The Acute Immunological Response to Blood Transfusion Is Influenced by Polymicrobial Sepsis

Abstract

Blood transfusion is a well-established risk factor for adverse outcomes during sepsis. The specific mechanisms responsible for this effect remain elusive, and few studies have investigated this phenomenon in a model that reflects not only the clinical circumstances in which blood is transfused, but also how packed red blood cells (PRBCs) are created and stored. Using a cecal ligation and puncture model of polymicrobial sepsis as well as creating murine allogeneic and stored PRBCs in a manner that replicates the clinical process, we have demonstrated that transfusion of PRBCs induces numerous effects on leukocyte subpopulations. In polymicrobial sepsis, these responses are profoundly dissimilar to the proinflammatory effects of PRBC transfusion observed in the healthy mouse. Transfused septic mice, as opposed to mice receiving crystalloid resuscitation, had a significant loss of blood, spleen, and bone marrow lymphocytes, especially those with an activated phenotype. Myeloid cells behaved similarly, although they were able to produce more reactive oxygen species. Overall, transfusion in the septic mouse may contribute to the persistent immune dysfunction known to be associated with this process, rather than simply promote proinflammatory or anti-inflammatory effects on the host. Thus, it is possible that blood transfusion contributes to the multiple known effects of sepsis on leukocyte populations that have been shown to result in increased morbidity and mortality.

Sepsis is the leading cause of death in the critically ill with 750,000 cases and 210,000 deaths annually (¹); in fact, in the general surgery population, the incidence of sepsis is greater than the incidence of pulmonary embolism and myocardial infarction combined (²). Sepsis mortality has been attributed to derangements in the innate and adaptive immune systems (³). In addition, there are well known cofactors that are associated with an increased morbidity and mortality during sepsis, including blood transfusion (⁴). These associations have led to restrictive policies for transfusion in many hospitals in an attempt to improve outcomes (⁴).

Evidence increasingly demonstrates the adverse effects of blood transfusions in terms of subsequent patient infection and mortality. In trauma and critically ill patients, blood transfusions are associated with increased morbidity and mortality, especially with increased amounts of transfusion, as well as with the transfusion of blood that has been stored for longer periods (^{5, 6}). Blood transfusion is an independent risk factor for ventilator-associated pneumonia, perioperative/postinjury infection, and multiple organ dysfunction syndrome; in fact, the effect of blood transfusions being associated with poor outcomes is present even after adjustment for confounding factors such as age and severity of injury (^{6, 7}).

Although the cause of the increased morbidity and mortality associated with packed red blood cell (PRBC) transfusion is assuredly multifactorial, one of the major causes is believed to be the effect of PRBC transfusion on host protective immunity (^{8, 9}). The mechanisms behind blood transfusion-induced alterations in protective immunity remain unclear. Although debated, allogeneic, stored PRBCs are presumed to induce T-cell anergy, leukocyte dysfunction, and expansion of immunosuppressive cell populations (^{10, 11}).

Analysis of transfusion in a clinically relevant animal model is lacking. This is of importance because the contribution of transfusion to poor outcomes is generally associated with critically ill patients or individuals who have sustained a major systemic inflammatory response, whether from surgery or trauma. In fact, the only randomized controlled trials that illustrate worsened outcomes with PRBC transfusion (in certain subsets) have been carried out in intensive care unit populations (Transfusion Requirements in Critical Care Investigators and Transfusion Requirements After Cardiac Surgery trials) (^{12, 13}).

There are several challenges that have hindered translational research regarding PRBCs’ impact on this patient population. Difficulties exist in creating murine allogeneic stored PRBCs similar to what would be given to a human (¹⁴), as well as limitations in murine models of sepsis and trauma (¹⁵). Finally, there is a degree of anemia, especially in regard to hemorrhagic or traumatic shock where the immediate benefits of blood transfusion outweigh any risk (¹⁶). Therefore, isolating the mechanisms behind this well-known phenomenon has remained problematic.

