Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Mar 1.
Published in final edited form as: ASAIO J. 2021 Mar 1;67(3):290–296. doi: 10.1097/MAT.0000000000001230

Thrombosis and bleeding in extracorporeal membrane oxygenation (ECMO) without anticoagulation: a systematic review

Sven R Olson 1, Catherine R Murphree 1, David Zonies 2, Andrew D Meyer 3, Owen JT McCarty 4, Thomas G DeLoughery 1, Joseph J Shatzel 1,4
PMCID: PMC8623470  NIHMSID: NIHMS1749560  PMID: 33627603

Abstract

Extracorporeal membrane oxygenation (ECMO) causes both thrombosis and bleeding. Major society guidelines recommend continuous, systemic anticoagulation to prevent thrombosis of the ECMO circuit, though this may be undesirable in those with active, or high risk of, bleeding. We aimed to systematically review thrombosis and bleeding outcomes in published cases of adults treated with ECMO without continuous systemic anticoagulation. Ovid MEDLINE, Cochrane CENTRAL and CDSR, and hand search via SCOPUS were queried. Eligible studies were independently reviewed by two blinded authors if they reported adults (≥18 years) treated with either VV- or VA-ECMO without continuous systemic anticoagulation for ≥24 hours. Patient demographics, clinical data, and specifics of ECMO technology and treatment parameters were collected. Primary outcomes of interest included incidence of bleeding, thrombosis of the ECMO circuit requiring equipment exchange, patient venous or arterial thrombosis, ability to wean off of ECMO, and mortality. Of the 443 total publications identified, 34 describing 201 patients met our inclusion criteria. Most patients were treated for either acute respiratory distress syndrome or cardiogenic shock. The median duration of anticoagulant-free ECMO was 4.75 days. ECMO circuity thrombosis and patient thrombosis occurred in 27 (13.4%) and 19 (9.5%) patients, respectively. Any bleeding and major or “severe” bleeding was reported in 66 (32.8%) and 56 (27.9%) patients, respectively. 40 patients (19%) died. While limited by primarily retrospective data and inconsistent reporting of outcomes, our systematic review of anticoagulant-free ECMO reveals an incidence of circuity and patient thrombosis comparable to patients receiving continuous systemic anticoagulation while on ECMO.

Keywords: Extracorporeal membrane oxygenation, thromboembolism, hemostasis, heparin, hemorrhage

Introduction:

Despite aggressive supportive care and disease-specific treatments, in-hospital mortality of critically ill patients with severe cardiac and/or respiratory failure approaches 50%.1 A variety of ventilatory and mechanical circulatory support technologies have been explored as adjunctive therapies in these scenarios, including prone positioning, ventricular assist devices, and extracorporeal membrane oxygenation (ECMO). In particular, ECMO has the capability of replacing both pulmonary and cardiac functions, and the number of United States medical centers offering this technology has increased substantially over the past two decades from 120 to nearly 400.2

ECMO consists of an externalized circuit that provides blood oxygenation (venovenous, VV) or both oxygenation and circulatory support (venoarterial, VA). Traditionally, VV-ECMO and VA-ECMO are utilized for acute respiratory distress syndrome (ARDS) or cardiogenic shock, respectively, though these techniques have also been applied to patients undergoing extensive broncheoalveolar lavage, as a bridge to ventricular assist device placement or heart transplant, and perioperatively for major thoracic surgeries among other indications.3–5 A recent meta-analysis of two large randomized trials and 3 observational studies found that VV-ECMO provided a statistically significant mortality benefit in ARDS compared to conventional mechanical ventilation.6 Data for VA-ECMO in cardiac failure is limited to non-randomized cohort studies, though a meta-analysis similarly suggested a significant survival benefit.7

Yet data in favor of ECMO has historically been tempered by frequently encountered thromboembolic and bleeding complications. The rate of any bleeding during ECMO is reported to be as high 29%,8 with a 10% risk of major bleeding and 4–10% risk of intracranial hemorrhage (ICH).9,10 This high bleeding risk can be attributed to multiple patient and treatment-related factors associated with ECMO. First, patients requiring ECMO are often either critically ill or in the post-operative period after major surgery, both scenarios associated with heightened risks for bleeding complications.11,12 Second, the ECMO circuit and pump themselves impair multiple innate primary and secondary hemostatic mechanisms. Paradoxically, many of the same risk factors that contribute to bleeding during ECMO also confer an elevated risk of thromboembolism; critical illness, sedation, frequent blood product transfusion, non-pulsatile blood flow, and exposure of blood to the non-biologic ECMO circuit all contribute to a substantial prothrombotic milieu. This leads to clotting of the ECMO oxygenator in anywhere from 10–16% of cases, leading to reduced ECMO pump efficiency and potentially complete exchange of ECMO circuit components.13 For this reason, ECMO protocols have routinely included the use of systemic anticoagulation, most often with unfractionated heparin (UFH) with the primary goal of minimizing circuit thrombosis, a practice endorsed by international guidelines from the Extracorporeal Life Support Organization (ELSO).14

