The Clinical and Financial Cost of Preemptive Management of CMV Disease – Implications for Immunotherapy

Although CMV infection only rarely leads to direct infectious mortality, and early CMV reactivation may reduce relapse rates in patients with myeloid malignancies, the management of CMV by preemptive antivirals may itself induce toxicity and financial burden which need to be quantified to provide justification for antiviral cellular immunotherapy. In this study, we analyzed CMV reactivation in 134 (72 males, 62 females) patients undergoing allo-SCT at our institute between 2006 through 2012, comparing outcomes and cost after viral reactivation based upon the risk of CMV reactivation. 102 subjects were transplanted for malignant diseases; and 32 were transplanted for non-malignant diseases. 81 subjects received CD34+ selected myeloablative, 12 received UCBT, and 41 had T-replete nonmyeloablative transplant. Median age at transplant was 40 years and the median follow-up is 4.25 years. 119 (88.8%) were at risk for CMV by virtue of either the donor or recipient being seropositive prior to transplant. Of these, 90 (75.6%) had CMV reactivation in the blood (>250 copies/ml by PCR) including 4 (3.4%) who had CMV organ disease (involving the GI=3 , lung=2, CNS=1) and all received antivirals. Median time to first CMV reactivation was 34 days (range: 8-105 days). 11% reactivated before d14+ and 92% by d100+. For comparison, we defined two groups: "Antiviral group" consisting of the 90 subjects (reactivation and/or organ disease) and a “No therapy group" consisting of 44 subjects (29 at risk subjects who did not need antiviral treatment, and 15 who were not at risk). The Antiviral group averaged  12.6 days of IV antiviral therapy and 13.9 additional days of hospitalization (P<0.02). We calculated the additional cost for the Antiviral group to be between $55,000 to $71,000 per patient. We also evaluated the impact of CMV risk on overall survival and on NRM but found no statistically significant impact.  However, in multivariable modeling of NRM, CMV reactivation >250 copies/ml (OR=3, P<0.048), total duration of inpatient IV antiviral therapy (OR = 1.04 ,P<0.001), type of transplant (T-deplete vs. T-replete) (OR=4.65 ,P<0.017) , were found to be significantly associated; the model excluded age, sex, diagnosis (malignant vs. non malignant) and CMV risk status.Our findings document that although direct infectious mortality from CMV has been eliminated, there is a significant cost of therapy in terms of drug and inpatient hospitalization as well as a strong association with NRM. Implications for antiviral cellular therapy are that early administration even prior to day 14 is necessary to meaningfully impact reactivation and that cost savings exceeding $25,000 are possible with even moderate (50%) reduction in the rate of reactivation requiring antiviral therapy.