Coronary computerized tomographic angiography (CTA) has been used as a noninvasive method for ruling out high-grade stenoses. Even in the absence of such stenoses, analysis of coronary atherosclerosis may provide for important prognostic information, and this may be superior to exclusive coronary artery calcium scoring. We tested this hypothesis in patients undergoing CTA for clinical indications who had no stenoses requiring revascularization. From December 2004 to December 2006, 706 consecutive patients who underwent CTA but had no high-grade stenoses were included (58% men, mean age 59 ± 11 years). CTA and coronary artery calcium scoring (Agatston method) were performed using a 64-slice CT scanner with a gantry rotation time of 330 ms. CT angiograms were categorized as completely normal (group 1), showing minor plaque (group 2), or showing intermediate stenoses (group 3). Follow-up information was obtained in 670 patients (95%) over a mean of 3.2 years. There were 31 major adverse events (5%), namely 9 deaths (all noncoronary), 2 myocardial infarctions, 5 strokes, 13 coronary revascularization procedures (percutaneous or surgical >6 months after CTA), and 2 peripheral percutaneous interventions. Coronary status as defined by CTA was predictive of major events after adjustment for age and gender. In group 1, the probability of event-free survival at 3 years was 100%; in group 2, it was 96%; and in group 3, it was 91%. Compared to group 1, the risk in group 2 was increased 2.3-fold, and in group 3, it was increased 5.6-fold after adjusting for age and gender. However, after addition of the coronary artery calcium score to the regression analysis, CT angiographic status no longer appeared to be predictive. In conclusion, the risk of a major adverse cardiovascular event or death increased in a graded manner with degree of coronary atherosclerosis as defined by CTA even in the absence of high-grade coronary stenoses. However, in the absence of high-grade stenoses, we were unable to demonstrate a superior prognostic value of CTA compared to coronary artery calcium.
Coronary computed tomographic (CT) angiographic studies reporting on clinical outcomes have identified number of vessels harboring significant coronary stenosis as the main prognostic factor. Indeed, in symptomatic subjects, CT angiogram had a prognostic meaning over and above the coronary artery calcium score, the latter being much less precise in differentiating between stenotic and nonstenotic plaques. The value of detecting minor coronary stenoses or noncalcified nonstenotic plaques has not been conclusively defined. Classic invasive angiographic studies have suggested that in symptomatic patients, detecting nonstenotic plaques should have an important prognostic meaning compared to a completely normal coronary angiogram. A proof of principle for the prognostic role of coronary atherosclerosis in asymptomatic subjects has been provided by coronary calcium scanning. Independent of established cardiovascular risk factors, an increased amount of coronary calcium has been associated with an increased rate of cardiovascular events. It appears plausible that the analysis of noncalcified in addition to calcified coronary plaques would increase the prognostic power of computed tomography. However, head-to-head comparisons of CT angiography (CTA) and calcium scanning thus far have usually included symptomatic patients with some requiring revascularization. Against this background, we undertook the present study to analyze the outcome of patients undergoing CTA for clinical indications who also had a coronary calcium scan and in whom significant coronary stenoses were ruled out so that there was no need for revascularization.
Methods
All patients undergoing CTA for suspected coronary artery disease from December 2004 to December 2006 who did not have coronary stenoses requiring revascularization were included in a consecutive manner. All patients had sinus rhythm at time of CTA (otherwise it was not performed) and had no renal failure, hyperthyroidism, or allergy to iodine contrast medium. In April 2009, these patients were contacted by mail, inquiring for any cardiovascular events that might have taken place since CT angiographic examination. The mailing was kept very simple and merely asked a patient to state that there had been no event or else that a new diagnosis had been made by a doctor or that another event had occurred. In that case, the patient was called, and the patient’s medical records were acquired to ascertain the diagnosis. For patients not answering the mailing, efforts were made by telephone to establish contact. If this also failed, the primary physicians were contacted to acquire the respective medical information. If patients appeared to be lost to follow-up, we asked the respective citizen registries about a change of address or death of the patient. This enabled us to obtain a relatively complete follow-up in this clinical setting of approximately 95%. Events were verified according to patients’ records and were classified by consensus by 2 of the investigators (AS, TV). This approach was approved by the local ethics committee (Arztekammer Hessen, Frankfurt am Main, Germany).
