GUIDELINE ON VALIDATION OF THE LIMULUS AMEBOCYTE LYSATE TEST AS AN END-PRODUCT ENDOTOXIN TEST FOR HUMAN AND ANIMAL PARENTERAL

DRUGS, BIOLOGICAL PRODUCTS, AND MEDICAL DEVICES December 1987

Prepared

by:

Center Center Center Center

for for for for

Drug Evaluation and Research Biologic Evaluation and Research Devices and Radiological Health Veterinary Medicine

Maintained

by:

Division of Manufacturing Office of Compliance Center for Drug Evaluation Food and Drug Administration 5600 Fishers Lane Rockville, MD 20857

and Product and Research

Quality

(HFN-320)

CONTENTS

INTRODUCTION I. II. III. IV. Background Legal Effect ...................................................... .................................................... Requirements Drugs ...................................... and Biological Products ............... ...4 ...6 6 8 11 11 1 3

Regulatory

Human and Animal A. Validation B. Testing

of the LAL Test ................................... of Drugs by the LAL Test .............................

V.

Medical

Devices

................................................ of the LAL Test ..................................

A. Validation B. Testing VI, Appendices A. Initial

of Devices

by the LAL ;est..........................14 16 7 8

..................................................... Quality Control.....................................1 Endotoxins Relationship Test....................1 Between the Control Endotoxin.....lo

B. USP XXI/NF C. Determination Standard D. Maximum E. Maximum

XIV Bacterial of the Endotoxin

and the Reference

Standard

Valid

Dilution

Calculation..........................2 Limit Table......................2

2 5

Dose and Endotoxin

INTRODUCTION

This Limulus using

guideline Amebocyte this

sets

forth Lysate as

acceptable test. It also

conditions describes endotoxin products),

for

use

of

the for human

procedures test animal for

methodology drugs (including

an end-product biological The

injectable drugs, the and rabbit

injectable in lieu of

medical pyrogen

devices. test.

procedures

may be used

For

the

purpose refer for means

of only

this to

guideline, Limulus

the Amebocyte and

terms

lysate licensed The in a term

or

lysate by official States Drug the

reagent Center test

Lysate

Biologic that drug or

Evaluation a test is

Research.

referenced

United New Animal

Pharmacopeia Application

monograph,

a New Drug License.

Application,

a Biological

I
I.

BACKGROUND

In a notice of January 12, 1973 (38 FR 1404)) FDA announced that Limulus Amebocyte Lysate derived cells from circulating blood (La), (amebocytes) of the horseshoe crab, (Limulus polyphemus), is a biological product. As such, it is subject to licensing requirements as provided in section 351 of the Public Health Service Act (42 U.S.C. 262). Since 1973, LAL has proved to be a sensitive indicator of the presence of bacterial endo toxins Because of this demonstrated (eyrowns). sensitivity, I&L can be of value in preventing the administration or use of products which may produce fever, shock, and death if administered to or used in humans or animals when bacterial endotoxins are present. When the January 12, 1973 notice was published, available data and experience with LAL were not adequate to support its adoption as the final pyrogen test in place of the rabbtt pyrogen test, which had been accepted and recognized for many years. In order to establish a data permitted the base and gain experience with the use of LAL, that notice introduction of LAL into the marketpiace without a license. This was upon the condition that its use be limited to the in-process testing of drugs be reached and other products, that the decision to use it voluntarily by affected firms, and the labeling on LAL state that the test was not suitable as a replacement for the rabbit pyrogen test. Since that time, production techniques have been greatly improved and standardized so that consistently yield LAL with an endotoxin they sensitivity over 100 times greater than originally obtained. Moreover, it is widely recognized that the LAL test is faster, more economical, and requires a smaller volume of product than does the rabbit pyrogen test. In addition, the procedure is less labor intensive than the rabbit test, making it possible to perform many tests in a single day. In a notice published in the Federal Register of November 4, 1977 (42 FR 57749)) FDA described conditions for the use of LAL as an end-product test for endotoxins in human biological products and medical devices. The notice stated further that the application of LAL testing to human drug products would be the subject of a future Register Federal publication. The then Bureau of Medical Devices, now FDAs Center for Devices and Radiologic Health (CDRH), issued recommended procedures for the use of LAL testing as an end-product endotoxin test on March 26, 1979. These procedures were revised as a result of the comments received from interested parties . AS a direct result of CDRHs experience in approving petitions for the use of the LAL test in place of the rabbit pyrogen test, several procedures for using the LAL test have evolved and have been adopted for devices. In the FEDERAL REGISTER of January 18, 1980 (45 FR 3668)) FDA announced the availability of a draft guideline that set forth procedures for use of the LAL test as an end-product testing method for endotoxins in human This draft guideline was made and animal injectable drug products. -l-

