Mechanisms of chronic regional postischemic dysfunction in humans. New insights from the study of noninfarcted collateral-dependent myocardium.

BACKGROUND Even in the absence of a previous myocardial infarction, patients with coronary artery disease often present with chronic regional wall motion abnormalities that are reversible spontaneously or after coronary revascularization. In these patients, regional dysfunction has been proposed to result either from prolonged postischemic dysfunction (myocardial "stunning") or from adaptation to chronic hypoperfusion (myocardial "hibernation"). This study examines which of these two mechanisms is responsible for the chronic regional dysfunction often detected in patients with angina and noninfarcted collateral-dependent myocardium. METHODS AND RESULTS Twenty-six anginal patients (19 men; mean age, 60 +/- 9 years old) with chronic occlusion of a major coronary artery but without previous infarction were studied. Positron emission tomography was performed to measure absolute regional myocardial blood flow with 13N-ammonia at rest (n = 26) and after intravenous dipyridamole (n = 11). The kinetics of 18F-deoxyglucose and 11C-acetate were measured to calculate the rate of exogenous glucose uptake and the regional oxidative metabolism (n = 15). Global and regional left ventricular function was evaluated by contrast ventriculography at baseline (n = 26) and after revascularization (n = 12). Transmural myocardial biopsies from the collateral-dependent area were obtained in seven patients during bypass surgery and analyzed by optical and electron microscopy. According to resting regional wall motion, patients were separated into groups with and without dysfunction of the collateral-dependent segments. In patients with normal wall motion (n = 9), regional myocardial blood flow, oxidative metabolism, and glucose uptake were similar among collateral-dependent and remote segments. By contrast, in patients with regional dysfunction (n = 17), collateral-dependent segments had lower myocardial blood flow (77 +/- 25 versus 95 +/- 27 mL.min-1.100 g-1, p < 0.001), smaller k values (slope of 11C clearance reflecting oxidative metabolism, 0.049 +/- 0.015 versus 0.068 +/- 0.020 min-1, p < 0.001) and higher glucose uptake (relative 18F-deoxyglucose-to-flow ratio of 1.9 +/- 1.6 versus 1.2 +/- 0.2, p < 0.05) compared with remote segments. However, myocardial blood flow and k values were similar among collateral-dependent segments of patients with and without segmental dysfunction. After intravenous dipyridamole, collateral-dependent myocardial blood flow increased from 78 +/- 5 to 238 +/- 54 mL.min-1.100 g-1 in three patients with normal wall motion and from 88 +/- 17 to only 112 +/- 44 mL.min-1.100 g-1 in eight patients with regional dysfunction. There was a significant (r = -0.85, p < 0.001) inverse correlation between wall motion abnormality and collateral flow reserve. Analysis of the tissue samples obtained at the time of bypass surgery showed profound structural changes in dysfunctioning collateral-dependent areas, including cellular swelling, loss of myofibrillar content, and accumulation of glycogen. Despite these alterations, the regional wall motion score improved significantly in the patients studied before and after revascularization (from 3.8 +/- 1.3 to 0.8 +/- 0.9, p < 0.005). CONCLUSIONS In a subgroup of patients with noninfarcted collateral-dependent myocardium, immature or insufficiently developed collaterals do not provide adequate flow reserve. Despite nearly normal resting flow and oxygen consumption, these collateral-dependent segments exhibit chronically depressed wall motion and demonstrate marked ultrastructural alterations on morphological analysis. We propose that these alterations result from repeated episodes of ischemia as opposed to chronic hypoperfusion and represent the flow, metabolic, and morphological correlates of myocardial "hibernation."