INTRODUCTION
Therapeutic infusion of fluids to maintain systemic perfusion is a critical strategy for the management of acute hemorrhage. Previous studies have established that colloid solutions are superior to crystalloid solutions for fluid resuscitation (1-3). The primary therapeutic goal of fluid resuscitation is not only to reestablish an adequate circulation with sufficient arterial blood pressure and cardiac output, but also to restore tissue perfusion and oxygenation. Therefore, maintenance of the peripheral circulation is a very important factor. The peripheral circulation has previously been studied using various methods. Flow velocity has been measured by laser Doppler flowmetry (4, 5) and indirectly by monitoring partial oxygen pressure (6). The microcirculation has been microscopically examined in vivo during hemorrhagic shock (7, 8), as well as after resuscitation (9, 10). To our knowledge, however, the effects of colloid and crystalloid solutions on the peripheral microcirculation, hemodynamics, and colloidal osmotic pressure (COP) have not been directly compared microscopically during acute severe hemorrhage and resuscitation in vivo. We have focused on these issues to assess the adequacy of fluid resuscitation because subclinical peripheral tissue hypoperfusion occurs early in the course of hemorrhagic shock (11).
The rabbit ear chamber (REC) method developed at our laboratory allows a single blood vessel to be observed by intravital microscopy, directly and noninvasively on a real-time basis in models of shock. This method has been confirmed to be useful for observing the effects of various interventions on peripheral hemodynamics (12). The vasomotor activity, neural control, and response to drugs of vessels in an REC are similar to those of in situ vessels 6 weeks after attachment to a clear window (12). We have used this model to study the peripheral microcirculation in systemic agglutination anaphylaxis (13), as well as to examine the effects of nitrous oxide (14) and oxygen (15, 16) in inhaled air, and carbon dioxide (17) in arterial blood on the circulation.
This study tested the hypothesis that hydroxyethyl starch (HES), a colloid preparation, was superior to Ringer's lactate solution (LR), a crystalloid preparation, in terms of microcirculatory changes during acute excessive hemorrhage. The REC method was used in a rabbit model of acute severe hemorrhage to compare the effects of HES and LR on circulatory depression, changes in COP, and peripheral microcirculatory dysfunction. As stated above, a major advantage of this model is that it allows the peripheral microcirculation to be directly observed. This is important in the study of severe hemorrhage because the maintenance of blood pressure does not always ensure adequate peripheral microcirculation.
MATERIALS AND METHODS
Animal preparation
All experiments were done in accordance with the National Institutes of Health guidelines on the use of experimental animals. Approval from the Animal Use Committee of Tokyo Women's Medical University was obtained before initiating the experiments.
A total of 40 randomly selected Japanese albino rabbits (body weight, 2.5-3.0 kg) was studied. Transparent round chambers made of acrylic resin were inserted in the earlobes as described previously (12). New microvessels arose from the blood vessels of the dermis, and they covered the entire cavity within 6 weeks. Six weeks after insertion of the ear chamber, the RECs were observed microscopically at a magnification of 100×. Microcirculatory changes were recorded using a microscope-closed video camera (DXC 750; Sony, Tokyo, Japan) with a shutter speed of 1/10,000 of a second.
After intravenous injection of pentobarbital 30 mg kg-1, the trachea of the rabbit was intubated, and 1 mg kg-1 pancuronium was administered intravenously. Anesthesia was maintained with inhaled 0.5% isoflurane. Respiration was controlled with the use of a ventilator (ART-100; Acoma, Tokyo, Japan), and end-tidal CO2 was monitored (N-1000; Nellcor, Pleasanton, CA). Arterial blood gases were measured with a blood-gas analyzer (model ABL330; Radiometer, Tokyo, Japan) throughout the study, and arterial PO2 was maintained at more than 100 mmHg during the study to eliminate potential effects of systemic hypoxemia on our results. Oral temperature was continuously monitored, and body temperature was kept constant with the use of a heating pad. Urinary output was continuously monitored. Plasma COP was measured with a colloid osmometer (4420; Wescor, Logan, UT).
