After initiating cardiopulmonary bypass for a routine CABG you notice that the pre-membrane pressure is rising by 35 mmHg per minute although the arterial flow through your roller pump is constant at 4.5 L/min. After 15 mins the pre-membrane pressure is 700 mmHg and is still rising. The patient's radial pressure is 50 mmHg and the line pressure is 200 mmHg. The pre-bypass ACT was 780 seconds. No anti-fibrinolytics have been administered. The patient has not been on pre-operative heparin. The patient had aspirin ceased one week previously. The patient's nasopharyngeal temperature is 31.8 C and the arterial blood temperature is 28 C.

What is happening?
What are my options at this point?
How high is too high?
Could I have predicted this event?
How could I have avoided this event?

Discussion

Pathologic fibrin formation in the heat exchanger and/or membrane despite adequate anticoagulation as determined by ACT constitutes one of the few unpredictable and potentially serious incidents requiring oxygenator change-out during cardiopulmonary bypass. The first case report of trans membrane pressure excursions was in 1990. The incidence of severe fibrin deposition appears to be in the order of 1:800 cases where prophylactic measures are not undertaken. Lesser degrees of fibrin deposition are quite common and go largely unnoticed. Inlet pressures greater than 500 mmHg are considered to be abnormal.

Analysis of the cryoprecipitates in affected heat exchangers/oxygenators show a complex of platelets, factor VII, von Willebrand factor, fibronectin, fibrinogen and factor XIII. It is postulated that cryofibrinogen precipitates first and this reaction contains a large amount of fibrinogen which provides a template for platelet adhesion. Hypothermia increases intrinsic platelet reactivity and greater fibrinogen binding in vitro. The fibrin polymer formed in these cases has a much lower gel porosity as the fibrin forms a much tighter network. It has been noted that the abrupt cessation of aspirin type compounds will lead to a period where patients form this abnormal fibrin gel.

Two changes in the design of oxygenators may have contributed to reports of high pre-membrane pressures over the last ten years. The first hollow fibre membrane oxygenators had intra- capillary flow of blood with the gas passing outside of the fibres, later models had an extra- capillary flow of blood with bundle packaging designed to increase the stirring effect of the blood and hence created turbulent flow. This redesign improved the gas exchange performance about fourfold over the intra-capillary type and allowed a reduction in total area required for gas exchange and hence a concomitant reduction in priming volume. Highly efficient uncoated stainless steel and plastic heat exchangers were added allowing a reduction in cross sectional areas and blood path route. These design changes set up a scenario whereby a further obstruction to flow along this route can cause a critical increase in resistance and shunting in the oxygenator. Most manufacturers do not recommend exposing the membrane to pressures greater than 700 mmHg. In severe cases these pressure excursions and loss of gas exchange area can lead to a failure of oxygenation.

The choice of an oxygenator/heat exchanger with the least restrictive blood path may be enough to avoid critical pressure excursions or frank failure should this phenomena occur. Oxygenators manufactured by Avecor, Bard, Cobe (Optima and Duo), Terumo, Medtronic, Dideco, Shiley, Sarns, Sorin, Jostra and Polystan have all demonstrated the problem. The incidence appears to increase when a priming volume of less than 300 cc is employed. The requirement for ultra low priming volumes for oxygenators appears to have unearthed a far more serious consequence.

The use of heparin bonded and polymer coatings by manufactures will markedly decrease the incidence of excursions. The platelet drop seen following the initiation of bypass is sometimes greater than a dilutional effect alone, this can be attenuated by albumin, heparin and polymer coatings.

The routine use of albumin (as little as 0.035 mg per 10Omls prime) in studies designed to examine the phenomena demonstrated that the incidence can be reduced to near zero with a preservation of the platelet count. This finding as well as the preservation of the oncotic pressure explains albumins continued widespread use. Small volumes of autologous blood to protein coat the wetted surfaces in the priming process also appears to be prophylactic.

