Published papers

Ventilation with a small tidal volume (Vt) is associated with better clinical outcomes than with a large Vt particularly in critical settings, including acute lung injury. To determine whether V t influences the lipopolysaccaharide (LPS) recognition pathway, we studied CD 14 expression in rabbit lungs and the release of TNF-α by cultured alveolar macrophages after 240 min of ventilation with a large (20 ml/kg) vs. a small (5 ml/kg) Vt. We also applied small or large Vt to lungs instilled with 50 μg/kg of LPS. The alveolar macrophages collected after large Vt ventilation revealed a 20-fold increase in LPS-induced TNF-α release compared with those collected after small Vt ventilation, whereas TNF-α was undetectable without LPS stimulation. In animals ventilated with a large Vt, the expression of CD14 mRNA in whole lung homogenates and the expression of CD14 protein on alveolar macrophages, assessed by immunohistochemistry, were both significantly increased in the absence of LPS stimulation. A large V t applied to LPS-instilled lungs increased the pulmonary albumin permeability and TNF-α release into the plasma. These results suggest that mechanical stress caused by a large Vt sensitizes the lungs to endotoxin, a phenomenon that may occur partially via the upregulation of CD14.

In health, acute anemia is accompanied by changes in the distribution of blood flows at all of the central, regional, and microcirculatory levels. This redistribution in blood flows provides the capacity to maintain tissue oxygenation with hematocrit as low as 21%. What is not known with certainty is whether the capacity to maintain tissue oxygenation in the presence of acute anemia can be influenced significantly by concurrent disease such as sepsis and cardiac disease. The single clinical trial [2] found an apparent survival benefit by not exposing patients with sepsis to blood transfusions until the hemoglobin concentration was less than 70 g/L. The question remains as to whether this observation was the consequence of a protective effect anemia or an injurious effect of transfusing old stored blood.

Study objectives To investigate, in mastectomy patients, the effectiveness of continuous cervical epidural block using a low-dose fentanyl infusion in combination with local anesthetics. Design Prospective, observational study. Setting 450-bed, university-affiliated hospital. Patients 21 ASA physical status I and II female patients undergoing modified radical mastectomy. Interventions An epidural catheter was inserted at the C7-Th1 interspace before the induction of anesthesia. Anesthesia was maintained using a low concentration of sevoflurane with nitrous oxide-oxygen (N2O-O 2). A mixture of 100 μg fentanyl and 49 mL of 1% mepivacaine was prepared, and 7 mL of this solution was epidurally injected before the initial incision. This same solution was continuously infused at a rate of 7 mL/hr (fentanyl 17.5 μg/hr) throughout the anesthesia, and at 2 mL/hr (fentanyl 5 μg/hr) postoperatively. Measurements and main results Intraoperative mean arterial pressure (MAP) and heart rate (HR), postoperative pain and analgesic use, and the frequency of postoperative side effects of anesthesia, including nausea, dizziness, and respiratory depression, were recorded. The protocol described provided stable intraoperative hemodynamic control with no or low-dose nicardipine infusion. Sufficient postoperative analgesia was achieved in 18 of 21 patients. One patient reported postoperative nausea, and no other side effects were reported. Conclusions Continuous epidural infusion of the low-dose fentanyl mixture described above provides adequate intraoperative hemodynamic control and postoperative pain relief, with a low rate of side effects in mastectomy patients. © 2004 by Elsevier Inc.

Objective: Acute lung injury induced by lung overstretch is associated with neutrophil influx, but the pathogenic role of neutrophils in overstretch-induced lung injury remains unclear. Design: To assess the contribution of neutrophils, we compared the effects of noninjurious large tidal volume (VT) ventilation on lungs in normal and neutrophil-depleted animals. Setting: Research animal laboratory. Subjects: Twenty-six male Japanese white rabbits. Interventions: Animals were mechanically ventilated for 4 hrs with one of the three following protocols: large VT (20 mL/kg), small VT (8 mL/kg), and large VT (20 mL/kg) with neutrophil depletion achieved by a single dose of vinblastine injection (0.75 mg/kg) intravenously 4 days before the experiment. Measurements and Main Results: Large VT ventilation produced alveolar neutrophil influx compared with low VT (p = .002) without evidence of edema or increased epithelial permeability. The neutrophil influx was accompanied by increases in interleukin-8 in bronchoalveolar lavage fluid (p = .04). Immunohistochemistry of large VT lungs showed increased interieukin-8 staining in bronchial epithelial cells, alveolar epithelium, alveolar macrophages, and smooth muscles of pulmonary vessels. Neutrophil depletion attenuated the interleukin-8 increase in the lung. Large VT did not increase plasma interleukin-8 or tumor necrosis factor-α in plasma and bronchoalveolar lavage fluid. No expression of p-selectin or intercellular adhesion molecule-1 was observed. Conclusions: Cyclic overstretching of normal rabbit lungs with noninjurious large VT produced neutrophil influx and interleukin-8 increase in bronchoalveolar lavage fluid. Production of pulmonary interleukin-8 by lung overstretch might require the interaction between resident lung cells and migrated neutrophils. This study suggests that large VT ventilation potentiates the predisposed, subclinical lung injury, such as nosocomial pneumonia or aspiration of gastric contents.

