Stephen M. Cain

Supply Dependency of Oxygen Uptake in ARDS: Myth or Reality?

Several reports state that oxygen uptake changed in direct correlation with changes in total oxygen delivery to the tissues in the adult respiratory distress syndrome (ARDS). Oxygen uptake appeared to be limited by oxygen delivery even at normally adequate levels so that uptake was abnormally dependent on supply. These reports are discussed with respect to whether or not such a result could have been due to errors in measurement or to mathematical coupling by relating two quantities that shared a common variable. Having rejected that proposition, animal experiments are cited in which abnormal oxygen supply dependency was produced by microembolization. The accompanying loss of reactive hyperemia and inability to extract oxygen were consistent with a progressive loss of recruitable capillaries. Evidence is presented that the potential for embolization in ARDS is greatly enhanced by activation of the complement and arachidonic acid cascades as well as by the xanthine oxidase system. The resultant use of molecular oxygen by non-ATP producing oxidase systems might also account for the increase of supply dependent oxygen demand in ARDS.

Experimental models of pathologic oxygen supply dependency.

Pathologic oxygen supply dependency is an abnormal situation in which oxygen uptake (Vo2) varies directly with oxygen delivery. Its presence in patients with adult respiratory distress syndrome and/or sepsis has been associated with particularly high mortality rates that may be the result of tissue hypoxia that causes multiple organ failure. The evidence for this association has been indirect because we cannot use invasive methods that would be necessary to verify or disprove the hypothesis. Because further progress will depend on the development of adequate animal models of pathologic oxygen supply dependency, we have attempted to evaluate some of the available information in this area as well as the likelihood that tissue hypoxia will prove to be the precipitating factor. In anesthetized dogs injected or infused with endotoxin, many of the features of pathologic oxygen supply dependency have been successfully produced. These features include defective peripheral oxygen extraction, increased oxygen demand, and increased lactate levels. Regional measurements have shown that gut Vo2 decreases before other areas, particularly skeletal muscle. Lactate measurements alone were shown not to be sufficient proof of tissue hypoxia. More direct measurements of actual energy states and tissue Po2 are indicated for future research efforts.

Systemic and regional oxygen uptake and delivery and lactate flux in endotoxic dogs infused with dopexamine.

ObjectiveTo test whether dopexamine, a dopaminergic and β2-adrenergic receptor agonist, would: a) direct a greater share of cardiac output to gut than to muscle when used to increase systemic oxygen delivery (Do2) in endotoxic dogs; and b) enhance the ability of peripheral tissues to extract oxygen. DesignTwo groups of eight dogs infused for 1 hr with 2 mg/kg Escherichia coli endotoxin. One group was continually infused with dopexamine (12 μg/min.kg) and the other group was not (control group). After 2 hrs, oxygen extracting ability was challenged by changing inspired gas to 12% oxygen for 30 mins. SubjectsAnesthetized, paralyzed, pumpventilated mongrel dogs. InterventionsDonor RBCs and dextran used during endotoxin infusion to maintain cardiac output while preserving hematocrit near 40%. Measurements and Main ResultsIn the dopexamine-treated group, cardiac output, systemic Do2, and oxygen consumption (Vo2) were higher than in the control group during the first 90 mins, but were not thereafter. Gut and muscle blood flow did not differ between groups, but the fraction of cardiac output going to each region tended to be less in the dopexamine-treated dogs. Arterial lactate values increased to about 6 mmol/L in all dogs. In both groups, limb muscle first produced lactate but then took up lactate after the first hour. The gut in controls converted from lactate uptake in the first hour to producing about 20 μmol/min.100 g, whereas the gut never produced lactate in the dopexamine-treated group. During hypoxia, systemic Do2 and Vo2 decreased only in the dopexamine-treated group, even though oxygen extraction was only slightly above 40%. Oxygen extraction was not demonstrably improved by dopexamine treatment. ConclusionsDopexamine temporarily increased systemic Do2 and Vo2 in volume-expanded endotoxic dogs during normoxia and may have caused gut mucosa to be preferentially perfused and thus to be kept better oxygenated. (Crit Care Med 1991; 19:1552)

