This article is more than 5 years old. Information may no longer be current.
Optimal testing strategy for patients with suspected CAD remains unclear
The best strategy for management of low- and intermediate-risk patients with chest pain is an issue on which the cardiology community is divided. That is especially true for the question of which testing modality should be used, or whether any testing modality should be used at all.
New-onset stable chest pain in people without diagnosed heart disease affects approximately 4 million Americans per year, and concerns about the consequences of testing low-risk populations and catheterization of patients without obstructive CAD have been raised.
Further clouding the picture are a lack of large-scale studies linking testing modalities to clinical outcomes. That changed when results of the PROMISE study, which analyzed an initial strategy of coronary computed tomography angiography (CTA) compared with functional testing for symptomatic patients with suspected CAD, were presented at the American College of Cardiology Scientific Sessions in March and later published in The New England Journal of Medicine. The data revealed no significant differences in CV outcomes between patients assigned anatomical testing or functional testing.
This one study generated tremendous interest and controversy in the cardiology community. Through interviews with leading experts, Cardiology Today captured the varied reactions.
Source: Jackson Smith Photography; printed with permission.
“If you’re managing patients with new angina in the age range of 60 years with an average of two risk factors, the clinical outcomes are basically the same whether you use an existing noninvasive functional test or an anatomical imaging test like CT,” Cardiology Today Section Editor W. Gregory Hundley, MD, FACC, FAHA, professor in internal medicine and radiology and medical director of CV imaging at Wake Forest Health Services, Winston-Salem, North Carolina, said in an interview. “To me, these recent results indicate that your practice should utilize the anatomical or functional test which is both appropriate for the indication, and with which your group is proficient in performance. More data are needed to determine cost-effectiveness.”
With no definitive answer as to whether anatomical testing or functional testing is better for low-risk symptomatic patients, cardiologists will have to consider a number of factors when choosing which modality to use — or whether to perform testing at all — including cost, insurance coverage, radiation exposure, existing expertise and whether confirmation of CAD existence or severity of existing CAD is the required information.
Spotlight on PROMISE
When the PROMISE trial was initiated, “studies showing how imaging could affect hard cardiovascular outcomes were sorely needed,” Christopher M. Kramer, MD, FACC, FAHA, Ruth C. Heede Professor of Cardiology and professor of radiology at the University of Virginia Health System, wrote in an editorial accompanying the publication of PROMISE in NEJM.
“The trial was a very good attempt at addressing a crucial question in cardiology surrounding the best approach to patients with suspected CAD,” said João A.C. Lima, MD, MBA, professor of medicine and radiology at Johns Hopkins Medicine and Cardiology Today Editorial Board member.
Pamela S. Douglas, MD, MACC, FASE, FAHA, and colleagues randomly assigned 10,003 low-risk, symptomatic patients who had at least one CV risk factor and whose doctors called for nonurgent CV testing for evaluation of suspected CAD to CTA or functional testing including exercise ECG, nuclear stress testing or stress echocardiography, results of which were used to guide subsequent care decisions. The mean age of the patients was 60.8 years, 52.7% were women and 87.7% had chest pain or dyspnea on exertion.
As reported by Douglas at ACC 2015, results showed no improvement in clinical outcomes between the two testing strategies during a median follow-up of 2 years (see Table).
Reactions to the results and use of the different testing modalities overall were varied.
Kramer told Cardiology Today that the results “tell us that coronary CTA is a reasonable approach in patients ... with suspected coronary disease at low to intermediate risk and is a reasonable alternative to functional stress testing. It shows that CTA should be included in clinical protocols and decision trees, as it has been in appropriate use criteria.”
The low event rate in PROMISE was positive news for many.
“The study followed 10,000 patients with stable chest pain and there was an extremely low event rate,” said Douglas, who holds the Ursula S. Geller professorship for cardiovascular diseases at Duke Clinical Research Institute and Duke University School of Medicine. “That has huge implications for clinical care and testing strategies.”
