Further discussion in ECMO Adult

The application of extracorporeal life support (ECLS) to adults with respiratory failure, although more recent and less well established than neonatal ECLS, is undergoing continuing development and refinement. Consequently, criteria for the selection of adult patients for ECLS are also less well established. Although the adult experience is perhaps too small to subject to critical evaluation, selection criteria applied by adult ECLS centers are fairly uniform and predict substantial mortality with conventional therapy based on historical data. This chapter will review the rationale for ECLS in adults, as well as the current experience. Selection criteria used by various centers are presented, and formulated into inclusion and exclusion criteria. These criteria should be considered dynamic; as more experience is gained, they will continue to evolve.

Traditionally, ECLS has been used in adults with severe hypoxemic acute respiratory failure (ARF) due to the acute respiratory distress syndrome (ARDS) or other causes of diffuse pneumonitis. These causes still remain the most common setting for its use today. More recently however, ECLS has been applied to hypercapnic respiratory failure due to severe acute reactive airway disease. This chapter will emphasize the use of ECLS in ARDS and other causes of hypoxemic respiratory failure. Information on its application in reactive airway disease will also be included, since this is a rapidly expanding indication.

Pathophysiology of acute respiratory distress syndrome (ARDS)

The acute respiratory distress syndrome (ARDS) describes the clinical and radiographic manifestation of acute direct or indirect pulmonary inflammatory states. ARDS occurs subsequent to a large number of initiating disease processes, both of pulmonary and extrapulmonary origin (Table 1), but particularly sepsis and trauma. The focus of infection or tissue injury establishes a systemic inflammatory response which initiates the ARDS process. Cytokines such as tumor necrosis factor and interleukin-1 and the subsequent secondary mediators are released, activating a number of cascades (coagulation, kinin, complement) as well as neutrophil aggregation and adherence in the pulmonary microcirculation. The pulmonary endothelium is injured through the release of vasoactive substances, proteases, and toxic oxygen species. This subclinical initiating phase lasts several hours, and is followed by three pathophysiologic phases with clinical manifestations [1 ].

The exudative phase consists of the progression of injury to the interstitium and alveolar cells. Inflammatory cells migrate into the interstitium and alveolar spaces. Protein-rich fluid enters the alveolar spaces with formation of hyaline membranes. The resulting interstitial and alveolar edema leads to alveolar collapse with decreased compliance and intrapulmonary shunting. This phase lasts as long as the systemic response, typically up to a week. If the systemic response if prolonged, this phase can be extended. The proliferative or regenerative phase follows, characterized by gradual reduction of the inflammatory process, includes proliferation of type II pneumocytes as part of the reparative process. The interstitium remains edematous with the continued presence of inflammatory cells. The process of fibrosis starts in this phase, which may last for several days to weeks. The fibrotic phase includes more dense deposition of collagen in interstitial spaces, and in alveolar spaces of collapsed alveoli.

Table 1. Initiating events in acute respiratory distress syndrome (ARDS)



Infectious pneumonia (bacterial, viral)

Microbial sepsis

Pulmonary contusion

Multiple trauma

Pulmonary aspiration of gastric contents

Acute pancreatitis

Near drowning

Massive blood transfusions

Table 2. Clinical and radiographic manifestations of acute lung injury (ALI) and acute respiratory distress syndrome (ARDS)



Chest radiograph

Patchy infiltrates

Extensive diffuse infiltrates

PaO2 / FiO2

100 - 200

< 100

Compliance ml/cm H2O

30 - 50

< 30

ARDS can be self limited with a rapid recovery over several days with complete resolution. In these cases, the exudative and proliferative phases are short and recovery ensues with little residual fibrosis. If the systemic process that initiates the pulmonary response is severe and prolonged, or continues unabated, then a more marked pulmonary response ensues with a more prominent fibrotic phase leading to severe and largely irreversible lung injury.

Clinical manifestations

Acute lung injury is defined in terms of its clinical and radiographic findings. ARDS is the more severe form of the spectrum of acute lung injury, and is characterized by more severe derangements in oxygenation, lung compliance, and radiographic changes (Table 2). These changes necessitate mechanical ventilation in order to maintain adequate gas exchange.

Effects of mechanical ventilation

Traditional concepts of barotrauma were based on the radiographic appearance of air in extrapulmonary sites, such as pneumomediastium, pulmonary interstitial emphysema, and pneumothorax). These complications are presumed to be due to excessive pressures generated in the airways resulting in airway rupture and dissection of air into the tissue spaces.