Heart rate (HR) was monitored with an electrocardiograph. Systolic blood pressure, diastolic blood pressure, and mean arterial pressure (MAP) were monitored with a high-fidelity transducer-tipped catheter (Millar Microtip catheter pressure transducer, 6F SPC-360; Millar Instruments, Houston, TX), placed in the right femoral artery. A similar catheter was placed in the left artery for blood letting. A 20-G, 3.2-cm catheter (Terumo Co., Tokyo, Japan) was placed in the left auricular vein for infusion of HES or LR. A catheter was placed in the inferior vena cava via the right femoral vein to monitor central venous pressure (CVP). All signals were monitored continuously using a multichannel polygraph (360; NEC San-ei, Tokyo, Japan). All variables were recorded continuously throughout the experiments.
Study design and experimental protocol
After surgical preparation and 30 min of stabilization, baseline hemodynamic variables were recorded. The rabbits were divided into two groups of 20 each. One group was given LR as control and the other was given 6% HES. To induce shock, blood letting was performed. The target bleeding volume was equivalent to 40% to 50% of the circulating blood volume. The bleeding volume per time per minute was 20 mL (10%-13% of circulating blood volume). Blood was released a total of four times at 3-min intervals for a total bleeding volume of 80 mL.
In the control group, 40 mL of LR solution was rapidly infused intravenously 3 min after the last bleeding procedure. Additional LR was then given by intravenous infusion to a total volume of 200 mL, delivered at a rate of 320 mL h-1 for 30 min. In the HES group (n = 20), HES was given in a similar fashion. First, 20 mL of HES was immediately infused intravenously after the blood letting procedure. Additional HES was then administered by intravenous infusion to a total volume of 100 mL, delivered at a rate of 160 mL h-1 for 30 min. After the infusions, we measured all variables for 30 min.
During the stabilization period after surgical preparation, we selected arterioles with diameters of 20 to 100 μm, displayed on a video television screen. Blood vessel diameter, blood flow velocity, and blood flow rate throughout and after the blood letting and fluid infusion were compared with the baseline values. To analyze blood flow velocity, the play speed of the video recorder was set at 1/60 of a second. The distances between two erythrocytes at the center of the blood vessel were measured 10 times, and the values were averaged. Blood flow rate was calculated by multiplying the blood flow velocity by the blood vessel cross-sectional area.
Statistical analysis
All data are expressed as means ± SD. Statistical comparisons were performed using repeated-measure analysis of variance followed by Fisher's protected least significant difference test. A value of P < 0.05 was considered to indicate statistical significance.
RESULTS
Immediately after blood removal, there were no significant differences between the HES group and the LR group in any measured variable except for HR. Microscopic examination showed that mean arteriolar diameter decreased to 40.5% ± 14.8% of the baseline value in the HES group compared with 43.3% ± 13.1% in the LR group. Immediately after the completion of infusion, mean arteriolar diameter significantly recovered to 90.8% ± 10.2% of the baseline value in the HES group compared with 62.6% ± 10.7% in the LR group (P < 0.005; Fig. 1). Mean arteriolar diameter was significantly lower than the baseline value in both groups from immediately after blood withdrawal to the completion of infusion (i.e., throughout the experiment; P < 0.005).
Microscopic views of microvessels observed with the REC method are shown in Figures 2 and 3. In the both groups, the diameters of the arterioles and venules were reduced, and the capillary density decreased after blood withdrawal (Figs. 2B and 3B). After the infusion, the diameters of the arterioles and venules improved (Fig. 2C) in the HES group. In the LR group, they remained poor (Fig. 3C).
Blood flow velocity also differed significantly between the groups. Immediately after blood removal, mean blood flow velocity decreased to 15.5% ± 6.0% of the baseline value in the HES group compared with 15.9% ± 6.9% in the LR group. After the completion of infusion, mean blood flow velocity recovered to 93.5% ± 16.6% of the baseline value in the HES group compared with 57.6% ± 10.3% in the LR group (P < 0.005; Fig. 4). Blood flow velocity was significantly lower than the baseline value in both groups throughout the experiment (P < 0.05).
