Red blood cell storage: the story so far (original) (raw)

Introduction

Red blood cells are still the most widely transfused blood component worldwide and their story is intimately entwined with the history of transfusion medicine and the changes in the collection and storage of blood1,2.

At present, the most widely used protocol for the storage of red blood cells (for up to 42 days) is the collection of blood into anticoagulant solutions (typically citrate-dextrose-phosphate); red cell concentrates are prepared by the removal of plasma and, in some cases, also leukoreduction. The product is stored at 4 ± 2° C in a slightly hypertonic additive solution, generally SAGM (sodium, adenine, glucose, mannitol, 376 mOsm/L)1.

Despite this, a definitive protocol that reconciles long-term storage on the one hand and safety and efficacy of the transfusion therapy on the other is still the subject of intense debate and discussion. In fact, although the organisation of the blood system, through the achievement of self-sufficiency, currently enables ordinary requests of the transfusion ‘market’ to be met, in the case of a calamity, disaster, or emerging infections3, or in particular periods of the year, local reserves can sometimes reach a minimum. There is still an underlying concern about the real need to store blood components for as long as possible in order to obtain a gradual increase in the interval between the donation and the transfusion, and how much this elastic time span can be prolonged without definitively compromising the quality of the product and, in the final analysis, the recipients’ health2. Indeed, although the transfusion establishment initially pursued both objectives (product quality and prolongation of the storage period), recent retrospective studies (whose results are, therefore, weakened by all the statistical limitations of this type of analysis)5–8 have indicated the apparent irreconcilability of the two aims. These studies seem to suggest that the quality (in terms of safety and efficiency) of red blood cells decreases in proportion to the time the storage period is prolonged. Furthermore, there is extremely convincing molecular evidence9,10 which, together with the results of clinical studies11–33, appears to confirm the preliminary conclusions regarding the likely poorer quality of red blood cells stored for a long time. However, the statistical validity and methodological rigour, in terms of evidence-based medicine, of the clinical studies have recently been challenged, highlighting the need for prospective, double-blind, randomized studies, in like fashion to the one carried out by Walsh et al.34 in 2004, which led the authors to conclude “the data did not support the hypothesis that transfusing red blood cells stored for a long time has detrimental effects on tissue oxygenation in critically ill, anaemic, euvolumaeic patients without active bleeding”. The international scientific community now seems much more convinced of the need of prospective studies, since such studies, on large cohorts of subjects, are currently underway35,36.

The key point of the problem is probably the lack of universally accepted standard criteria that closely reflect the dramatic molecular changes that occur during prolonged storage of red blood cells and which, simply put, would enable ‘good’ blood to be distinguished from ‘no longer sufficiently good’ blood. The current standard requirements for patenting new additive solutions in the USA, and also suggested in the recommendations of the European Council37, are essentially based on two parameters: the level of haemolysis (below the threshold of 0.8% at the end of the storage period, following the introduction of the “95/95” rule38) and a survival rate of the transfused cells of more than 75% at 24 hours after transfusion. This latter parameter can be assessed by measuring the half-life of red blood cells labelled with 51 chromium prior to transfusion. These parameters are, however, fairly general and easily affected by the considerable biological variability between donors, given that it is known that blood from some donors resists storage better than that from other donors39.

Haemolysis is an easier parameter to monitor. Typically, between 0.2 and 0.4% of red blood cells stored in the presence of standard additive solutions are haemolysed after 5–6 weeks of storage, while pre-storage leukoreduction halves the incidence of this phenomenon40. These widely accepted and well-established parameters do not, however, reflect the profound molecular changes that affect red blood cells during their storage.