A newer concept of barotrauma is developing, in which overdistension of alveoli can result in a diffuse lung injury which is histologically indistinguishable from the diffuse alveolar damage ascribed to other causes of acute lung injury and ARDS [2-7]. It is also becoming recognized that mechanical ventilation can convert an otherwise mild, self-limited course of ARDS into a more prolonged and severe course. Studies in animals have suggested that it is excessive alveolar overdistension (excessive volume) more directly than excessive pressure which initiates or propagates the injury. This ventilator-induced lung injury has been termed volutrauma, in order to distinguish it from the more conventional concept of barotrauma. A major premise of ECLS is that reduction in mechanical ventilator support can prevent ventilator-induced acute lung injury or attenuate its course.

Table 3. Mortality of severe acute respiratory failure in patients 12 to 65 years of age (NHLBI-NIH additional data study)










Hydrostatic edema






COPD exacerbation






Natural history of severe respiratory failure

The impetus for the development of extracorporeal support of severe acute respiratory failure in adults is the exceptionally high mortality associated with the syndrome in spite of intensive therapy. Since any new therapy should improve outcome, it is necessary to examine the outcome of severe ARF when treated with conventional therapies, i.e. mechanical ventilation.

Survival with conventional treatment

A prospective study of severe respiratory failure under conventional mechanical ventilation was conducted in conjunction with the National Heart, Lung and Blood Institute of the National Institutes of Health (NHLBI-NIH) adult ECMO trial in 1975. Severe respiratory failure was defined as intubation and mechanical ventilation for over 24 hours, and a requirement for FiO2 of .5 or greater to prevent hypoxemia. The overall mortality rate was 66%, being slightly lower for patients under the age of 65 (Table 3) [8 ]. This mortality rate has changed little over the past 20 years. A prospective trial of outcome in 1426 patients with severe respiratory failure was conducted as part of the IVOX®multicenter study from 1990 through 1993 [9]. Using the same criteria as the NHLBI-NIH study, regardless of the resulting PaO2, mortality rate was 45%. In 375 patients in whom hypoxemia (PaO2 < 60) or hypercapnia (PaCO2 > 45) was documented on at least one occasion, the mortality rate was 67%, comparable to the NIH study.

Table 4. Mortality related to organ failure (NHLBI-NIH additional data study)

Age 12-65

Age > 65





Pulmonary failure only





Pulmonary + 1 organ





Pulmonary + 2 organ





Pulmonary + 3 organs





Pulmonary + 4 organs










Factors affecting survival

The NHLBI-NIH prospective epidemiological study examined characteristics of the patients which were related to outcome. Age was related, with mortality rate of 61% for age under 65 and 81% for patients over 65 (Table 4). The presence of extrapulmonary organ failure was also related to an increased mortality. Other studies have corroborated the above findings [10-12 ].

The reasons that mortality in severe ARF has not changed significantly in 20 years are not established. The vast majority of patients with severe ARF die from their underlying disease, sepsis, or multiple organ failure rather than from hypoxemia [10]. There is a growing amount of evidence that mechanical ventilation, when used in an attempt to support gas exchange during severe ARF, can cause worsening of acute lung injury and even the development of multiple organ dysfunction (see above). Since all patients with severe ARF are supported with mechanical ventilation, it is difficult to distinguish primary from secondary pulmonary injury. Much of the evidence to support the notion that mechanical ventilation alters the natural course of severe ARF comes from animal studies. It is currently accepted, however, that prevention of secondary pulmonary injury and organ failure in severe ARDS may improve survival.

Pre-ECLS management strategies in ARDS

Although the pathogenesis of ARDS and potential pharmacologic therapies are under intensive investigation, no such therapies have yet been identified. Thus the management of patients with ARDS remains supportive, and involves several strategies aimed at reducing the risk of ventilator-induced lung injury and other secondary pulmonary complications. These strategies include limitation of pulmonary distention, enhancing alveolar recruitment, and shifting from blood gas-oriented end-points of mechanical ventilation to ventilator setting-oriented end-points. This change in orientation has introduced the strategies of alveolar recruitment, permissive hypercapnia and limitation of lung inflation through the use of pressure-controlled ventilation, inverse ratio ventilation, low-frequency ventilation or combinations of these modes.

Improvements have been made in the understanding and application of mechanical ventilation in the support of acute respiratory failure. It is now recognized that CO2 removal and oxygenation result from two distinct effects of airway pressure therapy. Removal of CO2 depends upon ventilation of functional alveolar-capillary units, while oxygenation depends upon maintaining distention of alveoli. In severe ARF, ventilation and alveolar distention are present in non-dependent areas of the lung, partly present in mid regions, and absent in dependent regions. The goals of mechanical ventilation, therefore, are to recruit alveoli to the inflated state, then ventilate these areas with a volume appropriate to the effective lung volume.

Limitation of lung inflation

The lungs of patients with ARDS are characterized by areas of aerated lung, recruitable lung, and atalectatic lung (poorly responsive to inflation pressure), with the latter occurring at the more dependent lung regions. It has been estimated that as little as 20% of lung tissue in ARDS can be easily aerated with mechanical ventilation [13 ]. The use of typical tidal volumes may therefore may overdistend and injure the well aerated lung.

