Screening for coronary artery disease (CAD) in any population is controversial. That point was illustrated by this week's Detection of Ischemia in Asymptomatic Diabetics (DIAD) study in the Journal of the American Medical Association.
In the DIAD study article in JAMA (2009, Vol. 301:15, pp. 1547-1555), Young and colleagues report the results of the trial of 1,123 middle-aged diabetics, who were randomized to receive SPECT myocardial perfusion imaging (MPI) versus usual care. They showed that the event rates were low (0.6% per year) and were not reduced in the SPECT MPI group.
The problem of this underpowered study (the entire study population experienced a total of only 32 cardiovascular hard events, defined as heart attack or cardiovascular death) is that the authors screened the wrong patients with the wrong test. A 0.6% per year event rate fits into the standard definition of a low-risk group. Why, then, test them with a moderately expensive test?
A recent consensus statement of the American Diabetes Association (ADA) clearly points out the lack of a need for such screening of diabetic patients (Bax et al, Diabetes Care, 2007, Vol. 30:10, pp. 2729-2736). But should diabetics have any testing to determine whether they are in particularly high- or low-risk groups? The consensus statement goes on to say that nonimaging tests can be helpful in identifying high-risk cohorts who may benefit from an imaging test. Then the question arises as to which test to use in such patients.
A reasonable test is the CT coronary calcium scan. It is a noninvasive test performed with a very small amount of radiation and without an injection of contrast. The amount of calcium in the coronary arteries is related to the coronary artery plaque burden and has consistently been shown in numerous studies to be more predictive of cardiac events than the standard physical examination and blood testing done in the doctor's office.
What DIAD showed us is that screening for myocardial ischemia with the SPECT procedure reported in the study is not cost-effective. From the outset, the study was not adequately powered to look at event rates in this population.
In addition, the patients were not chosen to be diabetics at high risk. The researchers did find that 8% of the diabetics had at least moderate ischemia on SPECT MPI scans, but this is too low a prevalence to allow ischemia testing to be cost-effective.
Furthermore, any benefit of testing that might be present in the higher-risk patients may have been undetectable due to the low proportion of patients with these findings. In this regard, the study did show that in the patients with at least moderate ischemia, the hard cardiac event rate was 12%, compared with 2% in the normal scan group.
However, filtering out patients with screening using coronary calcium scanning (CCS) may well be cost-effective and could identify a high-risk group who then would be candidates for SPECT imaging.
With respect to CCS in diabetics, the work of Raggi et al (Journal of the American College of Cardiology, 2004, Vol. 43:9, pp. 1663-1669) is particularly illuminating. In a large cohort, these investigators demonstrated that for any degree of coronary calcification, diabetics had higher mortality risk than nondiabetics. Importantly, 39% had no or minimal coronary calcium, and their mortality rate was no different from the nondiabetic patients with these CCS findings. At the other end of the spectrum, diabetics with high scores had high mortality rates.
SPECT MPI is an excellent tool for predicting the likelihood of benefit with revascularization. But to be effective and cost-effective, this test has to be performed in a patient population that is at sufficiently high risk. The 20% of diabetics with CCS scores greater than 400 are such a group.
If Young and colleagues in the DIAD study had begun with diabetics with higher clinical risk, performed CCS as the screening examination, and then only used SPECT MPI in the patients with high CCS scores, the results of the trial may well have shown that SPECT MPI saves lives. Such a hypothesis is worthy of an appropriately powered randomized clinical trial.
In the meantime, such a sequential testing approach -- beginning with an inexpensive physical exam and blood testing, progressing to CCS when high clinical risk is found, and then to SPECT MPI in those at highest clinical risk -- is a reasonable clinical practice, supported by consensus statements and appropriate-use criteria of many prestigious professional organizations.
Dr. Daniel S. Berman
AuntMinnie.com contributing writer
April 17, 2009
Dr. Berman is chief of cardiac imaging and nuclear cardiology at Cedars-Sinai Medical Center's S. Mark Taper Foundation Imaging Center.
Related Reading
Screening type 2 diabetics for asymptomatic CAD not worthwhile, April 15, 2009
Copyright © 2009 AuntMinnie.com
