Hemodilution Accelerates the Passage of Plasma (Not Red Cells) Through Cerebral Microvessels in Rats
Background and Purpose Hemodilution lowers the total circulatory red cell mass and blood viscosity and thereby may alter the time of passage of red cells and plasma through cerebral microvessels. This study was designed to clarify this question.
Methods Adult Wistar-Kyoto rats, aged approximately 32 weeks, were divided into hemodilution and control groups. Local cerebral blood flow and microvascular red cell and plasma volumes in 14 brain structures were measured with the use of [14C]iodoantipyrine, 55Fe-labeled red cells, and [14C]inulin, respectively.
Results In the control group, the hematocrit in cerebral microvessels ranged from 0.29 to 0.45 with a mean of 0.36, which was 71% of the systemic hematocrit (0.51). The mean transit times of blood, red cells, and plasma through microvessels were 0.62 to 1.77 seconds (mean, 0.92 second), 0.44 to 1.15 seconds (mean, 0.65 second), and 0.78 to 2.5 seconds (mean, 1.25 seconds), respectively. In the hemodilution group, the mean hematocrit in microvessels was 0.28, which was 89% of the systemic hematocrit (0.32). Local cerebral blood flow was approximately 59% higher (P<.01) than that of the control animals. The rate of oxygen delivered to the brain was slightly increased (9%) after hemodilution. Blood volume in cerebral microvessels was similar to that of the control group. Mean transit time of blood was 0.62 second (68% of the control), transit time of red cells was 0.53 second (85% of the control), and transit time of plasma was 0.67 second (54% of the control).
Conclusions These findings indicate that isovolemic hemodilution accelerates the plasma (not red cell) flow velocity in cerebral microvessels.
Nutrient delivery to local brain tissue is proportionally related to LCBF.1 2 Once nutrients are delivered to the local microvascular system, the EF of the nutrients in the local brain tissue depends partly on the microvascular surface area and the blood flow velocity in microvessels.1 2 The larger the microvascular surface area, the higher the EF; in contrast, the faster the blood flow velocity, the lower the EF.1 The microvascular surface area of the brain is proportional to the functionally blood-perfused microvascular volume (Vb).1 2 3 On the other hand, the blood flow velocity in cerebral microvessels is inversely proportional to the transit time of blood through these microvessels (Tb).3 4 Therefore, the knowledge of LCBF, Vb, and Tb will lead to a better understanding of nutrient delivery and EF in local brain tissue.
Hemodilution is frequently used to treat neurological diseases such as cerebral infarction or cerebral vasospasm; however, its therapeutic effect is still controversial.5 6 Part of the reason for this controversy is that little is known about the simultaneous alterations of LCBF, Vb, and Tb after hemodilution. Moreover, since hemodilution lowers the number of red blood cells (RBCs) in blood, this may lead to changes in flow velocities of RBCs and plasma in cerebral microvessels, namely, the alterations of transit times of RBCs (Tr) and plasma (Tp) through cerebral microvessels. Tr is mainly related to the EF of oxygen, while Tp is responsible for the EF of substances such as glucose and amino acids.
Hematocrit in cerebral microvessels (mHct) is well known to be less than the systemic hematocrit (sHct) and is frequently assumed to be 85% of the sHct in clinical patients.6 However, mHct varies in various brain structures and ranges from 0.29 to 0.4 in normal awake rats.3 Hemodilution lowers systemic hematocrit (sHct) and thus may alter mHct. Therefore, Vb needs to be separately measured with RBC and plasma markers. Recently, Lin et al7 developed an RBC labeling technique using 55Fe. With this method, RBCs can be physiologically labeled with 55Fe (55Fe-RBCs) and used to measure RBC volume in microvessels (Vr) of local brain tissues.4 7 In this study the authors used 55Fe-RBCs and [14C]inulin to measure Vr and plasma volume in cerebral microvessels (Vp), respectively. Thus, Vb could be directly summed from Vr and Vp. LCBF was measured with [14C]4-iodo-N-methyl-antipyrine (IAP). With these data, mHct, Tb, Tr, and Tp in various brain structures were elucidated. The effects of hemodilution on LCBF, Vb, mHct, Tb, Tr, and Tp were clarified.
