No abstract

Since April 1988, we have been able to offer extracorporeal membrane oxygenation (ECMO) to patients who are either failing ventilation or who need cardiopulmonary support following cardiac surgery. During this time, we have supported 211 patients, the majority of whom have been supported with the Avecor (Affinity, Avecor Cardiovascular Inc., Minneapolis, MN, USA) spiral wound silicone-membrane oxygenator. Microporous hollow-fibre oxygenators have been used (18%) at our center. However, due to their early failure, they have necessitated conversion to a device that potentially would last for the entire ECMO course, i.e., Avecor and, subsequently, have not been used routinely until the introduction of the QuadroxD (Jostra Medizintechnik AG, Hirrlingen, Germany). The recent development of the poly-4-methyl-1-pentene diffusion membrane allows us to utilize hollow-fibre technology with a true (nonmicroporous) membrane. This has enabled us to use this low-resistance device with all its inherent advantages, without plasma leakage necessitating circuit changeover. From July 2000, we have used the QuadroxD oxygenator for our patients requiring ECMO support. This has been in patients with weights ranging from 2.16-51.0 kg (n = 23), with the longest support being for 1119 hours. This new technology has enabled us to utilize a single device for all patient sizes, so we now have an ECMO circuit set up at all times, minimizing the time required for support to be available, potentially improving survivor morbidity.

One challenge in providing extracorporeal circulation is to supply optimal flow while minimising adverse effects, such as haemolysis. To determine if the recent generation constrained vortex pumps with their inherent design improvements would lead to reduced red cell trauma, we undertook a study comparing three devices. Utilizing a simulated short-term ventricular assist circuit primed with whole human blood, we examined changes in plasma free haemoglobin values over a six-day period. The three pumps investigated were the Maquet Rotaflow, the Levitronix PediVAS and the Medos Deltastream DP3.This study demonstrated that all three pumps produced low levels of haemolysis and are suitable for use in a clinical environment. The Levitronix PediVAS was significantly less haemolytic than either the Rotaflow (p<0.05) or the DP3 (p<0.05). There was no significant difference in plasma free haemoglobin between the Rotaflow and the DP3 (p=0.71).

Centrifugal pumps are superior to roller pumps for extended support durations in terms of pump-induced haemolysis. In this study, we evaluated the commonly used Biomedicus BP 50 and compared it with the Jostra Rotaflow and a standard roller pump in an in vitro test circuit. Each circuit was run for a six-day period and repeated five times. Plasma haemoglobin values showed the roller pump to become more haemolytic than the Biomedicus (p = 0.022) and the Rotaflow. A statistically significant difference between the Biomedicus and the Rotaflow was observed on day six of the trial (p = 0.016), with the Rotaflow showing lower levels of haemolysis than the Biomedicus. These results support the use of the new generation centrifugal pump, the Rotaflow, as a suitable device for short-term ventricular assist.

We believe centrifugal ventricular assist is a realistic option in very small patients who require post-cardiopulmonary bypass support. It is relatively easy to establish and manage, the results, although showing no factors predictive of successful discharge, are encouraging.