Avoidance of large temperature gradients (> 10 C) between the heat exchanger and the blood during active cooling appears also to slow or avoid the onset of the phenomena. The practice of allowing the temperature to drift on bypass however does not eliminate the phenomena and indeed there are anecdotal reports of the phenomena occurring at normothermia. Cold reactive proteins are well described leading to morbidity in cardiac surgery. Cold agglutinins do not react with platelets and cryoglobulins only weekly so, these anti-bodies and proteins do not appear to be implicated in this phenomena as post incident laboratory examination has not turned up these proteins.

Of the known cold reactive proteins cryofibrinogen react strongly with platelets. It is known that smokers and patients with ischaemic heart disease often have higher levels of fibrinogen and higher platelet counts which may predispose this group to reactions. Streptokinase has been added to oxygenators that have demonstrated high membrane pressures and been changed out. This bench analysis has seen a rapid drop in the pressure as streptokinase cleaves the fibrin.

When pre-membrane pressure is not routinely measured hydraulic pressure separation of the tubing/connections or failure of oxygenation may be the first sign of a pre-membrane pressure excursion. These excursions regularly exceed 700 mmHg and can exceed 900 mmHg. Non- occlusive centrifugal type pumps being unable to generate the pressure required to defeat the increasing resistance will require increasing revolutions to maintain flow at the desired level. Reported critical reduction of flow in centrifugal pumps during these episodes can see an inability of the pump to flow greater than 1.5 litres/minute. This has erroneously led to a change- out of the centrifugal head during the case only to see a persistence of the problem.

Most episodes have an onset within minutes of initiating bypass and are self-limiting. The pre-membrane pressure stabilizes in under fifteen minutes at a high but acceptable level with an adequate ongoing gas exchange. Reducing flow whilst maintaining an adequate radial pressure and venous saturation will extend this conservative approach. Mild re-warming up to 33-35 C has been successfully employed to terminate the pressure rise. Vigilant monitoring of pre- membrane pressure during the onset of bypass can alert the team not to proceed with crossclamping and cardioplegia. Re-establishing ejection can allow oxygenator changeout without a period of circulatory arrest. Reports and descriptions of these events have shown that the next bypass run usually does not reproduce the same level of pressure excursion seen on the previous run. If not part of your usual prime albumin should be added if a whole new circuit is employed.

In severe cases the dilemma facing the perfusionist is whether to partially rewarm (to attempt to abort the process) or to cool further (to extend the circulatory arrest time required during oxygenator changeout should the heart be arrested).

In vivo and in vitro examinations of this phenomenon have not elucidated predictive markers. Patients who demonstrate this phenomenon are usually of a non O blood type. Patients - have higher levels factor VIII, von Willebrand factor, fibrinogen and lower D - dimmer levels.

Why has this phenomena occurred at some units but not at others. The simplistic explanation is that it requires a patient with an abnormal clotting tendency to have surgery at centers that use specific types of oxygenators without albumin prime or coated oxygenators and in whom there is a relatively large temperature gradient between the cooling circuit and the blood.

The phenomenon has been described in every oxygenator currently on the Australian market. Dr Carlo Vernier the Provisional Fellow in Cardiac Anesthesia at Liverpool in the first six months of this year surveyed every private and public hospital cardiac unit in Australia, at the time of this abstract he had replies from 80% of the units. This data is retrospective, reliant on recall and not from a database from most sites. Half the units stated that an oxygenators performance in relation to pre-membrane pressure was a factor in their purchase decision and yet only one third of units routinely monitor pre-membrane pressure. Over one third of units spread throughout Australia had experienced a pressure excursion incident some of which were significant enough to require an oxygenator changeout. Some oxygenators were thought to be particularly problematic. One quarter of sites used albumin in their prime as a specific prophylaxis against the phenomena. There was a large range of opinion as to what is a significant pressure rise and when a change-out would be considered on pressure grounds alone even when the pressure exceeded the manufacturers recommendations. As a group there was a fairly conservative attitude to what a significant pressure was (almost all were below 400 mmHg) and yet this same majority never measured the pressure.