Prognostic Value of Interleukin-1Receptor Antagonist in Patients

Undergoing PercutaneousCoronary Intervention

Giuseppe Patti, MD, Germano Di Sciascio, MD, Andrea D’Ambrosio, MD,Giordano Dicuonzo, MD, Antonio Abbate, MD, and Aldo Dobrina, MD

Elevated plasma levels of inflammatory markers, suchas C-reactive protein (CRP), have been associated withadverse outcome in selected patients with coronary ar-tery disease (CAD) treated with coronary angioplasty orstenting. The aim of this study was to evaluate thepredictive value of preprocedural interleukin-1 receptorantagonist (IL-1Ra) plasma levels for long-term majoradverse cardiac events (MACE) in a series of unselectedpatients with symptomatic CAD treated with percutane-ous coronary intervention (PCI). Seventy-three consecu-tive patients (62 men, aged 62 � 9 years) undergoingPCI were enrolled in a prospective follow-up study. IL-1Ra and CRP plasma levels were measured before theprocedure; 36 patients (49%) had unstable angina pec-toris on admission, 37 (51%) had chronic stable anginapectoris, and 30 (41%) had multivessel CAD, 15 of

whom underwent multivessel PCI. Success was achievedin all 73 patients, with coronary stenting performed in63 (86%). Follow-up clinical assessment included occur-rence of MACE at 3, 6, 12, and 18 months. Logisticregression analysis, performed to determine indepen-dent predictors of MACE, identified IL-1Ra levels in theupper quartile as the only independent predictive factorof MACE at 18 months (19% in the fourth quartile vs 0%in the first quartile; p � 0.032). Patients with highpreprocedural CRP levels (fourth quartile) had a nonsig-nificant increased risk of MACE (p � 0.09). Thus, pre-procedural IL-1Ra plasma levels appear to be a valuableindependent predictive factor of MACE in unselectedpatients undergoing PCI. �2002 by Excerpta Medica,Inc.

(Am J Cardiol 2002;89:372–376)

Identification of preprocedural factors predictive ofunfavorable outcome after percutaneous coronary

intervention (PCI) is the object of active investigation.Interleukin-1 is a potent proinflammatory cytokineinvolved in tissue inflammation, ischemia reperfusioninjury, and coronary atherosclerosis1–3; interleukin-1receptor antagonist (IL-1Ra) competitively blocks thebinding of IL-1 to its membrane receptors,1–4 modu-lating the inflammatory response. Even when plasmaIL-1 levels are low, IL-1Ra levels correlate well withseverity of inflammation, as in septic shock.1,5 C-re-active protein (CRP) is also a sensitive marker ofinflammation synthesized by hepatocytes in responseto proinflammatory cytokines,6 and high levels of thisreactant have been associated with adverse outcomeduring follow-up in selected patients treated withPCI.7–10 However, the predictive value of preproce-dural IL-1Ra plasma levels in the outcome after PCIhas not been previously characterized. This studyevaluates the correlation of preprocedural IL-1Ra andCRP plasma levels with the occurrence of late adversecardiac events in unselected patients with symptom-atic coronary artery disease (CAD) treated with PCI.

METHODSPatient population and study design: From January

1998, 73 consecutive patients with significant CADundergoing PCI in our institution were enrolled in aprospective follow-up study. All patients were symp-tomatic for myocardial ischemia, were free of myo-cardial infarction in the previous month, and gaveinformed consent to the study. They were all treatedwith aspirin, oral or transdermal nitrates, and a �blocker or calcium antagonist drug; intravenous ni-trates and heparin were administered to patients withunstable angina. Quantitative coronary angiographywas performed by electronic caliper, measuring min-imal lumen diameter before and after intervention.Significant stenosis was defined as a reduction of�70% of the lumen diameter; CAD was defined asdiffuse when a significant stenosis of �20 mm ormultiple significant stenoses in the same artery werepresent.11 PCI was performed the day after the admis-sion with conventional techniques by the femoral ap-proach. Coronary stents were implanted electively orfor suboptimal results after balloon dilatation. Angio-graphic success was defined as a reduction in stenosisto �50% residual narrowing after balloon inflations orstent implantation. Glycoprotein IIb to IIIa antagonistswere not used in our institution at the time of thestudy. Routine serial postprocedural creatine kinasedeterminations were not obtained if not clinically in-dicated. Clinical success was defined as improvementin ischemic symptoms associated with angiographic

From the Department of Cardiovascular Sciences, Campus Bio-Medico University, Rome, Italy. Manuscript received July 9, 2001;revised manuscript received and accepted October 15, 2001.

