By Melissa Stappen and David Hussong, Ph.D., ValSource, LLC
The United States Pharmacopeia (USP) harmonized Chapter <85> Bacterial Endotoxins Test describes the procedures for performing Limulus amebocyte lysate (LAL) assays for the detection of endotoxins in pharmaceutical products. The U.S. Food and Drug Administration (FDA) has established regulations for pyrogenicity of drug products, and had provided guidance relating to the use of the LAL assay as an indicator of pyrogenicity in lieu of the rabbit pyrogenicity test.
Compendial harmonization of the Japanese Pharmacopeia, European Pharmacopeia, and USP allowed the FDA to update its 1985 guidance for the use of the LAL. These updates were reflected in the withdrawal of the 1987 FDA guidance and publication in 2012 of a question and answer (Q&A) document intended to explain how to use the LAL assay in support of patient safety assays for product release and process development.
This article reviews the development and evolution of the LAL assay, its uses, and misunderstandings that have occurred.
Endotoxins And Pyrogens
Febrile responses by patients have been observed since medical observations of the infection process as far back as the 6th century BC.1 The general fever response was considered the result of pyrogens, and a subset of these pyrogens were from gram-negative bacteria.2 A great deal of interest evolved as studies of gram‑negative bacteria were found to induce a variety of responses in patients exposed to portions of their cell walls.1, 2 In 1923, studies by Siebert established the rabbit as the preferred model for pyrogens detection.3 Further studies showed this response was the same as those reactions following injections of drugs into patients.4
The concern about fever induction from injectable drugs was sufficiently established that the USP included a test for pyrogens in its 12th revision in 1942. This was the rabbit pyrogenicity test (RPT). Comparative studies on the effect of endotoxins on humans and rabbits were reported by Greisman and Hornick in 1969 showing some correlation between reactivity in man and rabbits, as well as differing degrees of reactivity between endotoxins from gram-negative bacterial species.5 However, they noted that febrile responses to endotoxins total doses in man exceeded the rise in fever response of rabbits.
Small amounts of endotoxins cause various physiologic responses. In addition to fever, small amounts enhance general resistance to infections and protect against radiation injury. Slightly greater doses can result in necrosis of tumors, and even greater doses induce shock and organ destruction with death. 6 The favorable attributes of endotoxins as therapy (tumor destruction and immune enhancement) were being explored into the 1970s, but potency testing was limited to the RPT.
Development Of The LAL
Unusual clotting properties of the blood of horseshoe crabs were noted in 1885 by Howell.7 Further studies of this phenomenon demonstrated that the amebocytes of the blood were necessary for coagulation when exposed to endotoxins, and that lysed amebocytes enhanced the reaction.8 The reactions of this lysate were further studied at the National Institutes of Health’s Division of Biologic Standards (later the FDA Bureau of Biologics, and then FDA CBER) as a potential replacement for the rabbit pyrogens test (RPT). The need for such a replacement was driven by the laboratory’s need to test at least four batches of biological products each day as part of their pre-release license criteria. The RPT is a resource-intensive activity, and the results are subject to a great deal of influences. At that time, the products tested included:
By 1972, FDA was preparing lysate batches from crabs that were bled about twice per year, and was receiving complaints about reagent variability from clinicians using commercially prepared lysates to determine endotoxin activity for patient studies.9 While this alternative assay seemed to be a very practical solution, several challenges to assuring reproducible results were evident. Additional studies reported in 1973 had shown the commercial lysates to exhibit remarkable variability.10, 11
To assure consistency of commercially provided reagents, in 1973 FDA published a notice in the Federal Register that proposed application of Section 351 of the Public Health Service Act to preparations of the LAL.11 Manufacturers of lysate were required to submit samples of each batch for testing by FDA, prior to release. It was quickly realized that lysates reacted differently to endotoxins from different strains of bacteria, and even different preparations of endotoxins from the same genus when tested on the basis of endotoxin weight. These variations could be as much as 1000‑fold.10, 12
It was apparent that test procedures and reagents needed to be “standardized,” including lysates and calibration standards.12 A large lot of lysate was purchased and standardized as “U.S. Reference Lysate, Lot 1.” An endotoxin “standard” was provided from another FDA lab, where pyrogens studies had been using endotoxin from Klebsiella pneumonia in RPT studies. This K. pneumonia endotoxin was “stabilized” with 0.1% human serum albumin (HSA) and lyophilized (Lot 1b).9
Lot 1b was replaced in 1974 with a bulk endotoxin (Lot EC) from E. coli 0113, extracted by hot aqueous phenol.9, 13 From this bulk, several sub-lots were prepared:
It is important to understand that these extracted, purified and formulated “endotoxin standards” were analytes intended for calibrating lysates. They were practical, empirically developed analytes with no intended relation to positive controls with products.