Similar trends were seen in blood flow rate. Immediately after blood removal, mean blood flow rate decreased to 2.90% ± 2.12% of the baseline value in the HES group compared with 3.50% ± 2.41% in the LR group. After the completion of infusion, mean blood flow rate significantly recovered to 79.1% ± 26.7% of the baseline value in the HES group compared with 23.5% ± 9.6% in the LR group (P < 0.005; Fig. 5). The blood flow rate was significantly lower than the baseline value in both groups throughout the experiment (P < 0.05).
HR did not differ significantly between the groups (Fig. 6A). In the HES group, HR was significantly lower than the baseline value after the withdrawal of 40 to 80 mL of blood and during infusion (P < 0.005). In the LR group, HR was significantly lower than the baseline value after the withdrawal of 60 to 80 mL and during infusion (P < 0.005).
MAP at the completion of blood letting fell to 21.4% ± 5.6% of the baseline value in the HES group and to 22.0% ± 7.0% in the LR group (Fig. 6B). After the completion of infusion, MAP was equivalent to 97.0% ± 7.3% of the baseline value in the HES group compared with 71.0% ± 12.6% in the LR group (P < 0.005). The recovery of arterial pressure was significantly better in the HES group than in the LR group. MAP was significantly lower than the baseline value in both groups throughout the experiment (P < 0.005).
Similar trends were seen in CVP. CVP decreased after blood removal in both groups (Fig. 6C). CVP was significantly lower in the LR group than in the HES group after the completion of infusion (P < 0.001). In the LR group, CVP was significantly lower than the baseline value throughout the experiment (P < 0.001). In the HES group, CVP after blood removal and infusion of 20 mL was significantly lower than the baseline value (P < 0.001).
Mean plasma COP after the completion of blood letting decreased to 12.4 ± 3.00 mmHg of the baseline value in the HES group compared with 11.2 ± 2.34 mmHg in the LR group. After the completion of infusion, mean plasma COP was significantly higher in the HES group (19.9 ± 2.38 mmHg) than in the LR group (7.92 ± 1.96 mmHg; P < 0.001; Table 1).
Mean base excess after the completion of blood letting decreased to 1.18 ± 1.12 mM L-1 in the HES group compared with 1.25 ± 1.39 mM L-1 in the LR group. After the completion of infusion, mean base excess was significantly higher in the HES group (1.8 ± 0.99 mM L-1) than in the LR group (-0.59 ± 1.88 mM L-1 ; P < 0.005). In both groups, mean base excess after the completion of blood letting and after the completion of infusion was significantly lower than the baseline value (P < 0.05; Table 1).
Mean hematocrit (Hct) after the completion of blood removal and after the completion of infusion was significantly lower than the baseline value in both groups (P < 0.05). The Hct after the completion of infusion in the HES group was slightly but not significantly lower than that in the LR group. The other differences between the groups were also not significant (Table 1).
Urinary output was significantly greater in the HES group (6.4 ± 2.3 mL kg-1 h-1) than in the LR group (3.1 ± 2.9 mL kg-1 h-1; P < 0.01).
DISCUSSION
As compared with LR, we found that HES better maintained arteriolar diameter, blood flow velocity, and blood flow rate, more promptly restored and effectively maintained blood pressure, more promptly restored CVP, was associated with a significantly greater urine output, and better maintained plasma COP.
Pathophysiologically, hemorrhagic shock is characterized by systemic microcirculatory failure (18, 19). Bleeding with an inadequate circulating blood volume leads to hypotension. In response, baroreceptors excite the sympathetic and adrenaline systems, causing peripheral vessels to contract and leading to so-called microcirculatory failure (20, 21). Microcirculatory dynamics during shock are characterized by contraction or disappearance of arterioles, with cessation of blood flow (22). Our study showed that increased blood loss was associated with progressive decreases in blood pressure and CVP, producing signs of circulatory failure in both groups. As for the microcirculation of the ear, initial removal of blood was accompanied by hemodilution, restoring the blood flow in both groups. However, with subsequent removal of blood, the blood flow gradually decreased, and the microcirculation in the LR group was similar to that during shock, even after the start of infusion. In the HES group, arteriolar diameter, blood flow velocity, and blood flow rate of the ear were significantly maintained after the start of infusion, and microcirculatory failure did not develop. The HES used in this experiment had a molecular weight of about 70,000 and is reported to remain in the blood as needed (23), helping to maintain circulating blood volume and to restore blood pressure, thereby preventing the onset of microcirculatory failure. The improvement in blood loss-induced metabolic acidosis after the infusion of HES also supports maintenance of the microcirculation.