Peripheral vascular responses to fluorocarbon administration

To detect the local effect of hyperoxia on skeletal muscle vasculature, 2.5-ml boluses of oxygenated or deoxygenated fluorocarbon emulsion (F-O2 or F-N2) were washed through the hindlimb of anesthetized dogs at prevailing arterial pressure. Instantaneous hematocrit changes at the outflow were registered and stored in digital form with the red cells serving as the nondiffusible tracer in the resulting washout curves. A gamma density function was fitted and the gamma index (1/square root of alpha) was derived as a measure of skewness or perfusion heterogeneity. After recovery from the initial hypotensive reaction to fluorocarbon emulsion, washout curves for F-O2 and F-N2 were registered and blood samples were taken during 40 min of normoxia followed by 40 min of hypoxic hypoxia. The initial reaction to fluorocarbon significantly increased the gamma index so that the experiments began with a high index of perfusion heterogeneity in the limb vasculature. No significant difference was seen between F-O2 and F-N2 in normoxia but F-O2 maintained greater heterogeneity during hypoxia. The increased heterogeneity observed after the fluorocarbon reaction correlated highly with the severity of the hypotensive reaction which was also found to correlate inversely with the ability of the limb musculature to increase the O2 extraction ratio with onset of hypoxia. This blunting of microcirculatory reactivity to hyperoxia and hypoxia was attributed, in part, to the initial transient fluorocarbon reaction, possibly mediated by complement activation.

Critical levels of O2 extraction following hemodilution with dextran or fluosol-DA☆

We compared the effect of two different blood diluents, dextran (DEX) and a perfluorochemical, on O2 delivery and use during progressive hemorrhage. Two groups of six dogs each were anesthetized, paralyzed, and pump-ventilated. One group was hemodiluted with Fluosol-DA 20% (FDA; Green Cross, Osaka, Japan) and the other with DEX to hematocrits of 26.1%. 1 ± 1.5% and 23.0% ± 2.8% (P > .05), respectively. Cardiac output and O2 uptake were measured as the animals were bled in stages to lower O2 delivery (cardiac output × arterial O2 concentration). Sufficient measurements were made to ascertain the plateau relationship, when O2 uptake was independent of delivery, and the downslope, when it was dependent. The intersection of the two straight lines that fit these parts of the curve was taken as the critical O2 delivery. The O2 extraction fraction at the critical level was significantly higher (P < .01) in the FDA group (0.79 ± 0.02) than in the DEX group (0.60 ± 0.05). Similarly, at the critical level, mixed venous PO2 was significantly lower (P < .05) in the FDA group (24.8 ± 2.5 torr) than in the DEX group (32.3 ± 2.2 torr). Significantly lower mixed venous PO2 and higher O2 extraction fractions at the critical delivery point in the FDA group suggested that FDA promoted the diffusion of O2 between the tissue site of use and RBCs.

Systemic and muscle oxygen uptake/delivery after dopexamine infusion in endotoxic dogs.

Background and MethodsThis study was designed to test whether dopexamine, a dopaminergic and β2-adrenergic agonist, would a) increase systemic oxygen delivery (Do2) in endotoxic dogs, and b) interfere with the ability of resting skeletal muscle to extract oxygen. There were three treatment groups (n = 6 in each group): control, endotoxin alone (E) 4 mg/ kg iv, and endotoxin + dopexamine (E + D) 12 μg/kg-min. Data were analyzed between and within groups by split-plot analysis of variance with significance of identified differences tested post hoc by Duncans multiple range test. Donor RBC and dextran were used after endotoxin to maintain adequate perfusion pressures, with Hct kept near 40%. Blood flow to left hindlimb muscles was decreased in controlled steps of 15 min each after stabilization. ResultsIn E group, cardiac output (Qt), mean arterial pressure (MAP), systemic Do2, and oxygen uptake (Vo2) decreased despite blood volume expansion. In E + D group with similar volume expansion, dopexamine maintained Qt, systemic Do2, and Vo2 near the control levels, although MAP and systemic vascular resistance were reduced. In comparison with control subjects, endotoxin increased critical Do2 in the isolated limb muscles from 4.6 to 7. mL/kg-min and decreased critical oxygen extraction from 81% to 68%. The pressure/flow relationship in the limb became flattened, indicating loss of vascular reactivity. In the E + D group, there was no further change in the pressure/flow curve nor in the critical oxygen extraction level. ConclusionsDopexamine provided hemodynamic support for endotoxic dogs, thereby increasing total DO2 and VO2, while not altering oxygen extraction in the muscle. (Crit Care Med 1991: 19:198)

Influence of Oxygen on Endothelium-derived Relaxing Factor/nitric Oxide and K+-dependent Regulation of Vascular Tone

