Elsevier

American Heart Journal

Volume 164, Issue 3, September 2012, Pages 320-326
American Heart Journal

Clinical Investigation
Acute Ischemic Heart Disease
A gender-specific blood-based gene expression score for assessing obstructive coronary artery disease in nondiabetic patients: Results of the Personalized Risk Evaluation and Diagnosis in the Coronary Tree (PREDICT) Trial

https://doi.org/10.1016/j.ahj.2012.05.012Get rights and content

Background

Currently available noninvasive tests to risk stratify patients for obstructive coronary disease result in many unnecessary cardiac catheterizations, especially in women. We sought to compare the diagnostic accuracy of presenting symptoms, noninvasive test results, and a gene expression score (GES) in identifying obstructive coronary artery disease (CAD) according to gender, using quantitative coronary angiography as the criterion standard.

Methods

The PREDICT trial is a prospective multicenter observational study designed to develop and validate gene expression algorithms to assess obstructive CAD, defined as at least one ≥ 50% diameter stenosis measured by quantitative coronary angiography. Patients referred for diagnostic cardiac catheterization with suspected but previously unknown CAD were enrolled. Noninvasive myocardial perfusion imaging (MPI) was available in 60% of patients. The GES, comprising gender-specific age functions and 6 gene expression terms containing 23 genes, was performed for all patients.

Results

A total of 1,160 consecutive patients (57.6% men and 42.4% women) were enrolled in PREDICT. The prevalence of obstructive CAD was 46.7% in men and 22.0% in women. Chest pain symptoms were a discriminator of obstructive CAD in men (P < .001) but not in women. The positive predictive value of MPI was significantly higher in men (45%) than in women (22%). An abnormal site-read MPI was not significantly associated with obstructive or severity of CAD. The GES was significantly associated with a 2-fold increase in the odds of obstructive CAD for every 10-point increment in the GES and had a significant association with all measures of severity and burden of CAD. By multivariable analysis, GES was an independent predictor of obstructive CAD in the overall population (odds ratio [OR] 2.53, P = .001) and in the male (OR 1.99, P = .001) and female (OR 3.45, P = .001) subgroups separately, whereas MPI was not.

Conclusions

Commonly used diagnostic approaches including symptom evaluation and MPI performed less well in women than in men for identifying significant CAD. In contrast, gender-specific GES performed similarly in women and men. Gene expression score offers a reliable diagnostic approach for the assessment of nondiabetic patients and, in particular, women with suspected obstructive CAD.

Section snippets

Patient selection and definitions

The PREDICT trial is a prospective multicenter observational study designed to develop and validate a gene expression algorithm to assess the likelihood of obstructive CAD.4 Subjects were eligible for enrollment in PREDICT if they were older than 20 years; had a history of chest pain, anginal symptoms suggesting myocardial ischemia, unstable angina, or asymptomatic with a high risk of CAD with no known prior CAD. Patients with known CAD, New York Heart Association class III or IV, left

Results

The development and validation cohorts from the PREDICT trial comprised 1,166 consecutive nondiabetic patients. Coronary angiograms of suitable quality for quantitative analysis were available in 1,160 of the 1,166 patients in these cohorts. Baseline demographics demonstrated that men and women had similar risk factors including age, body mass index, race distribution, and incidence of hypertension and dyslipidemia (Table I). Women were less likely to be on aspirin therapy but more likely to be

Discussion

This study demonstrates that in a contemporary US-based population referred to diagnostic angiography for suspected CAD, only 36.2% of patients and only 22.0% of women had obstructive CAD with a characteristic age-dependent rise in the prevalence of CAD after 60 years of age for women. Although symptoms on presentation were helpful in identifying obstructive CAD in men, symptoms of angina as defined by Diamond and Forrester5 were not helpful in predicting CAD in women. In this population

Conclusion

The use of a blood-based GES may be particularly helpful in the assessment of obstructive CAD in nondiabetic patients and, in particular, women for whom the use of symptoms and functional testing has proven unreliable. Further studies are needed to validate whether this test or other methods that use individualized genomic data will help promote more efficient and appropriate use of coronary angiography in women.

References (29)

  • M.R. Patel et al.

    Low diagnostic yield of elective coronary angiography

    N Engl J Med

    (2010)
  • S. Rosenberg et al.

    Multi-center validation of the diagnostic accuracy of a blood-based gene expression score for assessment of obstructive coronary artery disease in non-diabetic patients

    Ann Int Med

    (2010)
  • G.A. Diamond et al.

    Probability of CAD

    Circulation

    (1982)
  • R.D.C. Team

    A language and environment for statistical computing

    (2007)
  • Cited by (0)

    PREDICT: www.clinicaltrials.gov, NCT00500617.

    h

    For the PREDICT investigators

    View full text