LAL Testing As A Regulatory Procedure
FDA provided guidance for use of the LAL in 1987.14 Included in the guidance was a copy of USP <85>, which provided procedures for the test.15 Performance of the test included demonstrations of its ability to detect endotoxins in the drug product. A great deal of detail went into explaining that the reaction of the lysate only occurred under specific circumstances including pH, temperature, and appropriate ionic conditions. Often it was necessary to dilute products in water or buffer to attain the conditions necessary for the test to work. Based on the sensitivity of the lysate and the specified limit for the endotoxin content of the product, a maximum valid dilution (MVD) and minimum valid concentration (MVC) would be established, and recovery of endotoxin at that dilution/concentration needed to be demonstrated by adding a known level of activity in the form of the a reference standard for use as a “product positive control.” Generally, the added control endotoxin was a commercially prepared control standard endotoxin (CSE) used to calibrate the lysate that was being used. CSE was described in the 1985 USP <85> as, “A control standard endotoxin (CSE) is an endotoxin preparation other than the RSE that has been standardized against the RSE”.15 Most CSE preparations are prepared similarly to RSE by hot aqueous phenol extraction, which is a legacy of the original calibration standard.
FDA also needed to validate the LAL test in its laboratories and around 1980 began to distribute drug product samples containing added endotoxin to its regional laboratories for validation testing. Notwithstanding the addition of known quantities of endotoxins (CSE) to the drug products, many samples were found to yield much less endotoxin activity by assay than had been added. Although subsequent studies revealed that many factors related to the physical nature of extracted endotoxin contributed to the reduced recovery,16 FDA noted at that time that these interferences were reduced when crude extracts of gram-negative bacteria were used as the material to spike the products. Subsequent studies have noted structural and function differences between chemical extracts from endotoxins and physical extracts of endotoxins from microbial surface membranes that are typical of product contamination from gram‑negative bacteria. 17 Understanding these differences has been confounded by the historically interchangeable use of terms “endotoxin” and “lipopolysaccharide” (LPS). However, the reactive material purified from endotoxins should be described as LPS.
Regulatory Concerns Resulting In Citations Related To LAL Testing
Many procedural errors have been identified in LAL tests, as noted in FDA citations over the years.16 Additionally, FDA was challenged to keep current with the changing therapeutic dosing of products, which resulted in the need to change the endotoxins acceptance limits for products that have been listed in monographs and also were listed in Appendix E of the FDA’s 1987 guidance on the LAL.14 When USP harmonized chapter <85>, much of the material in the 1987 FDA guidance was unnecessary, allowing it and its Appendix E to be retired. However, certain technical issues remained in need of discussion to help users stay in regulatory compliance, so in 2012 the FDA put forward a Q&A document to help explain regulatory expectations.18
Products that interfere with LAL’s ability to react with LPS standard or that affect the LPS material itself will result in an observation of “inhibition” of the test. Common interferences may be chemical or physical. Chemical inhibition includes chelation, protein denaturation, or pH shifts, while physical inhibitors are ionic, adsorption (of endotoxin), or gelation.19, 20 To consider the impact of storage on the test process, the FDA Q&A document recommended a “hold time study” as it relates to endotoxins recovery following storage. These concerns were ultimately published as Question 3 of the 2012 Q&A document:
Is sample storage and handling important?
