Letermovir Prophylaxis Reduces the Risk of Cytomegalovirus Infection in HLA-Matched and Mismatched Allogeneic Hematopoietic Cell Transplant Patients Receiving Post-Transplant Cyclophosphamide

Introduction

Post-transplant cyclophosphamide (PTCy) is a widely adopted graft-versus-host disease (GVHD) prophylactic strategy for HLA-haploidentical donor hematopoietic cell transplant (HCT) and has recently been proven superior to calcineurin inhibitor-based GVHD prophylaxis in matched donor transplants (Bolaños-Meade. NEJM. 2023). Nonetheless, both HLA mismatched donor and PTCy are independent risk factors for cytomegalovirus (CMV) infections, which may result in morbidity and mortality (Goldsmith. Blood. 2021). Letermovir is effective in reducing the risk of clinically significant CMV infection (csCMVi) and improving survival, though data supporting its use in patients receiving PTCy is limited (Marty, NEJM. 2017). This study aims to assess the real-world effectiveness of letermovir in preventing csCMVi in this high-risk population.

Methods

This retrospective study included 331 adult patients who underwent allogeneic HCT with PTCy at the Moffitt Cancer Center between 2014 and 2022. The primary endpoint was the incidence of csCMVi, defined as CMV viremia with CMV PCR >650 international units (IU) or evidence of CMV organ disease. Secondary endpoints included adverse effects attributed to letermovir, time to neutrophil engraftment, incidence of chronic GVHD, non-relapse mortality (NRM), relapse free survival (RFS), relapse, and overall survival (OS). Endpoints were compared across three groups by recipient serostatus: 1) CMV negative (CMV-), CMV positive and did not receive letermovir (CMV+/No LET), and 3) CMV positive and received letermovir (CMV+/LET).

Results

Of 331 included patients, 61 (18%) were CMV+ and received letermovir due to high-risk categorization after adoption of letermovir as standard of practice. Of 270 patients who did not receive letermovir, 148 (45%) were CMV+ and either underwent HCT before letemovir was adopted or could not obtain due to cost/insurance issues, and 122 (37%) were CMV-. The median age was 61 and 78% of patients identified as Caucasian. Donors included haploidentical (40%), matched unrelated (38%), mismatched unrelated (12%), and matched sibling (11%), with 96% receiving peripheral blood stem cells. Median follow-up time was 24 months.

The cumulative incidence of csCMVi at day+100 was 11% (95% CI: 5-21%) in CMV+/LET, 45% (95% CI: 37-53%) in CMV+/No LET and 7% (95% CI: 4-13%) in CMV- patients (p<0.01). In multivariate analysis (MVA), CMV+/No LET had a higher risk of csCMVi compared to CMV+/LET (HR=2.96, 95% CI: 1.66-5.27, p<0.01). CMV- had a lower risk of csCMVi (HR=0.41, 95% CI: 0.18-0.92, p=0.03) compared to CMV+/LET. CMV disease was reported in 18 patients across the entire cohort. No grade 4 CMV disease was identified in CMV+/LET patients compared to 3 patients in CMV+/No LET. At day+100, the cumulative incidence of CMV disease was 2% (95% CI: 0.1-8%) in CMV+/LET, 10% (95% CI: 6-16%) in CMV+/No LET and 1% (95% CI: 0.1-4%) in CMV- patients (p<0.01).

For patients with csCMVi, the median duration of antiviral treatment was 22 days for CMV+/LET, 33 days for CMV+/No LET, and 30 days for CMV- patients (p=0.14). The median time from date of transplant to antiviral treatment was 97 days for CMV+/LET, 42 days for CMV+/No LET, and 54 days for CMV- patients (p=0.01). Of 61 patients receiving letermovir, 8 (13%) discontinued letermovir early due to treatment related adverse events (gastrointestinal toxicity, cytopenias) at a median of 82 days (range: 37-90) with no reported occurrences of csCMVi afterwards. Incidence of csCMVi at 1 year increased in CMV+/LET from 11% at day+100 to 21% at 1 year.

There were no associations between letermovir and recipient CMV serostatus on time to neutrophil engraftment, chronic GVHD, 1-year NRM, relapse, RFS, or OS.

Conclusion

Our findings support the use of letermovir prophylaxis during the first 100 days post-allogeneic HCT with PTCy in both HLA matched and mismatched donors to reduce the incidence of csCMVi. For patients receiving letermovir who developed csCMVi, time to csCMVi was delayed and the duration of therapy was reduced, thus decreasing exposure to antiviral therapy and minimizing associated adverse effects. Future studies should evaluate strategies to address late onset csCMVi beyond day+100, such as continued PCR monitoring or extended duration of prophylaxis. As PTCy is increasingly utilized, letermovir prophylaxis may address a major limitation of this transplant platform and improve patient outcomes.