HTTP 1

2
0
1
0
-2
O
ctober
2010
P
H
A
R
M
E
U
R
O
P
A

B
I
O
&

S
C
I
E
N
T
I
F
I
C

N
O
T
E
S
This issue is included in
the subscription to Pharmeuropa
2010 Council of Europe

ISSN 2075-2164
ISSN-L 2075-2164
Director of the Publication
Dr S. Keitel
Editorial Board
Prof. H. Altorfer
Dr S.R. Andersen
Dr T. Arvidsson
Dr F. Bailleul
Prof. Dr K.H.C. Baser
Dr A.F. Bristow
Dr L. Bruckner
Dr K.H. Buchheit
Prof. D.H. Calam
A. Costanzo
Dr C. Coune
Dr M. Ek
F. Ezrati
Prof. Dr G. Franz
Prof. S.H. Hansen
Dr C. Jarvis
Prof. H.J. de Jong
P. Jongen
K. Kreft
Prof. H.G. Kristensen
Dr G. Lee
M. Marre
Prof. B. Meier
Dr C. Milne
Dr J.M. Morris
Prof. A. Nicolas
C. Nicolas
Dr V. ppling
Prof. P.-P. Pastoret
Dr J.C. Rouby
S. Schaeffer
Prof. G. Scriba
Prof. Dr R. Seitz
Dr G. Silvester
J.-M. Spieser
Dr T. van der Stappen
Dr B. Strukelj
Dr E. Terao
Dr M. Terberger
L. Toumasson
Dr D. Wood
Page layout
I. Bylinski
I. Iankova
C. Knaup
A. Roberts
I. Vernay
Copyright
All rights conferred by virtue
of the International Copyright Convention are
specifically reserved to the Council of Europe
and any reproduction or translation requires
the written consent of the Publisher.
Contributions
The Editor is pleased to consider contributions
for publication in Pharmeuropa Scientific Notes
provided they have some relationship with the
work of the European Pharmacopoeia.
It is assumed that the copyright of material
submitted is that of the contributor or that the
necessary clearances have been obtained and
that material submitted has not been offered
simultaneously to another publication.
Authors alone are responsible for the views
expressed in signed contributions. Please see
Instructions for Authors (www.edqm.eu).
Mention of specific companies or
of certain manufacturers products is made only
for the convenience of the reader. This does not
imply that they are endorsed or recommended
by the European Pharmacopoeia Commission or
by the Council of Europe in preference to others
of a similar nature that are not mentioned.
European Directorate for the Quality of
Medicines & HealthCare (EDQM)
7 alle Kastner
CS 30026
F-67081 STRASBOURG
FRANCE
Tel: +33 (0)3 88 41 30 30
Fax: +33 (0)3 88 41 27 71
E-mail: info@edqm.eu
Internet: http://www.edqm.eu
Bank: Socit Gnrale
Account N: FR76
30003 02360 00550034256/76
Strasbourg
PH
A
R
M
EU
R
O
PA
BIO
&
SC
IEN
TIFIC
N
O
TES 2
0
1
0
-2
C
o
n
te
n
ts
Calibration of Hum
an Coagulation Factor VIII Concentrate
Ph. Eur. BRP Batch 4 for Use in Potency Assays
Collaborative Study for the Establishm
ent of Replacem
ent Batches
of Heparin Low
-M
olecular-M
ass for Assay Biological Reference
Preparations
Collaborative Study for Validation of a Serological Potency Assay
for Rabies Vaccine (inactivated) for Veterinary Use
ent of the Second
International Standard for Gram
icidin
ent of the Third International
Standard for Nystatin
Q
uality Control of M
etoprolol Extended-Release Form
ulations in
the Presence of Ethanol
Possible Am
biguities w
hen Testing Viscosity in Com
pendial
M
onographs - Characterisation of Grades of Cellulose Ethers
Instructions for Authors
O
c
t
o
b
e
r

2
0
1
0
P
H
A
R
M
E
U
R
O
P
A

B
I
O

&

S
C
I
E
N
T
I
F
I
C

N
O
T
E
S

2
0
1
0
-
2
The Editor is pleased to consider contributions for publication in
Pharmeuropa Bio & Scientific Notes provided they have some relationship
with the work of the European Pharmacopoeia. It is assumed that the
copyright of material submitted is that of the contributor or that the
necessary clearances have been obtained and that material submitted has
not been offered simultaneously to another publication. Authors alone are
responsible for the views expressed in signed contributions.
Please see instructions for authors published in this issue and available at
www.edqm.eu.
Articles published in Pharmeuropa Bio & Scientific Notes are indexed in the
PubMed database of the National Library of Medicine, available at
www.ncbi.nlm.nih.gov.material submitted has notl
How to contact us
Information and orders Internet: www.edqm.eu
European Directorate for the Quality of Medicines & HealthCare (EDQM)
Council of Europe - 7 alle Kastner
CS 30026, F-67081 STRASBOURG, FRANCE
Tel: +33 (0)*3 88 41 30 30
Fax: +33 (0)*3 88 41 27 71
*Do not dial 0 if calling from outside of France
Correspondence ....................................................................................................... Via the online Helpdesk (www.edqm.eu/hd)
How to place an order
Publications ...................................................................................................................................................https://www.edqm.eu/store
Reference standards ...................................................................................www.edqm.eu/site/EDQM_Reference_standards-649.html
Reference standards online order form ...............................................................www.edqm/eu/site/CRS-order-form-697.html
Further information, including answers to the most frequently asked questions regarding ordering, is available via the Helpdesk.
Submission of scientific articles ................................................................................................................ publishing@edqm.eu
All reference standards required for application of the monographs are available from the EDQM.
A catalogue of reference standards can be consulted on the EDQM website and printed directly. The list of newly released
reference standards (new reference standards and new batches) is available on the website http://crs.edqm.eu under the link
new.
Printed on acid-free paper by Bialec, Nancy, France

Members of the European Pharmacopoeia Commission: Austria, Belgium,
Bosnia and Herzegovina, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark,
Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia,
Lithuania, Luxembourg, Malta, Montenegro, Netherlands, Norway, Poland,
Portugal, Romania, Serbia, Slovak Republic, Slovenia, Spain, Sweden, Switzerland,
the former Yugoslav Republic of Macedonia, Turkey, United Kingdom and
the European Union.
Observers to the European Pharmacopoeia Commission: Albania, Algeria,
Argentina, Armenia, Australia, Belarus, Brazil, Canada, China, Georgia, Israel,
Madagascar, Malaysia, Moldova, Morocco, Republic of Kazakhstan, Russian
Federation, Senegal, Syria, Tunisia, Ukraine, United States of America and
WHO (World Health Organisation).
Pharmeuropa Bio & Scientic Notes, 2010-2 i
PHARMEUROPA BIO & SCIENTIFIC NOTES 2010-2
CONTENTS
Calibration of Human Coagulation Factor VIII Concentrate Ph. Eur. BRP Batch 4
for Use in Potency Assays
S. Raut, A. Costanzo, S. Daniels, A. Heath, K.-H. Buchheit .............................................................. 1
Collaborative Study for the Establishment of Replacement Batches of Heparin Low-
Molecular-Mass for Assay Biological Reference Preparations
E. Terao, A. Daas, G. Rautmann, K.-H. Buchheit. ........................................................................... 30
Collaborative Study for Validation of a Serological Potency Assay for Rabies Vaccine
(inactivated) for Veterinary Use
B. Krmer, L. Bruckner, A. Daas, C. Milne....................................................................................... 37
Collaborative Study for the Establishment of the Second International Standard for
Gramicidin
G. Rautmann, A. Daas, K.-H. Buchheit. .......................................................................................... 56
Collaborative Study for the Establishment of the Third International Standard for Nystatin
G. Rautmann, E. Charton, A. Daas, K.-H. Buchheit ........................................................................ 68
Quality Control of Metoprolol Extended-Release Formulations in the Presence of Ethanol
A. Amini, A. Dawood, A.-M. Hesselgren, H. Thor, S. Jnsson, T. Arvidsson, G. Ragnarsson,
M. Johansson ................................................................................................................................. 86
Possible Ambiguities when Testing Viscosity in Compendial Monographs -
Characterisation of Grades of Cellulose Ethers
E. Doelker ........................................................................................................................................ 92
Instructions for Authors .............................................................................................................. 100
ii Pharmeuropa Bio & Scientic Notes, 2010-2
Luropean DirecLoraLe or Lhe
QualiLy o Nedicines & HealLhCare
Un logiciel ddi lanalyse
statistique des rsultats de
titrages biologiques selon la
Pharmacope Europenne
What is CombiStats?
Who is CombiStats designed for?
Templates in just a few easy steps
Output delivered
Free demonstration version
available from the Internet
CombiStats is a user-friendly computer program for the
statistical analysis of data from biological dilution or
potency assays. It can perform calculations according to
Chapter 5.3 of the European Pharmacopoeia (5
th
and 6
th

Edition) including the following models:
parallel line models,
slope ratio models,
probit models,
ED50 calculations (and other percentiles),
4-parameter curve models,
single dose models,
combination of assays,
and many more of the family of generalised linear
models.
CombiStats is intended for use by those responsible for the
analysis of assay data but whose primary training is not in
statistics and who have no access to qualified statisticians
to assist them with the analysis. The software may be useful
in any situation where the analysis of data from dilution
assays is required but it is primarily intended for use in the
context of monographs of the European Pharmacopoeia.
Much effort has been invested to keep the software easy
to use and at the same time versatile enough to cover
a wide range of existing potency assays. With the help of
the wizard, it only takes a few simple steps to create your
own customised templates for routine assays. Just specify:
the number of preparations, doses and replicates,
the model,
the design,
a possible transformation,
the type of analysis of variance.
Once a template has been created you can simply open
the template, enter the observed data and click to start
the calculations. Individual observations, treatments or
preparations can be included or excluded with a simple click
of the mouse.
Standard results produced by the software are an analysis
of variance, potency estimates, and graphs of the fitted
models. The output is formatted in a pleasant and
consistent layout. The software also offers more advanced
features to allow the more experienced user to access less
commonly used options such as weighted regression and
export of estimated parameter vectors and associated
covariance matrices.
A free demonstration version can be downloaded from
www.combistats.eu. This demo-version includes the
complete manual and tutorial to help the user to get
acquainted with the most important features of the
software. Also included is a set of 24 example files offering
a wide range of possible configurations for specific
assay situations. For more information about system
requirements, prices, and how to order a user licence,
please consult the website.
Qu'est-ce que CombiStats ?
A qui s'adresse CombiStats ?
Crez vos propres masques
Les donnes de sortie
Une version de dmonstration
gratuite sur internet
CombiStats est un logiciel convivial ddi lanalyse
statistique des rsultats de titrages biologiques par dilution
ou titrages dactivit. Il peut raliser les calculs dcrits dans
le chapitre 5.3 de la Pharmacope Europenne (5
e
et 6
e

Editions) en utilisant les modles suivants :
modle en lignes parallles,
modle rapport de pentes,
modle des probits,
calculs de la DE50 (et autres pourcentiles),
modle de distribution 4 paramtres,
modle dose unique,
combinaison de dosages,
ainsi que de nombreux autres modles linaires
gnraliss.
CombiStats est loutil de tous ceux qui ont analyser des
rsultats de dosages sans tre statisticiens de formation
ou avoir accs lassistance dun statisticien. CombiStats
peut tre utile dans toute situation impliquant lanalyse
de rsultats de titrages par dilution, mais il a t conu
en premier lieu pour tre utilis dans le contexte des
monographies de la Pharmacope Europenne.
Des efforts considrables ont t faits pour que le logiciel
reste simple utiliser tout en tant suffisamment polyvalent
pour couvrir une large gamme de titrages rencontrs dans
la pratique. En vous laissant guider par le programme, vous
pouvez en quelques tapes simples crer vos propres masques
de calcul pour les analyses de routine. Il vous suffit de spcifier :
le nombre de prparations, doses et rplicats,
le modle,
le plan dessai,
une ventuelle transformation,
le type danalyse de variance.
Une fois le masque cr, il ne vous reste qu louvrir, entrer
les donnes observes et cliquer pour lancer les calculs.
Vous pouvez dun simple clic choisir dinclure ou dexclure
des observations, des prparations ou des traitements
particuliers.
CombiStats fournit classiquement une analyse de
variance, des estimations de lactivit et le graphe des
modles ajusts. La prsentation des donnes de sortie
est agrable et homogne. Le logiciel offre galement
aux utilisateurs expriments des fonctions avances
leur permettant daccder des options moins courantes
comme la rgression pondre et lexportation de vecteurs
de paramtres estims et de matrices de covariance
associes.
Une version de dmonstration gratuite est tlchargeable
sur le site www.combistats.eu. Elle comprend le manuel
intgral de l'utilisateur et un didacticiel permettant de
se familiariser avec le fonctionnement du logiciel. Sont
galement incluses 24 fiches dexemples offrant un large
panorama des configurations possibles dans des situations de
dosages spcifiques. Pour de plus amples informations sur la
configuration exige, les prix et les modalits de commande
de licences d'utilisateurs, rendez-vous sur le site internet !
A programme for the statistical
analysis of bio-assays
according to the European
Pharmacopoeia
Pharmeuropa Bio & Scientific Notes, 2010-2 1
Calibration of Human Coagulation Factor VIII Concentrate BRP
Calibration of Human Coagulation Factor VIII
Concentrate Ph. Eur. BRP Batch 4
for Use in Potency Assays
S. Raut, A. Costanzo, S. Daniels, A. Heath, K.-H. Buchheit
S. Raut. National Institute for Biological Standards and Control (NIBSC), Health Protection Agency, Potters Bar, EN6 3QG, Hertfordshire, UK.
A. Costanzo. (Corresponding author: e-mail: angele.costanzo@edqm.eu). European Directorate for the Quality of Medicines & HealthCare (EDQM),
7 alle Kastner, CS 30026, F-67081 Strasbourg, France.
S. Daniels. National Institute for Biological Standards and Control (NIBSC), Health Protection Agency, Potters Bar, EN6 3QG, Hertfordshire, UK.
A. Heath. National Institute for Biological Standards and Control (NIBSC), Health Protection Agency, Potters Bar, EN6 3QG, Hertfordshire, UK.
K.-H. Buchheit. European Directorate for the Quality of Medicines & HealthCare (EDQM), 7 alle Kastner, CS 30026, F-67081, Strasbourg, France.
ABSTRACT
The European Pharmacopoeia Biological Reference Preparation (Ph. Eur. BRP) Batch 4 was established as an
international common working standard for potency determination of human coagulation factor VIII (FVIII) preparations
to replace the dwindling stocks of the BRP Batch 3, the current European standard. Similarly, stocks of the current World
Health Organisation 7
th
International Standard (WHO 7
th
IS) were also running low. Therefore a project was jointly
organised by the European Directorate for the Quality of Medicines & HealthCare (EDQM, Council of Europe) and the
National Institute for Biological Standards and Control (NIBSC, UK) in order to replace both standards concomitantly.
The potency of the BRP Batch 4 was assigned during an international collaborative study involving 38laboratories with
reference to the WHO 7
th
IS and the BRP Batch 3.
Four candidate materials, 2 plasma-derived (samples A and C) and 2 recombinant (samples B and D) have been evaluated,
sample C being the specic candidate for the replacement of the BRP Batch 3. Participants were instructed to perform
8 independent assays following their own routine validated methods, by either the one-stage clotting assay or the
chromogenic assay, or both.
Laboratories returned 22 data sets for the clotting assay and 30 data sets for the chromogenic assay. This publication
reports the results obtained with both assays but only the results of the chromogenic assay are highlighted in the
conclusions, as it is the assay prescribed by the European Pharmacopoeia. Data were analysed separately for both assays.
The consensus potency value was calculated as the unweighted geometric mean of the unweighted geometric means of
each individual laboratory. For sample C, there was a signicant difference in potency estimate between the chromogenic
and the clotting assay. It was therefore not possible to reconcile both results. The chromogenic potencies however were
in very good agreement being 10.4 IU/ampoule (n = 30), when assessed against both standards. The inter-laboratory
geometric coefcient of variation (GCV) was 4.8 % and 7.1 % against the WHO 7
th
IS and the BRP Batch 3 respectively.
The Ph. Eur. BRP Batch 4 is a freeze-dried, plasma-derived concentrate. The material was lled in approximately
20 000 ampoules and lyophilised. The nal residual water content is 0.33 %. Based on accelerated degradation studies,
the stability of the material is suitable for a reference preparation. The candidate Ph. Eur. BRP Batch 4 was adopted at the
136
th
session of the European Pharmacopoeia Commission in March 2010. The standard will be available from the EDQM
with the catalogue number H0920000 upon exhaustion of the current batch.
KEYWORDS
Human coagulation factor VIII, reference standard, European Pharmacopoeia Biological Reference Preparation,
collaborative study, chromogenic assay, WHO 8
th
International Standard.
1. INTRODUCTION
The current European Pharmacopeia (Ph. Eur.) BRP Batch
3 (BRP3)/Mega 2 (US/FDA) for Factor VIII (FVIII) was
established in 2001 by calibration against the World Health
Organization (WHO) 6
th
International Standard (IS) for
Factor VIII Concentrate and is used as a working standard
in potency estimation of FVIII in plasma-derived and
recombinant therapeutic concentrates, which are primarily
used in the treatment of haemophilia A (FVIII deciency).
The BRP3/Mega 2 (US/FDA) had been established as working
standard by the Ph. Eur. and the FDA. Stocks of the BRP3
were running low and were expected to be exhausted
by 2010/11. The Steering Committee of the Biological
Standardisation Programme (BSP) thus decided to endorse
a project (BSP098) in January 2008 in order to qualify
a replacement batch for this BRP, calibrated against the
WHO 7
th
IS for FVIII Concentrate (NIBSC code: 99/678) and
against the BRP3.
Concomitantly, the National Institute for Biological
Standards and Control (NIBSC) was planning a project to
replace the WHO 7
th
IS for FVIII Concentrate. Therefore,
in the interest of harmonisation and continuity, a joint
collaborative study between the European Directorate
for the Quality of Medicines & HealthCare (EDQM) and
the NIBSC was undertaken to replace both standards.
Following trial ll studies on a number of therapeutic FVIII
concentrates, 4 candidate materials were selected, 2 plasma
derived (samples A and C) and 2 recombinant (samples B and
D), with sample C designated as specic candidate for the
proposed Ph. Eur. BRP Batch 4. The selection of materials
was based on ll characteristics, stability of FVIII potency,
minimal inter- and intra-assay variability and minimal
discrepancy between assay methodologies, as determined in
pre-qualication studies of the materials. Twenty thousand
(20,000) ampoules of each of the 4 candidate materials have
been prepared.
2 Pharmeuropa Bio & Scientific Notes, 2010-2
2. AIM
The objective of this collaborative study was to calibrate the
candidate Ph. Eur. BRP Batch 4 for Human Coagulation
FVIII Concentrate against the WHO 7
th
IS for FVIII
Concentrate and against the BRP3 for FVIII Concentrate
by the chromogenic method, although data from the one-
stage clotting assays were also assessed for comparison,
particularly as this joint study was also intended to assign
a calibration value to the replacement candidate for the
WHO 7
th
IS for FVIII Concentrate. A proposal for the
establishment of the WHO 8
th
IS for FVIII Concentrate has
been made in a separate report by NIBSC to the 2009 WHO
Expert Committee on Biological Standardisation (ECBS) [1].
Although all the data from the joint collaborative study are
presented, this report will primarily focus on the calibration
of the Ph. Eur. BRP Batch 4 for FVIII Concentrate.
3. PARTICIPANTS
Thirty eight laboratories participated in the study
(from 17 countries world-wide) and returned data for
analysis. They are listed in section 9. The participants
included 22manufacturers and 16 regulatory authorities.
Laboratories were coded for the study and the order of
listing in section 9 does not necessarily correspond with the
numerical codes. All raw data returned by the participants
were analysed at NIBSC.
4. MATERIALS AND METHODS
4.1. Sample A (07/350) - plasma derived candidate
preparation
The raw material for this candidate was a plasma-derived
high purity FVIII concentrate, prepared by chromatography
and containing von Willebrand factor. All donations used
to prepare this product were tested and found negative for
HBsAg, anti-HIV-1 and -2, anti-HCV, HCV-RNA (plasma
pools). Manufacturing of this product also included 2 viral
inactivation steps, solvent detergent and heat treatment
at 81 C for 72 hours. After reconstitution of the product,
the concentrate material was pooled and formulated in the
following buffer: 0.15M NaCl, 50mM Tris-HCl (pH 7.4), 1mM
CaCl
2
, 10mg/ml trehalose, 10mg/ml human albumin. The
formulated material was lled and freeze-dried in sealed
glass ampoules at NIBSC, under conditions required for
International Standards [2]. One ml of this material was
dispensed into each of approximately 20,000 ampoules.
The mean lling weight was 1.0054 g (range 1.0015 g to
1.0100g) and the coefcient of variation (CV) was 0.16 %
based on 624 check-weight samples. Mean residual moisture
after freeze-drying was 0.46 % (CV 20.6 %, n = 12) and mean
oxygen headspace was 0.44 % (CV 11.8 %, n = 11).
4.2. Sample B (07/352) - recombinant candidate
preparation
The raw material for this candidate was a recombinant
FVIII concentrate bulk. After thawing, the bulk concentrate
solution was pooled and formulated in the following buffer:
0.15M NaCl, 50mM Tris-HCl (pH 7.4), 1mM CaCl
2
, 10mg/
ml trehalose, 10mg/ml human albumin. The formulated
material was lled and freeze-dried in sealed glass ampoules
at NIBSC, under conditions required for International
Standards [2]. One ml of this material was dispensed into
each of approximately 20,000 ampoules. The mean lling
weight was 1.0054 g (range 1.0010 g to 1.0100 g) and the
coefcient of variation (CV) was 0.11 % based on 689 check-
weight samples. Mean residual moisture after freeze-drying
was 0.32 % (CV 21.9 %, n = 12) and mean oxygen headspace
was 0.37 % (CV 9.1 %, n = 12).
4.3. Sample C (08/106) - plasma derived candidate
preparation
The material was procured by the EDQM as the prime
candidate material for the proposed Ph. Eur. BRP Batch 4
for FVIII Concentrate from a European manufacturer and is
part of a batch that was also used to produce a therapeutic
product available on the European market. The raw material
for this candidate was a plasma-derived intermediate purity
FVIII concentrate, prepared by multiple precipitation
method and containing von Willebrand factor (vWF). All
donations used to prepare this product were tested and
found negative for HBsAg, anti-HIV-1 and -2, anti-HCV,
HCV-RNA (plasma pools). Manufacturing of this product
also included a viral inactivation pasteurisation process at
60 C for 10 hours in aqueous solution. After reconstitution
of the product, the concentrate material was pooled and
formulated in the following buffer: 50mM NaCl, 267mM
glycine (pH7.3), 18.7mM tri-sodium citrate dihydrate,
10mg/ml human albumin. The formulated material was
lled and freeze-dried in sealed glass ampoules at NIBSC,
under conditions required for International Standards
[2]. One ml of this material was dispensed into each of
approximately 20,000ampoules. The mean lling weight
was 1.0058 g (range 1.0015 g to 1.0085 g) and the coefcient
of variation (CV) was 0.22 % based on 686 check-weight
samples. Mean residual moisture after freeze-drying was
0.33% (CV 33.6 %, n = 12) and mean oxygen headspace was
0.28 % (CV 26.7%, n = 11). The material was sent to the
EDQM and stored at -20 C immediately upon receipt.
4.4. Sample D (08/104) - recombinant candidate
preparation
The raw material for this candidate was a recombinant
FVIII concentrate bulk. After thawing, the bulk concentrate
solution was pooled and formulated in the following buffer:
0.15M NaCl, 50mM Tris-HCl (pH 7.4), 1mM CaCl
2
, 10mg/
ml trehalose, 10 mg/ml human albumin. The formulated
material was lled and freeze-dried in sealed glass ampoules
at NIBSC, under conditions required for International
Standards [2]. One ml of this material was dispensed into
each of approximately 20,000 ampoules. The mean lling
weight was 1.0056 g (range 1.0010 g to 1.0085 g) and the
coefcient of variation (CV) was 0.12 % based on 739 check-
weight samples. Mean residual moisture after freeze-drying
was 0.57 % (CV 11.8 %, n=12) and mean oxygen headspace
was 0.27 % (CV 13.2 %, n=10).
4.5. Sample E - Ph. Eur. BRP Batch 3 (BRP3) for FVIII
Concentrate
This secondary working standard was supplied by EDQM
and is also the same standard as the Mega 2 US secondary
working standard. Sample E has been assigned a potency
of 8.6 IU/vial [3] for the chromogenic method through an
international collaborative study in 2001. During this study,
the BRP3 was also tested in the one-stage clotting assay and
the mean estimated potency was found to be 11.3 IU/vial
for this method. However, since the Ph. Eur. has adopted
the chromogenic assay as its sole method of determining
potency for FVIII preparations, no ofcial value was assigned
for the clotting assay to the BRP3. The FDA, on the other
hand, calibrated this same material as their Mega 2 US FVIII
Concentrate standard with assigned potencies for both the
chromogenic method (8.6 IU/vial) and the one-stage clotting
method (11.3 IU/vial).
4.6. Sample 7
th
IS - WHO 7
th
IS for FVIII Concentrate
This primary standard is a plasma-derived FVIII concentrate.
It was supplied by NIBSC and has a single assigned potency
of 11.0 IU/ampoule for all methods.
4.7. Materials dispatched for the study
Eight ampoules/vials of each of the above materials were
dispatched by NIBSC. Participants were asked not to use any
additional test materials in the assays, such as any internal
(in-house) standards.
4.8. Assay methods and study design
Participants were requested to carry out 8 independent
assays using their normal routine methodology, either
one-stage or chromogenic methods, preferably assays by
both methods. They were requested to carry out assays
on separate occasions (days) using a fresh ampoule/vial of
samples A, B, C, D, E and 7
th
IS for each assay, according to
balanced assay designs recommended in the study protocol.
A separate ampoule/vial of each material was provided
for each assay. The laboratories that used more than one
method were requested to use material from the same
ampoules/vials for both methods, provided this could be
done within 3 hours of reconstitution.
Participants were also requested to pre-dilute all samples
in congenital severe haemophilic plasma (SHP) (or
immunodepleted or chemically depleted plasma provided
it contained normal levels of vWF), to a nal FVIII
concentration of ~1.0 IU/ml. Detailed assay instructions
were provided in the study protocol. The details of the
methods, instruments and reagents used by the participants
are listed in Appendix I.
4.9. Statistical analysis
All assays were analysed as multiple parallel line bioassays
comparing log response to log concentration [4]. All
assays data were plotted, and in some cases individual
data points were excluded as possible data transcription
errors or potential technical errors. The statistical validity
of the assays was assessed by the usual ANOVA tests. The
parallelism of the assays was also assessed by comparing the
slopes of the dose-responses across the assays.
For each assay the potency estimates of samples A - E were
calculated relative to the concurrently tested WHO 7
th

IS for FVIII Concentrate (sample 7
th
IS). Potencies were
also calculated relative to the BRP3 for FVIII Concentrate
(sample E). Combined potency estimates for each laboratory
were obtained by taking unweighted geometric means of
results from all assays. Overall combined estimates were
obtained by taking unweighted geometric means of the
mean results from the different laboratories. Where a
laboratory performed more than one assay method, the
results for each method were analysed as if from separate
laboratories. Intra- and inter-laboratory variability is
expressed as the percentage geometric coefcient of
variation (%GCV) [5].
The mean potency estimates calculated by the participating
laboratories are presented (Appendix II). Differences in
potency estimates between laboratories were assessed using
a Duncan's multiple range test [6]. Differences in potency
estimates between assay method (one-stage or chromogenic)
were assessed by 2-sample t-tests, and conrmed by
Wilcoxon rank-sum tests.
5. RESULTS
5.1. Data received
Data were received from 38 laboratories in total.
Laboratory8 performed repeat one-stage assays with two
different machines. The results are analysed separately
and coded 8A (ACL Advance) and 8B (Coag-A-Mate X2),
see Appendix I. Laboratory 37 performed two different
chromogenic assays. They are analysed separately and coded
37A (chromogenic for plasma derived FVIII samples) and
37B (chromogenic for recombinant FVIII samples). Several
laboratories performed both one-stage and chromogenic
assays. In total there were 52 data sets (22 for one-stage
assays and 30 for chromogenic assays). Methods and
reagents used by the participants are summarized in
Appendix I.
5.2. Assay validity
When the individual assay data were plotted, all laboratories
had dose-responses that appeared to be a good t to the
parallel-line model, when the log transform of the response
was used. For laboratory 4, the untransformed response
was used. Some individual data points were excluded,
but all samples and assays were included in subsequent
analysis, with the exception of one assay from laboratory 28
(chromogenic) where sample C was excluded as non-linear.
5.3. Potency estimates relative to the 7
th
IS
The laboratory geometric mean estimates of potency relative
to the current 7
th
IS, the intra-laboratory %GCVs, the overall
geometric mean and inter-laboratory %GCVs, for both the
one-stage and chromogenic assay methods, and for all assays
combined are shown in Tables 1-5.
The potency estimates are also shown in the form of
stacking histograms (Figures 1-5). Each box represents
the laboratory mean, expressed as a percentage of the
overall geometric mean, calculated from results from both
assay methods combined. The boxes are labelled with the
laboratory code number. Results from one-stage assays are
shaded in grey, whilst the results from chromogenic assays
are unshaded (white).
The potency estimates calculated by the participating
laboratories are shown in Appendix II.
5.3.1. Sample A
From the histogram (Fig. 1a) and Table 1, it can be seen
that there is very good agreement between laboratories
and assay methods. The overall inter-laboratory GCV is
below 5 %, and the overall geometric means from the two
methods are identical at 9.44 IU/ml by one-stage (n = 22)
and chromogenic (n = 30) methods. There are no signicant
differences between methods and the combined mean value
is 9.44 IU/ml (n = 52; GCV = 5.0 %). Duncans multiple
range test did not indicate the presence of any outliers
(laboratories with results signicantly different from all
other laboratories) when applied to the results from both
methods combined. When applied to the chromogenic
results alone, laboratory 4 was signicantly higher than
other laboratories. Given the overall tight agreement
between laboratories, laboratory 4 was not excluded from
any subsequent calculations. Intra-laboratory GCVs ranged
from 0.64-12.81 % (n = 22) for the one-stage assays and
from 0.34-9.49 % (n = 30) for the chromogenic assays.
5.3.2. Sample B
that there is poorer agreement between laboratories for
sample B compared to sample A, when potency is expressed
relative to the current IS. The overall inter-laboratory GCV
is 14.48% with a combined geometric mean potency of
8.13 IU/ml (n=52). There were no signicant differences
between methods with geometric mean potencies of 8.26IU/
ml (n =22) and 8.03 IU/ml (n = 30) for one-stage and
chromogenic assays, respectively. No individual laboratory
was found to be an outlier, although laboratories 1, 3 and
26 formed a group that were signicantly different from the
other laboratories. Intra-laboratory GCVs ranged from
2.34-25.81 % (n = 22) for the one-stage assays and from
1.56-29.29 % (n = 30) for the chromogenic assays.
5.3.3. Sample C
From the histogram (Fig. 3a) and Table 3, it can be seen that
there is generally good agreement between laboratories,
with an overall GCV of 5.86 %. The one-stage assays
generally gave lower results than the chromogenic assays,
with the overall means being 9.59 (n = 22) and 10.35 IU/ml
(n = 30) respectively. The difference between methods was
statistically signicant (p<0.001). No individual laboratories
were found to be outliers. Intra-laboratory GCVs ranged
from 0.91-9.86 % (n = 22) for the one-stage assays and from
5.3.4. Sample D
that there is poor agreement between laboratories, with
an overall GCV of 19.74 %, with a combined geometric
mean potency of 5.09 IU/ml (n = 52). Some laboratories in
particular give low estimates, close to 50 % of the overall
mean. There were no signicant differences between
methods with geometric mean potencies of
5.14 IU/ml (n =22) and 5.06 IU/ml (n = 30) for one-stage
and chromogenic assays, respectively. No individual
laboratory was found to be an outlier, although laboratories
1 and 29 formed a group that were signicantly different
from the other laboratories. Intra-laboratory GCVs ranged
5.3.5. Sample E
From the histogram (Fig. 5a) and Table 5, it can be seen that
there is a marked difference in results from the one-stage
and chromogenic methods, as expected. The overall GCV was
19.57 %, but for the individual assay methods was 7.04%
for one-stage and 7.63 % for chromogenic assays. The
difference between methods was signicant (p<0.001). The
overall geometric means for sample E against the current
IS were 11.80 (n = 22) and 8.49 IU/ml (n = 30) by one-stage
and chromogenic assays respectively. This compares to the
assigned values of 11.3 and 8.6 IU/ml. Intra-laboratory GCVs
ranged from 2.84-16.24% (n = 22) for the one-stage assays
and from 1.11-20.02 % (n = 30) for the chromogenic assays.
5.4. Potency Estimates Relative to BRP3
The laboratory geometric mean estimates of potency
relative to BRP3 are shown in Tables 6-10 for the one-stage
and chromogenic assay methods respectively. The intra-
laboratory %GCVs are also shown, together with the overall
geometric mean and inter-laboratory %GCVs, for both the
one-stage and chromogenic assay methods, and for all assays
combined. The potency estimates are also shown in the form
of stacking histograms (Figures 1-5). The potency estimates
calculated by the participating laboratories are shown in
Appendix II.
5.4.1. Sample A
From the histogram (Fig. 1b) and Table 6, it can be seen
that there is good agreement between laboratories and assay
methods. The overall inter-laboratory GCV is below 8.44%
with a combined geometric mean potency of
9.32 IU/ml (n = 52). The overall geometric means are
9.06 (n=22) and 9.51 IU/ml (n = 30) for one-stage
and chromogenic assays respectively. This difference
is marginally signicant (p=0.034) using a t-test, but
marginally not signicant using the non-parametric
Wilcoxon rank test (p = 0.054). Duncans test did not
indicate the presence of any outliers when applied to the
results from both methods combined. When applied to the
individual assay methods alone, laboratories 13 and 29 were
signicantly higher than other laboratories. Intra-laboratory
GCVs ranged from 1.20-13.52% (n = 22) for the one-stage
assays and from 1.00-15.96 % (n = 30) for the chromogenic
assays.
5.4.2. Sample B
From the histogram (Fig. 2b) and Table 7, it can be seen that
there is poorer agreement between laboratories for sampleB
compared to sample A, when potency is expressed relative to
BRP3. The overall inter-laboratory GCV is 16.23 % and the
combined geometric mean potency is 8.04 IU/ml (n=52).
There were no signicant differences between methods
with geometric mean potencies of 7.88 IU/ml (n=22) and
8.16 IU/ml (n = 30) for one-stage and chromogenic assays,
respectively. No individual laboratory was found to be an
outlier. Intra-laboratory GCVs ranged from 0.90-36.32 %
(n= 22) for the one-stage assays and from 1.81-24.65 %
(n=30) for the chromogenic assays.
5.4.3. Sample C
that there is reasonable agreement between laboratories,
with an overall GCV of 10.08 %. The one-stage assays
generally gave lower results than the chromogenic assays,
with the overall means being 9.21 (n = 22) and 10.43 IU/ml
(n = 30) respectively. The difference between methods was
statistically signicant (p<0.001). No individual laboratories
were found to be outliers. Intra-laboratory GCVs ranged
5.4.4. Sample D
that there is poor agreement between laboratories, with
an overall %GCV of 19.07 % and a combined geometric
mean potency of 5.05 IU/ml (n = 52). Some laboratories
in particular give low estimates. There were no signicant
differences between methods with geometric mean
potencies of 4.91 IU/ml (n = 22) and 5.15 IU/ml (n = 30)
for one-stage and chromogenic assays, respectively. No
individual laboratory was found to be an outlier when both
assay methods were combined, although laboratory 29 was
signicantly lower than the other laboratories looking at
one-stage results alone. Intra-laboratory GCVs ranged from
1.93-34.09 % (n = 22) for the one-stage assays and from
5.4.5. Sample 7
th
IS
that there is generally good agreement between laboratories,
with the exception of laboratory 29, which gave a high
estimated potency. This reects the low estimate of potency
obtained for BRP3 against the IS by laboratory 29. There
is a slight difference between methods which is signicant
(p=0.007). This difference is still signicant if laboratory29
is excluded (p=0.005). Intra-laboratory GCVs (as for
sample E vs 7
th
IS) ranged from 2.84-16.24% (n= 22) for
the one-stage assays and from 1.11-20.02% (n = 30) for the
chromogenic assays.
5.5. Stability study
Investigation of long-term stability of the 4 candidate
preparations (samples A, B, C and D) has been performed
through accelerated degradation studies which allow the
prediction of degradation rates for ampoules stored at low
temperatures (e.g. -20 C) based on the observed loss in
samples stored at elevated temperatures (e.g. +4, +20, +37,
+45 C) [7]. This is an indirect method used to determine
rate of loss based on the relationship between reaction rates
and temperature given by an Arrhenius equation and where
a rst order reaction rate is assumed [8].
Ampoules of the 4 candidate preparations were placed into
elevated temperature storage in June 2008 at NIBSC. An
initial accelerated degradation study carried out at NIBSC
has indicated very low predicted degradation rates for all
4 candidate preparations. A more extensive accelerated
degradation study involving 3 laboratories was further
performed after a storage period of 8 or 12 months.
Three laboratories assayed samples that had been stored
for 8months, and one laboratory also assayed samples
stored for 12 months. Each laboratory carried out 3 repeat
assays (either one-stage or chromogenic). All laboratories
performed both types of assays except laboratory A which
performed only chromogenic assays.
The potencies of the samples stored at elevated temperature
were expressed relative to the -20 C sample. The gures
indicate good stability, with only a drop of around 10 % after
12 months storage at + 37 C for both candidate samples A
and B.
The Arrhenius model was tted to obtain predictions of the
expected loss in potency over time. For candidate standard
samples B and D, there was insufcient degradation to be
able to obtain any estimates of expected loss. For candidate
standard samples A and C the predicted % loss per year at
the different temperatures are shown in the Table I.
The results indicate that the predicted stability of sampleC
would allow it be a suitable material as the Ph. Eur. BRP
Batch 4 for FVIII Concentrate. Furthermore, the BRP will
be monitored regularly, i.e. yearly, throughout its life-
time. To this end, a subset of 30 ampoules of the candidate
material (sample C) was set aside at -80C, at the EDQM.
The ampoules were placed into aluminium bags containing
sufcient number of containers for one monitoring assay,
i.e. 3 ampoules, according to the internal Standard Opera-
ting Procedure. These samples will be used as a reference for
real-time stability monitoring of the BRP.
6. DISCUSSION
The current Ph. Eur. BRP Batch 3/Mega 2 (US/FDA) for FVIII
Concentrate was established in 2001 through an interna-
tional collaborative study and has an assigned potency of
8.6 IU/vial for the chromogenic assay [3]. It is used as a
working standard for the determination of FVIII potency in
plasma-derived and recombinant therapeutic products by
the chromogenic assay method. Stocks of this standard are
running low. Similarly, stocks of the WHO 7
th
IS for FVIII
Concentrate were also running low and both standards were
expected to be depleted by 2009/10. Therefore, it was decided
to organize a joint international collaborative study between
NIBSC and EDQM in order to calibrate replacement batches
of both standards.
The current Ph. Eur. BRP Batch 3 for FVIII Concentrate was
calibrated against the WHO 6
th
IS (97/616) [9] and the WHO
5
th
IS (88/640) [10] for FVIII Concentrate, the Mega 1 US and
the Ph. Eur. BRP Batch 2 [11] for FVIII Concentrate in both
the one-stage clotting and the chromogenic assays[3].
Similarly, for the sake of harmonisation and continuity of
the International Unit (IU), assays on candidate materials
in this study were carried out using both types of assays
and against the current WHO 7
th
IS for FVIII Concentrate
(99/678) (sample 7
th
IS) as well as against the current Ph.
Eur. BRP Batch 3 for FVIII Concentrate (sample E).
6.1. Intra- and inter-laboratory variability
The variability of assays within laboratories differed
considerably, with GCVs ranging from 0.34 % to 36.32 %
(Tables1-10). There was no obvious trend for one method to
give better or worse inter-assay variability; lower GCVs were
associated with particular laboratories rather than particular
methods, indicating that internal quality control procedures
within laboratories are probably the most important factor
Table I - Summary results of the accelerated degradation study
Candidate Material
Lab
Mean residual potencies after storage
Mean predicted % loss
per year
(% vs -20 C ampoules)
+4C +20C +37C +45C
Sample A (07/350)
A 95 93 82
0.01% (at -20C)
0.15% (at + 4C)
B 95 93 81
C 99 95 83
C* 98 98 91 79
Sample B (07/352)
A 104 104 98
Insufficient degradation to
obtain estimates.
B 102 102 88
C 101 98 89
C* --- --- ---
Sample C (08/106)
A 90 89 86
0.52% (at -20C)
2.22% (at + 4C)
B 96 95 90
C 99 93 90
C* 101 97 91 88
Sample D (08/104)
A 98 108 98
Insufficient degradation to
obtain estimates
B 101 105 94
C 102 104 92
C* --- --- ---
- tested after storage for 8 months; * - tested after storage for 12 months;
Results are the combined mean values from 3 independent one-stage and chromogenic assays

in determining reproducibility of FVIII assays, rather than
the method used. A summary of intra-laboratory varia-
bility for estimates of all samples by all methods is given
in Table11. For all candidate samples (A-D) and for both
methods, there was a higher percentage of laboratories with
GCVs below 5% for estimates relative to the 7
th
IS than for
estimates relative to sample E (BRP3/Mega 2 US). Further-
more, intra-laboratory GCVs tended to be greater for the
recombinant candidates (sample B and D) compared to the
plasma-derived candidates (samples A and C). This may not
be so surprising as recombinant concentrates can give in-
creased assay variability, as observed in previous studies[12].
Figures generally represented good reproducibility and for
candidate samples A and C, majority of the labs (>90 %) gave
intra-laboratory GCVs < 10 %.
A summary of inter-laboratory variability for estimates of all
samples by all methods is given in Table 12. Relative to the
7
th
IS, there was very good agreement between laboratories
for the plasma-derived candidates, samples A and C, with
overall combined method GCVs of 5.0 % and 5.9 % respec-
tively. Relative to sample E (BRP3/Mega 2 US), low inter-
laboratory variability was similarly obtained for the plasma-
derived candidates, samples A and C, with overall combined
method GCVs of 8.4 % and 10.1 % respectively. There was
poorer agreement between laboratories for the recombinant
candidates, samples B and D, when assayed relative to both
standards with GCVs between 15-19.7 %. This reects the
known higher variability when assaying recombinant mate-
rials. Sample E (BRP 3/Mega 2 US) is known to have a signi-
cant one-stage/chromogenic discrepancy, however within
methods there is good inter-laboratory agreement (Fig. 5).
6.2. Potency estimates of candidate materials
Mean potency estimates by the two methods for all samples
relative to both standards are summarized in Table 12. In the
study to calibrate the current WHO 7
th
IS for FVIII Concen-
trate (99/678), there was an overall method discrepancy
of < 2 % [13]. In this study, when assayed relative to the
7
th
IS, sample A gave identical mean estimates of potency of
9.44 IU/ml (with an overall inter-laboratory GCV of < 5 %),
and with less than 5% discrepancy between methods when
assayed relative to sample E (BRP3/Mega 2 US) (Tables 12-
13). The 2 recombinant samples B and D also gave low assay
discrepancy between methods (<5 %) when assayed relative
to either standards. In contrast, sample C gave signicantly
(p<0.001) lower one-stage potencies compared to the chro-
mogenic assays, by 7 % and 12 % when assayed relative to
the 7
th
IS and sample E (BRP3/Mega 2 US) respectively. The
chromogenic potencies for sample C however were in very
good agreement relative to the two standards, being
10.4 IU/ml (n = 30) for assays against both standards and
inter-laboratory GCVs of 4.8 % and 7.1 % relative to WHO 7
th

IS and sample E (BRP3) respectively.
Furthermore, as expected, when assayed relative to the
7
th
IS, sample E gave a signicantly (p<0.001) higher
one-stage potency compared to the chromogenic assay by
39 % (Tables12-13), with one-stage potency approximately
4% higher than the value of 11.3 IU/ml observed during
the establishment study and chromogenic potency approxi-
mately 1 % lower than the assigned value of 8.6 IU/ml. This
extremely close recovery of the original assigned value for
the BRP3 in this study validates the original calibration and
indicates that there should be excellent continuity into BRP
Batch 4.
Data from the initial accelerated degradation study at NIBSC
indicated very good predicted stability for all 4 candidate
preparations. The preliminary data have been conrmed in a
more extensive study in three laboratories.
7. CONCLUSION
The candidate BRP Batch 4 for FVIII Concentrate (sample C)
was shown to be suitable as a reference standard. Based on
the results of the study, it was proposed to establish it as the
Ph. Eur. BRP Batch 4 for FVIII Concentrate with an assig-
ned potency of 10.4 IU/ampoule for the chromogenic assay
as prescribed by the Ph. Eur. monograph Assay of human
coagulation factor VIII (01/2008:20704) [14].
This proposal was adopted by the Ph. Eur. Commission at its
136
th
session in March 2010. The BRP Batch 4 will be avai-
lable from the EDQM (catalogue number H0920000) upon
exhaustion of the current batch.
The data obtained in the accelerated degradation study show
that the stability of the lyophilised candidate preparation is
good. The BRP will be monitored throughout its lifetime at
regular intervals, i.e. once per year, against samples stored at
-80 C. The BRP will be stored and shipped at -20 C.
8. ACKNOWLEDGEMENTS
The contributions of all the participants in the study are
gratefully acknowledged. We are grateful to the colleagues
in the Standards Division (Paul Jefferson and his team at
CBRM and Paul Matejtschuk and his team at TDI), NIBSC,
for ampouling and processing the candidate preparations
and for carrying out trial lls. We would also like to express
our sincere thanks to CSL Behring (Marburg, Germany)
for kindly donating the material for the production of the
candidate BRP and to Baxter AG (USA/Switzerland), Bayer
(USA), Grifols (Spain) and Talecris (USA) for their kind
donation of other materials for the study. We thank Sally
Bevan for carrying out some of the initial potency testing for
pilot lls. We also thank the ISTH/SSC Factor VIII/Factor IX
Subcommittee (chaired by Professor Alok Srivastava) for its
guidance.
We are extremely grateful to Dr Andreas Hunfeld, Rabea
Nther & Renate Nawrot at the Paul-Ehrlich Institut (Ger-
many), and to Dr Guy Rautmann & Chantal Raphalen at the
EDQM, for kindly accepting to test the accelerated degrada-
tion samples.
The study was run under the aegis of the Biological Stan-
dardisation Programme, which is a joint programme of the
Council of Europe and the European Commission.
9. PARTICIPANTS (in alphabetical order by country)
Oliva L.M., INAME Instituto Nacional de Medicamentos,
Argentina
Loh C., Liu L., Therapeutic Goods Administration, Australia
Ban A., Baxter/ IBC, Austria
Grtner P., Baxter AG, Austria
Leitner M., AGES PharmMed, Austria
Pock K., Stadler M., Octapharma GmbH, Austria
Turecek P., Gritsch H., Baxter Innovations GmbH, Austria
Weinberger J., Octapharma GmbH, Austria
Chong H., Health Canada, Canada
Kjalke M., Novo Nordisk AS, Denmark
Krogh Rasmussen J., Novo Nordisk AS, Denmark
Sandberg E., Danish Medicines Agency, Denmark
Elmlinger D., Octapharma SA, France
Michalski C., LFB Biomdicaments, France
Mouillot L., Livre V., Agence Franaise pour la Scurit
Sanitaire des Produits de Sant, France
Rautmann G., EDQM, Council of Europe, Strasbourg,
France
Auer A., Biotest AG, Germany
Blessing H., CSL Behring GmbH, Germany
Hunfeld A., Paul-Ehrlich Institut, Germany
Gambelli D., Cini E., Kedrion SpA, Italy
Sardelli R., Istituto Superiore di Sanit, Italy
Kamimura K., Kaketsuken, Japan
Taneichi M., National Institute of Infectious Diseases, Japan
Watanabe Y., Baxter Ltd, Japan
Da Silva A., INFARMED, Portugal
Inder L., National Bio-products Institute, South Africa
Alonso C., Agencia Espanola de Medicamentos y Productos
Sanitarios, Spain
Hosta N., Instituto Grifols, Spain
Rosen S., Rossix AB, Sweden
Wermeille M., Chessa J-P., Baxter AG, Switzerland
Yang Y-C., Bureau of Food and Drug Analysis Department of
Health, Taiwan
Blande A., Bio Products Laboratory, UK
Raut S., Daniels S., National Institute for Biological Stan-
dards and Control, UK
Jankowski M., Wyeth BioPharma (CAD), USA
Khabbaz N., Allyn T., Bayer Healthcare, USA
Leerabhandh M., Baxter Thousand Oaks, USA
Wong L., Baxter Ltd, USA
Wood L., Lee T., CBER/U.S. FDA, USA
10. ABBREVIATIONS
BRP: Biological Reference Preparation; BSP: Biological Stan-
dardisation Programme; CI: Condence Interval; CL:Con-
dence Limits; CV: Coefcient of Variation; ECBS:WHO
Expert Committee on Biological Standardisation; EDQM:
European Directorate for the Quality of Medicines&Heal-
thCare; FVIII: Factor VIII; FDA: Food and Drug Administra-
tion; GCV: Geometric Coefcient of Variation; GM: Geo-
metric Mean; IS: International Standard; IU: International
Units; Lab: laboratory; log: logarithm; NIBSC: National
Institute for Biological Standards and Control; OMCL:
Ofcial Medicines Control Laboratory; Ph. Eur.: European
Pharmacopoeia; RT: Room temperature; US: United States of
America; WHO: World Health Organization.
11. REFERENCES
[1] Raut S, Daniels S, Heath A. An international collabora-
tive study to assign value to the 8
th
WHO International Stan-
dard for blood coagulation Factor VIII Concentrate (07/350).
Ref: BS/09.2117. WHO Expert Committee on Biological
Standardization; 2009.
[2] Campbell PJ. International biological standards and
reference preparations. 1. Preparation and presentation of
materials to serve as standards and reference preparations.
JBiol Standardisation 1974;2:249-67.
[3] Kirschbaum N, Wood L, Lachenbruch P, Weinstein M,
Daas A, Rautmann G, Spieser JM, Buchheit KH. Calibration
of the Ph. Eur. BRP Batch 3/Mega 2 (US/FDA) standard
for human coagulation factor VIII concentrate for use in
potency assay. Pharmeuropa Bio 2002(1):31-60.
[4] Finney DJ. Statistical methods in biological assay. 3
rd
edi-
tion. London, UK: Charles Grifn; 1978.
[5] Kirkwood TBL. Geometric means and measures of dis-
persion. Biometrics 1979;35:908-9.
[6] Duncan DB. T-tests and intervals for comparisons sug-
gested by the data. Biometrics 1975;31:339-59.
[7] Kirkwood TBL, Tydeman MS. Design and analysis of
accelerated degradation tests for the stability of biological
standards II. A exible computer program for data analysis.
J Biol Standardisation 1984;12:207-14.
[8] Kirkwood, TBL. Predicting the stability of biological
standards and products. Biometrics 1977;33(4):736-42.
[9] Raut S, Heath A, Barrowcliffe TW. Establishment of the
6
th
International Standard for FVIII concentrate. Thromb
Haemost 2001;85:1071-78.
[10] Barrowcliffe TW, Kemball-Cook G, Heath A. Proposed
5
th
International Standard for Factor VIII Concentrate. Re-
port to participants. An SSC working party study on behalf
of the SSC FVIII/IX subcommittee. Unpublished report 1994.
[11] Barrowcliffe TW. Establishment of a reference prepara-
tion and evaluation of a chromogenic method. Pharmeuropa
Bio 1996(1):39-47.
[12] Raut S, Sands D, Heath AB, Barrowcliffe TW. Variability
in factor VIII concentrate measurement: results from SSC
eld collaborative studies. J Thromb Haemost 2003;1:1927-
34.
[13] Raut S, Bevan S, Hubbard AR, Sands D, Barrowcliffe
TW. A collaborative study to establish the 7
th
International
Standard for Factor VIII Concentrate. J Thromb Haemost
2005;3:119-26.
[14] Assay of Human Coagulation Factor VIII, general chap-
ter 2.7.4. Ph. Eur. 6
th
edition. Strasbourg, France: Council of
Europe; 2008.

