Spatial heterogeneity of local blood flow and metabolite content in dog hearts.

Spatial variation (heterogeneity) of myocardial blood flow was studied under basal conditions in relation to four biochemical markers: creatine kinase (CK), lactate dehydrogenase (LDH), ATP, and glycogen. A total of 508 individual 0.5-g samples from the left ventricular free wall was studied in 12 dogs. Myocardial blood flow was measured by radioactive microspheres (15 micron diam) injected via a pigtail catheter into the left ventricle during light sedation (closed-chest flow measurements); following thoracotomy, a second set of microspheres was injected via a catheter into the left atrium (open-chest flow measurements, n = 5). In 27-54 samples/heart, myocardial blood flow, CK, LDH, protein, ATP, and glycogen were determined, permitting a direct correspondence between local blood flow and metabolic markers in each sample and an assessment of the spatial heterogeneity of flow and metabolite content. Correlations between myocardial blood flow per gram tissue and metabolite concentration per gram tissue in pooled data were weak but significant: partial correlation coefficients were 0.137 for CK (P = 0.024), 0.167 for LDH (P = 0.006), 0.341 for ATP (P less than 0.001), 0.123 for glycogen (P = 0.053), but not significant for protein vs. myocardial blood flow. The coefficient of variation, which defines the extent of spatial heterogeneity, averaged 20% for closed-chest flow measurements, 19% for open-chest flow measurements, 22% for CK, 17% for LDH, 15% for protein, 8% for ATP, and 18% for glycogen; these values are over and above the variability due to the technique error, indicating a definite physiological spatial variability. The correlation between local blood flow and the studied metabolites can only explain a minor portion of the spatial heterogeneity of myocardial blood flow. Although a physiological link between blood flow and metabolite content for small regions of the heart is demonstrated, the true local variability of blood flow may be modulated predominantly by other factors.