Author + information
- Kevin E. Boczar, BSc,
- Mohammed Alam, MD,
- Benjamin J.W. Chow, MD∗ ( and )
- Girish Dwivedi, MD, PhD
- ↵∗University of Ottawa Heart Institute, 40 Ruskin Street, Ottawa, Ontario, K1Y 4W7, Canada
Identifying characteristics that predict future adverse events can be used to facilitate patient monitoring or therapy. Although, computed tomographic coronary angiography (CTCA) measures of coronary atherosclerosis, coronary artery disease severity, and left ventricular (LV) ejection fraction are prognostic markers of major adverse events, other potential measures have not been fully explored (1). Previous studies have demonstrated that left ventricular end-diastolic volume (LVEDV) has incremental prognostic value (2). However, the adoption of prospective electrocardiogram (ECG)-triggered CTCA has eliminated the ability to directly measure LVEDV. It has previously been shown that LVEDV can be estimated using LV and left atrial (LA) volumes obtained during ventricular diastasis (3). The prognostic value of estimated LVEDV (LVEDVEstimated) is unknown. The objective of our study was to validate the predictive model for LVEDV and determine the incremental prognostic value of LVEDVEstimated.
Using the University of Ottawa Heart Institute Cardiac CT Registry, 97 patients who experienced major adverse events (all-cause mortality and nonfatal myocardial infarction [MI]) were identified (1). A matched (Morise score) control population of 98 patients was randomly selected from the same registry. The details of the CTCA procedure have been described previously (1). In brief, a triphasic intravenous contrast administration protocol was used for final image acquisition. Retrospective ECG-gated datasets were acquired with the GE Volume CT (GE, Milwaukee, Wisconsin) with 64 × 0.625 mm slice collimation. The 5% to 95% phases (10% increments) were reconstructed using a slice thickness of 1.25 mm and an increment of 0.625 mm. CTCA images were post-processed using the GE Advantage Volume Share Workstation and interpreted by expert observers blinded to all clinical data. LV and LA volumes were measured using the 75% phase and at end-diastole. Using a previously developed model for LVEDVEstimated, LVEDVEstimated was calculated from LV and LA volumes at the 75% phase [LVEDV = (1.021 × LV75% phase volume) + (0.259 × LA75% phase volume)] (3). LVEDVEstimated was indexed (LVEDVIEstimated) to the body surface area for analysis.
Continuous variables were expressed as mean ± SD and categorical variables as proportions or percentages. Statistical significance was defined as p < 0.05. The Student t test was used to compare continuous variables and the chi-square test was used for categorical variables. The prognostic value of LVEDVIEstimated was assessed using Cox proportional hazard models for the composite of all-cause death and nonfatal MI. A total of 195 consecutive patients were analyzed. The baseline characteristics and the imaging characteristics of the test and control populations were similar (Table 1). Mean follow-up duration was higher (p = 0.008) for the test population than for the control population (450 ± 283 days vs. 757 ± 371 days).
The LV and LA volume indices, at diastasis, were greater in the event cohort than in the control group (Table 1). Using these measures, LVEDVIEstimated was calculated and compared with “true” CTCA-measured LVEDV. The correlation between measured LVEDVI and LVEDVIEstimated was very good (r = 0.910, 95% confidence interval [95% CI]: 0.882 to 0.931; p < 0.0001). LVEDVIEstimated was significantly larger in the event population than in the control population (91.7 ± 4.3 ml/m2 and 69.7 ± 2.5 ml/m2, respectively; p ≤ 0.0001).
In a Cox proportional hazards regression analyses, previous coronary artery disease (hazard ratio [HR]: 1.89, 95% CI: 1.248 to 2.865; p = 0.003) and LVEDVIEstimated (HR: 1.008, 95% CI: 1.004 to 1.012; p = 0.001) emerged as univariate predictors of all-cause death and nonfatal MI. However, when multivariate survival analysis was performed using significant univariate predictors as variables, only LVEDVIEstimated (HR: 1.01, 95% CI: 1.003 to 1.012; p = 0.02) emerged as a predictor of all-cause death and nonfatal MI.
Previous studies have shown that LVEDV assessed by different cardiac imaging modalities has independent and incremental prognostic value in identifying patients at high risk of cardiac events (2,4). However, to the best of our knowledge, the incremental prognostic ability of LVEDV assessed by CTCA has not been demonstrated before. Moreover, to minimize patient radiation exposure, prospective ECG-triggered CTCA has been adopted into routine clinical practice. Because prospective ECG-triggered CTCA does not permit the direct assessment of LV ejection fraction and LVEDV, the ability to estimate LVEDV would be clinically desirable.
Our group has previously developed a model for estimating LVEDV from prospective ECG-gated CTCA datasets. In the current study, we were able to validate this model and additionally demonstrate the prognostic value of LVEDV measurements from this predictive formula.
Please note: Dr. Chow holds the Saul and Edna Goldfarb Chair in Cardiac Imaging; receives research and fellowship training support from GE Healthcare; and receives educational support from TeraRecon Inc. Dr. Dwivedi is supported by Banting post-doctoral fellowship, Dowager Countess Eleanor Peel trust (Rothwell-Jackson travelling fellowship [United Kingdom]) and Whit and Heather Tucker cardiology endowed research fellowship grants (Canada). All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation
- Chow B.J.,
- Wells G.A.,
- Chen L.,
- et al.
- Sharir T.,
- Germano G.,
- Kavanagh P.B.,
- et al.
- Khatri P.J.,
- Tandon V.,
- Chen L.,
- Yam Y.,
- Chow B.J.