Limitation of lung inflation by reducing tidal volumes from the traditional 10-15 ml/kg to 4-9 ml/kg has been proposed to reduce volutrauma. Although no controlled studies exist to support this premise, two uncontrolled studies do suggest that this approach may be improve survival in ARDS [14,15].

The use of pressure-controlled ventilation has characteristics that suggest a useful role in protection from ventilator-induced lung injury. A decrease in peak pressure and an increase in mean airway pressure are observed. It is likely that alveolar overdistension can be reduced without having to reduce mean airway pressure. Pressure controlled ventilation also results in reduced inspiratory flow rates at larger tidal volumes, which may improve gas distribution. The definitive advantages of pressure-controlled ventilation, however, have not been fully evaluated.

Alveolar recruitment

Recruitment of atalectatic alveoli has two beneficial effects. First, it decreases intrapulmonary shunting and can reduce inspired oxygen requirements. Second, it can decrease cyclic opening and closing of lung units and decrease the shear forces associated with mechanical lung injury. Methods used to recruit alveoli include pressure-controlled ventilation, and extending the inspiratory phase (inverse ratio ventilation).

Pressure-controlled (PC) ventilation results in an inspiratory waveform with a higher early flow which tapers over the inspiratory period. This results in a higher mean tidal volume for the same peak tidal volume, affording greater alveolar recruitment at no increased volume cost. Inverse ratio ventilation (PC-IRV) extends this concept further by further prolonging the inspiratory cycle, allowing for improved distribution of gas flow during inspiration in addition to further increasing the mean tidal volume without increasing the peak volume. The mechanisms involved, however, are not fully known, especially since this mode of ventilation at typical respiratory rates produces an interdependence of mean airway pressure, auto PEEP, tidal volume, and respiratory rate in which it is difficult to assess the physiologic mechanisms at play. As a result, considerable controversy exists as to the mechanism as well as the benefit achieved through its use [16].

The use of very low frequency pressure-controlled ventilation (LFPC-IRV) removes much of this interdependence. In this mode, with a respiratory rate or 5 to 6 breaths per minute, the expiratory period is sufficiently long in spite of the markedly inverted ratio to prevent the development of auto PEEP. The inspiratory period approaches 10 seconds, raising mean airway pressure close to peak airway pressure. These features allow a very low peak airway pressure (24 to 30 cm H2O) while sustaining a mean airway pressure for alveolar recruitment. Alveolar recruitment is thus largely shifted from the expiratory phase (PEEP) to the inspiratory phase, and ventilation (CO2 removal) is shifted from the inspiratory to the expiratory phase.

Permissive hypercapnia

Limitation of pulmonary distention as a means of reducing volutrauma, and the use of low ventilatory rates, frequently involves levels of alveolar ventilation which result in hypercapnia. The acceptance of hypercapnia in order to achieve the goals of reduced ventilatory support is termed permissive hypercapnia. Although experience continues to accumulate on the physiologic effects, potential benefits, and relative safety of this method, available data support its application in patients with ARDS who are not at risk.

Acute elevations in PCO2 have physiologic effects on several organ systems. Sympathetic nervous system activity is enhanced. Tachycardia, peripheral vasodilation, and increased cardiac output are typical cardiovascular responses. An increase in cerebral blood flow and volume, with a propensity for increased intracranial pressure in susceptible individuals can occur. An increase in pulmonary vascular resistance due to pulmonary vasoconstriction is a physiologic effect which can be additive to existing pulmonary hypertension related to the underlying disease.

Hypercapnia appears to be well tolerated in most patients. The problems with hypercapnia appear to be related to an increased respiratory drive and acute acidosis [17 ]. Sedation and usually neuromuscular blockers are required for management. Persistent neuromuscular blockade, peripheral neuropathy, and myopathy have been associated with the prolonged use of neuromuscular blockers in the ICU. The tolerability of acute respiratory acidosis in critically ill patients has not been well defined, although it appears to be well tolerated in most patients. Caution should be used in patients with cardiovascular instability, intracranial disease, tissue anoxia, or uncorrected metabolic acidosis until better characterization of adverse effects is achieved.

The degree of hypercapnia and acidosis which can be tolerated is likewise not well established. Although the need for correction of pH with sodium bicarbonate is unknown, Hickling et al [14] noted no adverse effects without buffering in patients in whom pH was as low as 7.0 and PCO2 was as high as 120 torr. Sodium bicarbonate may best be used selectively in whom induction of respiratory acidosis results in hemodynamic compromise or other adverse event.

