Michael W. Rooney

Tourniquet-induced exsanguination in patients requiring lower limb surgery. An ischemia-reperfusion model of oxidant and antioxidant metabolism.

Background Surgically induced ischemia and reperfusion is frequently accompanied by local and remote organ injury. It was hypothesized that this procedure may produce injurious oxidants such as hydrogen peroxide (H2 O2), which, if unscavenged, will generate the highly toxic hydroxyl radical (*symbol* OH). Accordingly, it was proposed that tourniquet‐induced exsanguination for limb surgery may be a useful ischemia‐reperfusion model to investigate the presence of oxidants, particularly H2 O2. Methods In ten patients undergoing knee surgery, catheters were placed in the femoral vein of the limb operated on for collection of local blood and in a vein of the arm for sampling of systemic blood. Tourniquet‐induced limb exsanguination was induced for about 2 h. After tourniquet release (reperfusion), blood samples were collected during a 2‐h period for measurement of H2 O2, xanthine oxidase activity, xanthine, uric acid (UA), glutathione, and glutathione disulfide. Results At 30 s of reperfusion, H2 O2 concentrations increased ([nearly equal] 90%) from 133+/‐5 to 248+/‐8 nmol *symbol* ml sup ‐1 (P < 0.05) in local blood samples, but no change was evident in systemic blood. However, in both local and systemic blood, xanthine oxidase activity increased [nearly equal] 90% (1.91+/‐ 0.07 to 3.93+/‐0.41 and 2.19+/‐0.07 to 3.57+/‐ 0.12 nmol UA *symbol* ml sup ‐1 *symbol* min sup ‐1, respectively) as did glutathione concentrations (1.27+/‐0.04 to 2.69+/‐0.14 and 1.27+/‐0.03 to 2.43+/‐0.13 micro mol *symbol* ml sup ‐1, respectively). At 5 min reperfusion, in local blood, H2 O2 concentrations and xanthine oxidase activity peaked at 796+/‐38 nmol *symbol* ml sup ‐1 ([nearly equal] 500%) and 11.69+/‐1.46 nmol UA *symbol* ml sup ‐1 *symbol* min sup ‐1 ([nearly equal] 520%), respectively. In local blood, xanthine and UA increased from 1.49 +/‐0.07 to 8.36+/‐0.33 nmol *symbol* ml sup ‐1 and 2.69 +/‐0.16 to 3.90+/‐0.18 micro mol *symbol* ml sup ‐1, respectively, whereas glutathione and glutathione disulfide increased to 5.13+/‐0.36 micro mol *symbol* ml sup ‐1 and 0.514+/‐ 0.092 nmol *symbol* ml sup ‐1, respectively. In systemic blood, xanthine oxidase activity peaked at 4.75+/‐0.20 UA nmol *symbol* ml sup ‐1 *symbol* min sup ‐1. At 10 min reperfusion, local blood glutathione and UA peaked at 7.08+/‐0.46 micro mol *symbol* ml sup ‐1 and 4.67 +/‐0.26 micro mol *symbol* ml sup ‐1, respectively, while the other metabolites decreased significantly toward pretourniquet levels. From 20 to 120 min, most metabolites returned to pretourniquet levels; however, local and systemic blood xanthine oxidase activity remained increased 3.76+/‐0.29 and 3.57+/‐0.37 nmol UA *symbol* ml sup ‐1 *symbol* min sup ‐1, respectively. Systemic blood H2 O2 was never increased during the study. During the burst period ([nearly equal] 5–10 min), local blood H2 O2 concentrations and xanthine oxidase activities were highly correlated (r = 0.999). Conclusions These studies suggest that tourniquet‐induced exsanguination for limb surgery is a significant source for toxic oxygen production in the form of H2 O2 and that xanthine oxidase is probably the H2 O2 ‐generating enzyme that is formed during the ischemia‐reperfusion event. In contrast to the reperfused leg, the absence of H2 O2 in arm blood demonstrated a balanced oxidant scavenging in the systemic circulation, despite the persistent increase in systemic xanthine oxidase activity.

Regional hemodynamics and oxygen supply during isovolemic hemodilution alone and in combination with adenosine-induced controlled hypotension.

