EuroELSO Journal Club

Interested in the latest scientific publications?

EuroELSO is dedicated to knowledge and education. Therefore, EuroELSO Steering Committee has launched the EuroELSO Journal Club. We are selecting on a quarterly basis recent interesting ECLS related manuscripts and provided summaries in a concise format. We hope this activity might turn to a useful tool for all physicians and specialists interested in ECLS/ECMO.

On behalf of EuroELSO Steering Committee
Jan Belohlavek, Roberto Lorusso, Justyna Swol, Matteo di Nardo and Simon Finney

02 / 2019

General survival of patients managed by veno-arterial ECMO (VA-ECMO) is as low as 30%. Due to the very high resource utilization, there is a growing need for a reliable model for identifying patients with favorable or alternatively poor prognosis. Several scoring systems have been introduced in the past, including the ENCOURAGE or the SAVE score, but are mostly relying on baseline clinical data at the time of cannulation and were not validated for extracorporeal cardiopulmonary resuscitation patients. A more flexible and dynamic prognostic model is needed. An international group of investigators, therefore, suggested and validated the point-of-care biomarker-based PREDICT VA-ECMO score, that can be used during the post-implantation period and can aid in the difficult decision making to either continue or withdraw care. PREDICT VA-ECMO is based on lactate, pH and bicarbonate concentrations only.
A derivation cohort included a total of 205 subjects, 51% received VA-ECMO during resuscitation and 43% had a severe shock. Two prediction models based on point-of-care biomarkers were developed using penalized logistic regression in an elastic net approach. SAVE, SAPS, SOFA and APACHE scores were recorded for comparison for the prediction of hospital survival. PREDICT VA-ECMO was superior to all comparators regarding the area under the receiver operation characteristic curve (AUC) in both 6 and 12 hours of survival. Performance of the scores in the external validation cohort was also encouraging.

The PREDICT VA-ECMO score was designed to reflect the current effectiveness of organ perfusion by VA-ECMO regardless of initial indication or baseline parameters. It’s easy-to-calculate nature is suitable for acute emergency situations. The authors commented, that interestingly, adding age and gender to the model did not improve further prognostication, suggesting that, during this most critical, first phase of VA-ECMO therapy, perfusion parameters may be more important than underlying clinical characteristics. The limitation of the score is a lack of its information for neuroprognostication, which according to the guidelines, should be provided after cardiac arrest until 72 hours after the collapse and cannot be replaced by any clinical score designed for the other purposes.


Dual-lumen veno-venous ECMO cannulas (DLC) demonstrated an advantage of easier ambulation for patients, lower sedation levels and earlier extubation in comparison with a conventional two site cannulation (usually 19-23 French cannulas in adults). The risks, on the other site, include bleeding considered as, intracranial hemorrhage (ICH), migration to the right ventricle and right ventricular rupture and are mostly related to the large cannulae size. ICH during ECMO occurs in around 7 % of patients with DLC according to the previous literature and is caused by thrombocytopenia, anticoagulation and possibly intracranial venous hypertension from obstruction of the internal jugular vein and superior vena cava by the ECMO cannula or a deep venous thrombus. Although 31 French dual-lumen cannula enables greater blood flow with better oxygenation, according to a recent retrospective cohort study from the ELSO registry it also poses a greater risk of ICH when compared with a 27 French equivalent. The outcome of an ICH in ECMO patients (with DLC) is critical – one-month mortality after the event reaches 81 %.

This study used a retrospective propensity score matching within the ELSO database cohort (n=744). Patients with a 31 French DLC had an ICH rate of 4.3% compared with 1.6% in patients cannulated with a 27 French DLC (p = 0.03). There was no difference in mortality, hemolysis or cannula complications and, thus, smaller cannulas may be preferable when feasible. Maximal blood flow achievable by 27 French DLC is 4,5L/min, by 31 French up to 6L/min. Authors of the study, therefore, hypothesized that a 31 French DLC offers a theoretical advantage in patients with severely impaired oxygenation. However, there was no statistical difference between the groups regarding ECMO blood flow at 4 and 24 hours suggesting that optimal flow can be achieved with either cannula in most patients.

The limitations of the study included observational design, pooled data from multiple centers and lack of data regarding the quality of life and functional outcome after decannulation. Although propensity score matching was performed, residual confounding cannot be ruled out. The incidence of ICH in the cohort study was rather low in comparison with overall literature and ICH might have been underdiagnosed within the registry patients.


