Secondary Logo

Journal Logo

Review Article

Hemolysis and Plasma Free Hemoglobin During Extracorporeal Membrane Oxygenation Support: From Clinical Implications to Laboratory Details

Dufour, Nicolas*; Radjou, Aguila*; Thuong, Marie*

Author Information
doi: 10.1097/MAT.0000000000000974
  • Free


Hemolysis and Physiologic Mechanisms to Circumvent Intrinsic Toxicity of Heme

Red blood cells, dominant in terms of cellular blood constituents, are potential silent and harmful weapons if destroyed in amount that overwhelmed management capability of natural systems existing in mammals. Intrinsic red blood cell toxicity is mediated by hemoglobin (Hb), a major metalloprotein constituted by four chains of globin (α2β2 tetramer) and four heme groups (each containing an iron ion). Each chain of globin chain welcomes one heme molecule by noncovalent (but strong) association.

When outside its protective reductive environment within the red blood cell, Hb (especially heme) can exert important oxidative damages on various organs (see. below).

Normal catabolism of red blood cells, after a mean life span of 120 days, takes place in bone marrow, liver, and spleen. Local macrophages are responsible for normal clearance of old red blood cells (erythrophagocytosis, physiologic intratissue hemolysis). Within the macrophages, degradation of Hb and heme via heme oxygenase, leads to production of amino acids (by-products of globin), iron, carbon monoxide (CO), and biliverdin (then converted in bilirubin).

In contrast, intravascular hemolysis (schematically, the uncontrolled and unwanted hemolysis), is very weakly present in normal situation. When happening, free Hb spontaneously dissociates in αβ-dimers which can cross the glomerular membrane and reach extravascular spaces thanks to its small size (32 kilodalton). Free heme is also released following spontaneous oxidative reactions.1 Normal management of such unwanted release of oxidant compounds involves a first step of neutralization by two plasma proteins: haptoglobin (HptG) and hemopexin (HpX). Haptoglobin rapidly interacts with extracellular Hb dimers and neutralizes its reactive ferric part whereas HpX takes over free heme. Once associated with their toxic moieties, HptG and HpX are retrieved by hepatocytes or splenocytes by dedicated receptors (CD163 for Hb-HptG complex and CD91 for heme-HpX complex), also present on the surface of monocytes/macrophages.1,2

Pathophysiology and Clinical Consequences of Hemolysis in the Extracorporeal Membrane Oxygenation Setting

Hemoglobin represents the main component of red blood cells (approximately 97% of its dry mass) and it is assumed that regarding its biochemical properties, hemolysis-related toxicity is attributable to Hb. Nonetheless, unknown compounds from the membrane or the cytoplasm of red blood cells could play an underestimated role in this pathophysiology.

When intravascular hemolysis is important, overwhelming scavenging capacity of HptG, free Hb dimers start to circulate freely in the bloodstream, diffuse and release free heme. We summarize here the main currently described consequences of red blood cell hemolysis as well as data available regarding the associated morbidity and mortality.

Pathophysiologic Consequences of Hemolysis

Three main pathophysiologic aspects dominate in the short-term ECMO setting.

  1. Increased vascular tone because of nitric oxide (NO) depletion: Free Hb is able to rapidly interact with NO and to catalyze its transformation into nitrate if oxygen-Hb is involved or into a ferric-NO complex if deoxy-Hb is involved.1 Subsequent NO depletion is then responsible for a severe loss in a major negative regulator of the vascular smooth muscle tone, inducing a potent vasoconstriction, supported by human and animal evidences.1,3 This low-NO dependent vasoconstriction is responsible for both an increase in systemic and in pulmonary vascular resistance.4
  2. Abnormal coagulation and platelet activation: The afford mentioned NO depletion is also responsible for an impaired control of platelet and endothelial behavior towards coagulation and platelet aggregation. In vitro, free Hb promotes platelet adhesion and microthrombi formation mediated by von Willebrand factor.5 Others mechanisms are involved and derived from hemolysis-associated inflammation. The net result is a procoagulant state6,7 with clinical evidences observed in chronic hemolytic disorders (such as sickle cell disease, paroxysmal nocturnal hemoglobinuria, etc.) which are now established to support a strong link between hemolysis and a prothrombotic state.4,8,9
  3. Renal tubular toxicity: Heme-containing proteins (Hb but also myoglobin in case of rhabdomyolysis) are known for decades to be toxic to kidneys. Histopathologic findings comprise acute tubular necrosis (ischemic damage) and in severe forms precipitates with casts formation which obstruct renal tubules.10,11 When bound by HptG, Hb forms a protein complex that is too large to cross the glomerular membrane, preventing its toxicity to downstream kidney structures. In contrast, unbound dimeric Hb is filtered by the glomerulus and penetrates into the proximal tubule cells using the megalin and cubilin receptors.12 Large uptake of heme-containing proteins is thought to saturate normal detoxifying capability of tubular cells, particularly the way tubular cells deal with high intracellular concentration of free iron, highly cytotoxic.13

Clinical Consequences of Hemolysis in the Extracorporeal Membrane Oxygenation Setting

With a good level of certainty, all the pathologic changes addressed above and caused by free Hb (FHb) irruption in the bloodstream seem to be associated with an increased morbidity (through renal impairment) and perhaps also mortality.