Crystalloid solutions are distributed to the intravascular blood and the interstitial fluid in a ratio of 1:3 shortly after infusion (24). Maintenance of blood volume by crystalloid solutions alone requires an infusion volume about four times greater than the bleeding volume in humans. This large volume of solution increases total body water and causes hypoalbuminemia. Retention of water and sodium in the lungs has also been reported (25). Infusion of large volumes of crystalloid solutions can cause pulmonary edema (26). In our preliminary experiments using rabbits, infusion of LR in a volume 3-fold higher than that of HES volume caused an acute rise in CVP and pulmonary edema, resulting in hypoxemia and circulatory insufficiency. The microcirculation could not be evaluated in the REC. Therefore, the volume of LR used in the present study was reduced to 2.5 times the blood loss volume. When colloidal solutions are used, some human studies have reported that an infusion volume 1.5 times greater than the blood loss is sufficient (24). In our experiment, the infusion volume of HES was reduced to 1.25 times the volume of blood removed to correspond to the reduced volume of LR used. The CVP was restored to greater than 90% of the baseline level, suggesting that an infusion volume approximating the bleeding volume is adequate to maintain blood pressure. Assessment of the microcirculation indicated that blood pressure was maintained without sacrifice of the peripheral circulation.
As for the rate of infusion, administration of the required amount of infusion solution within several minutes would have placed considerable stress on the circulation of the rabbit. Previous studies of infusion therapy in animal models of blood loss have assessed the effects of rapid intravenous injection and intravenous infusion (3, 27). On the basis of our preliminary experiments, we decided to initially give one-fifth of the required infusion volume by rapid intravenous injection, and the remaining portion by intravenous infusion to ensure the rabbits' survival.
The plasma osmotic pressure in the HES group was significantly higher than that in the control group. The osmotic pressure of the HES used in this experiment was similar to that of the plasma osmotic pressure (28). Therefore, the plasma volume could be increased without inducing transfer of interstitial fluid to the vascular compartment. Early distribution to the extravascular compartment was also minimal, enabling the restoration of blood pressure and the maintenance of blood volume without upsetting the fluid balance of the body. However, the administration of large volumes of plasma substitutes carries an increased risk of hemorrhage as an adverse reaction. Studies of the relation of the HES dose to coagulation test results and to clinical evidence of hemorrhage have reported that the tolerated dose of HES is about 30 mL kg-1 (29). In our experiment, the administered dose of HES was about 30 mL kg-1, within the clinically used range, and there was no evidence of problematic adverse effects.
Effects of plasma substitutes on the peripheral circulation include increased blood flow, resulting from reduced blood viscosity due to dilution. Intraoperative administration of HES has been reported to dilate small arteries in the kidney, increasing renal blood flow (30). Silk (31) found that direct administration of HES and Dextran into the renal artery improves systemic hemodynamics, increasing renal blood flow. Thus, these drugs were shown to directly cause renal vasodilatation. Yoshikawa et al. (32) measured organ blood flow after hemodilution with plasma substitutes and verified an increase in skin blood flow. Holbeck et al. (33) showed in a model of endotoxemia in the cat that hypovolemia is an important cause of disturbed intestinal blood flow and metabolism, even in systemic inflammatory response syndrome. They reported that the infusion of colloidal solution was useful for improving these disturbances. In all of these studies, the microcirculation was directly monitored by means of electromagnetic rheometers, laser Doppler rheometers, isotopes, or plethysmographs. Studies using mesenteric arteries of the rat (4) or pig (34) are invasive. Noninvasive studies of the microcirculation by intravital microscopy have been done using the bulbar conjunctiva of the dog (7) and a skin fold chamber of the hamster (8). These studies directly showed that the peripheral circulation was improved by the injection of blood substitutes. We also measured blood vessel diameter and blood flow rate by real-time direct monitoring and directly demonstrated that HES infusion was superior to LR infusion for maintenance of the peripheral circulation.
