INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ADEQUATE OXYGEN DELIVERY to the brain is important for its function because of its low capacity for anaerobic metabolism. The blood components relevant for the oxygen delivery to the brain are viscosity and oxygen content. Under physiological conditions, moderate changes of either blood viscosity or oxygen content are not followed by a change in oxygen delivery indicating a vascular compensation (10, 17, 22). According to Poiseuille's law, an increase in blood viscosity should result in a lowered blood flow to the brain resulting in decreased cerebral oxygen delivery. The unchanged cerebral blood flow (CBF) at an increased blood viscosity (2, 25) indicates a compensatory vasodilation that keeps cerebral oxygen transport constant.

Concerning the second component, the arterial oxygen content (Ca_O2), corresponding mechanisms of compensation appear to exist. During hemodilution, an inverse linear relationship between Ca_O2 and CBF has been observed (1, 10, 21, 27). This relationship is best explained by cerebral vasodilation at reduced Ca_O2 during hemodilution. The increase in CBF allows maintenance of cerebral oxygen delivery at a decreased Ca_O2. Changes in CBF that take place under these conditions may result from a mechanism sensitive to the Ca_O2 (10, 12).

From these findings, it can be postulated that an increase in blood viscosity as well as hemodilution induces compensatory vasodilation resulting in a maintained oxygen delivery to the brain. If this hypothesis is correct, compensatory vasodilation should become exhausted when both factors are combined. To test this hypothesis blood viscosity was increased in rats at different degrees of hemodilution. Oxygen delivery was calculated from the values of CBF and Ca_O2 measured during different degrees of hemodilution. Specifically, two groups of conscious rats were compared. In both groups Ca_O2 was gradually decreased by blood exchange and isovolemic volume replacement by a nonoxygen-carrying solution until set hematocrit (Hct) values were reached. One group received only blood replacement (BR) and served as a hemodilution control group. In the other group, blood viscosity was elevated by intravenous administration of high-molecular-weight polyvinylpyrrolidone (BR/PVP). In both groups, mean and local CBF were measured by the 4-iodo[N-methyl-¹⁴C]antipyrine method. To obtain cerebral oxygen transport, Ca_O2 was also determined.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals. The experiments were performed on 48 male Sprague-Dawley rats weighing 220-440 g (Zentralinstitut für Versuchstierzucht, Hannover, Germany). Experiments were in accordance with institutional guidelines. The animals were maintained under temperature-controlled environmental conditions on a 14 h:10 h light/dark cycle. The animals were fed a standard diet (Altromin C 1000: Lage, Germany) and allowed free access to food and water ad libitum until the experiments were started.

Experimental procedure. The experimental design is outlined in Fig. 1. Before surgery, the rats were anesthetized by inhalation of a gas mixture of halothane (1-2%), N₂O (60-70%), and oxygen (remainder). Anesthesia was maintained during surgery by inhalation of the gas mixture via nose cone. Body temperature was held at 37-37.5°C by a temperature-controlled heating pad. Polyethylene catheters (PE-50, Labokion; Sinsheim, Germany) were inserted into the right femoral artery and vein. After surgery, the animals were placed in a rat restrainer (Braintree Scientific; Braintree, MA) and were allowed 120 min to recover from anesthesia before blood exchange was started in conscious animals. Blood pressure was monitored continuously by a quartz pressure transducer (Hewlett-Packard; Palo Alto, CA).

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Flow chart of the experimental protocol. Hct, hematocrit; PVP, polyvinylpyrrolidone; BR, blood replacement; n = number of rats.

