartery changes brought about by successful renal transplan-tation, including diabetes, background cardiovascular dis-ease, smoking, rejection episodes, and use of mammaliantarget of rapamycin inhibitors, a class of drugs that mayper se have a favorable effect on atherosclerosis (21,22). Theflowchart of the selection process is shown in Figure 1. Weselected all patients with uneventful clinical course (no rejec-tion episodes). We excluded five patients on sirolimus oreverolimus, because these drugs per se prevent intimal hy-perplasia and have a different effect on vascular functioncompared with standard cyclosporin-based regimens(21,22) and four patients on a regimen including antithymo-cyte globulin (Muromonab-CD3), a drug that interferes in acomplex manner with several components of the inflamma-tion cascade (23). Five additional patients were excluded,because they had acute rejection episodes, and two patientswere excluded because of positive cross-match transplant(n=1) or blood group system–incompatible transplant (n=1).Furthermore, we excluded patients with diabetes mellitus(n=1), patients with history of cardiovascular disease(electrocardiogram-documented angina, myocardialinfarction documented by electrocardiogram, and bio-markers of myocardial necrosis, cerebral ischemia, or re-vascularization procedures; n=5), smokers (n=7), andthose taking angiotensin-converting enzyme inhibitors orangiotensin receptor blockers (n=10), active forms of vita-min D (n=13), erythropoietin (n=21), statins (n=2), and sup-plementary vitamin pills, including folate (n=2). Threepatients were lost to follow-up before the second study(i.e., the study after transplantation), six patients withdrewconsent, and seven patients were excluded because of startingangiotensin-converting enzyme inhibitors/angiotensin recep-tor blockers (n=4) or active vitamin D compounds (n=3).

Figure 1. | Flowchart of patients enrolled in the study. ABO, blood group system; ACEi, angiotensin-converting enzyme inhibitor; ATG,antithymocyte globulin; EPO, erythropoietin.

2 Clinical Journal of the American Society of Nephrology

As expected, the selected cohort (n=178) had a young meanage (3269 years); 159 patients in this cohort before trans-plantation had been treated by hemodialysis, and 19 pa-tients had been treated by peritoneal dialysis. Thirty-threepatients were anuric before transplantation. Before trans-plantation, patients received standard treatment for CKD-MBD at our institution (i.e., phosphate binders: calciumacetate [n=62], calcium carbonate [n=34], or sevelamer[n=21]). Of 178 patients, 124 patients had live donor trans-plantation with at least one matched haplotype, whereas 54patients had cadaveric transplantations. Phosphate binderswere stopped in all patients after transplantation.In all patients, immunosuppression was performed by a

standard regimen on the basis of calcineurin inhibitors(cyclosporin in 82 patients and tacrolimus in 96 patients)associated with mycophenolate mofetil and prednisolone.Target tacrolimus and cyclosporin levels were set accord-ing to standard recommendations. By the end of the 12thweek after transplantation, the maintenance dose of pred-nisolone was reduced to 5 mg/d. Five patients developedpost-transplant diabetes during the 6-month follow-upperiod, and all patients were on prednisolone treatmentduring the same period (dose range=5–60 mg/d).As a control group, we enrolled 96 healthy controls

recruited by in-hospital advertisement, and the vast majorityof these volunteers was hospital personnel. These healthyvolunteers were specifically selected tomatch the age and sexdistributions of patients in this study. Routine analyses docu-mented that healthy volunteers had normal renal function(GFR.90 ml/min), no hypertension, diabetes, or disordersof lipid metabolism, and no other relevant disease.

Study ProtocolIn living transplant recipients, baseline biochemical and

IMT assessments were performed within 4 weeks beforetransplantation. In cadaveric transplant recipients, thesame measurements were performed during the routineclinical evaluation performed at the time of renal trans-plantation. Both in living and cadaveric transplant recip-ients, biochemical and IMT assessments were repeated 6months after transplantation. eGFR was calculated accord-ing to the simplified version of the Modification of Diet inRenal Disease equation. Homeostatic model assessment(HOMA), an index of insulin sensitivity, was calculatedwith the formula HOMA=fasting plasma glucose (milligramsper deciliter)3immunoreactive insulin (microinternationalunits per milliliter)/405. The methods of measurement of25 hydroxy vitamin D (25OHVD), PTH, FGF23 (second gen-eration, two-site mAb ELISA; Kainos Laboratories, Tokyo,Japan), high-sensitivity C-reactive protein (hsCRP), andserum insulin were described in detail in a previous pub-lication (24).

Common Carotid B-Mode Doppler UltrasoundMeasurement of IMTScanning was performed using the recommendations by

the American Society of Echocardiography (12) using an in-strument generating a wide-band ultrasonic pulse with amiddle frequency of 12 MHz (ATL 5000; Advanced Tech-nology Laboratories Inc., Bothell, WA). After an initial over-view of vessel orientation, wall thickness, plaques, andsurrounding structures, in all participants, a high-resolution

B-mode ultrasound of the common carotid arteries withscanning of the longitudinal axis until the bifurcation andthe transversal axis was performed. Measurements weremade in triplicate on each side by tracing far-wall blood in-tima and media adventitia interfaces using the leading edgeto leading edge method. All patients and controls wereblindly examined by one experienced operator (a licensedradiology technologist). IMT was measured, always inplaque-free areas, at 1 cm proximal to the bifurcation oneach side, and the average value was taken as an estimateof the IMT. The intraoperator variability for IMT at ourlaboratory was 4%. The upper limit of the normal rangeof IMT in our laboratory (established in 93 healthy controlsof the study) is 0.71 mm.

Statistical AnalysesNon-normally distributed variables were expressed as

medians (interquartile ranges), and normally distributedvariables were expressed as means6SDs. A P value ,0.05was considered to be statistically significant. Variableswith a non-normal distribution were appropriately logtransformed before analysis. Comparisons between twogroups were assessed by the paired t and Wilcoxon ranksums tests as appropriate. Multiple regression analysiswas applied to identify the independent correlates ofIMT. Tested risk factors included all Framingham risk fac-tors except for diabetes and smoking (because smokersand patients with diabetes were excluded from the study),biomarkers of bone mineral disorders (P, Ca, 25OHVD,PTH, and FGF23), CRP, albumin and insulin resistance asmeasured by the HOMA index, and the type of renal trans-plantation (living/cadaveric). Multiple regression modelswere built by including into the models all significant bivar-iate correlates of the main outcome measure (IMT). Themodels had sufficient power to test the independent associ-ation of this outcome measure with relevant correlates (i.e.,at least 10 observations per covariate in the same models)(25). All statistical analyses were performed by using theSPSS 15.0 (SPSS Inc., Chicago, IL) statistical package.

ResultsThe flowchart of patients enrolled into this study is

presented in Figure 1. Living related and cadaveric kidneytransplants (Table 1) had similar ages and causes of CKD.Antihypertensive and immune-suppressive agents werealso similar along with the dialysis treatment modalitybefore transplantation. Therefore, additional analyseswere performed in the combined population of patientswith living and cadaveric transplants. The mean age inthe control group of healthy subjects (33610 years old)was very close to that of patients (3269 years old), andthe sex distribution was also similar (control group: 73%men and 27% women; patient group: 71% men and 29%women).As shown in Table 2, average eGFR after transplantation

was 86614 ml/min per 1.73 m2, and no patient had aGFR,60 ml/min per 1.73 m2. After transplantation, therewas a mild fall in systolic and diastolic BP (24 and22 mmHg, respectively), a modest (+1.1 kg/m2) increasein body mass index, and a rise in serum total and LDLcholesterol and triglycerides. HOMA index reduced after

Clin J Am Soc Nephrol 10: ccc–ccc, March, 2014 IMT, Atherosclerosis, MBD in Renal Transplantation Patients, Yilmaz et al. 3

transplantation. Likewise, biomarkers of nutrition and in-flammation underwent favorable changes as indicated byan 11% increase in serum albumin and a substantial reduc-tion in hsCRP. Among CKD-MBD biomarkers, FGF23 felldramatically after transplantation (from 6443 to 30 pg/ml).Serum calcium and phosphate reverted to the normal range,and PTH fell to 60 pg/ml. Before transplantation, all patientshad either vitamin D insufficiency (i.e., plasma 25OHVD

levels between 50 and 72 nmol/L [n=61; 34.3%]) or frankdeficiency (i.e., 25OHVD levels#50 nmol/L [n=117;65.7%]). After transplantation, 25OHVD increased by about22%, and 62 patients (38%) attained vitamin D sufficiency.

