For many years there has been ongoing debate in the United Kingdom (UK) about the necessity of using blood components from cytomegalovirus IgG seronegative (CMV neg) donors for CMV neg patients undergoing allogeneic hematopoietic stem cell transplantation, in the era of universal leukodepletion. From June 2012, our centre followed the guidance of the Advisory Committee on the Safety of Blood Tissues and Organs (SaBTO) in the United Kingdom, using CMV unselected leucodepleted components for all patients regardless of CMV serostatus. As part of an ongoing biovigilance commitment, we assessed the impact this policy had on rates of CMV seroconversion and viremia.

137 patients were included in the study, each allografted between March 2012 and December 2013. The median age was 49 years (range 2-71) and 62% were male. Diseases transplanted included acute leukemia or myelodysplastic syndrome (73.0%), lymphoma (11.0%), chronic leukemias (7.3%) and others (8.7%). 42 patients (30.7%) received myeloablative conditioning, 43 (31.4%) received total body irradiation, and 112 (81.7%) received in-vivo T-cell depletion with alemtuzumab. 119 (86.9%) received peripheral blood stem cells, 8 (5.8%) bone marrow, and 10 (7.3%) cord blood. 35 donors (25.5%) were siblings. 78 patients (56.9%) were CMV neg when first documented, of which 62 (79.5%) had a CMV neg donor. 59 patients (43.1%) were CMV IgG seropositive (CMV pos) when first documented, of which 36 (64.3%) had a CMV pos donor. All patients had CMV serology checked again within 28 days pre-transplant: of the 78 CMV neg patients, 14 (17.9%) were found to have changed their CMV status to seropositive. Of note, these patients had received significantly more platelet transfusions in the three months prior to transplant than those who did not seroconvert pre-transplant (median 13 vs 0 adult doses, p<0.001, see figure). Another such patient, who had not received platelet transfusions, had received multiple doses of intravenous immunoglobulin. There was also a trend towards increased packed red cell transfusions (median 8 vs 0 adult units, p=0.087) in these patients. None of these patients subsequently showed CMV viremia (by quantitative CMV DNA polymerase chain reaction) post-transplant, supporting the notion that this apparent seroconversion was due to passive transfer of CMV IgG antibody from seropositive blood donors, although CMV IgM or CMV DNA was not tested for at the time. By contrast, the rate of CMV viremia in patients found to be CMV pos at first contact was 89.3% at 100 days, with a median time to first detected viraemia of 22 days post-transplant. No CMV viraemia was seen in the 62 CMV neg patients with a CMV neg donor. Of the 15 CMV neg patients with a CMV pos donor, four patients (26.7%) had CMV viraemia, at 24, 46, 48, and 54 days respectively.


In 78 CMV IgG negative patients, there were only four incidences of proven primary CMV infection, all in patients receiving CMV pos stem cell donations. Excluding these cases, there were no cases of primary CMV infection associated with transfusion of CMV unselected blood components. However, we found a significant incidence of probable passive transfer of CMV IgG antibody, which was strongly associated with the use of platelet transfusions. Whilst this has no direct impact on the patient, it has significant implications for donor selection if it is incorrectly interpreted as representing prior CMV infection. It is conceivable that a CMV na•ve patient could be incorrectly ascribed as CMV positive due to detectable passive antibody: by selecting a CMV positive donor in these circumstances, the patient is thus subjected to a significant risk of primary CMV infection. Transplant centres must remain vigilant of this possibility, and be aware of the CMV selection policy for blood components in referring hospitals. Pre-transfusion CMV status should be obtained in all potential transplant candidates and, in those who appear to seroconvert, CMV IgM and DNA should be considered as indicators of recent seroconversion.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

This icon denotes a clinically relevant abstract