The worsening healthcare crisis has increased pressures to. Imaging

Hasonló dokumentumok
Correlation & Linear Regression in SPSS

Miskolci Egyetem Gazdaságtudományi Kar Üzleti Információgazdálkodási és Módszertani Intézet Factor Analysis

STUDENT LOGBOOK. 1 week general practice course for the 6 th year medical students SEMMELWEIS EGYETEM. Name of the student:

A stabil angina kivizsgálását befolyásoló tényezők a finanszírozási adatbázok alapján

Miskolci Egyetem Gazdaságtudományi Kar Üzleti Információgazdálkodási és Módszertani Intézet Nonparametric Tests

Correlation & Linear Regression in SPSS

Statistical Inference

Miskolci Egyetem Gazdaságtudományi Kar Üzleti Információgazdálkodási és Módszertani Intézet. Correlation & Linear. Petra Petrovics.

FAMILY STRUCTURES THROUGH THE LIFE CYCLE

Mark Auspitz, Fayez Quereshy, Allan Okrainec, Alvina Tse, Sanjeev Sockalingam, Michelle Cleghorn, Timothy Jackson

Gottsegen National Institute of Cardiology. Prof. A. JÁNOSI

OROSZ MÁRTA DR., GÁLFFY GABRIELLA DR., KOVÁCS DOROTTYA ÁGH TAMÁS DR., MÉSZÁROS ÁGNES DR.

Miskolci Egyetem Gazdaságtudományi Kar Üzleti Információgazdálkodási és Módszertani Intézet. Nonparametric Tests. Petra Petrovics.

Construction of a cube given with its centre and a sideline

Miskolci Egyetem Gazdaságtudományi Kar Üzleti Információgazdálkodási és Módszertani Intézet. Hypothesis Testing. Petra Petrovics.

Választási modellek 3

A rosszindulatú daganatos halálozás változása 1975 és 2001 között Magyarországon

Supporting Information

Statistical Dependence

EN United in diversity EN A8-0206/419. Amendment

Nuclear cardiology. Zámbó Katalin Department of Nuclear Medicine

Abigail Norfleet James, Ph.D.

Szív MR vizsgálat az ISZB diagnosztikájában.

Cluster Analysis. Potyó László

On The Number Of Slim Semimodular Lattices

Cashback 2015 Deposit Promotion teljes szabályzat

EPILEPSY TREATMENT: VAGUS NERVE STIMULATION. Sakoun Phommavongsa November 12, 2013

Sebastián Sáez Senior Trade Economist INTERNATIONAL TRADE DEPARTMENT WORLD BANK

Supplementary Table 1. Cystometric parameters in sham-operated wild type and Trpv4 -/- rats during saline infusion and

Performance Modeling of Intelligent Car Parking Systems

Computer Architecture

EGÉSZSÉGTUDOMÁNY, LVII. ÉVFOLYAM, SZÁM 2013/4

Gazdaságtudományi Kar. Gazdaságelméleti és Módszertani Intézet. Logistic regression. Quantitative Statistical Methods. Dr.

Angol Középfokú Nyelvvizsgázók Bibliája: Nyelvtani összefoglalás, 30 kidolgozott szóbeli tétel, esszé és minta levelek + rendhagyó igék jelentéssel

BKI13ATEX0030/1 EK-Típus Vizsgálati Tanúsítvány/ EC-Type Examination Certificate 1. kiegészítés / Amendment 1 MSZ EN :2014

Using the CW-Net in a user defined IP network

ELEKTRONIKAI ALAPISMERETEK ANGOL NYELVEN

ANGOL NYELV KÖZÉPSZINT SZÓBELI VIZSGA I. VIZSGÁZTATÓI PÉLDÁNY

Miskolci Egyetem Gazdaságtudományi Kar Üzleti Információgazdálkodási és Módszertani Intézet. Correlation & Regression

First experiences with Gd fuel assemblies in. Tamás Parkó, Botond Beliczai AER Symposium

A jövőbeli hatások vizsgálatához felhasznált klímamodell-adatok Climate model data used for future impact studies Szépszó Gabriella

PIACI HIRDETMÉNY / MARKET NOTICE

FORGÁCS ANNA 1 LISÁNYI ENDRÉNÉ BEKE JUDIT 2

Report on the main results of the surveillance under article 11 for annex II, IV and V species (Annex B)

PROSPECTIVE ASSESSMENT OF THE RISK OF BACTEREMIA IN CIRRHOTIC PATIENTS AFTER EUS WITH AND WITHOUT FNA

Descriptive Statistics

Szívkatéterek hajlékonysága, meghajlítása

Report on esi Scientific plans 7 th EU Framework Program. José Castell Vice-President ecopa, ES

Tudományos Ismeretterjesztő Társulat

Étkezési búzák mikotoxin tartalmának meghatározása prevenciós lehetıségek

ANGOL NYELV KÖZÉPSZINT SZÓBELI VIZSGA I. VIZSGÁZTATÓI PÉLDÁNY

Decision where Process Based OpRisk Management. made the difference. Norbert Kozma Head of Operational Risk Control. Erste Bank Hungary

Az fmri alapjai BOLD fiziológia. Dr. Kincses Tamás Szegedi Tudományegyetem Neurológiai Klinika

Eladni könnyedén? Oracle Sales Cloud. Horváth Tünde Principal Sales Consultant március 23.

The Impact of Developmental Summer Bridge Programs on Students Success

EEA, Eionet and Country visits. Bernt Röndell - SES

Rezgésdiagnosztika. Diagnosztika

Utolsó frissítés / Last update: február Szerkesztő / Editor: Csatlós Árpádné

A BÜKKI KARSZTVÍZSZINT ÉSZLELŐ RENDSZER KERETÉBEN GYŰJTÖTT HIDROMETEOROLÓGIAI ADATOK ELEMZÉSE

HALLGATÓI KÉRDŐÍV ÉS TESZT ÉRTÉKELÉSE

Paediatrics: introduction. Historical data.

IP/09/473. Brüsszel, március 25

Emelt szint SZÓBELI VIZSGA VIZSGÁZTATÓI PÉLDÁNY VIZSGÁZTATÓI. (A részfeladat tanulmányozására a vizsgázónak fél perc áll a rendelkezésére.

A jövedelem alakulásának vizsgálata az észak-alföldi régióban az évi adatok alapján

IES TM Evaluating Light Source Color Rendition

Phenotype. Genotype. It is like any other experiment! What is a bioinformatics experiment? Remember the Goal. Infectious Disease Paradigm

SAJTÓKÖZLEMÉNY Budapest július 13.

MATEMATIKA ANGOL NYELVEN MATHEMATICS

Expansion of Red Deer and afforestation in Hungary

Genome 373: Hidden Markov Models I. Doug Fowler

Intézményi IKI Gazdasági Nyelvi Vizsga

Mapping Sequencing Reads to a Reference Genome

failed back surgery syndrome gerincműtét utáni krónikus fájdalom szindróma

FÖLDRAJZ ANGOL NYELVEN

Utolsó frissítés / Last update: Szeptember / September Szerkesztő / Editor: Csatlós Árpádné

ACTA CAROLUS ROBERTUS. Károly Róbert Főiskola Gazdaság és Társadalomtudományi Kar tudományos közleményei Alapítva: (1)

The Hungarian National Bibliography. Peter Dippold National Széchényi Library

Implementation of water quality monitoring

Report on the main results of the surveillance under article 11 for annex II, IV and V species (Annex B)

N É H Á N Y A D A T A BUDAPESTI ÜGYVÉDEKRŐ L

ELEKTRONIKAI ALAPISMERETEK ANGOL NYELVEN

ESBL termelő baktériumok okozta infekciók kezelési lehetőségei. Szalka András Szent Imre Kórház

Módszertani eljárások az időtényező vezetési, szervezeti folyamatokban betöltött szerepének vizsgálatához

Akut szívelégtelenség -a kardiológus szemével. Dr. Becker Dávid Semmelweis Egyetem Városmajori Szív- és Érgyógyászati Klinika

Hasznos és kártevő rovarok monitorozása innovatív szenzorokkal (LIFE13 ENV/HU/001092)

3. MINTAFELADATSOR KÖZÉPSZINT. Az írásbeli vizsga időtartama: 30 perc. III. Hallott szöveg értése

Dr. Sasvári Péter Egyetemi docens

Supplementary materials to: Whole-mount single molecule FISH method for zebrafish embryo

NSR Settlements. This session will discuss:

4 vana, vanb, vanc1, vanc2

Klaszterezés, 2. rész

Report on the main results of the surveillance under article 11 for annex II, IV and V species (Annex B)

Report on the main results of the surveillance under article 11 for annex II, IV and V species (Annex B)

Bird species status and trends reporting format for the period (Annex 2)

Report on the main results of the surveillance under article 11 for annex II, IV and V species (Annex B)

Összegyűjtöttük, a magyar radiológus és radiográfus közösség hogyan vesz részt aktívan az idei európai radiológiai kongresszuson.

Mr. Adam Smith Smith's Plastics 8 Crossfield Road Selly Oak Birmingham West Midlands B29 1WQ


Szükségletre korrigált egészségügyi ellátás igénybevételének egyenlôtlenségei Magyarországon

Lexington Public Schools 146 Maple Street Lexington, Massachusetts 02420

Átírás:

Imaging Incremental Prognostic Significance of Combined Cardiac Magnetic Resonance Imaging, Adenosine Stress Perfusion, Delayed Enhancement, and Left Ventricular Function Over Preimaging Information for the Prediction of Adverse Events Scott E. Bingham, MD; Rory Hachamovitch, MD, MSc Downloaded from http://circ.ahajournals.org/ by guest on April 4, 2017 Background Although cardiac magnetic resonance imaging (CMR) is capable of yielding extensive data in routine practice, the relative incremental prognostic value of adenosine stress perfusion, myocardial delayed enhancement (DE), and left ventricular volumes and function is unclear. Methods and Results We followed up 908 consecutive patients who underwent combined CMR for suspicion of coronary stenosis and/or ischemia at 2.6 1.2 years, during which 101 total cardiac events occurred (all-cause death, myocardial infarction, or late revascularization). Increase in Cox proportional-hazards model global 2 ( 2 ) with the addition of CMR data after adjustment for clinical data defined incremental prognostic value. Cardiac magnetic resonance imaging without abnormalities had a 2.4% event rate per year ( 1% cardiac death or myocardial infarction). Abnormal CMR was associated with event rates of 5.6% to 7.0% per year, varying with which and how many components were abnormal. After adjusting for the pre-cmr data (age, dyspnea, prior coronary artery disease, resting heart rate, renal disease, and diabetes mellitus, 2 :43.6, P 0.0001; C index 0.695), the addition of left ventricular ejection fraction, aortic flow, delayed enhancement, and stress perfusion data all incrementally increased 2 (55.2, 63.3, 68.0, and 68.9, respectively; all P 0.00001; C indices 0.717, 0.722, 0.747, and 0.736). The number of abnormal CMR domains both added incremental prognostic value and risk stratified patients with respect to risk of events. Conclusions CMR analysis of ventricular volume, aortic flow, myocardial viability, and stress perfusion all add incremental value for prediction of adverse events over pre-cmr data and can be combined to further enhance prognostication. Normal combined CMR confers a low risk of subsequent cardiac events. (Circulation. 2011;123:1509-1518.) Key Words: magnetic resonance imaging myocardial ischemia perfusion survival myocardial infarction The worsening healthcare crisis has increased pressures to limit the use of expensive imaging modalities. 1 These issues have been exacerbated by the rapid expansion of the imaging armamentarium used for the assessment of patients with known or suspected coronary artery disease (CAD). This has resulted in pressures to more thoroughly evaluate new and emerging technologies before their clinical acceptance and reimbursement, possibly requiring the development of a body of evidence to support their use. Clinical Perspective on p 1518 Prognostic assessment plays a central role in the evaluation and management of patients with known or suspected CAD. Various noninvasive cardiac imaging modalities that are designed to assess myocardial perfusion or ventricular wall motion have demonstrated incremental prognostic power compared with clinical risk assessment. 2 7 Of the newer modalities currently in use, cardiac magnetic resonance imaging (CMR) uniquely provides detailed diagnostic information on myocardial viability, vasodilator stress perfusion, ventricular wall motion and volumes, cardiac output, and valvular abnormalities. 8 Although stress CMR protocols often assess and report all of these parameters, previous studies have focused on the incremental prognostic value of 1 or 2 parameters. The relative incremental predictive value of the combination of all of these component indices of CMR for future cardiac events is unknown. The objective of the current study was to assess the incremental prognostic power of myocardial viability, vasodilator stress perfusion, and ventricular wall motion and volumes over patient clinical and historical data alone for the prediction of adverse cardiac events. Methods The current study is intended to be an effectiveness study (ie, a study in which referring physicians rather than the study design dictate the selection of patients for testing and the nature of post-cmr patient management). Received September 6, 2009; accepted January 28, 2011. From the Central Utah Clinic, Provo, UT (S.E.B.); and Section of Cardiovascular Imaging, Department of Cardiovascular Medicine, Cleveland Clinic, Cleveland, OH (R.H.). The online-only Data Supplement is available with this article at http://circ.ahajournals.org/cgi/content/full/circulationaha.109.907659/dc1. Correspondence to Scott E. Bingham, MD, Director of Imaging, Central Utah Clinic, 1055 N 500 W, Provo, UT 84604. E-mail sbingham@cut.net 2011 American Heart Association, Inc. Circulation is available at http://circ.ahajournals.org DOI: 10.1161/CIRCULATIONAHA.109.907659 1509

