Clinical and Economic Burden of Multiple Double-Stranded DNA Viral Infections after Allogeneic Hematopoietic Cell Transplantation

Abstract

Conditioning regimens for allogeneic hematopoietic cell transplantation (allo-HCT) are immunosuppressive and increase the risk for reactivation of and infection with double-stranded DNA (dsDNA) viruses, which contribute to morbidity and mortality after allo-HCT. This retrospective observational study evaluated the association of dsDNA viral infections with clinical outcomes, health resource utilization (HRU), and health care reimbursement after allo-HCT. Patients who underwent allo-HCT between 2012 and 2017 were identified from a US open-source claims database (Decision Resource Group Real-World Evidence Data Repository; n=13,363) and categorized according to the presence or absence of dsDNA viral infection, defined as having ≥1 diagnosis code for cytomegalovirus (CMV), adenovirus (AdV), human herpesvirus 6 (HHV-6), or BK virus (BKV)/Epstein-Barr virus (EBV)/John Cunningham virus (JCV) (grouped together given a lack of specific diagnoses codes) within 1 year after allo-HCT. Only first allo-HCT data were used in patients who underwent multiple procedures. Study outcomes included clinical outcomes (eg, time to all-cause mortality, new diagnosis of renal impairment), HRU (hospital and intensive care unit length of stay [LOS], readmission rates), and health care reimbursement (total, inpatient, and outpatient costs as reported reimbursements from insurance claims). For all outcomes, patients were stratified by the presence/absence of any dsDNA viral infection and number (none, 1, 2, or ≥3) and type(s) of infection. The effect of graft-versus-host disease (GVHD) was assessed as well. Twenty-nine percent of patients were diagnosed with CMV, 13% with BKV/EBV/JCV, 5% with AdV, and 4% with HHV-6 in the year following their first allo-HCT. A single dsDNA viral infection was documented in 30% of individuals, 2 in 8%, and ≥3 in 2%. Patients with no viral infections had an overall hospital LOS (index hospitalization plus readmissions) of 41.3 days and a total health care reimbursement of $266,345. These numbers increased for every additional viral infection, regardless of the presence or absence of GVHD; the overall hospital LOS was 61.4 days and total healthcare reimbursement was $431,614 in patients with 1 viral infection, 77.0 days and $639,097 in patients with 2 viral infections, and 103.3 days and $964,378 in patients with ≥3 viral infections. An increase in the number of dsDNA viral infections was associated with a significantly higher adjusted hazard of all-cause mortality (1 versus 0 dsDNA viral infections: hazard ratio [HR], 1.5; [95% confidence interval (CI), 1.3 to 1.6]; 2 versus 0: HR, 2.0 [95% CI, 1.7 to 2.3]; ≥3 versus 0: HR, 2.4 [95% CI, 1.8 to 3.3]) and a significantly higher incidence of new diagnosis of renal impairment, regardless of the presence of GVHD (35% of patients with ≥3 infections, 31% of patients with 2 infections, 26% of patients with 1 infection, and 19% of patients with no infection). These results indicate that more directed prevention and treatment strategies for dsDNA viral infections could substantially improve clinical outcomes and reduce HRU.