USP <1228.x>; An Evolving Series of Informational Chapters on Depyrogenation

• United States Pharmacopeia

Introduction

The first mention of depyrogenation in the United States Pharmacopeia (USP) can be traced back to an informational chapter that appeared in the Fifth Supplement to USP 20-NF 15 entitled, “Sterilization and Sterility Assurance of Compendial Articles” (USP, 1984). This chapter treated depyrogenation as a subset of sterilization because the focus was dry heat, and depyrogenation by dry heat will also sterilize. However, for other forms of depyrogenation and sterilization, processes are very diff erent and the terms are not synonymous. Since its original publication, the tenets of <1211>, most importantly the requirement that depyrogenation results in a three-log reduction of endotoxin activity for dry heat depyrogenation, have been adopted by both Food and Drug Administration (FDA, 2004) and Technical Report 3 (revised) from the Parenteral Drug Association (PDA, 2013)

In 2010, the USP General Chapters-Microbiology Expert Committee (EC) decided that it was time to incorporate the innovations made in the sterilization and depyrogenation sciences since 1984, and split these topics into two separate series of informational chapters. The <1229.x> series now consists of 15 subchapters describing various methods of sterilization. Recognizing that depyrogenation can take many forms, the EC decided to follow the example in the <1229. x> series example and provide guidance with respect to a number of common methods of depyrogenation in the evolving series of chapters. Table 1 is a listing of chapters in the <1228.x >series and their current status:

Table 1. Current Status of the <1228.x> series

In the process of defining the scope of the <1228.x> series, the EC recognized that advances in parenteral formulation and manufacturing, a renewed industry focus on raw material quality, a shift in active ingredients from small molecule to more complex biological entities, and the concepts of Quality by Design (QbD) and Risk Management all challenge the narrow and very prescriptive description and requirements for depyrogenation that were first proposed for dry heat in the original <1211>. One of the goals of the <1228.x>effort was to refl ect on the 30+ years of knowledge and experience gained since 1984, and where appropriate, propose some new ways to think about the design, execution and assessment of depyrogenation processes. Those topics listed in Table 2.

Table 2. Conventional thinking in <1211> and Alternative thinking in <1228.x>

Some definitions are helpful to the following discussion:

1. Depyrogenation is defined in <1228> as the direct and validated destruction or removal of pyrogens. For the purposes of the <1228>series, the term “depyrogenation” refers to the destruction or removal of bacterial endotoxins, the most prevalent and quantifiable pyrogen that may be found as a contaminant in parenteral preparations.
2. Endotoxins or Naturally Occurring Endotoxins (NOE) are components of the outer cell membrane of Gram negative bacteria, which are known to induce a febrile response in humans and other mammals and initiate the Limulus Amebocyte Lysate (LAL) cascade. The endotoxin complex contains many cell wall components including phospholipids, lipoproteins and lipopolysaccharides.
3. Lipopolysaccharide (LPS) is the biologically active portion of the endotoxin complex. LPS is an amphipathic molecule, with the hydrophobic Lipid A portion of the molecule buried within the outer leafl et of the Gram negative outer cell membrane and the hydrophilic O polysaccharide exposed to the cell’s external environment. Purified LPS will form aggregates in solution, with the extent of the aggregation dependent largely on the chemical composition of the matrix in which the LPS resides. Two preparations of purified LPS are important to the discussion: The USP Reference Standard Endotoxin (RSE) is the primary LPS standard, and Control Standard Endotoxins (CSE) provided by manufacturers of LAL reagents are generally used as secondary calibration standards and Positive Controls for the Bacterial Endotoxins Test. (USP, 2015d).
4. An Endotoxin Unit (EU) is a unit of activity of an endotoxin or LPS preparation.
5. An Endotoxin Indicator (EI) is an analytical tool, analogous to a biological indicator that may be used, where required or desired, in conjunction with any physical measurements needed to analyze the eff ectiveness of a depyrogenation process. An unprocessed EI is the positive control – an indicator that has not undergone depyrogenation. A processed EI is the test material – an indicator that has undergone the depyrogenation, regardless of the process technology.