1510 Circulation April 12, 2011 Downloaded from http://circ.ahajournals.org/ by guest on April 4, 2017 1009 Patients referred for clinically indicated Stress Perfusion CMR 932 Patients 7 Patients unable to complete protocol due to claustrophobia 70 Patients excluded due to significant valvular disease 908 eligible Patients with follow up 24 Patients lost to follow-up (2.6%) Figure 1. Derivation of study cohort. CMR indicates cardiac magnetic resonance imaging. Study Population All patients referred to stress CMR at Central Utah Clinic (Provo, UT) between July 7, 2002, and October 31, 2006 (inclusive), were prospectively enrolled in a preplanned and predefined data registry. Informed consent was obtained at the time of testing, and all patient information was collected and protected in compliance with the Health Insurance Portability and Accountability Act standards under investigational review board approval. We identified 1009 patients who were referred for stress CMR and underwent a protocol combining 4 components: adenosine stress perfusion, myocardial delayed enhancement, left ventricular volumes and function, and aortic blood flow CMR in an outpatient clinic setting. As shown in Figure 1,7 patients who were initially referred for stress CMR began, but did not complete, imaging because of claustrophobia and were excluded from analyses. Following American College of Cardiology/ American Heart Association definitions, 9 70 patients had moderate to severe valvular disease and were excluded from analysis. No patients were excluded for uninterpretable images. Successful follow-up was completed in 97.6% of these patients, leaving a final study population of 908 patients who were followed up for a median of 948 days (25th, 75th percentiles: 639, 1263). CMR Image Acquisition After informed consent, adenosine was infused via a peripheral vein for at least 2 minutes and 45 seconds and was discontinued only after left ventricular myocardial contrast arrival was visualized. The stress gadoversetamide dosage used was 0.1 mmol/kg given at 3 cc/s. Adenosine infusion was discontinued when left ventricular myocardial contrast arrival was noted. An ECG-triggered, interleaved, notched saturation gradient echo sequence was performed with an 8-channel cardiac receiver coil, array spatial and sensitivity technique (order 2) on a 1.5T magnet (Twinspeed Zoom Mode, General Electric Medical Systems, Milwaukee, WI). Subsequently, 5 to 10 short-axis left ventricular slices (dependent on heart rate) were obtained over every other cardiac cycle during the first pass of contrast in the left ventricle. In all, 40 stress perfusion images were acquired for every slice over 80 cardiac cycles. Immediately after stress first-pass perfusion imaging, a second dose of gadoversetamide 0.1 mmol/kg was given (total 0.2 mmol/kg). Late gadolinium enhancement imaging was begun 10 minutes later. Myocardial delayed enhancement (DE) imaging was performed using a segmented k-space inversion recovery sequence with optimized inversion time. All patients underwent cine functional imaging using a steady state free-precession sequence acquiring a short-axis stack through the entire heart with typical spatial resolution of at least 2.0 2.6 8.0 mm using retrospective ECG gating. Breath-hold segmented k-space velocity-encoded phase contrast imaging was performed perpendicular to the proximal ascending aorta using retrospective ECG gating and phantom correction of background phase offsets. CMR Image Analysis and Interpretation Image postprocessing and analysis were performed on a dedicated workstation using ReportCard version 3.7 (General Electric Medical Systems, Milwaukee, WI). Left ventricular end-diastolic and endsystolic frames were identified in the short-axis stack. Left ventricular contours were drawn to include the papillary muscles as part of the blood pool. The extent of myocardial late enhancement was expressed as quintiles of transmurality (normal 0, 1% to 25%, 26% to 50%, 51% to 75%, and 76% to 100%) for each of the 17 left ventricular segments by visual analysis. Stress perfusion defect was defined by visual analysis. Perfusion defects were considered abnormal if all of the following were observed: (1) The defect was subendocardial or transmural, (2) was present in at least 2 contiguous short-axis slices, (3) persisted at least 10 seconds after first myocardial contrast arrival, and (4) conformed to a coronary artery territory. Isolated epicardial or papillary muscle defects and susceptibility artifacts were considered nonischemic. Rest perfusion imaging was not performed. Stress perfusion defect was expressed in 3 categories (normal, subendocardial, or transmural) for each of the 16 left ventricular short-axis segments (perfusion imaging was not performed in the long axis). A composite CMR end point combining information from left ventricular function (left ventricular ejection fraction [LVEF] normal versus abnormal), DE (present versus absent), and stress perfusion (normal versus abnormal) was set equal to the number of these elements abnormal (minimum: 0, maximum: 3). Aortic blood flow was measured by phase-contrast velocity encoding perpendicular to the proximal aorta above the coronary artery ostia. Phantom correction for background velocity offsets was performed in all cases. Offline ventricular volume measurements were made by the Simpson method. All measurements conform to accepted cardiac imaging standards. 10 LVEF 50% was considered abnormal. Patient Follow-Up and Outcome Measures All patients were followed up for a minimum of 1 year after their index CMR study (mean follow-up 2.6 1.2 years). Outcomes were identified by a review of hospital and outpatient electronic medical records, a mail questionnaire completed by study subjects, and a prescripted telephone interview with each study subject. The primary end point of this study was the occurrence of cardiac events, defined as the aggregate of all-cause death, nonfatal myocardial infarction, or late ( 90 days postindex test) revascularization. To better understand the ability of normal CMR results to identify low risk, secondary end points were used solely for analyses of normal CMR results. These included cardiac death and hard events (cardiac death or myocardial infarction [MI]). In patients with multiple events, only the first event was considered for survival analysis. Event rates were annualized by dividing the observed events by the mean duration of follow-up for each group. Valve surgery without revascularization was not considered an event. Statistical Analysis Baseline patient characteristics were described in terms of median (25 th and 75th percentiles) for continuous variables and frequencies for categorical variables. Categorical variables were compared using a 2 test. A Bonferroni correction was used to correct for multiple comparisons. A 2-tailed value of P 0.05 was considered statistically significant. Multivariable Modeling and Survival Analysis Cox proportional hazards modeling was used to assess the association of CMR information and survival time free of cardiac events after adjusting for other potential confounders. 11 13 Because both the inclusion and exclusion (censoring) of patients undergoing early revascularization introduces biases that may distort our results, all survival models were stratified by the use of early revascularization. 12 The association of referral to early revascularization and clinical, historical, and testing data were modeled using multiple logistic regression. 12,13 Covariate selection for model entry was based on clinical experience and identification of known confounders. The models were carefully examined, when applicable, for proportional hazards assumption, multicollinearity, and the additive value of the terms. 12,13

Bingham and Hachamovitch Prognostic Significance of 4-Component CMR 1511 Downloaded from http://circ.ahajournals.org/ by guest on April 4, 2017 A ratio of 10 events per degree of freedom of the model was maintained. The event-free survival stratified by key CMR parameters was plotted using the Kaplan-Meier method, and was compared using log-rank tests. The S-PLUS 2000 (Release 2) software package (Insightful Corp, Seattle, WA) with supplemental libraries (Hmisc, Design) was used for all analyses. Determination of Incremental Value Incremental prognostic value was defined as a significant increase in the Wald statistic and likelihood ratio tests after the addition of imaging data to an optimized Cox model of preimaging data alone. We performed 2 sets of analyses. First, on the basis of pre-cmr data, we developed a model predicting the study end point. Second, we added the results of pre-gadolinium CMR data (LV volumes, LVEF, valvular data) to the first model (because these data elements are often known before testing and, if not, could be defined without proceeding with additional testing). Third, we added DE data to the model from the previous, second step. In a parallel third step to this, we evaluated the added value of stress perfusion data over the second model as well (because DE and stress perfusion could either be obtained independently of the other). Finally, we considered information from all CMR data combined. Incremental prognostic value was considered to be present for any of these steps when a significant increase in Wald global 2 was found. On the basis of these Cox proportional hazards models, risk-adjusted survival curves were developed. Regression models were validated by means of resampling (100 iterations), and biascorrected C indices were determined. Model calibration was assessed by constructing a bootstrap-corrected calibration curve by resampling the differences between average Cox proportional hazards predicted survival and Kaplan-Meier (observed) estimates at a single time point. Reclassification The impact of CMR results on reclassification of patient risk was determined using net reclassification improvement (NRI). 14 For each patient, the predicted risk of an adverse event was determined on the basis of each model, and we then assessed the impact on patient reclassification of the addition of information to these models. Reclassification to a higher-risk group was considered upward movement in classification, whereas reclassification to a lower-risk group was considered downward movement. Improvement in reclassification was estimated by taking the sum of differences in proportions of individuals reclassified upward minus the proportion reclassified downward for people who developed events and the proportion of individuals moving downward minus the proportion moving upward for those who did not develop events. NRI represents the net number of patients with improved reclassification, summing reclassification in patients with events and patients without events. Because there are no accepted thresholds defining low versus intermediate-to-high risk for total events, we examined reclassification using thresholds of 1.5% and 2.0% total events per year to define low-, intermediate-, and high-risk groups. Results Patient Characteristics and CMR Findings Patients in this study tended to be middle-aged and predominantly male, with the majority having known hypertension, a smaller number with a family history of premature CAD, and about a quarter with diabetes mellitus (Table 1). About half of patients presented with dyspnea and /or angina, but of the latter, 10% had typical angina at the time of presentation. Relatively small numbers of patients had significant comorbidities, and about half had a history of prior CAD. Of the latter, about one third had prior MI or percutaneous coronary intervention (PCI), and smaller numbers had coronary artery Table 1. Patient Demographics Follow-up time, d 948 (639, 1263) Demographics/cardiac risk factors Age 65 (55, 74) Male, % 58.6 Hypertension, % 63.8 Diabetes mellitus, % 25.4 Family history of premature CAD, % 41.7 Smoking, % 6.1 Framingham risk score 7.0 (5.0, 10.0) Symptoms, % Typical angina 9.0 Any angina 56.2 Dyspnea 51.7 Comorbidities, % Renal disease 7.7 COPD 10.1 PVD 6.2 CV history, % Prior CAD 49.4 Prior MI 34.9 Prior PCI 32.7 Prior CABG 15.0 Stress test Resting heart rate 66.0 (30.0, 74.0) Peak heart rate 77.0 (67.0, 88.0) LV function LVEDV 124.0 (102.0, 149.0) LVESV 45.0 (33.0, 65.0) LVEF 63.0 (56.0, 69.0) LVSV 74.0 (63.0, 87.0) Aortic flow 70.0 (60.0, 84.0) Stress CMR study Abnormal wall motion, % 44.2 Abnormal perfusion, % 32.8 Abnormal DE, % 37.6 Summed wall motion score 0.0 (0.0, 17.0) Summed perfusion score 0.0 (0.0, 2.0) Summed rest score 0.0 (0.0, 5.0) Continuous variables represented as median (25 th and 75th percentiles). CAD indicates coronary artery disease; COPD, chronic obstructive pulmonary disease; PVD, pulmonary vascular disease; CV, cardiovascular disease; MI, myocardial infarction; PCI, percutaneous coronary intervention; CABG, coronary artery bypass grafting; LV, left ventricular; LVEDV, LV end-diastolic volume; LVESV, LV end-systolic volume; LVEF, LV ejection fraction; LVSV, LV stroke volume; and DE, delayed enhancement. bypass grafting (CABG). About one third of patients had abnormal stress perfusion or delayed enhancement, and 44% had abnormal wall motion. Additionally, 276 patients had subendocardial segmental defects whereas 53 had transmural segmental defects. Transmural defects occurred more frequently among patients with versus without prior MI (12.3% versus 2.4%, respectively; P 0.001).