Main publications dealing with mortality and kidney-related morbidity are summarized below and in Table 1.

Table 1.:
Clinical Consequences of Hemolysis During Extracorporeal Membrane Oxygenation Support, Emanating From Main Publications Dealing With Mortality and Kidney-Related Morbidity

In a Chinese single-center retrospective study published in 2016 and dealing with venoarterial (VA) ECMO in adults, Lyu et al.19 observed a correlation between peak plasma free Hb (pFHb) and peak serum creatinine (first 3 days following ECMO onset). They found that an increased pFHb was a predictor of acute renal failure in this population. The authors also observed a lower survival rate in patients developing acute renal failure. The same team, focusing on pediatric patients treated by ECMO after cardiac surgery, had previously reported the same observation with maximum pFHb being an independent risk factor for acute renal failure.20

An Australian team,21 working retrospectively on an adult population treated with both venovenous (VV) and VA ECMO, noticed that severe hemolysis was almost exclusively present during VV ECMO. They observed an increase (nonstatistically significant) in in-hospital mortality, related to pFHb concentration.

In another retrospective single-center American study17 published in 2015 and conducted in adults treated mostly with VA ECMO (82%), severe hemolysis (pFHb > 500 mg/l) in the first 24 hours of ECMO onset was identified as an independent predictor of mortality.

Based on a VV ECMO retrospective cohort of 318 adults, Lehle et al.18 showed in 2015 that the occurrence of hemolysis was associated with an increased mortality (peak pFHb in survivors of 90 mg/l, interquartile range [60–142] vs. 148 mg/l [91–256] in non-survivors; p ≤ 0.001).

In a more recent study (2018), dealing with a pediatric population (n = 216) that exhibited a quite high degree of hemolysis compared with other studies (see Table 2), Dalton et al.23 observed a weak association between hemolysis and the development of renal failure. Using multivariable Cox modeling, they observed that daily pFHb was independently associated with the development of renal failure during ECMO, with a hazard ratio of 1.04 for each 100 mg/l increase in pFHb (95% confidence interval [CI], 1.02–1.06; p < 0.001). They did not observe any impact on mortality.

Table 2.:
Prevalence of Hemolysis During Extracorporeal Membrane Oxygenation Support (Only Publications Dealing With Retrospective Data With Sufficient Number of Patients and Where Hemolysis Data Were Reported Have Been Considered

Finally, still in a pediatric population (n = 96), Okochi et al.24 observed that patients with pFHb > 300 mg/l had a 10-fold increase in in-hospital mortality (95% CI, 3.4–32; p < 0.001). As well, hemolysis was associated with use of slow continuous ultrafiltration, but not continuous renal replacement therapy.

To conclude, even if definitive conclusions are hard to draw because of the heterogeneity of the available studies (whether it be in terms of population, treatment management, or statistical methodologies), hemolysis seems to be associated with an increased morbidity through renal impairment, most likely provisional. An important limit has nonetheless to be kept in mind: association is not causation. As perfectly reviewed by Kilburn et al.,26 aside from hemolysis, many others ECMO-associated factors play a role in the pathophysiology of acute kidney injury during ECMO (ischemia-reperfusion injury, systemic inflammatory response, renal micromacrocirculatory dysfunction, etc.). As well, the duration of the hemolytic stress is most likely an important parameter, perhaps more critical than the isolated value of pFHb concentration by itself: because of the time-limited buffering capability of the system, prolonged low/moderate-level of hemolysis could most likely prove to be more deleterious than a time-limited severe hemolytic event.

More clinical studies are needed to strengthen these observations, in addition to the existing incontestable in vivo data and the strong underlying pathophysiologic evidences.

Hemolysis During Extracorporeal Membrane Oxygenation: Prevalence, How and When?

Since the first extracorporeal circulations were applied in the early 1960s, hemolysis was rapidly recognized as a complication attributable to the global circuitry, namely its moving parts (the roller pumps) and area of high resistance flow (the oxygenator). Fifty years later, Mendler-concept centrifugal pumps and low-resistance membrane oxygenators as well as biocompatible coatings have clearly improved the mechanical part of the problem. But hemolysis still occurs, more or less silently. Two factors are recognized to promote and generate hemolysis: thrombosis (head pump and oxygenator) and excessive mechanical stress (head pump).16,18,27–31

Thrombosis As Risk Factor

Thrombosis is responsible for locally impaired flow conditions that may lead to mechanical stress and then hemolysis. Although oxygenator thrombosis is less likely involved in acute hemolysis because of the lower mechanical stress involved, head pump thrombosis is a critical situation usually associated with acute and severe hemolysis.18

Having said that, a relevant question is to know if a potential correlation exists between level of anticoagulation and hemolysis. Unfortunately, it is not possible to correctly answer this question as data available in the literature are not robust enough. Some authors report anticoagulation dose, others report the results of coagulation test, never both of these data… complexifying a valuable analysis, with opposing results regarding hemolysis. For example:

