The recent publication by the U.S. Food and Drug Administration (FDA) of Guidance for Industry Pyrogen and Endotoxins Testing: Questions and Answers, dated June of 2012, the Department of Health and Human Services updated the agency’s thinking regarding pyrogen and bacterial endotoxin testing requirements—and has (rightly so) prompted medical device manufacturers to take a new look at their endotoxin testing programs.

Fig 1. – Endotoxins testing or lab equipment utilized for testing.
Bacterial endotoxin is the cell wall from gram-negative bacteria. The cell wall is generally composed of lipid polysaccharide (LPS) material. Medical devices that contact circulating blood or cerebrospinal fluid, or are used in intraocular products or in vitro fertilization procedures, are required to be endotoxin-free. Endotoxinfree means that the device does not contain endotoxin levels high enough to elicit a fever response in a patient. This level is generally defined as 5EU/kg.

The majority of pyrogens found on medical devices are bacterial in nature. Some medical devices that are required by the FDA to be tested for endotoxins are:

• Intravenous catheters and tubing,

• Cardiac stents and angioplasty balloons,

• Certain implants such as pacemakers/defibrillators,

• Spinal catheters,

• Transfusion catheters,

• Intraocular lens implants,

• Cranial drills,

• Perfusion tubing,

• Dialysis tubing and filters,

• Vascular grafts, and

In vitro fertilization catheters.

Historically, the FDA has required end-product testing of medical devices to release the lot. Therefore, most quality control (QC) labs and quality assurance (QA) units would submit three to 10 samples to the lab for testing. This testing involves following a validation program that, when performed, indicates the product is compatible with the detector system (horseshoe crab blood) lysate. Some products contain chemicals that inhibit or enhance the ability of the lysate to form a clot, or in the case of turbidimetric or chromogenic testing, to develop a reaction to endotoxin. Small amounts (under ppm) of bacterial endotoxins cause the lysate to clot or create a reaction. Therefore, it was always important to test the materials for compatibility.

This procedure was more appropriately called inhibition/enhancement testing. Once validated using three lots of materials, in accordance with the FDA’s Guidance, June 2012, the products can be tested under United States Pharmacopeia (USP) 35 2012 chapter 85, Bacterial Endotoxin Test. So, the majority of medical device manufacturers would test under these requirements and release a product to market when the results indicated an acceptable response. (See Figure 1)

An acceptable response is based on the dilution factor and the sensitivity of the reagent or lysate. The current requirement under USP section 161 is 20 endotoxin units per device. Intrathecally (cerebrospinal fluid) communicating devices have a much lower limit of 2.15 EU per device. Most medical devices have some level of endotoxin, albeit a very small amount. Using this information can be helpful in controlling endotoxin levels and quality attributes.

The Sampling Plan

Table 1. Endotoxin Limits.
When establishing a sampling plan for in-process and finished products, note that the FDA leans on the recommendations in the Quality Systems Regulations’ CFR 820 current edition, sections 21 CFR 820.70 21 and 820.250 for appropriateness of the sampling plans. AANSI/ AAMI ST 72 addresses sampling plans and sampling frequency. The requirement for sampling plans is generally limited to a production batch. Alternative sampling plans besides finish product testing (i.e., pre-sterilization) are acceptable but must be validated with appropriate manufacturing controls and testing. The FDA may ask for statistical endotoxin sampling plans in future submissions. The risk assessment regarding potential for contamination in raw materials, in process materials, and finished product would be a great step forward for many. The FDA is recommending a 30-day notice for sampling plan changes. Alternative plans will require more information for wet processes. These are production processes that utilize water during the manufacturing of the product or components (i.e., extrusion/washing or in-process testing, such as leak testing for balloon catheters). The ANSI/ASTM ST72 2002/R2010, Bacterial Endotoxin Test-Test Methodologies, Routine Monitoring and Alternative to Batch Testing from the Association for the Advancement of Medical Instrumentation in Washington, DC, provides more information on alternative sampling plans.

Sample Storage and Handling

While sample storage and handing processes are certainly important, most medical device samples that come into the lab for testing are stored under room temperature conditions. In the past, the endotoxin stability of medical devices has never been questioned. Endotoxins are sticky substances and generally can be very stable once on medical devices. Liquid medical devices would be the only exceptions. Liquid medical devices are those that are used for perfusion and organ preservation. Combination products (whereby the principal mode of action is a medical device) are high risk and should be tested for endotoxin stability.

If a device company is experiencing issues with waterborne organisms (from extrusion tanks), then the source of the endotoxin and stability of that contaminate should be considered during the shipment to the lab.

Extrusion water samples should be sent to the lab controlled at 4°C to 8°C for testing. Products should be shipped to the lab under controlled conditions until data is obtained that supports the shipment conditions. Shipment temperature and humidity conditions can vary widely. Spiking samples with endotoxin and bracketing the shipment conditions may be required for future submissions.