1512 Circulation April 12, 2011 Downloaded from http://circ.ahajournals.org/ by guest on April 4, 2017 40% 35% 30% 25% 20% 15% 10% 5% 0% * 3.5% 11.2% * * No DE Patient Outcomes During the follow-up period, we identified 101 cardiac events (9.8%; death, MI, or late revascularization). These included 30 cardiac deaths, 36 noncardiac deaths, 5 nonfatal MIs, and 19 late revascularizations. These late revascularizations included 13 late PCI and 9 late CABG (3 patients underwent both PCI and CABG on follow-up, and were thus counted only once for late revascularization). In addition, 103 patients underwent revascularization (91 PCI, 13 CABG) early ( 90 days) after the index CMR study. CMR Results and Referral to Early Revascularization Referral to early revascularization after CMR increased as a function of both DE and stress perfusion results (Figure 2). Although revascularization rates were greater with abnormal compared to normal DE (20.5% versus 5.8%), they were also greater after abnormal compared with normal stress perfusion both in the setting of normal and abnormal DE. Further, revascularization rates were also greater with abnormal than with normal DE in the setting of normal perfusion (11.2% versus 3.5%). Logistic regression modeling of early revascularization ( 2 102, P 0.001) revealed that abnormal perfusion and DE were the most closely associated with early revascularization. With respect to the temporal distribution of early revascularizations, 87% occurred in the first 30 days, with 4.1 and 2.0% occurring between 30 to 60 days and 60 to 90 days, respectively. With respect to later revascularizations, 6.7% of them occurred between 90 days and 6 months, and the observed frequency of revascularizations increased after this time point, with 28%, 37%, and 28% occurring between 6 months and 1 year, between 1 and 2 years, and 2 years after testing. Univariable Predictors of Adverse Events Of baseline patient data, the variables most associated with cardiac events included patient age, hypertension, dyspnea, diabetes mellitus, Framingham risk score, and renal disease. Interestingly, prior CAD, PCI, or CABG were not predictors. Rest heart rate was a significant predictor of outcomes (Tables 2 and 3). DE 27.3% 24.3% 5.8% * 20.5% 512 98 55 243 567 341 Normal Perfusion Abnormal Perfusion All Figure 2. Frequency of referral to early revascularization after cardiac magnetic resonance imaging (CMR) with versus without delayed enhancement (DE) (white, black bars), with versus without stress perfusion abnormalities (left, center pair of bars), and overall (all). *P 0.01. N for each subgroup is underneath the bar. Table 2. Univariable Prediction of Adverse Events: Preimaging Data and CPH-Based HRs (95% CIs) Cardiac Events, HR (95% CI) Demographics/cardiac risk factors Age 1.05 (1.03, 1.07)* Man 1.10 (0.72, 1.62) Hypertension 2.62 (1.51, 4.57) Diabetes mellitus 1.72 (1.12, 2.65) Family history of CAD 0.70 (0.56, 1.18) Smoking 0.48 (0.45, 1.09) Framingham risk score 1.07 (1.01, 1.13) Symptoms Angina 1.01 (0.89, 1.15) Dyspnea 1.95 (1.27, 3.00) Comorbidities Renal disease 2.31 (1.25, 4.25) COPD 1.00 (0.61, 2.12) PVD 1.51 (0.46, 2.17) CV history Prior CAD 1.19 (0.78, 1.81) Prior MI 1.05 (0.69, 1.61) Prior PCI 1.31 (0.85, 2.00) Prior CABG 1.02 (0.55, 1.87) Stress test Resting heart rate 1.02 (1.00, 1.03) Peak heart rate 1.00 (0.98, 1.01) *Per year of age; per bpm. CPH indicates Cox proportional hazards model; HR, hazard ratio; CI, confidence interval; CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease; PVD, pulmonary vascular disease; CV, cardiovascular disease; MI, myocardial infarction; PCI, percutaneous coronary intervention; and CABG, coronary artery bypass grafting. CMR-defined LVEF and LV stroke volume were predictive of cardiac events. CMR-determined aortic flow was also predictive of cardiac events, as were abnormal wall motion, perfusion, and delayed enhancement, as well as summed scores of these variables. Event Rates as a Function of CMR Results Normal CMR results were associated with low risk irrespective of which component was examined (Figure 3). With respect to cardiac death rates, event rates were very low (0.6% to 0.7%), with the exception of postnormal EF, which was slightly higher (1.0%). With respect to hard events, (cardiac death or MI), event rates were low (0.6% to 0.9%) again, with the exception of postnormal EF (1.2%). A similar pattern was present with respect to cardiac events. Event rates after abnormal CMR results reveal similar cardiac event rates after abnormal DE, EF, or stress perfusion (5.6% to 6.1%). Event rates were greater in the setting of 2 abnormalities (6.5% to 7.0%). With increasing numbers of CMR components found abnormal, risk of adverse outcomes increased significantly (Figure 4; P 0.001). Interestingly, subendocardial perfusion defects were associated with greater risk than transmural perfusion defects, with the former a

Bingham and Hachamovitch Prognostic Significance of 4-Component CMR 1513 Downloaded from http://circ.ahajournals.org/ by guest on April 4, 2017 Table 3. Univariable Prediction of Adverse Events: CMR Variables and CPH-Based HRs (95% CIs) Events, HR (95% CI) LV function LVEDV 1.00 (0.97, 1.01) LVESV 1.00 (1.00, 1.01) LVEF 0.97 (0.96, 0.99) LVSV 0.98 (0.97, 0.99) Aortic flow 0.98 (0.97, 0.99) Stress CMR study Abnormal wall motion 2.03 (1.33, 3.01) Abnormal perfusion 2.00 (1.31, 3.04) Abnormal DE 2.17 (1.42, 3.30) Summed wall motion score 1.02 (1.01, 1.04) Summed perfusion score 1.04 (1.01, 1.08) Summed rest score 1.03 (1.01, 1.06) CMR indicates cardiac magnetic resonance imaging; CPH, Cox proportional hazards model; HR, hazard ratio; LV, left ventricular; LVEDV, LV end-diastolic volume; LVESV, LV end-systolic volume; LVEF, LV ejection fraction; LVSV, LV stroke volume; and DE, delayed enhancement. significant univariable predictor of risk, whereas the latter was not (hazard ratio [HR]: 1.14 [1.04 to 1.25], P 0.0036 versus 1.07 [0.86 to 1.34], P 0.5200) (HR expressed per number of subendocardial or transmural defects present). Of note, there were relatively limited numbers of patients with transmural defects. Multivariable Modeling and Survival Analysis We initially examined the association of pre-cmr data with cardiac events using a Cox proportional hazards analysis, thus identifying a pre-cmr chronic pulmonary hypertension model (model 1). The second model consisted of fixing these covariates and adding LVEF. This resulted in the addition of LVEF into the model with an increase in 2 (global 2 41.7 to 51.5; C index 0.695 to 0.717). Fixing the pre-cmr and LVEF data in the model and adding aortic flow data resulted in an additional increase in global 2 (57.5; C index 0.722; model 3), which was further increased with the addition of Figure 3. Frequency per year of cardiac death or myocardial infarction (MI), cardiac death, and cardiac event rate after normal stress perfusion (n 610), normal delayed enhancement (DE) (n 567), normal ejection fraction (EF) (n 770), or all 3 normal (n 488). ts (%) tal Event Tot 25% 20% 15% 10% 5% 0% 5.8% 9.9% 13.6% 21.0% 470 131 125 79 None One Two Three Number of Abnormal CMR Measures Figure 4. Frequency of total events as a function of the number of cardiac magnetic resonance imaging (CMR) test components abnormal (of stress perfusion, delayed enhancement [DE], and left ventricular ejection fraction [LVEF]). P 0.001 across categories. N for each subgroup is underneath the bar. DE information ( 2 63.5; C index 0.747; model 4). Finally, the addition of stress perfusion data to pre-cmr data, LVEF, and aortic flow added further value (global 2 62.5; C index 0.740; model 5) (Table 4). Because of potential bias introduced by patients referred to revascularization early after testing, we also repeated models 4 and 5 using early revascularization as a time-dependent covariate. In model 4, time to early revascularization but not early revascularization or the interaction between these covariates was significant (P 0.0002, P 0.2995, and 0.122, respectively). Similar results were present for model 5 using early revascularization as a time-dependent covariate (early revascularization, time to early revascularization, and the interaction between these covariates were significant: P 0.3084, P 0.0003, and P 0.1462, respectively). The final modeling step (addition of both DE and stress perfusion data) resulted in a further increase in value ( 2 69.1; C index 0.727), but neither of these variables were significant at a P 0.05 level. We believe that this was due to the presence of significant collinearity between the stress perfusion and DE covariates (correlation r 0.66, P 0.01). This was reflected by both DE and stress perfusion covariates being nonsignificant when both were included in the Cox model. To overcome this, we used the derived composite CMR end point previously defined. After adjusting for the pre-cmr data, the addition of composite CMR end point resulted in a significant increase in 2 (41.7 to 54.1, P 0.0001, C index 0.734). Risk-adjusted survival curves reveal significant stratification by the number of CMR components abnormal (Figure 5). In the absence of abnormalities, risk of cardiac events is relatively low; with all 3 components abnormal, adjusted risk is 30% at about 4 years. The final modeling step was also performed using extended Cox proportional hazards models, with early revascularization handled using a time-dependent covariable. This approach resulted in a marked decrease of the of delayed late enhancement from 0.7419 to 0.6275 ( 15.5% decrease; HR 2.10 to 1.87, Table 4, model 4). With respect to the model including stress perfusion (Table 4, model 5, last column), no meaningful change in the for stress perfusion occurred (0.5653 to 0.5437, 4% decrease, HR 1.76 to 1.72).

