A New Model of Severe Hemorrhagic Shock in Rats

Abstract

We here introduce a fixed-pressure model of hemorrhagic shock in rats that maximizes effects on mean arterial blood pressure (MAP) during shock and yet maintains high reproducibility and controllability. The MAP of rats was adjusted to 25 to 30 mm Hg by blood withdrawals during 30 min. After a shock period of 60 min, rats were resuscitated either with lactated Ringer solution (LR) only or with the collected blood 3-fold diluted with LR (LR + blood) and monitored for further 150 min. Throughout the experiment, vital parameters and plasma marker enzyme activities and creatinine concentration were assessed. Thereafter, liver, kidneys, small intestine, heart, and lung were harvested and evaluated histopathologically. Vital parameters, plasma marker enzyme activities, creatinine concentration, and histopathology indicated pronounced but reliable and reproducible systemic effects and marked organ damage due to hemorrhagic shock and resuscitation. In contrast to rats that received LR + blood, which survived the postresuscitation period, rats receiving LR only invariably died shortly after resuscitation. The hemorrhagic shock model we present here maximally affects MAP and yet is highly reproducible in rats, allowing the study of various aspects of hemorrhagic shock and resuscitation under clinically relevant conditions.

Hemorrhagic shock results from a life-threatening loss of blood and leads to tissue ischemia and insufficient evacuation of cellular metabolic degradation products. Mortality is linked directly to massive blood loss or occurs indirectly due to secondary multiple organ failure. In particular, loss of gastrointestinal, renal, hepatic, and pulmonary function is frequent after hemorrhagic shock.^2,14,18,46

Current guidelines for nonsurgical treatment of hemorrhagic shock recommend rapid volume resuscitation by using crystalloids to restore the intravascular volume.³³ However, this practice is controversial because aggressive restoration of intravascular volume with a rapid increase in blood pressure before controlling hemorrhage can lead to increased mortality.²⁰ Consequently, ‘controlled fluid resuscitation’ with small volumes before stopping bleeding has been advocated.^4,9 In particular, treating severe hemorrhagic shock that bears the risk of interruption of blood flow to vital organs remains clinically challenging.³⁸ To improve existing strategies in emergency medicine, the characteristics of and possible therapeutic interventions for severe hemorrhagic shock need to be studied in an effective, realistic, and reproducible animal model.

Numerous animal models intended to simulate and study hemorrhagic shock conditions have been established so far.^{1,6-8,12,16,23,25-30,35,37,45} In view of experimental reproducibility and reliability, fixed-pressure models are superior to fixed-volume approaches. In fixed-pressure animal models, hemorrhagic shock is achieved by withdrawing blood until the desired blood pressure is reached; during the shock period, the blood pressure is held at this value by further blood withdrawal or reinfusion of collected blood or other fluids. The circulating blood volume of mammals does not correlate strictly with body weight due to individual differences in the relation of blood volume to mass of fat and muscle tissue and does not linearly increase with body weight gain (heavier animals usually have more fat and therefore comparatively less blood). Therefore, unlike fixed-pressure models, fixed-volume models (which are based on body weight) yield variable decreases in mean arterial blood pressure (MAP) and thus in the intensity of hemorrhagic shock.

In existing fixed-pressure models, either the adjusted MAP remains high (35 mm Hg or above)^16,17,28 and thus shock-related tissue injury is rather low or the duration of the shock period is very short^8,12,39 or unrealistically long,^17,37 thus not reflecting the shock situations of 45 to 60 min typically seen in Western industrialized countries. In other studies using models based on fixed-pressure hemorrhagic shock, substantial variations in MAP are tolerated during shock episodes²³ or animals are not resuscitated,³⁷ a practice that is reasonable only for very mild models and that excludes the possibility to study resuscitation regimens for severe hemorrhagic shock.

Considering the advantages and disadvantages of the already existing hemorrhagic shock models in the context of clinical problems and characteristics of most severe hemorrhagic shock, we here describe a fixed-pressure model in rats that involves severe shock, is life-threatening, and is characterized by a well-controlled MAP that is just high enough to reliably ensure survival after a clinically realistic shock period of 60 min and for a reasonable period subsequent to a 30-min resuscitation. Moreover, we minimized artificial influences like heparinization, which we replaced with acid citrate dextrose solution A (ACDA), as is contained in human stored blood. Our new rat model likely is applicable to studies of various resuscitation regimens and potentially beneficial interventions during severe hemorrhagic shock.

