To the present time, the main indication for ECCO2R is severe ARDS secondary to a potentially reversible cause.74,75 Generally accepted contraindications to extracorporeal life support and ECCO2R include the presence of significant bleeding, surgery in the preceding 72 hours, severe brain damage, uncontrolled severe sepsis, unresolved malignancies, severe chronic systemic disease, and ARDS of a known irreversible origin.
In 1979, Gattinoni et al2 reported the use of LFPPV-ECCO2R in an adult patient with ARDS. The technique involved venovenous bypass: The common femoral and jugular veins were cannulated both distally and centrally through surgical cutdowns. The wounds and the multiple cannulation involved continuous oozing of blood and limitations in nursing care and patient mobility.76 Subsequently, we developed a double-lumen cannulation of the femoral vein that allowed a single cutdown.77 With saphenosaphenous bypass,78 surgery became very superficial and distal drainage was unnecessary. The number of cannulas therefore was reduced to two. The most significant improvement in cannulation, however, came with the springwire-reinforced percutaneous cannulas,9,79 which are placed by a modified Seldinger technique,80 with a shorter procedure time, practically no bleeding, a reduced risk of cannulation-site infection, and very simple decannulation.
Blood flow is normally kept at 15% to 30% of the patient’s cardiac output. The system must have a capability of running at 50 to 60 mL/kg/min should we need to substitute for the natural lung oxygenation (Fig. 21-5).
LFPPV-ECCO2R circuit. (Reproduced, with permission, from Gattinoni, et al. JAMA. 1986;256:881–886,. Copyright © 1986 American Medical Association. All rights reserved.)
Our current standard includes the use of two springwire-reinforced percutaneous femoral cannulas (20F to 28F), a centrifugal pump, and a plasma tight hollow-fiber polymethylpentene oxygenator. The entire circuit is surface heparinized to minimize the need for systemic anticoagulation.81–83
Following an initial 50 to 100 IU/kg intravenous heparin bolus at the time of cannulation, heparin infusion is started, aiming at the selected activated clotting time (150 to 200 seconds in the case of Jostra Bioline surfaces). Surface-heparinized circuits can even be run without any systemic anticoagulation84 for at least 12 to 48 hours as needed to stop or prevent incidental bleeding. Antithrombin III activity is maintained around 100% to promote surface-bonded and intravenous heparin function. Platelets are transfused when lower than 50,000/μL. When heparinized surfaces are not in use, activated clotting time and/or partial thromboplastin times of 1.5 to 2 times normal must be maintained at all times.
ECCO2R normally is started in a sedated, paralyzed patient. After the initial adjustments (which normally take 1 to 2 hours), the ventilator is set to provide a low-frequency sigh over a baseline constant PEEP (e.g., using intermittent mandatory ventilation or biphasic positive airway pressure). PEEP is adjusted to maintain mean airway pressure at the prebypass level and to prevent acute worsening of lung edema. A catheter inserted into the inspiratory line provides a constant oxygen supply and constant PEEP during the long expiratory pause. As soon as possible, attempts are made to reestablish spontaneous respiratory activity, most often in the form of pressure supported breathing with an intermittent sigh (e.g., biphasic positive airway pressure plus assisted breathing or intermittent mandatory ventilation plus pressure support).
Hemodynamics are not affected by the venovenous bypass. Changes in lung function can be followed with venous admixture measurements. Very high values of mixed venous saturation and arterial O2 saturation suggest decreased cardiac output with an increased proportion of extracorporeal blood flow/total cardiac output. When lung function improves, weaning is attempted by decreasing and PEEP and by decreasing the CO2 removal from the membrane lung.40 When necessary, the extracorporeal circuit setup allows low-flow venovenous ECCO2R to be converted into high-flow venovenous ECMO, and management then is focused on achieving a viable oxygenation despite an extremely reduced or even absent natural-lung oxygen transfer.
Complications of Venovenous ECCO2R
We have never stopped ECCO2R because of a technical accident. From day to day, however, various changes of circuit elements may be required, mainly involving the membrane lung and/or the centrifugal pump(s).
Bleeding always has been the major complication with extracorporeal life-support techniques. The NIH-ECMO study reported an average blood-product transfusion of 3575 mL/day.85 In 1986, we reported86 an average blood-product requirement of 1.8 L/day in our first forty-three patients. The use of percutaneous cannulation coupled with heparinized surfaces decreased the packed-red-cell requirement to 200 to 300 mL/day.75 Bleeding from the chest drainage tubes is a major complication that often demands a surgical approach.87 The major threat posed by extracorporeal support remains intracranial hemorrhage. We reported that a prebypass PaCO2 of greater than 75 mm Hg, disseminated intravascular coagulation, or a positive brain computed tomographic scan before bypass increase the risk of fatal intracranial hemorrhage during bypass.75
Clinical Results in Acute Respiratory Distress Syndrome Patients
It is difficult to isolate from the literature the experience with venovenous ECCO2R because of undefined boundaries between venovenous ECCO2R, venovenous ECMO, and partial extracorporeal CO2 removal (PECO2R).40 ECCO2R refers more to a way of managing the diseased lung rather than to technical details, and the flexible handling of the circuitry is one of the advantages of ECCO2R. The technique can shuttle back and forth between venovenous ECCO2R, venovenous ECMO, and PECO2R in the same patient.