ECMO is used in patients with the most severe conditions and is, therefore, accompanied by difficult decision making and profound ethical considerations. In a recent electronic, cross-sectional, scenario-based survey, 539 senior physicians from 39 countries across six continents were questioned about their daily clinical practice and attitudes toward the initiation, limitation, and withdrawal of venovenous ECMO (VV-ECMO) for severe respiratory failure. The survey stressed the situation of ECMO-dependent patients, who are unlikely to be transitioned from ECMO to conventional mechanical ventilation and are not eligible for transplant. In these scenarios, ECMO support has been referred to as a “bridge to nowhere”.

There was no significant difference in respondent’s age, years of experience, or center characteristics between those who would or would not initiate ECMO. As a factor contributing to the decision not to cannulate for ECMO, most physicians mentioned older patient age (46.9%), additional organ failures (37.7%), and prolonged mechanical ventilation (35.1%). Factors influencing the decision to withdraw from ECMO were patient comorbidities (70.5%), patient’s wishes (56.0%), and etiology of respiratory failure (37.7%). Although the decision process was shared with another physician in a vast majority of cases (90.5%), only 53.2%, 45.3%, and 29.5% of respondents involved surrogates, awake patients, or bedside nurses. 15% of respondents also rarely or never discussed the possibility of ECMO withdrawal with the patient or surrogate at the time of ECMO initiation.

Authors commented that this low involvement of patients and their loved-ones into shared decision making may pose a significant ethical hazard and a need for a better education and communication was identified. Furthermore, the existence of a hospital mechanism to enable making medical decisions against patients or surrogate’s wishes was reported by over half of the physicians, however, details about its invocation in a real practice remain unclear. In the present survey, some respondents were more likely to not to escalate rather than withdraw treatment in this clinical setting.


Acute respiratory distress syndrome (ARDS) is a common reason for ICU hospitalization with acute mortality of 35-46%. The cornerstone of ARDS management is lung by higher levels of positive end-expiratory pressure (PEEP). Lung recruitment maneuvers (LRM) during mechanical ventilation are still controversial. The effect of PEEP and LRM is variable and dependent on the potential for lung recruitment (PLR), which might guide the ventilatory strategy. The situation is more complicated in patients with ARDS managed by ECMO. The data in this clinical setting are limited. An aim of a single-center, retrospective, observational cohort study of the British authors was to investigate the pattern of lung recruitability of patients with severe ARDS requiring ECMO.

47 patients with severe ARDS mainly of pulmonary origin (86%) on ECMO were included in the study. PLR and the weight of nonaerated, poorly aerated, normally aerated, and hyperaerated lung tissue was assessed at low (5 cmH2O) and high (45 cmH2O) airway pressures. Patients were classified into the high or low PLR group based on the median potential for lung recruitment value of the study population. The percentage of potentially recruitable lung was defined as the proportion of the total lung weight accounted for by nonaerated lung tissue in which aeration was restored, according to CT, by an airway pressure of 45 cmH2O from an airway pressure of 5 cmH2O. To evaluate difference in pre-ECMO clinical and demographical variables and differences in outcomes, patients were subdivided into two PLR categories with a group with PLR below the median (PLRlow) and a group with PLR above the median (PLRhigh). The median PLR was relatively high (24.3%, interquartile range = 11.4–37%) with a large variation ranging from –2% to 76.3% of the total lung weight. Subjects with above the median PLR demonstrated shorter ECMO duration (8 vs 13 d; p = 0.013) and shorter ICU stay (15 vs 22 d; p = 0.028), mortality was not statistically different (24% vs 46%; p = 0.159).

Although PLR increases with total lung weight, that is, with ARDS severity, a higher PLR in this study was associated with a better prognosis even when recruitment was not attempted and mechanical ventilation was not adapted to the degree of PLR due to absence of clear data guiding mechanical ventilation during ECMO and observational nature of the study. All patients were ventilated with standard settings regardless of PLR and the authors commented that it is difficult to say whether patients with higher PLR would have benefitted from a more aggressive lung recruitment or higher PEEP levels.


ECMO is turning from an experimental to a well-established method in critical care and the first long-term outcome data are becoming available. Cognitive function of 38 patients treated with ECMO for respiratory failure between the years 1995 and 2009 was investigated in a recent Swedish single-center retrospective cohort study, assessing a combination of brain imaging, neurocognitive testing, and interview. Originally, 73 eligible patients were identified, but 23 died after discharge and 19 were lost to follow-up. The likely limitation of the study is, therefore, selection bias.

Besides the usual ICU impact on neurological function, the use of ECMO adds an extra burden of altered coagulation, hemodynamics, and risk of prolonged hospitalization and immobilization. Many patients with respiratory failure on ECMO are subjected to permissive hypoxemia (defined as a median peripheral saturation of less than 93% during the observation period) in order to protect them from ventilator-induced lung injury as per ELSO guidelines. The safety of permissive hypoxemia has been questioned due to the risk of cognitive dysfunction. The authors of the present study, however, hypothesized that it is unlikely for irreversible hypoxic cerebral dysfunction to occur when hypoxemia is present with preserved sufficient oxygen delivery to the brain. Indeed, the original treatment protocol was designed in order to compensate for a reduction in oxygen saturation by increasing the ECMO flow.