A brief list of the elements of the so-called “red blood cell storage lesion” includes10: morphological changes, slowed metabolism with a decrease in the concentration of adenosine triphosphate (ATP), acidosis with a decrease in the concentration of 2,3-diphosphoglycerate (2,3-DPG), loss of function (usually transient) of cation pumps and consequent loss of intracellular potassium and accumulation of sodium within the cytoplasm, oxidative damage with changes to the structure of band 341 and lipid peroxidation, apoptotic changes with racemisation of membrane phospholipids and loss of parts of the membrane through vesiculation9. Some of these changes occur within the first few hours of storage, for example, the decrease in pH or the increases in potassium and lactate; others, however, take days or weeks10. Together, these events risk compromising the safety and efficacy of long-stored red blood cells, reducing their capacity to carry and release oxygen, promoting the release of potentially toxic intermediates (for example, free haemoglobin can act as a source of reactive oxygen species) and negatively influencing physiological rheology (through the increased capacity of the red blood cells to adhere to the endothelium42,43 or through their enhanced thrombogenic44 or pro-inflammatory45 potential). These observations at a molecular level were supported by the results of a series of clinical studies (albeit retrospective and not randomised). These studies appeared to show a relationship between the duration of storage and a proportional increase in adverse events in the transfused patients, although the data available are preliminary and the statistically more reliable studies that conform more closely with the gold standard criteria represented by evidence-based medicine are considered necessary by many4 and are, indeed, underway5.

Clinical evidence of adverse effects following the transfusion of red blood cells stored for prolonged periods

Numerous clinical studies have been carried out throughout the world to identify a possible relationship between the duration of storage of red blood cells, the changes observed at a molecular level and side effects in the transfused patients, in order to determine whether and, if so, to what extent red blood cells stored for a long time lose safety and efficacy11–33. In 2009, Zimrin and Hess2 and Lelubre et al.8 conducted meticolous analyses of the data from the studies published so far. Despite the intrinsic statistical limitations of retrospective, non-randomised studies, the results of such studies are undeniably useful if they are considered as a warning bell, albeit debatable, but not to be ignored, of a potential increase in the negative effects of the transfusion of red blood cells in proportion to the duration of storage of the blood product in certain groups of patients such as those in intensive care11–16, those undergoing cardiac interventions17–24, those submitted to colorectal surgery25–27, or those with multiple trauma28–33.

The side effects described in these groups of patients following multiple transfusions of ‘old’ red cells are very varied, ranging from a decrease in gastric pH11 to an increase in mortality rate12, from multi-organ failure28 to an increased incidence of pneumonia in patients transfused following aorto-coronary artery bypass18,19,21, from an increased susceptibility to infections29 to major complications following heart surgery14,22,23, and from an increase in the duration of hospital admissions30,31 to the development of complications such as trasfusion-related acute lung injury (TRALI)46,47.

It is, however, worth stating that given the current lack of irrefutable statistical proof, it cannot yet be concluded “there’s no smoke without fire”, to mention Steiner and Stowell6. In fact, it worth remembering that a few years ago a series of retrospective, non-randomised clinical studies suggested a correlation between reduced efficacy of transfusions and lack of leukoreduction; the subsequent prospective, randomised studies did not, however, fully confirm these observations 48,49. The storage of red blood cells, however, has not (up to now35,36) been the focus of prospective, randomised studies similar to those needed to market a new drug50. For this reason, although it has now been ascertained and widely accepted that something more or less irreparable occurs during prolonged storage of red blood cells, it is currently impossible to conclude objectively and without preconceptions that these changes are accompanied by decreased efficacy and safety of the blood component.

Storage lesions

While from a clinical point of view there is only preliminary evidence, still to be confirmed, from the molecular point of view, the observations of changes that accumulate in red cells in proportion to the duration of their storage are numerous and indisputable, as described here. Although the average half-life of red blood cells in the circulation is 120 ± 4 days51, the standard maximum duration of storage of RCC is 42 days. This is because transfused red cells seem to have a notably shorter half-life. In fact, 25% of the cell components are removed from the recipient’s circulation within 24 hours of transfusion; in other words, of four units of red cells transfused, one is completely eliminated by the body already the day after the transfusion. There are probably two causes for this. The first, which is easily deducible, is that at the time of being donated, the unit of blood contains a percentage of already aged red blood cells which, during storage, do nothing other than complete their aging process and are too old by the time of transfusion. The second cause depends on the storage conditions, which are far from being normal, physiological conditions and which represent a greater and more long-lasting stress than the red blood cells are able to counteract, despite their well-supplied protein machinery ab origine. In fact, although red blood cells are anucleated and, therefore, lack a real genome and consequently protein synthesis, they do have their own armamentarium of proteins devoted to protecting and maintaining pre-existing protein functions through a “central core” of chaperone proteins, heat shock proteins and proteins involved in the detoxification of free radicals (peroxiredoxins, catalases, glutathione peroxidases) whose role is critical in the economy of the red blood cell proteome (the protein complement of the genome) and interactome (the system of protein-protein interactions)52.

