Dissecting Inflammation in ESRD: Do Cytokines and C-Reactive Protein Have a Complementary Prognostic Value for Mortality in Dialysis Patients?
Carmine Zoccali*,,
Giovanni Tripepi and
Francesca Mallamaci*,
* Nephrology, Hypertension & Renal Transplantation Unit, Ospedali Riuniti, Reggio Calabria, and CNR-IBIM Clinical Epidemiology of Renal Diseases and Hypertension, Calabria, Italy
Address correspondence to: Prof. Carmine Zoccali, Nephrology, Hypertension & Renal Transplantation Unit & CNR-IBIM, Clinical Epidemiology of Renal Diseases and Hypertension, c/o Ki Point Gransial SRL, Via Filippini n. 85, 89100 Reggio Calabria, Italy. Phone: +39-0965-397010; Fax: +39-0965-26879; E-mail: carmine.zoccali{at}tin.it
Because atherogenesis represents a type of chronic inflammationthat involves multiple elements of the inflammatory-immune response,the simultaneous prediction power for death of C-reactive protein(CRP) and proinflammatory cytokines (IL-1, IL-6, IL-18, andTNF-) was tested in a cohort of 217 patients with ESRD. Duringthe follow-up period (average 41 mo), 112 patients died. Inan analysis that was adjusted for other risk factors, the relativerisks for death of patients who were exposed to high levelsof one, two, three, and four or more inflammation biomarkerswere 1.48, 1.64, 2.76, and 3.05 times higher, respectively,than that of patients in the reference category (no inflammation).In this model, the explained variation in mortality that wasattributable to overall inflammation burden (+9.1%) was marginallyhigher (P = 0.06) than that provided by IL-6 alone (+6.1%).In an alternative analysis based on the Bayesian approach (receiveroperating characteristic curves analysis), the prediction powerof the combined inflammatory burden was identical to that providedby the sole IL-6 (0.59 ± 0.04 versus 0.59 ± 0.04).IL-6 captures almost entirely the prediction power of the overallinflammation burden in patients with ESRD. IL-6 seems to bean almost ideal indicator of the severity of inflammation. Theuse of this biomarker can be recommended in clinical studiesthat aim to better the understanding of inflammation or to modifyit in this population.
Substantial evidence has been accrued that inflammation is amajor factor in the high mortality of ESRD, and a variety ofexperimental and epidemiologic studies coherently indicate thatcytokines and other inflammatory proteins are not only conduciveto cardiovascular damage in experimental models but also predictiveof cardiovascular events in patients with ESRD. We recentlyreported that IL-6, a key player in the acute-phase response,is the strongest predictor of mortality among inflammation markersand that C-reactive protein (CRP) seems to be the most suitablemarker to grade the inflammation status in clinical practiceon cost-effectiveness grounds (1). Whereas the predictive powerof individual inflammatory proteins now is well established,the important question of whether the overall inflammation burdenthat is estimated by the combined measurement of these moleculesprovides additional prognostic information in comparison withthe measurement of individual cytokines and acute-phase reactantshas not been studied. The issue is relevant because the majorcytokines—IL-1, IL-6, IL-18, TNF-, and CRP, the pentraxinthat is synthesized by the liver under the stimulus of thesecytokines—have pleiotropic and redundant actions at leastin part reflecting complementary aspects of the inflammatory-immuneprocess (2,3).
In the present study, we therefore tested the prediction powerfor all-cause mortality of overall inflammation burden as estimatedby the measurement of plasma concentration of CRP and of IL-1,IL-6, IL-18, and TNF- in a cohort of patients with ESRD. Theresults show that IL-6 captures almost entirely the prognosticpower for death of the inflammatory process in patients withESRD.
Protocol
The protocol conformed to the ethical guidelines of our institution,and informed consent was obtained from each participant. Forhemodialysis (HD) patients, all studies were performed duringa midweek nondialysis day; for continuous ambulatory peritonealdialysis (CAPD) patients at empty abdomen, studies were performedbetween 8:00 a.m. and 1:00 p.m.
