How I approach membrane lung dysfunction in patients receiving ECMO

© The Author(s) 2020. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creat iveco mmons .org/publi cdoma in/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Introduction With improvements in circuit technology and expanding supportive evidence, extracorporeal membrane oxygenation (ECMO) use has grown rapidly over the past decade [1]. Advances in pump and membrane lung (ML) design have led to simpler and more efficient circuits. Circuitrelated complications, however, remain frequent and associated with considerable morbidity [2].


Introduction
With improvements in circuit technology and expanding supportive evidence, extracorporeal membrane oxygenation (ECMO) use has grown rapidly over the past decade [1]. Advances in pump and membrane lung (ML) design have led to simpler and more efficient circuits. Circuitrelated complications, however, remain frequent and associated with considerable morbidity [2].

Mechanisms of membrane lung dysfunction
The ML is responsible for oxygen uptake and carbon dioxide removal. The non-biologic surface of the ML activates inflammatory and coagulation pathways with thrombus formation, fibrinolysis, and leukocyte activation [3][4][5] leading to ML dysfunction. Activation of coagulation and fibrinolysis can precipitate systemic coagulopathy or hemolysis, while clot deposition can obstruct blood flow [6,7]. Additionally, moisture buildup in the gas phase and protein and cellular debris accumulation in the blood phase may contribute to shunt and dead-space physiology, respectively, impairing gas exchange [8,9]. These three categories-hematologic abnormalities, mechanical obstruction, and inadequate gas exchange-prompt the majority of ML exchanges.

Hematologic profile
Monitoring of hematologic variables, including coagulation and hemolysis labs, can help identify the development of an ECMO coagulopathy or hemolysis.

Pressure monitoring
The pressure drop across the ML (ΔP) is measured as (Additional file 1: Supplemental Figure): where P Pre = pre-ML pressure, P Post = post-ML pressure.
As clot forms in the ML, increases in resistance (R ML ) are reflected as increases in ΔP. To correct for changes in blood flow rate (BFR), monitoring of ΔP normalized for BF rate (ΔP/BFR) more directly reflects R ML .

Membrane lung gas transfer
Applying the Fick principle across the ML, oxygen (O 2 ) transfer may be calculated as: where V′O 2 = O 2 transfer across the ML (mL/min), BFR = blood flow rate (L/min), C x O 2 = O 2 content of (pre-/post-ML) blood (mL/L) for where Hb = hemoglobin (g/dL), S x O 2 = O 2 saturation of (pre-/post-ML) blood, P x O 2 = O 2 partial pressure of (pre-/post-ML) blood (mmHg).
Measurement of V′O 2 provides an objective measure of oxygen transfer and can confirm ML dysfunction, when clinically indicated.

Membrane lung dysfunction
Prompt recognition of ML dysfunction is vital for safety, allowing for elective replacement in a controlled manner. On the other hand, replacement of an adequately functioning device-requiring temporary cessation of ECMO Open Access *Correspondence: bzakhary@gmail.com 1 Division of Pulmonary and Critical Care Medicine, Oregon Health and Science University, Portland, OR, USA Full list of author information is available at the end of the article support-places the patient at unnecessary risk while consuming a limited and expensive resource.
Based on the pathophysiology of the ML, replacement may be required for one of three reasons: if there is (A) an associated hematologic abnormality, (B) an increasing obstruction to blood flow, or (C) inadequate gas exchange (Fig. 1).

Hematologic abnormalities
The presence of an ECMO coagulopathy, typified by elevated clotting times, hypofibrinogenemia, thrombocytopenia, and elevated D-dimer without alternate explanation raises concern for circuit-related coagulopathy (CRC). Alternatively, evidence of hemolysis with elevated plasma-free hemoglobin, without alternate explanation, is concerning for circuit-related hemolysis. In both cases, the diagnosis is presumptive and only confirmed when values normalize after circuit exchange [6].

