Lazaroid pretreatment preserves gas exchange in endotoxin-treated dogs

: The lazaroids are a new class of potent free-radical scavengers. We tested whether U-74389G, a lazaroid, could attenuate some of the adverse cardiopulmonary effects of sepsis.

: Dogs were randomized to receive either 10 mg/kg U-74389G (n = 10), or a saline control (n = 11). After baseline measurements of hemodynamics and gas exchange, they were then randomized to receive either 0.2 mg/kg endotoxin or a saline infusion. Measurements of hemodynamics and gas exchange were repeated. The study was concluded 70 minutes after endotoxin infusion and the lungs were then removed for histologic evaluation.

: In endotoxin-treated control animals, P0₂ decreased (278 ± 123 mm Hg to 67 ± 13 mm Hg, P < .05) and intrapulmonary shunt increased (12.9% ± 1.1% to 28.2% ± 11.4%, P < .05) after endotoxin. Pretreatment with U-74389G attenuated the decrease in PO₂ (476 ± 61 mm Hg to 226 ± 143) and the increase in intrapulmonary shunt (12.6% ± 6.1% to 14.3% ± 6.8%) observed after endotoxin. The extent of lung injury and systemic hemodynamics were similar between control or U-74389G-treated dogs.

: A free-radical-scavenger can attenuate the gas exchange defect commonly associated with endotoxin but it does not improve the derangement of systemic hemodynamics.     

Blood interleukin 10 levels parallel the severity of septic shock

The aim of this study was to investigate the relation between interleukin (IL) 10, tumor necrosis factor alpha (TNF), IL-1, and IL-6 levels in patients with septic shock and relate these cytokine levels to the development of organ failure.

In 11 patients with septic shock of recent onset, blood was sampled for determinations of TNF, IL-1, IL-6, and IL-10. The degree of organ failure was scored for four organ systems (respiratory, hepatic, renal, hematologic) in the first 48 hours of the study.

The APACHE II score was 21 ± 4. Three patients died. IL-10 levels were directly correlated with TNF levels (r = 0.73, P < .05) and IL-6 levels (r = 0.67, P < .05); and inversely correlated with total C3 (r = −0.73, P < .05) and CH50 (r = −0.68, P < .05). Both IL-10 and TNF levels were correlated to the organ failure score (r = 0.75 and r = 0.68, both P < .01). Six patients with high IL-10 levels (>60 pg/mL) had lower C3 (37 ± 11 v 62 ± 10 mg/dL) and CH50 (32 ± 7 v 68 ± 19%), and higher organ failure scores (5.7 ± 0.8 v 3.8 ± 1.3) than those with low IL-10 levels (all P < .05).

Although IL-10 has an inhibitory effect on the production of cytokines, it is released together with TNF and IL-6 in patients with septic shock. IL-10 blood levels are directly related to the severity of inflammation and the development of organ failure in septic shock.     

Nitric oxide metabolism in canine sepsis: relation to regional blood flow

To investigate the role of nitric oxide (NO) in early endotoxemia on the systemic and regional blood flow by measuring the plasma nitrite/nitrate (NOx) and blood nitrosyl-hemoglobin (NO-Hb) levels.

This was a prospective, controlled, experimental study conducted in an animal research laboratory on 15 male mongrel dogs. Escherichia coli endotoxin (1 mg/kg) was injected intravenously.

Hepatic, renal, and iliac blood flow and cardiac output (CO) were measured before and 15, 30, 45, 90 and 180 minutes after injection of Escherichia coli endotoxin (1 mg/kg) (n = 6). NOx efflux from the organs was calculated by measuring plasma NOx levels. The arterial blood levels of NO-Hb were also measured (n = 4). As control studies, blood samples from dogs (n = 5) without exposure to endotoxin were assayed at 180 minutes for NOx and NO-Hb. Following endotoxin injection, mean arterial pressure decreased and reached its lowest value at 90 minutes (baseline vs. 90 minutes: 119.1 ± 5.8 vs. 82.5 ± 16.7 mm Hg, P < .0001). Hepatic artery bloodflow increased significantly (baseline vs. 180 minutes: 23.6 ± 12.0 vs. 170.0 ± 68.4 mL/min, P < .0001). There were no significant changes in plasma levels of NOx, uptake or release of NOx across the measured vascular beds, NO-Hb levels at anytime point. In the portal system, the portal vein flow correlated with NOx release (R = 0.69, P < .0001).

