The Prognostic Impact of the Donor/Recipient CMV Serostatus in Patients with Different Chronic Hematological Disorders after Allogeneic Hematopoietic Stem Cell Transplantation: A Report from the Infectious Diseases Working Party of EBMT

Introduction

Cytomegalovirus (CMV) seropositivity of the recipient and/or the donor is still associated with a major overall survival (OS) decline in acute leukemia patients undergoing allogeneic hematopoietic stem cell transplantation (allo-HSCT) despite the fact that the incidence of CMV disease has markedly been reduced by preemptive treatment of CMV infection. However, CMV infection after allo-HSCT might also have a favourable effect by reducing the relapse incidence (RI), at least in acute myeloid leukemia. Less is known about the prognostic impact of the CMV serostatus in patients with different chronic hematological disorders undergoing allo-HSCT.

Methods

We analyzed the prognostic impact of the CMV serostatus in a large cohort of 6968 patients who underwent allo-HSCT between the year 2005 and the year 2016 and who were documented in the database of the European Bone Marrow Transplantation Group. Patients with the following diseases were included: Myelodysplastic syndrome (n=2326), multiple myeloma (n=1788), myeloproliferative neoplasm other than Philadelphia chromosome-positive chronic myeloid leukemia (Ph+ CML) (n=979), follicular lymphoma (n=673), Ph+ CML (n=458), mantle cell lymphoma (n=433) and B cell chronic lymphocytic leukemia (n=311). The primary endpoint was 2-year progression-free survival (PFS) of patients with CMV seronegativity of both the donor and the recipient vs the remaining patients.

Results

The analysis encompassed 2026 patients (29%) with CMV seronegativity of both the donor and the recipient, 2510 patients (36%) with CMV seropositivity of both of them in addition to 815 (12%) and 1617 patients (23%) with CMV seropositivity only of the donor or the recipient, respectively.

Patients with CMV seronegativity of both the donor and the recipient had a significantly superior 2-year PFS (52.5%) compared to patients with CMV seropositivity of the donor and/or the recipient (49.5%, p=0.03). In addition, 2-year OS was significantly shorter for patients with CMV seropositivity of the donor and/or the recipient (62.3%) than for those cases with CMV seronegativity of both of them (65.4%, p=0.01). The negative prognostic impact of CMV seropositivity of the donor and/or the recipient remained in a multivariate Cox regression analysis including other important variables (e.g., T cell depletion, donor type) with a significantly reduced PFS (HR 1.09, 95% CI 1.01-1.17, p=0.03) and OS (HR 1.13, 95% CI 1.04-1.23, p=0.003).

CMV-seropositive patients allografted from a CMV-seronegative donor showed a particular unfavorable OS compared to patients with CMV seronegativity of both of them (HR 1.21, 95% CI 1.10-1.34) whereas OS was only slightly reduced in patients with CMV seropositivity of only the donor (HR 1.03, 95% CI 0.91-1.17) or of both of them (HR 1.07, 95% CI 0.98-1.17) (p=0.001). Patients with CMV seropositivity of the donor and/or the recipient further showed a significantly increased non-relapse mortality (NRM) in a multivariate Cox regression analysis (HR 1.15, 95% CI 1.03-1.28, p=0.02). CMV-seropositive patients with a CMV-seronegative donor had the highest NRM (HR 1.22, 95% CI 1.07-1.39), followed by CMV-seropositive recipients with CMV-seropositive donors (HR 1.05, 95% CI 0.93-1.19) and finally CMV-seronegative recipients with CMV-seropositive donors (HR 0.99, 95% CI 0.83-1.18) (p=0.01).

The donor/recipient CMV serostatus had no significant impact on acute or chronic graft- vs -host disease, donor/recipient hematopoietic chimerism or RI. Interaction analyses revealed that there are not relevant differences on the prognostic impact of the CMV serostatus among different underlying diseases and donor types.

Conclusions

Our data show that the CMV serostatus still has a significant prognostic impact in patients with different chronic hematological disorders undergoing allo-HSCT, despite the implementation of sophisticated antiviral strategies in recent years. CMV-seronegative donors should be avoided whenever possible for CMV-seropositive recipients whereas a CMV-seropositive donor might be adequate for a CMV-seronegative recipient. The fact that CMV seropositivity of the donor and/or the recipient - as the most important risk factor for CMV infection - had no significant impact on the RI argues against a clinically relevant `virus vs malignancy´ effect in these disorders.