Tracheal gas insufflation

The introduction of fresh gas flow at 4 to 6 l/min into the major airways via a catheter at the level of the carina can reduce the CO2 load in the anatomic deadspace of the airways, thereby improving the efficiency of tidal volume ventilation [18, 19 ]. The flow can be phasic, synchronized to the ventilatory cycle [20 ], but this is probably offers no advantage over continuous flow. The efficacy of this method is improved with smaller tidal volumes and hypercapnic states, and thus is an effective supplemental method for reducing required tidal volumes and improving CO2 clearance in patients undergoing the above strategies.


Figure 1. Suggested algorithmic approach to application of ventilator support strategies prior to ECLS. PSV - pressure support ventilation, PCV - pressure controlled ventilation, PC-IRV - pressure controlled inverse ratio ventilation, LFPC-IRV - low frequency pressure controlled inverse ratio ventilation. Tracheal insufflation may be considered as a strategy during permissive hypercapnia. Excessive limits of FiO2, pressure and volume include FiO2 > 0.6, peak inspiratory pressure > 35 cm H2O (pressure control mode), Vt > 8-9 ml·kg-1, and PEEP > 15.

A suggested algorithm for selecting a strategy for mechanical ventilator support prior to use of ECLS is given in Figure 1.

Adult ECLS experience

In attempt to reduce mortality from severe ARF, a national study of adult ECLS sponsored by the National Heart, Lung, and Blood Institute (NHLBI) of the NIH was initiated in 1975 and completed in 1979 [21 ]. Patients with severe acute respiratory failure were randomized to either conventional mechanical ventilation alone (control), or mechanical ventilation supplemented by partial venoarterial ECMO (treatment). Patients qualified for either a slow entry set of criteria, or a rapid entry set of criteria if life-threatening hypoxemia was present (Table 5). Although 300 patients were to be entered, the study was discontinued after 90 patients, with an approximately 90% mortality in each of the control and treatment groups.

Table 5. 1975 NIH National ECMO Trial entry criteria

Fast entry criteria FiO2 1.0 continuously PEEP > 5 cm H2O PaCO2 30 - 45 PaO2 < 50 torr on 3 measurements over at least 2 hours Slow entry criteria FiO2 > .6 for at least 48 hours Qs/Qt > 30% (on FiO2 1.0) PEEP > 5 cm H2O PaCO2 30 - 45

Following publication of the results of the NIH trial, clinical research in adult ECLS all but ceased. Interest in adult ECLS returned following the publication by Gattinoni et al [22 ] of an increased survival in a series of 43 patients managed with low-flow venovenous ECLS for extracorporeal CO2 removal (ECCO2R) and low-frequency positive pressure ventilation. The technique of ECCO2R is conceptually different from venoarterial ECLS used in the NIH trial, emphasizing markedly reduced mechanical ventilation for lung rest, using apneic oxygenation, and employing an extracorporeal circuit primarily for removal of carbon dioxide. Gattinoni and colleagues used the same entry criteria as the 1975 NIH trial, plus an additional criteria of total static lung compliance (TSLC) less than 30 mlcm H2O-1. The addition of a low TSLC may have selected a more severe or advanced degree of respiratory failure which could conceivably have resulted in a higher risk group than the original NIH trial. In spite of this, they reported a 49% survival.

The differences in survival between the two studies might be attributed to several factors. The use of venoarterial perfusion reduces pulmonary blood flow, which may contribute to pulmonary microthrombosis and fibrosis. Venovenous ECLS (used in the Gattinoni study) maintains full pulmonary blood flow, which may be important in lung repair. Some the NIH centers had little or no prior experience with ECLS before starting the trial. A relatively small number of patients was entered at some centers. (The neonatal ECMO experience demonstrates a much improved survival after an initial learning period at a given center.) The NIH study was performed during an influenza epidemic, and the majority (57%) of the patients entering this trial had viral or bacterial pneumonia. Eight percent of patients had pulmonary embolism and not diffuse lung disease, all of whom died. Seven percent of patients had fluid overload due to congestive heart failure, in which irreversible cardiac dysfunction may have been the major determinant of outcome. A major purpose of ECLS is lung rest, yet this concept was not uniformly applied in the NIH trial, in that ventilator settings remained high in many of the ECLS patients after being placed on bypass. The patients were subjected to an average of five days of mechanical ventilation at high inflation pressures and FiO2 of 1.0 before entering the trial. There was significant incidence of complications. Bleeding in the ECLS group of the NIH study was high, averaging 2.5 liters per day, and the required massive transfusions could have contributed to further pulmonary injury.

A number of centers have corroborated the findings of Gattinoni et al., with approximately the same survival, including Gattinoni et al. in Milan, Lennartz et al. in Marburg, Falke et al in Dusseldorf, Bindslev et al. in Stockholm, and Anderson, Bartlett et al. in Ann Arbor [23 ,24 ], and Brunet in Paris [25] (Table 6). A number of improvements have been adopted with the experience gained from these trials. The use of percutaneous cannulation has markedly decreased bleeding complications from cannulation [26 ]. The use of surface-bonded circuits, eliminating the need for systemic anticoagulation, has undergone evaluation [27 ] and is in use in some clinical centers [26].