Materials and Methods

Chemicals and materials.

Formalin solution (10%, buffered) and hematoxylin (51260, Fluka Sigma-Aldrich, Steinheim, Germany), Ringer and lactated Ringer (LR) solutions (Fresenius, Bad Homburg, ketamine (10%, Ceva, Düsseldorf, Germany), lidocaine (xylocaine 1%, AstraZeneca, Wedel, Germany), and ACDA (Baxter, Deerfield, IL) were obtained from commercial sources. Catheters (inner diameter, 0.58 mm; outer diameter, 0.96 mm; Portex, Smiths Medical International, Hythe, UK), neonatal blood filters (200 µm; Impromediform, Lüdenscheid, Germany), paraffin (catalog no. 501006, Paraplast Tissue Embedding Medium, McCormick Scientific, St Louis, MO), surgical suture (Resorba, Nürnberg, Germany), medical oxygen (Air Liquide, Düsseldorf, Germany), and isoflurane (Forene 100%, Wiesbaden, Germany) were obtained from the listed vendors.

Animals.

Male Wistar rats (Rattus norvegicus; n = 16; weight, 400 to 450 g; age, 12 to 15 wk) were obtained from the Central Animal Unit of Essen University Hospital (Essen, Germany). Animals were kept under standardized conditions of temperature (22 ± 1 °C), humidity (55% ± 5%), and 12:12-h light:dark cycles with free access to food (Ssniff Spezialdiäten, Soest, Germany) and water. All animals received humane care according to the standards of Annex III of the directive 2010/63/EU of the European Parliament and of the Council of 22 September 2010 on the protection of animals used for scientific purposes.¹³ The experimental protocol has been approved according to the local animal protection act.

Anesthesia, analgesia, and catheter placement.

Experimental procedures are summarized in Figure 1. Rats were anesthetized with isoflurane (2% in 100% medical O₂ at 1 L/min for induction of anesthesia, 1% to 1.5% throughout the experiment) through face masks connected to a vaporizer (Isofluran Vet Med Vapor, Drägerwerk, Lübeck, Germany) and received ketamine (50 mg/kg SC) into the right chest wall for analgesia; additional ketamine (40 mg/kg SC) was administered as needed (that is, to prevent reappearance of blink and interdigital reflexes) throughout experiments. Lidocaine (5 mg/kg SC) was applied locally prior to creation of a skin-deep incision along the right groin, and catheters were placed within the femoral artery and vein. Thereafter, a skin-deep incision was made along the ventral cervical skin and a catheter placed within the right external jugular vein. All catheters were fixed with surgical suture (silk). At the end of the experimental procedure (that is, after the postresuscitation period), the small intestine, kidneys, liver, lung, and heart were harvested under deep isoflurane anesthesia, and thus the rats were euthanized.

Figure 1.

Schedule of the experimental procedures. 1, Start of anesthesia; 2, analgesia (ketamine, lidocaine); 3, catheter inserted into femoral artery (blood sample 1; start of biomonitoring); 4, all catheters inserted (blood sample 2); 5, start of shock induction (SI); 6, end of shock induction (MAP = 25 to 30 mm Hg; blood sample 3); 7, start of resuscitation (R; blood sample 4); 8, end of resuscitation (blood sample 5); 9, blood sample 6; 10, blood sample 7; 11, blood sample 8; 12, resection of organs (death of rat).

Induction of hemorrhagic shock and resuscitation.

After catheter insertions, rats were allowed to stabilize for 30 min. Hemorrhagic shock was induced by removing 2 mL blood every 3 min through the femoral artery catheter, by using a 2-mL syringe (Terumo, Leuven, Belgium) prefilled with 0.2 mL ACDA solution. Bleeding was continued until the MAP dropped to 25 to 30 mm Hg; this process typically took about 20 min. During the following 10 min, the MAP was ‘fine tuned’ by blood sampling of smaller volumes (0.5 to 1 mL). The blood was stored in sterile plastic conical tubes at 37 °C. For the next 60 min, the MAP remained between 25 to 30 mm Hg, typically without the need of any further intervention. In some individual cases, small amounts (0.1- to 0.5-mL aliquots) of 0.9% NaCl solution had to be administered or additional small blood samples (0.1- to 0.5-mL aliquots) withdrawn, to keep the MAP within the desired range. During the 30 min after the shock episode, group-specific resuscitation fluids were infused into the jugular vein by using a syringe pump (Perfusor-Secura FT; B Braun, Melsungen, Germany). Experiments were continued for 150 min more, unless the rat died earlier. To compensate for fluid loss over surgical areas and respiratory epithelium and to stabilize the blood pressure during the shock period, 0.9% NaCl solution (5 mL/kg/h, 37 °C) was infused through the femoral vein catheter throughout the experiment in all rats studied.