In the early phase of endotoxemia in the dog, the significant reduction in systemic vascular resistance and hepatic arterial resistance are not associated with any measurable NOx release in the systemic circulation or the liver.     

Treatment with N-acetylcysteine during acute respiratory distress syndrome: A randomized, double-blind, placebo-controlled clinical study

Intravenous N-acetylcysteine (NAC) has been reported to improve systemic oxygenation and reduce the need for ventilatory support in patients with an acute lung injury. In the more serious form, namely established adult respiratory distress syndrome (ARDS) (Pao₂/Fio₂ ≤ 200 mm Hg), we tested the hypothesis that treatment with intravenous NAC may be beneficial.

Respiratory dysfunction was graded daily according to the need for mechanical ventilation and Fio₂ and to the evolution of the lung injury score (LIS) and the Pao₂/Fi0₂ ratio in 42 patients with established ARDS receiving either NAC 190 mg/kg/day or placebo as a continuous intravenous infusion over the first 3 days of their clinical course.

NAC and placebo groups (22 and 20 patients, respectively) were comparable for demographic characteristics, ARDS categories, severity of illness (simplified acute physiology score [SAPS II]) LIS and Pao₂/Fio₂ ratio. Mortality rate was 32% for the NAC and 25% for the placebo group (difference not significant). At admission (day 1), 91% of patients in the NAC and 95% in the placebo group required ventilatory support; at days 2, 3, 5, and 7 after admission, the percentage of patients receiving ventilatory support was not significantly reduced for both groups in comparison with day 1. Moreover, there were no differences between the two groups at the same observation days. In both groups, the Fio₂ was significantly lower and the Pao₂/Fio₂ ratio was significantly higher than the initial values during the evolution (Fio₂ at day 3, P < .01 for NAC and P < .05 for placebo; Pao₂/Fio₂ at day 3: P < .01 for NAC and P < .02 for placebo), but this improvement was similar for both groups and, moreover, the between-group comparison was never significantly different at the various collection days. The LIS decreased significantly in NAC group between days 1 and 3 (2.23 ± 0.62 v 1.76 ± 0.17; P < .05), whereas no changes were observed in the placebo group; at day 5, there was a significant difference between the two groups (1.53 ± 0.21 for the NAC v 2.15 ± 0.19 for the placebo group; P < .05). In the prevalent sepsis category (10 patients in the NAC and 9 in the placebo group), the mortality rate, the need of ventilatory support, the intensive care unit stay, and the Pao₂/Fio₂ evolution did not differ significantly in both subgroups.

In this relatively small group of patients presenting with an established ARDS subsequent to a variety of underlying diseases, intravenous NAC treatment during 72 hours neither improved systemic oxygenation nor reduced the need for ventilatory support oxygenation nor reduced the need for ventilatory support.     

Increased cardiac output increases lung water in canine permeability pulmonary edema

We investigated the effects of increased cardiac output (CO) on oleic acid pulmonary edema in 14 open-chest, anesthetized, mechanically ventilated dogs. Pulmonary artery wedge pressure (Pawp) was adjusted to approximately 9 mm Hg via a left atrial balloon and CO to 1.7 L · m⁻¹ via systemic arteriovenous fistulas (AVF); five minutes after oleic acid (0.08 mL · kg⁻¹), dogs were randomly divided into two groups, high CO and low CO. In the high CO group, CO was increased by opening the AVFs. Pawp was maintained at 9 mm Hg for four hours in all dogs. The average CO time in the high CO group was 3.9 L · min⁻¹ and 1.3 L · min⁻¹ in the low CO group (P < .01). Lung water accumulation was significantly increased in the high CO group with a wet weight/ body weight ratio of 29 g ò kg⁻¹v 21 g · kg ⁻¹ in the low CO group (P < .004). With time, mean pulmonary artery pressure increased significantly (P < .05) in both groups, but was not different between groups at any time. While pulmonary vascular resistance remained constant in the high CO group, it increased markedly (P < .05) in the low CO group, possibly due to a decrease in pulmonary vascular surface area. The increase in lung water accumulation in the high CO group is probably due to prevention of pulmonary vascular derecruitment and therefore a greater perfused pulmonary vascular surface area.     