Table 1 - Mean laboratory estimates for sample A relative to WHO 7
th
IS FVIII Concentrate (sample 7
th
IS -
11.0IU/ml) by one-stage and chromogenic assays together with estimates of intra-laboratory
variability (GCV%) for individual laboratories and inter-laboratory variability for method estimates and
combined estimates
Number of Geometri c Intra-l ab Number of Geometri c Intra-l ab
Assays Mean GCV% Assays Mean GCV%
(n) (IU/ml ) (n) (IU/ml )
1 - - - 8 9.57 4.08
2 8 9.38 6.15 8 9.32 8.66
3 - - - 8 9.85 4.16
4 8 9.76 4.38 8 10.92 3.66
5 8 9.26 3.29 8 9.15 1.93
6 8 9.39 3.19 - - -
7 - - - 8 9.71 5.67
8A 8 9.76 3.19 - - -
8B 8 9.51 5.03 - - -
9 8 9.10 4.06 8 9.34 0.34
10 8 9.37 1.55 8 8.97 3.74
11 8 9.37 4.39 8 9.15 1.22
12 - - - 8 9.18 4.26
13 8 11.25 9.67 - - -
14 - - - 8 9.63 5.83
15 - - - 8 9.52 2.34
16 8 9.72 1.53 8 9.51 0.93
17 8 9.75 0.64 8 9.54 1.48
18 - - - 8 9.35 6.16
19 8 9.19 1.62 8 9.43 2.37
20 8 9.43 2.73 - - -
21 - - - 8 9.33 3.21
22 8 9.30 11.49 - - -
23 - - - 8 9.24 9.49
24 8 8.79 4.97 8 9.59 1.93
25 - - - 8 9.73 5.45
26 - - - 4 9.12 3.78
27 8 8.79 12.81 - - -
28 - - - 8 9.76 4.26
29 8 8.35 4.20 8 8.60 4.46
30 8 9.23 0.77 - - -
31 - - - 8 9.57 9.29
32 - - - 8 9.38 7.12
33 - - - 8 9.47 3.97
34 8 9.34 2.99 - - -
35 - - - 8 9.86 5.78
36 8 10.42 5.41 - - -
37A - - - 8 9.29 1.80
37B - - - 8 9.07 3.50
38 8 9.54 2.70 8 9.39 3.55
Inter-Lab GCV = 4.11%
One-Stage Chromogeni c
Lab No.
Combi ned Geometri c Mean = 9.44 IU/ml (n=52)
Combi ned Inter-Lab GCV = 4.98%
Geometri c Mean = 9.44 (n=22) Geometri c Mean = 9.44 (n=30)
Table 2 - Mean laboratory estimates for sample B relative to WHO 7
th
IS FVIII Concentrate (sample
7
th
IS - 11.0IU/ml) by one-stage and chromogenic assays together with estimates of intra-laboratory
variability (GCV%) for individual laboratories and inter-laboratory variability for method estimates and
combined estimates
1 - - - 8 5.30 6.57
2 8 8.70 5.28 8 7.80 13.69
3 - 8 5.85 10.67
4 8 9.06 7.26 8 8.81 10.46
5 8 8.11 25.81 8 7.90 5.30
6 8 8.22 5.90 - - -
7 - - - 8 6.80 4.99
8A 8 8.07 5.12 - - -
8B 8 7.72 5.58 - - -
9 8 8.05 9.98 8 9.16 4.59
10 8 6.83 3.47 8 8.00 5.67
11 8 7.75 7.45 8 7.95 4.28
12 - - - 8 8.88 6.43
13 8 9.39 4.75 - - -
14 - - - 8 8.08 5.84
15 - - - 8 9.73 4.36
16 8 9.06 2.96 8 8.86 1.56
17 8 9.17 4.42 8 9.91 9.05
18 - - - 8 7.88 17.72
19 8 8.84 4.78 8 8.89 3.16
20 8 7.64 16.77 - - -
21 - - - 8 8.03 2.98
22 8 7.08 10.59 - - -
23 - - - 8 8.67 10.08
24 8 7.28 2.53 8 8.31 6.30
25 - - - 8 9.55 6.15
26 - - - 4 5.30 5.79
27 8 7.92 10.58 - - -
28 - - - 8 6.87 10.73
29 8 7.49 3.22 8 8.24 4.27
30 8 8.62 2.34 - - -
31 - - - 8 8.49 5.87
32 - - - 8 9.27 4.69
33 - - - 8 8.72 29.29
34 8 8.97 6.82 - - -
35 - 8 8.05 6.25
36 8 10.21 3.06 - - -
37A - - - 8 8.18 15.46
37B - - - 8 8.60 10.46
38 8 8.34 4.05 8 7.49 11.70
Inter-Lab GCV = 10.48% Inter-Lab GCV = 17.00%
Lab No.
Table 3 - Mean laboratory estimates for sample C relative to WHO 7
th
7
th
IS - 11.0IU/ml) by one-stage and chromogenic assays together with estimates of intra-
laboratory variability (GCV%) for individual laboratories and inter-laboratory variability for
method estimates and combined estimates
1 - - - 8 10.67 6.65
2 8 9.33 3.66 8 11.12 11.32
3 - - - 8 10.46 8.90
4 8 9.41 8.61 8 11.19 6.67
5 8 9.51 2.32 8 10.21 3.73
6 8 9.41 3.27 -
7 - - - 8 10.47 7.63
8A 8 9.87 2.01 - - -
8B 8 9.77 4.57 - - -
9 8 9.84 4.93 8 10.10 3.36
10 8 9.67 2.69 8 9.96 6.84
11 8 9.47 4.73 8 10.17 2.15
12 - - - 8 10.54 4.46
13 8 10.43 8.71 - - -
14 - - - 8 10.94 5.43
15 - - - 8 10.10 4.76
16 8 9.85 0.91 8 10.56 1.69
17 8 9.92 2.76 8 10.24 0.63
18 - - - 8 10.19 4.21
19 8 9.37 3.13 8 10.19 2.29
20 8 9.72 1.55 - - -
21 - - - 8 10.48 2.58
22 8 9.18 5.79 - - -
23 - - - 8 10.54 5.91
24 8 9.49 7.88 8 10.15 2.52
25 - - - 8 10.61 4.75
26 - - - 4 10.34 3.03
27 8 9.17 9.86 - - -
28 - - - 7 10.87 6.49
29 8 8.84 1.31 8 9.05 6.35
30 8 9.55 1.13 - - -
31 - - - 8 10.46 8.41
32 - - - 8 10.68 3.62
33 - - - 8 10.70 8.93
34 8 9.57 3.79 - - -
35 - - - 8 10.48 3.20
36 8 10.14 7.05 - - -
37A - - - 8 10.52 1.85
37B - - - 8 9.83 1.31
38 8 9.68 3.18 8 9.07 30.31
Lab No.
Table 4 - Mean laboratory estimates for sample D relative to WHO 7
th
7
th
IS - 11.0IU/ml) by one-stage and chromogenic assays together with estimates of intra-
laboratory variability (GCV%) for individual laboratories and inter-laboratory variability for
method estimates and combined estimates
1 - - - 8 3.07 9.39
2 8 4.89 21.02 8 5.67 10.49
3 - - - 8 3.68 9.49
4 8 5.43 7.86 8 5.82 11.24
5 8 5.12 20.57 8 5.02 7.62
6 8 5.36 3.08 - - -
7 - - - 8 3.81 5.42
8A 8 5.60 6.22 - - -
8B 8 5.42 3.47 - - -
9 8 4.73 6.05 8 6.01 2.31
10 8 4.62 4.53 8 5.29 1.75
11 8 5.08 3.14 8 5.68 2.90
12 - - - 8 5.32 7.84
13 8 5.78 4.28 - - -
14 - - - 8 5.10 8.53
15 - - - 8 6.22 3.47
16 8 5.48 0.53 8 5.50 2.83
17 8 5.75 2.51 8 6.34 4.75
18 - - - 8 4.58 3.79
19 8 5.42 4.75 8 5.70 4.28
20 8 5.74 20.69 - - -
21 - - - 8 5.68 8.26
22 8 4.63 11.93 - - -
23 - - - 8 5.05 10.01
24 8 4.51 4.39 8 5.65 10.26
25 - - - 8 6.26 2.99
26 - - - 4 3.68 6.88
27 8 5.06 8.89 - - -
28 - - - 8 4.29 9.33
29 8 2.92 14.45 8 2.99 10.50
30 8 5.47 1.12 - - -
31 - - - 8 5.36 6.16
32 - - - 8 5.72 7.19
33 - - - 8 5.58 23.26
34 8 5.60 6.14 - - -
35 - - - 8 5.41 4.26
36 8 6.43 2.29 - - -
37A - - - 8 5.43 17.17
37B - - - 8 5.63 7.87
38 8 5.21 4.87 8 4.99 7.42
Chromogeni c One-Stage
Inter-Lab GCV = 16.54% Inter-Lab GCV = 22.15%
Lab No.
Table 5 - Mean laboratory estimates for sample E (Ph. Eur. BRP Batch 3/Mega 2 US) relative to WHO 7
th
IS FVIII
Concentrate (sample 7
th
IS - 11.0 IU/ml) by one-stage and chromogenic assays together with estimates of intra-
laboratory variability (GCV%) for individual laboratories and inter-laboratory variability for method estimates
and combined estimates
1 - - - 8 8.51 5.52
2 8 13.59 5.57 8 8.64 8.76
3 - - - 8 8.55 7.85
4 8 13.18 8.35 8 9.41 14.73
5 8 12.65 9.77 8 8.74 3.48
6 8 11.66 8.02 - - -
7 - - - 8 8.62 5.98
8A 8 11.20 5.85 - - -
8B 8 11.78 3.84 - - -
9 8 10.62 4.49 8 9.05 3.28
10 8 10.94 4.78 8 8.52 2.99
11 8 11.51 4.76 8 8.42 5.40
12 - - - 8 8.10 2.81
13 8 10.87 10.42 - - -
14 - - - 8 7.86 11.10
15 - - - 8 8.58 1.11
16 8 12.76 4.31 8 8.42 3.98
17 8 11.32 2.93 8 8.79 4.56
18 - - - 8 8.53 3.10
19 8 12.65 4.28 8 8.62 2.93
20 8 12.89 2.84 - - -
21 - - - 8 8.43 5.32
22 8 11.88 12.99 - - -
23 - - - 8 8.50 5.84
24 8 11.06 4.48 8 8.76 2.76
25 - - - 8 8.36 7.15
26 - - - 8 8.71 3.39
27 8 11.25 16.24 - - -
28 - - - 8 8.43 6.08
29 8 11.31 10.55 8 6.17 10.10
30 8 11.58 2.90 - - -
Lab No.
1 - - - 8 8.51 5.52
2 8 13.59 5.57 8 8.64 8.76
3 - - - 8 8.55 7.85
4 8 13.18 8.35 8 9.41 14.73
5 8 12.65 9.77 8 8.74 3.48
6 8 11.66 8.02 - - -
7 - - - 8 8.62 5.98
8A 8 11.20 5.85 - - -
8B 8 11.78 3.84 - - -
9 8 10.62 4.49 8 9.05 3.28
10 8 10.94 4.78 8 8.52 2.99
11 8 11.51 4.76 8 8.42 5.40
12 - - - 8 8.10 2.81
13 8 10.87 10.42 - - -
14 - - - 8 7.86 11.10
15 - - - 8 8.58 1.11
16 8 12.76 4.31 8 8.42 3.98
17 8 11.32 2.93 8 8.79 4.56
18 - - - 8 8.53 3.10
19 8 12.65 4.28 8 8.62 2.93
20 8 12.89 2.84 - - -
21 - - - 8 8.43 5.32
22 8 11.88 12.99 - - -
23 - - - 8 8.50 5.84
24 8 11.06 4.48 8 8.76 2.76
25 - - - 8 8.36 7.15
26 - - - 8 8.71 3.39
27 8 11.25 16.24 - - -
28 - - - 8 8.43 6.08
29 8 11.31 10.55 8 6.17 10.10
30 8 11.58 2.90 - - -
31 - - - 8 8.71 3.35
32 - - - 8 8.77 6.75
33 - - - 8 7.55 20.02
34 8 11.81 7.46 - - -
35 - - - 8 8.82 7.71
36 8 12.37 7.17 - - -
37A - - - 8 9.11 4.70
37B - - - 8 9.12 2.88
38 8 11.37 3.28 8 8.57 18.89
Lab No.
Table 6 - Mean laboratory estimates for sample A relative to Ph. Eur. BRP Batch 3/Mega 2 US (sample E: one-
stage potency - 11.3 IU/ml; chromogenic potency- 8.6 IU/ml) by one-stage and chromogenic assays together
with estimates of intra-laboratory variability (GCV%) for individual laboratories and inter-laboratory
variability for method estimates and combined estimates
1 - - - 8 9.50 3.52
2 8 7.76 6.18 8 8.78 12.72
3 - - - 8 9.91 7.53
4 8 8.61 5.67 8 10.03 8.91
5 8 8.25 1.73 8 9.18 4.16
6 8 9.22 8.48 - - -
7 - - - 8 9.61 2.83
8A 8 9.74 4.55 - - -
8B 8 9.05 5.11 - - -
9 8 9.65 5.02 8 8.77 3.52
10 8 9.37 2.94 8 8.88 5.00
11 8 9.08 2.88 8 9.19 4.73
12 - - - 8 9.84 1.96
13 8 11.70 8.89 - - -
14 - - - 8 10.17 7.41
15 - - - 8 9.55 2.51
16 8 8.36 1.20 8 9.42 1.00
17 8 9.62 2.16 8 9.15 2.55
18 - - - 8 9.58 2.52
19 8 8.21 4.89 8 9.37 4.61
20 8 8.27 4.59 - - -
21 - - - 8 9.17 3.92
22 8 9.78 9.20 - - -
23 - - - 8 9.57 5.59
24 8 8.89 7.24 8 9.33 3.94
25 - - - 8 10.09 9.95
26 - - - 4 9.03 6.71
27 8 9.35 9.74 - - -
28 - - - 8 10.32 8.87
29 8 8.20 13.52 8 12.27 14.34
30 8 8.92 2.80 - - -
Lab No.
1 - - - 8 9.50 3.52
2 8 7.76 6.18 8 8.78 12.72
3 - - - 8 9.91 7.53
4 8 8.61 5.67 8 10.03 8.91
5 8 8.25 1.73 8 9.18 4.16
6 8 9.22 8.48 - - -
7 - - - 8 9.61 2.83
8A 8 9.74 4.55 - - -
8B 8 9.05 5.11 - - -
9 8 9.65 5.02 8 8.77 3.52
10 8 9.37 2.94 8 8.88 5.00
11 8 9.08 2.88 8 9.19 4.73
12 - - - 8 9.84 1.96
13 8 11.70 8.89 - - -
14 - - - 8 10.17 7.41
15 - - - 8 9.55 2.51
16 8 8.36 1.20 8 9.42 1.00
17 8 9.62 2.16 8 9.15 2.55
18 - - - 8 9.58 2.52
19 8 8.21 4.89 8 9.37 4.61
20 8 8.27 4.59 - - -
21 - - - 8 9.17 3.92
22 8 9.78 9.20 - - -
23 - - - 8 9.57 5.59
24 8 8.89 7.24 8 9.33 3.94
25 - - - 8 10.09 9.95
26 - - - 4 9.03 6.71
27 8 9.35 9.74 - - -
28 - - - 8 10.32 8.87
29 8 8.20 13.52 8 12.27 14.34
30 8 8.92 2.80 - - -
31 - - - 8 9.10 9.67
32 - - - 8 9.25 7.71
33 - - - 8 10.56 8.94
34 8 8.84 5.84 - - -
35 - - - 8 9.27 2.70
36 8 9.75 12.57 - - -
37A - - - 8 8.99 3.21
37B - - - 8 8.57 3.80
38 8 9.59 1.93 8 9.51 15.96
Lab No.
Table 7 - Mean laboratory estimates for sample B relative to Ph. Eur. BRP Batch 3/Mega 2 US (sample E:
one-stage potency - 11.3 IU/ml; chromogenic potency- 8.6 IU/ml) by one-stage and chromogenic assays
together with estimates of intra-laboratory variability (GCV%) for individual laboratories and inter-
laboratory variability for method estimates and combined estimates
1 - - - 8 5.44 6.86
2 8 7.27 8.36 8 8.2 17.2
3 - - - 8 5.89 14.57
4 8 7.55 2.1 8 8.01 13.23
5 8 7.27 36.32 8 7.62 3.8
6 8 7.86 4.49 - - -
7 - - - 8 6.84 1.81
8A 8 8.24 7.75 - - -
8B 8 7.46 7.24 - - -
9 8 8.59 10.42 8 8.81 4.36
10 8 7.3 5.68 8 8.25 7.43
11 8 7.7 8.48 8 8.24 7.25
12 - - - 8 9.35 4.64
13 8 9.76 9.67 - - -
14 - - - 8 9.16 14.2
15 - - - 8 9.75 4.84
16 8 8.26 4.76 8 9.33 2.65
17 8 9.27 4.75 8 9.88 12.95
18 - - - 8 7.82 18.74
19 8 7.9 4.54 8 8.91 3.78
20 8 6.69 20.1 - - -
21 - - - 8 8.49 7.13
22 8 6.09 8.58 - - -
23 - - - 8 8.58 4.55
24 8 7.52 2.57 8 8.24 7.57
25 - - - 8 9.66 10.73
26 - - - 4 5.22 6.44
27 8 7.51 6.86 - - -
28 - - - 8 6.76 9.83
29 8 7.62 5.08 8 11.22 6.09
30 8 8 5 0 9 - - -
Lab No.
1 - - - 8 5.44 6.86
2 8 7.27 8.36 8 8.2 17.2
3 - - - 8 5.89 14.57
4 8 7.55 2.1 8 8.01 13.23
5 8 7.27 36.32 8 7.62 3.8
6 8 7.86 4.49 - - -
7 - - - 8 6.84 1.81
8A 8 8.24 7.75 - - -
8B 8 7.46 7.24 - - -
9 8 8.59 10.42 8 8.81 4.36
10 8 7.3 5.68 8 8.25 7.43
11 8 7.7 8.48 8 8.24 7.25
12 - - - 8 9.35 4.64
13 8 9.76 9.67 - - -
14 - - - 8 9.16 14.2
15 - - - 8 9.75 4.84
16 8 8.26 4.76 8 9.33 2.65
17 8 9.27 4.75 8 9.88 12.95
18 - - - 8 7.82 18.74
19 8 7.9 4.54 8 8.91 3.78
20 8 6.69 20.1 - - -
21 - - - 8 8.49 7.13
22 8 6.09 8.58 - - -
23 - - - 8 8.58 4.55
24 8 7.52 2.57 8 8.24 7.57
25 - - - 8 9.66 10.73
26 - - - 4 5.22 6.44
27 8 7.51 6.86 - - -
28 - - - 8 6.76 9.83
29 8 7.62 5.08 8 11.22 6.09
30 8 8.5 0.9 - - -
31 - - - 8 8.52 5.15
32 - - - 8 9.05 14.52
33 - - - 8 10.15 11.92
34 8 8.67 12.32 - - -
35 - - - 8 8.13 7.38
36 8 9.11 6.27 - - -
37A - - - 8 7.54 19.31
37B - - - 8 8.1 9.96
38 8 8.2 5.01 8 7.15 24.65
Lab No.
Table 8 - Mean laboratory estimates for sample C relative to Ph. Eur. BRP Batch 3/Mega 2 US (sample E: one-
1 - - - 8 10.60 1.26
2 8 7.71 8.36 8 10.49 11.95
3 - - - 8 10.52 8.56
4 8 8.30 7.83 8 10.28 13.43
5 8 8.47 3.30 8 10.25 5.17
6 8 9.23 9.12 - - -
7 - - - 8 10.36 2.78
8A 8 9.84 4.32 - - -
8B 8 9.29 4.92 - - -
9 8 10.44 5.80 8 9.49 4.23
10 8 9.66 3.17 8 9.86 6.27
11 8 9.18 5.20 8 10.22 3.57
12 - - - 8 11.30 5.31
13 8 10.85 2.54 - - -
14 - - - 8 11.55 7.69
15 - - - 8 10.13 5.52
16 8 8.47 1.54 8 10.46 3.27
17 8 9.78 3.01 8 9.82 1.73
18 - - - 8 10.44 2.06
19 8 8.37 4.60 8 10.12 4.37
20 8 8.53 2.66 - - -
21 - - - 8 10.30 0.92
22 8 9.65 6.41 - - -
23 - - - 8 10.91 0.74
24 8 9.59 5.12 8 9.87 5.33
25 - - - 8 10.69 8.77
26 - - - 4 10.23 4.04
27 8 9.76 10.85 - - -
28 - - - 7 11.25 1.69
29 8 8.68 16.47 8 12.92 16.93
30 8 9 22 2 58 - - -
Lab No.
1 - - - 8 10.60 1.26
2 8 7.71 8.36 8 10.49 11.95
3 - - - 8 10.52 8.56
4 8 8.30 7.83 8 10.28 13.43
5 8 8.47 3.30 8 10.25 5.17
6 8 9.23 9.12 - - -
7 - - - 8 10.36 2.78
8A 8 9.84 4.32 - - -
8B 8 9.29 4.92 - - -
9 8 10.44 5.80 8 9.49 4.23
10 8 9.66 3.17 8 9.86 6.27
11 8 9.18 5.20 8 10.22 3.57
12 - - - 8 11.30 5.31
13 8 10.85 2.54 - - -
14 - - - 8 11.55 7.69
15 - - - 8 10.13 5.52
16 8 8.47 1.54 8 10.46 3.27
17 8 9.78 3.01 8 9.82 1.73
18 - - - 8 10.44 2.06
19 8 8.37 4.60 8 10.12 4.37
20 8 8.53 2.66 - - -
21 - - - 8 10.30 0.92
22 8 9.65 6.41 - - -
23 - - - 8 10.91 0.74
24 8 9.59 5.12 8 9.87 5.33
25 - - - 8 10.69 8.77
26 - - - 4 10.23 4.04
27 8 9.76 10.85 - - -
28 - - - 7 11.25 1.69
29 8 8.68 16.47 8 12.92 16.93
30 8 9.22 2.58 - - -
31 - - - 8 10.15 9.29
32 - - - 8 10.52 4.55
33 - - - 8 11.93 20.21
34 8 9.06 8.58 - - -
35 - - - 8 9.86 3.04
36 8 9.48 10.46 - - -
37A - - - 8 10.18 1.47
37B - - - 8 9.28 2.61
38 8 9.72 0.77 8 9.59 24.27
Lab No.
Table 9 - Mean laboratory estimates for sample D relative to Ph. Eur. BRP Batch 3/Mega 2 US (sample E: one-
Number of Geometric Intra-l ab Number of Geometri c Intra-lab
1 - - - 8 3.16 9.85
2 8 4.09 26.85 8 5.96 14.53
3 - - - 8 3.70 14.94
4 8 4.52 5.29 8 5.29 13.35
5 8 4.59 34.09 8 4.84 8.01
6 8 5.13 3.34 - - -
7 - - - 8 3.83 5.20
8A 8 5.72 9.93 - - -
8B 8 5.24 3.94 - - -
9 8 5.05 9.14 8 5.78 1.79
10 8 4.93 4.02 8 5.45 3.73
11 8 5.06 3.98 8 5.89 6.73
12 - - - 8 5.62 6.84
13 8 6.01 10.22 - - -
14 - - - 8 5.78 9.94
15 - - - 8 6.23 3.86
16 8 5.00 3.95 8 5.79 4.64
17 8 5.81 3.02 8 6.32 9.08
18 - - - 8 4.55 4.82
19 8 4.84 6.79 8 5.72 6.64
20 8 5.03 17.07 - - -
21 - - - 8 6.01 4.44
22 8 3.98 11.51 - - -
23 - - - 8 4.99 7.72
24 8 4.66 4.32 8 5.60 11.04
25 - - - 8 6.71 5.70
26 - - - 8 3.62 8.19
27 8 4.80 7.50 - - -
28 - - - 8 4.22 8.14
Lab No.
Number of Geometric Intra-l ab Number of Geometri c Intra-lab
1 - - - 8 3.16 9.85
2 8 4.09 26.85 8 5.96 14.53
3 - - - 8 3.70 14.94
4 8 4.52 5.29 8 5.29 13.35
5 8 4.59 34.09 8 4.84 8.01
6 8 5.13 3.34 - - -
7 - - - 8 3.83 5.20
8A 8 5.72 9.93 - - -
8B 8 5.24 3.94 - - -
9 8 5.05 9.14 8 5.78 1.79
10 8 4.93 4.02 8 5.45 3.73
11 8 5.06 3.98 8 5.89 6.73
12 - - - 8 5.62 6.84
13 8 6.01 10.22 - - -
14 - - - 8 5.78 9.94
15 - - - 8 6.23 3.86
16 8 5.00 3.95 8 5.79 4.64
17 8 5.81 3.02 8 6.32 9.08
18 - - - 8 4.55 4.82
19 8 4.84 6.79 8 5.72 6.64
20 8 5.03 17.07 - - -
21 - - - 8 6.01 4.44
22 8 3.98 11.51 - - -
23 - - - 8 4.99 7.72
24 8 4.66 4.32 8 5.60 11.04
25 - - - 8 6.71 5.70
26 - - - 8 3.62 8.19
27 8 4.80 7.50 - - -
28 - - - 8 4.22 8.14
29 8 2.96 9.72 8 4.07 16.02
30 8 5.39 1.93 - - -
31 - - - 8 5.37 6.57
32 - - - 8 5.58 9.89
33 - - - 8 6.50 7.02
34 8 5.42 13.25 - - -
35 - - - 8 5.47 10.43
36 8 5.74 4.03 - - -
37A - - - 8 5.01 21.44
37B - - - 8 5.30 6.87
38 8 5.12 6.43 8 4.76 21.96
Combined Geometri c Mean = 5.05 IU/ml (n=52)
Combined Inter-Lab GCV = 19.07%
Lab No.
Geometri c Mean = 4.91 (n=22) Geometric Mean = 5.15 (n=30)
Table 10 - Mean laboratory estimates for sample 7
th
IS (WHO 7
th
IS - 99/678) relative to Ph. Eur. BRP Batch
3/Mega 2 US (sample E: one-stage potency - 11.3 IU/ml; chromogenic potency- 8.6 IU/ml) by one-stage and
chromogenic assays together with estimates of intra-laboratory variability (GCV%) for individual
laboratories and inter-laboratory variability for method estimates and combined estimates
1 - - - 8 11.11 5.52
2 8 9.14 5.57 8 10.95 8.76
3 - - - 8 11.06 7.85
4 8 9.43 8.35 8 10.05 14.73
5 8 9.83 9.77 8 10.83 3.48
6 8 10.66 8.02 - - -
7 - - - 8 10.97 5.98
8A 8 11.10 5.85 - - -
8B 8 10.55 3.84 - - -
9 8 11.70 4.49 8 10.45 3.28
10 8 11.36 4.78 8 11.11 2.99
11 8 10.80 4.76 8 11.23 5.40
12 - - - 8 11.68 2.81
13 8 11.44 10.42 - - -
14 - - - 8 12.03 11.10
15 - - - 8 11.02 1.11
16 8 9.74 4.31 8 11.23 3.98
17 8 10.98 2.93 8 10.76 4.56
18 - - - 8 11.09 3.10
19 8 9.83 4.28 8 10.98 2.93
20 8 9.64 2.84 - - -
21 - - - 8 11.22 5.32
22 8 10.46 12.99 - - -
23 - - - 8 11.13 5.84
24 8 11.24 4.48 8 10.80 2.76
25 - - - 8 11.32 7.15
26 - - - 8 10.86 3.39
27 8 11.05 16.24 - - -
28 - - - 8 11.22 6.08
29 8 10.99 10.55 8 15.33 10.10
30 8 10 73 2 90 - - -
Lab No.
1 - - - 8 11.11 5.52
2 8 9.14 5.57 8 10.95 8.76
3 - - - 8 11.06 7.85
4 8 9.43 8.35 8 10.05 14.73
5 8 9.83 9.77 8 10.83 3.48
6 8 10.66 8.02 - - -
7 - - - 8 10.97 5.98
8A 8 11.10 5.85 - - -
8B 8 10.55 3.84 - - -
9 8 11.70 4.49 8 10.45 3.28
10 8 11.36 4.78 8 11.11 2.99
11 8 10.80 4.76 8 11.23 5.40
12 - - - 8 11.68 2.81
13 8 11.44 10.42 - - -
14 - - - 8 12.03 11.10
15 - - - 8 11.02 1.11
16 8 9.74 4.31 8 11.23 3.98
17 8 10.98 2.93 8 10.76 4.56
18 - - - 8 11.09 3.10
19 8 9.83 4.28 8 10.98 2.93
20 8 9.64 2.84 - - -
21 - - - 8 11.22 5.32
22 8 10.46 12.99 - - -
23 - - - 8 11.13 5.84
24 8 11.24 4.48 8 10.80 2.76
25 - - - 8 11.32 7.15
26 - - - 8 10.86 3.39
27 8 11.05 16.24 - - -
28 - - - 8 11.22 6.08
29 8 10.99 10.55 8 15.33 10.10
30 8 10.73 2.90 - - -
31 - - - 8 10.86 3.35
32 - - - 8 10.79 6.75
33 - - - 8 12.53 20.02
34 8 10.52 7.46 - - -
35 - - - 8 10.73 7.71
36 8 10.05 7.17 - - -
37A - - - 8 10.39 4.70
37B - - - 8 10.37 2.88
38 8 10.93 3.28 8 11.04 18.89
Combi ned Geometric Mean = 10.88 IU/ml (n=52)
Lab No.
Table 11 - Intra-laboratory variability (GCV%) for estimates of samples A, B, C, D, E and 7
th
IS by all methods.
Figures indicate number of laboratories with GCVs < 5% or > 10%
Table 12 - Summary table showing the overall mean potency estimates (IU/ml) for the 2 methods (one-stage
and chromogenic assays) and the combined mean potency estimates (IU/ml) of samples A, B, C, D, E and 7
th

IS together with the inter-laboratory variability (GCV%)
Inter-laboratory GCVs are given in parenthesis.
*Difference between methods statistically signicant (p<0.001)
$
Difference between methods statistically signicant at p=0.007
Table 13 - Summary table showing the mean potency estimates for the 2 methods (one-stage and chromogenic
assays) as percentages of overall means
GCVs < 5% GCVs > 10% GCVs < 5% GCVs > 10% GCVs < 10%
A 39/52 (75%) 2/52 (4%) 27/52 (52%) 5/52 (10%) 97/104 (93%)
B 20/52 (39%) 14/52 (27%) 15/52 (29%) 15/52 (29%) 75/104 (72%)
C 33/52 (64%) 2/52 (4%) 27/52 (52%) 8/52 (15%) 94/104 (90%)
D 24/52 (46%) 12/52 (23%) 15/52 (29%) 14/52 (27%) 78/104 (75%)
E 27/52 (52%) 11/52 (21%) - - -
7th IS - - 25/48 (75%) 9/52 (17%) -
Sampl es Assay vs 7th IS Assays vs BRP3
Assays vs both
standards
Combi ned
Potency
Combi ned
Potency
(IU/ml ) (IU/ml ) (IU/ml ) (IU/ml ) (IU/ml ) (IU/ml )
(n=22) (n=30) (n=52) (n=22) (n=30) (n=52)
A 9.44 (6.09%) 9.44 (4.11%) 9.44 (4.98%) 9.06 (9.29%) 9.51 (7.19%) 9.32 (8.44%)
B 8.26 (10.48%) 8.03 (17.00%) 8.13 (14.48%) 7.88 (11.39%) 8.16 (19.18%) 8.04 (16.23%)
C 9.59* (3.65%) 10.35 (4.83%) 10.02 (5.86%) 9.21* (8.48%) 10.43 (7.10%) 9.89 (10.08%)
D 5.14 (16.54%) 5.06 (22.15%) 5.09 (19.74%) 4.91 (16.56%) 5.15 (20.73%) 5.05 (19.07%)
E 11.80* (7.04%) 8.49 (7.63%) 9.76 (19.57%) - - -
7th IS - - - 10.53
$
(7.04%) 11.14 (7.63%) 10.88 (7.89%)
Sampl es
Assay vs 7th IS Assays vs BRP3
One-Stage Chromogeni c One-Stage Chromogeni c
(n=22) (n=30) (n=22) (n=30)
A 100.0% 100.0% 97.2% 102.0%
B 101.6% 98.8% 98.0% 101.5%
C 95.7% 103.3% 93.1% 105.5%
D 101.0% 99.4% 97.2% 102.0%
E 120.9% 87.0% - -
7th IS - - 96.8% 102.4%
Samples
Assay vs 7th IS Assays vs BRP3
Figure 1 - Stacking histograms showing mean laboratory potency estimates (as a % of Overall Mean) for
sampleA relative to: (a) 7
th
IS and (b) BRP Batch 3. Results from one-stage assays are shaded in grey;
results from chromogenic assays are unshaded (white)
Sample A vs IS
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
% of overall mean potency
50 60 70 80 90 100 110 120 130 140 150
29
29
09
19
24
27
05
10
11
12
26
37B
02
05
06
08B
10
11
20
22
30
34
38
01
02
09
14
15
16
17
18
19
21
23
24
31
32
33
37A
38
04
08A
16
17
03
07
25
28
35
36 04 13
(a)
(b)
Sample A vs E
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
50 60 70 80 90 100 110 120 130 140 150
02 04
05
16
19
20
29
37B
08B
11
24
30
34
02
09
10
26
37A
06
10
27
01
05
11
15
16
17
19
21
24
31
32
35
38
08A
09
17
22
36
38
03
07
12
18
23
04
14
25
28
33 13 29
sampleB relative to: (a) 7
th
(a)
(b)
Sample B vs IS
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
50 60 70 80 90 100 110 120 130 140 150
01
26
03 10
22
07
28
24
29
38
08B
11
20
27
02
05
18
05
06
08A
09
10
11
14
21
24
29
35
37A
02
30
38
23
31
33
37B
04
16
19
34
04
12
16
19
13
17
09
25
32
15
17
36
Sample B vs E
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
50 60 70 80 90 100 110 120 130 140 150
26 01 22
03
20
07
28
02
05
10
38
04
08B
11
24
27
29
05
18
37A
06
08A
19
38
02
04
24
35
37B
09
16
30
10
11
21
23
31
34
09
19
17
36
12
14
16
32
13
15
25
17
33
29
sampleC relative to: (a) 7
th
(a)
(b)
Sample C vs IS
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
50 60 70 80 90 100 110 120 130 140 150
22
27
29
29
38
02
04
05
06
08B
10
11
19
20
24
30
34
38
08A
09
16
17
36
05
09
10
11
15
17
18
19
24
37B
13
01
03
07
12
16
21
23
25
26
31
32
33
35
37A
02
04
14
28
Sample C vs E
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
50 60 70 80 90 100 110 120 130 140 150
02 04
05
16
19
20
29
34
06
08B
11
24
30
36
09
37B
38
08A
10
17
22
27
38
10
15
17
19
24
35
09
01
02
03
04
05
07
11
16
18
21
26
31
32
37A
13
23
25
12
14
28
33 29
Figure 4 - Stacking histogram showing mean laboratory potency estimates (as a % of Overall Mean) for
sampleD relative to: (a) 7
th
(a)
(b)
Sample D vs IS
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
50 60 70 80 90 100 110 120 130 140 150
29 01
29
03
26
07 28 10
22
24
18
02
09
05
11
27
38
05
14
23
38
04
06
08B
19
30
10
12
31
35
37A
08A
16
34
02
11
16
19
21
24
32
33
37B
13
17
20
04
09
15
36
17
25
Sample D vs E
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
50 60 70 80 90 100 110 120 130 140 150
29 01 26 03
07
02
22
29
28 04
05
24
18
19
27
05
38
06
09
10
11
16
20
38
23
37A
08B
30
34
04
31
37B
10
12
24
32
35
08A
17
36
09
11
14
16
19
13
02
21
15
17
33 25
Figure 5 - Stacking histogram showing mean laboratory potency estimates (as a % of Overall Mean) for: (a)
sample E (BRP Batch 3) relative to 7
th
IS and (b) sample 7
th
IS relative to E (BRP Batch 3). Results from one-
stage assays are shaded in grey; results from chromogenic assays are unshaded (white)
(a)
(b)
Sample E vs IS
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
50 60 70 80 90 100 110 120 130 140 150
29 33 14 01
10
11
12
16
18
21
23
25
28
02
03
05
07
15
17
19
24
26
31
32
35
38
04
09
37A
37B
09
10
13
08A
17
24
27
29
38
06
08B
11
22
30
34
36 05
16
19
20
04 02
Sample IS vs E
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
50 60 70 80 90 100 110 120 130 140 150
02
04
05
16
19
20
36
04
08B
22
34
09
37A
37B
06
08A
11
17
27
29
30
38
01
02
03
05
07
10
15
17
18
19
23
24
26
31
32
35
38
10
13
24
11
12
16
21
25
28
09
14
33 29
APPENDIX I:
Methods and reagents used by the participants in the study
Lab Method
Pre-dilution:
Buffer or Def Plasma
Source of Factor VIII
Deficient Plasma
Analyser
APTT Reagent Used/
Chromogenic Kit
Used/Kit
1 Chromogenic Def Plasma (P) Diagnostica Stago iEMS Reader
COAMATIC FVIII
Chromogenix
2 Chromogenic P Dade-Behring Dade-Behring BCS
COAMATIC FVIII
Chromogenix
2 One-Stage P Dade-Behring Dade-Behring BCS
Dade-Behring; Dade
Actin FSL
3 Chromogenic P Werfen Austria ACL 3000
COAMATIC FVIII
Chromogenix
4 Chromogenic P Dade-Behring
Bio-TEK: Synergy
HT Reader
COAMATIC FVIII
Chromogenix
4 One-Stage P Dade-Behring
Trinity Biotechnology:
KC-40 Micro
TM

