Hospital Pharmacist Vol 7 No 5 p118-123
May 2000 Special Features

Blood and blood products

Providing a safe and cost-effective blood transfusion service

By D. O'Shaughnessy, MBA, DPhil

This first article of our special feature considers the risks associated with blood transfusion and how these are being overcome. The author also discusses the cost implications for the service. The second article covers the alternatives to allogenic blood transfusion

The first record of blood transfusions date back to 1665 when animal experiments were being conducted by Richard Lower, an Oxford physician. They started as dog-to-dog experiments and proceeded to animal-to-human in the next two years. Almost simultaneously in France, Jean Denis, physician to King Louis XIV, initiated his own trials, and, by 1667, had successfully transfused four patients with lamb or calf blood. (Both animals have blood groups similar to the blood group A, which is associated with Anglo-Saxon lineage.) An attempt at transfusing a fifth patient was accompanied by a reaction, resulting in the voiding of black urine and death of the patient (almost certainly because the recipient was group O, associated with Celtic lineage). The patient's widow threatened to start a lawsuit and, although it was subsequently found that the patient had been poisoned with arsenic, this resulted in the banning of blood transfusions in England and France. 1 A whole century passed before blood transfusion was revived by James Blundell, an obstetrician from Guy's hospital, London, who perceived that blood transfusion would be appropriate for severe postpartum haemorrhage. He initially experimented with animals and found that blood from one animal could not be substituted for the blood of another with impunity. He also showed that blood could be transfused with a syringe even after collection into a container, provided it was used with minimum delay (see Figure 1). He also came to appreciate the importance of withdrawing all air from the syringe before transfusion. During the remainder of the nineteenth century, efforts were directed towards the improvement of devices or surgical techniques for direct transfusion (usually from the artery of the donor to the vein of the recipient), but indiscriminate use resulted in an increased number of untoward incidents. And so, when the twentieth century dawned, blood transfusion was still as primitive as it was two centuries before. The elucidation of the ABO blood group system by Landsteiner in 1900 and the discovery of the anticoagulant properties of sodium citrate in 1914 were two of the most important factors that led to blood banking as we know it now. Further developments in blood preservatives and an increased knowledge of other blood group systems have enhanced the effective utilisation of allogeneic blood as a therapeutic resource.2

Figure 1: The Impellor, devised by James Blundell in 1629. Blood from the donor was poured into the metal funnel, which was surrounded by warm water. The syringe was used to inject the blood via the tubing into the recipient

The transfusion service In 1921, Percy Oliver and three other British Red Cross members became the first volunteers to give blood to patients in King's College hospital, London. Having realised that many patients had neither friends nor relations willing to donate blood, he rapidly formed what he called the Red Cross Blood Transfusion Service and built up a donor panel from his home.
At that time, it was accepted medical practice to extract blood by cutting into the donor's veins and transfusing directly into the patients. Volunteers, who were not paid for their donations and rarely claimed travelling expenses, became closely involved with the hospital they visited, knew which patient they were helping and were given details of the outcome of each transfusion. The advantages of voluntary blood donation were soon realised. In countries where professional donation was the rule, up to 30 per cent had to be rejected due to ill health, in particular syphilis.
At one point, the service lost many donors because of the contemptuous way in which they were treated by one of the institutions. With the help of a medical adviser, Oliver drew up a set of regulations to ensure good practice in dealing with donors, on the part of doctors and institutions using the service.
By 1938, however, war loomed and it was realised that blood would need to be collected and stored, rather than come from peripatetic donors. The Ministry of Health set up regional depots throughout the country which were administered by the Medical Research Council during the war. After the war, these became the National Blood Transfusion Service (NBS).

