Prognostic Significance. of Cardiac Cinefluoroscopy for Coronary Calcific Deposits in Asymptomatic Hi h Risk Subjects

Transcription

1 354 August 1994:354-8 CORONARY ARTERY DISEASE Prognostic Significance. of Cardiac Cinefluoroscopy for Coronary Calcific Deposits in Asymptomatic Hi h Subjects 19 ROBERT C. DETRANO, MD, PHD, NATHAN D. WONG, PHD, * WEIYI TANG, MD, WILLIAM J. FRENCH, MD, FACC, DEMETRIOS GEORGIOU, MD, EDDY YOUNG, MD, OLEH S. BREZDEN, MA, TERENCE M. DOHERTY, BA, KENNETH A. NARAHARA, MD, FACC, BRUCE H. BRUNDAGE, MD, FACC Torrance, Los Angeles and Irvine, California Objecdves. This remmh investigated the unnostic significance of radiographically detectable coronary calcific deposits. Background, Coronary calcific deposits are almost always associated with coronary atkrile rosis. We investigated the association between fluorescopically determined coronary calcium and coconary bmn disease end points at 1 year of follow-up. Me". This prospective population-based cohort study was conducted In the suburbs of Los Angeles. Fourteen hundred sixty-one asympsomailk adults with an estimated 2!109i risk of having a alronary heart disease event within 8 years underwent cardiac cinduoroseepy for assessment of coronary um at initiation of the study. Clinical status Including angina, documented myocardial infarction, myocardial revascularization and death from coronary heart disease were determined after 1 year. Results. The prevalence of UK deposits was high (47%). A follow-up examination at 1 year was successfully completed in 99.9% of subjects. Six subjects (0.4%) had died from coronary heart disease and 9 (0.6%) had had a nonfatal myocardial infarction. Thirty-seven subjects (2.5%) reported angina pectoris, and 13 (0.9%) had undergone myocardial revascularization. Fifty-three subjects had at least one event during the I-year period. Radiographically detectable calcium was associated with the presence of at least one 0 f these end points, with a risk ratio of 2.7 (confidence limits ). The presence of coronary calcium was an independent predictor of at least one end point when controlling for age, gender and risk factors. However, three deaths due to cormry heart disease and two nonfatal myocardial infarctions occurred in subjects without detectable coronary calcium. Conclusions. The presence of coronary calcific deposits incurs an increased risk of coronary heart disease events in asymptomatic high risk subjects at 1 year. This increased risk is independent of that Incurred by standard risk factors. U Am Coll Cardiol 1994;24 :354-8) Coronary heart disease is the most common cause of death and disability in industrialized nations (1,2). Autopsy studies show that coronary atherosclerosis, the pathologic substrate of coronary heart disease, is present in up to 15% of older children and young adults who die of violent causes (3), and in almost 100% of octogenarians (4). Pathologic and clinical studies (5-9) indicate that coronary calcific deposits are rare when atherosclerosis is absent. Blankenhorn (10) performed histopathologic studies of 89 randomly selected hearts. Ex- From the Saint John's Cardiovascular Research Center and Division of Cardiology, Harbor-University of California, Los Angeles. Medical Center, Torrance ; Department of Medicine, University of California, Los Angeles, School of Medicine, Los Angeles ; and 'Preventive Cardiology Program, Department of Medicine, University of California, Irvine, California. This study was supported by Grant 7R01 HL from the National Heart, Lung, and Blood Institute, National Institutes of Health. Bethesda, Maryland: and grants from Saint John's Hospital. Santa Monica, California and the Columbus Heart Center, Milan, Italy. Manuscript received November 8, 1993 ; revised manuscript received February 28, 1994, accepted March 2, Address for correspondence : Dr. Robert Detrano, Saint John's Cardiovascular Research Center, 1124 West Carson Street, RB2, Torrance, California OW by the American College of Cardiology amining 3,500 arterial segments microscopically and radiographically, he found that all calcified radiopaque lesions were atherosclerotic. Until recently, the only in vivo method for evaluating coronary atherosclerosis has been coronary angiography. However, because of its risks and expense, this procedure cannot be applied to large