Despite these recommendations, cases of ECMO run entirely without continuous systemic anticoagulation, both as a planned study protocol and in the setting of pre-existing major bleeding, have been reported with increasing frequency. Given the easily modifiable nature of anticoagulation as a potential strategy to mitigate bleeding, we sought to systematically evaluate the literature on ECMO performed without continuous systemic anticoagulation to determine the thrombotic and bleeding outcomes of this practice.

Methods

Search Strategy

A systematic literature search of Ovid MEDLINE, Cochrane Central Register of Controlled Trials (CENTRAL) and Cochrane Database of Systematic Reviews (CDSR), without time restriction, was undertaken. A full description of search strategies can be found in e-Figure 1. We also hand-searched for additional relevant articles via SCOPUS. Inclusion criteria were English-language articles reporting the outcomes of adults (≥18 years) treated with ECMO, but without continuous systemic anticoagulation, for a minimum of 24 hours. Studies reporting administration of an initial bolus of systemic anticoagulation, as well as subcutaneously administered, prophylactic-dose anticoagulation (unfractionated or low-molecular-weight heparin), were allowed. Bolus doses of UFH in the amount of 50–100 U/kg typically utilized for ECMO cannulation would be expected to clear from the circulation within hours,15 and prophylactic dose anticoagulants are recommended by major society guidelines for VTE prevention in critically-ill patients, with a “strong recommendation” based on moderate certainty in the evidence of effects;12 therefore, neither were considered confounders.

Study selection process

Two authors (SO and CM) independently identified studies eligible for inclusion based on an initial screen of reference titles and abstracts. Articles (including meeting abstracts) were included for further review if they met the pre-specified inclusion criteria listed above. Randomized controlled trials (RCTs), prospective and retrospective observational studies were included. Article records were independently reviewed for inclusion in duplicate, and discrepancies were resolved by consensus between the two reviewing authors (SO and CM). Bias assessment was deemed irrelevant, as the studies ultimately included for final review included mainly retrospective case reports and case series.

Data collection

Baseline characteristics of individual patient cases treated with anticoagulation-free ECMO included age; indication(s) for and type of ECMO (VV or VA) used; average blood flow rate; utilization of heparin or other anticoagulant-bonded ECMO circuit, type and frequency of other antiplatelet or anticoagulant medications used, frequency of blood product transfusion including packed red blood cells, platelets, plasma, cryoprecipitate, and prothrombin complex concentrates, indication(s) for withholding systemic anticoagulation during ECMO, and median duration of anticoagulation-free ECMO.

The primary outcome of interest obtained from included studies was thromboembolic complications, both within the patient and the ECMO circuit. Patient thrombosis was stratified by venous or arterial events, and use of VV- or VA-ECMO. Specific attention was paid to the incidence of ECMO circuit thrombosis that required exchange of any circuit component. In addition, bleeding outcomes were recorded, including location and severity. Effort was taken to stratify bleeding events by criteria established by the International Society of Thrombosis & Haemostasis (ISTH).16 When severity of bleeding events were unclear or authors used alternative definitions (e.g. “severe” rather than major bleeding), both reviewers (SO and CM) independently reviewed the study to come to a consensus. Finally, survival and ECMO-weaning outcomes were recorded. These results were summarized via descriptive statistics. Consideration was given to explore for associations between independent variables and the outcomes of interest using linear regression, though this was ultimately deferred given the incomplete reporting of the variables and outcomes, as well as the limitation of data to primarily case reports and series.

Results

A total of 443 publications met our initial search criteria. After removal of duplicates, 426 were screened for inclusion, of which 34 were ultimately selected for full text review. Given the lack of multi-arm cohort or randomized controlled trials, we were unable to perform meta-analysis of outcomes of interest. Studies examined for quantitative synthesis included 32 case reports or case series and 2 prospective, single-arm, observational studies. Details for study review, inclusion and exclusion can be found in e-Figure 1; details regarding the requisite PRISMA checklist for systematic reviews can be found in e-Figure 2. Bias assessment was not conducted due to the inclusion of only non-randomized studies.

