OUP user menu

Incremental prognostic significance of left ventricular dysfunction to coronary artery disease detection by 64-detector row coronary computed tomographic angiography for the prediction of all-cause mortality: results from a two-centre study of 5330 patients

James K. Min, Fay Y. Lin, Allison M. Dunning, Augustin Delago, John Egan, Leslee J. Shaw, Daniel S. Berman, Tracy Q. Callister
DOI: http://dx.doi.org/10.1093/eurheartj/ehq020 1212-1219 First published online: 2 March 2010

Abstract

Aims Early reports indicate a prognostic value of coronary artery disease (CAD) detection by coronary computed tomographic angiography (CCTA), although studies have been limited by small samples in single centres. Coronary computed tomographic angiographic measures of left ventricular ejection fraction (LVEF) to add incremental prognostic value beyond CAD detection have not been examined.

Methods and results We evaluated 5330 consecutive patients without known CAD undergoing CCTA at two centres. Stenosis severity by CCTA was graded as none (0%), mild (1–49%), moderate (50–69%), or obstructive (≥70%). Left ventricular ejection fraction was graded as normal (>50%) or reduced (≤50%). About 2.3 ± 0.6 year follow-up of patients for all-cause mortality was performed using multivariate and Cox proportional hazards models; 100 deaths occurred (1.9%). Detection of obstructive CAD correlated with mortality [hazards ratio (HR) 2.44, 95% confidence interval (CI) 1.61–3.72, P < 0.001]. Compared with those without obstructive CAD, individuals with increasing numbers of vessels with obstructive CAD experienced increased risk of death: 1-vessel (HR 2.23, 95% CI 1.34–3.72), 2-vessel (HR 3.29, 95% CI 1.62–6.71), or 3-vessel (HR 7.35, 95% CI 3.79–14.29) (P < 0.001 for all). Compared with those with LVEF >50%, those with LVEF ≤50% exhibited higher rates of death (HR 1.56, 95% CI 1.04–2.36, P = 0.03). Annualized mortality rates in those with non-obstructive CAD and LVEF >50% were low (0.51%) and increased accordingly for non-obstructive CAD and LVEF ≤50% (0.74%), obstructive CAD and LVEF >50% (1.76%), and obstructive CAD and LVEF ≤50% (3.97%) (log-rank test P < 0.001).

Conclusion In a large two-centre cohort of patients without known CAD, obstructive CAD detection by CCTA was related to incident death by the absolute presence of as well as increasing numbers of vessels with obstructive CAD. The addition of LVEF by CCTA enhanced risk correlation for death.

  • Computed tomography
  • Coronary disease
  • Angiography
  • Prognosis

Introduction

Sixty-four-detector row coronary computed tomographic angiography (CCTA) has emerged as a promising non-invasive modality for anatomic evaluation of coronary artery disease (CAD), demonstrating high diagnostic performance for detection and exclusion of CAD.13 At present, several studies have examined the prognostic value of CAD detection by CCTA. Nevertheless, these studies have been largely limited by small sample sizes, single centre evaluations, short follow-up periods, and use of older generation CCTA scanners.410 Furthermore, while retrospective ECG-gated helical CCTA acquisition permits the assessment of global left ventricular ejection fraction (LVEF), the incremental prognostic value of LVEF by CCTA has not yet been evaluated.1113

The goal of the present study was to determine in a large cohort of patients from two centres whether CCTA detection of CAD and LVEF evaluation would offer independent and additive prognostic information for the prediction of all-cause death.

Methods

Patients

We evaluated 5330 consecutive patients from October 2005 to November 2007 who underwent 64-detector row CCTA at two centres (Tennessee Heart and Vascular Institute, Hendersonville, TN, USA and Capital Cardiology Associates, Albany, NY, USA) who were without known CAD. Patients were referred for evaluation by CCTA for a variety of indications including evaluation of symptoms, signs of cardiac disease (abnormal rest or stress test), or asymptomatic patients with peripheral arterial disease, cerebrovascular disease, or multiple CAD risk factors. All patients were in normal sinus rhythm and were capable of the breath hold needed for CCTA. Patients with heart rates >70 b.p.m. were given 5 mg intravenous metoprolol at 5 min intervals to a total dose of 25 mg. If the patient's heart rate did not drop below 70 b.p.m., CCTA was performed at the lowest heart rate. Further, immediately prior to the CCTA image acquisition, patients were administered 0.4 mg of sublingual nitroglycerin, as was the laboratory protocol at both sites.