TTIs

There is the potential for a number of diseases to be transmitted by transfusion of blood or its components. They are referred to as transfusion transmitted infections (TTIs). Donor selection criteria and the subsequent testing of all donations are designed to prevent such disease transmission. Accurate estimates of risk are important so that informed decisions can be made on whether to use voluntarily donated allogeneic blood or the alternatives. Currently in the UK, it is mandatory that each unit be tested for hepatitis B and C, human immunodeficiency virus (HIV) 1 and 2 and syphilis (see Table 1). Some units are tested for cytomegalovirus (CMV) to identify a CMV-negative donor panel. It is not mandatory to test for low-frequency diseases such as malaria and human T cell lymphotropic virus (HTLV) 1 and 2. There is as yet no test for new variant Creutzfelt-Jacob disease (nvCJD). Higher frequency community infections such as parvovirus or hepatitis A (HAV) are also not tested for, even though transmission by transfusion can occur.3,4 Table 1: Laboratory tests Agent Test performed Laboratory method used Hepatitis B virus Hepatitis B surface antigen Third generation ELISA Hepatitis C virus Hepatitis C virus antibody Third generation ELISA Syphilis TPHA Haemagglutination (automated) CMV CMV antibody Latex agglutination assay HIV 1 and 2 HIV antibody Third generation ELISA Key: ELISA=enzyme-linked immunosorbent assay TPHA=Treponema pallidum haemagluttination In the USA, detection of HIV is by amplification of viral nucleic acid polymerase chain reaction (PCR) in all donors; in the UK this is reserved for testing plasma. The cost of taking all of these laboratory tests is approximately £1.25 in the UK.

Figure 2: The three main blood transfusion components

Although each unit of donated blood is tested for markers of viral infection, there are reasons why viral transmission still occurs. The primary reason is the "window" period when the donor is infectious but has not yet developed a detectable antigen or antibody. Another reason is that some chronic carriers of hepatitis B or C are undetectable by current assays.5 Occasionally, laboratory error, combined with a different manufacturing process (see Figures 2 and 3), contributes to transfusion transmitted infections, as in the hepatitis C infection of postpartum women receiving anti-D in Ireland up to 1994. Figure 3: The processing of plasma into its main components. This is no longer carried out in the UK

SHOT SHOT, or serious hazards of transfusion, is a confidential and voluntary record of transfusion-related deaths and diseases. In the first 24 months, there were 366 incidents (with 22 deaths), half as a result of giving the wrong blood to patients. There were 12 infections (four bacterial, seven viral and one protozoal-malaria) including one fatality.⁶

Current risk of TTIs

Table 2 shows the incidence of TTI in the UK and USA. The risk of transmitting infection applies to all components (except HTLV and plasma) and the recipient of one of these infected products has an 80 to 90 per cent chance of acquiring the infection. A reduction in the relative risk between 1992 and 1994 can be attributed to more sensitive screening tests which substantially reduced the "window" period. This period is currently estimated at 22 and 66 days for third generation assays of anti-HIV and anti-hepatitis C virus, respectively, and 14 days for nucleic acid testing (PCR of p24 antigen in HIV). The "at risk" groups are more likely to have more than one virus and therefore each positive test can act as surrogate markers for the other viruses.

Table 2: Risk of infection from screened blood
Agent	1992 (USA)	1994 (USA)	1994 (UK)
HCV	1 in 3,300	1 in 5,000	1 in 40,000
HTLV*	1 in 50,000	1 in 100,000	Not tested
HBV	1 in 200,000	1 in 200,000	1 in 200,000
HIV	1 in 225,000	1 in 410,000	1 in 5,000,000
*Risk only in cellular products less than 14 days old

HIV In the USA, the risks of transfusion-associated HIV infection rose rapidly from its first occurrence in 1978 to a peak of 1.1 per cent of transfused units in 1982. With the introduction of HIV education, there was a progressive reduction in the risk, mostly as a result of "at risk" people declining to donate (see Figure 4).
Anti-HIV screening, introduced in early 1985 in the USA (later in the UK) practically eliminated the residual 10 per cent of positive donors. Unfortunately, 201 cases of acquired immune deficiency syndrome (AIDS) were reported, due to blood transfusions between 1986 and 1992. Since then, it has been estimated that, of the 18m products transfused annually, about 43 recipients are likely to receive contaminated blood, out of which only five to 10 will develop AIDS before dying of the illness which necessitated transfusion. This is 2,000 times lower than at the peak of the epidemic (1982 to 1984).

Figure 4: Risk of HIV contamination

Hepatitis C From the 1970s to the 1990s, approximately 12 per cent of transfusion recipients in the USA and 0.5 to 1 per cent in the UK developed post-transfusion non-A non-B (NANB) hepatitis. Virtually all of those receiving pooled plasma products at that time had NANB hepatitis, as did those receiving over 100 units of red cells in chronic transfusion programmes. It is now clear that at least 85 per cent of all cases of NANB were due to hepatitis C virus (HCV) infection. With the introduction of third generation HCV tests, the residual risk appears to be between 1 in 20,000 and 1 in 600,000 units.