Table 6. Experience with ECLS for pulmonary support in severe ARF













Ann Arbor









Data compiled from [22] and [24].

A randomized, controlled trial comparing computer protocol-driven mechanical ventilation (PC-IRV) with LFPPV-ECCO2R was reported by Morris et al [28 ] in 1994. There was no statistically significant difference in survival in the two groups. Survival in the control group was 42%, suggesting that a structured approach to ventilator management may improve survival. Survival in the LFPPV-ECCO2R group was 33%. Several factors may have contributed to the lower survival in the ECLS group. The experience at the center was severely limited prior to initiation of the trial. Bleeding was excessive, averaging 1760 ml/day. Lung rest was not achieved, with peak inflation pressures averaging 45 cm H2O in all treated patients, and over 52 to 55 cm H2O for the last 9 patients treated. These excessive levels may have been required because low-flow ECCO2R may not provide sufficient gas exchange for all patients undergoing ECLS.


Extracorporeal life support is indicated when the patient has an acute, potentially reversible, life-threatening form of respiratory failure which is unresponsive to conventional therapy [29]. Acute refers both to the rapidity of development of respiratory failure, and to its duration prior to consideration for ECLS. Potentially reversible indicates that the disease process causing respiratory failure can be treated or has a reasonable expectation for recovery with supportive measures. Life-threatening refers to severe respiratory failure with an unacceptably high expected mortality. Since expected survival with ECLS is at least 50%, ECLS is appropriate when expected survival is less than 20% [30 ]. Unresponsive to conventional therapy indicates that measures to improve gas exchange with aggressive mechanical are unsuccessful in achieving adequate gas exchange, or result in levels of support which place the patient at risk of ventilator-induced lung injury.

Patient selection

Selection criteria vary somewhat between adult ECLS centers worldwide. All centers are in agreement with regard to severity of respiratory failure, but differ in other inclusion and exclusion criteria. Although each center has established selection criteria, each patient is considered on a case-by-case basis, and may be included or excluded in spite of what is indicated by the established criteria. Experience plays a major role in patient selection, and cannot be characterized or quantitated.

Failure of conventional therapy

Adults with severe ARF due to diffuse parenchymal disease should fail a trial with mechanical ventilation with strategies such as PC-IRV or LFPC-IRV before being considered for ECLS. Failure is defined as persistent hypoxemia with > 30% transpulmonary shunt on an FiO2 > .6 and inability to improve compliance above 0.5 mlcm H2O-1kg-1 [24]. The trial of PC-IRV or LFPC-IRV should be attempted for 12 hours or more, since early improvement in lung function is only modest, with continued improvement over the subsequent 12 to 24 hours.

Severity of respiratory failure

There are no uniformly accepted criteria for characterizing severity of hypoxemic respiratory failure in considering extracorporeal support. Each center, however, has established criteria which are largely based on the 1974 NIH National ECMO Trial criteria. The original criteria had a fast entry arm and a slow entry arm (Table 5). The fast entry criteria allowed for as little of two hours of mechanical ventilation, while the slow entry criteria were applied after at least 12 hours. Since it may take several hours to fully assess the response to ventilator manipulation (PEEP, PC-IRV, etc.) the fast entry criteria may not be appropriate as a basis for establishing entry criteria in most patients. If hypoxemia is severe and life-threatening, however, early initiation of ECLS may be life-saving.

The NIH criteria were based primarily on the presence of severe hypoxemia, requiring a PaO2 < 50 on an FiO2 > 0.60 as well as an intrapulmonary shunt > 30%. Gattinoni et al., in their original study, added the additional criteria of reduced pulmonary compliance (< 30 ml/cm H2O with 10 ml/kg tidal volume) in order to select patients with lung injury which might be less responsive to conventional mechanical ventilation. A reduced pulmonary compliance has subsequently been widely accepted as an inclusion criterion. Bartlett et al. [31 ] define reduced compliance as < 0.5 mlcm H2O-1. Gattinoni et al. [22] have more recently introduced a lack of recruitment, defined as a lack of PaO2 response to increasing PEEP from 5 to 15 cm H2O, as a third criterion.

Etiology of respiratory failure

A challenge in selecting adults for ECLS is the determination of reversibility of the disease process. Diffuse parenchymal disease is the hallmark of acute lung injury and ARDS. A number of disease states progress to this final common pathway, which has a fairly uniform prognosis regardless of etiology or initiating event.