  1. Okochi et al.24 observed that hemolysis (defined as pFHb > 300 mg/l) was present in 52% of their patients if anti-Xa activity was ≤ 0.2 unit/ml, but only in 18% if > 0.2 unit/ml;
  2. Dalton et al.23 observed that lower heparin dose adjusted for body weight was independently associated with higher daily pFHb;
  3. Lou et al.16 did not observe any correlation between heparin dose and level of hemolysis; and
  4. Pan et al.21 observed that aPPT (activated Partial Thromboplastin Time) tends to be longer in patient with high level of hemolysis (51.7 seconds for high-level hemolysis, 50.2 seconds for low level, and 47.5 seconds for normal pFHb; p = 0.18). In this population, patients with severe hemolysis had also a significantly higher level of fibrinogen (5.6 g/l vs. 3.6; p = 0.03).

Furthermore, an important limit has to be highlighted to correctly interpret these data (and to complexify the problem): the hypercoagulability state associated with ECMO may be responsible for a lengthening of coagulation times (consumption of clotting factors) which are therefore no longer exclusively related to the pharmacological effect of the anticoagulant treatment.

Excessive Mechanical Stress, Rotation Speed, Cavitation and Blood-Air Interface

Clinical experience as well as experimental data agree on the fact that head of centrifugal pump could be a major source of hemolysis when rotating in unsuitable conditions. A high negative pressure in the inlet pipe is not, by itself, a sufficient cause of hemolysis32–34 but is susceptible to promote hydrodynamic cavitation in the head pump when the latter is spinning at relatively high speed (unsuitable settings between flow and rotation speed). Hydrodynamic cavitation is the process leading to vaporization, bubble generation, and bubble implosion which occurs in a liquid as a result of a decrease and subsequent increase in local pressure. Empirically observed on helix-driven boat propellers at high speed where bubbles arise and disappear within the water around blades, hydrodynamic cavitation is an unwanted event in a centrifugal pump carrying blood! Because it generates a temporary high shear stress together with a blood-air interface known to be a key factor,33,34 the worse mechanical conditions are gathered to promote a severe and acute hemolysis.

In other words, a pump set to a high rotation speed with an optimal preload and afterload will be able to provide a high flow without intense mechanical stress within the pump head, resulting in an absence of (or a weak) hemolysis. The same speed setting with inadequate preload (e.g., smaller cannula) and/or increased afterload (e.g., membrane thrombosis) may result in a high mechanical stress, promoting hydrodynamic cavitation and a high level of hemolysis.

In their respective studies, Lou et al.16 and Pan et al.21 did not find any association between the pump speed and the level of pFHb, but the first authors observed that the higher was the severity of hemolysis, the lower was the pressure in the inlet line (respectively –6 mm Hg when no hemolysis, –9 mm Hg for mild hemolysis, –10 mm Hg for moderate hemolysis, and –13 mm Hg for severe one; p = 0.016).

Extracorporeal Membrane Oxygenation and Severe Hemolysis: When Does It Occur?

As summarized by expert authors,27,28 some circumstances predispose to such high level of hemolysis and could be prevented. But hemolysis may also be present more silently, generating a low noise damage.

First, when the venous line chatters or is momentarily occluded (line kinking, patient cough, insufficient flow in the drainage cannula) and if a high rotation speed is set up (> 3,000 rpm), negative pressure within the head pump can exceed –700 mm Hg, causing cavitation and severe hemolysis. Inappropriate action facing such decrease in pump flow is often to increase the pump speed, which worsen the problem. Reduce the pump speed and correct the triggering factors (hypovolemia, venous cannula displacement, twisted line, etc.) are the only correct responses to such scenario.

Second, head pump thrombosis should be suspected when a rapid and substantial increase in lactate dehydrogenase (LDH) and pFHb is observed as well as the perception of abnormal noise or vibration in the pump head.18,30,31 Decrease in pump mechanical efficiency (drop in blood flow) is a very late sign of malfunction and should not be awaited to decide an urgent replacement.

Prevalence of Hemolysis During Extracorporeal Membrane Oxygenation

Hemolysis is a common phenomenon during ECMO support but in most cases, its intensity is low, generating tolerable pFHb concentration. If we focus on the limited number of studies dealing with hemolysis during ECMO performed with a centrifugal pump,16–19,21,23,24 severe hemolysis (pFHb > 500 mg/l) appeared to be a rare (but not exceptional) event (Table 2), observed in ≈2–20% of the patients (having excluded the study by Dalton et al.,23 displaying an unusual prevalence of severe hemolysis, i.e., 67%). Furthermore, after exclusion of a study reporting a particularly low prevalence of hemolysis,18 only ≈30% of the patients display a low-level hemolysis (pFHb < 100 mg/l), considered as not clinically relevant, meaning that the largest number of patients experience a “significant” hemolysis, with the associated pathophysiologic consequences.

There is no data robust enough to draw tendencies regarding hemolysis propensity in case of VV ECMO compared to VA ECMO, which can be a relevant question as blood flow is usually higher during VV setting.