In this report, we have created a murine model of blood transfusion that better recapitulates the treatment of typical human PRBCs, including storage and allogenicity, and then determine the effect of transfusion on the immune system in both healthy animals, as well as in a clinically relevant model of murine sepsis. We propose that transfusion of PRBCs will result in the alteration of the phenotype and functionality of specific immune effector cell populations dependent on whether the animal is healthy or septic. In the septic individual, these transfusions alter immune function, possibly leading to further risk for infection, end-organ damage, and mortality.

MATERIALS AND METHODS

Mice

All experiments were approved by the Institutional Animal Care and Use Committee at the University of Florida. Specific pathogen-free male C57BL/6 and BALB/c mice were purchased from Jackson Laboratory (Bar Harbor, Me) at 6 to 7 weeks of age and allowed to acclimatize for 1 week before being used for experimental procedures. Mice were maintained on standard rodent food and water ad libitum.

Allogeneic and stored PRBCs

BALB/c mice were placed in a CO₂ chamber until they become unresponsive, and then blood was immediately collected by intracardiac puncture using heparinized syringe. Whole blood was placed in sterile 1.6-mL tubes containing CPDA (citrate-phosphate-dextrose anticoagulant preservative) solution (1:6), mixed gently, and centrifuged at 4,000 revolutions/min for 5 min at 4°C. The plasma and buffy coat layers were carefully aspirated, and Adsol additive solution (Fenwal, Inc., Lake Zurich, IL) was added to the PRBCs (1:6) and stored at 4°C in the dark until use (14 days). Hematocrit was measured at different time points (60% at day 0), and no significant changes were identified up to 14 days (57%).

Cecal ligation and puncture

For induction of polymicrobial sepsis in C57BL/6 (B6) mice, cecal ligation and puncture (CLP) was performed under isoflurane anesthesia as previously described (¹⁷). Briefly, cecum was exposed after a laparotomy, ligated with 2-0 silk suture, and punctured through and through with a 25-gauge needle. The cecum was returned in the abdomen, and the incision was closed using surgical clips. After the procedure, the mice were administered 0.05 to 2 mg/kg buprenorphine in 1 mL 0.9% saline, returned to their respective cages, and closely monitored for any signs of distress. The Institutional Animal Care and Use Committee requires euthanasia for moribund mice that are then considered as nonsurvivors.

Transfusion

B6 mice underwent cannulation of one of the femoral arteries under isoflurane anesthesia as previously described (¹⁸). Total blood volume for the mouse was estimated to be 70 mL/kg or 1.75 mL of blood for a 25-g mouse. Animals were transfused with sterile room-temperature Ringer’s lactate (LR, 1 mL) solution or 600 to 700 μL of stored allogeneic PRBCs (derived from BALB/c mice), which is equivalent to 3 to 4 U of human PRBCs. The volume difference between the crystalloid and PRBCs delivered to the mouse was to compensate for the differences in intravascular volume between colloid and crystalloid. In addition, mice were not bled before transfusion or infusion of crystalloid as we did not wish to add shock to one of the variables to consider or to create significant anemia in the crystalloid group of mice as a confounder. Mice that underwent CLP were transfused on postoperative day 2. Again, mice were not bled before transfusion or crystalloid infusion as septic mice require resuscitation (and they would be at risk for death if we remove further intravascular volume) as well as not wishing to create severe anemia in the crystalloid group. Mice were subsequently killed 1 day after transfusion, and their blood, spleen, and bone marrow were harvested as previously described (¹⁹).

Flow cytometry

Spleens, whole blood, and bone marrow were harvested, and single-cell suspensions were created by passing the cells through 70-μm-pore cell strainers (BD Falcon, Durham, NC). Erythrocytes were then lysed using ammonium chloride lysis buffer and washed two times using phosphate-buffered saline (PBS) without calcium, phenol red, or magnesium. Cells were stained with the following antibodies for flow cytometric studies: fluorescein isothiocyanate anti-CD8, fluorescein isothiocyanate anti-CD11c, phycoerythrin (PE) anti-CD69, PE anti-IA/IE, allophycocyanin anti-CD4, PE Cy7 anti-CD25, PE Cy7 anti-CD11b, allophycocyanin anti–Gr-1, and pacific blue anti-CD19, and anti-Ly6G (BD Pharmingen, Billerica, Mass). Samples were acquired and analyzed using an LSRII flow cytometer (BD Biosciences, San Jose, Calif) and FACSDiva (BD Biosciences (¹⁹).