IMT: Effect of Renal TransplantationBefore transplantation, IMT in patients on dialysis

(0.960.2 mm) was higher than in age- and sex-matched

Table 1. Demographic and clinical characteristics of the study population according to the type of transplant received

Patient Characteristics Living (n=124) Cadaveric (n=54) P Value

Sex (men/women) 93/31 33/21 0.06Age (yr) 31610 3368Etiology of chronic renal failure, n (%) 0.44GN 29 (23) 10 (18)Hypertension 23 (18) 6 (11)ADPKD 8 (6) 5 (9)Chronic pyelonephritis 5 (4) —Reflux nephropathy 8 (6) 5 (9)Amyloidosis 4 (3) 3 (5)Unknown 47 (38) 25 (46)

Antihypertensive drugs, n (%) 0.52a-Blockers 5 (4) 1 (2)Calcium-channel blockers 19 (15) 5(9)b-Blockers 6 (5) 1 (2)ab-Blockers 6 (5) 4 (7)

Immunosuppressant agents, n (%) ,0.001Cyclosporin 66 (53) 16 (30)Tacrolimus 58 (47) 38 (70)

Dialysis modality, n (%) 0.63Hemodialysis 111 (89) 47 (87)Peritoneal dialysis 13 (11) 7 (13)

ADPKD, autosomal dominant polycystic kidney disease.

Table 2. Clinical and hemodynamic data before and after transplantation

Parameters Before Transplantation After Transplantation P Value

Systolic BP (mmHg) 136 (110–195) 132 (90–155) ,0.001Diastolic BP (mmHg) 84 (65–110) 82 (50–97) ,0.001Body mass index (kg/m2) 25.462.6 26.562.3 ,0.001Total cholesterol (mg/dl) 188.5623.1 207.5626.2 ,0.001Triglycerides (mg/dl) 132.9622.3 140.2619.1 ,0.001LDL cholesterol (mg/dl) 117.4617.3 119.8617.3 0.18HDL cholesterol (mg/dl) 41.165.9 43.866.1 ,0.001Serum glucose (mg/dl) 86.7610.1 84.4610.7 0.04Insulin (IU/L) 7.161.9 5.961.4 ,0.001Homeostasis model assessment IR 1.560.5 1.260.3 ,0.001eGFR (ml/min per 1.73 m2) 6.364.0 85.8613.7 ,0.001Serum albumin (g/dl) 3.760.3 4.160.3 ,0.001High-sensitivity C-reactive protein (mg/L) 25 (4–53) 3.0 (1.0–9.2) ,0.001Serum calcium (mg/dl) 8.260.6 8.860.5 ,0.001Serum phosphate (mg/dl) 6.561.3 4.160.9 ,0.001Intact parathyroid hormone (pg/ml) 234 (89–776) 60 (21–132) ,0.00125 Hydroxy vitamin D (nmol/L) 44.869.9 54.4611.1 ,0.001Fibroblast growth factor 23 (pg/ml) 6443 (378–77,812) 30.4 (10.1–59.5) ,0.001Logn fibroblast growth factor 23 (pg/ml) 9.060.9 3.360.3 ,0.001Intima media thickness (mm) 0.960.2 0.760.1 ,0.001

Data are means6SDs or medians (ranges) as appropriate. IR, insulin resistance; logn, natural logarithm.

4 Clinical Journal of the American Society of Nephrology

healthy controls (0.660.1 mm) and exceeded the upperlimit of the normal range in 83.1% of patients. IMT de-creased by 22% after transplantation to 0.6860.13 mm(P=0.001).

Cross-Sectional Analyses of the Relationship between RiskFactors and IMT before and after TransplantationBefore transplantation, FGF23, hsCRP, systolic BP, and

HOMA index (P,0.001) as well as eGFR and to a weakerextent, 25OHVD and serum phosphate were all related toIMT (Table 3, pretransplantation IMT bivariate analyses). TheGFR, FGF23, and CRP were the strongest correlates of IMTbefore transplantation (Figure 2). Similarly, post-transplantation(Table 3, post-transplantation IMT bivariate analyses) FGF23,eGFR, PTH, and phosphate associated strongly with IMT,whereas diastolic BP, hsCRP, HOMA, 25OHVD, and serumcalcium showed weaker correlations with the same parameter.Of note, hsCRP and FGF23 before transplantation were directlyrelated each other (r=0.30), and this relationship was highlysignificant (P,0.001).Multiple regression models were built and included var-

iables associated with IMT at bivariate analyses (see above) aswell as other established risk factors for IMT, including age,sex, body mass index, LDL cholesterol, and serum albumin.Before transplantation (Table 3, pretransplantation IMT mul-tiple regression analyses), FGF23 (b=0.28; P,0.001), eGFR(b=0.23; P=0.002), systolic BP (b=0.17; P=0.01), and hsCRP(b=0.15; P=0.03) were all independent correlates of IMT.After transplantation (Table 3, post-transplantation IMT

multiple regression analyses), FGF23 and to a very slightdegree, eGFR, serum calcium and phosphate, and PTHmaintained an independent association with IMT. Thetype of renal transplantation (living/cadaveric) had noindependent relationship with IMT both pre- and post-transplantation (Table 3).