Effects of N-acetylcysteine in endotoxic shock

: The release of oxygen-free radicals has been implicated in both peripheral vascular and myocardial alterations of septic shock. N-Acetylcysteine (N-AC), a substrate for the production of glutathione, has potent antioxidant effects. As a nitrosothiol, it may also improve capillary blood flow. We studied the effects of N-AC in a dog model of endotoxic shock.

: Ten pentobarbital-anesthetized, mechanically ventilated dogs were randomly assigned to receive either N-AC (150 mg/ kg loading dose in 1 hour, followed by 20 mg/kg · h maintenance dose) or D5W. After the loading dose, each dog received 3 mg/kg Escherichia coli endotoxin intravenously. After 30 minutes, saline infusion was started to restore and maintain baseline filling pressures.

: The loading dose of N-AC increased Do₂ significantly (from 661 ± 54 to 914 ± 190 mL/min, P < .05), but Vo₂ remained stable. After the administration of endotoxin, fluid challenge restored cardiac output to baseline, in both groups. Hemoglobin and, thus, Do₂ were slightly lower in the N-AC-treated dogs, but Vo₂ was similar in both groups. At the end of the study, O₂ER was significantly higher in the N-AC-treated dogs than in the control dogs. Blood lactate levels fell more rapidly in the N-AC dogs than in the control dogs. Blood lactate levels returned to normal in the N-AC dogs but not in the control dogs. Tumor necrosis factor (TNF) also decreased significantly in the N-AC dogs but remained elevated in the control dogs.

: These data indicate that N-AC administration in endotoxic shock is well tolerated, may increase oxygen availability to the tissues, and is associated with an attenuation of TNF release.     

Role of neurosympathetic pathways in the vascular response to sepsis

The aim of this study was to determine the role of sympathetic neural activity in the hemodynamic adaptations to sepsis in pigs with an emphasis on circuit adaptations. A fall in resistance to venous return (RVR) was predicted in contrast to what was previously observed in sympathetically intact animals that had no change in RVR.

We anesthetized and ventilated 13 pigs and gave 5 mg/kg of indomethacin. We measured cardiac output (CO) by thermodilution and measured pulmonary arterial (PAP), pulmonary capillary wedge (Pcw), right atria) pressure (Pra), and arterial pressure (MAP). Intermittent inflation of a 50-mL balloon in the right atrium was used to transiently arrest the circulation for the measurement of mean circulatory filling pressure (MCFP). RVR was calculated from (MCFP — Pra)/CO. Animals were divided into two groups; 6 received 10 mg/kg of the ganglionic blocker, hexamethonium and norepinephrine to maintain MAP; 7 had their spinal cords cut at C-2. After baseline measurements, all animals received 10 μg/kg/h of endotoxin for 2 hours, and hemodynamic measurements were repeated. Plasma samples were obtained for measurements of immunoreactive endothelin-1 (ET-1), which was assayed by a radioimmunoassay.

Hexamethonium had no significant effect on hemodynamics except for an increase in heart rate. After endotoxin, MAP and SVR fell, PAP rose, and CO and RVR did not change. Spinal section resulted in an increase in heart rate and small increase in PAP and MCFP After endotoxin, there was a further increase in heart rate, PAP, and MCFP with a marked fall in MAP and CO. RVR increased from 2.1 ± 0.46 after spinal section to 3.6 ± 54 mm · min/L (P < .05). ET -1 in the hexamethonium group (n = 2) rose from 2.21 ± .14 to 11.5 ± 2.1 pg/ml at 2 hours, and in the spinal group (n = 7) from 2.04 ± 0.77 to 6.85 ± 3.9 pg/mL at 45 minutes.

Spinal section resulted in a more profound fall in blood pressure and less increase in MCFP than in previously studied animals with sympathetic nervous system intact, but there was still an increase in RVR and PAP ET-1 is a possible mediator of the increase in RVR and PAP.     

Measurement of gastric mucosal carbon dioxide tension by saline and air tonometry

This study compares the balloon air tonometry method of measuring gastric mucosal CO₂ to standard saline tonometry. Also, this study investigates the effect of histamine-2 receptor blockade on the precision of tonometric measures of gastric mucosal Pco₂ (P_tco₂).

