Prognostic Value of CT Angiography in Coronary Bypass Patients

Download Prognostic Value of CT Angiography in Coronary Bypass Patients

Post on 28-Nov-2016




4 download

Embed Size (px)


  • PC





    B A










    R E















    J A C C : C A R D I O V A S C U L A R I M A G I N G V O L . 4 , N O . 5 , 2 0 1 1

    2 0 1 1 B Y T H E A M E R I C A N C O L L E G E O F C A R D I O L O G Y F O U N D A T I O N I S S N 1 9 3 6 - 8 7 8 X / $ 3 6 . 0 0

    P U Brognostic Value of CT Angiography inoronary Bypass Patients

    njamin J. W. Chow, MD,* Osman Ahmed, BSC,* Gary Small, MBCHB,*dul-Aziz Alghamdi, MBBS,* Yeung Yam, BSC,* Li Chen, MSC, George A. Wells, PHD

    tawa, Ontario, Canada

    B J E C T I V E S We sought the incremental prognostic value of coronary computed tomography

    giography (CTA) in coronary artery bypass graft (CABG) patients.

    C KG ROUND Coronary CTA is a noninvasive and accurate tool for the detection of obstructive

    ronary artery disease, and coronary CTA appears to have prognostic value in patients without previous

    ascularization. However, the prognostic value of coronary CTA to predict major adverse cardiac

    ents in CABG patients is unclear.

    E THOD S Consecutive CABG patients were prospectively enrolled and cardiac risk was calculated

    ing the National Cholesterol Evaluation Program/Adult Treatment Panel III. Using the severity of native

    ronary artery disease and graft disease, the number of unprotected coronary territories (UCTs) (0, 1,

    or 3) was calculated. Patients were followed for cardiac death and nonfatal myocardial infarction. All

    ents were conrmed with death certicates or medical records and reviewed by a clinical events


    S U L T S Between February 2006 and March 2009, 250 consecutive patients were enrolled and

    lowed for a mean of 20.8 10.1 months. At follow-up, 23 patients (9.2%) had major adverse cardiac

    ents (15 cardiac deaths and 8 nonfatal MI). The absence of UCTs conferred a good prognosis with an

    nual event rate of 2.4%. Conversely, patients with 1, 2, and 3 UCTs had annualized event rates of 5.8%,

    .1%, and 21.7%, respectively. Multivariable analysis showed that UCTs (hazard ratio: 2.08; 95%

    ndence interval: 1.40 to 3.10; p 0.001) was a predictor of major adverse cardiac events when

    justed for clinical variables. Examining the receiver-operator characteristic curves, the area under the

    rve increased from 0.61 to 0.76 when UCTs was combined with clinical variables (p 0.001).

    NC L U S I O N S Assessing UCTs with coronary CTA appears to have prognostic value in CABG

    tients and is incremental to clinical variables. Coronary CTA appears to be a promising tool for risk

    atication of CABG patients. Further multicenter studies using large CABG cohorts are needed to

    nrm our ndings. (J Am Coll Cardiol Img 2011;4:496502) 2011 by the American College of

    rdiology Foundation

    m the *Division of Cardiology, University of Ottawa Heart Institute, Ottawa, Ontario, Canada; Department of Radiology,e Ottawa Hospital, Ottawa, Ontario, Canada; and the Cardiovascular Research Methods Centre, University of Ottawaart Institute, Ottawa, Ontario, Canada. This study was supported in part by the Imaging for Cardiovascular Therapeutics

    L I S H E D B Y E L S E V I E R I N C . D O I : 1 0 . 1 0 1 6 / j . j c m g . 2 0 1 1 . 0 1 . 0 1 5ject RE02-038 and the Canada Foundation for Innovation No. 11966. Dr. Chow is supported by CIHR New Investigatorard MSH-83718. He receives research support from GE Healthcare, Pfizer Inc., and AstraZeneca; fellowship trainingport from GE Healthcare; and educational support from TeraRecon Inc. The other authors have reported that they haverelationships to disclose.

    nuscript received November 10, 2010; revised manuscript received January 14, 2011; accepted January 18, 2011.

  • Cass(Cvalad









    P III National

    ol Education Program/

    atment Program III

    protected coronary

    J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 4 , N O . 5 , 2 0 1 1

    M A Y 2 0 1 1 : 4 9 6 5 0 2

    Chow et al.

    Prognosis of CTA in CABG Patients

    497oronary computed tomography angiogra-phy (CTA) is a noninvasive diagnostictool useful for the detection of obstructivecoronary artery disease (CAD) and the

    essment of coronary artery bypass graftsABG) (113). Coronary CTA has prognosticue and predicts all-cause mortality and majorverse cardiac events (MACE) such as cardiac

    See pages 492 and 503

    ath and nonfatal myocardial infarction (MI)19). However, previous studies have excluded

    tients with a history of coronary revascularization19). To further support the utility of coronaryA in patients with CABG, the prognostic value64-slice coronary CTA in this population re-

    ires further investigation.Studies using invasive coronary angiography havemonstrated that the completeness of revascular-tion is an important predictor of outcome andt the number of arterial territories lacking a

    tent graft is a key determinant of mortality (20).though coronary CTA can assess coronary andft disease, its ability to assess the native arteriesCABG patients has been questioned (3,21).us, demonstrating the prognostic value of coro-

    ry CTA in the CABG population would furtherport its clinical use.