Prior to the initiation of the scan, we (T.Q.C. and A.D.) prospectively collected information on the presence of categorical cardiac risk factors in each individual. Systemic arterial hypertension was defined as a documented history of high blood pressure or treatment with anti-hypertensive medications. Diabetes mellitus was defined by diagnosis of diabetes made previously by a physician and/or use of insulin or oral hypoglycaemic agents. Dyslipidaemia was defined as known but untreated dyslipidaemia or current treatment with lipid-lowering medications. A positive smoking history was defined as current smoking or cessation of smoking within 3 months of testing. Family history of coronary heart disease was determined by patient query.

Scan protocol and image reconstruction

All scans were performed using a 64-detector row CT scanner (Lightspeed VCT, GE Healthcare, Milwaukee, WI, USA). Imaging of a test bolus of contrast was performed at 2 mm superior to the take-off of the left main coronary artery for precise timing of contrast injection. During the CCTA angiography acquisition, 100 cc of iodinated contrast (Isovue 370, Bracco Diagnostics, Princeton, NJ, USA or Visipaque, GE Healthcare, Princeton, NJ, USA) was injected, followed by a 50 cc saline flush. Contrast timing was performed to optimize uniform contrast enhancement of the coronary arteries. The scan parameters were: 64 × 0.625 mm collimation, tube voltage 120 mV, and effective 400–650 mA. Estimated radiation doses ranged from 10 to 20 mSv.

Helical scan data were obtained with retrospective ECG gating. Images were reconstructed immediately after completion of the scan in a consistent manner to identify motion-free coronary artery images. Electrocardiographically gated data sets were reconstructed at 70, 75, and 80% of the cardiac cycle after the QRS complex to identify central diastole, with additional data sets reconstructed at 40, 45, and 50% of the cardiac cycle to identify central early diastole. Optimal phase reconstruction was assessed by comparison of different phases, and the phase with the least amount of coronary artery motion was chosen for analysis. Multiple phases were used for image interpretation if minimal coronary artery motion was different for different arteries. Coronary computed tomographic angiographies were evaluated on two-dimensional maximum intensity projections in oblique cardiocentric views that focused on coronary arterial segments for optimal viewing. Two orthogonal thin maximal intensity projection views approximating traditional coronary angiography angles were used for the left anterior descending, left circumflex, and right coronary artery circulations, respectively. Three-dimensional rotation was performed, when necessary, to focus on diagonal and marginal branch vessels.

For CCTA with suboptimal image quality, multiphase reconstruction was employed for additional points within the cardiac cycle for the identification of phases with the least amount of cardiac motion artefact. In other cases, multisector reconstruction algorithms were employed to optimize image quality. Three-dimensional views using curved multiplanar reformation and short-axis cross-sectional viewing techniques were additionally used to enhance detection of obstructive coronary plaque, if necessary. In all individuals, irrespective of image quality, every arterial segment was scored. If a coronary artery segment within an artery was uninterpretable despite these multiple techniques, the unevaluable segment was scored similar to the most proximal segment which was evaluable.

Coronary computed tomographic angiographies were also reconstructed at 10% increments of the cardiac cycle beginning with 10–100% for LV function (LVF) assessment.

Non-invasive coronary artery analysis by coronary computed tomographic angiography

All scans were analysed by a level III-certified cardiologist with experience interpreting several thousand CCTA scans (T.Q.C. and A.D.). Coronary arteries were visually scored for the presence of coronary plaque. Coronary arteries were divided into the left main artery, left anterior descending artery, left circumflex artery, and right coronary artery. Diagonal and septal branches were considered as part of the left anterior descending artery; obtuse marginal branches were considered as part of the left circumflex artery; and right ventricular marginal branches, posterior descending artery, and right posterolateral branches were considered as part of the right coronary artery. In the cases of left coronary dominance, the posterior descending artery was considered as part of the left circumflex artery. For the purposes of grading, coronary arteries were graded in accordance to the plaque within it exhibiting the greatest luminal diameter stenosis severity.