I
available to interested parties to permit manufacturers, who had used the LAL test in parallel with the rabbit submit data that could be considered in the preparation guideline. especially those pyrogen test, to of any final

In response to comments received on the January 18 draft guideline, FDA made several significant changes (i.e. Endotoxin limits changed and deletion of section on Absence of Non-endotoxin Pyrogenic Substances), and many minor editorial changes. The agency also determined that a single document should be made available .covering all FDA regulated products that may be subject to LAL testing. Primarily because of the addition of biological products and medical devices to the guideline, the the agency made, in the FEDERAL REGISTER of March 29, 1983 (43 FR 13096), another draft of the guideline available for public comment. Based on the comments received on the March 29 draft guideline, FDA has The comments used in made several changes in this final guideline. of these changes may be viewed at FDAs Dockets Management support Branch, Room 4-62, 5600 Fishers Lane, Rockville, MD between 9 am and 4 pm Briefly, the significant changes made are: Monday through Friday. A. B. Inclusion of endpoint-turbidimetric validation criteria and kinetic-turbidimetric for the chromogenic, LAL techniques. can be a gel-clot must be used. of a

Any technique (gel-clot, chromogenic or turbidimetricj used in testing a product for endotoxin. However, if lysate is used in a different technique the results interpreted using the criteria for the technique being Elimination of the rabbit pyrogen testing requirement colony. to test the sensitivity

C. D.

The Center for Devices and Radiological Health (CDRH) has adopted the !JSP Endotoxin Reference Standard and revised the limit expressions from ng/mL to EU/mL. The new limit for medical. devices is 0.5 EU/mL except for devices in contact with cerebrospinal fluid for which the limit is 0.06 EU/rnL. These limits for devices are equivalent to those for drugs for a 70 Kg man when consideration is given to the following: 1. In the worst case situation, all endotoxin present in the combined rinsings of 10 devices could have come from just one device. A wide variation in bioburden is common to some devices, Published FDA studies indicate added endotoxin is recovered non-pyrogenic water rinse. that from less than half devices using of a

2.

E.

The Center for Drug Evaluation and Research (CDER) has added a the hour and the listing of maximum doses per Kg per endotoxin limits for most of the aqueous corresponding This injectable drugs and biologics currently on the market. companies making listing was added to promote uniformity among the same product. -2-

II.

LEGAL EFFECT OF THE GUIDELINE

This guideline is issued under section 10.90(b) (21 CFR 10.90(b)) of FDAs administrative regulations, which provides for use of guidelines to outline procedures or standards of general applicability that are acceptable to FDA for a subject matter within its statutory authority. Although guidelines are not legal requirements, a person, who follows an agency guideline may be assured that the procedures or standards will be acceptable to FDA. The following guideline has been developed to inform manufacturers of human drugs (including biologicals), animal drugs, and medical devices of procedures FDA considers necessary to validate the use of LAL as an end-product endotoxin test. A manufacturer who adheres to the guideline would be considered in compliance with relevant provisions of the applicable FDA current good manufacturing practice regulations (CGMP) for drugs and devices and other applicable requirements. As provided in 21 CFR 10.90(b), persons who use methods and techniques not provided in the guideline should be able to adequately assure, through validation, that the method or technique they use is adequate to detect the endotoxin limit for the product.

- 3 -

III.

REGULATORY PROVISIONS THAT PERMIT INITIATION OF END-PRODUCT TESTING WITH LAL

The regulatory provisions that a firm must meet before using the LAL test as an end-product test are not the same for all categories of products different provisions because of the applicable statutory and These provisions are as follows: regulations. A. subject to New Drun Human Drugs ADDliCatiOnS (NDAs) or Abbreviated New Drum ADDlications (ANDAs). Antibiotic Drug ADDliCatiOnS. and animal drugs subject to New Animal Drug (NADAs). and Animal Drug ADDliCatiOnS Abbreviated New Auolication. For these classes of drugs, manufacturers are to submit a However, supplemental application to provide for LAL testing. for human use and 21 CFR under 21 CFR 314.70(c) for drugs 514.8(d)(3) for drugs for animal use various changes may be made before FDA approval. Under these sections changes in testing of a human or animal drug that give increased assurance that the drug will have the characteristics of purity it purports or is represented to possess should be placed into effect at the earliest possible time. Therefore, if a firm validates the LAL test for a particular drug product covered by a new drug in this guideline using a LAL application by the procedures reagent licensed by the Center for Biologic Evaluation and Research (OBER) for the technique being used, the change may be made concurrently with the submission of the supplement The supplement should contain initial quality providing for it. control data, inhibition/enhancement data and the endotoxin limit for the drug product. B. Biolopical Droducts for human use.

