Comparison Of Ultrasound Versus Magnetic Resonance Venography For Diagnosis Of Catheter-Related Thrombosis In Children: A Multicenter Multinational Study

Abstract

BackgroundVenous thrombotic events (VTE) in children are related to central venous catheters and occur predominantly in the upper venous system. In routine clinical practice, ultrasound (US) is the most frequently used imaging technique for diagnosis of VTE in children because of its ease of use and non-invasiveness. However, US is known to be relatively insensitive for detection of VTE in the central upper venous system. Magnetic resonance venography (MRV) is a promising alternative as MRV can comprehensively image the central venous system, is minimally invasive and does not involve radiation. However, MRV involves high technical and logistic demands, and requires sedation in young children. ObjectivesTo compare the feasibility and diagnostic accuracy of US and MRV for diagnosis of central venous catheter related VTE in children. MethodsStudy Design: Cross-sectional diagnostic study. Study population: Children 0-<18 years of age with a central venous catheter in place for any reason were recruited prospectively and consisted of i) children who are asymptomatic for VTE: children who were screened for central venous catheter-related VTE within 20-60 days of catheter placement ii) children who are symptomatic for VTE: children presenting with signs and symptoms of central venous catheter related VTE within 7 days of symptom onset (all study-related imaging tests were completed within this time period). Children were excluded in case of i) systemic anticoagulant or antiplatelet therapy, and ii) inability to undergo contrast MRV. Imaging tests: Each participant had three imaging tests performed i) Doppler compression US ii) MRV without contrast iii) MRV with gadolinium contrast. All imaging studies were performed within 48 hours of each other. Data collection included clinical and demographic information on each subject, imaging and video documentation of each US and MRV. Central Adjudication Committee: All imaging studies were blinded and independently reviewed and interpreted by a central adjudication committee for the presence or absence of VTE. Outcomes: The two main study outcomes were 1) Feasibility of performing each test: assessed by a) the number of children who completed each diagnostic study and b) the number of evaluable studies; and 2) Diagnostic accuracy of each diagnostic test: assessed by the proportion of VTE identified by the respective test in relation to the total number of VTE identified by any imaging modality. ResultsA total of 152 children were enrolled from 24 centers in 9 countries: Argentina, Brazil, Mexico, United States, Canada, United Kingdom, Netherlands, Austria and Germany. The median age (range) for children was 11 years (2 months to 17 years). Fourteen (9.2%) children had clinical symptoms of VTE. One hundred and thirty three (88%) children had US completed, 113 (74%) had MRV without contrast and 113 (74%) had MRV with contrast. 18 subjects withdrew from the study before any test was performed for various reasons. Adjudication and analysis are ongoing, and will be completed by October 14th, 2013. Therefore, final results regarding the feasibility and diagnostic accuracy of each test will be available at the time of presentation. ConclusionWe have completed the largest and most comprehensive study to date comparing the feasibility and diagnostic accuracy of currently available non-invasive imaging modalities for detecting central venous catheter related VTE in children. Ultrasound demonstrated better feasibility than MRV, however, MRV still proved feasible in nearly three-quarters of pediatric subjects and was performed without sedation in individuals as young as 2 months old. The overall feasibility and diagnostic accuracy of each test will be reported after final adjudication and analysis are complete. The study findings will be used to formulate an effective pediatric diagnostic algorithm for diagnosis of central venous catheter-related VTE. The algorithm will be used as guidance for clinical practice as well as outcome assessments for clinical trials of antithrombotic agents in children. Disclosures:Male:Bristol Myer Squibb: Consultancy. Yee:Bristol Myer Squibb: Research Funding. Loewe:Bristol Myer Squibb: Consultancy. Krishnamurthy:Koninklijke Philips NV: Research Funding; Bristol Myer Squibb: Consultancy. Chalmers:Bristol Myer Squibb: Consultancy. Newburger:Merck: Consultancy; Janssen Pharmacutical: Consultancy; Bristol Myer Squibb: Consultancy. Ramirez:Bristol Myer Squibb: Employment. Portman:Bristol Myer Squibb: Employment. Mitchell:Bristol Myer Squibb: Consultancy; Eisai: Consultancy; Boehringer Ingelheim: Consultancy.