Yes. The ability to detect endotoxins can be affected by storage and handling. Firms should establish procedures for storing and handling (which includes product mixing) samples for bacterial endotoxins analysis using laboratory data that demonstrate the stability of assayable endotoxins content. Protocols should consider the source of endotoxins used in the study, bearing in mind that purified bacterial endotoxins might react differently from native sources of endotoxins.
The experience of the Q&A authors caused them to add the note in the last sentence that the purified “standards” may not be representative of the bacterial endotoxin contamination found in a process stream or finished product. Although this was empirically observed in FDA’s validation studies around 1980, there was no further investigation of the biochemical cause of this “interference.” However, since the hold time interference using CSE was first described in 2013, 21 a great number of concerns have been voiced about the failure to recover the LPS standard added to certain products formulated with chelators (e.g., citrate buffer) and polysorbate solutions.20, 21 This failure of the control arm of the suitability study, known as low endotoxin recovery (LER), appeared to suggest that the user should revert to the RPT.
However, consistent with the position that purified LPS does not behave like native contaminating endotoxins, naturally occurring endotoxin (NOE) has been described for use as a control analyte in biopharmaceutical formulations that contain chelators or polysorbate and previously interfered with recovery of nominal spike activity. This “interference” appears to be caused by dissociation of the aggregates common to purified LPS in the presence of dispersing agents and chelators that bind cations needed for the molecules’ stability.20
Bolden et al also described the effective use of a NOE preparation that, when calibrated against CSE, provides favorable recovery of activity in hold time studies.20 Also noted was that while CSE activity during hold time studies was diminished in the LAL assay, its potency was also diminished in the RPT. Significantly, the NOE retained its activity in the LAL assay, and potency in the RPT was retained. This favorable recovery relates to the larger and balanced structure of the amphipathic natural molecule of endotoxin present on the cell’s outer membrane. This demonstrates that the initially reported control arm of the suitability study, or hold time study for storage, was flawed due to an inappropriate challenge material when hot phenol extracted LPS was used as the analyte. Whereas interference from a “hold time effect” was not apparent as LER using NOE, this appears to demonstrate a simple solution for the perceived LER effect.
USP has considered data describing the complexity of endotoxin in the outer membrane of gram‑negative bacteria and concluded that NOE is better representative of product contamination22, 23 and conforms to the definition of CSE provided in the original <85>.15 Use of the NOE represents a practical solution to a control (hold time study) test failure caused by inappropriate design of experiment using an analyte that was not intended for product challenge, but rather for lysate standardization. We feel this practical solution is a step forward in demonstrating the suitability of the assay for “complex product” formulations and that NOE will better support control tests for assurance of patient safety.
About The Authors:
Melissa Stappen, consultant with ValSource, LLC, has over 20 years of experience in the pharmaceutical, medical device, biotech, and clinical/healthcare provider settings. During her career, she has provided support to quality assurance, quality control, and compliance departments as a subject matter expert relating to endotoxin and microbial contamination control. Her current role of validation consultant emphasizes risk management based programs for laboratory instrumentation, methods, and equipment.
Melissa holds an AAS in medical laboratory technology, with ASCP board certification, and has been a member of the Parenteral Drug Association since 2005. You can contact her at firstname.lastname@example.org.
David Hussong has 46 years of professional microbiology experience. Since 2015, he has been a senior consultant with ValSource, LLC, where he specializes in regulatory microbiology issues. David is retired from the Commissioned Corps of the U.S. Public Health Service after 30 years with the FDA. While at FDA, he led the effort to replace the 1987 guidance for use of the LAL with the 2012 Question and Answer guidance. Also, David is the chair of the USP Microbiology Expert Committee for the 2015 – 2020 cycle. He has been a member of the Parenteral Drug Association since 1993, and a member of the American Society for Microbiology since 1975.
David earned his Ph.D. in microbiology from the University of Maryland (UM) and has also served as a research microbiologist at UM, the U.S. Department of Agriculture, and the U.S. Naval Medical Research Institute. He can be reached at email@example.com.