Lipopolysaccharide stimulation

Splenocytes, 10⁶, were placed in 24-well culture plates (Corning Life Sciences, Lowell, Mass) with 500 μL Dulbecco modified Eagle medium (CellGro, Manassas, Va) supplemented with 10% fetal calf serum, 10 mM HEPES, glutamine, and penicillin/streptomycin (complete Dulbecco modified Eagle medium). Cultured cells were stimulated with 1 μg/mL lipopolysaccharide (LPS) (Escherichia coli O26:B6; Sigma-Aldrich, St Louis, Mo). Culture supernatant was harvested 24 h later and stored at −80°C until assayed.

Cytokine production

One day after transfusion, whole blood was harvested by intracardiac puncture. The plasma was collected and stored at −80°C until the time of analysis. Assessments of cytokine concentrations from the mouse plasma and culture supernatant were performed using a commercially available multiplexed Luminex kit (MILLIPLEX MAP, Mouse Cytokine/Chemokine Panel; Millipore, Billerica, Mass). Cytokines evaluated included interleukin 1β (IL-1β), IL-6, IL-12 (p70), interferon-inducible protein 10 (IP-10), keratinocyte-derived chemokine (KC), monocyte chemoattractant protein 1 (MCP-1), macrophage inflammatory protein 1α (MIP-1α), and tumor necrosis factor α (TNF-α). All assays were performed according to the manufacturer’s protocols. Cytokine concentrations were determined using BeadView software (Millipore).

Reactive oxygen species detection

Splenocytes were prepared using a Ficoll density gradient (1.104 specific gravity) and washed using PBS without calcium, phenol red, or magnesium. Cells were then labeled for surface markers as described and washed twice with PBS. Reactive oxygen species production was determined using dihydrorhodamine 123 (DHR123; Invitrogen, Carlsbad, Calif). Subsequently, cells were stimulated with phorbol myristate acetate at 37°C and evaluated by flow cytometry analysis at various points over the subsequent 30 min period. A minimum of 1 × 10⁴ live, non-debris cells were collected for analysis.

Statistical analysis

Continuous variables were first tested for normality and equality of variances. Differences among groups in flow cytometric analyses were evaluated using Student t test. All statistical analyses were performed using GraphPad Prism 5.0 (GraphPad software, La Jolla, Calif).

RESULTS

Transfusion of stored allogeneic PRBCs in a naive mouse acutely induces a proinflammatory response

We initially sought to determine what the baseline immune response was to transfusion of aged, allogeneic PRBCs in a healthy animal. Twenty four hours after transfusion, analysis of the spleen demonstrated that there were significantly greater numbers of splenic CD4⁺ T cells, as well as a trend for increased CD8⁺ T cells and CD19⁺ B cells (Fig. 1A). Almost all the lymphocytes had an increased activation status as determined by CD69 expression (Fig. 1B). There was also a trend toward an increase in the number of CD11b⁺ myeloid cells (data not shown). There was no significant difference in the total number of leukocytes present in the spleens of animals that received LR solution or PRBCs (data not shown).

Fig. 1: Absolute numbers and activation status of splenocytes after PRBC transfusion in naive mice. Spleens from C57Bl/6 mice were harvested 1 day after receiving either allogeneic PRBCs (n = 3) or LR solution (n = 3). Cellular phenotypes were analyzed by flow cytometry using anti-CD4 (T-helper cells), anti-CD8 (cytotoxic T lymphocytes), and anti-CD19 (B cells). The activation status was determined by using anti-CD69. Data are shown in absolute numbers and are representative of three separate experiments, with an n = 3 per experiment (*P < 0.05).

