In this issue:DepyrogenationCalendar

Volume 11, No.5

p~LALUPDATE®

December 1993

Depyrogenationby Michael E. Dawson, Ph.D.

Dear LAL User,

This issue ofthe LAL UPDATEreturns to a subject we havecovered in several previous issues ofthis publication (LAL UPDATEvst», No.1, Vo/.6 No.2, and Vol.8,No.1). The current article by Dr.Michael E. Dawson addresses thevalidation ofdepyrogenation inmore detail than the earlier ones.

As Associates ofCape Cod, Inc.enters its 20th year in the LALbusiness and the LAL UPDATE getsready to celebrate its lith birthday,users should expect to see somechanges in the newsletter. With thenext publication of the LAL UP­DATE another publication, the LALUPDATE International Edition willmake its debut. The two publicationswill be identical except that theInternational Edition will alsoinclude international market issues,such as changes in the European orJapanese Pharmacopeias. Thecalendar section and new productservice announcements will also bemore territory-specific. For those ofour readers who need to follow LALregulations abroad, I suggest youwrite us for a subscription to bothissues. We also hope to publish theInternational Edition in Spanish,German and French.

Sincerely,

~~~/~~Thomas J. Novitsky, Ph.D.Editor

Depyrogenation is the removal or de­struction of pyrogens, and particularlyendotoxin. Endotoxin removal from anarticle is validated by showing a multi­log reduction in the endotoxin load froman initial concentration. Gram-negativebacterial endotoxin is the most signifi­cant pyrogen in most situations and isalso the most refractory. Since condi­tions required to destroy endotoxin willdestroy other pyrogens, the term depyro­genation is appropriate.

Depyrogenation can be accomplishedin a number of ways, but thermal destruc­tion of endotoxin is the most commonand effective method for heat stable arti­cles. This LAL UPDATE addresses thequestion of depyrogenation by dry heatand its validation.

In order to set up a depyrogenationprocedure, an appropriate temperatureand exposure time must be determined.Temperatures in excess of 180°C effec­tively destroy endotoxin. Below 180°C,depyrogenation may be incomplete evenafter extended periods. The time requiredto achieve a given level of endotoxin de­struction decreases as temperature in­creases, e.g., a three log reduction re­quires 65.4 minutes at 190°C but only1.5 minutes at 250°C (Tsuji and Lewis,1978). Unfortunately, different rates ofendotoxin destruction have been foundin different studies. For example, basedon results from three studies, Avis et al.(1987) cite times of 10,130 and 500 min­utes for equivalent destruction of endo­toxin at 250°C. The USP BET states

"such as 250°C or above for sufficienttime" and references the informationalchapter <1211> "Sterilization and Ste­rility Assurance of Compendia! Articles".

Depyrogenation studies have adoptedconcepts and terminology fromsterilization and an excellent practicalmonograph published by the PDA dealswith both topics. (Validation of Dry HeatProcesses Used for Sterilization andDepyrogenation, PDA Technical ReportNo.3, PDA, Philadelphia, PA). Threeparameters appropriated from the theoryof sterilization that are frequentlyencountered in the discussions ofdepyrogenation are D value. z value andFvalue.

D value: the time required to give aone log (90%) destruction of endot­oxin at a given temperature - thus atime of 3D will give a three log reduc­tion.

z value: the temperature incrementrequired to change the D value by onelog. It requires the D value to beknown at two temperatures and as­sumes linear kinetics for the destruc­tion of endotoxin.

F value: the time required to giveequivalent destruction at differenttemperatures. An F value is specificfor a stated degree of destruction, forexample, two log or three logs. Fvalues are calculated using a referencetemperature and a z value.

'G) 1993. Associates of Cape Cod. Inc.. Woods Hole. Massachusetts All rights reserved

These parameters are useful for theoret­ical discussions ofendotoxin destructionand particularly for comparisons madewithin a single study. However, signifi­cant differences between the measuredparameters are evident when studies arecompared. Ludwig and Avis discussmarked discrepancies between F valuesin different studies. These authors sug­gest that a primary cause for the discrep­ancies is differences in the control of thesample temperature, the rate of heatingand rates of introduction and removal ofsample. Another difference can be thetype of vessel being depyrogenated(endotoxin carrier). These authors havealso demonstrated clearly that differentendotoxin preparations (particularlythose with fillers) show markedly differ­ent rates of depyrogenation. Finally, Tsujiand Harrison (1987) report that the de­struction of endotoxin does not follow thesimple logarithmic decline exhibited insterilization studies on spore suspensions.Hecker and coworkers (1993) reiteratethis point and state that two D values arerequired to describe endotoxin destruc­tion. If it is not possible to determine aD value that is valid for the whole curve,z and F values are similarly uncertain.These authors also recommend that endo­toxin preparations with fillers should notbe used for depyrogenation studies be­cause of their relative ease of destruction.