Structure and Content of the <1228.x>series

The 1228.x series of chapters is structured to provide a robust introduction to depyrogenation, supplemented by individual chapters detailing specific methods of depyrogenation, testing, and control. Because endotoxins can only come from the normal growth cycle of Gram Negative Bacteria (GNB), the control of endotoxins as contaminants in parenteral products, water systems, raw materials, product contact materials and in process samples is really a matter of controlling conditions that would support the proliferation of Gram negative bacteria during all parts of the manufacturing process. Although the 1228.x chapters provide information on depyrogenation processes and validation, the series emphasizes that prevention and control of GNB must be the primary focus for avoiding endotoxin contamination.

<1228>, “Depyrogenation”

The Introductory chapter to the <1228.x> series provides an overview of depyrogenation, including a discussion of the choice of an appropriate depyrogenation method, principles of validation of a depyrogenation process, and routine process control. This chapter emphasizes the importance of understanding and controlling variability in any depyrogenation process to obtain consistent and accurate results. These variables include:

• The choice of challenge material (native endotoxins or purified LPS)
• Characteristics of the material being depyrogenated
• Levels of endotoxins activity needed to demonstrate depyrogenation in the study
• Preparation of test samples
• Endotoxin or LPS recovery methods
• Test method (gel clot or quantitative methods)
• Choice of depyrogenation method

The introductory chapter makes a clear distinction between native endotoxins (outer membrane vesicles or cell wall fragments that contain peptidoglycan, membrane proteins, phospholipids and lipopolysaccharides) and purified lipopolysaccharide. While the LPS molecule is the biologically active portion of the endotoxin complex, its behavior as an embedded, adaptable moiety of the living Gram negative cell envelope and its behavior as a highly purified chemical entity can be very different, depending on the extracellular matrix. The choice of analyte used for the depyrogenation study may have a significant impact on the outcome. Chapter <1228.5>, “Endotoxin Indicators” provides guidance on the choice of analyte and methods of preparing native endotoxins should the analyte of choice be a native preparation.

For any depyrogenation process validation, the log reduction is calculated using the formula

Log reduction = (Log₁₀ activity of the unprocessed EI) – (Log₁₀ activity of the processed EI)

For a study with an unprocessed EI activity level of 2947 EU and processed EI activity level of 0.529 EU, the log reduction would calculate to:

Log reduction = (Log₁₀ 2947) - (Log₁₀ 0.529) = (3.469) – (0.284) = 3.974

<1228> challenges the long held requirements that 1) 1000 of recoverable EU is the minimum de facto starting point for depyrogenation studies and 2) the acceptance criterion for any depyrogenation study is a reduction in endotoxins or LPS activity of at least three logs. For example, when looking at the endotoxin reduction efficiency of a postGram negative fermentation depyrogenation process where upwards of 108 EU/mL may be present, does it make sense to execute depyrogenation studies on the product stream by adding an additional 1000 EU of purified LPS? Likewise, where a product stream or incoming glass vials historically contain <1 EU/mL or <1 EU/vial, what is gained by loading the study with a level of endotoxin activity that is three orders of magnitude greater than what is naturally present?

What was the genesis of the 1984 three-log reduction requirement? The three log reduction was recommended because at that time, glass vials came packaged in cardboard and frequently contained cardboard dust that often resulted in large and variable quantities of endotoxin. OR, the sensitivity of the gel clot reagents at the time (maximum sensitivity 0.125 EU/mL), and the lack of availability of high concentration CSEs were analytical limiting factors. OR, perhaps the suggestion of the three log reduction in <1211> could have been arbitrary, perhaps another application of the “rule of three.”

