Clinical Outcomes of Neonatal Central Line-Related Thrombosis: A Multicentric Retrospective Cohort Study

Introduction: Neonates are particularly at risk of thromboembolism (TE), especially neonates requiring central lines. Management of neonatal TE is particularly complex, as the higher risk of bleeding must be balanced with the thrombotic risks, including organ dysfunction, post-thrombotic syndrome and death. The objectives of this study are to describe the treatment modalities and to compare the effectiveness and safety between antithrombotic treatment modalities among neonates with central line-related TE. Methods: A multicentric retrospective cohort study enrolled neonates ≤28 days of life requiring a central line with a radiologically confirmed TE in the anatomical territory of the central line, admitted in one of eight Canadian NICUs (2013-2018). Data from the Canadian Neonatal Network registry were linked to clinical outcomes of interest in individual medical records, namely TE resolution and TE progression within three months, major bleeding (MB) and clinically relevant non-major bleeding (CRNMB) defined using ISTH criteria. Logistic regression explored predictors of anticoagulation use and whether treatment modality predicted clinical outcomes. Institutional review boards of all sites approved the study. Results: Overall, 417 neonates sustained a TE diagnosed at a median of 12 days (25-75th percentile: 6-27) after birth. Of those, 81% (n=338) had a venous thrombosis and 19% (n=73) had an arterial thrombosis. Neonates had a median gestational age of 33 weeks (range: 27-38) with a median birth weight of 2.2 kg (range 0.9-3.2). Associated medical conditions were common in these children, such as congenital heart defects (51%), respiratory distress syndrome (47%), necrotizing enterocolitis (16%), and sepsis (9%). While umbilical venous catheters (n=72%), picclines (n=52%) and umbilical arterial catheters (n=31%) were the most common types of central line associated with TE, several patients had more than one central line in place during their admission. Expectative management with or without central line removal, anticoagulation and other treatment strategies (thrombolysis or antiplatelets) were used in 59%, 39% and 2% of patients with venous TE, respectively. For arterial TE, anticoagulation was more commonly used (52%) followed by expectative management (38%), antiplatelets therapy (6%) and thrombolysis (4%). In addition, older gestational age (p=0.002), male sex (p=0.04), occlusive TE (p<0.001), and TE location (p<0.001) were independently associated with anticoagulation use for venous TE. However, for arterial TE, only neonates with older gestational age were being treated significantly more often with anticoagulation (p=0.02). Complete TE resolution and progression occurred in 43.5% and 2.3% of patients, while MB and CRNMB happened in 7.4% and 4.3% of neonates. Clinical outcomes for venous TE did not significantly differ based on treatment modality (TE resolution: p = 0.26, TE progression: p = 0.3, MB: p = 1.0, CRNMB: p = 0.051). For arterial TE, there was no difference in the rate of MB (p=0.43) based on the treatment modality. However, there was a significant improvement of the TE resolution in the anticoagulation group (p=0.003). Conclusion: Anticoagulation was often withheld in neonates with perceived lower thrombotic risk or higher bleeding risk, and expectative management was not associated with unfavourable outcomes. Further prospective studies are needed to tailor the treatment of neonatal central line-related TE using a patient-centred approach.