201 individual patients with a median age of 42.5 years were treated with anticoagulation-free ECMO among the 34 reviewed studies; see Table 1 for full details of patient and ECMO characteristics. Cases were fairly evenly split between VV- and VA-ECMO (111 and 90, respectively). Cardiogenic shock and ARDS were the two most common indications for ECMO. ECMO as a bridging therapy peri-operatively was also frequent in the included studies. Approximately half (55.9%) of included studies described the use of biocompatible ECMO tubing; for a large number of studies, details on whether such technology was employed was not reported. The most common reason for performing ECMO without continuous systemic anticoagulation was due to a planned study protocol in the two prospective, single-arm cohort studies (72 patients, 35.8% of all patients). Other common reasons included thrombocytopenia, ICH and diffuse alveolar hemorrhage (DAH). 44.5% of patients not receiving therapeutic-dose, systemic anticoagulation were receiving other antiplatelet or prophylactic-dose anticoagulant medications during ECMO. ECMO was run without any systemic anticoagulation for a median of 4.75 days, though was extended up to 130 days in one case. The total duration of anticoagulant-free ECMO in the included studies was 304.7 days.

Table 1.

ECMO details and patient characteristics from included studies

Patient/ECMO characteristic Outcome
Individual patients reported, n 201
Median patient age, n (range) 42.5 (16–75)
Patients receiving VV or VA ECMO (n/n) 111/90
Studies reporting use of anticoagulant-bonded ECMO circuitry, n (%) 19 (55.9)
Justification for ECMO Cardiogenic shock (80)
Acute respiratory distress syndrome (69)
  -Reason unspecified (7)
  -PNA (bacterial) (22)
  -PNA (viral) (17)
  -Aspiration (10)
  -Extra-pulmonary sepsis (3)
  -Post-cardiac surgery (8)
  -Goodpasture syndrome (1)
  -Granulomatosis with polyangiitis (1)
Traumatic lung injury (20)
  -Hemopneumothorax (7)
  -Lung contusion (12)
  -Flail chest (1)
Diffuse alveolar hemorrhage (6)
  -No reason identified (1)
  -Lupus pneumonitis (3)
  -Anticoagulation-associated (1)
  -Bronchial-arterial fistula (1)
Acute interstitial PNA (1)
Bronchial fistula (2)
Septic shock (1)
Pulmonary embolism (1)
CPB wean failure (4)
Extracorporeal CPR (5)
Bridge to lung transplant (9)
Post-pneumonectomy (3)
Justification for anticoagulant-free ECMO (patients may have had multiple) Study protocol (72)
Institutional protocol (57)
Thrombocytopenia (25)
Intracranial hemorrhage (16)
“Bleeding”, “diffuse bleeding”, “surgery”, or not described (8)
Diffuse alveolar hemorrhage (7)
Hemoptysis (7)
Hemothorax (6)
Elevated ACT while on heparin (6)
Bronchial bleeding (3)
Mediastinal bleeding (3)
Liver laceration (2)
Bleeding from chest tube (1)
Splenic laceration (1)
Gastrointestinal bleed (1)
Retroperitoneal bleeding (1)
Femoral arterial bleed (1)
ECMO flow rate, average (range), L/min 3.8 (2.5–10)
Administration of other anticoagulants/antiplatelet agents, n (%) Total: 89 (44.3)
Aspirin (7)
Clopidogrel (8)
Naphamostat (1)
LMWH (prophylactic dose) (69)
UFH (prophylactic dose) (4)
Median number of ECMO days without systemic anticoagulation, n (range) 4.75 (1–130)
Patients receiving blood products, n (%) 105 (52.2)
Open in a new tab

The total incidence of thrombosis (patient or circuit requiring exchange) was 22.9% (n=46). Venous or arterial thrombosis occurred in 19 patients (9.5%), the majority (n=14, 7%) of which were arterial events occurring almost exclusively in patients on VA-ECMO. Arterial events included intracardiac thrombus (n=8), leg ischemia (n=5), and ischemic stroke (n=1). Venous events comprised 5 patients (2.5% of all patient thrombosis), and included catheter-associated thrombosis (n=2), lower extremity deep vein thrombosis (n=3) and inferior vena cava thrombosis (n=1). There were 27 described cases (13.4%) in which one or more components of the ECMO circuit required exchange due to circuit thrombosis during anticoagulation-free ECMO. The majority of circuit thrombosis occurred within the oxygenator. All cases that required exchange of any component of ECMO circuit occurred despite the use of biocompatible tubing.

66 patients (32.8%) experienced bleeding of any severity during systemic anticoagulation-free ECMO, with 56 (27.9%) experiencing severe or major bleeding. Bleeding severity was not routinely reported according to ISTH standards, nor was bleeding location, with several studies reporting “severe” bleeding without clear definitions. Surgical site bleeding was the most common, occurring in 34 patients (51% of all bleeding patients), many of which prompted re-operation to achieve hemostasis. Only one patient developed new ICH during anticoagulant-free ECMO on day 4 of treatment, with eventual withdrawal of care and death due to worsening ICH several days later.17 Of patients who bled, 27.3% were receiving antiplatelet agents and/or prophylactic dose anticoagulants. Successful weaning from ECMO was reported in 85 patients (42.3%). 40 patient deaths (19.9%) were reported.