Under 21 CFR 601.12 significant changes in the manufacturing methods of biological products are required to be reported to the agency and may not become effective until approved by the Director, OBER. Therefore, a manufacturer of a biological product shall obtain an approved amendment to its product license before changing to the use of LAL in an end-product test, irrespective of the validation procedure used. C. Drugs not subject
A manufacturer

to oremarket

aunroval.

of an injectable drug for human or animal use is not subject to premarket approval would be able to use LAL test as an end-product test for endotoxins without any information to the agency. CGMPs require the submitting manufacturer to have data on file to validate the use of the LAL test for each product for which it is being used. that the
D. Medical On the devices Devices. basis since in of extensive experience November 1977, CDRH believes - 4 review of LAL data that the LAL test, on

when validated according to this least guideline, is at equivalent to the rabbit pyrogen test as an end-product test for medical devices. A manufacturer labeling a device as non-pyrogenic must validate the LAL test for that device in the test laboratory to be used for end-product testing before using the LAL test as an end-product endotoxin test for any device. The data discussed under Section V of this guideline may be expressed graphically or in tabular form and should be on file at the manufacturing site; no preclearance prior to use of the LAL test as an end-product test is required if it is used according to this FDA guideline. Voluntary submission of LAL validation and inhibition data obtained following issuance of this guideline will be accepted for CDRH review and comment. When a manufacturer plans to use LAL test procedures that deviate from significantly the LAL guideline, a premarket notification under section 510(k) of the Federal Food, Drug, and Cosmetic Act (the Act) or a Premarket Approval Application (PMA) supplement under section 515 of the Act should be submitted. Significant deviations would -include-but not necessarily be limited to-higher endotoxin concentration release criteria, sampling from fewer than three lots for inhfbition/en.hancement testing, lesser sensitivity to endotoxin, rabbit retest when the LAL method shows endotoxin above the recommended allowable endo toxin dose, and a device rinsing protocol resulting in greater dilution of endotoxin than that recommended in this guideline. CDRH will also consider submissions in the form of a premarket notification or PMA supplement for another deviation from this draft guideline; process control of endotoxin contamination with reduced end-product testing, i.e., a decrease in the number of devices lot undergoing end-product testing. The per manufacturer must demonstrate adequate control of the production process by the use of routine checks for endotoxin at key stages of production except where it has been shown that no possibility of contamination exists. To facilitate subsequent investigational devices subject encouraged to use this guideline to be manufactured. PMA review, providers to 21 CFR part 812 or 813 when a non-pyrogenic device of are is

- 5 -

IV.

HUMAN AND ANIMAL

DRUGS AND BIOLOGICAL

PRODUCTS

GENERAL REQUIREMENT Manufacturers validation, A. shall in-process, use an LAL reagent licensed and end-product LAL tests. by CBER in all

VALIDATION

OF THE LAL TEST

Validation of the LAL test as an endotoxin test for the release of drugs includes following: human and animal the (1) initial qualification of the laboratory, and (2) inhibition and enhancement tests. 1. INITIAL OUALIFICATION OF THE LABORATORY

Various methodologies have been described for the detection of using limulus amebocyte lysate. Currently, endotoxin, available licensed lysates use commercially the clot, gel endpoint-turbidimetric or kinetic-turbidimetric chromogenic, Other methods which have been reported show techniques. potential for increasing further the sensitivity of the LAL method . should assess testing Manufacturers the variability of the laboratory before any offical tests are performed. Each analyst using a single lot of LAL and a single lot of endotoxin should for confirmation perform the test of labeled LAL reagent sensitivity or of performance criteria. Appendix A gives the procedures and test criteria for the current licensed techniques. 2. INHIBITION AND ENHANCEMENT TESTING

The degree of product inhibition or enhancement of the LAL procedure should be determined for each drug formulation before the LAL test is used to assess the endotoxin content of any drug. All validation tests should be performed on undiluted dilution. drug product or on an appropriate Dilutions should not exceed the Maximum Valid Dilution (MVD) (see Appendix D). At least three production batches of each finished product should be tested for inhibition and enhancement. a) GEL-CLOT TECHNIOUE