We obtained hourly measurements of P_tco₂ from two gastric tonometers inserted orally in 19 healthy volunteers. One tonometer measured P_tco₂ by the intermittent saline method, whereas the other measured P_tco₂ using a newer continuous air method. Subjects received intravenous 5% dextrose during the first 6 hours of the experiment followed by a continuous infusion of a solution of ranitidine in 5% dextrose for another 6 hours. The ranitidine infusion was titrated to maintain gastric fluid pH ≥ 4.

Comparison of air to saline tonometry yielded a bias of −1.3 mm Hg with a limit of agreement of 6.6 mm Hg under optimal conditions of optimal gastric fluid pH (gastric fluid pH ≥ 5.0). Measures of P_tco₂ were lower with ranitidine for either group, 45.3 ± 1.3 mm Hg versus 39.7 ± 0.5 mm Hg for saline (P < .01) and 45.9 ± 1.0 versus 41.3 ± 0.5 for air (P < .01). The mean Pco₂ gap (P_tco₂ - P_arterialco₂) at gastric fluid pH ≥ 5.0 was 1.4 mm Hg, with a standard deviation of 2.7 mm Hg. A span of three standard deviations yields a normal limit for Pco₂ gap of 9.5 mm Hg.

Measures of P_tco₂ with the air tonometer method are similar to those obtained with saline tonometry. The reliability of P_tco₂ measurements with either method improved with the use of ranitidine to maintain gastric fluid pH ≥5.     

Reversal of muscle fatigue in intact rabbits by intravenous potassium chloride

Skeletal muscle fatigue has been associated with potassium efflux from the myocytes, resulting in endogenous increases in blood potassium concentration ([K⁺]). Conversely, exogenous increases in extracellular [K⁺] potentiates contraction in isolated muscle preparations. The mechanisms responsible for these contradictory effects of [K⁺] on skeletal muscle function are unknown. Moreover, little is known about the effect of exogenous increases in [K⁺] on force generation by intact animals, given potassium's deleterious effect on cardiac function.

We compared the response to exogenous increases in blood [K⁺] in rabbits given an infusion of potassium chloride (KCl) intravenously (IV) (0.2 mol/L; KCl group; N = 7) to a group given 0.9% sodium chloride (NaCl) (control; N = 7). The rabbits underwent low-frequency, isometric twitch stimulation of the left hindlimb (square wave pulses 100 microseconds, 40V, 0.25 Hz) throughout the experiment. Both groups received 0.9% NaCl (25 mL/h) during the first hour of twitch stimulation and experienced similar decreases in hindlimb forces to 70% of initial force. A continuous infusion of KCl or of saline (60 mL/h) was started, and hindlimb stimulation continued for 2 hours.

There were no changes in [K⁺] in the control group, and twitch forces progressively declined during the next 2 hours (369 ± 47 g to 279 ± 34 g, P < .01). Arterial [K⁺] increased in the KCI group from 2.6 ± 0.1 to 10.1 ± 0.5 mmol/L (P <.01), and hindlimb twitch forces almost doubled (418 ± 49 g to 756 ± 55 g, P < .01). Force frequency curves showed improved contractility in the KCI group at stimulation frequencies below 30 Hz.

Exogenous increases in blood [K⁺] potentiate skeletal muscle contraction in intact animals and reverse low-frequency twitch fatigue. A possible mechanism may be the maintenance of intracellular [K⁺] by hindering K⁺ efflux from skeletal muscle cells.     

Central venous pressure, pulmonary artery occlusion pressure, intrathoracic blood volume, and right ventricular end-diastolic volume as indicators of cardiac preload

Central venous pressure (CVP), pulmonary artery occlusion pressure (PAOP) and right ventricular end-diastolic volume (RVEDV) are often regarded as indicators of both circulating blood volume and cardiac preload. To evaluate these relationships, the response of each variable to induced volume shifts was tested. The relationships between these variables and cardiac index (CI) and stroke volume index (SVI) was also recorded to assess the utility of each variable as an indicator of cardiac preload. The responses of the new variable intrathoracic blood volume (ITBV) to the same maneuvers was also tested. To examine the effects of changes in cardiac output alone on ITBV, the effects of infusing dobutamine were studied.

Ten anesthetized piglets were studied during conditions of normovolemia, hypovolemia, and hypervolemia. The effects of an infusion of dobutamine were examined under normovolemia and hypovolemia. Cardiac output was measured by thermodilution, and ITBV was measured by double-indicator dilution.