Studies were performed in ten pentobarbital-anesthetized, open chest dogs to evaluate regional circulatory effects of isovolemic hemodilution alone, and in combination with adenosine-induced controlled hypotension. Regional blood flow measured with 15-μm radioactive microspheres was used to calculate regional oxygen supply. Hemodilution with 5% dextran (40,000 molecular weight) reduced arterial hematocrit and oxygen contentby approximately one-half and caused heterogeneous changes in regional blood flows; flow decreased in the spleen, was unchanged in the renal cortex, liver, skeletal muscle and skin, and increased in the duodenum, pancreas, brain and myocardium; however, only inthe brain and myocardium were increases in flow sufficient to preserve oxygen supply. Intravenous infusion of adenosine reduced aortic pressure by 50% and reduced flow in most tissues (renal cortex, pancreas, liver, spleen, skin, and brain), with the result that oxygen deficits were produced or accentuated in these organs. The magnitude of flow reductions in the renal cortex (−73%) and cerebral cortex (−37%) were noteworthy. In themyocardium, direct coronary vasodilation by adenosine caused parallel increases in blood flow and oxygen supply to levels exceeding prevailing metabolic requirements. It is concluded that 1) during isovolemic hemodilution alone, oxygen supply to the brain and myocardium is maintained at the expense of oxygen supply to less critical organs and, 2) during combined isovolemic hemodilution and adenosine−induced hypotension, oxygen is oversupplied to the myocardium but undersupplied to the brain and kidney. These latter effects suggest the need for extensive clinical monitoring of patients in whom combined isovolemic hemodilution and adenosine−induced hypotension is utilized.

Myocardial blood flow and oxygen consumption during isovolemic hemodilution alone and in combination with adenosine-induced controlled hypotension.

Recent reports have proposed combining isovolemic hemodilution and controlled hypotension tolimit blood loss during surgery. Before such a technique can be considered for clinical use, it must be demonstrated that it does not endanger maintenance of adequate myocardial oxygenation. Accordingly, measurements of left ventricular myocardial blood flow and oxygen consumption were obtained during isovolemic hemodilution alone and in combination with adenosine-induced controlled hypotension in ten pentobar-bital-anesthetized, open chest dogs with normal coronay circulation. Hemodilution to a hematocrit of 21.7% was produced by isovolemic exchange of whole blood for 5% dextran. In the presence of hemodilution, adenosine was infused intravenously at a rate sufficient to decrease mean aortic pressure to 51 mm Hg. Myocardial blood flow was measured with radioactive microspheres and used to calculate global left ventricular myocardial oxygen consumptionand oxygen supply. Hemodilution alone increased aortic blood flow (+43%) but had no effect on aortic pressure, left atrial pressure, heart rate, or left ventricular dP/dtmax an increase in myocardial blood flow (+130%) maintained oxygen supply and consumption at the baseline level. Adenosine-induced hypotension during hemodilution decreased heart rate (−35%), left ventricular dP/dt max (−28%), and aortic blood flow (−14%). These systemic responses were accompanied byreduced myocardial oxygen consumption (−29%) and increased myocardial blood flow (+54%) and myocardial oxygen supply (+72%). These latter effects resulted in reduction in the coronary arterio-venous oxygen content difference and in an attendant rise in coronary sinus Po2 (+66%), which are signs of luxuriant myocardial perfusion. The present study demonstrated in anesthetized dogs that 1) myocardial oxygenation is well maintained during isovolemic hemodilution alone and, 2) myocardial oxygenation is influenced favorably when isovolemic hemodilution is combined with adenosine-induced controlled hypotension. Further studies are required to evaluate the safetyof the latter condition in hearts with stenotic coronary arteries.

Regional hemodynamics and oxygen supply during isovolemic hemodilution in the absence and presence of high-grade β-adrenergic blockade

Studies were performed in 16 pentobarbital-anesthetized dogs to evaluate regional circulatory effects of isovolemic hemodilution in the absence (group 1) and presence (group 2) of high-grade beta-adrenergic blockade with propranolol. Regional blood flow measured with 15 microm radioactive microspheres was used to calculate regional oxygen supply. In group 1, hemodilution with 5% dextran (40,000 molecular weight) reduced arterial hematocrit and oxygen content by approximately one half and had heterogeneous effects on regional blood flows. Blood flow was unchanged in the renal cortex, liver, and spleen, and it increased in the pancreas, duodenum, brain, and myocardium; however, only in the brain and myocardium were increases in blood flow sufficient to maintain oxygen supply at baseline (pre-hemodilution) levels. In group 2, intravenous administration of propranolol (1 mg/kg) itself decreased blood flow in the spleen and myocardium and had no other regional effects. Hemodilution after propranolol caused regional circulatory changes that were essentially similar to those in the absence of propranolol. It is concluded that (1) during isovolemic hemodilution, oxygen supply to the brain and myocardium is maintained at the expense of oxygen supply to less critical organs, and (2) this pattern of regional circulatory response during hemodilution remains intact in the presence of high-grade beta-adrenergic blockade with propranolol.