Clinical and Biologic Diagnosis of Hemolytic Event

Clinical manifestations related to chronic low-intensity hemolysis in patients with ECMO support are very rare and nonspecific.

Conversely, acute and severe hemolytic events may be detected when a change in urine color happens in patients with preserved urine output (hemoglobinuria). As well, in patients treated by continuous hemofiltration, an acute change in the color of the effluent (that turns pinkish-reddish) may happen35 and should be considered as an alarm message. Associated general clinical signs comprise shivering, hyperthermia, hypotension, and delayed jaundice.

Biologically, the classical hemolysis-related changes are present: LDH and phosphatemia increase, decline in Hb, HptG decrease, unconjugated bilirubin increase, and hyperkalemia in severe forms. Nonetheless, as 1) lactate dehydrogenase is totally unspecific, 2) haptoglobin is often already reduced, and 3) increase in unconjugated bilirubin is delayed, the assessment of pFHb is a key biologic marker to estimate the severity of hemolysis.

Besides, endogenous production of CO is also significantly increased during severe hemolytic events, linked to the normal catabolic pathway of Hb through heme oxygenase. As CO is rapidly catch by Hb because of its high affinity, this parameter could be readily assessed by routine point of care co-oximeter through carboxyhemoglobin (HbCO) level with a simple venous or arterial blood sample. Many authors have underlined the interest of this indirect marker assessed in blood35,36 as well as in exhaled gas.37

Assessing Plasma Free Hemoglobin: The Laboratory Side (Where the Devil Is in the Details)

Normal and Pathologic Concentrations in Plasma Free Hemoglobin

Old data emanating from healthy adult volunteers found pFHb concentrations below 50 mg/l38 and many hospital laboratories have adopted this cut-off value. Nonetheless, team dealing with ECMO devices agree to consider as significant (thus pathologic) pFHb concentration which are higher than 100 mg/l (sometimes 150 mg/l).21,28 Quite consensually, severe hemolysis is defined when pFHb is > 500 mg/l17,21,39 or > 1,000 mg/l.16,18,23 Between these two extreme ranges (100 to 500–1,000 mg/l), hemolysis is considered as moderate/mild.

It should be noticed that such classification has been set quite empirically and is not associated with the occurrence of clinical manifestations such as the prevalence of acute renal failure or a circuitry dysfunction. The time component of such parameter (how long is a given concentration of pFHb able to persist, how long to exhaust the buffering capacities of HtpG and HpX) is of a high importance, precluding any simple and direct relationship between a given concentration of pFHb and the associated morbimortality.

Preanalytical Requirements

As the purpose is to assay a marker of hemolysis, any factors that could increase hemolysis should be strictly avoided. Classical recommendations for blood sampling and tubes management should be applied such as a limited time of tourniquet application (in the ECMO setting, most of the patients have an arterial catheter that rules out this recommendation), an immediate transfer to the laboratory, the fact of avoiding mechanical stress during transport (e.g., pneumatic tube without correct packaging leading to violent impacts) and a quick centrifugation upon reception.

Available Methods

Five main methods are currently used to assess pFHb, each of them having their own advantages and limitations.

The first method, the simplest one, relies on the direct quantification using direct spectrophotometric assays: the absorbance of plasma is measured at various discrete wavelengths representative of the Hb absorption peaks (main peak at 415 nm) as well as bilirubin (450 nm) and/or lipid peaks (600–700 nm) for correction purposes in case of icteric and/or lipemic samples. Absorbance is then function of pFHb concentration, deduced from known standard calibrators. Numerous authors have established their own formula with variable results.38

The two next methods rely on the conversion of Hb and its derivatives into a stable end product, that possesses a peak absorption at a given wavelength, allowing its quantification on a spectrophotometer: this is the alkaline hematin detergent method (also called the Triton/NaOH method)40,41 and the Drabkin-derived assays.42

The fourth method is a tetramethylbenzidine (TMB)-based assay that rely on the pseudoperoxidase activity of heme (original method described by Standefer and Vanderjagt43 in 1977). This peroxidase activity is used to catalyze a reaction between hydrogen peroxide and TMB. The oxidized TMB has a blue color and its concentration is assessed at a wavelength of 650 nm.

The last method is to use enzyme-linked immunosorbent assay (ELISA)-based approaches in its sandwich version: a first monospecific or monoclonal antibody, covalently linked to a support, allows to catch the free Hb molecule. Quantification of pFHb is then performed by using a second specific antibody that is linked to an enzyme and that binds to the immobilized antigen (the Hb molecule). A substrate is then added and converted by the enzyme into a colored product, whose amount (spectrophotometrically determined) is proportional to the amount of pFHb. This method is then the most specific (and costly) one. Depending on the antibody used, targeted epitope on the Hb molecule may be masked or not by HptG, allowing such method to only assess the real free concentration of unbound Hb in plasma or the sum of total plasma Hb (bound and unbound to HptG).