Address for reprints: Germano Di Sciascio, MD, Department ofCardiovascular Sciences, Campus Bio-Medico University, Via E. Lon-goni, 83, 00155 Rome, Italy. E-mail: g.disciascio@unicampus.it.

372 ©2002 by Excerpta Medica, Inc. All rights reserved. 0002-9149/02/$–see front matterThe American Journal of Cardiology Vol. 89 February 15, 2002 PII S0002-9149(01)02254-8

improvement, without major in-hospital complica-tions (myocardial infarction, urgent bypass surgery, ordeath). After the procedure, aspirin (100 mg/day) andstatins, if indicated, were indefinitely prescribed,whereas ticlopidine (250 mg twice daily) was contin-ued for 4 weeks in patients undergoing stent implan-tation. At the time of enrollment no patient had clin-ical and laboratory evidence of malignancy, infec-tions, inflammatory diseases or recent (�2 months)surgery or trauma. Follow-up data were collected byoffice visit at 3, 6, 12, and 18 months in all patients.Follow-up assessment included occurrence of majoradverse cardiac events (MACE), defined as angina,

myocardial infarction, cardiac death, or coronary re-vascularization; incidence of inflammatory and im-mune diseases was recorded as well. Repeat hospital-ization was advised for recurrent symptoms and/orsigns of myocardial ischemia and repeat coronaryangiography was performed if indicated.

Venous blood samples were drawn 3 to 6 hoursbefore the procedure to measure plasma levels ofIL-1Ra and CRP. Patients were kept fasting for atleast 8 hours and at complete rest for at least 10minutes before blood collection. Blood samples weredrawn in all patients from 6:00 to 8:30 A.M. with a16-G needle keeping the venous stasis to a minimum.All procedural and laboratory data were prospectivelycollected in a computerized database.

Laboratory assays: Venous blood samples werecollected in syringes and anticoagulated with ethyl-enediaminetetraacetic acid and sodium citrate for de-tection of IL-1Ra and CRP levels, respectively. Theblood was refrigerated at 4°C until it was centrifugedat 2000 g for 30 minutes; the plasma obtained wasfrozen at �80°C, until it was essayed. IL-1Ra kitswere purchased from R&D System (Minneapolis,Minnesota). The assays used the quantitative sand-wich enzyme immunoassay technique. The microplatereader was an LP400 from Sanofi-Diagnostics Pasteur(Milan, Italy); the lower level of detection for the

TABLE 1 Patient Demographic Characteristics

VariableTotal

(n � 73)Asymptomatic

(n � 61)MACE

(n � 12)Relative Risk

(95% CI) p ValueCorrectedp Value

Men 62 (85%) 53 (85%) 9 (15%) 0.80 0.66 0.80(0.20–3.21)

Age (yrs) 62 � 9 62 � 9 59 � 11 0.33 0.45Diabetes mellitus 14 (19%) 11 (79%) 4 (21%) 1.40 0.69 0.82

(0.44–4.52)Systemic hypertension 36 (49%) 30 (83%) 6 (17%) 1.03 0.79 0.87

(0.37–2.89)Hypercholesterolemia 42 (56%) 35 (83%) 7 (17%) 1.03 0.80 0.78

(�200 mg/dl) (0.36–2.95)

Cigarette smoking 40 (55%) 32 (80%) 8 (20%) 1.65 0.56 0.42(0.54–5.00)

TABLE 2 Clinical and Angiographic Characteristics

VariableTotal

(n � 73)Asymptomatic

(n � 61)MACE

(n � 12)Relative Risk

(95% CI) p ValueCorrectedp Value

Stable angina pectoris 37 (51%) 33 (89%) 4 (11%) 0.49 0.32 0.30(0.16–1.47)

Unstable angina pectoris 36 (49%) 28 (78%) 8 (22%) 2.06 0.32 0.30(0.68–6.23)