Composite Samples

Fig. 2 – Vials of cell culture material.
The FDA recommends that finished product samples for analysis of bacterial endotoxin can be pooled into a composite sample. Pooled extracts can be kept at stability temperatures until tested. Generally, this temperature is required by the practitioner to validate the stability storage and handling procedures. The extracts are generally aqueous (water for injection) washings from the internal blood contact surface of intravenous or transfusion assemblies. Bacterial endotoxins are not expected to be affected by 2°C to 8°C storage. There are many references to the theory that endotoxins do not rinse off medical devices completely with water. A water rinse generally requires vigorous mixing and flushing to remove endotoxins from the polymers. The FDA recommends 30 seconds of vigorous mixing from each of the product containers. They also request that individual containers be stored and available for retest. (See Figure 2)

Transitioning between Alternative Bacterial Endotoxins Tests (BETs)

When determining the best process for transitioning from one BET to another, note that the recombinant Horseshoe Crab factor C assay requires validation under the USP section , Validation of Compendial Procedures. It must be demonstrated that equivalent or better results, as compared to a compendial test, can be achieved. Additional work may be required based on one’s risk management program. Some of these are recommendations regarding sampling and testing of medical devices. Other procedures or testing regimens can be used. However, the appropriateness of the procedures must be justified.

The FDA states: “For devices, a 30-day notice may be appropriate for changes to quality control testing used in incoming components, raw materials, inprocess devices, or finished devices, including performing end-product pyrogen testing on non-sterile samples prior to sterilization.”

Endotoxin Limits

The limits for endotoxin stated in the USP 35 2012 chapter 161, Transfusion and Infusion Assemblies and Similar Medical Devices, depend on the route of administration, the intended use of the device, and what the device contacts: blood, cardiovascular system, lymphatic system, cerebrospinal fluid, intrathecal routes of administration, permanently implanted devices, and devices implanted subcutaneously. (See Table 1)

The FDA has added some information that can be confusing to the non-practitioner. The process of preparing eluate/ extract from the devices can vary widely with each device configuration. This can include:

A. Flushing the device: catheters that the internal surfaces are labeled non-pyrogenic

B. Disassembly or cutting up devices into pieces: devices that are too big to be adequately rinsed

C. Immersion: implantable devices

The QC lab generally will recommend the appropriate eluate/extract technique. The guidance indicates that labs can test the surface area that contacts the patient. One can use an adjustment factor to determine limits. However, generally 40 mL of non-pyrogenic water is used to rinse the device. Ten units are recommended in their guidance document. Using more than 40 mL per device requires a submission.

The FDA has also indicated that one should evaluate multiple units of the same device used in a procedure. The devices must meet the same endotoxin limit as a single device; for example: 20 EU/device or 0.5 EU/mL.

The FDA is also requesting 510k premarket notification submissions when the firm is deviating from the new guidance or recognized standard. Significant deviations include: sampling fewer than three lots for inhibition enhancement testing; using a lesser sensitivity to endotoxins; using more than 40 mL per device for rinsing without adjusting the dilution factor, which would cause a greater dilution than the guidance document allows.

Quality by Design (QBD)

Quality by design (QBD) concepts can support endotoxin limits. The goals of product and process risk assessments are to provide a consistently non-endotoxic product. Firms need to evaluate raw material or in-process testing. When the process is considered a wet process, QA should review the critical control points and validate the propensity for endotoxin contamination. QA teams may be required to add this as part of QBD programs. That means quantitative testing (i.e., setting alert and action limits for in-process materials) is preferred by the FDA over limit testing. This type of approach will prevent excursions before they happen and prevent a recall. The FDA will be evaluating the endotoxin limit approach submitted with each new device. A strategy for setting limits is justified. QBD programs should be instituted in most production environments.

In the agency’s new Guidance document, the FDA provides direction on what is acceptable and reveals some color on the sampling requirements and quality- by-design concepts. QBD risk management, according to ISO 10993-1:2009 Biological Evaluation of Medical Devices Part 1: Evaluation and Testing in the Risk Management Process, would require device management to increase testing and sampling targets for wet manufacturing processes. Utilizing the quantitative approach to endotoxin testing will, for the majority of affected medical device manufacturers, provide the most beneficial route to compliance.

This article was written by Steven G. Richter, PhD, President and Chief Scientific Officer of Microtest Laboratories, Inc., Agawam, MA. He founded Microtest after a distinguished career at the U.S. Food & Drug Ad ministration. For more information, email This email address is being protected from spambots. You need JavaScript enabled to view it. or visit See Microtest Laboratories at BIOMEDevice, Boston, Booth 1010.

Medical Design Briefs Magazine

This article first appeared in the April, 2013 issue of Medical Design Briefs Magazine.

Read more articles from this issue here.

Read more articles from the archives here.