Introduction: CMV infection is associated with direct and indirect morbidities in BMT recipients. Understanding cumulative morbidity associated with infections is difficult, as time-to-first event analyses do not represent recurrent events. We characterized infectious outcomes and non-relapse mortality (NRM) in a large cohort of haploidentical BMT recipients at Johns Hopkins Hospital (JHH) using novel analytic methods to detail recurrent events and estimate cumulative risks.

Material and Methods: Demographics, transplant variables, CMV reactivation (DNAemia) and disease, and other outcomes (infections, GVHD and hospitalizations) were collected on sequential haploidentical BMT recipients at JHH (2004-2015). Infectious events were longitudinally depicted. Cox proportional hazards models were used to estimate effects of infections on 1-year survival, adjusted for age, disease related index (DRI), GVHD, time to engraftment, and transplant year. Survival and NRM differences were assessed by Kaplan Meier and competing risks functions, respectively.

Results:683 people who received non-myeloablative haploidentical BMT using PTCy, mycophenolate and tacrolimus were included. CMV reactivation occurred in 268 (39%), with organ disease documented in 19 (2.8%). Recurrent reactivation was recorded in 83 (12% of cohort), predominating in seropositive recipients of negative donors (D-/R+). Neutropenia occurred in 24/123 (28%) of subjects with ganciclovir-treated 1st reactivation, increasing to 41% and 60% with 2nd and 3rd episodes. Patients who had CMV reactivation had higher mean numbers of bacteremias (0.48 vs. 0.36, p=0.006), clinical pneumonias (0.24 vs. 0.18, p= 0.03) invasive fungal infections (IFI, 0.23 vs. 0.15, p= 0.05), non-CMV viral infections (0.53 vs. 0.39, p=0.008), and hospitalizations (1.68 vs. 1.27, p=0.001). Patients with CMV reactivation had lower 1-year survival (68% vs. 76%, p=0.016), with higher NRM in competing risk analyses. Associations between CMV reactivation and other infections were appreciated in longitudinal analyses. In regression analysis, risks for death were associated with DRI (HR 4.3 - 2.5, p<0.001), older age (HR 1.1, p=0.02), IFI (HR 1.7, p=0.003), pneumonias (HR 1.9, p<0.001), and bacteremias (HR 1.7, p<0.001).

Conclusion: CMV reactivation and recurrence is common in seropositive haploidentical BMT recipients, especially with seronegative donors, but disease is rare. Longitudinal modeling of reactivation and treatment-associated neutropenia suggests that reactivation is implicated in secondary infections that increase risks for death. These findings may instruct better-targeted prophylaxis strategies.

Disclosures

Marr:MycoMed Technologies: Consultancy, Equity Ownership, Patents & Royalties.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution