The controversy pertaining to low recovery of purified endotoxin (CSE) and natural endotoxin (NOE) now dominate BET discussions. Does this recovery problem impact the work in your setting, and have you lost confidence in an LAL reagent being able to detect endotoxin in a pharmaceutical contamination event?
Jay Bolden, Associate Sr. Consultant Biologist, Global Quality Laboratories, Eli Lilly and Company: The recent requirement to conduct endotoxin recovery studies for certain products has impacted quality control and development laboratory resources. The concern is understandable in light of certain disclosed data, but other disclosed data suggests the concern may be largely theoretical. Based on the data we’ve generated to date, we have confidence that the Bacterial Endotoxins Test is suitable for intended purpose for our products and we continue to have confidence in it.
Dr. Constanze Reinhard, Head of QC Microbiology, Lonza: Yes – ongoing discussions with costumers with respect to ET specification, endotoxin hold times and clinical relevance of the study outcome. I still have confidence in the assay however not in CSE as the “right” standard for hold studies.
Dr. Tim Sandle, Head of Microbiology, Bio Products Laboratory Limited: I think the difference between purified LPS and natural endotoxin complexes is established, and they will behave differently under conditions of the test.
The issue of low endotoxin one is important, although the subject is causing some confusion for LAL users. It would be good to get a consensus between the providers of LAL regents together with a clear signal from the regulators.
Jenny Farrington, PhD, Associate Director, Regulatory Affairs, Associates of Cape Cod, Inc.: The vast majority of our customers do not experience low recovery of purified endotoxin (CSE) and natural endotoxin (NOE) because it is only related to therapeutic products formulated with Polysorbate 80 and citrate. Samples outside of this category pose no issues. The LAL methodologies provide the most sensitive and rapid endotoxin quantitation for all other samples such as water testing, in-process, final product and raw material analysis.
Cheryl Platco, Merck: I have full confidence that the LAL test can detect contamination of bacterial endotoxins because the true contamination analyte is a native endotoxin. The purified CSE standards do not represent native endotoxins in terms of the stability of the analyte of concern. The CSE standards are purified LPS molecules which were designed as standards for activity measurements. The purified LPS molecules were not designed to be used as stability analyte surrogates for real contamination events.
Most BET labs use a kinetic LAL method and an incubating microplate reader for endotoxin testing. Do these methods have substantial limitations that warrant the development of new endotoxin test technologies?
Bolden: The current BET using LAL and incubating microplate readers have detection capability to femtogram levels of detection; no other modern chemical analysis can deliver that sensitivity with the broad specificity of LAL. The downside is that accuracy and precision are relatively loose compared to chemical methods, but comparable to other biological assays. Analytical capability improvements are always warranted as long as they can demonstrate end-user value over existing technology.
Reinhard: Yes – only detection of endotoxins; a simple / comparable assay to detect pyrogenes in general would be beneficial. Sandle: First off, as well as microplate readers, tube readers remain a popular option as well. The limitations of the technology is in terms of ‘hot wells’ and occasional erroneous results. These are attributable to the way the software reads wells or as a result of an external impact, such as vibration. One technology driver is for greater stability and hence test reliability. There are others, of course, such as improved accuracy and a faster time-to-result.
Farrington: Feedback from our customers indicates an incubating microplate reader is a reliable method for BET. When there is a need to add in test samples during the test or a customer is experiencing hot wells and would prefer a sample container that can be depyrogenated, we recommend the use of a tube reader. An incubating tube reader allows flexibility in adding samples while meeting regulatory requirements.
Platco: The kinetic LAL methods are valuable assays for quantitatively determining endotoxin activity levels. When compared to the gel clot semi quantitative method, the kinetic and endpoint methods have taken giant steps forward in determining endotoxin activity measurements. The kinetic methods are tests for activity. The MAT test is also a test for activity. Some newer tests which have been discussed are tests for measuring endotoxin as an impurity. Therefore, some of the newer testing would measure pyrogenically inactive and active forms of LPS. LPS as an impurity is yet to be determined to be clinically relevant in terms of pyrogenicity. These topics are still controversial and under discussions. There may be some special circumstances where it would be good to have an assay which can physically remove or not be affected by either biological or chemical interferences from the sample matrix.
Do your routine test methods enable you to meet most or all of your BET responsibilities?
Bolden: Our routine test methods, using LAL, have enabled us to meet all of our BET responsibilities for over 30+ years.
Reinhard: Yes
Sandle: For sample-by-sample testing, yes to a degree, although capacity is sometimes an issue. This is fine for batch release. For greater process control, however, then real-time endotoxin of water systems represents an enhancement of what can currently be achieved through the LAL test.
Farrington: Yes, at Associates of Cape Cod, Inc., we often test to the very limit of quantification for endotoxin and find our routine test methods enable us to meet all of our BET responsibilities in a USP compliant manner. When using the BET test to analyze our water systems, raw materials and in-process materials as well as customer samples, we find the key to a successful routine test is perform proper method validation upfront.
Platco: The kinetic LAL methods are valuable assays for quantitatively determining endotoxin activity levels. When compared to the gel clot semi quantitative method, the kinetic and endpoint methods have taken giant steps forward in determining endotoxin activity measurements. The kinetic methods are tests for activity. The MAT test is also a test for activity. Some newer tests which have been discussed are tests for measuring endotoxin as an impurity. Therefore, some of the newer testing would measure pyrogenically inactive and active forms of LPS. LPS as an impurity is yet to be determined to be clinically relevant in terms of pyrogenicity. These topics are still controversial and under discussions. There may be some special circumstances where it would be good to have an assay which can physically remove or not be affected by either biological or chemical interferences from the sample matrix.
Do you believe that inhibition problems in the BET are a significant limitation of LAL reagent? Is glucan enhancement a limitation in your lab?
Bolden: No. Inhibition/enhancement testing is described in the BET, and has been a central tenet of BET method development since the inception of the compendia chapter. The vast majority of inhibition mechanisms are easily overcome to allow the end user to answer the question “Can the method reliably detect endotoxin in the test solution?” We have tools to overcome glucan interference including the non-compendial recombinant Factor C reagent which is not susceptible to the potential glucan false-positive results.
Masakazu Tsuchiya, Senior Research Scientist, Charles River Laboratories: No. It is important to find causes to the problems. Causes of inhibition can be elucidated for most of the cases. Once the causes are elucidated, solutions for the inhibition can be found. Betaglucan is not a problem because we use beta-glucan blocking buffer when beta-glucan contamination is suspected.
Reinhard: No – detection principle of oligomers seems to be representative for the in vivo situation. Glucan enhancement not an issue in the lab and blocking of the enhancement effect only required in very rare cases.
Sandle: This is very much product dependent. For testing water, LAL has always been fine. For some products, inhibition can be an issue. The most effective way remains dilution, but this can be restricted by the MVD. The availability of buffers normally helps.
With glucan, this has been a problem in the past when linked to cellulosic filters. The extent to which lysate with a glucan blocker should be used depends on how concerned a particular organization is about beta-glucan: is it something of no medical significance and thus an annoyance to be screened out? Or is it something with the potential to cause physiological harm and for which pharmaceutical companies need to know about? While this is product dependent to a degree, I don’t think the industry has fully addressed the significance of glucan in products.
Dr. James Cooper: During the past 40 years, the LAL industry has greatly increased the sensitivity, reaction rate and stability of LAL reagents. With few exceptions, modern day reagents have sufficient sensitivity to enable substantial dilution to avoid inhibitory test conditions. It is not uncommon for end-product testing to yield safety factors that are a hundred times greater than a given product's endotoxin limit. The fact that no pyrogenic reactions have occurred in the LAL era attests to the remarkable adequacy of LAL reagents to protect patients from endotoxin contamination.
Beta glucans interfere with endotoxin tests because they activate LAL reagent by an alternative pathway. Glucan blocking agents are permitted by the USP to avoid false-positive endotoxin test results. Recognition that beta-glucans primarily arise from cellulosic sources enables our industry to anticipate and control their presence; for example, cellulosic depth filters are a common source of glucans. All unexpected endotoxin OOS results should be screened for beta glucans and resolved as invalid if their presence is verified.
Farrington: Inhibition problems in the BET are present but not a significant limitation of the LAL reagent. We find that the dilution of the sample and/or the use of a buffer will yield a valid test in the majority inhibitory samples. (1,3) – BD Glucan (BG) is known to cause an enhancement but can easily be overcome by use of a glucan blocking buffer. Since BG may be present in the raw materials and processed equipment used in manufacturing, it is valuable to understand this contaminant not only as a potential source of endotoxin specification excursions but to prevent adverse events especially with immunologically active biologicals.
Platco: I do not believe that inhibition of the BET test is a significant limitation of the LAL reagents. There are different types and causes of inhibition. The inhibition of the activation of the LAL reaction and detection mechanisms is one type of interference which is addressed by resolving test interferences (enhancement/inhibition). An LER type of inhibition is the inability of the molecule to activate the lysate because it has been rendered unable to perform the activation of the lysate. This may be due to a variety of causes from what we have term masking through actual destruction of the endotoxin by harsh product ingredients like preservatives.
The presence of glucans has long been determined to be a nonclinically relevant analyte which can falsely activate the lysate clotting cascade. New lysate formulations as well as glucan blocking reagents have resolved these testing issues.
Do LAL reagents have sufficient sensitivity to meet your BET responsibilities?
Bolden: Yes. As current sensitivity capabilities are pushed, it may allow simpler interference removal options in the future.
Reinhard: Yes but borderline; for optical devices the assay should be more sensitive.
Sandle: Yes, lysates going down to 0.001 EU are sensitive enough, provided that interfering factors can be overcome to the point that the product can be tested within its endotoxin limit.
Farrington: Yes, we have found that our products with a sensitivity of 0.001 EU/mL meet both our and our customers BET needs. A sensitivity of 0.001 EU/mL is equivalent to testing to the level of 0.1 ppt. This sensitivity is often more than what is needed for samples tested without dilution. The availability though of LAL with such a high sensitivity allows one to dilute many folds to achieve a valid test using the BET.
Platco: The LAL reagent sensitivity far surpasses most chemical based analytical tests due to the activation pathway mechanisms. The sensitivity of the test allows one to dilute out sample interferences and retaining an ability to report a clinically relevant LAL value with several margins of safety in most cases.
Do you believe that BET labs should use quantitative BET, whenever possible, instead of over-the-limit testing?
Bolden: Quantitative testing enables the laboratory, quality and manufacturing the information needed to assess and mitigate atypical results ahead of a failing one, so yes, quantitative testing is advantageous over limit testing.
John Dubczak, Director of Operations, Charles River Laboratories: Quantitative LAL testing is unquestionably the preferred approach to any LAL lab. A quantitative test necessarily involves instrumentation. Software running the instruments analyzes and interprets the data. Reports are automatically created thereby providing efficiency to laboratory operations. With CFR compliant software, access is secured via passwords, user levels are defined and managed. Audit trails and data security verify all of the actions that have been taken. All of these objective measures are not possible with the manual semi-quantitative gel clot test.
Reinhard: Yes – better for case by case assessment in case of excursions.
Sandle: Yes, being able to quantify is important for it enables better trend analysis and trouble shooting. Being able to pick up, say, a problem with a water system through a gradual rise in endotoxin below the threshold is very useful and it enables a faster response.
Farrington: While quantitative BET methods allow for better trending and metrics, it is still important to maintain a well-documented, well defined and robust over-the –limit reagent as a gold standard.
Platco: Quantitative BET testing allows the user to detect contamination events long before the events become test failures. For this reason, it is valuable to test product at a noninhibitory concentration where the PPC controls are valid, yet the reported values are clinically relevant and meaningful in terms of manufacturing process control.
Have the FDA and other regulatory/advisory agencies provided meaningful insight and guidance to industry detailing what they need to see regarding endotoxin testing and detection studies?