In summary, in this prospective randomized study, we compared the analgesic effects of intrathecal plus intravenous fentanyl and intravenous fentanyl in 36 cesarean section patients. There were no differences between the two methods in the duration of analgesia and motor block, postoperative pain scale scores, or the frequency of side effects such as nausea/vomiting and hypotension. © JSA 2003.

Background: In the partial CO2 rebreathing method, monitored changes in CO2 elimination and end-tidal CO2 in response to a brief rebreathing period are used to estimate cardiac output. However, dynamic changes in CO2 production during ischemia and reperfusion may affect the accuracy of these estimates. This study was designed to compare measurements of cardiac output as produced by the partial CO2 rebreathing (NICO), bolus (BCO), and continuous thermodilution (CCO) methods of monitoring cardiac output. Methods: Cardiac output was continuously monitored using both NICO and CCO in 28 patients undergoing aortic reconstruction. BCO measurements were taken at the following intervals when hemodynamic stability was achieved: (1) after anesthetic induction; (2) during aortic cross-clamp; (3) at reperfusion of the iliac artery; and, (4) during peritoneal closure. Results: The bias and precision (1 SD) derived from all the measurements between NICO and BCO was -0.58 ± 0.9 1/min, whereas for CCO and BCO it was 0.38 ± 1.17 1/min. The bias between NICO and BCO was small after anesthetic induction and during cross-clamp, but increased following reperfusion. The bias between CCO and BCO was relatively small until reperfusion but increased significantly at peritoneal closure. Conclusions: Results indicate that in aortic reconstruction surgery the performance of NICO monitoring is comparable with that of CCO; however, the direction of bias in these continuous measurement devices is the opposite.

Background: Preservation of gut integrity has become a therapeutic goal to obviate bacterial translocation in the critically ill. The authors examined whether olprinone, a phosphodiesterase III inhibitor, protected functional and structural integrity of gut mucosa against acute progressive hypoxia. Methods: Thirty-two animals were randomly allocated to a control group (n = 12), a low-dose group (0.2 μg · kg-1 · min-1 olprinone; n = 10), or a high-dose group (0.6 μg · kg-1 · min-1 olprinone; n = 10) after preparatory surgery. Ascending aortic and portal blood flow, intramural pH of the ileum, and portal endotoxin levels were measured at normoxia and through three stages of progressive hypoxia (fraction of inspired oxygen = 0.17, 0.13, and 0.10). Results: At normoxia, ascending aortic flow in the high-dose group was approximately 20% higher than in the control and low-dose groups. During progressive hypoxia, both ascending aortic and portal flow in the control group were depressed, whereas olprinone infusion attenuated such alterations and redistributed blood to the splanchnic area in a dose-dependent manner. On the contrary, the reduction of intramural pH of the ileum and the elevation of portal endotoxin levels observed in the control group were significantly minimized in both the low-and high-dose groups to a similar extent during acute hypoxia. Histopathologic alterations of gut mucosa observed in the control group were minimized by olprinone infusion dose-independently, accompanied by reduction of mortality rate of the animals. Conclusions: Olprinone slows progression of intestinal mucosal acidosis and gut barrier dysfunction, concurrently with preservation of microscopic structures, through both flow-dependent and -independent mechanisms under acute hypoxia. Such properties of olprinone may serve to protect the host under insult.

In summary, although further studies are obviously required, the present results using arterial HbCO level suggest that the early elevation of bilirubin is caused to some extent by production excess, whereas late hyperbilirubinemia is mainly due to its metabolism or excretion retardation in patients undergoing esophagectomy and reconstruction surgery.

