Many studies have evaluated the impact of cytomegalovirus (CMV) reactivation after allogeneic hematopoietic stem cell transplantation (allo-HSCT) showing a significant association with reduced risk of relapse while other studies failed to show the same results. On the other hand, other frequent viral infections or reactivations like Epstein-Barr virus (EBV), Human Herpesvirus 6 (HHV6) and BK virus (BK-V) have not been evaluated in the same context and/or together with CMV. The aim of this study is to evaluate the impact of CMV, EBV, HHV6 and BK-V reactivation in acute leukemia (AL) patients receiving allo-HSCT in complete remission (CR) and evaluate their impact alone or associated on disease relapse.

Patients and methods:

We evaluated 158 consecutive AL patients who received allo-HSCT at our center between January 2008 and June 2015; 95 (60%) were males, median age was 46 years (range: 18-67), 101 (64%) had acute myeloid leukemia (AML) and 57 (36%) had acute lymphoblastic leukemia (ALL). At transplantation, 114 (72%) patients were in first complete response (CR1) and 44 (28%) were in second CR (CR2). Conditioning regimen was myeloablative for 97 (61%) patients and of reduced intensity in 61 (39%) patients. DNA levels of CMV, EBV, HHV6 and BK-V were detected by quantitative RQ-PCR after weekly monitoring up to 3 months after allo-HSCT. CMV-DNA, EBV-DNA, HHV6-DNA or BK-V-DNA was considered positive when the copies exceeded 1000 copies/ml.


Among 158 patients, 45 (28%) patients had CMV reactivation after a median time of 40 days (6-72) after allo-HSCT; 53 (34%) patients had EBV reactivation after a median time of 42 days (18-91) after allo-HSCT; 33 (21%) patients had HHV6 reactivation after a median time of 24 days (6-69) and 30 (19%) patients had BK-V reactivation after a median time of 32 days (6-60) after allo-HSCT. There was 61 (39%) patients who did not have any viral reactivation, 46 (29%) patients who had only one reactivation (no matter the type), and 51 (32%) patients who had at least two or more co-reactivations (group 1). When evaluating relapse incidence, we found that patients who had only one reactivation were not significantly different from those who did not have any reactivation, with a comparable relapse incidence, thus, we analyzed those patients as one group (group 2, n=107). The cumulative incidence of relapse at 6 months for group 1 and group 2 was 2% (95% CI: 0.2-9.4) and 15.9 % (95%CI: 9.7-23.5) respectively, this result remained significantly different when stratifying on the type of conditioning, in RIC: 15.8% (95%CI: 3.7-35.7) vs 21.4% (95%CI: 10.5-35) respectively; and in MAC: 0 % vs 15.4% (95%CI: 7.8-25.3) respectively, p=0.04. This finding was confirmed independently of the presence of GVHD or not.


We demonstrate here for the first time, in this homogeneous population, that co-reactivation of multiple viruses after allo-HSCT could lead to a lower incidence of relapse in AL patients in CR; this can be explained as a stimulation of the immune system via both the functional activation and amplification of NK compartment. This could also explain some conflicting results in studies evaluating single viral reactivation alone without consideration of other reactivations. These encouraging results should be validated at a larger multicenter level.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.