Healthcare Resource Utilization in Transplant Patients Who Are at a Higher-Risk to Develop Cytomegalovirus Infection during Their Primary Transplant-Related Hospitalization

Cytomegalovirus reactivation commonly referred to as CMV infection (CMVi) is a frequent event after allogeneic hematopoietic cell transplantation (HCT), with studies associating CMVi within the first 100 days post-HCT with higher risk of non-relapse mortality (NRM) and decreased overall survival (OS). In addition, understanding the impact of CMVi on resource utilization during the primary HCT admission is critical. Together, this knowledge of epidemiology and resource utilization may be used to inform preventive strategies to minimize CMVi, e.g., use of antiviral agent letermovir.

After receiving IRB approval, we retrospectively reviewed institutional electronic medical records and CMVi database from 824 patients who underwent their first allogeneic HCT between 2011 and 2016 at City of Hope (pre-letermovir era). Patients were censored at death, disease relapse or lost to follow up. Data collected: demographics, HCT indication, conditioning regimen, CMV serostatus of the donor and recipient (D/R), length of stay (LOS) for primary HCT admission (all allo HCT were performed as inpatient), readmission rates in first 100 days, and use of supportive care. CMV viral load of >250 genomic copies/ml constituted a diagnosis of CMVi. CMV viral load surveillance in MUD recipients began at engraftment or day +21 post-HCT, whichever occurred earlier. For Haplo and cord blood (CB) HCT, CMV viral load surveillance started on day +14. The primary endpoint of the study was LOS for HCT admission. Supportive care use, transfusions, growth factors and antiviral usage were secondary endpoints. The differences in resource utilization between different groups were examined by CMVi during the primary HCT admission period, using Wilcoxon test or chi-square test whenever appropriate.

Median age of patients at the time of HCT was 52 years (range: 1-78), with 57% of patients being male. The most common diagnoses included: AML (39%), ALL (21%) and MDS/MPN (17%). Patients underwent MUD (n=627, 76%), Haplo (n=102, 12%), or CB-HCT (n=95, 12%), and 44% of patients received myeloablative conditioning regimen. Majority of the patients were CMV seropositive (83.7%). Graft source was peripheral blood stem cells in 75% of the recipients. Most commonly used graft-versus-host disease prophylaxis consisted of post-transplant cyclophosphamide (100%), Tacrolimus/sirolimus (83%), and cyclosporine/cellcept (78%) in Haplo, MUD, and CB-HCT recipients, respectively.

During the primary HCT admission, rate of CMVi was 7%, 36% and 28% in all of MUD, Haplo, and CB-HCT, respectively (compared to 25%, 71.6%, and 50.5% in MUD, Haplo and CB-HCT respectively in the first 100 days after HCT). Rate of CMVi in CMV+ recipients was 8.2% in MUD, 41.6% Haplo and 34.2% in CB-HCT (Table 1). Majority of patients with CMVi received antiviral therapy (85.8%), with Haplo and CB-HCT more likely to be treated than MUD (p=0.023). LOS was longer among CMVi patients compared to no CMVi patients in each donor type, median of 59 vs. 36 days for the overall cohort (p<0.01). The difference in LOS by CMVi remained significant (p<0.001) in the multivariable regression model including donor type, graft source, primary diagnosis and conditioning intensity (Table 2). Filgastrim use was higher among CMVi patients than no CMVi patients in MUD (p<0.001), but not in Haplo or CB-HCT (p>0.2). Transfusion of packed red blood cells (PRBC) and platelet units were significantly higher among CMVi recipients of MUD and Haplo (p<0.02), but not CB-HCT (p>0.82). There was no significant difference in hospital readmission by CMVi across donor type in the first 100 days (p>0.5).

In conclusion, the rate of CMVi during primary HCT admission was high, particularly in the Haplo and Cord HCT (>50% of the CMVi occurring within 100 days of HCT). Given the relatively high CMV viral load cut-off values and later CMV surveillance initiation, the rate could, in fact, have been underestimated in our cohort. CMVi during primary HCT admission was associated with significantly higher health care resource utilization; longer hospital LOS and supportive care utilization (CMV specific antiviral usage, transfusion and growth factors use). Prophylactic strategies to prevent early CMVi in alloHCT should be considered to decrease NRM and improve value based care delivery.