Analysis of these splenocytes further reinforced the concept that PRBC transfusion in a healthy animal has a similar effect to other inflammatory stimuli. Splenocytes from transfused mice produced significantly more proinflammatory cytokines and chemokines in response to ex vivo LPS stimulation (Fig. 2). Exposure of splenocytes to 2-week-old PRBCs or preservative solutions, CPDA and Adsol, in vitro was not able to independently induce an exaggerated cytokine/chemokine response (Fig. 3). Thus, although splenocytes from mice that received blood transfusion were more responsive to LPS stimulation, coculturing of blood, blood products or preservatives with naive spleen cells did not induce a similar increase in the production of cytokines/chemokines, indicating the absence of LPS contamination in the transfusate. Although not significant, proinflammatory cytokines in the plasma were detected at higher concentrations in PRBC-transfused healthy mice compared with mice that received LR solution (Fig. 4).

Fig. 2: Splenocyte cytokine and chemokine secretion of naive mice after PRBC transfusion. Spleens from C57Bl/6 mice were harvested 1 day after receiving either allogeneic PRBCs (n = 3) or LR solution (n = 3). One million spleen cells were treated with LPS (1 μg/mL). Culture supernatants were collected 24 h later and assessed for cytokine/chemokine production using multiplex Luminex kit. Data were confirmed by a second experiment (*P < 0.05, **P < 0.01, ***P < 0.001).

Fig. 3: Splenocyte cytokine and chemokine secretion after exposure to PRBC, preservative solution, or LPS. One million spleen cells from naive mice were treated with either LPS (1 μg/mL), PRBCs (30% volume), or CPD and Adsol solution (30% volume). Cells in medium served only as untreated controls. Culture supernatants were collected 24 h later and assessed for cytokine/chemokine production using multiplex Luminex kit. ****P < 0.0001, one-way analysis of variance.

Fig. 4: Plasma cytokine and chemokine production of naive mice after PRBC transfusion. Blood was collected from mice 1 day after receiving either allogeneic PRBCs (n = 9) or LR solution (n = 9) in heparinized syringe by intracardiac puncture. Plasma cytokine/chemokine levels were measured using multiplex Luminex kit. Data shown are the mean from three separate experiments.

Proinflammatory events associated with exposure to microbial products or danger signals can also lead to the eventual downregulation of cell surface markers that are upregulated in the early acute inflammatory period. One prototypical response to an inflammatory event is the loss of major histocompatibility complex class II (MHCII⁺) expression on circulating monocytes as well as other myeloid-derived cells (²⁰). Interestingly, 1 day after transfusion, there was a significant decrease in the mean fluorescence intensity of circulating MHCII⁺ expression on both Ly6G⁻CD11b⁺ monocytic (Fig. 5A) and CD11c⁺ dendritic cells (Fig. 5B) consistent with reduced antigen presentation capacity.

Fig. 5: MHCII⁺ status of circulating monocytes and dendritic cells in naive mice after PRBC transfusion. Blood was collected from C57Bl/6 mice 1 day after receiving either allogeneic PRBCs (n = 3) or LR solution (n = 3) in heparinized syringe by intracardiac puncture. The MHCII expressions (IA/IE⁺) of monocytes (Ly6G⁻CD11b⁺) and dendritic cells (CD11c⁺) were analyzed by flow cytometry. Data shown are representative of three separate experiments (*P < 0.05, **P < 0.01).

The response of leukocytes to allogeneic stored PRBCs during sepsis is different to that of healthy mice

Surprisingly, PRBC transfusion following polymicrobial sepsis induced alterations that are dissimilar to the effects observed in a healthy mouse. Interestingly, PRBC administration was associated with a significant loss of CD8⁺ T cells and CD19⁺ B cells in the spleen, in contrast to an increase seen in naive mice. In addition, the activation status of most lymphocytes, as determined by CD69 expression, was decreased, rather than increased as compared with animals receiving crystalloid solution (Fig. 6 and Table 1). In addition, there was also an absolute loss of splenic, myeloid-derived leukocytes, including Ly6G⁺CD11b⁺ neutrophils, Ly6G⁻CD11b⁺ monocytes, Gr-1⁺CD11b⁺ myeloid derived suppressor cells, and CD11c⁺ dendritic cells (Fig. 7). Associated with this was a further overall loss in the total number of leukocytes present in the spleen (data not shown).

