Predictive Value of Cardiac Computed Tomography and the Impactof Renal Function on All Cause Mortality (from Coronary Computed

Tomography Angiography Evaluation for Clinical Outcomes)

Girish Dwivedi, PhDa, Myra Cocker, PhDa, Yeung Yam, BSca, Stephan Achenbach, MDb,Mouaz Al-Mallah, MD, MScc, Daniel S. Berman, MDd, Matthew J. Budoff, MDe,

Filippo Cademartiri, MD, PhDf,g, Tracy Q. Callister, MDh, Hyuk-Jae Chang, MD, PhDi,Victor Cheng, MDj, Kavitha M. Chinnaiyan, MDk, Augustin Delago, MDl, Allison M. Dunning, MScm,Martin Hadamitzky, MDp, Jörg Hausleiter, MDp, Philipp A. Kaufmann, MDq, Troy M. LaBounty, MDd,Fay Lin, MDn,o, Erica Maffei, MDf,g, Gilbert L. Raff, MDk, Leslee J. Shaw, PhDr, Todd C. Villines, MDs,James K. Min, MDd, and Benjamin J.W. Chow, MDa,*, on Behalf of the CONFIRM Investigators

Patients with chronic kidney disease have a worse cardiovascular prognosis than those

aDepartment oHeart Institute, OUniversity of ErlWayne State UndDepartment of ICalifornia; eDepaLos Angeles, CalHospital, MonastiMedical Center, RoInstitute, HendersCardiovascular HoOklahoma Heart IRoyal Oaks, MichDepartments of mPMedical CollegeNew York; pDivi

0002-9149/13/$ -http://dx.doi.org/1

without. The aim of this study was to determine the incremental prognostic value of coro-nary computed tomographic angiography in predicting mortality across the entire spectrumof renal function in patients with known or suspected coronary artery disease (CAD). A largeinternational multicenter registry was queried, and patients with left ventricular ejectionfraction (LVEF) and creatinine data were screened. National Cholesterol EducationProgram Adult Treatment Panel III risk was calculated. Coronary computed tomographicangiographic results were evaluated for CAD severity (normal, nonobstructive, or obstruc-tive) and an LVEF <50%. Patients were followed for the end point of all-cause mortality.Among 5,655 patients meeting the study criteria, follow-up was available for 5,572 (98.9%;median follow-up duration 18.6 months). All-cause mortality (66 deaths) significantlyincreased with every 10-unit decrease in renal function (hazard ratio [HR] 1.23, 95%confidence interval [CI] 1.07 to 1.41). All-cause mortality occurred in 0.33% of patientswithout coronary atherosclerosis, 1.82% of patients with nonobstructive CAD, and 2.43% ofpatients with obstructive CAD. Multivariate Cox proportional-hazards models revealed thatimpaired renal function (HR 2.29, 95% CI 1.65 to 3.18), CAD severity (HR 1.81, 95% CI 1.31to 2.51), and an abnormal LVEF (HR 4.16, 95% CI 2.45 to 7.08) were independent predictorsof all-cause mortality. In conclusion, coronary computed tomographic angiographicmeasures of CAD severity and the LVEF provide effective risk stratification across a widespectrum of renal function. Furthermore, renal dysfunction, CAD severity, and the LVEFhave additive value for predicting all-cause death in patients with suspected obstructiveCAD. � 2013 Elsevier Inc. All rights reserved. (Am J Cardiol 2013;111:1563e1569)

Previous single-center and multicenter trials have estab-lished the utility of coronary computed tomographic angi-ography (CCTA) for the diagnosis, management, andprognosis of patients with established or suspected coronary

f Medicine, Division of Cardiology, University of Ottawattawa, Ontario, Canada; bDepartment of Medicine,

angen, Erlangen, Germany; cDepartment of Medicine,iversity, Henry Ford Hospital, Detroit, Michigan;maging, Cedars Sinai Medical Center, Los Angeles,rtment of Medicine, Harbor UCLA Medical Center,ifornia; fCardiovascular Imaging Unit, Giovanni XXIIIer di Treviso, Italy; gDepartment of Radiology, Erasmustterdam, TheNetherlands; hTennesseeHeart andVascularonville, Tennessee; iDivision of Cardiology, Severancespital, Seoul, Korea; jDepartment of Cardiovascular CT,nstitute, Tulsa, Oklahoma; kWilliam Beaumont Hospital,igan; lCapitol Cardiology Associates, Albany, New York;ublic Health, nMedicine, and oRadiology, Weill Cornelland the New York Presbyterian Hospital, New York,sion of Cardiology, Deutsches Herzzentrum München,

see front matter � 2013 Elsevier Inc. All rights reserved.0.1016/j.amjcard.2013.02.004

artery disease (CAD).1e5 Unless receiving renal replace-ment therapy, patients with end-stage renal failure are notconsidered ideal candidates for CCTA because of therisk for contrast agenteinduced nephrotoxicity.6 However,

Munich, Germany; qCardiac Imaging, University Hospital, Zurich,Switzerland; rDepartment of Medicine, Emory University School of Medi-cine, Atlanta, Georgia; and sDepartment of Medicine, Walter Reed MedicalCenter, Washington, District of Columbia. Manuscript received November22, 2012; revised manuscript received and accepted February 3, 2013.