numbers of asymptomatic persons. Therefore, assessment of the prevalence of coronary atherosclerosis has been subject to biases inherent in autopsy and angiographic studies (11,12). Cardiac cinefluoroscopy has been used as a diagnostic test for coronary atherosclerosis (13-21) ; however, its application as a prognostic tool in asymptomatic subjects has never been tested. This is a report of the association of coronary calcium detected by 6neth-roscopy with coronary heart disease events in asymptomatic high risk subjects. Methods Recruitment. The South Bay Heart Watch screening clinic evaluated 5,023 subjects between December 1990 and December These subjects were X45 years of age and had at least one coronary risk factor. They were recruited /94/$7.00

2 DETRANO ET AL. CINEFLUOROSCOPY FOP. CORONARY CALCIFIC DEPOSITS 355 through newspaper, radio, television advertising and uifeci mail campaigns. A trained nurse (O.S.B.) administered an angina questionnaire and recorded cardiac history (prior myocardial infarction or revascularization), smoking (yes if currently smoking ; no if not currently smoking) and hypertension and diabetes histories (yes if being treated with diet or medications, or both ; no otherwise). The same nurse measured the height and weight and the systolic blood pressure twice after the subject had rested in a sitting position for 10 min. In addition, he performed a phlebotomy to obtain serum for analysis of total and high density lipoprotein cholesterol by a California Association of Pathology-certified laboratory. Twelve-lead electrocardiograms (ECGs), recorded by the nurse, were reviewed independently without knowledge of other data by three boardcertified cardiologists who used the Minnesota code to exclude subjects with prior myocardial infarction and a Romhilt-Estes score a4 to diagnose left ventricular hypertrophy. Differences were settled at quarterly meetings by majority rule. The angina questionnaire, adapted from that of Rose et al. (22), contained the following live questions : 1) Have you experienced chest pain or discomfort? 2) Is this discomfort provoked by exertion? 3) Is this discomfort relieved by rest or nitroglycerin? 4) Does relief of the discomfort occur within 10 min? 5) Does this discomfort ever last >30 min? We excluded subjects (4%) who answered yes to at least three of these questions, as well as those whose ECG showed diagnostic Q waves (5%) by the Minnesota criteria (23). All subjects reporting a history of a prior myocardial infarction were also excluded. The nurse used the risk factor information to calculate the risk of coronary heart disease events during an 8-year period according to the Framingham risk calculation algorithm (24) and excluded subjects found to have a <10% risk. The remaining 1,461 subjects (29%), underwent cinefluoroscopy to detect coronary calcification and were invited to return for a follow-up visit after I year. All subjects gave written informed consent as volunteer participants in this investigation, which had been approved by the Harbor-University of California, Los Angeles Human Subjects Committee on October 5, Fluoroscopic protocol. We performed cinefluoroscopy using a 60 left anterior oblique projection in all 1,461 subjects. We also used a 30 right anterior oblique projection in the first 451 subjects until we ascertained that this procedure increased the number of subjects with detectable coronary calcium by <I%. The imaging utilized a Philips X-ray tube, image intensifier and digital cardiac imaging (DCI) unit with a 512 x 512 pixel matrix format. Exposure factors included pulse width 16 ms, kilovolt (peak) (kvp) 50 to 80 and ma 50 to 90. The imaging unit adjusts kvp according to the output of the image intensifier so as to ensure optimal X-ray -enetration of the chest. The image intensifier light output is split into two parts, of which one is focused on 35-mm Kodak tine film moving at 15 frames/s and the other on a television camera whose signal is digitized, logarithmically amplified and subtracted. The results regarding the digitized images will be reported elsewhere. The average radiation dose was 800 mrads. image interpretation. Two observers (R.C.D., W.T.) reviewed the images without knowledge of each other's reading or the risk factor information. The distributions of the right coronary artery, the