Leucocyte Depletion Of The Blood Supply – How Will Patients Benefit?

Following a decision by the Department of Health, all blood components manufactured by the four UK transfusion services have been subjected to a leucocyte-depletion (LD) step since November 1999. This was introduced as a precautionary step against the theoretical possibility of transmission of abnormal prion protein (PrPsc), the probable causative agent of new-variant Creutzfeldt–Jakob disease (nvCJD). In nvCJD patients, PrPsc has been found in lymphoid tissue, and in experimental transmissible spongiform encephalopathies (TSE) infectivity has been associated with the buffy coat. LD is performed on blood components within 2 d of collection by filtration or by selected apheresis techniques; the objective is to attain a residual leucocyte count of < 5 × 106 leucocytes/component in 99% of components produced, with > 95% statistical confidence. Studies are ongoing to define the profile of LD components with regard to leucocyte subsets, cellular fragments and residual normal prion protein (PrPc). Other potential benefits to patients include removal of leucocyte-associated viruses and bacteria and abrogation of certain side-effects, such as febrile reactions, human leucocyte antigen (HLA) alloimmunization and immunomodulation. For cytomegalovirus (CMV), several small-scale studies in which prestorage LD has been used have totally prevented CMV transmission; this is not the case for bedside filtration. Many sources, including the British Committee for Standards in Haematology, now regard rigorously performed LD as being equivalent to CMV testing for provision of CMV-safe components. However, UK transfusion services will continue to provide CMV seronegative components as required. LD is likely to provide a reduction in the risk of human T-cell leukaemia/lymphoma virus (HTLV) transmission, but this cannot be directly compared to the benefits of HTLV testing. The risk of bacterial transmission may be reduced, provided filtration is carried out after a 2- to 8-h period of leucocyte/bacterial contact to allow phagocytosis. Many studies have suggested that leucocyte-derived cytokines are a major cause of non-haemolytic febrile reactions, particularly to platelets, but no formal studies have been carried out to examine the febrile transfusion reaction (FTR) rate in the context of universal LD. LD undoubtedly reduces the rate of HLA alloimmunization in multitransfused patients with haematological disorders. The clinical benefit of this is established for patients with severe aplastic anaemia, with improved allograft outcome, but for patients with acute leukaemia the clinical advantages are less certain. Transfusion-induced immunomodulation is an area of considerable research activity. There is no convincing evidence that cancer recurrence is adversely affected by transfusion at the time of primary tumour removal, but there is accumulating evidence for an association between transfusion and post-operative infections in patients undergoing elective surgery. It is less clear whether LD can totally or even partially abrogate this effect, but the question is of huge health economic importance. Post-operative infection is a strong predictor of length of hospital stay and hence overall surgical costs. The £40–50 million/year cost of universal LD in the UK would be entirely offset by only a 1–2% reduction in the post-operative infection rate, and a large cohort study is ongoing in an attempt to answer the above question. In the spring of 1998, the British Committee for Standards in Haematology published guidelines which outlined the patient groups likely to benefit from leucocyte-depleted (LD) blood components ( Murphy et al, 1998a ). These were estimated to require no more than 5–10% of red cells and 50% of platelets to be manufactured as LD. Four months later, acting on advice from the Spongiform Encephalopathy Advisory Committee, the Department of Health announced that UK transfusion services would be required to leucocyte deplete all blood components by 1 November 1999 because of a theoretical risk of blood-borne transmission of the causative agent of new-variant Creutzfeldt–Jakob disease (nvCJD) ( Department of Health, 1998). The same policy will also be followed in Ireland. Universal LD has also been introduced in several other countries (Portugal, France and Canada), and others are giving the issue serious consideration (including the USA, Switzerland, Germany and The Netherlands). Although cases of nvCJD are so far restricted to the UK (with the exception of one case in France), the presence of bovine spongiform encephalopathy (BSE) in a number of European countries is undoubtedly a factor in this trend. However, in North America, where BSE/nvCJD is not such a pressing issue, removal of the possible adverse immunomodulatory effects of transfusion and a reduction in transmission of cell-associated viruses have been major