TECHNOLOGY/APPLICATIONS

Preparation and Use of Endotoxin Indicators for Depyrogenation Process Studies

LAL Users GroupL>

Preface charged membrane surfaces, hydrophobic attraction, or various other methods. Depyrogenation by inactivation on This article was developed by a subcommittee of the heat stable surfaces is usually accomplished by long expo­ LAL Users Group, an organization of scientists dedicated sure to high temperature, and otherwise by acid/base to the development and application of Limulus Amebo­ hydrolysis, oxidation, alkylation, moist heat, or ionizing cyte Lysate technology. Comprised of members from vari­ radiation. There are several excellent references available ous pharmaceutical and medical device manufacturers which include detailed discussions on the particulars of and other related industries, the primary objective of th~ each method (1-3). group is to guide the membership into the routine usage of Verification of dry heat depyrogenation may be sup­ LAL methodology. This includes the coordination of com­ ported by data showing the relationship of thermal treat­ pendia! requirements, vendor ;user relationships, routine ment to endotoxin inactivation. Studies by Tsuji, Robert­ handling procedures, and reagent evaluation. son and co-workers ( 4-7) demonstrated that the inactiva­ LAL Users Group Endotoxin Indicator Subcommittee: tion kinetics of lipopolysaccharide derived from various John Shirtz (Chairman), Burroughs Wellcome Co. gram-negative organisms were nearly identical to each Ed Seelig, Wyeth/ Ayerst Laboratories other, and that dry-heat D-values of intact whole cells at Karen McCullough, LAL Consultant 210°C could be calculated. They further demonstrated Betsy Waldheim, Organon, Inc. that the inactivation curves for purified endotoxin could Tim Schimmel, Merck & Co. be made linear and that inactivation for any dry heat Joan Craig, Smith Kline & French Laboratories process could be predicted given the product heating John Dubczak, Baxter Healthcare curve. These studies served to demonstrate the correlation between purified endotoxin and whole cells and helped Background establish the foundation for using purified endotoxin in depyrogenation studies as a predictor of whole cell (both The introduction of endotoxin from gram-negative bac­ intact cells and cellular debris) inactivation. teria into the bloodstream has long been recognized as a Akers and co-workers (8, 9) carried this calculation potentially serious condition. These bacterial endotoxins, or pyrogens. are fever-producing substances present in the concept one step further by developing F-value require­ ments for the destruction of endotoxin. This approach is outer cell membrane which retain their potency upon lysis also detailed in the Parenteral Drug Association Techni­ of the cell. The health industry employs production proce­ cal Report #3 (10). In a more recent study, Ludwig and dures that include a means for depyrogenation to insure that parenteral products, medical devices. and compo­ Avis (II) reported F-values which were significantly dif­ ferent from those reported by Akers, and concluded that nents are free of these pyrogens. In addition, quality con­ substantial endotoxin reduction can be achieved using trol procedures usually include analytical methods for temperatures as low as 170°C. These experiments, howev­ process validation and monitoring to assure that pyrogens er, utilized a specially designed heating cell with precision have been eliminated during processing. Depyrogenation is defined as the elimination of all py­ control and were not intended to mimic a routine produc­ rogenic substances, including bacterial endotoxin, and is tion depyrogenation cycle. None of these studies, howev­ er, includes any detailed technique for preparing and re­ generally achieved either by removal or inactivation. The covering endotoxin challenges analogous to the traditional specific method to be used is determined by both the Biological Indicator which has been a valuable tool for physiochemical nature of the endotoxin and the nature of many years in sterilization validation and monitoring. the material to be depyrogenated. Depyrogenation by re­ The employment of an endotoxin challenge in depyro­ moval may be accomplished by reverse osmosis, ultrafil­ genation validation is an important complement to ther­ tration, affinity chromatography, dilution, through ad­ mal profile monitoring. In their chapter on Sterilization sorption to a specfic medium such as charcoal or positively and Sterility Assurance of Compendia! Articles, USP ( 12) recommends an endotoxin challenge be included with Received October 17, 1988. Accepted for publication January 19 any dry heat process intended to render glassware or 1989. . containers free of pyrogens. They also note that the depyr­ "Author to whom correspondence should be addressed: Mr. John T. Shirtz. Burroughs Wellcome Co .• P.O. Box 1887. Greenville. NC 27834- ogenation process should demonstrate a minimum 3 log 1887. cycle reduction in the difference between the recoverable

Vol. 43, No. 3 I May-June 1989 109 endotoxin level and the residual endotoxin level.' The but may also be performed with an endotoxin-free pipet.· specifics of preparation and recovery. however. are left up syringe, or with any other similar equipment which can to the user. accurately and consistently deliver the desired volume. It Considering the numerous variables that the prepara­ has been suggested that the use of a surfactant may also be tion, handling. and evaluation of an endotoxin challenge included in the inoculum diluent if it has been demonstrat­ can involve. the following is presented as a guide to facili­ ed to enhance recoverability of the endotoxin yet not tate the use of such challenges. The primary application of interfere with the depyrogenation process. the challenge, or "Endotoxin Indicator," would be for dry It is generally recognized that the number of recover­ heat depyrogenation, but the concepts could be applied able endotoxin units may vary from the theoretical inocu­ and/or modified for various other endotoxin reduction lated level, depending on a variety of factors such as techniques. endotoxin concentration in the original inoculum (the more potent the inoculum, the higher the recovery), glass Endotoxin Indicator type and shape, drying method. recovery process, etc. Drying or fixing of endotoxin onto representative chal­ An Endotoxin Indicator is a carrier to which a known lenge components may be conducted in one of several amount of endotoxin is added, and is used to measure ways. The easiest is to subject the inoculated units to endotoxin reduction or inactivation in a depyrogenation ambient laminar flow, although the use of an elevated process. It serves i.n a capacity similar to a Biological temperature (not exceeding 50°C), such as in an incuba­ Indicator in a sterilization process and is used to further tor, may be preferred to hasten the process. Vacuum dry­ demonstrate heat penetration and distribution data estab­ ing or lyophilization have been reported to be very effi­ lished by physical measurement. Endotoxin Indicators are cient fixation methods for both the quick drying of endo­ prepared from a purified lipopolysaccharide which has toxin onto challenge units, and for excellent recovery of been adequately characterized against the current Refer­ endotoxin once fixed ( 13, 14 ). ence Standard Endotoxin (RSE). This material may be obtained through a number of sources including licensed Preliminary Testing lysate manufacturers or bacteriological media distribu­ tors as a lyophilized stock. It is incumbent upon the user to confirm the validity of the Endotoxin Indicator before its use as a tool for the Carrier evaluation of routine depyrogenation processes: 1) The endotoxin concentration both in the spiking Different surfaces and materials may contribute in solution (the "initial spike") as well as in the dried varying degrees towards the effectiveness of the depyro­ reconstituted control vial ("recoverable endotoxin") genation process. Therefore, the unit which serves as a must be determined. It is also recommended that carrier for the endotoxin should be the same as or chemi­ each laboratory determine the percent recoverable cally similar to that which is being depyrogenated. A endotoxin from each type of container used as an universal carrier, such as a small colored vial or planchet, indicator, as recovery may vary widely depending on may be developed for numerous applications provided the the chemical composition and configuration of the thermal characteristics of the universal carrier have been unit employed (13, 14). shown to be at least as resistant to the depyrogenation 2) A key issue in the fixing process remains the ques­ process as the item being studied. Depyrogenation of the tion of whether any additives, such as surfactants carrier prior to its use ( 12) is necessary to minimize the used in the spiking solution to increase endotoxin possibility of false positives due to intrinsic endotoxin recovery, are likely to enhance the depyrogenation contamination. All inoculated units must be clearly iden­ capability of the process being studied (thus in a tified for easy retrieval after processing. false negative sense enhancing the overall outcome of the validation exercise on the challenge units Inoculation only). This concern would usually impact only those The lyophilized endotoxin should be reconstituted and processes wherein the properties of the additive may diluted with a sufficient amount of endotoxin-free water be effective. Comparative studies could be per­ so as to provide a concentration of at least 20,000 EU / mL. formed both with and without the additive to ex­ This figure is suggested in order to achieve I 000 EU in 0.1 plore for this possibility. mL assuming a 50% recoverable endotoxin, and may be 3) Some carrier surfaces may leach interfering sub­ adjusted up or down. The diluted endotoxin solution must stances during long exposure to high temperature. always be thoroughly vortexed prior to use. As a preliminary screen, each type of carrier surface To aid in the fixation process, the challenge inoculum may be tested for interference by adding a known volume should be as small as practically possible; a vol­ amount of endotoxin in solution to unspiked, pro­ ume of 0.1 mL or less is usually adequate. Inoculation cessed carriers. These carriers are not dried, but are itself is best accomplished using a calibrated mechanical subjected to the same extraction and recovery pro­ pipetter fitted with a disposable endotoxin-free pipet tip, cesses as any other carrier. If the amount of recover­ able endotoxin is within one two-fold dilution of the 1 Endotoxin reduction by methods other than dry heat may not neces­ endotoxin spike, it may be assumed that interfer­ sarily achieve a 3-log cycle reduction. ence, if any, is minimaL