Median follow-up time was 9.0 years after treatment. Median full-scale intelligence quotient, memory index, and executive index were 97, 101, and 104, respectively (normal, 100±15) and cognitive function was not reduced in the group which experienced prolonged hypoxemia. There was no overall significant difference between patients who received veno-venous (VV-ECMO) and veno-arterial (VA-ECMO) treatment, however, in VA-ECMO group, brain imaging showed cerebrovascular lesions more often (64% in VA-ECMO, 26% in VV-ECMO). In patients with cerebrovascular lesions, memory function and executive function were significantly reduced. Permissive hypoxemia was not correlated with long-term cognitive dysfunction. Return to work or school occurred in half of the cohort within one year after discharge. The authors concluded that patients treated with ECMO for respiratory failure may have normal cognitive function years after treatment, if not affected by cerebrovascular lesions.


Early revascularization remains standard treatment of acute myocardial infarction complicated by cardiogenic shock (AMI-CS) . Medical therapies such as inotropes failed to improve outcomes. The comparison of the intra-aortic balloon pump (IABP), and best medical care, was published in IABP-SHOCK II clinical trial in 2012 and showed no survival benefit. The need for an improved device design is, therefore, prudent and great expectations were put on Impella device, a catheter-based micro-axial flow pump. The group of European investigators built a multi-national database of AMI-CS cases treated with an Impella device in order to compare the outcome to patients from the IABP-SHOCK II trial in a matched fashion. 237 patients with cardiogenic shock (definition in accordance with IABP-SHOCK II trial, systolic blood pressure below 90 mm Hg or need of inotropes, and clinical signs of pulmonary congestion, and signs of impaired end-organ perfusion12) complicating acute myocardial infarction (either ST-segment–elevation myocardial infarction or non–ST-segment–elevation myocardial infarction) who were treated with an Impella device and who underwent early revascularization were matched to 237 subjects from the IABP-SHOCK II cohort. All patients underwent early revascularization and received optimal medical treatment per current guidelines.
There was no significant difference regarding the primary outcome of 30-day all-cause mortality when Impella was compared with IABP or best medical treatment (48.5% vs. 46.4%, p=0.64), but the incidence of adverse events was higher in the Impella group (severe or life-threatening bleeding 8.5% vs. 3.0%, p<0.01, peripheral vascular complications 9.8% vs. 3.8%, p=0.01.) The increase in peripheral vascular and bleeding complications might be related to the larger lumen of the vascular access needed for the Impella devices – the Impella 2.5 and CP are placed via 12 and 14 French sheaths, respectively, as compared to the 7-8 French sheath or sheath-less insertion for the IABP device. There were no increased risks for stroke, re-infarction and stent thrombosis in the Impella group. The analysis is limited to IABP-treated patients as a control group, but itdid not change the results.

In this retrospective analysis of patients with AMI-CS, the use of an Impella device was not associated with lower 30-day mortality compared with matched patients from the IABP-SHOCK II trial treated with an IABP or medical therapy. The limitation of the present study is its population in profound CS and inclusion of a large proportion of post-resuscitation cases with a poor prognosis. Also, according to recent evidence, early Impella implantation seems superior to later intervention, however, this effect was not observed in the present analysis. Prospective, randomized clinical trial, preferably with an early Impella implantation, is, therefore, warranted.


Heparin-induced thrombocytopenia (HIT) is caused by antibodies binding to the complex of heparin and platelet factor and typically occurs four days after heparin administration. Without immediate discontinuation, almost half of the patients develop thrombosis and the mortality rate increases to 30%. According to previous literature, the prevalence of HIT while on ECMO is approximately 8%. Change to an alternative of systemic heparin is paramount, data on a continuation of using heparin-bonded ECMO circuit after HIT diagnosis are limited. In a recent retrospective single-center analysis published in Perfusion, further use of a heparin-bonded ECMO system did not lead to adverse clinical outcomes.

Out of 455 patients who underwent ECMO during a nine-year period, 3,1% tested positive on HIT. Highly specific serotonin release assay was used because the cause for thrombocytopenia in ECMO patients could be the ECMO system itself. Systemic heparin was discontinued and heparin-bonded ECMO circuit maintained in all 14 subjects. 78,6% survived to discharge which did not significantly differ from the rest of the study population. Mean platelet count increased from 59.8k/µL at the time of HIT diagnosis to 280.2k/µL 14 days after heparin discontinuation, when it did not differ from the rest of the cohort. There was no decrease in platelet count after systemic heparin discontinuation. Platelet count further increased after decannulation from ECMO device and it is difficult to evaluate if that happened because of the general negative ECMO effect on platelet count or discontinuation of exposure to the heparin-bonded circuit. Out of 14 patients with HIT, 3 died during hospitalization (two due to multiorgan failure, one due to intracranial hemorrhage).