1514 Circulation April 12, 2011 Table 4. Cox Proportional Hazards Model Predicting Cardiac Events Model 1, Clinical and Historical* Model 2, LV Function Model 3, Aortic Flow Model 4, DE Model 5, Stress Perfusion Downloaded from http://circ.ahajournals.org/ by guest on April 4, 2017 HR 95% CI P HR 95% CI P HR 95% CI P HR 95% CI P HR 95% CI P Age 1.59 1.30 1.93 0.0001 1.61 1.32 1.95 0.0001 1.51 1.24 1.84 0.0001 1.50 1.23 1.83 0.0001 1.49 1.22 1.83 0.0001 Dyspnea 1.70 1.10 2.62 0.0167 1.55 1.01 2.41 0.049 1.68 1.08 2.63 0.0219 1.70 1.09 2.65 0.0199 1.72 1.10 2.69 0.0173 Prior CAD 1.11 0.72 1.70 0.6421 0.98 0.63 1.52 0.9336 1.05 0.67 1.63 0.8325 0.75 0.45 1.24 0.2574 0.89 0.56 1.42 0.6273 Rest heart rate 1.17 1.01 1.37 0.0421 1.14 0.99 1.32 0.0769 1.07 0.91 1.25 0.4312 1.10 0.93 1.29 0.2653 1.10 0.93 1.30 0.2495 Renal disease 1.74 0.93 3.25 0.0822 1.68 0.90 3.14 0.1024 1.70 0.90 3.18 0.0998 1.65 0.88 3.10 0.1203 1.75 0.93 3.29 0.0807 Diabetes mellitus 1.41 0.91 2.19 0.1212 1.52 0.98 2.36 0.0624 1.62 1.04 2.51 0.0335 1.59 1.02 2.47 0.0394 1.55 0.99 2.42 0.0532 LVEF NA NA NA 0.78 0.67 0.92 0.0028 0.78 0.67 0.92 0.0026 0.86 0.72 1.03 0.0942 0.84 0.71 1.00 0.0504 Aortic flow NA NA NA NA NA NA 0.88 0.77 1.00 0.0300 0.88 0.77 1.00 0.0548 0.88 0.77 1.00 0.0541 DE presence NA NA NA NA NA NA NA NA NA 2.10 1.22 3.62 0.0075 NA NA NA Abnormal stress perfusion NA NA NA NA NA NA NA NA NA NA NA NA 1.76 1.08 2.87 0.0243 *Global 2 41.7, C index 0.695, P 0.00001; Global 2 51.5, C index 0.717, P 0.00001; Global 2 57.5, C index 0.722, P 0.00001; Global 2 63.5, C index 0.747, P 0.00001; Global 2 62.5, C index 0.736, P 0.00001. Hazard ratio expressed per 10 years of age, per 10 bpm of rest heart rate, per 10% LVEF, and per 10 cc/min aortic flow. HR indicates hazard ratio; CI, confidence interval; DE, delayed enhancement; CAD, coronary artery disease; and LVEF, left ventricular ejection fraction. Finally, graphic presentation of model calibration (observed versus predicted survival) is shown in Figure 6. Predicted event rates appear to approximate observed event rates through most of the range of values. Risk Reclassification Reclassification of risk (classification groups were set at 0 to 1.5%, 1.5% to 2.0%, and 2.0%) was assessed separately for the addition of DE and stress perfusion data to pre-de and prestress perfusion data (model 4 to model 3 and model 5 to model 3, respectively; online-only Data Supplement Table I). The addition of DE data to the pre-de model resulted in 2 correct (up) reclassifications and 4 incorrect (down) reclassifications in patients who experienced events (NRI events, 0.022). Additionally, 18 incorrect (up) reclassifications and 109 correct (down) reclassifications occurred in patients who did not have events (NRI nonevents, 0.111). Overall, 8.9% of patients were correctly reclassified by the addition of the delayed enhancement information (NRI net, 0.089). The addition of stress perfusion data to the model of preimaging data resulted in no correct (up) reclassifications and 2 incorrect (down) reclassifications in patients who experienced events (NRI events, 0.023). Additionally, 22 incorrect (up) reclassifications and 71 correct (down) reclassifications occurred in patients who did not have events (NRI nonevents, 0.058). Overall, 3.5% of patients were correctly reclassified by the addition of the stress perfusion information (NRI net, 0.035). Discussion In the current study, we found that in a series of 908 patients with short- to intermediate-term follow up after index 4 component CMR (assessment of stress perfusion, viability, cardiac flow, and LV structure and function), the CMR results added incremental prognostic information over pre- CMR data and were incremental over each other as well. Hence, they provided complementary rather than overlapping information. After normal combined (4 component) CMR, patients had a low risk of future adverse events and cardiac death ( 0.4% annual cardiac mortality rate). Conversely, even with a single CMR-defined abnormality (eg, low aortic Survival Proba ability 8 0.9 1.0 0.7 0. 0.6 None One or Two All Three Figure 5. Risk-adjusted survival free of total events in patients as a function of the number of cardiac magnetic resonance imaging (CMR) test components abnormal (zero, 1 to 2, or 3 of stress perfusion, delayed enhancement [DE], and left ventricular ejection fraction [LVEF]). P 0.001 across categories as per Cox proportional hazards model results. N remaining at each time point is shown above x axis. 0.5 0 250 500 750 1000 1250 1500 Time (Days)

Bingham and Hachamovitch Prognostic Significance of 4-Component CMR 1515 A Fractio n Surviving 3 365 Days 0.90 0.92 0.88 0 0.98 1.00 2 0.94 0.96 0.95 0.96 0.97 0.98 0.99 Predicted 365 Day Survival Downloaded from http://circ.ahajournals.org/ by guest on April 4, 2017 B 365 Days n Surviving 3 Fractio C 0.98 1.00 0.96 2 0.94 0.90 0.92 0.88 0 0.98 1.00 0.95 0.96 0.97 0.98 0.99 Predicted 365 Day Survival Figure 6. Model calibration. A, Model 4 (baseline data, left ventricular ejection fraction [LVEF], aortic flow, and delayed enhancement [DE]); B, model 5 (baseline data, LVEF, aortic flow, and stress perfusion); and C, pre-cmr (cardiac magnetic resonance imaging) data and the derived composite CMR end point. Predicted 365-day survival (x axis) versus observed 365-day survival (y axis). Open circles represent apparent calibration; x symbols, bias-corrected calibration. Fractio n Surviving 3 365 Days 0.90 0.92 0.88 0 2 0.94 0.96 0.95 0.96 0.97 0.98 0.99 Predicted 365 Day Survival

1516 Circulation April 12, 2011 Downloaded from http://circ.ahajournals.org/ by guest on April 4, 2017 flow or increased LV end-systolic volume index), patients were found to be at increased risk. Further, additional abnormal findings identified by CMR incrementally increased risk, thus enhancing stratification. On the basis of results of Cox proportional hazards modeling, after adjusting for pre-cmr data, all components of the stress CMR (stress perfusion, DE, and LV function) were incremental over the pre-cmr data as well as over each other. When modeling total events, we found CMR-derived measures of LV volume (LV end-systolic volume index), and hemodynamics (aortic flow) added incremental prognostic value over pre-cmr data, and both stress perfusion and DE added further information over the combination of pre-cmr and pre stress test CMR data. Comparison With Previous Studies The current study extends the results of a number of previous CMR prognostic studies in stable patients with known or suspected CAD. Jahnke et al 5 followed up 302 patients after normal stress CMR perfusion and reported a low event rate comparable to that previously reported with nuclear and echocardiographic techniques. 4 Subsequently, Pilz et al 6 extended these results in a study of 218 patients with normal CMR, reporting similar results. However, not all studies are concordant on this point. 7 Although several studies reported outcomes in patients with normal versus abnormal stress perfusion CMR, with respect to incremental prognostic value, Steel et al 7 recently reported the additive prognostic value of DE and stress perfusion data over pre-cmr data and each other. The results of our study are consistent with previous reports indicating that the presence and extent of myocardial delayed enhancement provides incremental prognostic value for adverse events. 7,15 18 The current study extends the findings of previous articles by also examining the third component of the stress examination, resting LV function. In the current study, a number of these CMR parameters added incrementally to each other. Additionally, to the best of our knowledge, the current study is the largest prognostic stress CMR study reported to date. The presence of abnormalities in stress perfusion, DE, LVEF, and aortic flow each conferred incremental prognostic power to predict future adverse cardiovascular events, and were additive, not redundant, in predicting risk. Similar to nuclear and echocardiographic techniques, stress CMR provides estimates of ischemic burden, myocardial scar/viability, and LV volumes and function. The current study confirms the results of previous studies using these other modalities, that these parameters are prognostically complementary rather than redundant. 4,19 21 Cardiac magnetic resonance imaging is unique, however, in encompassing the advantages of both of these more widely used techniques. Like echocardiography, CMR can provide assessments of hemodynamics, ventricular structure and function, and vascular and valvular structure. Moreover, CMR, like nuclear techniques, can visualize perfusion abnormalities and discern areas of scar, ischemia, and normal myocardium. Finally, because of its greater temporal and spatial resolution, CMR may have advantages over these other modalities, but its comparative cost effectiveness is uncertain. Which Is the Most Predictive CMR Variable? Historically, the 2 associated with individual covariates in the Cox model has been put forward as the metric to compare covariate value. However, because this represents a measure of association, not of effect, the HR would serve as a clinically more relevant measure (although this is not always the case). Thus, the presence of DE is associated with more than a doubling of risk (Table 4, model 4). LVEF would need to decrease by 20% to be associated with the same level of risk. Thus, DE would appear to be a superior prognosticator than LVEF. Similar results can be found with stress perfusion compared to LVEF. As discussed above, a delayed enhancement stress perfusion comparison is challenging because of collinearity, and additional analyses would be required to sort out this comparison. Risk Reclassification Analysis of reclassification of risk was performed using the relatively new approach of calculating NRI and number of patients correctly reclassified. 14 The results of this analysis revealed a relatively low NRI and a moderate proportion of patients correctly reclassified. These results are limited because of several issues, including relatively short follow-up, the lack of an accepted threshold for total events (the end point used in the current study), as well as the intrinsic limitations of this approach. Post-CMR Patient Management We found that post-cmr revascularization referral is driven predominantly by CMR results. Specifically, the likelihood of referral is largely determined by the presence of ischemia and, to a lesser degree, scar, but not LVEF. In the setting of abnormal ischemia, physicians referred patients to revascularization in similar numbers whether or not DE was present. In the absence of ischemia, a greater revascularization rate was present with versus without DE. The role of ischemia in influencing revascularization decisions is similar to that previously reported with other stress imaging techniques whereas a role for both scar and LVEF has been reported. 4 This close relationship between CMR results and revascularization may have introduced a bias that would obfuscate the relationship between test results and subsequent patient risk because of the influence of revascularization in reducing post-cmr events. This would be consistent with previous publications using other modalities reporting this phenomenon. 12 To overcome this potential bias, rather than censoring patients referred to early revascularization, we retained them in our analyses but used them as a stratification factor in the survival modeling. However, to further examine this issue, we also evaluated extended Cox proportional hazards models using early revascularization as a time-dependent covariate. Finally, with respect to the temporal distribution of post- CMR revascularization, we found that revascularizations aggregated very early ( 30 days) after testing or 6 months after testing.