110 Journal of Parenteral Science & Technology · There is currently some discussion in the industry re­ developed and used to expeditiously extract the endotoxin garding the use of wet vs. dry Endotoxin Indicators in from the Endotoxin Indicator. For most applications, sev­ depyrogenation studies. In the case of a continuous wash/ eral minutes treatment by either or both devices should be dry fdepyrogenation process, the argument may be made adequate (13). that the units (for example, vials) enter the depyrogenat­ The extracted solution should then be tested for the ing apparatus while still wet, and therefore the use of a wet presence of bacterial endotoxin by the Limulus Amebo­ indicator is appropriate. However, if the vial processing is cyte Lysate (LAL) test. The relative simplicity and speci­ considered in its entirety, another approach may be to ficity of the LAL test makes this method an ideal candi­ prepare dry Endotoxin Indicators, and introduce them to date for evaluation of the Endotoxin Indicator. Many the wash/dry fdepyrogenation process as any other raw more challenge units can be tested in less time than with material vial. In this case, the cumulative effect of the the traditional rabbit test, and semi-quantitative or quan­ entire continuous cycle to eliminate endotoxin will be titative comparisons using processed vs. non-processed evaluated, more closely approximating the "real life" pro­ units can be performed. In addition, unlike the manipula­ duction situation. Where a continuous process for the tions necessary for testing many inhibitory pharmaceuti­ preparation of vials is not employed, it is likely that only cal products, the Endotoxin Indicator uses (non-inhibi­ the heat treatment of the vials will be evaluated for endo­ tory) endotoxin-free water which generally requires no toxin reduction. In this situation, the use of dry vials might special pre-treatment. also be considered, as the small volume of liquid in a wet Extracts may be tested for residual endotoxin by one of vial would evaporate well before the depyrogenating tem­ the following LAL methods: perature is reached, leaving a dry vial to be subjected to I) The gel clot test may be used as a limits test, and the the process. results reported as