Separation of Myocardial Versus Peripheral Effects of Calcium Administration in Normocalcemic and Hypocalcemic States Using Pressure-Volume (Conductance) Relationships

This study used left ventricular pressure-volume (conductance) relationships to separate the effects of calcium administration on myocardial performance and peripheral vasoconstriction in normocalcemic and hypocalcemic states. Hypocalcemia was produced in anesthetized dogs with intravenous citrate-phosphate-dextrose until serum [Ca2+] was approximately 0.7 mmol/L. Calcium (CaCl2) bolus (5 mg/kg) was administered during normocalcemia (n = 6) and hypocalcemia (n = 6), and data were collected at 1, 5 and 10 min after CaCl2 administration. During normocalcemia, CaCl2 administration increased [Ca2+] 19% at 1 min and was accompanied by a 47% (P < 0.05) decrease in left ventricular contractility (i.e., end-systolic elastance or E(lves)) and a 13% (P < 0.05) increase in systemic vascular resistance. At 5 and 10 min, serum [Ca2+] and the hemodynamic variables began to return to the baseline values. During hypocalcemia, E(lves) decreased 25% (P < 0.05), but after CaCl2 bolus, it increased to baseline levels and remained there during the 10-min period. Hypocalcemia and the CaCl2 bolus did not significantly affect SVR. In conclusion, these studies suggest that the indications for the use of calcium should depend on the initial serum level of ionized calcium.

Lack of Increased Cardiac Output During Hemoglobin Hemodilution Can be Reversed with Sodium Nitroprusside

To investigate a possible connection between EDRF or nitric oxide (NO) and the unchanged cardiac output (CO) during hemoglobin-hemodilution we infused nitroprusside (NP) in eight Hb-diluted dogs (Hct approximately 20%). Normal hypotensive doses of NP were not effective and supranormal doses (133.0 micrograms/kg/min) were needed to induce even a modest decrease in mean AoP (approximately 25 mmHg). With these NP doses, cardiac output increased 177%, diastolic AoP (afterload) decreased 30%, while systolic AoP and LVEDP (preload) were unchanged. Heart rate, LV contractility (pressure-volume function) and blood volume were not changed throughout the study. Normally, NP alone decreases both preload and afterload resulting in unchanged CO. In the Hb + NP dogs, CO increased because only afterload decreased suggesting a selective effect of Hb on venous and arterial smooth muscle relaxation. In hemodilution with nonhemoglobin colloids, CO increases primarily because the diluted blood offers less viscous resistance to ventricular ejection. It appears that in order for cardiac output to increases in the presence of Hb, some decrease in arteriolar resistance is needed, presumably to unmask the effects of reduced viscosity. These results suggest the unchanged CO during Hb-dilution is related to a selective effect of Hb on venous and arteriolar nitric oxide (EDRF) function.

On the Interaction of the Liposomal Membrane with Blood Components

Liposome-encapsulated hemoglobin (LEH) has been shown to be a viable candidate as a blood replacement. However, few data have been presented as to how LEH interacts with normal blood components. Liposomes were prepared from egg lecithin, cholesterol, and dicetyl phosphate or phosphatidic acid, and mixed with fresh blood plasma or whole blood. Erythrocyte osmotic fragility, prothrombin time (extrinsic coagulation efficiency), activated partial thromboplastin time (intrinsic coagulation efficiency), plasma clot stability in urea (fibrin stabilizing factor), and clot retraction (platelet activation) were measured. Although liposomes were found to bind extensively to erythrocytes, all tests indicated that the liposomes had no significant adverse effects, provided that normal levels of plasma Ca++ were maintained. The ability of liposomes to absorb Ca++ from the plasma was related directly to the amount of dicetyl phosphate or phosphatidic acid present and thus, presumably, to the presence of negatively charged species in the membrane. The mechanics of deformation of the LEH membrane were investigated by encapsulating Hemoglobin S in liposomes. Liposomes containing Hemoglobin S were found to sickle when deoxygenated, but not liposomes containing normal hemoglobin. Shape analysis of sickled liposomes yielded a deforming stress of 10(6) dynes/cm2, about 50 times greater than the reported limit for shear elasticity of the erythrocyte membrane.

Influence of nifedipine on systemic and regional hemodynamics during adenosine-induced hypotension in dogs.