Bolden: Yes, both the 2012 FDA Guideline and the very recent Ph.Eur 5.1.10 contain thoughtful insight and guidance derived from a broader or unique perspective that industry may not have. Regulatory agencies and compendias will admit that the time it takes to author or revise and publish guidance is significant, so it does take a certain amount of vision and willpower to move things forward. Agencies and industry should view each other as partners and collaborate on guidance; in this way, patient safety needs, deep technical expertise and practical experience will intersect to provide the best available guidance.
Reinhard: Partially – however, the knowledge / background between inspectors is quite different and thus discussions sometimes not on a professional, biochemical level. The expectations of the inspectors vary a lot.
Sandle: Although FDA put out their Q&A a few years back it didn’t completely replace their LAL guidance from 1987, which industry found very useful. In Europe there is no clear guidance from regulators at all. In short, there is a gap in terms of regulatory guidance. This extends to areas like sample size; number of representative finished products; low endotoxin recovery and so on.
One unresolved issue is the status of the Monocyte Activation Test (MAT) – a whole blood test for pyrogens – and how this fits in with general endotoxin testing.
Farrington: Several guidance’s are available providing detailed insight into BET testing. The most notable being the FDA’s Guidance for Industry, Pyrogen and Endotoxins Testing: Questions and Answers Guidance for Industry which supplements the topics not covered in the United States Pharmacopeia (USP) ChapterBacterial Endotoxins Test, USP ChapterMedical Devices – Bacterial Endotoxin and Pyrogen test, and the Association for the Advancement of Medical Instrumentation (AAMI) ST72:2011, Bacterial Endotoxins—Test Methodologies, Routine Monitoring, and Alternatives to Batch Testing. The majority of questions are resolved within these documents.
Platco: The BET tests are compendial tests and their execution, control, and sample qualification requirements are clearly spelled out in the compendial chapters. Also, the manufacturer’s product inserts for use are considered valid documents for execution of the testing.
The FDA has not provided industry with guidance for how much raw data they want to see in product filings. The degree of granularity and amount of data seems to vary between FDA divisions. I have observed regulatory submissions with charts of raw data for both samples and standards as well as LAL test results with %PPC recoveries only.
What is the role of technology and service providers to assist in the testing and detection of endotoxins? Are they viewed as partners? Are they driving the technology forward?
Bolden: The technology seems to have been limited by the conservative nature of industry, regulators and the interpretation of the pharmacopoeias. From a regulatory perspective, it helps to have consistent application of a stable technology; from an industry perspective compendia harmonization was widely welcomed for efficiency purposes. The downside is there is a real and perceived barrier to entry for a new technology outside the scope of the harmonized compendia BET chapter(s).
Dubczak: Because US LAL manufacturers are licensed by the FDA, we are the service providers and it is our responsibility to drive technology. Technology, however, has to have benefits that satisfy a true need within the industry. To paraphrase an old axiom, “necessity is the mother of invention”. An open dialog with the regulatory authorities as well customers is required in order to understand the industry needs and regulatory expectations.
Reinhard: We regard them as partner and expect that they drive the technology to the next level.
Sandle: I think the connection was strong a few years ago and industry leaders were regularly asked for their views. In recent years this seems to have dried up a little and I think the LAL industry needs to reconnect with the pharmaceutical sector and this will help to accelerate product development.
Farrington: At Associates of Cape Cod, Inc., we believe our role as your service provider should be a partner. Our technical, customer and contract test service departments take pride in working alongside you to develop method validation, determining what LAL reagent sensitivity is best for your product and resolving any issues that may arise. Our technical expertise is shared with our partners through onsite hands-on private workshops, webinars, seminars, onsite trainings for compounding pharmacies, consulting services, BEST® seminar, and public BET workshops as well as the LAL updates.
Platco: The lysate manufacturers are extremely valuable in helping the users understand the test mechanisms as well as the limitations of the test and testing capabilities. In many cases, there are some assumptions that industry has made which have been made clearer by the technology and service providers, each of which have their proprietary formulations and testing software. All US vendors have been quite open to discuss the pros and cons of their lysate formulations and testing techniques. Most are also interested in driving technology forward by making more robust systems and software for compliance to data integrity. One of the aspects of LAL testing that industry has not been clear about is that the product formulations from different test vendors of the lysate may react differently with product even though the same test “method” is being used.