Study Cohort
A total of 217 patients with ESRD (120 men and 97 women) whohad been on regular dialysis treatment for at least 6 mo (medianduration 43 mo; interquartile range 20 to 101 mo) without historyof congestive heart failure and without intercurrent inflammatoryillnesses were considered eligible for the study. The main demographicand clinical characteristics of the study population are detailedin Table 1. The prevalence of diabetes in this cohort was 15%(33 of 217 patients).
Table 1. Clinical and biochemical parameters of the study populationa
HD patients were treated thrice weekly with standard bicarbonatedialysis (138 mmol/L Na, 35 mmol/L HCO3, 1.5 mmol/L K, 1.25mmol/L Ca, and 0.75 mmol/L Mg) either with cuprophane or semisyntheticmembranes (dialysis filters surface area 1.1 to 1.7 m2). Theaverage urea Kt/V in these patients was 1.22 ± 0.27.All patients who were on CAPD were on four exchanges per daywith standard dialysis bags. The average weekly Kt/V in thesepatients was 1.69 ± 0.29. Eighty-six patients were habitualsmokers (22 ± 17 cigarettes/d). A total of 110 patientswere on treatment with erythropoietin. Ninety-seven patientswere being treated with antihypertensive drugs (72 on monotherapywith angiotensin-converting enzyme inhibitors, AT-1 antagonists,calcium channel blockers, or and blockers, and 25 were ondouble or triple therapy with various combinations of thesedrugs).
Follow-Up
After the initial assessment, patients were followed up foran average of 41 mo (range 0.8 to 70.0 mo). During the follow-up,fatal cardiovascular events (myocardial infarction, electrocardiogram-documentedarrhythmia, heart failure, stroke, and other thrombotic eventsexcept for arteriovenous fistula thromboses) and death wererecorded accurately. Each death was reviewed and assigned anunderlying cause by a panel of five physicians. As a part ofthe review process, all available medical information abouteach death was collected. This information always included studyand hospitalization records. In the case of an out-of-hospitaldeath, family members were interviewed by telephone for betterascertainment of the circumstances surrounding death.
Laboratory Measurements
Blood sampling was performed after an overnight fast between8:00 a.m. and 10:00 a.m. always during a midweek nondialysisday for HD patients and at empty abdomen for CAPD patients.After 20 to 30 min of quiet resting in semirecumbent position,samples were taken into chilled EDTA Vacutainers, placed immediatelyon ice, and centrifuged within 30 min at –4°C, andthe plasma was stored at –80°C before assay. Serumlipids, albumin, calcium, phosphate, and hemoglobin measurementswere made using standard methods in the routine clinical laboratory.The plasma concentrations of asymmetric dimethylarginine andhomocysteine were determined as reported elsewhere (4,5). SerumCRP was measured by using a commercially available kit (immunonephelometricmethod, lower limit of detection 3.5 mg/L; Behring, Scoppito,L’Aquila, Italy). Serum levels of IL-6, IL-1, IL-18, andTNF- were measured by ELISA with the use of Quantikine HighSensitivity kits (intra-assay coefficient of variation for thesesubstances 1.6 to 10%; interassay coefficient of variation 3.3to 10.2%; R&D Systems, Minneapolis, MN).