In each coronary artery, coronary atherosclerosis was defined as any tissue structures >1 mm2 that existed either within the coronary artery lumen or adjacent to the coronary artery lumen that could be discriminated from surrounding pericardial tissue, epicardial fat, or the vessel lumen itself. Coronary atherosclerotic lesions were quantified for stenosis by visual estimation. Luminal diameter stenosis severity was graded as none (0% luminal stenosis), mild (1–49% luminal stenosis), mild (50–69% luminal stenosis), moderate (50–69% luminal stenosis), or obstructive (≥70% luminal stenosis). Per cent obstruction of coronary artery lumen was based on a comparison of the luminal diameter of the segment exhibiting obstruction to the luminal diameter of the most normal appearing site immediately proximal to the plaque. In instances in which plaque was highly calcified, two-dimensional oblique images were also visualized without maximal intensity projection (i.e. 0.625 mm isotropic voxel resolution) or multiplanar reformats with cross-sectional views to minimize any partial volume averaging or beam hardening artefact of calcium.

We evaluated the presence or absence of obstructive CAD at a per-patient level and a per-vessel level. The presence of obstructive CAD at the per-patient level was defined as any coronary artery segment exhibiting obstructive CAD. The presence of obstructive CAD was also graded on a per-vessel basis from 0 to 3, summating obstructive CAD for the left anterior descending, left circumflex, and right coronary arteries. The presence of left main artery obstructive CAD was categorized as two-vessel disease for left main arteries, which gave rise to the left anterior descending artery and left circumflex artery.

Non-invasive assessment of left ventricular ejection fraction by coronary computed tomographic angiography

Left ventricular ejection fraction was quantitated in the following manner. By use of a dedicated semi-automated Hounsfield unit threshold level-based LV segmentation software with manual correction via a three-dimensional post-processing workstation (CardIQ Function, Advantage AW Workstation, GE Healthcare), end-diastolic volume (EDV), end-systolic volume (ESV), and LVEF were calculated by the following formula: (EDV−ESV/EDV). ED and ES were defined as the phase with the largest LV cavity before aortic valve opening and the phase with the smallest LV cavity before mitral valve opening, respectively. This method calculates the LVEF according to the following formula: EDV−ESV/EDV, using a per-pixel classification after appropriate definition of the mitral valve plane, LV endocardium, and inclusion of the LV outflow tract.

Readers were also permitted to use the three-dimensional post-processing workstation to manipulate three-dimensional data to prescribe two-chamber, three-chamber, four-chamber, and short-axis reformations. An initial start plane based on the mid-point of the maximal mitral annular plane in the sagittal view was used to generate double-oblique short axis measures, as we have described previously (reference). Using an orthogonal view of the LV to the sagittal plane, a four-chamber view was created, from which two- and three-chamber views were identified. Images were then interactively manipulated by the reader to assess the entirety of these respective views. This enabled comprehensive assessment of all myocardial segments in at least four views. Left ventricular ejection fraction was judged as normal (LVEF > 50%) or reduced (LVEF ≤ 50%) in binary fashion.

Follow-up

The primary endpoint was time to death from all causes. Follow-up procedures were approved by the study center's institutional review board. Death status was ascertained by querying the Social Security Death Index in all patients (i.e. 100% follow-up).

Statistical analysis

SPSS 12.0 (www.spss.com, Chicago, IL, USA) was used for all statistical analyses. Categorical variables are presented as frequencies and continuous variables as means ±1 standard deviation. Variables were compared with χ2 statistic for categorical variables and by Student's unpaired t-test for continuous variables. All continuous variables were found to be normally distributed. Time to death from all causes was calculated using univariable Cox proportional hazards models. Risk-adjusted models were also devised including multivariable stepwise models adjusting for baseline cardiac risk factors. Multivariable models were limited to not more than 10 variables to avoid model overfitting. Relative risk ratios were calculated with 95% confidence intervals (CIs) based on binomial distributions. A two-tailed P < 0.05 was considered statistically significant. In each case, the proportional hazards assumption was met. Finally, we also evaluated effect modification using interaction terms in Cox regression models.

Results

Clinical characteristics of the coronary computed tomographic angiography cohort predicting near-term mortality

About 5330 consecutive patients ≥40 years without known CAD comprised the study cohort. Of the 5330 patients, 499 (9%) were recruited from Albany, NY, USA; the remaining 4831 (91%) were recruited from Tennessee. With analyses performed excluding Albany patients, no significant differences were observed in reported results. The average age of the study group was 56.1 ± 13.9 years, with 45.6% male. Additional characteristics of the entire patient cohort can be seen in Table 1. Survival was evaluated after a mean follow-up period of 2.3 ± 0.7 years (27.2 ± 7.7 months). At the completion of follow-up, a total of 100 (1.9%) deaths were reported. The majority of individuals was referred for chest pain or tightness: among the study cohort, 711 (13.3%) were asymptomatic, whereas 1436 (26.9%), 1718 (32.2%) and 1465 (27.5%) experienced non-cardiac pain, atypical angina, and typical angina, respectively. Frequencies of symptoms and typicality of angina in relation to number of vessels with obstructive CAD are listed in Table 2.