Table 1: Absolute cell counts in spleen 1 day after LR solution or PRBC transfusion*

Fig. 6: Absolute numbers and activation status of splenocytes after PRBC transfusion in septic mice. Male C57BL/6 mice received either allogeneic PRBCs (n = 11) or LR solution (n = 9) 2 days after CLP. Spleens were harvested 1 day after transfusion and analyzed by flow cytometry using anti-CD4 (T-helper cells), anti-CD8 (cytotoxic T lymphocytes), and anti-CD19 (B cells). The activation status was determined by using anti-CD69. Data shown are the mean from three separate experiments (*P < 0.05, **P < 0.01).

Fig. 7: Absolute numbers of splenic myeloid cells after PRBC transfusion in CLP mice. Myeloid-derived leukocytes from spleens of septic mice were analyzed by flow cytometry using anti-CD11b and anti-Ly6G (neutrophils), anti–Gr-1, and anti-CD11b (myeloid derived suppressor cells), and anti-CD11c (dendritic cells). Monocytes were CD11b⁺ and Ly6G⁻. Data shown are the mean of three separate experiments (*P < 0.05, **P < 0.01).

These effects associated with the transfusion of PRBCs during sepsis were not limited to the spleen. There was also a relative and absolute loss of myeloid precursors in the bone marrow of septic mice after PRBC transfusion, as compared with crystalloid resuscitation (Fig. 8). The percentage of circulating lymphocytes also decreased, as well as the activation status of the circulating CD4⁺ T cells (Fig. 9A). In addition, the percentage of circulating dendritic cells declined (Fig. 9B).

Fig. 8: Percent and absolute numbers of GR-1⁺11b⁺ cells in the bone marrow of mice after CLP and then resuscitated with PRBCs or LR solution. The percentage (left) and absolute number (right) of Gr-1⁺CD11c⁺ cells in the bone marrow of septic mice 1 day after receiving either PRBCs (n = 10) or LR solution (n = 9) are shown as mean from three separate experiments (*P < 0.05).

Fig. 9: Percentage of circulating leukocytes in septic animals after resuscitation with PRBCs or LR solution. Blood was collected from septic mice 1 day after receiving either allogeneic PRBCs (n = 11) or LR solution (n = 9) in heparinized syringe by intracardiac puncture. Leukocytes were analyzed by flow cytometry using anti-CD4 (T-helper cells), anti-CD8 (cytotoxic T lymphocytes), and anti-CD19 (B cells). The activation status was determined by using anti-CD69 (A). Dendritic cells were identified using anti-CD11c (B). Data shown are the mean from three separate experiments (*P < 0.05, **P < 0.01).

Allogeneic stored PRBC transfusion during sepsis is associated with a dysregulated inflammatory state

The functional response of splenocytes in septic animals after transfusion reflects a different immune environment. Packed red blood cell–transfused mice secreted more IP-10, MIP-1α, TNF-α, and IL-10 in response to LPS as compared with crystalloid-resuscitated, septic mice, similar to that seen in healthy naive animals, but produced less KC and IL-6, which was opposite of what was seen in naive mice (Fig. 10A and Table 2). A similar trend was seen in plasma cytokine levels of these mice (Fig. 10B). In addition, although there were fewer splenic neutrophils and monocytes, transfusion during sepsis increased the capacity of these cells to produce reactive oxygen species (Fig. 11A) compared with mice that received LR solution, a phenomenon not demonstrated in naive mice (Fig. 11B). Finally, mortality was modestly increased in septic mice that received blood transfusion, although this did not reach statistical significance (Fig. 12).