CI was correlated to CVP with r² = .42 (P ≤ .01), to PAOP with r² = .43 (P ≤ 5.01), to RVEDV index with r² = .21 (P ≤ .01), and to ITBV with r² = .78 (P ≤ .01) (pooled absolute values). Bias (mean difference of the percent changes with NORMOVOLEMIA = 100%) ± 1 SD; for SVI - ITBV index was 1 ± 22%, for SVI — CVP it was −128 ± 214%; for SVI — PAOP it was −36 ± 46%; and for SVI -RVEDV index it was 1 ± 29%. Dobutamine infusion increased heart rate (to about 190 × min⁻¹) and CI by 30% in normovolemia and hypovolemia, while ITBV remained basically unchanged.

Under the experimental conditions choosen neither CVP, PAOP, nor RVEDV reliably indicated changes in circulating blood volume, nor were they linearly and tightly correlated to the resulting changes in SVI. ITBV reflected both changes in volume status and the resulting alteration in cardiac output. The possibility that ITBV might be cardiac output-dependent was not supported. ITBV, therefore, shows potential as a clinically useful indicator of overall cardiac preload.     

Depletion of myocardial adenosine triphosphate during prolonged untreated ventricular fibrillation: effect on defibrillation success

We studied left ventricular endomyocardial adenosine triphospate levels in 13 large mongrel dogs before and during ventricular fibrillation induced cardiac arrest to assess whether myocardial adenosine triphosphate content could predict successful cardiopulmonary resuscitation. Endomyocardial biopsies were performed during sinus rhythm (control), after 15 min of ventricular fibrillation or 10 min of ventricular fibrillation and 5 min of open chest cardiopulmonary resuscitation, after 20 min of ventricular fibrillation and 10 min of open chest cardiopulmonary resuscitation and after 40 min ventricular fibrillation and 15–20 min open chest cardiopulmonary resuscitation. Myocardial adenosine triphosphate was measured utilizing a bioluminescence method adapted for use with endomyocardial biopsies and normalized to protein content. Left ventricular endomyocardial adenosine triphosphate content fell significantly over time from a control level of 8.88 ±0.9 pg/mg protein to 5.73 ± 0.5 pg/mg protein at 15 min of cardiac arrest, to 3.4 ± 0.4 μg/mg protein after 30 min of cardiac arrest and to 1.98 ± 0.3 μg/mg protein after 60 min of cardiac arrest (P < 0.001). Adenosine triphosphate levels were significantly different between animals that received 10 min of ventricular fibrillation and successful open chest cardiopulmonary resuscitation and those that received 40 min of ventricular fibrillation and unsuccessful open chest cardiopulmonary resuscitation (4.35 ± 0.48 vs. 2.11 ± 0.43 μg/mg protein; P < 0.025). Endomyocardial adenosine triphosphate levels falling below 3.5 μg/mg protein were associated with only 2/6 animals being successfully resuscitated, while 6/7 successfully resuscitated animals had adenosine triphosphate levels 3.5 μg/ mg protein (Positive Predictive Value = 0.75, Negative Predictive Value = 0.80). Myocardial adenosine triphosphate content diminishes significantly during prolonged ventricular fibrillation and once levels fall below 3.5 μg/mg protein, successful resuscitation is rare.     

Influence of preoperative mechanical bone quality and bone mineral density on aseptic loosening of total hip arthroplasty after seven years

Objective. To test mechanical bone quality and bone mineral density of the femoral head at the day of implantation as indicators for femoral prosthesis loosening.

Methods. Mechanical bone quality of a femoral head slice was assessed by destructive compression testing combined with bone mineral density measurements using peripheral quantitative computed tomography. Fourteen patients with walking pains were attainable for a radiographical follow-up mean 7.1 years after implantation.

Results. Radiolucent lines along the stem were evident in 11 of 14 femurs, most of them seen in Gruen zones 7, 6, 1, 3, 14, and showed strong correlations to preoperative bone strength (r=−0.80; P<0.001) and axial stiffness (r=−0.75; P=0.002), yet not to bone mineral density (r=−0.67; P=0.009). Slight varus deviations <3° were noted in six femurs. Preoperative strength was reduced in this femurs to 54% (P=0.006), and stiffness to 61% (P=0.038), while bone mineral density did not differ significantly.

Conclusions. Femoral prosthesis loosening after seven years can be predicted by mechanical bone quality of the femoral head at the time of implantation. Bone mineral density measurements may also indicate future stem loosening but have to interpreted carefully, keeping in mind a poorer predictive value.

Relevance Indications and choice of type of hip arthroplasty should be balanced in osteoporotic bones in particular. While preoperative bone mineral density measurement allows the prediction of mechanical bone quality, its relevance in predicting failure in arthroplasty treatment remains unclear.     

Ischemia/reperfusion injury to the ileum does not account for the ileal Vo2-Do2 alterations induced by endotoxin

Endotoxin (lipopolysaccharide [LPS])-induced systemic organ injury leads to disruption of normal systemic organ metabolic processes, which are manifest clinically by signs of accelerated anaerobic metabolism (eg, tissue acidosis and hyperlactatemia) and altered Vo₂-Do₂ relationships. The association of increased anaerobic metabolism with Vo₂-Do₂ alterations has led to the notion that ischemia/reperfusion (I/R) injury may be a prerequisite for the development of Vo₂-Do₂ alterations during endotoxemia. However, in contrast to sepsis, in which oxygen consumption is often increased, oxygen consumption is severely decreased after I/R injury. Based on these observations, we hypothesized that I/R injury would result in systemic organ Vo₂-Do₂ alterations, which are distinct from those that occur in sepsis.

We used the in situ autoperfused feline ileal preparation to simultaneously examine microvascular permeability, reflected as the ileal lymph to plasma protein concentration ratio (C_L/C _P), and ileal Vo₂-Do₂ relationships after either intravenous LPS (2.0 mg/kg; N = 5) or I/R injury (n = 5), and in matching controls (n = 5).

As expected, all LPS-treated and I/R-injured animals were found to have extensive ileal histological damage and marked increases in the C_L/C_P compared with controls (0.315 ± 0.009 and 0.329 ± 0.034, respectively, v 0.097 ± 0.009; P < .001, both comparisons). In addition, the critical Do₂ (Do₂c) was elevated, and the critical oxygen extraction was decreased in both the I/R and LPS groups relative to controls. However, as initially hypothesized, the Vo₂ at the critical Do₂ was markedly decreased in the I/R group compared with that of the LPS group.

These data indicate that I/R injury is insufficient to account for the systemic organ Vo₂-Do₂ alterations that occur with LPS injury.     

Noninvasive assessment of regional albumin extravasation in porcine septic shock

In order to study noninvasively the regional distribution of changes in microvascular albumin flux in septic shock, ¹³¹I human serum albumin and ^98mTc red blood cells were injected IV into anesthetized pigs randomized to receive saline (n = 4) or live E. coli bacteria 3 × 10⁸ × kg⁻¹ IV (n = 8). Images of thorax, abdomen, and left hindlimb were obtained using a gamma camera and a computer. Septic pigs developed pulmonary hypertension and low cardiac output circulatory shock. Hematocrit rose and plasma colloid osmotic pressure fell. Regions of interest were drawn in the ^99mTc images. For each region we calculated an albumin leak index. The albumin leak index over the lungs was higher in sepsis than in controls (1.62 ± 0.42 × 10⁻³ v 0.45 ± 0.05 × 10-3 × min⁻¹, P < .005), even when divided by the pulmonary intravascular filtration pressure, incorporating hydrostatic and colloid osmotic pressures (P < .01). The index was also higher in abdominal regions (for the peripheral abdomen, 6.07 ± 2.98 × 10⁻³ v 0.40 ± 0.72 × 10⁻³ × min⁻¹, P < .005), but not in the hindlimb. In sepsis, the albumin leak index increased more in the abdomen than in the lungs (P < .05). We conclude that in porcine septic shock pulmonary microvascular permeability is increased, but the microvascular albumin flux and thus plasma extravasation increased more in the abdomen than in the lungs.     

Heat shock protein (HSP70) expression in septic patients

: This study investigates heat shock protein 70 (HSP₇₀) expression by peripheral blood mononuclear cells (PBMCs) of septic patients admitted to an intensive care unit and examines the possibility of a correlation between HSP₇₀ levels and plasma tumor necrosis factor alpha (TNF-) concentrations. Additionally, we evaluated whether the HSP₇₀ production could be regarded as a prognostic factor for the development of septic shock as well as for patient survival.