In an interview with Cardiology Today, Steven E. Nissen, MD, FACC, said he thinks the results should be interpreted as negative and cast doubt on the appropriateness of CTA in this population.
Source: Douglas PS, et al. N Engl J Med. 2015;doi:10.1056/NEJMoa1415516.
“The PROMISE trial basically showed that CTA offers no value for patients for assessment of possible ischemic heart disease. The study was, in my view, almost entirely negative,” said Nissen, Cardiology Today Editorial Board member and chairman of the Robert and Suzanne Tomsich Department of Cardiovascular Medicine, Sydell and Arnold Miller Family Heart and Vascular Clinic, Cleveland Clinic. “The investigators prespecified they would test for superiority, and it failed that. They also prespecified that they would test for noninferiority; they were looking for no worse than a 10% worse outcome with CTA compared with conventional testing, and it failed noninferiority. I would advise people to avoid performing CTA in patients with suspected coronary disease.”
Moreover, the results also suggest that functional testing has no benefit in this specific population, according to Nissen. “When you have a rate of events that is this low, you can’t make a case that functional testing is providing any value. We have now moved to the point where patients with suspected angina can be treated medically without ischemia imaging of any kind, as we learned from the COURAGE trial. I don’t think there is value provided by doing so,” he said.
Other recent research
Other smaller-scale trials have examined potential benefits for CTA in low-risk patients with new-onset angina.
In 2012, two NEJM studies reported benefits for CTA as a triage mechanism in the ED. Litt and colleagues randomly assigned 1,370 low- to intermediate-risk patients presenting to the ED with possible ACS to undergo coronary CTA. The CTA group had a higher rate of discharge from the ED (49.6% vs. 22.7%), shorter length of stay (median, 18 hours vs. 24.8 hours; P < .001) and higher rate of CAD detection (9% vs. 3.5%). Hoffmann and colleagues randomly assigned 1,000 patients presenting to the ED with symptoms suggestive of ACS but without ischemic electrocardiographic changes or an initial positive troponin test to early coronary CTA or to standard evaluation. Those assigned CTA had a reduced length of stay compared with controls by 7.6 hours (P < .001) and were more likely to be discharged directly from the ED (47% vs. 12%, P < .001).
The SCOT-HEART study also was presented at ACC 2015 and simultaneously published in The Lancet. It included 4,146 patients referred for assessment of suspected angina due to CHD randomly assigned standard care plus coronary CTA or standard care alone. Assignment to CTA was more likely to change planned investigations (15% vs. 1%; P < .0001) and treatments (23% vs. 5%; P < .0001), but did not affect symptom severity at 6 weeks or subsequent admittances for chest pain. At 1.7 years, assignment to CTA was associated with a nonsignificant reduction in fatal and nonfatal MI (26 vs. 42; HR = 0.62; 95% CI, 0.38-1.01).
According to Lima, the SCOT-HEART study showed a “very positive outcome for introducing CTA in the algorithm” and “when taken together with the PROMISE trial, it supports that CTA can be an appropriate option in these patients.”
Douglas noted that SCOT-HEART participants in the CTA group had both exercise testing and CTA, which is not the preferred pathway in U.S. clinical practice. “Most people would like to have one test that is definitive, and not to commit themselves to doing two tests,” she said.
In the FACTOR-64 trial, Muhlestein and colleagues studied 900 patients with diabetes and no symptoms of CAD who were randomly assigned CAD screening with CTA or standard diabetes care. At 4 years, the primary outcome of all-cause mortality, nonfatal MI or unstable angina requiring hospitalization was similar between the CTA and control groups (6.2% vs. 7.6%; HR = 0.8; 95% CI, 0.49-1.42). In addition, the CTA group had higher rates of diagnostic coronary angiography (13.3% vs. 5.1%) and PCI (6% vs. 1.8%). The results were published in JAMA in December 2014.