Statistical Analyses
Data are reported as mean ± SD (normally distributeddata), median, and interquartile range (nonnormally distributeddata) or as percentage frequency, and comparisons among groupswere made by a P for trend. The combined prognostic power ofinflammatory proteins (inflammatory burden) for all-cause mortalitywas analyzed by a score that was calculated as the sum of fivebiomarkers (CRP, IL-1, IL-6, IL-18, and TNF-), each definedin categorical terms (1 = third tertile of the correspondingdata distribution; 0 = others two tertiles). We used the thirdtertile of each biomarker because in a previous article, thisfunctional form of variable provided adequate risk stratificationin the same dialysis cohort (1). We constructed six separateCox models on the basis of either the inflammation score orinflammation biomarkers considered individually. These modelsincluded all covariates that were associated to all-cause deathwith P < 0.10 at univariate Cox regression analysis as wellas variables that resulted to be associated with the inflammatoryscore with P < 0.10 at univariate analysis (Table 1). Thetreatment modality (HD/CAPD) was always introduced into theseCox models. The Cox models then were compared by using the –2log likelihood test (6). Furthermore, we performed an alternativeanalysis by the Bayesian approach (i.e., by comparing the areasunder the receiver operating characteristic [ROC] curves ofthe combined inflammatory load with that of IL-6 in the predictionof death [7]).
Hazard ratios and their 95% confidence intervals were calculatedwith the use of the estimated regression coefficients and theirstandard errors in the Cox regression analysis. All calculationswere made using a standard statistical package (SPSS for WindowsVersion 9.0.1; SPSS, Chicago, IL).
The cutoff values of serum CRP, IL-1, IL-6, IL-18, and TNF-that identified the third tertile of the relative data distributionwere 12.9 mg/L, 0.87 pg/ml, 8.5 pg/ml, 676 pg/ml, and 9.0 pg/ml,respectively. In the aggregate, 187 (86%) of 217 patients hadat least one biomarker within the corresponding third tertile(one biomarker in 74 cases; two biomarkers in 65 cases; threebiomarkers in 32 cases, and four or more biomarkers in 16 cases),whereas a minority of patients had all cytokines within thefirst two tertiles of the relative data distribution (n = 30;14%). As shown in Table 1, the number of altered (third tertile)inflammatory biomarkers was related directly to age and inverselyto serum albumin and tended to be inversely associated withserum cholesterol (P = 0.07) and homocysteine (P = 0.10). Asexpected, the average plasma concentration of each biomarkerrose in a stepwise manner from the reference group (no inflammation)to the group that showed four or more biomarkers in the correspondingthird tertile.
Overall Inflammatory Burden and Mortality: –2 Log-Likelihood Statistics
During the follow-up period, 112 patients died, 65 of them (58%of total deaths) of cardiovascular causes. The independent prognosticvalue of overall inflammation burden for all-cause mortalitywas analyzed in a multiple Cox regression analysis that adjustedfor a series of traditional and nontraditional risk factors.In this analysis, the hazard ratios of all-cause mortality increasedin a dose-response manner according to the number of altered(third tertile) biomarkers so that the relative risks of patientswho were exposed to one, two, three, and four or more increasedlevels of these inflammatory proteins were 1.48, 1.64, 2.76,and 3.05 times higher, respectively, than in those in the referencecategory (Table 2). However, the explained variation in mortalitythat was attributable to overall inflammatory burden was marginally(P = 0.06) higher than that provided by IL-6 (9.1 versus 6.1%;P = 0.06), which by far resulted to be the biomarker that showedthe highest (P < 0.01) prognostic power for death (CRP +3.0%;IL-18 +2.9%; IL-1 +0.5%; and TNF- +0.5%).
Table 2. Multiple Cox regression analysis of all-cause deatha
ROC Curves Analysis
The prediction power for mortality of overall inflammation burdenand of IL-6 was compared further by the analysis of ROC curves.By this approach, the areas under the corresponding ROC curvesboth were significantly superior to the threshold of diagnosticindifference (P = 0.03 and P = 0.04, respectively) but identicalbetween them (0.59 ± 0.04 versus 0.59 ± 0.04).
An inflammation score that was composed of CRP, IL-6, IL-1,IL-18, and TNF- predicts death no better than the sole IL-6in patients with ESRD. This finding indicates that IL-6 capturesalmost entirely the high risk for inflammation in ESRD and furtherhighlights the importance of high levels of this cytokine inpatients who are on long-term dialysis.