To the best of our knowledge, it must be highlighted that there is no recommended method nor standardized kit actually available on the market having an approval from competent authorities (e.g., In vitro Diagnostic label from United State Federal Drug Administration or CE (Conformité Européenne) label from European Community). None of the major leader in clinical biochemistry automation (Roche diagnostics, Siemens, Beckman Coulter, DiaSorin, Abbott, Mindray, Ortho-Clinical Diagnostics, etc.) offers comprehensive assays to clinical laboratories. Thus, each laboratory promotes its own technic, adapted from those previously described and subject to limitations.

Missing Points and General Limitations: Do We Really Assess Free Hemoglobin?

First, as rigorously demonstrated by Oh et al.44 in a remarkable work, all methods reviewed here (except ELISA-based ones) are not selective regarding heme and Hb, as they rely on biochemical properties that are similar for both cell-free heme and Hb. Thus, one should rather speak about a “total heme-based species” concentration. The clinical relevance of such distinction is unidentified yet but has to be known because heme and Hb do not elicit the same pathophysiologic pathways.1

Second, it has to be emphasized that most of these methods, except ELISA-based methods, are not able to distinguish free or bound (to HptG and HpX) forms of Hb and cell-free heme.44 The measured concentration is then the quasi-sum of bound and unbound heme-based species. Again, such distinction is of unknown impact on clinical appraisal, but it remains certain that heme-derived toxicity is mediated by unbound compounds, once scavenging properties of HptG and HpX are overwhelmed.1

With such clarification, it clearly appears that the so-called designation “plasma free hemoglobin” is a very approximate and imprecise one.

In addition to the two major conceptual limitations addressed above, others limits have to be mentioned: as virtually all the analytical process of these methods have a common end way that is based on spectrophotometric measurements, icteric specimens may provide invalid results38,44,45 with more or less artifacts. As well, hypertriglyceridemia (parenteral nutrition, high dose propofol infusion, etc.) could be highly detrimental when using direct spectrophotometric methods46,47 (that is not the case with ELISA-based ones).

Hemolysis and Routine Lab Tests: Drawbacks and Interferences

Hemolysis is known to interfere with many chemistry lab tests as well as providing an increase in compounds emanating from red blood cell cytoplasm (K+, Mg2+, PO42–, LDH, etc.).

Blood potassium concentration and hemolysis: no interference but a critical question related to its origin.

Regarding assessment of critical parameters susceptible to be life threatening, only one should be considered with a particular attention: the potassium concentration. An important point is here to distinguish an in vitro hemolysis (postsampling, generated in the collection tube) from an in vivo hemolysis. In the first case (the most frequent one) potassium concentration does not reflect the real in vivo concentration. In the second case, the potassium concentration is real, as lysis of K+ intracellular-rich red blood cells occurred before sampling, releasing K+ in the bloodstream.

Nonetheless, the way to deal with such information has to be totally different. In the first case (in vitro hemolysis), K+ concentration measured is overestimated (and clinicians manage the case with full knowledge of the facts). In the second case (in vivo hemolysis), the K+ concentration measured is real (same concentration in the tube and in the patient) and has to be managed as is.

Thus, a problem could emerge from the laboratory staff that is potentially not aware of a potential ECMO-induced hemolysis and considers that such increase in K+ is not real resulting from an in vitro hemolysis (actually, a routine situation and real-life problem). If the clinician is not warned, a second sample is demanded by the laboratory for control purpose, delaying urgent and lifesaving actions with potential fatal outcome, as reported by Ismail et al.48

Hemolysis as the biochemist’s devil that could impairs accuracy of virtually all spectrophotometry-based analysis.

Hemolysis, because of heme absorption spectrum, is able to interfere with many spectrophotometry-based assays if significant. Mild hemolysis is not a real problem in routine practice, causing weak interferences, but concentration of pFHb > 500 mg/l may be annoying, impairing an accurate assessment of numerous laboratory parameters in a wide field of discipline (biochemistry, toxicology, hemostasis).49 Of note, in the ECMO setting, aPPT and anti-Xa monitoring could be falsely shortened if hemolysis is present. However, the amplitude of the artifact may be considered as nonclinically relevant unless a severe hemolysis is present (pFHb > 500 mg/l).50–52

Conclusions and Outlooks

To date and with the currently available technologies, hemolysis is an unavoidable side effect of any extracorporeal circulation, especially when a high blood flow is required to assist a respiratory or heart failure, like during ECMO support.

However, whereas hemolysis is present in most cases as a low background noise, severe and acute hemolysis may occur with tangible consequences in terms of morbidity (essentially kidney injury) and mortality. Circumstances that lead to severe hemolysis, primarily associated with circuitry malfunction (head pump thrombosis, inadequate flow/speed setting, etc.), have to be known and anticipated by clinician and nurse staff involved in daily care of patients receiving an ECMO support. As well, as summarized in Figure 1, detection and management of hemolytic events should be part of the routine practice.

Figure 1.:
Global picture of ECMO-induced hemolysis-related issues and suggested management. ECMO, extracorporeal membrane oxygenation; HbCO, carboxyhemoglobin; LDH, lactate dehydrogenase; pFHb, plasma free hemoglobin.

