Red blood cell storage and transfusion-related immunomodulation

Introduction

Red blood cell (RBC) transfusion is a life-saving treatment for patients suffering severe blood loss or anaemia due to trauma injury, surgery, haemorrhage, haematological disease or malignancy. RBCs for transfusion are stored refrigerated in a preservative solution, which extends their shelf-life. Most of the preservative solutions in common use, such as saline-adenine-glucose-mannitol (SAG-M), enable refrigerated storage of RBCs for up to 42 days following collection. This expiry is based on criteria set by the United States of America Food and Drug Administration, which requires that 75 percent of transfused RBCs must be recoverable in the peripheral blood circulation 24 h after transfusion¹.

During refrigerated storage of RBC units, the RBCs undergo numerous physicochemical changes, collectively referred to as the RBC storage lesion, which affects the quality, function and in vivo survival of the transfused RBCs^2–4. The implications for the transfusion recipient of these storage-related changes to RBCs are currently a matter of considerable interest and debate in the clinical community.

In addition to their primary function to transport oxygen from the lungs to the tissues, RBCs are important regulatory components of haemorheology, the dynamics of blood flow^5,6. In doing so, RBCs interact with the other blood elements, including white blood cells (WBCs), platelets and vascular cells. Many of the physical changes that occur to stored RBCs appear to be similar to those that occur to diseased RBCs (such as in malaria, sickle cell disease, thalassemia), in which disturbances of vascular function are key morbidities^7,8. These changes include altered membrane surface receptors and cytoskeletal structures, which control RBC shape, flexibility (deformability) and aggregability. Knowledge gained from the study of diseased RBCs may provide insight into understanding the effect of storage on normal healthy RBCs and how these changes could influence the interaction of transfused RBCs with the recipient’s own cells and tissues.

Transfusion-related immunomodulation (TRIM) has emerged as a concept to potentially explain numerous clinical observations that suggest that RBC transfusion is associated with increased proinflammatory or immunosuppressive effects that may increase morbidity in at least some patient groups^9,10. The predominant mechanism of TRIM is likely to depend on an interplay of transfusion effects with the genetic predisposition and the intercurrent illnesses in the patient. Platelets and vascular endothelial cells also potentially contribute to the “response” as both cell types are highly responsive to inflammatory signals and when activated, release significant quantities of potent bioactive mediators. Thus, in situations of heightened inflammation or breach of vascular integrity, the immune and thrombotic systems are likely to be intricately linked in a complex network of signalling and response. This article aims to provide a perspective of the potential relationship between the RBC storage lesion and the concept of TRIM in its broader sense along with a brief overview of some of the research findings that could support this perspective. The role of proteomics in advancing our understanding of the RBC storage lesion as well as to provide insight into the biological mechanisms of TRIM is also discussed.

Clinical studies and the consequences of RBC transfusion

The role of RBC transfusion in poorer outcomes for transfusion recipients is currently a topic of active debate and controversy in the clinical community. Some clinical studies have identified older transfused RBCs as an independent risk factor for poorer outcomes in certain patient groups, such as cardiac surgery patients¹¹ and trauma patients^12,13, whilst no association was found in other studies^14,15. A number of reports were retrospective, observational studies, and suffer from various limitations that may have influenced the statistical findings. Specifically designed, prospective, randomised controlled clinical studies are currently underway to address the concerns about the role of RBC transfusion and the age of RBCs in poor outcomes of patients and the results are awaited with much interest.

RBC storage and the storage lesion

During refrigerated storage of RBC units, RBCs undergo progressive biochemical and morphological changes, referred to as the RBC storage lesion^2,4. Many of the changes are the consequence of oxidative stress, leading to the generation of reactive oxygen species, altered proteins and lipids, loss of membrane and cell constituents in the form of shed microparticles, changes to the RBC cytoskeleton resulting in RBC shape change and increased cell rigidity (Table I)¹⁶. These changes share some features of normal RBC aging and/or apoptosis¹⁷, as well as changes seen in certain diseases that affect RBCs, including thalassemia, sickle cell anaemia and malaria^7,8. However, a notable difference is that, unlike in vivo circulating RBCs, stored RBCs are also exposed to the cell debris that accumulates in the suspension fluid (supernatant) during storage, which may contain reactive constituents, such as denatured, aggregated or oxidised proteins and lipids (Table I)¹⁸.

Table I.

RBC storage lesion: biochemical and morphological changes

Progressive RBC changes	Accumulation in the supernatant
Acidosis ↓ intracellular pH	Acidotic ↓ pH, ↑ lactate
Slowed metabolism ↓ ATP	↑ Extracellular K+
↓ 2,3-DPG, ↓ O2 off-loading	↑ Haemolysis (plasma Hb)
Shape change, cytoskeletal damage, band 3 denaturation	↑ Damaged, oxidized proteins & lipids
Loss of cation pumping	↑ RBC microparticles
Oxidative damage; ↑ lipid peroxidation	↑ Cell debris, bioactive factors
Loss of cell membrane (shedding of microparticles)
→ Cell shrinkage, ↓ Deformability
Loss of membrane phospholipid asymmetry
RBC lysis

The widespread introduction of pre-storage filtration of RBC units to reduce the number of contaminating WBCs has provided some improvement of the quality of RBC units¹⁹, and has lowered the incidence of alloimmunisation and non-haemolytic allergic transfusion reactions potentially caused by WBC-derived bioactive factors²⁰, but has not eliminated the biochemical and morphological changes that occur to RBCs as a consequence of aging and storage.