Dr. Dwivedi is supported by the Dowager Countess Eleanor Peel Trust(Peel and Rothwell Jackson Postgraduate Travelling Fellowship), London,United Kingdom; the Whit & Heather Tucker Endowed Research Fellow-ship in Cardiology, Ottawa, Ontario, Canada; and the Banting PostdoctoralFellowship, Ottawa, Ontario, Canada. Dr. Kaufman is supported by theSwiss National Science Foundation, Bern, Switzerland. Dr. Chow issupported by New Investigator Award MSH-83718 from the CanadianInstitutes of Health Research, Ottawa, Ontario, Canada.

See page 1568 for disclosure information.*Corresponding author: Tel: 613-761-4044; fax: 613-761-4929.E-mail address: bchow@ottawaheart.ca (B.J.W. Chow).

www.ajconline.org

Table 1Patient characteristics of patients with follow-up (coronary arteries �12)

Variable All Patients (n ¼ 5,572) eGFR (ml/min/1.73 m2) p Value

�90 (n ¼ 2,130) 60e89 (n ¼ 2,827) 30e59 (n ¼ 580) <30 (n ¼ 35)

Age (yrs) 55.4 � 12.7 50.1 � 11.7 57.0 � 11.7 66.5 � 11.3 59.3 � 13.5 <0.001Men 2,881 (51.7%) 1,143 (53.7%) 1,530 (54.1%) 190 (32.8%) 18 (51.4%) <0.001Body mass index (kg/m2) 28.7 � 5.9 28.6 � 6.0 28.7 � 5.7 29.0 � 6.5 28.3 � 6.6 0.358Creatinine (mg/dl) 0.9 � 0.6 0.7 � 0.1 1.0 � 0.2 1.2 � 0.2 6.5 � 3.2 <0.001Smoker/ex-smoker 2,181 (39.1%) 833 (39.1%) 1,107 (39.2%) 229 (39.5%) 12 (34.3%) 0.945Hypertension 2,748 (49.3%) 827 (38.8%) 1,468 (51.9%) 422 (72.8%) 31 (88.6%) <0.001Dyslipidemia 2,783 (49.9%) 887 (41.6%) 1,499 (53.0%) 376 (64.8%) 21 (60.0%) <0.001Diabetes 692 (12.4%) 228 (10.7%) 332 (11.7%) 113 (19.5%) 19 (54.3%) <0.001Family history of CAD 2,204 (39.6%) 847 (39.8%) 1,149 (40.6%) 198 (34.1%) 10 (28.6%) 0.016Indication for studyChest pain 3,635 (65.2%) 1,520 (71.4%) 1,745 (61.7%) 352 (60.7%) 18 (51.4%) <0.001

Nonanginal chest pain 2,396 (43.0%) 1,199 (56.3%) 1,026 (36.3%) 166 (28.6%) 5 (14.3%) <0.001Atypical angina pectoris 776 (13.9%) 217 (10.2%) 445 (15.7%) 107 (18.4%) 7 (20.0%) <0.001Typical angina pectoris 463 (8.3%) 104 (4.9%) 274 (9.7%) 79 (13.6%) 6 (17.1%) <0.001

Dyspnea 584 (10.5%) 148 (6.9%) 329 (11.6%) 100 (17.2%) 7 (20.0%) <0.001Coronary artery narrowingNone 2,752 (49.4%) 1,333 (62.6%) 1,231 (43.5%) 179 (30.9%) 9 (25.7%) <0.001Nonobstructive 1,872 (33.6%) 555 (26.0%) 1,044 (36.9%) 260 (44.8%) 13 (37.1%) <0.001Obstructive 948 (17.0%) 242 (11.4%) 552 (19.5%) 141 (24.3%) 13 (37.1%) <0.001

Abnormal LVEF (<50%) 413 (7.4%) 94 (4.4%) 234 (8.3%) 77 (13.3%) 8 (22.9%) <0.001LVEF (%)* 63.2 � 11.4 62.21 � 10.5 63.4 � 10.9 64.0 � 13.9 60.8 � 15.3 0.026

Data are expressed as mean � SD or as number (percentage).* Analysis of the LVEF was performed in 2,896 patients (51.97%).

1564 The American Journal of Cardiology (www.ajconline.org)

emerging prospective clinical investigations from high-riskpatient populations receiving intravenous contrast media forcomputed tomography suggest that the incidence andserious negative clinical outcomes are much less commonthan previously believed.6 Moreover, the risk for nephro-toxicity can be reduced with proper patient preparation.6

Globally, a significant number of patients with varyingdegree of renal impairment undergo CCTA as part of theircardiac risk evaluation. Despite the significant numbers andobvious clinical implications, to the best of our knowledge,the prognostic value of CCTA across the different stages ofrenal impairment has not been evaluated in a large-scalemulticenter trial. The main aim of this study was to deter-mine the incremental prognostic value of CCTA in pre-dicting mortality across the entire spectrum of renal functionin a large prospective cohort of patients with known orsuspected CAD recruited in Coronary Computed Tomog-raphy Angiography Evaluation for Clinical Outcomes: AnInternational Multicenter Registry (CONFIRM).

Methods

Details of CONFIRM have been described previously.5

Briefly, centers with 64-slice computed tomographic scan-ners participating in this registry prospectively contributedto coronary computed tomographic angiographic databasesas part of a large multicenter observational registry. Quali-fying sites contributed baseline demographics, cardiac riskfactors, drug history, findings on CCTA, and outcomes.From February 2003 to December 2009, 27,125 consecutivepatients underwent CCTA at 12 enrolling centers in 6countries (Canada, Germany, Italy, Korea, Switzerland, and

the United States) and were prospectively entered. Eachenrolling center contributed 499 to 4,912 patients for anal-ysis. For the present study, patients with left ventricularejection fraction (LVEF) and serum creatinine assessmentswere screened for inclusion. Patients with histories ofcoronary revascularization (coronary artery bypass graftingand/or percutaneous coronary intervention), congenital heartdisease, or cardiac transplantation were excluded from thepresent investigation. Follow-up procedures were approvedby the institutional review board at each site.

Medical histories were obtained for all patients. Patients’pretest probability for obstructive CAD was calculated usingage, gender, and symptoms.3,7e9 Each patient’s risk forfuture cardiac events was assessed using National Choles-terol Education Program (NCEP) Adult Treatment Panel(ATP) III guidelines.3,10

The coronary computed tomographic angiographicprotocol, data acquisition, image postprocessing, and inter-pretation for the present study cohort were performed inaccordance with the guidelines of the Society of Cardio-vascular Computed Tomography.11,12

Coronary diameter stenoses were graded using a 3-pointscale (normal, nonobstructive [<50%], or obstructive[�50%]).3,13 Abnormal LVEFs were calculated using end-diastolic and end-systolic volumes, and an LVEF <50%was considered abnormal.14

Serum creatinine was used to assess renal function in allstudy participants. Estimated glomerular filtration rate (eGFR)was calculated for each patient using the Levey Modificationof Diet in Renal Disease (MDRD) formula (eGFR ¼ 175 �standardized serum creatinine � 1.154 � age � 0.203 �0.742 [if female]).15 The eGFR values were categorized as

Table 2Univariate analysis of clinical characteristics for all-cause death (coronary arteries �12)

Variable Death HR (95% CI) p Value

No (n ¼ 5,506) Yes (n ¼ 66)

Age (yrs) 55.3 � 12.6 64.7 � 14.7 1.06 (1.04e1.09) <0.001Men 2,847 (51.7%) 34 (51.5%) 1.01 (0.62e1.63) 0.984Body mass index (kg/m2) 28.7 � 5.9 26.9 � 6.6 0.94 (0.89e0.99) 0.014Cardiac risk factorsDiabetes 671 (12.2%) 21 (31.8%) 3.33 (1.98e5.58) <0.001Dyslipidemia 2,757 (50.1%) 26 (39.4%) 0.63 (0.38e1.03) 0.063Hypertension 2,695 (48.9%) 53 (80.3%) 4.26 (2.32e7.81) 0.001Family history of CAD 2,183 (39.6%) 21 (31.8%) 0.73 (0.43e1.22) 0.232Smoker/ex-smoker 2,151 (39.1%) 30 (45.5%) 1.38 (0.85e2.25) 0.189

eGFR (ml/min/1.73 m2) 2.97 (2.13e4.13) <0.001�90 2,114 (38.4%) 16 (24.2%)60e89 2,804 (50.9%) 23 (34.8%)30e59 560 (10.2%) 20 (30.3%)<30 28 (0.5%) 7 (10.6%)

eGFR (10-unit decrease) 78.3 � 20.3 70.0 � 38.1 1.23 (1.07e1.41) 0.003NCEP ATP III score 2.88 (1.97e4.22) <0.001Low risk 1,462 (26.6%) 8 (12.1%)Intermediate risk 3,084 (56.0%) 27 (40.9%)High risk 960 (17.4%) 31 (47.0%)

Coronary artery narrowing 2.40 (1.77e3.27) <0.001None 2,743 (49.8%) 9 (13.6%)<50% 1,838 (33.4%) 34 (51.5%)�50% 925 (16.8%) 23 (34.8%)

Abnormal LVEF (<50%) 391 (7.1%) 22 (33.3%) 6.64 (3.98e11.10) <0.001LVEF (10% decrease)* 63.3 � 11.2 55.3 � 18.6 1.54 (1.31e1.82) <0.001

Data are expressed as mean � SD or as number (percentage).* Analysis of the LVEF was performed in 2,896 patients (51.97%).

Coronary Artery Disease/Prognosis of Cardiac Computed Tomography 1565

<30, 30 to 59, 60 to 89, or �90 ml/min/1.73 m2, on the basisof the Kidney Disease Outcomes Quality Initiative classifi-cation of renal function.16

Patient follow-up for all-cause mortality was performedby each local institution by telephone interview, with vali-dation of reported death through medical records wheneverpossible and/or a national death registry.

Statistical analysis was performed using SAS version 9.2(SAS Institute Inc., Cary, North Carolina), and statisticalsignificance was defined as p <0.05. Categorical data arepresented as percentage frequencies. Continuous variablesare expressed as mean � SD. To compare patient charac-teristics, Wilcoxon’s rank-sum test was used to comparecontinuous variables and the chi-square test was used forcategorical variables. Prognostic values of eGFR, CADseverity, and the LVEF were assessed for univariate andmultivariate associations with all-cause mortality using Coxproportional-hazard models. For risk-adjusted analysis, theindependent prognostic value of eGFR, CAD severity, andthe LVEF was assessed by controlling for clinical predictors(NCEP ATP III score) and creating adjusted survival curves.Model overfitting was considered, and the proportional-hazards assumption was met. The incremental value ofeGFR, CAD severity, and the LVEF was calculated bydefining the clinical predictor model, followed by theaddition of eGFR, CAD severity, and the LVEF. Receiver-operator characteristic curves were constructed for modelsof clinical predictors only, clinical predictors plus eGFR,clinical predictors plus eGFR plus CAD severity, and

clinical predictors plus eGFR plus CAD severity plus theLVEF. Areas under the receiver-operating characteristiccurves (with 95% confidence intervals [CIs]) werecompared to evaluate the discriminative ability of eGFRover clinical predictors, CAD severity over clinical predic-tors plus eGFR, and the LVEF over clinical predictors pluseGFR plus CAD severity to predict all-cause mortality.