Authors pointed out that so far, there is no existing data proving if heparin, chemically bonded to the ECMO system, affects the generation of HIT antibodies. The bonded heparin is biologically inactive and has almost no anti-factor Xa activity. However, small parts of the heparin can leak out and eventually become biologically active. The present study did not show any signals of a detrimential effect of keeping HIT patients on heparin-bonded circuit though. The appropriate management of HIT while on ECMO is, therefore, immediate discontinuation of systemic heparin, but the continuation of heparin-coated extracorporeal membrane oxygenation circuits appears to be safe. Larger prospective trials will be needed to confirm that.


Systemic anticoagulation mostly with unfractionated heparin is necessary to prevent thrombosis during ECMO. Titration and monitoring of an anticoagulation therapy are challenging due to the combined impact of the underlying disease (e.g. sepsis), the drug and circuit effects. Thromboelastography (TEG) is widely used in the setting of major hemorrhage management. Adaptation of those methods for the monitoring of systemic anticoagulation therapy during ECMO was evaluated in the multicenter pilot trial in comparison with an activated partial thromboplastin time (aPTT). A total of 42 patients with acute respiratory failure and veno-venous ECMO were randomized either to TEG (target 16–24 min of the R parameter) or aPTT-based protocol (target 1.5–2 of aPTT ratio.)

Heparin dosing was lower in the TEG group compared to the aPTT group (11.7 (9.5–15.3) IU/kg/h vs. 15.7 (10.9–21.3) IU/kg/h, respectively, p = 0.03) and the TEG group experienced a lower incidence of bleeding from surgical sites. Otherwise, the two groups did not significantly differ neither in a number of hemorrhagic or thrombotic events and transfusions given nor in the duration of ECMO. The TEG-based protocol triggered heparin infusion rate adjustments more frequently (p < 0.01) and results were less frequently in the target range compared to the aPTT-based protocol (p < 0.001). One of the major concerns of using a low level of anticoagulation during ECMO is the development of consumptive coagulopathy. Since the D-dimer levels were comparable in the two groups, this was not of a concern in the present study.

aPTT is a standard hemostatic profile used to monitor heparin treatment. It is performed on plasma samples in an absence of the cellular components and may not, therefore, provide a realistic picture of the in vivo hemostatic capacity. TEG is a whole blood point-of-care assay, that reflects the state of the entire coagulation cascade. The TEG-based protocol was safe and feasible and resulted in lower consumption of heparin with a statistically insignificant trend towards the reduction of bleeding events. However, the use of TEG-based protocol also resulted in a higher cost compared to the aPTT-based one.


01 / 2019

Refractory cardiogenic shock can occur after cardiac surgery, as an exacerbation of chronic heart failure with reduced ejection fraction, or from de novo heart failure, most commonly caused by acute coronary syndromes. V-A ECMO might be a game-changer in support of patients in such conditions. Despite being able to offer blood flow of more than 5 L/min of oxygenated blood, VA-ECMO can cause some adverse effects as increased afterload, LV dilation, myocardial ischemia, elevated pulmonary pressures and blood stasis with consecutive thrombus formation. A team of researchers from Ohio, USA, reviewed concomitant use of percutaneous left ventricular decompression with Impella Pump. The largest US-based retrospective study included data collected between 2014 and 2016. Two cohorts of patients were compared – V-A ECMO supported cohort  (n = 36) and ECPELLA cohort supported with a combination of V-A ECMO and Impella (n = 30). The primary outcome was all-cause mortality within 30 days of V-A ECMO implantation. Patients in the ECPELLA cohort experienced a higher incidence of ST-elevation myocardial infarction (STEMI) and percutaneous coronary intervention (PCI), otherwise, the baseline characteristics were similar.

Thirty-day all-cause mortality was significantly lower in the ECPELLA cohort (57% vs. 78%; hazard ratio [HR] 0.51 [0.28–0.94], log rank p = 0.02), and this difference remained intact after correcting for STEMI and PCI. The secondary outcomes, including duration of support, stroke, major bleeding, hemolysis, and cardiac recovery, were comparable between the cohorts. The only difference was found in an inotropic score, which was greater in the V-A ECMO group by day 2 (11 vs. 0; p = 0.001). The addition of Impella to V-A ECMO for patients with refractory cardiogenic shock was associated with lower all-cause 30-day mortality, lower inotrope use, and comparable safety profiles as compared with V-A ECMO alone. 