Overlapping and Complementary Role of Inflammatory Proteins and Atherosclerosis
Mainly because of the intellectual drive by Ross (3), atherosclerosisnow is seen as an inflammatory disease, and substantial evidencehas been accrued that the inflammation is a major contributorto arterial disease in patients with ESRD (1,8). Proinflammatoryand anti-inflammatory cytokines are crucial elements in suchprocess because the balance and the interplay of these proteinscontrol the direction, amplitude, and duration of inflammationas well as tissue remodeling. Cytokines and acute-phase responseproteins intervene at various levels in the atherosclerosisprocess. Thus, CRP seems to be implicated in endothelial dysfunction,foam cell formation, and inhibition of progenitor cell survivaland differentiation (2). IL-6, a ubiquitous cytokine that isfundamental for leukocyte and endothelial cell activation, ishighly expressed at the shoulder of the atherosclerotic plaque,where it may induce plaque instability by driving the localsynthesis of matrix metalloproteinases, monocyte chemotacticprotein-1, and TNF- (9). Furthermore, IL-6 stimulates fibrinogensynthesis in the liver via IL-6–responsive sequences inthe promoter region of the fibrinogen gene, thereby enhancingthe risk for thrombosis (10). IL-1 (11) and IL-18 (12) playa relevant role in the inflammatory component of atherosclerosisin experimental models. TNF-, besides being involved in plaqueinstability, is implicated in myocardial dysfunction and remodelingafter an acute ischemic insult (13). The disparate interferencesof cytokines and CRP with atherothrombosis suggest that thecombined measurement of these biomarkers may provide prognosticinformation that is superior to what can be derived by individualbiomarkers.
Individual and Combined Prediction Power of Inflammation Biomarkers in ESRD
In patients who have ESRD without clinically apparent inflammatorycomplications, independent of exposure to dialysis membranesand/or contaminated dialysis fluid, plasma levels of IL-6, IL-1,and TNF- and of CRP are raised substantially (8). The detrimentaleffect of inflammation in ESRD seems particularly prominentin the cardiovascular system because high CRP and high cytokinelevels signal a high risk for mortality and incident cardiovascularcomplications (1,14–19). We previously showed that theplasma concentrations of IL-6, CRP, and other cytokines aremutually interrelated in dialysis patients without overt inflammatoryprocesses (1). The hypothesis that the overall inflammationburden that is estimated by the combined measurement of cytokinesand CRP has not been examined so far in patients with ESRD.To address this problem, we constructed a composite score basedon the categorization of each inflammation marker into tertiles.We did so because this categorization allowed adequate riskstratification in our study cohort (1). Somewhat unexpected,this analysis showed that the combined use of these biomarkersprovides very small, if any, gain in prediction power for deathin comparison with that provided by the sole IL-6. The "inflammationscore" was expected to capture the prognostic power of nonoverlappingor partially overlapping pathogenetic pathways that lead toarterial injury better than IL-6 or other individual cytokines.In other words, we hypothesized that this score reflects severalcritical aspects of the offending potential of inflammation,from endothelial dysfunction and plaque formation to plaqueinstability and atherothrombosis. Contrary to our expectations,the sole measurement of IL-6 was almost as informative as thecomposite inflammation score. IL-6 perhaps is the most investigatedcytokine in human diseases (20). This protein has a very markedpleiotropy, being involved not only in inflammation but alsoin the regulation of several fundamental organ functions. Apartfrom the action on multipotent hematopoietic cell progenitorsand on B cells or from its stimulatory role for the synthesisof other cytokines and acute-phase reactants in the liver, IL-6affects myocardial cell growth, stimulates thermogenesis, suppressesthe thyroid axis, and induces growth hormone secretion (20).This large variety of effects makes IL-6 unique among proinflammatorycytokines and indicates that it represents not only a potentialagent of organ damage in pathophysiologic conditions but alsoa potent marker of the overall severity of the inflammationprocess. Our observation that IL-6 captures almost entirelythe predictive power of inflammation in ESRD is in keeping withfindings in the general population, in which IL-6 emerged asthe sole inflammation marker to predict progression of atherosclerosisin an analysis that considers also CRP, adhesion molecules,and E-selectin (21). The marked pleiotropy of IL-6 makes itthe best inflammation marker in ESRD, its prediction power beingalmost equal to what can be derived from simultaneous measurementsof several other cytokines and CRP.