Advantages, disadvantages
Given disparities in the literature, cardiologists must carefully consider the advantages and disadvantages of the testing modalities when deciding on a strategy for low-risk patients, or when deciding to use any test at all.
“If I have a relatively low-risk patient and want to know whether CAD is present, coronary CTA defines that very well. The functional tests, for me, become somewhat more useful when I know that the disease is present and I want to know its importance, because it’s going to prompt me to move down a particular care pathway,” Hundley said. However, a technology that does both extremely well does not yet exist. “Whether that will end up being an MRI technology or a combined CT/nuclear imaging technology, we don’t yet know,” he said.
Other advantages of CTA include the ability to identify mild nonobstructive CAD, its wide availability worldwide and easy interpretation. “It is very easy to train people to use CTA as a first diagnostic strategy,” Lima said.
However, there are several disadvantages of CTA compared with functional testing, including difficulty interpreting results in older patients who may have extensive coronary calcification and lack of the kind of prognostic information that is obtained with the treadmill portion of a stress test, Kramer said.
Concerns of radiation
“Radiation should play a major role in influencing what test should be chosen. Particularly, the younger the person, radiation in the mind of the physician should be of paramount importance in choosing the test,” Lima told Cardiology Today.
CTA has less radiation than nuclear stress testing, but exercise ECG and stress echocardiography use no radiation. Hundley noted that radiation levels associated with CT technology have declined in recent years and efforts to further decrease levels are being made.
In PROMISE, median cumulative exposure per patient was lower in the CTA group, but overall exposure was higher because one-third of patients assigned functional testing were not exposed (see Table).
Hundley said physicians must think about this issue in terms of the patient’s entire medical history. “I don’t think about it as a test in isolation; I think about what the patient will be exposed to, the other comorbidities and where I might send them. It may not just be a single radiation exposure that you’re dealing with in many of these patients,” he said.
Confounding the issue, Kramer said, is a lack of studies on the risk for adverse outcomes from radiation exposure. “This sense that a single CT scan is going to lead to cancer down the road is not based on hard data. I think we just don’t understand the risk of radiation from CT or stress single-photon emission computed tomography (SPECT),” he said.
Cost considerations
Whether the results of PROMISE or other studies will prompt payers to change coverage policies is not yet known.
“The next studies starting to look at utilization and expenses and cost will be very informative, especially as we move to ... accountable care organizations, and whether one modality will be stressed over others will start to depend on costs associated,” Hundley said.
As one example, an economic substudy of the PROMISE trial was also presented at ACC 2015. Researchers estimated initial testing costs at $404 per person for CTA, $501 to $514 per person for stress echocardiography, $174 per person for exercise ECG and $946 to $1,132 per person for nuclear stress testing. After calculating total costs by intention to treat and mean cost difference, total costs per person for CTA were slightly more but not significantly different than costs for functional testing: $279 at 3 months, $358 at 12 months, $388 at 24 months and $694 at 36 months. The jump from 24 months to 36 months was primarily explained by some outlier patients in the CTA group who required non-CV care, and the other differences were explained by increased revascularization in the CTA group, Daniel B. Mark, MD, MPH, professor of medicine at Duke Clinical Research Institute, said during a press conference.
“The results of one study haven’t revolutionized practice, in part because there are economic and political barriers to the use of CTA more broadly,” Kramer said. “Depending on one’s location, insurance may not pay for CTA for this indication. The hope would be that with the results of studies such as PROMISE, insurers might change their mind about the utility of this test. But things are slow to change.”
Nissen said he expects such barriers to remain in place. “Most payers don’t [cover] it now, and I think they will not start paying for it, certainly not with a result that is a bad outcome.”
According to Hundley, the major impact on cost and utilization will come from treatment of high-risk patients, not those at low risk. “Patients with existing disease are where the high utilization is, and applying all these tests to them is where it becomes more problematic,” he said.
More research needed
A number of avenues warrant further research, experts told Cardiology Today.
