Factors Associated with Prolonged Hospital Stay during Induction Chemotherapy for High Risk Pediatric Acute Lymphoblastic Leukemia

Introduction

Pediatric patients with High Risk (HR) or Very High Risk (VHR) acute lymphoblastic leukemia (ALL) receiving intensive induction chemotherapy often have a complicated and prolonged hospital stay. The aim of this study was to identify risk factors that are associated with a complicated hospital course or prolonged inpatient admission during induction chemotherapy, and investigate if these factors influence disease-free or overall survival.

Methods

All patients (N=73) (age<21 years) diagnosed from 2009-2016 with de novo HR (leukocyte count at presentation≥50,000/mm³ or age≥10 years, T cell ALL) or VHR (≥13 years old, presence of Philadelphia chromosome, hypodiploidy, MLL rearrangement, intrachromosomal amplification of chromosome 21, high minimal residual disease, or induction failure) ALL, treated at Riley Children's Hospital, Indianapolis, were included for chart review. Univariate and multivariate logistic regression was used to test for statistical significance. A P-value<0.05 was considered to be statistically significant. Variables with significance in the univariate models were considered for multivariate analysis. The overall survival (OS) and disease (leukemia)-free survival (LFS) were analyzed using the Kaplan-Meier method and log-rank test.

Results

N=42 (57%) patients (group A) had an uncomplicated course (discharged at day 4 after receiving pegylated-asparaginase, no readmission or ICU stay) during induction chemotherapy. N=31 (43%) patients (group B) experienced a prolonged admission (median hospital stay=8 days vs 4 days (group A), P=0.02) or an ICU stay (N=13 (18%)), due to infectious or metabolic complications. Only 2 patients (3%) died during induction chemotherapy.

In univariate logistic regression, patients with a platelet count<20,000 /mm³ (P=0.04), low serum bicarbonate (P=0.002), and a high blood urea nitrogen (BUN) (P=0.03) on day 4 were more likely to have a prolonged admission (OR=4.52, 8.21, and 3.02, respectively). Age, Hispanic ethnicity, or severity of anemia or neutropenia were not associated with a prolonged hospital stay or ICU stay. Multivariate stepwise logistic regression identified low serum bicarbonate (P=0.002) and a platelet count<20,000/mm³ (P=0.02) on day 4 of admission to be predictive of a prolonged hospital stay. LFS was significantly worse in group B compared to group A (5 year LFS=58% vs 89%, respectively, P=0.03). No difference in the OS was observed in the two groups (A and B) (5 year OS=67% (group B) vs 83% (group A), P=0.9).

Conclusion

A significant proportion of HR or VHR pediatric ALL patients experience a prolonged inpatient hospital stay during induction chemotherapy. We identified factors that may be predictive of a prolonged length of hospitalization due to complications, and noted an inferior leukemia-free survival in this group of patients.