There are no methods to identify reversibility in acute lung injury. Experience with lung biopsy is too limited to draw conclusions or propose biopsy results as a selection criterion. Since acute lung injury of any etiology is marked by diffuse inflammation and fibrosis, patients are considered for ECLS on the basis of severity of respiratory failure and not etiology. Chronic lung disease complicated by acute lung injury deserves critical evaluation, since the chronic component diminishes chances of recovery. Therefore, existence of moderate to severe chronic lung disease is considered a contraindication to the use of ECLS.

Etiologies of acute respiratory failure supported with ECLS include ARDS of multiple etiologies, diffuse pneumonia or pneumonititis, and more recently, reactive airways disease (Table 7).

Table 7. Etiologies of adult ARF supported with ECLS

Acute respiratory distress syndrome Sepsis Aspiration pneumonitis Pancreatitis Postpartum ARDS Pneumonia Bacterial pneumonia Viral pneumonia Legionnaire's disease Reactive airways disease Status asthmaticus

Duration of respiratory failure and mechanical ventilation

The likelihood of lung recovery and survival in severe ARF appear to be related to the duration of mechanical ventilation prior to instituting ECLS. Anderson et al. [24] reported a decreasing survival with increasing duration or prior treatment. Because of poor survival of patients who received six or more days of mechanical ventilation prior to ECLS, they consider a period of six or more days a contraindication to ECLS. Results of the IVOX®trial support the relationship of duration of mechanical ventilation and outcome (Figure 2), however, a survival rate of 20% was reported for patients with 8 or more days of mechanical ventilation. Although selection criteria for the IVOX®trial were different, these patients still had severe advanced acute respiratory failure. The relationship between survival and duration of pre-ECLS mechanical ventilation supports the concept that mechanical ventilation at the levels of support required to maintain gas exchange in diseased lungs can contribute to the worsening of parenchymal lung injury.

Some centers do not consider prolonged duration of mechanical ventilation prior to ECLS as an exclusion criteria [26]. The intensity of mechanical ventilation in the pre-ECLS period is perhaps a more important determinant of outcome, and this should be considered in conjunction with the duration of mechanical ventilation. A prolonged duration (> 7 days) is therefore only a relative contraindication to ECLS.

Coexisting disease

In the NIH prospective trial, the largest influence on survival was the presence of extrapulmonary organ failure. Patients with respiratory failure alone had the best prognosis, with a mortality rate of 40%. Mortality increased substantially with increases in the number of organ failures (Table 4). Each type of organ system failure was approximately equally contributory to mortality rate. Multiple organ dysfunction must be carefully evaluated prior to ECLS. Early multiple organ dysfunction may be secondary to mechanical ventilation, is often reversible, and frequently responds to ECLS. Late or advanced multiple organ failure, however, should be considered a contraindication to ECLS.

Venovenous ECLS requires native cardiac output for oxygen delivery, therefore normal cardiac function is necessary for this form of support. Severe pulmonary hypertension frequently results in refractory right heart failure, and is a contraindication to venovenous support.

Hypercarbic respiratory failure

The role of ECLS in hypercapnic states is more recently recognized. ECLS promises to be an effective support method for severe reactive airway disease, since the disease is largely reversible, and most deaths are due to complications of therapy with mechanical ventilation. The first successful application of venoarterial ECLS in status asthmaticus was reported by MacDonnell et al. [32 ] in 1981, and of venovenous ECLS by Shapiro et al [33 ] in 1993. Since these cases, the number of applications has increased, with an exceptionally low mortality rate. ECLS should be considered as early as possible in severe reactive airway disease when conventional therapy fails, as delays can result in an increased rate of complications and a poorer outcome.

Selection criteria for including and excluding ECLS in adult patients with severe acute respiratory failure are summarized in Table 8 and Table 9, respectively.

Selection of technique

Venovenous ECLS

For most adult patients meeting the criteria for ECLS, venovenous support is the method of choice. The technique is referred to as both extracorporeal CO2 removal (ECCO2R) and venovenous ECMO (V-V ECMO). Although the terms have been used interchangeably and the extracorporeal system is largely identical, the goals of support with these two methods differ.

Extracorporeal CO2 removal emphasizes carbon dioxide removal through the use of low-flow (1 to 2 L·min-1) bypass, utilizing apneic oxygenation via the natural lungs. Near

Table 8. Selection of patients with respiratory failure for ECLS: Inclusion criteria

Hypoxemic respiratory failure Failure of mechanical ventilation (PC-IRV) to reverse hypoxemia and improve lung compliance Diffusely abnormal chest radiograph Transpulmonary shunt > 30% on FiO2 > 0.6 Total static lung compliance < 0.5 ml/cm H2O/kg/ (or < 30 ml/cm H2O at Vt 10 ml/kg) Lack of PEEP recruitment response (PEEP 5-->15 cm H2O) Hypercarbic respiratory failure Uncorrectable hypercarbia with pH < 7 and PIP > 45, or PaCO2 > 45 despite Ve > 200 mlkg-1min-1