left main-left anterior descending coronary artery and the left circumflex coronary artery were evaluated. The two observers assessed these regions as to the presence of calcific deposits. Follow-up. After I year, surviving subjects returned to the South Bay Heart Watch clinic. We assessed coronary heart disease status using questions concerning intervening hospital admissions, the angina questionnaire used for the initial visit and a 12-lead ECG recorded in fashion identical to that used I year earlier. The same three cardiologists using the same Minnesota criteria reviewed these tracings. This committee met quarterly to decide by majority rule on all cases for which at least one member had diagnosed a new silent myocardial infarction. Subjects who were unable or unwilling to return to the clinic were contacted by telephone by the nurse coordinator (0. S. B.) and either were visited in their home by staff personnel or were mailed a questionnaire to obtain information regarding angina and hospital admissions. We obtained medical records for all subjects admitted to the hospital. We also obtained all medical records for nonsurviving subjects after first contacting next of kin to obtain information about the circumstances of the subject's death. A committee of three board-certified cardiologists reviewed these records without knowledge of other data to determine the occurrence of myocardial infarction or myocardial revascularization (coronary artery bypass surgery or percutaneous transluminal coronary angioplasty). Definitions. Myocardial infarction was defined as 1) prolonged chest pain prompting hospital admission with either evolutionary diagnostic ECG changes (ST segment elevation 2!2 mm or evolving Q waves, or both), and/or elevation of serum creatine kinase to twice the upper limit of normal or a positive serum creatine kinase MB fraction ; or 2) evolutionary, diagnostic ECG changes and the preceding enzyme elevations in the absence of prolonged chest pain. Coronary heart disease death was determined by committee review of medical records as having occurred if proved by autopsy or if the subject died within I h after the onset of prolonged severe chest pain or during hospital admission for an acute myocardial infarction. Angina pectoris was defined by a score of ~3 on the angina questionnaire. Coronary heart disease was considered present if any of the following end points occurred during the 1-year period : coronary heart disease death, myocardial infarction (including silent), angina pectoris or myocardial revascularization. Statistical analysis. We &.ompared cardiac end points with the presence of coronary calcium using a likelihood ratio,

3 356 DETRANO ET AL. CINEFLUOROSCOPY FOR CORONARY CALCIFIC DEPOSITS 100 Prevalence (%) 51h 6th 7th 8th 9th Table l. Coronary Heart Disease Events in 1,459 Subjects* With or Without Coronary Calcium Eventt *Follow-up data were not obtained for 2 subjects (both in the survivor group) of the total study group of 1,461 subjects. tsome subjects had more than one event. Calcium (n = 691) No Calcium (n = 768) p Value I (coronary heart 3(0.4%) 3(6.4%) 1.00 NS disease death) 11 (nonfatal myocardial 2 (0.3%) infarction) III fit->z;!'~'y 10 (1.5%) 3 (0.4%) revascularization) IV (angina) 270.9%) 10(1.3%) ,11,111 or IV 370.4%) 16(2.1%) Decade of Life Ftwe 1. Bar graph showing the prevalence of coronary calcium in the indicated decade of life in the 1,461 subjects, chi-square analysis or Fisher exact test when appropriate. We used multivariable logistic regression to determine independent contributors to the presence of one or more end points. Results Subject characteristics. Twelve hundred eighty-one men ( %)and 180 women (12%) were found eligible. Their mean age ± SD was 63.0 ± 7.8 years. Eighteen percent had diabetes mellitus (treated with diet or medications, or both), 52% reported a history of hypertension, 21% were smokers and 44% reported a history of coronary heart disease in a fist-degree relative, The mean systolic blood pressure was 144 t 18 mm Hg. The mean serum cholesterol and high density lipoprotein cholesterol was 241 ± 49 mg/dl and mgldl, respectively. lire of ium. The prevalence of coronary calcium was 47%. The prevalence in individual vessels was 44.5% for the left main-left anterior descending coronary arteries,18.1% for the