This is the largest reported cohort that evaluated the practicality of PCRTO as the primary weaning strategy for V-A ECMO in the adult population. During the recruited period, 41.9% of all patients supported with V-A ECMO had weaning by PCRTO. Nearly all patients who completed a PCRTO went on to have eventual successful ECMO decannulation. We demonstrated that PCRTO was a feasible and reproducible strategy for trialing off V-A ECMO with a low risk of adverse events. Laboratory parameters that reflect the degree and recovery of end-organ ischemic injury, including serum ALT and creatinine levels, were useful for predicting a successful PCRTO.
Over the recent decades since the widespread utilization of V-A ECMO, multiple weaning strategies exist, of which decremental pump flow reduction and arteriovenous bridging recirculation technique are the mainstream approaches2–10. Nevertheless, variations in practice exist among institutions, and there are no consensus guidelines to standardize the weaning algorithms. Pump flow reduction is simple to implement, usually involving reducing ECMO flow to 33–50% of full support2, 4, 25, 26. However, as ECMO is still providing a small amount of veno-arterial flow, assessment of the native heart function may be inadequate, especially of right ventricular function in light of a reduced preload13, 27. Arteriovenous bridging recirculation technique could resemble a complete trial off of V-A ECMO support, yet it requires extra reconfiguration of the ECMO circuit by adding a parallel limb between the inflow and outflow cannula, and its utilization is limited by a short trial time and an increased risk of clot formation9, 28.
The concept of PCRTO originated in pediatric ECMO12, and has some obvious advantages over the previous two methods discussed. It mimics a complete cessation of V-A ECMO support while obviating the need to clamp off the circuit. PCRTO merely involves reducing the pump head revolutions in a controlled fashion until blood flows in a retrograde manner from the arterial to the venous cannulae. During retrograde blood flow, the systemic cardiac output is entirely taken over by native heart function, and the centrifugal pump serves as a braking system to prevent excessive arterial to venous shunting of the retrograde flow. In this physiological state, the effects of restoring normal cardiopulmonary circulation including increased RV filling, increased LV preload, and reduced LV afterload can be tested.
The feasibility of PCRTO was broadly evaluated in one of the largest reported series of adult patients supported with V-A ECMO in this study. We showed that PCRTO is highly practicable, safe, and has excellent prognostic value for ECMO weaning success. Importantly, it was able to discriminate patients who go on to be successfully decannulated. In order to routinely adopt PCRTO as a weaning strategy, ECMO caregivers need to acquire the skills in managing the distal limb reperfusion catheter to prevent clotting and the catastrophic complication of distal embolism when veno-arterial flow is reinstituted. Moreover, the process of PCRTO requires intensive monitoring to detect and manage adverse events promptly. The utility of repeated echocardiographic assessment, such as documentation of LVOT VTI during PCRTO, may differentiate weaning success and failure in patients with marginal cardiac recovery. Other hemodynamic assessment tools include insertion of the Swan-Ganz catheter have also been proposed as adjunctive techniques to facilitate determination of readiness to wean2–5, 25, 26. Future studies should aim to delineate the optimal number and duration of trials, and the feasibility of PCRTO in complex ECMO configurations such as V-AV and the presence of LV venting4, 25, 29.
The determination of readiness to wean is another controversy in ECMO weaning, and multiple studies have suggested that stabilization of end-organ function including hemodynamic profile, respiratory parameters, and liver function is essential for successful weaning2, 7, 25. Other reports advocate the use of surrogate markers such as serum lactate level and mixed venous saturation to adjudicate adequacy of oxygen delivery13. Perhaps most importantly, echocardiographic features including aortic VTI > 10 cm, lateral mitral annulus peak systolic velocity > 6 cm/s, and a LVEF > 20% have been shown to have high specificity for weaning success2, 7, 25. In our cohort, serum levels of hepatic transaminases and cardiac enzymes were satisfactory predictors of successful PCRTO, suggesting that weaning attempts should only be made after allowing adequate time for organ recovery.
This study had several inherent limitations of a retrospective cohort design. PCRTO was used entirely as the weaning approach in all recruited patients, and a matched control group was not included. Second, this was a single-center study, and subsequent generalizability of PCRTO to other ECMO configurations and care settings remains to be examined. Third, the sample size was limited in the cohort, especially in the subgroup of patients with detailed echocardiographic data, possibly precluding detection of significant differences.