Extracorporeal membrane oxygenation (ECMO) is increasingly used as rescue therapy for critically ill patients who are expected to die. The high ECMO morbidity and mortality1 warrant awareness of determinants of poor outcome. Among the factors associated with higher in-hospital mortality during ECMO is bleeding events.2 Hemorrhage is frequent in critically ill patients on ECMO3 as they are exposed to anticoagulation—to prevent circuit clotting—and are susceptible to coagulopathy and platelet dysfunction. Intracranial hemorrhage (ICH) during ECMO is associated with worse outcome that significantly impacts patients’ survival and quality of life.4 Most studies on ICH during ECMO were done in the pediatric population.5 Herein, we aim to determine the incidence and predictors of ICH occurring during ECMO support in adults.
An IRB-approved retrospective review was performed on adults >18 years who received ECMO from 2007 to 2013. The main outcome variable was the onset of imaging-confirmed ICH occurring during ECMO. Because risk factors for ischemic and hemorrhagic stroke are different, we excluded cases with ischemic stroke. Cases with ICH after ECMO removal were also excluded from the analysis. The ECMO pumps used during the study period was mainly Bio-Medicus BP80 (Medtronic, Inc. MN) or less commonly Thoratec CentriMag (Thoratec, Pleasanton, CA) pump. Anticoagulation during ECMO was mainly achieved by heparin infusion guided by activated partial thromboplastin time (aPTT) and activated clotting time (ACT). Using Shapiro–Wilk test, all continuous variables were not normally distributed and accordingly were compared using the Mann–Whitney U test and categorical variables were compared using χ2 test. To examine independent predictors for ICH during ECMO, we used multivariate logistic regression analysis with backward stepwise elimination. The model included age, mean ECMO duration, mean highest ACT level and lowest platelet count during EMCO, and the requirement for intra-aortic balloon counter pulsation before ECMO support.
One hundred fifty-four patients (mean age = 51 years and 75.3% males) received ECMO for cardiac (n = 115) or pulmonary (n = 39) indications. Eighty-one percent of the cases received venoarterial ECMO and 19% received venovenous ECMO. Twelve subjects (7.8%) developed imaging-confirmed ICH during ECMO support. The majority of cases (98.1%) were on heparin intravenous infusion during ECMO support to prevent circuit clotting. There was no significant difference between cases with and without ICH according to ECMO indication. Compared with ECMO cases without ICH, those with ICH had longer mean ECMO duration (9.41 ± 7.51 vs. 5.37 ± 6.64 days, p = 0.007), and a higher mean ACT levels (647 ± 301 vs. 442 ± 258 seconds, p = 0.016)—but no significant difference in aPTT, INR, and platelet count. Patients who developed ICH also experienced significantly higher frequency of other non-ICH bleeding episodes (75% vs. 36.6%, p = 0.017), and required more frequent platelet transfusion (114 ± 158 vs. 43 ± 66 units, p = 0.016) during ECMO support (Table 1). Ten of the 12 cases (83.3%) with ICH died during index hospitalization. In the multivariate model (N = 139, p from Hosmer and Lemeshow = 0.829), a longer ECMO duration (odds ratio [OR] = 1.074, 95% confidence interval [CI] = 1.005–1.148, p = 0.035) and ACT level (OR = 1.003, 95% CI 1.001–1.005, p = 0.012) were independently associated with the risk of ICH after controlling for patients’ age, mean lowest platelet value during ECMO and the requirement for intra-aortic balloon counter pulsation before ECMO support (Figure 1; Table 2).
The current study shows that ECMO duration in those with ICH is approximately double those without ICH and is an independent risk for ICH occurring during ECMO support. This is probably related to the longer exposure to anticoagulant medications, which are crucial for all patients on ECMO. The foreign surface in ECMO circuit activates platelets and clotting system, and therefore, anticoagulation is mandatory to prevent thromboembolism originating in this nonendothelialized circuit. Moreover, a proportion of ECMO patients have preexisting platelet dysfunction and coagulopathy due to various reasons, including sepsis and hepatic dysfunction. In addition, most of these patients are cardiac patients on single or dual antiplatelet therapy. The fact that ACT is significantly higher in patients who developed ICH suggests that a longer ECMO duration as a risk for ICH may be driven by longer exposure to anticoagulant medications and therefore higher chance of exposure to periods of supratherapeutic levels of anticoagulants.
Most ECMO studies in adults combined ischemic and hemorrhagic stroke in a common risk factor analysis.6 The incidence of ICH as reported to ELSO is 4% and 2% in adults undergoing ECMO for respiratory and cardiac indications, respectively.7 In a recent nationwide analysis of 23,950 subjects receiving ECMO, 3.6% had ICH with mortality of 59.7% and morbidity of 9.5%, and this incidence almost doubled from 2.4% in 2001–2006 to 4.5% in 2007–2011.4 The incidence of ICH in our study was 7.8%, which is higher and may be related to complexity of our patient population because our hospital is a tertiary care facility serving the Tampa Bay and the surrounding counties.
Intracranial hemorrhage has traditionally been considered a contraindication for systemic anticoagulation as it may worsen hemorrhage and neurological outcome. Although ICH occurring after withdrawal of ECMO may not have such a grave outcome8 and may not promote limitations in patient care, ICH occurring during ECMO support is more serious as these patients have to be maintained on anticoagulation. These patients usually have withdrawal of care due to futility or may continue ECMO with a less aggressive anticoagulation protocol but remain at risk of extension of bleed. We found a significantly high in-hospital mortality in patients who experienced ICH (83.3%) and we suspect that the lack of significant mortality difference between groups is related to the small sample size. The main study limitations are the small number of cases with main outcome variable, retrospective design, and being a single-center study. Based on our findings, we recommend more frequent neurological checks and monitoring of coagulation parameters (especially ACT) in subjects who remain on ECMO for longer duration and to attempt earlier rather than later weaning from ECMO whenever feasible to avoid risk of ICH.
1. Omar HR, Mirsaeidi M, Socias S, et al: Plasma free hemoglobin is an independent predictor of mortality among patients on extracorporeal membrane oxygenation support. PLoS One 2015.10: e0124034.
2. Aubron C, Cheng AC, Pilcher D, et al: Factors associated with outcomes of patients on extracorporeal membrane oxygenation support: A 5-year cohort study. Crit Care 2013.17: R73.
3. Brogan TV, Thiagarajan RR, Rycus PT, Bartlett RH, Bratton SL: Extracorporeal membrane oxygenation in adults with severe respiratory failure: A multi-center database. Intensive Care Med 2009.35: 2105–2114.
4. Nasr D, Rabinstein A: Stroke outcomes in patients receiving extracorporeal membranous oxygenation. Abstract W MP25.
5. Hardart GE, Fackler JC: Predictors of intracranial hemorrhage during neonatal extracorporeal membrane oxygenation. J Pediatr 1999.134: 156–159.
6. Lan C, Tsai PR, Chen YS, Ko WJ: Prognostic factors for adult patients receiving extracorporeal membrane oxygenation as mechanical circulatory support—A 14-year experience at a medical center. Artif Organs 2010.34: E59–E64.
8. Le Guennec L, Bertrand A, Laurent C, et al: Diffuse cerebral microbleeds after extracorporeal membrane oxygenation support. Am J Respir Crit Care Med 2015.191: 594–596.