considerations in the debate on whether to undertake this costly health care intervention. This article will review the rationale for the decision in the UK to proceed with universal LD of the blood supply to reduce or prevent the theoretical possibility of transfusion-associated nvCJD, and will discuss the possible impact of this policy on other adverse and beneficial effects of transfusion. In 1996, the first description appeared of a new rapidly progressive spongiform encephalopathy apparently unique to the UK, although one case has subsequently been reported in France. The disease bore some similarities to classic CJD, but affected much younger adults, had predominant behavioural features as well as neurological deficit and showed brain pathology strikingly like kuru ( Will et al, 1996 ). The disease was given the title 'new-variant CJD' (nvCJD) and its incidence, along with that of sporadic CJD, was monitored by the UK CJD Surveillance Centre. New-variant CJD is the latest in a large group of animal and human prion diseases with common pathology and causation. In each, there is conversion of normal cellular prion protein (PrPc) to an insoluble form (PrPsc). Normal PrPc is widely distributed, although its function is unclear (for more detail, see Turner & Ironside, 1998). Accumulation of amyloid-like PrPsc in neurological tissue leads to the spongiform appearance of the brain and the rapidly progressive neurological features, although the exact mechanism is unknown. Prion diseases are transmissible within and, in certain cases, between species. The PrPsc itself is believed to be the infectious agent, although the issue of whether infectivity also includes a small amount of nucleic acid continues to be debated ( Finkel, 1996). Because of the temporal association with the epidemic of bovine spongiform encephalopathy (BSE) estimated to have affected 750 000 British cattle ( Anderson et al, 1996 ), urgent studies were undertaken to establish whether BSE had 'jumped species' into the human population through consumption of BSE-infected beef and other cattle derivatives. Different prion strains can be identified by their biochemical characteristics, incubation period and brain distribution of pathological lesions when injected into susceptible experimental animals. Two studies using these techniques were published in the autumn of 1997, suggesting that BSE and nvCJD were indeed caused by the same agent, which was different from that causing sporadic CJD ( Bruce et al, 1997; Hill et al, 1997 ). Because of steps which have been taken to eliminate BSE from the food chain, it is presumed that the nvCJD outbreak from bovine sources will be self-limiting. However, the final number of nvCJD cases cannot be predicted with any certainty, and may range from 75 cases to 80 000 cases ( Cousens et al, 1997 ). Equally, the time-scale over which new clinically apparent cases will present is uncertain as the incubation period of prion diseases is measured in decades and is not fully determined yet for nvCJD. A further complication is that susceptibility to sporadic, iatrogenic and nvCJD is determined by a common polymorphism in the normal prion protein. All cases of nvCJD reported so far are homozygous for methionine at codon 129, a phenotype found in 37% of the normal population ( Zeidler et al, 1997 ). It is as yet unclear whether heterozygotes and valine homozygotes will be totally resistant to the disease or whether they will present after a longer incubation period. This issue has recently been reviewed in detail ( Turner, 1999; Williamson, 1999). Prion diseases can be spread parenterally from neural tissues, as demonstrated by cases of classic CJD in recipients of pituitary-derived human growth hormone ( Buchanan et al, 1991 ). Many epidemiological studies of different design have shown convincingly that there is no increased risk of classic CJD associated with previous transfusion of either blood components or pooled plasma products (reviewed in Turner & Ironside, 1998). However, total reassurance regarding the safety of transfusion with regard to nvCJD would be premature. A blood-borne phase is certainly a possibility not excluded by current knowledge. Transmission by blood transfusion is therefore theoretically possible. The following relevant pieces of evidence were available to the Spongiform Encephalopathy Advisory Committee during 1997 and 1998, when this issue was being considered. Normal PrPc is widely distributed in human blood, including T and B lymphocytes, monocytes and platelets, from which it is released upon activation ( Perini et al, 1996a , b; Dodelet & Cashman, 1998). At various stages of the infection, nvCJD appears to have a propensity for lymphoid tissue not seen in classic CJD nor, interestingly, in BSE. The abnormal conformer, PrPsc, has been demonstrated in the lymphoid follicles of tonsils in all nvCJD patients examined ( Hill et al, 1999 ) and in the appendix