BACKGROUND: The present investigation was designed to obtain an absolute measurement of myocardial blood flow and of its transmural distribution in ischemic heart disease and idiopathic dilated cardiomyopathy and to provide a reference standard for cardiac imaging in nuclear cardiology. METHODS AND RESULTS: Regional myocardial blood flow and its transmural distribution were estimated by the reference microsphere method in eight patients with idiopathic dilated cardiomyopathy (n = 4) or ischemic heart disease (n = 4) during heart transplant procedure. Before aortic clamping, 99mTc-labeled human albumin microspheres were injected into the left atrium while arterial blood was sampled from the aorta at a constant rate. No complications were observed during or after the procedure. From the excised heart, myocardial slices for gamma camera imaging and well counting analysis were obtained. Myocardial blood flow was assessed by a well counter, correlated with the extent of fibrosis expressed as collagen per total tissue proteins obtained from 4-hydroxyproline and glycine as determined by high-performance liquid chromatography. Microsphere distribution, as seen by gamma camera images in a different slice, was correlated with the extent of fibrosis assessed by histological analysis of the same myocardial specimen. Mean transmural myocardial blood flow was 0.49 +/- 0.17 and 0.38 +/- 0.15 mL.min-1 x g-1 in idiopathic dilated cardiomyopathy and ischemic heart disease, respectively (P < .01). Endocardial-to-epicardial flow ratio was lower in ischemic heart disease than in idiopathic dilated cardiomyopathy patients (0.99 +/- 0.33 versus 1.16 +/- 0.30, P < .05). Mean myocardial fibrosis was 9 +/- 6% in idiopathic dilated cardiomyopathy and 25 +/- 28% in ischemic heart disease. In both groups, no correlation was found between myocardial blood flow values and the extent of fibrosis. In ischemic heart disease, regional myocardial blood flow was not significantly affected by the severity of coronary stenosis (< or = 70% or > 70%) either in the endocardium (0.44 +/- 0.24 versus 0.36 +/- 0.16 mL.min-1 x g-1, P = NS) or in the epicardium (0.50 +/- 0.33 versus 0.38 +/- 0.33 mL.min-1 x g-1, P = NS). By gamma camera imaging, transmural microsphere distribution appeared more homogeneous in idiopathic dilated cardiomyopathy than in ischemic heart disease (mean coefficient variation, 18% and 27%, respectively; P < .02); the severity of perfusion impairment did not correlate with the extent of fibrosis evaluated by histological criteria. CONCLUSIONS: Heart transplant surgery offers a valuable model to assess absolute myocardial perfusion in human heart failure. Myocardial blood flow is markedly depressed in failing hearts of both ischemic heart disease and idiopathic dilated cardiomyopathy patients; a different transmural myocardial blood flow distribution is observed in ischemic heart disease than in idiopathic dilated cardiomyopathy, with prevalent endocardial perfusion in the latter but not the former condition. In patients with end-stage heart failure, myocardial blood flow appears to be similarly impaired in fibrotic and viable regions. Mechanisms other than myocardial fibrosis and coronary lesions appear to operate in determining myocardial blood flow impairment in heart failure.