Risk factor information was ascertained as documented in medical records. The respective risk factor was present if previously diagnosed by a treating physician. Otherwise, systemic hypertension was present if ≥2 values >140/90 mm Hg were documented or a patient received antihypertensive medication. Diabetes was present if fasting blood glucose values were >126 mg/dl on 2 occasions, 200 mg/dl if nonfasting, or if a patient received antidiabetic medication. Hypercholesterolemia was present if low-density lipoprotein cholesterol was >160 mg/dl, total cholesterol was >240 mg/dl, or a patient received lipid-lowering treatment. Smoking status was recorded. Because we were unable to consistently differentiate between former and current smoking in all patients, we distinguished only smoking versus not smoking.
CTA was performed using a 64-detector-row CT scanner (Somatom Sensation Cardiac 64; Siemens Medical Solutions, Forchheim, Germany) with a gantry rotation time of 330 ms (collimation 64 × 0.6 mm, reconstruction increment 0.3 mm). Image acquisition was performed during inspiratory breath-hold. To familiarize a patient with the protocol, breath-holding was practiced before the examination. Patients whose heart rates were >60 beats/min received bisoprolol 5 mg or atenolol 100 mg orally 1 hour before CT examination. For insufficient decrease in heart rate, up to 6 vials (30 mg) of metoprolol were injected intravenously. Before CTA, a coronary calcium scan was performed using prospective electrocardiographic triggering usually at 70% of the RR interval. The collimation was set to 30 × 0.6 mm, and the reconstructed slice thickness was 3 mm (adapted field of view depending on heart size, matrix 512 × 512, pixel size usually 0.5 × 0.5 mm). Calcium was defined as the presence of >2 contiguous pixels with >130 HU. These lesions were automatically identified and marked in color by the workstation. All lesions were added to calculate the total Agatston score.
Immediately before the actual CT angiographic examination, an oral premedication of 2 jets of Nitrolingual (glycerol trinitrate; Pohl-Boskamp, Hohenlockstedt, Germany) was applied in all patients. To estimate the transit time of the contrast agent, Imeron 350 (iodine 350 mg/ml; Bracco Altana Pharma, Konstanz, Germany) 10 ml was injected into an antecubital vein through an 18-gauge catheter with an injection speed of 5 to 6 ml/s followed by a saline bolus (50 ml, flow 5 ml/s). The total contrast dosage for CTA was adapted to the calculated scan duration (5 to 6 ml/s + 5 ml, total 65 to 80 ml, infusion rate 5.0 to 6.0 ml/s, saline bolus 50 ml, flow 5 to 6 ml/s).
Two separate datasets were reconstructed during different time points of the cardiac cycle (65% and 70% of RR interval). If deemed necessary for improved image quality, further reconstructions were performed in small steps (25% to 85% of RR interval). Additional reconstructions were calculated if stenoses were detected in the 65% and 70% reconstructions to rule out artifacts. After reconstruction, CT raw data were transferred to a personal computer–based workstation (Wizard; Siemens Medical Solutions, Erlangen, Germany). Image interpretation was based on axial source slices and on multiplanar reconstructions and thin maximum intensity projection images. Presence of a coronary artery stenosis within the 15 coronary segments was visually assessed irrespective of image quality. No coronary segments were a priori excluded from analysis. Any decrease in lumen diameter >50% was classified as significant stenosis. For severe calcification precluding analysis of luminal narrowing, high-grade stenosis was assumed. If invasive coronary angiography was performed within 5 days of CTA and did not—together with clinical information—confirm the presence of a high-grade stenosis or stenosis requiring revascularization, that patient remained in the study, because no interventional or surgical revascularization had to be undertaken.
Descriptive statistics include absolute and relative frequencies for categorical variables and means ± SDs and medians and quartiles for continuous variables. Mann-Whitney U test was used for comparison of continuous variables in 2 different groups. For time-to-event analysis, Kaplan-Meier curves were plotted. Log-rank tests were performed for analyzing different groups. To analyze several risk factors at a time, Cox proportional hazards regression models were calculated. All analyses are exploratory and no correction for multiple testing was carried out. Therefore, all p values given are descriptive only and cannot be considered statistically significant at any level. All statistical analyses were done using SPSS 15 or R (SPSS, Inc., Chicago, Illinois).
CT angiographic findings were initially classified in 4 categories according to plaque extent and localization. These categories were (1) completely normal, (2) minor plaque and/or calcification (luminal narrowing <30%), (3) intermediate stenoses (luminal narrowing 30% to 50%) in the main coronary arteries (American Heart Association segments 1 to 3, 5 to 8, 11 to 13), and (4) high-grade stenoses but only in side branches or distal portions of the main coronary arteries and thus not requiring revascularization. Because category 4 was rarely observed (only 17 patients, 2%) and to increase statistical power, the final classification consisted of 3 categories with increasing severity of coronary atherosclerosis: (1) completely normal, (2) minor plaque and/or calcification (luminal narrowing <30%), and (3) intermediate stenoses (luminal narrowing 30% to 50%) in the main coronary arteries or high-grade stenoses in other coronary segments.