In the sub-population of patients who underwent ECMO without anticoagulation as part of a prospective protocol (n=72 among two studies), all utilizing a biocompatible circuit, the total incidence of thrombosis (patient or circuit requiring exchange) was 11% (n=8); of these, 1 patient (1.3%) suffered a basilar artery stroke and was subsequently discovered to have a hematologic malignancy, and 7 patients (9.7%) experienced circuit thrombosis. Bleeding occurred in 38.9% (n=28) of patients; of these, at least 25% (n=18) consisted of major bleeding, with the remaining patients experiencing “any” bleeding according to the study. Of note, much of the bleeding in this latter group of patients was felt to be expected post-operative bleeding from the sternotomy site. Mortality was reported at 34–54%. Full details of pooled thrombotic and bleeding outcomes from all included studies in this systematic review, as well as comparisons with previously published analyses, can be found in Table 2.

Table 2.

Pooled thrombotic and bleeding outcomes in ECMO patients with or without continuous anticoagulation in our systematic review, as compared with previous systematic reviews.

No continuous anticoagulation Continuous anticoagulation
Olson et al Sy et al, 201734 Vaquer et al, 20178 Zangrillo et al, 201333
 Total patients 201 50 1042 1763
 Thrombosis Pooled estimate No. (%) patients Point estimate, % Point estimate, % Point estimate, %
Total thrombosis (patient + circuit) 46 (22.9%) NA NA NA
Patient thrombosis, any type 19 (9.5%) 24% 4.6% 10%
  Venous 5 (2.5%)
 • Upper-extremity catheter-associated (2)
 • Lower extremity DVT (3); concurrent with catheter-associated clot in one patient
 • IVC thrombosis (1)		 0 4.6% 10%
  Arterial 14 (7.0%)
 • Intracardiac thrombus (8)
 • Ischemic basilar artery stroke (1)
 • Lower extremity ischemia (5)		 24% 0 10%
  Thrombosis incidence among VV ECMO cases 4 (3.6%) NA 4.6% NA
  Thrombosis incidence among VA ECMO cases 13 (14.8%) 24% NA NA
Patients requiring exchange of one or more ECMO circuit components 27 (13.4%)
 • Oxygenator thrombosis (40)
 • Circuit thrombosis (3)
 • Pump thrombosis (4)
 • Plasma leakage (3)
 • “Worsening oxygenation” (4)		 NA 12.8% 29%
Bleeding Olson et al
No. (%) of patients		 Sy et al, 201734
Point estimate, %		 Vaquer et al, 20178
Point estimate, %		 Zangrillo et al, 201333 Point estimate, %
Any bleeding, and confirmed sources 66 (32.8%)
 • Surgical site: 34
 • Cannula site: 1
 • ICH: 1
 • Pulmonary hemorrhage: 2		 NA 29.3% 33%
“Severe” or major bleeding 56 (27.9%) 43% 10.4% NA
New or worsening intracranial hemorrhage 1 (0.5%) NA 5.4 NA
Proportion of bleeding patients with confirmed receipt of antiplatelet and/or prophylactic dose anticoagulants during ECMO 18 (27.3%) NA NA NA
ECMO outcomes
Able to wean from ECMO 85 (42.3%) NA NA NA
Mortality 40 (19.9%) 56% 37.7% 54%
Open in a new tab

Discussion

To the best of our knowledge, this systematic review is the first to synthesize thrombotic and bleeding outcomes for patients on VV and VA-ECMO without continuous systemic anticoagulation. We found an overall incidence of thrombosis (patient or circuit) of 21.9%, and an overall incidence of bleeding of 32.8%. Patient thrombosis occurred in 8.5% of patients and was predominantly arterial in the setting of VA-ECMO. A minority of patients (13.4%) experienced thrombosis of the EMCO circuit mandating exchange of one or more circuit components, though these still occurred more frequently than patient thrombosis. The majority of bleeding events were classified as major or “severe,” though only one patient experienced new or worsening ICH during anticoagulant-free ECMO.

