Researchers have developed and validated a weighted scoring system that incorporates patient- and scan-related factors to predict pre-coronary CT angiography (CCTA) risk of an uninterpretable result in symptomatic patients, according to a study published online July 26 in Circulation: Cardiovascular Imaging.
While CCTA is the fastest growing cardiovascular imaging modality in the U.S., its utility hinges on high-quality images of the entire coronary tree. Scans with one or more uninterpretable coronary segments may preclude accurate assessment and create uncertainty regarding severe coronary artery disease, ultimately leading to a nondiagnostic study and potentially spurring additional testing.
Thomas E. Vanhecke MD, department of cardiovascular medicine at Genesys Regional Medical Center in Grand Blanc, Mich., and colleagues developed and tested a scoring system to identify factors associated with uninterpretable CCTA results.
The researchers retrospectively identified the strongest variables associated with the presence of at least one major interpretable coronary segment on the CCTA scans of 8,585 symptomatic patients (mean age 55.6 years; 47.8 percent men) participating in the Advanced Cardiovascular Imaging Consortium multicenter CCTA registry between July 2, 2007, and June 30, 2009.
These data were used to define the uninterpretable risk score (URS), which was then prospectively applied to a validation cohort of 915 patients (mean age 55.3 years; 47.2 percent men) between July 1, 2009 and Oct. 28, 2009.
Although the primary outcome of the study was the presence of at least one uninterpretable segment, Vanhecke et al emphasized that such studies are not necessarily rated nondiagnostic and thus they selected frequency and predictors of “complete nondiagnostic” CCTA results as a secondary outcome.
Although coronary artery calcium scores (CACS) and heart rate prior to CCTA were not included in the primary analysis, the researchers conducted an additional analysis of these factors. Finally, they also reviewed data regarding downstream testing and outcome events at three months.
In the final development cohort, comprised of 8,536 scans, 6.4 percent of studies were characterized as completely nondiagnostic. Among scans with reasons provided for the nondiagnostic classification, motion artifacts (63.4 percent), excessive image noise (32.6 percent), excessive calcification (29.9 percent) and poor intravascular contrast (20.6 percent) were cited as the most common reasons for poor results.
The researchers added that 16.3 percent of the 8,585 development scans had at least one uninterpretable coronary segment.
Among the 4,835 patients with coronary artery calcium score, increased scores by group category were linked with higher likelihood of uninterpretable results.
Vanhecke and colleagues identified seven parameters as independent predictors of an uninterpretable coronary segment in the development cohort: diabetes, hypertension, chronic obstructive pulmonary disease (COPD), sedentary lifestyle, BMI (by categories of <30 kg/m 2, 30-40 kg/m 2 and >40 kg/m 2), age (by categories of <65, 65-75 and > 75 years) and history of atrial fibrillation.
Among participants with calcium scores and heart rate data, two additional predictors were found: coronary artery calcium score (by categories of <100, >100 to <400, >400 to <1,000 and >1,000 AU) and heart rate (by categories <60, >60 to <70 and >80 beats per minute).
“In both cohorts, both scores (with and without CACS and heart rate) the frequency of attaining uninterpretable coronary segments increased about 1.5 fold with each four-point rise in the URS,” stated the researchers. In addition, patients in the development cohort with one or more uninterpretable segments also demonstrated increased frequency of subsequent hospitalization, stress testing, invasive coronary angiography, percutaneous coronary intervention, coronary artery bypass grafting and myocardial infarction at three months.
The study, explained Vanhecke and colleagues, extends ROMICAT findings by identifying BMI, sedentary lifestyle and COPD as predictors of uninterpretable segments and offers a successful prospective validation. However, they cautioned, “the optimal score above which CCTA should be avoided and an alternate testing strategy pursued remains unknown.”
Limitations to the study