MMF Is Associated with Worse OS in CMV Seropositive AML Patients Undergoing MUD HCT with Calcineurin-Inhibitor-Based Gvhd Prophylaxis

Background: Calcineurin-inhibitor (CNI) with mycophenolate mofetil (MMF) or methotrexate (MTX) is commonly used as graft-versus-host disease (GVHD) prophylaxis for HLA-matched hematopoietic cell transplantation (HCT) but MMF, which targets both T- and B-lymphocytes, is associated with higher risk for cytomegalovirus (CMV) reactivation [Blood Adv.2023;7(8):1394-1403]. We investigated whether overall survival (OS) differed with MMF- vs MTX-based GVHD prophylaxis according to recipient CMV serostatus and underling HCT disease indication.

Methods: We evaluated 3105 patients with acute myeloid leukemia (AML, n=1572), acute lymphoblastic leukemia (ALL, n=453) or myelodysplastic neoplasms (MDS, n=1080) who underwent HLA-matched unrelated donor (MUD) peripheral blood (PB) HCT from 2008-17 using a publicly available Center for International Blood and Marrow Transplant Research (CIBMTR) dataset. Four groups were compared: MTX/ CMV+ (n=1527), MTX/ CMV- (n=916), MMF/ CMV+ (n=395), and MMF/ CMV- (n=267). All patients also received tacrolimus; none received post-transplant cyclophosphamide (PTCy). Primary outcome was OS.

Results: The median (interquartile range) age was 57 (45-64), 56 (45-64), 62 (54-68), and 61 (50-67) years in the respective 4 groups. Within each group, most patients had AML or MDS. Disease status was mostly early or intermediate (51-63%). About half had HCT-CI > 2. In vivo T cell depletion was used in 41%, 65%, 48% and 52%, respectively. More patients in the MTX groups received myeloablative conditioning (58% and 55%, respectively), while reduced intensity/non-myeloablative conditioning was more common in the MMF groups (73% and 65%, respectively).

In multivariate models stratified by disease, we noted significantly worse OS in CMV+ patients who received MMF, but this effect was restricted to AML. Among AML patients, OS was only 31% (95% confidence interval [CI] 25-37) among the MMF/CMV+ group, but was higher and not significantly different for the MMF/CMV- (54%, 95% CI 45-63) and MTX/CMV+ (51%, 95% CI 48-55) groups compared to the MTX/CMV- group (58%, 95% CI 53-63) [Figure 1, Table 1] In ALL, GVHD prophylaxis type/CMV serostatus had no significant association with OS. In MDS, CMV+ patients had worse OS (HR 1.2, 95% CI 1.01-1.4, p=0.03) than CMV- patients irrespective of GVHD prophylaxis (MMF/CMV+ vs MTX/CMV+: HR=1.1, 95% CI 0.8-1.4, p=0.5). Donor CMV serostatus had no impact on these associations. Data on CMV reactivation and causes of death were not available; however, our data suggest that the inferior OS in AML was driven mostly by higher non-relapse mortality and some increase in relapse (data not shown).

To evaluate how the use of PTCy prophylaxis might impact the interaction between MMF and recipient CMV+ serostatus, we next analyzed a separate CIBMTR MUD PB-HCT cohort that received PTCy/CNI/MMF prophylaxis (N=242; AML, n=136, ALL, n=42; MDS, n=64) and assessed the effect of CMV serostatus on OS. A majority (95%) received HCT from 2015-18. Our results showed that CMV+ recipients had a significantly worse OS than CMV- recipients, but only in AML (HR 2.7, 95% CI 1.1-6.4, p=0.03) and not in ALL (HR 0.4, 95% CI 0.1-1.5, p=0.2), or MDS (HR 0.7, 95% CI 0.3-1.5, p=0.3). As all patients in this cohort received PTCy/CNI/MMF, we were unable to assess the independent impact of MMF and CMV serostatus in this setting. Nevertheless, these data suggest similar trends as seen in the CNI cohort (without PTCy), where the inferior OS in the MMF/CMV+ group was noted only in AML patients.

Conclusion: Our data suggest that the use of MMF with CNI-based prophylaxis (without PTCy) should be avoided in AML CMV+ patients. Why this adverse effect of CMV seropositivity and MMF prophylaxis is restricted to AML patients, and not seen in ALL, and whether the same applies to PTCy-prophylaxis with/without MMF, needs further investigation but could be due to CMV-induced expansion of NK cells [Béziat. Blood. 2013; Foley. Blood. 2012] which mediates alloreactivity against AML but not ALL [Ruggeri Blood 1999], and is suppressed by MMF more than MTX [Ohata. BBMT. 2011].