Materials and Methods
This study is in compliance with institutional guidelines for the use of animals. Thirty-six male Wistar-Kyoto rats (mean age, 32 weeks; mean weight, 394 g) were anesthetized with a gas mixture of halothane and oxygen. Cannulas were placed in the femoral arteries and veins bilaterally. The surgical wounds were infiltrated with lidocaine and closed with sutures. A plaster cast was fitted below the midthorax of each rat to immobilize the hindlimbs and protect the catheters. Anesthesia was discontinued, and each rat was allowed 1.5 hours to recover before the cerebral circulating experiment was started. Body temperature was continuously monitored and maintained at approximately 37°C with a heat lamp. The rats were then randomly divided into control and hemodilution groups. The latter were treated with isovolemic hemodilution by replacing 10 to 12 mL of whole blood with the same volume of 3% modified fluid gelatin (Roger Bellon France) in 2 minutes. With this procedure, the sHct could be lowered from approximately 0.51 to 0.3-0.35.
The arterial blood pressure and heart rate of each rat were continuously monitored and recorded throughout the entire experimental procedure, and the following physiological data were determined (either 20 minutes after hemodilution or 2 hours after anesthesia was discontinued for the control animals) by measuring arterial blood gases, plasma glucose level, systemic hemoglobin, and sHct levels. Immediately after completion of the measurements of physiological variables, the animals were subjected to the following cerebral hemodynamic studies.
Measurement of Local Cerebral Blood Flow
LCBF was measured in 12 rats (6 for each group) using the IAP technique developed by Sakurada et al.8 For each rat, approximately 12 μCi of IAP (Amersham International) in 1 mL of normal saline was infused at a constant rate via the femoral venous catheter for 30 seconds, during which time arterial blood samples were collected every 5 seconds for assay of arterial concentration of IAP. The animals were decapitated at 30 seconds, and their brains were quickly removed and then immersed in isopentane chilled to −40°C. The brains were placed on a dissecting plate kept in a freezer with the temperature at −10°C.
Blood samples were centrifuged, and plasma 14C radioactivity was determined by beta counting. By referring to a standard anatomic atlas,9 we dissected brain tissue samples of bilateral hemispheres from the following 14 areas. Eight areas were in the forebrain including the frontal cortex, the sensorimotor cortex, the temporal cortex, the caudate-putamen, the occipital cortex, the thalamus, the hypothalamus, and the hippocampus; three areas were in the rostral hindbrain including the superior colliculus, the inferior colliculus, and the ventral midbrain; the other three areas were in the caudal hindbrain including the pons, the medulla oblongata, and the cerebellar vermis. The brain tissue samples were weighed, dissolved in 0.5 mL of 1.0N sodium hydroxide, vortexed, hydrolyzed for approximately 15 hours at 45°C in a shaking water bath, and neutralized with 1.0N hydrochloric acid before scintillant was added. The 14C radioactivity of each brain tissue sample was determined by beta counting. LCBF was calculated from the brain tissue and the plasma radioactivity data with the use of the Kety-Sokoloff equations.
Measurement of Red Cell Volume in Cerebral Microvessels
Vr was measured in each group of animals (6 rats for each group) by the use of 55Fe-RBCs. The method of labeling RBCs with 55Fe was described elsewhere by Lin et al.7 In brief, [55Fe]FeCl2 solution (3 mCi, Amersham International) diluted with 6 mL of saline was injected intraperitoneally into each donor rat after the peritoneal cavity was opened. Seven days later, blood from the donor rat that contained 55Fe-RBCs was harvested. The ratio of the radioactivity between plasma and RBCs of the donor rat blood was less than 1%. The blood (1 to 1.4 mL) was injected intravenously into each Wistar-Kyoto rat. One minute after injection of the radioactive blood, a small blood sample (15 μL) was collected, and the 55Fe-RBC radioactivity (Rr) was derived from the 55Fe radioactivity data of blood (Rb) and plasma (Rp), sHct, and the equation Rr=[Rb−Rp(1−sHct)]/sHct. Vr was calculated from the Rr and the equation Vr=Ar/Rr, where Ar was the 55Fe activity of brain tissue.