Table 2: LPS-stimulated spleen cytokine production*

Fig. 10: Splenocyte function of septic mice after PRBC transfusion. A, Spleens from septic mice were harvested 1 day after receiving either allogeneic PRBCs (n = 7) or LR solution (n = 7). One million spleen cells were treated with LPS (1 μg/mL). Culture supernatants were collected 24 h later and assessed for cytokine/chemokine production using multiplex Luminex kit. B, Blood was collected from septic mice 1 day after receiving either allogeneic PRBCs (n = 13) or LR solution (n = 11) in heparinized syringe by intracardiac puncture. Plasma cytokine/chemokine levels were measure using multiplex Luminex kit. Data shown are the mean from three separate experiments.

Fig. 11: Reactive oxygen species generation of splenic neutrophils and monocytes after PRBC transfusion during sepsis. Spleens from septic (A) and naive (B) mice were harvested 1 day after receiving either PRBCs or LR solution were stimulated ex vivo with phorbol myristate acetate in the presence of DHR123 at 37°C for up to 30 min. Mean fluorescence intensity of DHR123 was measured from neutrophil- and monocyte-gated cells (*P < 0.05).

Fig. 12: Survival of septic mice with blood versus crystalloid resuscitation. C57Bl/6 mice underwent CLP with a 25-gauge needle. Survival of septic mice receiving either PRBCs (n = 23) or LR solution (n = 18) on day 2 was followed up to 7 days after CLP. The figure is the combination of two separate survival studies performed.

DISCUSSION

During sepsis, blood transfusion appears to modify and, in places, amplify the known immunological effects of sepsis, including further exaggerating inflammatory cytokine production, while simultaneously inducing immune-suppressive phenotypes known to affect mortality to primary and secondary infections. By using a relatively more clinically relevant model of murine PRBC creation and storage as well as transfusion during murine sepsis, our data indicate that the effects of blood on the recipient depend on the inflammatory status of the individual (⁴). Using the CLP model of polymicrobial sepsis to reflect a human condition known to have increased morbidity and mortality with blood transfusion (¹⁷), we have demonstrated that the response by the recipient’s leukocytes do not reflect those of a healthy mouse that has been transfused PRBCs or a septic mouse that has received crystalloid resuscitation, which is known to improve outcomes in septic mice. Thus, the illustrated dissimilarities in leukocyte numbers, phenotype, and function in transfused septic mice reflect an immune environment that may place the host at an increased risk for poor outcomes.

Previous data from our laboratory and others have demonstrated that “emergency myelopoiesis” and an appropriate innate immune response are required for early survival during sepsis (^{21, 22}). In addition, Hotchkiss et al. (²³) have demonstrated that a loss of lymphocytes and defects in adaptive immunity during the septic response result in increased mortality. Follow-up studies have revealed that functioning activated B cells are vital to the immune system and to survival early in sepsis (²⁴). Our work demonstrates that PRBC transfusion exacerbates the loss of both T- and B-cell populations after CLP, as well as reducing their activation status both in the circulation and in secondary lymphoid tissue.

In addition, these transfusion-associated alterations are not limited to lymphoid populations. Our laboratory has demonstrated the importance of functioning myeloid-derived cells to outcome during both primary and secondary infections, especially in the early septic period (<3 days) (^{19, 21, 22, 25}). Mice transfused in the early phase of sepsis have an absolute loss of almost all their splenic myeloid cell populations. Transfusion of stored allogeneic PRBCs appears to interfere with “emergency myelopoiesis,” as reflected by the relative and absolute deficiency in myeloid precursors present in the bone marrow of septic recipients.

Our ex vivo data demonstrate that the increased sensitivity of splenocytes of healthy transfused mice exposed to LPS concurs with the in vivo data of Hod et al. (⁹). In both cases, transfusion resulted in a significant increase in cytokine/chemokine production after LPS stimulation. The response of septic mice to similar transfusions and stimulation is not as robust.