Study groups.

We compared 3 experimental groups. The sham group (n = 4 rats) experienced no shock or resuscitation. The shock LR group (n = 6) was resuscitated with LR solution (37 °C) equal to 3 times the total volume of blood collected. The shock LR + blood group (n = 6) underwent resuscitation by autotransfusion of the collected ACDA-treated blood plus LR at twice the volume of the collected blood. The collected blood was mixed gently with the necessary amount of LR immediately before resuscitation, and the prewarmed (37 °C) blood–LR solution was administered through a neonatal (200-µm) blood filter to exclude microclots, which may have formed during blood storage. Samples of the resuscitation fluid were taken before and after resuscitation and centrifuged (3000 × g for 15 min at 25 °C). The free hemoglobin content and LDH activity of the resulting supernatant was quantified to control for hemolysis.

Biomonitoring.

For close regulation of our model, we monitored several systemic and vital parameters throughout the experimental procedure. Systolic, diastolic, and mean arterial blood pressure were measured continuously by using the femoral artery catheter, which was connected to a pressure transducer, and displayed on a monitor. Ringer solution was infused at 3 mL/h to keep the catheter functional. Heart rates were determined from systolic blood pressure spikes. The rectal temperature of the rats was monitored continuously by using a rectal sensor and maintained at 37 ± 1 °C by means of an underlying heated operating table and by covering the animal with aluminum foil. The breathing rate per minute was determined according to the number of ventilatory movements in 10-min intervals.

Assessment of blood and plasma parameters.

By using 2-mL syringes (Pico50, Radiometer Medical, Brønshøj, Denmark) containing 80 IU electrolyte-balanced heparin, blood samples (0.7 mL) for blood gas analysis and the assessment of marker enzyme activities were taken from the femoral artery catheter immediately after its insertion, after insertion of all catheters, at the end of shock initiation, immediately before the beginning of resuscitation, and at the end of the resuscitation period and 30, 90, and 150 min thereafter (Figure 1). After each blood sampling, 0.7 mL 0.9% NaCl solution was administered via the femoral artery.

Arterial oxygen and carbon dioxide partial pressures, oxygen saturation, pH, acid–base status, hemoglobin concentration, hematocrit, electrolytes (Na⁺, K⁺, Cl^–, Ca²⁺), metabolic parameters (lactate, glucose), and osmolality were assessed by using a blood gas analyzer (ABL 715, Radiometer, Copenhagen, Denmark).

Blood plasma was obtained by centrifugation (3000 × g for 15 min at 25 °C) and stored at 4 °C until its use (within 4 h). The plasma activities of LDH as a general indicator of cell injury, AST and ALT as indicators of liver injury, and creatine kinase to assess muscle cell injury and the concentration of creatinine as an indicator of kidney injury and function were determined by using a fully automated clinical chemistry analyzer (Vitalab Selectra E; VWR International, Darmstadt, Germany).

Histopathologic evaluation of organs.

Samples of the small intestine (beginning, middle, and end of jejunum), the entire left kidney, the left lobe of the liver, and the entire lung and heart were fixed for 24 to 48 h in formalin (10% neutral buffered). Paraffin-embedded sections (3 µm) were stained with hematoxylin–eosin and histopathologic changes documented photographically.

Control for hemolysis in blood for resuscitation.

The concentration of free hemoglobin within the resuscitation fluid of the shock LR + blood group was assessed to exclude infusion of partially hemolyzed blood. Hemoglobin in the supernatant of the resuscitation fluid was determined from the absorption of the hemoglobin Soret band; to this end, the absorption maximum between 400 and 420 nm was determined spectrophotometrically, with LR serving as a blank. Values were corrected for nonspecific absorption or turbidity at 475 nm. Hemoglobin concentrations (µmol/L) were calculated according to a molar extinction coefficient of ϵ = 131,000 M^–1cm⁻¹. Hemolysis also was assessed based on LDH activity. If the concentration of free hemoglobin or the activity of LDH within the resuscitation fluid exceeded the preshock plasma value of the respective rat by 30% or more, the fluid was not used, and the rat was euthanized.