Longitudinal Analyses of Changes in Risk Factors afterTransplantation and Changes in IMTIn these analyses (Table 4), we tested the association

between changes (i.e., the difference between values ofeach variable after transplantation minus the value of thesame variable before transplantation) in individual riskfactors with simultaneous changes in IMT brought aboutby renal transplantation.Changes in FGF23 and CRP were the strongest bivariate

correlates of IMT (Figure 3, Table 4). In these unadjustedanalyses, IMT was also associated with eGFR, serum phos-phate and calcium, and to a weak extent, 25OHVD andPTH.Onmultiple regression analysis, changes in FGF23 (b=0.26;

P,0.001) and hsCRP (b=0.25; P=0.001) emerged as thestrongest independent correlates of IMT changes. Changesin eGFR and serum phosphate maintained an independentassociation with changes in IMT, but these relationshipswere fairly weak (Table 4). Similar to cross-sectional analy-ses, the type of renal transplantation (living/cadaveric) hadno independent relationship with changes in IMT after renaltransplantation (Table 4).

Table 3. Bivariate and multivariate standardized correlation coefficients (b) and P values between intima media thickness andpertinent variables before and after transplantation

Parameters

Pretransplantation IMT b (P Value) Post-Transplantation IMT b (P Value)

Bivariate Multipleregression Bivariate Multiple

regression

Age (yr) 0.06 (0.43) 0.06 (0.45) 20.04 (0.63) 20.13 (0.10)Sex 20.08 (0.28) 0.01 (0.94) 0.01 (0.96) 0.05 (0.55)Body mass index (kg/m2) 20.02 (0.82) 20.09 (0.22) 0.09 (0.26) 0.10 (0.19)LDL cholesterol (mg/dl) 0.08 (0.32) 0.10 (0.20) 0.04 (0.62) 0.12 (0.13)Systolic BP (mmHg) 0.27 (,0.001)a 0.17 (0.01)a 0.02 (0.79) 0.01 (0.90)Diastolic BP (mmHg) 0.07 (0.36) 20.05 (0.50) 0.15 (0.04)a 0.05 (0.49)High-sensitivity C-reactiveprotein (mg/L)

0.34 (,0.001)a 0.15 (0.03)a 0.16 (0.04)a 0.10 (0.18)

Serum albumin (g/dl) 20.06 (0.42) 20.07 (0.33) 0.07 (0.39) 0.11 (0.15)Homeostasis modelassessment index

0.26 (,0.001)a 0.04 (0.59) 0.15 (0.05)a 0.01 (0.88)

eGFR (ml/min per 1.73 m2) 20.40 (,0.001)a 20.23 (,0.001)a 20.29 (,0.001)a 20.16 (0.03)a

Fibroblast growthfactor 23 (pg/ml)

0.41 (,0.001)a 0.28 (,0.001)a 0.36 (,0.001)a 0.31 (,0.001)a

25 Hydroxy vitaminD (nmol/L)

20.22 (0.003)a 20.04 (0.64) 20.19 (0.01)a 20.09 (0.25)

Serum calcium (mg/dl) 20.09 (0.24) 20.07 (0.39) 20.22 (0.03)a 20.15 (0.03)a

Serum phosphate (mg/dl) 0.20 (,0.01)a 0.11 (0.17) 0.29 (,0.001)a 0.14 (0.04)a

Intact parathyroidhormone (pg/ml)

0.13 (0.09) 0.08 (0.29) 0.31 (,0.001)a 0.15 (0.03)a

Living/cadaverictransplantation

20.02 (0.81) 0.02 (0.77) 0.03 (0.69) 20.04 (0.56)

IMT, Intima media thickness.aSignificant association.

Clin J Am Soc Nephrol 10: ccc–ccc, March, 2014 IMT, Atherosclerosis, MBD in Renal Transplantation Patients, Yilmaz et al. 5

