Abstract
Introduction Central venous lines (CVLs) are crucial to the treatment of acute lymphoblastic leukemia (ALL). They are utilized for lab draws, chemotherapy, and blood product administration. Two of the most common CVLs are Port-a-Cath (Port) lines and peripherally inserted central catheters (PICC). Ports are percutaneously inserted into the jugular or subclavian veins and remain underneath the skin while not in use. PICCs, short-term CVLs, are inserted into smaller caliber veins and are always exposed outside of the skin. Ports, preferred by patients, enable patients to continue routine activities, such as showering, without the risk of infection or loss of central access. Additionally, PICCs require daily flushing to maintain line patency while Ports require monthly flushing.
At Nemours Children's Hospital Delaware (NCH-DE), proceduralists opt for PICCs in newly diagnosed ALL patients citing risks for complications related to Port placement (poor wound healing, hematoma formation, greater infection risk) due to neutropenia and thrombocytopenia. However, evidence that neutropenia and thrombocytopenia pose these risks in CVL placement is insufficient. Anecdotally, patients appear to have more complications, especially thrombosis, related to PICCs rather than Ports. To study this, we performed a retrospective chart review to compare the incidence of complications related to CVL placement in newly diagnosed ALL patients at NCH-DE.
Methods A retrospective chart review of all patients with newly diagnosed ALL between January 2015 and June 2025 was conducted. Demographic data (age, gender, ALL type), CVL characteristics (type, duration in place, and anatomical location), and complications (bloodstream infections [CLABSI], thrombosis, superficial infection, hematoma formation, and new fevers), white blood cell count (WBC), absolute neutrophil count (ANC), and platelet counts (Plt) at diagnosis were analyzed. Patients were divided into two groups based on the type of CVL placed at diagnosis (PICC vs. Port). The incidence of adverse events was compared by age, gender, WBC, ANC, and Plt counts at diagnosis, ALL type, number of lumens (PICC group only), and date of CVL placement (before or after July 2019) using Chi square and non-parametric statistical analyses such as Fisher Exact and Kruskal Wallis testing where appropriate. Statistical analysis was performed via Stata Version 17.
Results Between January 2015 and June 2025, 167 patients (90 male, 70 female) were diagnosed with ALL (B-ALL = 142, T-ALL = 20) at NCH-DE. One patient had Biphenotypic Leukemia (B-ALL + AML) and another had Blastic Plasmacytoid Dendritic Neoplasm, leukemia type. In total, 27 Ports, 123 PICCs, and 8 tunneled CVLs were placed at the time of diagnosis. In the Port group, average blood counts were WBC 6.4K/µL, ANC 1000/µL, Plt 84K /µL. In the PICC group, average blood counts were WBC 11.4K /µL, ANC 700/µL, Plt 43K /µL. No statistically significant difference among average blood counts was identified between the groups. A total of 3 CLABSIs and 2 thromboses were identified in the Port group versus 11 CLABSIs and 11 thromboses in the PICC group (p = 0.5 and 0.4 respectively). Twenty-four total complications, including CLABSI, thrombosis, superficial infection, and fever within 24 hours, were noted in those with PICC placement compared to 5 in the Port group (p=0.08). Patients who had a single lumen PICC placed were more likely to develop a fever within 24 hours of placement (p = 0.02). No patients developed a hematoma, superficial infection, or skin breakdown following CVL placement.
Conclusion No significant difference was identified in the incidence of complications between newly diagnosed ALL patients who have a PICC versus a Port placed. We found a trend towards statistical significance when evaluating the incidence of overall complications with PICC placement compared to port placement. This suggests that newly diagnosed ALL patients are likely to tolerate Port placement without an increased risk of complications. This study is limited due to the nature of its retrospective chart review and limited sample size. Our next aim is to develop a protocol to streamline CVL placement at diagnosis of ALL and then perform a prospective study comparing outcomes based on CVL placement at diagnosis and confirm the findings of this study.