View this table:
Table 1

Baseline demographics

Entire sample (n = 5330)Alive (n = 5230)Dead (n = 100)P-value
Age (years)56.1 ± 13.955.8 ± 13.869.4 ± 10.9<0.001
Male gender (%)45.645.554.00.09
Hypertension (%)57.757.664.00.20
Diabetes (%)19.419.232.00.001
Dyslipidaemia (%)56.556.843.00.006
Family history (%)51.451.545.00.19
Current smoking (%)31.231.234.30.50
Non-cardiac pain10741046280.048
Atypical angina13841364200.17
Typical angina16771642350.44
Other696680160.38
View this table:
Table 2

Comparison of risk factors by number of vessels obstructed

No CAD, mean (SD) or %1VD, mean (SD) or %P-value2VD, mean (SD) or %P-value3VD, mean (SD) or %P-value*
Number of Patients47454514258
Age (years)55.0 ± 13.963.1 ± 11.1<0.00164.3 ± 11.0<0.00167.6 ± 10.4<0.001
Male sex (%)43.659.0<0.00159.2<0.00172.4<0.001
Hypertension (%)56.266.5<0.00173.9<0.00170.70.03
Diabetes (%)17.731.1<0.00132.4<0.00132.80.003
Dyslipidaemia (%)54.669.6<0.00173.2<0.00169.00.03
Family history (%)51.847.10.0657.70.1641.40.12
Current smoking (%)31.131.30.9330.70.9141.40.09
Non-cardiac pain (%)21.824.10.1226.30.0931.00.03
Atypical angina (%)35.330.60.2132.10.8531.00.91
Typical angina (%)28.529.90.2429.90.3925.80.97
Other symptoms (e.g. palpitations) (%)14.415.40.3511.70.5512.10.84
All-cause mortality (%)1.264.66.319.0
  • *All P-values represent comparisons to no CAD.

Univariable coronary computed tomographic angiography models estimating death from all causes

In univariable Cox models, traditional CAD risk factors were associated with death during the study period, including older age and diabetes. Similarly, the presence of any obstructive CAD was significantly associated with mortality [hazards ratio (HR) 4.46, 95% CI 2.99–6.66, P < 0.001]. Further, in comparison to those with no obstructive CAD, increasing all-cause mortality was identified for individuals with 1-vessel (P < 0.001), 2-vessel (P < 0.001), and 3-vessel (P < 0.001) obstructive CAD (Table 2). Vessel-based gradations of obstructive CAD identified individuals who tended to be older (P < 0.001), male (P < 0.001), hypertensive (P < 0.001), diabetic (P < 0.001), dyslipidaemic (P < 0.001), and smokers (P < 0.001) (Table 2 and 3). Similarly, those individuals with moderate or obstructive CAD within the left main artery exhibited higher rates of mortality during the follow-up period (HR 3.48, 95% CI 1.94–6.23, P < 0.001). In univariate analysis, reduced LVEF was associated with increased rates of all-cause death (HR 2.07, 95% CI 1.38–3.11, P < 0.001).

View this table:
Table 3

Univariate survival analysis

Unadjusted hazards ratio (95% confidence interval)P-valueAdjusted hazards ratio (95% confidence interval)P-value
Age1.08 (1.07–1.10)<0.0011.08 (1.06–1.10)<0.001
Gender1.40 (0.95–2.08)0.090.66 (0.44–0.98)0.04
Hypertension1.31 (0.88–1.97)0.20N/AN/A
Diabetes1.97 (1.29–3.00)0.0021.79 (1.17–2.75)0.007
Dyslipidaemia0.58 (0.39–0.86)0.0060.39 (0.26–0.59)<0.001
Current smoking1.15 (0.76–1.75)0.50N/AN/A
Obstructive CAD4.46 (2.99–6.66)<0.0012.44 (1.61–3.72)<0.001
Reduced LVEF2.07 (1.38–3.11)<0.0011.56 (1.04–2.36)0.03

Multivariable coronary computed tomographic angiography models estimating death from all causes

In multivariable Cox regression analysis, considering age, hypertension, dyslipidaemia, diabetes, and family history, both obstructive CAD (P < 0.001) and reduced LVEF (P = 0.01) significantly predicted all-cause mortality. Risk of death increased with numbers of vessels involved with obstructive CAD for 1-vessel (HR 2.23, 95% CI 1.34–3.72, P = 0.005), 2-vessel (HR 3.29, 95% CI 1.62–6.71, P = 0.001), and 3-vessel (HR 7.35, 95% CI 3.79–14.29, P < 0.001).