The discrepancies surrounding pub­lished values for these parameters pro­vide little guidance for selection of a tem­perature/time regime for depyrogenation.However, Tsuji and Lewis (1978), Lud­wig and Avis (1990), and Hecker et al.(1993) are in general agreement that athree log destruction of endotoxin is ac­complished in two minutes or less at 250 0

C. Thus, the "classical regime" of250°Cfor 30 minutes offers a high level of as­surance of at least a three log reduction.This regime also has the advantage ofprecedent, having been widely used inthe industry for many years.

In order to ensure depyrogenation, itis important to note that every article inthe oven must be exposed to at least thestated temperature for no less than thestated time. The total time over whichthe oven is operating will always belonger than the time stated at temperature

LAL UPDATE Vol. II. Number 5, page 2

to allow for the coldest part of the loadto heat up.

Validation of Dry HeatDepyrogenation Processes

Before validation can begin, theprocess temperature and time must beselected. It is common to add a safetyfactor of either time or temperature. Atleast three hours at a minimum of 180°Cor 30 minutes at 250°C are commonlyused. Consider the type of material tobe depyrogenated. For example, silicontubing can be effectively depyrogenatedby dry heat, but temperatures of 250°Cmake it brittle and prone to split or break.Extremes oftime and temperature shouldbe avoided. An hour at a processtemperature of 250°C is reasonable.Caution: excessive periods at hightemperatures are at best a waste of energyand may actually damage materials beingdepyrogenated.

An important consideration is whetherthe process is being validated for all typesof articles that might be introduced intothe oven, or whether depyrogenation ofa particular article is being validated.Generally the process is validated for alltypes of articles. It is therefore importantto perform the validation using worst caseconditions, which usually translates tothe greatest possible load. However, itmay be necessary to validate a secondtime/temperature regime for particulararticles, such as the silicon tubingmentioned above.

Validation of endotoxin destructionunder the selected conditions requirestwo phases. First, the physical heatingcharacteristics of the oven must be es­tablished and cold spots identified. Then,the proposed cycle must be challengedwith endotoxin. At least a three log re­duction of endotoxin must be demonstrat­cd.

Phase 1Use calibrated temperature probes

(thermistors or thermocouples) in vials/vessels throughout the oven under max­imum load. Include probes in the mid­dle of the load. A common minimumprobe configuration is an X pattern offive probes on the top, middle and bot­tom shelves. Additional probes may be

necessary in large ovens while fewer maybe justified for small ones. Use calibrat­ed multichannel recorders to record thedata. Run the oven to identify the coldspots and to determine the time requiredto reach temperature at that point in theload. It may be necessary to try differentloads to determine the one with the slow­est heating to ensure that the validationis conducted for the worst case. Deter­mine the run time and temperature toensure that the process temperature ismaintained for the required time, with asafety margin.

Depyrogenation that takes place dur­ing heat up and cool down is not consid­ered when the process time is selected,thus adding a safety factor of time. Also,the oven temperature may be set abovethe selected process temperature in or­der to reduce the time for the coldest partof the load to reach the required temper­ature. In this case, the temperature inthe oven will continue to rise above thestated process temperature until equilib­rium is reached, adding a safety factor oftemperature.

The advantage of setting the oven tem­perature higher than that required fordepyrogenation is illustrated in FiguresI and 2. For example, with the oven tem­perature set to 250°C, the cold spot doesnot reach the process temperature of250°C for more than one and a halfhours(Figure I). With the oven temperatureset at 300°C the cold spot in the loadreaches temperature in less than I hour(Figure 2).

Phase 2Prepare or obtain challenge articles to

