20 Years of Experience with Allogeneic Hematopoietic Cell Transplantation in the Outpatient Setting

Abstract

Reduced intensity conditioning with fludarabine and low dose total body irradiation before allogeneic hematopoietic stem cell transplantation (HCT) was developed to treat older or medically infirm patients with advanced hematologic malignancies. Given the regimen's low toxicity, it lends itself to be an outpatient procedure. Nevertheless, some patients experienced adverse events, leading to hospitalization during the first 100 days after HCT. The current study analyzed data from 957 patients with hematological malignancies who received either related (n=441) or unrelated (n=516) HLA-matched HCT between 1997 and 2017 (10% of patients mismatched for 1 HLA antigen or allele), with a median follow-up of 7.1 (range, 0.3-19.9) years. Median age of related recipients was 55 (range 18-79) years and 60 years for unrelated recipients (range 18-77). HCT-CI≥ 2 was similar for both groups (68% and 74%, respectively). The objective of the study was to determine how many patients required hospitalizations, the reasons for the hospitalizations, and whether hospitalizations adversely affected non-relapse mortality (NRM) and overall HCT outcome. We found that 54% of patients were not hospitalized or only had an overnight hospital admission for PBSC infusion. The main reasons for hospital admission (36% of related, 56% of unrelated recipients) were infection and regimen-related toxicity (Figure 1). Most admissions occurred in the first 3 weeks after HCT (60%), and these were mainly for neutropenic fever and regimen-related toxicity. The 5-year risk of NRM was significantly higher among hospitalized patients (27%) than among those who were not hospitalized (14%, Figure 2). Of note, the cumulative incidence curves between the two patient groups continued separating throughout the first 5-years. Table 1 shows an HR for NRM of 1.69 (95%CI; P=0.001) with the first hospital admissions. The HR for NRM increased with each subsequent admission by 1.3 (P=0.003). Univariate analysis also showed higher NRM among unrelated compared to related recipients. Five-year relapse mortality among hospitalized and non-hospitalized patients was similar (26% and 28%, respectively). In conclusion, 54% of current patients lived at home or in apartments throughout the HCT course while 46% had at least one hospital admission. Even one hospital admission adversely affected long-term outcome by increasing the risk of NRM. The events connecting hospital admission and increased NRM are subjects of ongoing analyses.