Results

In the whole registry, 27,125 patients were screened withCCTA at 12 participating centers. Of the 5,864 patients withclinical variables, creatinine value, CAD severity, andLVEF assessments (normal vs abnormal LVEFs), 209patients were excluded because of histories of coronaryrevascularization, congenital heart disease, or cardiactransplantation. The final study population included 5,655patients. Follow-up was available for 5,572 (98.9%; meanage 55.4 � 12.7 years, 52% men), with a median follow-upduration of 18.6 months (range 13.8 to 24.2). Table 1 liststhe baseline characteristics of the study cohort. Patients withimpaired renal function were older and had a higher prev-alence of hypertension, diabetes, and dyslipidemia. Higherincidence of chest pain and dyspnea as indications forCCTA were observed in patients with impaired renalfunction. Absolute LVEF measurements were available in52% of the analyzed population. The mean LVEF whenassessed as a continuous variable was 63.2 � 11.4%, with

Table 3All-cause death grouped according to estimated glomerular filtration rate

eGFR (ml/min/1.73 m2) n All Cause Death

All patients 5,572 66 (1.18%)�90 2,130 16 (0.75%)60e89 2,827 23 (0.81%)30e59 580 20 (3.45%)<30 35 7 (20.0%)

Figure 1. All-cause mortality (percentage) grouped according to eGFR andCAD severity.

Table 4Cox models for all-cause death

Model HR (95% CI) p Value

Clinical variablesNCEP ATP III score 2.88 (1.97e4.22) <0.001

Clinical variables plus eGFRNCEP ATP III score 2.42 (1.64e3.57) <0.001eGFR (ml/min/1.73 m2) 2.48 (1.79e3.43) <0.001

�90 1.0 —

60e89 0.94 (0.50e1.79) 0.85730e59 3.48 (1.79e6.78) <0.001<30 17.72 (7.05e44.56) <0.001

Clinical variables plus eGFR plus CADseverity

NCEP ATP III score 2.08 (1.38e3.12) <0.001eGFR (ml/min/1.73 m2) 2.29 (1.65e3.18) <0.001

�90 1.0 —

60e89 0.82 (0.43e1.56) 0.54330e59 2.91 (1.49e5.69) 0.002<30 13.16 (5.16e33.56) <0.001

Coronary narrowing 1.81 (1.31e2.50) <0.001None 1.0 —

<50% 4.19 (1.98e8.83) <0.001�50% 4.28 (1.94e9.47) <0.001

Clinical variables plus eGFR plus CADseverity plus abnormal LVEF

NCEP ATP III score 1.86 (1.24e2.79) 0.003eGFR (ml/min/1.73 m2) 2.01 (1.44e2.81) <0.001

�90 1.0 —

60e89 0.74 (0.39e1.41) 0.36130e59 2.33 (1.18e4.61) 0.015<30 8.74 (3.32e23.05) <0.001

Coronary narrowing 1.81 (1.31e2.51) <0.001None 1.0 —

<50% 4.02 (1.91e8.48) <0.001�50% 4.24 (1.91e9.37) <0.001

Abnormal LVEF (<50%) 4.16 (2.45e7.08) <0.001

Figure 2. Cox risk-adjusted (for clinical predictor) all-cause mortalityefreesurvival by eGFR for patients with eGFRs �90 ml/min/1.73 m2 (blue line),60 to 89 ml/min/1.73 m2 (green line), 30 to 59 ml/min/1.73 m2 (red line),and <30 ml/min/1.73 m2 (black line) (p <0.001).

1566 The American Journal of Cardiology (www.ajconline.org)

significantly lower values in patients with impaired renalfunction.

In total, 2,820 patients (51.6%) had abnormal results onCCTA, with 948 patients (17%) demonstrating obstructiveCAD. The prevalence of abnormal results on CCTAincreased with worsening renal function (37.4% and 74.2%in patients with eGFRs �90 vs <30 ml/min/1.73 m2,respectively; Table 1).

At follow-up, all-cause mortality was observed in 66patients (1.18%), with an annualized mortality rate for theentire study cohort of 0.62%. All-cause mortality signifi-cantly increased with every 10-unit decrease in eGFR(hazard ratio [HR] 1.23, 95% CI 1.07 to 1.41). An eGFR�90 ml/min/1.73 m2 conferred excellent survival, with only16 of the total 2,130 patients (0.75%) dying in this group(Table 2). Decrease in eGFR was associated with a signifi-cant increase (HR 2.97, 95% CI 2.13 to 4.13) in all-causemortality, with 0.75%, 0.81%, 3.45%, and 20% patientsdying during study follow-up for eGFRs �90, 60 to 89, 30to 59, and <30 ml/min/1.73 m2, respectively (Table 3).

The absence of coronary atherosclerosis conferred anexcellent prognosis, with only 9 deaths (0.32%) in thisgroup. Patients with nonobstructive and obstructive CADhad all-cause death rates of 1.82% and 2.43%, respectively.An abnormal LVEF (HR 6.64, 95% CI 3.98 to 11.10) anda 10% decrease in the LVEF (HR 1.54, 95% CI 1.31 to1.82) were associated with significant increase in all-causemortality. A significant increase in all-cause mortality wasalso seen with progressive decrease in eGFR and increase inCAD severity (Figure 1).