Statistics.

Data are expressed as mean ± SEM. Comparisons among multiple groups were performed using one-way ANOVA either for nonrecurring or for repeated measures followed by Fisher least significant difference posthoc analysis. A P value of less than 0.05 was considered statistically significant.

Results

Blood pressure and other vital parameters.

In the sham group, MAP remained around 100 mm Hg throughout the experiment (Figure 2). In the shock groups, MAP reliably was decreased from 102 ± 5 mm Hg to a stable value of 27 ± 1 mm Hg by the withdrawal of 29.5 ± 1 mL blood/kg during 30 min. In rats receiving LR + blood for resuscitation, the MAP increased to around 90 mm Hg and slightly but nonsignificantly decreased during the remainder of the experiment. In rats of the shock LR group, MAP reached only 50 to 60 mm Hg during resuscitation and immediately dropped to 31 ± 3 mm Hg thereafter; the rats died short time later (Figure 2). The average time of death in this group was 70 ± 20 min subsequent to resuscitation.

Figure 2.

Effect of severe hemorrhagic shock and resuscitation on mean arterial blood pressure (MAP). Rats experienced hemorrhagic shock for 60 min and then were resuscitated either with lactated Ringer solution (LR) or LR + blood, as shown in Figure 1. Values shown represent the mean ± SEM of 4 rats for the sham group and of 6 rats per shock group. The curve of the LR group ends with the death of the first rat of this group. *, P < 0.05 compared with sham group (since shock induction); +, P < 0.05 compared with shock LR + blood group (since resuscitation).

The heart rate of the sham group animals remained constant around 340 bpm throughout the experiment (Table 1). In the shock animals, heart rate decreased to about 280 bpm during the shock period. Resuscitation with LR + blood increased the heart rate to basal values at the end of the experiment, whereas in the shock LR group, heart rate remained at approximately 300 bpm until death.

Table 1.

Effects of severe hemorrhagic shock and resuscitation on heart rate, breathing rate, and rectal temperature in rats

	Group	Basal	End of shock	End of resuscitation	2 h after resuscitation
Heart rate (bpm)
	Sham	342 ± 4.9	336 ± 13.8	318 ± 9.3	354 ± 14.7
	Shock LR	351 ± 11.7	282 ± 11.2a,b	300 ± 10.2a
	Shock LR + blood	345 ± 8.7	279 ± 12.5a,b	309 ± 7.6a	357 ± 8.8
No. of respirations per minute
	Sham	59 ± 3.0	63 ± 3.6	66 ± 1.6	58 ± 0.9
	Shock LR	55 ± 3.0	59 ± 3.2	59 ± 4.8
	Shock LR + blood	56 ± 1.5	60 ± 1.8	75 ± 2.1a,b	63 ± 4.0
Rectal temperature (°C)
	Sham	36.5 ± 0.1	37.2 ± 0.1a	37.2 ± 0.1a	37.3 ± 0.1a
	Shock LR	36.8 ± 0.1	36.4 ± 0.2b	36.6 ± 0.2b
	Shock LR + blood	36.7 ± 0.1	36.7 ± 0.1b	37.0 ± 0.1	37.0 ± 0.1

Animals experienced hemorrhagic shock for 60 min and then were resuscitated either with lactated Ringer solution (LR) or LR + blood, as shown in Figure 1. Data are shown as mean ± SEM. Basal values were obtained immediately after insertion of the catheter into the femoral artery.

^aP < 0.05 compared with basal value of the same group.

^bP < 0.05 compared with value for sham group at the same time point.

Rats in the shock LR + blood group exhibited a significant increase in breathing rates during the resuscitation period. At all other time points, values of both shock groups were not significantly different from those of sham group animals (Table 1).

The rectal temperature of the sham group animals slightly increased during the basal equilibration period and then remained constant throughout the experiment (Table 1). Time-dependent comparisons of these rats with those of both shock groups indicated a decrease in rectal temperature during the shock period and its increase during resuscitation in the shock LR + blood group but not the shock LR rats.

Blood hemoglobin.

In the sham group animals, blood hemoglobin concentration did not change significantly throughout the experiment (Figure 3). In both shock groups, blood hemoglobin decreased to about 9.5 g/dL at the end of shock induction and to around 8.0 g/dL at the end of the shock period. In the shock LR group rats, blood hemoglobin further decreased during resuscitation before they died. In rats receiving LR + blood, basal hemoglobin concentration was not restored until 30 min after resuscitation but finally exceeded the basal value. No signs of coagulation or hemolysis were apparent in the blood used for resuscitation, and no hemolysis occurred in rats during or after shock.

Figure 3.

Effect of severe hemorrhagic shock and resuscitation on blood hemoglobin concentration. Rats experienced hemorrhagic shock for 60 min and then were resuscitated either with lactated Ringer solution (LR) or LR + blood, as shown in Figure 1. Values shown represent the mean ± SEM of 4 rats for the sham group and of 6 rats per shock group. The curve of the LR group ends with the death of the first rat of this group. *, P < 0.05 compared with sham group; +, P < 0.05 compared with shock LR + blood group.

Plasma marker enzymes activities and creatinine concentration.

In sham group rats, plasma activities of creatine kinase, LDH, AST, and ALT and creatinine concentration remained fairly low throughout the experiment (Table 2). In the shock LR group rats, these values increased significantly during the shock period (approximately 6-fold for LDH; 4-fold for creatine kinase; 2-fold for AST, ALT, and creatinine). That increase continued during resuscitation, in which another doubling of all parameters occurred, except in creatinine concentration which remained roughly constant during that phase. Until the end of resuscitation, all values of the shock LR + blood group were comparable to those of the shock LR group. In the LR + blood group, the values of all parameters except creatinine concentration increased 6- to 7-fold during the postresuscitation period.

Table 2.

Effects of severe hemorrhagic shock and resuscitation on plasma markers in rats

		Basal	End of shock	End of resuscitation	2 h after resuscitation
LDH (U/L)
	Sham	78.0 ± 20.2	115.6 ± 10.6	130.7 ± 16.1a	124.4 ± 19.0
	Shock LR	79.0 ± 19.3	417.4 ± 89.7ab	1278.4 ± 241.9ab
	Shock LR + blood	54.2 ± 4.7	454.8 ± 110.1ab	1153.8 ± 144.8ab	6541.9 ± 1464.2ab
AST (U/L)
	Sham	47.3 ± 3.0	59.3 ± 7.6	59.5 ± 11.4	58.5 ± 11.1
	Shock LR	50.8 ± 3.3	110.8 ± 6.9ab	188.0 ± 24.4ab
	Shock LR+ blood	47.8 ± 3.1	117.0 ± 17.8ab	164.2 ± 17.3ab	735.0 ± 126.3ab
ALT (U/L)
	Sham	61.8 ± 4.3	85.6 ± 12.4	88.5 ± 15.8	80.0 ± 13.1
	Shock LR	64.2 ± 5.6	125.9 ± 4.6ab	246.7 ± 35.3ab
	Shock LR + blood	56.5 ± 3.1	112.8 ± 15.2a	233.8 ± 23.1ab	1346.9 ± 127.6ab
Creatine kinase (U/L)
	Sham	109.3 ± 30.4	214.0 ± 36.8a	188.8 ± 45.0	175.7 ± 15.2
	Shock LR	97.8 ± 29.2	415.7 ± 90.2a	607.8 ± 81.8ab
	Shock LR + blood	76.3 ± 15.4	486.0 ± 142.4ab	912.5 ± 247.7ab	5294 ± 770.9ab
Creatinine (mg/dL)
	Sham	0.47 ± 0.03	0.51 ± 0.02	0.51 ± 0.03	0.50 ± 0.04
	Shock LR	0.50 ± 0.02	0.90 ± 0.03ab	0.82 ± 0.02ab
	Shock LR + blood	0.47 ± 0.03	0.94 ± 0.07ab	1.17 ± 0.09ab	1.77 ± 0.10ab

Animals experienced hemorrhagic shock for 60 min and then were resuscitated either with lactated Ringer solution (LR) or LR + blood, as shown in Figure 1. Data are shown as mean ± SEM. Basal values were obtained immediately after insertion of the catheter into the femoral artery.

^aP < 0.05 compared with basal value of the same group.

^bP < 0.05 compared with value for sham group at the same time point.