ECMO significantly perturbs the normal balance of hemostasis. Thrombosis is fueled by ECMO biomaterial-mediated activation of coagulation, complement and inflammatory cascades, as well as increased platelet activation and release of prothrombotic granules and microparticles.13 At the same time, the ECMO circuit can produce a potent bleeding diathesis via near-universal thrombocytopenia, loss of key platelet surface molecules GPVI, GPIb, CD63 and P-selectin, shear-mediated loss of high molecular weight von Willebrand multimers, and hypofibrinogenemia.13,18,19 The presence of competing risks for pathologic thrombosis and bleeding during ECMO might suggest a form of “rebalanced hemostasis”, though in reality this balance is exceedingly difficult to predict and has thus driven efforts to refine both ECMO technology and clinical practice.

Newer centrifugal pumps and modifications to the intraluminal surface of the ECMO circuit can reduce local coagulation, inflammation and platelet activation.20,21 In particular, heparin-bonded ECMO circuit is now widely utilized, supported by clinical data suggesting reduced blood transfusion requirements, redo sternotomy, length of mechanical ventilation and hospital length of stay compared to standard tubing.22,23 However, clinical data showing improvements in thrombotic outcomes with this technology have remained elusive.24 ECMO blood flow rate was recently shown to significantly affect platelet and leukocyte prothrombotic protein expression in vitro, though will require confirmation in clinical studies.25 Strategies such as antithrombin supplementation, heparin alternatives such as DTIs, antifibrinolytics, exogenous coagulation factors, and regional anticoagulation all have conflicting or limited evidence, or suffer from many of the same limitations as heparin.26,27 Inhibition of contact pathway factors XI and XII has shown promise in non-human primate models of extracorporeal blood flow as well as early phase human trials, though safety and efficacy will need to be confirmed in larger clinical studies.28–31 The use of continuously administered, systemic anticoagulation with heparin during ECMO has thus remained a key recommendation by ELSO.14

Driven by the accumulation of published ECMO cases run without any systemic anticoagulation, and the practicality of this potential intervention, we aimed to pool thrombotic and bleeding outcomes with this strategy and compare them to historical data from patients treated with standard-of-care, continuous therapeutic anticoagulation.

Thrombosis

Historically reported rates of thrombosis and bleeding in ECMO with the use of continuous systemic anticoagulation can be discerned from the annual ELSO registry reports and meta-analyses. The most recent ELSO report in 2017 listed a 22.1% and 15.6% incidence of ECMO circuit thrombosis in cases of VV- and VA-ECMO, respectively. Of these, 13.4% and 8.2% consisted of oxygenator thrombosis.32 An important caveat to the ELSO registry data is the lack individual patient data and anticoagulation practices, and therefore data from clinical trials may be more informative for comparison. In a recent meta-analysis by Vaquer et al. of 12 mostly non-randomized cohort studies of VV-ECMO for ARDS, the pooled rate of oxygenator failure requiring equipment exchange was 12.8%,8 while in a meta-analysis by Zangrillo et al of 12 retrospective studies of mostly VA-ECMO, oxygenator dysfunction requiring replacement occurred in 29%.33 The 13.4% incidence of ECMO circuit thrombosis requiring equipment exchange revealed by our review is within the range reported by these previous studies that used systemic anticoagulation.

In the two meta-analyses discussed previously, VTE occurred in 4.6% of VV- and 10% of VA-ECMO patients, respectively. Leg ischemia as a form of arterial thrombosis occurred in 10% in the meta-analysis of exclusively VA-ECMO studies, and was not reported in the meta-analysis of VV-ECMO studies.8,33 While the 2017 ELSO report does not list rates of VTE, arterial thrombosis was reported in 3% and 7.4% of patients treated with VV and VA-ECMO, respectively.32 Finally, pooled rates of thromboembolism specifically in VA-ECMO treated without continuous systemic anticoagulation have been synthesized by Sy et al in a systematic review and proportional meta-analysis (including three of the studies in our systematic review), with arterial thrombosis occurring in 24%.34 In our review of all cases of ECMO (VV and VA) run without continuous systemic anticoagulation, the total incidence of in vivo thrombosis was 9.5%, the majority of which (7%) were arterial events, predominantly intracardiac thrombus and lower extremity ischemia in those on VA-ECMO with femoral artery cannulation. This latter finding is perhaps illustrative of the inherent risk in VA-ECMO of cannulating the femoral artery; evidence supports the use of distal perfusion catheters to ensure adequate blood supply to the cannulated limb,35 though the frequency of this practice was not uniformly reported by studies included in this analysis.