Bingham and Hachamovitch Prognostic Significance of 4-Component CMR 1517 Downloaded from http://circ.ahajournals.org/ by guest on April 4, 2017 Study Limitations The current study has all the inherent limitations associated with nonrandomized, single-site, observational studies. Although the impact of selection biases, spurious observations, and missing data cannot be ignored, patients in observational studies better represent those seen in practice compared with those in randomized trials. The current study was designed as an effectiveness study in that post-cmr patient management was dictated by their physicians clinical practice. A single reader (S.E.B.) interpreted all CMR studies, and no consensus interpretation was performed. However, our group has previously shown that intraobserver variability was minimal. 22 Finally, it would be difficult to state to what extent the current results are generalizable to studies using dobutamine stress CMR, an alternative form of stress CMR. Several statistical issues are also relevant. First, survival analyses indicated that the added value of CMR variables often did not manifest a dramatic increase in global Wald 2. It is important to note that, despite this, the risk-adjusted HR associated with the CMR covariates in this setting were quite large. For example, the addition of DE information to the survival model increased the 2 from 57.5 to 63.5. Although this is not a dramatic gain, the HR associated with DE (2.10) indicates a doubling of risk in the presence of this CMR finding, even after adjusting for other factors. Thus, although the change in 2 was not large, the clinical impact was significant. In order to avoid overfitting our models, total cardiac events was the primary analytic end point of the second analysis. A larger study population or a higher incidence of adverse events would be required to overcome this limitation. Importantly, when using a composite end point, the interpretation of the results should be made in the context of what type of event was predominant in the study. 12 Conclusions Consideration of all components of stress perfusion CMR enhances posttest risk stratification. These parameters are incremental to pre-cmr data and to each other. Further studies to define how these parameters may be best applied will be needed. Acknowledgments The authors thank Benjamin Nutter, MSc (Department of Quantitative Health Sciences, Cleveland Clinic, Cleveland, OH), for his professional statistical support. Source of Funding Funding for this study came from the Central Utah Imaging Research Fund. Disclosures Dr Bingham served as consultant to the advisory board of General Electric Healthcare. Dr Hachamovitch serves as a consultant for Lantheus Medical Imaging. References 1. Winter A, Ray N. Paying accurately for imaging services in Medicare. Health Aff (Millwood). 2008;27:1479 1490. 2. Bodi B, Sanchis J, Lopez-Lereu MP, Nunez J, Mainar J, Monmeneu JV, Husser O, Dominguez E, Chorro FJ, Llacer A. Prognostic value of dipyridamole stress cardiovascular magnetic resonance imaging in patients with known or suspected coronary artery disease. J Am Coll Cardiol. 2007;50:1174 1179. 3. Dall Armellina E, Morgan TM, Mandapaka S, Ntim W, Carr JJ, Hamilton CA, Hoyle J, Clark H, Clark P, Link KM, Case D, Hundley WG. Prediction of cardiac events in patients with reduced left ventricular ejection fraction with dobutamine cardiovascular magnetic resonance assessment of wall motion score index. J Am Coll Cardiol. 2008;52: 279 286. 4. Hachamovitch R, Beller GA. Critical review of imaging approaches for diagnosis and prognosis of CAD. In: Di Carli MF, Kwong R, eds. Novel Techniques for Imaging the Heart: Cardiac MR and CT. Oxford, UK: Blackwell Publishing; 2008. 5. Jahnke C, Nagel E, Gebker R, Kokocinski T, Kelle S, Manka R, Fleck E, Paetsch I. Prognostic value of cardiac magnetic resonance stress tests: adenosine stress perfusion and dobutamine stress wall motion imaging. Circulation. 2007;115:1769 1776. 6. Pilz G, Jeske A, Klos M, Ali E, Hoefling B, Scheck R, Bernhardt P. Prognostic value of normal adenosine-stress cardiac magnetic resonance imaging. Am J Cardiol. 2008;101:1408 1412. 7. Steel K, Broderick R, Gandla V, Larose E, Resnic F, Jerosch-Herold M, Brown KA, Kwong RY. Complementary prognostic values of stress myocardial perfusion and late gadolinium enhancement imaging by cardiac magnetic resonance in patients with lnown or suspected coronary artery disease. Circulation. 2009;120:1390 1400. 8. Poon M, Fuster V, Fayad Z. Cardiac magnetic resonance imaging: a one-stop-shop evaluation of myocardial dysfunction. Curr Opin Cardiol. 2002;17:663 670. 9. Bonow RO, Carabello BA, Chatterjee K, de Leon ACJ, Faxon DP, Freed MD, Gaasch WH, Lytle BW, Nishimura RA, O Gara PT, O Rourke RA, Otto CM, Shah PM, Shanewise JS. Focused update incorporated into the ACC/AHA 2006 Guidelines for the Management of Patients With Valvular Heart Disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation. 2008. 2008;118:e523 e661. 10. Cerqueira MD, Weissman NJ, Dilsizian V, Jacobs AK, Kaul S, Laskey WK, Pennell DJ, Rumberger JA, Ryan T, Verani MS, American Heart Association Writing Group on Myocardial Segmentation and Registration for Cardiac Imaging. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart: a statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. Circulation. 2002;105:539 542. 11. Cox D. Regression models and life tables (with discussion). J Roy Statist Soc Ser B Metho. 1972;34:187 220. 12. Hachamovitch R, Di Carli MF. Methods and limitations of assessing new noninvasive tests: part II: outcomes-based validation and reliability assessment of noninvasive testing. Circulation. 2008;117:2793 2801. 13. Harrell FJ. Regression Modeling Strategies. New York, NY: Springer- Verlag; 2001. 14. Pencina MJ, D Agostino RB Sr, D Agostino RB Jr, Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med. 2008;27:157 172. 15. Kwon DH, Halley CM, Carrigan TP, Zysek V, Popovic ZB, Setser R, Schoenhagen P, Starling RC, Flamm SD, Desai MY. Extent of left ventricular scar predicts outcomes in ischemic cardiomyopathy patients with significantly reduced systolic function: a delayed hyperenhancement cardiac magnetic resonance study. J Am Coll Cardiol Cardiovasc Imaging. 2009;2:34 44. 16. Kwong RY, Chan AK, Brown KA, Chan CW, Reynolds HG, Tsang S, Davis RB. Impact of unrecognized scar detected by cardiac magnetic resonance imaging on event-free survival in patients presenting with signs or symptoms of coronary artery disease. Circulation. 2006;113: 2733 2743. 17. Yan AT, Shayne AJ, Brown KA, Gupta SN, Chan CW, Luu TM, Di Carli MF, Reynolds HG, Stevenson WG, Kwong RY. Characterization of the peri-infarct zone by contrast-enhanced cardiac magnetic resonance imaging is a powerful predictor of post myocardial infarction mortality. Circulation. 2006;114:32 39. 18. Yokota H, Heidary S, Katikireddy CK, Nguyen P, Pauly JM, McConnell MV, Yang PC. Quantitative characterization of myocardial infarction by cardiovascular magnetic resonance predicts future cardiovascular events in patients with ischemic cardiomyopathy. J Cardiovasc Magn Reson. 2008;10:17 23. 19. Hachamovitch R, Hayes SW, Friedman JD, Cohen I, Berman DS. Comparison of the short-term survival benefit associated with revasculariza-

1518 Circulation April 12, 2011 tion compared with medical therapy in patients with no prior coronary artery disease undergoing stress myocardial perfusion single photon emission computed tomography. Circulation. 2003;107:2900 2907. 20. Hachamovitch R, Hayes S, Friedman JD, Cohen I, Berman DS. Relative role of inducible ischemia versus ejection fraction in the prediction of survival benefit with revascularization compared to medical therapy in patients with no prior revascularization undergoing stress myocardial perfusion SPECT. J Nucl Cardiol. 2006;13:768 778. 21. Dorbala S, Hachamovitch R, Curillova Z, Thomas D, Vangala D, Kwong RY, Di Carli MF. Incremental prognostic value of gated Rb-82 positron emission tomography myocardial perfusion imaging over clinical variables and rest LVEF. J Am Coll Cardiol Cardiovasc Imaging. 2009;2:846 854. 22. Bingham SE, Johnson C, Mitchell S, Smith N, Smith D. Adenosine stress myocardial perfusion magnetic resonance imaging: initial experience in a community imaging center. J Cardiovasc Magn Reson. 2004;6:128. Abstract. Downloaded from http://circ.ahajournals.org/ by guest on April 4, 2017 CLINICAL PERSPECTIVE Cardiac magnetic resonance imaging (CMR) is unique in its ability to provide assessment of hemodynamics, ventricular structure and function, valvular structure and function, myocardial viability, and stress perfusion. Various CMR-derived data are known to provide important prognostic information. However, the relative incremental prognostic value of 4 CMR components (vasodilator stress perfusion, myocardial delayed enhancement, aortic blood flow, and left ventricular volumes) is unclear. We followed up 908 patients for a mean of 2.6 1.2 years after they underwent combined 4-component CMR for suspicion of coronary stenosis and/or ischemia. In total, 101 cardiac events were observed, including 30 cardiac deaths, 36 noncardiac deaths, 5 nonfatal myocardial infarctions, and 19 late revascularizations. Normal 4-component CMR was followed by a 2.4% annual cardiac event rate ( 0.4% for cardiac death) whereas abnormal CMR had an event rate of 5.6% to 7.0%. The risk of cardiac events increased significantly with increasing number of abnormal CMR components. After we adjusted for pre-cmr data, the addition of left ventricular ejection fraction, aortic flow, delayed enhancement, and stress perfusion data all added incremental prognostic power. CMR components were additive, not redundant, in predicting risk. Future studies may define how each CMR component is best applied in clinical practice.

Incremental Prognostic Significance of Combined Cardiac Magnetic Resonance Imaging, Adenosine Stress Perfusion, Delayed Enhancement, and Left Ventricular Function Over Preimaging Information for the Prediction of Adverse Events Scott E. Bingham and Rory Hachamovitch Downloaded from http://circ.ahajournals.org/ by guest on April 4, 2017 Circulation. 2011;123:1509-1518; originally published online March 28, 2011; doi: 10.1161/CIRCULATIONAHA.109.907659 Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231 Copyright 2011 American Heart Association, Inc. All rights reserved. Print ISSN: 0009-7322. Online ISSN: 1524-4539 The online version of this article, along with updated information and services, is located on the World Wide Web at: http://circ.ahajournals.org/content/123/14/1509 Data Supplement (unedited) at: http://circ.ahajournals.org/content/suppl/2011/03/24/circulationaha.109.907659.dc1 http://circ.ahajournals.org/content/suppl/2016/04/13/circulationaha.109.907659.dc2 Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: http://www.lww.com/reprints Subscriptions: Information about subscribing to Circulation is online at: http://circ.ahajournals.org//subscriptions/

Appendix 1 Cardiovascular Risk Reclassification with the Addition of Delayed Enhancement and of Stress Perfusion Models to a Model of Baseline Patient Characteristics Predicted Risk 0-1.5%- 1.5% 2.0% 1.5% 2.0% >2.0% N 117 0 0 119 9 2 Survival (%) 96.2% NA NA 96.3% 100% 100% Pre- Imaging Data 0-1.5% 1.5%- 2.0% >2.0% Addition of Stress Perfusion Addition of Delayed Enhancement Predicted Risk 0-1.5%- >2.0% N 70 0 20 35 13 11 Survival (%) 98.3% NA 96.0% 98.9% 100% 100% N 43 0 658 7 31 704 Survival (%) 99.1% NA 95.6% 100% 98.7% 95.7% Values represent number of patients in each risk category (0-1.5%, 1.5%-2.0%, >2.0%) as defined by the initial Cox proportional hazards-based pre-imaging survival model and the initial model with imaging data added. Survival rates are based on Kaplan-Meier estimates. 1