Dade-Behring; Dade
Actin FSL
5 Chromogenic P Siemens
Siemens: Behring
Coagulation Timer
COAMATIC FVIII
Chromogenix
5 One-Stage P Siemens Schnitger-Gross
Siemens; Pathromtin
SL
6 One-Stage P Dade-Behring Sysmex: Coagrex-100S
Dade-Behring; Dade
Actin Activated
Cephaloplastin Reagent
7 Chromogenic Buffer Not Used Not Given
COAMATIC FVIII
Chromogenix
8A One-Stage P HRF Inc.
Beckman Coulter ACL
Advance
Beckman Coulter
APTT SP lyophilized
8B One-Stage P HRF Inc.
General Diagnostics:
Coag-A-Mate X2
BioMerieux Automated
APTT
9 Chromogenic P Dade-Behring ACL Advance
COAMATIC FVIII
Chromogenix
9 One-Stage P Dade-Behring Dade-Behring BCS-XP
Dade-Behring; Actin
FS
10 Chromogenic P George King
Tecan Robotic
Processor
COATEST SP4 FVIII
Chromogenix
10 One-Stage P George King
BioMerieux: Coag-a-
Mate MTX
BioMerieux Micronized
silica/Platelin L
11 Chromogenic P Helena Biosciences
ACL 200/ACL 200
Plus/ACL 3000 Plus
COAMATIC FVIII
Chromogenix
11 One-Stage P Helena Biosciences
ACL 200/ACL 200
Plus/ACL 3000 Plus
IL; Micronized silica +
bovine brain (lyo)
VersaMax Molecular
Devices
COAMATIC FVIII
Chromogenix
13 One-Stage P Dade-Behring ACL Elite Pro Ilex APTT reagent
Spectra Max Plus micro
plate reader
COAMATIC FVIII
Chromogenix
15 Chromogenic P Dade-Behring ACL Advance
COAMATIC FVIII
Chromogenix
16 Chromogenic P Dade-Behring Plate Reader
COAMATIC FVIII
Chromogenix
16 One-Stage P Dade-Behring
STAGO Diagnostica
STA-R
IL; HemosIL APTT-SP
(liquid)
17 Chromogenic P Dade-Behring ACL Futura
COATEST SP4 FVIII
Chromogenix
17 One-Stage P Dade-Behring ACL Futura
IL; HemosIL APTT-SP
(liquid)
18 Chromogenic P Dade-Behring Not Given
COAMATIC FVIII
Chromogenix
Analyser
Used/Kit
19 Chromogenic P Dade-Behring Automate Sysmex CA 540
Dade-Behring Chromogenic
Kit
19 One-Stage P Dade-Behring Automate Sysmex CA540
Dade-Behring; Dade Actin
FSL
20 One-Stage P Siemens Amelung KC-10A micro
Siemens; Dade Actin Activated
Cephaloplastin
21 Chromogenic P Technoclone iEMS Reader
COAMATIC FVIII
Chromogenix
22 One-Stage P Dade-Behring ACL 9000 IL; SynthasiL
23 Chromogenic P Dade-Behring iEMS Reader MF
COAMATIC FVIII
Chromogenix
24 Chromogenic P Baxter Plate Reader Baxter; Immunochrom FVIII C
24 One-Stage P Baxter Amelung KC-4A micro Technoclone; DAPTTIN
25 Chromogenic P Dade-Behring ACL ?
COAMATIC FVIII
Chromogenix
26 Chromogenic P Diagnostic Grifols ACL 3000 & ACL Elite Pro
COAMATIC FVIII
Chromogenix
27 One-Stage P HRF Inc. ACL 7000 BioMerieux Automated APTT
28 Chromogenic P Diagnostic Grifols Amelung: AMAX CS-190
COAMATIC FVIII
Chromogenix
29 Chromogenic P HRF Inc. ACL Advance
COAMATIC FVIII
Chromogenix
29 One-Stage P HRF Inc. ACL Advance Trinity
30 One-Stage P Vital Products Sysmex CA-50 Dade-Behring
Molecular Devices: T-Max
reader
COATEST SP4 FVIII
Chromogenix
32 Chromogenic P Dade-Behring ACL 3000 Plus
COATEST SP4 FVIII
Chromogenix
33 Chromogenic P Dade-Behring Plate Reader
Kit
34 One-Stage P HRF Inc. Dade-Behring BCS
Pacific Hemostasis;
KONTACT
35 Chromogenic P Dade-Behring Dade-Behring BCS-XP
Kit
36 One-Stage P In-house SHP Dade-Behring BCS
Pacific Hemostasis;
KONTACT
37A Chromogenic P Not Given Not Given Not Given
37B Chromogenic P Not Given Not Given Not Given
38 Chromogenic P Baxter Plate Reader
COATEST SP4 FVIII
Chromogenix
38 One-Stage P Baxter Dade-Behring BCS-XP Technoclone; DAPTTIN
Lab Method
Pre-dilution:
Buffer or Def
Plasma
Source of Factor VIII
Deficient Plasma
APTT Reagent Used/
Chromogenic Kit
APPENDIX II:
Laboratories Own Calculations / Potency relative to IS (= 11.0 IU/ml) / One-stage
assays
A B C D E
Lab Method
2 9.72 8.70 9.41 4.87 13.61
4 9.76 9.06 9.46 5.44 13.18
5 9.30 7.10 9.50 4.60 12.70
6 9.60 8.30 9.60 5.40 11.60
08A 9.80 8.10 9.90 5.50 12.00
08B 9.50 7.80 9.80 5.50 12.60
9 9.10 8.07 9.82 4.74 10.60
10 9.35 6.83 9.66 4.62 10.94
11 9.37 8.22 9.47 5.08 11.54
13 10.99 9.23 10.24 5.67 10.85
16 9.70 9.10 9.80 5.50 12.70
17 9.54 8.64 9.84 5.57 11.52
19 9.20 8.80 9.40 5.40 12.80
20 9.35 7.62 9.67 5.77 12.80
22 10.71 7.28 10.88 7.97 12.17
24 9.07 7.27 9.51 4.63 10.58
27 8.97 8.15 9.48 5.20 11.67
29 9.20 8.00 9.90 5.60 11.70
30 9.23 8.59 9.56 5.46 11.53
34 9.34 8.97 9.58 5.61 11.84
36 9.55 9.38 9.31 5.91 10.83
38 9.56 8.35 9.69 5.22 11.09
Sampl e
One-Stage
Laboratories Own Calculations / Potency relative to IS (= 11.0 IU/ml) / Chromogenic
assays
A B C D E
Lab Method
1 9.4 5.3 10.6 3.1 8.5
2 9.74 7.78 11.25 5.58 8.74
3 9.93 6.46 11.12 3.91 8.66
4 11.01 8.53 11.32 5.67 9.34
5 9.1 7.9 10.2 5 8.7
7 9.5 6.75 10.37 3.82 8.52
9 9.37 9.15 10.11 6 8.98
10 8.97 8.05 10.01 5.29 8.53
11 9.14 8.38 10.14 5.68 8.55
12 9.6 9.2 10.9 5.5 8.9
14 . . . . .
15 9.6 9.9 10.5 6.2 8.7
16 9.5 8.8 10.2 5.5 8.5
17 9.5 9.94 10.19 6.34 8.74
18 9.37 7.71 10.02 4.58 8.43
19 9.6 8.9 10.2 5.9 8.7
21 9.1 8.1 10.2 5.7 8.2
23 9.26 8.59 10.48 5.08 8.5
24 9.65 8.39 10.2 5.74 8.92
25 9.82 9.32 10.35 6.14 8.38
26 9 5.26 10.23 3.62 8.66
28 10.2 7.4 11 4.4 8.5
29 9.4 9.1 9.8 5.4 8.5
31 9.71 8.6 10.6 5.44 8.73
32 9.09 9.11 10.73 6.07 8.98
33 9.9 9.5 10.65 5.43 7.7
35 9.74 8.05 10.48 5.41 8.82
37A 9.42 8.3 10.64 5.53 8.19
37B 9.18 8.76 9.91 5.71 8.16
38 9.66 7.62 10.9 5.07 9.07
Chromogeni c
Sampl e
Laboratories Own Calculations / Potency relative to BRP3 (= 11.3 IU/ml) / One-stage
assays
A B C D IS
Lab Method
02 7.90 7.45 7.88 4.03 9.11
4 8.42 7.39 8.15 4.45 9.23
5 9.20 6.80 9.40 4.40 10.80
6 9.30 8.20 9.30 5.30 10.70
08A 9.20 7.60 9.30 5.20 10.30
08B 8.60 6.90 8.80 4.90 9.90
9 9.70 8.62 10.47 5.06 11.74
10 9.16 7.18 9.45 4.84 11.10
11 9.17 8.12 9.25 5.00 10.99
13 11.43 9.66 10.63 5.92 11.23
16 8.50 8.50 8.60 5.10 9.90
17 9.17 8.69 9.46 5.60 10.38
19 8.40 8.10 8.60 5.00 10.00
20 8.27 6.77 8.52 5.07 9.56
22 10.06 6.94 9.87 7.55 10.87
24 9.28 8.03 9.67 5.21 11.06
27 9.51 7.73 9.73 4.91 11.15
29 . . . . .
30 9.00 8.59 9.31 5.45 10.82
34 8.93 8.81 9.16 5.50 10.61
36 10.09 9.40 9.80 5.91 11.39
38 9.62 8.28 9.75 5.17 10.96
One-Stage
Sampl e
Laboratories Own Calculations / Potency relative to BRP3 (= 8.6 IU/ml) /
Chromogenicassays
Sampl e
A B C D IS
Lab Method
1 9.4 5.3 10.6 3.2 11.1
2 9.13 8.41 10.31 5.86 11.12
3 9.93 6.47 11.1 3.97 10.97
4 9.52 7.39 9.8 4.91 9.55
5 9.1 7.6 10.3 4.8 10.8
7 9.65 6.78 10.53 3.84 11.11
9 8.81 8.94 9.5 5.87 10.55
10 8.36 7.82 9.58 5.16 10.36
11 9.19 8.43 10.23 5.72 11.06
12 9.6 9 10.8 5.5 11.2
14 10.3 9.5 11.7 6 12.4
15 9.7 10 10.5 6.3 11.2
16 9.4 9.2 10.1 5.7 11.2
17 9.21 9.94 9.87 6.34 10.8
18 9.76 7.7 10.42 4.56 11.19
19 8.8 8.2 9.3 5.4 10.2
21 8.6 8 9.6 5.6 10.5
23 8.89 7.91 10.18 4.7 10.36
24 9.39 8.45 9.94 5.76 11.01
25 10.29 10.18 11.04 6.83 11.39
26 8.85 5.17 10.1 3.54 10.7
28 10.8 7.3 11.5 4.4 11.1
29 . . . . .
31 9.55 8.66 10.48 5.37 10.85
32 8.35 7.95 10.13 5.35 9.79
33 10.74 10.45 11.77 6.5 12.54
35 9.16 8.19 9.86 5.52 10.78
37A 9.14 7.71 10.32 5.14 9.38
37B 8.72 8.27 9.43 5.4 9.39
38 8.71 6.64 9.86 4.45 9.89
Chromogeni c
Heparin Low-Molecular mass for Assay BRP
1. INTRODUCTION
Low-molecular-mass heparins are polysaccharides
fractionated from heparin of animal origin, which
are widely used in many clinical indications for
their highly efficient anti-coagulant properties.
The Ph.Eur. monograph Heparin, low-molecular-
mass(0828) requires that the anti-coagulant activity of
low-molecular-mass heparin products is determined by
2in vitro assays measuring the ability of the product to
accelerate the inhibition of factorXa (anti-Xa assay) and
factorIIa/thrombin (anti-IIa assay) by antithrombinIII[1].
These 2chromogenic assays have to be performed using a
reference standard such as the Heparin low-molecular-mass
for assay BRP to express activities inIU/volume.
The current batch of Heparin low-molecular-mass for assay
BRP (batch5; BRP5) was established in March2004, with
assigned anti-Xa and anti-IIa activities of 1281IU/ampoule
and 301IU/ampoule respectively[2]. Due to dwindling
stocks of BRP5, a collaborative study (codedBSP096) was
launched in the context of the BSP under the auspices of
the Council of Europe and the European Commission to
establish replacement batches.
This current publication reports the results of the study
which aimed at calibrating 3candidate materials against
the current 2
nd
IS for Heparin, low-molecular-weight,
using the chromogenic assays described in the Ph.Eur.
texts. The current Heparin low-molecular-mass for
assay BRP was included in the sample panel to assure
the continuity between subsequent BRP batches.
2. PARTICIPANTS
Fourteen European and non-European laboratories,
including Official Medicines Control Laboratories (OMCLs)
and manufacturers from 10countries (Australia, Austria,
Canada, Denmark, France, Germany, Portugal, Sweden,
Switzerland, UK) and the Council of Europe/EDQM
laboratory enrolled for this study; 13participants returned
results. In this publication each of these laboratories is
referred to by an arbitrarily assigned number not necessarily
representing the order of the listing in section8.
3. MATERIALS
3.1. Starting material
The candidate BRPs (cBRPs) were prepared from a
batch of starting material of porcine origin, generously
donated by a European manufacturer, that was
compliant with the specifications of the Ph.Eur.
monograph Heparin, low-molecular-mass(0828)[1].
3.2. Production of candidate batches
Three bulk aqueous solutions of low-molecular-mass
heparin were filled in type I white glass vials and freeze-dried
using similar protocols. The mean filling volumes for the
3 cBRPs (samplesC, D &E) were respectively: 1.004, 1.005
and0.997g, with a coefficient of variation of0.08, 0.03
and 0.54% (n=34, n=26, n=9). A batch homogeneity study
performed prior to the collaborative study confirmed the
homogeneity of each batch. The candidate batches were kept
at4C.
Accelerated degradation studies indirectly assessing the
stability of the cBRPs were launched. Results collected
Collaborative Study for the Establishment of
Replacement Batches of Heparin Low-Molecular-Mass
for Assay Biological Reference Preparations
E. Terao, A. Daas, G. Rautmann, K.-H. Buchheit
ABSTRACT
A collaborative study was run by the European Directorate for the Quality of Medicines & HealthCare (EDQM) in the
context of the Biological Standardisation Programme (BSP), under the aegis of the Council of Europe and the European
Commission, to establish replacement batches for the dwindling stocks of the Heparin low-molecular-mass for assay
European Pharmacopoeia Biological Reference Preparation (BRP). The replacement batches of BRP are intended to be
used in the assays for anti-Xa and anti-IIa activities, as described in the European Pharmacopoeia (Ph. Eur.) monograph
Heparins, low-molecular-mass(0828).
Three freeze-dried candidate batches were calibrated against the current International Standard (IS) for Heparin, low-
molecular-weight (2
nd
IS, 01/608). For the purpose of the continuity check between subsequent BRP batches, the current
Heparin low-molecular-mass for assay BRP (batch5) was also included in the test panel. Thirteen official medicines
control and manufacturers laboratories from European and non-European countries contributed data. A central statistical
analysis of the datasets was performed at the EDQM. On the basis of the results, the 3candidate materials were assigned
a potency of 104IU/vial for the anti-Xa activity and 31IU/vial for the anti-IIa activity. Taken into account the preliminary
stability data and the results of this collaborative study, the 3batches of candidate BRP were adopted in June 2010 by the
Commission of the Ph.Eur. as Heparin low-molecular-mass for assay BRP batches6, 7 and8.
KEYWORDS
Low-molecular-mass Heparin, anti-Xa, anti-IIa, chromogenic assay, Biological standardisation, collaborative study,
Biological Reference Preparation, European Pharmacopoeia.
E. Terao. (corresponding authors e-mail: eriko.terao@edqm.eu). European Directorate for the Quality of Medicines & HealthCare (EDQM), Council
of Europe, 7 Alle Kastner, F-67081 Strasbourg, France
A. Daas. European Directorate for the Quality of Medicines & HealthCare (EDQM), Council of Europe, Strasbourg, France
G. Rautmann. European Directorate for the Quality of Medicines & HealthCare (EDQM), Council of Europe, Strasbourg, France
K.-H.Buchheit. European Directorate for the Quality of Medicines & HealthCare (EDQM), Council of Europe, Strasbourg, France
after1 and 6months supported the stability of the cBRPs.
Indication on the high stability of low-molecular-mass
heparins was provided by the accelerated degradation studies
performed for earlier calibration studies (e.g.BSP060 in
2003) and from the monitoring scheme in place at EDQM
for the current and former batches of BRP. There was thus
no indication that the cBRPs were not stable.
3.3. Study samples
The test panel included the following references and 3coded
batches of candidate BRP:
Sample A:
2
nd
International Standard for Heparin, low-molecular-
weight (01/608; 2
nd
IS)
assigned potencies: anti-Xa: 1097IU/ampoule, anti-IIa:
326IU/ampoule
Sample B:
Heparin low-molecular-mass for assayBRP (batch5,
EDQM code: H0185000)
assigned potencies: anti-Xa: 1281IU/ampoule, anti-IIa:
301IU/ampoule
Samples C, D, E:
3freeze-dried batches of low-molecular-mass heparin for
assay (cBRPs)
indicative potency ranges: anti-Xa: 90-100IU/vial, anti-IIa:
25-35IU/vial.
Four vials of each sample(A-E) were provided to all
participants. The contents had to be reconstituted in 1mL
distilled water and kept at4C if needed. Indicative potency
ranges for both anti-Xa vand anti-IIa activities were given to
participants in order to allow adjustment of sample dilutions
for the assays.
4. METHODS, DESIGN AND STATISTICAL ANALYSIS
4.1. Methods
Participants were requested to use their routine anti-Xa
and anti-IIa chromogenic assays with methods preferably
based on the Ph. Eur. monograph Heparins, low-molecular-
mass(0828)[1].
4.2. Study design
Participants were asked to perform at least 3 independent
assays for each method, testing at least 4dilution points.
Each assay had to be carried out on separate days, using
one freshly reconstituted set of vials per assay. The 4
th
set of
samples provided could be used for pre-testing purposes.
Participants were requested to include the 2
nd
IS in each
run and to use a balanced sample order scheme for each
assay. Examples of schemes for10-, 6-, 4- and 3-place
assays were given as information. Depending on the number
of samples that could be tested at once (i.e.in a run), it
was in some cases necessary to perform several runs for
each assay. The suitable design had to be defined and be
adhered to, from beginning to end of the whole testing.
4.3. Data reporting and statistical analysis
Participants were asked to submit raw data, calculated
results as well as deviations from the pharmacopoeial
methods using the electronic datasheets provided. They were
also encouraged to provide electronic CombiStats data files
if available.
All raw data were submitted to central statistical analysis
at EDQM using the CombiStats
software[3]. Each run

was treated as a separate assay. The parallel line model[4]
was used for the majority of laboratories. In some cases
a log transformation of the responses was found to
produce better linearity. There were also cases where a
4-parametric logistic curve fit or 3-parametric exponential
curve fit was more suitable. In one case the slope-ratio
model was applied. The same model was used for all
assays from a given laboratory. The complete computer
output containing all data and model specifications
was made available to all participants of the study.
5. RESULTS AND DISCUSSION
5.1. Assay data returned
Thirteen laboratories reported results for analysis. All
laboratories carried out at least 3independent assays as
requested. All participants performed their assays according
to the methods described in the Ph.Eur. or slightly modified
methods derived thereof, except Laboratory1 which used
an in-house chromogenic method. Most laboratories used
the absorbance measured at 405nm as end-point, except
Laboratories1, 6, 7 and9 which used a kinetic end-point.
All laboratories submitted results of their own calculations.
Six laboratories used CombiStats
for calculations and

7laboratories used other software. Laboratory13 reported
that 1of their anti-IIa assays was found invalid for
sampleB and another one was found invalid for SampleE.
A complete overview of the reported activities per assay
is given in Table1a for anti-FactorXa and in Table2a for
anti-FactorIIa. Also shown are the geometric mean(GM)
and the geometric coefficient of variation(GCV)
1
.
5.2. Assay validity
Validity of the assays was assessed by analysis of variance to
check for significant deviations from parallelism and/or
linearity at the 1%level of significance. Significant
deviations from linearity were not considered as a reason to
reject an assay if it could be explained from a quadratic
curvature where the sum of squares for quadratic regression
was less than 1/100
th
the sum of squares for linear
regression[5,6]. The data were also submitted to visual
inspection of the plots. In no case was it found necessary to
reject an assay.
5.3. Anti-Xa and anti-IIa activities estimates
The activities with respect to the 2
nd
IS as calculated at
the EDQM are displayed in Table1b for anti-FactorXa and
in Table2b for anti-FactorIIa, together with theGM and
theGCV. The difference with the activities calculated by
the participants (Table1a and2a) was in general small,
although differences of more than3% could be observed
in a few individual cases. A graphical representation
of the results is provided in Figure1. The histograms
show the estimated activity per assay, as calculated at
the EDQM, in percentage of the assigned activity. For
SamplesC, D andE, the assigned activity was taken
to be the proposed value at the end of this report.
The overall GM of the 13laboratory means is shown at
the bottom of Tables1a, 1b, 2a and2b, together with the
GCV and the 95%confidence limits. It can be seen that the
difference between the overall outcome for results based on
1
There co-exist 2 different denitions of the GCV in the published literature. They are not equivalent and caution is advised when comparing the
GCVs from different studies. The denition used in this report is GCV=(exp(v)-1)
0.5 .
100%, whereas some other publications use the denition
GCV=(exp(v
0.5
)-1
.
100%. In both equations, v is the sample variance of the log-transformed activities.
participants calculations and central calculations was very
small. The inter-laboratoryGCV is between3.4% and5.5%.
The results for anti-FactorIIa activity of SampleB from
Laboratory13 were about 20%higher than the overall
mean of the other laboratories and should be considered
for exclusion as an outlier. If this laboratory was excluded
the GCV was reduced from more than5% to less than2%.
5.4. Confirmation of the assigned activity
of the current BRP (BRP5).
In order to assure batch continuity, the current Heparin
low-molecular-mass for assay BRP was included in this
study to check that the assigned activity is confirmed.
TheGM for anti-FactorXa was 1297IU/ampoule with
95%confidence limits of1269 and 1324IU/ampoule, which
encompassed the assigned activity of 1281IU/ampoule.
TheGM for anti-FactorIIa was 305IU/ampoule (excluding
Laboratory13) with 95%confidence limits of 302 and
309IU/ampoule, which did not encompass the assigned
activity of 301IU/ampoule. However, it encompassed the
overall GM of 307IU/ampoule as found in the establishment
report of the BRP5 when all experimental methods were
included in the calculations[2]. The assigned activities of
BRP5 were therefore considered as confirmed in this study.
5.5. Comparison between the candidate batches
The 3candidate batches were prepared from the same
batch of starting material, using a similar procedure.
The overall means per batch appeared very similar. It was
therefore interesting to check if any statistically significant
differences could be observed between the batches.
To this end, an analysis of variance was performed
allowing for systematic laboratory effects and batch effects.
Calculations were performed with the SAS System[7] and
a summary of the output is shown in Tables3 and4.
The probabilities (p-values) for linear batch effect were
0.646 and0.967 respectively, confirming that there was no
indication for systematic differences between batches.
5.6. Activities of the candidate batches
The overall GM of the 3candidate batches were respectively
104.1IU/vial, 104.7IU/vial and 104.3 IU/vial for
anti-FactorXa activity. The overallGM of the 3candidate
batches were respectively 31.2IU/vial, 31.2IU/vial and
31.2IU/vial for anti-FactorIIa activity (Tables1b and2b).
Since the small differences could be entirely explained from
the normal assay variation, it was proposed to assign the
candidates the same activity, after rounding to the nearest
integer.
6. CONCLUSIONS
From the study, it appeared that SamplesC, D andE are
suitable as BRPs for use in the anti-Xa and anti-IIa assays,
as described in the Ph.Eur. monograph Heparins, low-
molecular-mass(0828). In line with the results of the
annual monitoring scheme performed at EDQM, this study
also confirmed the assigned anti-Xa and anti-IIa activities
of the current batch of Heparin low-molecular-mass for
assayBRP.
Based on the results of this collaborative study, the
3candidate batches were adopted in June2010 by the
European Pharmacopoeia Commission as Heparin low-
molecular-mass for assay BRP batches6, 7 and8 with the
following assigned potencies:
anti-Xa activity: 104 IU/vial
anti-IIa activity: 31 IU/vial
The BRP will be stored and shipped at 4C. The stability of
the anti-Xa and anti-IIa activities of the 3batches of BRP
will be monitored annually by the EDQM throughout their
lifetime.
7. ACKNOWLEDGEMENTS
We would like to acknowledge the European manufacturer
who generously donated material for their kind
collaboration as well as all participants for contributing
data to this collaborative study. MsSally Woodward ensured
excellent secretarial support throughout the study.
Table 3 Analysis of variance for anti-Factor Xa activity
Effect Degrees of freedom Sum of Squares Mean Square F value
Probability
(Pr > F)
Laboratories 12 1018 84.8 37.2 <0.0001
Batches 2 2.03 1.02 0.45 0.646
Residual Error 24 54.7 2.28
Total 38 1075
Table 4 Analysis of variance for anti-Factor IIa activity
Effect Degrees of freedom Sum of Squares Mean Square F value
Probability
(Pr > F)
Laboratories 12 57.0 4.75 19.1 <0.0001
Batches 2 0.02 0.008 0.03 0.967
Residual Error 24 5.97 0.249
Total 38 63.0
8. LIST OF PARTICIPANTS
(in alphabetical order of country)
Dr. Allison JONES, Dr. Chong LOH, Therapeutic Goods
Administration, Australia
Dr. Andreas WEILER, Dr. Georg DECRISTOFO,
SandozGmbH, Austria
Ms. Hing CHONG, Health Canada, Canada
Dr. Guy RAUTMANN, Mrs Chantal RAPHALEN,
MrsGwenalle COZIC, EDQM, Council of Europe
Dr. Eva SANDBERG, Danish Medicines Agency, Denmark
Dr. Bjrn FEHRMANN, LEO Pharma, Denmark
Dr. Laurence MOUILLOT, Dr. Valrie LIEVRE, Agence
Franaise de Scurit Sanitaire des Produits de Sant,
France
Mr. Patrick ROUSSEAU, Mrs Sabrina VIGER,
GlaxoSmithKline, France
Dr. Cline MARTINEZ, Sanofi-Aventis, France
Dr. Hinrich NAGEL, Nordmark Arzneimittel GmbH Co KG,
Germany
Dr. Luis Miguel MEIRINHOS SOARES, Dr. Clia
FIGUEIREDO, Infarmed IP, Portugal
Dr. Kristina ERLANDSSON-PERSSON,
Medical Products Agency, Sweden
Dr. Roger MEIER, Mr. Urs MEYER, Swissmedic, Switzerland
Dr. Elaine GRAY, Mr. John HOGWOOD, National Institute
for Biological Standards and Control, Health Protection
Agency,U.K.
9. ABBREVIATIONS
BRP: Biological Reference Preparation
BSP: Biological Standardisation Programme
cBRP: candidate Biological Reference Preparation
EDQM: European Directorate for the Quality of
Medicines&HealthCare
GCV: Geometric Coefficient of Variation
GM: Geometric Mean
IS: International Standard
IU: International Unit
OMCL: Official Medicines Control Laboratory
Ph. Eur.: European Pharmacopoeia
WHO: World Health Organization
10. REFERENCES
[1] Heparins, low-molecular-mass, monograph0828.
Ph.Eur. 6
th
Edition. Strasbourg, France:
Council of Europe; 2008 (vol2).
[2] Gray E, Rigsby P, Behr-Gross ME. Collaborative study
to establish the Low-molecular-mass heparin for
assay: European Pharmacopoeia Biological Reference
Preparation. Pharmeuropa Bio 2004;1:59-76.
[3] CombiStats v.4.0. Strasbourg, France: EDQM
Council of Europe [available at: www.combistats.eu].
[4] Finney DJ. Statistical Method in
Biological Assay, 3
rd
Edition. London,
UK: Charles Griffin&Co.Ltd.;1978.
[5] Bliss CI. The calculation of microbial assays.
Bacteriol Rev 1956;20(4):243-58.
[6] Hewitt W. Influence of curvature of response
lines in antibiotic agar diffusion assays. J Biol
Stand1981;9(1):1-13.
[7] SAS Institute Inc., SAS OnlineDoc, Version8,
Cary, NC: SAS Institute Inc., 1999.
L
a
b
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
1
2
5
8
1
2
4
7
1
2
7
0
1
2
5
8
0
.
9
1
0
8
1
1
5
1
1
7
1
1
3
4
.
4
1
0
7
1
1
5
1
1
7
1
1
3
5
.
0
1
0
9
1
1
3
1
1
4
1
1
2
2
.
5
2
1
2
9
0
1
2
6
7
1
2
6
9
1
2
7
6
1
.
0
1
0
5
1
0
2
1
0
5
1
0
4
1
.
9
1
0
8
1
0
3
1
0
7
1
0
6
2
.
6
1
0
8
1
0
3
1
0
7
1
0
6
2
.
8
3
1
2
5
4
1
3
3
2
1
2
7
5
1
2
8
7
3
.
1
1
0
6
9
7
1
0
5
1
0
3
4
.
9
9
1
1
0
3
1
0
4
9
9
7
.
5
1
0
4
9
9
9
9
1
0
1
2
.
8
4
1
3
7
5
1
3
9
0
1
4
1
9
1
2
4
7
1
3
5
6
5
.
7
1
0
3
1
0
0
9
6
9
5
9
8
3
.
6
1
0
1
9
6
9
8
9
3
9
7
3
.
3
9
7
9
6
9
7
9
5
9
6
1
.
4
5
1
3
4
8
1
2
0
1
1
3
4
0
1
2
9
5
6
.
5
1
0
8
1
0
1
1
0
9
1
0
6
4
.
5
1
1
0
1
0
6
1
1
4
1
1
0
3
.
7
1
0
4
1
0
6
1
1
6
1
0
8
6
.
0
6
1
2
7
1
1
2
8
1
1
2
2
7
1
2
5
9
2
.
3
9
9
1
0
2
1
0
1
1
0
1
1
.
5
1
0
0
1
0
0
1
0
2
1
0
1
1
.
1
1
0
5
1
0
0
1
0
0
1
0
2
2
.
8
7
1
3
3
0
1
2
7
9
1
2
7
9
1
2
9
6
2
.
3
1
1
3
1
1
0
1
0
9
1
1
1
1
.
9
1
1
2
1
1
0
1
1
1
1
1
1
0
.
9
1
1
2
1
1
0
1
1
1
1
1
1
0
.
9
8
1
2
5
3
1
2
3
6
1
2
6
0
1
2
5
0
1
.
0
1
0
2
1
0
3
1
0
5
1
0
3
1
.
5
1
0
4
1
0
2
1
0
6
1
0
4
1
.
9
1
0
3
1
0
2
1
0
4
1
0
3
1
.
0
9
1
3
5
5
1
4
6
5
1
3
4
5
1
3
8
7
4
.
7
1
0
9
1
0
7
1
0
2
1
0
6
3
.
4
1
0
1
1
0
9
1
1
2
1
0
7
5
.
4
1
0
6
1
0
4
1
0
7
1
0
6
1
.
4
1
0
1
2
7
4
1
3
1
4
1
3
1
6
1
3
0
1
1
.
8
1
0
9
1
0
3
1
0
9
1
0
7
3
.
1
1
1
2
1
0
8
1
0
6
1
0
9
2
.
7
1
0
9
1
0
9
1
1
3
1
1
0
2
.
1
1
1
1
2
6
2
1
2
1
3
1
2
7
7
1
2
5
0
2
.
7
1
0
1
1
0
0
1
0
3
1
0
1
1
.
2
9
7
9
8
1
0
1
9
9
2
.
3
1
0
2
9
7
1
0
3
1
0
1
3
.
0
1
2
1
2
8
8
1
3
4
4
1
2
7
3
1
3
0
1
2
.
8
1
1
3
1
1
0
1
0
7
1
1
0
2
.
8
1
0
8
1
0
6
1
0
8
1
0
8
1
.
1
1
0
3
1
0
8
1
0
7
1
0
6
2
.
8
1
3
1
3
3
4
1
3
4
1
1
4
2
0
1
3
6
4
3
.
5
8
7
1
0
1
9
9
9
5
8
.
1
9
5
1
1
1
9
5
1
0
0
9
.
0
9
5
1
0
4
9
1
9
7
6
.
8
(
1
2
7
2

-
1
3
2
4
)
(
1
0
1
.
5

-
1
0
7
.
3
)
(
1
0
1
.
7

-
1
0
7
.
8
)
(
1
0
1
.
3

-
1
0
7
.
3
)
A
s
s
a
y
A
s
s
a
y
A
s
s
a
y
A
s
s
a
y
G
M
G
C
V
S
a
m
p
l
e

B
S
a
m
p
l
e

C
S
a
m
p
l
e

D
S
a
m
p
l
e

E
G
M
G
C
V
G
M
G
C
V
G
M
G
C
V
9
5
%

c
o
n
f
i
d
e
n
c
e

l
i
m
i
t
s
G
e
o
m
e
t
r
i
c

c
o
e
f
f
i
c
i
e
n
t

o
f

v
a
r
i
a
t
i
o
n

(
G
C
V
)
G
e
o
m
e
t
r
i
c

m
e
a
n

(
G
M
)
5
.
0
%
1
0
4
.
3
5
.
0
%
1
0
4
.
7
4
.
8
%
1
0
4
.
3
3
.
4
%
1
2
9
8
T
a
b
l
e

1
a

A
n
t
i
-
X
a

a
c
t
i
v
i
t
i
e
s

e
x
p
r
e
s
s
e
d

i
n

I
U
/
v
i
a
l

r
e
l
a
t
i
v
e

t
o

t
h
e

2
n
d

I
S

(
a
s
s
a
y

v
a
l
u
e
s

c
a
l
c
u
l
a
t
e
d

b
y

p
a
r
t
i
c
i
p
a
n
t
s
)
T
a
b
l
e

1
b

A
n
t
i
-
X
a

a
c
t
i
v
i
t
i
e
s

e
x
p
r
e
s
s
e
d

i
n

I
U
/
v
i
a
l

r
e
l
a
t
i
v
e

t
o

t
h
e

2
n
d

I
S

(
a
s
s
a
y

v
a
l
u
e
s

c
a
l
c
u
l
a
t
e
d

b
y

E
D
Q
M
)
L
a
b
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
1
2
5
8
1
2
4
2
1
2
7
1
1
2
5
7
1
.
2
1
0
8
1
1
4
1
1
7
1
1
3
4
.
2
1
0
7
1
1
5
1
1
7
1
1
3
4
.
7
1
0
9
1
1
3
1
1
4
1
1
2
2
.
4
2
1
2
9
1
1
2
6
7
1
2
7
2
1
2
7
7
1
.
0
1
0
6
1
0
3
1
0
7
1
0
5
1
.
9
1
0
9
1
0
4
1
0
8
1
0
7
2
.
5
1
0
9
1
0
4
1
0
8
1
0
7
2
.
6
3
1
2
5
4
1
3
3
2
1
2
7
5
1
2
8
6
3
.
1
1
0
6
9
7
1
0
5
1
0
3
4
.
9
9
1
1
0
3
1
0
4
9
9
7
.
2
1
0
4
9
9
9
9
1
0
1
3
.
2
4
1
4
1
2
1
4
1
6
1
4
3
4
1
2
5
1
1
3
7
6
6
.
4
1
0
5
1
0
0
9
5
9
5
9
9
4
.
7
1
0
3
9
7
9
8
9
3
9
8
4
.
1
9
9
9
6
9
7
9
5
9
7
2
.
0
5
1
3
4
0
1
1
8
0
1
3
1
3
1
2
7
6
6
.
8
1
0
9
1
0
1
1
1
0
1
0
7
4
.
5
1
1
1
1
0
6
1
1
5
1
1
1
3
.
9
1
0
4
1
0
6
1
1
6
1
0
9
6
.
0
6
1
2
6
4
1
2
7
3
1
2
0
7
1
2
4
8
2
.
9
9
9
1
0
1
1
0
2
1
0
1
1
.
4
1
0
0
9
9
1
0
3
1
0
1
1
.
7
1
0
8
9
9
1
0
1
1
0
3
4
.
4
7
1
3
2
3
1
2
6
4
1
2
7
0
1
2
8
6
2
.
5
1
1
3
1
0
9
1
0
9
1
1
0
1
.
9
1
1
2
1
0
9
1
1
1
1
1
1
1
.
4
1
1
2
1
1
0
1
1
1
1
1
1
1
.
1
8
1
2
5
2
1
2
3
5
1
2
6
0
1
2
4
9
1
.
0
1
0
2
1
0
3
1
0
5
1
0
4
1
.
4
1
0
4
1
0
3
1
0
6
1
0
4
1
.
5
1
0
4
1
0
2
1
0
4
1
0
3
1
.
0
9
1
3
5
5
1
4
6
5
1
3
4
5
1
3
8
7
4
.
7
1
0
9
1
0
7
1
0
2
1
0
6
3
.
7
1
0
1
1
0
9
1
1
2
1
0
7
5
.
2
1
0
6
1
0
4
1
0
7
1
0
6
1
.
4
1
0
1
2
8
3
1
3
0
2
1
3
0
9
1
2
9
8
1
.
0
1
1
0
1
0
3
1
0
7
1
0
6
3
.
3
1
1
2
1
0
7
1
0
7
1
0
9
2
.
8
1
0
9
1
0
9
1
1
3
1
1
1
2
.
3
1
1
1
2
6
2
1
2
1
3
1
2
7
7
1
2
5
0
2
.
7
1
0
2
1
0
0
1
0
3
1
0
2
1
.
4
9
7
9
8
1
0
1
9
9
2
.
3
1
0
2
9
7
1
0
3
1
0
1
3
.
0
1
2
1
3
0
8
1
3
6
9
1
3
0
1
1
3
2
6
2
.
8
1
1
2
1
0
8
1
0
6
1
0
9
2
.
8
1
0
8
1
0
6
1
0
8
1
0
7
1
.
4
1
0
2
1
0
7
1
0
6
1
0
5
2
.
8
1
3
1
3
3
7
1
3
2
7
1
3
8
2
1
3
4
9
2
.
1
8
8
9
6
9
1
9
2
4
.
8
9
4
1
0
6
9
1
9
7
7
.
7
9
6
9
7
9
1
9
4
3
.
5
(
1
2
6
9

-
1
3
2
4
)
(
1
0
0
.
9

-
1
0
7
.
4
)
(
1
0
1
.
5

-
1
0
7
.
9
)
(
1
0
1
.
2

-
1
0
7
.
5
)
A
s
s
a
y
A
s
s
a
y
A
s
s
a
y
A
s
s
a
y
S
a
m
p
l
e

B
S
a
m
p
l
e

C
S
a
m
p
l
e

D
S
a
m
p
l
e

E
G
e
o
m
e
t
r
i
c

m
e
a
n

(
G
M
)
1
2
9
7
1
0
4
.
1
1
0
4
.
7
G
C
V
1
0
4
.
3
G
e
o
m
e
t
r
i
c

c
o
e
f
f
i
c
i
e
n
t

o
f

v
a
r
i
a
t
i
o
n

(
G
C
V
)
3
.
6
%
5
.
4
%
5
.
3
%
5
.
2
%
G
M
G
C
V
9
5
%

c
o
n
f
i
d
e
n
c
e

l
i
m
i
t
s
G
M
G
C
V
G
M
G
C
V
G
M
T
a
b
l
e

2
a

A
n
t
i
-
I
I
a

a
c
t
i
v
i
t
i
e
s

e
x
p
r
e
s
s
e
d

i
n

I
U
/
v
i
a
l

r
e
l
a
t
i
v
e

t
o

t
h
e

2
n
d

I
S

(
a
s
s
a
y

v
a
l
u
e
s

c
a
l
c
u
l
a
t
e
d

b
y

p
a
r
t
i
c
i
p
a
n
t
s
)
T
a
b
l
e

2
b

A
n
t
i
-
I
I
a

a
c
t
i
v
i
t
i
e
s

e
x
p
r
e
s
s
e
d

i
n

I
U
/
v
i
a
l

r
e
l
a
t
i
v
e

t
o

t
h
e

2
n
d

I
S

(
a
s
s
a
y

v
a
l
u
e
s

c
a
l
c
u
l
a
t
e
d

b
y

E
D
Q
M
)
L
a
b
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
3
0
5
3
1
6
3
0
7
3
0
9
1
.
8
3
4
3
3
3
2
3
3
3
.
7
3
5
3
3
3
2
3
3
3
.
9
3
4
3
3
3
2
3
3
3
.
3
2
3
0
6
3
0
5
3
0
4
3
0
5
0
.
4
3
1
3
2
3
1
3
1
1
.
5
3
2
3
2
3
1
3
1
1
.
9
3
2
3
2
3
1
3
2
1
.
9
3
3
0
8
2
9
0
3
0
2
3
0
0
3
.
1
3
1
3
0
3
0
3
0
1
.
9
3
1
2
8
3
1
3
0
5
.
9
3
2
2
8
3
6
3
2
1
2
.
6
4
3
1
3
2
9
7
3
0
5
3
0
1
3
0
4
2
.
2
3
1
3
1
3
0
3
1
3
0
1
.
4
3
1
3
0
3
0
2
9
3
0
2
.
6
3
0
3
0
3
0
3
0
3
0
1
.
5
5
3
0
4
3
1
0
3
0
1
3
0
5
1
.
5
3
1
3
0
3
0
3
0
1
.
9
3
1
3
1
3
0
3
1
1
.
1
2
9
3
0
3
0
3
0
1
.
7
6
3
0
0
2
9
2
3
0
4
2
9
9
2
.
1
3
0
3
0
3
2
3
1
3
.
7
3
1
3
1
3
1
3
1
0
.
0
3
0
3
0
3
0
3
0
0
.
0
7
3
0
3
3
0
6
3
0
0
3
0
3
1
.
0
3
2
3
2
3
2
3
2
0
.
0
3
2
3
2
3
2
3
2
0
.
0
3
2
3
1
3
1
3
1
1
.
8
8
3
0
9
3
0
3
3
1
7
3
1
0
2
.
3
3
1
3
1
3
1
3
1
0
.
0
3
1
3
1
3
2
3
1
1
.
8
3
1
3
1
3
1
3
1
0
.
0
9
3
2
3
3
2
8
3
0
6
3
1
9
3
.
7
3
2
3
1
3
1
3
1
1
.
5
2
9
3
2
3
2
3
1
4
.
6
2
9
3
1
3
1
3
0
3
.
4
1
0
3
0
8
3
1
4
2
9
5
3
0
5
3
.
2
3
0
3
0
2
9
3
0
1
.
9
3
0
3
0
3
0
3
0
0
.
3
2
9
3
1
2
8
2
9
5
.
3
1
1
3
0
4
3
0
8
2
9
2
3
0
1
2
.
7
3
1
3
1
3
0
3
1
1
.
7
3
1
3
2
3
1
3
1
1
.
7
3
0
3
1
3
1
3
1
1
.
5
1
2
3
0
3
3
1
0
2
9
1
3
0
1
3
.
3
3
0
3
1
3
0
3
0
2
.
8
3
1
3
1
3
1
3
1
0
.
3
3
0
3
1
2
9
3
0
2
.
5
1
3
I
n
v
a
l
i
d
3
8
3
3
5
5
3
6
9
5
.
4
3
3
3
4
3
6
3
4
4
.
4
3
1
3
2
3
6
3
3
7
.
9
3
2
3
4
I
n
v
a
l
i
d
3
3
4
.
3
(
3
0
0

-
3
2
0
)
(
3
0
.
4

-
3
1
.
9
)
(
3
0
.
5

-
3
1
.
8
)
(
3
0
.
2

-
3
1
.
6
)
9
5
%

c
o
n
f
i
d
e
n
c
e

l
i
m
i
t
s

(
e
x
c
l
.

L
a
b

1
3
)
(
3
0
2
-
3
0
8
)
O
v
e
r
a
l
l

G
C
V
5
.
5
%
4
.
1
%
O
v
e
r
a
l
l

G
M

(
e
x
c
l
u
d
i
n
g

L
a
b

1
3
)
3
0
5
O
v
e
r
a
l
l

G
C
V

(
e
x
c
l
u
d
i
n
g

L
a
b

1
3
)
1
.
8
%
3
.
4
%
4
.
0
%
9
5
%

c
o
n
f
i
d
e
n
c
e

l
i
m
i
t
s
G
C
V
O
v
e
r
a
l
l

G
M
3
1
0
3
1
.
2
3
1
.
2
3
0
.
9
G
C
V
A
s
s
a
y
G
M
G
C
V
A
s
s
a
y
G
M
G
M
S
a
m
p
l
e

B
S
a
m
p
l
e

C
S
a
m
p
l
e

D
S
a
m
p
l
e

E
A
s
s
a
y
G
M
G
C
V
A
s
s
a
y
L
a
b
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
3
0
7
3
1
4
3
0
9
3
1
0
1
.
0
3
5
3
3
3
2
3
3
4
.
1
3
5
3
3
3
2
3
3
4
.
0
3
5
3
3
3
3
3
3
3
.
3
2
3
0
6
3
0
4
3
0
3
3
0
5
0
.
5
3
2
3
3
3
2
3
2
1
.
7
3
2
3
3
3
2
3
2
1
.
6
3
2
3
3
3
2
3
3
1
.
4
3
3
0
8
2
9
0
3
0
2
3
0
0
2
.
9
3
1
3
0
3
0
3
0
1
.
5
3
1
2
8
3
1
3
0
5
.
8
3
2
2
8
3
6
3
2
1
2
.
0
4
3
1
3
2
9
7
3
0
5
3
0
3
3
0
4
2
.
0
3
1
3
1
3
0
3
0
3
0
1
.
6
3
1
3
0
3
0
3
0
3
0
2
.
6
3
0
3
0
3
0
3
0
3
0
0
.
8
5
2
9
3
3
1
0
3
0
1
3
0
1
2
.
7
3
0
3
0
3
0
3
0
0
.
5
3
0
3
1
3
0
3
0
1
.
3
3
0
3
0
3
0
3
0
1
.
1
6
2
9
7
2
9
3
3
0
3
2
9
7
1
.
8
3
0
3
0
3
3
3
1
5
.
3
2
9
3
1
3
1
3
1
3
.
7
2
9
3
0
3
0
3
0
1
.
3
7
3
0
2
3
0
6
3
0
1
3
0
3
0
.
8
3
2
3
2
3
2
3
2
0
.
4
3
2
3
2
3
1
3
2
0
.
7
3
2
3
1
3
1
3
1
0
.
8
8
3
1
2
3
0
9
3
2
3
3
1
5
2
.
3
3
1
3
1
3
1
3
1
1
.
1
3
1
3
1
3
2
3
2
2
.
5
3
1
3
1
3
1
3
1
0
.
5
9
3
2
3
3
2
8
3
0
6
3
1
9
3
.
7
3
2
3
1
3
1
3
1
1
.
5
2
9
3
2
3
2
3
1
4
.
6
2
9
3
1
3
1
3
0
3
.
4
1
0
3
0
7
3
1
1
2
9
8
3
0
6
2
.
2
3
1
3
0
2
9
3
0
3
.
2
3
0
3
0
3
0
3
0
0
.
9
2
9
3
1
2
8
3
0
4
.
2
1
1
3
0
4
3
0
8
2
9
2
3
0
1
2
.
7
3
1
3
1
3
0
3
1
1
.
6
3
1
3
2
3
1
3
1
1
.
7
3
0
3
1
3
0
3
1
1
.
5
1
2
3
0
6
3
1
3
2
8
7
3
0
2
4
.
6
3
0
3
2
3
0
3
0
3
.
3
3
1
3
1
3
1
3
1
0
.
3
3
0
3
1
2
9
3
0
2
.
5
1
3
3
5
6
3
8
3
3
5
5
3
6
5
4
.
3
3
3
3
4
3
6
3
4
5
.
0
3
1
3
3
3
6
3
3
8
.
0
3
2
3
4
3
7
3
4
7
.
9
(
3
0
0

-
3
1
9
)
(
3
0
.
5

-
3
2
.
0
)
(
3
0
.
6

-
3
1
.
9
)
(
3
0
.
3

-
3
2
.
0
)
G
C
V
A
s
s
a
y
G
M
G
C
V
A
s
s
a
y
G
M
S
a
m
p
l
e

B
S
a
m
p
l
e

C
S
a
m
p
l
e

D
S
a
m
p
l
e

E
A
s
s
a
y
G
M
G
C
V
O
v
e
r
a
l
l

G
M
3
0
9
3
1
.
2
3
1
.
2
3
1
.
2
G
C
V
A
s
s
a
y
G
M
O
v
e
r
a
l
l

G
C
V
5
.
3
%
4
.
1
%
3
.
7
%
4
.
6
%
9
5
%

c
o
n
f
i
d
e
n
c
e

l
i
m
i
t
s
9
5
%

c
o
n
f
i
d
e
n
c
e

l
i
m
i
t
s

(
e
x
c
l
.