DiscussionThis study shows that IMT of the carotid arteries, an

intermediate phenotype and a valid surrogate biomarker ofatherosclerosis denoting intimal fibroplasia and muscle cellhypertrophy in the vascular wall, reduces after renaltransplantation. Reduction in IMT after renal transplanta-tion parallels the decline in CRP levels and the marked fallin FGF23 associated with restored renal function. Classicrisk factors and other risk factors apparently do notcontribute to explain the improvement in IMT after trans-plantation.Successful renal transplantation substantially reduces the

risk for cardiovascular complications in patients with trans-plants compared with well matched patients on dialysis onthe waiting list (1). In a study by Kasiske et al. (26) in the1990s, no relationship was observed between Framinghamrisk factors, including smoking, hypertension, and hyperlip-idemia, and the risk of myocardial infarction, cardiac death,

and coronary interventions in recipients of renal transplants(26). Among nonclassic risk factors, CRP (9) and biomark-ers of CKD-MBD (10) undergo dramatic changes aftertransplantation, and FGF23 and 1,25(OH)2 vitamin levelsapproach the normal range in most patients, whereas se-rum PTH often remains mildly to moderately elevated (27).Numerous low-powered cross-sectional ultrasound studiesof carotid arteries have been performed in patients withrenal transplants (28–32). The vast majority of these studiesconsidered isolated biomarkers (mainly biomarkers of in-flammation) (29,31,32) or just a limited set of purported riskfactors. Longitudinal studies have several advantages overcross-sectional studies for exploring causation. Indeed,these studies provide information about individual changesin the variables of interest, exclude between-subjects vari-ation from error, and allow investigation of the relationshipbetween predictor variables with relevant study endpoints (33).

Figure 2. | Relationship of the intimate media thickness before transplantation to the levels of fibroblast growth factor 23 (A) and high-sensitivity C-reactive protein (hsCRP) (B).

Table 4. Bivariate and multivariate associations between changes (D) in intima media thickness after renal transplantation andrelevant parameters

ParametersDIMT b (P Value)

Bivariate Multiple regression

DSystolic BP (mmHg) 0.11 (0.16) 0.03 (0.74)DDiastolic BP (mmHg) 0.04 (0.64) 20.06 (0.45)DLDL cholesterol (mg/dl) 0.02 (0.79) 0.04 (0.61)DAlbumin (g/L) 0.06 (0.44) 0.01 (0.91)DhsCRP (mg/L) 0.29 (,0.001)a 0.25 (0.001)a

DHOMA 0.23 (,0.001)a 0.13 (0.10)DeGFR (ml/min/1.73 m2) 20.21 (,0.01)a 20.16 (0.02)a

DLog FGF23 (pg/ml) 0.33 (,0.001)a 0.26 (,0.001)a

D25OHVD (nmol/L) 20.16 (0.04)a 20.08 (0.28)DSerum calcium (mg/dl) 20.23 (,0.001)a 20.14 (0.06)DSerum phosphate (mg/dl) 20.27 (0.001)a 0.15 (0.05)a

DiPTH (pg/ml) 0.16 (0.04)a 20.17 (0.34)Living/cadaveric renal transplantation 0.02 (0.82) 20.06 (0.46)

Data are standardized correlation coefficients (b) and P values. IMT, intima media thickness; hsCRP, high-sensitivity C-reactiveprotein; HOMA, homeostasis model assessment index; FGF23, fibroblast growth factor 23; 25OHVD, 25 hydroxy vitamin D; iPTH,intact parathyroid hormone.aStatistically significant relationship.

6 Clinical Journal of the American Society of Nephrology

With 178 patients, our longitudinal study had sufficientpower for testing the independent contribution of a large setof potential risk factors to the variance of IMT, includingclassic (Framingham) risk factors and risk factors relatedwithCKD, such as inflammation, insulin resistance, low albumin,and biomarkers of CKD-MBD. Even brief periods of systemicinflammation by infectious diseases determine a relevantincrease in IMT (+17%) within 3 months of the infectiousprocess, and such an increase is attenuated by appropriatedtreatment with antibiotics (34). Furthermore, this study is thefirst testing the evolution of the same parameters and FGF23along with simultaneous IMT measurements before and af-ter renal transplantation. FGF23 is strongly associated withwhole-body atherosclerosis burden as measured by nuclearmagnetic resonance imaging in a large study in the generalpopulation in Sweden (35). It is, therefore, of relevance toinvestigate whether similar links also exist in patients ondialysis before and after renal transplantation. In this study,we tried to minimize the confounding effects of preexistingcardiovascular damage (myocardial infarction, peripheralvascular disease, or cerebrovascular disease), diabetes, andsmoking on the evolution of arterial damage by excludingpatients with these risk factors. Although this approach maybe difficult to apply in the aging dialysis population with ahigh cardiovascular burden in most centers in northern Eu-rope and the United States, in Turkey, 65% of transplants aremade in patients on dialysis in the age range of 20–44 yearsold (36) (i.e., an age range substantially lower than in north-ern Europe and the United States). Likewise, we tried tolimit the confounding effect of drug treatment by excludingpatients on renin angiotensin blockers, statins, erythropoie-tin, active forms of vitamin D, and inhibitors of mammaliantarget of rapamycin, all drugs that per se may influence vas-cular disease. Our analyses in this selected population showthat FGF23 and serum CRP are the strongest independentcorrelates of IMT in both baseline analyses before transplan-tation and longitudinal analyses incorporating measure-ments made before and after transplantation. Ourobservations suggest that attenuation of inflammation andalmost complete correction of high FGF23 levels may play arelevant role in the amelioration of arterial disease after