circulation_2011_02_5tord-imp:circulation 11/18/11 2:20 PM Page 183 Képalkotó eljárások Az adenozinterheléssel, perfúziós vizsgálattal, késői kontraszthalmozással és balkamra-funkciós vizsgálattal kombinált szív-mr-vizsgálatok prognosztikai többletjelentősége a nem kívánt események előrejelzésében a képalkotó eljárások nélkül nyerhető adatokkal összehasonlítva Scott E. Bingham, MD; Rory Hachamovitch, MD, MSc A vizsgálat háttere Bár a szív mágnesesrezonancia- (MR-) vizsgálata széles körű adatokkal szolgálhat a mindennapos gyakorlatban, az adenozinterheléses perfúziós vizsgálat, a késői myocardialis kontraszthalmozás (DE, delayed enhancement), továbbá a bal kamrai térfogat és funkció vizsgálatának relatív prognosztikus többletértéke nem tisztázott. Módszerek és eredmények Kilencszáznyolc beteget követtek 2,6±1,2 éven keresztül, akik kombinált szív-mr-vizsgálaton estek át koszorúér-szűkület és/vagy ischaemia gyanúja miatt. A követési idő alatt összesen 101 cardialis esemény (összhalálozás, myocardialis infarctus vagy késői revascularisatio) következett be. A Cox-féle proporcionális kockázati modell globális χ 2 (χ 2 ) értékének a klinikai adatokkal történt korrekciót, majd a szív-mr-vizsgálati eredmények hozzáadását követő növekedését határozták meg prognosztikai többletértékként. A szív-mr-vizsgálat negatív eredménye esetén az éves eseménygyakoriság 2,4% volt (<1% cardialis halálozás vagy myocardialis infarctus). Ha a szív-mr-vizsgálat kóros eltérést mutatott ki, az eseménygyakoriság évi 5,6%-7,0% volt, a pontos érték az eltérések fajtájától és számától függően változott. A szív-mrvizsgálat előtti adatokkal (életkor, megelőző koszorúér-betegség, nyugalmi szívfrekvencia, vesebetegség és cukorbetegség, χ 2 : 43,6, P<0,0001; C index 0,695) történt korrekciót követően a bal kamrai ejekciós frakció, az aortaáramlás, a késői halmozás és a terheléses perfúziós vizsgálat eredményeinek hozzáadása esetén a χ 2 -érték további növekedése volt megfigyelhető (55,2, 63,3, 68,0 és 68,9; minden esetben P<0,00001; C-index 0,717, 0,722, 0,747 és 0,736). A szív-mr-vizsgálaton észlelt kóros paraméterek száma egyrészt prognosztikus többletértékkel bírt, másrészt hozzájárult a betegek rizikófaktorok szerinti csoportosításához az események bekövetkeztének kockázatát illetően. Következtetések A kamrai térfogat, az aortaáramlás, a myocardiumviabilitás és a terheléses perfúzió MR-vizsgálata prognosztikus többletértékkel bír a nem kívánt események előrejelzésében a képalkotó eljárások nélkül nyerhető adatokhoz képest, és a vizsgálatok kombinálhatóak a prognózis további pontosítása érdekében. A kombinált szív-mr-vizsgálatok negatív eredménye esetén a későbbiekben bekövetkező cardialis események kockázata alacsony. (Eredeti megjelenés: Circulation. 2011;123:1509 1518.) Kulcsszavak: mágneses rezonanciavizsgálat myocardialis ischaemia perfúzió túlélés myocardialis infarctus Az egészségügyben is tapasztalható fokozódó válság miatt megerősödtek a költséges képalkotó módszerek használatának korlátozására irányuló törekvések. 1 A problémát nehezíti, hogy az ismert vagy feltételezett koszorúér-betegségben szenvedő betegek vizsgálatára rendelkezésre álló képalkotó eszközök és módszerek köre gyorsan bővül. Ennek eredményeként születtek az újonnan kifejlesztett technológiák klinikai gyakorlatba történő befogadása és finanszírozása előtti alaposabb értékelésre irányuló törekvések. Ennek részeként szükségessé válhat egy olyan bizonyítékhalmaz létrehozása, amely igazolja alkalmazásukat. A prognózis meghatározása központi szerepet játszik az ismert vagy feltételezett koszorúér-betegségben szenvedő betegek vizsgálata és kezelése során. Számos, a szívizomzat perfúzióját vagy a kamrafalmozgást vizsgáló nem invazív kardiológiai képalkotó módszerről igazolódott, hogy prognosztikus többletértékkel bír a klinikai kockázatbecsléssel összehasonlítva. 2 7 A jelenleg alkalmazott újabb módszerek közül a szív-mrvizsgálat egyedülálló módon részletes diagnosztikai adatokat szolgáltat a szívizom-viabilitásról, a vasodilatator terheléses perfúzióról, a kamrafalmozgásról és a kamrai térfogatokról, a perctérfogatról, illetve a billentyű-rendellenességekről. 8 Bár a terheléses szív-mr-vizsgálati protokollok során gyakran meghatározzák és megadják az összes fenti paramétert, a korábbi tanulmányok során csupán egy vagy két paraméter prognosztikus többletértékét vizsgálták. Nem ismert, hogy a szív-mrvizsgálat során megállapítható összes fenti paraméter kombinációja milyen mértékű relatív prognosztikus többletértékkel bír a jövőbeli cardialis események előrejelzésében. Jelen tanulmány célja, hogy felmérje a szívizom-viabilitás, a vasodilatator ter- Érkezett 2009. szeptember 6-án, elfogadva 2011. január 28-án. Munkahelyi háttér: Central Utah Clinic, Provo, UT (S. E. B.); és Section of Cardiovascular Imaging, Department of Cardiovascular Medicine, Cleveland Clinic, Cleveland, OH (R. H.). A cikkhez tartozó csak online megjelenő melléklet elérhetősége: http://circ.ahajournals.org/cgi/content/full/circulationaha.109.907659/dc1. Levelezési cím: Scott E. Bingham, MD, Director of Imaging, Central Utah Clinic, 1055 N 500 W, Provo, UT 84604. E-mail: sbingham@cut.net 2011 American Heart Association, Inc. 1. évfolyam, 2 3. szám, 2011. június szeptember Circulation Magyar Kiadás 183 192 183

circulation_2011_02_5tord-imp:circulation 11/18/11 2:20 PM Page 184 184 Circulation Magyar Kiadás 2011. június szeptember 1009, klinikailag indikált terheléses perfúziós szív-mr-vizsgálatra utalt beteg 932 beteg 908, elemzésre alkalmas, követett beteg Hét beteg klausztrofóbia miatt nem fejezte be a vizsgálatot 70 beteget szignifikáns billentyûbetegség miatt zártunk ki az elemzésbôl 24 beteg (2,6%) esetében a követés sikertelen volt 1. ábra. A vizsgálati kohorsz származtatása CMR: szív mágneses rezonanciavizsgálat. heléses perfúzió, illetve a kamrafalmozgás és kamrai térfogatértékek prognosztikus többletértékét a beteg klinikai és kórtörténeti adataihoz képest a nem kívánt cardialis események előrejelzésében. Módszerek Jelen tanulmány típusát tekintve hatékonysági vizsgálat, amelyben nem a vizsgálati protokoll határozza meg, hanem a vizsgáló orvosok választják ki a vizsgálatra kerülő betegeket és a szív-mr-t követően alkalmazott kezelés fajtáját. A vizsgált populáció Minden olyan beteget, akit a Central Utah Klinikán (Provo, Utah) 2002. július 7. és 2006. október 31. között terheléses szív-mr-vizsgálatra utaltak, prospektíve besoroltak egy előzetesen felállított és meghatározott adatbázisba. A betegek a vizsgálat elvégzésekor tájékozott beleegyezést adtak, és minden betegadat felvétele és védelme az Amerikai Egyesült Államok egészségbiztosítási törvényének (Health Insurance Portability and Accountability Act) előírásai szerint történt, a kutatást felügyelő vizsgálóbizottság jóváhagyásával. Ezerkilenc olyan beteget azonosítottunk, akit a járóbeteg-ellátás keretein belül terheléses szív-mr-vizs gá - lat ra utaltak, és a következő négy összetevőből álló vizsgálaton esett át: adenozinterheléses perfúziós vizsgálat, a késői myocardialis kontraszthalmozás, a bal kamrai térfogatok és funkció, illetve az aortaáramlás szív-mr-rel történő vizsgálata. Az 1. ábrán látható, hogy hét beteg, akit terheléses szív-mr-vizsgálatra utaltak, megkezdte ugyan, de klausztrofóbia miatt nem fejezte be a vizsgálatot, és így az elemzésből kizárásra került. Hetven beteg az Amerikai Kardiológiai Kol lé gium/ Amerikai Szív Egyesület (American College of Cardiology/American Heart Association) definíciói 9 szerinti középsúlyos vagy súlyos billentyűbetegségben szenvedett, és emiatt kizárásra került az elemzésből. Nem került kizárásra beteg értékelhetetlen felvételek miatt. A követés a betegek 97,6%-ában sikeres volt, így a tanulmány 908 beteg adatait dolgozta fel, akiknek medián követési ideje 948 nap volt (25-ös és 75-ös percentilis: 639 nap és 1263 nap). A szív-mr-képalkotás módja A betegek a tájékozott beleegyezést követően adenozininfúziót kaptak egy perifériás vénán keresztül, legalább 2 perc 45 másodpercig, amelyet csak a kontrasztanyagnak a bal kamrai izomzatban történő megjelenését követően függesztettek fel. A terhelés során alkalmazott gadoverse tamid dózisa 0,1 mmol/kg volt, 3 cm 3 /s sebességgel adagolva. A vizsgálat során egy EKG-trig gerelt, átfedéses, fogazott szaturációs gra diens-echo-szekvenciát készítettek egy nyolccsatornás kardiológiai vevőtekerccsel, array spatial and sensitivity technikát alkalmazva (má so dik elrendezés), egy 1,5 T erősségű mágnes segítségével (Twinspeed Zoom Mode, General Electric Medical Systems, Milwaukee, WI). Ezt követően (a szívritmustól függően) 5-10 rövid tengelyi bal kamrai szeletről készült felvétel minden további szívciklus során, a kontrasztanyagnak a bal kamrán történő első áthaladása alatt. Összesen 40 terheléses perfúziós felvétel készült minden szeletről 80 szívcikluson át. Közvetlenül az első áthaladáskor rögzített ( first pass ) terheléses perfúziós felvételek után ismét gadoversetamid került beadásra 0,1 mmol/ kg dózisban (összesen 0,2 mmol/kg). A késői gadolíniumhalmozásról készült felvételek rögzítése 10 perccel később kezdődött. A késői myocardialis kontraszthalmozás vizsgálata egy optimalizált inverziós idejű szegmentált k-space inversion recovery -szekvencia alkalmazásával történt. Minden betegről készült funkcionális dinamikus (cine) MRvizsgálat egy steady-state szabad precessziós szekvencia alkalmazásával, amely során retrospektív EKG-kapuzás segítségével egy típusos (legalább 2,0 2,6 8,0 mm) felbontású rövid tengelyi képhalmazt nyertek a teljes szívről. A lélegzet-visszatartás alatti szegmentált k-space sebességkódolt fáziskontrasztos felvételeket a proximalis felszálló aortára merőleges síkban készítették el retrospektív EKG-kapuzást és a háttérfázis-eltolódás fantomkorrekcióját alkalmazva. Szív-MR-képelemzés és értékelése A képek utólagos feldolgozása és elemzése egy hitelesített berendezéssel történt a ReportCard 3.7-es verziójának alkalmazásával (General Electric Medical Systems, Milwaukee, Wisconsin). A rövid tengelyi képhalmazban azonosították a bal kamrai végdiasztolés és végszisztolés felvételeket. A bal kamrai kontúrok megadásakor a papillaris izmokat a vértartalom részének tekintették. A késői myocardialis halmozás mértékét a transmuralitas kvintiliseiben fejezték ki (normális=0, 1 25%, 26 50%, 51 75% és 76 100%) a 17 kamrai szegmentum vonatkozásában, vizuális elemzéssel. A terheléses perfúziós defektust vizuális elemzéssel állapították meg. A perfúziós defektust akkor minősítették rendellenesnek, ha a következő feltételek mindegyike tel - jesült: 1. a defektus subendocardialis vagy transmuralis volt, 2. legalább két szomszédos rövid tengelyi szeletben megfigyelhető volt, 3. a kontrasztanyagnak a szívizomzatban történő első megjelenését követően legalább 10 másodpercig fennállt, és 4. megfelelt egy koszorúér-ellátási területnek. Az izolált epicardialis vagy szemölcsizmot érintő defektusokat és a szuszceptibilitási műtermékeket nem ischaemiás eredetűnek minősítették. Nyugalmi perfúziós felvételeket nem készítettek. A terheléses perfúziós defektus típusát három kategóriába (normális, subendocardialis, illetve transmuralis) sorolták mind a 16 rövid tengelyi bal kamrai szegmentum esetében (a hosszanti tengelyben nem történt perfúziós vizsgálat). A kompozit szív-mr-végpontot a balkamra-funkcióról (bal kamrai ejekciós frakció [LVEF]: normális vagy kóros), a késői kontraszthalmozásról (kimutatható vagy sem) és terheléses perfúzióról (normális vagy kóros) nyert adatok kombinációjaként határozták meg, számszerű értéke a fenti vizsgálatok során észlelt kóros eredmények száma (minimum: 0, maximum: 3). Az aorta véráramlását fáziskontrasztos sebességkódolt méréssel határozták meg, a proximalis aorta koszorúér-szájadékok feletti részére merőleges vetületben. Minden esetben a háttér sebességeltolódás fantomkorrekcióját alkalmazták. Az offline (vizsgálat utáni) kamrai térfogatméréseket a Simpson-féle módszerrel végezték. Minden mérés az elfogadott kardiológiai képalkotó irányelveknek megfelelően történt. 10 Az 50% alatti bal kamrai ejekciós frak ciót tekintették kórosnak. Betegkövetés és az értékelés szempontjai Minden beteget legalább egy évig követtek a kezdeti szív-mr-vizs - gálatot követően (átlagos követési idő 2,6±1,2 év). A kimenetelt a kórházi és járó beteg elektronikus egészségügyi nyilvántartási adatok áttekintésével egy, a vizsgálati alanyok által kitöltött és postán beküldött kérdőív, illetve egy előírt, minden vizsgálati alannyal felvett telefonos interjú alapján határozták meg. Jelen vizsgálat elsődleges végpontja a cardialis események előfordulása volt, amelyet az összhalálozás, a nem fatális myocardialis infarctus és a késői (a kezdeti vizsgálatot követően eltelt több mint 90 nap után végzett) revascularisatio összetettjeként határoztak meg. Azért, hogy megállapíthassák a nem kóros szív-mr-eredmények jelentőségét az alacsony kockázatú betegek azonosításában, a másodlagos végpontokat kizárólag a nem kóros szív-mr-eredmények elemzésére használták. Ezek közé a cardialis halálozást és a kemény végponti eseményeket (cardialis halálozás vagy myocardialis infarctus) sorolták. Azoknál a betegeknél, akiknél több esemény következett be, csak az első eseményt vették figyelembe a túlélési elemzés során. Az események éves előfordulását a megfigyelt események számának a követési idő átlagos értékével történő elosztásával állapították meg a különböző csoportokban. A revascularisatio nélküli szívbillentyűműtéteket nem sorolták az események közé. Statisztikai elemzés A betegek kiindulási jellemzőit folytonos változók esetében medián értékekkel (25. és 75. percentilis), kategorikus változók esetében gyakorisági értékekkel írták le. A kategorikus változókat χ 2 -próbával hasonlítot-