Measurement of Plasma Volume in Cerebral Microvessels
Vp was also measured in each group of animals (6 rats for each group) with [14C]inulin. For each animal, 15 μCi of inulin (Amersham International) in 1 mL of saline was injected intravenously. One minute after the injection, a small blood sample (20 μL) was collected, and the plasma 14C radioactivity (Pi) was measured by beta counting. At the end of blood sampling the rats were decapitated, and their brains were removed and frozen. The radioactivity of the 14 brain structures was measured by beta counting. Vp was derived from the [14C]inulin radioactivity data of brain tissue (Ai) and plasma (Pi) by the equation Vp=Ai/Pi.
Vb was calculated as follows: Vb=Vr+Vp. mHct was determined by mHct=Vr/Vb. Mean Tb was determined by Tb=Vb/LCBF. Mean Tr and Tp were calculated by Tr=Tb×(mHct/sHct) and Tp=Tb×[(1−mHct)/(1−sHct)], respectively.4 The rate of oxygen delivered to a local brain (OD) was derived by multiplying LCBF by volume fraction of oxygen in arterial blood (VO2, expressed in milliliters oxygen per milliliters blood).
All physiological data are presented as mean±SEM. Differences in these data between groups were assessed with unpaired Student’s t test.
Data of LCBF, Vr, and Vp are presented as mean±SEM. Differences between the control and hemodilution groups were assessed by MANOVA for the 14 structures.10 When the MANOVA assessments between groups differed significantly, the unpaired Student’s t test was used to identify the brain structures when the differences in data between groups were large.10 Vb, Tb, Tr, and Tp were derived values from LCBF, Vr, and Vp and therefore were reported as estimated mean±SEE.4 The unpaired Student’s t test was again used to identify the area when the differences in Vb, Tb, Tp, and Tr between groups were large.
The difference between groups was assumed to be significant at P≤.05.
sHct and hemoglobin levels were 0.512±0.007 and 15.3±0.3 g/dL, respectively, for the control group and were 0.321±0.004 and 10.3±0.2 g/dL, respectively, for the hemodilution group. Hemodilution decreased sHct and hemoglobin levels and thereby significantly lowered VO2 (P<.01) compared with the data of the control group (0.15±0.001 mL/mL blood versus 0.22±0.003 mL/mL blood). Hemodilution also led to a significant decrease in systemic blood pressure (103±5 mm Hg versus 125±3 mm Hg, P<.01) and an increase in heart rate (445±6 beats per minute versus 378±12 beats per minute, P<.01). However, hemodilution did not result in any significant change of the arterial pH (7.46±0.01 versus 7.45±0.01), Po2 (155±1 mm Hg versus 150±3 mm Hg), Pco2 (35.2±1.3 mm Hg versus 37.8±1 mm Hg), or plasma glucose (185±11 mg/dL versus 163±6 mg/dL) as data were compared between the hemodilution and the control groups.
Local Cerebral Blood Flow
LCBF ranged from 59 mL/100 g per minute in the medulla to 142 mL/100 g per minute in the sensorimotor cortex for the control group and 93 mL/100 g per minute in the medulla to 235 mL/100 g per minute in the frontal cortex for the hemodilution group (Table 1⇓).
Hemodilution resulted in a general increase (mean, 59%) in LCBF compared with data of the control group (P<.01 by MANOVA). Analyzed with uncorrected Student’s t test, differences in LCBFs between the two groups were significant (P<.05) in all 14 structures (Table 1⇑).
Red Cell Volume in Cerebral Microvessels
Vr ranged from 3.5 to 6.6 μL/g in the 8 forebrain areas and from 3.3 to 6.5 μL/g in the 6 hindbrain structures of the control group (Table 2⇓). The Vr values of the hemodilution group ranged from 3.2 to 5.4 μL/g (Table 2⇓) and were generally less (P<.05 by MANOVA) than those of the control group; however, the decrease in Vr was not significant (P>.05) for each brain area as analyzed by t test.