In univariate analysis, clinical parameters (age, bodymass index, diabetes, hypertension, and NCEP ATP IIIscore), eGFR, and coronary computed tomographic angio-graphic parameters (severity of CAD and an abnormalLVEF) were significant predictors of all-cause mortality

(Tables 2 and 4). For the risk-adjusted analysis, the NCEPATP III score was used as the clinical predictor to determinethe incremental value of eGFR and coronary computedtomographic angiographic measures because it combines

Figure 3. Cox risk-adjusted (for clinical predictor and eGFR) all-cause mortalityefree survival by CAD for patients without coronary atherosclerosis (blue line),nonobstructive CAD (green line), and obstructive CAD (red line) (p <0.001).

Coronary Artery Disease/Prognosis of Cardiac Computed Tomography 1567

age, gender, and cardiac factors into a single measure. Amultivariate Cox model of eGFR, CAD severity, and anabnormal LVEF was tested. The eGFR, CAD severity, andan abnormal LVEF remained independent predictors of all-cause mortality after adjusting for the clinical predictor(Figures 2 and 3, Table 4). A significant increase (p <0.001)in the global chi-square value for the Cox model wasobserved after the addition of eGFR (global chi-square ¼50.64) to the clinical predictor (global chi-square ¼ 29.56).Additions of CAD severity and an abnormal LVEF to modelswith clinical predictor plus eGFR (global chi-square ¼61.85) and clinical predictor plus eGFR plus CAD severity(global chi-square ¼ 98.78) as variables, respectively, alsodemonstrated significant increases in p values (<0.001).

A receiver-operating characteristic curve was first createdfor clinical predictor (area under the curve 0.67). Subse-quently, eGFR (area under the curve 0.71) and each coro-nary computed tomographic angiographic measure (CADseverity and an abnormal LVEF) were sequentially added inthe model to create additional curves. The area under thecurve was compared to assess the discriminative abilityof each additional measure. The addition of coronarycomputed tomographic angiographic measures resulted ina significant improvement in the area under the curve to0.78 (p <0.001).

Discussion

This is the first multicenter study to evaluate the prog-nostic value of coronary computed tomographic angio-graphic measures in patients with varying degrees of renalimpairment. Additionally, we have also established theimpact of renal function in predicting all-cause mortality inpatients who undergo CCTA. One of the main strengths ofthis study is the demonstration of the prognostic value ofCCTA and renal impairment in a large international multi-center registry. The other strength of our study is the use ofeGFR rather than serum creatinine alone to estimate the truelevel of normal renal function, because eGFR is consideredamong the best measures of overall kidney function inhealth and disease.17

Chronic kidney disease is one of the most importantcardiovascular risk states and is associated with overall poorrates of optimal cardiovascular risk factor control.18,19

Practice guidelines from the National Kidney Foundation in2002 and an American College of Cardiology and AmericanHeart Association task force in 2004 recommended thatchronic renal impairment be considered a coronary heartdisease risk equivalent.18,19 There is now clear evidence thateven mild to moderate renal impairment is associated witha substantial increase in cardiovascular risk.20 Our study hada substantial number (n ¼ 580) of patients with eGFRs of 30to 60 ml/min/1.73 m2, which constitutes the bulk of thepatients with chronic renal impairment in developed coun-tries.21 We found significant mortality in this group (3.45%)compared with patients with eGFRs �90 ml/min/1.73 m2

(0.75%). Findings of CAD on CCTA provided incrementalvalue in this group. Although only 0.55% of the patients inthis group with no CAD experienced mortality, 2.51% and6.38% died in the presence of nonobstructive and obstruc-tive CAD, respectively. These findings could expand theprognostic role of CCTA to this important patient cohort.The exact cause responsible for increased cardiovascularmortality has not been established, but multiple mecha-nisms, such as anemia, increased oxidative stress, derange-ments in calcium-phosphate homeostasis, inflammation,conditions promoting coagulation, and ventricular and atrialcardiomyopathy, have been proposed.22 Our data suggestthat coronary artery atherosclerosis (obstructive and non-obstructive) assessed by CCTA might represent anotherimportant potential mechanism.

Our study confirms that increasing CAD severity onCCTA is associated with a progressive decrease in survival.Furthermore, CAD severity was a predictor of mortalityacross the entire spectrum of renal function. Indeed, thistrend became even more obvious in patients with worseningeGFRs, and the effect persisted after adjustment for clinicalparameters. Impaired renal function and CAD severity wereindependent predictors in the present study population.Because CCTA has the potential to assess nonobstructiveCAD as well, we were able to analyze results according tothe absence of CAD, nonobstructive CAD, and obstructive

1568 The American Journal of Cardiology (www.ajconline.org)

CAD, which is not possible with competing techniques suchas myocardial perfusion imaging and stress echocardiog-raphy. For almost all stages of impaired renal function, thenonobstructive group showed intermediate risk for all-causemortality compared with the corresponding no-CAD andobstructive-CAD groups (Figure 1). We have shown that theLVEF measured during CCTA retains its incremental value.True 3-dimensional reconstructions with high-resolutionendocardial border definition using computed tomographyallow the accurate estimation of LVEF.23 Moreover, newercomputed tomographic scanners have the ability to performCCTA with low patient radiation exposure levels bydecreasing the tube current to 4% during the non-reconstructed phase (MinDose; Siemens Healthcare, Erlan-gen, Germany) in all scanned phases.24 On the basis of ourresults, assessment of the LVEF should be considered ifend-systolic and end-diastolic data sets are available.