L
a
b

1
3
)
O
v
e
r
a
l
l

G
M

(
e
x
c
l
u
d
i
n
g

L
a
b

1
3
)
3
0
5
O
v
e
r
a
l
l

G
C
V

(
e
x
c
l
u
d
i
n
g

L
a
b

1
3
)
2
.
0
%
(
3
0
2
-
3
0
9
)
9
4
14 1
4 10 11
8 11 7 8
7 7 6 7
2 3 5 5
11 2 12 13 11 3 12 12
8 1 10 3 6 2 10 10
6 6 5 13 9 4 2 3 1 8
6 8 3 10 10 7 12 4 12 6 11 2 8 9 13
5 11 1 1 2 12 5 9 13 4 4 9 3 5 7 1 4 9 13 13
8
4
8
6
8
8
9
0
9
2
9
4
9
6
9
8
1
0
0
1
0
2
1
0
4
1
0
6
1
0
8
1
1
0
1
1
2
1
1
4
9
0
9
2
9
4
9
6
9
8
1
0
0
1
0
2
1
0
4
1
0
6
1
0
8
1
1
0
1
1
2
1
1
4
1
1
6
1
1
8
1
2
0
1
2
2
1
2
4
1
2
6
1
2
8
12
11 12 11
9 10 5 14 9 7
6 8 4 3 8 2
10 2 7 6 10 11 12
6 9 12 5 4 8 9 6
4 11 5 3 4 7 5 12 3 10 7 2
4 4 11 8 3 5 10 12 6 11 4 9 1 1
13 13 13 3 6 8 2 2 1 9 7 1 1 10 5 8 3 2 7 13 13 1 13
8
4
8
6
8
8
9
0
9
2
9
4
9
6
9
8
1
0
0
1
0
2
1
0
4
1
0
6
1
0
8
1
1
0
1
1
2
1
1
4
9
0
9
2
9
4
9
6
9
8
1
0
0
1
0
2
1
0
4
1
0
6
1
0
8
1
1
0
1
1
2
1
1
4
1
1
6
1
1
8
1
2
0
1
2
2
1
2
4
1
2
6
1
2
8
12
11
6
3
12 9
12 8 7
10 2 6 2
3 8 10 10 5 11
4 11 3 13 9 9 14 5 13 9
11 6 2 12 7 10 10 4 12 7 8
13 4 4 6 8 8 5 12 7 5 9 10 4 11 7 1 2
3 13 11 6 9 4 1 2 7 5 1 1 3 6 5 3 8 4 2 1 13 1 13
8
4
8
6
8
8
9
0
9
2
9
4
9
6
9
8
1
0
0
1
0
2
1
0
4
1
0
6
1
0
8
1
1
0
1
1
2
1
1
4
9
0
9
2
9
4
9
6
9
8
1
0
0
1
0
2
1
0
4
1
0
6
1
0
8
1
1
0
1
1
2
1
1
4
1
1
6
11
6 9
8 9 5 8
9 2 14 4 12 7
4 6 11 2 12 10 6 4 11 7 2
13 4 3 8 8 12 6 7 12 5 12 10 13 1
4 11 6 12 5 5 2 7 1 10 10 10 6 11 9 8 3 2
13 13 3 4 11 3 9 1 10 7 1 5 3 9 5 4 8 7 2 1 13 1 3 13
8
4
8
6
8
8
9
0
9
2
9
4
9
6
9
8
1
0
0
1
0
2
1
0
4
1
0
6
1
0
8
1
1
0
1
1
2
1
1
4
9
0
9
2
9
4
9
6
9
8
1
0
0
1
0
2
1
0
4
1
0
6
1
0
8
1
1
0
1
1
2
1
1
4
1
1
6
1
1
8
1
2
0
1
2
2
1
2
4
1
2
6
1
2
8
Sample B - anti-Factor Xa Sample B - anti-Factor IIa
Potency (as % of assigned activity) Potency (as % of assigned activity)
Sample C - anti-Factor Xa Sample C - anti-Factor IIa
Sample E - anti-Factor Xa Sample E - anti-Factor IIa
Sample D - anti-Factor Xa Sample D - anti-Factor IIa
Figure 1 Histograms of anti-Xa and anti-IIa activities per assay (in percentage of assigned activity)
Numbers in the boxes are the laboratory codes. Each box represents 1 assay.
Validation of a Serological Potency Assay for Rabies Vaccine (inactivated) for Veterinary Use
Collaborative Study for Validation of a Serological
Potency Assay for Rabies Vaccine (inactivated)
for Veterinary Use
B. Krmer, L. Bruckner, A. Daas, C. Milne
ABSTRACT
European Pharmacopoeia (Ph. Eur.) monograph 0451 on Rabies vaccine (inactivated) for veterinary use describes
an in vivo batch potency test that is based on the NIH test. This assay uses a large number of mice and results in a
significant degree of suffering. In the interest of replacement, reduction and refinement of animal tests (3R) a serological
potency assay for Rabies vaccine (inactivated) for animal use, developed and validated at the Paul-Ehrlich-Institut, has
been assessed in a collaborative study organised by the EDQM (European Directorate for the Quality of Medicines &
HealthCare). The goal was to demonstrate the wider transferability of the proposed assay and confirm its suitability. The
study involved 13 laboratories and assessed 4 different vaccines from the EU market. Results of the study confirm that
a limit test using a relatively small number of animals in a serological assay is possible, reproducible and reliable. The
optimal number of animals per vaccine is product specific but may roughly be indicated to be between 8 and 10 for the
products included in this study. Non-responders should be included in the analysis because they may reflect sub-potent
vaccines. However, there may be a need to impose a maximum on the number of non-responders allowed for the reference
vaccine as a monitor for assay validity. This assay provides a significant 3R improvement in terms of both the number of
animals used and the amount of suffering entailed and provides a more reliable and reproducible assay format than the
vaccination challenge assay. It also reduces the time required as compared to the vaccination challenge assay. It has been
recommended to the Ph. Eur. group of experts 15V that this assay be included as an alternative to the batch potency assay
in the Ph. Eur. monograph 0451.
KEYWORDS
Rabies vaccine (inactivated) for veterinary use, NIH test, 3R alternative, serological assay, batch potency test, European
Pharmacopoeia.
B. Krmer, Paul-Ehrlich-Institut, Paul-Ehrlich-Strasse 51-59, D63225 Langen, Germany
L. Bruckner. Institute of Virology and Immunoprophylaxis (IVI), Sensemattstrasse 293, CH3147 Mittelhusern, Switzerland
A. Daas. European Directorate for the Quality of Medicines & HealthCare (EDQM), Council of Europe, 7 Alle Kastner, CS 30026, F67081
Strasbourg, France
C. Milne (corresponding authors e-mail: catherine.milne@edqm.eu). European Directorate for the Quality of Medicines & HealthCare (EDQM),
Council of Europe, 7 Alle Kastner, F-67081 Strasbourg, France.
1. AIM
The aim of the collaborative study was to demonstrate the
wider transferability of the proposed assay and confirm
its suitability for the potency assay of inactivated rabies
vaccines for veterinary use.
2. INTRODUCTION
The Ph. Eur. monograph 0451 on Rabies vaccine
(inactivated) for veterinary use [1] describes in section 3-5
an in vivo batch potency test that is based on the NIH test
and involves a vaccination challenge assay with test and
reference vaccines.
Section 2-4-3 of monograph 0451 allows the replacement
of the vaccination challenge assay for batch potency by a
validated alternative method and briefly describes a serology
based assay.
The vaccination challenge assay is well known to be
both heavy on animal use and a very harsh assay for the
animals. It is also known to have high variability and can
be problematic in terms of reaching all of the validity
criteria[2, 3]. However, despite the possibility to use an
alternative assay, the vaccination challenge assay is still
widely used for batch release of inactivated rabies vaccines
for veterinary use as attempts to use the serological assay
presently outlined in the Ph. Eur. monograph have not been
convincing.
In the interest of applying the 3R principle for humane
alternatives to animal tests [4] and the goal of advancing
harmonisation of an assay with potentially better
reproducibility, the Paul-Ehrlich-Institut (PEI) has
developed and validated an alternative serology assay in
their laboratory [5]. The published assay involves the
immunisation of groups of mice with the test vaccine diluted
appropriately or the reference standard vaccine preparation
which is adjusted to the minimum potency allowed. 14days
after immunisation blood samples are taken and the sera
are tested individually for rabies antibody using a virus
neutralisation assay for the detection of antibodies against
rabies with a fluorescence detection method. Dilutions
of the sera that reduce the number of fluorescent cells by
50per cent are calculated. The vaccine passed if the antibody
titres obtained with the test vaccines were greater than
or equal to the antibody titres obtained for the reference
standard (P=0.95). Two similar assays, the Rapid Fluorescent
Focus Inhibition Test (RFFIT) and a Fluorescent Antibody
Virus Neutralisation test (FAVN) are described in the World
Organisation of Animal Health (OIE) Manual of Diagnostics
Tests and Vaccines in a chapter concerning rabies as
standard approved techniques [6]. Dr.Krmer convincingly
presented the vaccine batch potency assay outlined in the
publication [5] and the validation data from her laboratory
at the annual meeting for OMCLs involved in batch release
of veterinary vaccines in May 2008, Strasbourg, and to
the Biological Standardisation Steering Committee of the
European Pharmacopoeia in June2008, Strasbourg. As a
result the Biological Standardisation Steering Committee
endorsed a collaborative study to validate the wider
transferability of the proposed assay and to confirm its
suitability for inactivated rabies vaccines on the European
market. A small scale feasibility phase showed that the assay
could be successfully transferred to another laboratory and
the following collaborative study was then carried out.
3. PARTICIPANTS
13 laboratories from 10 countries including EU Member
States, Canada and the USA participated in the collaborative
study. 8 were official control laboratories of Regulatory
Authorities and 5 were manufacturers. Participants are listed
in alphabetical order by country in section 9 of this report.
Throughout the report the participants are referred to by a
code number which is unrelated to the order of listing.
4. STUDY OUTLINE
4.1. Principle of the assay
The serological assay used involves the immunisation of
groups of 6 mice with approximately 1/5
th
the recommended
dose volume of the test vaccine diluted appropriately, or
of the reference standard vaccine preparation which is
adjusted to the minimum potency allowed in the Ph. Eur.
14 days after immunisation blood samples are taken and
the sera are tested individually for rabies antibody using the
described virus neutralisation assay. Briefly, sera are titrated
on 96well microtitre plates and incubated for 1h with
rabies virus. After adding BHK cells and incubating for 48h
the presence of un-neutralised rabies virus is revealed by
immunofluorescence. Dilutions of the sera that reduce the
number of fluorescent cells by 50 per cent are calculated.
4.2. Assay design
Laboratories were asked to test the potency of 4 different
inactivated rabies vaccines for veterinary use using the
candidate serological method. Samples were coded (A-D)
so as not to reveal their identities. The vaccines represent
a range of products available on the EU market and are
produced by different manufacturers. Three independent
repetitions of the assay with the 4 test vaccines and the
reference preparation were to be performed. The serum
neutralisation activity of the test vaccines were assessed for
compliance by comparing against activity for the reference
vaccine which was set at the minimum allowed potency.
In addition, participants were provided with a panel of 3
prepared test sera labelled High, Medium and Low to be
tested in the virus neutralisation assay. The sera panel was
to be tested on 3 separate occasions. All assays included a
negative serum sample.
In addition to the main study, 1 laboratory accepted to retest
each of the test samples using the Ph. Eur. vaccination
challenge assay for Rabies vaccine (inactivated) for veterinary
use (Ph. Eur. monograph 0451) to confirm the potency of the
test samples.
4.3. Materials
4.3.1. Material provided by the EDQM
4.3.1.1. Test Samples
Sample A
A commercially produced current lot of Rabdomun (Essex
Animal Health-Pfizer). A liquid presentation rabies vaccine
(inactivated) of the Flury LEP strain. 1 mL/ dose. Approved
specification >1 IU/dose. Manufacturers estimated potency
at release: 7.3 IU/mL.
Sample B
A commercially produced current lot of Rabigen (Virbac).
A liquid presentation rabies vaccine (inactivated) of the
PRV strain. 1 mL/ dose. Approved specification >1 IU/dose.
Manufacturers estimated potency at release: 15 IU/mL.
Sample C
A commercially produced expired lot (exp. date 06, 2005)
of Biocan R inj. a.u.v. (Bioveta, a.s.). A liquid presentation
rabies vaccine (inactivated) of the SAD Vnukovo
strain. 1mL/ dose. Approved specification >2 IU/dose.
Manufacturers estimated potency at release in 2003:
3 IU/mL. Potency as determined at PEI in 2009: 0.25 IU/mL.
Sample D
A commercially produced current lot of Nobivac Rabies
(Intervet/Schering-Plough Animal Health). A liquid
presentation rabies vaccine (inactivated) of the PRV
strain. 1mL/dose. Approved specification >3 IU/dose.
Manufacturers estimated potency at release: 8 IU/mL.
All test samples were kindly donated by the respective
manufacturers.
Each laboratory received 6 vials (2 per test) of each test
sample to be stored at 4C upon receipt.
4.3.1.2. Control Sera (CS)
Control sera were prepared and pre-tested in the laboratory
of the co-project leader at PEI.
3 vials of 100 L were provided for each sera type to be
stored at -20 C upon receipt (1 vial per test).
Serum low
Mean estimated activity from 10 assays: 2.4 IU/mL, SD 0.98.
Serum medium
Mean estimated activity from 10 assays: 5.8 IU/mL, SD 1.12.
Serum high
Mean estimated activity from 10 assays: 28.1 IU/mL, SD10.04.
Negative sera
Negligible activity.
4.3.1.3. Reference Material
Ph. Eur. BRP batch 4
A freeze dried preparation of 11 IU/vial to be reconstituted
in 1 mL phosphate buffered saline pH 7.1. To be stored at
-70C upon receipt.
Each laboratory received 3 vials.
2
nd
WHO IS for rabies immunoglobulin (RAI)
30 IU/ampoule
This material was to be reconstituted and diluted in 15 mL
of PBS. Once completely resuspended the material should be
aliquoted in 150 L volumes and stored at 70C until use.
Each laboratory received 1 vial.
4.3.1.4. Specic reagent
Fluorescein labelled antibody
Rabies antibody fluorescein conjugate (FDI, Microtest), to
be dissolved in 5 mL Aqua dest. 4 aliquots of 1.25 mL were
prepared and stored at 20C upon receipt.
Each laboratory received 1 vial.
4.3.2. Other required materials supplied by the
participant
Rabies is a zoonotic disease and is pathogenic for humans.
While working with rabies virus, the local biosafety rules
for this material must be strictly applied. Tests have to be
performed in laboratories with adequate containment. The
staff of the laboratory should be familiar with the handling
of rabies virus. Only persons showing a sufficient anti-rabies
antibody titre should perform rabies testing.
4.3.2.1. Instruments, Material
Laminar ow
CO
2
incubator
Fluorescence microscope, lter 450-490 nm
Water bath (37C)
Test tube shaker (e.g. Vortex)
96 wells microtiter plates, at bottom (Nunc, Greiner)
Pipettes, variable (e.g. 10-100L, 100-1000L)
Pipettes, multi-channel (50-200 L)
Syringes (1mL), needles (0.45 x 25 mm, 26Gx1) for
vaccination, (0.60 x 25 mm, 23Gx1) for taking blood
Serum tubes (1 mL, Eppendorf, Sarstedt)
4.3.2.2. Reagents, media, cells
BHK-21 cells, 2.5 x 10
5
cells/mL
Rabies challenge virus (adapted to BHK-21 cells),
diluted to 100 CCID
50
/50L MEM-Earle (Biochrom AG
Cat.No. F0315)
Cell growth medium with 10% FCS (1l MEM-Earle
with 10 mL L-glutamine, 10 mL NEA, 100 mL FCS)
Dilution medium without FCS
PBS, pH 7.1
80 % Acetone (diluted with H
2
O)
Mouse anesthetic (e.g. Ketamin 10 mg/mL with
Xylazin 0.4 mg/mL; application: 0.1 mL intraperitoneal
per 10g animal weight)
4.3.2.3. Animals
NMRI mice, female, 18-20 g; 6 mice per vaccine and
reference (30 mice per assay)
4.4. Method
Participants were provided with a common protocol to
perform the assay, details of which are provided below.
4.4.1. Sera preparation
For each test: 6 animals per test vaccine and 6 animals per
reference standard.
Animals were to be allowed an adaptation period of 2-3 days
before immunisation.
4.4.1.1. Preparation of samples for injection
BRP4 standard was dissolved and diluted to 1 IU/mL using
PBS (1 mL BRP4 + 10 mL PBS).
Sample A: no dilution
Sample B: no dilution
Sample C: diluted 1:2 using PBS (1 mL vaccine + 1 mL PBS)
Sample D: diluted 1:3 using PBS (1 mL vaccine + 2 mL PBS)
4.4.1.2. Immunisation
6 mice were vaccinated respectively with a test vaccine or
the BRP4 standard.
Each mouse was injected intraperitoneally with 0.2 mL of
the vaccine or the BRP (prepared as noted above).
4.4.1.3. Blood collection and preparation of sera
Blood was collected 14 days after immunisation by heart
puncture after anaesthetising the animals (post-mortem
puncture and jugular bleeding were accepted as apossible
alternative to heart puncture). In either case thestaff
members should be well trained and experienced with the
method they used. The anaesthetic used should be tested in
2-3 control mice prior to the test.
Serum was extracted after centrifugation (5,500 x g) and
kept at -20C until the neutralisation test was carried out.
4.4.2. Neutralisation Test
Each assay required:
2 control plates
2 plates per test sample
2 plates for the reference standard (BRP 4)
4.4.2.1. Plate set up and neutralisation
Prepare:
standard serum RAI 30 IU/mL (IS) diluted to 2 IU/mL
BHK-21 cells, 2.5 x 10
5
cells/mL
working dilution of CVS = Rabies challenge virus
(adapted to BHK-21 cells) diluted to 100 CCID
50
/50 L
Virus titration: prepare 10-fold dilution steps (10
-1
-10
-4
) from
working dilution for the control plate; use 50 L of each
virus dilution (10
0
-10
-4
) per well.
CONTROL PLATE 1 (REFER TO PLATE LAYOUT)
Titration of RAI
Place 80 L of the medium in wells A1-A6, 50 L of the
medium in wells B1-G6 and 200 L of the medium in wells
H1-H6. Add 20 L of RAI (2 IU/mL) in wells A1-A6. Transfer
50 L of each well from row A to H, discard 200 L from
wells H1-H6.
Virus titration
Place 50 L of the medium in wells A7-E12. Add 50 L of
each virus dilution (10
0
-10
-4
) as indicated.
Negative control
Place 40 L of the medium in wells G7-G12 and add 10 L of
negative serum in each well.
Cell control
Place 100 L of the medium in wells H7-H12.
CONTROL PLATE 2 AND SERUM TITRATION PLATES (REFER TO PLATE
LAYOUT)
Place 80 L of the medium in wells A1-A12.
Place 50 L of the medium in wells B1-G12.
Place 200 L of the medium in wells H1-H12.
Perform 2-fold dilution steps with 4 replicates per
control serum or test serum: pipette 20 L of each of
the 1
st
test serum into wells A1-A4, 20 L of the 2
nd
test
serum into wells A5-A8, etc. Transfer 50 L of each
of the serum dilutions from row A to B to C until H.
Discard 200 L from row H.
Add 50 L of CVS working dilution per well for all
wells in control plate 2 and the serum titration plates
and to the RAI and NC wells of control plate 1.
Incubate the plates for 1h at 37C in a CO
2
incubator.
Add 100 L of BHK-21 cell suspension (2.5 x 10
5
cells/
mL) to each well in the serum titration plates and to
each of the wells of the control plates.
Incubate plates at 37C in a CO
2
incubator for 48 h
1h.
4.4.2.2. Staining of infected cells (day 2) (48 h 1h)
Discard medium and wash MTP once by adding 100 L
of PBS in each well.
Discard PBS, add 100 L of acetone (80% at room
temperature) per well.
Incubate MTP for 30 min at room temperature.
Discard acetone and dry MTP in the air.
Dilute an aliquot of the rabies conjugate 1:100 with
PBS (e.g. 1.2 mL conjugate +118.8 mL PBS).
Add 90 L of conjugate per well and incubate at 37C
for 30 min.
Discard conjugate, wash twice with 150 L of PBS and
once with 150 L of distilled water.
Dry MTP in the air.
4.4.2.3. Test evaluation
Read the plates under UV-microscope (100-fold
magnification) and note the results in the test protocol:
Virus is neutralised Negative immunofluorescence
(sign)
Virus is not neutralised Positive immunofluorescence
(sign +)
The test is valid if:
all cell controls (CC) are free from fluorescing cells;
the titre of the CVS used is between 1.5 and
2.5log
10
ID
50
/50 L.
A vaccine complies if a one-sided limit test (Wilcoxon-Mann-
Whitneys exact test) shows that antibody titres obtained
with the test vaccines are greater than the antibody titres
obtained for the standard (P=0.95).
5. RESULTS
13 laboratories submitted results from assays. In this report
they are referred to by their code numbers (1 to 13) allocated
at random and not necessarily corresponding to the order
of listing in the list of participants. 11 of the 13 laboratories
carried out 3 fully independent assays using 6 mice per assay
and per vaccine, as requested. Laboratory 10 reported results
from 1 assay only and laboratory 13 performed 3 assays
using 6 mice per assay per vaccine but all the sera were
evaluated in the same experiment. Statistical evaluation was
performed at the EDQM.
5.1. Serum activities by serology
The virus neutralisation ratios were submitted to statistical
calculations using the CombiStats software [7]. The dose/
response curves were estimated using the probit model [8].
In cases where no slope could be estimated, the Spearman/
Kaerber method was used. The relative activity per serum
was calculated against the 2
nd
International Standard for
RAI. An example sheet is provided in Appendix 1. Activities
below the quantification limit were set to 0 IU/mL and these
are designated as non-responders in the remainder of this
report. Activities above the quantification limit were set to
999 IU/mL which is an arbitrarily chosen value higher than
the highest measurable activity with the chosen design.
The exact value is irrelevant since subsequent analyses are
based on rank-orders. As it is more convenient to work with
activities on an arithmetic scale they were transformed
CONTROL PLATE 1
1 2 3 4 5 6 7 8 9 10 11 12
RAI CVS
A 1:5 1:5 1:5 1:5 1:5 1:5 10
0
10
0
10
0
10
0
10
0
10
0
B 1:10 1:10 1:10 1:10 1:10 1:10 10
-1
10
-1
10
-1
10
-1
10
-1
10
-1
C 1:20 1:20 1:20 1:20 1:20 1:20 10
-2
10
-2
10
-2
10
-2
10
-2
10
-2
D 1:40 1:40 1:40 1:40 1:40 1:40 10
-3
10
-3
10
-3
10
-3
10
-3
10
-3
E 1:80 1:80 1:80 1:80 1:80 1:80 10
-4
10
-4
10
-4
10
-4
10
-4
10
-4
F 1:160 1:160 1:160 1:160 1:160 1:160
G 1:320 1:320 1:320 1:320 1:320 1:320 NC NC NC NC NC NC
H 1:1600 1:1600 1:1600 1:1600 1:1600 1:1600 CC CC CC CC CC CC
RAI=immunoglobulin standard, CVS=rabies virus, NC=negative control, CC=cell control
CONTROL PLATE 2 AND SERUM TITRATION PLATES
1 2 3 4 5 6 7 8 9 10 11 12
Medium
Sera/
transfer
Dilution CS / Low CS /Medium CS / High
80 L 20 L 1:5 A
50 L 50 L 1:10 B
50 L 50 L 1:20 C
50 L 50 L 1:40 D
50 L 50 L 1:80 E
50 L 50 L 1:160 F
50 L 50 L 1:320 G
200 L 200 L 1:1600 H
CS=control sera (high, medium or low)
to logarithms after addition of 1 IU/mL to avoid problems
with non-responders. All transformed titres are therefore in
the range 0 (for non-responders) to 3 (for full responders).
A complete overview is given in Table 1 and a graphical
rendering of the data is provided in Figure 1.
In order to assess the intra- and inter-laboratory variation
due to the serological method itself, 3 centrally provided
control sera were included in each assay. All laboratories
provided data from 3 independent assays as requested, except
for laboratory 13 which provided only 1 set of data. The log
transformed activities are listed in Table 2 and a graphical
representation of the mean value per laboratory is given
in Figure 2. The overall mean of the sera is 0.44 IU/mL,
0.85IU/mL and 1.38 IU/mL for the low, medium and high
control sera respectively, with respective standard deviations
of 0.13IU/mL, 0.18 IU/mL and 0.23 IU/mL. This implies
that there is an overall satisfactory separation between the 3
sera and that the vast majority of the individual results stay
within a 2-fold range around the overall mean.
5.2. Potency of the vaccines
The potency of the test vaccines A, B, C and D was assessed
by comparison of the induced serum activities with those
induced by reference preparation BRP Batch 4 (assigned
potency 11 IU/vial). The dilution scheme was carefully
designed for optimal discrimination (pass/fail) at the critical
level of the labelled potency. The BRP had to be diluted
to 1 IU/mL and 0.2 mL had to be injected into the mice,
which is the dose that was expected to induce measurable
activities at the low end of the dose-response regression.
The test vaccines had to be administered at a dose of 0.2 IU
on the basis of the labelled potency. For vaccines A and B,
which both had a labelled potency of 1 IU/mL this implied
0.2 mL of undiluted vaccine. The rationale is that if the
test vaccines induce higher activities than the BRP at this
dose, the conclusion is justified that the injected dose was
actually more than 0.2 IU and therefore the original sample
contained at least 1 IU/mL. For vaccines C and D, which had
labelled potencies of 2 IU/mL and 3 IU/mL respectively, this
implied a 1:2 and 1:3 dilution prior to injection of 0.2 mL
into the mice. The rationale is similar: if higher activities are
induced than the BRP, the injected dose was actually more
than 0.2 IU and therefore the diluted samples contained
more than 1 IU/mL. It then follows that the undiluted
samples contained at least 2 IU/mL and 3 IU/mL respectively.
The activities were compared with the one-sided exact
test of Wilcoxon-Mann-Whitney [9] using the CombiStats
software. This test is a non-parametric test based on rank-
orders and is robust against non-normality and outliers. If
the test is significant (P<0.05) and the mean response of
the test vaccine is higher than the mean response of the
BRP, the vaccine is said to pass the release test. Otherwise
it fails or additional testing would have to be performed in
practice. It should be noted that a significant p-value alone
is not sufficient as it may also indicate that the test-vaccine
is significantly less potent than the required limit. This is
because the software decides on the direction of the one-
sidedness of the test on the basis of the mean response. An
example sheet is provided in Appendix 2. An overview of
the resulting p-values is given in Table 3a for the individual
assays (6 mice per vaccine) and in Table 3b for the pooled
results per laboratory (18 mice).
As part of the study, 1 laboratory carried out the challenge
assay in mice to confirm the expected potencies. They were
found to be 9.2 IU/vial, 3.9 IU/vial, 2.1 IU/vial and 19.2IU/vial
for vaccines A, B, C and D respectively. Before the study the
project leader also determined the potency of vaccine C
using the in vivo assay, with an averaged result of 0.25 IU/mL
from 2 assays.
Vaccine C was included in this study because it was known
to be sub-potent and was therefore expected to fail the test.
The other vaccines were expected to pass the test. As can be
seen from Table 3a this is confirmed in most of the assays
with correct decisions in 90%, 90%, 97% and 74% of the
assays for the 4 test vaccines respectively. When the results
of the 18 mice per laboratory are pooled the number of
correct decisions is 100% for vaccines A and B. Laboratory
2 has a repeatable problem with the correct assessment of
vaccine C. Laboratory 5 has a repeatable problem with the
correct assessment of vaccine D. As the pass/fail decision
is very clear and reproducible in all other laboratories it
would be worthwhile to investigate why these 2 laboratories
deviated so largely from the overall outcome.
5.3. Inclusion or exclusion of non-responders
It is known that some animals tend to show no immune
reaction, even for relatively high doses. To overcome this
problem it had been suggested to exclude non-responders
from the statistical analyses. The resulting p-values upon
exclusion of non-responders are also shown in Tables 3a
and 3b. However, as can be seen from Table 1 the number
of non-responders tends to depend on the potency of the
vaccines. The rate of non-responders is 5.9% for the BRP,
3.6% for vaccine A, 2.3% for vaccine B, 7.2% for vaccine C
and 2.7% for vaccine D. The rate for the sub-potent vaccine
C is markedly higher than for the high potency vaccines A,
B and D. This observation indicates that exclusion of non-
responders can bias the outcome of the assay as the non-
responsiveness may not be entirely intrinsic to the animal,
but may reflect a sub-potent vaccine. Moreover, as can be
seen from Tables 3a and 3b exclusion of non-responders
does not improve the number of correct pass/fail decisions.
To reflect the true potency non responders should therefore
be included in the analysis. However, there may be a need
to impose a maximum on the number of allowed non-
responders for the reference vaccine, e.g. not more than
1 non-responder in a set of 6, or not more than 2 in a set
of 8, in order to monitor assay validity. An example of how
calculations could be performed to determine the number is
provided in appendix 3.
5.4. Number of animals required
As can be expected, the number of correct pass/fail decisions
increases with the number of animals included in the assay.
A number of 6 animals per vaccine may be sufficient for
vaccines that are well above the labelled potency, but for
vaccines closer to this limit a higher number of animals
may be required. As can be seen from Table 3b the number
of correct pass/fail decisions becomes in many laboratories
almost 100% when the results of 18 mice are pooled. The
optimal number of animals will probably be less than 18. In
order to get an idea about the optimal number of animals
for the vaccines included in this study, the results were
submitted to bootstrap resampling for sample sizes between 3
and 12 drawn from the set of 18 mice, with replacement and
ignoring inter-assay variation. The estimated rate of pass
decisions is listed in Table 4. Each percentage is the result of
100.000 bootstrap simulations. Ideally the percentages should
be 100% for vaccines A, B and D and should be 0% for vaccine
C. It can be seen that a quite satisfactory discriminative power
can in many cases already be achieved with 8-10 animals. The
use of only 6 animals may be sufficient in some cases, but in
general it seems recommendable to use at least 8 animals.
Using more than 10 animals does not noticeably increase the
power of the test anymore.
6. DISCUSSION
The Ph. Eur. monograph 0451 on Rabies vaccine
(inactivated) for veterinary use [1] describes in section3-5
an in vivo batch potency test that is based on the NIH
test and involves a vaccination challenge assay with test
and reference vaccines. A single in vivo assay for a given
test vaccine requires the use of at least 100 animals. Due
to the variable nature of the assay and the difficulties
encountered meeting the validity criteria it is in addition
often necessary to repeat the assay [2,3]. Section 2-4-3 of the
monograph0451 allows the replacement of the vaccination
challenge assay for batch potency by a validated alternative
method and briefly describes a serology based assay. Despite
the possibility of using a serological assay until present this
has not been put into practice.
B. Krmer and colleagues at PEI have developed and
validated a serological potency assay for Rabies vaccine
(inactivated) for veterinary use which greatly reduces
the number of animals required and refines considerably
the amount of suffering entailed [5]. The aim of this
collaborative study was to demonstrate the wider
transferability of the proposed assay and confirm its
suitability for the potency assay of inactivated rabies vaccines
for veterinary use with the ultimate goal of encouraging its
use as a replacement method for the vaccination challenge
assay for batch potency. The study involved 13 laboratories
who tested 4 different vaccines from the European market
with different strain composition. One of these samples
was known to be of insufficient potency. Centrally provided
control sera of different activity was also provided to help
evaluate the assay in vitro aspects of the method.
Results using commonly provided control sera of high,
medium and low potency (Table 2 and Figure 2) demonstrate
good intra- and inter-laboratory variation of the in vitro
component of the method as evidenced by the overall
satisfactory separation between the 3 sera and the vast
majority of the individual results which stay within a 2-fold
range around the overall mean. The traditional vaccination
challenge assay is known to be highly variable[2, 3]
and as a reflection of this specifications in the Ph.Eur.
monograph0451 are set with confidence limits of not less
than 25per cent and not more than 400 per cent of the
estimated potency. This study suggests that the proposed
method would provide a significant improvement with
respect to assay repeatability and reproducibility. It should
also be noted that the vaccination challenge assay requires
at least 4 weeks for completion not including the assay
preparation and data analysis. The proposed assay could be
performed in just under 3 weeks thus providing a reduction
in the time required to complete the analysis.
Potency of the test vaccines were compared to the reference
standard (BRP4) using a non-parametric limit test based
on rank-orders which is robust against non-normality
and outliers. Three vaccines of sufficient potency (A, B
and D) and 1 of insufficient potency (C) were evaluated
against the BRP. As noted in Tables 3a and 3b the vast
majority of individual assays (6 mice) resulted in the
appropriate outcome. When results per laboratory were
pooled (18mice), 100 per cent of correct results were
found for vaccines A and B. For vaccines C and D all
laboratories but 1 found the correct result. Laboratory 2 had
a repeatable problem with vaccine C and Laboratory 5 had a
repeatable problem with vaccine D. It is unclear why these
2laboratories had discrepant results in these specific cases.
Although the method was new to all users the technique
seems to have been well transferred as evidenced by the
results for the other vaccines in those labs. There may have
been unaccounted for problems with dilution for individual
assays however this is difficult to confirm. Participants were
all requested to use female NMRI mice and no one reported a
deviation from this strategy. However as always with animal
experiments it remains possible that the source of strains
used could be a contributing factor to observed differences.
In the strategy used in this study 1/5
th
the recommended
dose volume of the test vaccine diluted appropriately, or
of the reference standard vaccine preparation which is
adjusted to the minimum potency allowed in the Ph. Eur.
was administered to the mice. The Ph. Eur. monograph0451
minimum requirement for potency of an inactivated rabies
vaccine for veterinary use is 1 IU/dose. The reference
material (BRP4) was thus diluted to 1 IU/mL (on the basis of
a 1 mL dose) and 0.2 mL was injected into each mouse. For
test vaccines A and B the approved specification is
1 IU/mL so 0.2 mL of the vaccines were administered
undiluted. However some of the vaccines tested had
approved specifications above 1 IU/mL (2 IU/mL for
vaccineC and 3 IU/mL for vaccine D). As the activity of
the test vaccines was to be compared to the reference
(at 1IU/mL) the test vaccines were diluted appropriately
(1:2 or 1:3 respectively) before administration for a valid
comparison. As evidenced by the results, this strategy
worked effectively in the given circumstances. Application
of this strategy for a given vaccine should be validated on an
individual basis. It may also be interesting when developing
a standard strategy for a vaccine with a specification higher
than 1 IU/mL to consider applying the reference vaccine at
the approved cut off dose (e.g. 2 IU/mL or 3 IU/mL) and then
administering the test vaccine undiluted. This variation in
the approach would have to be appropriately validated for
the individual case.
In the assay set up of this study, 6 animals were used per
sample. While this resulted in the majority of individual
assays achieving the anticipated result it can be expected
that the number of correct pass/fail decisions increases
with the number of animals used in the assay. An analysis
of all the available data using a simulation to determine the
results for sample sizes between 3 and 12 was performed
and the results are provided in Table 4. It appears that
satisfactory results with sufficient discriminating power can
be obtained in many cases with 8 -10 animals. While the use
of 6 animals may be sufficient in some cases and particularly
if the vaccines are well above the labelled potency, in general
at least 8 animals would be recommended. Depending on
the vaccine 10 animals may be required however more
than 10 does not appear to provide significant advantages.
In any case the use of 8-10 animals per test group would
significantly reduce the number of animals required per
assay to 16-20 (reference and test vaccine/assay) as compared
to 100 for the vaccination challenge assay. The final choice
of group size should be established based on evidence
collected during the validation for individual vaccines.
In any animal experiment it may be expected to encounter
some animals that are non-responders i.e. that show no
immune reaction even at very high doses. It had been
suggested that animals which show no response in this assay
could be considered non-responders and excluded from the
final analysis. To evaluate the impact of this proposal results
were calculated both with and without the presumed non-
responders. The results are shown in Tables 3a and 3b. The
number of animals without measurable response appears to
depend on the potency of the vaccine. The number is lower
for vaccines A, B and D which have potencies above the cut
off and higher for vaccine C which has reduced potency.
This higher number of animals which lack a measurable
response in the case of vaccine C is presumably due to
the presence of both real non-responders and a lack of
response related to sub-potency of the vaccine. As such the
exclusion of non-responders could bias the assay outcome.
It is also noted that the exclusion of non-responders does
not improve the number of correct pass fail decisions. It is
therefore unadvisable to remove non-responders from the
final calculations. Nevertheless it may be useful to impose
a maximum on the number of allowed non-responders for
the reference vaccine in order to keep surveillance on the
validity of the assays once assay conditions are established
and validated e.g. not more than 1 non-responder in a set
of 6 or not more than 2 in a set of 8 (example calculations
provided in appendix 3).
The assay outlined in the collaborative study is a limit
test which is able to discriminate vaccine samples that
have a potency value above the approved minimum.
As noted in section 5 and throughout the discussion,
individual validation for specific vaccines can fine tune the
discriminatory power to specific situations by choosing
the most suitable conditions in terms of reference dose,
dilutions used and animal number. A survey of typical
vaccine batches from EU manufacturers suggests that in
general the vaccines are formulated with a comfortable
margin above the minimum allowed potency. As such
this assay could be an effective tool as a batch potency
test in confirming the compliance with the specifications.
It provides a real advantage in terms of animal welfare
and assay consistency. However, in cases where the batch
potency of the test vaccine is at the limit of acceptable
potency values, significantly more animals would have
to be used in order to determine if the batch were just
above, or just below the line. Alternatively in such cases
the vaccination challenge assay might be used to assess the
potency, although the vaccination challenge assay has also
its limitations, particularly with respect to its precision.
Although the final result of the assay is a limit test which
does not involve a final potency value for the test vaccine,
the relative activity per serum is calculated against a
common standard for rabies immunoglobulin (2
nd
IS RAI)
and the assay includes the calculation of the overall mean
of the sera from the test and reference preparations. These
values if charted over time can be used to help monitor both
the assay consistency (with respect to the values obtained
for the BRP) and production consistency (by observing the
values obtained for successive batches of the test vaccine).
7. CONCLUSIONS AND RECOMMENDATIONS
The study has shown that a limit test using a relatively small
number of animals in a serological assay as compared to the
full vaccination challenge in vivo potency test is possible,
reproducible and reliable. The optimal number of animals
per vaccine is product specific but may roughly be indicated
to be between 8 and 10 for the products included in this
study. Non-responders should be included in the analysis
because they may reflect sub-potent vaccines. However,
there may be a need to impose a maximum on the number
of non-responders allowed for the reference vaccine as a
monitor for assay validity. In this study it was necessary
to fix a given dose for the reference that would be used to
compare with all test vaccines but the optimal dose for
reference and test vaccine should be determined during the
in-house validation of the method to further improve its
discriminative power.
This assay provides a significant 3R improvement for the
batch potency testing of Rabies vaccine (inactivated) for
veterinary use in terms of both the number of animals used
and the amount of suffering entailed and as evidenced by
the study provides a more reliable and reproducible assay
format than the vaccination challenge assay. It also reduces
the time required as compared to the vaccination challenge
assay.
The Ph. Eur. already allows for replacement of prescribed
assays by 3R alternatives provided sufficient validation has
been carried out. This collaborative study is a major step
in the validation of such an assay and all users are strongly
encouraged to take the remaining necessary steps to validate
and implement the assay in their own specific situations
without delay.
It is also strongly recommended that details of this assay
be taken up by the group of experts 15V of the Ph. Eur. for
inclusion as an alternative assay to the batch potency test
in monograph 0451. This would be an important step to
facilitate the regulatory implementation of the assay.
Additionally, international partners are encouraged to
review the data and consider acceptance of the assay in their
own systems as this is an important contributing factor to
successful implementation of alternative methods in light of
the global nature of the veterinary vaccine industry.
8. ACKNOWLEDGEMENTS
The organisers would like to thank Dr. B. Krmer (PEI,
Germany) and Dr. L. Bruckner (IVI, Switzerland) for acting
as project leader as well as Dr. Krmers colleagues at PEI
for their additional support. Sincere thanks are extended
to all of the participants for their important contribution
to the project. The organisers are grateful to Bioveta a.s.
(Czech Republic), Essex Animal Health Pfizer (Germany),
Intervet/Schering Plough Animal Health (The Netherlands)
and Virbac S.A. (France) for providing the study samples
for the project and to the National Institute for Biological
Standards and Control (United Kingdom) for providing the
international standard of rabies immunoglobulin.
The collaborative study was run in the framework of the
Biological Standardisation Programme of the EDQM with
the support of the Council of Europe and the European
Commission (internal project number BSP105). The
statistical evaluation was performed by Mr. A. Daas (EDQM)
and the project was coordinated by Dr. C. Milne (EDQM).
9. PARTICIPANTS (LISTED IN ALPHABETICAL
ORDER BY COUNTRY)
Els Goossens, Pieter-Jan Serreyn, Veterinary and
Agrochemical Research Centre, Belgium
Aline Rinfret, Gayle Pulle, Health Canada, Canada
Ivana Benesova, Institute for State Control of Veterinary
Biologicals and Medicaments, Czech Republic
Michael Caras, Petra Strnadov, Bioveta, a.s., Czech Republic
Claire Chouvet, Jacques Lchenet, Mrial, France
Florence Cliquet, Alexandre Servat, Agence Franaise de
Securit Sanitaire des Aliments, Nancy, France
Franois-Xavier Deramoudt, Virbac S.A., France
Beate Krmer, Paul-Ehrlich-Institut, Germany
Karin Stnkel, Schering-Plough; Essex Animal Health,
Germany
Gbor Kulcsr, Central Agricultural Office, Directorate of
Veterinary Medicinal Products, Hungary
Imke Kross, Chistian Weiss, Wim Hesselink, Intervet/
Schering-Plough Animal Health, The Netherlands
Lukas Bruckner, Institute of Virology and
Immunoprophylaxis, Switzerland
Christian Borgneit, Catrina Stirling, Pfizer, United Kingdom
(Correspondence only)
Donna M. Gatewood, Alethea Fry, United States Department
of Agriculture, Animal and Plant Health Inspection Service,
Center for Veterinary Biologics, USA
10. ABBREVIATIONS
3R: Replacement, reduction, refinement of animal tests,
BHK: Baby hamster kidney cell line, BRP: Biological
Reference Preparation, BSP: Biological Standardisation
Programme, CCID
50
: Cell culture infectious dose 50%,
CS: Control serum, CVS: Challenge virus strain, EDQM:
European Directorate for the Quality of Medicines &
HealthCare, EU: European Union, FAVN: Fluorescent
Antibody Virus Neutralization, FCS: Ftal calf serum, FDI:
Fujirebio Diagnostics, Inc., IS: International standard,
IU: International unit, IVI: Institute of Virology and
Immunoprophylaxis, LEP: Low egg passage, MEM: Minimum
Essential Medium, MTP: Microtitre plate, NEA: Non-essential
amino acids, NIH: National Institute of Health, NMRI: Naval
Medical Research Institute, OIE: World Organisation for
Animal Health (Office International des Epizooties), OMCL:
Official Medicines Control Laboratory, PBS: Phosphate-
buffered saline, PEI: Paul-Ehrlich-Institut, Ph. Eur.:
Pharmacope Europenne / European Pharmacopoeia, PRV:
Pseudo-rabies virus, RAI: Rabies immunoglobulin, RFFIT:
Rapid Fluorescent Focus Inhibition Test, SAD: Street
AlabamaDufferin, SD: Standard deviation, USA: United
States of America, UV: Ultraviolet
11. REFERENCES
[1] Rabies vaccine (inactivated) for veterinary use,
monograph 0451. Ph. Eur. 6th Edition. Strabourg,
France: Council of Europe; 2008.
[2] Daas A, Milne C. Establishment of Batch 4 of the
Biological Reference Preparation (BRP) for Rabies
Vaccine (Inactivated) for Veterinary Use. Pharmeuropa
Bio 2004(1):17-22.
[3] Servat A. et al. In vivo Potency Tests of Rabies
Inactivated Vaccines for Veterinary Use A 2-year
retrospective analysis of data according to the criteria
of the European Pharmacopoeia. Pharmeuropa
2008;20(4): 655-64.
[4] Russel W.M. S, Burch R.L. The principles of humane
experimental technique. London, UK: Methuen; 1959.
[5] Krmer B et al. The rapid fluorescent focus
inhibition test is a suitable method for batch potency
testing of inactivated rabies vaccines. Biologicals
2009;37(2):119-26.
[6] World Organisation for Animal Health (OIE). Manual
of Diagnostic Tests & Vaccines for Terrestrial Animals,
Chapter 2.1.13. Rabies. OIE 6th Ed.; 2008.
[7] EDQM, Council of Europe. CombiStats (4.0) [software].
Web: www.combistats.eu
[8] Finney D.J. Statistical Method in Biological Assay. 3rd
Edition. London, UK: Griffin; 1978.
[9] Mann H. B, Whitney D. R. On a Test of Whether one
of Two Random Variables is Stochastically Larger
than the Other. The Annals of Mathematical Statistics
1947;18(1): 5060.
Table 1 Estimated activities per serum (log IU/mL)
Lab Mouse Assay 1 Assay 2 Assay 3 Assay 1 Assay 2 Assay 3 Assay 1 Assay 2 Assay 3 Assay 1 Assay 2 Assay 3 Assay 1 Assay 2 Assay 3
1 0.740 0.368 1.090 1.207 1.137 1.160 1.278 0.621 1.373 0.368 0.507 0.583 0.802 1.349 1.373
2 0.802 0.000 0.895 0.679 1.422 1.230 1.349 1.278 1.373 0.999 0.933 0.131 0.413 1.207 1.302
3 0.000 0.932 0.699 1.422 1.422 1.760 1.422 0.413 1.373 0.621 0.810 0.257 1.422 0.999 1.373
4 0.802 0.679 0.131 0.413 0.461 0.097 0.999 0.999 1.373 0.092 0.106 0.583 0.740 0.999 0.760
5 0.802 0.125 0.000 1.422 0.999 1.230 0.254 1.884 1.302 0.802 0.862 0.435 0.932 0.413 1.373
6 0.125 1.137 0.627 1.278 0.932 1.230 1.207 1.422 1.373 0.621 0.792 0.301 0.999 0.999 1.373
1 0.626 0.755 0.617 1.884 3.000 2.801 3.000 0.823 1.449 0.866 1.090 1.199 1.810 1.739 1.795
2 0.175 0.524 0.921 3.000 2.111 2.801 1.472 1.534 2.104 1.278 1.022 1.274 3.000 0.954 2.131
3 0.679 0.645 0.164 3.000 1.739 2.017 1.244 1.742 1.711 0.000 1.373 1.133 3.000 2.111 1.712
4 1.147 0.894 0.995 3.000 1.592 1.941 0.645 2.053 1.859 0.000 1.373 0.000 3.000 2.036 2.029
5 0.417 0.699 1.071 1.810 2.185 2.801 1.919 2.252 1.859 1.884 0.760 1.274 0.866 1.961 2.704
6 0.289 0.875 1.206 1.278 3.000 2.355 1.919 2.040 1.940 1.884 1.160 1.274 1.207 2.260 2.704
1 0.602 0.437 0.461 1.113 1.183 0.932 1.520 1.183 0.999 0.277 0.277 0.461 0.863 0.607 0.866
2 0.602 0.437 0.511 1.183 1.113 0.999 1.520 1.113 1.137 0.000 0.354 0.327 0.974 0.773 0.932
3 0.494 0.538 0.327 1.183 1.183 1.067 1.520 1.183 0.999 0.277 0.354 0.368 0.863 0.652 0.866
4 0.494 0.538 0.565 1.325 1.113 0.866 1.395 1.637 1.588 0.138 0.160 0.168 0.789 0.652 0.679
5 0.602 0.538 0.621 1.254 0.000 0.000 1.218 1.113 1.588 0.398 0.243 0.289 1.053 0.719 0.802
6 0.602 0.437 0.368 1.113 1.113 0.999 1.520 1.113 1.137 0.314 0.277 0.368 1.053 0.587 0.565
1 0.836 0.634 1.137 1.397 1.090 1.422 1.239 1.387 0.917 0.398 0.954 0.560 0.000 1.287 1.611
2 1.256 0.000 0.999 1.933 1.519 3.000 1.659 1.713 1.614 0.977 1.090 0.718 1.470 1.452 1.852
3 0.905 0.881 1.207 0.781 1.519 1.422 1.179 1.170 0.177 1.470 0.152 1.263 0.102 1.713 1.611
4 0.655 1.090 0.327 1.113 1.519 1.137 1.329 1.310 1.210 1.183 0.888 1.215 1.470 1.309 1.254
5 0.559 0.583 0.866 1.933 1.373 1.422 0.751 1.354 1.416 0.102 1.090 1.391 1.958 1.153 1.611
6 0.851 0.747 0.092 3.000 1.373 1.422 1.179 1.480 1.614 0.659 1.908 0.103 1.325 1.713 1.611
1 0.665 0.621 0.866 1.477 1.422 0.289 1.671 3.000 1.884 0.000 0.802 0.415 1.479 1.137 1.278
2 0.848 0.000 1.207 1.477 1.349 0.254 1.360 1.349 1.137 0.000 0.145 0.000 0.726 1.349 1.278
3 0.975 0.621 0.000 1.671 1.067 1.207 1.332 1.958 1.278 0.216 0.461 0.255 0.855 0.327 1.137
4 0.917 0.511 0.222 1.671 1.422 1.137 1.195 1.958 1.207 0.991 0.679 0.462 0.604 0.327 1.207
5 1.118 0.866 1.137 0.000 1.422 1.278 1.466 1.349 0.289 0.668 0.565 0.289 0.593 0.000 0.461
6 1.449 0.932 0.461 1.195 1.422 1.207 2.144 1.349 1.278 0.668 0.511 0.195 1.196 1.349 0.679
1 0.454 0.508 1.466 1.509 1.137 1.807 1.783 1.067 1.058 0.871 1.325 0.736 1.654 1.662 3.000
2 0.513 0.794 0.619 1.652 1.067 1.697 1.944 1.662 0.998 0.624 1.552 1.204 1.848 1.588 1.697
3 0.000 0.461 0.737 1.950 1.736 3.000 1.629 0.999 1.465 1.075 0.289 1.204 1.968 3.000 1.466
4 1.286 0.863 0.253 1.430 3.000 3.000 1.644 1.137 1.240 1.141 0.562 1.204 1.645 3.000 3.000
5 0.874 0.571 1.466 2.052 1.588 1.697 1.644 1.067 1.240 0.680 1.039 0.805 1.968 3.000 3.000
6 1.258 0.680 0.927 1.876 1.736 1.466 1.277 0.999 1.807 1.507 1.015 0.508 1.968 1.588 3.000
1 0.261 0.232 0.000 0.000 0.634 1.233 0.000 0.823 3.000 0.071 0.115 0.066 0.383 0.712 0.039
2 0.633 0.383 0.123 1.090 1.220 0.516 1.090 1.251 1.298 0.000 0.081 0.068 0.700 0.168 0.461
3 0.423 0.208 0.515 1.022 0.998 0.914 0.954 0.000 1.298 0.203 0.201 0.257 1.088 0.460 1.144
4 0.232 0.000 0.515 0.888 1.220 0.694 0.954 3.000 0.000 0.051 0.097 0.583 0.760 0.998 0.913
5 0.000 0.051 0.034 1.090 0.790 0.629 1.090 0.699 0.740 0.071 0.071 0.095 0.768 0.301 1.232
6 0.301 0.154 0.016 1.022 0.790 0.000 0.954 0.823 1.227 0.000 0.020 0.146 0.642 0.267 0.461
1 0.598 0.621 0.935 1.048 1.137 1.278 1.126 1.475 1.207 0.669 0.743 0.518 1.050 1.005 1.215
2 0.548 0.999 0.687 1.197 1.207 1.278 1.265 1.005 1.207 0.212 0.743 1.008 1.305 1.161 1.008
3 0.187 0.565 1.280 1.048 1.207 1.278 0.975 1.475 1.278 0.182 0.743 0.628 1.114 0.808 1.008
4 0.598 0.740 0.451 1.285 1.137 1.207 1.191 1.475 1.137 0.314 0.796 0.875 1.305 0.869 0.874
5 0.143 0.802 1.146 1.520 1.067 1.278 1.050 1.475 1.278 0.498 1.237 0.419 0.783 0.000 0.874
6 0.117 0.621 1.173 1.285 1.278 1.278 1.126 1.475 1.278 0.317 0.416 0.679 1.177 0.944 0.874
1 0.699 0.490 0.646 1.467 3.000 3.000 0.000 1.075 3.000 0.422 1.297 0.099 1.377 3.000 3.000
2 0.477 0.472 0.535 0.904 1.292 1.229 1.658 1.294 1.066 0.383 3.000 0.675 1.135 1.291 1.216
3 0.301 0.348 0.698 3.000 3.000 3.000 1.147 1.294 3.000 0.824 0.074 0.998 0.000 1.505 0.239
4 0.131 1.075 0.000 1.135 0.000 1.229 1.511 1.294 1.228 1.397 1.297 0.306 1.377 1.627 0.770
5 0.477 0.116 0.061 1.467 1.292 0.891 0.000 1.631 1.066 0.477 1.297 0.348 0.829 1.505 1.216
6 0.344 0.714 0.206 1.135 1.515 3.000 1.464 1.631 3.000 0.346 0.000 1.219 1.377 3.000 1.216
1 0.602 0.977 0.845 0.547 1.637
2 0.910 1.113 1.113 0.659 1.183
3 0.000 1.637 1.045 0.494 1.045
4 0.547 1.183 1.045 0.277 1.183
5 0.910 0.977 1.113 0.212 1.183
6 0.781 1.113 1.113 0.212 0.977
1 0.468 0.174 0.608 0.852 0.699 1.161 1.507 1.251 3.000 0.257 0.000 0.000 0.578 0.112 0.968
2 0.534 0.260 0.663 1.378 0.760 1.370 1.248 1.322 1.158 0.288 0.150 0.487 0.000 0.723 0.000
3 0.517 0.292 0.663 1.265 0.760 1.370 1.376 1.251 1.350 0.418 0.106 0.487 0.696 0.844 0.968
4 0.578 0.657 0.663 1.265 1.394 1.161 1.507 3.000 0.723 0.036 0.042 0.785 0.854 0.805 1.160
5 0.853 0.451 0.952 1.051 0.823 1.370 3.000 0.699 1.350 0.022 0.430 0.396 0.629 0.686 1.160
6 0.569 0.468 0.865 1.265 0.823 1.161 1.507 1.394 3.000 1.051 0.423 0.598 0.696 0.743 1.365
1 1.067 0.977 0.954 1.422 0.977 1.373 1.349 1.183 1.373 0.088 0.119 0.199 0.999 1.048 1.302
2 0.866 1.254 1.090 1.137 1.183 1.230 1.349 1.254 1.373 0.000 0.102 0.477 1.349 1.124 0.954
3 0.145 0.000 1.160 1.067 0.000 1.373 1.278 1.254 1.302 0.324 0.000 0.232 1.067 1.282 1.090
4 0.621 1.254 0.888 1.349 1.254 1.373 1.422 1.183 1.022 0.508 0.398 0.583 0.932 1.124 1.373
5 1.067 0.910 0.699 1.067 0.977 0.000 1.278 1.325 1.230 0.862 0.845 0.335 1.137 1.124 1.022
6 1.137 0.977 1.022 1.207 1.045 0.954 1.422 1.325 1.373 0.000 0.243 0.301 1.067 0.991 1.302
1 0.823 1.022 0.699 1.286 1.519 1.983 1.160 1.446 1.519 1.160 0.421 0.000 2.132 1.446 1.519
2 0.888 0.823 1.090 1.375 1.519 1.373 1.160 0.823 0.823 0.131 0.301 0.529 2.057 1.983 1.983
3 0.428 0.640 1.160 1.982 2.057 1.230 0.888 1.519 1.519 1.446 1.415 1.302 1.519 1.519 1.519
4 0.383 1.160 0.176 1.375 1.519 1.446 1.446 1.373 1.519 0.954 1.230 0.529 1.373 2.132 1.090
5 0.888 0.760 0.265 0.954 1.519 1.519 1.446 1.446 1.519 1.230 1.438 1.230 1.983 2.132 1.519
6 1.230 0.265 0.176 0.823 1.422 2.057 1.446 1.090 1.160 1.230 0.630 1.090 3.000 2.132 2.057
9
10
Vaccine D Vaccine C Vaccine B Vaccine A BRP Batch 4
1
13
11
2
3
4
5
6
7
12
8
Explanations: grey cells indicate non-responders.
Figure 1 Scatter plots of serum activities per vaccine
0.0
0.5
1.0
1.5
2.0
2.5
3.0
BRP Vac A Vac B Vac C Vac D
A
c
t
i
v
i
t
y

