Controversies in ECMO

Parag Gharde, Sandeep Chauhan 

Controversy is defined as “Public debate about a matter which arouses conflicting opinion.” (Oxford English Dictionary)

ECMO or ECLS-

To start with there is controversy regarding the name itself. There are two abbreviations that we commonly come across in literature –

ECMO- extracorporeal membrane oxygenation

ECLS- extracorporeal life support

This support system is used in different clinical settings where the reversible organ dysfunction has affected the lung, the heart or both. When lung rest is needed the veno-venous ECMO system is used and for providing cardiac rest veno-arterial system is used. The term extracorporeal membrane oxygenation (ECMO) came into being when Dr. J. Donald Hill first used it successfully in a road traffic accident victim who developed acute respiratory distress syndrome after multiple blood transfusions. Since lung rest and oxygenation was the main therapeutic requirement, the life support system was named ECMO. But since this support system is now often being used in patients requiring cardiac support where the native lung is functioning well, the term ECMO is not suitable and hence the term ECLS is now in common use in clinical practice. The ECMO society itself is named as “Extracorporeal Life Support Organization-ELSO”. Therefore the reader should not get confused when they come across the terms- ECMO and ECLS, both meaning the same but to the author the term ECLS appears to be more appropriate as it can be used interchangeably whether support is required for the lung, the heart or for both. The use of the term ECMO is justified only if it is being used to support pulmonary function with the aim of giving rest to the lung.

Does ELSO work?

After the first successful use of ECLS in an adult, which was reported by Hill etal in 1972 (NEJM 1972; 26: 629-34.), the National Institute of Health sponsored a multicenter randomized study by Zapol etal in 1974, which showed 90% mortality in both ECLS and conventional care group. The anticipated enrolment for the study was 300 patients but the study was stopped after just 92 patients, as the death rates were similar in both the groups. This prospective randomized trial deflated the initial euphoria that was generated by the report of Hill etal. Was this supposed to be a certain miscarriage of a support system devised to deal with patients who are non-responders to conventional medical practice, even before seeing the light of the day? This study was conducted in adults with hypoxic respiratory failure and published in 1979. (JAMA 1979; 242:2193-2196). This first prospective multicenter randomized clinical trial demands a close scrutiny. During the study period a nation wide epidemic of influenza pneumonia broke out, which might have affected the results. Though the protocol included lung rest but the inflation pressures were high compared to the present recommendation, which may have caused lung injury. VA ECLS was used rather than VV ECLS, which may be responsible for high incidence of pulmonary micro-thrombosis due to decreased pulmonary blood flow. ECMO was instituted after a mean duration of 9 days of mechanical ventilation, which in the present era has been reduced to 7 days.

Morrios etal (Am J Respir Crit Care Med-1994;149(3);88) randomized 40 patients with sever ARDS to either pressure controlled inverse ratio ventilation or extracorporeal carbon dioxide removal. Survival at 30 days was not significantly different (42% in mechanical ventilation group and 33% in ECLS group).

After these two randomized trials, showing failure of ECLS in the adult setup had put ECLS into disrepute despite many anecdotal mini case series and isolated case reports showing the benefits were published. The much awaited CESAR trial (Efficacy and economic assessment of Conventional ventilator support versus Extracorporeal membrane oxygenation for Severe Adult Respiratory failure), a multicenter randomized controlled trial by Peek etal was published in Lancet 2009. (Lancet. 2009:374:1351-1363.) 180 patients were randomized and 90 patients were managed at the participating tertiary care centers while 90 were referred to Glenfield hospital. Five patients in the ECMO referral group died during transportation and thus the ECMO group consisted of 68 patients of which 43 survived to 6 months (63%). This study showed that survival apart from appropriate time of institution of ECLS also significantly depends on the expertise of the ECLS conducting center.

There is no consensus on the optimal time for institution of ECLS. Waiting too long will result in danger of end organ dysfunction and poor outcome, while too early institution without optimizing medical therapy will expose the patient to the inherent risk of ECLS.

The success in ECLS has been with neonatal acute hypoxemic respiratory failure with survival to discharge rates reaching 80%. The success story started with Bartlett etal reporting the first successful use of ECLS in a neonate in 1976.

Extracorporeal carbon-dioxide removal (ECCO₂R)

A membrane lung is used to remove carbon dioxide in conditions of acute exacerbation of chronic obstructive pulmonary disease. A low flow is required and is perfused by femoral artery-venous shunt. Low blood flow is not adequate for performing oxygenation. Morris etal conducted a randomized control trial using this device to eliminate CO₂. This trial showed no difference between ECCO₂R and the conventional treatment and the study was stopped after enrolling 40 patients only. The ECCO₂R arm used low flow in a group of patients with severe lung disease, which warranted higher ECMO flows. Before starting the trial, the trial team had limited experience on sheep and one patient.(Am J Respir Crit Care Med,1994:149; 295-305.)