Previous studies with myocardial perfusion imaging andstress echocardiography have revealed similar prognosticabilities of these techniques.25,26 However, these techniquesmay have their own limitations in this patient group.25,26

CCTA too has a few limitations in these groups of patients,such as blooming artifacts (from coronary calcification) andrisk for radiation exposure. Recent technological advances,however, have reduced radiation exposure significantly.27

One previous study that collected data from 2 centersevaluated the prognostic role of CCTA in patients withmoderate renal impairment (eGFR 30 to 59 ml/min/1.73 m2)and showed its incremental value.28 Our study populationwas from a large multicenter registry, included patients withsevere renal impairment (eGFR <30 ml/min/1.73 m2), andalso demonstrated the incremental role of CCTA-determinedLVEF. A recent study showed a reduction in the incidenceof major atherosclerotic events with the use of simvastatinand ezetimibe in patients with chronic renal disease with noprevious myocardial infarctions or coronary revasculariza-tion.29 Because the use of statin therapy has been postulatedto reduce adverse events in patients with nonobstructivedisease by “stabilizing” underlying CAD, our study findingsmay also be useful for consideration in future recommen-dations about the use of this group of drugs in patients withimpaired renal function with nonobstructive CAD.

Cardiac death data were not universally available, so all-cause mortality was used as the outcome. The true propor-tion of deaths attributable to cardiac or cardiovascular eventsin the study population is therefore not known. However,capture of all-cause mortality has its own advantages, suchas an ability to capture adverse deaths that could be relatedto complications (such as pneumonia) arising from preex-isting cardiac (heart failure from CAD) or renal (fluidoverload) conditions. The clinical risk predictor (NCEPATP III score) used in this analysis is used for predictingcardiovascular events, not all-cause mortality. It may havetherefore underestimated the prognostic power of clinicalvariables and overestimated that of eGFR and CCTA. Notall centers collected information regarding early revascu-larization. It is possible that patients diagnosed withobstructive CAD on CCTA could have undergone revas-cularization, resulting in superior survival and underesti-mation of the true prognostic power of eGFR and CCTA.Serum creatinine was not available in a large number of

patients in the registry, so selection bias cannot be excluded.Last, because not all CONFIRM centers reported absolutepercentage LVEFs, the LVEF could not be analyzed asa continuous variable for the entire cohort. However, datawere available at most centers regarding the presence ofa normal or an abnormal LVEF, defined volumetrically at athreshold of 50%, and the LVEF was thus analyzed asa categorical variable. However, in a subset of 2,896patients with absolute percentage LVEFs, the main studyfindings were confirmed.

Acknowledgment: We extend our gratitude to the inves-tigators at each participating center.

Disclosures

Dr. Achenbach receives grant support from SiemensHealthcare, Erlangen, Germany, and Bayer Schering PharmaAG, Berlin, Germany, and is a consultant for Servier,Neuilly-sur-Seine, France. Dr. Budoff is on the speaker’sbureau of GE Healthcare. Dr. Cademartiri receives grantsupport from GE Healthcare, is a consultant for Servier, andis on the speaker’s bureau of Bracco, Milan, Italy.Dr. Chinnaiyan receives grant support from Bayer PharmaAG and Blue Cross Blue Shield Blue Care Network ofMichigan, Grand Rapids, Michigan. Dr. Kaufmann receivesgrant support from the Swiss National Science Foundation,Bern, Switzerland, and GE Healthcare. Dr. Maffei receivesgrant support from GE Healthcare. Dr. Raff receives grantsupport from Siemens Healthcare, Blue Cross Blue ShieldBlue Care Network of Michigan, and Bayer Pharma AG.Dr. Min receives research support from and is on thespeaker’s bureau of GE Healthcare. Dr. Chow receivesresearch support from GE Healthcare, Milwaukee, Wiscon-sin; Pfizer, Inc., New York, New York; and AstraZeneca,Wilmington, Delaware. Dr. Chow receives fellowshipsupport from GE Healthcare and educational support fromTeraRecon Inc, Foster City, California.

1. Hamon M, Biondi-Zoccai GG, Malagutti P, Agostoni P, Morello R,Valgimigli M, Hamon M. Diagnostic performance of multislice spiralcomputed tomography of coronary arteries as compared with conven-tional invasive coronary angiography: a meta-analysis. J Am CollCardiol 2006;48:1896e1910.

2. Min JK, Shaw LJ, Devereux RB, Okin PM, Weinsaft JW, Russo DJ,Lippolis NJ, Berman DS, Callister TQ. Prognostic value of multi-detector coronary computed tomographic angiography for prediction ofall-cause mortality. J Am Coll Cardiol 2007;50:1161e1170.

3. Chow BJ, Wells GA, Chen L, Yam Y, Galiwango P, Abraham A, ShethT, Dennie C, Beanlands RS, Ruddy TD. Prognostic value of 64-slicecardiac computed tomography: severity of coronary artery disease,coronary atherosclerosis and left ventricular ejection fraction. J Am CollCardiol 2010;55:1017e1028.

4. Chow BJ, Small G, Yam Y, Chen L, Achenbach S, Al-Mallah M,Berman DS, Budoff MJ, Cademartiri F, Callister TQ, Chang HJ, ChengV, Chinnaiyan KM, Delago A, Dunning A, Hadamitzky M, HausleiterJ, Kaufmann P, Lin F, Maffei E, Raff GL, Shaw LJ, Villines TC, MinJK; on Behalf of the CONFIRM Investigators. Incremental prognosticvalue of cardiac computed tomography in coronary artery disease usingCONFIRM: Coronary Computed Tomography Angiography Evalua-tion for Clinical Outcomes: an International Multicenter Registry. CircCardiovasc Imaging 2011;4:463e472.