IL-6 captures almost entirely the prediction power of the overallinflammation burden in patients with ESRD. IL-6 seems to bean almost ideal indicator of the severity of inflammation. Theuse of this biomarker can be recommended in clinical studiesthat aim to better our understanding of inflammation or to modifyit in this population.
Tripepi G, Mallamaci F, Zoccali C: Inflammation markers, adhesion molecules, and all-cause and cardiovascular mortality in patients with ESRD: Searching for the best risk marker by multivariate modeling.
J Am Soc Nephrol 16 [Suppl 1]
: S83
–S88, 2005[CrossRef]
Armstrong EJ, Morrow DA, Sabatine MS: Inflammatory biomarkers in acute coronary syndromes: Part I—Introduction and cytokines.
Circulation 113
: e72
–e75, 2006[Free Full Text]
Ross R: Atherosclerosis: An inflammatory disease.
N Engl J Med 340
: 115
–126, 1999[Free Full Text]
Zoccali C, Bode-Boger S, Mallamaci F, Benedetto F, Tripepi G, Malatino L, Cataliotti A, Bellanuova I, Fermo I, Frolich J, Boger R: Plasma concentration of asymmetrical dimethylarginine and mortality in patients with end-stage renal disease: A prospective study.
Lancet 358
: 2113
–2117, 2001[CrossRef][Medline]
Mallamaci F, Zoccali C, Tripepi G, Fermo I, Benedetto FA, Cataliotti A, Bellanuova I, Malatino LS, Soldarini A: Hyperhomocysteinemia predicts cardiovascular outcomes in hemodialysis patients.
Kidney Int 61
: 609
–614, 2002[CrossRef][Medline]
Collet D:
Modelling Survival Data in Medical Research. Tests in Statistical Science, London, Chapman & Hall, 1994
, pp 72
–85
Altman DG: Two further ways for looking at diagnostic tests: The likelihood ratio and the ROC curve. In:
Practical Statistics for Medical Research, edited by Altman DG, London, Chapman & Hall, 1991
, pp 416
–418
Zoccali C, Mallamaci F, Tripepi G: Inflammation and atherosclerosis in end stage renal disease.
Blood Purif 21
: 29
–36, 2003[CrossRef][Medline]
Armstrong EJ, Morrow DA, Sabatine MS: Inflammatory biomarkers in acute coronary syndromes: Part II—Acute-phase reactants and biomarkers of endothelial cell activation.
Circulation 113
: e152
–e155, 2006[Free Full Text]
Verschuur M, de Jong M, Felida L, de Maat MP, Vos HL: A hepatocyte nuclear factor-3 site in the fibrinogen beta promoter is important for interleukin 6-induced expression, and its activity is influenced by the adjacent –148C/T polymorphism.
J Biol Chem 280
: 16763
–16771, 2005[Abstract/Free Full Text]
Merhi-Soussi F, Kwak BR, Magne D, Chadjichristos C, Berti M, Pelli G, James RW, Mach F, Gabay C: Interleukin-1 plays a major role in vascular inflammation and atherosclerosis in male apolipoprotein E-knockout mice.
Cardiovasc Res 66
: 583
–589, 2005[Abstract/Free Full Text]
Elhage R, Jawien J, Rudling M, Ljunggren HG, Takeda K, Akira S, Bayard F, Hansson GK: Reduced atherosclerosis in interleukin-18 deficient apolipoprotein E knockout mice.
Cardiovasc Res 59
: 234
–240, 2003[Abstract/Free Full Text]
Nian M, Lee P, Khaper N, Liu P: Inflammatory cytokines and postmyocardial infarction remodeling.