(
l
o
g

I
U
/
m
l
)
Laboratory 1
0.0
0.5
1.0
1.5
2.0
2.5
3.0
A
c
t
i
v
i
t
y

(
l
o
g

I
U
/
m
l
)
Laboratory 2
0.0
0.5
1.0
1.5
2.0
2.5
3.0
A
c
t
i
v
i
t
y

(
l
o
g

I
U
/
m
l
)
Laboratory 3
0.0
0.5
1.0
1.5
2.0
2.5
3.0
A
c
t
i
v
i
t
y

(
l
o
g

I
U
/
m
l
)
Laboratory 4
0.0
0.5
1.0
1.5
2.0
2.5
3.0
A
c
t
i
v
i
t
y

(
l
o
g

I
U
/
m
l
)
Laboratory 5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
A
c
t
i
v
i
t
y

(
l
o
g

I
U
/
m
l
)
Laboratory 6
0.0
0.5
1.0
1.5
2.0
2.5
3.0
A
c
t
i
v
i
t
y

(
l
o
g

I
U
/
m
l
)
Laboratory 7
0.0
0.5
1.0
1.5
2.0
2.5
3.0
A
c
t
i
v
i
t
y

(
l
o
g

I
U
/
m
l
)
Laboratory 8
0.0
0.5
1.0
1.5
2.0
2.5
3.0
A
c
t
i
v
i
t
y

(
l
o
g

I
U
/
m
l
)
Laboratory 9
0.0
0.5
1.0
1.5
2.0
2.5
3.0
A
c
t
i
v
i
t
y

(
l
o
g

I
U
/
m
l
)
Laboratory 10
0.0
0.5
1.0
1.5
2.0
2.5
3.0
A
c
t
i
v
i
t
y

(
l
o
g

I
U
/
m
l
)
Laboratory 11
0.0
0.5
1.0
1.5
2.0
2.5
3.0
A
c
t
i
v
i
t
y

(
l
o
g

I
U
/
m
l
)
Laboratory 12
0.0
0.5
1.0
1.5
2.0
2.5
3.0
A
c
t
i
v
i
t
y

(
l
o
g

I
U
/
m
l
)
Laboratory 13
Explanations: each plot shows the activities of the sera per vaccine of the pooled assays
(i.e. 18 mice per vaccine). The fat horizontal bars are the median activity.
Table 2 Estimated activities of the control sera (log IU/mL)
Titre Mean Titre Mean Titre Mean
1 0.41 0.79 1.28
2 0.55 0.86 1.13
3 0.58 0.82 1.23
1 0.36 1.00 >q.l.
2 0.38 1.23 2.04
3 0.29 1.20 1.86
1 0.45 0.60 1.52
2 0.40 0.66 1.11
3 0.33 0.62 1.14
1 0.44 0.91 1.47
2 0.53 1.03 1.34
3 0.45 1.06 1.61
1 0.55 0.92 1.67
2 0.62 0.80 1.35
3 0.41 0.74 0.93
1a 0.61 1.37 1.81
1b 0.45 1.13 1.86
2 0.41 1.00 1.66
3 0.51 1.07 1.46
1 0.43 0.77 1.28
2 0.54 0.71 >q.l.
3 0.47 0.57 >q.l.
1 0.66 0.91 1.33
2 0.60 0.81 1.08
3 0.62 0.80 1.28
1 0.20 0.70 1.47
2 0.18 0.84 >q.l.
3 0.24 1.01 1.55
10 1 0.44 0.44 0.65 0.65 1.11 1.11
1 0.26 0.57 1.27
2 0.23 0.38 1.32
3 0.25 0.73 1.16
1 0.80 1.00 1.35
2 0.50 0.85 1.20
3 0.64 0.82 1.37
13 1 0.34 0.34 0.95 0.95 1.52 1.52
0.44 0.85 1.39
0.13 0.18 0.23
Explanations: >q.l. = above quantification limit. These results are not included
in the mean.
0.53
0.48
1
Lab
Standard deviation
Mean
12
11
9
8
High
0.39
1.14
0.82
1.00 4
3
2
5
0.50
Assay
0.63
1.14
0.83 0.52
Low Medium
0.35 1.95
1.32
1.47
1.26
1.70
1.21
0.63
0.48
1.31
1.25
1.51
1.23
1.28 7
6
0.89
0.56
0.85
0.84
0.68
0.65
0.25
0.21
F
i
g
u
r
e

2

M
e
a
n

a
c
t
i
v
i
t
i
e
s

o
f

t
h
e

c
o
n
t
r
o
l

s
e
r
a

p
e
r

l
a
b
o
r
a
t
o
r
y
E
x
p
l
a
n
a
t
i
o
n
s
:

t
h
e

d
a
s
h
e
d

l
i
n
e
s

i
n
d
i
c
a
t
e

t
h
e

o
v
e
r
a
l
l

m
e
a
n

o
f

t
h
e

1
3

l
a
b
o
r
a
t
o
r
i
e
s
.

T
h
e

g
r
e
y

a
r
e
a
s

c
o
v
e
r

1

s
t
a
n
d
a
r
d

d
e
v
i
a
t
i
o
n

f
r
o
m

t
h
e

o
v
e
r
a
l
l

m
e
a
n
.
1
.
3
0
2
.
0
0
2
.
3
0
c t | v | t y ( | o g I U ] m | )
S
e
r
u
m

c
o
n
L
r
o
l

"
P
l
g
h
"
S
e
r
u
m

c
o
n
L
r
o
l

"
M
e
d
l
u
m
"
0
.
0
0
0
.
3
0
1
.
0
0
1
2
3
4
3
6
7
8
9
1
0
1
1
1
2
1
3
A c
L
a
b
o
r
a
t
o
r
y
S
e
r
u
m

c
o
n
L
r
o
l

"
L
o
w
"
Table 3a P-values of the one-sided exact test of Wilcoxon-Mann-Whitney per assay
Explanations: grey cells indicate vaccines that fail the release test. Cells with a value of >0.500 indicate that the mean response of
the test was lower than the mean response of the standard.
vac A vac 8 vac C vac u vac A vac 8 vac C vac u
1 0.062 0.012 0.476 0.083 0.117 0.024 0.476 0.083
2 0.029 0.047 0.330 0.030 0.036 0.089 0.333 0.061
3 0.019 0.001 >0.300 0.004 0.039 0.002 >0.300 0.009
1 0.001 0.004 0.227 0.002 0.001 0.004 0.010 0.002
2 0.001 0.004 0.004 0.001 0.001 0.004 0.004 0.001
3 0.001 0.001 0.064 0.001 0.001 0.001 0.009 0.001
4 0.001 0.004 0.111 0.002 0.001 0.004 0.024 0.002
1 0.001 0.001 >0.300 0.001 0.001 0.001 >0.300 0.001
2 0.017 0.001 >0.300 0.001 0.002 0.001 >0.300 0.001
3 0.030 0.001 >0.300 0.003 0.002 0.001 >0.300 0.003
1 0.019 0.042 >0.300 0.184 0.019 0.042 >0.300 0.036
2 0.002 0.001 0.064 0.001 0.004 0.002 0.121 0.002
3 0.003 0.063 0.242 0.001 0.003 0.063 0.242 0.001
1 0.040 0.008 >0.300 0.294 0.004 0.008 >0.300 0.294
2 0.001 0.001 >0.300 0.361 0.002 0.002 >0.300 0.317
3 0.130 0.031 >0.300 0.073 0.229 0.061 >0.300 0.134
1a 0.001 0.001 0.469 0.001 0.001 0.001 0.469 0.001
1b 0.001 0.002 0.242 0.001 0.002 0.004 0.396 0.002
2 0.001 0.001 0.120 0.001 0.001 0.001 0.120 0.001
3 0.003 0.111 >0.300 0.003 0.003 0.111 0.309 0.003
1 0.023 0.023 >0.300 0.004 0.004 0.004 >0.300 0.009
2 0.001 0.002 >0.300 0.021 0.002 0.004 >0.300 0.041
3 0.023 0.023 0.228 0.041 0.004 0.004 >0.300 0.080
1 0.001 0.001 0.391 0.001 0.001 0.001 0.391 0.001
2 0.001 0.001 0.277 0.013 0.001 0.001 0.277 0.013
3 0.024 0.036 >0.300 0.430 0.024 0.036 >0.300 0.430
1 0.001 0.038 0.133 0.029 0.001 0.003 0.133 0.002
2 0.028 0.002 0.194 0.001 0.002 0.002 0.038 0.001
3 0.001 0.001 0.133 0.006 0.002 0.002 0.268 0.013
10 1 0.001 0.003 >0.300 0.001 0.002 0.006 >0.300 0.002
1 0.002 0.001 >0.300 >0.300 0.002 0.001 >0.300 0.032
2 0.001 0.001 >0.300 0.032 0.001 0.001 >0.300 0.032
3 0.001 0.004 >0.300 0.002 0.001 0.004 >0.300 0.002
1 0.018 0.001 >0.300 0.128 0.000 0.000 >0.030 0.001
2 0.421 0.029 >0.300 0.110 0.018 0.001 >0.300 0.128
3 0.093 0.003 >0.300 0.033 0.013 0.003 >0.300 0.033
1 0.013 0.011 0.081 0.001 0.013 0.011 0.081 0.001
2 0.001 0.009 0.330 0.001 0.001 0.009 0.330 0.001
3 0.001 0.003 0.236 0.003 0.001 0.003 0.100 0.003
8

lncludlng all mlce 8emovlng non-responders
7
6
3
4
3
2
1
Lab Assay
13
12
11
9
Table 3b P-values of the one-sided exact test of Wilcoxon-Mann-Whitney of pooled assays
vac A vac 8 vac C vac u vac A vac 8 vac C vac u
0.000 0.000 >0.300 0.000 0.001 0.000 >0.300 0.001
0.000 0.000 0.003 0.000 0.000 0.000 0.000 0.000
0.000 0.000 >0.300 0.000 0.000 0.000 >0.300 0.000
0.000 0.000 0.126 0.000 0.000 0.000 0.184 0.000
0.001 0.000 >0.300 0.132 0.000 0.000 >0.300 0.198
0.000 0.000 0.062 0.000 0.000 0.000 0.096 0.000
0.000 0.000 >0.300 0.000 0.000 0.000 >0.300 0.001
0.000 0.000 >0.300 0.000 0.000 0.000 >0.300 0.000
0.000 0.000 0.079 0.000 0.000 0.000 0.062 0.000
0.000 0.000 >0.300 0.014 0.000 0.000 >0.300 0.001
0.014 0.000 >0.300 0.006 0.002 0.000 >0.300 0.011
0.000 0.000 0.036 0.000 0.000 0.000 0.023 0.000 13
12
11
9
8
2
1
lncludlng all mlce 8emovlng non-responders
Lab
7
6
3
4
3
Table 4 Estimated probabilities to pass the vaccines using different numbers of mice
Explanations: the percentages give the estimated probability that a vaccine would pass the release test in a given laboratory using
N mice per preparation. Each percentage is based on 100.000 bootstrap resamples of size N with replacement from the pool of sera
per laboratory and vaccine, ignoring inter-assay variation..
Lab Vaccine N=3 N=4 N=5 N=6 N=7 N=8 N=9 N=10 N=11 N=12
A 42% 44% 62% 69% 76% 82% 86% 89% 92% 94%
B 54% 55% 76% 81% 86% 92% 94% 96% 97% 98%
C 4% 3% 4% 4% 4% 4% 4% 4% 4% 4%
D 37% 43% 61% 69% 77% 83% 87% 90% 92% 94%
A 100% 100% 100% 100% 100% 100% 100% 100% 100% 100%
B 71% 72% 91% 94% 97% 99% 99% 100% 100% 100%
C 29% 26% 30% 40% 46% 50% 53% 55% 61% 64%
D 79% 85% 98% 99% 100% 100% 100% 100% 100% 100%
A 70% 63% 56% 86% 82% 94% 93% 91% 97% 96%
B 100% 100% 100% 100% 100% 100% 100% 100% 100% 100%
C 0% 0% 0% 0% 0% 0% 0% 0% 0% 0%
D 78% 86% 97% 99% 100% 100% 100% 100% 100% 100%
A 68% 77% 92% 96% 98% 99% 100% 100% 100% 100%
B 53% 55% 71% 80% 87% 91% 93% 95% 97% 98%
C 11% 9% 14% 15% 17% 19% 20% 21% 24% 24%
D 65% 60% 63% 84% 83% 91% 93% 94% 97% 97%
A 40% 37% 46% 58% 61% 70% 73% 78% 81% 84%
B 65% 66% 80% 88% 92% 96% 97% 98% 99% 99%
C 0% 0% 0% 0% 0% 0% 0% 0% 0% 0%
D 11% 9% 12% 14% 15% 17% 18% 20% 20% 21%
A 80% 89% 98% 99% 100% 100% 100% 100% 100% 100%
B 52% 56% 81% 85% 91% 95% 96% 98% 99% 99%
C 13% 11% 16% 18% 20% 23% 23% 25% 28% 29%
D 96% 100% 100% 100% 100% 100% 100% 100% 100% 100%
A 66% 61% 62% 85% 83% 93% 93% 94% 97% 97%
B 58% 48% 49% 74% 69% 86% 83% 85% 91% 90%
C 2% 1% 1% 1% 1% 1% 1% 1% 0% 0%
D 41% 45% 66% 73% 80% 87% 89% 92% 94% 96%
A 68% 69% 83% 91% 94% 97% 98% 99% 99% 100%
B 65% 67% 86% 91% 95% 97% 98% 99% 99% 100%
C 4% 3% 3% 3% 3% 3% 3% 3% 3% 2%
D 35% 31% 42% 50% 56% 63% 66% 71% 75% 78%
A 80% 78% 76% 95% 94% 98% 99% 99% 100% 100%
B 65% 61% 57% 84% 82% 91% 93% 91% 97% 96%
C 13% 11% 17% 19% 21% 24% 26% 28% 30% 32%
D 67% 63% 70% 86% 87% 94% 95% 96% 98% 98%
A 100% 100% 100% 100% 100% 100% 100% 100% 100% 100%
B 70% 82% 94% 96% 98% 99% 100% 100% 100% 100%
C 1% 0% 0% 0% 0% 0% 0% 0% 0% 0%
D 100% 100% 100% 100% 100% 100% 100% 100% 100% 100%
A 70% 77% 94% 96% 98% 99% 100% 100% 100% 100%
B 88% 94% 99% 100% 100% 100% 100% 100% 100% 100%
C 0% 0% 0% 0% 0% 0% 0% 0% 0% 0%
D 26% 23% 27% 34% 39% 45% 46% 50% 54% 57%
A 19% 21% 27% 33% 38% 43% 47% 51% 54% 57%
B 71% 82% 95% 98% 99% 100% 100% 100% 100% 100%
C 0% 0% 0% 0% 0% 0% 0% 0% 0% 0%
D 21% 23% 34% 40% 46% 52% 56% 61% 64% 68%
A 76% 83% 96% 98% 99% 100% 100% 100% 100% 100%
B 48% 64% 81% 89% 94% 96% 98% 99% 99% 100%
C 15% 14% 19% 23% 25% 29% 31% 33% 35% 37%
D 92% 97% 100% 100% 100% 100% 100% 100% 100% 100%
5
6
7
8
9
10
12
1
2
3
4
11
13
Appendix 1

Appendix 2
Appendix 3
Follow-up of BSP105
In the study report of BSP105 a suggestion is made to impose a requirement on the maximum allowed number of non-
responders in the group of mice immunised with the standard vaccine (BRP), for example not more than 1 in a group of
6or not more than 2 in a group of 8.
The overall percentage of non-responders observed in this study ranges from 2.3% for the high potency vaccine B to 7.2%
for the sub-potent vaccine C. For the BRP this percentage is 5.9%. Assuming that these percentages represent the normal
expected frequency of non-responders, it is possible to calculate the probability that a given number of non-responders
occurs in a group of size n.
Table 1 shows the probability of m or more non-responders in a group of n mice, assuming an expected frequency of 6%
non-responders (the cumulative binomial distribution with p=0.06). For example, in a group of 10 mice we may expect 2
or more non-responders in 11.8% of the assays. An assay should therefore not be declared invalid when 2 non-responders
occur because this will occur rather frequently. However, in only 1.9% of the assays we may expect 3 or more non-
responders in a group of 10, so that could be a reasonable criterion to declare the assay invalid.
If we set the cut-off for the probability level at 5% we see that the example given in the study report is conrmed by this
table: not more than 1 non-responder in a group of 6 mice and not more than 2 non-responders in a group of 7 to 12 mice.
Tables for other values of p are given on the next page for information only.
Table 1 Expected frequencies of at least m non-responders in a group of n mice (p=0.06)
p=0.06
Total number of mice tested (n)
n=6 n=7 n=8 n=9 n=10 n=11 n=12
m0 100% 100% 100% 100% 100% 100% 100%
m1 31.0% 35.2% 39.0% 42.7% 46.1% 49.4% 52.4%
m2 4.6% 6.2% 7.9% 9.8% 11.8% 13.8% 16.0%
m3 0.4% 0.6% 1.0% 1.4% 1.9% 2.5% 3.2%
m4 0.0% 0.0% 0.1% 0.1% 0.2% 0.3% 0.4%
m5 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
m6 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
m7 - 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
m8 - - 0.0% 0.0% 0.0% 0.0% 0.0%
m9 - - - 0.0% 0.0% 0.0% 0.0%
m10 - - - - 0.0% 0.0% 0.0%
m11 - - - - - 0.0% 0.0%
m12 - - - - - - 0.0%
Proposed criterion for maximum allowed number of non-responders in the group of animals
immunised with the standard vaccine (BRP)
P
r
o
b
a
b
i
l
i
t
y

o
f

m

o
r

m
o
r
e

n
o
n
-
r
e
s
p
o
n
d
e
r
s
p=0.03
n=6 n=7 n=8 n=9 n=10 n=11 n=12
m0 100% 100% 100% 100% 100% 100% 100%
m1 16.7% 19.2% 21.6% 24.0% 26.3% 28.5% 30.6%
m2 1.2% 1.7% 2.2% 2.8% 3.5% 4.1% 4.9%
m3 0.1% 0.1% 0.1% 0.2% 0.3% 0.4% 0.5%
m4 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
m5 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
m6 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
m7 - 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
m8 - - 0.0% 0.0% 0.0% 0.0% 0.0%
m9 - - - 0.0% 0.0% 0.0% 0.0%
m10 - - - - 0.0% 0.0% 0.0%
m11 - - - - - 0.0% 0.0%
m12 - - - - - - 0.0%
p=0.08
n=6 n=7 n=8 n=9 n=10 n=11 n=12
m0 100% 100% 100% 100% 100% 100% 100%
m1 39.4% 44.2% 48.7% 52.8% 56.6% 60.0% 63.2%
m2 7.7% 10.3% 13.0% 15.8% 18.8% 21.8% 24.9%
m3 0.9% 1.4% 2.1% 3.0% 4.0% 5.2% 6.5%
m4 0.1% 0.1% 0.2% 0.4% 0.6% 0.9% 1.2%
m5 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.2%
m6 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
m7 - 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
m8 - - 0.0% 0.0% 0.0% 0.0% 0.0%
m9 - - - 0.0% 0.0% 0.0% 0.0%
m10 - - - - 0.0% 0.0% 0.0%
m11 - - - - - 0.0% 0.0%
m12 - - - - - - 0.0%
P
r
o
b
a
b
i
l
i
t
y

o
f

m

o
r

m
o
r
e

n
o
n
-
r
e
s
p
o
n
d
e
r
s
P
r
o
b
a
b
i
l
i
t
y

o
f

m

o
r

m
o
r
e

n
o
n
-
r
e
s
p
o
n
d
e
r
s
Establishment of Second International Standard for Gramicidin
Collaborative Study for the Establishment of the Second
International Standard for Gramicidin
G. Rautmann, A. Daas, K.-H. Buchheit
ABSTRACT
An international collaborative study was organised by the European Directorate for the Quality of Medicines
& HealthCare (EDQM) to establish the WHO second International Standard (IS) for gramicidin as the stocks of the 1
st
IS,
established in 1964, were close to depletion. The candidate material did not show any sign of potency loss when kept at
elevated temperatures of + 4 C, + 20 C, + 37 C and + 45 C for 3 months. Six laboratories from 5 countries as well
as the EDQM laboratory participated in the collaborative study. Potencies of the candidate material were estimated by
microbiological assays with sensitive micro-organisms. To ensure continuity between consecutive batches, the 1
st
IS for
gramicidin was used as standard.
Based on the results of the study, the 2
nd
IS for gramicidin was adopted at the meeting of the WHO Expert Committee on
Biological Standardization (ECBS) in 2008 with an assigned potency of 1070 International Units per mg (IU/mg). The
2
nd
IS for gramicidin is available from the EDQM.
KEYWORDS
Gramicidin, WHO International Standard, collaborative study, microbiological assay.
G. Rautmann. (corresponding authors e-mail: guy.rautmann@edqm.eu). European Directorate for the Quality of Medicines & HealthCare (EDQM),
Council of Europe, 7 Alle Kastner, F-67081 Strasbourg, France
A. Daas. European Directorate for the Quality of Medicines & HealthCare (EDQM), Council of Europe, Strasbourg, France
K.-H.Buchheit. European Directorate for the Quality of Medicines & HealthCare (EDQM), Council of Europe, Strasbourg, France
1. INTRODUCTION
Gramicidin is a heterogeneous mixture of linear polypeptides
isolated from the fermentation broth of Brevibacillus brevis
Dubos. The polypeptides exhibit -helix structures which
dimerize to span lipid bilayer thus creating ion channels.
The resulting increased bacterial cell wall permeability to
small inorganic ions is the basis of the antimicrobial activity.
The 1
st
International Standard (IS) for gramicidin was
established by the World Health Organization (WHO) in
1964 on the basis of an international collaborative study [1].
It was assigned with a potency of 1000 International Units
per mg (IU/mg), each ampoule containing approximately
55mg.
As stocks of the 1
st
IS were becoming exhausted, the Euro-
pean Directorate for the Quality of Medicines & HealthCare
(EDQM), was requested by the WHO Expert Committee on
Biological Standardization (ECBS) to undertake appropriate
steps for its replacement by the establishment of a new
batch.
2. PARTICIPANTS
A total of 6 laboratories from 5 different countries
including Argentina, France, Norway, Portugal, Thailand
and the laboratory of the EDQM participated in the study.
A list of participants in alphabetical order by surname
is given in Section 9. Each participant is referred to
in this report by an arbitrarily assigned number, not
necessarily reflecting the order of listing in the Appendix.
3.1. Bulk material, processing and stability
Candidate bulk material was kindly donated by Alpharma
ApS, Copenhagen (Denmark). Several separate bottles
each containing 0.5 kg of gramicidin of the same batch of
current pharmaceutical grade appropriate for therapeutic
use were received by the EDQM in April 2006. Upon receipt,
the bulk material was stored at + 4C before processing.
The candidate material was claimed by the manufacturer
to comply with the quality standards of the relevant
monographs of both the European Pharmacopoeia (Ph. Eur.)
and the United States Pharmacopeia (USP). A certicate of
analysis was provided in the batch documentation.
3.2. Filling
All powder weighing and filling were carried out in glove
boxes under a controlled atmosphere by use of argon gas.
The powder of one bottle was allowed to equilibrate at room
temperature and subsequently submitted to homogenisation
in a Turbula mixer. Twelve weighings of 30 g were
distributed in separate containers which were subsequently
sealed, protected from light and stored at 20C. The filling
campaign was organised over 5 consecutive days from
5to 9 November 2007. Prior to any further processing of
the gramicidin powder, containers were stored overnight
unopened under the glove box to enable room temperature
equilibration.
3.3. Production of a suitable reference
standard for monitoring purposes
IS are primary reference materials and as such cannot
be tested against higher order reference standards. As
a consequence, real time stability studies are not usual
practice and in many cases, stability of IS was assessed only
by means of accelerated degradation studies.
In the documentation supplied with the batch, stability
was confirmed for a period of 36 months. Over this period,
no decrease in microbiological activities was reported.
Nevertheless, it was decided to store some of the powder
at 80C and to use it, at regular intervals in the future, to
assess the potency of vials stored at 20C, the customary
storage temperature of the IS batch for gramicidin.
Consequently, it was decided to start the filling process by
weighing 100 mg in white glass ampoules from the first
30g container. A total of 220 ampoules were prepared. They
were immediately sealed by fusion and tested for leaks by
methylen blue uptake. The entire batch was stored at 80C
and is identified under the EDQM internal number 32665.
3.4. Production of the 2
nd
WHO IS for gramicidin
candidate batch
The lling was carried out in the EDQM facilities. Fillings
were organised in morning and afternoon sessions and
traced by sub-batch numbering. A total of 2020 vials were
lled with 100 mg of gramicidin, closed with an inert rubber
stopper and sealed. All vials were stored at 20C.
3.5. Quality control on bulk and nal batch
3.5.1. Conformity of the bulk
Prior to starting the lling process, 20 g were sampled
from the bulk. This material was submitted to physico-
chemical analysis according to the Ph. Eur. monograph
Gramicidin(0907), to conrm compliance. The results
obtained using the analytical methods described under
Ultraviolet and Visible Absorption Spectrophotometry, Thin
Layer Chromatography, Composition, Related Substances,
Loss on Drying and Sulfated Ash were in good agreement
with those of the certicate of analysis provided by the
manufacturer. The bulk was therefore considered suitable
for further processing.
3.5.2. Homogeneity of powder llings
Filling was carried out to a nominal content of 100mg
with a balance connected to a recording device. Mean
lling weights and relative standard deviations (RSD) were
calculated for each session. No signicant variability was
observed between sessions and the overall mean lling
weight was 100.78mg (RSD 0.11%). These results supported
the assumption that there was no heterogeneity in weight
between vials as a result of lling sessions fractionated into
half days.
3.5.3. Sampling design
During each half day filling session, 3 filled vials were
sampled at the beginning, middle and end of the session.
Each vial was individually tagged for unambiguous
identification and stored at 20C until subsequent analysis.
3.5.4. Quality control
From the above pool of samples, the 3 vials collected at the
middle of each filling session were tested for their water
content by the loss on drying (LOD) method, as described in
the Ph. Eur. chapter 2.2.32. Loss on drying.
Three assays were carried out (1 to 3). Samples from different
half day filling sessions (Monday morning to Friday afternoon)
are numbered 1 to 10. Each assay included 1sample from
each session. To eliminate as much as possible the effect of the
order of weighing, the samples were tested using a rotating
design. The order of weighing was 3, 10, 7, 4, 1, 8, 5, 2, 9, 6
in the 1
st
assay and the same order (but rotated) was used for
assays 2 and 3, starting with session 5 and 4 respectively. A
summary of results is given in Table 1.
Table 1 Water content of the 2
nd
IS by the LOD method
(in %)
Session Assay 1 Assay 2 Assay 3 Mean SD
1 0.573 0.414 0.761 0.583 0.174
2 0.643 0.621 0.764 0.676 0.077
3 0.869 0.515 0.755 0.713 0.181
4 0.757 0.816 0.974 0.849 0.112
5 0.585 0.663 0.952 0.733 0.193
6 0.777 0.605 0.825 0.736 0.116
7 0.792 0.714 0.808 0.771 0.050
8 0.504 0.577 0.712 0.598 0.106
9 0.794 0.694 0.887 0.792 0.097
10 0.746 0.668 0.842 0.752 0.087
Mean 0.704 0.629 0.828 0.720
SD 0.119 0.111 0.087 0.083
An analysis of variance was carried out in which the difference
between assays (1 to 3) and between sessions (1to10) were
modeled as classes and the order of weighing was modeled as
a linear effect. The results are given in Table2.
Table 2 Analysis of variance
Source DF SS Mean Square F-value Probability
Order of
weighing
1 0.01269200 0.01269200 1.96 0.1796
Session 9 0.18214013 0.02023779 3.12 0.0207
Assay 2 0.20255527 0.10127763 15.63 0.0001
DF: Degrees of Freedom, SS: Sum of Squares
The order in which the samples were tested does not have
a signicant effect. The largest variation is explained by
systematic differences between assays. The largest difference
is observed between assay 2 and assay 3 where the mean
LOD was 0.629% and 0.828% respectively. There also
appears to be a systematic difference between samples from
different sessions, but this effect is less important than the
difference between assays. The largest difference is found
between samples from session 1 (Monday morning) and
session 4 (Tuesday afternoon) with mean losses of 0.583%
and 0.849% respectively. This difference is practically
irrelevant so it can be considered that the batch is
sufciently homogeneous for its intended purpose.
3.6. Environmental control
During the entire lling process, the glove box was
continuously monitored for oxygen content and hygrometry.
Measurements were carried out at the beginning and end of
each half day session. Values recorded did not signicantly
deviate from settings.
3.7. Stability studies on the product in the nal container
An accelerated degradation study was carried out at the
EDQM by storing vials of the candidate batch of the 2
nd
IS for
gramicidin at + 20C, + 37C and + 45C in several climatic
chambers (Binder, KBF 720 model).
3.7.1. Accelerated Degradation Assessed by
Microbiological Assay
The potency of the candidate IS in these vials was estimated
relative to the material kept in ampoules at 20C. Two vials
were analysed by 2 independent assays for each temperature.
The data are presented in Annex 1 in both tabular and
graphic format after 1, 3 and 6 months of storage. In
addition, a vial stored at 80C was also used to estimate the
potency of vials stored at 20C to generate some baseline
data for future monitoring purposes. No decrease in potency
of the vials stored at 20C was observed in this latter test.
Assuming that the expected recovery should be 100%
in the absence of any degradation, all these values were
within the 6% acceptance criterion set to account for the
variability of the analytical method based on the long history
of monitoring data collected at the EDQM for gramicidin.
However samples stored for a longer period at each of the
elevated temperatures appeared to exhibit a systematic
lower potency (compare curve in Annex 1 at 3 and 6 months
respectively).
Extrapolation by means of a model based on the Arrhenius
equation [2] shows that no predictable loss of potency (less
than 0.1%) is expected over a period of 10 years when stored
at 20 C. From these data it is anticipated that the stability
of the 2
nd
IS for gramicidin is satisfactory.
3.7.2. Accelerated Degradation Assessed by Liquid
Chromatography
EDQM has a long record of experience in monitoring
the stability of ofcial Ph. Eur. reference standards
for antibiotics. Due to the inherent variability of the
microbiological assay methods, it was decided some years
ago to replace them by stability indicating methods such as
reversed-phase high performance liquid chromatography
(LC) for monitoring the stability of the Ph. Eur. reference
standards. It was therefore considered to be of benet to
estimate the degradation at elevated temperature by LC in
addition to microbiological assays with the aim of collecting
data for future replacement of the method.
One vial of each of the 3 elevated storage temperatures was
analysed using the liquid chromatography analytical method
described under Test. Related substances of the Ph. Eur.
monograph Gramicidin (0907).
Individual peaks were identied on each chromatogram and
their contents expressed as the mean areas in per cent by
normalisation calculated from triplicate injections for each
vial. The data are presented in Annex 2 in both tabular and
graphic format after 1, 3 and 6 months of storage.
After 1, 3 or 6 months of storage, no dramatic modications
were observed in the impurity proles of samples stored
at any of the elevated temperatures when compared to the
chromatogram recorded with the vials stored at 20C.
Individual peak area changes were considered within method
variability and thus not signicant. Furthermore, the
impurity prole of the vial stored at 80C was comparable
to the impurity proles recorded for samples stored for
6months at any elevated temperature.
These results are in agreement with the results generated by
the microbiological assay method and conrm the absence
of any signicant degradation resulting in potency loss when
vials of the proposed 2
nd
IS for gramicidin are stored at an
elevated temperature for up to 6 months, considering the
variability of the methods respectively.
3.7.3 Conclusion of the stability studies
Vials of the proposed 2
nd
IS for gramicidin were submitted
to an accelerated degradation study to predict the stability
at the customary storage temperature of 20C. The results
obtained with 2 orthogonal analytical methods demonstrated
that the vials did not exhibit any reduction in potency or any
change in the impurity prole. It is therefore concluded that
the stability of the batch at 20C is satisfactory.
Considering that the precision of the liquid chromatography
method is much better than the precision of the
microbiological assay, it is believed that with respect to the
variability of these methods, any signicant change in the
impurity prole will be detected ahead of any signicant
loss of potency. It is therefore decided to monitor in the
future the stability of the WHO 2
nd
IS for gramicidin on an
annual basis by means of liquid chromatography and to
assess the impact of any signicant modication (decrease
in percentage of the principal peak / increase in the level of
impurity or appearance of the new impurity peak) on the
potency by the microbiological assay.
The vials of the proposed 2
nd
IS for gramicidin are stored
at the EDQM at 20 C. Users of the 2
nd
IS for gramicidin
are advised that the material should be stored at 20C.
Once dissolved, the solution should be used within 1 day.
Solutions should always be made fresh and never stored
frozen prior to use.
4. Collaborative study
4.1. Samples
Each laboratory was provided with:
3 vials of WHO 1
st
IS for gramicidin (64/10), 1000 IU/mg
containing approximately 55 mg of powder per ampoule
(EDQM internal code: 28117)
7 vials of candidate WHO 2
nd
IS for gramicidin batch
containing approximately 100 mg of powder per vial
(activity about 1000 IU/mg) (EDQM internal code: 32786)
4.2. Assay method and study design
The participants were asked to estimate the potency of
the gramicidin candidate batch by a microbiological
turbidimetric assay method using the WHO 1
st
IS for
gramicidin (64/10) as reference preparation. It was requested
that any analytical method used be in compliance with
requirements set in regional compendia in particular with
respect to method validity criteria. A total of 6independent
assays were to be carried out by each participant.
Prior to carrying out the study an enquiry was carried out
which demonstrated that participants were going to use very
similar testing procedures. Based on this enquiry, a pilot
assay was performed in the EDQM laboratory in order to
develop and provide details for the study protocol, taking the
Ph. Eur. as the example.
Participating laboratories were requested to follow the study
protocol as far as possible and according to the prescription
given in the Pharmacopoeia which is their usual reference.
5. RESULTS
5.1. Statistical methods
The experimental data obtained in this study were analysed
as parallel line assays [3], using the SAS-System [4] (GLM
procedure) and CombiStats [5]. Both programmes give
identical outcomes, but the output is somewhat easierto
transform to tables with the SAS-System, whereas
CombiStats provides a more streamlined output for
individual assays.
All assays were submitted to visual inspection of the plots
to check for unusual features. Validity of the assays was
assessed according to the ow chart in Figure 1.
In routine situations where decisions are based only on
one assay or only on a few assays, the level of signicance
is usually taken to be P=0.05. In collaborative studies with
many participants, however, a more conservative level of
signicance is often used. This is because the level of P=0.05
leads to about 10 per cent errors of the rst kind (incorrect
rejection of assays), whereas errors of the second kind
(incorrect acceptance of assays) will not inuence the global
outcome of the study much because of the large amount of
data available. Hence, the level of signicance in this study
is taken to be P=0.01 which would imply an expectation of
about 2 per cent incorrect rejections. A slight but signicant
curvature was not considered reason for rejection if the
mean square for quadratic regression was less than 1/100
of the mean square for linear regression and the difference
between preparations was small [6, 7].
Whenever a laboratory performed several assays based on the
same weighings, yielding several non-independent estimates
of potency, a weighted mean potency of the valid sub-assays
was calculated using weights proportional to the reciprocal
of the variance. Only the valid assays per laboratory were
combined using the same method of weighted combination,
but a semi-weighted combination was used whenever the
condence intervals of the independent potency estimates
did not satisfactorily overlap each other by means of a
2
test
for homogeneity (P<0.10). The estimates (one for each of the
participants) were then combined into one single estimate
with a 95 per cent condence interval using the same
method of semi-weighted combination.
5.2. Results
Six laboratories reported results from assays. In this report
they are referred to by their randomly assigned code-
numbers (1 to 6), not necessarily corresponding to the
order of listing in the list of participants. Four laboratories
used the Ph. Eur. method and 2 used another method.
All participants carried out at least 6 assays as requested.
Laboratory 5 carried out 7 assays but reported having
problems obtaining valid assays. A total of 33 valid assays
were reported or 1559 turbidity readings.
The complete computer output of the parallel line analyses
as performed at EDQM was made available in PDF format
to participants of the study (59 pages generated by
CombiStats). A summary of the results, as generated by
the SAS-System is given in Table 3. The potency estimates
and associated 95 per cent condence intervals are shown,
together with the relevant P-values of the analysis of
variance. P-values below the signicance level of 0.01
are printed on a grey background. The potency estimates
and condence intervals based on calculations by the
participants are also listed.
A graphical representation of the potencies and condence
intervals of each individual assay is shown in Figure2
(EDQM calculations) and in Figure 3 (Participants
calculations). Potency estimates ranged from 965 IU/mg
(Lab 1) to 1252 IU/mg (Lab 6), based on EDQM calculations.
Laboratory 1
The 6 assays were statistically valid and the potency
estimates were homogeneous (P=0.155). The weighted
combined estimate is 1011 IU/mg ( 2.5%).
Laboratory 2
estimates were heterogeneous (P<0.001). The semi-weighted
Laboratory 3
This laboratory used 5 doses for each assay, but only the
3 middle doses were used in the calculation. The 6 assays
were statistically valid and the potency estimates were
heterogeneous (P=0.039). The semi-weighted combined
estimate is 1059 IU/mg ( 1.1%).
Laboratory 4
Laboratory 5
This laboratory reported having problems obtaining
statistically valid assays. A total of 7 assays were performed
but only 3 were judged valid by the participant and therefore
they initially reported only data from these 3 assays.
However, upon request from the EDQM, data from the other
4 assays were also submitted. Three of the 7 assays were
found to be statistically valid at EDQM. This laboratory used
5 doses for each valid assay, but only the 3 middle doses
were used in the calculation. The potency estimates were
heterogeneous (P<0.001). The semi-weighted combined
estimate is 1044 IU/mg ( 3.1%).
Laboratory 6
combined estimate is 1102 IU/mg (2.9%).
A histogram of all potency estimates per assay is shown in
Figure 4 and a histogram of the mean results per laboratory
is shown in Figure 5. The nal potency estimates and
condence intervals per laboratory are summarised in
Table4 and a graphical representation is given in Figure 6.
The
2
value for between-laboratory variation is highly-
signicant (P<0.001) so a semi-weighted combination was
made, which yields 1068 IU/mg ( 1.7%).
Table 4 Combined potency estimates per laboratory

Final potency estimates (IU/mg)
Lab 95% Lower conf.
limit
Estimated po-
tency
95% Upper
conf. limit
1 987 1011 1036
2 1119 1142 1166
3 1047 1059 1070
4 1033 1053 1072
5 1013 1044 1077
6 1071 1102 1133
Comb. 1050 1068 1086
6. COMMENTS FROM PARTICIPANTS
None of the participants opposed to the proposal to assign a
potency of 1070 IU/mg to the 2
nd
IS for gramicidin.
1
The 2
nd
IS for Gramicidin is available from the EDQM (catalogue number ISA_28168)
7. RECOMMENDATION
It was proposed to the ECBS that the Second WHO IS for
Gramicidin be assigned an antimicrobiological activity of
1070 IU per milligram of substance. The potency of this
batch was estimated by the participants of the collaborative
study using the 1
st
WHO IS for gramicidin as the reference
standard in the microbiological assay. Therefore, the
continuity of the International Unit for gramicidin
originating with the establishment of the 1
st
WHO for
gramicidin, is maintained. The 2
nd
IS for gramicidin was
adopted at the WHO ECBS meeting in October 2008
1
.
8. ACKNOWLEDGEMENTS
The organisers express their sincere thanks to all
participants for their valuable contributions to this study.
Special thanks go to the donator of the gramicidin drug
substance, Alpharma ApS, Copenhagen (Denmark), for
their greatly appreciated contribution. Sally Woodward is
acknowledged for skilful assistance. The study was organised
by the EDQM (project code ISA004) as the custodian
laboratory for the WHO.
9. LIST OF PARTICIPANTS (listed in alphabetical order
by surname)
N. BELIXAN, Instituto Nacional Medicamentos,
Administracin Nacional de medicamentos alimentos y
tecnologa medica, Buenos Aires - Argentina
Y. CORTEZ, Agence Franaise de Scurit Sanitaire des
Produits de Sant, Vendargues - France
L.M. MEIRINHOS SOARES, INFARMED Autoridade
Nacional do Medicamento e Produtos de Sade, I. P.,
Lisbon- Portugal
G. RAUTMANN, C. RAPHALEN & J. SANDRIN, European
Directorate for the Quality of Medicines & HealthCare -
Council of Europe
N. RUANGRITTINON, Bureau of Drug and Narcotic,
Nonthaburi - Thailand
R. SANDIK, Alpharma AS, Oslo - Norway
10. ABBREVIATIONS
ECBS: Expert Committee for Biological Standardization
EDQM: European Directorate for the Quality of Medicines &
HealthCare
IS: International Standard
IU: International Unit
LOD: Loss On Drying
Ph. Eur.: European Pharmacopoeia
LC: reversed-phase High Performance Liquid
Chromatography
RSD: Relative Standard Deviation
SD: Standard Deviation
USP: United States Pharmacopeia
WHO: World Health Organization
11. REFERENCES
[1] Technical Report Series 293, 17th report. WHO Expert
Committee on Biological Standardisation; 1964 Sept-Oct.
[2] Kirkwood, TB. Predicting the stability of biological
standards and products. Biometrics 1977;33(4):736-742.
[3] Finney DJ. Statistical Method in Biological Assay. 3rd
edition. London, UK: Charles Grifn & Co. Ltd.; 1978.
[4] SAS Institute Inc., SAS OnlineDoc, Version 8, Cary,
NC, USA: SAS Institute Inc.; 1999.
[5] CombiStats v4.0. Strasbourg, France: EDQM, Council
of Europe [available at: www.combistats.eu].
[6] Bliss CI. The calculation of microbial assays, Bacteriol
Rev. 1956;20(4):243-58.
[7] Hewitt W. Inuence of curvature of response lines
in antibiotic agar diffusion assays, J Biol Stand.
1981;9(1):1-13.
Lab Assay
Calculated by participants
(IU/mg)
Calculated at EDQM (IU/
mg)
p-values Analysis of variance
Quad.
Regr. /
Lin.
95%
Lower
conf.
limit
Esti-
mated
potency
95%
Upper
conf.
limit
95%
Lower
conf.
limit
Esti-
mated
potency
95%
Upper
conf.
limit
Non-
paralle-
lism
Non-
linea-
rity
Qua-
dratic
regres-
sion
Lack of
quadra-
tic t
1
1 967 1013 1062 967 1013 1062 0.521 0.087 0.047 0.318 0.002
2 944 1033 1131 944 1033 1131 0.648 0.523 0.349 0.522 0.001
3 916 965 1016 916 965 1016 0.694 0.169 0.097 0.366 0.001
4 1008 1083 1164 1008 1083 1164 0.740 0.445 0.219 0.774 0.001
5 931 999 1071 931 999 1071 0.894 0.071 0.024 0.702 0.005
6 959 1032 1110 959 1032 1110 0.825 0.108 0.045 0.524 0.004
2
1 1048 1086 1125 1048 1086 1125 0.630 0.293 0.157 0.511 0.001
2 1110 1172 1240 1110 1172 1240 0.370 0.487 0.334 0.483 0.001
3 1044 1082 1122 1044 1082 1122 0.484 0.159 0.084 0.404 0.002
4 1101 1162 1228 1101 1162 1228 0.386 0.356 0.177 0.646 0.003
5 1158 1196 1238 1158 1196 1238 0.161 0.527 0.280 0.765 0.001
6 1140 1167 1194 1140 1167 1194 0.338 0.224 0.088 0.926 0.001
3
1 1039 1064 1090 1039 1064 1090 0.847 0.136 0.048 0.911 0.001
2 1058 1077 1097 1058 1077 1097 0.051 0.172 0.063 0.970 0.001
3 1021 1044 1067 1021 1044 1067 0.330 0.970 0.808 0.978 0.000
4 1009 1036 1063 1009 1036 1063 0.154 0.914 0.766 0.766 0.000
5 1054 1077 1101 1049 1077 1106 0.679 0.438 0.203 0.962 0.001
6 1016 1044 1072 1013 1043 1075 0.824 0.933 0.830 0.764 0.000
4
1 1009 1055 1105 1009 1055 1105 0.686 0.000 0.000 0.864 0.010
2 989 1042 1097 989 1042 1097 0.113 0.072 0.023 0.848 0.003
3 991 1062 1139 991 1062 1139 0.136 0.616 0.445 0.540 0.001
4 1000 1047 1096 1000 1047 1096 0.960 0.004 0.001 0.421 0.005
5 1021 1060 1101 1021 1060 1101 0.901 0.132 0.139 0.165 0.001
6 1004 1049 1097 1004 1049 1097 0.394 0.000 0.000 0.316 0.009
5
1 1072 1107 1144 1053 1091 1133 0.131 0.691 0.808 0.414 0.000
2 1021 1059 1100 1007 1050 1097 0.028 0.242 0.369 0.154 0.002
3 1028 1112 1214 1028 1112 1214 0.416 0.508 0.251 0.932 0.011
4 999 1011 1024 999 1011 1024 0.613 0.005 0.001 0.830 0.002
5 1051 1069 1088 1051 1069 1088 0.658 0.001 0.008 0.002 0.003
6 1046 1059 1073 1046 1059 1073 0.007 0.000 0.001 0.001 0.003
7 1036 1063 1090 1036 1063 1090 0.212 0.000 0.000 0.734 0.042
6
1 1011 1041 1071 1011 1041 1071 0.388 0.530 0.299 0.681 0.000
2 1057 1101 1147 1057 1101 1147 0.310 0.391 0.176 0.982 0.001
3 1036 1061 1087 1036 1061 1087 0.055 0.286 0.420 0.173 0.000
4 1118 1143 1168 1118 1143 1168 0.118 0.526 0.391 0.464 0.000
5 1144 1252 1374 1144 1252 1374 0.117 0.136 0.054 0.641 0.006
6 1043 1148 1266 1043 1148 1266 0.268 0.112 0.039 0.902 0.008

p-values below the signicance level of 0.01 are printed on a grey background.