right coronary artery and 12.0%0 for the left circumflex coronary artery. Twenty-seven percent of subjects had calcium in only one vessel, 13% in two vessels and 7% in all three vessels. Figure I shows the dependence of coronary calcium prevalence on age. There was a significant increase in coronary calcium prevalence from the 5th to the 9th decade of life (p < 0.001). I ever discordance. A second independent observer reread the fluoroscopic studies in blinded manner. Disagreement concerning the presence of calcium arose in 7.7% of cases. These two investigators then reviewed the discordant studies in blinded fashion and made a consensus decision. Completeness of follow-up. The research team successfully obtained follow-up information for 1,442 (99.9%) of the 1,444 survivors and for all 17 nonsurvivors. Thirteen hundred ninety-six participants (96.8%) returned to the South Bay Heart Watch clinic ; of the remaining subjects, 35 (2.4%) were interviewed by telephone, 10 (0.7%) were interviewed in their home by staff personnel and I subject (0.1%) was followed-up by mailed questionnaire. Seventeen subjects died in the year after their fluoroscopic examination. Medical records were obtained and reviewed for all 17. Six of the 17 deaths were ascertained to have been caused by coronary heart disease. Prediction of events. Table I summarizes the coronary heart disease events that occurred in a 1-year period and their relation to the detection of coronary calcium. Fiftythree subjects (3.6%) experienced a coronary heart disease end point within 1 year after their cinefluoroscopic examination. Thirty-seven of these events occurred in subjects with detectable coronary calcium, and 16 in those without coronary calcium (p = 0.001). Six subjects (0.4%) died from coronary heart disease in this period. Of these, three had detectable coronary calcium on fluoroscopic study. Two of the six died suddenly within I h after the onset of severe prolonged chest discomfort, three died in the course of acute myocardial infarction and one died from congestive heart failnlre shortly after coronary artery bypass surgery. Of the 11 subjects who died from noncoronary causes, 4 died from a cerebrovascular accident and one from cardiomyopathy of undetermined etiology. Of the five subjects with a noncoronary but cardiovascular death, four had coronary calcium on cinefluoroscopic examination. Nonfatal myocardial infarction (one silent) occurred in 10 subjects. These infarctions were more common among subjects with (1.2%) than among those without (0.3%) calcium (p = 0.03). Thirteen subjects underwent myocardial revascularization with either coronary bypass surgery or percutaneous transluminal coronary angioplasty. Ten of these had demonstrated coronary calcium on their initial visit (p 0.03). Of 37 subjects reporting angina on follow-up examination, 27 had demonstrated coronary calcium during the initial cinefluoroscopic study (p = 0.001). Logistic regression using age, gender, coronary calcium status and risk factors

4 August 1994:354-3 DETRANO ET AL. 357 CINEFLUOROSCOPY FOR CORONARY CALCIFtC DEPOSITS Table 2. Bivariate Analysis of Any Coronary Heart Disease End Point With Each Factor Mean ± SD or Frequency 95% Confidence Interval, Left ventricular hypertrophy (present) Fluoroscopic coronary 47% calcium (present) Hypertension (present) Family history (present) 44% Age (yr) 63 ± * Male gender 88% Smoking status 21% (presently smoking) Serum cholesterol (mgfdll 241 ± : HDL cholesterol (mg/dl) 44 ± Cholesterol/HDL ratio 5.9 ± , Diabetes (present) 19,70 1, Systolic blood pressure 144 ± 18 1, (mm Hg) Body mass index (kg/m 2 ) 30 ± *Per decade. tper 30. mg/dl Per 10 mg/dl. Pcr 10 mm Hg. HDL = high density lipoprotein. revealed that only history of hypertension, left ventricular hypertrophy, family history and calcium status were independent risk predictors of at least one coronary heart disease end point. Detection of coronary calcium in at least one, two or three vessels incurred an overall event risk of 5.4%, 5.6% and 6.2%, respectively. Thus, the increase in risk incurred by more extensive calcium deposits was not high. Table 2 shows mean values and frequencies as well as the risk ratios with 95% confidence intervals for each risk factor variable considered alone. The most powerful predictor of an adverse outcome was