Circ Res 94
: 1543
–1553, 2004[Abstract/Free Full Text]
Zoccali C, Benedetto FA, Mallamaci F, Tripepi G, Fermo I, Foca A, Paroni R, Malatino LS: Inflammation is associated with carotid atherosclerosis in dialysis patients.
J Hypertens 18
: 1207
–1213, 2000[CrossRef][Medline]
Stenvinkel P, Heimburger O, Jogestrand T: Elevated interleukin-6 predicts progressive carotid artery atherosclerosis on dialysis patients: Association with Chlamydia pneumoniae seropositivity.
Am J Kidney Dis 39
: 274
–282, 2002[Medline]
Zimmermann J, Herrlinger S, Pruy A, Metzger T, Wanner C: Inflammation enhances cardiovascular risk and mortality in hemodialysis patients.
Kidney Int 55
: 648
–658, 1999[CrossRef][Medline]
Yeun JY, Levine RA, Mantadilok V, Kaysan GA: C-reactive protein predicts all cause and cardiovascular mortality in hemodialysis patients.
Am J Kidney Dis 35
: 469
–476, 2000[Medline]
Kimmel PL, Phillips TM, Simmens SJ, Peterson RA, Weihs KL, Alleyne S, Cruz I, Yanovski JA, Veis JH: Immunological function and survival in hemodialysis patients.
Kidney Int 54
: 236
–244, 1998[CrossRef][Medline]
Stenvinkel P, Wanner C, Metzger T, Heimburger O, Mallamaci F, Tripepi G, Malatino LS, Zoccali C: Inflammation and outcome in end stage renal failure: Does female gender constitute a survival advantage?
Kidney Int 62
: 1791
–1798, 2002[CrossRef][Medline]
Papanicolaou DA, Wilder RL, Manolagas SC: Chrousos GP: The pathophysiologic roles of interleukin-6 in human disease.
Ann Intern Med 128
: 127
–137, 1998[Abstract/Free Full Text]
Tzoulaki I, Murray GD, Lee AJ, Rumley A, Lowe GD, Fowkes FG: C-reactive protein, interleukin-6, and soluble adhesion molecules as predictors of progressive peripheral atherosclerosis in the general population: Edinburgh Artery Study.
Circulation 112
: 976
–983, 2005[Abstract/Free Full Text]
This article has been cited by other articles:
M. Tong, J. J. Carrero, A. R. Qureshi, B. Anderstam, O. Heimburger, P. Barany, J. Axelsson, A. Alvestrand, P. Stenvinkel, B. Lindholm, et al. Plasma Pentraxin 3 in Patients with Chronic Kidney Disease: Associations with Renal Function, Protein-Energy Wasting, Cardiovascular Disease, and Mortality
Clin. J. Am. Soc. Nephrol.,
September 1, 2007;
2(5):
889 - 897.
[Abstract][Full Text][PDF]
Top
Abstract
Introduction
, IL-6, IL-18, and TNF-
) was tested in a cohort of 217 patients with ESRD. During the follow-up period (average 41 mo), 112 patients died. In an analysis that was adjusted for other risk factors, the relative risks for death of patients who were exposed to high levels of one, two, three, and four or more inflammation biomarkers were 1.48, 1.64, 2.76, and 3.05 times higher, respectively, than that of patients in the reference category (no inflammation). In this model, the explained variation in mortality that was attributable to overall inflammation burden (+9.1%) was marginally higher (P = 0.06) than that provided by IL-6 alone (+6.1%). In an alternative analysis based on the Bayesian approach (receiver operating characteristic curves analysis), the prediction power of the combined inflammatory burden was identical to that provided by the sole IL-6 (0.59 ± 0.04 versus 0.59 ± 0.04). IL-6 captures almost entirely the prediction power of the overall inflammation burden in patients with ESRD. IL-6 seems to be an almost ideal indicator of the severity of inflammation. The use of this biomarker can be recommended in clinical studies that aim to better the understanding of inflammation or to modify it in this population. 
