Risk of cytomegalovirus transmission by blood products to immunocompromised patients and means for reduction

Cytomegalovirus (CMV) transmission remains a particular concern for immunocompromised patients, such as those undergoing bone marrow transplantation or receiving chemotherapy. It is of major importance for CMV seronegative patients that the risk of CMV transmission with blood products is minimized. The currently available techniques, use of CMV-seronegative blood products and leucocyte-reduction by filtration, have limitations. Pathogen inactivation is the newest approach to making blood safer. Several technologies are already available or in development. The final choice of a technique will most likely depend on the risk for severe CMV disease in the patient population as well as efficacy, safety, and costs of the chosen techniques. Cytomegalovirus (CMV) is a highly cell-associated virus, which normally causes an asymptomatic infection in the immune competent host. After a primary infection, the virus persists in a latent state, from which it can be reactivated under certain conditions. Since CMV can cause severe illness and even death in immunocompromised patients, such as patients undergoing haematopoietic stem cell transplantation, solid organ transplantation, and patients suffering from leukaemia or lymphoma, the spread of CMV through blood products should be prevented. The risk for a viral infection through blood products can be limited by an improved selection of the donor-population. However the high prevalence of CMV seropositive individuals in the donor-population of many countries represents a particular problem since an increasing demand for CMV-free blood products may be difficult to meet if CMV positive donors are excluded. In addition, the cost of maintaining a CMV-negative blood supply can often be high. White blood cell (WBC) reduction has been shown to reduce the risk of transmission of CMV via blood products to a level within the same order of magnitude as by selection of CMV negative blood products (Bowden et al, 1995; Ljungman et al, 2002). There is no consensus whether the techniques are equivalent, what conclusions should be drawn regarding CMV testing of blood products and whether different techniques should be used in patient groups at varying risks for CMV disease. The aim of this review is to discuss the risk for CMV infection through blood products in different immunosuppressed patient populations to compare different methods for preventing CMV transmission including viral detection, leucoreduction and pathogen inactivation. The clinical symptoms of a CMV infection are generally mild or absent in a normal immune competent individual. A mononucleosis-like syndrome with fever, malaise, myalgia, liver dysfunction and heterophil antibody negative lymphocytosis may occur several weeks after exposure (Klemola & Kaariainen, 1965). In contrast, CMV infection in transplant patients may lead to severe illness (Ljungman, 2002) including hepatitis, thrombocytopenia, haemolytic anaemia and pneumonia (Neiman et al, 1977; Meyers et al, 1982, 1990; Paya et al, 1989; Sayage et al, 1989) with potentially lethal outcome, despite modern antiviral therapy (Ljungman et al, 1992, 2001; Einsele et al, 2000; Machado et al, 2000; Limaye et al, 2001; Boeckh et al, 2003). The severity of the CMV infection correlates directly with the degree of immunosuppression for each transplant setting and/or patient condition, as well as with the serological status of patients and, in the transplant situation, of their donors (Smyth et al, 1991; Egan et al, 1998; Hertz et al, 1998; Ljungman et al, 1998; Lowance et al, 1999; Einsele et al, 2000; Hebart et al, 2001; Junghanss et al, 2002). Data on the importance of CMV infection in non-transplant cancer patients are limited. Among adults with acute leukaemia, CMV pneumonia was reported to occur in 2·9% of patients with a case-fatality rate of 57% (Nguyen et al, 2001). There is an increased risk for CMV infections and also for CMV-associated disease in lymphoma patients receiving T-cell suppressive therapy, such as fludarabine (Anaissie et al, 1998) or anti-CD52 antibodies (Campath) (Bowen et al, 1997; Lundin et al, 2002; Nguyen et al, 2002). Recent data indicate that infection with CMV may not only cause a viral disease syndrome, but also a more general impairment of the cellular immune response (Grigoleit et al, 2002). This hypothesis is supported by the observation that infection of dendritic cells with CMV may not only cause an immune response sufficient to control infection with CMV, but also a down regulation of major histocompatibility complex (MHC) class I and II, which could explain why bone-marrow recipients