Vol. 43, No. 3 I May-June 1989 111 Acceptance Criteria Processed An Endotoxin Indicator which has un- · Indicator dergone a depyrogenation procedure. For the laboratory analysis to be considered valid: Non-processed An Endotoxin Indicator which has not l) All criteria for a valid LAL test must be met ( 15 ). Indicator undergone a depyrogenation proce­ 2) The positive control must demonstrate a minimum dure. 1000 recoverable Endotoxin Units (EU), and Recoverable The number of Endotoxin Units which 3) The negative control should not exhibit any measur­ Endotoxin can be detected in a non-processed able level of endotoxin. Endotoxin Indicator. Residual The number of Endotoxin Units which Summary Endotoxin can be detected in a processed Endo­ Biological Indicators (B.I.'s) have traditionally been toxin Indicator. employed in the validation and routine monitoring of ster­ Pyroburden The resident level of naturally occuring ilization processes used for the manufacture and control endotoxin on an item prior to depyro­ of pharmaceuticals and medical devices. In this paper, the genating processing. phrase "Endotoxin Indicator" has been coined to describe a tool analogous to the Bl which can be used in the References validation and routine control of endotoxin reduction pro­ I. Weary, M .. and Pearson. F .. "A manufacturer's guide to depyrogen­ cesses. Like the BI.'the Endotoxin Indicator provides the ation," Biopharm .• l, 4 ( 1988). 2. Pearson, F. C. Pyrogens. Endotoxins. LAL Testing. and Depyrogen­ user with an in vitro biological test to complement physi­ azion, Marcel Dekker, Inc .. New York, 1985. cal measurement used to control the manufacturing pro­ 3. Parenteral Drug Associaiton. "Depyrogenation." Technical Reporz cess. No. 7 ( 1985). 4. Tsuji, K .. and Harrison. S. J.. "Limulus Amebocyte Lysate-A It must be remembered that the total pyroburden in an Means to Monitor Inactivation of Lipopolysaccharide," in Biomedi­ aseptically processed drug is the sum of the pyroburden of cal Applicazions of zhe Horseshoe Crab (Limulidae), Cohen E .. each factor contributing to the manufacture of that prod­ Ed .. Alan R. Liss. New York. 1979. 5. Robertson. J. H., Gleason. D., and Tsuji, K .. "Dry-heat destruction uct. This includes raw materials, packaging components, of lipopolysaccharide design and construction of dry-heat destruc­ the environment and the manufacturing process itself. tion apparatus." Appl. Em•iron. Microbial.. 36,705 ( 1978). Therefore, all aspects of the process must be considered 6. Tsuji, K .. and Harrison, S. J.. "Dry-heat destruction of lipopolysac­ charide: Dry-heat destruction kinetics," App/. Em•iron. Microbial .. for a complete validation. 36, 710 ( 1978). This pyroburden may also change during processing of 7. Tsuji, K .. and Lewis. A. R .. "Dry-heat destruction of lipopolysac­ a drug. Fluctuations in pH, solvent content, and tempera­ charide: Mathematical approach to process evaluation," Appl. En­ t•iron. Microbial.. 36, 715 ( 1978). ture may decrease pyroburden while introduction of some 8. Akers. M. J.. Avis. K. E .. and Thompson, B.. "Validation studies of raw materials, especially water, may increase endotoxin the Fostoria Infrared Tunnel Sterilizer," J. Parenzer. Drug Assoc., content. (16) Regularly scheduled monitoring of all asep­ 34, 330 ( 1980). 9. Akers, M. J .• Ketron, K. M., and Thompson, B. R., "F Value tic processes should identify and address each step of the requirements for the destruction of endotoxin in the validation of process considered to have potential for the possible intro­ dry-heat sterilizationfdepyrogenation cycles," J. Parenrer. Sci. duction or removal of endotoxin in the final product. Most Techno/., 36, 23 ( 1982). 10. Parenteral Drug Association. "Validation of dry heat processes used of these situations can be thoroughly studied using Endo­ for sterilization and depyrogenation," Technical Report No. 3 toxin Indicators to challenge routine production condi­ ( 1981). tions, thereby providing insight to the assurance of endo­ II. Ludwig. J. D .. and Avis. K. E .. "Validation of a heating cell for precisley controlled studies on the thermal destruction of endotoxin toxin-free final products. in glass," J. Parenzer. Sci. Techno[ .. 42, I ( 1986). 12. U. S. Pharmacopeia XXI. United States Pharmacopeia! Conven­ Glossary tion. Inc .. 1985, p. 1349. 13. Novitsky. T. J., Schmidt-Gengenback, J., and Remillard, J. F., D-value Process time at a specified temperature "Factors affecting recovery of endotoxin absorbed to container sur­ faces," J. Parenrer. Sci. Techno/., 40, 6 ( 1986). which will reduce a microbiological 14. Jensch, U. E .• Gail, L., and Klavehn. M .. "Fixing and Removing of challenge by 90%. Bacterial Endotoxin from Glass Surfaces for Validation of Dry F-value Mathematically derived calculation of Heat Sterilization," in Detection of Bacterial Endotoxins With the Limu/us Amebocyte Lysate Test, Alan R. Liss, New York, 1987. the equivalent time relative to a refer­ I 5. "Guideline on validation of the Limulus Amebocyte Lysate Test as ence temperature. an end-product endotoxin test for human and animal parenteral Initial Spike The number of Endotoxin Units added drugs, biological products, and medical devices," Food and Drug Administration, Dec. 1987. to an Endotoxin Indicator prior to a 16. McCullough, K. Z., "Process control: In-Process and raw material depyrogenation process. testing using LAL." Pharm. Techno/. 12.5 ( 1988).

112 Journal of Parenteral Science & Technology