Outpatient Vyxeos Induction without Planned Admission for Select Patients with Secondary Acute Myeloid Leukemia (sAML) Is Safe and Yields Healthcare Resource Savings

INTRODUCTION: With the FDA approval of CPX-351 (Vyxeos; Jazz Pharmaceuticals, Inc.; Palo Alto, CA), the prospect of inducing select patients with acute myeloid leukemia (AML) with a regimen that does not require continuous intravenous in the outpatient (OP) setting became a reality. With the recognition that an OP-induction strategy has the potential benefits of reduced healthcare resource utilization (HRU),^1,2 improved quality of life³ and decreased risk of nosocomial infections, some centers have opted to administer Vyxeos in the OP setting and to admit patients only if complications which necessitate inpatient (IP) management occur. The safety and decreased HRU of this approach during consolidation has been demonstrated⁴ in the study population treated on the pivotal phase III trial (NCT01696084).⁵ We present the first real-world experience of OP AML induction with Vyxeos without planned hospital admission.

METHODS: We performed a retrospective analysis of all patients at our institution treated with OP-Vyxeos without planned admission (OP-Vyxeos) between August, 2017 and June, 2020. Patients were deemed safe for OP induction if they met the criteria outlined in Table 1 at diagnosis or could be stabilized in the hospital and discharged when these criteria were met. Primary and secondary objectives were to evaluate the safety and HRU, respectively, of this approach. The primary endpoint was 30-day mortality. Secondary endpoints included 60-day mortality, rate of adverse effects (AEs), number of hospitalizations, number of hospital days, number of intensive care (ICU) days, and number of days receiving intravenous (IV) antibiotics. When possible, primary and secondary endpoints were compared to patients who received Vyxeos induction in the inpatient setting (IP-Vyxeos) on the phase III trial.^5,6

RESULTS: Twenty-five patients received OP-Vyxeos at our institution between August, 2017 and June, 2020 and all were evaluable for the primary endpoint. Baseline characteristics of the OP- and IP-Vyxeos populations are presented in Table 2. The mean age of OP-Vyxeos patients was less than IP-Vyxeos patients (65.4 years vs. 67.7 years) and a higher proportion of OP-Vyxeos patients were <60 years old (20% vs 0%). ECOG Performance Status (PS) was similar between the two groups. Baseline white blood cell count was similar between the two groups but a lower proportion of OP-Vyxeos patients had a platelet count </= 50 x10⁹/L (32% vs. 62%).

Clinical efficacy and safety outcomes are presented in Table 3. Complete response (CR) rate for Vyxeos was similar between the OP- and IP-Vyxeos groups (48% vs. 37%), as was time to neutrophil recovery (34 vs. 35 days). There were no early deaths (before 30 days) in the OP-Vyxeos group compared to 9 (5.9%) in the IP-Vyxeos group. Sixty-day mortality was similar between the two groups (4.5% vs. 13.7%). The cause of death in the one OP-Vyxeos patient who died was fungal pneumonia in the setting of presumed refractory leukemia. Febrile neutropenia (FN) with or without documented infection and bleeding (any grade) were the most common AEs in both groups, however, the rates varied significantly between OP- and IP-groups (FN: 88% vs. 68%; bleeding: 28% vs. 74.5%). FN was by far the most common AE in the OP-Vyxeos group and was the admission diagnosis for 91% (32/35) of admissions.

HRU-metrics are presented in Table 4. The majority of OP-Vyxeos patients required hospitalization (88%); however, despite an increased mean number of hospitalizations (1.4 vs. 1.0), the median total number of hospital days in the OP group was decreased (22 vs. 28). Three patients (12%) never required hospitalization during OP-Vyxeos induction. The median days spent OP prior to first admission was 8 and 18% (4/22) of patients remained OP for >14 days.

CONCLUSIONS: Outpatient Vyxeos induction without planned admission for select patients was safe and resulted in decreased healthcare resource utilization in our cohort of patients. There were no early deaths in the OP-Vyxeos group. Despite an increased mean number of hospitalizations, patients who received OP-Vyxeos had a decreased number of total hospital days. Limitations of this analysis include that this is a small, single-center, retrospective analysis. In the future, we hope to validate our criteria for OP-induction in a larger, multi-institution patient cohort.