Coronary collateral circulation and diastolic function

Comparison of treadmill exercise echocardiography before and after exercise in the evaluation of patients with known or suspected coronary artery disease

OBJECTIVES

We sought to compare the feasibility and accuracy of peak treadmill exercise echocardiography versus postexercise echocardiography imaging.

BACKGROUND

Although peak exercise echocardiography has been reported for both supine and orthostatic bicycle exercise and has shown higher sensitivity than postexercise imaging, acquiring images at peak exercise with treadmill has not been explored.

METHODS

Peak and post-treadmill exercise echocardiography and coronary angiography were performed on 89 patients with known or suspected coronary artery disease. Positive exercise echocardiography was defined as necrosis or ischemic response. Positive coronary angiography was defined as >/=1 diseased vessels (>/=50% luminal narrowing). Images were analyzed in a blind manner by an expert observer.

RESULTS

Postexercise images were acquired within 80 seconds after exercise (40 +/- 14). Mean heart rate (bpm) was 139 +/- 22 at peak versus 118 +/- 25 at postexercise imaging (P <.001). Interpretable peak and postexercise images were obtained for all 89 patients. Of the 72 classified as having positive exercise echocardiography, 23 had new regional wall motion abnormality at peak (21 with positive angiography), which resolved at postexercise imaging. Sensitivity was higher with peak than with postexercise imaging (94% vs 73%, P <.001). Specificity was similar (68% vs 79%), as was predictive positive value (92% vs 93%). Negative predictive value was again higher with peak imaging (76% vs 44%, P <.05). Total accuracy was higher with peak imaging (89% vs 74%, P <.05).

CONCLUSIONS

Peak treadmill exercise echocardiography is technically feasible and has higher sensitivity and accuracy than post-treadmill exercise echocardiography. Therefore in the clinical setting peak exercise echocardiography should be performed to diagnose ischemia.

Patients with coronary collaterals and normal left ventricular systolic function: clinical, hemodynamic, and angiographic characteristics

One hundred and twenty consecutive patients with significant coronary artery disease, normal left ventricular systolic function, and coronary collaterals were compared with 111 patients with the same characteristics but with no collaterals. No significant differences were found between the two groups in hypertension, diabetes mellitus, and smoking. The left ventricular end diastolic pressure was 16.4+/-7 in the study group and 16.9+/-6.9 in the controls (NS). Significantly more diseased vessels were observed in the study group than in the control group (2.1+/-0.6 versus 1.7+/-0.6, p=<0.001). One hundred and one totally occluded vessels were found in the study group but only two in the control group. The richest collateral supply was to the right coronary artery: 94 sources to 85 diseased vessels (111%) including 66 sources to 52 totally occluded arteries (127%); to the left anterior descending: 59 sources to 89 diseased vessels (66%) including 37 sources to 33 totally occluded arteries (112%). The poorest supply was to the left circumflex: 17 sources to 69 diseased vessels (25%), including nine sources to 16 totally occluded arteries (56%). No collaterals were observed in 14 totally occluded vessels in the study group and in two of the controls, while the systolic function at rest was still normal. It is suggested that coronary collaterals are important in preserving left ventricular systolic and diastolic performance at least at rest. Not readily visible collaterals may also prevent systolic dysfunction.