Proteomic approaches have started to be used to elucidate the changes that occur to RBCs during storage²¹. Oxidative damage to RBCs during storage has been shown to be a significant factor^18,22. Other proteomic studies have identified the accumulation of altered proteins in the supernatant of stored RBC units²³, or in the microparticles that are shed from RBCs during storage^18,24,25. The packaging of damaged or altered proteins, such as band 3 and haemoglobin, as well as procoagulant phosphatidylserine into the microparticles shed by RBCs during storage presents an opportunity for insightful proteomic investigation into the mechanisms of the RBC storage lesion.

Other studies have shown changes to cell surface molecules of RBCs during storage, including cell adhesion receptors such as CD47^26,27, carbohydrate receptors²⁸ and complement regulatory molecules (unpublished observations). Such cell surface molecules are known to be involved in cell-cell interactions or to protect the cell from clearance by phagocytic cells. How changes to the cell surface receptors on RBCs during storage affect the behaviour of the RBCs when transfused is an important area of further investigation.

RBCs and TRIM effects

The concept of TRIM, in particular how transfused stored RBCs interact with the recipient’s own cells and tissues and whether such interactions elicit responses in the recipient that contribute to increased morbidity is an area that requires more investigation. In vitro and in vivo models of transfusion provide a way forward, although are inevitably challenged by the complexity of the systemic biology involved. In vitro models over-simplify the biology, whilst animal-based in vivo models are difficult to correlate with human biology and clinical complexities that co-exist with the need for transfusion.

The “two-insult” model of post-transfusion injury proposes that the first insult (i.e. the patient’s underlying inflammatory condition) primes the patient’s immune cells or endothelium, and frank inflammation is triggered by a second inflammatory insult, resulting in full-scale activation^29,30. Transfusion has been proposed as a potential second insult. This model provides a feasible basis to begin to understand the biological mechanisms at play in various clinical settings in which transfusion has been implicated as a risk factor for poor outcome, including transfusion-related lung injury (TRALI). In the broadest sense, such a model could explain the dynamics between proinflammatory versus immunosuppressive responses and the role of the coagulopathy/thrombosis and vascular activation in TRIM. Developing appropriate working models to test this is the challenge.

Responses by allogeneic WBCs

A few groups have investigated the response of allogeneic WBCs to stored RBCs. Different models and immune response read-outs have been used, which makes comparison of the results difficult. Some studies have used whole blood assays, whilst others have used isolated WBC populations (i.e. neutrophils, mononuclear cells or T lymphocytes). Prestorage leucocyte reduction of RBC units appears to mitigate some WBC responses, but not others. For example, we and others have shown that normal allogeneic mononuclear cells can be induced to release cytokines by supernatant from leucocyte-reduced RBC units^31,32, whilst cytokine release by allogeneic neutrophils appears to be mitigated³³. The predominance of a proinflammatory versus an immunosuppressive cytokine response may also be influenced by whether or not the RBC unit was prestorage leucocyte-reduced. Supernatant from leucocyte-reduced RBC units has recently been shown to induce regulatory T cells, and this effect was not related to storage duration of the RBC unit³⁴. Together these results suggest a complex and dynamic interplay of effects. All of these in vitro experiments have been performed using allogeneic “responder” WBCs from normal healthy donors. The response of WBCs from transfusion recipients with co-existing morbidities that modulate their immune and/or coagulation or vascular systems is not known and adds further dimensions to the complexity of the potential effects of TRIM.

Adhesion to endothelial cells

In healthy individuals, RBCs do not appreciably adhere to the vascular endothelium, thus maintaining smooth blood flow. Using an in vitro continuous flow perfusion model to simulate blood flow, we and others have demonstrated adhesion of stored RBCs to vascular endothelial cells and that the number of adhered RBCs increases with prolonged storage^35–37. Prestorage leucocyte-reduction reduced the number of adherent RBCs, but did not eliminate the effect, suggesting that storage-related changes to the RBCs are implicated in the mechanism of adhesion. Pretreatment of the endothelial cells with endotoxin to mimic infection resulted in increased strength of adhesion of RBCs to the endothelial cells³⁸. Increased adhesion of RBCs to vascular endothelium may affect blood flow and oxygen delivery in certain patients, particularly those with microvascular dysfunction and diseases that alter RBC physicochemical properties, such as sickle cell anaemia and thalassemia^7,8. Further studies are required to better understand the mechanisms of adhesion of stored RBCs to endothelial cells and whether these in vitro findings correlate with the behaviour of transfused, stored RBCs in vivo. Models to investigate the interaction of WBCs and platelets with stored RBCs under flow conditions are also needed to explore the potential dynamics of TRIM.

Key questions and role of proteomics

Although a great deal of knowledge has been generated over decades of research into the effects of storage on RBCs, there is a significant absence of understanding about certain key questions including 1) does the storage age of transfused RBCs matter in critically ill patients; 2) at what point are the changes that occur to RBCs during storage irreversible; 3) how does the transfusion recipient’s own cells and tissues respond to the transfusion of storage-affected RBCs; 4) does a prior insult to the patient, as proposed by the two-insult model, predispose the patient to TRIM or other adverse consequences and 5) can improved storage conditions for RBC units mitigate these effects?

Newer generation experimental preservative solutions that provide buffering capacity and maintain an increased pH appear to delay some of the significant storage-related changes that occur to RBCs and therefore may provide an opportunity for improved quality of RBC units³⁹.

Proteomics, in conjunction with cell biology, offers a powerful tool to better understand the biological mechanisms and consequences of the RBC storage lesion. With the application of these advanced tools, new light may be shed to address some of the important outstanding questions in transfusion medicine.