Platelet Activation and Severe Bleeding During Extracorporeal Carbon Dioxide Removal in Chronic Obstructive Pulmonary Disease Patients

A recent randomized controlled trial of extracorporeal CO2 removal (ECCO2R) in chronic obstructive pulmonary disease (COPD) patients reported bleeding as a major concern.1 While different mechanisms, such as thrombocytopenia, coagulation factor consumption, acquired von Willebrand disease,2 severe endothelial dysfunction,3 and anticoagulation-related side effects, could explain bleeding during implantation, preimplantation hemostatic status has not been fully explored. We investigated the association between platelet activation before ECCO2R initiation and the occurrence of severe bleeding during ECCO2R therapy in 16 COPD patients. We found that severe bleeding during ECCO2R was associated with a higher expression of CD40 Ligand (sCD40L) before ECCO2R. Extracorporeal CO2 removal, which is based on old concepts but with continuous technical improvements,4 has the potential to shift our current approaches to treating both acute hypercapnic failure (mainly acute exacerbation AE of COPD AE-COPD patients) and acute hypoxemic respiratory failure (mainly acute respiratory distress syndrome patients). However, hemorrhagic and thrombotic side effects have been reported, along with hemolysis.4–7 Platelet activation or hyper-reactivity have been reported in stable COPD and AE-COPD patients.8 Therefore, we searched for associations between corresponding biologic status and further occurrences of hemorrhagic complications during ECCO2R. We explored these points using a formalized ECCO2R register (NCT02965079) combined with a study project (Hector: Hemostasis and ECCO2R).3 Methods This study included consecutive AE-COPD patients with an ECCO2R decision therapy in a single center. We recorded severe hemorrhagic events (types 3–5 of the Bleeding Academic Research Consortium standardized classification)9 and analyzed classic hemostasis parameters before the implantation of ECCO2R devices and during ECCO2R courses. Two medical devices were used: the Hemolung device, ALung, Pittsburgh, with 15.5 Fr veno–venous catheters (either femoral or right internal jugular) or the iLA Activve device, Xenios/Novalung, Heilbronn, Germany, with veno–venous catheters, either femoral (24 Fr) or right internal jugular (18 Fr). All patients were treated with unfractionated heparin with an aXa goal of 0.3 to 0.6 IU/ml. We quantified plasma platelet activation markers—sCD40L and soluble P-selectin (sP-sel)—and endothelial activation markers—E-selectin (sE-sel) and Angiopoietin-2 (Ang-2). For the descriptive analysis, data are expressed as median (interquartile range) for continuous variables and frequencies and percentages for categorical variables. To compare continuous data, we used the Mann-Whitney U test, while Fisher's exact test was used for categorical variables. All statistical analyses were conducted as two-sided tests, with a significance level of p < 0.05. We used R Studio software, including R version 4.3.1. Results Sixteen acute AE-COPD patients were included. Their main demographics and clinic-biologic characteristics are described in Table 1 and were, in part, reported in a previously published study focusing on the course of endothelial dysfunction during ECCO2R.3 Eight patients were treated with the Hemolung device, and eight patients were treated with the iLA Activve device. A severe hemorrhagic event was observed in five patients. The characteristics of these events are reported in Table 2. The duration of ECCO2R was 5 4, 8 days for 11 patients without severe hemorrhagic complication and 8 8, 8 days for five patients with severe hemorrhagic complication (p < 0.001). Patients with severe bleeding during support displayed no differences in leukocytes, hemoglobin, and platelets counts before ECCO2R implantation (respectively, with a p value of 0.336, 0.212, and 0.193). No difference in daily unfractionated heparin (UFH), daily aXa, or daily platelet count during ECCO2R therapy was observed.3 Table 1. - Main Demographics, Clinical and Biologic Characteristics at Baseline (Before ECCO2R Initiation) in 16 AE-COPD Patients, Separated Between 11 Patients Without or Five Patients With Severe Hemorrhagic Complication During ECCO2R No Severe Hemorrhage Severe Hemorrhage p Demography and clinical course n 11 5 NA Age (years) 67 60, 70 64 63, 76 0.530 Male/female, n/ n 5/6 2/3 1 SAPS2 33 27, 47 40 32, 45 0.649 Charlson score 3 1, 5 4 2, 4 0.954 BMI (kg/m2) 27 24, 39 25 24, 25 0.533 Platelet aggregation inhibitor (n) 1 (clopidogrel–aspirin) 0 1 ECCO2R duration (days) 5 4, 8 8 8, 8 0.038 ICU survival, n (%) 9 (82%) 3 (60%) 0.755 ICU length of stay (days) 21 16, 35 15 15, 16 0.172 Biologic values before ECCO2R implantation PaO2 (mm Hg) 67 62, 76 69 66, 76 0.734 PaCO2 (mm Hg) 70 67, 73 67 62, 70 0.280 pH 7.27 7.22, 7.32 7.30 7.28, 7.30 0.776 Leucocytes (G/L) 13 8.45, 16.80 10.20 9.70, 11.40 0.336 Hemoglobin (g/100 ml) 12.50 10.65, 13.25 10.80 7.30, 11.10 0.212 Platelets (G/L) 261 191, 359 177 154, 247 0.193 Fibrinogen (g/L) 4.9 4.4, 6.2 5.3 5.0, 5.7 0.955 INR 1.30 1.05, 1.35 1.20 1.10, 1.20 0.455 aPTT ratio 0.93 0.91, 1.08 1.07 0.90, 1.20 0.621 D-dimer (ng/ml) 929 668, 1625 1,654 755, 3,319 0.462 sCD40L (ng/ml) 1,420 1,196, 1,510 2,624 2,574, 2,673 0.033 sP-sel (ng/ml) 28,311 25,124, 32,081 61,931 49,531, 74,330 0.059 sE-sel (ng/ml) 35,299 19,268, 48,207 63,507 45,260, 81,753 0.475 Ang-2 (pg/ml) 7,658 5,456, 12,505 15,969 11,406, 20,532 0.637 Ang-2, angiopoietin-2; aPTT, activated partial thromboplastin clotting time; BMI, body mass index; ICU, intensive care unit; INR, international normalized ratio; SAPS, simplified acute physiology score; sCD40L, CD40 ligand; sE-sel; E-selectin; sP-sel, soluble P-selectin. Table 2. - Characteristics of the Severe Bleeding Events Observed in Five Patients Patient # 1 # 2 # 3 # 4 # 5 Timing of bleeding (ECCO2R days) D7 D1 D7 D1 D6 ECCO2R cessation No No Yes No No BARC classification 3a 3b 3b 3a 3b Site of bleeding ENT ECCO2R catheter insertion* Retroperitoneal ECCO2R catheter insertion* Gastrointestinal Number of packed red blood cells 0 2 3 2 2 ECCO2R device Hemolung Hemolung iLA Activve iLA Activve iLA Activve Catheter size (Fr) 15.5 15.5 18 24 18 Access point Right internal jugular Right internal jugular Right internal jugular Right femoral Right internal jugular No patient suffered from more than a severe bleeding event.*No difficulties during catheter insertion.BARC, Bleeding Academic Research Consortium; ECCO2R, extracorporeal CO2 removal. Regarding hemostasis, patients with severe bleeding during support had similar levels of D-dimer and fibrinogen before ECCO2R implantation (respectively, with a p value of 0.462 and 0.955). No differences in prothrombin time (PT) and activated partial thromboplastin clotting time (aPTT) ratios were observed (Table 1). We then analyzed plasma levels of platelet and endothelial activation markers. In patients with further bleeding during support, sCD40L was significantly increased (p = 0.033), while the difference in sP-sel values did not reach statistical significance (p = 0.059). Endothelial activation markers sE-sel and Ang-2 levels were comparable between groups before ECCO2R implantation. Since blood acidification has been recently described to modify platelet activation during ECCO2R,10 we compared arterial blood gas analysis before ECCO2R implantation in patients with or without bleeding during support and did not find any differences for pH and arterial partial pressure of carbon dioxide (PaCO2). Discussion Overall, our results obtained in a homogeneous group of very severe AE-COPD patients suggest that increased platelet activation before ECCO2R implantation may be one mechanism that contributes to bleeding during support, besides other factors such as the type of device, the blood flow rate, the types of vascular access, and the anticoagulation regimen. Platelet activation was more pronounced in the subgroup of patients further suffering from severe bleeding during ECCO2R. We hypothesize that platelet activation observed before ECCO2R implantation could impair their ability to aggregate. Platelets are a major constituent of almost all thrombi, independent of their location and composition in red blood cells and fibrin, suggesting that they are actors impossible to circumvent in thrombus formation during extracorporeal support. Accordingly, monitoring platelet parameters before ECCO2R but also ECMO implantation could be of value for predicting whether a patient may bleed during support. However, positive validation studies would be a necessary step before implementing such biologic monitoring in the clinical practice. Additionally, investigating bleeding in patients treated with low-flow rotational blood pumps ECCO2R devices seems mandatory, considering their major differences with centrifugal pumps ECCO2R devices. Such investigations should include the evaluation of the influence of platelet activation on bleeding.