We investigated the effect of hypoxia on acetylcholine (ACh) stimulated, endothelium-derived relaxing factor/nitric oxide (EDRF/NO)-dependent relaxation, and on basal tension in rat aortic rings. ACh (10(-9)-10(-6) M)-mediated relaxation at high [95%, Emax -76.2 +/- 4.5% of phenylephrine (PE)-induced constriction] and normal (20%, Emax -81.2 +/- 3.6%) O2 levels was inhibited by hypoxia (5%, Emax -36.2 +/- 7.2%); residual hypoxic relaxation was blocked by the K+ channel antagonist glibenclamide. To address whether O2 influenced EDRF/NO and K+ channel contributions to basal tone, the effect of stepwise reduction of available O2 (95, 20, 5, and 0%) was studied in intact and endothelial cell (EC)-denuded rings. The effects in these rings were compared with results of the same progressive reduction in O2 in the presence of the NO-synthase inhibitor N omega-nitro-L-arginine methyl ester (L-NAME) (10(-4) M) or glibenclamide (10(-4) M). EC-intact and EC-denuded rings constricted to 0.80 +/- 0.10 and 1.41 +/- 0.15 g, respectively. Reducing O2 to 20% had no significant effect on vascular tension, but 5% caused constriction (p < 0.05) in EC-intact rings (0.90 +/- 0.15 g). This hypoxic vasoconstriction was blocked by L-NAME, but not by glibenclamide, suggesting that hypoxic vasoconstriction was mediated by withdrawal of EDRF/NO. In contrast, EC-denuded rings showed a significant relaxant response at 5% O2. When O2 was then reduced further (95% N2/5% CO2), both EC-intact and EC-denuded rings relaxed, and this relaxation reached baseline tension (0.10 +/- 0.1 g).(ABSTRACT TRUNCATED AT 250 WORDS)

Effects of time and microembolization on O2 extraction by dog hindlimb in hypoxia

Abstract Microembolization of peripheral vascular beds has been suggested as a mechanism for the pathological O2 supply dependency seen in adult respiratory distress syndrome (ARDS). To test the feasibility of this, we injected 19 million 14pm microspheres per 100 g of muscle into the arterial supply of the left hindlimb in anesthetized dogs (n = 8). Venous outflow from the limb was isolated and flow, O2 concentration, and blood gas tensions of arterial and venous blood were measured before and after embolization (EMB) and then during 60 minutes of hypoxia induced by ventilating the paralyzed animal with 9% O2-91 % N2. Another group (n = 8) was not embolized and was treated the same for comparison. Limb blood flow increased transiently with EMB but venous PO2 remained elevated (P

Oxygen Supply Dependency in the Critically Ill — A Continuing Conundrum

There was little dispute that endotoxin treatment of experimental animals could recreate the O2 extraction defect that had been observed in critically ill patients. The remaining question was whether or not this necessarily signified pervasive tissue hypoxia. Some limitation to O2 diffusion in the tissues had been postulated because of known effects of endotoxin that ultimately result in damage to endothelium. We were unable to alter the critical DO2 or 0(2)ER in endotoxic dogs by manipulating the arterial PO2. This tended to rule against there being a diffusion limitation created by the endotoxin as a result of endothelial disruption or microvascular dysfunction. The results of the DCA and dopexamine experiments served to remind us that arterial lactate measurements may or may not indicate widespread tissue hypoxia. Sepsis, as emulated by endotoxin infusions, is also a metabolic disease that can cause inactivation of PDH and thus cause lactacidosis without tissue hypoxia. Regional measurements of lactate flux indicated that gut was hypoxic in spite of DO2 above critical because of maldistribution of blood flow between muscularis and mucosa. The questions persist of how much tissue hypoxia is caused by sepsis or endotoxin when DO2 is supported at supposedly adequate levels and whether there are marked regional differences. Such questions still await answers. Newer technological advances that permit assessment of tissue oxygenation by noninvasive methods, such as near infrared spectrophotometry or nuclear magnetic resonance measurement of tissue energy potential, may soon be feasible in critically ill patients. This kind of information will be of vast importance in designing the most effective therapeutic regimen.

Oxygen Delivery and Intentional Hemodilution

Normovolemic hemodilution can progress to a classic form of hypoxia called anemic anoxia in the terminology first used by Barcroft (1920). It differs from anoxic anoxia, or hypoxic hypoxia in more modern terminology, in that arterial PO2 can be quite normal but the arterial O2 concentration is rower than normal. The question naturally arises, therefore, of why one would intentionally hemodilute in a clinical setting. The reason becomes clearer when consideration isgiven to the desirability of sparing as much as possible the use of allogeneic blood because of its additional risk factors. Furthermore, hemodilution within specified limits may actually be desirable by virtue of increased blood fluidity and consequently better tissue oxygenation under some circumstances. There is even evidence that tissue PO2 in various organ systems may actually increase with mild hemodilution (Messmer et al., 1973). Given the desirability of preoperative autologous blood donation and/or volume expansion with cell-free diluents such as dextran, albumin, and hetastarch, it is worthwhile to ask what the lower limits of hemodilution might be. Such limits are set to some extent by the ability of the body to compensate for the decrease in oxygen carrying capacity. The nature of those compensatory actions, their effectiveness, and the net effect upon tissue oxygenation in health and disease states are the topics that will be discussed.