Previous myocardial infarction 40 (55%) 32 (80%) 8 (20%) 1.65 0.56 0.98(0.54–5.00)

Left ventricular ejection fraction �50% 22 (30%) 16 (73%) 6 (27%) 2.32 0.17 0.78(0.84–6.40)

Multivessel PCI 15 (21%) 13 (87%) 2 (13%) 0.91 1.00 0.19(0.22–3.72)

PCI of diffuse CAD 31 (42%) 22 (71%) 9 (29%) 4.06 0.030 0.12(1.20–13.79)

TABLE 3 Procedural Characteristics (n � 73)

Treated coronary vessels 89Left anterior descending 44Left circumflex 21Right 21Grafts 3

Vessels/patient 1.22Treated lesions 98Patients with stent 63 (86%)Implanted stents 75Stent length (range) (mm) 16 (8–35)Minimal lumen diameter after procedure

(range) (mm)3.2 (2.0–4.0)

CORONARY ARTERY DISEASE/IL–1 RECEPTOR ANTAGONIST AND CORONARY INTERVENTION 373

assay was 49 pg/ml. CRP was determined by highsensitivity rate nephelometry (DADE-Behring CRP;Behring Institute, Scoppito, L’Aquila, Italy).

Statistical analysis: Values are expressed as mean �SD for normally distributed variables and median andinterquartile ranges (IR) for IL-1Ra and CRP plasmalevels. The nonparametric Mann-Whitney U test wasused to compare continuous nonpaired data, and theWilcoxon test for paired data. Categorical variableswere compared by chi-square or Fisher’s exact test, asindicated. IL-1Ra and CRP levels were divided intoquartiles and compared by the Kruskal-Wallis test orby analysis of variance for multiple comparisonswithin all 4 groups. Logistic regression analysis wasperformed to determine independent predictors ofMACE. Relative risk, with 95% confidence intervals(CI), was calculated. Statistical analysis was per-formed using the Statistical Package for Social Sci-ences software (SPSS 10.0 for Windows, SPSS Inc.,Chicago, Illinois). Actuarial event-free survival curveswere obtained with the Kaplan-Meyer method. A pvalue �0.05 was considered significant.

RESULTSClinical and procedural characteristics: Demo-

graphic and clinical/angiographic characteristics ofthe 73 patients are indicated in Table 1 and Table 2,respectively. In particular, 49% of patients had unsta-ble angina on admission (Braunwald class IIB orIIIB), 30% had left ventricular dysfunction, and 41%had multivessel CAD.

Procedural characteristics are indicated in Table 3.Clinical success was achieved in all 73 patients andangiographic success in all 89 treated coronary arter-ies; multivessel interventions were performed in 15patients (21%); 63 patients (86%) underwent coronarystenting, and 31 of those (43%) underwent stenting ofdiffuse CAD. There were no in-hospital acute majorcomplications: no patient died, had acute myocardialinfarction, or required emergency bypass surgery.

Baseline IL-1Ra and CRP plasma levels were 130pg/ml (interquartile range 105 to 172 pg/ml) and 0.25

mg/dl (0.01 to 0.59 mg/dl), respectively.During the 18-months follow-up, 61 pa-tients (84%) remained free of cardiacsymptoms and MACE, and had no induc-ible myocardial ischemia at functionaltests. Twelve patients (16%) experiencedMACE during follow-up: 4 (6%) patientshad a target vessel-related myocardial in-farction (fatal in 2 patients), 5 (7%) hadunstable angina, and 3 (4%) had recur-rence of effort angina. Repeat angiogra-phy was performed in 6 of these 12 patientsand repeat angioplasty was performed in 5(7%) (target vessel revascularization in 4and angioplasty on a different coronary ar-tery in the fifth patient). During the entirefollow-up period, no patient had evidenceof systemic inflammatory or immune dis-eases or malignancy. At multivariate anal-ysis, among several clinical and angio-

graphic parameters, only coronary stenting of diffuseCAD was independently correlated with the occurrenceof MACE during follow-up (relative risk 4.06 [95% CI1.20 to 13.79], p � 0.020; Tables 1 and 2).