Table 9. Selection of patients with respiratory failure for ECLS: Exclusion criteria

Absolute Contraindication to systemic anticoagulation (except surface-coated systems) Terminal disease with short expected survival Underlying moderate to severe chronic lung disease Advanced multiple organ failure syndrome Unresponsive septic shock Uncontrolled metabolic acidosis Central nervous system injury Severe immunosuppression Relative: venovenous ECLS Mechanical ventilation > 7 days Myocardial dysfunction (CI < 3.5) on inotropic therapy Severe pulmonary hypertension (MPAP > 45 or > 75% systemic) Cardiac arrest Age > 60 years Relative: venoarterial ECLS Mechanical ventilation > 7 days Irreversible or chronic cause of cardiac dysfunction Age > 60 years

Disadvantages of venoarterial ECLS include reduced pulmonary blood flow (which may negatively impact lung repair), arterial discharge of emboli (which can reach vital organs), further impairment of left ventricular function via LV volume overload, and cardiac dependence on ECLS performance (which enhances the adverse sequelae in case of bypass circuitry failure). Advantages include lack of dependence on good cardiac function to maintain oxygenation. If a patient placed on venoarterial support for cardiac and pulmonary support subsequently develops an improved and adequate cardiac function, then the patient is typically converted from venoarterial to venovenous bypass.

Arteriovenous ECLS

ECLS in the arteriovenous configuration has been proposed as a pumpless means of achieving extracorporeal circulation. Since arterial blood is largely oxygenated, little oxygen transfer could take place except in the presence of marked hypoxemia. It is not likely that arteriovenous ECLS will play a role in hypoxemic respiratory failure. Its potential for CO2 transfer is large, however, and it may have a role in hypercarbic respiratory failure without hypoxemia (reactive airways disease). Theoretical and animal studies support the concept as viable [34 ], but clinical studies have not emerged as of yet.


1. Tomashefski TF: Pulmonary pathology of the adult respiratory distress syndrome. Clinics Chest Med 1990; 11:593-619.

2. Greenfield LJ, Ebert PA, Benson DW: Effect of positive pressure ventilation on surface tension properties of lung extracts. Anesthesiology 1964; 25:312-316.

3. MacDonald GF, Dines DE, Sessler AD, Titus JL: Pulmonary abnormalities associated with prolonged oxygen therapy and mechanical ventilation. Am Rev Respir Dis 1968; 98:155-156.

4. Barsch J, Birbara C, Eggers GWN, et al: Positive pressure as a cause of respirator induced lung disease. Ann Intern Med 1970;72:810.

5. Dreyfuss D, Basset G, Soler P, Saumon G: Intermittent positive-pressure hyperventilation with high inflation pressures produces pulmonary microvascular injury in rats. Am Rev Respir Dis 1985; 132:880-884.

6. Kolobow T, Moretti MP, Fumagalli R, Mascheroni D, Prato P, Chen V, Joris M: Severe impairment in lung function induced by high peak airway pressure during mechanical ventilation. Am Rev Respir Dis 1987; 135:312-315.

7. Tsuno K, Prato P, Kolobow T: Acute lung injury from mechanical ventilation at moderately high airway pressures. J Appl Physiol 1990; 69:956-961.

8. Bartlett RH, Morris AH, Fairley HB, Hirsch R, O'Connor N, Pontoppidan H: A prospective study of acute hypoxic respiratory failure. Chest 1986; 89:684-689.

9. Vasilyev A, Schaap RN, Mortensen JD: Hospital survival rate of patients with acute respiratory failure in modern intensive care units: an international multicenter prospective survey. Chest 1994 (In press).

10. Bell RC, Coalson JJ, Smith JD, Johanson WG: Multiple organ system failure and infection in adult respiratory distress syndrome. Ann Intern Med 1983; 99:293-298.

11. Fowler AA, Hamman RF, Zerbe GO, Benson KN, Hyers TM: Adult respiratory distress syndrome: prognosis after onset. Am Rev Respir Dis 1985; 132:472-478.

12. Montgomery AB, Stager MA, Carrico J, Hudson LD: Causes of mortality in patients with adult respiratory distress syndrome. Am Rev Respir Dis 1985; 132:485-489.

13. Gattinoni L, Pesanti A, Avalli L, Rossi F, Bombino M: Pressure-volume curves of total respiratory system in acute respiratory failure. Computed tomographic scan study. Am Rev Respir Dis 1987; 136:730-736.

14. Hickling KG, Henderson SJ, Jackson R: Low mortality associated with low volume pressure limited ventilation with permissive hypercapnia in severe adult respiratory distress syndrome. Intensive Care Med 1990; 16:372-377.