who become infected with CMV show a delay in immune reconstitution (Grigoleit et al, 2002). This is illustrated by the fact that CMV seronegative recipients of stem cell transplants (SCTs) from CMV seropositive donors are at a significantly increased risk for death due to bacterial and fungal infections. Mortality due to bacteraemia or invasive fungal infection was 18·3% for CMV seronegative stem cell recipients of seropositive donations, and 9·7% for seronegative stem cell recipients receiving from seronegative donors. These numbers indicate a possible immunomodulatory effect of primary CMV infection independent of the direct CMV associated morbidity (Nichols et al, 2002). The importance of a primary infection in immunosuppressed individuals varies between different patient populations. Primary infection is an important risk factor for CMV disease and also for rejection in solid organ transplant recipients (Rubin, 1990; Gane et al, 1997; Flechner et al, 1998; Lowance et al, 1999). The major risk factor for CMV disease in allogeneic SCT patients is CMV seropositivity of the patient, however primary infection is also associated with a significant risk for development of CMV disease. The importance of primary infection in other patient populations, such as acute leukaemia patients, undergoing induction chemotherapy and autologous SCT recipients is less well known. The main source of CMV transmission to a seronegative organ or SCT recipient is the donor infectivity (Bowden et al, 1986, 1991a; Balfour et al, 1989; Smyth et al, 1991; Bailey et al, 1992; Merigan et al, 1992; Snydman et al, 1993; Conti et al, 1994; Rostaing et al, 1997; Ljungman et al, 1998; Lowance et al, 1999). However, transmission from blood products has also been documented to be an important source of primary CMV infection unless preventive measures are taken (Bowden et al, 1986, 1991b). An incidence of primary CMV infection of 23% has been reported following the administration of CMV unscreened blood to autologous SCT recipients (Reusser et al, 1990). Seronegative patients receiving grafts from CMV seronegative stem cell donors have a similar risk, with incidence figures between 32% and 37% (Bowden et al, 1986; Miller et al, 1991). However, the use of CMV seronegative blood products has no, or a very limited, impact on the risk for CMV infections in CMV seronegative allogeneic SCT patients with CMV seropositive donors (Bowden et al, 1986; Miller et al, 1991). The risk in less immunosuppressed patients than SCT recipients might be lower. In CMV seronegative organ transplant recipients receiving organs from CMV seronegative donors, the risk for CMV transmission from unscreened, non-leucocyte-depleted blood products is rather low (Preiksaitis et al, 2002) and the consequences might be less severe than after CMV infection transmitted from the organ donor (Falagas et al, 1996). In seronegative children with malignancies, no transfusion-associated CMV infection was found regardless of whether the patients received leucocyte-depleted or standard blood products (Preiksaitis et al, 1997). The risk in non-immunocompromised patients of transfusion-associated CMV disease is low (0·9%) (Preiksaitis et al, 1988). What is the risk of CMV transmission to an already CMV seropositive individual? Contact with blood products containing CMV may cause reactivation of latent CMV virus in a state of immunosuppression. Another possibility is reinfection with a new strain of CMV, as demonstrated in patients with the human immunodeficiency virus (HIV) and organ transplant recipients (Chou, 1987; Grundy et al, 1988). A higher risk for a severe CMV infection has been suggested following infection with a second strain of CMV (Grundy et al, 1988). It is very difficult to assess with certainty the importance of CMV transmission by blood products into already CMV seropositive patients. Whether CMV disease occurring in seropositive patients is more or less frequently caused by a reactivation of a previous strain or by reinfection of a new strain is unknown but is likely to be strongly influenced by the existing CMV immunity in the patient (and the donor) (Nichols et al, 2002; Ljungman et al, 2003). Leucocytes are the most likely vehicle of transmission for the highly cell-associated CMV in blood products (Einhorn & Öst, 1984; Rice et al, 1984). CMV transcripts have been demonstrated in a subset of CD33+ cells of the bone marrow of naturally infected CMV seropositive subjects (Kondo et al, 1994; Kondo & Mocarski, 1995). Granulocyte transfusions have also been associated with a high risk of CMV transmission (Hersman et al, 1982a). CMV can be demonstrated in granulocytes either by culture, immunostaining or by direct DNA detection (Gerna et al, 1991, 1992; Boland et al, 1992). Following incubation with high titres of CMV, in vitro monocytes and CD8 lymphocytes can be infected (Einhorn & Öst, 1984; Braun & Reiser, 1986). Monocytes seem to be a very important source of CMV in blood products of healthy seropositive individuals (Taylor-Wiedeman et al, 1991), particularly in the terminally differentiated CD14+ monocytes of the myeloid/granulocyte lineage (Bolovan-Fritts et al, 1999). It has been difficult to isolate CMV from fresh blood or leucocytes from healthy blood donors (Hersman et al, 1982b). The high risk for transmission through blood products is therefore likely to be caused by factors stimulating CMV reactivation. One such effect is the allogeneic effect. In vitro data indicate that allogeneic stimulation can reactivate CMV from both seropositive and seronegative blood donors (Söderberg-Naucler et al, 1997). This effect is most likely to be mediated through cytokine stimulation in the allogeneic situation. In the immune competent CMV seropositive host, CMV reactivation is most likely to be directly controlled by the pre-existing immune response. However, in severely immunocompromised individuals, this control function might fail. The risk for CMV disease when blood products transmit a primary infection is difficult to determine. It will most likely depend on the immunosuppressed state of the patient, the viral load transmitted by the blood product, and the preventive measures taken against CMV (antiviral prophylaxis, monitoring and pre-emptive therapy). In a randomized study comparing CMV seronegative blood products to leucocyte-depleted blood products, Bowden et al (1995) found a similar risk of CMV infection and disease. The risk for CMV transmission by erythrocyte or platelet transfusion is probably related to the number of leucocytes present in these products. Although the concentration of leucocytes in unfiltered erythrocyte preparations is generally 10 times higher than in platelet preparations, due to multiple transfusion of platelet preparations the risk of CMV transmission is probably greater for platelet recipients (Bowden, 1995). Fresh frozen plasma does not seem to transmit CMV infection, possibly because the few leucocytes found in the plasma are disrupted during the freezing procedure (Bowden & Sayers, 1990; Bowden, 1995). It is important to differentiate between the detection of pre-existing CMV infection, i.e. seropositivity, and ongoing CMV infection. The important risk factor for transmission of CMV is existing seropositivity in the blood donor population. The prevalence of CMV antibodies in a healthy donor population varies greatly, and figures between 20% and 100% have been presented (Bowden, 1995). Approximately 50–80% of the blood donor populations in the USA are CMV seropositive (Krajden et al, 1996). Since primary CMV infection in the immune competent host rarely causes any characteristic symptoms, the history of CMV infection in a blood donor is unreliable. CMV antibody was detected in 37% of a volunteer donor population that did not give a history of CMV infection (Zhang et al, 1995). The most common techniques for screening for a previous CMV infection are serological assays based on passive particle agglutination and latex agglutination. Both are quite sensitive and quick, and are therefore suitable for screening. There are some limitations to serological testing. One is the possibility of a ‘window period’ between becoming infected with CMV and seroconversion. Analysis of IgM antibodies might shorten the duration of this ‘window period’ by enabling earlier detection of a primary infection before IgG seronconversion has occurred. The second limitation existing for all screening tests is the risk of false negative results. This might be due to genotypic variation of the virus giving rise to antibodies not detected by the assay used or waning of the antibodies giving titres that are too low for detection by the assay used (Laupacis et al, 2001). Different assays have different levels of sensitivity. For example, the risk of false negative results is approximately 4% for the ‘immunofluorescent antibody test’ and 25% for the complement fixation test (Bowden, 1995). The risk for incorrect serological assignation to seropositive or seronegative of course exists both for patients, organ or stem cell donors, or blood donors. The risk is most probably highest for immunocompromised patients, since these might have an impaired capacity to mount an antibody response to CMV. Finally, it has been suggested that abortive infections with CMV exist that do not give rise to a serological immune response and unless the individuals are exposed to replicating virus, this