key: cord-0788580-gil5alsy
authors: Cercone, Jessica L; Kram, Shawn J; Trammel, Morgan A; Rackley, Craig R; Lee, Hui-Jie; Merchant, James; Kram, Bridgette L
title: Impact of Initial Anticoagulation Targets on Bleeding and Thrombotic Complications for Patients with Acute Respiratory Distress Syndrome Receiving Extracorporeal Membrane Oxygenation
date: 2022-05-13
journal: J Cardiothorac Vasc Anesth
DOI: 10.1053/j.jvca.2022.05.012
sha: 9d6ad6c3de2012568844bfe43df861a29446d172
doc_id: 788580
cord_uid: gil5alsy

Objective : To evaluate the impact of anticoagulation targets and intensity on bleeding events, thrombotic events and transfusion requirements in patients with acute respiratory distress syndrome (ARDS) receiving venovenous extracorporeal membrane oxygenation (ECMO) and continuous infusion heparin Design : Retrospective cohort study Setting : Single-center, large academic medical center Participants : 136 critically ill patients Interventions : Three therapeutic targets were implemented over time and evaluated: (1) no protocol (September 2013-August 2016): no standardized anticoagulation protocol or transfusion thresholds; (2) < 50 sec (September 2016-January 2018): standardized activated partial thromboplastin time (aPTT) goal < 50 sec, maximum heparin infusion rate of 1200 units/hr, transfusion threshold of Hgb < 8; and (3) 40-50 sec (February 2018-December 2019): aPTT goal of 40-50 sec, no maximum heparin infusion rate, transfusion threshold of Hgb < 7. Measurements and Main Results : Continuous variables were compared using Kruskall-Wallis tests and categorical variables were compared using Fisher's exact tests. The primary endpoint, incidence of at least one bleeding event, was highest in the no protocol group, though not statistically different between groups (39.3% vs. 26.7% vs. 34%, p=0.5). Thrombotic complications were similar. The median units of packed red blood cells transfused was highest in the no protocol group (3 vs. 2 vs. 0.5, p<0.001). Conclusions : Anticoagulation protocols standardizing aPTT goals to < 50 or 40-50 sec may be a reasonable strategy for patients receiving venovenous ECMO for ARDS. More restrictive hemoglobin transfusion thresholds in combination with lower aPTT targets may be associated with a reduction in transfusion requirements.

 This retrospective cohort conducted at an experienced ECMO center assessed the impact of both initial anticoagulation strategy and intensity on bleeding and thrombotic complications, defined by ELSO criteria, in patients with acute respiratory distress syndrome (ARDS) receiving venovenous ECMO and continuous infusion heparin.

 For patients receiving heparin, a standardized anticoagulation protocol targeting an aPTT goal of < 50 or 40-50 sec was associated with a reduction in bleeding without increasing thrombotic risk as compared to having no standardized anticoagulation protocol.

 Coupled with standardized transfusion thresholds, packed red blood cell transfusions decreased when targeting these lower aPTT goals.

Acute respiratory distress syndrome (ARDS) is associated with significant morbidity and inhospital mortality rates ranging from 34-46%. 1, 2 Management of ARDS includes low tidal volume ventilation, prone positioning, judicious fluid management, and, in selected cases, neuromuscular blocking agents. 3 Utilization of extracorporeal membrane oxygenation (ECMO) in adult patients has steadily risen since 2009. 4 Venovenous (VV) ECMO serves as a rescue therapy for patients with refractory ARDS when conventional therapies fail to adequately support oxygenation and/or ventilation. 3, 5, 6 Complications of ECMO are primarily related to bleeding or thromboembolism. 7, 8 Introduction of the ECMO circuit into the circulatory system results in an inflammatory and hypercoagulable response. To balance this pro-thrombotic state, a reduction in coagulation factors may induce bleeding disorders following cannulation. [7] [8] [9] Systemic anticoagulation is routinely administered to prevent ECMO circuit thromboses. 10 Bleeding is the most common complication of ECMO, with rates estimated between 29.3-60%. 2, 6, 9, 11 Bleeding complications may include intracranial hemorrhage, pulmonary hemorrhage, surgical site bleeding, and gastrointestinal bleeding. 4, 8, 9 Higher severity of illness, activated partial thromboplastin time (aPTT) levels above goal on systemic anticoagulation, and initiation of ECMO in post-surgical patients have been identified as risk factors for bleeding. 9 Bleeding complications have been independently associated with reduced survival. 9 Neurologic events such as ischemic stroke and intracranial hemorrhage are associated with increased mortality among ECMO patients. [12] [13] [14] [15] Thrombotic complications, including both within the circuit (e.g. oxygenator thrombosis) and the patient (e.g. venous thromboembolism), occur frequently in patients supported with ECMO. 16 Thus, the risk of bleeding and thrombotic complications must be balanced when using systemic anticoagulation.