Hepatitis B In the past, blood transfusion offered a significant risk of hepatitis B virus (HBV) infection. Nowadays, the actual rates of transmission are low and in the order of what would be expected in the population not transfused (six per one million individuals). Vaccination is recommended in patients requiring repeated administration of blood components.

Bacterial infection Recently, there has been renewed concern about the risk of transmitting bacterial infection by blood transfusion. The main source of bacteria is the donor arm, with most isolates being Staphylococcus aureus, S epidermidis, Pseudomonas and Yersinia spp. They survive in stored blood because they can use citrate as their source of carbon. The risk is greatest with platelets which are stored for up to five days at room temperature and from where bacteria have been cultured in up to 0.4 per cent of bags of platelet concentrates. It is unknown whether these in vitro data represent true contamination of the unit or whether small amounts of transfused bacteria cause recipient morbidity. The majority of red blood cell (RBC) units implicated in bacterial contamination cases have been stored over 25 days. The contaminating organism are those that flourish in the cold, such as Pseudomanas or Yersinia spp.
It is unknown if leucodepletion (removal of leucocytes by filtration, see below) has an effect on bacterial contamination. A number of studies, mainly on red cells, has suggested that there should be an eight-hour incubation period prior to leucodepletion to allow donor granulocytes to phagocytose and self-sterilise the unit from contaminating bacteria, but experience with apheresis donor platelets, in which leucodepletion occurs during collection, has not yet shown an increase in bacterial contamination.
More stringent antisepsis procedures (including iodine) are being introduced by transfusion services in an attempt to reduce bacterial contamination of collected blood.

Syphilis Syphilis testing continues despite only seven donors being identified in 10 years at a centre in London. The main reasons are that it also acts as a surrogate marker for possible HIV and hepatitis C virus. Fortunately, it is a low cost test (10 to 25 pence per test). In the USA and Europe, there is no testing for other parasites such as Plasmodium (malaria) trypanosomes (Chagas) and Borrelia (Lyme disease).

New variant Creutzfelt-Jacob Disease (nvCJD) Transmissible spongiform encephalopathies (TSE) are clinical diseases characterised by spongiform changes in the brain. They encompass a spectrum of neurological diseases affecting humans which are related to prion proteins or PrPs (CJD, nv CJD, Kuru) and animals (bovine spongiform encephalopathy, BSE, in cattle and scrapie in sheep). The prevalence of human TSE is 0.7 per million of the population. TSE may be sporadic (87 per cent), genetic (8 per cent), infectious or iatrogenic in aetiology, but all have the potential for infectivity.
Exposure to temperatures in excess of 300C for one hour are required to kill the causative agent, the prion (proteinaceous infectious particle) or PrP. BSE occurs in cattle fed with offals containing nervous tissue which was rendered by a different sterilisation technique and in which extraction using organic solvents has been excluded. Over a third of British dairy herds have been affected since the first recognised cases in 1986. In 1996, stringent guidelines were issued to control BSE. The government slaughtered 166,000 cattle, and bovine offals were banned from human consumption in July, 1998. Prior to the ban, it was assumed that BSE posed no public health problem because of the previous experience with scrapie in sheep, which has been endemic for many years. Unfortunately, it is believed that 700,000 infected cattle have entered the human food chain.
It was in 1995 that nvCJD was first discovered in two teenagers and it is now believed to be BSE transmitted to humans. There have been a total of 29 cases to date, but models of infectivity predict 100 to 35,000 cases in the next 10 to 20 years. Transmission studies in rodents suggest that infection with PrP via the gastrointestinal tract (as opposed to direct innoculation into the central nervous system) requires a phase of peripheral lymphoreticular (within white blood cells) replication, prior to neurological replication. There are also anecdotal reports of classic CJD being contracted subsequent to albumin and blood transfusion from donors who have later died of CJD. However, case control studies have not shown an association between classic CJD and prior blood transfusion to date.
The uncertain incubation period of nvCJD (from five to 30 years), means that we are unlikely to see secondary cases of nvCJD from blood transfusion and organ donation until well into this millenium.

