MANUFACTURING
Innovation in Endotoxin Detection Moving Out of the Stone Age Endotoxin Detection – a Brief History With the threat of the H1N1 virus and the upcoming flu season upon us, vaccine manufacturers are working diligently to provide an H1N1 vaccine to help protect us from a large outbreak of illness. In addition to safety and effectiveness, part of the quality testing for products such as vaccines and pharmaceuticals includes ensuring that the product passes specifications for sterility and endotoxin.
“In addition to safety and effectiveness, part of the quality testing for products such as vaccines and pharmaceuticals includes ensuring that the product passes specifications for sterility and endotoxin.” The history of endotoxin detection has always included the use of an animal. In the 1920s, a test for the presence of pyrogens in a solution was developed by Florence Seibert using rabbits. During World War II, there was a high demand for safe intravenous solutions, and Florence Seibert’s rabbit pyrogen test was used to check the products for contamination. The United States Pharmacopeia published the first pyrogen test in 1942. This test involved injecting the pharmaceutical into a rabbit and monitoring the animal for an increase in temperature or a fever. If the rabbit spiked a fever, it indicated that the sample contained an unacceptable level of a pyrogen, such as endotoxin, and the batch of product could not be sold. In the late 1960s, researchers studying the blood clotting system of the horseshoe crab found that it was extremely sensitive to endotoxin. Frederick Bang and Jack Levin 86 INTERNATIONAL PHARMACEUTICAL INDUSTRY
developed an endotoxin test that involved mixing the blood-clotting factors that are in the amebocyte with a drug sample in a test tube (1). If sufficient endotoxin was present, they found that the liquid in the tube would clot in such way that when the tube was inverted top to bottom, the clot stayed in the bottom of the tube. This was the first “LAL” test for endotoxin. LAL stands for Limulus Amebocyte Lysate and contains the factors inside the amebocytes (blood cells) of the horseshoe crab, Limulus polyphemus. In 1977, the FDA approved the use of the LAL test as a replacement for the Rabbit Pyrogen Test to detect endotoxin in human and animal injectable pharmaceuticals and biologicals, and implantable medical devices. The first commercialised LAL method, the gel clot LAL method, is not very different from the original test made by Bang and Levin. This test provides a simple yes or no answer as to whether or not the product being tested is contaminated with endotoxin. In the 1980s and 1990s, photometric methods were developed that would allow a manufacturer to determine the amount of endotoxin that may be in their product. In these tests, the mixture
of LAL and test sample turns the colour yellow or becomes turbid in the presence of endotoxin. The degree of colour change or how fast the change occurs makes it possible to calculate how much endotoxin is in the test sample. A New Technology for a New Century LAL-based methods continue to dominate the QC laboratories that conduct endotoxin detection tests. The horseshoe crab ancestors and their blood clotting system date back hundreds of millions of years. So, the evolution of the science behind the LAL-based methods was established a long time ago. It has taken humans some time to catch up. The horseshoe crab has also provided us with the endotoxin detection technology of the future. In LAL is the crab’s Factor C, the enzyme that is sensitive to the presence of endotoxin. Using a clone of the horseshoe crab’s DNA, a recombinant form of Factor C (rFC) was produced and an endotoxin detection assay has been developed from it. The rFC assay is the evolution of the LAL assay. It is equivalent to other photometric endotoxin detection methods that use LAL to detect endotoxin
Figure 1: Endotoxin standard curves using eight different lots of rFC. The log net fluorescence is proportional to the log endotoxin concentration and is linear in the 0.01 to 10 EU/ml range. Lot-tolot standard curves exhibit excellent reproducibility. www.ipimedia.com
according to the parameters listed in United States Pharmacopeia (USP) chapter “Validation of Compendial Procedures” (2). In 2009, the US Food and Drug Administration approved 510(K) submissions that use the rFC method as the final release test. Predictable Assay Performance Due to the biological nature of LAL, utilising blood from the horseshoe crab introduces lot-to-lot variability. Differences between the blood of animals, seasonal changes and environmental factors all can contribute to the variability of the final product. Because rFC is a recombinant form of the horseshoe crab Factor C, it retains all of the endotoxin reactivity that
LAL is known for. However, the cloning and downstream manufacturing procedures result in a reagent that contains only Factor C and none of the other factors found in the traditional LAL cascade. The end result is greater control, leading to a product that has greater lot-to-lot consistency. Standard curves generated from different lots of rFC show remarkable consistency. This translates into more reproducible results and less variable recovery values for your PPCs (Positive Product Controls) (Figure 1). The rFC Assay is Endotoxin-Specific Endotoxin detection methods that rely on LAL suffer from the fact that the LAL enzyme cascade includes two pathways, one triggered by endotoxin and the other
Figure 2: Comparison of glucan activity between kinetic chromogenic LAL and rFC. The false positive signal from the LAL assay is reduced in the presence of a glucan blocker. The rFC assay is endotoxin-specific.