Table 3 Overview of assay results generated by the SAS System
Figure 1 Flow chart for assay validity check
Figure 2 Individual potency estimates and 95% condence intervals per assay (calculated at EDQM)
The numbers below the 95 % condence intervals are the laboratory codes. Invalid assays are shown with an empty dot.
Figure 3 Individual potency estimates and 95% condence intervals per assay (calculated by participants)
The numbers below the 95 % condence intervals are the laboratory codes. Invalid assays are shown with an empty dot.
Figure 4 Histogram of nal potency estimates per vial
Figure 5 Histogram of nal potency estimates per laboratory
Numbers in the boxes are the laboratory codes
Figure 6 Potency estimates and 95% condence intervals per laboratory (Combined assays)
Annex 1: Accelerated Degradation, Microbiological Assay Results
Relative Potency in per cent versus samples stored at -20C
Storage temperature +20C +37C +45C
1 month
Vial 1 100.7 102.9 105.7
CI 95% -1.9 +1.9 -2.0 +2.1 -2.0 +2.0
Vial 2 103.6 105.2 103.5
CI 95% -2.0 +2.1 -1.7 +1.7 -1.5 +1.5
Mean 102.1 104.1 104.5
CI 95% +2.4 +2.5 -2.0 +2.1 -1.9 +1.9
3 months
Vial 1 102.2 105.0 102.8
CI 95% -2.8 +2.9 -2.3 +2.3 -2.4 +2.5
Vial 2 104.7 100.7 100.0
CI 95% -1.7 +1.7 -2.3 +2.3 -2.4 +2.5
Mean 104.0 102.8 101.4
CI 95% -1.4 +1.5 -3.3 +3.4 -2.6 +2.6
6 months
Vial 1 103.4 95.7 96.3
CI 95% -2.0 +2.1 -4.3 +4.5 -5.1 +5.4
Vial 2 102.6 103.8 97.8
CI 95% -2.4 +2.5 -4.3 +4.5 -5.1 +5.4
Mean 103.0 99.2 96.7
CI 95% -1.5 +1.5 -6.0 +6.3 -2.5 +2.6
Mean: geometric mean
CI: condence intervalle P=0.95
80
85
90
95
100
105
110
+20C +37C +45C
P
e
r

C
e
n
t
Storage Temperature
Relative Potency
1 month
3 months
6 months
P
e
a
k

1
P
e
a
k

2
P
e
a
k

3
i
m
p
u
r
i
t
y

D
P
e
a
k

5
g
r
a
m
i
c
i
d
i
n

C
1
i
m
p
u
r
i
t
y

A
g
r
a
m
i
c
i
d
i
n

C
2
i
m
p
u
r
i
t
y

B
g
r
a
m
i
c
i
d
i
n

A
2
g
r
a
m
i
c
i
d
i
n

B
2
P
e
a
k

1
3
A
f
t
e
r

1

m
o
n
t
h
-
2
0
C
0
.
0
1
0
.
0
3
0
.
1
9
0
.
1
1
0
.
2
1
9
.
3
0
.
7
7
2
.
9
7
0
.
3
7
8
.
5
3
2
.
8
7
0
.
1
4
+
2
0
C
0
.
0
1
0
.
0
3
0
.
1
9
0
.
1
1
0
.
2
9
.
2
9
0
.
7
6
2
.
9
2
0
.
3
7
8
.
5
2
2
.
8
8
0
.
2
6
+
3
7
C
0
.
0
1
0
.
0
3
0
.
2
1
0
.
1
2
0
.
2
2
9
.
3
0
.
7
5
2
.
8
5
0
.
3
6
8
.
5
4
2
.
8
6
0
.
1
4
+
4
5
C
0
.
0
1
0
.
0
3
0
.
2
2
0
.
1
3
0
.
2
2
9
.
3
0
.
7
5
2
.
8
5
0
.
3
6
8
.
5
6
2
.
8
3
0
.
0
7
A
f
t
e
r

3

m
o
n
t
h
-
2
0
C
0
.
0
1
0
.
0
3
0
.
1
9
0
.
0
9
0
.
1
2
9
.
2
1
2
.
8
7
0
.
8
1
7
.
1
5
2
.
9
7
0
.
1
8
+
2
0
C
0
.
0
1
0
.
0
3
0
.
1
8
0
.
0
9
0
.
1
2
9
.
1
8
0
.
9
9
2
.
8
8
0
.
9
1
7
.
1
6
2
.
8
4
0
.
1
8
+
3
7
C
0
.
0
1
0
.
0
3
0
.
1
9
0
.
0
9
0
.
1
2
9
.
1
8
0
.
9
7
2
.
9
4
0
.
9
2
7
.
0
8
2
.
8
5
0
.
1
4
+
4
5
C
0
.
0
1
0
.
0
3
0
.
1
8
0
.
1
0
.
1
1
9
.
1
7
0
.
9
7
2
.
9
5
0
.
8
7
7
.
0
5
2
.
9
9
0
.
1
9
A
f
t
e
r

6

m
o
n
t
h
-
8
0
C
0
.
0
1
0
.
0
3
0
.
1
3
0
.
0
7
0
.
1
5
8
.
4
0
.
9
6
2
.
7
9
1
.
1
6
9
.
5
5
2
.
6
0
.
2
4
-
2
0
C
0
.
0
1
0
.
0
4
0
.
1
9
0
.
0
6
0
.
1
5
8
.
1
5
1
.
2
2
3
.
1
6
1
.
0
9
9
.
9
3
2
.
7
1
0
.
2
4
+
2
0
C
0
.
0
1
0
.
0
2
0
.
2
1
0
.
0
8
0
.
1
5
8
.
1
3
1
.
1
3
3
.
1
3
1
.
4
1
0
.
7
2
2
.
5
9
0
.
2
2
+
3
7
C
0
.
0
1
0
.
0
2
0
.
2
1
0
.
0
8
0
.
1
4
8
.
1
9
1
.
0
8
2
.
9
5
1
.
5
9
1
0
.
2
9
2
.
5
2
0
.
2
5
+
4
5
C
0
.
0
1
0
.
0
3
0
.
1
9
0
.
0
7
0
.
1
4
8
.
3
1
.
0
2
2
.
7
7
1
.
2
4
9
.
8
2
2
.
6
3
0
.
2
4
Annex 2: Accelerated Degradation, Liquid Chromatography Results Mean Impurity Peak Areas in Per Cent
I
m
p
u
r
i
t
y

P
r
o

l
e
s

a
f
t
e
r

1

M
o
n
t
h
I
m
p
u
r
i
t
y

P
r
o

l
e
s

a
f
t
e
r

3

M
o
n
t
h
s
I
m
p
u
r
i
t
y

P
r
o

l
e
s

a
f
t
e
r

6

M
o
n
t
h
s
Establishment of Third International Standard for Nystatin
Collaborative Study for the Establishment of the
Third International Standard for Nystatin
G. Rautmann, E. Charton, A. Daas, K.-H. Buchheit
ABSTRACT
Due to the depletion in stocks of the World Health Organization (WHO) 2
nd
International Standard (IS) for nystatin, an
international collaborative study was organised by the European Directorate for the Quality of Medicines & HealthCare
(EDQM) to establish a replacement batch. Seventeen laboratories participated in the collaborative study, performing the
microbiological diffusion assay to estimate the potency of the candidate 3
rd
International Standard for nystatin. The 2
nd

International Standard for nystatin was used as a standard to ensure the continuity of the unitage.
Follow-up accelerated degradation studies demonstrated that the IS is stable when at the customary storage temperature
of 20 C.
The 3
rd
IS for nystatin was adopted by the WHO Expert Committee on Biological Standardization (ECBS) in 2006 with an
assigned potency of 5710 International Units per mg (IU/mg). The 3
rd
IS for nystatin is available from the EDQM.
KEYWORDS
Nystatin, microbiological assay, WHO International Standard, collaborative study.
G. Rautmann. (Corresponding author: e-mail: guy.rautmann@edqm.eu). European Directorate for the Quality of Medicines & HealthCare, Council
of Europe, 7 Alle Kastner, CS0036, 67081 Strasbourg, France.
E. Charton. European Directorate for the Quality of Medicines & HealthCare, Council of Europe, 7 Alle Kastner, CS0036, 67081 Strasbourg,
France.
A. Daas. European Directorate for the Quality of Medicines & HealthCare, Council of Europe, 7 Alle Kastner, CS0036, 67081 Strasbourg, France.
K.-H. Buchheit. European Directorate for the Quality of Medicines & HealthCare, Council of Europe, 7 Alle Kastner, CS0036, 67081 Strasbourg,
France.
1. INTRODUCTION
Nystatin is obtained from Streptomyces noursei. A mixture
of antifungal polyenes, it is often used intravenously in
humans for the treatment of systemic fungal infections.
The 1
st
International Standard (IS) for nystatin was
established by the World Health Organization (WHO) in
1963 on the basis of an international collaborative study [1].
It was assigned with a potency of 3000 International Units
per mg (IU/mg). A 2
nd
IS for nystatin was established in 1982
with an assigned potency of 4855 IU/mg [2].
As supplies of the 2
nd
IS were becoming exhausted, the
European Directorate for the Quality of Medicines & Health-
Care (EDQM) was requested by the WHO Expert Committee
on Biological Standardization (ECBS) to undertake
appropriate steps for its replacement by the establishment of
a new batch.
2. PARTICIPANTS
A total of 17 laboratories from 16 different countries around
the world volunteered to participate in the study. Each
participant is referred to hereafter by an arbitrarily assigned
number, not necessarily reflecting the order of listing in
section 10.
3.1. Bulk material and processing
Candidate bulk material was kindly donated by DSM Capua
S.p.A. (Italy). Two separate bottles each containing 1 kg of
nystatin of the same batch of current pharmaceutical grade
appropriate for therapeutical use were received by the EDQM
in December 2005. Upon receipt and prior to processing,
the bulk material was stored at + 4 C. The candidate
material was claimed by the manufacturer to comply with
the quality standards of the relevant monographs of both the
European Pharmacopoeia (Ph. Eur.) and the United States
Pharmacopeia (USP). A certificate of analysis was provided in
the batch documentation.
3.2. Filling
All powder weighing and filling were carried out in glove
boxes under a controlled atmosphere by use of argon gas.
The powder of one bottle was allowed to equilibrate at
room temperature and was submitted to extensive mixing
in a Turbula mixer. Ten weighings of 50 g were distributed
in separate containers subsequently sealed, protected
from light and stored at 20 C. The filling campaign was
organised over 5 consecutive days from 27-31 March 2006.
Prior to any further processing of the nystatin powder, the
containers were stored overnight unopened under a glove
box to enable room temperature equilibration.
3.3. Production of a suitable reference standard for
monitoring purposes
WHO IS are primary reference materials and as such
cannot be tested against higher order reference standards.
As a consequence, real-time stability studies are not usual
practice and in many cases stability of WHO IS are assessed
by means of accelerated degradation studies.
In the documentation supplied with the batch, stability was
confirmed over a period of 36 months. Data indicated no
sign of decrease for parameters including among others,
microbiological activities and impurity profiles by liquid
chromatography after an incubation at 25 2 C over the
3-year period.
It appeared judicious however to store some of the powder
at 80 C and to use it, at regular intervals in the future,
to assess the potency of vials stored at 20 C, the storage
temperature of the WHO IS for nystatin. Consequently it was
decided to start the filling process by weighing 100 mg in
white glass ampoules from the first 50 g container. A total of
190 ampoules were prepared. They were immediately sealed
by fusion and tested for leaks by methylene blue uptake. The
entire batch was stored at 80 C and was identified under
the number 06/03-103.
3.4. Production of the WHO IS for nystatin candidate
batch 3
To increase throughput, two gloves boxes were used for
the filling of the vials. The filling sessions were organised
into morning and afternoon sessions, with clear sub-batch
numbering to allow traceability of both the glove box used
and the filling session concerned. A total of 3819 vials were
filled with 100 mg of nystatin, closed with an inert rubber
stopper and sealed. All vials were stored at 20 C.
3.5. Quality control on bulk and final batch
3.5.1. Conformity of bulk
During the first weighing session on the first morning, 2 lots
of 5 g of nystatin powder were set aside to enable subsequent
analysis according to the Ph. Eur. monograph [3], to
confirm the compliance declared by the manufacturer. The
results obtained by use of the analytical methods described
under Absorbance, Composition, Heavy metals, Loss on
drying and Sulfated ash were in good agreement with those
of the certificate of analysis provided by the manufacturer.
The bulk was therefore considered suitable for further
processing.
3.5.2. Homogeneity of powder fillings
Filling was carried out to a nominal content of 100 mg
with a balance connected to a recording device. Mean
filling weights and relative standard deviations (RSD)
were calculated for each session and for each glove box.
No significant variability was observed between sessions
and the overall mean filling weights were 103.17 mg (RSD
1.75%) and 104.26 mg (RSD 2.33 %) for the 2 gloves boxes
respectively. These results supported the assumption that
there was no heterogeneity in weight between vials as a
result of the use of 2 glove boxes.
3.5.3. Sampling design
During each half-day filling session and for each glove
box, 3 filled vials were sampled at the beginning,
middle and end of the session. Each vial was
individually tagged for unambiguous identification
and stored at 20 C until subsequent analysis.
3.5.4. Quality control
From the above pool of samples, the 3 vials collected at the
middle of each filling session were tested for their water
content by the European Pharmacopoeia loss on drying
(LOD) method (2.2.32) [4]. Samples were assayed in a design
that would allow elimination of a possible trend in the order
of weighing and the inter-assay variation, while at the same
time enabling the measurement of these effects, should there
be any. In short this means that the samples were handled
in chronological order in half of the assays and in anti-
chronological order in the other half and that samples from
each day were included in each assay.
An analysis of variance was carried out in which 4 effects were
modelled:
difference between day of assay (5 classes);
difference between glove boxes (2 classes);
trend of preparation (10 sessions ordered on a
chronological scale);
trend of weighing (10 weighings ordered on a
chronological scale).
The calculations were performed with the SAS-System[6].
The analysis of variance (see Table 1 below) shows a highly
significant difference between assays (p<0.0001) and a
significant trend in the order of weighings (p<0.02). No
trend is observed for the order of sampling (p=0.70) and no
significant difference between the glove boxes can be observed
(p=0.10).
In conclusion, there is no significant indication for
heterogeneity of the fillings between days and glove boxes.
Two randomly chosen vials of the filling campaign are not
likely to differ by more than 0.13% from each other
(
ts 2 2.020 0.00205 1.414 0.13
).
The results therefore demonstrated that the 2 batches were
to be considered homogeneous. The mean results per group
of samples are presented in Table 2.
The estimated water content is 4.09 % with 95 % confidence
limits of 4.04 % and 4.14 % where the mean results per
assay are used to estimate the uncertainty (n=6). These
results are not in disagreement with the figure of 4.3 %
stated on the certificate of analysis.
Table 1 Analysis of variance
Source Degrees of
Freedom
Sum of Squares Mean Square F-value Probability
Assay 5 0.0910 0.0182 8.87 <0.0001
Glove box 1 0.0057 0.0057 2.79 0.1022
Session 1 0.0003 0.0003 0.15 0.7011
Weighing 1 0.0142 0.0142 6.93 0.0118
Residual error 42 0.0862 0.00205
Total 50 0.2016
3.6. Environmental control
During the entire filling campaign, the glove boxes were
continuously monitored for oxygen content and hygrometry.
Measurements were carried out at the beginning and end of
each half-day session. Values recorded did not significantly
deviate from settings.
4. COLLABORATIVE STUDY
4.1. Samples
Each participating laboratory received:
7 vials of Nystatin candidate standard (EDQM code:
27979) containing approximately 100 mg of powder per
vial (activity about 6000 IU/mg);
3 vials of Nystatin 2
nd
International Standard
(80/508), 4855 IU/mg (EDQM code: 28389) containing
approximately 100 mg of powder per ampoule.
4.2. Assay method and study design
The participants were asked to estimate the potency of the
nystatin candidate IS by a microbiological diffusion assay
method using the WHO 2
nd
IS for nystatin (80/508) as
reference preparation. It was requested that any analytical
method used be in compliance with requirements set in
regional compendia in particular with respect to method
validity criteria. A total of 6 independent assays were to be
carried out by each participant.
Prior to the study, an enquiry was carried out which
demonstrated that participants were going to use very similar
testing procedures (see Annex 1). All but one participant used
saccharomyces cerevisiae. All participants used one of the
pharmacopoeias prescribed media (medium F/ medium19).
Based on this enquiry, a pilot assay was performed in the
EDQM laboratory in order to elaborate the study protocol,
taking the European Pharmacopoeia as the example.
Participating laboratories were requested to follow the study
protocol as far as possible and according to the prescription
given in the pharmacopoeia which is their usual reference.
5. RESULTS AND STATISTICAL ANALYSIS
5.1. Statistical methods
The experimental data obtained in this study were analysed
as parallel-line assays [5], using the SAS-System (GLM
procedure) [6] and CombiStats [7]. Both programmes
give identical outcomes but the output is somewhat easier
to transform into tables with the SAS-system, whereas
CombiStats provides a more streamlined output for
individual assays.
All assays were submitted to visual inspection of the plots
to check for unusual features. Validity of the assays was
assessed according to the flow chart in Figure1. In routine
situations where decisions are based on only 1 assay or only
a few assays, the level of significance is usually taken to be
P=0.05. In collaborative studies with many participants,
however, a more conservative level of significance is often
used. This is because the level of P=0.05 leads to about
10 per cent errors of the first kind (incorrect rejection
of assays), whereas errors of the second kind (incorrect
acceptance of assays) will not influence the global outcome
of the study much because of the large amount of data
available. Hence, the level of significance in this study is
taken to be P=0.01 which would imply an expectation of
about 2 per cent incorrect rejections. A slight but significant
curvature was not considered reason for rejection if the
mean square for quadratic regression was less than 1/100
of the mean square for linear regression and the difference
between preparations was small [8, 9].
Whenever a laboratory performed several assays based on the
same weighings, yielding several non-independent estimates
of potency, a weighted mean potency of the valid sub-assays
was calculated using weights proportional to the reciprocal
of the variance. The valid assays per laboratory were
combined using the same method of weighted combination,
but a semi-weighted combination was used whenever the
confidence intervals of the independent potency estimates
did not satisfactorily overlap each other by means of an
2

test for homogeneity (P<0.10). The estimates (one for each
of the participants) were then combined into one single
estimate with a 95 per cent confidence interval using the
same method of semi-weighted combination.
5.2. Results
Seventeen laboratories reported results from assays. In
this report they are referred to by their randomly assigned
code-numbers (1 to 17), not necessarily corresponding
with the order of listing in the list of participants. Thirteen
laboratories used the European Pharmacopoeia method,
1 used the British Pharmacopoeia (= European), 1 used
the United States Pharmacopeia, 1 used the Chinese
Pharmacopoeia, and 1 used the Argentinean Pharmacopoeia.
Table 2 Mean result per group
Day
(part)
Mean n Assay Mean n Weighing Mean n
Glove
box
Mean n
Mo (AM) 4.053 3 1 4.100 8 1 4.102 6 A 4.112 21
Mo (PM) 4.080 3 2 4.014 8 2 4.145 6 B 4.081 30
Tu (AM) 4.063 3 3 4.086 9 3 4.110 6
Tu (PM) 4.098 6 4 4.083 9 4 4.085 6
We (AM) 4.123 6 5 4.151 7 5 4.082 6
We (PM) 4.082 6 6 4.131 10 6 4.108 6
Th (AM) 4.092 6 7 4.075 6
Th (PM) 4.088 6 8 4.060 5
Fr (AM) 4.118 6 9 4.043 3
Fr (PM) 4.100 6 10 4.130 1
Overall mean: 4.094 (n=51)
All participants carried out at least 6 assays as requested.
Laboratory 6 carried out 7 assays, one for each vial provided.
Laboratory 17 also tested all 7 vials, but performed 2 assays
on the first 2 vials and 1 assay for the remaining 5 vials.
Laboratory 4 carried out all assays by 2 operators, but only
the results of 2 assays from the 2
nd
operator were reported
because the other results were considered invalid by the
laboratory. The results from this 2
nd
operator are coded
as Laboratory 4b. Laboratories 4, 7, 9 and 12 carried out
multiple sub-assays per vial.
For the calculations, all sub-assays were analysed as
individual assays after which they were combined into 1
potency estimate per vial. If all sub-assays are counted as
individual assays a total of 192 assays were reported or 9032
zone-diameter readings.
A summary of the results, as generated by the SAS-System
is given in Table 3 (See Annex 2 for the essential SAS-scripts
used). The potency estimates and associated 95 per cent
confidence intervals are shown, together with the relevant
p-values. P-values below the significance level of 0.01 are
printed on a grey background. The confidence intervals
based on calculations by the participants are also listed.
Some laboratories reported the potency estimate and/or
confidence intervals in terms of percentage of the estimated
or assumed potency. In these cases the percentages were
converted to IU/mg at the EDQM. Laboratory 13 did
not report confidence intervals. Laboratory 14 reported
confidence intervals which were so discrepant from the
results found at EDQM that the laboratory was contacted
in order to find an explanation. After investigation the
participant concluded having made a mistake and that their
recalculations concurred with the EDQM calculations.
A graphical representation of the confidence intervals of
each individual (sub)-assay is shown in Figure 2 (EDQM
calculations) and in Figure 3 (participants calculations).
Potency estimates ranged from 5045 IU/mg (laboratory 4) to
6162 IU/mg (laboratory 13) disregarding the extremely high
results from laboratory 14 (up to 7656 IU/mg).
Laboratory 1
combined estimate is 5829 IU/mg ( 1.6 %).
Laboratory 2
Laboratory 3
Laboratory 4
This laboratory carried out 3 sub-assays per vial using one
Latin square plate per sub-assay, resulting in 18 potency
estimates. In addition this laboratory carried out assays
by a second operator who also performed 3 sub-assays per
vial, but only the results of 2 vials (i.e. 6 sub-assays) were
reported because the other results were considered invalid
by the laboratory. The results from this 2
nd
operator are
coded as Laboratory 4b. Of the assays from the first operator
6 plates gave invalid results, 3 of which were part of assay
6 and therefore yielding no valid potency estimate from
assay 6. Of the assays from the second operator 5 plates gave
invalid results, leaving only one statistically valid result.
Since this result is also of borderline quality with a p-value
for non-parallelism of 0.015, it may be argued whether it
would not be better to disregard all results from the second
operator and only include the valid results from the first
operator. If this is done, the 5 resulting confidence intervals
show significant heterogeneity (P<0.001), so a semi-
weighted combination was taken. This results in a potency
estimate of 5310 IU/mg ( 2.0 %).
Laboratory 5
Laboratory 6
This laboratory carried out 7 assays, 1 for each vial provided.
Laboratory 7
estimates. All sub-assays were statistically valid and could be
combined by taking the weighted mean potency, resulting
in 6 potency estimates which were homogeneous (P=0.794).
The weighted combined estimate is 5981 IU/mg ( 0.4 %).
Laboratory 8
The 6 assays were statistically valid but the potency
Laboratory 9
This laboratory used 6 Latin square plates per assay, but
reported only the results of those plates they considered
valid, resulting in a total of 32 sub-assays. This laboratory
used 2 dose levels per preparation with 2 independent
weighings per dose. It was therefore not possible to check
for parallelism in the individual assays, but the laboratory
reported that the linearity of the dose-response over the dose
range used had previously been established. None of the sub-
assays showed significant deviations from parallelism and
visual inspection of the plots showed that the assays were
of satisfactory quality. All assays were therefore retained
as being valid and were combined by taking the weighted
mean potency, resulting in 6 potency estimates which were
heterogeneous (P<0.001). The semi-weighted combined
estimate is 5299 IU/mg ( 0.7 %).
Laboratory 10
The 6 assays were statistically valid but had rather wide
confidence limits (about 10 % on average). Taken as
individual assays the results should be rejected because
of lack of precision. However, the potency estimates are
remarkably homogeneous (P=1.000) and visual inspection
of the plots did not reveal any features which could indicate
fundamental invalidity. The weighted combined estimate
is 5612 IU/mg ( 3.4 %). Since the combined potency
is sufficiently precise to be acceptable, the assays are
a-posteriori accepted as being valid despite their individual
lack of precision.
Laboratory 11
confidence limits (about 7.5 % on average). Taken as
individual assays the results should be rejected because of
lack of precision. However, visual inspection of the plots did
not reveal any features which could indicate fundamental
invalidity and the confidence intervals showed satisfactory
homogeneity (P=0.964). The weighted combined estimate
lack of precision.
Laboratory 12
estimates. One sub-assay was statistically invalid due to
significant deviations from parallelism. The valid sub-assays
could be combined by taking the weighted mean potency,
resulting in 6 potency estimates which were heterogeneous
(P=0.006). The semi-weighted combined potency estimate is
5937 IU/mg ( 0.5 %).
Laboratory 13
confidence limits (about 6.0 % on average). Taken as
individual assays the results should be rejected because of
lack of precision. However, visual inspection of the plots did
not reveal any features which could indicate fundamental
invalidity and the confidence intervals showed satisfactory
homogeneity (P=0.282). The weighted combined estimate
lack of precision.
Laboratory 14
One assay was statistically invalid due to significant deviations
from parallelism. Three other assays were statistically invalid
due to significant deviations from linearity. Overall, all
assays gave much higher responses for the test preparation
than for the standard. The potency was on average much
higher than expected (about 7450IU/mg). At first it was
thought that a dilution error might have been made, or
that a dilution step was overlooked in the calculations but
neither appeared to be the case. The participant suggested
that a probable cause could be that the temperature
fluctuations experienced during transit may have
compromised the integrity of either one of the materials.
It seemed therefore prudent to exclude the outlying results
from this laboratory from the final calculation for the
assignment of a potency value.
Laboratory 15
Laboratory 16
The laboratory initially submitted results from only 4 assays
because they judged assay 1 and assay 6 to be statistically
invalid. Upon request these 2 assays were also submitted.
Assay 6 was indeed found to be statistically invalid due to
significant deviations from parallelism and linearity, but
assay 1 was found to be acceptable whereas assay 4 had to
be declared invalid because of significant deviations from
linearity. The 4 valid potency estimates were heterogeneous
(P<0.001). The semi-weighted combined potency estimate is
5621 IU/mg ( 2.4 %).
Laboratory 17
This laboratory submitted results from 9 assays. All 7 vials
were tested. The first 2 vials were assayed twice and the
other 5 vials were assayed once. All assays were statistically
valid. The duplicate results for the first 2 vials were first
combined using a weighted combination. The resulting 7
potency estimates were homogeneous (P=0.501) and the
weighted combination was 5337 IU/mg ( 1.1 %).
A histogram of all potency estimates per assay is shown in
Figure 4 and a histogram of the mean results per laboratory
is shown in Figure 5. The final confidence intervals per
laboratory are summarised in Table 4 and a graphical
representation is given in Figure 6.
The
2
value for between-laboratory homogeneity is highly
significant (P<0.001) so a semi-weighted combination was
made excluding the result from laboratory 14, which yields
5710 IU/mg with 95 % confidence limits of 5673 to
5747 IU/mg (which is 0.6 %). If the result from laboratory
14 is included the outcome is 5780 IU/mg, but it is not
easy to give confidence limits for this figure because the
assumption of normality is violated.
For the purpose of comparison the general unweighted
geometric mean potency was calculated for the
104independent assay estimates giving 5775 IU/mg if
laboratory14 is included, and 5684 IU/mg with 95 %
confidence limits of 5557 to 5814 IU/mg (which is 2.3 %)
if laboratory 14 is excluded. If all statistically invalid results
are also excluded the unweighted geometric mean potency is
5686 IU/mg with 95 % confidence limits of 5560 to
5816 IU/mg (which is 2.3 %).
6. COMMENTS FROM PARTICIPANTS
A draft report was sent to the 17 study participants. Fourteen
of them did not oppose the conclusions and the proposed
assignment of 5710 IU per milligram of nystatin.
One participant commented that the overall mean should
be calculated using weights inversely proportional to the
variance based on the width of the confidence limits. This
would result in a potency estimate of 5831 IU/mg and
therefore this laboratory concludes that the proposal of
5710IU/mg is an underestimation.
The organisers replied that if confidence intervals between
laboratories do not sufficiently overlap, as is the case in this
study, the inter-laboratory variation should also be taken
into account to establish weights given to the potency
estimates. The weighted mean of 5831 IU/mg does not give
due consideration to the observed inter-laboratory variation
and gives disproportional weight to laboratories 2, 7, and
12. The semi-weighted mean of 5710 IU/mg is therefore
considered more appropriate.
Two participants returned editorial comments.
7. CONCLUSION
It was proposed to the WHO ECBS that the 3
rd
International
Standard for Nystatin (EDQM internal code ISA_29384)
be assigned an antimicrobiological activity of 5710 IU per
milligram of substance. The 3
rd
IS for nystatin was adopted
at the WHO ECBS meeting in 2006.
8. STABILITY STUDIES
During the discussions at the WHO ECBS meeting in 2006,
the experts were of the opinion that additional accelerated
degradation tests should be carried out to gather some
insight into the predicted stability of the batch in addition to
the proposed continuous assessment by real-time studies.
An accelerated degradation study was therefore carried
out at the EDQM by storing vials of the 3
rd
International
Standard for nystatin at 20 C, + 4 C, + 20 C, + 37 C and
+ 45 C in several climatic chambers. Potencies of stressed
vials were estimated by microbiological assays and changes
in impurity profiles by high pressure liquid chromatography.
8.1. Microbiological Assay
The remaining antibiotic potency of individual vials stored at
elevated temperature was estimated by assaying the growth
inhibition of the sensitive micro-organism Saccharomyces
cerevisiae (IP 1432-83). The diffusion method was used.
The potencies were estimated as the relative potencies
against ampoules kept at 80 C. Two vials were analysed
in 2 independent assays for each temperature. Data
was analysed by use of the parallel-line model to assess
compliance with validity criteria and calculate the potencies.
The data are presented in Table 5 and Figure 7 in tabular and
graphic format after 1 and 3 months of storage.
The results from the 3 month time point at 37 C failed to
meet the criteria for parallelism and linearity and were not
considered valid and thus not taken into account in the
assessment of the data. All other assays were statistically
valid. Assuming that the expected recovery should be 100%
in the absence of any degradation, all these values were
within the 6 % acceptance criterion set to account for the
variability of the analytical method based on the long history
of monitoring data collected at the EDQM. Moreover the
data points at the higher storage temperature of + 45 C did
not show any convincing evidence of degradation.
Extrapolation by means of a model based on the Arrhenius
equation [10] shows that a loss of potency of 2.5 % is
expected over a period of 10 years when stored at 20 C.
However, this extrapolation is quite uncertain due to the
inclusion of the result of 90.2 % at 37 C after 3 months.
Indeed, Grubbs test for outliers [11] indicates that this value
is to be treated with suspicion. If it is excluded from the
model, the expected loss of potency is 0.05 % after 10 years
at 20 C which is negligible. Because of the outlying value
it was decided to run a retest. This retest yielded a result of
100.2 % which is more in line with the other result obtained
at 37 C, but unfortunately this test had to be declared
invalid due to significant deviations from parallelism
(p-value = 0.026). It was noted that the residual error of
this test was rather low compared to that obtained in other
assays and that the mean square for non-parallelism was not
as high as obtained in one other valid assay. It was therefore
decided to calculate F-ratios with respect to a pooled
estimate of the residual error based on all assays which gave
a p-value of 0.075. In addition, visual inspection of a plot
of the assay data revealed no unusual features so it may be
justified to retain the result for inclusion in the Arrhenius
model. If this is done, the expected loss of potency is 0.03 %
after 10 years at 20 C.
From these results it is anticipated that the stability of the
3
rd
International Standard for nystatin is satisfactory.
8.2. Liquid Chromatography
In addition to microbiological assays, it was believed to be of
benefit to estimate the degradation at elevated temperature
by rp-HPLC.
One vial of each of the 4 storage temperatures was analysed
by triplicate injections. Individual peaks were identified on
each chromatogram and their contents expressed as the
mean areas by normalisation for each vial.
The data are presented in Table 6 and Figures 8 and 9 in
tabular and graphic format after 1 and 3 months of storage.
After 1 or 3 months of storage, no dramatic modifications
were observed in the impurity profiles of samples stored
at any of the elevated temperatures when compared to the
chromatogram recorded with the vials stored at 20 C.
Individual peak area changes were considered within method
variability and thus not significant.
These results are in good agreement with the results
generated by the microbiological assay method and confirm
the absence of any significant degradation resulting in
potency loss when vials of the 3
rd
International Standard for
nystatin are stored at elevated temperature for 3 months.
Considering that the precision of the liquid chromatography
method is higher than the microbiological assay, it is
believed that with respect to the variabilities of these
methods, any significant change in the impurity profile will
be detected ahead of any significant loss of potency. It was
therefore proposed and accepted by the ECBS to monitor
in the future the stability of the 3
rd
International Standard
for nystatin by means of liquid chromatography and to
assess the impact of any significant modification (decrease
in percentage of the principal peak / increase in the level
of impurity or appearance of new impurity peak) on the
potency by the microbiological assay.
8.3. Conclusion of the stability studies
The results obtained with 2 orthogonal analytical methods
demonstrated that the vials did not exhibit any reduction
in potency nor any change in the impurity profile when
submitted to an accelerated degradation. Extrapolation by
means of a model based on the Arrhenius equation indicated
an antimicrobiological activity loss of about 0.05 % over a
period of 10 years.
It is therefore concluded that the stability of the batch at the
customary storage temperature of 20 C is satisfactory.
As detailed in the leaflet provided with the IS, it is
recommended that the powder contained in the vial is
dissolved into the appropriate concentration and that the
resulting solution is used within 1 day. Solutions should
always be freshly made and never stored frozen prior to use.
The EDQM will continue to monitor the impurity profile of
the vials stored at 20 C by means of high pressure liquid
chromatography on an annual basis.
9. ACKNOWLEDGEMENTS
The organisers express their sincere thanks to all
participants for their valuable contributions to this study.
Special thanks go to the donator of the nystatin drug
substance, DSM CAPUA, Italy. The study was organised by
the EDQM (project code ISA002) as the WHO custodian
centre for antibiotics.
10. PARTICIPANTS (in alphabetical order of country)
Hireche A, Laboratoire National de Contrle des Produits
Pharmaceutiques, Hydra Alger, Algeria
Belixan N, Instituto Nacional Medicamentos, Buenos Aires,
Argentina
Hawke D, Therapeutic Goods Administration, Woden ACT,
Australia
Bittencourt C and Laporta L, Farmacopia Brasileira, Santa
Maria RS, Brazil
Shangchen Y, China National Institute for the Control of
Pharmaceutical and Biological Products, Beijing, China
Novotna P, Institute for State Control of Veterinary
Biologicals and Medicaments, Brno, Czech Republic
Overballe Petersen C, Alpharma, Copenhagen, Denmark
Charton E and Rautmann G, European Directorate for the
Quality of Medicines & HealthCare, Council of Europe,
France
Cortez Y, Agence franaise de scurit sanitaire des produits
de sant, Vendargues, France
Sensale M.R, DSM Food Specialities, Capua-CE, Italy
Moideen S.V.K, National Pharmaceutical Control Bureau,
Petaling jaya, Malaysia
Tyski S, Polish Institute of Public Health, Warsaw, Poland
Meirinhos Soares L.M, Autoridade Nacional do Medicamento
e Produtos de Sade, Lisbon, Portugal
Bystrova L, Scientific Centre for Evaluation of Medicinal
Products, Moscow, Russia
Mirkovic B, Medicines and Medical Devices Agency of Serbia,
Belgrade, Serbia
Ruangrittinon N, Bureau of Drug and Narcotic, Nonthaburi,
Thailand
Crivellone M, United States Pharmacopeia, Rockville, USA
11. ABBREVIATIONS
ECBS: Expert Committee on Biological Standardization;
EDQM: European Directorate for the Quality of
Medicines & HealthCare; IS: International Standard;
IU: International Units; LOD: Loss on Drying; Ph. Eur.:
European Pharmacopoeia; RSD: relative standard deviation;
USP: United States Pharmacopeia; WHO: World Health
Organization
12. REFERENCES
[1] Lightbown JW, Kogut M, Uemura K. The International
Standard for Nystatin. Bull World Health Organ
1963;29:87-94.
[2] Thomas AH, Dixon H. Ref: BS/82.1350. WHO Expert
Committee on Biological Standardization; 1982.
[3] Nystatin, monograph 0517. Ph. Eur. 7
th
Edition.
Strasbourg, France: Council of Europe; 2010(vol 2).
[4] Loss on drying, general chapter 2.2.32. Ph. Eur. 7
th