Purpose. This study aimed to compare low-molecular weight hydroxyethyl starch containing 1% dextrose (HES) infusion and lactated Ringer's solution (LR) in the prevention of hypotension associated with spinal anesthesia for cesarean section. Methods. Sixty-seven patients scheduled for cesarean section under spinal anesthesia were randomly allocated to receive either LR (n = 35) or HES (n = 32) infusion before cesarean delivery. Infusion of the fluid was started immediately after arrival at the operating room, through two fully open i.v. routes of 18 or 16 gauge. The two groups were compared in terms of the incidence of hypotension; ephedrine dose; cord and maternal blood gas, hemoglobin, and glucose; and Apgar scores. Results. Intravenous fluid volume until delivery in the LR group was significantly greater than that in the HES group (1298 ± 503 and 973 ± 339 ml, respectively) in spite of the similar periods of intravenous infusion (18.1 ± 3.9 and 18.2 ± 4.1 min). The incidence of hypotension, and the ephedrine dose, blood gas analyses, and Apgar scores were not significantly different between the groups. The ephedrine dose correlated with the anesthesia level by spinal anesthesia (P < 0.05). Conclusion. This study did not show an advantage of HES compared with LR in the prevention of hypotension or in the reduction of ephedrine dose during cesarean section under spinal anesthesia. The anesthesia level, rather than the choice of intravenous fluid solution, might be related to the ephedrine dose.

Background. Recent evidence indicates that sevoflurane treatment before prolonged ischaemia reduces infarct size in normal hearts, mimicking ischaemic preconditioning. We examined whether exposure to sevoflurane before brief ischaemia, inducing a 'stunned myocardium', provided such protective effects in an isolated working heart from normal or septic rats. Methods. With institutional approval, 91 rats were randomly allocated into one of either caecal-ligation and perforation (CLP: n=50) or sham (Sham: n=41) procedure groups 24 h before the study. After determination of baseline measurements, including cardiac output (CO), myocardial oxygen consumption (mVO2) and cardiac efficiency (CE; COpeak systolic pressure/mVO2), each isolated heart was perfused with or without 2% sevoflurane for 15 min before global ischaemia (pre-ischaemia). After 15 min ischaemia and 30 min reperfusion, all hearts were assessed for functional recovery of myocardium (post-reperfusion). Results. During the pre-ischaemia period, 2% sevoflurane caused a significant reduction of CO in the CLP group compared with the Sham group. During the post-reperfusion period, both CO (16.9 vs 11.0 ml min-1) and CE (11.2 vs 7.7 mm Hg ml-1 (μI O2)-1) was higher in the sevoflurane-treated vs -untreated hearts from CLP rats, and was accompanied by lower incidence of reperfusion arrhythmia compared with control hearts (8 vs 32%). In contrast, 2% sevoflurane did not provide cardioprotective effects in normal rats. Conclusions. The current study demonstrates that pre-treatment with sevoflurane minimizes myocardial dysfunction and the incidence of reperfusion arrhythmia after brief ischaemic insults in septic hearts.

Background: Although precise mechanisms remain to be determined, recent studies show that heme oxygenase-1 (HO-1), providing endogenous carbon monoxide (CO) and bilirubin, serves as an antiinflammatory enzyme. This study aimed to clarify roles of CO in regulation of microvascular adhesion of platelets and leukocytes in endotoxemia. Methods: Rats pretreated with or without hemin were anesthetized with pentobarbital and received continuous infusion of endotoxin. Platelets labeled with carboxyfluorescein diacetate succinimidyl ester and leukocyte behavior in mesenteric venules were visualized using intravital ultra-high-speed intensified fluorescence videomicroscopy. To examine the mechanisms for the effects of HO-1 on platelet and leukocyte behavior during endotoxemia, these studies were repeated with superfusion of either CO, bilirubin, or zinc protoporphyrine-IX. Results: Endotoxin caused a marked depression of platelet velocity traversing along periendothelial regions, accompanied by augmented rolling and adhesion of leukocytes in venules. The endotoxin-elicited changes were attenuated by the HO-1 induction with hemin and restored by blockade of the enzyme activity with zinc protoporphyrine-IX, a potent inhibitor of HO-1. Such an inhibitory action of HO-1 on microvascular cell adhesion was reproduced by local superfusion of the buffer containing CO at micromolar concentrations. Such antiadhesive actions of CO on leukocytes disappeared under immunoneutralization of glycoprotein Ibα, an adhesion molecule against platelets, but not against leukocytes. Platelets isolated from hemin-treated rats increased their ability to generate CO and displayed lesser sensitivity of agonist-induced aggregation than those from controls. Conclusions: These results suggest that CO desensitizes endotoxin-induced adhesive responses of leukocytes, mainly through its ability to ameliorate platelet activation in vivo.