: Blood samples of 29 patients were taken 24 hours after the diagnosis of sepsis. HSP₇₀ expression and TNF- level were measured using indirect immunofluorescent analysis and a commercially available enzyme-linked immunosorbent assay method, respectively.

: PBMCs expressed significantly high levels of HSP₇₀ (11.9 ± 5.6 [sd]) compared with those of the healthy control group (3.2 ± 2.1 % positive cells). Such enhanced levels were correlated to plasma TNF- concentrations (r = .99, P < .01). This study failed to demonstrate a relationship between HSP₇₀ production and clinical outcome.

: These findings give further evidence that also in humans, heat shock response is activated during sepsis. The correlation observed between HSP₇₀ overproduction and TNF- plasma concentrations suggests that HSP₇₀ exerts a possible protective effect against TNF- cytotoxicity. Such hypothesis has not been confirmed by our clinical data.     

Advantages of dextran 70 over ringer acetate solution in shock treatment and in prevention of adult respiratory distress syndrome. A randomized study in man after traumatic-haemorrhagic shock

Twenty-three patients aged 20–58 years in a serious state of shock and suffering from major pelvic and femoral fractures sustained in traffic accidents were studied. As initial fluid management, either dextran 70 (1000–1500 ml) together with Ringer's acetate solution (2000–3000 ml), or Ringer's acetate alone (5000–8000 ml), was administered on a random basis. Thus, 12 patients received dextran and 11 patients crystalloid treatment to counteract shock. Both groups were given whole blood. The resuscitation time-i.e. the interval from the start of fluid therapy until a stable circulatory condition was achieved-was significantly shorter (P < 0.001) in the dextran group (108 ± 18 min; mean ± S.D.) than in the Ringer group (170 ± 43 min). During the post-resuscitative observation period of 6 days the dextran patients were given 500 ml of dextran daily, while the Ringer group did not receive any colloidal solution. To maintain a stable circulation and a urinary output above 50 ml/h the Ringer patients required significantly more (P < 0.001) crystalloid solution (910 ± 300 ml) daily than the dextran patients (460 ± 400 ml).

The frequency of adult respiratory distress syndrome (ARDS) was significantly lower (P < 0.05) in the dextran group (0 of 12) than in the Ringer group (4 of 11). Thus, in the initial treatment of traumatic-haemorrhagic shock and in the post-resuscitative period dextran 70 would seem of advantage over Ringer's acetate both in shortening the shock period and in reducing the frequency of adult respiratory distress syndrome.     

Canine biventricular performance during acute progressive pulmonary microembolization: Regional myocardial perfusion and fatty acid uptake

The cardiac output during acute pulmonary artery hypertension (PAH) may be compromised by right ventricular (RV) outflow obstruction, myocardial ischemia, or adverse ventricular interactions. To study the relative contributions of these mechanisms to impaired biventricular function during PAH, we injected 75 to 105-μm glass beads into the pulmonary vasculature of 11 dogs and correlated the hemodynamic alterations with changes in biventricular function. Biventricular ejection fractions and interventricular septal motion were evaluated by gated blood pool angiocardiograms. Regional myocardial blood flow (n = 8 dogs) was measured with microspheres, and regional myocardial fatty acid metabolism during PAH (n = 8 dogs) was assessed by measuring the regional myocardial extraction of the radioiodinated fatty acid analog, 15-(p-iodophenyl)-3-methylpentadecanoic acid (3m-PIP/PDA). Following microembolization, the mean pulmonary artery pressure increased from 14 ± 1 to 48 ± 2 mm Hg (P < .001) and cardiac output decreased from 3.5 ± 0.3 to 2.6 ± 0.3 1 /min (P < .05). Right ventricular volume increased and RV ejection fraction decreased progressively. Abnormal septal motion during PAH occurred in 5 of 11 dogs and was associated with decreased LV diastolic compliance. Right ventricular myocardial blood flow remained increased during PAH while the regional extraction of 3m-PIP/PDA was uniform throughout the heart. These findings suggest that during moderate PAH the canine cardiac output is limited by RV outflow obstruction and decreased LV diastolic compliance rather than by ischemia.     