Considering the current absence of specific treatment of severe hemolytic events, several outlooks that might preclude pessimistic conclusions are however existing. Especially, significant hopes are expected from HptG replacement or supraphysiologic dosing, in order to annihilate the toxicity of free Hb. Persuasive in vitro and in vivo data in animals exist to seriously deepen such therapeutic direction in human.53,54 Thus, although purified HptG is available in Japan for patients exposed to acute hemolysis (extracorporeal circulation, thermal injury, or massive transfusion) to limit kidney injury,55,56 research is also ongoing to target chronic hemolytic diseases.1,57

A less specific approach could comprise pFHb removal from the bloodstream using high permeability filters during hemodialysis.58,59

Whatever the future therapeutic options considered, pathophysiology of heme-related toxicity (HRT) clearly emphasizes the role of natural scavenger proteins as powerful systems but also stresses their limited buffering capability, quickly overwhelmed in case of mild-chronic or severe-acute hemolytic events.

On the laboratory side, substantial progresses in our ability to assess the level of HRT has to be achieved. Biochemists and industrials have to develop fully automatized and specific tests to assess pFHb as well as ways to routinely differentiate “real” free heme-containing species from total ones (bound and unbound to HptG/HpX), the first one being more relevant in terms of pathophysiology and toxicity. Indirect markers of HRT such as peroxidated lipids, like oxidized low-density lipoprotein60–62 could perhaps be also an interesting way to quantify the resultant oxidative stress.

Such improved heme-related stress appraisal seems to be fundamental to guide potential future preventive or curative treatments for patients at high risk to develop severe-acute hemolysis.