Edition. Strasbourg, France: Council of Europe; 2010.
[5] Finney DJ. Statistical Method in Biological Assay. 3rd
Edition. London, UK: Charles Griffin & Co. Ltd.;1978.
[6] SAS Institute Inc., SAS OnlineDoc, Version 8, Cary,
NC, USA: SAS Institute Inc.; 1999.
[7] CombiStats v3.1. Strasbourg, France: EDQM, Council
of Europe [available at: www.combistats.eu].
[8] Bliss CI. The calculation of microbial assays. Bacteriol
Rev. 1956;20(4):243-58.
[9] Hewitt W. Influence of curvature of response lines
in antibiotic agar diffusion assays. J Biol Stand.
1981;9(1):1-13.
[10] Kirkwood TB. Predicting the stability of
biological standards and products. Biometrics
1977;33(4):736-742.
[11] Grubbs FE. Procedures for detecting outlying
observations in samples. Technometrics 1969;
11(1):1-21.
Figure 1 Flow chart for assay validity check
Table 3 Overview of assay results
p-values < 0.01 are printed on a grey background
Lab Assay
95% Lower
conf. limit
Estimated
potency
95% upper
conf. limit
95% Lower
conf. limit
Estimated
potency
95% upper
conf. limit
Non-
parallelism
Non-
linearity
Lack of
quadratic fit
Quadratic
regression
Quad. regr./
Lin. regr.
1 5660 5930 6220 5659 5934 6222 0.496 0.696 0.753 0.433 0.001
2 5630 5880 6140 5628 5880 6142 0.278 0.255 0.105 0.812 0.000
3 5460 5720 6000 5456 5723 6001 0.875 0.311 0.322 0.243 0.002
4 5500 5770 6050 5499 5767 6045 0.197 0.400 1.000 0.180 0.002
5 5520 5800 6090 5519 5797 6087 0.325 0.581 0.463 0.463 0.001
6 5680 5840 6010 5681 5842 6007 0.085 0.340 0.909 0.147 0.001
1 5872 5943 6015 5872 5943 6015 0.097 0.400 0.923 0.179 0.000
2 5901 5957 6013 5901 5957 6013 0.657 0.936 0.797 0.797 0.000
3 5844 5916 5990 5844 5916 5990 0.933 0.257 0.663 0.113 0.000
4 5842 5930 6020 5842 5930 6020 0.459 0.063 0.785 0.020 0.001
5 5967 6026 6086 5967 6026 6086 0.289 0.785 0.623 0.623 0.000
6 5902 5967 6032 5902 5967 6032 0.152 0.595 0.698 0.347 0.000
1 5631 5890 6155 5632 5889 6157 0.254 0.180 0.218 0.163 0.001
2 5434 5663 5905 5432 5665 5906 0.053 0.202 0.075 0.962 0.000
3 5555 5814 6079 5553 5812 6082 0.910 0.115 0.425 0.054 0.002
4 5561 5785 6009 5562 5783 6012 0.796 0.180 0.486 0.087 0.001
5 5652 5797 5953 5650 5799 5951 0.694 0.838 0.649 0.705 0.000
6 5535 5727 5918 5535 5725 5920 0.958 0.713 0.414 0.976 0.000
1.1 5627 5748 5871 5578 5748 5922 0.105 0.738 0.521 0.668 0.000
1.2 5469 5547 5625 5381 5547 5715 0.015 0.380 0.170 0.914 0.000
1.3 5498 5611 5726 5440 5611 5786 0.646 0.188 0.091 0.492 0.000
2.1 5376 5459 5542 5295 5459 5625 0.007 0.012 0.279 0.005 0.005
2.2 5280 5350 5420 5235 5350 5465 0.000 0.001 0.463 0.000 0.005
2.3 5285 5366 5448 5199 5366 5535 0.078 0.012 0.751 0.003 0.005
3.1 5002 5093 5185 4945 5092 5241 0.485 0.015 0.954 0.004 0.004
3.2 5221 5316 5413 5179 5316 5454 0.180 0.009 0.076 0.008 0.003
3.3 4989 5112 5238 4877 5112 5350 0.845 0.648 0.625 0.433 0.001
4.1 5317 5417 5520 5296 5418 5541 0.168 0.070 0.030 0.418 0.000
4.2 5252 5371 5493 5225 5371 5519 0.460 0.014 0.008 0.168 0.001
4.3 5214 5323 5433 5114 5322 5535 0.198 0.037 0.130 0.032 0.004
5.1 5036 5124 5212 4969 5123 5278 0.350 0.163 0.515 0.075 0.002
5.2 4934 5022 5111 4890 5021 5154 0.908 0.367 0.740 0.173 0.001
5.3 5024 5103 5183 4961 5103 5246 0.522 0.377 0.272 0.390 0.000
6.1 4877 4951 5024 4818 4950 5084 0.365 0.002 0.076 0.002 0.005
6.2 4969 5045 5122 4884 5045 5207 0.519 0.004 0.059 0.004 0.005
6.3 5023 5083 5143 4973 5083 5194 0.347 0.001 0.023 0.002 0.003
1.1 5102 5189 5279 5024 5189 5356 0.497 0.000 0.434 0.000 0.015
1.2 5234 5332 5431 5180 5332 5486 0.779 0.000 0.155 0.000 0.013
1.3 5186 5279 5374 5161 5279 5398 0.051 0.000 0.071 0.000 0.01003
2.1 5098 5182 5267 5038 5182 5327 0.000 0.001 0.511 0.000 0.008
2.2 4847 4946 5046 4777 4945 5115 0.055 0.000 0.299 0.000 0.013
2.3 5052 5137 5223 4995 5137 5280 0.015 0.025 0.398 0.010 0.003
1 5499 5640 5781 5499 5639 5782 0.408 0.285 0.473 0.159 0.001
2 5480 5697 5924 5337 5613 5899 0.197 0.080 0.528 0.032 0.007
3 5643 5805 5974 5639 5806 5977 1.000 0.975 1.000 0.825 0.000
4 5443 5611 5785 5444 5614 5787 0.720 0.248 0.679 0.109 0.001
5 5518 5730 5948 5519 5730 5948 0.870 0.522 0.637 0.304 0.001
6 5505 5687 5874 5502 5684 5870 0.092 0.082 0.234 0.054 0.002
1 5277 5541 5818 5277 5542 5818 0.061 0.141 0.064 0.504 0.001
2 5362 5607 5863 5362 5608 5863 0.935 0.150 0.864 0.055 0.004
3 5322 5548 5785 5322 5550 5785 0.274 0.136 0.276 0.090 0.003
4 5429 5671 5923 5429 5672 5923 0.093 0.681 0.968 0.387 0.001
5 5133 5380 5639 5170 5422 5682 0.778 0.205 0.460 0.106 0.003
6 5589 5806 6031 5589 5807 6031 0.767 0.215 0.425 0.119 0.002
7 5369 5630 5903 5319 5576 5842 0.790 0.256 0.766 0.108 0.003
1.1 5754 5975 6196 5882 5978 6076 0.351 0.110 0.369 0.057 0.000
1.2 5765 5901 6037 5768 5904 6043 0.085 0.388 0.606 0.206 0.000
1.3 5837 5962 6087 5845 5978 6114 0.252 0.150 0.893 0.055 0.000
1.4 5713 5926 6139 5713 5929 6153 0.576 0.471 0.628 0.264 0.000
1.5 5751 6028 6305 5752 6031 6323 1.000 0.898 0.885 0.664 0.000
1.6 5836 6016 6196 5840 6019 6203 0.653 0.643 0.604 0.438 0.000
2.1 5716 5893 6070 5656 5840 6029 0.725 0.125 0.544 0.053 0.001
2.2 5781 5929 6077 5797 5927 6060 0.788 0.148 0.099 0.284 0.000
2.3 5863 6013 6163 5888 6037 6190 0.382 0.247 0.448 0.137 0.000
2.4 5706 5925 6144 5703 5923 6151 0.323 0.136 0.445 0.065 0.001
2.5 5767 5958 6149 5782 5970 6163 0.854 0.443 0.596 0.249 0.000
2.6 5851 6013 6175 5850 6011 6176 1.000 0.235 0.633 0.105 0.000
3.1 5822 6071 6320 5825 6069 6322 0.197 0.169 0.933 0.063 0.001
3.2 5912 6089 6266 5895 6071 6253 0.833 0.152 0.399 0.080 0.001
3.3 5737 5939 6141 5739 5951 6172 0.725 0.125 0.544 0.053 0.001
3.4 5791 5970 6149 5785 5967 6154 0.419 0.246 0.639 0.111 0.000
3.5 5825 6030 6235 5827 6028 6236 0.344 0.189 0.912 0.072 0.001
3.6 5879 6061 6243 5879 6058 6243 0.398 0.236 0.806 0.096 0.000
4.1 5798 5934 6070 5800 5940 6084 0.118 0.155 0.894 0.057 0.000
4.2 5768 5910 6052 5770 5916 6065 0.514 0.083 0.175 0.070 0.000
4.3 5806 5973 6140 5813 5979 6150 0.177 0.150 0.571 0.064 0.001
4.4 5921 6067 6213 5930 6073 6218 0.637 0.109 0.282 0.067 0.000
4.5 5795 5938 6081 5798 5944 6093 0.532 0.140 0.904 0.051 0.000
4.6 5813 5987 6161 5820 5993 6171 0.589 0.142 0.603 0.058 0.001
5.1 5788 5973 6158 5791 5976 6168 0.601 0.159 0.615 0.066 0.001
5.2 5740 5985 6230 5744 5988 6243 0.667 0.179 0.934 0.067 0.001
5.3 5722 5985 6248 5723 5988 6265 0.497 0.152 0.694 0.060 0.001
5.4 5750 5971 6192 5755 5974 6201 0.524 0.249 0.712 0.107 0.001
5.5 5744 5971 6198 5747 5974 6211 0.878 0.196 0.930 0.075 0.001
5.6 5768 5959 6150 5773 5962 6157 0.478 0.589 1.000 0.310 0.000
6.1 5888 6027 6166 5896 6034 6176 0.807 0.133 0.328 0.078 0.000
6.2 5948 6094 6240 5948 6088 6232 0.237 0.071 0.219 0.047 0.000
6.3 5807 5907 6007 5811 5914 6019 0.344 0.264 0.854 0.109 0.000
6.4 5781 5972 6163 5786 5980 6180 1.000 0.144 0.839 0.053 0.001
6.5 5771 5986 6201 5775 5993 6220 0.752 0.164 0.469 0.080 0.001
6.6 5821 6038 6255 5856 6084 6320 0.271 0.842 0.629 0.747 0.000
1 5871 5947 6025 5871 5947 6025 0.056 0.079 0.492 0.032 0.000
2 5663 5909 6165 5663 5909 6165 0.804 0.822 0.668 0.652 0.000
3 5471 5614 5760 5471 5614 5760 0.061 0.150 0.242 0.116 0.001
4 5616 5792 5973 5616 5792 5973 0.818 0.546 0.882 0.279 0.001
5 5646 5775 5907 5646 5775 5907 0.962 0.190 0.754 0.074 0.001
6 5595 5712 5830 5595 5712 5830 0.796 0.105 0.661 0.039 0.001
6
7
8
Calculated by participants (IU/mg)
4b
5
Calculated at EDQM (IU/mg) p-Values Analysis of variance
1
2
3
4
Table 3 Overview of assay results (continued)
p-values < 0.01 are printed on a grey background
n.a: not applicable
Lab Assay
95% Lower
conf. limit
Estimated
potency
95% upper
conf. limit
95% Lower
conf. limit
Estimated
potency
95% upper
conf. limit
Non-
parallelism
Non-
linearity
Lack of
quadratic fit
Quadratic
regression
Quad. regr./
Lin. regr.
Calculated by participants (IU/mg) Calculated at EDQM (IU/mg) p-Values Analysis of variance
1.1 5293 5364 5437 5293 5365 5437 0.179 . . . .
1.2 5292 5349 5407 5294 5349 5406 0.264 . . . .
1.3 5133 5260 5389 5128 5260 5395 0.756 . . . .
1.4 5270 5318 5366 5270 5319 5367 0.558 . . . .
1.5 5194 5258 5323 5193 5258 5324 0.348 . . . .
2.1 5114 5189 5264 5115 5188 5262 0.054 . . . .
2.2 5107 5181 5257 5107 5180 5255 0.157 . . . .
2.3 5142 5197 5252 5143 5196 5249 0.073 . . . .
2.4 5126 5173 5219 5120 5172 5225 0.091 . . . .
2.5 5057 5141 5226 5054 5140 5227 0.163 . . . .
2.6 5121 5219 5320 5122 5219 5317 0.157 . . . .
3.1 5164 5248 5333 5167 5249 5331 0.530 . . . .
3.2 5177 5265 5355 5179 5266 5355 0.856 . . . .
3.3 5192 5297 5404 5190 5298 5409 0.338 . . . .
3.4 5174 5240 5307 5177 5241 5307 0.076 . . . .
3.5 5095 5188 5284 5098 5189 5282 0.364 . . . .
3.6 5097 5220 5346 5099 5221 5345 0.831 . . . .
4.1 5259 5336 5413 5261 5337 5413 0.060 . . . .
4.2 5241 5327 5415 5244 5328 5414 0.206 . . . .
4.3 5233 5297 5362 5236 5298 5360 0.093 . . . .
4.4 5238 5321 5405 5241 5322 5404 0.082 . . . .
4.5 5233 5311 5389 5238 5312 5387 0.109 . . . .
5.1 5209 5284 5360 5213 5285 5358 0.071 . . . .
5.2 5167 5238 5310 5170 5238 5308 0.700 . . . .
5.3 5225 5293 5361 5226 5293 5362 0.327 . . . .
5.4 5192 5254 5317 5192 5255 5318 0.350 . . . .
5.5 5158 5210 5263 5156 5211 5265 0.597 . . . .
5.6 5150 5242 5336 5152 5243 5334 0.800 . . . .
6.1 5394 5462 5531 5396 5463 5532 0.807 . . . .
6.2 5453 5509 5566 5456 5511 5567 0.692 . . . .
6.3 5446 5526 5608 5450 5528 5608 0.246 . . . .
6.4 5380 5455 5530 5384 5456 5529 0.398 . . . .
1 5124 5599 6113 5126 5600 6113 0.599 0.059 0.094 0.080 0.005
2 5008 5600 6251 5011 5601 6252 0.696 0.135 0.107 0.224 0.003
3 5032 5607 6238 5035 5608 6238 0.708 0.132 0.175 0.130 0.005
4 5151 5605 6095 5152 5606 6095 0.168 0.060 0.157 0.051 0.005
5 5189 5632 6108 5190 5633 6109 0.902 0.148 0.830 0.054 0.005
6 5247 5614 6005 5247 5615 6006 0.895 0.129 0.170 0.130 0.002
1 5088 5573 6107 5095 5570 6092 0.450 0.197 0.754 0.079 0.005
2 5268 5713 6198 5266 5710 6195 0.217 0.522 0.552 0.334 0.001
3 5257 5525 5809 5259 5523 5801 0.598 0.171 0.246 0.136 0.001
4 5093 5467 5870 5091 5465 5868 0.907 0.221 0.788 0.089 0.003
5 5099 5471 5871 5098 5468 5869 0.912 0.178 0.655 0.073 0.003
6 5162 5594 6062 5160 5591 6060 0.230 0.973 0.848 0.898 0.000
1.1 5826 5945 6067 5851 5936 6022 0.207 0.050 0.272 0.027 0.001
1.2 5785 5887 5992 5797 5875 5955 0.246 0.081 0.358 0.040 0.000
1.3 5725 5835 5947 5732 5823 5914 0.307 0.067 0.085 0.104 0.000
2.1 5821 5969 6142 5849 5956 6064 0.268 0.038 0.273 0.020 0.001
2.2 5894 5997 6101 5907 5979 6051 0.904 0.090 0.944 0.030 0.000
2.3 5868 5978 6090 5898 5967 6038 0.157 0.015 0.091 0.014 0.001
3.1 5719 5954 6199 5776 5939 6107 0.746 0.163 0.810 0.062 0.002
3.2 5782 5943 6108 5783 5922 6065 0.316 0.234 0.559 0.112 0.001
3.3 5685 5967 6262 5793 5978 6169 0.449 0.081 0.487 0.033 0.003
4.1 5760 5984 6216 5770 5969 6174 0.715 0.230 0.981 0.090 0.002
4.2 5729 5927 6131 5743 5903 6067 0.454 0.827 0.804 0.578 0.000
4.3 5865 6030 6199 5924 6038 6155 0.118 0.094 0.555 0.037 0.001
5.1 5722 5856 5992 5733 5841 5951 0.547 0.402 0.584 0.221 0.000
5.2 5822 5907 5993 5823 5893 5964 0.002 0.604 0.751 0.346 0.000
5.3 5733 5882 6034 5734 5867 6003 0.482 0.094 0.569 0.037 0.001
6.1 5893 5983 6075 5897 5965 6034 0.714 0.049 0.944 0.015 0.000
6.2 5844 6019 6199 5841 5972 6106 0.813 0.302 0.945 0.127 0.001
6.3 5832 6017 6207 5894 6001 6111 0.939 0.799 0.965 0.509 0.000
1 6162 6162 6162 5789 6149 6534 0.273 0.903 0.750 0.750 0.000
2 5940 5940 5940 5583 5954 6348 0.064 0.787 0.827 0.513 0.001
3 6078 6078 6078 5838 6093 6359 0.259 0.140 0.661 0.054 0.004
4 5952 5952 5952 5538 5935 6358 0.546 0.358 0.167 0.727 0.000
5 5760 5760 5760 5440 5774 6125 0.263 0.556 0.516 0.388 0.001
6 5694 5694 5694 5419 5705 6002 1.000 0.067 0.147 0.065 0.005
1 n.a. 7471 7656 7851 0.096 0.001 0.000 0.267 0.000
2 n.a. 7093 7301 7521 0.696 0.000 0.000 0.652 0.000
3 n.a. 7384 7632 7897 0.001 0.053 0.018 0.624 0.000
4 n.a. 7072 7321 7588 0.734 0.011 0.003 0.844 0.000
5 n.a. 7131 7349 7580 0.434 0.002 0.001 0.821 0.000
6 n.a. 7246 7433 7630 0.036 0.060 0.023 0.492 0.000
1 5614 5870 6136 5614 5870 6136 0.702 0.398 0.275 0.421 0.001
2 5620 5857 6103 5620 5857 6103 0.887 0.227 0.171 0.291 0.001
3 5516 5809 6114 5516 5809 6114 0.058 0.776 0.484 0.937 0.000
4 5774 5956 6144 5774 5956 6144 0.573 0.467 0.332 0.449 0.000
5 5545 5822 6111 5545 5822 6111 0.307 0.641 0.375 0.766 0.000
6 5717 6008 6313 5717 6008 6313 0.407 0.631 0.731 0.375 0.001
1 5812 5976 6145 5812 5976 6145 0.519 0.469 0.344 0.657 0.000
2 5646 5719 5793 5646 5719 5793 0.686 0.257 0.404 0.119 0.000
3 5356 5431 5506 5356 5431 5506 0.418 0.038 0.019 0.748 0.000
4 5434 5534 5634 5434 5534 5634 0.398 0.007 0.005 0.220 0.000
5 5361 5440 5521 5361 5440 5521 0.012 0.724 0.669 0.483 0.000
6 5419 5522 5627 5419 5522 5627 0.000 0.000 0.000 0.000 0.006
1.1 5108 5410 5712 5113 5410 5717 0.680 0.454 0.932 0.213 0.003
1.2 5181 5376 5572 5183 5377 5575 0.535 0.268 0.106 0.994 0.000
2.1 4927 5291 5655 4984 5334 5692 0.947 0.393 0.280 0.403 0.002
2.2 5065 5253 5441 5065 5252 5441 0.188 0.698 0.840 0.413 0.000
3 5235 5392 5549 5235 5391 5549 0.576 0.910 0.836 0.703 0.000
4 5129 5261 5394 5128 5260 5393 0.887 0.161 0.061 0.722 0.000
5 4981 5234 5486 4986 5234 5492 0.681 0.201 0.113 0.402 0.001
6 4989 5268 5547 4994 5268 5550 0.094 0.293 0.513 0.155 0.003
7 5291 5415 5540 5292 5415 5541 0.922 0.645 0.596 0.442 0.000
15
9
17
10
11
16
12
13
14
F
i
g
u
r
e

2

I
n
d
i
v
i
d
u
a
l

e
s
t
i
m
a
t
e
s

p
e
r

a
s
s
a
y

(
c
a
l
c
u
l
a
t
e
d

a
t

E
D
Q
M
)

(
n
u
m
b
e
r
s

b
e
l
o
w

t
h
e

9
5

%

c
o
n

d
e
n
c
e

i
n
t
e
r
v
a
l
s

a
r
e

t
h
e

l
a
b
o
r
a
t
o
r
y

c
o
d
e
s
;

i
n
v
a
l
i
d

a
s
s
a
y
s

a
r
e

s
h
o
w
n

w
i
t
h

a
n

e
m
p
t
y

d
o
t
)
1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
b 4
b4
b4
b4
b 4
b
5
5
5
5
5
5
6
6
6
6
6
6
6
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
8
8
8
8
8
8
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
1
01
0 1
0 1
0 1
0 1
01
1 1
11
11
1 1
1 1
1 1
21
21
2 1
2 1
21
21
2 1
2 1
21
21
2 1
21
2 1
21
2 1
21
2 1
21
31
3 1
31
31
31
3 1
41
4 1
41
4 1
4 1
41
5 1
51
5 1
51
5 1
5 1
61
61
6 1
61
6 1
61
7 1
71
7 1
7 1
71
71
7 1
7 1
7
4
0
0
0
4
5
0
0
5
0
0
0
5
5
0
0
6
0
0
0
6
5
0
0
7
0
0
0
7
5
0
0
8
0
0
0
P o t e n c y ( I U / m g )
L
a
b
o
r
a
t
o
r
y
F
i
g
u
r
e

3

I
n
d
i
v
i
d
u
a
l

e
s
t
i
m
a
t
e
s

p
e
r

a
s
s
a
y

(
c
a
l
c
u
l
a
t
e
d

b
y

p
a
r
t
i
c
i
p
a
n
t
s
)

(
n
u
m
b
e
r
s

b
e
l
o
w

t
h
e

9
5

%

c
o
n

d
e
n
c
e

i
n
t
e
r
v
a
l
s

a
r
e

t
h
e

l
a
b
o
r
a
t
o
r
y

c
o
d
e
s
;

e
m
p
t
y

d
o
t
s

i
n
d
i
c
a
t
e

a
s
s
a
y
s

c
o
n
s
i
d
e
r
e
d

t
o

b
e

i
n
v
a
l
i
d

b
y

t
h
e

p
a
r
t
i
c
i
p
a
n
t
s

t
h
e
m
s
e
l
v
e
s
)
1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
6
6
6
6
6
6
6
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
8
8
8
8
8
8
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
1
01
0 1
0 1
0 1
0 1
01
1 1
11
11
1 1
1 1
1 1
21
21
2 1
2 1
21
21
2 1
21
2 1
21
2 1
21
2 1
21
2 1
21
21
2 1
31
3 1
31
31
31
3
1
5 1
51
5 1
51
5 1
5 1
61
61
6 1
61
6 1
61
7 1
71
7 1
7 1
71
71
7 1
7 1
7
4
0
0
0
4
5
0
0
5
0
0
0
5
5
0
0
6
0
0
0
6
5
0
0
7
0
0
0
7
5
0
0
8
0
0
0
P o t e n c y ( I U / m g )
L
a
b
o
r
a
t
o
r
y
Figure 5 Histogram of nal potency estimates per laboratory (numbers in the boxes are the laboratory codes)
Figure 4 Histogram of nal potency estimates per independent assay (numbers in the boxes are the laboratory codes)
Lab
Final potency estimates (IU/mg)
95% Lower conf. limit Estimated potency 95% upper conf. limit
1 5736 5829 5924
2 5937 5963 5990
3 5694 5776 5860
4 5207 5310 5415
5 5614 5684 5754
6 5523 5610 5699
7 5955 5981 6007
8 5732 5801 5870
9 5262 5299 5337
10 5426 5612 5805
11 5392 5540 5693
12 5906 5937 5968
13 5810 5940 6073
14 7244 7394 7547
15 5801 5898 5997
16 5485 5621 5760
17 5277 5337 5398
Comb. 5673 5710 5747
The combined results do not include Laboratory 14.
Table 4 Combined potency estimates per laboratory
F
i
g
u
r
e

6

9
5

%

c
o
n

d
e
n
c
e

i
n
t
e
r
v
a
l
s

p
e
r

l
a
b
o
r
a
t
o
r
y

(
c
o
m
b
i
n
e
d

a
s
s
a
y
s
)
Table 5 Accelerated degradation study: microbiological assay results
Relative Potency in per cent vs sample stored at - 80C
Storage temperature - 20C + 4C + 20C + 37C + 45C
1 month
Vial 1 104.7 102.1 97.5 102.7 100.7
CI 95% 100.1-109.6 97.8-106.6 92.2-103.2 98.5-107.1 96.0-105.7
Vial 2 105.6 106.8 101.5 99.8 101.0
CI 95% 99.6-112.1 101.5-112.4 97.4-105.7 94.8-105.1 95.7-106.5
Mean 105.1 104.0 100.1 101.6 100.9
CI 95% 101.5-108.8 100.8-107.4 96.9-103.4 98.4-104.8 97.4-104.4
3 months
Vial 1 99.5 102.6 99.2 90.2 100.9
CI 95% 96.2-103.0 96.6-109.1 93.0-106.0 85.7-94.9 95.8-106.4
Vial 2 100.6 102.1 99.0 97.1 99.2
CI 95% 95.6-105.9 97.5-106.9 94.7-103.4 93.9-100.4 95.1-103.5
Mean 99.8 102.3 99.1
n.c.
99.9
CI 95% 97.1-102.7 98.6-106.0 95.6-102.6 96.7-103.2
CI: condence interval; n.c.: not calculated
Table 6 Accelerated degradation study: liquid chromatography results (mean peak areas in per cent)

One Month Three months
Temp. -20C +4C +20C +37C +45C -20C +4C +20C +37C +45C
Peak 1 0.12 0.11 0.12 0.12 0.14 0.12 0.13 0.13 0.15 0.13
Peak 2 0.08 0.08 0.08 0.08 0.09 n.d. n.d. 0.09 0.09 n.d.
Peak 3 0.33 0.32 0.32 0.32 0.32 0.33 0.32 0.33 0.32 0.32
Peak 4 0.75 0.76 0.77 0.78 0.78 0.77 0.83 0.77 0.78 0.76
Peak 5 0.49 0.5 0.51 0.5 0.57 0.52 0.62 0.48 0.49 0.43
Nystatin 89.46 89.22 89.48 89.13 89.06 89.58 89.15 90.05 90.64 91.13
Peak 6 0.83 0.82 0.83 0.83 0.81 0.81 0.85 0.74 0.64 0.62
Peak 7 0.34 0.35 0.396 0.36 0.32 0.34 0.35 0.3 0.2 0.19
Peak 8 0.8 0.82 0.85 0.92 0.82 0.8 0.82 0.69 0.55 0.51
Peak 9 1.3 1.3 1.29 1.24 1.07 1.29 1.28 1.12 0.76 0.71
Peak 10 0.85 0.86 0.86 0.84 0.73 0.88 0.86 0.79 0.6 0.6
Peak 11 1.17 1.18 1.18 1.16 1.04 1.18 1.18 1.14 1.07 1.06
Peak 12 0.25 0.3 0.31 0.36 0.42 0.26 0.3 0.34 0.38 0.17
Peak 13 0.43 0.39 0.36 0.46 0.44 0.44 0.45 0.42 0.43 0.62
Peak 14 0.77 0.86 0.87 0.68 0.89 0.76 0.76 0.73 0.8 0.8
Peak 15 0.22 0.23 0.2 0.39 0.22 0.23 0.23 0.21 0.25 0.24
Peak 16 0.16 0.16 0.14 0.18 0.16 n.d. n.d. 0.13 0.15 n.d.
Peak 17 0.24 0.27 0.24 0.33 0.28 0.46 0.41 0.3 0.34 0.49
Peak 18 0.21 0.23 0.2 0.22 0.25 0.22 0.23 0.23 0.23 0.22
Peak 19 0.52 0.54 0.49 0.53 0.61 0.5 0.55 0.51 0.55 0.52
Peak 20 0.21 0.21 0.19 0.19 0.29 0.11 0.13 0.11 0.11 0.11
Peak 21 0.1 0.09 0.07 0.08 0.15 0.13 0.2 0.14 0.18 0.14
Peak 22 0.24 0.25 0.22 0.22 0.42 0.25 0.34 0.26 0.28 0.24
Peak 23 0.13 0.13 0.07 0.07 0.15 n.d. n.d. n.d. n.d. n.d.
n.d.: not detected
Figure 8 Impurity proles after 1 month
Figure 9 Impurity proles after 3 months
Figure 7 Accelerated degradation study: microbiological assay results
A
n
n
e
x

1
:

M
e
t
h
o
d
s

u
s
e
d

b
y

t
h
e

p
a
r
t
i
c
i
p
a
n
t
s
L
a
b
o
r
a
t
o
r
y

c
o
d
e
1
2
3
4
5
6
7
8
9
1
0
1
1
1
2
1
3
1
4
1
5
1
6
1
7
E
x
p
e
r
i
m
e
n
t
a
l

d
e
s
i
g
n
R
a
n
d
o
m
i
s
e
d

b
l
o
c
k
R
a
n
d
o
m
i
s
e
d

b
l
o
c
k
R
a
n
d
o
m
i
s
e
d

b
l
o
c
k
L
a
t
i
n

s
q
u
a
r
e

L
a
t
i
n

s
q
u
a
r
e

(
6
X
6
)
L
a
t
i
n

s
q
u
a
r
e

(
6
X
6
)
L
a
t
i
n

s
q
u
a
r
e
R
a
n
d
o
m
i
s
e
d

b
l
o
c
k
L
a
t
i
n

s
q
u
a
r
e

(
8
X
8
)
R
a
n
d
o
m
i
s
e
d

b
l
o
c
k
R
a
n
d
o
m
i
s
e
d

b
l
o
c
k
L
a
t
i
n

s
q
u
a
r
e
R
a
n
d
o
m
i
s
e
d

b
l
o
c
k
R
a
n
d
o
m
i
s
e
d

b
l
o
c
k
R
a
n
d
o
m
i
s
e
d

b
l
o
c
k
L
a
t
i
n

s
q
u
a
r
e
R
a
n
d
o
m
i
s
e
d

b
l
o
c
k
S
u
b
s
t
a
n
c
e

d
i
s
s
o
l
v
e
d

i
n
D
M
F
D
M
F
D
M
F
D
M
F
D
M
F
D
M
F
D
M
F
D
M
F
D
M
F
D
M
F

u
n
t
i
l

1
0
0
0

I
U
/
m
l
D
M
F

u
n
t
i
l

1
0
0
0

I
U
/
m
l
D
M
F
D
M
F
D
M
F
D
M
F
D
M
F
D
M
F
S
t
o
c
k

s
o
l
u
t
i
o
n

d
i
l
u
t
e
d

w
i
t
h
0
.
0
5

M

p
h
o
s
p
h
a
t
e

b
u
f
f
e
r

p
H
6
.
0
0
.
0
5

M

p
h
o
s
p
h
a
t
e

b
u
f
f
e
r

p
H
6
.
0

5
%

v
/
v

D
M
F

0
.
0
5

M

p
h
o
s
p
h
a
t
e

b
u
f
f
e
r

p
H
6
.
0

5
%

v
/
v

D
M
F

0
.
0
5

M

p
h
o
s
p
h
a
t
e

b
u
f
f
e
r

p
H
6
.
0

5
%

v
/
v

D
M
F

P
o
t
a
s
s
i
u
m

p
h
o
s
p
h
a
t
e

b
u
f
f
e
r

p
H
6
.
0
0
.
0
5

M

p
h
o
s
p
h
a
t
e

b
u
f
f
e
r

p
H
6
.
0

5
%

v
/
v

D
M
F

0
.
0
5

M

p
h
o
s
p
h
a
t
e

b
u
f
f
e
r

p
H
6
.
0

5
%

v
/
v

D
M
F

0
.
0
5

M

p
h
o
s
p
h
a
t
e

b
u
f
f
e
r

p
H
6
.
0
0
.
0
5

M

p
h
o
s
p
h
a
t
e

b
u
f
f
e
r

p
H
6
.
0

5
%

v
/
v

D
M
F

0
.
0
5

M

p
h
o
s
p
h
a
t
e

b
u
f
f
e
r

p
H
6
.
0
0
.
0
5

M

p
h
o
s
p
h
a
t
e

b
u
f
f
e
r

p
H
6
.
0

5
%

v
/
v

D
M
F

D
M
F
0
.
0
5

M

p
h
o
s
p
h
a
t
e

b
u
f
f
e
r

p
H
6
.
0
0
.
0
5

M

p
h
o
s
p
h
a
t
e

b
u
f
f
e
r

p
H
6
.
0

5
%

v
/
v

D
M
F

0
.
0
5

M

p
h
o
s
p
h
a
t
e

b
u
f
f
e
r

p
H
6
.
0

5
%

v
/
v

D
M
F

0
.
0
5

M

p
h
o
s
p
h
a
t
e

b
u
f
f
e
r

p
H
6
.
0

5
%

v
/
v

D
M
F

0
.
0
5

M

p
h
o
s
p
h
a
t
e

b
u
f
f
e
r

p
H
6
.
0

5
%

v
/
v

D
M
F

T
e
s
t

m
i
c
r
o
o
r
g
a
n
i
s
m
S
a
c
c
h
a
r
o
m
y
-
c
e
s

c
e
r
e
v
i
s
i
a
e

A
T
C
C

2
6
0
1
S
a
c
c
h
a
r
o
m
y
-
c
e
s

c
e
r
e
v
i
s
i
a
e

A
T
C
C

9
7
6
3
S
a
c
c
h
a
r
o
m
y
-
c
e
s

c
e
r
e
v
i
s
i
a
e

A
T
C
C

9
7
6
3
S
a
c
c
h
a
r
o
m
y
-
c
e
s

c
e
r
e
v
i
s
i
a
e

A
T
C
C

9
7
6
3
C
a
n
d
i
d
a

t
r
o
p
i
c
a
l
i
s

A
T
C
C

1
3
8
0
3
S
a
c
c
h
a
r
o
m
y
-
c
e
s

c
e
r
e
v
i
s
i
a
e

A
T
C
C

9
7
6
3
S
a
c
c
h
a
r
o
m
y
-
c
e
s

c
e
r
e
v
i
s
i
a
e

A
T
C
C

9
7
6
3
S
a
c
c
h
a
r
o
m
y
-
c
e
s

c
e
r
e
v
i
s
i
a
e

C
C
M

8
1
9
1
S
a
c
c
h
a
r
o
m
y
-
c
e
s

c
e
r
e
v
i
s
i
a
e

N
C
Y
C

8
7
S
a
c
c
h
a
r
o
m
y
-
c
e
s

c
e
r
e
v
i
s
i
a
e

A
T
C
C

2
6
0
1
S
a
c
c
h
a
r
o
m
y
-
c
e
s

c
e
r
e
v
i
s
i
a
e

A
T
C
C

9
7
6
3
S
a
c
c
h
a
r
o
m
y
-
c
e
s

c
e
r
e
v
i
s
i
a
e

A
T
C
C

9
7
6
3
S
a
c
c
h
a
r
o
m
y
-
c
e
s

c
e
r
e
v
i
s
i
a
e

A
T
C
C

9
7
6
3
S
a
c
c
h
a
r
o
m
y
-
c
e
s

c
e
r
e
v
i
s
i
a
e

A
T
C
C

2
6
0
1
S
a
c
c
h
a
r
o
m
y
-
c
e
s

c
e
r
e
v
i
s
i
a
e

A
T
C
C

9
7
6
3
S
a
c
c
h
a
r
o
m
y
-
c
e
s

c
e
r
e
v
i
s
i
a
e

A
T
C
C

9
7
6
3
S
a
c
c
h
a
r
o
m
y
-
c
e
s

c
e
r
e
v
i
s
i
a
e

A
T
C
C

9
7
6
3
N
u
m
b
e
r

o
f

s
u
b
c
u
l
t
u
r
e

(
J
a
n
u
a
r
y
-
M
a
r
c
h

2
0
0
6
)
/
/
5
t
h

N
e
w

o
n
e
4
t
h

p
a
s
s
a
g
e
2
n
d

p
a
s
s
a
g
e
N
o
t

m
o
r
e

t
h
a
n

5
1
2

g
e
n
e
r
a
t
i
o
n
s

f
r
o
m

m
a
s
t
e
r

c
u
l
t
u
r
e
/
/
N
o
t

m
o
r
e

t
h
a
n

3

p
a
s
s
a
g
e
s
/
/
?
/
A
b
o
u
t

8

o
f

s
u
b
c
u
l
t
u
r
e
s
D
i
l
u
t
i
o
n

s
t
e
p
s
1
.
5
1
.
5
2
.
0
1
.
5
1
.
5
1
.
5
2
.
0
1
.
5
3
.
0
2
.
0
2
.
0
1
.
5
1
.
5
1
.
5
1
.
5
1
.
2
5
1
.
5
M
a
t
e
r
i
a
l
P
e
t
r
i

d
i
s
h
e
s
P
e
t
r
i

d
i
s
h
e
s
P
e
t
r
i

d
i
s
h
e
s
P
l
a
t
e
s
P
l
a
t
e
s
P
l
a
t
e
s
P
l
a
t
e
s
P
e
t
r
i

d
i
s
h
e
s
P
l
a
t
e
s
P
e
t
r
i

d
i
s
h
e
s
P
e
t
r
i

d
i
s
h
e
s
P
e
t
r
i

d
i
s
h
e
s
P
l
a
t
e
s
P
e
t
r
i

d
i
s
h
e
s
P
e
t
r
i

d
i
s
h
e
s
P
l
a
t
e
s
P
e
t
r
i

d
i
s
h
e
s
N
r
.

o
f

a
s
s
a
y
s
6
6
6
6

(
3
)
,

r
e
p
e
a
t
e
d

b
y

2
n
d

o
p
e
r
a
t
o
r
6
7
6

(
6
)
6
7

a
s
s
a
y
s

b
u
t

o
n
l
y

6

r
e
s
u
l
t
s

p
r
o
v
i
d
e
d
6
6
6

(
3
)
6
6
6
6
7

+

2
N
r
.

o
f

d
i
s
h
e
s

o
r

p
l
a
t
e
s

u
s
e
d

p
e
r

a
s
s
a
y
6
1
6
9
3
1
1
6
8
4

t
o

6
6
6
3
9
9
6
1
9

t
o

1
5
A
p
p
l
i
c
a
t
i
o
n
s

p
e
r

d
i
s
h

o
r

p
l
a
t
e
6
6
6
3
6
3
6
3
6
3
6
6
6
4
6
6
3
6
6
6
6
6
4
6
C
o
n
c
e
n
t
r
a
t
i
o
n
s

a
p
p
l
i
e
d

p
e
r

d
i
s
h

o
r

p
l
a
t
e
6
6
6
6
6
6
6
6
4

(
2
)
6
6
6
6
6
6
8
6
R
e
p
l
i
c
a
t
e
s

o
f

c
o
n
c
e
n
t
r
a
t
i
o
n
s

p
e
r

d
i
s
h

o
r

p
l
a
t
e
1
1
1
6
6
6
6
1
8
1
1
6
1
1
1
8
1
I
n
c
u
b
a
t
i
o
n

t
e
m
p
e
r
a
t
u
r
e
3
1
C
3
2
C
3
2
C
3
0
C
3
2
C
3
0
C
3
5
-
3
7
C
3
2
C
3
7
C
2
9
-
3
1
C
3
7
C
3
0

C
3
2
C
2
9
-
3
1
C
3
2
C
2
8
.
5
-
3
2
.
5
C
3
1
C
I
n
c
u
b
a
t
i
o
n

t
i
m
e
1
8

h
o
u
r
s
1
8

h
o
u
r
s
1
8

h
o
u
r
s
4
8

h
o
u
r
s
1
8

h
o
u
r
s
1
8

h
o
u
r
s
1
8

h
o
u
r
s
1
8

h
o
u
r
s
1
8

h
o
u
r
s
1
8

h
o
u
r
s
2
0

h
o
u
r
s
1
8

h
o
u
r
s
1
8

h
o
u
r
s
1
6
-
1
8

h
o
u
r
s
2
4

h
o
u
r
s
1
6
-
1
8

h
o
u
r
s
1
8

h
o
u
r
s
I
n
h
i
b
i
t
i
o
n

z
o
n
e
s

m
e
a
s
u
r
e
d

b
y
P
B
I
-
R
e
a
d
B
i
o
t
i
c
A
n
t
i
b
i
o
t
i
c

Z
o
n
e

R
e
a
d
e
r

F
i
s
h
e
r
-
L
i
l
l
y
A
n
t
i
b
i
o
t
i
c

Z
o
n
e

R
e
a
d
e
r

F
i
s
h
e
r
-
L
i
l
l
y
N
a
k
e
d

e
y
e
,

m
m

r
u
l
e
r
A
u
t
o
m
a
t
i
c

Z
o
n
e

M
e
a
s
u
r
i
n
g

E
q
u
i
p
m
e
n
t
M
a
c
r
o

L
e
n
s

m
o
u
n
t
e
d

o
n

a

d
i
g
i
t
a
l

C
a
m
e
r
a

a
n
d

s
o
f
t
w
a
r
e

"
L
E
I
C
A

Q
w
i
n
"
Z
o
n
e

R
e
a
d
e
r
S
y
s
t
e
m

f
o
r

i
m
a
g
e

p
r
o
c
e
s
s
i
n
g

a
n
d

a
n
a
l
y
s
i
s
-
L
U
G
I
A

G

-

v
e
r
s
i
o
n

4
.
7
1
A
u
t
o
m
a
t
i
c

r
e
a
d
e
r

e
q
u
i
p
m
e
n
t
M
i
t
u
t
o
y
o

d
i
g
i
t
a
l

c
a
l
i
p
e
r
P
B
I
-
R
e
a
d
b
i
o
t
i
c
R
e
a
d
b
i
o
t
i
c

v
i
e
w
e
r
S
l
i
d
e

c
a
l
i
p
e
r
A
n
t
i
b
i
o
t
i
c

Z
o
n
e

R
e
a
d
e
r

F
i
s
h
e
r
-
L
i
l
l
y
P
B
I
-
R
e
a
d
b
i
o
t
i
c
M
a
n
u
a
l

r
e
a
d
i
n
g

0
.
1
m
m

p
r
e
c
i
s
i
o
n
?
Annex 2: SAS-Script used for the calculations
/* This is the essential script to perform the analysis of variance.
It expects a SAS-dataset Nystatin with the following elds:
prep: 1 for standard, 2 for test.
dose: on log-scale with the primary dose set to 0.
cdose: copy of dose.
row: indicates the row in Latin square designs.
block: the column in Latin square designs or the petri-dish in randomised block designs.
block and/or row are set to 1 if not applicable for their respective designs.
obs: the value of the observation (possibly transformed).
*/
ods select none;
proc glm data=Nystatin;
/* Perform the Anova by progressively relaxing model assumptions */
class prep cdose block row;
model obs=block row prep dose dose(prep) dose*dose cdose*prep / ss1;
ods output OverallAnova=OverallAnova ModelAnova=ModelAnova;
data Anova(keep=source df ss ms fvalue probf);
/* Non-linearity has to be calculated in a separate datastep */
retain dfLin ssLin;
set ModelAnova OverallAnova;
if df>0 then output;
if Source=dose*dose then do; dfLin=df; ssLin=SS; end;
if Source=prep*cdose then do; dfLin=dfLin+df; ssLin=ssLin+ss; end;
if Source=Error then do;
Source=Non-linearity; FValue=(ssLin/dfLin)/ms; ProbF=1-ProbF(FValue,dfLin,df);
ss=ssLin; df=dfLin; ms=ss/df;
if df>0 then output;
end;
ods select all;
proc print data=Anova noobs;
run;
/* This is the essential script to perform the potency calculations.
It expects a SAS-dataset info with the following elds:
Assigned: The assigned potency of the standard
mgS: weight taken of the Standard
mlS: Dilution used to prepare the primary dose of the Standard.
mgT: weight taken of the Test
mlT: Dilution used to prepare the primary dose of the Test.
*/
ods select none;
proc glm data=Nystatin;
/* Fit the parallel line model and output the parameter estimates and covariance matrix */
class block row;
model obs=prep dose block row / inverse solution;
ods output InvXPX=CovB ParameterEstimates=ParmEst;
data Estimate(keep=Low Est High);
/* calculate the relative potency (m) */
set ParmEst; where Parameter=prep; a=Estimate;
set ParmEst; where Parameter=dose; b=Estimate;
m=a/b;
/* Use Fiellers theorem to compute the condence limits */
set CovB; where Parameter=prep; v11=prep;
set CovB; where Parameter=dose; v12=prep; v22=dose;
set Anova; where source=Error; t=tinv(0.975,df); s=sqrt(ms);
g=(t*t*s*s*v22)/(b*b);
root=v11-2*m*v12+m*m*v22-g*(v11-v12*v12/v22);
mL=(m-g*v12/v22-t*s/b*sqrt(root))/(1-g);
mU=(m-g*v12/v22+t*s/b*sqrt(root))/(1-g);
/* Transform the relative potency to IU by correcting for the pre-dilutions */
set info; Correction=Assigned*mgS/mlS*mlT/mgT;
Low=Correction*exp(mL); Est=Correction*exp(m); High=Correction*exp(mU);
output;
stop;
ods select all;
proc print data=Estimate noobs;
run;
Quality Control of Metoprolol Extended-Release Formulations with Ethanol
Quality Control of Metoprolol Extended-Release
Formulations in the Presence of Ethanol
A. Amini, A. Dawood, A.-M. Hesselgren, H. Thor, S. Jnsson,
T. Arvidsson, G. Ragnarsson, M. Johansson
A. Amini. (Corresponding author: e-mail: ahmad.amini@mpa.se), T. Arvidsson, M. Johansson. Medical Products Agency, P.O. Box 26, SE-751 03
Uppsala, Sweden and Division of Analytical Pharmaceutical Chemistry, Department of Medicinal Chemistry, Uppsala University, Biomedical Centre,
Box 574, SE-751 23, Uppsala, Sweden
A. Dawood. Division of Analytical Pharmaceutical Chemistry, Department of Medicinal Chemistry, Uppsala University, Biomedical Centre, Box 574,
SE-751 23, Uppsala, Sweden
A.-M. Hesselgren, H. Thor, S. Jnsson, G. Ragnarsson. Medical Products Agency, P.O. Box 26, SE-751 03 Uppsala, Sweden
1. INTRODUCTION
The in-vitro dissolution testing at optimized conditions
has often been used as a surrogate for the assessment
of bioequivalent, and is an essential requirement for
registration and quality control of solid oral formulations.
The in-vivo stability of extended-release (ER) formulations
is a prerequisite in order to get a constant and controlled
release of the drug which is mainly absorbed from small
intestine. As a consequence, the plasma concentration
level is a function of dissolution rate of a drug in the ER
formulations. The release rate from the extended-release
tablets (a common ER tablet principle) depends on many
factors, e.g. the drug diffusion coefficient, drug/polymer
ratio, drug solubility and dose, porosity and shape of
the matrix, heterogeneity and molecular mass of release
controlling polymers, solubility of excipients and processing
techniques [1-3]. In an ER formulation, matrix-forming
polymers such as hypromellose and ethyl-cellulose, in
contact with aqueous media and biological fluids hydrates
and forms a protective or barrier gel layer on the matrix
surface. The tablet matrix forms a network consisting of
polymer molecules which are kept together by chemical
or physical bonds.The porosity of this network influences
the dissolution rate of the drug [4, 5]. Soluble drugs in
the ER tablets are released primarily by diffusion through
the barrier-gel while drugs with low solubility are released
primarily by gel erosion. However, it can be concluded
that the drug release occurs through a combination of the
two mentioned mechanisms [1]. Disintegrating tablets
comprising a large number of barrier-coated pellets
represent another ER tablet principle, where the drug
release is dependent on the permeability of the coating layer.
Everything which affects the release rate of the drug
may induce a change in the plasma drug concentration
that sometimes causes undesired consequences [6]. For
instance, ethanol in alcoholic beverages can interact with
ER formulations and thereby change the release profile of
the drug [7]. Therefore, investigation on the in-vitro release
behaviour of ER formulations in the presence of alcohol
is necessary to explore any adverse effect of alcohol on the
drug release.
However, when judging the potential in-vivo impact of the
in-vitro drug release in the presence of alcohol there are a
number of factors that need to be considered, apart from the
effect of alcohol on drug release, such as gastric emptying
rate, drug solubility in the gastrointestinal fluid and the
surface area exposed to the dissolution medium or lumenal
fluids [8, 9]. A large impact of alcohol on the in-vitro drug
release indicates however that alcohol consumption in
relation to drug intake may be a potential problem to be
addressed in the labeling or to be studied in suitable in-vivo
bioequivalence studies. The food and drug administration
(FDA) in 2005 issued an alert to inform the healthcare
professionals regarding the potential adverse effect of
ethanol on ER formulations [10]. The relevance of the
in-vitro drug release testing in the presence of alcohol (at
concentrations up to 40%) has recently been reviewed [9].
Metoprolol (Figure 1) is a cardio-selective
1
-adrenergic
antagonist [11] which has been widely used for treatment
of essential hypertension [12] and angina pectoris [13].
Metoprolol is completely absorbed from the gastrointestinal
tract but because of extensive first-pass metabolism; the
bioavailability of the drug is approximately 50% [11].
The clearance half-life of metoprolol varies between 3
and 4 hours, therefore a twice daily administration of
conventional tablets will be required for effective 24-
hour period treatment [14]. In order to improve patient
comfort and compliance, ER tablets of metoprolol succinate
and metoprolol tartrate have been formulated. These
formulations release the metoprolol content in a controlled
and predictable manner to accomplish a flat plasma drug
concentration time curve and cause fewer short-term
side effects than immediate-release (IR) solid oral dosage
forms, which generally release a greater initial amount
ABSTRACT
This paper presents in-vitro metoprolol release from four different extended-release (ER) formulations, i.e. Metoprolol
GEA
Retard, Logimax
forte, Metoprolol Sandoz
and Seloken ZOC
in the presence of 10 to 40% (v/v%) ethanol

at pH1.2 and pH 6.8. The assay of metoprolol in the dissolution media was performed by reversed phase liquid
chromatography (RP-LC) using a mixture of methanol and 100 mM phosphate buffer (pH 3.5) in 40:60 ratio as eluent.
The dissolution data showed that the metoprolol contents of Metoprolol Sandoz
and Seloken ZOC
were released fast in

the presence of 20% ethanol at the investigated conditions, while the other products demonstrated much more stability
against ethanol. Unexpectedly it was discovered that the release of metoprolol from Metoprolol GEA
Retard and to some

extent also from Logimax
forte decreased in the ethanol containing media.