Inhibition/enhancement testing should be conducted according to the directions in the preparatory section of the USP Bacterial Endotoxins Test (see Appendix B). Briefly, the concent rat ion containing method involves taking a drug
varying concentrations of a standard endotoxin that bracket the sensitivity of the lysate and comparing it to a series of the same endotoxin concentrations in water alon:. The with endotoxin and then diluted drug product is spiked the drug concentration with additional drug product (so that constant) to the same endotosin concentrations in remains

- 6 -

Results of endotoxfn determination in water and the should Fall product within plus/minus a twofold dilution of the labeled sensitivity. If the undiluted drug product shows inhibition, the drug product can be diluted, not to exceed the MVD, with the same diluent that will be used in the release testing procedure and the above repeated. Negative controls (diluent plus lysate) should be included in all inhibition/enhancement testing.
water.
drug

b)

CHROMOGENICAND ENDPOINT-TURBIDXMETRIC TECHNIOUES In inhibition/enhancement testing by these techniques, a drug concentration containing 4 lambda concentration of the RSE or CSE (lambda is equal to the lowest endotoxin concentration used to generate the standard curve) is tested in duplicate according to the lysate manufacturers methodology. The standard curve for these techniques shall consist of at least four RSE or CSE concentrations in water that extend over the desired range. If the desired range is greater than one log, additional standards concentrations should be included. The standard curve must meet the criteria for linearity as outlined in Appendix A(2). The detected amount of endotoxia in the spiked drug must be within plus or minus 25% of the 4 lambda concentration for the drug concentration to be considered to neither enhance nor inhibit the assay. If the undiluted drug product shows inhibition, the drug product can be diluted, not to exceed the MVD, and t:. test repeated. An alternate procedure may be performed as described above except the RSE/CSE standard curve is prepared in LAL negative drug product, i.e. no detectable endotoxin, instead of LAL negative water. The standard curve must meet the test for linearity, i.e. r equal to or greater than 0.980, and in addition the difference between the O.D. readings for the lowest and highest endotoxin concentrations must be greater than 0.4 and less than 1.5 O.D. units. If the standard curve does not meet these criteria, the drug product cannot be tested by the alternate procedure.

c) KINETIC-TURBIDIMETRIC

TECHNIOUE

In inhibition/enhancement testing by this technique, a drug concentration containing 4 lambda concentration of the RSE or CSE (lambda is equal to the lowest endo toxin concentration used to generate the standard curve) is tested in duplicate according to the lysate manufacturers methodology. The standard curve shall consist of at least four RSE or CSE concentrations. If the desired range is greater than one log, additional standard concentrations should be included. The standard curve must meet the criteria outlined in Appendix A(3). The calculated mean amount of endotoxin in the spiked drug product, when referenced to the standard curve, must be within plus or minus 25% to be considered to neither enhance nor inhibit
the assay. If the -7undiluted drug product shows

inhibition or enhancement, the drug product not to exceed the MVD, and the test repeated,

can

be

diluted,

An alternate procedure may be performed whereby the RSE/CSE curve is standard prepared in drug product or product dilution instead of water. The drug product cannot have a background endo toxin concentration of more than 10% (estimated by extrapolation of the regression line) of the lambda concentration (lambda equals the lowest concentration used to generate the standard curve). The standard curve must meet the test for linearity, i.e. r equal to or less and in addition the slope of the regression than -0.980, must be less than -0.1 and greater than -1.0. If the standard curve does not meet these criteria, the drug product cannot be tested by the alternate procedure. when In those instances the drug is manufactured in various concentrations of active ingredient while the other components of the formulation remain constant, only the highest and lowest concentration If there is a significant difference, i.e. greater need be tested. between the inhibition endpoints or if the drug than twofold, solutions is different, then each concentration, per mL, in the test remaining concentrations should be tested. If the drug product shows inhibition or enhancement at the MVD, when tested by the procedures in the above sections, and is amenable to rabbit testing, then the rabbit will still be the appropriate test for that drug. If the test inhibiting or enhancing substances can be neutralized without is more affecting the sensitivity of the test or if the LAL test sensitive than the rabbit pyrogen test the ML test can be used. For those drugs not amenable to rabbit pyrogen testing, the manufacturer should determine the smallest quantity of endotoxin that can be detected. This data should be submitted to the appropriate FDA Office for review. The inhibition/enhancement tests must be repeated on one unit of the product if the lysate manufacturer is changed. If the lysate technique is changed, the inhibition and enhancement tests must be using three batches. repeated When the manufacturing process, the product formulation, the sou,rce of a particular ingredient of the drug formulation, or lysate lot is changed, the positive product control can be used to reverify the validity of the LAL test for the product. Firms that are obtaining an ingredient from a new manufacturer are encourged to include as part of their vendor qualification the rabbit to determine that the ingredient does not contain pyrogen test non-endotoxin pyrogens. B. Routine Testinv of Drugs by the LAL Test. from the Samples, should be

based on data End-product testing is to be inhibition/enhancement testing as outlined in Section A(2). standards, positive product controls and negative controls tested at least in duplicate.