1. Schaer DJ, Buehler PW, Alayash AI, Belcher JD, Vercellotti GM. Hemolysis and free hemoglobin revisited: Exploring hemoglobin and hemin scavengers as a novel class of therapeutic proteins. Blood 2013.121: 1276–1284.
2. Smith A, McCulloh RJ. Hemopexin and haptoglobin: Allies against heme toxicity from hemoglobin not contenders. Front Physiol 2015.6: 187.
3. Minneci PC, Deans KJ, Zhi H, et al. Hemolysis-associated endothelial dysfunction mediated by accelerated NO inactivation by decompartmentalized oxyhemoglobin. J Clin Invest 2005.115: 3409–3417.
4. Rother RP, Bell L, Hillmen P, Gladwin MT. The clinical sequelae of intravascular hemolysis and extracellular plasma hemoglobin: A novel mechanism of human disease. JAMA 2005.293: 1653–1662.
5. Da Q, Teruya M, Guchhait P, Teruya J, Olson JS, Cruz MA. Free hemoglobin increases von Willebrand factor-mediated platelet adhesion in vitro: Implications for circulatory devices. Blood 2015.126: 2338–2341.
6. L’Acqua C, Hod E. New perspectives on the thrombotic complications of haemolysis. Br J Haematol 2015.168: 175–185.
7. Wurzinger LJ, Blasberg P, Schmid-Schönbein H. Towards a concept of thrombosis in accelerated flow: Rheology, fluid dynamics, and biochemistry. Biorheology 1985.22: 437–450.
8. Peacock-Young B, Macrae FL, Newton DJ, Hill A, Ariëns RAS. The prothrombotic state in paroxysmal nocturnal hemoglobinuria: A multifaceted source. Haematologica 2018.103: 9–17.
9. Wun T, Brunson A. Sickle cell disease: An inherited thrombophilia. Hematology Am Soc Hematol Educ Program 2016.2016: 640–647.
10. Nath K, Murali N. Greenberg A. Myoglobinuric and hemoglobinuric acute kidney injury Primer on Kidney Diseases. 2009, pp. Philadelphia, PA, Elsevier Health Sciences, 298–304.
11. Tracz MJ, Alam J, Nath KA. Physiology and pathophysiology of heme: Implications for kidney disease. J Am Soc Nephrol 2007.18: 414–420.
12. Gburek J, Verroust PJ, Willnow TE, et al. Megalin and cubilin are endocytic receptors involved in renal clearance of hemoglobin. J Am Soc Nephrol 2002.13: 423–430.
13. Deuel JW, Schaer CA, Boretti FS, et al. Hemoglobinuria-related acute kidney injury is driven by intrarenal oxidative reactions triggering a heme toxicity response. Cell Death Dis 2016.7: e2064.
14. Betrus C, Remenapp R, Charpie J, et al. Enhanced hemolysis in pediatric patients requiring extracorporeal membrane oxygenation and continuous renal replacement therapy. Ann Thorac Cardiovasc Surg 2007.13: 378–383.
15. Gbadegesin R, Zhao S, Charpie J, Brophy PD, Smoyer WE, Lin JJ. Significance of hemolysis on extracorporeal life support after cardiac surgery in children. Pediatr Nephrol 2009.24: 589–595.
16. Lou S, MacLaren G, Best D, Delzoppo C, Butt W. Hemolysis in pediatric patients receiving centrifugal-pump extracorporeal membrane oxygenation: Prevalence, risk factors, and outcomes. Crit Care Med 2014.42: 1213–1220.
17. Omar HR, Mirsaeidi M, Socias S, et al. Plasma free hemoglobin is an independent predictor of mortality among patients on extracorporeal membrane oxygenation support. PLoS One 2015.10: e0124034.
18. Lehle K, Philipp A, Zeman F, et al. Technical-induced hemolysis in patients with respiratory failure supported with veno-venous ECMO - Prevalence and risk factors. PLoS One 2015.10: e0143527.
19. Lyu L, Long C, Hei F, et al. Plasma free hemoglobin is a predictor of acute renal failure during adult venous-arterial extracorporeal membrane oxygenation support. J Cardiothorac Vasc Anesth 2016.30: 891–895.
20. Lv L, Long C, Liu J, et al. Predictors of acute renal failure during extracorporeal membrane oxygenation in pediatric patients after cardiac surgery. Artif Organs 2016.40: E79–E83.
21. Pan KC, McKenzie DP, Pellegrino V, Murphy D, Butt W. The meaning of a high plasma free haemoglobin: Retrospective review of the prevalence of haemolysis and circuit thrombosis in an adult ECMO centre over 5 years. Perfusion 2016.31: 223–231.
22. Lehle K, Lubnow M, Philipp A, et al. Prevalence of hemolysis and metabolic acidosis in patients with circulatory failure supported with extracorporeal life support: A marker for survival? Eur J Heart Fail 2017.19(suppl 2): 110–116.
23. Dalton HJ, Cashen K, Reeder RW, et al.; Eunice Kennedy Shriver National Institute of Child Health and Human Development Collaborative Pediatric Critical Care Research Network (CPCCRN): Hemolysis during pediatric extracorporeal membrane oxygenation: Associations with circuitry, complications, and mortality. Pediatr Crit Care Med 2018.19: 1067–1076.
24. Okochi S, Cheung EW, Barton S, et al. An analysis of risk factors for hemolysis in children on extracorporeal membrane oxygenation. Pediatr Crit Care Med 2018.19: 1059–1066.
25. Borasino S, Kalra Y, Elam AR, et al. Impact of hemolysis on acute kidney injury and mortality in children supported with cardiac extracorporeal membrane oxygenation. J Extra Corpor Technol 2018.50: 217–224.
26. Kilburn DJ, Shekar K, Fraser JF. The complex relationship of extracorporeal membrane oxygenation and acute kidney injury: Causation or association? Biomed Res Int 2016.2016: 1094296.
27. Lequier L, Horton SB, McMullan DM, Bartlett RH. Extracorporeal membrane oxygenation circuitry. Pediatr Crit Care Med 2013.14(5 Suppl 1): S7–S12.
28. Toomasian JM, Bartlett RH. Hemolysis and ECMO pumps in the 21st century. Perfusion 2011.26: 5–6.
29. Vercaemst L. Hemolysis in cardiac surgery patients undergoing cardiopulmonary bypass: A review in search of a treatment algorithm. J Extra Corpor Technol 2008.40: 257–267.
30. Lubnow M, Philipp A, Foltan M, et al. Technical complications during veno-venous extracorporeal membrane oxygenation and their relevance predicting a system-exchange–retrospective analysis of 265 cases. PLoS One 2014.9: e112316.