    The objective of this prospective cohort study isunderstand the potential incremental prognosticue of coronary CTA in CABG patients.

    E T H O D S

    tween February 2006 and March 2009, 269nsecutive CABG patients undergoing coronary

    A were prospectively enrolled in a Cardiac CTAgistry and were followed for all-cause death andACE (cardiac death and nonfatal MI) (13,14).e study was approved by the Institutional Hu-n Research Ethics Board, and all patients pro-ed written informed consent.nical predictors. A detailed medical history andoratory results were recorded at the time of

    ronary CTA. Although a validated prognosticnical model for CABG patients is lacking, age,, hypertension, and diabetes appear to be predic-

    s of late cardiac events in CABG patients,23). Therefore, patients age, sex, cardiac risktors, and symptoms were used to estimate patientk using the National Cholesterol Education Pro-

    ginunartstem/Adult Treatment Panel III (NCEP/ATP III)idelines and the Morise score (14,24,25).ronary CTA. Before image acquisition, metoprololdiltiazem (oral and/or intravenous) was admin-

    ered, targeting a heart rate of65 beats/min, androglycerin 0.8 mg was administered sublingually,26,27). A biphasic timing bolus (15 to 25 ml of

    ntrast (Visipaque 320 or Omnipaque 350, GEalthcare, Princeton, New Jersey; 40 ml of salineution) was used (13,14). A triphasic protocol0% contrast, 40%/60% contrast/saline solutionml], and saline solution [40 ml]) was used to

    uire the final dataset. The volume and rate ofntrast were individualized according to scan timed patient body habitus (13,14).Retrospective electrocardiogram-gated datasetsre acquired with the GE Volume CT (GEalthcare, Milwaukee, Wisconsin) (64 0.625-mme collimation, 350-ms gantry rotation, 400 to

    0 mA, kilovolt peak 120, and pitch of6 to 0.24) (14). Images were recon-ucted using a slice thickness of 0.625

    with an increment of 0.4 mm usingcardiac phase(s) with the least amount

    cardiac motion (13,14).ronary CTA image analysis. Images werest-processed using the GE Advantagelume Share Workstation (GE Health-e) and interpreted by expert observersnded to all clinical data (14). A 4-pointding score (normal, mild [50%],derate [50% to 69%], severe [70%])s used for the evaluation of native CADd CABG (28). In segments that wereassessable, forced reading was per-med, and readers provided their best educatedess. Cases with 5 unassessable segments wereluded from analysis. Significant stenoses were

    fined as left main 50% diameter stenosis, othertive vessel stenosis 70%, or graft stenosis0%.

    Patients were categorized according to the num-r (0, 1, 2, or 3) of unprotected coronary territoriesCTs) (20). Each patient had 3 coronary territo-s, corresponding to each major epicardial arteryft anterior descending artery, circumflex artery orery supplying the posterior descending arteryght coronary artery or circumflex artery]) andir corresponding branches (diagonal and mar-

    A B B R

    A N D A



    CAD co

    CTA co







    Adult Tre

    UCT un

    territoryal arteries). A coronary territory was deemprotected if: 1) an ungrafted native coronery had a significant stenosis; 2) a significnosis in the native artery was distal to the gE V I A T I O N S

    C R O N YM S

    coronary artery bypass

    ronary artery disease

    mputed tomography


    major adverse cardiac

    cardial infarctionedaryantraft

  • inshaasssysinasigsysPa(attraeverecAofOuwaAlnothediaarrcodemaupStausiCadeprewe


    R E















    J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 4 , N O . 5 , 2 0 1 1

    M A Y 2 0 1 1 : 4 9 6 5 0 2

    Chow et al.

    Prognosis of CTA in CABG Patients

    498ertion; or 3) a native artery and its graft bothd significant stenoses (20). The left main wasigned 2 coronary territories in right dominanttem, but 3 coronary territories when left dom-nt. Similarly, the circumflex artery was as-ned 2 territories when the native coronarytem was left dominant.tient follow-up. Patient follow-up was performed

    6-month intervals) by telephone interview byined research staff blinded to all clinical data. Allnts were confirmed with death records, hospitalords, or correspondence with treating physicians.clinical events committee (blinded to the resultsthe coronary CTA) reviewed all events.tcome measures. The primary outcome measures a composite of cardiac death and nonfatal MI.l deaths were reviewed and classified as cardiac orncardiac. Deaths were considered cardiac when

    primary cause of death was related to myocar-l ischemia/infarction, heart failure or cardiachythmia, and when a noncardiac cause of deathuld not be identified (29). Nonfatal MI wasfined as myocardial ischemia resulting in abnor-l cardiac biomarkers (99th percentile of theper normal limits) (30).tistical analysis. Statistical analyses were performedng SAS software version 9.2 (SAS Institute Inc.,ry, North Carolina), and statistical significance wasfined as p 0.05. Continuous variables weresented as means and SDs, and categorical variablesre presented as frequencies with percentages.The prognostic value of UCTs was assessed forivariable association as well as multivariable as-iation with MACE. All unadjusted comparisonscardiac events were performed using log-rankts. For risk-adjusted analysis, a multivariable Coxportional hazard model was used to assess theependent prognostic value of UCTs adjusted for