15. Hickling KG, Walsh J, Henderson S, Jackson R: Low mortality rate in adult respiratory distress syndrome using low-volume pressure-limited ventilation with permissive hypercapnia: a prospective study. Crit Care Med 1994; 22:1568-1578.

16. Shanholtz C, Brower R: Should inverse ratio ventilation be used in adult respiratory distress syndrome? Am J Respir Crit Care Med 1994; 149:1354-1358.

17. Bidani A, Cardenas VJ, Tzouanakis AE, Zwischenberger JB: Permissive hypercapnia. In Bone RC (ed.): Pulmonary and Critical Care Medicine, Mosby, 1995, Ch. 21, 1-17.

18. Nahum A, Burke WC, Ravenscraft SA, et al: Lung mechanics and gas exchange during pressure control ventilation in dogs: augmentation of CO2 elimination by an intratracheal catheter. Am Rev Respir Dis 1992; 146:965-973.

19. Ravenscraft SA, Burke WC, Nahum A, Adams AB, Nakos G, Marcy TW, Marini JJ: Tracheal gas insufflation augments CO2 clearance during mechanical ventilation. Am Rev Respir Dis 1993; 148:345-351.

20. Burke WC, Nahum A, Ravenscraft SA et al: Modes of tracheal gas insufflation: comparison of continuous and phase specific gas injection in normal dogs. Am Rev Respir Dis 1993; 148:562-568.

21. Zapol WM, Snider MT, Hill JD, Fallat RJ, Bartlett RH, Edmunds LH, Morris AH, Peirce EC, Thomas AN, Proctor HJ, Drinker PA, Pratt PC, Bagneiwski A, Miller RG: Extracorporeal membrane oxygenation in severe acute respiratory failure. JAMA 1979; 242:2193-2196.

22. Gattinoni L, Pesenti A, Mascheroni D, Marcolin R, Fumagalli R, Rossi F, Iapichino G, Romagnoli G, Uziel L, Agostoni A, Kolobow T, Damia G: Low-frequency positive-pressure ventilation with extracorporeal CO2 removal in severe acute respiratory failure. JAMA 1986; 256:881-886.

23. Anderson HL, Delius RE, Sinard JM, McCurry KR, Shanley CJ, Chapman RA, Shapiro MB, Rodriguez JL, Bartlett RH: Early experience with adult extracorporeal membrane oxygenation in the modern era. Ann Thorac Surg 1992; 53:553-563.

24. Anderson H, Steimle C, Shapiro M, Delius R, Chapman R, Hirschl R, Bartlett RH: Extracorporeal life support for adult cardiorespiratory failure. Surgery 1993; 114:161-173.

25. Brunet F, Belghith M, Mira J-P, Lanore JJ, Vaxelaire JF, Santucci JD, Dhainaut JF: Extracorporeal carbon dioxide removal and low-frequency positive-pressure ventilation. Chest 1993; 104:889-898.

26. Gattinoni L, Pesenti A, Bombino M, Pelosi P, Brazzi L: Role of extracorporeal circulation in adult respiratory distress syndrome. New Horizons 1993; 1:603-612.

27. Knoch M, Köllen B, Dietrich G, Müller E, Mottaghy K, Lennartz H: Progress in veno-venous long-term bypass techniques for the treatment of ARDS. Int J Artific Organs 1992; 15:103-108.

28. Morris AH, Wallace CJ, Menlove RL et al: Randomized clinical trial of pressure-controlled inverse ration ventilation and extracorporeal CO2 removal for adult respiratory distress syndrome. Am J Respir Crit Care Med 1994; 149:295-305.

29. Bartlett RH: Extracorporeal life support for cardiopulmonary failure. Curr Probl Surg 1990; 27(10):623-705

30. Bartlett RH: Extracorporeal membrane oxygenation revisited, revisited. Ann Thorac Surg 1992; 738-742.

31. Chapman RA, Bartlett RH: Extracorporeal Life Support Manual. University of Michigin Medical Center, 1991.

32. MacDonnell KF, Moon HS, Sekar TS, Ahluwalia MP: Extracorporeal membrane oxygenator support in a case of severe status asthmaticus. Ann Thorac Surg 1981; 171-175.

33. Shapiro MB, Kleaveland AC, Bartlett RH: Extracorporeal life support for status asthmaticus. Chest 1993; 103:1651-54.

34. Grier LR, Blount J, Conrad SA, Brown EG, Mortensen JD: Extracorporeal CO2 removal via a pumpless arteriovenous hollow fiber membrane system. Crit Care Med 1992;20 (Suppl):S60.

35. Conrad SA, Bagley A, Bagley B, Schaap RN: Major findings from the clinical trials of the intravascular oxygenator (IVOX®). Artificial Organs 1994; 18:846-863.

back to top