circulation_2011_02_5tord-imp:circulation 11/18/11 2:20 PM Page 185 Bingham és Hachamovitch Kombinált szív-mr-vizsgálatok 185 ták össze. A többszörös összehasonlítások korrekciójára Bonferroni-féle korrekciót alkalmaztak. A p<0,05 (kétoldalú elemzés) értéket minősítették statisztikailag szignifikánsnak. Többváltozós modellek és túlélési vizsgálat A szív-mr-vizsgálati eredmények és cardialis eseményektől mentes túlélési idő összefüggésének vizsgálatára Cox-féle arányos kockázati modellt alkalmazták az egyéb lehetséges módosító tényezőkkel történt korrekciót követően. 11 13 Mivel a korai revascularisation áteső betegek bevonása, illetve kizárása is olyan torzításokhoz vezet, amelyek meghamisíthatják az eredményeket, ezért minden túlélési modellt a korai revascularisatio alapján alcsoportokra osztottak. 12 A korai revascularisatióra történő utalás, illetve a klinikai és anamnesztikus adatok és vizsgálati eredmények közötti összefüggést többszörös logisztikus regressziós modellel határozták meg. 12,13 A modellbe bevont kovariánsok kiválasztása a klinikai tapasztalatok és az ismert módosító tényezők azonosítása alapján történt. A modelleket, ahol lehetséges volt, gondosan megvizsgálták az arányos kockázatbecslés, a multikollinearitás, illetve a vizsgált tényezők additív értékének tekintetében. 12,13 A modellben >10 esemény/szabadságfok arányt tartottak fenn. A legfontosabb szív-mr-paraméterek alapján rétegzett eseménymentes túlélést a Kaplan Meier-módszerrel ábrázolták és log-rank teszttel hasonlították össze. Minden elemzéshez az S-PLUS szoftvercsomagot (Insightful Corp., Seattle, Washington) használták kiegészítő mappákkal (Hmisc, Design). A többletérték meghatározása A prognosztikus többletértéket a Wald-féle statisztikai és valószínűségi vizsgálatok során tapasztalható szignifikáns növekedésként definiálták, amelyet a képalkotó vizsgálati eredményeknek a képalkotó eljárások nélküli adatok optimalizált Cox-féle modelljéhez adva észleltek. Két elemzéssorozatot végeztek. Elsőként a szív-mr-vizsgálat előtti adatok alapján felállítottak egy modellt a vizsgálati végpontok előrejelzésére. Ezt követően az első modellhez hozzáadták a gadolínium beadása előtti szív-mr-vizsgálatok (bal kamrai térfogatok, bal kamrai ejekciós frakció, billentyűparaméterek) eredményeit (mivel ezek az adatok gyakran ismertek a vizsgálat előtt, ha pedig nem, abban az esetben meghatározhatók további kiegészítő vizsgálatok nélkül). Ez után az előző, második lépésben kapott modellhez hozzáadták a késői kontraszthalmozás adatait. Emellett párhuzamosan, szintén harmadik lépésként a terheléses perfúziós vizsgálat eredményeinek többletértékét is meghatározták a második modellhez képest (mivel a késői kont raszthalmozás vizsgálata és a terheléses perfúziós vizsgálat egymástól függetlenül is megtörténhet). Végül az összes szív-mr-vizsgálati eredmény együttes figyelembevételével is készítettek számításokat. Ha bármelyik lépésben szignifikáns növekedést találtak a Wald-féle globális χ 2 -értékben, a prognosztikus többletértéket igazoltnak tekintették. A Cox-féle proporcionális kockázati modellek alapján kockázatkiigazított túlélési görbéket ábrázoltak. A regressziós modelleket az újbóli mintavétel (100 ismétlés) módszerével validálták, és meghatározták a torzításkorrigált (bias corrected) C-indexeket. A modellkalibráció egy bootstrap- (mintából való mintavételen alapuló) eljárással korrigált kalibrációs görbe előállításával történt, amelyhez az átlagos Cox-féle arányos kockázatok által előre jelzett túlélés és a Kaplan Meier-módszer szerinti (megfigyelt) becslések különbségeinek egy időpontbeli ismételt összehasonlítását végezték el. Újraosztályozás A szív-mr-vizsgálati eredmények hatását a betegeket érintő kockázat újraosztályozására a nettó újraosztályozási javulás (NRI, net reclassification improvement) segítségével határozták meg. 14 Minden beteg esetében meghatározták a nem kívánt események várható kockázatát az egyes modellek alapján, majd felmérték a különböző adatoknak a modellekhez történő hozzáadásának hatását a kockázati csoportba történő besorolásra. Magasabb rizikójú csoportba történő újraosztályozás esetén felfelé, alacsonyabb rizikójú csoportba történő újraosztályozás esetén lefelé történő elmozdulásról van szó az osztályozásban. Az osztályozás javulását úgy határozták meg, hogy azokban a betegekben, akikben nem várt esemény következett be, kivonták a felfelé újraosztályozott egyének hányadából a lefelé újraosztályozott betegek hányadát; azoknál pedig, akiknél nem következett be esemény, a lefelé soroltak hányadából a felfelé soroltak hányadát, és a két különbséget összeadták. Az NRI azoknak a betegeknek a nettó számát adja meg, akik az újraosztályozás során pontosabban kerültek besorolásra, összesítve az újraosztályozást mindazon betegek vonatkozásában, akiknél jelentkezett és akiknél nem jelentkezett nem várt esemény. 1. táblázat. Demográfiai adatok Követési idô, nap 948 (639, 1263) Demográfiai/cardialis rizikófaktorok Életkor 65 (55, 74) Férfi, % 58,6 Magasvérnyomás-betegség, % 63,8 Cukorbetegség, % 25,4 Korán kialakult koszorúér-betegség a családi anamnézisben, % 41,7 Dohányzás, % 6,1 Framingham-rendszer szerinti pontérték 7,0 (5,0, 10,0) Tünetek, % Típusos angina 9,0 Bármilyen angina 56,2 Nehézlégzés 51,7 Társbetegségek, % Vesebetegség 7,7 COPD 10,1 PVD 6,2 Cardiovascularis anamnézis, % Korábbi CAD 49,4 Korábbi MI 34,9 Korábbi PCI 32,7 Korábbi CABG 15,0 Terheléses vizsgálat Nyugalmi szívfrekvencia 66,0 (30,0, 74,0) Csúcsfrekvencia 77,0 (67,0, 88,0) Balkamra-funkció LVEDV 124,0 (102,0, 149,0) LVESV 45,0 (33,0, 65,0) LVEF 63,0 (56,0, 69,0) LVSV 74,0 (63,0, 87,0) Aortaáramlás 70,0 (60,0, 84,0) Terheléses szív-mr-vizsgálat Kóros falmozgás, % 44,2 Kóros perfúzió, % 32,8 Kóros késôi kontraszthalmozás, % 37,6 Összegzett falmozgáspontérték 0,0 (0,0, 17,0) Összegzett perfúziós pontérték 0,0 (0,0, 2,0) Összegzett nyugalmi pontérték 0,0 (0,0, 5,0) A folytonos változók medián értékekben kifejezve (25. és 75. percentilis). CAD: (coronary artery disease) koszorúér-betegség; COPD: (chronic obstructive pulmonary disease) krónikus obstruktív tüdôbetegség; PVD: (pulmonary vascular disease) pulmonalis érbetegség; MI: (myocardial infarction) myocardialis infarctus; PCI: (percutaneous coronary intervention) perkután koszorúérintervenció; CABG: (coronary artery bypass grafting) koszorúér-bypassmûtét; LV: (left ventricular) bal kamrai; LVEDV: (LV end-diastolic volume) bal kamrai végdiasztolés térfogat; LVESV: (LV end-systolic volume) bal kamrai végszisztolés térfogat; LVEF: (LV ejection fraction) bal kamrai ejekciós frakció; LVSV: (LV stroke volume) bal kamrai verôtérfogat. Mivel nincs általánosan elfogadott érték az alacsony és a közepes, illetve a magas rizikó meghatározására az összes bekövetkezett esemény tekintetében, vizsgálatukban az újraosztályozás során 1,5, illetve 2%-os éves eseménygyakoriságot határoztak meg az alacsony, közepes és magas rizikójú csoportok elkülönítésére.