prior infection may interfere with seroconversion (Zhang et al, 1995). Instead of detecting antibodies, another option is to try to detect viral antigens or viral nucleic acids in the peripheral blood. It has been shown that it is possible to detect CMV DNA or RNA in both seropositive and seronegative healthy blood donors (Table I) It is not known whether blood products from CMV seropositive donors, from which CMV DNA or RNA can be detected, give a different risk of CMV transmission than blood products that are CMV DNA or RNA negative. However, it has been shown that the likelihood of detecting CMV is highest early after primary infection. CMV viraemia, as indicated by positive viral culture, was detected in leucocyte samples of one out of six patients within 9–10 weeks after infection (Zanghellini et al, 1999), while CMV DNA was detected in the leucocytes of five of the six patients during the first 16 weeks of infection with a successive decline, although one of six patients were still positive at 48 weeks after infection. One study of 52 immunocompetent individuals (including 40 pregnant women) with a symptomatic primary CMV infection showed that CMV DNA was detected in leucocytes in all patients during the first month after an acute CMV infection. A steady decline of the viral DNA was thereafter demonstrated during 6 months follow-up. Thereafter no viral DNA could be detected (Revello et al, 1998). The detection of CMV DNA or RNA in blood products from CMV seronegative donors could explain the observation that seroconversion occurs in 4–6% of recipients of CMV negative blood products (Bowden et al, 1986; Miller et al, 1991). However, it should be noted that the studies showing a high detection frequency of CMV DNA or RNA from seronegative blood component donors are the and more studies have shown a very low frequency in both seronegative and seropositive blood component donors. This might be due to the development in for nucleic It is also likely that the risk for transmission from a seronegative donor is on the serological assay used for in that a more assay might a donor with a low antibody as Since the risk for transmission by blood products of CMV is several have been to the These use of CMV seronegative blood donors, reduction of the number of leucocytes in the blood products, and of the blood products to the It should be that each of these has including of products, incorrect of the of product, of techniques, and the of the new techniques based on of the blood products. It should also be as that in all studies could as well be caused by of the serological testing of the patient the donor in a transplant as the of the of blood products. The first was to blood donors for CMV and CMV seronegative donors for CMV seronegative patients. The with this technique is that it might be difficult to CMV seronegative products from donor populations the CMV prevalence is and the risk for false negative screening assays is still as Seronegative blood products strongly reduce the incidence of infection in CMV seronegative SCT patients with CMV seronegative stem cell donors to (Bowden et al, 1986, 1995; et al, 1988). results can be after solid organ transplantation et al, Since CMV is a cell-associated virus and exists in peripheral blood to reduce the number of leucocytes might reduce the risk for transmission of CMV. Several studies were in different patient populations and showed that this was the A reduction of CMV levels in vitro has been demonstrated following leucocyte reduction et al, 1998). a reduction of CMV DNA in and in blood following has been reported et al, In clinical a significant risk reduction of CMV transmission was after the of leucocytes from blood products. of blood products from seropositive donors the risk to levels similar to when products from seronegative donors are used (Bowden et al, 1995; et al, 1990; et al, 1998; & 2001; Ljungman et al, 2002). The of blood on the of the leucocyte reduction and the number of leucocytes at have suggested at a reduction in the of virus et al, 1998; et al, 2001). different factors the of leucocyte including the of the blood to be and the or et al, 1998; et al, 2001). Although the number of infected leucocytes to transmit CMV infection is not the leucocytes a risk of viral In addition, of may be by with different although the of has been shown to be between and 10 et al, 2000; et al, leucocyte are not against CMV or the of infected leucocytes et al, 1997). However, the detection rate of CMV DNA is in plasma than in blood or leucocytes (Table the of this might be limited. There are other possible of leucocyte such as reduction of the risk for transmission of other a possible risk for and for transfusion that might