Progression of TTIs

Recipients who acquire HIV infection from blood transfusion progress to clinical disease at a rate similar to that of other infected individuals, with a mean incubation period of 10 years. In the UK, 1,346 people were thus infected, all but 11 being male. Most of those infected had haemophilia and were receiving concentrates of factors VIII or IX. The plasma used to produce these concentrates was neither screened for HIV nor heat-treated in the years prior to 1985. Almost half of those who were HIV positive 20 years ago have progressed to full-blown AIDS and 61 per cent of these have died, one third due to non-AIDS causes, such as liver and cardiovascular diseases. The preponderant cause of death in the AIDS group has been opportunistic infections. Kaposi's sarcoma has not been reported in any cases of transfusion-related HIV.

Hepatitis C The natural history of transfusion-acquired HCV is similar to that acquired through other modes of transmission. Genotypes 1,2 and 3 predominate in Northern Europe and North America. Among those infected with hepatitis C, 20 per cent eliminate the virus and become PCR negative. Of the 80 per cent who have persistent hepatitis, 20 per cent will have chronic progressive liver disease at 10 years while 1 to 9 per cent will progress to liver failure. Co-infection with HIV exerts an adverse effect, as does moderate alcohol. Prior to hepatitis C testing in the UK, at least 1 per cent of the transfused population were infected with HCV. This includes all those with severe haemophilia, and all those receiving more than 100 units of blood such as patients with thalassemia major .

Hepatitis B This is rare, due to the screening of blood donors and vaccination of patients. In addition, most recipients who acquire HBV resolve the initial infection and develop immunity. Vaccination also confers protection against Hepatitis D, a ribonucleic acid (RNA) virus that requires B to be present. Those unfortunate to have contracted hepatitis D virus via coagulation factor concentrates, or because it is endemic (Africa and Middle East), develop rapidly progressive liver disease.

CMV As 50 to 60 per cent of the UK population are anti-CMV positive, post transfusion CMV is quite common after platelet or red cell transfusion if the cells are not leucodepleted. However, with the exception of those at risk of pneumonitis, most cases are subclinical. Transmission of virus with fresh frozen plasma (FFP) or plasma derivatives has not been convincingly reported. The following groups of patients are at risk of infection from CMV:

• Foetuses and premature babies weighing less than 1500g
• Bone marrow, stem cell and organ transplant recipients
• Immunosuppressed individuals

Transfusion-related lung injury (TRALI) Strictly speaking, this is not a TTI, but a transfusion-related disease where donors (who have previously been pregnant or transfused) have in their plasma antibodies to human leucocyte antigens (HLA) or granulocyte antigens. Transfusion of plasma or platelets suspended in plasma with these antibodies results in the patients' granulocytes aggregating in the pulmonary circulation, leading to hypoxaemia with shadows visible on the chest x-ray.

Risk reduction

With heightened public awareness of blood transfusion safety, policy formulation is driven by the need to reduce the risk to zero. However, any one method is unlikely to be absolutely effective and several overlapping methods are likely to be expensive. The solution is thus likely to involve minimising the risk at an acceptable cost which could indicate minimising exposure to blood products in the first instance.

Selection of donors Safety begins with donor selection and the altruistic nature of voluntary blood donation is essential to this. The construction of a good pre-donation questionnaire is essential. The knowledge that prison inmates tend to spread hepatitis means that blood collection teams should not collect in prisons. Since the HIV epidemic in male homosexuals and drug abusers, those "at risk" groups have always been excluded. Similarly, those recently exposed to live vaccines, who have just returned from endemic malarial areas or who are febrile at donation, are excluded from donating their blood. It is also important to exclude donors from intercurrent illnesses such as dental abscesses and diarrhoea or vomiting. TRALI could be prevented by excluding all previously transfused or pregnant donors, but this would exclude at least 20 per cent of the donor pool and is therefore not practical.

Special processes A number of processes exist which can help to reduce the risks associated with transfused blood.

Heat treatment
Viral inactivation of cellular products is not possible with heat. However, pooled plasma products can be pasteurised at 80C for 72 hours (60C and a shorter period of time for albumin) with little change in plasma protein activity. The temperatures required to clear plasma of nvCJD would destroy the proteins and render it inactive.

