Feasibility of Tissue Plasminogen Activator Dwell Therapy to Reduce Central Venous Catheter Associated Thrombosis: A Randomized Controlled Pilot Trial

Backround: To evaluate the feasibility of a randomized controlled trial (RCT) of the brinolytic tissue plasminogen activator (t-PA) vs unfractionated heparin (UFH) central venous catheter (CVC) dwell therapy to reduce risk of CVC-associated deep venous thrombosis (CADVT) in critically ill children. Methods: This single center quadruple blinded pilot RCT enrolled children ≤ 18 years of age with CVC placed within 72 hours of admission to the pediatric intensive care unit (ICU) Weight-adjusted dose of study drug dwell (t-PA vs UFH) was installed to alternating lumen of CVC every 3 days for 10 doses, CVC removal or ICU discharge. Ultrasound with doppler was performed at study completion. Main Results: Of 426 children screened from April-Dec 2019, 86 (20%) were eligible with 20 enrolled and randomized. Primary outcome measure of enrollment rate was 23%. One child was withdrawn immediately after randomization due to development of exclusion criterion. Secondary feasibility outcome measures were proportion of children who received study drug within 24 hours of consent (100%), proportion with ultrasound (100%), and proportion completing the study (95%). Eighteen of 19 children received the rst dose within 48 hours of CVC placement. All children missed some dose days because of lumen specied to be in continuous use. Median dwell time for doses received was >2 hours. There were no protocol violations. Six of 19 patients (31.6%) developed CADVT, 1 of which was occlusive. There were no catheter-associated blood stream infection or signicant bleeding. Conclusion: Critically ill children requiring CVC are at high risk for CADVT. A future multicenter, blinded, RCT to determine the effectiveness of t-PA vs UFH dwell in reducing CADVT is feasible.


Background
Venous thromboembolism (VTE), which mainly manifests as deep venous thrombosis (DVT), is a top contributor to harm and excess costs in hospitalized children. 1,2 Critical illness and central venous catheter (CVC) are the most important risk factors for DVT in children. [3][4][5] CVC-associated DVT (CADVT) can serve as a nidus of infection and increases the risk of CVC-associated bloodstream infection (CABSI). 6 Tissue plasminogen activator (t-PA), a brinolytic enzyme, is an established therapy to restore CVC patency after dysfunction from intra-catheter thrombi. [7][8][9] For prophylaxis, t-PA infused to dwell in dialysis catheters is associated with lower risk of CADVT than unfractionated heparin (UFH) dwell. [10][11][12][13] t-PA may also prevent CADVT in non-dialysis CVC, but evidence is limited. t-PA dwell has only been shown to reduce the risk of CVC dysfunction in children with long term CVC dependency, as this study was underpowered to evaluate the effect of t-PA on CADVT or CABSI. 14 A multicenter, randomized clinical trial (RCT) is needed to determine the e cacy of t-PA in preventing CADVT in children. This pilot RCT aimed to test the feasibility of a RCT of t-PA dwell against CADVT in critically ill children.

Study design
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.

Population
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:

Study Drug
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.

Study Procedures
Within 24 hours after enrollment, the bedside nurse administered the study drug. The CVC was ushed 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 e cacious 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 ushed 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.

Outcomes
The primary outcome measure was enrollment rate de ned 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 e cacy were CADVT as diagnosed by the systematic ultrasound, CABSI as diagnosed by the clinical team, and CVC dysfunction de ned as inability to draw or ush CVC, and for safety, any clinically overt bleeding.

Statistical Considerations
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.

Eligibility and consent
A total of 426 children with a CVC were screened April-Dec 2019 (Fig. 1). Of these, 86 (20%) were eligible.
The most frequent reasons for exclusion were expected CVC removal in less than 48 hours (N = 68) and concurrent receipt of therapeutic anticoagulation (N = 58). Of the eligible children, 30 (35%) were approached for consent. Lack of available parent/guardian was the most common reason for failure to approach for consent (N = 41). Enrollment rate was 23% of eligible patients. All consented patients were randomized. However, one child was excluded after randomization, but prior to receiving the study drug, because of immediate use of therapeutic anticoagulation. All other patients received the study drug within 24 hours of enrollment.

Patient and CVC characteristics
Median age of enrolled children was 6 years (IQR: 1.4, 11 years) ( Table 1)

Outcomes
Only aggregate results are presented as recommended for pilot RCTs. 15 Six of 19 children (31.6%) developed CADVT, 1 of which was occlusive. There were no episodes of CABSI. Episodes of CVC dysfunction occurred in 7 CVC (36.8%). Most episodes were categorized as an inability to draw blood from a lumen. There were 5 non-major nor clinically signi cant bleeding events in the RCT. None required intervention.

Discussion
This pilot RCT has shown that a blinded RCT of t-PA vs UFH dwell to decrease the risk of CADVT in critically ill children is feasible, with 23% of parents of eligible patients agreeing to participate in the RCT and randomized within 72 hours of CVC placement. Fifty-six of the 86 eligible patients were not approached for consent. The most common reason (41/56) was lack of family availability. Of those approached, 67% consented to study participation. Expansion of research team availability in a future study to include nights and weekends may improve the enrollment rate. No violations in study protocol were reported. All children missed some doses of study dwell as lumens requiring continuous infusion of vasoactive medication did not receive study drug, which was pre-speci ed in the protocol.
This RCT was not designed nor powered to answer the question if t-PA is e cacious in reducing CADVT compared to UFH dwell in critically ill children with recent CVC placement. However, the results con rm that this is a population at risk for CADVT, with over 30% developing CADVT. Similarly, this RCT was not designed to determine safety. However, there were no clinically signi cant bleeding events. As expected, from the long history of using t-PA for CVC occlusion and UFH to maintain CVC patency, we con rmed that both study drugs have favorable safety pro les in this critically ill population.
A large, multicenter pediatric RCT will be needed to determine the effectiveness of t-PA dwells in reducing the risk of CADVT. Based on a risk of CADVT of 31.6%, 20% relative risk reduction of CADVT, type 1 error of 0.05 and power of 0.8, we will need 800 children per arm to complete this RCT. Patient factors including non-CVC risks for VTE (age, diagnosis, mechanical ventilation, inotropes use) and CVC factors (type, diameter, duration) may convey additive risk and need to be incorporated in the trial design.

Conclusion
Critically ill children requiring CVC are at high risk for CADVT. A future multicenter, blinded, RCT that will determine the effectiveness of t-PA vs UFH dwell in reducing CADVT is feasible, with favorable consent rates of approached patients and good adherence to study protocol. SJH was responsible for study conception and design, analysis and data interpretation, drafting and critical revision and nal approval of manuscript. SS contributed to study design, data acquisition and interpretation, critical revision and nal approval of the manuscript. EVSF contributed to study design, analysis and data interpretation, critical revision and nal approval of the manuscript.