Subsequently, we developed a multivariate model accounting for combined assessment of obstructive CAD and reduced LVEF, categorizing individuals into one of four groups by a summary variable. In comparison to those with none or non-obstructive CAD with normal LVEF, individuals with none or non-obstructive CAD with reduced LVEF (HR 2.30, 95% CI 1.25–4.26, P = 0.008); obstructive CAD with normal LVEF (HR 3.73, 95% CI 2.19–7.25, P < 0.001); and obstructive CAD with reduced LVEF (HR 4.26, 95% CI 2.48–7.35) demonstrated increased hazards for all-cause death. A Cox proportional hazards survival analysis was conducted to evaluate the interaction between reduced LVEF and obstructive CAD. The analysis found no significant interaction between reduced LVEF and obstructive CAD.

The Kaplan–Meier survival curves in Figure 1 depict the differences in survival among individuals with obstructive vs. none or non-obstructive CAD; none, 1-vessel, 2-vessel, or 3-vessel obstructive CAD; obstructive vs. none or non-obstructive left main stenosis; and normal or reduced LVEF. Further discrimination of these individuals into the summary variable accounting for CAD and LVF demonstrated incremental reduction in survival for individuals with no obstructive CAD and reduced LVEF; obstructive CAD and normal LVEF; and obstructive CAD and reduced LVEF when compared with those with no obstructive CAD and normal LVEF (Figure 2).

Figure 1

Kaplan–Meier curves illustrating survival for patients with (A) no obstructive vs. obstructive coronary artery disease; (B) 0-, 1-, 2-, or 3-vessel obstructive coronary artery disease; (C) no obstructive vs. obstructive left main stenosis; and (D) normal vs. reduced left ventricular ejection fraction.

Figure 2

Kaplan–Meier curves illustrating survival for patients with no obstructive coronary artery disease and normal left ventricular function; no obstructive coronary artery disease and reduced left ventricular function; obstructive coronary artery disease and normal left ventricular function; and obstructive coronary artery disease and reduced left ventricular function.

Discussion

The primary results of this study demonstrate the prognostic value of 64-detector row CCTA in a large two-centre registry by measures of CAD stenosis severity, extent, and location. Further, we establish the incremental prognostic value of reduced LVEF by CCTA beyond measures of angiographic coronary artery stenosis.

To date, limited evidence exists to support the prognostic significance of CCTA. We previously reported the predictive value of CCTA measures of plaque severity, extent, location, and distribution in a 1127 patient cohort undergoing 16-detector row CCTA for the prediction of all-cause mortality.6 In a larger study with a longer 6.5-year follow-up period, Ostrom et al.7 reported the ability of CTA performed on electron beam CT scanners to predict incident death. Nevertheless, these studies were limited to 16-detector row and electron beam CT scanners, respectively, and large-scale studies with current generation 64-detector row scanners have to date not been reported.

Furthermore, all studies to date that have examined the prognostic value of 64-detector row CCTA have done so at single centres with relatively small sample sizes. In a recent analysis by Carrigan et al.,5 227 individuals were followed for 2.3 years. Coronary computed tomographic angiographic measures of coronary artery stenosis severity were predictive of 18 major adverse cardiac events (MACE); however, only four ‘hard’ coronary events (death or myocardial infarction) were observed. Similarly, Gaemperli et al.10 studied 220 patients undergoing 64-detector row CCTA followed for 14 months, for whom adverse CAD events occurred in 59 patients. Of these events, deaths (n = 4) and myocardial infarction were few (n = 3), and incident CAD events were driven largely by target vessel revascularization. Most recently, van Werkhoven et al.9 examined the incremental prognostic value of anatomic detection of CAD by 64-detector row CCTA in 541 individuals undergoing both CCTA and myocardial perfusion scintigraphy (MPS). In a 1.8-year follow-up, CCTA measures of CAD severity demonstrated incremental prognostic value for the prediction of adverse CAD events above and beyond measures of myocardial perfusion by MPS. Uniform to all of these prior studies is the finding that not only can obstructive CAD prognosticate adverse outcomes, but also the absence of obstructive coronary artery plaque by CCTA confers a very favourable prognosis. In the present study, we identified a similar relationship and those individuals without evident CAD by CCTA experienced a 2.3-year death rate of 1.3% (annualized death rate 0.6%), thus establishing a low-risk group.