Solvent and detergent (SD) treatment
Solvent-treated viral inactivation can only be applied to pooled plasma of up to 1,000 donations. The solvent and detergent are removed by oil extraction and chromatography. The technique inactivates lipid-coated viruses and has been used to treat factor concentrates for 10 years. More than three million units of SD-treated FFP has also been used in Europe. It has no cellular debris and is licensed in the UK for the same indications as FFP.
First generation immunoglobulins (not SD-treated) were always thought to be safe as a result of the Kohn fractionation process, until 1993 when there was an outbreak of HCV in the USA. Since then, only SD-treated immunoglobulins have been used. Both first generation and second generation immunoglobulins are licensed in the UK. Viral inactivation of cellular products is not possible with solvent and detergent.

Methylene blue dye
It is possible to inactivate viral cellular products with methylene blue and ultraviolet B (UVB) radiation, or psoralens and ultraviolet A (UVA) radiation. The addition of methylene blue to single units of plasma, which are then exposed to light (UVB) in a light box, means that there is no need to pool plasma, and over one million units of dye-treated plasma have been used in Europe. Unfortunately, the current generation of light boxes available in the UK do not have a high throughput facility and there has been one case report of HCV in Germany, which means that this is not an option currently used in the UK.

Leucodepletion
Modern leucocyte filters reduce the leucocyte count to less than 1x106. CMV tends to be associated with CD14+ monocytes and filtration of fresh CMV positive blood results in a product negative for CMV DNA (This is however, not the case for blood filtered at the bedside). The Council of Europe has endorsed the use of blood leucocyte-depleted in the transfusion centre as a safe substitute for CMV-negative blood. This does not apply to HIV, where evidence would suggest that transmission may occur with very low doses of leucocytes.
It is expected that leucodepletion of red cells will reduce the risk of nvCJD since all studies of PrP expression support the contention that leucocytes represent the main source of infectivity in the peripheral blood of nvCJD-infected individuals. However, platelets express PrP and they could be a source of infectivity even when leucodepleted.

Irradiation
Some blood components may need to be modified to meet patients' special needs. Thus gamma irradiation of 25Gy is required after a bone marrow transplant and in other immunodeficient states to ensure that no viable donor lymphocytes are present and reduce the risk of transfusion-related graft-versus-host disease. The most recent cases have been in neonates given fresh blood for an exchange transfusion. Leucocyte depletion removes insufficient leucocytes to guarantee protection.⁷
Specific indications for irradiation are:

• Intrauterine transfusions
• Congenital immunodeficiency states
• Matching HLA antigens, including products from relatives
• Bone marrow transplant patients
• Hodgkin's disease and therapy with purine analogues

Minimising donor exposure
To minimise exposure to neonates, the NBS provides units that have been separated into four smaller units (quad packs) or eight minipacks (octa packs) which can be reserved for a named neonate and used for sequential transfusions over 42 days. For oncology patients, donor exposure from multiple platelet transfusions can be reduced by using apheresis packs.
Donor exposure can also be reduced by lowering transfusion triggers as part of education and continually reviewing practice. This is facilitated by having an active and commited multidisciplinary blood transfusion committee within the hospitals. It is also facilitated by research into practices that exist and comparing the outcomes. Thus, the Sanguis project, which audited practices in 43 European hospitals, found a dramatic variation in the use of red cells, FFP and albumin from zero to over 85 per cent, demonstrating that by employing the best practice, unnecessary transfusions can be reduced and outcomes improved (see Figure 5).

Figure 5: Morbidity associated with blood transfusions. The incidence of post-operative complications in patients undergoing abdominal aortic surgery

Use of American plasma
Fractionated plasma products are a particular concern, since up to 20,000 donations are pooled per batch. Although animal experiments would suggest that plasma is unlikely to be a source of high infectivity for nvCJD, there is still concern because PrP is shed from leucocytes and platelets into plasma. There have been two occasions in which all products in a batch to which a patient with nvCJD has contributed were recalled. As a result, the British fractionation plant underwent extensive sterilisation and plasma has since been sourced from outside the UK, until a test is developed to screen for potential infection.

Recombinant products From 1997 onwards, the UK Haemophilia directors (doctors who direct haemophilia centres) have pushed for the use of recombinant products in haemophilia. Restricted availability world-wide has meant that this only applies to date for children and previously untreated cases. The cost implications have been immense.

Dr O'Shaughnessy is a consultant haematologist at Southampton General hospital

References