Due to lack of data supporting standardized anticoagulant goals and monitoring, anticoagulant lab targets during ECMO are variable and may include aPTT, ACT (activated clotting time), thromboelastography (TEG), and anti Xa levels. 16, 17 The accompanying inflammatory response to the device itself may complicate anticoagulant titration to laboratory targets due to diminished sensitivity of aPTT to unfractionated heparin (UFH). 10 Previous studies evaluating anticoagulation targets in patients receiving VV ECMO are limited by small sample sizes and heterogeneous cohorts inclusive of venoarterial (VA) ECMO, which carries a different risk profile for both bleeding and thrombosis. 11, 16, [18] [19] [20] The purpose of this study is to evaluate the association of different initial anticoagulation targets and intensity with bleeding and thrombotic events in patients receiving VV ECMO for ARDS.

This retrospective, observational cohort study was conducted at a large tertiary and quaternary care academic medical center. Patients ≥ 18 years of age were included if they received VV ECMO in the medical intensive care unit for ARDS from September 1, 2013 to December 31, 2019. Included patients received continuous infusion heparin and aPTT monitoring with at least two consecutive aPTTs. Patients were excluded if they received ECMO as a bridge to lung transplant, received ECMO for < 24 hours, were placed on ECMO >24 hours at outside hospital prior to admission to our institution, received ECMO for more than 30 days, were initially anticoagulated with a direct thrombin inhibitor, or had an additional indication for anticoagulation at the time of cannulation (e.g. pulmonary embolism or atrial fibrillation).

This study was IRB-approved with a waiver of informed consent.

Although continuous infusion heparin has remained the standard anticoagulant during the study period, therapeutic targets have changed over time at our institution. From January 2009 to August 2016, aPTT goals were not standardized and were selected at the discretion of the multidisciplinary team. In September 2016, anticoagulation goals were standardized to an aPTT < 50 seconds (sec) with a maximum heparin infusion rate of 1200 units/hour. Additional hematologic targets were also standardized and included hemoglobin ≥ 8 g/dL, platelets ≥ 75 x 10 9 /L, fibrinogen ≥ 100 mg/dL, INR < 2, and antithrombin III > 60% if heparin resistance. Heparin resistance was defined as a heparin rate ≥ 35 units/kg/hr with or without achieving therapeutic targets or clot formation. In February 2018, the ECMO protocol was amended to target a standardized aPTT goal of 40-50 sec with no maximum heparin infusion rate. Hematologic targets were revised based on best available evidence and institutional experience to target hemoglobin ≥ 7 g/dL, platelets ≥ 50 x 10 9 /L, fibrinogen ≥ 100 mg/dL, INR < 2, and the antithrombin III was removed. The aPTT of 40 sec for the lower target goal was added to provide clarification and consistency for providers. During the entire study period, nursing staff collected aPTTs every 6 hours while the patient was receiving heparin. Samples suspected of contamination were redrawn at the discretion of the treating team. The normal reference range for aPTT values at our institution is 26.8-37.1 sec. CardioHelp™ (Maquet/Getinge; Rastatt, Germany) and Rotaflow with a Quadrox-D® membrane oxygenator (Maquet/Getinge; Rastatt, Germany) with both Bioline® and Softline® circuit surface coating were used during the study period.

The primary endpoint was the occurrence of any bleeding event while receiving continuous infusion heparin for VV ECMO at the initial aPTT goal ordered after cannulation during the Evaluable Period. This Evaluable Period was defined as the time from heparin infusion initiation while on ECMO until one of the following occurred: discontinuation of ECMO therapy, change in aPTT goal, discontinuation of heparin for greater than 48 hours, switch to a direct thrombin inhibitor, or death. A bleeding event was defined as a hemoglobin drop of ≥ 2 g/dL in 24 hours, transfusion of more than 3 units of packed red blood cells (pRBCs) in 24 hours, or intracerebral hemorrhage (ICH) identified by 8 computed tomography or magnetic resonance imaging findings. 17 Secondary endpoints included the occurrence of any thrombotic event, blood product administration, time to first bleeding event, time to first thrombotic event, and process indicators including the proportion of aPTT values above, below, and at goal while receiving heparin. Thrombotic events were defined as imaging-confirmed deep vein thrombosis (DVT), pulmonary embolism (PE), ischemic stroke or clot formation requiring an ECMO circuit change. 17 To exclude events attributable to the cannulation procedure, time to first bleeding or thrombotic event was calculated from the time the heparin infusion was initiated. Additional secondary endpoints included duration of mechanical ventilation, duration of ECMO cannulation, length of stay and mortality.