the presence of left ventricular hypertrophy on the rest ECG (risk ratio 3.2). Fluoroscopic calcium considered alone had a risk ratio of 2.7. A history of hypertension had a similar risk ratio, 2.6, and a history of coronary heart disease in a first-degree relative had a risk ratio of 1.8. Neither smoking status, family history, diabetes, systolic blood pressure nor serum lipids affected risk in this sample. Table 3 shows the risk ratios calculated from the logistic regression coefficients for detectable calcium, age and gender and the three risk factors that were significant predictors by the bivariate analysis presented in Table 2. Left ventric- Table 3. Logistic Regression Coefficients Relating Probability of End Points to Factors and Coronary Calcium Status 95% Confidence Interval Coronary calcium (present) Age (per decade) ,67 Male gender Hypertension (present) Left ventricular hypertroph) (present) Family history (present) ular hypertrophy, hypertension, coronary calcium and family history were significant independent risk predictors with similar risk ratios (1.36 to 1.67). Age and gender were not significant predictors. The logistic equation, = logit (fixi l) had its argument, QJ, equal to Calcium Age Male Hypertension (Left ventricular hypertrophy) (Family history) (Age in years ; dichotomous variables either I or -1). Discussion Our results indicate that the cinefluoroscopic detection of coronary calcific deposits provides prognostic information in subjects at high risk for coronary heart disease events. This information is independent of that incurred by elevated levels of standard risk factors such as serum lipids, smoking and systolic blood pressure. Ample theoretic and experimental justification exists for such a predictive association (16,25,26). However, coronary calcium was not detected in five subjects who died from coronary heart disease death or had a nonfatal myocardial infarction. The mean age of these five subjects was 61 years and four of them were men, making the age and gender distributions similar to that of the entire sample. Moreover, their risk factor profiles were not remarkable except that none were smokers and all had hypertension. We have noted an increased prevalence of coronary calcium in smokers and no significant relation between coronary calcium and history of hypertension (27). In view of the relatively high prevalence of coronary calcium in this population, the presence of these false negatives may limit the applicability of cinefluoroscopy as a screening tool. Strengths and limitations of the study. The 1-year event rate in this cohort is likely to be accurate because of complete and comprehensive follow-up (>99%). Although the funding agency required that the fluoroscopic results be presented to the subjects, these were presented as research data whose meaning was to be determined and not as significant results requiring referral or further diagnostic evaluation. Thus, we do not believe that the knowledge of the results had great impact on the medical care of these subjects. Confirmation of this rests on the following facts. 1) Although reported smoking frequency decreased by 2%, there was no greater tendency for subjects with coronary calcium than for subjects without coronary calcium to quit smoking. 2) Mean systolic blood pressure did not change significantly in either group of subjects during the ]-year period. 3) If the presence of calcium had influenced the subjects with coronary calcium to report angina or undergo revascularization more frequently, we would have noted an especially high risk ratio for these "soft" end points. This was not noted. For these reasons, we do not believe that the knowledge of the test results significantly influenced the coronary end points. The participants were all self-referred volunteers known to be at high risk for coronary heart disease. This