31. Neal JR, Quintana E, Pike RB, Hoyer JD, Joyce LD, Schears G. Using daily plasma-free hemoglobin levels for diagnosis of critical pump thrombus in patients undergoing ECMO or VAD support. J Extra Corpor Technol 2015.47: 103–108.
32. Chambers SD, Ceccio SL, Annich GA, Bartlett RH. Extreme negative pressure does not cause erythrocyte damage in flowing blood. ASAIO J 1999.45: 431–435.
33. Chambers SD, Laberteaux KR, Merz SI, Montoya JP, Bartlett RH. Effects of static pressure on red blood cells on removal of the air interface. ASAIO J 1996.42: 947–950.
34. Pohlmann JR, Toomasian JM, Hampton CE, Cook KE, Annich GM, Bartlett RH. The relationships between air exposure, negative pressure, and hemolysis. ASAIO J 2009.55: 469–473.
35. Kai Man C, Koon Ngai L. Endogenous carbon monoxide production in extracorporeal membrane oxygenation-related hemolysis: Potential use of point-of-care CO-oximetry carboxyhemoglobin to detect hemolysis. Clin Case Rep 2018.6: 346–349.
36. Hermans G, Wilmer A, Knockaert D, Meyns B. Endogenous carbon monoxide production: A rare and detrimental complication of extracorporeal membrane oxygenation. ASAIO J 2008.54: 633–635.
37. Tidmarsh GF, Wong RJ, Stevenson DK. End-tidal carbon monoxide and hemolysis. J Perinatol 2014.34: 577–581.
38. Fairbanks VF, Ziesmer SC, O’Brien PC. Methods for measuring plasma hemoglobin in micromolar concentration compared. Clin Chem 1992.38: 132–140.
39. Extracorporeal Life Support Organization: ELSO Guidelines for Cardiopulmonary Extracorporeal Life Support. 2017.Ann Arbor, MI, Extracorporeal Life Support Organization.
40. Clegg JW, King EJ. Estimation of haemoglobin by the alkaline haematin method. Br Med J 1942.2: 329–333.
41. Zander R, Lang W, Wolf HU. Alkaline haematin D-575, a new tool for the determination of haemoglobin as an alternative to the cyanhaemiglobin method. I. Description of the method. Clin Chim Acta 1984.136: 83–93.
42. Moore GL, Ledford ME, Merydith A. A micromodification of the Drabkin hemoglobin assay for measuring plasma hemoglobin in the range of 5 to 2000 mg/dl. Biochem Med 1981.26: 167–173.
43. Standefer JC, Vanderjagt D. Use of tetramethylbenzidine in plasma hemoglobin assay. Clin Chem 1977.23: 749–751.
44. Oh JY, Hamm J, Xu X, et al. Absorbance and redox based approaches for measuring free heme and free hemoglobin in biological matrices. Redox Biol 2016.9: 167–177.
45. Hayes D Jr, McConnell PI, Preston TJ, Nicol KK. Hyperbilirubinemia complicating plasma-free hemoglobin and antifactor Xa level monitoring on venovenous extracorporeal membrane oxygenation. World J Pediatr Congenit Heart Surg 2014.5: 345–347.
46. Kroll MH, Elin RJ. Interference with clinical laboratory analyses. Clin Chem 1994.40(11 pt 1): 1996–2005.
47. Venado A, Wille K, Belott SC, Diaz-Guzman E. Unexplained hemolysis in patients undergoing ECMO: Beware of hypertriglyceridemia. Perfusion 2015.30: 465–468.
48. Ismail A, Shingler W, Seneviratne J, Burrows G. In vitro and in vivo haemolysis and potassium measurement. BMJ 2005.330: 949.
49. Lippi G, Plebani M, Di Somma S, Cervellin G. Hemolyzed specimens: A major challenge for emergency departments and clinical laboratories. Crit Rev Clin Lab Sci 2011.48: 143–153.
50. Kostousov V, Nguyen K, Hundalani SG, Teruya J. The influence of free hemoglobin and bilirubin on heparin monitoring by activated partial thromboplastin time and anti-Xa assay. Arch Pathol Lab Med 2014.138: 1503–1506.
51. Laga AC, Cheves TA, Sweeney JD. The effect of specimen hemolysis on coagulation test results. Am J Clin Pathol 2006.126: 748–755.
52. Lippi G, Montagnana M, Salvagno GL, Guidi GC. Interference of blood cell lysis on routine coagulation testing. Arch Pathol Lab Med 2006.130: 181–184.
53. Graw JA, Mayeur C, Rosales I, et al. Haptoglobin or hemopexin therapy prevents acute adverse effects of resuscitation after prolonged storage of red cells. Circulation 2016.134: 945–960.
54. Lipiski M, Deuel JW, Baek JH, Engelsberger WR, Buehler PW, Schaer DJ. Human Hp1-1 and Hp2-2 phenotype-specific haptoglobin therapeutics are both effective in vitro and in guinea pigs to attenuate hemoglobin toxicity. Antioxid Redox Signal 2013.19: 1619–1633.
55. Hashimoto K, Nomura K, Nakano M, Sasaki T, Kurosawa H. Pharmacological intervention for renal protection during cardiopulmonary bypass. Heart Vessels 1993.8: 203–210.
56. Kubota K, Egi M, Mizobuchi S. Haptoglobin administration in cardiovascular surgery patients: Its association with the risk of postoperative acute kidney injury. Anesth Analg 2017.124: 1771–1776.
57. Quimby KR, Hambleton IR, Landis RC. Intravenous infusion of haptoglobin for the prevention of adverse clinical outcome in Sickle Cell Disease. Med Hypotheses 2015.85: 424–432.
58. Cucchiari D, Reverter E, Blasco M, et al. High cut-off membrane for in-vivo dialysis of free plasma hemoglobin in a patient with massive hemolysis. BMC Nephrol 2018.19: 250.
59. Hulko M, Kunz M, Yildirim M, Homeyer S, Amon O, Krause B. Cell-free plasma hemoglobin removal by dialyzers with various permeability profiles. Sci Rep 2015.5: 16367.
60. Grinshtein N, Bamm VV, Tsemakhovich VA, Shaklai N. Mechanism of low-density lipoprotein oxidation by hemoglobin-derived iron. Biochemistry 2003.42: 6977–6985.
61. Higdon AN, Benavides GA, Chacko BK, et al. Hemin causes mitochondrial dysfunction in endothelial cells through promoting lipid peroxidation: The protective role of autophagy. Am J Physiol Heart Circ Physiol 2012.302: H1394–H1409.
62. Jeney V, Balla J, Yachie A, et al. Pro-oxidant and cytotoxic effects of circulating heme. Blood 2002.100: 879–887.

extracorporeal membrane oxygenation; hemolysis; plasma free hemoglobin; haptoglobin; heme-related toxicity

Copyright © 2019 by the ASAIO