Consistent with the principles of QbD and Risk Management, the EC believes that the more relevant and pragmatic indicator of depyrogenation efficiency should be based on process capability as it relates to patient safety. In other words, based on historical data, what is the “worst case” levels of endotoxins activity that could naturally contaminate the product, and what measurement of reduction capability is needed to achieve safe levels of endotoxins? Since there is no test currently available that can measure “0” endotoxins activity, a “safe” level for a product would be based on the most conservative product-specific endotoxin limit calculated from information in the package insert.

<1228.1>, “Depyrogenation by Dry Heat”

The only depyrogenation method discussed in the original <1211> was depyrogenation by dry heat. Dry heat inactivates endotoxins on heat stable materials essentially by incineration. Regardless of the depyrogenation technology used (batch oven or continuous tunnel), the effectiveness of this treatment is totally dependent on four basic parameters: run time, run temperature, load configuration, and the use of qualified/calibrated equipment. If one of the four basic parameters changes, the change must be assessed via the change control process to determine whether re-validation is required. Once validated, if there is no change to load configuration, run time, run temperature or equipment qualification, the chapter suggests that the use of endotoxin indicators may not be a value-added component of the re-validation effort.

<1228.3>, “Depyrogenation by Filtration”

The depyrogenation of product streams including LVPs and antibiotics by charge modified filtration and ultrafiltration were first described about 40 years ago (Nolan, et al, 1975; Sweadner, et al, 1977). Since that time, the technology has advanced and the use of various methods of filtration and chromatography, especially for depyrogenating complex biological molecules, have become commonplace. Chapter <1228.3> describes the following methods of filtration and chromatography: microporous membrane filtration, reverse osmosis, ultrafiltration, charge modified depth filters (endotoxins and LPS are negatively charged) and membrane adsorbers (ion exchange chromatography). The chapter also suggests that depyrogenation of a product stream may require combinations or series of technologies to assure consistently safe levels of endotoxin activity.

Although not originally defined, it is assumed that these methods of depyrogenation by filtration will reduce endotoxin activity by at least three logs. But what is the endotoxin indicator for product streams? For depyrogenation of liquids, endotoxins or LPS is added upstream of the filtration process, so the spiked product itself becomes the unprocessed endotoxin indicator. Product downstream of the filtration process is the equivalent of the processed indicator. Log reductions are calculated by measuring both upstream and downstream activity of the endotoxins and using the formula provided above (Introduction).

<1228.4>, “Depyrogenation by Rinsing”

Rinsing in hot highly purified water is the most common way of eliminating or reducing endotoxins activity on heat labile materials such as elastomeric closures and plastics. However, there are a number of variables that must be strictly controlled to assure success:

• The rinsing water must be of consistently high quality, and must be low in endotoxins. Water for Injection (WFI) is most commonly used.
• The water temperature is important (>60°C), and must be carefully controlled throughout the process.
• Load size must be identified.
• Once rinsed, materials must be dried prior to use or storage, as moist conditions may provide an environment for Gram negative organisms to proliferate, which could result in an endotoxin re-contamination event.
• Once validated, all equipment and processes must remain in a validated state, meaning control over load size, physical measurements (volume/quality/temperature of water used to rinse), preventive maintenance/calibration of all equipment and periodic measurement of the endotoxin levels on incoming materials to assure that baselines have not shifted.

<1228.5>, “Endotoxin Indicators”

Endotoxin Indicators were first described by the LAL Users’ Group in 1989 (Users’ Group, 1989). An endotoxin indicator (EI) serves the same purpose in a depyrogenation study that a biological indicator (BI) serves in a sterilization study, but the target of the action in an EI is purified lipopolysaccharide or native bacterial endotoxins, not bacterial spores. However, in principle the EI is the same as a BI: The activity level of the LPS or endotoxin on the EI, as measured by the Bacterial Endotoxins Test (BET), is compared before processing and after processing, and a log reduction is calculated.