5 358 DETRANO ET AL. CINEFLUOROSCOPY FOR CORONARY CALCIFIC DEPOSITS selection procedure was necessary to ensure an adequate number of coronary events in a relatively short period of time. However, the results for these subjects may not hold for subjects at lower risk chosen randomly from the general population. The comments and review of George Diamond, MD and the editorial assistance of Rebecca A. Moore are deeply appreciated. We recognize and appreciate the computational advice of A. Wow, PhD. eferteiw 1. Vital Statistics of the United States 1987, Vol. 11. Mortality Part A, Publication No Washington, D.C., Public Health Service Higgins M, Thom T. Trends in coronary heart disease in the United States, lot I Epidemiol 1989 ;18: PDAY Research Group. Relationship of atherosclerosis in young men to serum lipoprotein cholesterol concentrations and smoking : a preliminary report from the Pathobiological Determinants of Atherasclerosis in Youth (PDAY) Research Group. JAMA 1990 ;261: Lie IT, Hammond Pl. Pathology of the senescent heart: anatomic observations of 237 autopsy studies of patients 90 to 105 years old. Mayo Clin Proc 1908 ;63 :552-64, 5. Anderson HC. Calcific diseases : a concept. Arch Pathol Lab Med 1983 ;107 : Arnett EN, Isner IM, Redwood DR, et al. Coronary artery narrowing in coronary heart disease : comparison of cineangiographic and necropsy findings. Ann Intern Med 1979;91 : Eggen DA, Strong JP, McGill HC. Coronary calcification : relationship to clinically significant coronary lesions and race, sex, and topographic distribution. Circulation 1%5,32 :948-55, 8. Frink RJ, Achor RWP, Brown AL, Kincaid OW, Brandenburg RO. Significance of calcification of the coronary arteries. Am J Cardiol 1970 ;26, Warburton RK, Tampas JP, Soule AS, Taylor HC. Coronary artery calcification : its relationship to coronary artery stenosis and myocardial infarction. Radiology 1968 ;91 :109-I Blankenhom D. Coronary arterial calcification, a review. Am J Med Sci 1961 ;July :1-9. It. BeggCB, Biases in the assessment of diagnostic tests. Stat Med 1 7;6: Phllbrick W, Norwitz R, Feinstein A. Methodologic problems of exercise testing for coronary artery disease : groups, analysis and bias. Am J Cardiol 1 ;46: Hamby R, Tabrah E, Wisoff K, Hartstein J. Coronary artery calcification : clinical implications and angiographic correlates. Am Heart J 1974 ;87 : Bartel A, Chen J, Peter R, Behar V, Kong Y, Lester L. The significance of coronary calcification detected by fluoroscopy. A report of 360 patients. Circulation 1974;49: Aldrich R, Brensike J, Battaglini J, et al. Coronary calcifications in the detection of coronary artery disease and comparison with electrocardiographic exercise testing. Results from the NHLBI's Type 11 Coronary Intervention Study. Circulation 1979;59 : Margolis J, Chen J, Kong Y, Peter R, Behar V, Kisslo J. The diagnostic and prognostic significance of coronary artery calcification. A report of 800 cases. Diagn Radial 1980;137: Hung J, Chaitman B, Lam J, et al. Noninvasive diagnostic test choices for the evaluation of coronary artery disease in women : a multivariate comparison of cardiac fluoroscopy, exercise ECG and exercise thallium myocardial perfusion scintigraphy. J Am Coll Cardiol 1 :4: Carboni L, Celli M, D'Ermo A, Santoboni G, Zanchi M. Combined cardiac cinefluoroscopy. exercise testing and ambulatory ST-segment monitoring in the diagnosis of coronary artery disease : a report of 104 symptomatic patients. Int J Cardiol 1985 ;9: Detrano R, Salcedo E. Hobbs R. Yiannikis J. Cardiac cinefluoroscopy as an inexpensive aid in the diagnosis of coronary artery disease. Am J Cardiol 1986;57: Uretsky B, Rifkin R, Sharma S. Reddy P. Value of fluoroscopy in the detection of coronary stenosis : influence of age, sex and number of vessels calcified on diagnostic efficacy, Am Heart J 1 ;115 : Loecker T, Schwartz R. Cotta C, Hickman J. Fluoroscopic coronary artery calcification and associated coronary disease in asymptomatic young men. J Am Coll Cordial 1992;19: Rose G, McCartney J, Reid H. The Rose questionnaire. Br J Prev Soc Med 1977;31 : Blackburn H, Keys A, Simonson E, Rautaharju P, Punsar S. The electrocardiogram in population studies : a classification system. Circulation 1%0;21 : United States National Heart, Lung and Blood Institute. Framingham Study : An Epidemiological Investigation of Cardiovascular Disease. Section 37. U.S. Dept. of Commerce, Nat Bur of Information Services 01 71, 25. Detrano R, Froelicher VF. A logical approach to screening for coronary artery disease. Ann Intern Med 1987 ;106 : Witteman J, Kannel W, Wolf P. et al. Aortic calcified plagac= and cardiovascular disease (the Framingham Study). Am J Cardiol 1990;66 : Detrano R, Wong N. Prevalence of coronary artery calcium in high risk adults. Am Heart J. In press.