Performance of CCTA by retrospective ECG-gated helical acquisition permits assessment and accurate quantification of LVEF with high diagnostic performance. In a recent meta-analysis employing 4–16-detector row CT scanners, excellent agreement was noted between CCTA and cardiac magnetic resonance for measures of LV volumes and function, with <2% differences in LVEF.1113 These high levels of agreement have also been demonstrated for 64-detector row CCTA, although given its limited temporal resolution and non-simultaneous image acquisition of myocardial segments, regional wall motion assessment by CCTA has been less robustly proved. To date, prior studies examining the risk-predictive ability of CCTA have not evaluated the potential incremental value of LVEF. A wealth of evidence indicates the prognostic value of LVEF measures for the prediction of future CAD events by echocardiography, cardiac magnetic resonance imaging (CMR), monoplane ventriculography, and gated single-photon emission computed tomography.14,15 In the present study, reduced LVEF was associated with increased risk of death above and beyond measures of obstructive CAD by CCTA.

Limitations

While this study addresses many of the shortcomings of prior studies examining the prognostic potential of CCTA, it is not without limitations. Although this study included a large cohort of patients from more than one single centre, only two centres were included, and the applicability of the study findings to other centres may not be necessarily generalizable.

Further, the major endpoint of the present study was all-cause mortality. No composite endpoint, which included ‘softer’, albeit still clinically important, cardiac endpoints such as myocardial infarction, unstable angina, or CAD-related hospitalization, was available in our data set. Nevertheless, the assessment of all-cause death as an endpoint is advantageous as it is completely free of ascertainment and verification bias.

The treatment of individuals following CCTA is unknown for the present study. Whether percutaneous or surgical coronary revascularization occurred or whether enhanced medical therapy was implemented in the post-CCTA setting is unknown. Prior data have suggested that in both academic and private practice sites, medical therapy for CAD is newly initiated or increased in direct correlation with CCTA findings.16 Similarly, rates of referral to invasive catheterization also increase the following identification of CAD by CCTA. While both the actions may result in improved or at least equivalent clinical outcomes when compared with no imaging, the potential differential treatments of individuals in this non-randomized registry may have affected the study results. Further, it remains conceivable that higher rates of death for individuals with severe CAD by CCTA may have occurred as a direct complication of revascularization procedures or medical therapy. For proper addressing of this issue, large-scale trials examining prescribed treatment algorithms are needed.

We graded coronary artery lesion severity using mild (1–49%), moderate (50–69%), and severe (≥70%) luminal diameter stenosis categorizations. As 64-detector row CCTA is limited in its spatial resolution to approximately 0.5–0.75 mm, the ability to discriminate lesions of somewhat similar severity (e.g. 60 vs. 75%) is often difficult. Nevertheless, we employed this classification scheme, as it is the one that is commonly employed in clinical practice and has been advocated by current societal guidelines.

Given the non-negligible radiation doses associated with retrospective ECG-gated helical CCTA, many centres now favour prospective axial triggering techniques that reduce radiation but preclude measurement of LVEF. Use of prospective triggering for CCTA can result in effective biological radiation doses, which are approximately 80% lower than conferred by retrospective ECG gating techniques. Importantly, this study does not advocate for the use of retrospective ECG gating, despite the potential prognostic value of the information derived. In fact, the majority of CCTAs performed in the present study occurred at a stage in the evolution of 64-detector row CCTA, which preceded the development of technologies that permit prospective axial triggering. Indeed, in both the CCTA laboratories of the present study, prospectively triggered CCTA has now become the preferred image acquisition protocol. Future studies should examine the incremental cost-to-benefit ratio of the added radiation dose required to assess LVEF to determine whether retrospectively gated CCTA should be performed routinely.

Conclusions and clinical implications

Coronary computed tomographic angiography performed in a large cohort at two centres successfully identified individuals at higher risk of all-cause death at a follow-up of 2.3 years. The presence of obstructive CAD in increasing numbers of vessels or the left main artery portends a particularly poor prognosis. In addition, measures of LVEF add incremental prognostic values above and beyond CAD detection. Importantly, those without obstructive CAD (0.56%) and those with normal LVEF (0.66%) have a low risk of death.

Acknowledgements

This work was supported, in part, by a gift from the Michael Wolk Foundation.

Conflict of interest: J.K.M. receives research support and serves on the Speaker's Bureau for GE Healthcare.

References

View Abstract