removed from a nvCJD patient some months before the onset of symptoms ( Hilton et al, 1998 ). Buffy coat (containing leucocytes and platelets) from the blood of individuals with classic CJD could transfer the infection by direct intracerebral inoculation to mice ( Manuelidis et al, 1985 ; Tateishi, 1985), but not to primates ( Brown et al, 1994 ). Similar findings were obtained using blood components prepared from mice infected with human Gerstmann–Straussler–Scheinker syndrome (GSS) and hamster-passaged scrapie ( Brown et al, 1998 ). Infectivity was concentrated in the buffy coat and was also found in plasma. The relevance of these findings to human intravenous transmission is unclear. Transfer of prion from the periphery to the brain in susceptible mice appeared to depend on the presence of circulating B lymphocytes, which could therefore be a source of infectivity ( Klein et al, 1997 ). It should be noted that this study examined transfer within an animal but not between animals, and its relevance to the transfusion setting is unclear. However, taken with the observations on buffy coat and the lymphoid distribution of PrPsc in nvCJD, the data raised the possibility that circulating B cells had the potential to provide a means of blood-borne spread of the disease. Because of the young age of the nvCJD cases, there was also the very real possibility that many thousands of healthy blood donors might be incubating the disease and be capable of passing it on through transfusion. Indeed, of the first 25 nvCJD cases who presented during 1997 and 1998, at least three had donated blood on at least one occasion. This raised the alarming possibility that although BSE-infected beef as a source of nvCJD might disappear transfusion could perpetuate this disease into the future. This could continue virtually indefinitely as up to 10% of blood donors have themselves been transfusion recipients. An independent risk assessment conducted on behalf of the Department of Health created mathematical models of risk based on assumptions regarding infectivity, based on published data from animal studies. This study concluded that leucocyte depletion to < 5 × 106 leucocytes/unit had the potential to reduce the number of transfusion-associated nvCJD cases by up to 36% ( Det Norske Veritas Report, 1999). Following this, in July 1998, the Department of Health announced that all blood components would, 'as a precautionary measure', be manufactured as LD by 1 November 1999 ( Department of Health, 1998). It must be emphasized that the nvCJD risk from blood components manufactured from individuals in the preclinical phase of the condition remains unknown. Equally, the level and distribution of infectivity in unmodified blood have not been quantified, so the benefit of leucocyte depletion remains uncertain. Inevitably, further evidence has since become available which demonstrates some of the difficulties in generating experimental models of prion diseases relevant to nvCJD. For example, the role of the B cell in prion transfer from periphery to brain does not appear to involve direct carriage of PrPc on its surface ( Klein et al, 1998 ), throwing into question the relevance of B-cell removal from blood transfusions. Klein et al (1998) demonstrated a role for follicular dendritic cells, which do not circulate peripherally. In addition, new studies to determine the distribution of PrPc in blood have shown plasma and platelets to be the major source (68·5% and 26·5% respectively), with only 3·2% on leucocytes and 1·8% on red cells ( MacGregor et al, 1999 ). These findings are confirmed by flow cytometric examination of the cellular compartment, in which expression on platelets accounts for 96% of cellular PrPc, with a three- to fourfold increase in the number of surface molecules when platelets are activated ( Barclay et al, 1999 ). This study also found that leucocyte-associated PrPc accounted for only 3·7% of the cellular portion. Both methodologies suggest a 3:1 distribution of this small portion between mononuclear cells and granulocytes. The distribution in blood of PrPsc in individuals with nvCJD remains unknown. However, if the distribution of PrPsc is the same as that of normal PrPc, then universal LD, combined with suspension of red cells in additive solution, will minimize the amount of PrP to which transfusion recipients are exposed. As the degree of risk reduction of nvCJD transmission potentially achieved by different levels of leucocyte removal is not known with any certainty, the specification for universal LD in the UK is based on the current specification of < 5 × 106/unit of red cells or the adult therapeutic dose of platelets. This is consistent with studies in the literature on other benefits of leucocyte depletion which are also based on this specification. As it is not possible to perform individual leucocyte counting on the 3 million blood components/year produced in the UK, the