Bottom Line: The study, therefore, suggests that the addition of Impella to V-A ECMO is associated with improved survival in this clinical setting. Randomized controlled trials are required to confirm these findings.

Patel, Sandeep M., et al. “Simultaneous Venoarterial Extracorporeal Membrane Oxygenation and Percutaneous Left Ventricular Decompression Therapy with Impella Is Associated with Improved Outcomes in Refractory Cardiogenic Shock.” ASAIO Journal, vol. 65, no. 1, 2019, pp. 21–28., doi:10.1097/mat.0000000000000767.

V-A ECMO is still associated with a relatively high incidence of vascular complications and especially femoral cannulation increases the risk of lower-limb hypoperfusion and ischemia. Those can lead to severe irreversible conditions, including a limb loss. Although clinical evaluation of the perfusion is recommended, technological advances offer comprehensive and quantitative monitoring. A team from the Columbia University Medical Center in New York, USA, evaluated use of non-invasive lower-limb oximetry, using near-infrared reflectance spectroscopy (NIRS) in this clinical setting.

Data were collected from 25 patients (age 22-78) receiving femoral V-A ECMO between June 2016 and January 2017. Subjects were continuously monitored using the CASMED Fore-Sight Elite tissue oximeter. A retrospective pilot study was conducted to review the correlation between NIRS tissue saturations (StO2) and clinical indications of limb ischemia. Evaluated events included StO2s less than 50% for more than four minutes or StO2 differentials between the cannulated and non-cannulated legs greater than 15%.

Clinical signs of lower-limb ischemia, e.g. cold limb, mottled skin and absent Doppler signal, were observed in six patients from the group. All of those events were matched with StO2s below 50% that persisted for longer than four minutes. One patient had a false-positive device indication of hypoperfusion with StO2 below 50% for more than four minutes due to a venous saturation below 30% without the localized clinical counterpart and another had a false-positive absent Doppler signal caused by high doses of vasopressors. Five patients had StO2s below 50% for less than four minutes and none of these patients had clinical indications of lower-limb hypoperfusion, suggesting the importance of a time frame. Cannula-related obstruction of flow to the distal portion of the leg was associated with StO2 differentials greater than 15% in all affected patients and conversely no patients without cannula-related obstruction to flow had StO2 differentials greater than 15%. Sensitivity and specificity were 100% for diagnosing cannula-related obstruction to flow when the StO2 in the cannulated leg was below 50% for longer than four minutes and a differential between the cannulated and non-cannulated legs was greater than 15%. This has made NIRS monitoring especially valuable during the initiation of ECMO with femoral cannulation and for the period immediately following initiation when these patients are at the greatest risk for developing lower-limb ischemia.

Bottom line: Continuous NIRS monitoring might be used in the future as a warning system for identifying lower-limb ischemia and can also increase the confidence of bedside support staff to care for ECMO patients.

Patton-Rivera, Killian, et al. “Using near-Infrared Reflectance Spectroscopy (NIRS) to Assess Distal-Limb Perfusion on Venoarterial (V-A) Extracorporeal Membrane Oxygenation (ECMO) Patients with Femoral Cannulation.” Perfusion, vol. 33, no. 8, 2018, pp. 618–623., doi:10.1177/0267659118777670.

ECMO has been used to stabilize patients with acute PE in order to prolong valuable time for possible interventions, such as systemic thrombolysis, catheter-directed therapy or surgical embolectomy. Because these patients are often too unstable to transfer to the operating room or catheterization laboratory, a bridge to definitive therapy is needed (bridge-to-advanced therapy). The recommendation of the European Society of Cardiology for ECMO in the setting of a massive PE is based on case reports and case series and, therefore, is not assigned a class or level of evidence. Moreover, the latest American Heart Association scientific statement on the management of PE did not include the use of ECMO, citing a lack of data. At the current time, decisions need to be made on a case-by-case basis. However, one of the new additions to the slowly growing evidence in this clinical setting is a case series and review of the literature, published in Perfusion journal in December last year. A consecutive cohort of patients with confirmed PE was described in a pulmonary embolism response team (PERT) registry. Data were captured prospectively for up to one year of follow-up.

In total, 13 patients were identified, with the mean age of 49 ± 19 years, 46% were female. All of them experienced a cardiac arrest, either as an initial presentation or as in-hospital event. Right ventricular (RV) dilation on echocardiogram with RV hypokinesis was present in the whole cohort. 62% of subjects received systemic thrombolysis with intravenous tissue plasminogen activator (tPA), 23% underwent catheter-directed thrombolysis therapy using the EKOS system and 31% underwent surgical embolectomy. Mean ECMO duration was 5.5 days, ranging from 2-18 days. Thirty-day mortality was 31% and one-year mortality was 54%. The majority of patients treated with ECMO after PE in the PERT registry cohort series survived for 30 days, despite the fact that most had suffered at least one cardiac arrest event.