The most evident changes affecting red blood cells during the storage period are alterations of the cell phenotype, which varies from a smooth discoid shape to a phenotype characterised by various membrane protrusions or spicula (echinocyte) and finally to a spheroid-shaped cell (spheroechinocyte) 53. The reversibility of these changes is inversely proportional to the duration of storage.

The storage lesion also involves the fluxes of sodium ions (massive entry into the cell) and potassium ions (exit from the cell), since the Na+/K+ pump is inactive at 4°C 10. Although this is a reversible process (it takes 24 hours to restore the physiological gradients for sodium, and up to 4 days for potassium54), this phenomenon means that blood stored for a prolonged period should not be used for neonates or paediatric patients, unless first washed or the potassium removed from the storage medium55.

Another biochemical effect is a clear decrease in the levels of 2,3-DPG (which is consumed already within the first week), translating into increased affinity of haemoglobin for oxygen and, consequently, decreased capacity of the red blood cells to release oxygen according to local metabolic needs. The decrease in 2,3-DPG levels is also a reversible event, and completely normal levels can be restored within 3 days after the transfusion56.

The experimental evidence on the role of S-nitrosothiol-haemoglobin is, on the other hand, controversial. It was thought that reduced levels of this form of haemoglobin would be related to ‘old’ blood having a lesser vasodilatory effect in recipients57,58; however, recent molecular biology studies seem to suggest that this is not the case59.

Alongside these reversible changes, various irreversible events occur during the storage process, including fragmentation and aggregation of proteins and lipids, activated by radical species generated by prolonged, continuous oxidative stress60–62. In this way oxygen constantly leaves one molecule of haemoglobin to bind to another. It is known that, occasionally, an oxygen leaving the haemoglobin molecule carries with it an electron, forming a superoxide ion (O2−) and (ferric) methaemoglobin. Normally, the methaemoglobin is reduced by cytochrome b5 reductase63 and the superoxide is dismutated without consequences. However, during prolonged storage, the superoxide ion can interact with iron and water in a Fenton reaction, resulting in the formation of hydroxyl radicals capable of attacking and damaging both proteins and lipids, leading to their fragmentation and the formation of aggregates. For example, haemoglobin can be converted into hemichromes (haemoglobin whose cysteine residues have been oxidised, leading to the formation of aggregates). Eligible targets of the radical species generated in a cascade from the hydroxyl radical are membrane phospholipids (with the formation of lysophospholipids and malondialdehyde64), and proteins within (or closely related to) the cell membrane, such as the band 3 ion exchanger41 (which plays a fundamental role in maintaining the oxygen transport function of red blood cells65 and acts as an anchor for a series of key glycolytic enzymes66,67) and spectrin. These membrane alterations end up causing the previously-described echinocyte or spheroechinocyte phenotypes. Finally, it is known that the cell activates a process of vesiculation, in order to eliminate proteins and lipids that have been altered by oxidative stress, as to protect the cell from a further chain reaction of stress and consequent removal from the circulation68. In fact, aggregates of band 3 appear at the membrane during both in vivo and in vitro aging68, constituting membrane signals to “remove” the cell, through IgG- or complement-mediated phagocytosis by the recipients’ Kuppfer cells. These membrane neoantigens, by stimulating the immune system, seem to be related to the onset of pro-inflammatory events, which are often harmful if not fatal in critically ill patients undergoing transfusion therapy5,8. Alongside these signals, which are particular to red cell aging, a series of other markers appear; these markers are common in other physiological phenomena associated with programmed cell death or apoptosis, such as exposure of phosphatidylserine on the external leaflet of the lipid bilayer of the cell membranes, whose presence in microvesicles increases in proportion to the duration of storage69. This same phenomenon of vesiculation through membrane protrusions (blebs) has contributed to strengthening the parallels between the processes of red blood cell aging and apoptosis69, leading Lang and colleagues70 to coin the term “eryptosis” to describe this physiological phenomenon, which is exacerbated during the storage of the red cells. The increase in the number of vesicles (0.5 μm) with the duration of storage is noteworthy, as is the increased content of proteins (band 3 and ankyrin) and lipids (stomatin), again proportional to the duration of storage. In contrast, the variability in proteins in the red cell cytoplasm and membrane decreases gradually71–73.