be by leucocyte virus human virus and study has the or importance of leucocyte on infections with these and are causes of transfusion cells both class I and antigens to Since leucocytes be from a blood an of leucocytes of has been suggested to caused by However, the risk of or might with more cellular including and can be used as screening or assay of the of the et al, 1998). Since both the use of CMV seronegative blood products and leucocyte-depleted blood products are to reduce the risk for CMV the remains which is most One randomized study has CMV negative blood products with leucocyte-depleted blood products in CMV seronegative SCT recipients receiving grafts from CMV seronegative donors (Bowden et al, 1995). In bone marrow transplant patients, no significant could be found between the The risks for CMV infection were and while the risks for CMV disease were and of the patients that were infected CMV disease and the was with of CMV disease in five of six patients. The that a reduction of the number of leucocytes is as as the use of seronegative blood. The results of this study have been for several It was to detect a in CMV disease. of the patients received autologous grafts and therefore were at a risk for CMV disease. The blood products were and was no control regarding the of However, the study has influenced the of many to CMV negative with leucocyte-depleted blood et al the results from studies in SCT and acute leukaemia patients, including the randomized study by Bowden et al The risks for CMV infection were for patients receiving CMV seronegative blood products, for patients receiving leucocyte-depleted blood products, and for patients receiving unscreened, non-leucocyte-depleted blood products et al, 1999). After this studies different in SCT patients have been The results are in The of the studies were as Ljungman et al a of patients that only received leucocyte-depleted blood products with a control the of leucocyte-depleted and blood products from CMV seronegative donors. used monitoring and therapy was if the patients positive et al a of CMV seronegative patients receiving a seronegative transplant and cells from CMV seronegative donors with unscreened, leucocyte-depleted of these patients primary CMV infection et al, 2002). The this to a of patients receiving CMV seronegative blood products et al, 1998). et al a of CMV seronegative SCT recipients receiving different CMV seronegative blood products were used if blood products were by were used (Nichols et al, 2003). The risk for CMV infection was higher in the second (Table while CMV disease only in one patient was not reported whether this patient was during or the few techniques have been that blood products in vitro with the aim to reduce the risk for transmission of different These blood products as well as many different viral and virus, CMV These methods could be for the risk of CMV The that has been most is based on the inactivation of nucleic acids by DNA and RNA from blood products, and A The effect on CMV infection has been demonstrated in vitro and in an The of and was to of CMV in platelet In an immunocompromised in transfusion with with CMV of CMV and with with prior to with no of CMV disease and healthy 2001). technologies that aim to blood are currently and use nucleic or 2002; et al, 2002). This for pathogen inactivation in platelet has been in clinical for and et al, in the USA and is for use on platelet in These studies were not to detect any in transmission of infections including CMV the of the against CMV is based on in vitro data and The and technique for of plasma and a similar for blood cells are both in clinical in the while the technologies for of blood cells or are currently under development. these techniques also have other on blood products, including the inactivation of and could reduce the risk for transfusion and However, no such effect was documented in a controlled with and in et al, 2003). CMV transmission remains a particular concern for immunocompromised patients, such as those undergoing bone marrow transplantation or receiving chemotherapy. It is of major importance for CMV seronegative patients that the risk for transmission of CMV with blood products is minimized. The currently available techniques, use of CMV-seronegative blood products and leucocyte-reduction by filtration, both have limitations. Pathogen inactivation is the newest approach to making blood safer. Several technologies are already available or in development and their in CMV has to on CMV testing in the The final choice of a technique will most likely depend on the risk for severe CMV disease in the patient population as well as the efficacy, safety, and costs of the chosen techniques.