AIM: We sought to quantify the location and volume of thrombus in used hollow-fiber membrane oxygenators and correlate the volume of thrombus with patient demographics, flow characteristics and anticoagulation parameters. METHODS: Hollow-fiber membrane oxygenators (Quadrox D, Maquet, Rastatt, Germany) were collected after clinical use in ECMO circuits and divided into sections. Each section was digitally imaged and analyzed using ImageJ software. The location and total volume (cm3) of thrombus was calculated for different sections. In an effort to predict thrombus formation, we correlated thrombus volume with possible aggravating and mitigating variables. RESULTS: We collected 41 oxygenators from 27 patients. Twenty-seven (66%) were configured in the veno-venous mode and 14 (34%) in the veno-arterial mode. The median duration of use was 131 hours (interquartile range 61-214 hours). Eighteen (44%) were removed when the patient recovered, six (15%) were removed after withdrawal of care and seventeen (41%) were exchanged. The median volume of thrombus was 11.4 cm3 (interquartile range 2.2-44.5 cm3). CONCLUSIONS: A multivariable linear regression model suggested that the combination of median flow, configuration of ECMO and visible thrombus partially predicted internal thrombus volume (adjusted R2=0.39).
AIM: We sought to quantify the location and volume of thrombus in used hollow-fiber membrane oxygenators and correlate the volume of thrombus with patient demographics, flow characteristics and anticoagulation parameters. METHODS: Hollow-fiber membrane oxygenators (Quadrox D, Maquet, Rastatt, Germany) were collected after clinical use in ECMO circuits and divided into sections. Each section was digitally imaged and analyzed using ImageJ software. The location and total volume (cm3) of thrombus was calculated for different sections. In an effort to predict thrombus formation, we correlated thrombus volume with possible aggravating and mitigating variables. RESULTS: We collected 41 oxygenators from 27 patients. Twenty-seven (66%) were configured in the veno-venous mode and 14 (34%) in the veno-arterial mode. The median duration of use was 131 hours (interquartile range 61-214 hours). Eighteen (44%) were removed when the patient recovered, six (15%) were removed after withdrawal of care and seventeen (41%) were exchanged. The median volume of thrombus was 11.4 cm3 (interquartile range 2.2-44.5 cm3). CONCLUSIONS: A multivariable linear regression model suggested that the combination of median flow, configuration of ECMO and visible thrombus partially predicted internal thrombus volume (adjusted R2=0.39).
Authors: Andreas Kaesler; Freya Lilli Rudawski; Mark Oliver Zander; Felix Hesselmann; Isaac Pinar; Thomas Schmitz-Rode; Jutta Arens; Ulrich Steinseifer; Johanna Charlotte Clauser Journal: Cardiovasc Eng Technol Date: 2021-09-16 Impact factor: 2.305