Abstract
Background: The use of letermovir (LTV) in the allogeneic hematopoietic cell transplant has dramatically changed the incidence of cytomegalovirus reactivation (CMVr) with direct effect also on overall survival (OS). Such effect has been retrospectively observed also in the setting of haploidentical alloHCT (Haplo) using post-transplant cyclophosphamide (PTCy), a very-high risk landscape for CMVr and disease. Our hypothesis is that LTV could be associated in this high-risk setting to a reduction of non-relapse mortality (NRM), and to further indirect improvements such as graft-versus-host-disease (GVHD) or other viruses' reactivation. Moreover, in case of CMVr, a clinically different pattern and type of CMVr is expected.
Methods: We performed a monocentric, retrospective study using clinical data from the Institut Català d'Oncologia (Barcelona, Spain). We included adult patients (> 18y.o), with oncohematological diseases, receiving a first Haplo, using myeloablativemor reduced intensity conditioning (RIC) based on thiotepa/fludarabine/busulphan (TBF), peripheral blood stem cells as graft source, a GVHD prophylaxis based on PTCy. Inclusion period was from January 2019 to December 2023. The primary endpoint was NRM. Secondary endpoints were progression-free survival (PFS), OS, relapse/progression, acute grade 2-4 GVHD, chronic GVHD (all grade), CMV/EBV/BK/JC/ADV/HHV6 reactivations. Probabilities of OS and PFS were calculated using Kaplan-Meier estimator method. NRM, relapse/progression, GVHD and viral reactivations were calculated as cumulative incidences. To better describe the clinical phenotype of CMVr, a further sub analysis restricted to patients with CMVr was pre-planned comparing: days of first CMVr from Haplo, duration of first CMVr, number of CMVr, days of hospitalization due to CMV, number of CMV treatments, cytopenias due to CMV treatments, renal insufficiency due to CMV treatments, CMV disease prevalence, CMV-related deaths.
Results: Seventy consecutive patients were included. Median age at Haplo was 56y.o (range 23-70), female patients were 49% of the population, complete response at Haplo was reached in 71%, median donor age was 36y.o (range 17-63), ECOG>0 35%, HCT-CI> 2 49% and RIC TBF 71%. Acute leukemias were the most prevalent disease (48%) followed by non-Hodgkin lymphomas (27%) and myelodiplastic/myeloproliferative syndromes (16%). Serological positivity for CMV was 100% for patients and 26% for donors. No significant differences in clinical characterisstics were found between LTV (n=17) and no LTV (n=53) groups. Median follow-up among survivors was 48 months (range 16-75m). At +36m, no differences were reported between LTV and no LTV groups in terms of NRM (18% vs 28%, p=0.55), PFS (75% vs 57%, p=0.24), OS (70% vs 57%, p=0.40), relapse/progression (12% vs 19%, p=0.50), acute grade 2-4 GVHD (35% vs 32%, p=0.65), chronic all-grade GVHD (19% vs 17%, p=0.97). ECOG >0 at Haplo was the only factor associated with worse NRM (HR: 3.93, 95%CI: 1.36-11-35, p=0.01), shorter PFS (HR: 2.31, 95%CI:1.07-4.98, p=0.03) and decreased OS (HR: 3.52, 95%CI:1.64-7-54, p<0.01). At +6m post Haplo, LTV was associated with a significant reduction in CMVr (35% vs 79%, p<0.01), confirmed also in multivariate analysis (HR: 0.27, 95%CI:0.12-0.61, p<0.01). No effect on EBV/JC/BK/ADV/HHV6 reactivations was observed. Restricting the analysis only to patients with CMVr (LTV n=6, no LTV n=42), we found those who received LTV prophylaxis had a delayed first CMVr day (+58 vs +35, p=0.02). However, no significant differences were observed in terms of duration of first CMVr (33 vs 30 days, p=0.70), total episodes of CMVr (1.33 vs 1.57, p=0.55), days of hospitalization required for CMV therapy (7 vs 13, p=0.44), total number of anti-CMV drugs used (1.3 vs 1.6, p=0.56), and presence of cytopenias (83% vs 86%, p=0.84) or renal insufficiency at time of CMV treatment initiation (33% vs 27%, p=0.75). CMV disease was observed only in 5 patients not treated with LTV. Of these, 2 patients died of CMV disease. None of those who received prophylaxis developed CMV disease.
Conclusions: LTV is associated to a significant reduction in CMVr also in a high-risk population (Haplo with PTCy) without observing significant differences in terms of NRM. For those patients experiencing CMVr, the previous use of LTV was associated to a delayed reactivation and absence of CMV disease or death. Larger cohort of patients will be necessary to confirm these results