Endotoxin Indicators are usually thought of as glass vials “spiked” with purified LPS. However, anything can be an endotoxin indicator. If the depyrogenation load is a dry heat process for tools for use in the cleanroom, wrenches can be spiked and used as endotoxin indicators. For liquids, the upstream volume can be spiked with endotoxin and that becomes the indicator. For rinsing processes used for heat labile materials (e.g. rubber stoppers), a representative sample of the material is spiked and is considered to be the endotoxin indicator.

Given the discussion of “Low Endotoxin Recovery” over the past five years, the EC thought carefully about the analytes (native endotoxins or purified LPS) that might be used for depyrogenation studies (Bolden, et al, 2015). While the standard method is to use purified LPS, we know that purified LPS cannot always be recovered with 100% efficiency from many undiluted products, and we know that this recovery efficiency can be reduced even more if the LPS is held in contact with the product for even a short period of time. If we detect little or no measureable activity post filtration on one of these products, how do we know if the reduction was due to the process or the impact of the product matrix? We don’t. In order to separate depyrogenation from “LER” when looking at product streams, the EC suggests that an endotoxin that is stable and can be recovered upstream of the treatment may be used as the analyte. Chapter <1228.5> provides guidance on the production and control of such preparations. Three other chapters are in progress. These chapters address the topics Chemical Depyrogenation, Endotoxin Control and Monitoring, and a chapter of the deyprogenating efficiency of methods such as moist heat and ionizing radiation.

Summary

The EC’s decision to create a series of chapters on depyrogenation resulted from over thirty years of knowledge, experience and technological advances in manufacturing process and control, including the control of endotoxin contamination. Our proposals for alternate thinking rely heavily on the principles of QbD and Risk Management to reduce risk and implement prudent continuous control measures. Our focus is on data to demonstrate patient safety, and less on using traditional approaches to “cookie cutter” depyrogenation validation and monitoring. We hope that the release of current and future chapters will encourage industry and regulators to engage in an evolving paradigm on depyrogenation.

References

1. Bolden, Jay, Cheryl Platco, John Dubczak, James F. Cooper and Karen Zink McCullough. 2015. Stimulus to the Revision Process: The Use of Endotoxin as an Analyte in Biopharmaceutical Product Hold Time Studies. Pharmacopeial Forum 41(5).
2. Food and Drug Administration, United States Department of Health and Human Services. 1987. Guideline on Sterile Drug Products Produced by Aseptic Processing.
3. Food and Drug Administration, United States Department of Health and Human Services. 2004. “Guidance for Industry: Sterile Drug Products Produced by Aseptic Processing: Current Good Manufacturing Practice.”
4. LAL Users’ Group. 1989. Preparation and Use of Endotoxin Indicators for Depyrogenation Process Studies. J Parent Sci Tech. 43(3): 109-112.
5. Nolan, J.P., J.J. McDevitt, G.S. Goldmann. 1975. Endotoxin binding by charged and uncharged resins. Proc. Soc. Exp. Biol. Med. 149: 766-770.
6. Parenteral Drug Association. 2013. Technical Report 3, “Validation of Dry Heat Processes Used for Depyrogenation and Sterilization.”
7. Sweadner, Kathleen, M. Forte Lita L. Nelsen. 1977. Filtration Removal of Endotoxin (Pyrogens) in Solution in Different States of Aggregation. Applied and Environmental Microbiology. 34(4): 382-385
8. United States Pharmacopeia. 1984. <1211>, “Sterilization and Sterility Assurance of Compendial Articles”
9. United States Pharmacopeia. 2017a. <1228>, “Depyrogenation”.
10. United States Pharmacopeia. 2017b. <1228.1>, “Dry Heat Depyrogenation”.
11. United States Pharmacopeia. 2017c. <1228.3>, “Depyrogenation by Filtration.”
12. United States Pharmacopeia. 2017d. <1228.5>, “Endotoxin Indicators for Depyrogenation”.
13. United States Pharmacopeia. Pharmacopeial Forum. 2017. <1228.4>, “Depyrogenation by Rinsing. 43(5)