BackgroundThe mechanism by which small amounts of myofiber shortening lead to extensive wall thickening is unknown. When isolated fibers shorten, they thicken in the two orthogonal directions. In situ fibers, however, vary in their orientation through the wall, and each is tethered to near or distant neighbors, which allows shortening to occur both in the direction of the fibers and also perpendicular to them. This “cross-fiber” shortening may enable the wall to shorten in two directions and thereby thicken extensively in the third.Methods and ResultsNuclear magnetic resonance tagging is a noninvasive method of labeling and tracking myocardium of the entire heart in three dimensions that does not interfere with myocardial motion. To investigate the presence and importance of cross-fiber shortening in the intact left ventricle, 10 closed-chest dogs were studied by nuclear magnetic resonance tagging. Five short-axis and four long-axis images were acquired to reconstruct 32 cubes of myocardium in each dog at end diastole and end systole. Pathological dissection was performed to determine the fiber direction at the epicardium, midwall, and endocardium of each cube. Strain was computed from the three-dimensional cube coordinates in the fiber and cross-fiber directions for epicardial and endocardial surfaces, and thickening of the full wall and its epicardial and endocardial halves was determined. Shear deformations were also calculated. Fiber strain at the epicardium and endocardium was −6.4±0.7% and −8.5±0.6% (mean±SEM), respectively (difference,P> .05). Cross-fiber strain at epicardium and endocardium was −0.6±0.5% and −25±0.6%, respectively (difference,P< .05). Thickening of the full wall reached 32.5±1.0%, composed of epicardial thickening of 25.5±0.6% and endocardial thickening of 43.3±1.0% (difference,P< .05). Fiber/cross-fiber shear strain was small (<3%). Significant regional differences were present in all strains. A significant correlation was found between the extents of regional thickening and cross-fiber shortening.ConclusionsCross-fiber shortening at the endocardium, therefore, far exceeds cross-fiber shortening at the epicardium and fiber shortening at both epicardium and endocardium. Since no active shortening can occur locally in cross-fiber direction, the extensive endocardial cross-fiber shortening must result from interaction with differently aligned fibers at a distance. The correlation between regional thickening and cross-fiber shortening supports the hypothesis that this interaction is the mechanism for amplifying small amounts of fiber shortening to cause extensive endocardial thickening.