What changes do you foresee in the next decade for endotoxin testing?
Bolden: The wider use of the recombinant Factor C reagent as an alternative to the traditional LAL reagent is coming, and provides end-users the opportunity to use a more sustainable reagent that offers equivalent or superior quality improvements. The harmonized compendia should incorporate it into the general test chapters sooner rather than later. Newer analytical and instrument technologies will improve the precision, accuracy and throughput of this biological test that has protected patient safety for over 30 years.
Dubczak: The next decade (and decades to come) the biomedical industry will create products that are targeted and highly personalized. Individualistic biotherapeutics, will require extremely sophisticated yet adaptive production processes. Towards that end, rapid microbiological methods and rapid endotoxin testing will be regularly used to ensure that the processes are controlled for a timely delivery to patients.
Reinhard: Midterm: NOE & NOE standards will become more important in endotoxin hold studies
Longterm: BET assay will be replaced by a straight forward pyrogen assay
Sandle: I think there will be a further push towards portability and portable test units. With larger, bench top LAL testing there will be developments in relation to robotics. Finally, with LAL reagents, the issue of horseshoe crab conservation and need for a robust recombinant product will come to the fore.
Cooper: Parenteral industries will increase their investment in automation for endotoxin testing. Automation is a critical strategy for reducing human error and maximizing objectivity in documenting test results. Errors may occur with test dilutions, particularly those relating to creation of a standard curve. The expense of automation is somewhat offset by eliminating the cost of human error. Automation increases lab efficiency by off-loading routine testing, such as robotic systems for management of water testing.
The LAL cartridge system will continue to grow because point-ofuse testing is enabled by its simplicity, flexibility and automated features. This handheld system eliminates tasks for manipulation of LAL reagents and standards and minimizes the principal sources of variability in the endotoxin-test lab. Weber and coworkers (American Pharmaceutical Review, Endotoxin Supplement 2014) reported that this rapid endotoxin testing platform was regulatory compliant while reducing sample handling, testing time and deviations. Further, implementation of rapid test methods upstream gave additional assurance of product quality and allowed timely processing decisions. Wireless capability for remote system access and data export will foster more decentralization of endotoxin testing.
The NOE (naturally occurring endotoxin) proposed by the USP will provide a powerful new tool for fully understanding the endotoxin risks in new products. As exemplified by the current controversy regarding LLR (low lipopolysaccharide recovery) in certain biologics, low recovery that affects LPS is far less likely to impact recovery of natural endotoxin. With increased awareness that LPS often fails to represent endotoxin that is present in a complex biopharmaceutical process, a well-characterized NOE of sufficient concentration may be added to starting production matrices and followed throughout a given process.
Farrington: We foresee improvements to the technology including hardware and software innovations to meet the required regulatory and laboratory management needs. In addition, we expect the endotoxin and glucan regulations to become more similar since modern pharmaceutical processes include a number of upstream and downstream unit operations that entail risk of BG contamination. Since binding of BG leads to diverse immune effects, burdens should be characterized, controlled, and safe levels should be established.
Platco: One of the most interesting topics which industry may need to resolve in the next decade is whether or not the presence or level of microorganism debris from many types of potentially viable sources is going to be considered an impurity that must be measured through some type of analytical assay. At the moment, there is no clinically relevant data nor widely published adverse reactions due to a potential plethora of non-viable bacterial or other debris that may be present but not able to be measured. It is not known if these potential “impurities” will have to be dealt with in raw materials, water systems, and ultimately finished products as we strive to manufacture purer and cleaner products that may contain foreign substances that we have not yet discovered.
On a more realistic path, I would expect to see that products which cannot be easily tested in the LAL test to obtain meaningful pyrogenic activity result will find their way to testing via the monocyte activation testing. At the present time, MAT testing is not as simple as the LAL test and has its own set of limitations and controls. I believe that MAT can replace the remaining rabbit pyrogen testing once a set of activity standards between the pyrogenicity of the product in the rabbit and the activity in the MAT are calibrated against each other. The difficulty is that this calibration should be performed for each product.