The primary endpoint and all secondary endpoints, except thrombotic events, were collected during the Evaluable Period. Thrombotic events were collected for the duration of hospitalization. 21, 22 Collection of aPTT values began after a 6-hour washout period following heparin bolus administration during cannulation.

Patients meeting inclusion were identified by an institutional data repository, from which demographics, hospitalization data, ECMO-specific data and blood product administration were obtained. Heparin infusion rates and aPTT values were collected by manual chart review. Data were managed by REDCap electronic data capture tool. 23 

Descriptive statistics were used to summarize patient demographics and clinical characteristics between the three groups. Continuous variables were compared using Kruskall-Wallis tests and categorical variables were compared using Fisher's exact tests. For the primary endpoint, the proportion of patients with at least one bleeding event for each intervention group was estimated with the exact binomial 95% confidence interval (CI) using the Clopper-Pearson method. The mean aPTT values and heparin infusion rates while on ECMO were calculated for each patient, and boxplots were used for visualizing the distribution between groups. The proportions of times aPTT values were below, above, or at goal were assessed for each patient based on their anticoagulation targets, then the median was aggregated for each group. To summarize and evaluate this endpoint in the no protocol group, actual aPTT goals documented in the heparin order were used for these patients.

A sensitivity analysis was repeated for the primary and secondary endpoints using the actual aPTT goals specified in the heparin order for patients in the no protocol group. Patients were reassigned into three groups based on the intensity of the aPTT goal range selected: low-intensity (< 50, 40-50, 30-40 sec), moderate-intensity (40-60 or 45-55 sec), and high-intensity (50-70, 50-80, 60-80 sec).

To account for differences in baseline characteristics that may confound the results, a post-hoc logistic regression was performed for both the main analysis and sensitivity analysis using age, sex, duration of the Evaluation Period, duration of ECMO cannulation and aPTT goals as covariates based on differences identified in univariate analysis. All statistical tests were two-sided and assessed at an alpha=0.05 without accounting for multiple testing using R 4.0.0 (R Core Team, Vienna, Austria).

A total of 136 patients met inclusion criteria with 56 patients (41%) included prior to standardized aPTT goals (no protocol), 30 patients (22%) with an initial aPTT goal < 50 sec, and 50 patients (37%) with an initial aPTT goal 40-50 sec group. (Figure 1 ) In the no protocol group, the initial aPTT goal prescribed varied [40-60 sec (n=28), 60-80 sec (n=18), 40-50 sec (n=5), 50-70 (n=2), 30-40 sec (n=1), 45-55 sec (n=1), 50-80 (n=1)]. Baseline characteristics are outlined in Table 1 . Patients in the no protocol group were more like to be male with a longer duration of mechanical ventilation prior to ECMO cannulation. A change in aPTT goal occurred more frequently in the no protocol group and was primarily attributed to a bleeding event. In the study cohort, concomitant administration of antiplatelet medications was infrequent: 5 patients (3.7%) received aspirin monotherapy and 1 patient (0.7%) received both aspirin and clopidogrel. Four patients (2.9%) received antithrombin III and 1 patient (0.7%) received protamine. Laboratory-confirmed HIT was observed in three patients in the study cohort. In addition to clinical criteria, one case was diagnosed by platelet factor 4 assay alone, whereas the other two were confirmed by serotonin release assay.

In all groups, the median intravenous bolus dose of heparin administered during the cannulation procedure was 5000 units. Average heparin infusion rates and aPTT values per patient are summarized in Figure 2 and were higher in the no protocol group compared to the <50 sec and 40-50 sec groups Primary and secondary endpoints are summarized in Table 2 . Overall, 47 patients (34.6%) had at least one bleeding event during the Evaluation Period. The incidence of a bleeding event was highest in the no protocol group (39.3%, 95% CI 28.3-51.1%), followed by the 40-50 sec group (34%, 95% CI 23-46.5%), and was lowest in the < 50 sec group (26.7%, 95% CI 14-43%); however, the difference was not Table 1 .

Results of the sensitivity analysis are summarized in Table 3 . The incidence of at least one bleeding event was not statistically different in the low-, medium-, and high-intensity groups [32.6% (95% CI 24.2 to 41.8%); 31% (95% CI 17.2 to 47.9%); 47.6% (95% CI 28.6 to 67.2%), p=0.39]. The incidence of a thrombotic event was similar between groups. Patients in the moderate-and highintensity groups received more pRBCs compared to the low-intensity group. Demographic and anticoagulation data for the sensitivity analysis are reported in the Supplemental Table 2 .