Measurement of local CBF. Thirty minutes after the reduction of Hct and infusion of PVP, the local CBF (LCBF) was measured according to the method of Sakurada et al. (18). The operational equation for measurement of LCBF on the basis of the principles of inert gas exchange modified for nonvolatile tracers is presented below
where t is the variable time; C_i(T) the tissue concentration of the tracer at the time t when blood flow is terminated; C_a, the tracer concentration in the arterial blood; and  the brain-blood partition coefficient for iodo[¹⁴C]antipyrine. The constant K equals F/W , where F/W is the rate of blood flow per unit mass of tissue. Because the tissue concentrations and the time course can be derived, LCBF can be calculated from the value of K if the brain-blood partition coefficient for the tracer is known. For the measurement of LCBF, iodo[¹⁴C]antipyrine (specific activity 54 mCi/mM, DuPont-NEN) dissolved in 1 ml of saline was continuously infused at a progressively increasing infusion rate for 1 min via a femoral venous catheter. The progressively increasing infusion rate was a modification of the method described earlier (18). It was chosen to minimize equilibration of rapidly perfused tissues with arterial blood during the period of measurement. During the 1-min infusion period, 12-18 timed blood samples were collected in drops from the free-flowing arterial catheter directly onto filter paper disks (1.3 cm diameter) previously placed in small plastic beakers and weighed. The samples were weighed and radioactivity estimated with a liquid scintillation counter (Tri-Carb 4000 series; Canberra Packard; Frankfurt, Germany) after extraction of the radioactive compound with ethanol. After the 1-min infusion and sampling period, the animal was decapitated, and the brain was removed as quickly as possible and frozen in 2-methylbutane chilled to 40°C to 50°C with dry ice. The frozen brains were coated with chilled embedding medium (Lipshaw; Detroit, MI), stored at 80°C in plastic bags, and then sectioned into 20-µm slices at 20°C in a cryostat. Autoradiographic images were converted to digitized optical density images by an image processing system (MCID, Imaging Research; St. Catharine's, Canada). Tissue optical densities were converted to ¹⁴C concentration ([¹⁴C]) by comparison with the precalibrated standards. LCBF was calculated from local concentrations of [¹⁴C] and the time course of the blood iodo[¹⁴C]antipyrine concentrations corrected for the lag and washout in the arterial catheter. The washout correction rate constant was 100/min. The brain-blood partition coefficient for iodo[¹⁴C]antipyrine was 0.9 in our rats. With this technique, LCBF is highly correlated with local glucose consumption (13).

Statistical evaluation. Statistical differences were evaluated by the unpaired t-test comparing the groups at the same Hct. Data were presented as means ± SD.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Physiological variables. Table 1 shows the physiological variables immediately before infusion of iodo[¹⁴C]antipyrine. For reasons of clarity, the baseline values are not shown because differences could not be detected among groups in all physiological variables.

Viscosimetric data. Figure 2 summarizes the viscosimetric measurements of whole blood samples taken after hemodilution to Hct of 30, 22, 19, or 15% with or without PVP at different shear rates. We observed a shear rate-dependent blood viscosity at all Hct levels, which is in accordance with the well-known non-Newtonian fluid behavior. Hct reduction resulted in a gradual decrease of blood viscosity. In the groups with the combination of hemodilution and PVP (BR/PVP) a significantly higher whole blood viscosity was measured than in untreated animals (baseline, Hct >40). Quantitatively blood viscosity increased twofold with PVP at Hct of 30% compared with baseline. With progressive Hct reduction the PVP induced blood viscosity difference from the hemodiluted value increased four- to sixfold.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Whole blood viscosity at different shear rates depending on Hct. Values are means ± SD. Blood viscosity was determined at baseline (Hct >40; ) and after hemodilution to a Hct of 30, 22, 19, or 15% with either only hemodilution or with addition of PVP. Hct 30% (), Hct 22% (), Hct 19% (), Hct 15% (). Hemodilution + PVP was tested versus baseline and tested versus hemodilution at the same Hct level. *P < 0.05 BR vs. BR/PVP at the same Hct level.

CBF and oxygen transport. Mean CBF values of the different groups are shown in Fig. 3. The corresponding values of LCBF are listed in Table 2. LCBF was measured in 34 different brain structures. At the Hct levels of 30 and 22%, no differences in mean CBF and LCBF could be detected between the groups (BR versus BR/PVP). Although the mean CBF did not differ significantly between the BR and BR/PVP group, at the Hct level of 19%, 8 of 34 brain structures showed a significantly lower LCBF in the BR/PVP than in the BR group. At an Hct of 15%, mean CBF as well as LCBF in 15 of 34 brain structures were lower in the BR/PVP group than in the BR. Figure 4 shows the oxygen delivery to the brain in the different groups. Oxygen delivery was calculated from the product of mean CBF and arterial oxygen content. During the progressive decrease of Hct, cerebral oxygen transport was maintained in both groups except for the lowest Hct. At the Hct of 15%, oxygen transport to the brain was significantly lower in the BR/PVP group than in the BR group.

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Mean cerebral blood flow (CBF) depending on Hct of the different experimental groups tested. Values are means ± SD. CBF was determined at baseline (Hct 43%) and after isovolemic blood exchange to a Hct of 30, 22, 19, or 15% (BR) and the addition of PVP (BR/PVP). Mean CBF was measured by averaging the area-weighted means of coronal sections taken from the whole brain. *P < 0.05 BR vs. BR/PVP at the same Hct level.