Systemic oxygen delivery and consumption during acute lung injury in dogs

When oxygen delivery (QO₂ = Qt × CaO₂) is reduced in healthy animals, oxygen uptake by tissues (VO₂) is maintained by an increase in the O₂ extraction ratio ((CaO₂ − CvO₂)/CaO₂) until a critical level of delivery QO₂C is reached, below which VO₂ becomes dependent on QO₂. In patients with adult respiratory distress syndrome (ARDS), VO₂ becomes dependent on O₂ delivery even at high levels of QO₂ implying a systemic O₂ extraction defect. To determine whether lung injury, or its management with positive end-expiratory pressure (PEEP) and high inspired oxygen fractions (F1O₂), might disrupt O₂ extraction by peripheral tissues, we compared the critical O₂ delivery and extraction ratio in 30 anesthetized and paralyzed dogs in four groups. One group (n = 7) received oleic acid to produce an acute lung injury (F_lO₂ = 1.0) and a second group (n = 7) was maintained with PEEP (F_lO₂ = 1.0) after receiving oleic acid. A third group (n = 8) received no oleic acid (F_lO₂ = 1.0), and a fourth group received no oleic acid and was ventilated with room air. The critical O₂ delivery was determined in each animal as QO₂ was lowered in a stepwise manner by controlled blood removal. Neither the critical QO₂ nor critical O₂ extraction ratios were different (P> .05) among oleic acid, PEEP-treated, or hyperoxic control animals. Similarly, the critical QO₂ determined from the point at which arterial lactate began to increase showed no difference among groups. Thus, neither acute lung injury produced with oleic acid nor its treatment with PEEP was associated with a peripheral O₂ extraction defect. However, a linear regression across all groups demonstrated a positive correlation between the arterial PO₂ and corresponding critical O₂ delivery (r = < P < .05). These results suggest that high arterial O₂ tensions may contribute to an impairment of peripheral O₂ extraction.     

Effects of respiratory therapy on intracranial pressure

To determine whether respiratory therapy can be deleterious in patients with intracranial hypertension, we investigated the effects of passive and active respiratory therapy on intracranial pressure (ICP) and cerebral perfusion pressure in acute neurosurgical patients with normal (< 15 mm Hg) and increased (> 15 mm Hg) ICP. During passive respiratory therapy, moving the patients from a 30-degree head-up dorsal position to the lateral decubitus position increased ICP by 10 mm Hg (P < .001) in normal ICP patients (n = 26) and by 6 mm Hg (P < .001) in high ICP patients (n = 6). Cerebral perfusion pressure increased by 5 mm Hg (P < .05) in both cases. Clapping did not further change ICP or cerebral perfusion pressure. During active respiratory therapy, ICP decreased by 2 to 3 mm Hg (P < .05) in normal ICP patients (n = 11) and by 4 to 7 mm Hg (P < .01) in high ICP patients (n = 7). Cerebral perfusion pressure decreased by 1 to 5 mm Hg and remained unchanged in normal and high ICP patients, respectively. The ICP reduction was attributed mainly to the decrease in PaCO₂ and the decrease in intrathoracic pressure during active hyperventilation. These results indicate that passive and active respiratory therapy can be performed safely in acute neurosurgical patients with normal or increased ICP, provided the therapy can be done in a position that does not deteriorate ICP.     

Passive compliance and length of clinically short hamstring muscles of healthy men

This study examined the passive compliance and length of short hamstring muscles in relation to hamstring muscles not considered short in healthy men (ages 18–37). The right hamstrings of 30 men with straight-leg raising of 73.8° (group I) and 24 men with straight-leg raising of 61.0° (group 11) were compared. Subjects were positioned on their left sides with the pelvis stabilized and the right thigh fixed at 90°. Subjects received three maximal passive knee extension trials for data collection. Muscle activity was monitored with surface electromyography and passive resistance was measured with a dynamometer as the limb was photographed at force-dependent positions. Passive compliance was computed as the ratio of change in the knee angle to change in passive torque. Hamstring lengths were measured simultaneously. Results showed that the passive compliance curves for group II were shifted left compared to group I. revealed that the initial knee angles for group 11 were greater than for group I (P = 0.001), as were the maximal knee angles (P < 0.001). Passive compliance ratios for group 11 (1.29) were less than for group I (1.45), but not significantly different. Maximal passive torques were not different between groups. The change from initial muscle lengths to maximal lengths was less for group 11 than group I for the: (1) absolute length change (P = 0.027), (2) per cent change beyond initial length (P = 0.005), and (3) length change standardized as a percentage of femur length (P = 0.011).