Effective CO₂ clearance is achieved with blood flow as little as 10-15ml/kg/min, while oxygenation requires at least 50-60 ml/kg/min of blood flow. Nova lung (Germany) produces a membrane lung, which can be perfused using femoral arterio-venous shunt, enough blood flow for CO₂ removal.

ECLS Vs Ventricular assist device (VAD)

Patients with failing heart and waiting for heart transplant need a bridge to transplant till the time a donor heart is available. If these patients develop acute exacerbation of cardiac dysfunction, which route should be chosen- ECLS or VAD? There are no guidelines regarding this situation. The main advantage ECLS holds over VAD is in patients with bi-ventricular dysfunction, which will require a bi-VAD. VA-ECLS on other hand can support both the ventricles along with respiratory failure and refractory pulmonary artery hypertension. The only limitation is the duration of support, which at most is 4-6 weeks.

Cardiac ECLS

The controversy is regarding patient selection. Though ECLS guideline clearly defines presence of a reversible condition before institution of ECLS, there are situation when it’s difficult to predict if the condition is reversible and most often the decision is surgeon dependent who may have a biased opinion regarding the true indication for instituting ECLS. This usually results in financial burden, resource and manpower wastage, bad outcome, lowers the moral of the team and loss of faith in the support system (ECLS).

When is the ideal time to institute ECLS in cardiac surgical patient? Is it ideal to institute ECLS directly from CPB or to initiate it later in the intensive care unit after giving a trial? It has been shown that survival benefits were more in patients who went on ECLS directly from CPB thus avoiding the ill effects of prolonged low cardiac output state or subsequent to cardiopulmonary resuscitation (CPR) in the intensive care unit.

Antegarde or retrograde ECLS

Retrograde ECLS via femoral artery cannulation especially for cardiac support in a failing left ventricle is not a good option. The failing heart with this form of support has to compete with retrograde ECLS flow from the femoral artery cannula. This causes increase in LV wall stress and may even result in mitral regurgitation both of which increase left atrial (LA) pressure, thus affecting gas exchange due to pulmonary congestion. This may impair oxygenation and delivery of inadequately oxygenated blood to coronary and cerebral artery circulation will further worsen cardiac function and delay recovery. Therefore central ECLS with aortic cannulation has some advantage over peripheral ECLS, especially in case of cardiac support. But central ECLS takes time to initiate and is ideal in post cardiac surgery setup. In an emergency situation peripheral ECLS is easy to institute. Serial echocardiography needs to be done to rule out LA distention. If LA distention is present then placement of an LA vent is necessary via percutaneous atrial septostomy. Some prefer retrograde ECLS because of the fact that 10-15% patients suffer from stroke when carotid artery was used for arterial cannulation in profound hypotension or arrest. Neck vessels are used in children up to 5-6 years and femoral access in older patients. Limb ischemia is common with femoral artery cannulation and may require additional distal limb perfusion. Is axillary artery cannulation a better option? This can provide sufficiently oxygenated blood to the upper body, which is lacking with retrograde flow. The issue of limb ischemia is also addressed. The decreased pulmonary blood flow in full support VA ECLS may increase the risk of thrombus formation, in the pulmonary circulation, because of lower levels of anticoagulation.

Does the type of oxygenator influence outcome?

In vitro studies reported problems with Biomedicus (Medtronic, USA) pump heads. Thiara etal demonstrated improved circuit durability and reduced hemolysis when changing from Biomedicus (Minimax oxygenator) circuit to a Rotaflow (Lilliput 2 oxygenator), but failed to demonstrate survival improvement. (Perfusion, 2007;22:323-26). The earlier oxygenators were spiral wound silicone membrane oxygenator (Affinity, Avecor Cardiovascular Inc, USA), and the recent multiple hollow fibers <0.5 mm diameter are coated with polymethyl pentene which allows gas diffusion but not liquid. (Quadrox_D- Maquet, Jostra, Germany). It causes less platelet and plasma protein consumption, more effective gas exchange; offers lower resistance to blood flows and has smaller priming volume. Due to its less resistance to blood flow it is suitable to be used with centrifugal pump (Rotaflow- Maquet, Jostra, Germany). Siarajan etal (Interact CardioVasc Thorac Surg, 2010; 11 (4): 400-405.) have shown that with improvements in technology (centrifugal pump, hollow fiber oxygenators) there has been an improved outcome in pediatric extracorporeal support. But the study was a retrospective one.

When to start?

The guidelines are clear for institution of ECLS in respiratory failure patients, based on oxygenation index. But there are no such common guidelines for patients with low cardiac output states. Do we wait till we add all the inotropes in our arsenal? Waiting too long increases the risk of end organ damage. There is lack of consensus and thus every institution follows their own protocol. Rising lactate levels, falling mixed venous O2 saturation, wide core-peripheral temperature gap and poor urine output are signs of inadequate O2 delivery. Some consider mechanical support when no improvement is seen in signs of low cardiac output at epinephrine dose of 0.3μg/kg/min. fluid responsiveness should be seen, as these patients may be fluid responsive, in which case mechanical support will be useless.