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Cerebral oxygen delivery (DO2) depending on Hct. Values are means ± SD. Cerebral DO2 was calculated from the product of CBF and CaO2 at baseline (Hct 43%) and in both BR and BR/PVP groups. *P 0.05 BR vs. BR/PVP at the same Hct.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

This study shows the maintenance of oxygen delivery to the brain when blood viscosity is doubled compared with untreated controls and when, in addition, Hct is decreased by hemodilution to values as low as 22%. Oxygen delivery is lowered in some brain structures at Hct values of 19%, whereas it is globally reduced at Hct values of 15%. The results are compatible with the hypothesis that both an increase in blood viscosity and a decrease in Hct induce compensatory dilation of brain vessels, which results in a maintained oxygen delivery. The capacity to compensate for the changes in viscosity and Hct is exhausted when the Hct is decreased to less than one-half its normal value at a doubled blood viscosity.

The two main blood components that determine cerebral oxygen delivery are Ca_O2 and blood viscosity. At moderate degrees of hemodilution, the cerebral oxygen delivery is maintained by an increase in CBF (6, 11). To explain the hemodilution-related increase in CBF, each of these components has been favored as the main causal factor. Ca_O2 has been postulated as the crucial factor by several groups (2, 4, 10, 21, 20, 25), whereas others have preferred blood viscosity (7, 14, 16).

Apart from hemodilution, the question of whether Ca_O2 or blood viscosity is the major determinant of cerebral oxygen delivery is of more general interest. A major role of Ca_O2 has been postulated by several groups. In a clinical study Brown et al. (3) observed a significant correlation between Ca_O2 and CBF in patients suffering from paraproteinemia in combination with anemia. CBF was not related to differences in blood viscosity in these patients. In experimental studies different approaches have been used to separate the effects of oxygen-carrying capacity from blood viscosity during hemodilution. One approach to separate oxygen-carrying capacity from Hct was to use carbon monoxide (CO) to change Ca_O2 at constant Hct. When Ca_O2 was reduced by the addition of CO at constant Hct, an increase in CBF was observed suggesting Ca_O2 as the primary regulation factor for the cerebral perfusion (5). In addition, moderate loading of Hb by CO increased cerebral oxygen delivery by a change of the oxygen affinity of Hb (12). This finding supports the hypothesis that an oxygen-sensitive mechanism participates in the regulation of the cerebral circulation. Other approaches supporting the importance of Ca_O2 for the regulation of CBF were based on a complete substitution of blood by a cell-free oxygen-carrying Hb solution. The advantage of such a solution is its constant and defined viscosity, which, in contrast to normal blood does not vary with the shear rate. After a near-total exchange transfusion to a Hct <3% with such cell-free solutions, Waschke et al. (25) found that CBF increased but that increasing blood viscosity did not reduce CBF from this elevated level. Consequently, oxygen delivery to the brain of conscious rats was preserved at increased viscosity when most of the cell-based Hb was replaced by cell-free Hb. It was concluded that the oxygen content of the blood primarily regulates CBF and that changes in plasma viscosity are compensated by vascular adjustments. Concurrent with these findings, Ulatowski et al. (24) found that increasing Ca_O2 at a reduced Hct by adding a cell-free Hb solution to the blood resulted in a decrease in CBF. This finding is consistent with a cerebral vasoconstriction as a response to the increased oxygen-carrying capacity (1). The hypothesis that Ca_O2 is the major determinant of CBF during hemodilution was also supported by a study of Cole et al. (4), who compared two groups of rats in which Hct was reduced to identical values, but Ca_O2 was varied by the addition of an oxygen-carrying Hb solution or of albumin. When Hct was moderately decreased by one-third, CBF was increased by a corresponding amount. A further reduction of Hct at a maintained Ca_O2 resulted in an increased CBF indicating that under these conditions, blood viscosity has an effect on CBF in addition to Ca_O2. A dominating, but not exclusive, role of Ca_O2 in the regulation of CBF has also been demonstrated in rats hemodiluted to an Hct of 17% by the infusion of 6% hetastarch solution (23). When Ca_O2 was restored to normal values by hyperbaric oxygenation, CBF did not return to normal control values indicating that at this low Hct blood viscosity has some effect on CBF in addition to Ca_O2.