specification was modified to indicate the percentage of components in which this degree of leucocyte removal could be achieved and the statistical confidence with which this could be guaranteed. The final agreed specification is shown in Table I, along with corresponding specifications from France and the Council of Europe ( Council of Europe, 1999). The 5 × 106/unit level was established some years ago, based on observations from experimental and human studies on HLA alloimmunization ( Claas et al, 1981; Fisher et al, 1985 ). As discussed in detail below, subsequent clinical studies in alloimmunization and other transfusion complications suggest that it has limited biological significance. However, it remains a useful pragmatic specification, achievable in virtually every unit of blood processed, and is within the linear range of the counting techniques used to monitor the process. For whole blood donations, LD is achieved by passage of either the whole donation or processed components through third generation filters capable of 3–4 log leucocyte reduction. Whole blood filtration is convenient, but cannot be applied to donations from which platelets are to be manufactured as filters also remove at least 2 logs of platelets. Whole blood or processed red cells are filtered either at ambient temperature on the day of collection (designated day 0) or after an overnight hold at 4°C, whereas platelets are filtered during manufacture on day 1. No unit is filtered after day 2 because leucocytes gradually fragment during storage. Newer apheresis equipment for platelet collection performs LD without the use of a filter, using either particle bed separation or elutriation. Following on from earlier observations ( Bodensteiner, 1994), we have systematically examined the behaviour of blood from haemoglobin S heterozygote donors during filtration. Preliminary data suggest that 50% of units either cause filter blockage or LD is unsuccessful ( Williamson et al, 1999a ) . This does not appear to be simply due to sickling in the filter, and its precise cause remains unresolved. The system to be used for quality monitoring of the LD process is based broadly on guidelines published by the Biomedical Excellence for Safer Transfusion (BEST) group of the International Society for Blood Transfusion ( Dumont et al, 1996 ). Evaluation and acceptance of a filtration system is followed by local validation, with routine filtration then performed under closely defined optimal conditions. The mean leucocyte values achievable with current LD technology are at least 1 log lower than the 5 × 106/unit cut-off, so routine quality control can be carried out using statistical process monitoring (SPM). Data from a random sample of 1–5% of units processed are plotted on SPM charts so that any upwards drift in values is detected well before the 5 × 106/unit value is approached. The leucocyte counts in LD components are well below the sensitivity of standard haematology analysers. Techniques suitable for use in large-scale leucocyte counting are flow cytometry and static laser fluorimetry using the IMAGN 2000 machine ( Adams et al, 1997 ). Both techniques depend on detection of a nucleic acid-bound fluorescent dye in leucocytes and are linear in the range of interest. A national internal quality assurance scheme for leucocyte counting has been established, as well as a new external scheme within the UK National external Quality Assessment Scheme (NEQAS) structure, using a novel method for stabilizing leucocytes ( Barnett et al, 1998 ). There is not yet a national or international standard for WBC levels which could be used as a reference for different counting The possibility of this is under has been regarding the possibility that filtration could to be or even increase nvCJD if the process caused generation of leucocyte A study that this does not to any degree ( et al, 1999 ). In of the platelet association of PrPc, it may be relevant to platelet removal by filters and removal or generation of platelet during filters for whole blood or red cells remove 2 logs of platelets and have been shown not to platelets to any degree ( et al, 1997 ) . are present in small in the and have been demonstrated in blood components manufactured by standard means ( et al, 1997 ). A large study to PrPc, platelet and red cell and leucocyte in blood components that have been leucocyte by all techniques in use in the UK has been by the Department of Health, and will the role of different leucocyte in infectious and it would be to be to perform a leucocyte on LD components. for using and flow cytometry have been ( et al, 1998 ), but may through leucocyte before An to whole blood is of This linear over a log range ( & Williamson, and may useful in of different The of any filtration process is the of up to 10% of the in the The on red cell components can be by whole blood, so that the is between plasma and red However, for donations