successful renal transplantation. CRP and FGF23 are associ-ated with each other in patients with CKD (37), and theserisk factors were directly related in baseline analyses (beforetransplantation) in this study. However, both of these bio-markers resulted in independent correlates of the evolutionof IMT after transplantation, suggesting that amelioration ofinflammation and correction of high FGF23 may attenuatecarotid atherosclerosis by separate biologic pathways. In linewith observational studies in patients with transplantsshowing no relationship between classic risk factors andall-cause and cardiovascular risk factors (26), in this study,hypertension, other Framingham risk factors, and insulin re-sistance were not associated with longitudinal changes inIMT after transplantation. Overall, our data are in keepingwith the hypothesis that the reduction of CRP levels andabatement of FGF23 levels brought about by restored renalfunction contribute to reverse arterial damage in patientswith transplants.Our study has limitations. First, although useful to explore

causal hypotheses, longitudinal studies do not prove causal-ity. Second, as briefly alluded to before, patients in our cohortwere all Turkish and younger than patients with transplantsin American and European registries of dialysis and trans-plantation, which limits the generalizability of our findings.Third, although we tried to limit the effect of drug treatmenton data interpretation, calcineurin inhibitors per se have anadverse effect on arterial disease in patients with transplants.Patients in our cohort were all treated with these drugs.Therefore, being evenly administered, it is unlikely that theseimmunosuppressant agents could have disturbed (to an im-portant extent) the appreciation of the relationships emergedin this study.In conclusion, IMT improves after renal transplantation,

and such an improvement parallels the reduction in CRPlevels and the dramatic fall in FGF23 associated withrestored renal function. These findings are in keeping withthe hypothesis that nonclassic risk factors may play a rolein vascular disease in patients with coeval stage 5 CKD ondialysis and that the reduction in cardiovascular risk aftertransplantation is, at least in part, driven by an ameliora-tion of these risk factors, like CRP and CKD-MBD.

Figure 3. | Relationship of the alteration of the intimate media thickness 2 weeks after the transplantation with the alteration of fibroblastgrowth factor 23 (A) or high-sensitivity C-reactive protein (hsCRP) levels (B).

Clin J Am Soc Nephrol 10: ccc–ccc, March, 2014 IMT, Atherosclerosis, MBD in Renal Transplantation Patients, Yilmaz et al. 7

AcknowledgmentsThe authors thank the Familial Mediterranean Fever Arthritis

Vasculitis andOrphanDiseasesResearch (FAVOR;www.favor.org.tr) web registries at Gulhane Medical Academy, Institute of HealthSciences for statistical analysis and contribution to the preparationof this manuscript.

DisclosuresNone.

References1. Wolfe RA, Ashby VB, Milford EL, Ojo AO, Ettenger RE, Agodoa

LY, Held PJ, Port FK: Comparison of mortality in all patients ondialysis, patients on dialysis awaiting transplantation, andrecipients of a first cadaveric transplant.NEngl JMed 341: 1725–1730, 1999

2. Hunsicker LG: A survival advantage for renal transplantation. NEngl J Med 341: 1762–1763, 1999

3. vanDijk PC, Jager KJ, de Charro F, Collart F, Cornet R,Dekker FW,Gronhagen-Riska C, Kramar R, Leivestad T, Simpson K, Briggs JD;ERA-EDTA registry: Renal replacement therapy in Europe: Theresults of a collaborative effort by the ERA-EDTA registry and sixnational or regional registries. Nephrol Dial Transplant 16:1120–1129, 2001

4. Kasiske BL, Chakkera HA, Roel J: Explained and unexplainedischemic heart disease risk after renal transplantation. J Am SocNephrol 11: 1735–1743, 2000

5. Stenvinkel P, Wanner C, Metzger T, Heimburger O, Mallamaci F,Tripepi G, Malatino L, Zoccali C: Inflammation and outcome inend-stage renal failure: Does female gender constitute a survivaladvantage? Kidney Int 62: 1791–1798, 2002

6. Benedetto FA, Tripepi G, Mallamaci F, Zoccali C: Rate of ath-erosclerotic plaque formation predicts cardiovascular events inESRD. J Am Soc Nephrol 19: 757–763, 2008