Other studies have supported the hypothesis that CBF follows the changes in blood viscosity. This hypothesis is based on the application of Poiseuille's law. If the increase in CBF observed during hemodilution is caused by the reduction in whole blood viscosity, an increase in blood viscosity should result in a decrease in CBF. Combining CO loading to reduce Ca_O2 with hemodilution to lower blood viscosity, Paulson et al. (16) observed a dependency of CBF on blood viscosity rather than on Ca_O2. They concluded that the decrease in blood viscosity is the primary mediator of the increase in CBF during hemodilution. Hurn et al. (9) measured the changes in microvascular pressure in pial arteries during an isovolemic hemodilution and observed that the microvascular pressure remained unchanged when CBF increased during hemodilution. Because there was no evidence of a change in arteriolar diameter, these authors (9) concluded that the hyperemia accompanying hemodilution was largely caused by changes in blood viscosity rather than by vasoactive mechanisms. Direct observation of the pial vascular response to changes in blood viscosity showed pial vasoconstriction when blood viscosity was decreased and vasodilation with increased blood viscosity (15). Hudak et al. (7) altered the Hct independent from oxygen-carrying capacity in a hemodilution study by using methemoglobin containing red blood cells in exchange with normal red blood cells. When the decrease in Hct induced by hemodilution was abolished by infusion of red blood cells containing methemoglobin at constant Ca_O2 CBF was decreased. However, CBF declined even further when Ca_O2 was also increased. These authors concluded that the change in blood viscosity only partly accounted for the changes in CBF. In addition, observation of pial vasoconstriction and an increase in CBF during hemodilution also resulted in the conclusion that the CBF response occurred by both passive (consequent to Poiseuille's law) and active mechanisms (8).

In conclusion, in the present study the hypothesis was tested that both an increase in blood viscosity and a decrease in Ca_O2 can trigger compensatory vascular mechanisms. By a combination of an increased blood viscosity and a gradual decrease of Ca_O2, the hypothesis of a compensatory capacity of cerebral vessels to both factors could be tested. Furthermore, this approach also allowed testing of the extent to which a combination of both factors could be compensated. The present results show that CBF was unchanged when blood viscosity was increased and simultaneously Hct was moderately decreased. From these data it can be concluded that in an experimental model of increasing plasma viscosity and under nonischemic conditions, the cerebral vessels are able to compensate for both an increase in viscosity and a decrease in Ca_O2. The present study also defines the limits of these vascular compensatory mechanisms. The lack of a further increase in CBF in the high-viscosity groups at Hct values of 19 and 15% shows the outstripping of vascular compensatory reactions to the increased blood viscosity at these low Hct values. In this situation, the combination of low Hct and high blood viscosity resulted in a decreased cerebral oxygen delivery.

The arterial acidemia seen with PVP transfusion at a Hct of 15% when normocapnia was maintained is presumed to be a consequence of peripheral tissue hypoxia rather than a direct toxic effect of PVP. The same dose of PVP infused at a Hct of 19% or greater did not cause acidemia. Moreover, PVP transfusion did not cause acidemia when red blood cell-based Hb was replaced by cell-free Hb (25). Because of the blood-brain barrier, arterial acidemia at constant PCO₂ has little effect on CBF (13, 26) unless cerebral tissue acidosis did accompany the decrease in CBF with PVP transfusion at a Hct of 15%. In this case, the decrease in CBF might have been more dramatic without the presumed tissue acidosis.

In the present study, the range of vascular compensation to a combined increase in blood viscosity and decrease in Hct could be defined for nonischemic conditions. It can be expected that compensatory mechanisms are less effective during pathophysiological conditions. In an experimental model of middle cerebral artery occlusion using mannitol for altering blood viscosity, Muizelaar et al. (14) investigated the changes in CBF to a combined increase in blood viscosity by mannitol and a decrease in blood pressure. Whereas CBF remained fairly constant in the nonischemic hemisphere despite the decrease in blood pressure or increase in blood viscosity, CBF was decreased in the ischemic areas of the brain. The authors (14) concluded that CBF changes passively as a response to the alterations in blood viscosity if vascular adjustment is disturbed. In addition, Cole et al. (4) compared the hemodilution-related increase in CBF in the nonischemic rat brain to the reaction in the ischemic hemisphere during focal ischemia. In the ischemic brain areas Cole et al. (4) observed a dependency of CBF on blood viscosity during hemodilution, which was not related to Ca_O2. However, in the nonischemic hemisphere, CBF was influenced by both Ca_O2 and viscosity, suggesting that ischemia already induced maximal vasodilatation (19) and the flow behavior should be passively related to blood viscosity. These results are in accordance with those of the present study, which also shows that vascular compensatory mechanisms exist and can be compromised under extreme conditions. In this context, the present study defines the limits at which vascular compensatory mechanisms become exhausted.

The present study shows that CBF can be maintained in a normal nonischemic brain when blood viscosity is increased and simultaneously the oxygen content of the blood is decreased. This shows the existence of compensatory mechanisms, which result in the maintenance of oxygen delivery to the brain. Furthermore, the limits of such compensatory mechanisms are defined. A doubling of blood viscosity still results in a maintained oxygen delivery to the brain as long as Hct is not decreased to less than half of its normal value. Below such Hct values (19% and lower) the vasodilating capacity is exhausted resulting in a decrease in oxygen delivery to the brain. We conclude that compensatory mechanisms exist, which allow the maintenance of oxygen delivery to the brain during an increase in viscosity and a decrease in Hct over a wide range.