This single center quadruple blinded RCT was approved by Children’s Hospital of Wisconsin Institutional Review Board (IRB# 1293254-5). The RCT was conducted from April to December 2019. Informed consent and assent, as appropriate, were obtained prior to starting any study procedures. Children were randomized 1:1 to t-PA dwell (treatment arm) or UFH dwell (control arm). Randomization was performed by the investigational pharmacist using a computerized algorithm. Participants, care providers and study team were blinded to assignment until after completion of analysis.
All children admitted to the pediatric ICU with a CVC placed prior to or during admission to the pediatric ICU were screened for eligibility. Patients > 2 weeks post-gestational age to ≤ 18 years of age with a CVC placed within 72 hours of enrollment were included. Exclusion criteria were:
Non-English-speaking subjects and/or parent/guardian
Active CVC infection; defined as positive blood culture from the in-situ CVC at time of enrollment
Current radiographically confirmed VTE
CVC diameter < 1.9 Fr
Current or previous diagnosis of heparin induced thrombocytopenia or allergy to UFH or t-PA
Med-a-port catheters or hemodialysis catheters
Currently receiving treatment doses of anticoagulation (UFH infusion > 15U/kg/h, enoxaparin injections ≥ 2mg/kg/day or ≥ 60mg/day)
Active internal bleeding
Recent intracranial or intraspinal surgery
Serious head trauma
Intracranial or other conditions that may increase the risk of bleeding
Coagulopathy or bleeding diathesis (includes platelet count < 20,000/mm3, INR > 2.0)
Expected CVC removal within 48 hours
Expected transfer or discharge within 48 hours
Exclusion based on bleeding risk aligned with local criteria for administration of systemic t-PA.
Children randomized to treatment arm received weight-adjusted dose of t-PA (Alteplase; Activase®, Genentech, 1mg/mL concentration). Children weighing < 10 kg, 10–20 kg or > 20 kg received 0.5 mg (0.5 mL), 1 mg (1 mL) or 2 mg (2 mL) of t-PA, respectively. Children randomized to control arm received the equivalent volume per weight of 10 U/mL of UFH. The study drug was dispensed in indistinguishable syringes labeled with the child’s unique study ID.
Within 24 hours after enrollment, the bedside nurse administered the study drug. The CVC was flushed with normal saline, then the study drug was infused, dwelling for 30 minutes to 4 hours. Dwells were stopped based on the clinical need for the lumen. Longer dwell times may be more efficacious but may not be practical due to clinical needs. After the dwell, the study drug was withdrawn, the CVC checked for blood return and then flushed with normal saline. Each lumen was treated, as possible, every 3 days until discharge from ICU, removal of CVC, or a maximum of 10 doses of study drug were administered. Study drug was not administered to lumens used for continuous infusion of vasoactive medication.
All children who received ≥ 1 dose of the study drug were followed for CADVT, CABSI and bleeding for 7 days after CVC removal, or for 30 days if the CVC remained in place.
For children with CVC still in place at the time of hospital discharge, medical records were reviewed to capture events up to 7 days after discharge. At the end of the study period, ultrasound with doppler was performed to assess for CADVT in the site of CVC placement. The reading radiologist was blinded to treatment assignment.
The primary outcome measure was enrollment rate defined as proportion of eligible children who were randomized. Secondary feasibility outcome measures were proportion of children who received study drug within 24 hours of consent, proportion with ultrasound, and proportion of enrolled patients completing the study. Other secondary outcomes for efficacy were CADVT as diagnosed by the systematic ultrasound, CABSI as diagnosed by the clinical team, and CVC dysfunction defined as inability to draw or flush CVC, and for safety, any clinically overt bleeding.
Baseline characteristics and outcomes were reported as a single cohort to avoid over-interpretation of unstable estimates from a limited sample size.15 Data was presented as median (interquartile range, IQR) for continuous variables and count (percentage) for categorical variables. A sample size of 20 was planned based on available resources.