circulation_2011_02_5tord-imp:circulation 11/18/11 2:20 PM Page 186 186 Circulation Magyar Kiadás 2011. június szeptember Nincs késői kontraszthalmozás Van késői kontraszthalmozás 2. táblázat. A nem kívánt események egyváltozós elôrejelzése: képalkotó vizsgálatok elôtti adatok és Cox-féle arányos kockázati modell alapján meghatározott kockázati arányok (95%-os konfidenciaintervallum) Normális perfúzió Kóros perfúzió Összesítve 2. ábra. A szív-mr-vizsgálatot követô korai revascularisatióra történô utalás gyakorisága késôi kontraszthalmozás fennállásakor és annak hiányában (fekete és fehér oszlopok), terheléses perfúziós rendellenességek fennállásakor és hiányukban (középsô és bal oldali oszloppár), illetve összesítve (jobb oldali oszloppár). *p<0,01. Az egyes alcsoportok elemszáma (n) az oszlopok alatt látható. Eredmények Betegjellemzők és a szív-mr-vizsgálatok eredményei A jelen vizsgálatba bevont betegek jellemzően középkorúak voltak, túlnyomórészt férfiak, többségük ismert magasvérnyomás-betegségben szenvedett, kisebb részüknél a családi anamnézisben korán kialakult koszorúér-betegség volt fellelhető, és körülbelül negyedük szenvedett cukorbetegségben (1. táblázat). A betegek mintegy fele számolt be légszomjról és/vagy mellkasi panaszról, de az utóbbiak csak kevesebb mint 10%- ának volt típusos anginája a vizsgálatba történő bevonáskor. Aránylag kevés beteg esetében állt fenn jelentős társbetegség, és körülbelül 50%-uk anamnézisében volt fellelhető korábbi koszorúér-betegség. Az utóbbiak megközelítőleg egyharmada szenvedett el korábban myocardialis infarctust vagy esett át perkután koszorúér-intervención (PCI), és kevesebben estek át koszorúér-bypassműtéten (CABG). A betegek körülbelül egyharmadában volt kimutatható terheléses perfúziós zavar vagy késői kontraszthalmozás, és 44%-ukban állt fenn falmozgászavar. Továbbá 276 betegben volt kimutatható subendocardialis szegmentális perfúziós defektus, míg 53 betegben transmuralis szegmentális defektus ábrázolódott. A transmuralis perfúziós defektus gyakrabban fordult elő azoknál a betegeknél, akik korábban átestek myocardialis infarctuson, mint azoknál, akik nem (12,3% vs. 2,4%; p <0,001). Kimenetel A követési idő alatt 101 cardialis eseményt észleltek (9,8%; halálozás, myocardialis infarctus vagy késői revascularisatio). A fenti érték 30 cardialis eredetű halált, 36 nem cardialis eredetű halált, öt nem fatális myocardialis infarctust és 19 késői revascularisatiót foglal magában. A késői revascularisa - tiók esetében 13 késői PCI és kilenc késői CABG történt (három beteg PCI-n és CABG-n is átesett a követés során, de csak egy alkalommal vettük őket számításba a késői revascularisatio vonatkozásában). Ezenkívül 103 beteg esett át korai revascularisatión (kevesebb mint 90 nappal) a kezdeti szív- MR-vizsgálatot követően (91 PCI, 13 CABG). Demográfiai/cardialis rizikófaktorok Cardialis esemény, HR (95%-os CI) Életkor 1,05 (1,03, 1,07)* Férfi 1,10 (0,72, 1,62) Magasvérnyomás-betegség 2,62 (1,51, 4,57) Cukorbetegség 1,72 (1,12, 2,65) Korán kialakult koszorúér-betegség a családi anamnézisben 0,70 (0,56, 1,18) Dohányzás 0,48 (0,45, 1,09) Framingham-rendszer szerinti pontérték 1,07 (1,01, 1,13) Tünetek Angina 1,01 (0,89, 1,15) Nehézlégzés 1,95 (1,27, 3,00) Társbetegségek Vesebetegség 2,31 (1,25, 4,25) COPD 1,00 (0,61, 2,12) PVD 1,51 (0,46, 2,17) Cardiovascularis anamnézis Korábbi CAD 1,19 (0,78, 1,81) Korábbi MI 1,05 (0,69, 1,61) Korábbi PCI 1,31 (0,85, 2,00) Korábbi CABG 1,02 (0,55, 1,87) Terheléses vizsgálat Nyugalmi szívfrekvencia 1,02 (1,00, 1,03) Csúcsfrekvencia 1,00 (0,98, 1,01) * életévre vonatkoztatva; ütés/percre vonatkoztatva. HR: (hazard ratio) kockázati arány; CI: (confidence interval) konfidenciaintervallum; CAD: (coronary artery disease) koszorúér-betegség; COPD: (chronic obstructive pulmonary disease) krónikus obstruktív tüdôbetegség; PVD: (pulmonary vascular disease) pulmonalis érbetegség; MI: (myocardial infarction) myocardialis infarctus; PCI: (percutaneous coronary intervention) perkután koszorúér-intervenció; CABG: (coronary artery bypass grafting) koszorúér-by - pass mû tét. A szív-mr-vizsgálat eredményei és a korai revascularisatióra történő utalás A szív-mr-vizsgálatot követően a korai revascularisatióra történő utalás gyakorisága mind a késői kontraszthalmozás, mind a terheléses perfúziós vizsgálat eredményeinek függvényében növekedett (2. ábra). Bár a revascularisatio aránya nagyobb volt kóros késői kontraszthalmozás fennállása esetén, mint normális késői halmozás esetén (20,5% vs. 5,5%), ugyanakkor kóros terheléses perfúzió esetén gyakrabban történt revascularisatio, mint normális perfúzió esetén, mind normális, mind kóros késői kontraszthalmozás fennállásakor. Továbbá, normális perfúzió esetén a revascularisatio gyakoribb volt kóros késői kontraszthalmozás kimutatásakor, mint normális halmozás kimutatásakor (11,2% vs. 3,5%). A korai revascularisatio logisztikus regressziós modellezése során (χ 2 102, P<0,001) igazolódott, hogy a kóros perfúzió és késői kontraszthalmozás függ össze legszorosabban a korai revascularisatióval. A korai revascularisatiók időbeli megoszlását tekintve 87%-uk az első 30 nap során, 4,1%-uk a 30. és 60. nap, míg 2,0%-uk a 60. és 90. nap között történt. A késői revascularisatiók 6,7%-a a 90. nap és a hatodik hónap között történt, ezt az időpontot követően a revascularisatio gyakorisága megnőtt, 28% a hatodik hónap és az első év vége között, 37% a második évben, 28% pedig a vizsgálatot követően több mint két évvel történt.

circulation_2011_02_5tord-imp:circulation 11/18/11 2:20 PM Page 187 Bingham és Hachamovitch Kombinált szív-mr-vizsgálatok 187 3. táblázat. A nem kívánt események egyváltozós elôrejelzése: szív-mrvizsgálati változók és Cox-féle proporcionális kockázati modell alapján meghatározott kockázati arányok (95%-os konfidenciaintervallum) Balkamra-funkció Események, HR (95%-os CI) LVEDV 1,00 (0,97, 1,01) LVESV 1,00 (1,00, 1,01) LVEF 0,97 (0,96, 0,99) LVSV 0,98 (0,97, 0,99) Aortaáramlás 0,98 (0,97, 0,99) Terheléses szív-mr-vizsgálat Kóros falmozgás 2,03 (1,33, 3,01) Kóros perfúzió 2,00 (1,31, 3,04) Kóros késôi kontraszthalmozás 2,17 (1,42, 3,30) Összegzett falmozgáspontérték 1,02 (1,01, 1,04) Összegzett perfúziós pontérték 1,04 (1,01, 1,08) Összegzett nyugalmi pontérték 1,03 (1,01, 1,06) HR: (hazard ratio) kockázati arány; CI: (confidence interval) konfidenciaintervallum; LV: (left ventricular) bal kamrai; LVEDV: (LV end-diastolic volume) bal kamrai végdiasztolés térfogat; LVESV: (LV end-systolic volume) bal kamrai végszisztolés térfogat; LVEF: (LV ejection fraction) bal kamrai ejekciós frakció; LVSV: (LV stroke volume) bal kamrai verôtérfogat. A nem kívánt események egyváltozós előjelzői A betegek kiindulási adatai közül a cardialis eseményekkel leginkább összefüggő változók a betegek életkora, a magasvérnyomás-betegség, a fulladásos panaszok, a cukorbetegség, a Framingham-pontrendszer szerinti besorolás és a vesebetegség voltak. Érdekes módon a korábbi koszorúér-betegség, PCI vagy CABG nem bizonyult előjelzőnek. A nyugalmi szívfrekvencia a kimenetel szignifikáns előjelzője volt (2. és 3. táblázat). A szív-mr-vizsgálat során meghatározott bal kamrai ejekciós frakció és bal kamrai verőtérfogat előre jelezte a cardialis eseményeket. A szív-mr-vizsgálat során meghatározott aortaáramlás szintén előre jelezte a cardialis eseményeket, ugyanígy a falmozgászavar, a perfúziós zavar és a kóros késői kontraszthalmozás, illetve e változók összegzett pontértéke is. Cardialis eredetű halálozás vagy myocardialis infarctus Cardialis eredetű halálozás Cardialis esemény gyakoriság Mind normális 3. ábra. A cardialis eredetû halálozás vagy myocardialis infarctus (MI) éves gyakorisága, a cardialis eredetû halálozás éves gyakorisága, illetve a cardialis eseménygyakoriság normális terheléses perfúzió (n=610), normális késôi kontraszthalmozás (n=567) és normális ejekciós frakció (n=770), illetve mindhárom vizsgálat normális eredménye (n=488) esetén. Normális ejekciós frakció Normális késői kontraszthalmozás Normális terheléses perfúzió Összes esemény (%) Egy sem Egy Kettő Három Kóros szív-mr-paraméterek száma 4. ábra. Az össz-eseménygyakoriság a kóros szív-mr-paraméterek (terheléses perfúzió, késôi kontraszthalmozás és bal kamrai ejekciós frakció) számának függvényében. p<0,001 a kategóriák között. Az egyes alcsoportok elemszáma (n) az oszlopok alatt látható. Eseménygyakoriság a szív-mr-vizsgálati eredmények függvényében A szív-mr-vizsgálatok során észlelt normális eredmények esetén a kockázat alacsonynak bizonyult, attól függetlenül, hogy melyik paramétert vizsgálták (3. ábra). A cardialis eredetű halálozás vonatkozásában az eseménygyakoriság nagyon alacsony volt (0,6% és 0,7% közötti), kivéve a normális ejekciós frakció esetét, ahol ennél kissé magasabbnak bizonyult (1,0%). A kemény végponti események (cardialis halálozás vagy myocardialis infarctus) vonatkozásában az eseménygyakoriság szintén alacsony volt (0,6% és 0,9% közötti), kivéve a normális ejekciós frakció esetét (1,2%). Hasonló arányokat találtunk a cardialis események vonatkozásában is. Kóros szív-mr-vizsgálati eredmények esetén az eseménygyakoriság hasonlónak bizonyult mind kóros késői kontraszthalmozás, mind kóros ejekciós frakció, illetve kóros terhelési perfúzió esetében (5,6% és 6,1% között). Két kóros paraméter fennállásakor az eseménygyakoriság magasabb volt (6,5% és 7,0% közötti). A kórosnak mért szív-mr-paraméterek számának növekedésével a nem kívánt kimenetel kockázata szignifikáns mértékben emelkedett (4. ábra, p<0,001). Érdekes módon a subendocardialis perfúziós defektusok jelenléte nagyobb kockázattal járt, mint a transmuralis perfúziós defektusok jelenléte: az előbbi a kockázat szignifikáns egyváltozós előjelzőjének bizonyult, míg az utóbbi nem (kockázati arány [HR, hazard ratio]: 1,14 [1,04 1,25], p=0,0036 vs. 1,07 [0,86 1,34], p=0,5200) (a kockázati arányt a subendocardialis vagy transmuralis defektusok számával megadva). Megjegyzendő, hogy aránylag kevés betegben állt fenn transmuralis defektus. Többváltozós modellek és túlélési vizsgálat Elsőként a szív-mr-vizsgálat előtti adatok összefüggését vizsgálták a cardialis eseményekkel Cox-féle arányos kockázati analízist alkalmazva, ezáltal egy szív-mr-vizsgálat előtti krónikus pulmonalis hipertenziós modellt azonosítva (első modell). A második modellt e változók rögzítésével és a bal kamrai ejekciós frakció hozzáadásával állították fel. Ennek eredményeként a bal kamrai ejekciós frakciónak a modellhez történő hozzáadásával a χ2 értékének növekedését tapasztalták (a globális χ2 41,7-ről 51,5-re, a C-index 0,695-ről 0,717- re nőtt). A szív-mr-vizsgálat előtti adatok és a bal kamrai ejekciós frakció értékének rögzítése, majd az aortaáramlási paraméterek hozzáadása a globális χ 2 további növekedését eredményezte (57,5; C-index 0,722; harmadik modell), amelyet a késői kontraszthalmozás-vizsgálat eredményeinek hozzáadása tovább növelt (χ 2 63,5; C-index 0,747; negyedik modell). Végezetül, a terheléses perfúziós vizsgálat eredményei-