Abstract 3110


Monitoring and (preemptive) treatment of cytomegalovirus (CMV) infection or disease have been implemented in clinical routine algorithms in patients after allogeneic stem transplantation (allo-SCT) in the early nineties and resulted in a dramatic decrease of CMV disease from 20–30% to less than 5% in recent years. However, a significantly reduced relapse incidence (RI) in acute myeloid leukemia (AML) patients with vs. without early CMV replication has been reported just recently. Larger analyses evaluating the prognostic impact of recipient and donor CMV serostatus - as one of the most important risk factors for CMV infection and disease - are lacking at the era of preemptive treatment of CMV infection.

Design and methods:

We analyzed the prognostic impact of the (pretransplant) recipient and donor CMV serostatus in 16 628 adult de novo acute leukemia patients who underwent allo-SCT between 1998 and 2009 and are documented in the European Bone Marrow Transplantation (EBMT) database (Cohort I). Additionally, 544 patients with secondary AML (e.g. following myeloproliferative neoplasms) who underwent allo-SCT and were reported to the German Registry for Stem Cell Transplantation (DRST) were analyzed according to the prognostic impact of the CMV serostatus (Cohort II). Finally, 97 acute leukemia patients allografted at the Charité University Hospital (Berlin, Germany) were studied more in detail including other parameters such as kinetics of CMV infection and hematopoietic recipient/donor chimerism (Cohort III).


Recipient and/or donor CMV seropositivity was associated with a reduced 2-year leukemia-free survival (LFS) (44% vs. 49%, p<0.001) and overall survival (OS) (50% vs. 56%, p<0.001) in de novo acute leukemia patients (Cohort I). These observations were mainly due to an increased non-relapse mortality (NRM) (23% vs. 20%, p<0.001) and strongest in recipients allografted from a non-HLA identical sibling. However, also higher RI was noted especially for acute lymphoblastic leukemia (ALL), resulting in a markedly decreased 2-year OS (46% vs. 55%, p<0.001) in these patients. The negative prognostic impact of recipient and/or donor CMV seropositivity with respect to LFS, OS, NRM and RI remained in a multivariate Cox regression model including different other parameters (e.g. conditioning type). Conversely, the donor CMV serostatus had no significant prognostic impact among CMV seropositive recipients. Recipient CMV seropositivity was further associated with a significantly increased frequency of death due to interstitial pneumonitis, but decreased incidence of both chronic and II-IV° acute graft-versus-host disease (28% vs. 31%, p<0.001). However, it had no significant impact on the incidence of graft failure. Also in secondary AML (Cohort II) recipient and donor CMV seronegativity was associated with a superior 2-year LFS (48% vs. 41%) and OS (54% vs. 44%), but decreased RI (27% vs. 30%) and NRM (25% vs. 31%). The negative impact of CMV seropositivity of the recipient and/or the donor with respect to OS likewise remained in this cohort of secondary AML in a multivariate Cox regression model (Hazard ratio 1.4, p=0.03). Finally, analysis of cohort III showed that CMV seropositive vs. CMV seronegative patients had a significantly reduced OS even in the absence of CMV antigenemia and less frequently had a complete day +30 hematopoietic chimerism (68% vs. 92%, p=0.02).


The recipient and donor CMV serostatus still has an important prognostic effect in acute leukemia patients after allo-SCT - particularly in ALL and those allografted from a non-HLA identical sibling - despite the implementation of most sophisticated techniques for monitoring and (preemptive) treatment of CMV in recent years. The increased RI in cases with CMV seropositivity of the recipient and/or the donor argues against a clinically important ‘virus-versus-leukemia effect′ which has been suggested just recently.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.