Bleeding

The cumulative incidence of total bleeding described in the 2017 ELSO registry report was 39.4% and 51% for VV- and VA-ECMO, respectively.32 In the meta-analysis by Vacquer et al, the pooled incidence of bleeding for VV-ECMO was 29.3%, of which 10.4% was classified as “significant.”8 In the meta-analysis by Zangrillo et al of predominantly VA-ECMO studies, the cumulative incidence of any bleeding was 33%.33 Our review found an overall bleeding incidence of 32.8%, of which 27.9% was significant. Important to note is the non-negligible proportion of patients who bled during ECMO who were also receiving antiplatelet and/or prophylactic-dose anticoagulants during the procedure (27.3%), as well as the frequent occurrence of post-operative surgical site bleeding. Given the lack of consistent reporting of bleeding outcomes and the fact that post-surgical bleeding after sternotomy is not uncommon, it is difficult to make meaningful comparisons in bleeding risk.

In summary, the rate of circuit thrombosis summarized by our systematic review is comparable with historically reported rates while using continuous systemic anticoagulation, and lower than rates reported by previous analysis of anticoagulant-free VA-ECMO by Sy et al.8,32,34 The reason for these similar rates remains largely speculative at this point, though an optimistic appraisal could attribute these favorable outcomes to more widespread use of biocompatible ECMO circuit and pump technology, and improved overall care of critically ill patients. The incidence of patient specific thrombosis (9.5%) revealed by this review was also within the range of previously reported studies, and is encouraging in its support of the notion that systemic anticoagulation is primarily meant to mitigate circuit thrombosis.

This systematic review has several limitations worth highlighting. First, the majority of data was retrospective in nature, consisting primarily of case reports and series. Consideration must be given to both selection and publication bias, as patient cases with more favorable outcomes may have been more likely to be published. While a large number of cases were drawn from prospective, single-arm observational studies, these, too, suffer from a lack of comparator arm and strict enrollment criteria to reduce risk of bias. Details of ECMO flow rate, use of biocompatible circuit, administration of other anticoagulants/antiplatelet agents, and administration of blood products were not routinely reported. In addition, details of thrombotic and bleeding events were inconsistently reported or defined. Thrombotic events within the ECMO circuit were not always accompanied by details of whether circuit exchange was necessary or patient oxygenation was impaired; these clinically meaningful outcomes deserve special emphasis when reporting patient outcomes, since minor oxygenator thrombosis may be common and remain clinically silent. Similarly, bleeding events were inconsistently stratified by severity, with many of these events potentially representing somewhat expected post-operative bleeding rather than caused by the pathophysiologic mechanisms previously described. Many patient cases also included major surgeries which could confound the detected rates of thrombosis and bleeding, as surgery itself is associated with substantial risks of thrombosis and bleeding.

Conclusion

While limited by the nature and quality of included studies, the results of our systematic review are compelling by suggesting comparable rates of circuit and patient thrombosis regardless of the administration of continuous systemic anticoagulation. Inconsistency in the nature and reporting of bleeding outcomes makes drawing conclusions on this outcome difficult. The ELSO has included recommendations for consistent reporting of thrombotic and bleeding outcomes in their anticoagulation guidelines; adhering to these reporting guidelines may help to more accurately synthesize clinical data in the future.14 In the absence of this kind of rigorously collected, prospective clinical data, we do not advocate for routine omission of systemic anticoagulation with ECMO, though we do feel our conclusions are hypothesis-generating; given the many potential causes of bleeding during ECMO, omission of systemic anticoagulation represents a potentially simple and attractive evolution of ECMO practice that could improve patient outcomes. This consideration is particularly relevant in patients with active bleeding or at high risk of bleeding such as patient with ICH, DAH, or those suffering from trauma.

Supplementary Material

1

Acknowledgements:

Funding:

This work is supported by grants from the National Institutes of Health (R01HL101972, R01HL144113).

Abbreviations:

ARDS

Acute respiratory distress syndrome

DAH

Diffuse alveolar hemorrhage

ECMO

Extracorporeal membrane oxygenation

ELSO

Extracorporeal life support organization

ICH

Intracranial hemorrhage

ISTH

International Society of Thrombosis & Hemostasis

RCT

Randomized controlled trial

UFH

Unfractionated heparin

VA

Venoarterial

VV

Venovenous

Footnotes

Guarantor statement: SRO takes responsibility for the content of this manuscript, including the data and analysis.

