Objectives. The purpose of this study was to test the hypothesis that endothelial dysfunction occurs in humans before the development of structural coronary atherosclerosis when risk factors for this disease are present.

Background. Animal studies have demonstrated that known risk factors for coronary atherosclerosis (hyperlipidemia, hypertension, diabetes) result in impaired endothelium-dependent vascular reactivity before the development of structural atherosclerosis. Previous studies in patients have been unable to distinguish early structural atherosclerotic disease from dysfunctional endothelium.

Methods. Twenty-six patients with angiographicalty normal coronary arteries were studied at cardiac catheterization. The epicardial arteries were imaged using high resolution intravascular ultrasound to detect early structural changes and to determine changes in lumen size during pharmacologic provocation. A selective intracoronary Doppler velocity catheter was subsequently used to determine coronary blood flow velocity changes in response to the same pharmacologic provocation. Group I (9 patients) had no risk factors for atherosclerosis. Group II (17 patients) had one or more risk factors present.

Results. Although both Groups I and II had a normal microvascular vasodilator response to adenoslne or papaverine infusion (estimated coronary flow increase 396 ± 200% vs. 326 ± 161% [mean ± SD], respectively, p = 0.103), only Group I patients had an intact response to acetylcholine infusion (378 ± 203% vs. 75 ±93% in Group II, p = 0.001). Group II patients had an abnormal epicardial artery cross-sectional area vasoconstriction response to acetylcholine infusion (−16.6 ± 12.4% [13 patients] vs. 1.3 ±11.5% in Group I, p < 0.0007). An additional four Group II patients had severe spasm during acetylcholine infusion. Epicardial vasodilator response to nitroglycerin infusion, however, was preserved in Group II (14.6 ± 4.3% vs. 9.6 ± 3.5% in Group I, p = 0.212). All Group I patients had normal vessels by intravascular ultrasound. Of the 17 patients in Group II, 7 had minimal disease on ultrasound (intimal thickening or small eccentric plaque) in the study vessel. These patients did not respond differently from the 10 Group II patients without demonstrable disease on ultrasound.

Conclusions. Patients with risk factors for coronary artery disease, normal coronary angiograms and no measurable disease by intracoronary ultrasound exhibit selective endothelial dysfunction at both the epicardial and microvascular levels. These findings may have implications for the treatment of “preclinical” coronary atherosclerosis.