328 - Optimizing Central Line Selection to Minimize Catheter-Associated Complications in Pediatric Patients with Acute Leukemia

Poster Number: 328
Publication Number: 328.113

Elise Rio, University of Rochester School of Medicine and Dentistry, rochester, NY, United States; Matthew Custer, UCSF, San Francisco, CA, United States; Hongyue Wang, University of Rochester, Rochester, NY, United States; Jeffrey R. Andolina, Golisano Children's Hospital at The University of Rochester Medical Center, Rochester, NY, United States; Eric Snyder, Wilmot Cancer Institute, Rochester, NY, United States; Carol Fries, University of Rochester, Rochester, NY, United States

Background: Pediatric patients with acute leukemia require central venous catheter (CVC) insertion for chemotherapy administration, putting them at risk of complications such as bacteremia and deep vein thrombosis (DVT). While some institutions opt for tunneled CVC (TCVC) or implanted venous access device (IVAD) placement at diagnosis, others use a temporary peripherally inserted central catheter (PICC) to mitigate infection risk before later converting to TCVC or IVAD. Further, patients with low absolute neutrophil count (ANC) are at increased risk of infection, but it is not fully understood whether initial ANC corresponds to the risk of CVC-associated bacteremia. It is of clinical importance to establish guidelines to optimize CVC selection based on individual patient characteristics.

Objective: 1) To test whether initial CVC type contributes to the risk of bacteremia and/or DVT in pediatric patients with acute leukemia; 2) To test whether initial ANC corresponds to the risk of CVC-associated bacteremia 

Design/Methods: We assessed clinical characteristics, initial CVC type, and complications for all pediatric patients with acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML) treated with induction chemotherapy at the University of Rochester between 2010 and 2020. We used multiple logistic regression analysis to test the relationship between clinical variables and outcomes of interest, with age as a covariate. 

Results: Patient characteristics by outcome for 110 patients with acute leukemia are detailed in Tables 1 and 2. The overall incidence of bacteremia was 10%; bacteremia was not significantly associated with CVC type in overall analysis. In subgroup analysis of patients with ANC≥0.5x103/µL, having an IVAD or TCVC (vs. PICC) was associated with a greater incidence of bacteremia (Odds Ratio (OR) 18.25, 95% CI 1.016-327.6). This trend was not observed in patients with ANC < 0.5 who already had a greater incidence of bacteremia regardless of CVC type (OR 4.626, 95% CI 1.002-21.36; Figure 1). Likewise, patients who experienced bacteremia had a lower overall presenting ANC than those who did not (P=0.042). The overall incidence of DVT was 8.3%, with an insignificant trend toward greater incidence in the PICC cohort (10.5% vs. 5.1%, P=NS). Conclusion(s): Our findings suggest that initial catheter type impacts the risk of bacteremia in patients with ANC≥0.5, while patients with ANC < 0.5 are already at increased risk regardless of CVC type. To mitigate this risk, temporary PICC placement may be a worthwhile consideration for most patients undergoing initial treatment for acute leukemia.
Figure 1: Risk of bacteremia.After accounting for other clinical variables by multiple logistic regression analysis, Odds Ratios (OR) and their corresponding 95% confidence intervals are shown for the following cohorts: Patients with IVAD, patients with other TCVC, and patients with an initial ANC < 0.5x10^3/µL. Initial ANC < 0.5 was associated with a significantly greater incidence of bacteremia, regardless of CVC type.
Table 1. Patient characteristics by incidence of bacteremia*Wilcoxon Rank Sum
†Fisher exact
‡2 patients in the ‘No Bacteremia’ group lacked detailed data and were eliminated from analysis (Total analyzed: Nf108)