7. Hruska KA, Mathew S, Lund RJ, Memon I, Saab G: The patho-genesis of vascular calcification in the chronic kidney diseasemineral bone disorder: The links between bone and the vascu-lature. Semin Nephrol 29: 156–165, 2009

8. London GM, Guerin AP, Verbeke FH, Pannier B, Boutouyrie P,Marchais SJ, Metivier F: Mineral metabolism and arterial functionsin end-stage renal disease: Potential role of 25-hydroxyvitamin Ddeficiency. J Am Soc Nephrol 18: 613–620, 2007

9. Simmons EM, LangoneA, SezerMT, Vella JP, Recupero P,MorrowJD, Ikizler TA, Himmelfarb J: Effect of renal transplantation onbiomarkers of inflammation and oxidative stress in end-stagerenal disease patients. Transplantation 79: 914–919, 2005

10. Molnar MZ, Naser MS, Rhee CM, Kalantar-Zadeh K,Bunnapradist S: Bone and mineral disorders after kidney trans-plantation: Therapeutic strategies. Transplant Rev (Orlando) 28:56–62, 2014

11. Kidney Disease: Improving Global Outcomes (KDIGO) CKD-MBD Work Group: KDIGO clinical practice guideline for thediagnosis, evaluation, prevention, and treatment of ChronicKidneyDisease-Mineral and BoneDisorder (CKD-MBD). KidneyInt Suppl 113: S1–S130, 2009

12. Stein JH, Korcarz CE, Hurst RT, Lonn E, Kendall CB, Mohler ER,Najjar SS, Rembold CM, Post WS; American Society of Echo-cardiography Carotid Intima-Media Thickness Task Force; En-dorsed by the Society for Vascular Medicine: Use of carotidultrasound to identify subclinical vascular disease and evaluatecardiovascular disease risk: A consensus statement from theAmerican Society of Echocardiography Carotid Intima-MediaThickness Task Force. J Am Soc Echocardiogr 21: 93–111, quiz189–190, 2008

13. Tardif JC,Heinonen T,Orloff D, Libby P: Vascular biomarkers andsurrogates in cardiovascular disease. Circulation 113: 2936–2942, 2006

14. European Medicines Agency: Pre-Authorisation Evaluation ofMedicines for Human Use. Available at: http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2009/09/WC500003242.pdf. Accessed January 10, 2010

15. SzetoCC,ChowKM,WooKS, Chook P, Ching-Ha KwanB, LeungCB, Kam-Tao Li P: Carotid intima media thickness predicts

cardiovascular diseases in Chinese predialysis patients withchronic kidney disease. J Am Soc Nephrol 18: 1966–1972, 2007

16. Benedetto FA, Mallamaci F, Tripepi G, Zoccali C: Prognosticvalue of ultrasonographic measurement of carotid intima mediathickness in dialysis patients. J Am Soc Nephrol 12: 2458–2464,2001

17. Suwelack B, Gerhardt U, Witta J, Hillebrandt U, Hohage H:Effect of parathyroid hormone levels on carotid intima-mediathickness after renal transplantation. Am J Hypertens 14: 1012–1018, 2001

18. Wolf M: Update on fibroblast growth factor 23 in chronic kidneydisease. Kidney Int 82: 737–747, 2012

19. Meier-Kriesche HU, Schold JD, Srinivas TR, Reed A, Kaplan B:Kidney transplantation halts cardiovascular disease progressionin patientswith end-stage renal disease.Am J Transplant 4: 1662–1668, 2004

20. Yilmaz MI, Sonmez A, Saglam M, Yaman H, Kilic S, Turker T,Unal HU, Gok M, Cetinkaya H, Eyileten T, Oguz Y, Caglar K,Vural A, Mallamaci F, Zoccali C: Longitudinal analysis of vas-cular function and biomarkers of metabolic bone disorders be-fore and after renal transplantation. Am J Nephrol 37: 126–134,2013

21. Joannides R, Etienne I, Iacob M, Hurault de Ligny B, Barbier S,Bellien J, Lebranchu Y, Thuillez C, GodinM: Comparative effectsof sirolimus and cyclosporin on conduit arteries endothelialfunction in kidney recipients. Transpl Int 23: 1135–1143, 2010

22. Seckinger J, Sommerer C, Hinkel UP, Hoffmann O, Zeier M,Schwenger V: Switch of immunosuppression from cyclosporineA to everolimus: Impact on pulse wave velocity in stable de-novorenal allograft recipients. J Hypertens 26: 2213–2219, 2008