Urine volume was significantly higher in the HES group than in the LR group. The kidneys are the one of the first organs affected by shock. Renal blood flow also decreases in response to massive hemorrhage. Administration of HES increased circulating blood volume, restored renal blood flow, and directly dilated arterioles in the kidney, thereby improving renal blood flow. Urine volume was consequently maintained.
In conclusion, microscopic in vivo observation of rabbit ear microcirculation showed that intravenous infusion of HES more effectively maintains the peripheral circulation, hemodynamics, and COP in a model of acute severe hemorrhage than does LR.
ACKNOWLEDGMENTS
The authors thank Professor Chiyoji Ohkubo and Professor Emeritus Makishige Asano (National Institutes of Public Health, Japan). They also thank the Japanese Heart Pressure Research Laboratory and Japanese Heart Pressure Research Promotion Society for supplying the facilities and equipment for the present study.
REFERENCES
1. Rafie AD, Rath PA, Michel MW, Kirschner RA, Deyo DJ, Prough DS, Grady JJ, Kramer GC: Hypotensive resuscitation of multiple hemorrhage using crystalloid and colloids. Shock 22:262-269, 2004.
2. Boldt J: Fluid choice for resuscitation of the trauma patient: a review of the physiological, pharmacological, and clinical evidence. Can J Anaesth 51:500-513, 2004.
3. Krausz MM, Bashenko Y, Hirsh M: Crystalloid and colloid resuscitation of uncontrolled hemorrhagic shock following massive splenic injury. Shock 16:383-388, 2001.
4. Paes-da-Silva F, Gonzalez AP, Tibirica E: Effects of fluid resuscitation on mesenteric microvascular blood flow and lymphatic activity after severe hemorrhage shock in rats. Shock 19:55-60, 2003.
5. Mittlmeier T, Vollmar B, Menger MD, Schewior L, Raschke M, Schaser K-D: Small-volume hypertonic hydroxyethyl starch reduces acute microvascular dysfunction after closed soft-tissue trauma. J Bone Joint Surg Br 85:126-132, 2003.
6. Boura C, Caron A, Longrois D, Mertes PM, Labrude P, Menu P: Volume expansion with modified hemoglobin solution, colloids or crystalloid after hemorrhagic shock in rabbits: effects in skeletal muscle oxygen pressure and use versus arterial blood velocity and resistance. Shock 19:176-182, 2003.
7. Cheung ATW, Driessen B, Jahr JS, Duong PL, Ramanujam S, Chen PCY, Gunther RA: Blood substitute resuscitation as a treatment modality for moderate hypovolemia. Artif Cells Blood Substit Immobil Biotechnol 32:189-207, 2004.
8. Wettstein R, Tsai AG, Erni D, Winslow RM, Intaglietta M: Resuscitation with polyethylene glycol-modified human hemoglobin improves microcirculatory blood flow and tissue oxygenation after hemorrhagic shock in awake hamsters. Crit Care Med 31:1824-1830, 2003.
9. Gierer P, Vollmar B, Schaser K-D, Andreas C, Gradl G, Mittlmeier T: Efficiency of small-volume resuscitation in restoration of disturbed skeletal muscle microcirculation after soft-tissue trauma and haemorrhagic shock. Langenbecks Arch Surg 389:40-45, 2004.
10. Paxian M, Keller SA, Huynh TT, Clemens MG: Perflubron emulsion improves hepatic microvascular integrity and mitochondrial redox state after hemorrhagic shock. Shock 20:449-457, 2003.
11. Fiddian-Green RG, Haglund U, Gutierrez G, Shoemaker WC: Goals for the resuscitation of shock. Crit Care Med 21:S25-S31, 1993.
12. Asano M, Ohkubo C: Physiological and pathophysiological events of cutaneous microcirculation observed in the rabbit ear chamber. In Tsuchiya M, Wayland H, Oda M, Ozaki I (ed): Intravital Observation of Organ Microcirculation. Amsterdam: Excerpta Medica, 1983, pp 31-48.
