CT Angiography Bests Functional Testing in Diabetic Patients With Chest Pain: PROMISE

Screening diabetic patients with CT angiography reduces the risk of CV death and MI by 62% compared with functional stress testing, report researchers.

CT Angiography Bests Functional Testing in Diabetic Patients With Chest Pain: PROMISE

For patients with diabetes and stable coronary artery disease, use of CT angiography is associated with fewer adverse cardiovascular events when compared with screening that employs functional stress testing, according to a new analysis of the PROMISE study.

Patients with diabetes who underwent CT angiography had a 62% lower risk of cardiovascular death and MI than those who had functional stress testing (HR 0.38; 95% CI 0.18-0.79), report investigators. Comparatively, there was no difference in clinical outcomes in nondiabetic patients randomized to either of the noninvasive imaging tests.

“Of the 20% of the population who were diabetic, they had a significantly better outcome,” senior investigator Pamela Douglas, MD (Duke University School of Medicine, Durham, NC), told TCTMD. “It was late after randomization, which is what we would expect to the case. . . . It was a huge reduction, 62%, in some pretty hard endpoints, cardiovascular death and MI. We’re not talking about squishy endpoints. The obvious conclusion is that these patients are sicker and the ability to better identify coronary artery stenosis and do something about it in these patients, whether it’s revascularization or statins and aspirin, really can have an impact on subsequent events.”

In the PROMISE trial, which was published in 2015, investigators randomly assigned 10,003 patients with chest pain to anatomical testing with CT angiography or to functional testing with exercise electrocardiography, nuclear stress testing, or stress echocardiography. After 2 years of follow-up, CT angiography was associated with fewer catheterizations showing no obstructive CAD, but there was no difference in the primary composite endpoint of death, MI, hospitalization for unstable angina, or major procedural complications.

Published online February 25, 2019, ahead of print in the Journal of the American College of Cardiology, the latest analysis is based on 2,144 patients with diabetes at baseline. As a whole, individuals in the diabetic subgroup were more likely than nondiabetic patients to have cardiovascular risk factors, such as hypertension, dyslipidemia, depression, and increased body mass index. Among the diabetic patients randomized to CT angiography versus functional stress testing, however, baseline characteristics (including age, sex, and risk factors) were balanced. Among the diabetic patients randomized to functional testing, the majority (73%) underwent nuclear stress testing. 

In the diabetic subgroup, the rate of cardiovascular death/MI at 2 years was 1.1% among those randomized to CT angiography and 2.6% among those randomized to stress testing (P = 0.01). There was no significant difference in outcomes among the nondiabetic patients.

In terms of processes of care, patients with diabetes were more likely than nondiabetic patients to be referred for invasive coronary angiography within 90 days of the noninvasive test. Patients undergoing CT angiography had increased referral to invasive angiography within 90 days, but the presence of diabetes did not modify the relationship between the noninvasive testing modality and referral. A similar finding was observed with referral for revascularization: more patients undergoing CT angiography were referred for revascularization but diabetes did not modify the relationship.

Overall, patients with diabetes in PROMISE were more likely to have a prescription at 60 days for aspirin, statins, a beta-blocker, and an ACE inhibitor/angiotensin receptor blocker (ARB), but again, the presence of diabetes did not modify the relationship between the testing modality and a 60-day prescription for aspirin or a statin.

To TCTMD, Douglas said there is not a lot of difference in the processes of care between diabetic and nondiabetic patients with positive results for CAD on CT angiography or stress testing, noting that the two patient groups go to the catheterization lab in roughly similar rates. However, when the researchers drilled down into the data, they did find that diabetic patients with a positive CT result were nearly 40% more likely to take a statin and 25% more likely to take aspirin at 60 days than were nondiabetic patients with a positive CT scan.

For Douglas, given the higher-risk nature of diabetic patients, CT angiography should be the preferred noninvasive imaging test for those suspected of having CAD.

