American College of Cardiology
American College of Cardiology
March 28, 2018
4 min read

Imaging modalities offer different capabilities, benefits for identifying HF

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

Benjamin J. W. Chow

ORLANDO, Fla. — Cost-effectiveness of new strategies should be taken into consideration when determining the best imaging modality to identify HF in patients, according to two presentations at the American College of Cardiology Scientific Session.

Imaging methods

In the CTA-HF trial, Benjamin J. W. Chow, MD, professor of medicine and radiology at the University of Ottawa in Canada, and colleagues conducted a randomized controlled trial that enrolled patients from 11 centers in Canada and Finland. Patients had left ventricular dysfunction, HF, NYHA class II to IV symptoms or recent admission to a hospital or ED for HF. Those with significant renal insufficiency were excluded from the trial.

Patients were randomly assigned to coronary CTA (n = 120; mean age, 58 years; 72% men) or invasive coronary angiography (n = 126; mean age, 58 years; 71% men).

The primary outcome measure of interest was cost at 12 months. Researchers also calculated quality-adjusted life-years at 12 months and incremental cost-effectiveness ratio.

Patients were followed for approximately 550 days.

Most baseline characteristics were similar in both groups, although more patients assigned invasive coronary angiography had dyslipidemia and diabetes compared with patients assigned coronary CTA. Medication use was also similar in both groups, except for beta-blockers, which were more common in the coronary CTA group. Creatinine and LV ejection fraction were similar in patients assigned coronary CTA vs. invasive coronary angiography.


The prevalence of obstructive CAD and the severity of CAD were similar in both groups. The number of patients with an ischemic HF etiology were similar in both groups in the intention-to-diagnose and the as-tested analyses.

Patients who underwent coronary CTA received 26.9% lower radiation exposure compared with those who underwent invasive coronary angiography.

During follow-up, 30 patients in the coronary CTA group were referred for invasive coronary angiography. Both groups had similar rates of ischemia and viability testing.

No significant differences were seen in both groups for quality of life at baseline and at 12 months and various composite clinical outcomes.

Cardiac death occurred more often in patients who underwent coronary CTA compared with those who underwent invasive coronary angiography (6 vs. 1; P = .0499). Three of four patients in the coronary CTA group who died from HF had nonischemic HF.

Initial costs were lower for coronary CTA compared with invasive coronary angiography ($496 vs. $2,562; P = 1). Total cost was approximately $800 lower in those who received coronary CTA (P = .69).

In the intention-to-diagnose analysis, the coronary CTA group had fewer QALYs vs. the invasive coronary angiography group (0.847 vs. 0.88; P = .04), which equates to 12 more days of perfect health, according to the presentation. Coronary CTA was more cost-effective than invasive coronary angiography, and the incremental cost-effectiveness ratio was $26,032 per QALY for coronary CTA.

The as-tested analysis also found that total cost was lower in the coronary CTA group compared with the invasive coronary angiography group ($6,524 vs. $9,456; P = .95), although there was no difference in QALYs between the two groups.

“For our population, we conclude that patients with HF with unknown etiology, an invasive coronary angiography as a first strategy might be more cost-effective than CT,” Chow said. “In situations where patient and physician preference are factors in test selection, this might be a situation where CT may be cost-effective.”


In a separate presentation, routine cardiac MR did not improve the clinical diagnosis of nonischemic HF.

Ian Paterson, MD, FRCPC, cardiologist at University of Alberta in Edmonton, Canada, and colleagues analyzed data from 518 patients with newly diagnosed or established HF, a working diagnosis of nonischemic cardiomyopathy or HF with preserved ejection fraction and NYHA class II to IV symptoms within the last 12 months. Exclusion criteria included factors such as significant CAD, previous STEMI or non-STEMI, prior CMR without change in clinical condition and life expectancy less than 3 months.

Patients were randomly assigned to routine (n = 258; mean age, 59 years; 70% men) or selective (n = 260; mean age, 58 years; 68% men) CMR with echocardiography.     

“In this case, selective CMR was allowed for a clinical suspicion of [arrhythmogenic right ventricular cardiomyopathy], infiltrative cardiomyopathy, congenital or pericardial disease,” Paterson said. “This was based on Canadian guidelines at the time.”

Patients were followed for up to 12 months. During clinical assessments at 3 and 12 months, the treating team determined whether patients had either a specific or a nonspecific HF etiology according to prespecified criteria. The imaging interpreters for CMR and echocardiography also determined the HF etiology based on imaging findings.

The primary outcome was the clinical assessment of HF etiology at 3 months. Secondary outcomes included imaging-based HF etiology and clinical events, including death or CV hospitalization.

Baseline patient characteristics and medications were similar in both groups. Before any imaging was performed, clinicians identified a specific HF etiology in 36.9% of patients from the routine arm and 38% of patients from the selective arm (P = .88).

Most patients in the routine arm underwent CMR (n = 224), and in the 54 patients from the selective arm who underwent CMR, only three were performed per protocol for investigation of infiltrative or adult congenital heart disease.

Imaging identified a specific HF etiology in 34.1% of patients in the routine group and 29.7% of patients in the selective group (P = .27). When analyzed per protocol, a specific HF etiology was identified in 25.7% of patients (P = .04 vs. routine) from the selective group.

In the routine arm, more patients were identified as having a specific HF etiology with CMR compared with echocardiography (35.7% vs. 20.1%; P < .001).

“The test itself does seem to provide more specific diagnoses,” Paterson said.

However, the primary outcome was negative at 3 months, as the number of patients with specific HF etiologies as determined by the treating team did not significantly differ in the routine and selective groups (44.4% vs. 49.8%). This was also seen at 12 months (P = .14).

Survival was also not significantly different between the two groups at 12 months. In a post hoc analysis, survival did not differ in patients with specific and nonspecific HF etiologies determined clinically (P = .65), although patients with nonspecific HF etiologies identified through imaging had improved 12-month survival compared with those with specific HF etiologies (P = .02).

“We interpret these findings to mean that CMR does increase specific imaging diagnoses, but does not change specific clinical diagnoses,” Paterson said. “Imaging-based diagnoses enables stratification of risk. Greater attention to the use of CMR and HF diagnoses from imaging in general should be considered.”

Both studies were part of the IMAGE-HF project. – by Darlene Dobkowski


Chow BJW, et al.

Paterson I, et al. Featured Interventional Clinical Research III. Both presented at: American College of Cardiology Scientific Session; March 10-12, 2018; Orlando, Fla.

Disclosures: IMAGE-HF was supported by an unrestricted research grant from GE Healthcare. Chow reports he received research support from CV Diagnostix and other support from TeraRecon. Paterson reports no relevant financial disclosures.