Conflicts of interest: Dr. Shatzel reports receiving consulting fees from Aronora Inc. The remaining authors have no conflicts to disclose

References

  • 1.Bellani G, Laffey JG, Pham T, Fan E, Brochard L, Esteban A, et al. Epidemiology, Patterns of Care, and Mortality for Patients With Acute Respiratory Distress Syndrome in Intensive Care Units in 50 Countries. Jama. 2016;315(8):788–800. [DOI] [PubMed] [Google Scholar]
  • 2.Organization ELS. ECLS Registry Report International Summary 2019. 2019.
  • 3.Gaffey AC, Chen CW, Chung JJ, Goldberg LR, Bermudez CA, Acker MA, et al. Extracorporeal Membrane Oxygenation (ECMO) as a Bridge to Heart Transplantation: Impact on Post Transplantation Outcomes. The Journal of Heart and Lung Transplantation. 2017;36(4):S152–S153. [Google Scholar]
  • 4.Krecmerova M, Mosna F, Bicek V, Petrik F, Grandcourtova A, Lekes M, et al. Extracorporeal membrane oxygenation to support repeated whole-lung lavage in a patient with pulmonary alveolar proteinosis in life threatening dyspnoe--a case report. BMC Anesthesiol. 2015;15:173. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Barac YD, Bryner B, Bonadonna D, Wolfe C, Reynolds J, Haney JC, et al. Bilateral Pneumonectomy with Veno-Arterial Extracoporeal Membrane Oxygention as a Bridge to Lung Transplant. The Journal of Heart and Lung Transplantation. [DOI] [PubMed]
  • 6.Munshi L, Walkey A, Goligher E, Pham T, Uleryk EM, Fan E. Venovenous extracorporeal membrane oxygenation for acute respiratory distress syndrome: a systematic review and meta-analysis. The Lancet Respiratory medicine. 2019;7(2):163–172. [DOI] [PubMed] [Google Scholar]
  • 7.Ouweneel DM, Schotborgh JV, Limpens J, Sjauw KD, Engstrom AE, Lagrand WK, et al. Extracorporeal life support during cardiac arrest and cardiogenic shock: a systematic review and meta-analysis. Intensive care medicine. 2016;42(12):1922–1934. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Vaquer S, de Haro C, Peruga P, Oliva JC, Artigas A. Systematic review and meta-analysis of complications and mortality of veno-venous extracorporeal membrane oxygenation for refractory acute respiratory distress syndrome. Annals of intensive care. 2017;7(1):51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Noah MA, Peek GJ, Finney SJ, Griffiths MJ, Harrison DA, Grieve R, et al. Referral to an extracorporeal membrane oxygenation center and mortality among patients with severe 2009 influenza A(H1N1). Jama. 2011;306(15):1659–1668. [DOI] [PubMed] [Google Scholar]
  • 10.Pham T, Combes A, Roze H, Chevret S, Mercat A, Roch A, et al. Extracorporeal membrane oxygenation for pandemic influenza A(H1N1)-induced acute respiratory distress syndrome: a cohort study and propensity-matched analysis. American journal of respiratory and critical care medicine. 2013;187(3):276–285. [DOI] [PubMed] [Google Scholar]
  • 11.Gould MK, Garcia DA, Wren SM, Karanicolas PJ, Arcelus JI, Heit JA, et al. Prevention of VTE in Nonorthopedic Surgical Patients: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. CHEST. 2012;141(2):e227S–e277S. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Schunemann HJ, Cushman M, Burnett AE, Kahn SR, Beyer-Westendorf J, Spencer FA, et al. American Society of Hematology 2018 guidelines for management of venous thromboembolism: prophylaxis for hospitalized and nonhospitalized medical patients. Blood Adv. 2018;2(22):3198–3225. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Doyle AJ, Hunt BJ. Current Understanding of How Extracorporeal Membrane Oxygenators Activate Haemostasis and Other Blood Components. Front Med (Lausanne). 2018;5:352. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Organization ELS. ELSO Anticoagulation Guideline. 2014.
  • 15.Hirsh J, Anand SS, Halperin JL, Fuster V. Guide to anticoagulant therapy: Heparin : a statement for healthcare professionals from the American Heart Association. Circulation. 2001;103(24):2994–3018. [DOI] [PubMed] [Google Scholar]
  • 16.Kaatz S, Ahmad D, Spyropoulos AC, Schulman S. Definition of clinically relevant non-major bleeding in studies of anticoagulants in atrial fibrillation and venous thromboembolic disease in non-surgical patients: communication from the SSC of the ISTH. Journal of thrombosis and haemostasis : JTH. 2015;13(11):2119–2126. [DOI] [PubMed] [Google Scholar]