Most of these irreversible events seem to be favoured by prolonged oxidative stress in the non-physiological conditions of storage60–62. However, as far as proteins are concerned, the first signs of fragmentation and aggregation begin to appear in the third week of storage2, which is a remarkable temporal synchronisation with the findings, albeit in non-randomised, retrospective clinical studies, of adverse effects of red blood cells stored for more than 2 weeks in patients undergoing heart surgery23, and with the onset of conformational changes of band 341. It does, therefore, seem wise to prevent this type of irreversible lesion in the early period of storage, rather then intervening a posteriori (for example, through the addition of rejuvenating solutions74).

Anaerobic storage of red blood cells

Molecular and retrospective clinical studies have recently compelled the international scientific community to reconsider the validity of the current protocols for red blood cells storage. Randomised, prospective studies of unquestionable statistical rigour, are, however, yet to be completed35,36. If, however, future data confirm the numerous already available retrospective clinical observations, research and new storage strategies will need to be focused on avoiding the possible side effects of prolonged storage. Furthermore, ad hoc studies will have to be carried out on the impact that a change in the duration of storage of red cells could have on the self-sufficiency of the blood system.

From a molecular point of view, most of the changes occurring during storage are already well-known. Some of the changes are reversible, through the addition of new additive solutions1 or rejuvenating solutions74, while others are irreversible and must, therefore, simply be prevented. The former changes include alterations in the levels of small molecules, such as ATP and 2,3-DPG, pH, sodium and potassium, while the latter changes include irreparable denaturation of proteins following fragmentation and aggregation catalysed by free radicals. The underlying cause of these phenomena is the prolonged oxidative stress to which the red cells are exposed during storage60–62. To contrast this oxidative stress, it was proposed, in the past, to treat donors with antioxidants (vitamins E and C, beta-carotene62), although this type of treatment tends to limit oxidative stress rather than prevent it. Yoshida’s group75–77, however, suggested a storage protocol that tackled the problem at its source. They proposed storing blood directly in an atmosphere of inert gas at a pO2<4%, using a method that they patented (WO/1996/039026). The clinical outcome of this protocol has been tested with respect to the classical standards (haemolysis and red blood cell survival at 24 hours post-transfusion) with positive results; slowing in the decreases of 2,3-DPG and ATP was also observed. Furthermore, Yoshida’s group showed that addition of a standard rejuvenation solution at day 63 of storage restored the levels of 2,3-DPG and ATP, making storage for as long as 120 days theoretically possible76. In support of this strategy, but independently, Zolla’s group72 also analysed a model of anaerobic storage, using classical proteomic methods to compare the total protein profile of the red blood cells stored using this protocol with that of control units stored according to a standard protocol. No signs of fragmentation or aggregation were found in the blood stored in inert gas in the medium term (during the first 2 weeks); these phenomena began to be seen, albeit to a reduced extent, towards the end of the storage period (42 days), thus, from a molecular point of view, the blood provided to recipients was of better quality. It should, however, be appreciated that these results are drawn from preliminary studies and further investigations, both clinical and molecular, are needed and are currently in fieri.

In conclusion, given that anaerobic storage can prevent the above-described irreversible phenomena of fragmentation and/or aggregation72, as well as slower the decreases in 2,3-DPG and ATP levels75 (although these latter processes can be reversed in any case by the addition of rejuvenating solutions74), such a form of storage could be an excellent solution to the clinical problems observed in the preliminary retrospective studies on the efficacy and safety of the current protocols of blood storage, while awaiting the definitive clinical data, which will be provided by prospective, methodologically incontrovertible studies. The task remaining is to bridge the gap between basic research and large scale application of its results, a goal that current translational research must meet in order that anaerobic storage can be adopted in daily transfusion practice.

References