Preprocedural IL-1Ra and CRP levels and follow-upevents: Patients with MACE had significantly higherpreprocedural IL-1Ra levels than those without eventsduring the follow-up (189 pg/ml [IR 140 to 361] vs126 pg/ml [IR 99 to 167] p � 0.007) (Figure 1); thisdifference was also present, although not significant,for CRP levels (0.37 mg/dl [95% CI 0.24 to 0.94] vs0.21 (95% CI 0.00 to 0.56), p � 0.08). Moreover,patients with preprocedural IL-1Ra levels in the fourthquartile had the highest clinical event rate (33%),whereas those with baseline levels in the lowest quar-tile had no occurrence of MACE at 18 months (fourthvs first quartile: p � 0.008) (Table 4, Figure 2A), withan actuarial event-free survival of 66% versus 100%,respectively (Figure 3). Patients with CRP levels inthe fourth quartile had a nonsignificant increased riskof MACE (relative risk 5.28 (95% CI 0.68 to 40.92],p � 0.09) (Table 4, Figure 2B). At multivariate anal-ysis, after adding IL-1Ra and CRP levels to clinical,angiographic, and procedural parameters (includingstenting of diffuse CAD), IL-1Ra levels in the upperquartile resulted as the only independent predictivefactor of adverse events during follow-up (19% vs 0%in the first quartile, p � 0.032) (Table 4). Finally,among 6 patients with preprocedural IL-1Ra levels inthe fourth quartile who had MACE during follow-up,only 1 also had elevated (fourth quartile) values ofCRP.

DISCUSSIONThis study indicates that high preprocedural IL-

1Ra levels may be an independent predictor of MACEafter PCI; patients with the lowest preprocedural IL-1Ra levels (first quartile) experienced no MACE dur-ing 18 months of follow-up. Previous studies havedemonstrated that local release of IL-1 within coro-nary plaques stimulates intimal hyperplasia12; IL-1Ramodulates the activity of IL-1 and, therefore, may

FIGURE 1. Preprocedural distribution (with mean and SD) of IL-1Ra levels in pa-tients with and without MACE during the follow-up.

374 THE AMERICAN JOURNAL OF CARDIOLOGY� VOL. 89 FEBRUARY 15, 2002

modulate local inflammatory response also at the siteof atherosclerotic plaques.13,14 Thus, circulating levelsof IL-1Ra may in part reflect local plaque “activity”and may identify patients with unstable plaques15,16

that could be more prone to subsequent events afterPCI. A correlation of IL-1Ra levels with restenosis

after coronary stenting was also suggested by a sig-nificant association observed between IL-1Ra genepolymorphism and restenosis.17 Previous reports pro-posed CRP as a useful inflammatory marker to predictadverse events in selected subgroups of patients un-dergoing percutaneous or surgical coronary revascu-larization,7–10,18–21 although in a recent series, CRPhad no prognostic value on thrombotic events andtarget lesion revascularization after coronary stent-ing.21 In our study, elevated preprocedural CRP levelswere correlated only with a trend toward increasedrisk of late cardiac events after PCI; conversely, theassociation of elevated levels of IL-1Ra with long-term MACE was highly significant and did not corre-late with a clinical syndrome on admission. Moreover,unlike previous studies, this correlation was observedin a consecutive series of unselected patients treated inan experienced interventional laboratory, thus repre-senting a real-life “scenario” that includes patientswith different clinical, angiographic, and proceduralcharacteristics. Our results also suggest that preproce-dural IL-1Ra determination may have an importantprognostic value for managing patients during fol-low-up after PCI.22 Treatment modalities with prava-statin,23 clopidogrel,24 and abciximab25 have alreadybeen shown in preliminary studies to reduce CRPplasma levels and to attenuate the risk associated withbaseline CRP status after PCI. Accordingly, such ad-junctive therapies may result in clinical benefit inpatients with high preprocedural IL-1Ra levels under-going PCI, even when CRP levels are not elevated.The cost (approximately $800 for 50 assays) andcomplexity of IL-1Ra determination are, at present, arelative limitation for its routine clinical use, and lessexpensive assays are hoped for. Further studies in agreater number of patients are required to definitivelyconfirm the complementary predictive value of IL-1Ra and CRP after coronary interventions.

FIGURE 2. Occurrence of MACE after PCIs during 18 months offollow-up according to preprocedural IL-1Ra (A) and CRP (B)quartiles.