Table 21-3 is an attempt to collect the available experience on ECCO2R, defined as such by authors, including our own data (see also Table 21-4). At the time of writing, patients with severe ARDS are still starting on ECCO2R later and later in their illness. Whether this is wise, we have our doubts. Moreover, the time spent on bypass gets increasingly longer (139 days for our longest survival run), probably indicating increases in unmeasured elements of disease severity at the time of connection. Despite these considerations, many patients with ARDS suffer from complications related to ventilator-induced lung injury and are doomed to an unfavorable outcome. ARDS carries a substantial mortality, 31% to 39.8% in the ARDS Network studies.66
Table 21-3: Venovenous ECCO2R for ARDS: International Experience |Favorite Table|Download (.pdf)
Table 21-3: Venovenous ECCO2R for ARDS: International Experience
|Author (Ref.)||Year||Center||No. Patients||Survivors||% Survival|
|Wagner (87)||1990||Marburg (Germany)||76||38||50%|
|Bindslev (82)||1991||Stockolm (Sweden)||14||6||43%|
|Brunet (121)||1993||Paris (France)||23||12||52%|
|Morris (88)||1994||Salt Lake City (USA)||21||7||33%|
|Guinard (122)||1997||Paris (France)||10||4||40%|
|Gattinoni, Pesenti||2008||Milan, Monza (Italy)||124||49||40%|
Table 21-4: Outcome with ECCO2R in Milan-Monza (1979 to 2008; Adult Patients) |Favorite Table|Download (.pdf)
Table 21-4: Outcome with ECCO2R in Milan-Monza (1979 to 2008; Adult Patients)
|No. Patients||49 (40%)||75 (60%)|
|Age||33.9 ± 14.4||34.6 ± 15.2||NS|
|Days from intubation||11.9 ± 13.5||11.6 ± 9.1||NS|
|PaO2/FIO2||92.9 ± 44.6||76.3 ± 40.2||0.0353|
|Qs/Qt||0.46 ± 0.12||0.51 ± 0.10||0.0179|
|PaCO2 mm Hg||54.8 ± 17.2||63.9 ± 20.2||0.0106|
|PEEP cm H2O||11.8 ± 5.5||13.2 ± 4.2||NS|
|PIP cm H2O||41.8 ± 8.0||45.9 ± 10.5||0.0279|
|Cardiac index (L/min)||5.3 ± 1.6||4.9 ± 1.3||NS|
|Heart rate (bpm)||132 ± 19||125 ± 20||NS|
|BP mm Hg||84.5 ± 14.1||80.9 ± 13.7||NS|
|CVP cm H2O||11.1 ± 5.7||11.3 ± 5.1||NS|
|PAP mm Hg||33.1 ± 8.9||35.3 ± 7.9||NS|
|WP mm Hg||13.1 ± 6.5||13.9 ± 5.0||NS|
|On vasopressor||6 (12%)||25 (32%)||0.0174|
|Days of ECCO2R||15.7 ± 21.9||15.7 ± 16.3||NS|
Only one controlled, randomized trial has been conducted on the effect of LFPPV-ECCO2R in patients with severe ARDS.88 The investigators enrolled forty patients meeting the original NIH-ECMO entry criteria. Nineteen patients were randomized to LFPPV-ECCO2R and twenty-one to control mechanical ventilation. Survival was equivalent in the two groups: seven survivors in the ECCO2R and eight in the control group. The investigators and accompanying editorialists89 concluded that extracorporeal support is not recommended in ARDS.
A more balanced interpretation of this study, however, must take into account several considerations.90–92 The incidence of uncontrollable bleeding, leading to premature interruption of the treatment in seven of nineteen patients, was extremely high. More surprisingly, of seven ECCO2R survivors, five had been disconnected as an emergency because of severe bleeding. Average transfused blood products (packed red cells plus fresh-frozen plasma) was 3.39 L/day (4.79 L/day in the survivors). A problem in the management of blood clotting or the surgical procedure to control bleeding appears obvious when related to contemporary published experience. This study suggests that despite a high rate of catastrophic bleeding in the ECCO2R group, the net outcome was not worse in the ECCO2R group. As such, ECCO2R may prove beneficial, provided that the associated bleeding problems are handled effectively.