Header Artwork
Header Artwork


Extraction of Medical Devices for Bacterial Endotoxin Testing

 When performing the Bacterial Endotoxin test, we observe changes within a solution that occur as a reaction to the presence of endotoxin. As such, it is necessary that the test medium be a liquid. For pharmaceutical products, the sample is typically already a liquid or is in a powder form which can be reconstituted for testing. However, when testing a medical device, we need to perform an extraction to analyze the presence of endotoxins.  

Guidance on endotoxin tests on medical devices can be found in AAMI ST72, ISO11737-3 and USP <161>. Each of these standards offers the same general guidelines. The samples should be immersed in water that contains no detectable endotoxin at 37°C. Alternatives, however, are permitted if equivalency is demonstrated. 

 At Nelson Labs, various extraction methods are utilized depending on sample type and needs of the sponsor.  

Per ST72 and ISO 11737-3, commonly the extraction is performed by immersion of the test article in water at an ambient temperature that is incubated for a minimum of 1 hour at 37°C with agitation.  

USP 161 states: “The standard extraction method is to soak or immerse the device or flush the fluid pathway with extracting fluid that has been heated to 37 ± 1.0°, keeping the extracting fluid in contact with the relevant surface(s) for NLT 1 h at controlled room temperature. Alternate extraction or rinsing methods may be used, but must be demonstrated to be equivalent to or better 

 It is acknowledged that no matter what extraction method is being performed, there will inherently be some inefficiency and all endotoxins present may not be represented in the sample extract. In order to account for this inefficiency of extraction more stringent limits are imposed on medical devices.  

 Endotoxins are challenging. Extract and extraction methods suffer from poor extraction efficiencies and reproducibility. This fact is recognized by AAMI ST72, stating that “investigations have demonstrated that the extraction of endotoxin from spiked device materials might not achieve complete recovery”. Therefor limits for medical devices are set much more stringentto account for any possible inefficiency in the extraction method.”  

Furthermore, in contrast to bioburden determination where often extraction of micro-organisms is required, when extracting endotoxins from a medical device, there is no requirement for performing efficiency testing 

Peter Cornelis, Ir.

Senior Expert Microbiology

Peter Cornelis graduated from the Catholic University of Leuven (Belgium) in 2000 as a Master in Applied Biological Sciences (Major Biotechnology). In 2003 he started working for Toxikon Europe (now Nelson Labs) as a study director Microbiology and in-vitro Toxicology. From 2007 until 2016, he was department supervisor for Microbiology and in-vitro Toxicology. Since 2016, he is responsible for research, validation, and development of new microbiological and in-vitro toxicological methods. Peter is a member of the ISO committee TC 194 WG5, Cytotoxicity and WG 8, Irritation and sensitization. As an expert, he was involved in the adaptation of ISO/TC 194 ISO/DTS 11796:2022(E) Biological evaluation of medical devices — Guidance for interlaboratory studies to demonstrate the applicability of validated in vitro methods to assess the skin sensitization of medical devices.

Nathan Pett

Study Director BET

Nathan Pett graduated from Utah State University in 2018 with a bachelors in Biochemistry. In 2019 he began work for Nelson Labs as a Lab Analyst and in 2021 transitioned to the role of Study Director in the BET department.

Emily Spackman, B.S., RM(NRCM)

BET Consulting Study Director

Emily Spackman has worked for Nelson Laboratories in the Bacterial Endotoxins Lab for over 16 years. She graduated from the University of Utah and holds a bachelors degree in Biology. She is a member of The National Registry of Certified Microbiologists and has experience with both medical devices and pharmaceutical products.