TABLE 4 Interleukin-1 Receptor Antagonist and C-Reactive Protein Plasma Levels

Asymptomatic(n � 61)

MACE(n � 12)

Relative Risk(95% CI) p Value

Correctedp Value

IL-1Ra (pg/ml)

Median (interquartile range) 126 (99–167) 189 (140–361)Quartiles distribution

1 19 (31%) 0 — 0.008 vs 4 0.032 vs 42 15 (25%) 3 (25%) 1.00 0.10 vs 13 15 (25%) 3 (25%) 1.00 0.10 vs 14 12 (19%) 6 (50%) 2.00 0.44 vs 2 or 3

(0.59–6.79)CRP (mg/dl)

Median (interquartile range) 0.21 (0–0.56) 0.37 (0.24–0.94)Quartiles distribution

1 18 (30%) 1 (8%) 1.00 0.09 vs 4 0.14 vs 42 16 (26%) 2 (17%) 2.11 0.60 vs 1

(0.21–21.32)3 14 (23%) 4 (33%) 4.22 0.18 vs 1

(0.52–34.29) 0.59 vs 24 13 (21%) 5 (42%) 5.28 0.40 vs 2

(0.68–40.92) 1.00 vs 3

CORONARY ARTERY DISEASE/IL–1 RECEPTOR ANTAGONIST AND CORONARY INTERVENTION 375

1. Dinarello CA. Biological basis for interleukin-1 in disease. Blood 1996;87:2095–2147.2. Loppnow H, Werdan K, Reuter G, Flad HD. The interleukin-1 and interleu-kin-1 converting enzyme families in the cardiovascular system. Eur CytokineNetw 1998;9:675–680.3. Rothwell N, Allan S, Toulmond S. The role of interleukin-1 in acute neuro-degeneration and stroke: pathophysiological and therapeutic implications. J ClinInvest 1997;100:2648–2652.4. Arend WP. Interleukin 1 receptor antagonist. A new member of the interleukin1 family. J Clin Invest 1991;88:1445–1451.5. Fischer E, van Zee KJ, Marano MA, Rock CS, Kenney JS, Poutsiaka DD,Dinarello CA, Lowry SF, Moldawer LL. Interleukin-1 receptor antagonist circu-lates in experimental inflammation and in human disease. Blood 1992;79:2196–2200.6. Baumann H, Gauldie J. Regulation of hepatic acute phase plasma protein genesby hepatocyte stimulating factors and other mediators of inflammation. Mol BiolMed 1990;7:147–159.7. Walter DH, Fichtlscherer S, Sellwig M, Auch-Schwelk W, Schachinger V,Zeiher AM. Preprocedural C-reactive protein and cardiovascular events aftercoronary stent implantation. J Am Coll Cardiol 2001;37:839–846.8. Buffon A, Liuzzo G, Biasucci LM, Pasqualetti P, Ramazzotti V, Rebuzzi AG,Crea F, Maseri A. Preprocedural serum levels of C-reactive protein predict earlycomplications and late restenosis after coronary angioplasty. J Am Coll Cardiol1999;34:1512–1521.9. Gaspardone A, Crea F, Versaci F, Tomai F, Pellegrino A, Chiariello L, GioffrePA. Predictive value of C-reactive protein after successful coronary-artery stent-ing in patients with stable angina. Am J Cardiol 1998;82:515–518.10. Versaci F, Gaspardone A, Tomai F, Crea F, Chiariello L, Gioffre PA.Predictive value of C-reactive protein in patients with unstable angina pectorisundergoing coronary artery stent implantation. Am J Cardiol 2000;85:92–95.11. Di Sciascio G, Patti G, Nasso G, Manzoli A, D’Ambrosio A, Abbate A. Early