Organ histopathology.

Hemorrhagic shock and resuscitation resulted in pronounced histopathologic changes to liver, kidneys, and small intestine. These changes were most pronounced in the shock LR + blood group (Figure 4), presumably because they formed during the prolonged survival of these rats subsequent to resuscitation. No obvious histologic alterations were observed in samples from lung and heart under these conditions. Organ samples of sham animals showed no injury at all.

Figure 4.

Effects of severe hemorrhagic shock and resuscitation with lactated Ringer solution (LR) + blood on liver, kidney, and small intestine histology. Rats experienced hemorrhagic shock for 60 min and then were resuscitated with LR + blood, as shown in Figure 1. Histologic changes were assessed at the end of the postresuscitation period by using light microscopy. The images shown are representative of samples from 6 shock and 4 sham rats. Note the numerous erythrocytes (black arrows) in liver and kidney tissue and the abnormal nuclear ruffling and chromatin condensation within hepatocytes and kidney cells (white arrows). Pronounced necrotic hemorrhages can be noted in the small intestine. Hematoxylin and eosin stain; original magnification: 400× (A through D); 100× (E and F).

Discussion

The methods described and experimental results presented here create a severe but reliable and realistic acute hemorrhagic shock model in rats. During and after a hemorrhagic shock period of 60 min during which MAP was monitored closely and maintained at just 25 to 30 mm Hg (Figure 2), rats exhibited marked organ damage, as demonstrated histologically (Figure 4) and indicated by remarkable release of marker enzyme activity (Table 2). Rats receiving LR + blood for resuscitation reliably survived the entire experimental period, whereas those that were resuscitated with LR alone died shortly thereafter.

To date, several hemorrhagic shock–resuscitation models have been established, but none of them covers all important aspects related to severity, clinical reality, and reproducibility. For example, severe shock with MAP of 30 to 35 mm Hg has been achieved by many groups,^35,44,45 but shock periods of 2 h or longer do not reflect human clinical practice. Likewise, a MAP of 25 to 30 mm Hg for only 20 min does not result in life-threatening organ injury and is not clinically relevant,³⁹ whereas 60 min of 45 mm Hg¹⁹ does not yield severe hemorrhagic shock.

In addition to its high severity, the model we describe here is closely orientated toward human clinical practice. Our experimental schedule of shock induction within 30 min, a hemorrhagic shock period of 60 min, and subsequent resuscitation for 30 min appears to reflect most emergency situations realistically²⁰ (Figure 5).

Figure 5.

Comparison of timelines of the hemorrhagic shock model presented (upper bar) and the current practice for treatment after severe trauma and blood loss in humans according to the Trauma Register of the Deutsche Gesellschaft für Unfallchirurgie (lower bar).²⁰

The administration of anesthetics and analgesics can decisively influence the effects of hemorrhage and trauma.^3,25,26 Humans in hemorrhagic shock subsequent to traumatic events often receive isoflurane or ketamine–xylazine.²¹ Ketamine seems to be an ideal emergency analgesic, in that it positively influences blood pressure, whereas isoflurane, unlike ketamine, improves microcirculation. In line with this practice, the rats in our model received isoflurane and ketamine. In preliminary studies, xylazine itself led to a drop in MAP in sham animals and thus was not used. In addition, although often provided to human shock patients, we did not use opiates, because they can decrease peristalsis and thus may influence injury of the gut; in addition, some opiates decrease tissue injury due to ischemia–reperfusion.⁴² With the use of ketamine and lidocaine, we were able to decrease the isoflurane concentration needed (especially during placement of incisions and catheters) and thus its hypotensive effects. Lidocaine was used, in addition to ketamine, for local analgesia because of its rapid onset of action.