Bottom line: Although ECMO may be helpful in many patients who have a low chance of surviving massive PE without hemodynamic support, it does come with complications, in particular, major bleeding, that was suffered by 54% of patients from the PERT registry.

Al-Bawardy, Rasha, et al. “Extracorporeal Membrane Oxygenation in Acute Massive Pulmonary Embolism: a Case Series and Review of the Literature.” Perfusion, vol. 34, no. 1, 2018, pp. 22–28., doi:10.1177/0267659118786830.

Use of extracorporeal life support (ECLS) has a major impact on a physiological systemic response and organ function of the organism. However, whether use of ECLS results in immune dysregulation of an innate immune response remains unclear. Animal studies demonstrate that exposure of blood to the ECLS circuit induces an inflammatory response, but its extent in humans has not been yet well described. The team based mostly at the University of Michigan, USA, obtained blood samples from a total of 19 patients, 7 adults and 12 children, before, during, and after ECLS. Median ECLS duration was 10 days (range: 3–108) and nine patients died during the ICU stay. A function of the innate immune system was measured by ex vivo lipopolysaccharide (LPS)-induced tumor necrosis factor-α (TNF-α) and plasma cytokine levels (interleukin [IL]- 6, IL-8, IL-10, and TNF-α). Immunoparalysis was defined as ex vivo TNF-α levels less than 200 pg/ml. After stratifying the cohort by the presence of immunoparalysis before ECLS, those immunoparalyzed showed increased response to LPS on days 1 and 3 (p = 0.016). Those without pre-ECLS immunoparalysis showed a transient decrease in response on day 3 (p = 0.008). Plasma IL-10 levels were elevated in those with pre-ECLS immunoparalysis and dropped significantly by day 1 (p = 0.031). The number treated with steroids was similar in the two groups.

In summary, the study demonstrated that patients who were immunoparalyzed before going on ECLS showed a gradual increase in levels of innate immune function (as indicated by increased ex vivo TNF-α production) over time. This increased responsiveness may be secondary to the decrease in the anti-inflammatory cytokine IL-10 between ECLS initiation and day 1, or could be due to the better oxygen delivery to different vital organs including the immune system. In addition, the data also showed a transient drop in innate immune system responsiveness in the group without pre-ECLS immunoparalysis.

Bottom line: These data suggest that the responsiveness of the immune system is changing during ECLS and that there exists a critical need for larger studies designed to obtain a more complete understanding of the relationship among ECLS, inflammation, and the immune system.

Beshish, Asaad G., et al. “The Functional Immune Response of Patients on Extracorporeal Life Support.” ASAIO Journal, vol. 65, no. 1, 2019, pp. 77–83., doi:10.1097/mat.0000000000000748.

The role of V-A ECMO in severe pediatric septic shock remains controversial. Maintaining circulatory failure in a severe septic shock is challenging. However, whether the benefits of V-A ECMO outweigh the risks remains to be determined. The analysis by a multinational team led from Melbourne, Australia, represents the largest cohort study comparing extracorporeal circulatory support and conventional therapy in severe pediatric sepsis. Newly published data from a prospective cohort study followed a total of 164 children admitted to intensive care between the years 2006 and 2014. Children aged 30 days and more were treated in tertiary PICUs in Australia, New Zealand, the Netherlands, United Kingdom, and the United States. Majority of subjects was receiving conventional therapy only (n = 120), V-A ECMO was used in approximately one quarter (n = 44). Although the survival rate demonstrated the trend of superiority in V-A ECMO group (50 vs. 40%), it did not reach the statistical significance (p = 0.25; CI, –0.3 to 0.1). However, in children who suffered an in-hospital cardiac arrest, survival to hospital discharge was 18% with conventional therapy and 42% with V-A ECMO (p = 0.02; CI, 2.5–42%).  In this scenario, extracorporeal cardiorespiratory support was associated with a 24% survival advantage. Moreover, survival was significantly higher in patients who received ECMO flows higher than 150mL/kg/min compared with children who received standard flows or no ECMO (82%, 43%, and 48%; p = 0.03; CI, 0.1–0.7 and p < 0.01; CI, 0.2–0.7, respectively). The survival rate of children who received these very high-flow rates in two of the seven participating PICUs was almost 50% higher that seen in all other patients in the study, which supports the suggestion that maximizing extracorporeal circulatory output is fundamental to survival.

Lengths of ICU and hospital stay were significantly longer for children who had V-A ECMO, but this association was not present in survivors, indicating that VA ECMO resulted in non-survivors being kept alive for longer.