In the multivariable logistic regression analysis (Table 4) , after accounting for age, sex, duration of Evaluable Period, and duration of ECMO cannulation, the odds of at least one bleeding event for the < 50 sec group was 51% lower (odds ratio [OR] 0.49, 95% CI 0. 16-1.4 ) and for the 40-50 sec group was 33% lower (OR 0.67, 95% CI 0.25-1.75) than the no protocol group; however, the difference was not statistically different.

Among patients with ARDS receiving VV ECMO, a clinically meaningful, though not statistically significant, reduction in the rate of bleeding events was observed after standardizing aPTT to < 50 sec or 40-50 sec. The incidence of at least one thrombotic event during ECMO cannulation was similar between the three groups. Standardized aPTT goals resulted in reduced aPTT variablility and overall heparin exposure. In both the main analysis and sensitivity analysis, lower aPTT goals coupled with standardized transfusion thresholds were associated with a reduction in pRBC transfusions. Similar to other recent retrospective evaluations, our study endpoints reflect an overall improvement in outcomes for patients with ARDS receiving VV ECMO as observed by reductions in length of stay, mortality, duration of mechanical ventilation, and duration of ECMO over time. 24, 25 Though the Extracorporeal Life Support Organization guidelines do not endorse a standardized anticoagulation approach for patients receiving ECMO, several studies have evaluated the impact of aPTT targets on outcomes. 17 A systematic review evaluated bleeding and thrombotic events when comparing aPTT targets of < 60 or > 60 sec. In 5 studies with an aPTT target of > 60 sec, a 56% rate of bleeding and a 7% rate of thrombosis were reported. In contrast, three studies with an aPTT target < 60 sec found a lower incidence of bleeding (8%) and a higher rate of thrombosis (32%). 16 In a registry review of 192 patients who received VV ECMO, a mean aPTT > 53 sec was associated with more pRBC transfusions without a difference in oxygenator changes. 19 In a randomized, unblinded study (n=32), Aubron et al. compared low-dose (aPTT < 45 sec) and therapeutic heparin (aPTT 50-70 sec). Patients receiving therapeutic anticoagulation received more units of heparin per day, achieved higher mean aPTT values, and were more likely to receive at least one unit of pRBCs. No difference in rates of bleeding and thrombosis were observed in this pilot study. Notably, this study included both VA and VV ECMO and crossover between groups, thus limiting extrapolation to a homogenous VV ECMO cohort. 18 Deatrick et al. compared heparin using aPTT titration (goal 45-55 sec) to weight-based dosing (10 units/kg/hr) without titration and found that the titration arm had a higher median aPTT (48 versus 38 sec) but reported no difference in transfusion requirements, bleeding, or need for circuit changes. 20 Despite minor differences in bleeding definitions and transfusion thresholds, these studies align with our findings and support that ECMO centers define a maximum aPTT target to mitigate bleeding risk.

The rates and timing of bleeding events in our study are comparable to those reported in other ECMO centers. 2, 11, 16, 18 In an observational cohort of patients receiving VV ECMO, bleeding was observed in 41.4% of patients when targeting aPTT goal of 40-50 sec, and transfusion of pRBCs was independently associated with mortality. 11 In a retrospective cohort of patients receiving VV ECMO, ICH occurred in 11% of included patients with a median of 4 days after ECMO cannulation. 15 Heparin was titrated to a goal aPTT 50-75 sec in this study, though anticoagulation intensity was not evaluated as a risk factor for ICH. Aubron et al. reported a median time to the first bleeding event of 4 days, with a previous day aPTT > 70 as a risk factor independently associated with bleeding in a mixed cohort of VA-and VV ECMO. 9 This study has limitations related to retrospective design. Bleeding endpoints were evaluated while the patient was receiving heparin at a pre-specified aPTT goal; therefore, rates of bleeding reflect only the initial aPTT goal during the Evaluable Period and may underestimate bleeding that occurred during the entire ECMO course if a change in aPTT goal occurred. The patient population was restricted to VV ECMO and may not be generalizable to patients on VA ECMO. While there was no standard transfusion protocol prior to 2016, patients were typically transfused to target Hgb of 8-10 g/dL. Therefore, the primary endpoint had the potential to be confounded by the implementation of lower transfusion thresholds during the study period, however the proportion of patients with a bleeding event characterized by receipt of > 3 units of pRBCs within 24 hours was low in the no protocol group and highest in the 40-50 sec group.