and long-term results of stenting of diffuse coronary ar-tery disease. Am J Cardiol 2000;86:1166–1170.12. Rectenwald JE, Moldawer LL, Huber TS, Seeger JM,Ozaki CK. Direct evidence for cytokine involvement inneointimal hyperplasia. Circulation 2000;102:1697–1702.13. Dewberry R, Holden H, Crossman D, Francis S.Interleukin-1 receptor antagonist expression in humanendothelial cells and atherosclerosis. Arterioscler ThrombVasc Biol 2000;20:2394–2400.14. Nicklin MJ, Hughes DE, Barton JL, Ure JM, DuffGW. Arterial inflammation in mice lacking the Interleukin1 receptor antagonist gene. J Exp Med 2000;191:303–312.15. Biasucci LM, Liuzzo G, Fantuzzi G, Caligiuri G,Rebuzzi AG, Ginnetti F, Dinarello CA, Maseri A. Increas-ing levels of interleukin (IL)-Ra and IL-6 during the first2 days of hospitalization in unstable angina are associatedwith increased risk of in-hospital coronary events. Circu-lation 1999;99:2079–2084.16. Patti G, Abbate A, D’Ambrosio A, Manzoli A, Dic-uonzo G, Guarnieri G, Dobrina A, Di Sciascio G. Inter-leukin-1 receptor antagonist: a more sensitive marker ofunstable coronary syndromes than C-reactive protein (ab-str. Eur Heart J 2000;21:650.17. Kastrati A, Koch W, Berger PB, Mehilli J, StephensonK, Neumann FJ, von Beckerath N, Bottiger C, Duff GW,Schomig A. Protective role against restenosis from aninterleukin-1 receptor antagonist gene polymorphism inpatients treated with coronary stenting. J Am Coll Cardiol2000;36:2168–2173.18. Milazzo D, Biasucci LM, Luciani N, Martinelli L,

Canosa C, Schiavello R, Maseri A, Possati G. Elevated levels of C-reactiveprotein before coronary artery bypass grafting predict recurrence of ischemicevents. Am J Cardiol 1999;84:459–461.19. Cannon CP, Weintraub WS, Demopoulos L, Vicari R, Bazzino O, Ali N,DeLucca P, DiBattiste P, Rifai N, Braunwald E, for the TACTICS-TIMI 18investigators. High-sensitivity C-reactive protein (hs-CRP) to predict 6 monthmortality and relative benefit of invasive vs conservative strategy in patients withunstable angina: primary results of the TACTICS-TIMI 18 C-reactive proteinsubstudy (abstr). J Am Coll Cardiol 2001;37:315A.20. Heyde GS, de Winter RJ, Fischer J, Koch KT, Schouten Y, Wulfert FM, PiekJJ. Prognostic value of C-reactive protein in patients undergoing elective PCI(abstr). J Am Coll Cardiol 2001;37:27A.21. Sceller B, Hennen B, Markwirth T, Stall HP. C reactive protein has noprognostic impact on thrombotic events and target lesion re-intervention rate aftercoronary stenting (abstr). J Am Coll Cardiol 2001;37:39A.22. Patti G, Abbate A, D’Ambrosio A, Manzoli A, Dicuonzo G, Guarnieri G,Dobrina A, Di Sciascio G. Long-term variations of Interleukin-1 receptor antag-onist and C-reactive protein serum levels after successful coronary angioplasty(abstr). Eur Heart J 2000;21:164.23. Jialal I, Stein D, Balis D, Grundy SM, Adams-Huet B, Devaraj S. Effect ofhydroxymethyl glutaryl coenzyme A reductase inhibitor therapy on high sensitiveC-reactive protein levels. Circulation 2001;103:1933–1935.24. Robbins MA, Chew DP, Bhatt DL, Mukerjee D, Roffi M, Schneider JP, TopolEJ, Ellis SG. ADP-antagonists therapy added to aspirin prior to coronary inter-vention markedly attenuates the risk associated with baseline CRP status (abstr).J Am Coll Cardiol 2001;37:28A.25. Buffon A, Liuzzo G, Angiolillo DJ, Santamaria M, Pelliccioni S, Ginnetti F,Rebuzzi AG, Crea F, Biasucci LM, Maseri A. Abciximab decreases cytokineproduction by circulating monocytes in patients with unstable angina and ele-vated inflammatory markers (abstr). J Am Coll Cardiol 2001;37:27A.

FIGURE 3. Actuarial survival curves in patients with lowest (1st quartile) and highest(4th quartile) preprocedural IL-1Ra levels.

376 THE AMERICAN JOURNAL OF CARDIOLOGY� VOL. 89 FEBRUARY 15, 2002