2105. Liposomal Amphotericin B Use Before and After Implementation of Voriconazole Prophylaxis in Cancer Patients

Abstract

BackgroundInvasive fungal infections (IFI) are life-threatening complications of prolonged neutropenia in hematologic cancer or after hematopoietic stem cell transplantation (HSCT). Guidelines recommend mold prophylaxis (ppx) for patients at high risk of IFI. Patients receiving ppx with new signs of infection are often escalated to Liposomal Amphotericin B (L-AmB) for concerns of breakthrough mold infections. We describe the impact of implementing voriconazole (VZL) ppx in cancer patients.MethodsWe performed a quasi-experimental study of all adult patients prescribed L-AmB for ≥1 dose in Cancer Center at the University of Maryland Medical Center. VZL ppx was implemented for patients with hematologic cancer with anticipated prolonged neutropenia (≥ 7 days) in 4/2017. HSCT patients routinely received posaconazole ppx for ≥ 1 year during study period. Comparisons were made pre (November 2015–June 2017) and post (July 2017–December 2018) implementation of VZL ppx allowing for 3-month wash-in period. Cancer center-specific L-AmB days of therapy (DOT) per 1,000 patient-days (PD) were compared using segmented regression and Student t-test. Comparison of patient characteristics, mortality, nephrotoxicity and hospital length of stay (LOS) among patients receiving L-AmB in pre vs. post periods was done using Χ 2 and Student t-test.ResultsThere were 87 (24 pre, 63 post) unique patients included in the analysis, translating to a total of 17.6 L-AmB DOT per 1,000 PD for the study period. Mean L-AmB utilization in cancer center was 9.9 and 24.4 DOT per 1,000 PD (P = 0.0037) for pre and post-implementation, respectively. There was an average 16% increase of L-Amb quarterly (P = 0.93). Among patients receiving L-AmB, most had acute myelogenous leukemia (63% vs. 60%) with lung source (71% vs. 73%, P = 0.8). More patients had proven IFI pre-implementation (42% vs. 29%, P = 0.3). Nephrotoxicity (46% vs. 48%, P = 0.9), median LOS (17 vs. 28, P = 0.4) and inpatient mortality (30 % vs. 38%, p = 0.5) all increased without statistical significance.ConclusionAfter implementation of VZL ppx there was a significant increase in L-AmB use, and associated non-significant increases in LOS, mortality and nephrotoxicity for those receiving L-AmB. Larger, robust longitudinal studies are needed to better understand the implications of VZL ppx on this population. Disclosures All authors: No reported disclosures.