When to stop?

There is lack of information on the time for lung recovery in case of acute lung injury or ARDS. Therefore it is difficult to set a time limit to continue with ECLS support. In case of cardiac stunning, myocardial function generally improves within 72-96 hours, while myocarditis may require 10-14 days for the ventricle to recover. But there are no guidelines to define how long is long enough? The end comes with complications like bleeding, infection or multi-organ dysfunction, but do we need to stretch this far or its just about running the pump? The answer is not simple, it’s ethical and financial as well, but we shouldn’t forget that the body also deserves respect.

Anticoagulation-

Universally, unfractionated heparin (UFH) has been used to maintain activated clotting time in range of 160-200 seconds. The incidence of heparin-induced thrombocytopenia is 10-15% in ECLS. Bleeding and thromboembolic events during ECLS are one of the major causes of mortality. Ranucci etal (Crit Care 2011;15, R275) have shifted from conventional heparin based protocol to bivalirudin based protocol after two of their cases developed HIT during ECLS and were successfully treated with stopping of heparin and starting bivalirudin infusion. The study had two arms- bivalirudin group and UFH group. The study found significantly lower bleeding in the bivalirudin group with significantly increased use of platelet concentrate and purified anti-thrombin in the UFH group. Bivalirudin unlike heparin is a direct thrombin inhibitor and therefore does not require anti-thrombin for its activity. It has a short half-life of 25 minutes. Does this study from Ranucci etal warrant a paradigm shift in the way we practice anticoagulation for the conduct of ECLS? At present what best we can take from this study is that in patients with heparin resistance and HIT, or with excessive post operative bleeding on ECMO, bivalirudin can be a savior. More experience is required with this drug before we take a step towards change. Also bivalirudin is costly, lacks an antidote and is difficult to monitor with the standard ACT machine.

Bleeding-

In a patient on ECLS who is bleeding the guidelines suggests of blood transfusion along with component therapy (platelet concentrate, fresh frozen plasma, cryo precipitate). Coagulation monitoring with instruments like Sono-clot, thromboelastography gives us the insight into coagulation factor deficiency so that we can tailor our transfusion to meet the deficiency. The guideline also suggests temporarily withholding anticoagulation in patients with severe hemorrhage. But the lowest effective level of anticoagulation is not known. There are case reports of safe ECLS support for several days without any systemic anticoagulation. But one needs to be careful during periods of low flow ECLS (<2 L/min).

Can recombinant FVIIa be used in patients on ECLS who are bleeding? Again there are no guidelines on this and there have been few case reports of its use in ECLS showing both positive and negative results.

Extracorporeal support during cardiopulmonary resuscitation (E-CPR)

This is a new evolving dimension of ECLS. In emergency settings it’s often impossible to assess patient’s neurological status and predict the reversibility of the clinical picture. How can we select patients who will benefit- it’s a difficult call. Since E-CPR is a new extension of ECLS there are no large studies to look at for guidance and implementation. It is still in its infancy. The survival to discharge rate for in-hospital cardiac arrest is 10-15%. ECPR in witnessed arrest showed early return of spontaneous rhythm in the E-CPR group due to active decompression of the heart by ECLS and oxygen supply to the myocardium. The American Heart Association guidelines for CPR recommend consideration for ECLS to aid cardiopulmonary resuscitation in patients with refractory cardiac arrest (persistence of circulatory arrest despite more than 30 minutes of appropriate CPR), who have an easily reversible event and have had an excellent CPR. These guidelines are for witnessed in- hospital cardiac arrest. There are no guidelines for out of hospital cardiac arrest. Guen etal (Crit. Care 2011; 15: R29) in their study used ECLS on 51 patients of out-of-hospital cardiac arrest. 90% (n-46) died within 48 hours. At present in a developing country like India where health resources, health infrastructure and well-trained paramedics are limited E-CPR appears to be a distant reality, especially in the out-of-hospital cardiac arrest scenario.

Suggested reading

1. Gray BW, Shaffer AW, Mychaliska GB. Advances in neonatalextracorporealsupport: the role ofextracorporeal membrane oxygenationand the artificial placenta. Clin Perinatol. 2012; 39(2):311-29.
2. Chai PJ, Jacobs JP, Dalton HJ, Costello JM, Cooper DS, Kirsch R, Rosenthal T, Graziano JN, Quintessenza JA. Extracorporealcardiopulmonary resuscitation for post-operative cardiac arrest: indications, techniques,controversies, and early results–what is known (and unknown). Cardiol Young. 2011; 21 Suppl 2:109-17.
3. Stulak JM, Dearani JA, Burkhart HM, Barnes RD, Scott PD, Schears GJ. ECMOcannulationcontroversiesand complications. Semin Cardiothorac Vasc Anesth. 2009;13(3):176-82
4. Ford JW. NeonatalECMO: Currentcontroversiesand trends. Neonatal Netw. 2006;25(4):229-38.