Plasma Volume in Cerebral Microvessels
In the control group, Vp values were smallest in the thalamus (6.3±0.1 μL/g) and largest in the hypothalamus (12.3±1.1 μL/g) (Table 2⇑). Hemodilution also led to a general increase (P<.05 by MANOVA) in Vp (range, 6.9 to 15.5 μL/g) compared with data of the control group (Table 2⇑). The increases were significant (P<.05 by uncorrected t test) in 7 of the 14 areas.
Hematocrit in Cerebral Microvessels
mHct values ranged from 0.29 to 0.45 in the control animals, which were 56% to 87% (mean, 71%) of the sHct (0.51) of the control group (Table 2⇑). In the hemodilution group, mHct values ranged from 0.22 to 0.32 (mean, 0.28), which were 69% to 100% (mean, 89%) of the sHct (0.32) of the hemodilution group (Table 2⇑).
Blood Volume in Cerebral Microvessels
Vb values were 10.7 to 18.4 μL/g and 10.1 to 20 μL/g for control and hemodilution groups, respectively (Table 1⇑). Vb values tended to be slightly increased (mean, 8%) in most of the 14 areas in the hemodilution group; only in the sensorimotor cortex and the pons were the increases in Vb significant (P<.05, uncorrected t test).
Mean Transit Time of Blood Through Cerebral Microvessels
Tb values were shortest in the sensorimotor cor- tex (0.62±0.1 second) and longest in the medulla (1.77±0.26 seconds) for the control group (Table 1⇑). For the hemodilution group, Tb values ranged from 0.43 to 1.16 seconds and were significantly shorter (mean decrease, 32%; P<.05, uncorrected t test) than those of the control group in 12 of 14 areas (Table 1⇑).
Mean Transit Times of Red Cells and Plasma Through Cerebral Microvessels
For the control group, Tr values were 0.44 to 1.15 seconds; Tp values were much longer than Tr values and ranged from 0.78 to 2.5 seconds (Fig 1⇓). For the hemodilution group, Tr values were 0.41 to 0.92 second and were similar (P>.05) to those of the control group in all 14 areas (Fig 1⇓). In contrast, hemodilution significantly shortened (P<.05) Tp values in all 14 areas compared with those of the control group (Fig 1⇓). The mean decrease in Tp in all 14 areas by hemodilution was 46%.
Rate of Oxygen Delivered to a Local Brain
OD values ranged from 12.7 mL/100 g per minute in the medulla to 30.7 mL/100 g per minute in the sensorimotor cortex for the control group and from 13.7 to 34.9 mL/100 g per minute for the hemodilution group (Fig 2⇓). Hemodilution tended to result in a slightly general increase (mean, 9%) in OD in all 14 brain areas (Fig 2⇓).
The present results reveal that isovolemic hemodilution performed in awake rats leads to a modest increase in LCBF, slight increases in Vb and OD, and a significant shortening in Tb in both the forebrain and hindbrain structures. The shortening in Tb is mainly due to the shortening in Tp (not Tr), namely, an acceleration of plasma (not RBC) flow velocity in cerebral microvessels.
Our result displayed that hemodilution induced an increase in LCBF by 59%, which is compatible with previous reports.11 12 Whether the increase in LCBF is due to increases in either blood flow velocity in cerebral microvessels, the number of blood-perfused microvessels, or both is not well studied. Rosenblum13 observed that hemodilution induced increases in flow velocities of RBCs and plasma in pial arteries of anesthetized mice using high-speed microcinematography. However, with this technique the flow velocities of RBCs and plasma in the downstream microvessels in various deep brain structures cannot be measured.