Bottom line: Based on these results, V-A ECMO did not appear to offer an advantage over conventional therapy but increased ICU and hospital resource use. However, this negative finding should be treated with some caution. Despite failing to reach statistical significance, the absolute difference in mortality of 10% between the two cohorts may be deemed clinically important, particularly in light of the study’s limits with regard to sample size and the relatively wide CI.

Oberender, Felix, et al. “Venoarterial Extracorporeal Membrane Oxygenation Versus Conventional Therapy in Severe Pediatric Septic Shock.” Pediatric Critical Care Medicine, 2018, p. 1., doi:10.1097/pcc.0000000000001660.

The use of LV unloading in V-A ECMO patients is controversial and currently a matter of a thorough debate. The implementation of percutaneous LV unloading modalities is obviously attractive and apparently efficient, although costs and peculiar adverse events may undermine its utilization. Axial continuous flow and transaortic devices may represent an advanced modality of combined LV unloading and circulatory support, making V-A ECMO + transaortic suction device an intriguing association for patients supported in cardiogenic shock.

The series of 106 patients from Hamburg is extremely interesting since reports about a robust clinical experience but shed some light also on the beneficial effects of the combination of assist systems.

Successful weaning was achieved in almost 52% of the patients, with a survival to discharge of 35.8%, which was higher than the predicted SAVE score (20%).
The performance of right heart catheterization clearly indicated a marked reduction of pulmonary capillary wedge pressure after the introduction of the transaortic device during the V-A ECMO run.

Bottom line: From this series, it appears that the combination of LV unloading with V-A ECMO is beneficial. However, the ultimate impact of such an association with respect to patient survival remains to be elucidated

Schrage, Benedikt, et al. “Unloading of the Left Ventricle During Venoarterial Extracorporeal Membrane Oxygenation Therapy in Cardiogenic Shock.” JACC: Heart Failure, vol. 6, no. 12, 2018, pp. 1035–1043., doi:10.1016/j.jchf.2018.09.009.

One of the most important controversies and a major limiting factor of ECMO use, particularly in relation to venoarterial support for cardiogenic shock and cardiac arrest, relates to the long-term results. A single-center study by Finnish authors assessed retrospectively 133 V-A ECMO patients with a follow-up of 10 years. Overall, almost 50% of the subjects were successfully weaned from ECMO, whereas 12% were bridged to heart transplantation, 11.3% to a ventricular assist device, and 0.8% to total artificial heart. Survival to discharge was 64%. A short Form Health Survey was also conducted among hospital survivors, and findings showed better emotional well-being and equal energy, pain and general health perception as compared to the general population. Limitations were perceived only in physical health with 56% of the patients younger than 60 years returning to work.

Bottom line: This single-center series provides an interesting picture of the in-hospital outcome, particularly in relation to the access to more advanced circulatory and heart therapies. Favorable post-discharge quality of life was noticed, slightly in contrast with previous publications showing a high impact on neuro-psychological and physical aspects of ECMO survivors.

Jäämaa-Holmberg, Salla, et al. “Extracorporeal Membrane Oxygenation for Refractory Cardiogenic Shock: Patient Survival and Health-Related Quality of Life.” European Journal of Cardio-Thoracic Surgery, 2018, doi:10.1093/ejcts/ezy374.

The concept of “prophylactic ECMO” to improve critical preoperative modalities is not new and can be used in order to enhance perioperative course and outcome. This limited series of 12 patients in cardiogenic shock, who were evaluated for cardiac surgery, shows how the use of peripheral ECMO in the presence of a surgical diagnosis may significantly improve the preoperative conditions. Only 1 patient did not proceed to surgical correction of the underlying cardiac disease because of unknown anoxic brain injury. All the remaining patients survived to hospital discharge, with 2 death in the post-discharge period (at 76 and 230 days respectively) and with 6 patients surviving over 1 year.

 This outstanding experience, in patients who would otherwise have been excluded from surgery or most likely, would have a very complicated course and high perioperative mortality, confirm that ECMO may be effective. It could be used as a prophylactic tool to improve overall and organ conditions, enhance perioperative ICU stay, and successfully lead the patient to discharge with limited complications (in this series only 1 permanent stroke).

Bottom Line: ECMO can be used as a bridge-to-surgery in patients with mechanical complications of acute myocardial infarction and decompensated valvular disease in the presence of general conditions, which otherwise would define these patients inoperable or to transition patients from rescue to elective operation.

Watkins, A. Claire, et al. “Preoperative Venoarterial Extracorporeal Membrane Oxygenation Slashes Risk Score in Advanced Structural Heart Disease.” The Annals of Thoracic Surgery, vol. 106, no. 6, 2018, pp. 1709–1715., doi:10.1016/j.athoracsur.2018.07.038.