from which platelets are to be produced, red cells are during buffy coat removal and filtration. red cells platelets appear to be by passage over filters ( et al, ) . cell may even be by removal of ( et al, 1997 ). LD plasma may be from whole blood filtration or by filtration of plasma after A of four whole blood filters and one plasma filter of with < removal of factor ( et al, ). All filters the level of the this is clinically useful is unknown. There have been in the literature the complications of LD components. have appeared of in patients platelets or red cells through bedside which a surface ( et al, 1996 ). of appeared particularly The likely is generation of on the filter surface during to the with the which by the ( et al, 1997 ). The and have than filtration to minimize this complication because in normal plasma is in This complication should be in the UK prestorage leucocyte depletion is fully of a new syndrome over recipients of filtered red cells in the have also been by the ( and Blood Advisory Committee 1998). acute and within of their transfusion. All patients had blood filtered through different of a filter from one The precise cause of this complication is not yet but the the of a system for and to transfusion. The of Transfusion scheme has in its the of any adverse after so that any new complications of LD are ( Williamson et al, ). As with other primary infection with CMV as human is followed by the of CMV ( & 1996). The of include cells in the ( et al, et al, 1997; et al, 1997 ) and a circulating cell population ( & et al, which may be a source of infectious virus in CMV blood CMV testing of donors is used to components for transfusion to and is the standard with which other should be Although this has CMV disease in individuals from a mean of the risk of patients seronegative components CMV is not In a of recipients CMV seronegative patients although CMV disease ( et al, ). may be due either to donors in the seronegative during primary infection or to donors who have gradually ( et al, 1994 ). It has been estimated that < of leucocytes in CMV blood donors are infected ( et al, 1999 ). random removal of such cells by > 3 log < infected cells would in a LD unit of reaction of CMV removal by LD to a in LD components ( et al, ). The number of leucocytes required to a susceptible individual is not known with However, a number of clinical studies of LD performed under have demonstrated the that in 3 out of studies the degree of leucocyte removal was to less than is achieved ( Table In the studies in out of with leucocyte-depleted blood CMV ( et al, ; et al, ) compared with out of with blood which was leucocyte CMV ( et al, ). studies of recipients have been In the studies in which LD was performed on blood under with quality out of patients including patients in a UK study of LD platelets in allograft recipients ( et al, 1999a ). is leucocyte depletion as as CMV It is that the one large study which directly compared the in recipients performed filtration of blood components at the which is as ( et al, ). This study is the only published of any CMV transmission to CMV in recipients of LD components. In the primary of CMV in study was found 2 out of patients with CMV testing and 3 out of patients with LD components. The filters used in this study were capable of 3–4 log leucocyte but as no quality monitoring was possible the CMV cannot be In this a previous study of bedside filtration has demonstrated to provide LD components ( Williamson et al, 1994 ). In addition, the presence of CMV has been demonstrated in the of but not blood, suggesting from leucocytes ( et al, 1997 ). This could have been a further for the The Council of Europe, the of Blood and the British Committee for Standards in Haematology all now that components that have been leucocyte at source are equivalent in safety to as CMV However, this has not yet been by either the and in the or by the UK Transfusion for Standards Standards Committee, so CMV seronegative components will continue to be available in the using LD components produced under would be required to determine whether CMV seronegative and LD components are However, these require to patients to LD than seronegative components. in the LD carried out by the UK Transfusion may to before such which would to involve There is accumulating evidence that this virus is in the of It appears to be transmissible of ( et al, 1997 ) and ( et al, ). virus has been from the mononuclear cells of a healthy blood ( et al, 1997 ), the possibility that the agent could be transfusion. A small study of patients to in their mononuclear cells ( et al, 1997 ). These patients were entirely on LD so it is unclear whether of transmission to removal of the virus source or through transfusion. studies in this area are in T-cell leukaemia/lymphoma virus (HTLV) all the of HTLV random of a of its into the HTLV

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.