“We have no indication that CT angiography is the wrong answer,” she said. “We would like more positive indications that it’s the right answer.” While PROMISE showed no difference overall, data from SCOT-HEART, which looked at a higher-risk patient population, and from their latest diabetic subgroup analysis are very much congruent. The higher the risk of the patient, the greater the benefit of CT angiography, said Douglas.   

The 5-year results from SCOT-HEART, which were published in August 2018, showed that the risk of coronary heart disease death or nonfatal MI was reduced by a 41% when CT angiography was used instead of standard care alone to guide the management of patients with chest pain. That difference was driven by a lower rate of nonfatal MI.

David Newby, MD, PhD (University of Edinburgh, Scotland), who led SCOT-HEART, said the reduction in cardiovascular death/MI among the diabetic patients in PROMISE is an intriguing finding in light of the overall trial being neutral. “The fact that they saw a bigger effect in the diabetic patients says several things, I think,” said Newby. “For one, the diabetic patients were much higher risk. They had more risk factors. The problem with PROMISE is that the patient population was relatively low risk [overall], but the diabetics were higher risk, had more events, and arguably more to gain.”

In addition, Newby noted that diabetic patients often present with different symptoms than nondiabetic patients. They often don’t have chest pain and are more likely than nondiabetic patients to have a false-positive result on functional stress testing. “So I think there is rationale for seeing why diabetics might see more of a benefit than nondiabetic patients [with CT angiography],” he said.

Even in SCOT-HEART, which included a smaller percentage of diabetic patients, those with diabetes had a greater reduction in cardiovascular death and MI with CT angiography than did nondiabetic patients, although the P value for interaction was not significant. “The trend was that the diabetic patients seemed to benefit more,” he said.   

‘Preeminent Role of CT Angiography

In an editorial accompanying the study, Michael Blaha, MD, and Miguel Cainzos-Achirica, MD (Johns Hopkins Medical Institutions, Baltimore, MD), write that the “present analysis adds to a growing body of high-quality, high-impact evidence suggesting a preeminent role of coronary CT angiography for the assessment of patients with new-onset chest pain of suspected coronary cause.”

They add that while there was an increase in the early referral for invasive angiography and revascularization in the CT angiography arm, there is limited evidence for improved clinical outcomes with coronary intervention in stable chest pain. Instead, Blaha and Cainzos-Achirica believe the greater use of optimal medical therapy, particularly in high-risk diabetic patients, played the larger role.

“The overall outcome may be the consequence of clinicians starting preventive medications after visualization of coronary plaque, even in the absence of obstructive disease—the frequency of which was very low in PROMISE,” they write. “In the nondiabetic group, which had lower baseline CVD risk, this more intensive pharmacological management was unlikely to make a short-term difference in terms of incident events—longer follow-up would be needed.”

To TCTMD, Newby said that the visualization of CAD on CT angiography—rather than an assessment of the probability of disease with functional stress testing—may drive physicians to treat patients more intensively with medical therapy. “Definitely, we tend to use more appropriate treatments following CT,” he said.

In their editorial, Blaha and Cainzos-Achirica recommend that the next set of US guidelines follow the lead of the United Kingdom’s. The National Institute for Health and Care Excellence (NICE) recommends CT angiography as the first-line investigation for patients with stable chest pain with typical or atypical symptoms, as well as those with nonanginal chest pain with ECG changes suggestive of coronary artery disease. The editorialists say that the value of CT angiography is likely to grow in the coming years and that it is the most cost-effective option for the assessment of patients with new-onset, stable cardiac chest pain. 

Photo Credit: Michael Blaha, MD. Routine coronary CT angiography showing minimal calcified plaque in the left anterior descending artery just after the takeoff of the first diagonal branch.   

Michael O’Riordan is the Associate Managing Editor for TCTMD and a Senior Journalist. He completed his undergraduate degrees at Queen’s…

Read Full Bio
  • Douglas reports grant support from HeartFlow and serving on a data and safety monitoring board for GE Healthcare.
  • Newby was the lead investigator of SCOT-HEART. He reports receiving research grants from Siemens outside of the trial
  • Blaha and Cainzos-Achirica report no relevant conflicts of interest.