13. Komori M, Takada K, Tomizaw Y, Uezono S, Ozaki M: Urinary trypsin inhibitor improves peripheral microcirculation and bronchospasm associated with systemic anaphylaxis in rabbits in vivo. Shock 20:189-194, 2003.
14. Nakata T, Takada K, Komori M, Taguchi A, Fujita M, Suzuki H: Effects of inhalation of nitrous oxide on rabbit ear chamber microvessels. In Vivo 12:375-378, 1998.
15. Komori M, Takada K, Ozaki M: Effects of inspired oxygen concentration on peripheral microcirculation studied by the rabbit ear chamber method. In Vivo 15:303-308, 2001.
16. Shiotani M, Takada K, Komori M, Nakata T, Taguchi A, Suzuki H: The effect of ulinastatin on cutaneous microcirculation during inhalation of 100% oxygen in a rabbit ear chamber. In Vivo 11:173-178, 1997.
17. Komori M, Takada K: Microcirculatory study of a rabbit ear chamber for anesthetic management. J Jpn Soc Biorheol 14:48-61, 2000.
18. Miyazaki M, Nakamichi S: Peripheral microcirculation. Jpn J Anesthesiol 20:5-11, 1971.
19. Granger DN, Rutili G, McCord JM: Superoxide radicals in feline intestinal ischemia. Gastroenterology 81:22-29, 1981.
20. Zweifach BW, Lowenstein BE, Chambers R: Response of blood capillaries to acute hemorrhage in the rat. Am J Physiol 142:80, 1944.
21. Ninomia I, Irisawa H: Non-uniformity of sympathetic nerve activity in response to baroceptor inputs. Brain Res 87:313-322, 1975.
22. Lillehei RC, Longerbeam JK, Bloch JH, Manax WG: The nature of irreversible shock: experimental and clinical observations. Ann Surg 160:682-710, 1964.
23. Wiedersheim M: An investigation of oxyethylstarch as a new plasma volume expander in animals. Arch Int Pharmacodyn 111:353-361, 1957.
24. Imm A, Carlson RW: Fluid resuscitation in circulatory shock. Crit Care Clin 9:313-333, 1993.
25. Arieff AI: Fatal postoperative pulmonary edema: pathogenesis and literature review. Chest 115:1371-1377, 1999.
26. Ley SJ, Miller K, Skov P, Preisig P: Crystalloid versus colloid fluid therapy after cardiac surgery. Heart Lung 19:31-40, 1990.
27. Abu-Hatum O, Bashenko Y, Hirsh M, Krausz MM: Continuous fluid resuscitation for treatment of uncontrolled hemorrhagic shock following massive splenic injury in rats. Shock 18:574-579, 2002.
28. Haupt MT, Rackow EC: Colloid osmotic pressure and fluid resuscitation with hetastarch, albumin, and saline solutions. Crit Care Med 10:159-162, 1982.
29. Saito Y, Uchida K, Yamamoto H, Ichiishi N, Matsushita F: The maximum safe dosage of the low molecular weight hydroxyethyl starch, Hespander (HES), estimated from HES induced changes in coagulation parameters and clinical bleeding. Jpn J Anesthesiol 48:238-243, 1999.
30. Lee WH, Cooper N, Weidner MG, Murner ES: Clinical evaluation of a new plasma expander, hydroxyethyl starch. J Trauma 8:381-393, 1968.
31. Silk MR: The effect of Dextran and hydroxyethylstarch on renal hemodynamics. J Trauma 6:717-723, 1966.
32. Yoshikawa H, Yamamura H, Yamaguchi Y, Kosugi I, Okada K: Redistribution of cardiac output during acute normovolemic anemia with plasma expanders in the dogs. Jpn Clin Physiol 5:442-447, 1975.
33. Helbeck S, Grande PO: Hypovolemia is a main factor behind disturbed perfusion and metabolism in the intestine during endotoxemia in cat. Shock 18:356-373, 2002.
34. Jonas J, Heimann A, Strecker U, Kempski O: Hypertonic/hyperoncotic resuscitation after intestinal superior mesenteric artery occlusion: early effects on circulation and intestinal reperfusion. Shock 14:24-29, 2000.
KeyWords-: Rabbit ear chamber; microvascular change; microvascular flow rate; vasoconstriction; direct observation