In the present experiments, rats (male Wistar, 400 to 450 g) reliably survived a hemorrhagic shock-induced decrease in MAP to 25 to 30 mm Hg for at least 60 min. Under the same experimental settings, a MAP of 20 mm Hg or lower resulted in death of the rats in 20 min and a MAP of 21 to 24 mm Hg led to highly variable survival times in preliminary studies. MAP adjustment to pressures as low as 25 to 30 mm Hg have the additional advantage that the rats’ blood pressure usually does not increase again once the MAP has been adjusted. At higher MAP (for example, 35 to 45 mm Hg), regulatory mechanisms are still effective, making repeated intervention (further withdrawal of blood) necessary, which will negatively affect reproducibility. Consequently, the most severe hemorrhagic shock model that maximally affects MAP in rats and yet is highly reproducible should neither be based on a MAP below 25 mm Hg (because the effects are no longer reproducible) nor on much higher MAP values, because the effects are no longer severe. Infusion of 0.9% NaCl at 5 mL/kg/h, administered to compensate for evaporation through the respiratory epithelium and surgery wounds, was necessary to ensure stabilization of the MAP throughout the shock period. Pharmacologic agents and potentially protective substances can be added to this infusion to study their effects during and after hemorrhagic shock. Hemorrhagic shock between 25 to 30 mm Hg for 60 min allowed a reasonably prolonged postresuscitation period—and therefore opportunity to study resuscitation regimens—and approximates the most physiologically demanding condition.

Due to the severity of the shock phase, marked tissue injury became apparent shortly after resuscitation (Table 2; Figure 4). The tissue injuries and short survival time in the LR group were not due to decreased blood hemoglobin concentration (Figure 3), which was sufficient to ensure tissue oxygenation⁴¹ especially at 100% inspiratory oxygen, but to hypovolemia and altered or interrupted tissue perfusion. During both the hemorrhagic shock and resuscitation phases, massive fluid shifts occurred in both directions between the interstitial space and vessels, as indicated by the changes in hemoglobin concentration (Figure 3).

The composition of the fluids used for resuscitation (LR compared with LR + blood) decisively affected rat survival times. Survival times were prolonged significantly by resuscitation with blood compared with pure LR, consistent with other studies²² and clinical experience.³¹ The model therefore allows 100% survival of subjects, which is a particular advantage for studies requiring organ histopathology and endpoint blood analyses. Alternatively, survival time can be assessed as an important and meaningful parameter for evaluation of protective approaches or various resuscitation protocols.

In many models, heparin-based anticoagulants are administered to the animals prior to shock induction or are added to the collected blood to prevent its coagulation until its use for resuscitation.^24,25,39 However, under comparable human clinical situations, heparins are not used and may even be contraindicated.¹¹ Moreover, heparins have been reported to mediate protective effects during hemorrhagic shock,^5,32,43 and this result is unfavorable when, for example, the effects of drugs are intended to be studied. A more appropriate anticoagulant in this setting is ACDA, a citrate-based solution, because it is already contained in human stored blood that is used for resuscitation, and because it provides no protection against hemorrhagic shock.¹⁹ Therefore, we used ACDA instead of heparin in the current study.

Activation of coagulation and mild hemolysis are risks associated with blood sampling through catheters and during storage of collected blood; these side effects cannot be excluded completely. Infusion of hemolyzed blood could lead, for example, to artificially high plasma LDH activities and oxidative stress and kidney injury due to free hemoglobin. The intravenous application of invisible clots can lead to microembolism, in vivo activation of the coagulation cascade, and sudden unexpected death of the animals.³⁴ To verify that the blood used for resuscitation was not hemolytic or coagulated, the plasma hemoglobin concentration and LDH activity in the collected blood were quantified before and after resuscitation, and blood filters used to avoid infusion of microclots; none of our rats demonstrated any indication of abnormal coagulation or hemolysis.

In our model, ACDA-treated blood was mixed with LR (1:2) prior to infusion. The mixture or coinfusion of LR and blood introduces a risk that the calcium in LR exceeds the Ca²⁺-chelating capability of anticoagulants like ACDA. This situation could result in clot formation, especially at ambient temperature.^10,36 We minimized this risk by prewarming the blood–LR and by immediately infusing the mixture at 37 °C. However, care must be taken to exclude coagulation (for example, through examination of blood smears). Alternatively, LR and the collected blood can be administered through separate catheters. Normal saline must not be used for resuscitation, because it can lead to hyperchloremic metabolic acidosis at high rates of infusion.^15,40

In conclusion, the hemorrhagic shock model described here represents a severe but reproducible and realistic model in rats. The model allows reliable assessment of parameters such as survival times, plasma marker enzyme activity, and histopathologic alterations. Despite the limitations of animal studies in mimicking clinical reality, the timeline of our shock model reflects the current practice of treatment after severe trauma and blood loss in humans. Moreover, our methodic approach minimizes artificial and misleading influences like heparinization and the risk of unnoticed hemolysis and coagulation.