CMV: persistent nemesis for SCT

In this issue of Blood, Teira et al have shown, in a large database analysis, that cytomegalovirus (CMV) still kills stem cell transplant recipients but does not prevent relapse of malignancy.¹ 

CMV has been perceived as a less threatening viral cause of morbidity and mortality after allogeneic stem cell transplantation (SCT) in recent years, due to improved treatment options that include new pharmacologic treatment approaches²  and development of adoptive T-cell transfer.^3,4  Conflicting results have been published regarding whether CMV infection itself or the T-cell response against CMV influence the graft-versus-leukemia (GVL) effect post-SCT. In general, CMV is likely to influence the T-cell function because it is a persistent infection and has potent stimulatory and inhibitory effects on the immune system, especially on T cells. T cells are the most potent effectors of graft-versus-malignancy effects that decrease the risk of posttransplantation malignancy relapse. The present registry study involved 9469 patients from the Center for International Blood and Marrow Transplant Research (CIBMTR) database. The authors investigated how CMV serostatus and early reactivation until day 100 post-SCT has an influence on overall survival (OS), disease relapse, and nonrelapse mortality (NRM). The analysis included the 4 disease categories acute myelogenous leukemia (AML), acute lymphoblastic leukemia (ALL), myelodysplastic syndrome, and chronic myeloid leukemia. CMV reactivation resulted in increased NRM and decreased disease-free survival (DFS), translating into poorer OS. Positive CMV serology in donor and/or recipient also increased NRM and decreased DFS or OS in patients with AML or ALL. Early CMV reactivation or positive CMV serology had no effect on the risk of hematologic disease relapse except in patients with ALL in whom positive CMV serology was associated with an increased risk of disease relapse. That CMV reactivation increases NRM is common knowledge among transplant doctors. The finding that CMV disease still results in transplant-related mortality is not novel, but it strongly emphasizes that treatment possibilities are far from satisfactory. This study emphasizes the need for clinical research in this field. New treatment approaches such as adoptive T-cell transfer have been shown to be feasible for several viral pathogens,^5,6  but advance-phase clinical efficacy studies are rare and commercial interest is relatively limited. Therefore, investigator-driven clinical trials are needed to improve these treatment options.

This study by Teira et al advances the field by analyzing the relation of CMV reactivation and relapse of malignancy. In theory, the influence of CMV on the relapse rate could be positive or negative. Several earlier studies suggested that CMV reactivation improved the relapse rate. In contrast, Teira et al demonstrate that CMV reactivation decreases DFS. Why? The mechanism(s) by which latent viral infections such as CMV might actually interact with leukemia relapse have not been identified, but numerous effects of CMV on the human immune system have been described (see figure). CMV is a lytic infection and has a direct cytopathic effect. CMV has immunosuppressive effects as demonstrated by an influence on HLA expression and clinically by an increased rate of other infections associated with CMV reactivation. CMV is a potent activator of the first-line defense and may induce expansion of mature NK cells and γδT cells, cross-reactive to CMV and leukemia cells. In humans, CMV infects APCs and could therefore take over the control of the adaptive immune response.⁷  Because pp65, the most immunodominant CMV antigen, can induce strong T-cell responses, efficient host protection as well as immunopathology may be due to the T-cell response against CMV. However, due to the strong and persistent T-cell stimulation, T-cell exhaustion and senescence may be an issue, especially in the elderly patient. All of these effects may play a role in dictating how CMV infection has an influential effect on relapse of malignant disease.

This large database analysis provides tremendous descriptive strength and reveals sound information regarding the relative risks. But it also demonstrates the weakness of this type of retrospective study to identify the mechanisms underlying the described findings. The differences in the disease groups in this study show that an interaction may exist between viral infection and leukemia relapse. But we still do not understand the interplay of how this occurs. Future research will need to explore these interactions. The concepts and hypotheses of infection and the GVL response will have to be revisited and we may need to ask different questions to reveal the mechanisms of GVL. Clearly, while looking for answers about CMV, only more questions are raised.