Recently, Lin et al7 and Fenstermacher and coworkers3 4 developed methods for measuring RBC and plasma distribution spaces in cerebral microvessels. The volume of blood-perfused microvessels in various brain structures can thus be accurately measured. With the data of blood volume and LCBF, mean Tb, Tr, and Tp can be calculated. Since Tb is inversely proportional to the linear velocity of blood in cerebral microvessels, Tb can be used to indicate the blood flow velocity in cerebral microvessels.4 10 This is also true for both Tr and Tp. Using these methods, Fenstermacher and coworkers4 found that the increase in LCBF by hypercapnia is mainly due to increases in both RBC and plasma flow velocities in the already perfused cerebral microvessels, and the decrease in LCBF by pentobarbital is due to decreases in RBC and plasma flow velocities in cerebral microvessels and not due to a decreased percentage of perfused capillaries.10 Our results revealed that hemodilution resulted in a slightly general increase (8%) in the Vb of most brain structures (12 of 14 areas) studied. The mild increase in Vb may be due to either perfusion of more microvessels with blood or induced dilation of cerebral microvessels by hemodilution. Since all cerebral microvessels are perfused with blood in normal conscious rats,14 15 the increased Vb in conscious rats by hemodilution is likely due to mild dilation of cerebral microvessels. With little change in the number of blood-perfused cerebral microvessels, the LCBF increase by hemodilution is likely due to the increase in blood flow velocity in cerebral microvessels. This notion is confirmed by the present results. The mean Tb of the hemodilution group is only 68% (mean) of that of the control group; in other words, the blood flow velocity in cerebral microvessels is much faster in the hemodilution group than in the control group.
The flow pattern and velocity of RBCs in cerebral microvessels are different from those of plasma. The latter flows continuously throughout the vessel lumen, while the former flows intermittently and quickly in the smallest cerebral microvessels. Hemodilution lowering the blood viscosity leads to an acceleration of blood flow velocity in cerebral microvessels. We further demonstrated that Tr values were not significantly (P>.05) altered by hemodilution in all 14 brain structures; in contrast, the mean Tp values were significantly (P<.05) shortened (mean decrease, 46%) by hemodilution. These results indicate that the increase in LCBF by hemodilution is mainly due to an acceleration in plasma flow velocity in cerebral microvessels. The RBC flow velocity in cerebral microvessels is not significantly altered by hemodilution.
Hematocrit in cerebral microvessels varies in different brain structures. In normal awake rats, Tajima et al3 reported that mHct values ranged from 0.29 to 0.4 in various brain areas, which were 60% to 83% (mean, 72%) of the sHct. Our results showed that mHct in normal awake rats was 0.29 to 0.45, which was 56% to 87% (mean, 71%) of the sHct. mHct values also vary under different physiological conditions. Pentobarbital anesthesia (50 mg/kg) markedly increases mHct from 0.33 to 0.45, although it does not alter sHct.10 Our results showed that hemodilution that lowered sHct from 0.51 to 0.32 induced a mild decrease in mHct from 0.29-0.45 (mean, 0.36) to 0.22-0.32 (mean, 0.28). However, the proportion of mHct to sHct rose markedly from 71% (mean) to 89% (mean) by hemodilution. These results indicate that hemodilution cannot proportionally lower mHct as it does sHct. Accordingly, it is speculated that the driving force for RBCs passing through cerebral microvessels may not be related to blood viscosity but may be mainly due to vascular pressure, while that for plasma is related to blood viscosity as well as vascular pressure.
Hemodilution with plasma substitutes is frequently used to replace blood loss during surgery or to treat cerebral ischemia. Although hemodilution leads to an increase in LCBF, it concomitantly decreases the oxygen-carrying capacity of the diluted blood. To avoid a decease in OD, a theoretical optimal degree of hemodilution is considered to be 0.3.16 17 Our result revealed that the OD was slightly increased (9%) by hemodilution in awake rats when sHct was lowered to 0.32.
