BACKGROUND Papillary muscle dysfunction (PMD) has been implicated in the pathogenesis of ischemic mitral regurgitation (MR). We hypothesized that ischemic MR is not caused by PMD and/or dysfunction of the myocardial regions from where the papillary muscles arise but is related to reduction in global left ventricular (LV) function. To test this hypothesis, three groups of dogs were studied. RESULTS In group 1 dogs (n = 8), varying degrees of regional and global LV dysfunction were produced. In group 2 dogs (n = 7), the circulation to the papillary muscles was isolated from that of the rest of the LV. Dysfunction of one or both papillary muscles was produced without producing global LV dysfunction. Global LV dysfunction was also produced while keeping papillary muscle function intact. The degree of MR (assessed using contrast echocardiography) was correlated in both groups of dogs with thickening of the papillary muscles and regional and global LV function. In the group 3 dogs (n = 6), the spatial distribution of blood flow within each papillary muscle was determined during ischemia by using radiolabeled microspheres. Thickening of the papillary muscles was assessed at three different levels along their lengths and was correlated with average blood flow at these levels. In group 1 dogs, MR was noted only when global LV function was affected and its severity correlated inversely with global LV function (r = -0.84 with peak positive LV dP/dt and r = -0.95 with global LV thickening, respectively). In comparison, there was poor correlation between MR and anterior and posterior papillary muscle thickening (r = -0.38 and r = -0.49, respectively). In group 2 dogs, MR did not occur in the presence of either PMD or akinesia of the immediately adjacent LV myocardium. MR occurred only when global LV dysfunction was produced (with the papillary muscle function intact), and its severity correlated inversely with global LV function (r = -0.92 with LV dP/dt and r = -0.86 with global LV thickening, respectively). There was poor correlation between the degree of MR and thickening of the anterior and posterior papillary muscles (r = -0.24 and r = -0.38, respectively). In both groups of dogs, MR was associated with incomplete mitral leaflet closure (IMLC), and the severity of MR correlated linearly with the degree of IMLC (r = 0.98). MR was never associated with mitral valve prolapse. In the group 3 dogs, despite more inhomogeneous flow during ischemia to the anterior compared with the posterior papillary muscle, mean thickening of these muscles was similar (3 +/- 10% and 3 +/- 4%, respectively). Furthermore, there was minimal variability in thickening between different parts of the muscles (3 +/- 2% and 5 +/- 3%, respectively). CONCLUSIONS It is concluded that PMD and/or dysfunction of the immediately adjacent LV myocardium does not result in MR. MR occurs during ischemia only when global LV function is affected, even when thickening of the papillary muscles and the immediately adjacent LV remains intact. MR in this situation is related to IMLC; the greater the degree of IMLC, the greater the MR. These findings suggest that the mechanism of ischemic MR is not related to PMD. There may also be important therapeutic implications of these findings.