MATERIAL MATTERS COLUMN
Haemolysis and medical devices
Three recent events relating to the biocompatibility of materials and implantable medical devices combine to raise some questions over the relevance of some tests used to determine biological safety. I refer to these three events without specifying sources because I want to draw attention to the principles rather than highlight specific cases. The first event was the publication by regulatory bodies of a small number of cases of haemolysis identified in patients undergoing haemodialysis. The haemolysis was caused by kinking in some of the tubing and at least one patient died. Haemolysis is the phenomenon whereby red blood cells are damaged or age prematurely so that their ability to transport haemoglobin is reduced, which leads to haemolytic anaemia. There are, of course, many other causes of haemolytic anaemia, including the genetic conditions of sickle cell anaemia and thalassaemia, acquired conditions related to autoimmunity, and cases induced by infection and drugs. It has long been known that some medical devices could be responsible for damage to red blood cells and some levels of anaemia are sometimes seen with prosthetic heart valves and extracorporeal circulatory systems.
This is, therefore, not a new problem and obviously it is highly unfortunate that injuries are sustained through the use of the haemodialysis equipment mentioned above. The issue I wish to address here is how this damage to red cells occurs and how damage to these cells relates to the testing of devices and materials. The most relevant factor is that the damage to red blood cells caused by medical devices is primarily the result of mechanical interaction between blood and the device. These cells are not accustomed to interacting with synthetic materials, plastic tubing and metal valves; they are far more used to the softer endothelial lining of blood vessels. Thus, it should be no surprise that they may, under some conditions, interact unfavourably with harder, rougher or irregularly shaped devices. Quite a lot is known about the relationship between the haemodynamic properties of devices and risk of haemolysis, and intravascular or extracorporeal devices are normally tested with respect to these risks.
Haemolysis testing
In vitro testing for haemolysis is technically straight-forward because red cell preparations can be exposed to devices or materials under defined conditions and the amount of haemoglobin that is released is easily measured and used as a surrogate for red cell membrane damage. This is, in fact, so straightforward, that this test is widely used, but, in my opinion, not always for the best of reasons. Haemolysis testing of biomaterials has been advocated for, and used in, standard biological safety testing of materials for more than 30 years. This leads to the second event of relevance here. Recently financial media coverage was given to an announcement, intended for investors, of the successful passing of haemolysis tests of a new hydroxyapatite-based coating. Part of this text stated that successful completion of the haemolysis test “marks another critical step in the drive towards commercialisation,” and passing the test “indicated that the exceptional coating technology has the potential to redefine the current boundaries of passive and drug-eluting coatings.”
Obviously there is nothing wrong in using good test results for marketing strategies and investor-relations, especially when the tests are essentially mandated by standards, but we have to consider what this testing actually means. Haemolysis testing can involve direct contact between the red cell preparation and a material surface or, more usually, with an extract from the material obtained after incubation with saline. There is usually only short-term contact with the red cells. If the material is to be used within a device that will be employed in dynamic contact with circulating human blood such as in a valve or tubing it is difficult to see how the results of these static, short-term in vitro tests can be relevant to the prediction of the performance of the device; with this testing any risks of haemolysis will be far more related to the design and the haemodynamics than the extraction of leachables from the material. With the hydroxyapatite-based coating material mentioned above, it is far from clear how good results from this test can actually mean anything about the performance of the material in any realistic hydroxyapaptite application.
This leads to the third event, which in reality is a sequence of similar events during the past few years. The standard haemolysis test is sometimes referred to as a haemocompatibility test, alongside tests to evaluate the effects of materials on platelets or the blood clotting cascade. I do not believe that haemolysis has anything to do with haemocompatibility in the normal meaning of that word. Of more relevance is the fact that the test can be considered as a type of cytotoxicity test, with the target mechanism being destructive effects on the cell membrane rather than on any intracellular components such as the cell nucleus or the mitochondria. This is more relevant, but it often leads to a problem. Time and time again I have been asked to explain inconsistencies in haemolysis test data obtained during routine preclinical test programmes. I have seen some standard biomaterials with many years of successful clinical use being determined to be haemolytic by these procedures (undertaken by well qualified laboratories under good laboratory practice conditions), and where repeat testing reverses the results. In many of these cases there have been no problems with the standard cytotoxicity tests, nor with any of the other basic in vitro or in vivo preclinical tests.
 |
David Williams Clinical Engineering Department, Royal Liverpool University Hospital, Liverpool L69 3BX, UK, tel. +44 151 706 5606 fax +44 151 706 5803, e-mail: dfw.ce@liverpool.ac.uk |
Haemolysis and the prediction of clinical performance
In other words, haemolysis tests can be inconsistent as well as irrelevant. I am strongly of the opinion that biological safety tests should not be used if they lead to data that can be described as either biological or safe. The first of the haemolysis events I mentioned above clearly demonstrates how haemolysis can be an important and fatal factor in the biological safety of medical devices; but no value can be derived from haemolysis tests that are unreliable. Indeed, the uncertainty over false positive or negatives can lead to serious difficulties with regulatory approval and the prediction of clinical performance.
Professor David Williams DSc, FREng is Professor of Tissue Engineering at the University of Liverpool and Director of the UK Centre for Tissue Engineering located in the Universities of Liverpool and Manchester. He is Editor-in-Chief of Biomaterials, the leading journal in the biomaterials field. He is Scientific Director of STEPS, the European Commission Framework VI Programme on a Systems Approach to Tissue Engineering Products and Processes. Professor Williams is also a Managing Partner of Morgan & Masterson LLC, a consulting partnership that focusses on global health-care issues.