    nical variables (NCEP/ATP III) and createdjusted survival curves. Model overfitting was con-ered, and the proportional hazards assumptions met. The incremental value of UCTs wasculated by defining the clinical variables modellowed by the addition of UCTs. The area undereiver-operator characteristic curves (95% confi-nce intervals) was compared to evaluate the dis-mination ability of UCTs over clinical variablespredict MACE.S U L T S

    dy population. Over an enrollment period of 37nths, 4,536 consecutive patients underwent cor-


    Edary CTA with a total of 4,508 patients (99.4%)spectively enrolled in the University of Ottawaart Institute Cardiac CTA Registry. Of these,

    9 had previous CABG surgery and met thelusion criteria for this study. A total of 10

    tients were excluded from analysis for 5 unas-sable segments. Follow-up was available for 250tients (96.5%) with 9 patients lost to follow-upable 1). The baseline characteristics of the pa-nts lost to follow-up were similar to those withlow-up, except those lost to follow-up wereunger and were more likely to have atypical chestin or dyspnea.ivariable analysis of baseline characteristics withCE. Univariable analysis of baseline characteris-

    s is summarized in Table 2. Patients with

    able 1. Characteristics of Patients With Follow-Up (n 250)

    ean follow-up, months 20.8 10.1

    ge, yrs 65.4 9.5

    en 200 (80.0)

    ody mass index, kg/m2 29.5 5.7

    ardiac risk factors

    Smoker/ex-smoker 175 (70.0)

    Hypertension 157 (62.8)

    Dyslipidemia 235 (94.0)

    Diabetes 87 (34.8)

    Family history of CAD 136 (54.4)

    dications for study

    Chest pain 150 (60.0)

    Nonanginal chest pain 22 (8.8)

    Atypical angina 29 (11.6)

    Typical angina 99 (39.6)

    Dyspnea 55 (22.0)

    orise score 14.9 2.8

    CEP/ATP III risk

    Low risk 3 (1.2)

    Intermediate risk 59 (23.6)

    High risk 188 (75.2)

    nprotected coronary territories

    0 128 (51.2)

    1 74 (29.6)

    2 37 (14.8)

    3 11 (4.4)

    eft ventricular ejection fraction, %* 57.9 15.1

    aging parameters

    Imaging heart rate, beats/min 57.9 6.9

    Contrast infusion rate, ml/s 6.2 0.9

    Total contrast volume, ml 125.3 20.4

    Effective dose, mSv 23.2 5.2alues are mean SD or n (%). *Left ventricular ejection fraction could onlye accurately measured in 233 patients. Effective dose (mSv) dose lengthroduct 0.014.CAD coronary artery disease; NCEP/ATP III National Cholesterolucation Program/Adult Treatment Program III.

  • Ma ghigUCwiobof(TproveraliresRisanclicarvarard3.1MTaIncityusiThon0.6coad















    M 0.081

    N 0.061







    J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 4 , N O . 5 , 2 0 1 1

    M A Y 2 0 1 1 : 4 9 6 5 0 2

    Chow et al.

    Prognosis of CTA in CABG Patients

    499ACE were older, had a higher Morise score, andreater proportion were in the NCEP/ATP IIIh-risk category.Ts and MACE. The rates of MACE increasedth the increasing number of UCTs and wereserved in a total of 3.9%, 9.5%, 16.2%, and 45.5%patients with 0, 1, 2, and 3 UCTs, respectivelyable 3). The absence of UCTs conferred a goodgnosis with an annual event rate of 2.4%. Con-sely, patients with 1, 2, and 3 UCTs had annu-zed event rates of 5.8%, 11.1%, and 21.7%,pectively.k-adjusted Cox models. For the risk-adjustedalysis, the NCEP/ATP III was used as thenical variable because it combined age, sex, anddiac risk factors into a single measure. A multi-iable Cox model demonstrated that UCTs (haz-

    ratio: 2.08; 95% confidence interval: 1.40 to0; p 0.001), was an independent predictor for

    ACE adjusted for the clinical variables (Fig. 1,ble 4).remental value analysis. The discrimination abil-of UCTs over NCEP/ATP III was evaluated

    ng receiver-operator characteristic curves (Fig. 2).e area under the curve for the clinical variablely was 0.61 (95% confidence interval: 0.56 to6), with a significant increase to 0.76 (95%

    nfidence interval: 0.66 to 0.86) when UCTs wasded (p 0.001).

    I S C U S S I O N

    our knowledge, the prognostic value of coro-ry CTA in CABG patients has not beenviously reported. The results of our study

    ggest that coronary CTA assessment of UCTsof prognostic value and is incremental to

    nical measures.gnostic value of coronary CTA. The diagnosticuracy and prognostic value of coronary CTA

    ve been well studied but have focused on pati...


View more >