“The ideal study would be a three-pronged study in PROMISE-like patients with COURAGE-like design except with diagnostic testing. Take low-risk patients, randomize them to optimal medical therapy, or optimal medical therapy plus stress testing with ECG and imaging, or optimal medical therapy plus CTA,” Kramer said.
“There is some validity in the no-testing strategy, but it should be emphasized that we have not yet tested that hypothesis in a clinical trial since all the patients in PROMISE were tested. Maybe there is more equipoise now than there was at the time we started PROMISE,” Douglas said. “We really wanted to have that no-testing arm, but we didn’t feel that people would be comfortable not testing.”
Several experts said the pre-test probability of CAD plays a major role in determining diagnostic strategy, and noted that more research related to that could be useful.
“A triage using coronary artery calcium scoring, as recommended by the British National Institute for Health and Care Excellence (NICE) guidelines for those with a pre-test probability of less than 30%, as the initial test is an interesting proposal that they came to very thoughtfully,” Douglas said.
According to Hundley, “the next specific areas of research are to look at pre-test probabilities and start to match up these modalities with testing of the outcomes for evaluating different groups.”
In addition, the examination of newer stress imaging methods, such as magnetic resonance imaging and positron-emission tomography, should also be a focus of research.
Kramer also said the ISCHEMIA trial, in which randomized invasive or medical therapy is driven by the presence of extensive ischemia on functional stress testing, may provide further answers.
“The cardiovascular imaging field is delivering comparative-effectiveness studies with results that are likely to change clinical practice. This is indeed a PROMISE kept,” Kramer wrote in NEJM.– by Erik Swain
- References:
- Boden WE, et al. N Engl J Med. 2007;doi:10.1056/NEJMoa070829.
- Douglas PS, et al. N Engl J Med. 2015;doi:10.1056/NEJMoa1415516.
- Hoffmann U, et al. N Engl J Med. 2012;doi:10.1056/NEJMoa1201161.
- Kramer CM. N Engl J Med. 2015;doi:10.1056/NEJMe1501924.
- Ladapo JA, et al. Ann Intern Med. 2014;161:482-490.
- Litt HI, et al. N Engl J Med. 2012;doi:10.1056/NEJMoa1201163.
- Mark DB, et al. Joint ACC/JAMA Late-Breaking Clinical Trials. Presented at: American College of Cardiology Scientific Sessions; March 14-16, 2015; San Diego.
- Muhlestein JB, et al. JAMA. 2014;doi:10.1001/jama.2014.15825.
- The SCOT-HEART Investigators. Lancet. 2015;doi:10.1016/S0140-6736(15)60291-4.
- For more information:
- Pamela S. Douglas, MD, MACC, FASE, FAHA, can be reached at 2400 Pratt St., Durham, NC 27715; email: pamela.douglas@duke.edu.
- W. Gregory Hundley, MD, FACC, FAHA, can be reached at Cardiovascular Medicine Division, Wake Forest Health Sciences, 1 Medical Center Blvd., Winston-Salem, NC 27157; email: ghundley@wakehealth.edu.
- Christopher M. Kramer, MD, FACC, FAHA, can be reached at P.O. Box 800170, Charlottesville, VA 22908; email: ckramer@virginia.edu.
- João A.C. Lima, MD, MBA, can be reached at Blalock 524, Johns Hopkins Hospital, 600 N. Wolfe St., Baltimore, MD 21287; email: jlima@jhmi.edu.
- Steven E. Nissen, MD, FACC, can be reached at Cleveland Clinic Main Campus, Mail Code J2-3, 9500 Euclid Ave., Cleveland, OH 44195; email: nissens@ccf.org.
Disclosures: Douglas is principal investigator on a CT fractional flow reserve study sponsored by HeartFlow. Hundley is co-founder of Albus Inc. and Prova, reports receiving grants from the NIH and Siemens, and reports conducting studies for which contrast is donated by Bracco Diagnostics. Lima reports receiving grant support from Toshiba Medical Systems. Kramer and Nissen report no relevant financial disclosures.