Safety of leukoreduced, cytomegalovirus (CMV)-untested components in CMV-negative allogeneic human progenitor cell transplant recipients.

Transfusion-transmitted cytomegalovirus (TT-CMV) infection can lead to significant morbidity and mortality in CMV-negative (CMV-N) hematopoietic progenitor cell (HPC) transplant patients. In 1995, Bowden and colleagues demonstrated the efficacy of leukoreduced components to reduce TT-CMV in most high-risk populations, although there remained safety concerns in CMV-N allogeneic HPC transplant recipients. As a result, several transplant programs recommended both leukoreduced and CMV-N components for CMV-N allogeneic HPC patients receiving transplants from CMV-N donors (CMV). A more recent study, however, has challenged the clinical benefit of requiring CMV-N, in addition to leukoreduction, in CMV HPC patients. In a 10-year study, Thiele and colleagues found no cases of TT-CMV in 23 CMV HPC patients transfused with 1847 CMVuntested (CMV-U), leukoreduced components. We would like to share our institution’s experience in 100 CMV allogeneic HPC patients transfused with 6465 CMV-N and CMV-U cellular components. Before July 2006, the University of Michigan provided, when available, CMV-N components to all CMV HPC transplant recipients. A preliminary 12-month retrospective review showed that 52% (14/27, 2004) had received leukoreduced, CMV-U blood components due to shortages in CMV-N products, with no cases of TT-CMV. As a result, the transfusion policy was changed in mid-2006 to provide leukoreduced, CMV-U products for all HPC patients, regardless of pretransplant CMV status. For quality assurance, we monitored the CMV conversion rate over a 36-month period (January 2005 to December 2007) covering an 18-month period before and after the change in transfusion policy (Table 1). Per institutional practice, CMV-N patients were screened for CMV IgG every 1 to 2 months pretransplant, followed by regular testing for CMV nucleic acid testing (NAT) after transplant. The minimum posttransplant follow-up was 12 months. All blood products were provided by Southeast Michigan American Red Cross (Detroit, MI) and were leukoreduced after storage (Leukotrap RC system, Pall Corp., Port Washington, CA). As shown in Table 1, 100 patients were available for analysis and included both adult and pediatric patients. All patients were CMV-N before transplant, received a CMV allogeneic HPC transplant, and underwent weekly posttransplant CMV NAT monitoring. Except for sex, there were no significant differences in patient demographics, transplant type, or transfusion support in the two study cohorts. In the CMV-N policy period, only 11% to 15% of cellular components were CMV-N, with most patients receiving a mix of CMV-N and CMV-U. Only five patients received 100% CMV-N products. All five patients had low transfusion needs, requiring 2 to 10 red blood cell (RBC) and two to three platelet (PLT) transfusions. Two adult male patients had a single positive test for CMV IgG at 3 and 5 weeks after transplant, respectively (Table 1). Both patients tested negative for CMV IgM and CMV NAT and had no evidence of clinical CMV infection. Each patient received between 42 and 45 CMV-U cellular components in the weeks before seroconversion: neither patient had received intravenous immune globulin before CMV IgG testing. There were no CMV seroconversions in the CMV-U period. The overall CMV IgG seroconversion rate was 2% per patient and 0.03% per unit, which is comparable to the findings by Bowden and coworkers (2.4% per patient, 0.023% per component). The rate of confirmed TT-CMV was 0%, consistent with the study by Thiele and coworkers and lower than that reported by Wu and coworkers (6.5% per patient, 0.23% per CMV-positive component). As discussed by Thiele and Wu, the CMV IgG detected in our patients likely represents passive antibody from recent transfusions. Our findings confirm those of Thiele and affirm the equivalent safety of CMV-U, leukoreduced components in CMV allogeneic HPC patients. The absence of clinical TT-CMV infection in our study and that by Thiele and coworkers, despite the combined transfusion of nearly 8000 CMV-U, leukoreduced components, contradicts sentiments from a past multivariate analysis, which advocated continued provision of CMV-N and leukoreduced components for CMV transplant patients. The improved safety of CMV-N over CMV-U, leukoreduced is also not supported by a recent large prospective study of 34,000 blood donors. Ziemann and colleagues found CMV viremia only among newly seroconverted donors and a few CMV-N donors, arguing that CMV-N components may present the higher risk of TT-CMV due to passive transfusion of free CMV DNA. In summary, policies stipulating leukoreduced, CMV-N components in CMV allogeneic HPC patients do not confer additional safety and are limited by product shortages and significant transfusion support required by many allogeneic HPC patients.