5. Min JK, Dunning A, Lin FY, Achenbach S, Al-Mallah MH, BermanDS, Budoff MJ, Cademartiri F, Callister TQ, Chang HJ, Cheng V,Chinnaiyan KM, Chow B, Delago A, Hadamitzky M, Hausleiter J,

Coronary Artery Disease/Prognosis of Cardiac Computed Tomography 1569

Karlsberg RP, Kaufmann P, Maffei E, Nasir K, Pencina MJ, Raff GL,Shaw LJ, Villines TC. Rationale and design of the CONFIRM(Coronary CT Angiography Evaluation for Clinical Outcomes: AnInternational Multicenter) registry. J Cardiovasc Comput Tomogr2011;5:84e92.

6. Katzberg RW, Lamba R. Contrast-induced nephropathy after intrave-nous administration: fact or fiction? Radiol Clin North Am 2009;47:789e800.

7. Diamond GA, Forrester JS. Analysis of probability as an aid in theclinical diagnosis of coronary-artery disease. N Engl J Med 1979;300:1350e1358.

8. Chaitman BR, Bourassa MG, Davis K, Rogers WJ, Tyras DH, BergerR, Kennedy JW, Fisher L, Judkins MP, Mock MB, Killip T. Angio-graphic prevalence of high-risk coronary artery disease in patientsubsets (CASS). Circulation 1981;64:360e367.

9. Gibbons RJ, Chatterjee K, Daley J, Douglas JS, Fihn SD, Gardin JM,Grunwald MA, Levy D, Lytle BW, O’Rourke RA, Schafer WP,Williams SV, Ritchie JL, Gibbons RJ, Cheitlin MD, Eagle KA,Gardner TJ, Garson A Jr, Russell RO, Ryan TJ, Smith SC Jr. ACC/AHA/ACP-ASIM guidelines for the management of patients withchronic stable angina: executive summary and recommendations:a report of the American College of Cardiology/American HeartAssociation Task Force on Practice Guidelines (Committee onManagement of Patients With Chronic Stable Angina). Circulation1999;99:2829e2848.

10. National Cholesterol Education Program Expert Panel on Detection,Evaluation, and Treatment of High Blood Cholesterol in Adults (AdultTreatment Panel III) Final Report. Bethesda, Maryland: NationalInstitutes of Health, 2009.

11. Abbara S, Arbab-Zadeh A, Callister TQ, Desai MY, Mamuya W,Thomson L, Weigold WG. SCCT guidelines for performance ofcoronary computed tomographic angiography: a report of the Societyof Cardiovascular Computed Tomography Guidelines Committee.J Cardiovasc Comput Tomogr 2009;3:190e204.

12. Raff GL, Abidov A, Achenbach S, Berman DS, Boxt LM, Budoff MJ,Cheng V, DeFrance T, Hellinger JC, Karlsberg RP, Society ofCardiovascular Computed Tomography. SCCT guidelines for theinterpretation and reporting of coronary computed tomographic angi-ography. J Cardiovasc Comput Tomogr 2009;3:122e136.

13. Hoffmann U, Moselewski F, Cury RC, Ferencik M, Jang IK, Diaz LJ,Abbara S, Brady TJ, Achenbach S. Predictive value of 16-slice mul-tidetector spiral computed tomography to detect significant obstructivecoronary artery disease in patients at high risk for coronary arterydisease: patient-versus segment-based analysis. Circulation 2004;110:2638e2643.

14. Lin FY, Devereux RB, Roman MJ, Meng J, Jow VM, Jacobs A,Weinsaft JW, Shaw LJ, Berman DS, Callister TQ, Min JK. Cardiacchamber volumes, function, and mass as determined by 64-multi-detector row computed tomography: mean values among healthy adultsfree of hypertension and obesity. JACC Cardiovasc Imaging 2008;1:782e786.

15. Levey AS, Coresh J, Greene T, Marsh J, Stevens LA, Kusek JW, VanLente F. Expressing the Modification of Diet in Renal Disease studyequation for estimating glomerular filtration rate with standardizedserum creatinine values. Clin Chem 2007;53:766e772.

16. Levey AS, Coresh J, Balk E, Kausz AT, Levin A, Steffes MW, HoggRJ, Perrone RD, Lau J, Eknoyan G. National Kidney Foundationpractice guidelines for chronic kidney disease: evaluation, classifica-tion, and stratification. Ann Intern Med 2003;139:137e147.

17. Smith HW. Comparative physiology of the kidney. In: Smith HW, ed.The Kidney: Structure and Function in Health and Disease. New York,New York: Oxford University Press, 1951:520e574.

18. National Kidney Foundation. K/DOQI clinical practice guidelines forchronic kidney disease: evaluation, classification, and stratification. AmJ Kidney Dis 2002;39:S1.

19. Antman EM, Anbe DT, Armstrong PW, Bates ER, Green LA, Hand M,Hochman JS, Krumholz HM, Kushner FG, Lamas GA, Mullany CJ,Ornato JP, Pearle DL, Sloan MA, Smith SC Jr, Alpert JS, Anderson JL,Faxon DP, Fuster V, Gibbons RJ, Gregoratos G, Halperin JL, HiratzkaLF, Hunt SA, Jacobs AK; American College of Cardiology/AmericanHeart Association Task Force on Practice Guidelines (WritingCommittee to Revise the 1999 Guidelines for the Management of

Patients With Acute Myocardial Infarction). ACC/AHA guidelines forthe management of patients with ST-elevation myocardial infarction—executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines(Writing Committee to Revise the 1999 Guidelines for the Manage-ment of Patients With Acute Myocardial Infarction). Circulation2004;110:588e636.