The EF of substances from cerebral microvessels is partly related to the flow velocity of blood in these microvessels. Oxygen is mainly carried by RBCs, and therefore the EF of oxygen is related to the flow velocity of RBCs in cerebral microvessels, namely, to the Tr.17 The present results show that Tr is altered very little by hemodilution, and therefore the EF of oxygen in the brain may be unaltered by hemodilution. In contrast, since hemodilution markedly reduces the Tp, the EF of plasma-carrying substances, such as glucose and amino acids, may be reduced in the brain by hemodilution. However, the reduced EF of plasma-carrying substances does not imply a decreased influx. An accelerated plasma flow velocity by hemodilution may also enhance the clearance rates of waste products, such as lactic acid and urea, in the brain.18
In conclusion, isovolemic hemodilution with a plasma substitute leads to an acceleration of plasma and blood flow velocities in cerebral microvessels. The flow velocity of RBCs and hematocrit in cerebral microvessels and the amount of oxygen delivered to the brain are altered very little by hemodilution of this magnitude. These results are compatible with many studies of other organ beds.19 20
Selected Abbreviations and Acronyms
|LCBF||=||local cerebral blood flow|
|mHct||=||hematocrit in cerebral microvessels|
|OD||=||rate of oxygen delivered to a local brain|
|RBCs||=||red blood cells|
|Tb||=||transit time of blood through cerebral microvessels|
|Tp||=||transit time of plasma through cerebral microvessels|
|Tr||=||transit time of red blood cells through cerebral microvessels|
|Vb||=||blood volume in cerebral microvessels|
|VO2||=||volume fraction of oxygen in arterial blood|
|Vp||=||plasma volume in cerebral microvessels|
|Vr||=||red blood cell volume in cerebral microvessels|
This study was supported by the National Science Council of the Republic of China (NSC 80-0412-B-016-119, NSC 81-0412-B-016-559, and NSC 82-0412-B-016-147).
Reprint requests to Shinn-Zong Lin, MD, PhD, Division of Neurosurgery, Department of Surgery, Tri-Service General Hospital, 8, Section 3, Ting-Chou Rd, Taipei, Taiwan, ROC.
- Received April 11, 1995.
- Revision received July 10, 1995.
- Accepted August 7, 1995.
- Copyright © 1995 by American Heart Association
Fenstermacher JD, Blasberg RG, Patlak CS. Methods for quantifying the transport of drugs across brain barrier system. In: Boldman ID, ed. Membrane Transport of Antineoplastic Agents: The International Encyclopedia of Pharmacology and Therapeutics. Oxford, England: Pergamon Press, Inc; 1986:113-146.
Tajima A, Nakata H, Lin S-Z, Acuff V, Fenstermacher J. Differences and similarities in albumin and red blood cell flows through cerebral microvessels. Am J Physiol. 1992;262:H1515-1524.
Bereczki D, Wei L, Ostuka T, Hans FJ, Acuff V, Patlak C, Fenstermacher J. Hypercapnia slightly raises blood volume and sizably elevates flow velocity in brain microvessels. Am J Physiol. 1993;264:H1360-1369.
Mast H, Marx P. Neurological deterioration under isovolemic hemodilution with hydroxyl starch in acute cerebral ischemia. Stroke. 1991;22:680-683.
Lin S-Z, Nakata H, Tajima A, Gruber K, Acuff V, Patlak C, Fenstermacher J. Quantitative autoradiographic assessment of 55Fe-red cell distribution in rat brain. J Appl Physiol. 1990;69:1637-1643.
Sakurada D, Kennedy C, Jehle J, Brown JD, Carbin GL, Sokoloff L. Measurement of local cerebral blood flow with iodo[14C]antipyrine. Am J Physiol. 1978;234:H59-H66.
Paxnos G, Watson C. The Rat Brain in Steoreotaxic Coordinates. New York, NY: Academic Press Inc; 1982.
Hartmann A, Dettmers C, Beyenburg S. Effect of hemodilution on regional cerebral blood flow. Acta Neurol Scand. 1989;127:36-48.
Rosenblum WI. Effects of reduced hematocrit on erythrocyte velocity and fluorescein transit time in the cerebral microcirculation of the mouse. Circ Res. 1971;29:96-103.
Vetterlein F, Demmerle B, Bardosi A, Gobel U, Schmidt G. Determination of capillary perfusion pattern in rat brain by timed plasma labeling. Am J Physiol.. 1990;258:H80-H84.