demonstrate that the process time andtemperature are sufficient to effect at leasta three log destruction of endotoxin.These can be prepared by adding a smallvolume of a high endotoxin concentra­tion to give at least 1000 EU (stated inUSP chapter <1211» to challenge arti­cles. Sufficient endotoxin should be add­ed to give 1000 EU recoverable. If 10%is the lowest acceptable recovery, an ad­dition of 10,000 EU is appropriate.

Associates of Cape Cod, Inc. 's highpotency control standard endotoxin (CSEcatalog number 800-3) is intended for usein depyrogenation studies. Each vial

Figure 2. Equilibration of oven thermostat and cold spot of an oven at 300°C

Figure 1. Equilibration of oven thermostat and cold spot of an oven at 250°C

Ensure that the test or reagent is sensi­tive enough to detect at least a three logreduction from the measured endotoxinconcentration in the untreated vialextract. The endotoxin detected in thetreated article extract should be at leastthree logs less than that in the extractfrom the untreated controls. A four logreduction is recommended. If the gel-clotmethod is used, an endpoint test can beused such that, if the extract testsnegative, at least a three log reductionhas been demonstrated. Only if a positiveresult is obtained, is it necessary to test aseries of dilutions to determine whetherat least a three log reduction has beenobtained. Include positive productcontrols for the processed articles tocontrol for the possibility that negativeresults are due to inhibition rather thanabsence of endotoxin.

300 Oven Set at 250·C

250

200

Temp.

('C!

100

25

0.5 Time (hr) 1.5

Oven Set at 300"C

300

250

200

Temp.

rei

100

25

0.5 1.5Time [hr]

If recoveries are very variable, a largernumber should be tested.

Recover endotoxin from the untreat ­ed and heat-treated articles by adding aknown volume of LAL reagent water(LRW). A volume of 1 ml is convenientand simplifies calculations. Endotoxinrecovery can be effected by vigorousvortexing or sonication. Usuallyvortexing in the same way as is used toreconstitute a vial of CSE is sufficient,i.e., vortex for one minute initially andthen every 10 minutes for 30-60 minutes.If recovery is poor e.g., less than 20%,the period of vortexing can be extendedor sonication can be attempted. Novitskyet al. (1986) reported on the effectivenessof recovery of endotoxin from differentsurfaces by vortexing and sonication.

Perform an LAL test on the extracts.

contains 125~g endotoxin, or 25 ug/rnlwhen the contents are reconstituted in 5ml. Assuming a potency of 10 EU/ng,this is equivalent to 250,000 EU/m!.Thus 40~1 will contain 10,000 EU andcan be used to spike the challenge arti­cles. Dry the added endotoxin on thearticles (air dry or lyophilize). Suitablearticles include glass vials or ampoules.A vessel that allows a recovery volumeof 1 ml is convenient. Test the recoveryof added endotoxin. Recovery of at least10% of the added endotoxin is desirable.If depyrogenation of large vessels is tobe validated, challenge vials can beplaced inside the large vessel.

A paper published by the LAL Users'Group (1989) discusses the preparationof endotoxin indicators (challenge arti­cles) from vials that can then be used indepyrogenation studies. Using amberglass makes it easy to find the challengevials in a load. The authors note that ifchallenge articles are to be stored for lateruse, it should be demonstrated that thearticles do not lose potency during stor­age.

Alternatively, vials of lyophilizedendotoxin of an appropriate concentra­tion can be used directly as challenge ar­ticles. Associates of Cape Cod, Inc. 's0.5~g endotoxin/vial, (catalog number800-1) can be used directly as challengearticles provided that stoppers and labelsare removed. This is a convenient wayto perform depyrogenation studies andavoids the problem of poor recovery ofendotoxin. Associates of Cape Cod, Inc. 'scontrol standard endotoxins are regis­tered as medical devices under a 510Ksubmission for use in depyrogenation val­idation.

Distribute challenge articles through­out the oven, including the cold spot, andinclude them as part ofa full load. Onceagain, five vials per shelf in an X pat­tern, on at least the top, middle and bot­tom shelves, is a common arrangement.Additional vials should be used at thecold spots if these are not already cov­ered by the other vials. At least six vialsshould be left out of the oven as untreat­ed controls. If the challenge article usedis a well defined product with little vialto vial variability, it may not be neces­sary to test as many as six control vials.

LAL UPDATE Vol. I L Number S. page ~

Calendar

For example:If 10 ml vials are used as challenge

articles and are spiked with 40J.l1 of highpotency endotoxin, the nominal endo­toxin challenge is 10,000 ED, assumingthat the potency of the endotoxin is IOED/ng with the LAL lot being used. Ifendotoxin is recovered by adding 1 mlLRW to the article, then ED/ml is equiv­alent to ED/article. Ifthe mean measuredconcentration of the untreated controlsis 5,000 El.l/ml, the recovery rate is 50%.