  • 17.Factora FN, Bustamante S, Spiotta A, Avitsian R. Intracranial hemorrhage surgery on patients on mechanical circulatory support: a case series. J Neurosurg Anesthesiol. 2011;23(1):30–34. [DOI] [PubMed] [Google Scholar]
  • 18.Kalbhenn J, Wittau N, Schmutz A, Zieger B, Schmidt R. 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. Perfusion. 2015;30(8):675–682. [DOI] [PubMed] [Google Scholar]
  • 19.Kalbhenn J, Schlagenhauf A, Rosenfelder S, Schmutz A, Zieger B. Acquired von Willebrand syndrome and impaired platelet function during venovenous extracorporeal membrane oxygenation: Rapid onset and fast recovery. The Journal of heart and lung transplantation : the official publication of the International Society for Heart Transplantation. 2018;37(8):985–991. [DOI] [PubMed] [Google Scholar]
  • 20.Thomas J, Kostousov V, Teruya J. Bleeding and Thrombotic Complications in the Use of Extracorporeal Membrane Oxygenation. Semin Thromb Hemost. 2018;44(1):20–29. [DOI] [PubMed] [Google Scholar]
  • 21.Ontaneda A, Annich GM. Novel Surfaces in Extracorporeal Membrane Oxygenation Circuits. Front Med (Lausanne). 2018;5:321. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Mangoush O, Purkayastha S, Haj-Yahia S, Kinross J, Hayward M, Bartolozzi F, et al. Heparin-bonded circuits versus nonheparin-bonded circuits: an evaluation of their effect on clinical outcomes. Eur J Cardiothorac Surg. 2007;31(6):1058–1069. [DOI] [PubMed] [Google Scholar]
  • 23.Ranucci M, Balduini A, Ditta A, Boncilli A, Brozzi S. A Systematic Review of Biocompatible Cardiopulmonary Bypass Circuits and Clinical Outcome. The Annals of Thoracic Surgery. 2009;87(4):1311–1319. [DOI] [PubMed] [Google Scholar]
  • 24.Sohn N, Marcoux J, Mycyk T, Krahn J, Meng Q. The impact of different biocompatible coated cardiopulmonary bypass circuits on inflammatory response and oxidative stress. Perfusion. 2009;24(4):231–237. [DOI] [PubMed] [Google Scholar]
  • 25.Meyer AD, Rishmawi AR, Kamucheka R, Lafleur C, Batchinsky AI, Mackman N, et al. Effect of Blood Flow on Platelets, Leukocytes, and Extracellular Vesicles in Thrombosis of Simulated Neonatal Extracorporeal Circulation. Journal of thrombosis and haemostasis : JTH. 2019. [DOI] [PMC free article] [PubMed]
  • 26.Annich GM, Zaulan O, Neufeld M, Wagner D, Reynolds MM. Thromboprophylaxis in Extracorporeal Circuits: Current Pharmacological Strategies and Future Directions. Am J Cardiovasc Drugs. 2017;17(6):425–439. [DOI] [PubMed] [Google Scholar]
  • 27.Wong JJ-M, Lam JCM, Mok YH, Lee JH. Anticoagulation in extracorporeal membrane oxygenation. Journal of Emergency and Critical Care Medicine. 2018;2(2). [Google Scholar]
  • 28.Matafonov A, Leung PY, Gailani AE, Grach SL, Puy C, Cheng Q, et al. Factor XII inhibition reduces thrombus formation in a primate thrombosis model. Blood. 2014;123(11):1739–1746. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Larsson M, Rayzman V, Nolte MW, Nickel KF, Bjorkqvist J, Jamsa A, et al. A factor XIIa inhibitory antibody provides thromboprotection in extracorporeal circulation without increasing bleeding risk. Sci Transl Med. 2014;6(222):222ra217. [DOI] [PubMed] [Google Scholar]
  • 30.Cheng Q, Tucker EI, Pine MS, Sisler I, Matafonov A, Sun MF, et al. A role for factor XIIa-mediated factor XI activation in thrombus formation in vivo. Blood. 2010;116(19):3981–3989. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.DeLoughery EP, Olson SR, Puy C, McCarty OJT, Shatzel JJ. The Safety and Efficacy of Novel Agents Targeting Factors XI and XII in Early Phase Human Trials. Semin Thromb Hemost. 2019;45(5):502–508. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Organization ELS. ECLS Registry Report International Summary 2017. 2017.
  • 33.Zangrillo A, Landoni G, Biondi-Zoccai G, Greco M, Greco T, Frati G, et al. A meta-analysis of complications and mortality of extracorporeal membrane oxygenation. Crit Care Resusc. 2013;15(3):172–178. [PubMed] [Google Scholar]
  • 34.Sy E, Sklar MC, Lequier L, Fan E, Kanji HD. Anticoagulation practices and the prevalence of major bleeding, thromboembolic events, and mortality in venoarterial extracorporeal membrane oxygenation: A systematic review and meta-analysis. J Crit Care. 2017;39:87–96. [DOI] [PubMed] [Google Scholar]
  • 35.Makdisi G, Makdisi T, Wang IW. Use of distal perfusion in peripheral extracorporeal membrane oxygenation. Ann Transl Med. 2017;5(5):103. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1749560-supplement-1.pdf (365KB, pdf)

RESOURCES