It is clear that blood transfusion is associated with increased infections, morbidity, and mortality in the surgical and critically ill populations (⁴). A recent meta-analysis that reviewed the efficacy of human blood transfusion in the critically ill revealed that almost all studies (42 of 45) indicate that the risks of PRBC transfusion outweigh the benefits; two of the remaining studies were neutral, and the final remaining study demonstrated a benefit to transfusion in a subgroup of a single retrospective study (elderly patients with an acute myocardial infarction and a hematocrit <30%) (²⁶). The same report also determined that 17 of the 18 studies that appropriately reviewed mortality demonstrated that PRBC transfusion was an independent predictor of death. Finally, all of the 22 studies that analyzed PRBC association with nosocomial infection concluded that blood transfusion was an independent risk factor for infection (²⁶). Recently, Juffermans et al. determined that there was an association between the development of nosocomial bacterial infection and the transfusion of PRBCs in the initial phase of sepsis (²⁷). We wished to use a mouse model that better reflected this human data to determine how transfusion of murine PRBCs affected specific systemic and local leukocytes known to affect outcome and infection during the septic response.

Similar to other animal models, rodent blood banking is not without some difficulty in accurately recapitulating the human process. Previous research has demonstrated that rat PRBCs, as compared with human blood, undergo premature degeneration and in fact may be equivalent to human blood that had been stored for much longer periods (²⁸). However, subsequent studies illustrated that, if processed properly, murine PRBCs can be stored for up to a month and be used for transfusion research (²⁹). Regardless, there are still other confounding factors in regard to murine PRBC storage, including but not limited to the pH, potassium level, and lactate level of the stored blood that would make studying PRBCs stored beyond 14 days impractical (¹⁴). In addition, although using fresh syngeneic blood can be useful as a rodent model for massive transfusion and resuscitation research (³⁰), our initial preliminary research has illustrated differences in the host response when transfused fresh syngeneic PRBCs are compared with stored allogeneic PRBCs (data not shown). In fact, fresh blood can reverse some of the adverse effects of stored PRBCs (³¹). Virtually no blood transfused in the critically ill is autologous, and although debated, much data would indicate that allogeneic blood is more likely to be associated with poor outcomes than autologous blood (³²). In addition, on average, most PRBCs that are transfused in the United States are greater than 14 days old (²⁶), the storage age at which most analyses demonstrate an increased risk for morbidity and mortality after PRBC transfusion (as compared with fresh blood) (^{5, 33}). Other studies verify the immunological differences between fresh and stored murine PRBCs (^{9, 31}). Therefore, to best recapitulate the human condition, we elected to utilize allogeneic murine PRBCs that had been stored for 14 days, which is more similar to human blood that has been stored for 4 weeks (^{14, 29}).

Although we have attempted to reasonably recreate an animal model of blood transfusion, this study does have specific weaknesses. It cannot be excluded that hyperviscosity and initial hypervolemia may have contributed to some of the immune changes displayed. Unfortunately, limitations in murine models do not allow an exact replication of human transfusions in anemic patients. In addition, the type of crystalloid used for resuscitation, specifically LR solution in our experiment, can contribute to immunological alterations (³⁴). Although researches should continually strive to recapitulate the human condition, this model of murine transfusion (i.e. delivering PRBC to a non-anemic mouse) and delivery of LR solution to a mouse as a control comparison has been published in the literature (^{9, 30, 31}).

In conclusion, PRBC transfusion alters the immune response during sepsis. These effects appear to be specific to the mammalian response to infection/inflammation and are not necessarily present in naive recipients of blood. The downstream effects of stored allogeneic PRBCs on immunity are complex and not specific to simple proinflammatory responses or immune suppression, but rather a combination of different alterations of murine leukocyte responses. Increasingly, data demonstrate that the mammalian response to severe trauma or infection places the host in a persistent, inflammatory/immunosuppressed catabolic syndrome that can lead to poor outcomes (^{19, 35}). The displayed systemic effects of blood transfusion on immunity and myelopoiesis in the setting of infection may play a role in exacerbating persistent, inflammatory/immunosuppressed catabolic syndrome, which, in part, might explain some of the worsened outcomes displayed by critically ill transfused patients.

ACKNOWLEDGMENTS

The authors thank LifeSouth Community Blood Centers for supplying/donating the CPD and Adsol solutions for blood preservation.