Early Discharge and Out Patient Management After AML Induction Chemotherapy: Determinants of Safety

No clear consensus exists for the setting of supportive care after induction chemotherapy given for newly diagnosed acute myeloid leukemia (AML). While many centers provide care exclusively via in-patient settings, others including the Penn State Hershey Cancer Institute select suitable patients for discharge following chemotherapy and provide supportive care in out-patient infusion centers. Our single center practice offers 7 day a week transfusion and antibiotic support from skilled subspecialty nurses and midlevel practitioners. For ambulatory, afebrile AML patients with appropriate caregivers and social support allowing ∼q48 hour visits we encourage early discharge after administration of chemotherapy but before recovery from marrow aplasia. Patients are then readmitted if febrile or if other circumstances mandate hospitalization. This study presents results of this approach.

We analyzed results of management including the option of early discharge in a series of 347 consecutively treated newly diagnosed adult AML patients receiving cytarabine based, aplasia inducing chemotherapy between 2003 and 2010. 334 patients received 7 days of infusional cytarabine with 3 days of anthracycline or anthracedione (“7+3” – in a few cases with additional agents), while the other 13 received bolus high or intermediate dose cytarabine with other agents. Patients were ages 18–85 years (median age 60). All received induction chemotherapy in the hospital. We classified cases into those discharged prior to recovery of neutrophils (ANC) to >500/ul or with ANC>500 but before the ANC nadir was reached (“Early” discharges), those kept in hospital throughout the period of ANC<500 but discharged after this point (“Late”), and those dying while in the hospital without discharge. Primary outcomes were overall survival (OS) at 28, 60, and 365 days, and number of days of hospitalization of the initial 28 days from initiation of induction chemotherapy. Clinical parameters including age, marital status, white blood count (WBC), platelets, albumin, creatinine, bilirubin were analyzed for their effects on site of supportive care and outcomes.

Early discharge (after chemotherapy but prior to recovery of ANC>500) was employed in 259/347 patients of median age 60 years (range 18–85). They were outpatients for a median of 13 (range 1–23 days), of the initial 28 days from the start of chemotherapy (defined as day 0). 185/259 required readmission for sepsis or other complications prior to recovery of neutrophils. 234/259 recovered to ANC>500 at days 15–132. OS for this early discharge group was 96.1%, 90.0%, and 59.5% at 28, 60, and 365 days respectively. We analyzed predictive factors for survival at day 60 (OS60) in this group. While pretreatment WBC, platelets and albumin values were not related to OS60, greater age and unmarried status correlated with death prior to day 60. Mean age for those with survival at D60 was 55.89 and for those dying by D60 was 69.70 (95%CI 7.909 to 19.72 for the difference between means and p<0.001.) Unmarried status carried an odds ratio of 3.087 (95% CI 1.37–6.942, p=0.0069). Fifty-eight of 347 patients, median age 55 (range 20–79) were discharged late ie after achieving ANC>500, all surviving to day 60. Thirty patients of 347 died during the first hospitalization.

In this retrospective review, early discharge following AML induction chemotherapy was generally safe and was offered to 259 of 347 patients treated. Risk factors for death at day 60 in this group included advanced age and unmarried status. Prospective studies will be needed to confirm these data. This work has implications for safe and cost effective management of AML. Additional analysis is underway.

No relevant conflicts of interest to declare.