Figure 3: Endotoxin potency of four purified lipopolysaccharide preparations as tested by multiple methods. www.ipimedia.com
triggered by glucans. A positive LAL reaction due to glucans is a false positive for an endotoxin detection assay. As recombinant Factor C is the only enzyme from the cascade in the rFC assay, the test method is specific for endotoxin. Figure 2 illustrates the comparison between the reactivity of the kinetic chromogenic LALbased method (Chromogenic) and the rFC method relative to the presence of glucan in the test product. The positive signal with the Chromogenic method is reduced when a glucan blocker is added into the reaction mixture. The rFC method does not react with glucans, thereby reducing false positive results. Comparable Detection of Endotoxin Studies were conducted to demonstrate the ability of the rFC assay to detect endotoxin from different bacterial species. Figure 3 includes the results of a comparison between the responses of the rFC, kinetic chromogenic LAL and kinetic turbidimetric LAL photometric methods to different sources of purified endotoxin. Three assays were performed for each endotoxin and method combination. From the results shown in Figure 3, the rFC method recognises endotoxin from different source materials similar to other photometric procedures, demonstrating comparability to the LAL-based methods. Endotoxin Recovery from Products Similar to Other Methods Comparisons of the recoveries of known amounts of endotoxin also were conducted. Our global, multi-centre study included testing of ten pharmaceutical or related products with the kinetic chromogenic LAL and rFC methods. The products tested in the multi-centre study included items such as Water for Injection, Lactated Ringer’s Injection USP, 0.9% Sodium Chloride Injection USP, Erythropoietin, Albumin (Human) USP 25% Solution, Vancomycin HCl USP, and Hemodialysate. Each product was spiked with 0.1 EU/ml of endotoxin and tested by the kinetic chromogenic LAL and rFC methods by three analysts at six different locations. Figure 4 illustrates the results of the recovery of the spike. The percent recovery of the 0.1 EU/ml spike in all products (except Human Albumin) was within the current USP specification of 50 – 200% for both methods (3). In an additional study, the recovery of an endotoxin spike of 0.1 EU/ml in Water for Injection was measured by the kinetic turbidimetric LAL photometric method for comparison. The turbidimetric assays were conducted by three analysts at one location. The results in Table 1 show the similarity in the detected level of the endotoxin spike between the methods. In summary, the recovery of endotoxin INTERNATIONAL PHARMACEUTICAL INDUSTRY 87
from water and the other tested products using the rFC method is comparable to that of the LAL-based methods. Standard Curves meet Linearity Requirements An acceptable USP method must elicit results that are proportional to the concentration of the analyte being tested (2). For an endotoxin test, the analyte is endotoxin and linearity is demonstrated by the correlation
coefficient. Table 2 summarises the linear regression statistics from eighteen rFC standard curves generated by the multicentre study. Each curve was comprised of four standards (0.01, 0.1, 1, and 10 EU/ ml) in replicates of three. The correlation coefficients for all eighteen standard curves were greater than the US FDA required 0.980 specification for a quantitative endotoxin detection test (4).
Evolution The speed and ease with which the current endotoxin detection tests can be run in comparison with the rabbit pyrogen test has given the vaccine and pharmaceutical industries the ability to quickly, efficiently and confidently test their products for endotoxin contamination. The rFC assay provides an improvement that diminishes the use of animals for endotoxin detection as well as moving the technology out of the Stone Age. As products using recombinant technology are developed to improve the lives of humans and animals, the same technology should be applied to the quality testing of those products. The rFC assay is the evolution of the LAL test. Combining 21st century technology with the horseshoe crab’s endotoxinsensitive protein, Lonza has developed an equivalent, reliable and sustainable endotoxin detection method for the future. Lonza is dedicated to bringing further innovative rapid testing systems to the healthcare industry to help manufacturers deliver their new products to the consumer and patient more quickly, while helping to make sure the products are safe. References: (1) Levin, J. and Bang, F.B. Clottable protein in Limulus: Its localization and kinetics of its coagulation by endotoxin, Thromb. Diath. Haemorrh, 19:186 (1968) (2) United States Pharmacopiea, Chapter Validation of Compendial Procedures (3) Unites States Pharmacopiea, Chapter Bacterial Endotoxins Test (4) U.S. Department of Health and Human Services, Public Health Service, Food and Drug Administration, Interim Guidance for Human and Veterinary Drug Products and Biologicals (1991)
Figure 4: Recovery of a 0.1 EU/ml endotoxin spike in multiple products is comparable between methods.
Table 1: Recovery of a 0.1 EU/ml endotoxin spike in Water for Injection as tested by multiple methods.
MARIBETH DONOVAN JANKE, PH.D. is the Sr. Product Manager for Endotoxin Detection in the Rapid Testing Systems unit of Lonza Walkersville, Inc., Walkersville, MD, USA. For more information, please visit www.lonza. com/lal or email the author at
[email protected].
Table 2: Linear regression statistic summary, based on 18 standard curves, by six analysts at six sites. www.ipimedia.com
INTERNATIONAL PHARMACEUTICAL INDUSTRY 89
