BACKGROUND: Concurrent cytomegalovirus infection (CMVi) and gastrointestinal graft-versus-host disease (GI-GVHD) poses significant risks for increased morbidity and mortality in patients undergoing allogeneic hematopoietic cell transplantation (allo-HCT). Primary prophylaxis with letermovir therapy has been shown to decrease the risk of CMV reactivation, but studies examining this relationship after GI-GVHD are lacking. We reviewed our center's outcomes associated with concomitant CMVi and GI-GVHD before and after our adopting the use of letermovir therapy for CMV prophylaxis in 2017.

METHODS: This was a single-center, retrospective study of allo-HCT patients who developed GI-GVHD and CMVi between June 2013 and June 2021. CMVi was defined as detection of CMV in the blood or detection of CMV in sampled tissue. CMV colitis was defined as biopsy-proven tissue-invasive CMV disease. The primary outcome was one year survival.

RESULTS: We evaluated 43 allo-HCT patients who had concomitant CMVi and GI-GVHD. Out of 43 patients, 40 of them (93 %) had a high risk serostatus for CMVi (recipient seropositive). CMV colitis was confirmed by biopsy in 18 patients (42 %) and was clinically suspected in 20 patients (47 %). Twenty-five patients (58 %) developed CMV viremia, and 18 of them developed concomitant biopsy proven CMV colitis. Nine patients (21 %) received letermovir therapy for primary CMV prophylaxis, and 7 of these 9 patients (78 %) developed breakthrough CMVi while receiving prophylaxis. The median time from HCT to CMVi was similar between the group that received letermovir prophylaxis and the group that did not. Median peak CMV levels were lower in the prophylaxis group. Overall survival rates at one and five years were 65 % and 21 %, respectively, with the mortality reaching 25 % at 164 days and 50 % at 480 days. There was no statistically significant difference in one year survival between patients with CMV viremia but without colitis compared to those with CMV viremia and colitis (p = 0.648, 95 % CI 0.3-1.57). One-year survival was also not statistically different between patients who received letermovir prophylaxis compared to those who did not (p = 0.250, 95 % CI 0.60-6.97) or between patients with high grade GI-GVHD (grade 3-4) and low-grade GI-GVHD (grade 1-2; p = 0.277, 95 % CI 0.64-4.83).

CONCLUSION: In this high-risk cohort with GI-GVHD and CMVi, the peak viral load was lower in the group that received letermovir prophylaxis, but the median time to onset of CMVi and survival were not statistically different. The majority of patients in the letermovir prophylaxis group developed CMVi while on letermovir. Further investigations with larger sample sizes may better assess the impact of letermovir therapy on patient survival and the development of CMVi outcomes in the setting of GI-GVHD in the era of pre-emptive prophylactic antiviral therapy.