Strengths of this study are numerous. Objective criteria were utilized to define clinically significant major bleeding. By evaluating bleeding events only after heparin initiation and aPTT values after a 6-hour washout period, any bleeding event or elevated aPTT due to the cannulation procedure itself was excluded. Bleeding outcomes and aPTT values reported in this study are correlated with the pre-specified aPTT goal because the study was designed to cease data collection upon a change in aPTT goal, discontinuation of heparin, or decannulation. Additionally, the three groups had similar proportion of aPTT values above goal, so we do not suspect the times outside of the therapeutic range to be a cause of differences in rates of bleeding. From an operational standpoint, targeting an aPTT of 40-50 sec may be advantageous as it defines explicit instructions for nursing and providers when titrating heparin infusions.

Among patients with ARDS receiving VV ECMO, a clinically meaningful, though not statistically significant, reduction in the rate of bleeding events was observed after standardizing the aPTT goal to < 50 sec or 40-50 sec. Lower aPTT targets and restrictive transfusion thresholds pursuant to a protocol may reduce the need for pRBC transfusion. Anticoagulation protocols targeting lower aPTT goals of < 50 or 40-50 sec may be a reasonable strategy for patients receiving VV ECMO for ARDS. hours, switch to a direct thrombin inhibitor, or death.

Patterns of Care, and Mortality for Patients With Acute Respiratory Distress Syndrome in Intensive Care Units in 50 Countries

Systematic review and meta-analysis of complications and mortality of veno-venous extracorporeal membrane oxygenation for refractory acute respiratory distress syndrome

An Official American Thoracic Society/European Society of Intensive Care Medicine/Society of Critical Care Medicine Clinical Practice Guideline: Mechanical Ventilation in Adult Patients with Acute Respiratory Distress Syndrome

Extracorporeal Life Support Organization Registry Report

Efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR): a multicentre randomised controlled trial

Extracorporeal Membrane Oxygenation for Severe Acute Respiratory Distress Syndrome

Extracorporeal membrane oxygenation-hemostatic complications

Identification of acquired coagulation disorders and effects of target-controlled coagulation factor substitution on the incidence and severity of spontaneous intracranial bleeding during veno-venous ECMO therapy

Predictive factors of bleeding events in adults undergoing extracorporeal membrane oxygenation

Extracorporeal membrane oxygenation in the adult: a review of anticoagulation monitoring and transfusion

Anticoagulation and Transfusions Management in Veno-Venous Extracorporeal Membrane Oxygenation for Acute Respiratory Distress Syndrome: Assessment of Factors Associated With Transfusion Requirements and Mortality

Brain Injury Is More Common in Venoarterial Extracorporeal Membrane Oxygenation Than Venovenous Extracorporeal Membrane Oxygenation: A Systematic Review and Meta-Analysis

CNS Complications in Adult Patients Treated With Extracorporeal Membrane Oxygenation

Modifiable Risk Factors and Mortality From Ischemic and Hemorrhagic Strokes in Patients Receiving Venoarterial Extracorporeal Membrane Oxygenation: Results From the Extracorporeal Life Support Organization Registry

Timing, Outcome, and Risk Factors of Intracranial Hemorrhage in Acute Respiratory Distress Syndrome Patients During Venovenous Extracorporeal Membrane Oxygenation

Anticoagulation Practices during Venovenous Extracorporeal Membrane Oxygenation for Respiratory Failure. A Systematic Review

Extracorpor Life Support Organ

Low-Dose Versus Therapeutic Anticoagulation in Patients on Extracorporeal Membrane Oxygenation: A Pilot Randomized Trial

Comparison of Coagulation Parameters, Anticoagulation, and Need for Transfusion in Patients on Interventional Lung Assist or Veno-Venous Extracorporeal Membrane Oxygenation

Pilot study evaluating a non-titrating, weight-based anticoagulation scheme for patients on veno-venous extracorporeal membrane oxygenation

Prevalence and Risk Factors for Thrombotic Complications Following Venovenous Extracorporeal Membrane Oxygenation: A CT Scan Study

Incidence and Risk Factors for Cannula-Related Venous Thrombosis After Venovenous Extracorporeal Membrane Oxygenation in Adult Patients With Acute Respiratory Failure

The REDCap consortium: Building an international community of software platform partners

Comparison of Anticoagulation Strategies in Patients Requiring Venovenous Extracorporeal Membrane Oxygenation: Heparin Versus Bivalirudin

Extracorporeal Life Support Organization Registry International Report