Abstract 5079: Predictive Accuracy of Cardiopulmonary Exercise Stress Testing in Subjects With Chest Pain
Cardiopulmonary exercise testing (CPET) is widely used to grade the severity of heart failure and to assess its prognosis. However, it is unknown whether CPET may identify coronary artery disease (CAD) or exclude it in patients with chest pain. We prospectively studied 1,265 consecutive patients (55±8 years, 156 women, 628 retired from work, 234 white collar, 244 blue collar) who were evaluated with ECG stress testing (TE) for chest pain. No one had a previous myocardial infarction or documented CAD. All patients performed an incremental CPET on an electronically-braked cycle ergometer (Ergometrics 800S) until volitional fatigue. ECG was recorded during the test. Cardiovascular risk factors were present in 65%. Criteria for a positive CPET were: flattening in VO2/work rate slope with an inflection point more than 1 minute earlier than peak exercise and a flattened slope from the inflection point to peak exercise (<4 ml/min/W); the time from the inflection point to peak exercise should be similar in O2pulse. These two last criteria have been recently validated by our group in 202 patients with documented CAD (R.Belardinelli et al, Eur Heart J, 2003;24:1285). Follow-up lasted 12 months. Of 1,265 patients, 73 had a positive CPET and 1,192 had a negative CPET. Cardiac events occurred in 32 patients with a positive CPET and 10 out of 1,192 (positive predictive accuracy(PPV): 76%; negative predictive accuracy (NPV): 95%). Coronary angiography was performed in 73 patients with a positive CPET and revealed a CAD in 64. As compared with ET, sensitivity, specificity, PPV and NPV were all improved (ET: 48%,55%,33%,95%; CPET: 88%, 98%, 73%,99%, respectively). Patients with a peak VO2≥91% of predicted VO2max had no CAD in 100% of cases. In conclusion, CPET shows a predictive accuracy for CAD greater than TE in patients with chest pain. Its use should be encouraged among physicians in clinical practice.