This retrospective study analyzed the very unique experience of the Karolinska hospital during a 21-year period. 908 patients were transported to the Stockholm hospital. Severe complications occurred in 20% of the patients, strongly associated with V-A configuration and fixed-wing transport. The occurrence of severe complications was not, however, associated with increased mortality. On-transport mortality occurred in only 2 patients.

 Bottom line: ECMO patient transport is a complex and dangerous situation. This unique series, however, demonstrates how experience and organization may allow a very safe provision of such a service, which, unfortunately, is not void of unexpected and dangerous events. Fatal cases were fortunately rare and severe complications were not associated with dismal outcome.

Fletcher-Sandersjöö, Alexander, et al. “A Single-Center Experience of 900 Interhospital Transports on Extracorporeal Membrane Oxygenation.” The Annals of Thoracic Surgery, vol. 107, no. 1, 2019, pp. 119–127., doi:10.1016/j.athoracsur.2018.07.040.

ECMO is routinely used to treat patients after cardiac arrest. This study of French authors analyzed a single-center experience with ECMO used in the setting of in-hospital (IHCA) and out-of-hospital (OHCA) cardiac arrests during a 10-year period. In total, 45 (34,4%) IHCA and 86 (65.6%) OHCA cases were included, mean age was 43.2 years and 72% were male. Patient profiles were comparable with an obvious exception of no flow and low-flow times, reaching 85.3 minutes in the OHCA group.

82.4% of the patients died on ECMO (79% for IHCA and 84% in OHCA, no significant difference). Also, the neurologic outcome at hospital discharge was not different between groups, and the presence of shockable rhythm was associated with better outcome.

 Bottom line: The study shows that even with appropriate ECMO program organization, the outcome of IHCA and OHCA do not differ. Another important issue, however, accounts for futile interventions which, for the time being, represent the major issue in this setting

 Pozzi, Matteo, et al. “Extracorporeal Life Support for Refractory Cardiac Arrest: A 10-Year Comparative Analysis.” The Annals of Thoracic Surgery, vol. 107, no. 3, 2019, pp. 809–816., doi:10.1016/j.athoracsur.2018.09.007.

Prolonged bed rest may lead to a deterioration in an overall physical condition. The retrospective study of the authors from the University of Maryland School of Medicine assessed the feasibility and outcome of patients cannulated for V-A ECMO peripherally, who were able to ambulate. Among 104 patients submitted to V-A ECMO in a 2-year period, 15 were able to ambulate at least once while on support. In 46% of the patients, the indication was decompensated heart failure, in 54% acute pulmonary embolism and 26% of the group had V-A ECMO during CPR. Mean time from ECMO implant-to-ambulation was 4 days (range 1-42 days) with a mean post-cannulation walking distance of 300 feet. No complication or dysfunction regarding patient-related or device-related function and state was observed. One-year survival was 100% for this patient group

Bottom line: This limited series demonstrates that peripheral ECMO with femoral cannulation does not preclude active movement and ambulation. In selected patients, this objective might be pursued and does not apparently induce any major complications

Pasrija, Chetan, et al. “Ambulation with Femoral Arterial Cannulation Can Be Safely Performed on Veno-Arterial Extracorporeal Membrane Oxygenation.” The Annals of Thoracic Surgery, 2018, doi:10.1016/j.athoracsur.2018.10.048.

Use of ECMO in the setting of a high-risk pulmonary embolism remains controversial. This multicenter study followed 180 patients over a period of 2 years, data were analyzed retrospectively. 128 patients were treated without ECMO, circulatory support was used in 52 subjects. Overall 30-day mortality was 48.3% with 43% in those treated without ECMO, and 61.5% in those with ECMO. In the patients who underwent ECMO+ fibrinolysis 30-day mortality was 76.5%, whereas 29.4% was the mortality for patients with ECMO +embolectomy, and 77.7% for those who had ECMO alone. In the patients who received ECMO, 38.5% had a major in-hospital bleeding event, without significant difference across groups.

The authors stated that based on the results, stand-alone ECMO does not appear as an effective strategy in this clinical setting. However, the combination of ECMO and surgical embolectomy provided a high success rate and might be the subject of dedicated investigation, especially with regards to recent favorable series.

 Bottom line: Apparently the use of ECMO in acute pulmonary embolism may account for significant mortality as compared to medical treatment alone. But, obviously, the patients undergoing ECMO were in more critical conditions and often after a cardiac arrest.

Meneveau, Nicolas, et al. “Outcomes after Extracorporeal Membrane Oxygenation for the Treatment of High-Risk Pulmonary Embolism: a Multicentre Series of 52 Cases.” European Heart Journal, vol. 39, no. 47, 2018, pp. 4196–4204., doi:10.1093/eurheartj/ehy464.