It is recommended that the extractsfrom treated articles (processed in thedepyrogenation cycle) be tested in a lim­its test. If they show no detectable endo­toxin, this represents destruction by afactor of at least 5000 EU/ml -i- 0.03(125) ED/ml =160,000 or > 5 log reduc­tion. The extract from the treated arti­cles should be tested both spiked andunspiked in the normal way to ensure thatnegative results are due to absence ofendotoxin and not inhibition.

If the treated extracts test positive inthe limits test it is still possible that atleast a three log destruction of endotoxinhas been achieved and extracts should beassayed. Provided that the endpoints areall obtained at extract dilutions of lessthan 1:160, destruction of endotoxin isat least three logs for all articles. No sin­gle article should have an endpoint at1:I60 for the depyrogenation process tobe considered valid. Note: twofold dilu­tions from 1:2.5 will include 1:160 in thedilution scheme.

Write up the report, file it and have itavailable for inspection. The SOP shouldstate the conditions that will necessitaterevalidation. If the physical data (tem­perature and time) assures that specifiedminimum conditions are met or exceed­ed, it should not be necessary to repeatthe endotoxin challenge study (Phase II).However, an annual test using a chal­lenge article placed at the cold spot is areasonable check. Be sure the tempera­ture probes are calibrated regularly andthat the original "worst case" loadingcondition is not exceeded.

Each individual oven must bevalidated. With some thought thismethod can be.adapted to tunnel ovensand the same principles may be appliedto stopper washers.

LAL UPDATE VoL II, Number 5, page 4

ReferencesAkers, M. J., I(. M. Ketron, and B. R.Thompson. 1982. F value requirementsfor the destruction of endotoxin in thevalidation of dry heat sterilization/depy­rogenation cycles. 1. Parenter. Sci. Tech­nol. 36: 23.Avis, I(. E., R. C. Jewell, and J. D.Ludwig. 1987. Studies on the thermaldestruction of Escherichia coli endo­toxin. 1. Parenter. Sci. Technol. 41: 49­56.Baird, R. 1988. Validation of dry heattunnels and ovens. Pharm. Eng. 8 (2):31-33.Hecker, W., D. Witthauer, and A.Staerk.1993. Validationofheatinacti­vation of bacterial endotoxins. Proceed­ings of Second International PDA Con­gress "Validation of Pharmaceuticals andBiopharmaceuticals." Basel, Switzer­land.LAL Users' Group. 1989. Preparationand use ofendotoxin indicators for depy­rogenation process studies. 1. Parenter.Sci. Technol. 43 (3): 109-112.Ludwig, J. D., and I(. E. Avis. 1990.Dry heat inactivation ofendotoxin on thesurface of glass. 1. Parenter. Sci. Tech­nol. 44: 4-12.

The Center for Professional Advancement"LAL Testing: Medical Devices

and Biotechnology"Amsterdam April 25-27, 1994

Course DirectorMichael E. Dawson, PhD.Assistant Director, ACC

National Association ofNephrology Technologists

Marriott HotelLaGuardia Airport

May 16,1994"Endotoxin Contamination in

Hemodialysis Systems"by Marilyn J. Gould, PhD.Associate Director, ACC

94th American Society for MicrobiologyMay 23-27, 1994

Las Vegas Convention CenterVisit Associates of Cape Cod, Inc.

at Booth # 1153

Ludwig, J. D., and I(. E. Avis. 1989.Recovery of endotoxin preparations fromthe surface of glass capillary tubes. 1.Parenter. Sci. Technol. 43: 276-278.Novitsky, T. J., J. Schmidt-Gengen­bach, and J. F. Remillard. 1986. Fac­tors affecting endotoxin recovery fromcontainer surfaces. 1. Parenter. Sci. Tech­nol. 40: 284-286.Tsuji, 1(., and S. J. Harrison. 1978.Dry heat destruction of lipopolysaccha­ride: Heat destruction kinetics. Appl. En­viron. Microbiol. 36 (5): 710-714.Tsuji, 1(., and A. R. Lewis. 1978. Dry­heat destruction of lipopolysaccharide:Mathematical approach to process eval­uation. Appl. Environ. Microbiol. 36 (5):715-719.United States Pharmacopeia XXII.Mack Publishing Co., Easton, PA, 1483­1495, (1990).Validation of Dry Heat Processes Usedfor Sterilization and Depyrogenation.PDA Technical Report No.3, PDA, Phil­adelphia PAWeary, M., and F. C. Pearson. 1988.A manufacturer's guide to depyrogena­tion. Biopharm. April, 22-29.

IBC's Fourth Annual ConferenceThe Shift in R&D Strategies for the

Prevention & Treatment ofEndotoxemia & Sepsis

The Ritz-Carlton, Philadelphia, PA"Limulus Endotoxin Neutralizing Protein:

Potential for Human Therapy"by Thomas J. Novitsky, PhD.,

VPlDirector, ACCJune 13-15, 1994

Biomedical Focus VIIIConference and Exposition

Minneapolis Convention CenterMinneapolis, MNJuly 25-27, 1994

ACC Tabletop Exhibit

International Endotoxin SocietyThird Conference

Marina Congress CenterHelsinki, Finland

August 15-18, 1994