23. Beiras-Fernandez A,Walther S, Thein E,Muenzing S, Hammer C:Influence of polyclonal ATGs on expression of adhesion mole-cules: An experimental study. Transplant Proc 37: 1944–1946,2005

24. YilmazMI, Sonmez A, SaglamM, YamanH, Kilic S, Demirkaya E,Eyileten T, Caglar K, Oguz Y, Vural A, Yenicesu M, Zoccali C:FGF-23 and vascular dysfunction in patients with stage 3 and 4chronic kidney disease. Kidney Int 78: 679–685, 2010

25. Kleinbaum DG, Kupper LL, Muller KE,Nizam A: Applied Re-gression Analysis and Other Multivariable Methods, Boston,Duxbury Press, 1998, pp 389–390

26. Kasiske BL, Guijarro C, Massy ZA, Wiederkehr MR, Ma JZ: Car-diovascular disease after renal transplantation. J Am SocNephrol7: 158–165, 1996

27. Kalantar-Zadeh K, Molnar MZ, Kovesdy CP, Mucsi I,Bunnapradist S: Management of mineral and bone disorder afterkidney transplantation. Curr Opin Nephrol Hypertens 21: 389–403, 2012

28. Ca~nas L, Bayes B, Granada ML, Ibernon M, Porrini E, Benıtez R,Dıaz JM, Lauzurica R, Moreso F, Torres A, Lampreabe I, Serra A,Romero R: Is adiponectin a marker of preclinical atherosclerosisin kidney transplantation? Clin Transplant 26: 259–266, 2012

29. Turkmen K, Tonbul HZ, Erdur FM, Toker A, Biyik Z, Ozbiner H,Gaipov A, Gul EE, Kayrak M, Solak Y, Ozbek O, Turk S, Covic A:Soluble TWEAK independently predicts atherosclerosis in renaltransplant patients. BMC Nephrol 14: 144, 2013

30. Uchida J, Machida Y, Iwai T, Kuwabara N, Kabei K, Naganuma T,Kumada N, Nakatani T: Glucose intolerance is associated withincreased intimal-medial thickness of the carotid artery and in-creased pulse-wave velocity in renal transplant recipients.Transplant Proc 45: 1535–1539, 2013

31. CobanogluAK,GungorO, Kircelli F, Altunel E, AsciG,Ozbek SS,Toz H, Ok E: Role of asymmetric dimethylarginine in the pro-gression of carotid atherosclerosis in renal transplant patients. IntUrol Nephrol 45: 1463–1469, 2013

32. Gungor O, Kismali E, Sisman AR, Kircelli F, Carrero JJ, Tatar E,Asci G, Toz H: The relationships between serum sTWEAK, FGF-23 levels, and carotid atherosclerosis in renal transplant patients.Ren Fail 35: 77–81, 2013

33. Grootendorst DC, Jager KJ, Zoccali C, Dekker FW:Observationalstudies are complementary to randomized controlled trials.Nephron Clin Pract 114: c173–c177, 2010

34. Liuba P, Persson J, Luoma J, Yla-Herttuala S, Pesonen E: Acuteinfections in children are accompanied by oxidative modifica-tion of LDL and decrease ofHDL cholesterol, and are followed by

8 Clinical Journal of the American Society of Nephrology

thickening of carotid intima-media. Eur Heart J 24: 515–521,2003

35. MirzaMA,Hansen T, Johansson L, AhlstromH, Larsson A, Lind L,Larsson TE: Relationship between circulating FGF23 and totalbody atherosclerosis in the community.Nephrol Dial Transplant24: 3125–3131, 2009

36. Serdengecti K, Suleymanlar G, AltiparmakMR, Seyahi N: TurkishRegistry of Dialysis and Transplantation. Available at: http://tsn.org.tr/folders/file/registry%20kitap.pdf. Accessed March 25,2010

37. Munoz Mendoza J, Isakova T, Ricardo AC, Xie H, NavaneethanSD, Anderson AH, Bazzano LA, Xie D, Kretzler M, Nessel L,Hamm LL, Negrea L, LeonardMB, Raj D,Wolf M; Chronic RenalInsufficiency Cohort: Fibroblast growth factor 23 and In-flammation in CKD. Clin J Am Soc Nephrol 7: 1155–1162, 2012

Received: August 13, 2014 Accepted: November 24, 2014

Published online ahead of print. Publication date available at www.cjasn.org.

Clin J Am Soc Nephrol 10: ccc–ccc, March, 2014 IMT, Atherosclerosis, MBD in Renal Transplantation Patients, Yilmaz et al. 9