20. van der Velde M, Matsushita K, Coresh J, Astor BC, Woodward M,Levey A, van der Velde M, Matsushita K, Coresh J, Astor BC,Woodward M, Levey A, de Jong P, Gansevoort RT; Chronic KidneyDisease Prognosis Consortium, van der Velde M, Matsushita K, Cor-esh J, Astor BC, Woodward M, Levey AS, de Jong PE, GansevoortRT, Levey A, El-Nahas M, Eckardt KU, Kasiske BL, Ninomiya T,Chalmers J, Macmahon S, Tonelli M, Hemmelgarn B, Sacks F, CurhanG, Collins AJ, Li S, Chen SC, Hawaii Cohort KP, Lee BJ, Ishani A,Neaton J, Svendsen K, Mann JF, Yusuf S, Teo KK, Gao P, Nelson RG,Knowler WC, Bilo HJ, Joosten H, Kleefstra N, Groenier KH, AugusteP, Veldhuis K, Wang Y, Camarata L, Thomas B, Manley T. Lowerestimated glomerular filtration rate and higher albuminuria are associ-ated with all-cause and cardiovascular mortality. A collaborative meta-analysis of high-risk population cohorts. Kidney Int 2011;79:1341e1352.

21. Levey AS, Coresh J, Greene T, Stevens LA, Zhang YL, Hendriksen S,Kusek JW, Van Lente F; Chronic Kidney Disease EpidemiologyCollaboration. Using standardized serum creatinine values in themodification of diet in renal disease study equation for estimatingglomerular filtration rate [published correction appears in Ann InternMed 2008;149:519]. Ann Intern Med 2006;145:247e254.

22. Herzog CA, Asinger RW, Berger AK, Charytan DM, Díez J, Hart RG,Eckardt KU, Kasiske BL, McCullough PA, Passman RS, DeLoach SS,Pun PH, Ritz E. Cardiovascular disease in chronic kidney disease. Aclinical update from Kidney Disease: Improving Global Outcomes(KDIGO). Kidney Int 2011;80:572e586.

23. de Graaf FR, Schuijf JD, van Velzen JE, Nucifora G, Kroft LJ, de RoosA, Jukema JW, van der Wall EE, Bax JJ. Assessment of global leftventricular function and volumes with 320-row multidetector computedtomography: a comparison with 2D-echocardiography. J Nucl Cardiol2010;17:225e231.

24. Petersilka M, Bruder H, Krauss B, Stierstorfer K, Flohr TG. Technicalprinciples of dual source CT. Eur J Radiol 2008;68:362e368.

25. Hakeem A, Bhatti S, Dillie KS, Cook JR, Samad Z, Roth-Cline MD,Chang SM. Predictive value of myocardial perfusion single-photonemission computed tomography and the impact of renal function oncardiac death. Circulation 2008;118:2540e2549.

26. Karagiannis SE, Feringa HH, Elhendy A, van Domburg R, ChoncholM, Vidakovic R, Bax JJ, Karatasakis G, Athanasopoulos G, CokkinosDV, Poldermans D. Prognostic significance of renal function in patientsundergoing dobutamine stress echocardiography. Nephrol DialTransplant 2008;23:601e607.

27. Cordeiro MA, Miller JM, Schmidt A, Lardo AC, Rosen BD, Bush DE,Brinker JA, Bluemke DA, Shapiro EP, Lima JA. Non-invasive halfmillimetre 32 detector row computed tomography angiography accu-rately excludes significant stenoses in patients with advanced coronaryartery disease and high calcium scores. Heart 2006;92:589e597.

28. Yiu KH, de Graaf FR, Schuijf JD, van Werkhoven JM, van Velzen JE,Boogers MJ, Roos CJ, de Bie MK, Pazhenkottil A, Kroft LJ, BoersmaE, Herzog B, de Roos A, Kaufmann PA, Bax JJ, Jukema JW. Prog-nostic value of renal dysfunction for the prediction of outcome versusresults of computed tomographic coronary angiography. Am J Cardiol2011;108:968e972.

29. Baigent C, Landray MJ, Reith C, Emberson J, Wheeler DC, Tomson C,Wanner C, Krane V, Cass A, Craig J, Neal B, Jiang L, Hooi LS, LevinA, Agodoa L, Gaziano M, Kasiske B, Walker R, Massy ZA, Feldt-Rasmussen B, Krairittichai U, Ophascharoensuk V, Fellström B,Holdaas H, Tesar V, Wiecek A, Grobbee D, de Zeeuw D, Grönhagen-Riska C, Dasgupta T, Lewis D, Herrington W, Mafham M, Majoni W,Wallendszus K, Grimm R, Pedersen T, Tobert J, Armitage J, Baxter A,Bray C, Chen Y, Chen Z, Hill M, Knott C, Parish S, Simpson D,Sleight P, Young A, Collins R; SHARP Investigators. The effects oflowering LDL cholesterol with simvastatin plus ezetimibe in patientswith chronic kidney disease (Study of Heart and Renal Protection):a randomised placebo-controlled trial. Lancet 2011;377:2181e2192.