Results
In total 706 patients were included (demographic data presented in Table 1 ). Indications for performing CTA were angina pectoris in 275 patients (39%), suspected ischemic heart disease due to electrocardiographic (at rest) or echocardiographic changes in 124 (18%), pathologic exercise stress test in 105 (15%), dyspnea in 57 (8%), ventricular arrhythmias in 41 (6%), and other indications in 104 (15%). Risk factor distribution and demographic data were virtually identical for subjects with or without complete follow-up information (data not shown).
| Overall Group | Patients With No Major Adverse Event (n = 639) | Patients With Major Adverse Event (n = 31) | |
|---|---|---|---|
| Mean age (years) | 59 ± 11 | 59 ± 11 | 63 ± 10 |
| Men | 409 (58%) | 364 (57%) | 20 (65%) |
| Body mass index (kg/m 2 ) | 27 ± 5 | 27 ± 5 | 29 ± 5 |
| Smoker | 112 (16%) | 107 (17%) | 4 (13%) |
| Diabetes mellitus | 39 (6%) | 37 (6%) | 1 (3%) |
| Systemic hypertension | 371 (53%) | 336 (53%) | 20 (65%) |
| Hypercholesterolemia | 253 (36%) | 226 (36%) | 13 (43%) |
| Aspirin medication | 231 (33%) | 212 (34%) | 12 (39%) |
| Statin medication | 158 (23%) | 145 (23%) | 7 (23%) |
| Antihypertensive medication | 333 (48%) | 298 (47%) | 21 (68%) |
Mean coronary calcium score according to Agatston was 99 ± 296. Distribution of scores is presented in Table 2 . A normal CT angiographic finding with no coronary calcified or noncalcified plaques was seen in 165 patients (23%, “group 1”). Minor plaque was observed in 450 patients (64%, “group 2”). More extensive plaque and intermediate stenoses were seen in 90 patients (13%, “group 3”).
| All Patients (n = 706) | Men (n = 409, 58%) | Women (n = 297, 42%) | Patients With Major Adverse Events (n = 31) | |
|---|---|---|---|---|
| Total calcium score, mean ± SD | 99 ± 296 | 148 ± 377 | 33 ± 78 | 282 ± 426 |
| Total calcium score, median | 4 | 16 | 0 | 90 |
| 25th percentile | 0 | 1 | 0 | 20 |
| 75th percentile | 64 | 144 | 27 | 355 |
| Patients with calcium score 0 | 193 (27%) | 76 (19%) | 117 (39%) | 2 (7%) |
| Patients with calcium score 1–100 | 363 (52%) | 215 (53%) | 148 (50%) | 14 (45%) |
| Patients with calcium score 101–400 | 107 (15%) | 76 (19%) | 31 (10%) | 9 (29%) |
| Patients with calcium score >400 | 40 (6%) | 39 (10%) | 1 (<1%) | 6 (19%) |
Follow-up information was obtained in 670 patients (95%) over a mean of 3.2 years. There were 31 major adverse events (5%) consisting of 9 deaths, all noncoronary, 2 myocardial infarctions, 5 strokes (nonfatal), 13 coronary revascularization procedures (percutaneous or surgical >6 months after CTA), and 2 peripheral percutaneous interventions. Twenty of these events occurred in men, 11 in women. Cause of death was cancer in 4 patients (lung, colon, liver, and pancreas, respectively) and pulmonary embolism, liver cirrhosis, accident, stroke, and multiorgan failure after mitral valve reconstruction in the other patients, respectively. Probably due to preselection and the limited number of events, the prevalence of categorical risk factors (hypertension, diabetes, smoking, and hypercholesterolemia) was not significantly different in patients with or without major events. However, those with events were significantly older and received antihypertensive medication more frequently ( Table 1 ).
Mean (median) calcium score in patients with a major event was 282 ± 426 (90) versus 93 ± 289 (4) in patients with no major event (p <0.001, Mann-Whitney U test). Rates of major events were 2 (1%) in patients with a calcium score of 0, 14 (4%) in patients with a calcium score from 1 to 100, 9 (9%) in patients with a calcium score from 101 to 400, and 6 (15%) in patients with a calcium score >400. Figure 1 shows survival free of major events in the calcium score groups.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