Previous pharmacologic studies indicating competitive interactions between adenosine and nifedipine at the adenosine vascular receptor suggest that adenosine may he a less effective hypotensive drug after pretreatment with nifedipine. This hypothesis was tested in 18 pentobarbital-anesthetized, open-chest dogs by evaluating the hypotensive effects and regional hemodynamic responses to 60-minute intravenous adenosine infusions before and after bolus injection of nifedipine (20 μg/kg, IV). Regional blood flow was measured with 15-μm radioactive microspheres. Before nifedipine, infusion of adenosine at a rate of 126 ± 30 μmol/min caused a 50% reduction in mean aortic pressure that in the presence of no change in aortic blood flow was attributable to a proportional decrease in systemic vascular resistance. These systemic effects were associated with heterogeneous changes in regional blood flow; blood flow decreased in the renal cortex (−68%), pancreas (−50%), spleen (−77%), and skin (−61%); increased in the left (+112%) and right (+265%) ventricular myocardium; and did not change significantly in the duodenum, liver, skeletal muscle, or brain. Nifedipine did not alter the dose requirement or time course of the adenosine-induced hypotensive response or affect the associated systemic hemodynamic changes. Furthermore, nifedipine caused only minor alterations in the regional blood flow changes during adenosine-induced hypotension. Apparently the high plasma levels of adenosine required for controlled hypotension in the present study were sufficient to overcome the blocking influence of nifedipine at the adenosine vascular receptor. The study demonstrates that the hypotensive action of adenosine remains unimpaired after pretreatment with nifedipine. The present findings may have clinical relevance in patients with cardiovascular disease being treated with nifedipine who may need adenosine to lower blood pressure either to decrease bleeding or to decrease myocardial work and oxygen consumption.

Effect of jet ventilation on heart failure : Decreased afterload but negative response in left ventricular end-systolic pressure-volume function

OBJECTIVE To examine the mechanism of cardiac assist with systolic jet ventilation, specifically effects on loading conditions and left ventricular pressure-volume function. Both systolic and diastolic jet ventilation were compared in the absence and presence of heart failure. DESIGN Prospective, two-factor, repeated-measures study. SETTING Animal laboratory. SUBJECTS Ten anesthetized, closed-chest dogs. INTERVENTIONS The measurement protocol consisted of two phases: a) apnea, randomized jet ventilation (systole- and diastole-synchronized); b) postjet ventilation apnea, before and after heart failure, induced with a propranolol-imipramine-plasma expansion treatment. MEASUREMENT AND MAIN RESULTS Systolic and diastolic jet ventilation was associated with mean airway pressures of approximately 7 mm Hg and intrapleural pressures of approximately 3 mm Hg in both heart conditions. In normal hearts, jet ventilation (either mode) decreased transmural left ventricular end-diastolic pressure by 40% to 60% (p < .05), left ventricular end-diastolic volume 25 +/- 8%, and stroke volume by 28% to 30%. Heart failure was associated with decreases (41 +/- 6%) in end-systolic pressure-volume function (i.e., pressure change/volume change or elastance), transmural left ventricular end-systolic pressure (22 +/- 3%), and stroke volume (16 +/- 4%), and increased transmural left ventricular end-diastolic pressure (139 +/- 6%). Application of jet ventilation (either mode) during heart failure did not affect stroke volume but significantly (p < .05) attenuated transmural left ventricular end-diastolic pressure by 30% to 40%, left ventricular end-diastolic volumes by 33 +/- 9%, and transmural left ventricular end-systolic pressure by 11% to 19% (p < .05). After jet ventilation, left ventricular elastance was decreased 36 +/- 8% in normal hearts and 35 +/- 11% in failing hearts. Stroke volume, however, returned to baseline levels because of increases in transmural left ventricular end-diastolic pressure in both heart conditions, and also in failing hearts, because transmural left ventricular end-systolic pressure remained decreased approximately 30% (p < .05). CONCLUSIONS Jet ventilation did not decrease stroke volume in failing hearts because of the afterload-reducing benefit (decreased transmural left ventricular end-systolic pressure) of increased intrapleural pressure in dilated ventricles. Moreover, jet ventilation did not have positive effects on myocardial function and had negative effects on left ventricular elastance in the postjet ventilation period in both normal and failing hearts. Cardiac assist by jet ventilation was not cycle specific, suggesting no selective benefit of jet ventilation over conventional positive-pressure ventilation during heart failure. These studies demonstrate a negative inotropy associated with jet ventilation that, during heart failure, may compromise the general benefit of positive-pressure-mediated increases in intrapleural pressure.