Obstruction to blood flow
Increasing ΔP/BFR suggests increasing ML clot burden. As different MLs have different R ML , no cut-off values of ΔP define ML dysfunction and the trend should be carefully considered. A rapidly increasing ΔP, even if not associated with reduced gas exchange efficiency, is often a harbinger of impending ML failure and should prompt consideration of ML exchange. When ML pressures are not measured, an increasing pump speed requirement to maintain a stable BFR can serve as a surrogate for increasing ΔP, with the caveat that pump preload and afterload also affect this relationship. Fig. 1 Algorithmic approach to membrane lung monitoring for membrane lung dysfunction. Cut-off values for consideration of ML exchange are suggested values based on author experience and should be considered in the context of the patient and dependence on ECMO support. Please refer to text for details. ML, membrane lung; Plt, platelet count; INR, international normalized ratio; aPTT, activated partial thromboplastin time; fHb, free hemoglobin; LDH, lactate dehydrogenase; ΔP, Pressure drop across the ML; P Pre , pre-ML pressure; P Post , post-ML pressure; R ML , resistance within the ML; BFR, blood flow rate; V′O 2 , membrane lung oxygen uptake; C Pre O 2 , O 2 content of pre-ML blood; C Post O 2 , O 2 content of post-ML blood; P Pre O 2 , partial pressure of pre-ML O 2 ; P Post O 2 , partial pressure of post-ML O 2 ; P Pre CO 2 , partial pressure of pre-ML CO 2 ; P Post CO 2 , partial pressure of post-ML CO 2 . Flowchart is designed with adult ECMO patients in mind and may not be applicable to pediatric or neonatal patients. *Extent and frequency of coagulation and hemolysis lab monitoring is not well-established and will vary by center. Not all labs are required to diagnose coagulopathy or hemolysis. **When a ML fails, we recommend considering switching the entire circuit, rather than just the ML, if: (a) the ML and pump head are fused; (b) the ML dysfunction occurs in the setting of circuit-related coagulopathy; or (c) the ML dysfunction occurs in the setting of an older circuit (i.e., longer than 2 weeks). While the first is due to technical limitation, the latter aim to reduce the risk of ongoing or new circuit-related coagulopathy in circuits at risk for this phenomenon

Inadequate oxygen uptake
Worsening oxygenation during ECMO should prompt quantification of oxygen transfer. When the ML is no longer able to meet patient oxygen demand, ML exchange is indicated. There are three important considerations in making this decision.
First, it is necessary that measured V′O 2 is truly a maximal value. If circuit BFR is low, for example, the blood will be fully saturated early in the ML path and reserve will exist for additional oxygen transfer as BFR is increased. Similarly, if C Pre O 2 is artificially elevated, due to high recirculation fraction or impaired tissue extraction, or if the fraction of delivered oxygen in the sweep gas (F D O 2 ) is below 100%, the gradient driving oxygen transfer is reduced, and measured V′O 2 may not represent maximal capacity. As such, BFR should be sufficiently high that further increases do not increase arterial saturation, recirculation fraction should be minimized, and ML F D O 2 set to 100% to ensure an accurate assessment of maximal V′O 2 .
Second, though P Post-ML O 2 less than 200 mmHg can suggest a failing ML [6], it is vital to calculate V′O 2 for confirmation. In the setting of low C Pre O 2 or high circuit BFR, blood exiting the ML may not be fully saturated, with low P Post O 2 , despite normal V′O 2 . In this case, if the ML is exchanged, the patient is placed at risk without subsequent improvement in oxygen delivery.
Finally, no absolute values diagnose inadequate oxygen transfer and clinical context is important. In general, however, in a patient with hypoxemia and a ML with maximal V′O 2 < 100-150 mL/min, ML exchange is typically indicated.

Inadequate carbon dioxide clearance
ML dysfunction can also manifest as inadequate CO 2 clearance. Calculation of V′CO 2 is not typically performed as it varies in a nonlinear fashion with sweep gas flow rate and requires sampling ML exhaust CO 2 [10]. However, persistent P Post-ML CO 2 greater than 40 mmHg [6] and clearance of less than 10 mmHg PCO 2 between pre-and post-ML blood gases despite sweep gas flow rates of 10 L/min or greater is suggestive of ML dysfunction and ML exchange should be considered.

Sudden membrane lung failure
While serial monitoring of the ML may identify markers of dysfunction and allow for elective exchange, acute ML failure is a potentially life-threatening event with unique considerations. Mechanisms to ensure optimal management are provided in the Additional file 2.

Conclusion
The decision to exchange a ML is complex and without clear guidelines. In this manuscript, we outline a physiologic approach to troubleshooting this common yet high risk event.