Hypovolemic shock and cardiac contractility: assessment by end-systolic pressure-volume relations

Abstract

The end-systolic pressure-volume relation (ESPVR) is accepted as a load-independent measure of cardiac contractility. Potential curvilinearity of the ESPVR, dependency on coronary perfusion pressure (CPP) and sensitivity to the type of loading intervention might limit its use in hemorrhagic shock. This study compared ESPVRs obtained by caval and aortic occlusion under physiological loading conditions at baseline with those obtained during hemorrhagic shock (mean arterial pressure 45 mmHg). The left ventricular (LV) pressure (tip manometer) and volume (conductance catheter) were measured in ten anesthetized pigs. ESPVRs were fitted to linear and quadratic models. Within end-systolic pressure (Pes) ranges obtained under baseline conditions, ESPVR displayed only minimal curvilinearity (second-order coefficient a < 0.007) and could be accurately described by a linear model. However, nonlinearity of ESPVRs obtained over wider load ranges is suggested by negative volume axis intercepts of the linear model. Steeper ESPVR with aortic than with caval occlusion (2.28 +/- 0.22 vs 3.41 +/- 0.51 mmHg/ml, ns) could not be proven owing to the large interindividual variance of ESPVR slopes with both loading interventions. During shock the Pes range obtained by caval occlusion decreased to very low levels (from 49 +/- 2 to 34 +/- 1 mmHg), ESPVR did not adequately fit either of the two models (mean R < 0.66), and critical reduction of CPP induced negative ESPVR slope in four of ten experiments. In contrast, aortic occlusion at shock resulted in linear ESPVR (R = 0.927 +/- 0.029), Pes ranges (92 +/- 3 to 58 +/- 4 mmHg) comparable to the ones obtained by caval occlusion at control (113 +/- 5 to 73 +/- 6 mmHg), and steeper ESPVR than at control (3.41 +/- 0.51 to 7.38 +/- 1.0 mmHg/ml, P < 0.05). Interpretation of the increased ESPVR slope obtained with aortic occlusion as due to increased contractility in shock is, however, complicated by different Pes ranges. It is concluded that within Pes ranges obtained with caval or aortic occlusion in situ the ESPVR can be adequately fitted to a linear model. For assessment of the inotropic response to shock the ESPVR is of limited value because (1) caval occlusion is not suitable to generate ESPVR during shock, and (2) Pes ranges obtained with identical loading interventions differ greatly between baseline and shock and, therefore, apparent ESPVR changes are influenced by the potential nonlinearity of the ESPVR. Combining caval occlusion at baseline with aortic occlusion at shock would result in comparable Pes ranges. Interpretation of results is, however, complicated by diverging effects of the different loading interventions on the shape and slope of the ESPVR.