- 8 -

-i

For the have to endpoints be run at if there conditions run when controls must be the USP standard be noted

gel-clot technique, an endotoxin standard series does not be run with each set of tests if consistency of standard has been demonstrated in the test laboratory. It should least once a day with the first set of tests and repeated is change in lysate lot, endotoxin lot or test any during the day. An endotoxin standard series should be confirming end-product contamination. Positive product ( two lamda concentration of standard endotoxin in product) positive. If your test protocols state that you are using Bacterial Endotoxin Test, remember that it requires a series to be run with each test. The above deviation must in your test protocol.

For the chromogenic and endpoint-turbidimetric techniques, an endotoxin standard series does not have to be run with each set of tests if consistency of standard curves has been demonstrated in the test laboratory. It should be run at least once a day with the first set of tests and repeated if there is any change in lysate lot, endotoxin. lot or test conditions during the day. However, at least duplicates of a 4 lambda standard concentration in water and in each product (positive product control) must be included with each run of samples. The mean endotoxin concentration of the standard must be within plus/minus 25% of the actual concentration and the positive product control must meet the same criteria after subtraction of any endogenous endotoxin. An endotoxin standard series should be run when confirming end-product contamination. If the alternate procedure is used, a standard in product series must be conducted each time the product is tested. For the kinetic-turbidimetric test, it is not necessary to run a standard curve each day or when confirming end product contamination if consistency of standard curves has been demonstrated in the test laboratory.. However, at least duplicates of a 4 lambda standard concentration in water and in each product (positive product control) must be included with each run of samples. The mean endotoxin concentration of the standard when calculated using an archived standard curve (See Appendix C), must be within plus/minus 25% of the actual concentration and the positive product control must meet the same criteria after subtraction of any endogenous endotoxin. If the alternate procedure is used, a standard in product series must be conducted each time the product is tested. Before a new lot of lysate is used, the labeled lysate be or the performance criteria should laboratory, using the procedures in Appendix A. sensitivity confirmed of by the the

The sampling technique selected and the number of units to be tested should be based on the manufacturing procedures and the batch size. A minimum of three units, representing the beginning, middle, and end, from a lot. can be run should be tested These units individually or pooled. If the units are pooled and any endotoxin The LAL test may be is detected, repeat testing can be performed. repeated no more than twice. The first repeat consists of twice the initial number of replicates of the sample in question to examine the possibility that extrinsic contamination occurred in the initial -9-

assay procedure. On pooled samples, if any endotoxin is detected in the first repeat, proceed to second repeat. The second repeat consists of an additional 10 units tested individually, None of the 10 units tested in the second repeat may contain endotoxin in excess of the limit concentration for the drug product. The following should parenteral drugs to end-product endotoxin 1.
K/M:

be considered meet if the test:

the endotoxin LAL test is to

limit

be

for all used as an

any parenteral For drug except those administered intrathecally, the endotoxin limit for endotoxin is the amount of endotoxin defined as K/M, which equals (EU) allowed per ng or mL of product. K is equal to 5.0 EU/Kg. (SEE appendix D for definition of M).
K is

For parenteral administration, Drugs exempted

drugs that have an equal to 0.2 EU/Kg.

intrathecal

route

of

from the above endotoxin drugs for which

are: endotoxin limits have been

Compendia1 established.

other

2.

drugs covered by new Non-compendia1 drug applications, new animal drug applications, and antibiotic drug applications, biological product licenses where different limits have been approved by the agency. Investigational drugs or biologicals exemption has been filed and approved. Drugs or biologicals which cannot for which an IND or INAD

3. 4.

be tested

by the LAL method.

A batch which fails a validated LAL release test should not be retested rabbit test and released if it passes. Due to the high by the variability and lack of reproducibility of the rabbit test as an endotoxin assay procedure, we do not consider it an appropriate retest procedure for LAL failures.

10 -