Trial off venoarterial extracorporeal membrane oxygenation (VA-ECMO) is the process used to assess whether the patient can sustain lung or cardiac function (or both) without mechanical support. As the native lung or heart function improves, ECMO trial off can be considered.
The procedure commonly used consists of a sudden separation of the circuit from the patient by clamping both cannulae, the interposition of an arteriovenous (AV) bridge (if not already installed) to allow ECMO circuit recirculation, and the maintenance of anticoagulation to avoid clotting. This process introduces areas of stagnant blood within the circuit with the subsequent risk of clot formation. Clot formation can hamper the reinstitution of ECMO, should the trial off prove unsuccessful and risks propagation of clot into the patient.1
Pump Controlled Retrograde Trial Off (PCRTO) relies on the retrograde flow of blood through the ECMO circuit to maintain circuit integrity and allows a longer trial off from VA-ECMO support without circuit clot formation or significant patient hemodynamic compromise. This technique avoids the insertion of an AV bridge and the need to clamp the circuit. We present data on five neonatal patients who were trialled off using this method.
In infants, VA-ECMO is established by neck cannulation (using the right common carotid artery and right internal jugular vein). The cannulae are connected to the circuit, which (in our center) comprises a Thoratec Pedi-VAS centrifugal pump and a Medos Hi-lite 800 LT oxygenator.
In PCRTO, appropriate mechanical ventilation is established based on the lung mechanics of the child. The sweep gas is disconnected and the trial off is commenced.
At the start of the trial off, the flow probe is reversed and the pump revolutions per minute (RPM) is reduced. As the pressure generated by the pump falls below the systemic arterial pressure, a state of retrograde flow is induced. The RPM is further reduced until a retrograde flow of approximately 120 ml/min is achieved. The pump is acting as a brake to stop excessive retrograde flow with systemic steal of blood and the risk of increased pulmonary blood flow. Activated clotting time (ACT) is kept between 200 and 220 seconds.
Table 1 shows hourly ECMO circuit and hemodynamic data for five neonates trialled off VA-ECMO using our PCRTO technique. Table 2 shows demographic and ECMO run data. Figure 1 shows the stability in mean arterial pressure in patients during PCRTO. There were no circuit complications, including clot formation during any of the PCRTOs (Table 1). Patients 2 and 4 were able to be decannulated after 2 hours of PCRTO, patient 3 after 3 hours, and patients 1 and 5 after 5 hours.
During PCRTO in patient 1, the serum lactate increased from 1.6 (on ECMO) to 2.1, 3.8 after 1 and 2 hours, respectively, peaking at 3.8 after 4 hours. Mean arterial blood pressure was stable throughout. Patients 2, 3, 4, and 5 showed stable hemodynamics and a drop in lactate levels when the retrograde trial off was commenced.
Trialling off ECMO or any assist device is always very stressful for the patient because of the sudden increase in the cardiac preload and reduced afterload. The traditional trial off VA-ECMO requires an increase in heparinization of the patient and circuit, the weaning of the flow down to 50% (the minimum tolerated due to increased risk of clot formation), and clamping of the cannulae. An AV bridge is inserted into the circuit to allow blood to continue to circulate around the circuit and to prevent clotting. The cannulae are flushed every 15 minutes to avoid the stagnant blood from clotting.
In our experience, despite the further heparinization, the presence of the AV bridge to allow the recirculation of the ECMO and the routine flushing of the cannulae, the circuit still has a maximum “integrity” time of 2 hours before starting to clot, and in some patients despite apparent adequate anticoagulation (as measured by ACT), the circuit clots very quickly. By removing the bridge completely and relying on retrograde flow to maintain circuit integrity, a long trial off from ECMO support can be achieved without significant hemodynamic compromise and reduced risk of clot propagation.
Potential limitations of this technique are that there is a potential risk of microemboli in the ECMO circuit (normally filtered by the oxygenator) reaching the patient in state of retrograde flow or hemolysis caused by red blood cells flowing against the direction of the centrifugal pump. We have not seen evidence of this in the patients described here or in subsequent patients trialled off in this way. If there were to be microemoboli, these would enter the patient on the venous side in retrograde flow and would be of little clinical significance in the majority of patients. Some patients do not tolerate the afterload reduction and require vasopressors for the duration of the trail off to maintain adequate perfusion pressure or can only be trailed off using the classical clamping off technique.
Our experience is based solely on nonocclusive centrifugal blood pump systems, but theoretically it would be possible to use this technique in ECMO systems using occlusive roller pumps by walking the raceway backward into the pump head. Potentially, flow would be more constant and less dependent on patient blood pressure than in the centrifugal systems. This technique of PCRTO in VA-ECMO has become standard practice in our center.
1. Van Meurs K, Lally KP, Peek G, Zwischenberger JB ECMO (the ‘Red Book’).3rd ed. Ann Arbor, MI: Extracorporeal Life Support Organization;2005