KEYWORDS
Metoprolol, reversed phase liquid chromatography, dissolution, Metoprolol GEA
Retard, Logimax
forte, Metoprolol
Sandoz
, Seloken ZOC
, ethanol, hypromellose.
of the drug [15]. The objective of this paper has been to
investigate in-vitro metoprolol release from four different
ER dosage forms, i.e. Metoprolol GEA
Retard, Logimax

and Seloken ZOC
in the presence
of ethanol in the concentration range of 0 to 40% (v/v%).
The dissolution experiments were carried out at pH 1.2
and pH6.8 for four hours with and without the presence
of ethanol. The absorption of ethanol occurs in the small
intestine, thus the alcohol-formulation interaction will
be more intense in stomach [9]. The pH of the dissolution
media was adjusted to the aforementioned values to
approximately resemble different part of the digestive system
in their pH values.
In general quality testing of ER-formulations does not
include dissolution testing using ethanol in the dissolution
media. However, in cases when data is not available such
testing will be of value to investigate the effect of alcohol
containing dissolution media on the drug release. Therefore,
this study was performed as part of the market surveillance
program of medicines in Sweden.
Figure 1 - Chemical structure of metoprolol
2. EXPERIMENTAL
2.1. Chemicals and reagents
Extended-release formulations, i.e. Metoprolol GEA

Retard, Logimax
and Seloken
ZOC
containing 100 mg metoprolol per dosage form

were purchased from a local pharmacy (Uppsala, Sweden).
Metoprolol GEA
Retard contains metoprolol tartrate while

the others contain metoprolol succinate. Two metoprolol
reference standards, i.e. metoprolol tartrate and metoprolol
succinate were kindly supplied by Astra Zeneca (Sdertlje,
Sweden). Methanol (R1) was from Sigma Aldrich (Steinhem,
Germany), KH
2
PO
4
and NaOH were purchased from Merck
(Darmstadt, Germany). Ethanol (99.9%) came from Kemetyl
AB (Haninge, Sweden). Water was purified by Elga Maxima
LC water cleaning system. All samples to be analysed were
filtered through a 0.45 m PVDF filter (Pall, Ann Arbor,
USA). Buffer solutions for LC were filtered through 0.45 m
HVLP filter (Millipore, Cork, Ireland).
2.2 Equipment and methods
Reversed phase liquid chromatography was carried out using
a Waters Alliance LC system (Waters USA) equipped with an
UV detector, operating at 225 nm at ambient temperature.
The separation column was a Symmetry C18 column
(Waters, Milford, USA) of dimensions 150 x 39 mm packed
with 5 m particles. The flow rate was 0.7 mL/min. The
assay of metoprolol was performed using methanol: 100 mM
phosphate buffer at pH 3.5 (40:60 v:v) as eluent.
Eluent buffer was prepared by making a 100 mM KH
2
PO
4

solution and adjusting its pH to 3.5 with 85% ortho-
phosphoric acid (R). A gradient was used when Logimax

forte was analysed as this is a combination product
containing both metoprolol and felodipine. Metoprolol
was first eluted with the same eluent composition as in
the isocratic mode, then the column was washed with
methanol gradient to elute the more hydrophobic substance,
felodipine (compare Figure 2A and Figure 2B). The column
was thereafter equilibrated for ten minutes with the eluent.
2.3 Preparation of standard solutions
Two stock solutions (2 mg/mL) of metoprolol tatrate and
metoprolol succinate were prepared. Each stock solution was
then diluted with the eluent to the following concentrations:
6, 10, 30, 60 and 80 g/mL. The standard solutions
containing metoprolol tartrate were used when Metoprolol
GEA
Retard was analysed and those containing metoprolol

succinate when Logimax
and
Seloken ZOC
were analyzed. The concentration of the

standard solutions were chosen to give a measuring range
of 10-140% of metoprolol during the analyses as the sample
taken from the dissolution vessels were diluted with an equal
volume of water before being analyzed. All concentrations
determined, except those being less than 10%, were within
the concentration range of the calibration graph.
2.4 Method validation
The linearity of the method was conrmed by analysing
standard solutions in the concentration range of
6-101g/mL. The determination coefcient (R
2
) of 1.000
was obtained by plotting the peak area versus the metoprolol
concentration.
The accuracy of the measurements was determined to be
99.7% by analysing a standard solution with a concentration
of 30 g/mL of metoprolol. The relative standard deviations
(RSD %) of the peak areas and the retention times were
measured to be less than 1.0%, i.e., between 0.1% and
0.4% for the peak areas and between 0.1 and 0.8% for the
retention times. The repeatability of the separation system
was conrmed by performing six injections of a standard
solution of 30 g/mL metoprolol before each analysis.
2.5 Dissolution
In-vitro dissolution of the ER tablets was performed
according to the European Pharmacopeia [16] using
apparatus 2 (paddle). The dissolution apparatus was a
Erweka DT 80 (Germany) with the stirring speed of 100 rpm.
The dissolution media (phosphate buffer, pH 6.8 R and
HCl, pH 1.2) were prepared according to the European
Pharmacopoeias recipe [16]. Evaporation of the dissolution
medium was minimized by tightly closing the dissolution
vessels during the experiments. Six tablets were analyzed
simultaneously, i.e. three tablets per medium, according to
the following scheme:
Figure 2 - RP-LC analysis of metoprolol (A) and metoprolol
and felodipine (B). Experimental conditions as given in the
Experimental section. Peaks 1 and 2 represent metoprolol
and felodipine, respectively.
Experiment sets Ethanol (%) Number
of tablets
A
(pH = 1.2)
0 3
10 3
B
(pH = 1.2)
20 3
40 3
C
(pH = 6.8)
0 3
20 3
In each experiment 2x3 tablets at six different conditions
(i.e. sets A, B, and C) were tested. The total volume of each
vessel was 900 mL. The tablets were added into the vessels
with one minute interval. The sampling was carried out by
manually collecting 10 mL sample after 15, 30, 60, 120 and
240 minutes, using a syringe equipped with a 10 m lter
(poroplast lter, Erweka, USA). Three (3.0) mL of the sample
solution was diluted with 3.0 mL phosphate buffer (200 mM,
pH 3.5), before being analyzed by RP-LC. The dilution with
phosphate buffer was made to obtain a more efcient peak in
the chromatographic system.
2.6 Calculations
The concentration of metoprolol in the dissolution medium
(C
dis
) was determined converting the peak area of metoprolol
to concentration using a standard curve constructed with
metoprolol standard solutions at 6 to 80 g/mL. The percent
dissolved metoprolol at each time point was calculated by
Eq. (1):
% (Dissolved amount) = 100 (C
dis(t)
* V
dis(t)
/ A
tot
) Eq. (1)
where C
dis(t)
and V
dis(t)
are the determined concentration and
the total volume of the dissolution medium, respectively. A
tot

is the total amount of metoprolol in the dosage, i.e. 100mg.
V
dis(t)
was calculated by subtracting the volumes of the
samples taken (v) for analysis at each time point from the
original volume (V
tot
) of the dissolution medium.
Product name API Formulation Excipients
Metoprolol GEA
Retard Metoprolol tartrate Extended-release tablet (non-

dividable)
Hypromellose, povidone, colloidal silicon
dioxide anhydrous, magnesium stearate.
Logimax
forte Metoprolol succinate

and Felodipine
Extended-release tablet (non-
dividable) consisting of pellets
Hypromellose, HPC, cellulose, ethyl-
cellulose, propyl gallate, lactose,
non-colloidal silicon dioxide, hard
parafn, sodium aluminium silicate,
macrogol, sodium stearyl fumarate ,
macrogolglycerol hydroxystearate.
Metoprolol Sandoz
Metoprolol succinate Extended-release tablet (dividable

but non-crushable) consisting of
pellets
Hypromellose, cellulose, magnesium
stearate, povidone colloidal silicon dioxide
anhydrous, lactose, glucose, macrogol,
poly-acrylate, sacrose.
Seloken ZOC
Metoprolol succinate Extended-release tablet (dividable

but non-crushable) consisting of
pellets
Hypromellose, HPC, cellulose, ethyl-
cellulose, macrogol, hard parafn, non-
colloidal silicon dioxide anhydrous.
3. RESULTS
The investigated products, i.e. Metoprolol GEA
Retard,
Logimax
and Seloken ZOC
,
contain metoprolol (pK
a
= 7.9) as active pharmaceutical
ingredient (API) and hypromellose together with other
cellulose ethers as excipients (Table 1). Metoprolol GEA

Retard contains metoprolol tartrate while the others
contain metoprolol succinate. Both metoprolol tartrate
and succinate are freely soluble in water but as opposed
to metoprolol tartrate, metoprolol succinate is sparingly
soluble in ethanol [17-19], which can play a role in the
control of the metoprolol release in ethanol containing
media. Hypromellose, as opposed to ethyl-cellulose, is
insoluble in ethanol (95 %) but it is soluble in mixtures of
water and ethanol [4].
The dissolution proles of these delivery systems in different
dissolution media are discussed below.
3.1 In-vitro metoprolol-release from Metoprolol GEA

Retard formulation
The Metoprolol GEA
Retard formulation is a non-

dividable ER tablet which contains metoprolol tartrate as
API (Table1). The formulation was not disintegrated in
the dissolution media. A swollen gel lump was left in the
dissolution vessel after four hours.
The results from the drug-release in the presence and
absence of ethanol (Table 2) showed that the metoprolol
tartrate dissolution decreased signicantly (P = 0.05) in the
presence of ethanol at both pH 6.8 and pH 1.2. No signicant
difference in dissolution proles between these pH values
was observed. The release of the drug, however, showed a
tendency to decrease with increasing ethanol concentration
(Figure 3). This implies that ethanol suppress the release of
metoprolol from the formulation, which may depend on the
reduced hydration and/or porosity of the gel in the presence
of ethanol. The results demonstrated the consistency of the
gel structure at the examined conditions.
Time
(min)
pH = 6.8
(0% Ethanol)
pH = 6.8
(20% Ethanol)
pH = 1.2
(0% Ethanol)
pH = 1.2
(10% Ethanol)
pH = 1.2
(20% Ethanol)
pH = 1.2
(40% Ethanol)
15 7.2 % 5.9 % 7.7 % 7.1 % 6.5 % 5.9 %
30 10.7% 8.8 % 11.4 % 10.5 % 9.8 % 8.9 %
60 16.0 % 13.0 % 17.0 % 15.5 % 14.4 % 13.1 %
120 24.2 % 20.2 % 25.4 % 23.1 % 21.4 % 19.6 %
240 37.0 % 29.7 % 38.1 % 34.5 % 31.7 % 29.2 %
Table 1 - Analyzed extended-release tablets along with their active pharmaceutical ingredients (API) and excipients
Table 2 - Dissolution data from release of metoprolol from Metoprolol GEA
Retard tablets in the presence of 0, 10, 20

and 40% ethanol at pH 1.2 and pH 6.8 (n = 3)
The percentage of the released amount was calculated by equation 1.
Table 3 - Dissolution data from metoprolol release from Logimax
forte tablets in the presence of 0, 10, 20 and

40% ethanol at pH 1.2 and pH 6.8 (n = 3)
3.2. In-vitro metoprolol-release from Logimax
forte
formulation
The Logimax
forte formulation is a non-dividable ER

tablet which constitutes of pellets containing metoprolol
succinate (Table 1). Logimax
forte is a multiple-drug
delivery system which in addition to metopropol delivers
felopdipine, which also reduces blood pressure by blocking
the calcium channels. However, dissolution of felodipine was
not investigated.
The tablet did not immediately disintegrate in the
dissolution media, but sank to the bottom of the test vessel
after fteen minutes exposure. After approximately two
hours dissolution in the ethanol free media the gel eroded
and small particles appeared in the solution.
During the rst two hours of the dissolution process,
using 10 and 20% ethanol (pH 1.2), a slight decrease in the
liberated amount of metoprolol, compared to that in the
ethanol free medium (Table 3) was observed. However, an
increase in the ethanol concentration from 20% to 40%
resulted in a signicant increase in the release of metoprolol
after 30 min (Table 3). The results indicate that ethanol at
40% level may contribute to a faster erosion and dissolution
of the gel matrix, e.g. hypromellose and ethyl-cellulose. On
the other hand a decrease in the solubility of metoprolol
succinate in the presence of ethanol may slow down the
dissolution rate.
3.3. In-vitro metoprolol-release from Metoprolol Sandoz

formulation
Metoprolol Sandoz
delivering metoprolol succinate is a

disintegrating ER-tablet consisting of compressed depot
units (Table 1). In contact with dissolution medium the
tablet was immediately disintegrated into its constituent
units. The total surface-area of the pellets being exposed
to the surrounding medium will therefore be much larger
than that of a non-dividable ER-tablet, e.g. Metoprolol
GEA
Retard. The high surface area contributes to a higher

dissolution rate, since dissolution is directly proportional to
the exposed surface area. However, a comparison between
the results summarized in Tables 2 and 4 demonstrate
that the metoprolol dissolution from this ER-formulation
despite its larger exposed surface area is much smaller than
Metoprolol GEA
Retard. This illustrates the substantial

effect of excipient choice and formulation factors.
The results showed that the dissolution was slightly
affected by 20% ethanol during the rst 30 minutes of the
dissolution process both at pH 6.8 and pH 1.2. However,
the amount of metoprolol released after four hours in the
presence of 20% ethanol was found to be approximately
twice the amount released in the absence of ethanol at both
pHs. The same amount of metoprolol was released already
after 60 min in the presence of 40% ethanol (Table 4). The
alcohol sensitivity of this product has also been noted by the
manufacturer with a warning text implying that there will
be reinforcement of the metoprolol effect in the presence of
ethanol [20].
Time
(min)
pH = 6.8
(0% Ethanol)
pH = 6.8
(20% Ethanol)
pH = 1.2
(0% Ethanol)
pH = 1.2
(10% Ethanol)
pH = 1.2
(20% Ethanol)
pH = 1.2
(40% Ethanol)
15 0.4 % 0.1 % 0.5 % 0.3 % 0.5 % 0.6 %
30 0.8 % 0.5 % 1.2 % 0.8 % 0.9 % 1.6 %
60 2.4 % 2.0 % 3.2 % 2.4 % 2.3 % 7.3 %
120 6.4 % 8.1 % 8.6 % 7.2 % 7.8 % 18.1 %
240 15.3 % 23.4 % 18.4 % 17.3 % 23.1 % 32.7 %
Table 4 - Dissolution data from metoprolol release from Metoprolol Sandoz
in the presence of 0, 10, 20 and 40% ethanol

at pH 1.2 and pH 6.8 (n = 3)
Time
(min)
pH = 6.8
(0% Ethanol)
pH = 6.8
(20% Ethanol)
pH = 1.2
(0% Ethanol)
pH = 1.2
(10% Ethanol)
pH = 1.2
(20% Ethanol)
pH = 1.2
(40% Ethanol)
15 0.5 % 0.6 % 1.1 % 0.9 % 1.3 % 5.8 %
30 1.7 % 2.2 % 2.2 % 1.8 % 2.9 % 16.0 %
60 4.3 % 6.4 % 4.5 % 4.1 % 6.4 % 43.4 %
120 9.8 % 18.2 % 9.6 % 9.5 % 16.4 % 72.5 %
240 21.2 % 44.2 % 20.9 % 23.7 % 40.3 % 83.1 %
Figure 3 - In-vitro concentration-time proles after
metoprolol dissolution from Metoprorol GEA
Retard
tablets at different conditions. The concentration of
metoprolol was determined in the dissolution media at
pH 1.2 containing ethanol at 0%, 10%, 20% and 40% v/v
levels. The concentration of metoprolol was determined at
different time periods, i.e., 15 min (), 30 min (),
60 min (), 120 min () and 240 min ()
Time
(min)
pH = 6.8
(0% Ethanol)
pH = 6.8
(20% Ethanol)
pH = 1.2
(0% Ethanol)
pH = 1.2
(10% Ethanol)
pH = 1.2
(20% Ethanol)
pH = 1.2
(40% Ethanol)
15 2.4 % 2.9 % 3.6 % 3.5 % 3.7 % 8.5 %
30 4.9 % 7.2 % 6.7 % 6.6 % 8.0 % 36.5 %
60 8.9 % 18.8 % 11.2 % 11.3 % 18.0 % 71.5 %
120 14.5 % 47.3 % 17.5 % 20.7 % 43.2 % 84.8 %
240 25.0 % 78.1 % 28.2 % 40.8 % 76.1 % 87.8 %
Table 5 - Dissolution data from release of metoprolol from Seloken ZOC
tablets in the presence of 0, 10, 20 and

40% ethanol at pH 1.2 and pH 6.8 (n = 3)
3.4. In-vitro metoprolol-release from Seloken ZOC

formulation
This formulation is also a disintegrating ER-tablet which
consists of drug pellets coated with a mixture of ethyl-
cellulose and hydroxypropyl cellulose [21]. The pellets
contain metoprolol succinate as the active substance.
The ethyl-cellulose forms a water insoluble but ethanol
soluble barrier coating. The water soluble hydroxypropyl
cellulose is included to modify the release rate. These
tablets disintegrated after 3 minutes releasing its content
pellets into the dissolution medium. As given in Table 5, the
dissolution in the absence of ethanol was slightly faster at
pH 1.2 compared to that at pH 6.8. The metoprolol release
increased in the presence of ethanol in the dissolution
media. After four hours dissolution the metoprolol
concentrations at both pH 1.2 and pH 6.8 had increased,
compared to those in the ethanol free media, by about three
times in the presence of 20% ethanol. The time-period
required to release 50% of the metoprolol content (T
50%
) was
longer than four hours in the absence and approximately
2hours in the presence of 20% ethanol and between 30 and
60 min in the presence of 40% ethanol (Table 5).
4. DISCUSSION
Metoprolol GEA
Retard and Logimax
forte are non-

dividable matrix tablets which contain hypromellose as
matrix former. There are, however, other cellulose ethers,
such as HPC, present in Logimax
forte. Both of these

formulations showed to be less effected by the presence of
ethanol compared to the others. A comparison between the
dissolution proles summarized in Tables 2 and 3 shows that
the dissolution of metoprolol from Metoprolol GEA
Retard
was signicantly faster than that from Logimax
forte both
in the presence and absence of ethanol. The difference may
be explained by differences in formulation composition and
technology. The reduced dissolution of metoprolol from
these formulations in the presence of ethanol may be caused
by a reduced permeability of the gel matrix. In comparison
to Logimax
forte, Metoprolol GEA
Retard demonstrated to
be more stable at alcoholic conditions.
The in-vitro dissolution prole of Logimax
forte is much
more similar to Metoprolol Sandoz
and Seloken ZOC

than to Metoprolol GEA
Retard dosage form. The similarity

in the dissolution proles may be due to that the three
formulations consist of pellets. However, in contrast to
the Metoprolol Sandoz
and Seloken ZOC
formulations
the Logimax
forte was not disintegrated in contact with

dissolution media. Furthermore, the Logimax
forte
formulation demonstrated to be less effected by ethanol
containing media as compared to Seloken ZOC
and
Metoprolol Sandoz
.
The effect of alcohol on the release rate of Selocen ZOC

indicated that the pellet barrier coating was negatively
affected by ethanol resulting in a faster release of
metoprolol from the pellets. The proportion of the lm
coating polymers, ethyl-cellulose and hypromellose, in the
formulations may have inuenced the alcohol sensitivity of
the delivery systems, since the solubility of ethyl-cellulose in
ethanol is higher than that of hypromellose [5].
Metoprolol Sandoz
and Seloken ZOC
formulations were
demonstrated to be very similar from the dissolution and
ethanol sensitivity point of view. These formulations were
disintegrated in the dissolution media to release their
pellets. The total exposed surface area of these pellets is
much larger than the surface area of Metoprolol GEA

Retard. Despite this fact the dissolution from these
formulations, as compared to Metoprolol GEA
Retard, is
much smaller in the absence of ethanol (Tables 2, 4 and 5).
The salt forms, i.e. metoprolol tatrate (Metoprolol GEA

Retard) and metoprolol succinate, which inuence the
solubility of metoprolol in the ethanol containing media will
also affect the release rate. However, the release was mainly
controlled by the gel matrix composition, as demonstrated
by the Metoprolol GEA
Retard formulation. This

formulation, despite its highly ethanol-soluble metoprolol
tartrate content, demonstrated a slightly decreased
dissolution rate in the presence of ethanol.
Metoprolol Sandoz
and Seloken ZOC
, both containing
metoprolol succinate, are considered to be inter-
exchangeable, because of the formulation and dissolution
prole similarity. It should also be noted that the pellets
from Metoprolol Sandoz
and Seloken ZOC
will leave the

stomach, where the ethanol concentration is at its highest
level after ethanol intake, faster than the Metoprolol GEA

Retard and Logimax
forte ER-tablets do. This may, to a

certain degree, compensate for the higher ethanol sensitivity
of Metoprolol Sandoz
and Seloken ZOC
.
A comparison between the dissolution proles at pH 1.2
and pH 6.8 indicated that the dissolution rate was not pH
dependent.
This study clearly demonstrated that the release of
metoprolol was affected by intake of ethanol and therefore it
might not be safe to drink alcohol in combination with this
type of pharmaceutical formulations.
5. CONCLUSION
This study highlighted the effect of ethanol at different
levels on the dissolution of metoprolol from four different
extended release formulations. The results illustrated the
difference between these formulations with regard to the
dissolution of metoprolol in the absence and presence
of ethanol. Small differences in dissolution proles were
observed in media of different pHs, i.e. pH 1.2 and 6.8,
for the investigated formulations. The alcohol sensitivity
[13] Beneld PB, Clissold SP, Brogden RN. Metoprolol
an updated review of its pharmacodynamic and
pharmacokinetic properties, and therapeutic efcacy,
in hypertension, ischemic heart disease and related
cardiovascular disorders. Drugs 1986;31:376-429.
[14] Oosterhuis B, Jonkman JHG, Kerkhof FA.
Pharmacokinetic and pharmacodynamic comparison of a
new controlled-release formulation of metoprolol with a
traditional slow-release formulation. European Journal of
Clinical Pharmacology 1988;33:15-18.
[15] Kondo T, Tokinaga N, Suzuki A et al. Altered
pharmacokinetics and metabolism of valproate after
replacement of conventional valproate with the slow-
release formulation in epileptic patients. Pharmacology &
Toxicology 2002;90:135-138.
[16] Dissolution test for solid dosage forms, general chapter
2.9.3. Ph. Eur. 6th edition. Strasbourg, France: Council of
Europe; 2008.
[17] Clarkes Isolation and Identication of Drugs in
pharmaceuticals, body uids, and post-mortem material.
2nd edition. London, UK: The Pharmaceutical Press;
1986:779-780.
[18] Metoprolol succinate, monograph 1448. Ph. Eur. 6th
edition. Strasbourg, France: Council of Europe; 2008 (vol 2).
[19] Metoprolol tartrate, monograph 1028. Ph. Eur. 6th
edition. Strasbourg, France: Council of Europe; 2008 (vol 2).
[20] Warning regarding co-ingestion of ethanol and
Metoprolol Sandoz [available at: http://www.fass.se/LIF/
home/index.jsp].
[21] Sandberg A, Ragnarsson G, Jonsson UE, Sjgren J.
Design of a new multiple-unit controlled-release formulation
of metoprolol-metoprolol CR. European Journal of Clinical
Pharmacology 1988;33:3-7.
was neither affected by the pH of dissolution medium. The
Metoprolol GEA
Retard and Logimax
forte dosage forms

showed high tolerance to ethanol while Metoprolol Sandoz

and Seloken ZOC
formulations were more sensitive to

ethanol.
The dissolution data from the rst two hours of metoprolol
release from the examined products, except Seloken ZOC
,
reveal that the librated amount in the presence of 20%
ethanol may not be a concern. The released amounts are less
than 24%, being approximately of the same magnitude as
that released from Metoprolol GEA
Retard in the absence of

ethanol under the same time period.
6. ABBREVIATION
ER: Extended-Release formulation; IR: Immediate release;
HPC: Hydroxypropyl cellulose; g: Microgram; RP-LC:
Reversed Phase Liquid Chromatography; API: Active
Pharmaceutical Ingredient.
7. REFERENCES
[1] Siepmann J, Peppas NA. Modeling of drug release from
delivery systems based on hydroxypropyl methylcellulose
(HPMC). Advanced Drug Delivery Reviews 2001;48:139-157.
[2] Tahara K, Yamamoto K, Nishihata T. Overall mechanism
behind matrix sustained release (SR) tablets prepared with
hydroxypropyl methylcellulose 2910. Journal of Controlled
Release 1995;35:59-66.
[3] Levina M, Vuong H, Rajabi-Siahboomi AR. The Inuence
of Hydro-Alcoholic Media on Hypromellose Matrix Systems.
Drug Development and Industrial Pharmacy 2007;33:1125-
1134.
[4] Rowe RC, Sheskey PJ, Weller PJ, editors. Handbook
of Pharmaceutical Excipients, Washington, USA:
Pharmaceutical press and the American pharmaceutical
association. London UK; 2003;238-300.
[5] Rajabi-Siahboomi AR, Jordan MP. Slow release HPMC
matrix systems. European Pharmaceutical Review
2000;5:21-23.
[6] Lindenbaum J, Mellow MH, Blackstone MO, Butler
VP Jr. Variation in biologic availability of digoxin from
four preparations. New England Journal of Medicine
1971;285:1344-1347.
[7] Walden N, Nicholls FA, Smith K J. The effect of ethanol
on the release of opioids from oral prolonged-release
preparations. Drug Development and Industrial Pharmacy
2007;33:1101-1111.
[8] Hrter D, Dressman JB. Inuence of physicochemical
properties on dissolution of drugs in the gastrointestinal
tract. Advanced Drug Delivery Reviews 2001;46:75-87.
[9] Lennerns H. Ethanol-drug absorption interaction:
potential for a signicant effect on the plasma
pharmacokinetics of ethanol vulnerable formulations.
Molecular Pharmaceutics 2009;6:429-1440.
[10] FDA Alert for Healthcare professionals. [available at:
http://www.fda.gov 2005].
[11] Regrdh C-G, Borg KO, Johansson R et al.
Pharmacokinetic studies on the selective 1-receptor
antagonist metoprolol in man. Journal of Pharmacokinetics
and Biopharmaceutics 1974;2:347-364.
[12] Robertson JIS. State of the art review: -blockade and
the treatment of hypertension. Drugs 1983;25:11-15.
Possible Ambiguities when Testing Viscosity
Possible Ambiguities when Testing Viscosity in
Compendial Monographs
Characterisation of Grades of Cellulose Ethers
E. Doelker
E. Doelker. Corresponding author: e-mail: eric.doelker@unige.ch. Department of Pharmaceutical Sciences, Ecole de Pharmacie Genve-Lausanne,
University of Geneva, University of Lausanne, Quai Ernest-Ansermet 30, 1211 Geneva 4, Switzerland.
1. INTRODUCTION
Often, pharmacopoeial monographs prescribe a viscosity
test. The aim of such a test can be manifold:
(i) simply to identify and characterise a pure material and
accessorily provide the user with this useful property
(e.g. Ph. Eur. monographs isopropyl myristate, liquid
paraffin, macrogols, medium-chain triglycerides);
(ii) to determine indirectly the molecular mass of a
polymeric compound (e.g. Ph. Eur. monographs
powdered cellulose, microcrystalline cellulose,
cellulose acetate, cellulose acetate phthalate,
methacrylic acid and ethyl or methyl acrylate
copolymers, povidone, poly(vinyl acetate), poly(vinyl
alcohol));
(iii) to assess primarily a functionality of a material and
possibly define its grade (e.g. Ph. Eur. monographs
acacia, guar, guar galactomannan, xanthan gum,
methacrylic acid and ethyl acrylate dispersion 30per
cent, poly(vinyl acetate) dispersion 30 per cent,
microcrystalline cellulose and carmellose sodium).
Materials corresponding to case (i) are Newtonian liquids,
and thus defining or measuring their viscosity is not
an issue, and is generally achieved through the one-
point capillary viscometer method. For case (ii), which
corresponds to very dilute organic solutions of materials, the
same rheological behaviour is observed and thus capillary
(or falling ball) viscometry is generally used. In monographs,
the molecular mass may or may not be deduced from
the intrinsic viscosity, for instance from a table (e.g. see
Ph.Eur. monograph microcrystalline cellulose). Case (iii)
is much more of a concern in terms of both definition
and measurement of the viscosity. In practice, an aqueous
solution or suspension of the material is prepared at a
given concentration and these dispersions exhibit complex
rheological behaviour. This will be illustrated by the example
of the water-soluble cellulose ethers.
There are three issues when specifying a viscosity test in a
monograph of such a polymeric compound, relating to:
the rheological behaviour of the solutions of such
cellulose derivatives;
the measuring system used to determine the viscosity;
and
the way the material was originally defined by the
manufacturer.
The rheological behaviour depends on the molecular mass
of a given cellulose ether, the concentration of the aqueous
solution, and to a limited extent the temperature. At 2 per
cent m/m, the concentration used in Ph. Eur. monographs,
ABSTRACT
The European Pharmacopoeia (Ph. Eur.) monographs for the water-soluble cellulose ethers require viscosity
determination, either in the Tests section or in the non-mandatory Functionality-related characteristics section.
Although the derivatives are chemically closely related and used for similar applications, the viscosity tests strongly differ.
Some monographs generically speak of the rotating viscometer method (2.2.10) and a xed shear rate (e.g. 10 s
-1
), which
would necessitate an absolute measuring system, while others recommend the capillary viscometer method for product
grades of less than 600 mPas and the rotating viscometer method and given operating conditions for grades of higher
nominal viscosity. Viscometer methods also differ between the United States Pharmacopeia/National Formulary (USP/NF)
and the Japanese Pharmacopoeia (JP) monographs. In addition, for some cellulose ethers the tests sometimes diverge from
one pharmacopoeia to the other, although the three compendiums are in a harmonisation process. But the main issue
is that the viscometer methods originally employed by the product manufacturers are often not those described in the
corresponding monographs and generally vary from one manufacturer to the other.
The aim of this study was therefore to investigate whether such a situation could invalidate the present pharmacopoeial
requirements. 2 per cent solutions of several viscosity grades of hydroxyethylcellulose, hypromellose and methylcellulose
were prepared and their (apparent) viscosity determined using both relative and absolute viscometer methods. The
viscometer method used not only affects the measured viscosity but experimental values generally do not correspond
to the product nominal viscosities. It emerges that, in contrast to Newtonian solutions (i.e. those of grades of up to
ca. 50 mPas nominal viscosity), some of the viscometer methods currently specied in the monographs are not able
unambiguously to characterise the grades exhibiting non-Newtonian behaviour. It is also concluded that, unless the
various manufacturers redene their product viscosity grades using a single compendial test, two strategies could be
adopted, both based on the operating conditions specied in the labeling (i.e those of the manufacturer), the test appearing
either in the mandatory section if this is acceptable to the pharmacopoeia (like in some USP/NF monographs) or, for the
Ph. Eur., in the Functionality-related characteristics section.
KEYWORDS
Viscosity, shear rate, Newtonian, viscometer, rheometer, cellulose ethers, carmellose sodium, hydroxyethylcellulose,
hydroxypropylcellulose, methylhydroxyethylcellulose, hypromellose, methylcellulose.
aqueous solutions of cellulose derivatives behave like
Newtonian fluids for grades up to ca. 50 mPas, i.e. their
viscosity does not depend on the shear rate. The viscosity
is referred to as absolute viscosity. In contrast, solutions of
higher grades display a non-Newtonian behaviour, i.e. the
viscosity - here called apparent viscosity - depends on the
shear rate, except at very low shear rates. With this type of
derivatives, the viscosity actually decreases with increasing
shear rate (pseudoplastic or shear-thinning behaviour);
it can also be affected by the shearing time (thixotropy).
Figure 1 shows the rheogram (flow curve) of both a
Newtonian and a non-Newtonian (pseudoplastic) fluid.
Viscometers or rheometers together with their measuring
systems (measurement geometries) are distinguishable from
two technical standpoints [1].
(i) Possibility to vary the shear rate or not
- Apparatuses with fixed shear rate or one-point
apparatuses, e.g. capillary (Ostwald, Ubbelohde,
Cannon-Fenske) and falling ball (Hoeppler)
viscometers.
- Apparatuses with variable shear rate or multipoint
apparatuses, e.g. rotating and oscillatory
viscometers with coaxial, cone/plate or plate/plate
geometry.
(ii) Existence or not of defined flow within the measuring
system
- Measuring systems (geometries) with non-uniform
(not well-defined) shear rate, e.g. capillary and
falling ball viscometers, as well as rotating and
oscillatory viscometers with unspecified coaxial
geometry (small bobs in large cups) or with plate/
plate geometry; in such a case, relative viscosities
are measured, which are not comparable from one
apparatus to another.
- Measuring systems (geometries) with uniform
(well-defined) shear rate, e.g. rotating and
oscillatory viscometers with specified coaxial
geometry or cone/plate geometry; here, absolute
viscosities are measured, which are obtainable with
any such measuring system and will be the same at
a given shear rate.
It emerges that any viscometer/rheometer method (relative
or absolute) can be used without ambiguity to characterise
solutions (and thus grades) of materials showing Newtonian
behaviour (low-viscosity cellulose ether grades). However,
attention should be paid to the selection of the equipment
and operating conditions, which have to be adapted to
the viscosity range to be measured in order to avoid flow
turbulence and offsetting [2]. As for non-Newtonian liquids
(high-viscosity cellulose ether grades), the apparent viscosity
will be the same at a given shear rate (within experimental
variation) when measured using any absolute viscometer/
rheometer method, again provided the right conditions
are chosen. In contrast, the apparent viscosity will differ
according to the relative viscometer/rheometer method
used, even when careful calibration has been performed or,
for instance, the same rotational speed has been selected;
data are only comparable when obtained using the same
measuring geometry and operating conditions.
The third issue deals with the way the manufacturers
originally defined their material viscosity grades for the
end-users. With regard to water-soluble cellulose ethers,
it can be observed that only relative viscometry methods
were used and that even the temperature is variable from
one manufacturer to the other (Table 1). Thus, as an
example, the various hypromellose grades were defined
using either capillary viscometry at 20 C (Methocel
TM
A,
E, F, K), Hoeppler falling ball viscometry at 20 C (Tylose
MO), rotating (Brookfield LVT) viscometry at 25 C
(Benecel MP) or Ubbelohde or rotating viscometry at 20 C
(Pharmacoat, Metolose SH). As for hydroxyethylcellulose
products, they were defined using either Hoeppler falling
ball viscometry at 20 C (Tylose H) or rotating (Brookfield
LVT) viscometry at 25 C (Natrosol).
Regarding water-soluble cellulose ethers, the role played
by the type of rotating viscometer equipment used on the
measured viscosity has been examined for hypromellose
by Iza and Grossiord [2]. However, the viscosity data were
not discussed with reference to the viscometer method
and requirements specied in the specic pharmacopoeial
monographs, nor to the way the various manufacturers
dene their products. It is thus the scope of this article to
address these two issues and, in addition, to assess the effect
of the measurement temperature (20 C or 25 C) on the
measured viscosity, as some manufacturers use the latter to
characterise their products. The viscosity of various grades
of hydroxyethylcellulose, hypromellose and methylcellulose
will be determined for this purpose, using different
viscometer methods.
Figure 1 - Shear stress () vs. shear rate (D) graphs for Newtonian (left) and pseudoplastic (right) liquids. The apparent
viscosity (
app
) at shear rate D corresponds to the tangent of
2. EXPERIMENTAL
2.1. Materials
The Natrosol, Methocel
TM
and Metolose samples were
gifts of Ashland/Aqualon, Dow Chemical and Shin-Etsu
Chemical, respectively.
2.2. Methods
Hydrogels at 2.0 % m/m polymer were prepared based
on a dry basis. Viscosities were determined using the
following viscometers/rheometers: type 531 20 II Ubbelohde
viscometer mounted on the automatic measurement system
AVS 310 (Schott Instruments); Brookeld viscometers
model DV-II+ (Brookeld Engineering Laboratories,
Inc., Stoughton, USA); Haake RheoVisco L2 viscometer
(Haake GmbH, Karlsruhe, Germany); Epprecht-Rheomat
viscometers models 15, 15 FC and 15 T-FC (Contraves,
Zrich, Switzerland); and a stress-controlled Bohlin CVO
120 High-Resolution rheometer (Bohlin Reologi GmbH,
Mhlacker, Germany).
Two types of measuring systems (measurement geometries)
were used:
relative measuring systems: LV spindles # 3 or 4
Manufacturer Product names Method Temperature
Dow Chemical Company Methocel
TM
Ubbelohde viscometer
(ASTM D2363)
20 C
SE Tylose GmbH & Co. KG Tylose
Hoeppler falling ball
viscometer
20 C
Ashland/Aqualon
Natrosol, Klucel,
Blanose, Benecel
Rotating (Brookeld
LVT) viscometer
25 C (20 C for
Benecel)
Shin-Etsu Chemical Co., Ltd.
Pharmacoat,
Metolose
Ub belohde or rotating
viscometer
20 C
Nippon Soda Co., Ltd. Nisso HPC
Rotating (Brookeld
LVT) viscometer
20 C
with Brookeld DV-II+ or RheoVisco L2 viscometer,
cylindrical spindles # MS A to D with Rheomat 15 or
15 T-FC;
absolute measuring systems: co-axial cylinders
according to DIN 53019 (measurement bob diameter
14 or 24 mm) with Rheomat 15-FC, and CP 1/40 LS
cone/plate measuring system with Bohlin CVO 120
high-resolution rheometer.
The corresponding shear rates for the measuring system
were taken from the manufacturers technical information.
Apparent viscosities at given shear rates were obtained by
interpolation as described in the Ph. Eur. monographs.
2.3. Viscometers and conditions prescribed in the
Ph. Eur., USP and JP
The Ph. Eur., the USP and the JP each provide a general
monograph on viscosity [3-5] as well as specic monographs
for six water-soluble cellulose ethers. In the Ph. Eur. and
USP a viscosity test is systematically required, in contrast
to the JP, and in any case, the viscometer methods and
operating conditions specied are highly variable for
historical reasons, as shown in Table 2.
Table 1 - Some of the main manufacturers of cellulose ethers and the methods used to characterise the viscosity grades
Table 2 - Types of viscometers and conditions specied in the Ph. Eur., USP and JP monographs of
water-soluble cellulose derivatives (March 2010 status)
Derivative Ph. Eur. 6 USP 32/NF 27 JP XV
Carmellose sodium
Carboxymethylcellulose
sodium
Rotating (2.2.10),
20 C, 10 s
-1
[6]
Rotating <911>,
25 C [12]
No test [18]
Hydroxyethylcellulose Rotating (2.2.10), 25 C, < 100 mPas:
100 s
-1
, 100-20000 mPas: 10 s
-1
,
> 20000 mPas: 1 s
-1
[7]
Conditions specied in
the labeling [13]
No monograph
Hydroxypropylcellulose Rotating (2.2.10),
20 C, 10 s
-1
[8]
Conditions specied in
the labeling [14]
No test [19]
Methylhydroxyethylcellulose
Hymetellose
Rotating (2.2.10),
20 C, 10 s
-1
[9]
Rotating <911>,
20 C, 10 s
-1
[ 15]
No monograph
Hypromellose FRC
< 600 mPas: capillary (2.2.9), 20 C
600 mPas: rotating, single-cylinder
and rpm specied (2.2.10), 20 C
[10]*
< 600 mPas:
capillary, 20 C
600 mPas: rotating,
single-cylinder and
rpm specied, 20 C
[16]
< 600 mPas:
capillary, 20 C
600 mPas: rotating,
single-cylinder and
rpm specied, 20 C
[20]
Methylcellulose FRC
< 600 mPas: capillary (2.2.9), 20 C
600 mPas: rotating, single-cylinder
and rpm specied (2.2.10), 20 C [11]
Capillary, 20 C
(Ubbelhode) [17]
< 600 mPas:
capillary, 20 C
600 mPas: rotating,
single-cylinder and
rpm specied, 20 C
[21]
* The rotating viscometer method together with a shear rate of 10 s
-1
were the conditions specied in Ph. Eur. 5 for all viscosity grades.
3. RESULTS AND DISCUSSION
Two sets of data are reported and discussed to show the
difculties met when determining the apparent viscosity of
non-Newtonian solutions of cellulose ethers.
3.1. Effect of viscometer method, shear rate
and temperature on the apparent viscosity of
hydroxyethylcellulose grades
To illustrate the typical rheological behaviour of water-
soluble cellulose ethers, Figures 2 and 3 display the ow
curves (rheograms, shear stress vs. shear rate graphs) and
the viscosity curves (viscosity vs. shear rate graphs) at 25C
of 2 % solutions of two grades of hydroxyethylcellulose,
using the Epprecht-Rheomat 15 rheometer with the relative
cylindrical measuring systems A and C, respectively.
The linear ow prole of the low-viscosity
hydroxyethylcellulose grade (Natrosol 250 L of 15 mPas
nominal viscosity) is typically indicative of a Newtonian
behaviour (Figure2, left), with constant viscosity even
though the shear rate changes (Figure 3, left). In contrast,
the medium-viscosity grade (Natrosol 250 M of 6000 mPas
nominal viscosity) shows pseudoplastic (shear-thinning)
behaviour (Figure 2, right) with decrease of viscosity
with increasing shear rate (Figure 3, right). Solutions of
Natrosol 250 G (350 mPas nominal viscosity) and of
Natrosol 250 HX (30 000 mPas nominal viscosity) display
the same behaviour as Natrosol 250 M (not shown).
The rheograms were very close at 20 C (not presented
here). Similar results were obtained by Opota et al. for
hydroxypropylcellulose solutions [22]. At 2 % m/V, a
Newtonian behaviour was observed at 25 C for Klucel LF
(ca. 8 mPas typical viscosity), which became pseudoplastic
for Klucel GF and HF (ca. 270 mPas and 5250 mPas
typical viscosity, respectively). The conclusion was the same
with hypromellose solutions at 20 C: Newtonian behaviour
with Pharmacoat 606 (6 mPas nominal viscosity) and
pseudoplastic behaviour with Metolose 90 SH 15 000
(15000 mPas nominal viscosity). A transition behaviour
was observed with Metolose 60 SH 50 (50 mPas nominal
viscosity) [2].
The apparent viscosities, as interpolated at the shear rate
values specied in the hydroxyethylcellulose Ph.Eur.
monograph [7], are reported in Table 3. Data were obtained
using the Ubbelohde capillary viscometer method, a
Rheomat 15 viscometer equipped with the cylindrical
spindles relative measuring system and a Bohlin CVO
rheometer equipped with the CP 1/40 LS cone/plate
measuring system data, at both 20 C and 25 C. The
temperature of 25 C was selected for being that prescribed
by the hydroxyethylcellulose PhEur. monograph and also
used by the Ashland/Aqualon company for characterising
their products, while 20 C is the temperature used by the
SE Tylose company.
For the low-viscosity grade Natrosol 250 L with Newtonian
behaviour, it can be observed that, at the shear rate
specied in the Ph. Eur. monograph (100 s
-1
), the viscosities
measured at 25 C with the rotational viscometer method
meet the Ph.Eur. 75-140 % interval requirement, both
with the relative and the absolute equipments. However, the
rotational measuring systems used may technically not be
appropriate for determining such low viscosities, in contrast
to the capillary viscometer method that appears very suitable
(even though the shear rate is unknown), similarly to that
which was observed by Iza and Grossiord [2].
Figure 2 - Rheograms of 2 % solutions of Natrosol 250 L (left) and Natrosol 250 M (right) at 25 C
Figure 3 - Variation of viscosity with shear rate of 2 % solutions of Natrosol 250 L (left)
and Natrosol 250 M (right) at 25 C
Regarding the other Natrosol grades, which exhibit
non-Newtonian behaviour, the viscosities measured at the
specied shear rates comply with the Ph. Eur. 75 - 140%
limits for one of them (Natrosol 250 HX) and not for
the other two (Natrosol 250 G and Natrosol 250 M),
whatever the rotating viscometer used. For Natrosol 250G
the viscosity determined using the capillary viscometer
method does not fall within the nominal viscosity limits (no
measurement could be made with the Ubbelohde viscometer
used for the higher viscosity Natrosol grades). More
generally, when comparing the viscosities measured in the
conditions prescribed in the hydroxyethylcellulose Ph.Eur.
monograph [7] with the product nominal viscosities,
which were established by the manufacturer by means of
another viscometer method (the rotating Brookeld LVT
viscometer), it is important to realise that the shear rates
prescribed in the monograph were originally proposed
by trial and error. The shear rate dependent behaviour
of this type of derivatives is illustrated by the viscosities
interpolated at three shear rates obtained from the Bohlin
measurements (Table 4). It emerges that the best matching,
although not always satisfactory, is in fact obtained using the
shear rates specied in the hydroxyethylcellulose Ph. Eur.
monograph. Note for example that if a shear rate of 10 s
-1

would have been used for the Natrosol 250 HX sample, the
product could have been classied as a 10 000 mPas grade.
Another example of partly unsatisfactory matching is given
by the determination of the viscosity of 2 % m/m solutions
of some Tylose H products (SE Tylose GmbH & Co. KG) by
means of the method used by the Aqualon/Ashland company
for dening its various Natrosol grades, i.e. a Brookeld
LV type viscometer and dened spindles and rotation
speeds [23]. No matching could be observed between the
nominal viscosity of a Tylose H grade (300 mPas, based
on Hoeppler falling ball viscometer measurement of a
1.9 % m/m solution) and the measured value at 20 C
for the 2 % m/m solutions using a Brookeld LV and the
conditions specied by the Aqualon/Aashland company
(spindle 2, 60 rpm). In contrast, the value obtained using the
Brookeld LV and the conditions specied by the Aqualon/
Aashland company (spindle 4, 60 rpm) did fall within the
75-140 % limits of the nominal viscosity of a 4000 mPas
Tylose H grade. This conrms that no single universal
viscometer method can be used for characterising products
purchased from different manufacturers. Consequently,
the prescription of the hydroxyethylcellulose USP 32/NF 27
monograph [13] recommending the use of the conditions
specied in the labeling is probably the best way.
Finally, it appears that the test temperature only slightly
affects the viscosity. In any case, such little variation would
not lead to a possible redenition of the viscosity grades
if a temperature different from that used by the product
manufacturer would be prescribed.
3.2. Comparison of viscosity data regarding methyl-
cellulose and hypromellose grades using various
viscometer methods
The second set of data relates to apparent viscosities of ve
grades of hypromellose (substitution types 2910 and 2208)
and a methylcellulose grade. All products had a nominal
Natrosol
grade
Nominal
viscosity
(mPas)
Ph. Eur.
75-140%
limits (mPas)
Apparent viscosity (mPas) at
1 s
-1
10 s
-1
100 s
-1
250 L 15 11.25 - 21.0 - 24 15
250 G 350 262.5 - 490 - 238 186
250 M 6000 4500 - 8400 10070 4344 1137
250 HX 30000 22500 - 42000 38040 9840 1817
Table 4 - Apparent viscosities at 25 C of 2 % m/m solutions of hydroxyethylcellulose grades extrapolated at three different
shear rates (Bohlin CVO 120 with CP 1/40 LS)
Table 3 - Effect of the measuring system, shear rate and temperature on the apparent viscosities of 2 % solutions
of various hydroxyethylcellulose grades
Natrosol
grade
Nominal
viscosity*
(mPas)
Ph. Eur.
75-140 %
limits
(mPas)
Ubbelohde (ab-
solute or relative
viscosities)
(mPas)
Shear
rate
(s
-1
)
Rheomat 15 with cylindrical
spindles (relative viscosities)
(mPas)
Bohlin CVO with CP
1/40 LS (absolute
viscosities)
(mPas)
20 C 25 C 20 C 25 C 20 C 25 C
250 L 15 11.25 -
21.0
15 13 100 # A: 17 # A: 14 19 15
250 G 350 262.5 -
490
227 182 10 # B: 285 # B: 218 279 238
250 M 6000 4500 -
8400
- - 10 # C: 4495 # C: 3779 4921 4344
250 HX 30 000 22 500 -
42 000
- - 1 # D: 35 722 # D: 34 817 50 840 38 040
* Typical Brookeld LV viscosities at 2 % and 25 C, as dened by the manufacturer.
viscosity of more than 600 mPas. The apparent viscosities
were determined at 20 C according to the current Ph. Eur.
requirement for the hypromellose and methylcellulose
monographs [10, 11], i.e. using a single-cylinder type spindle
viscometer (the Brookeld LVDV-II+ or the LV Brookeld-
type Haake RheoVisco L2 viscometer), and the spindles and
rate of revolution specied in the monographs. Note that
the conditions are identical in the hypromellose USP 32/NF
27 and JP XV monographs [16, 20], as shown in Table 2. In
the case of methylcellulose, the same conditions also apply
to the Ph. Eur. and JP monographs [11, 21], but not to the
USP/NF monograph [17], where the capillary (Ubbelhode)
viscometer method is precribed. The viscosities were
also measured by means of a relative measuring system
(cylindrical spindles tted to a Rheomat 15 T-FC viscometer)
as well as an absolute measuring system (co-axial cylinders
according to DIN 53019 tted to a Rheomat 15-FC
viscometer). In the latter cases viscosities were interpolated
at 10 s
-1
, the shear rate recommended in the hypromellose
and methylcellulose Ph. Eur. 5 monographs for all viscosity
grades. Data are summarised in Table 5. The table also
lists the nominal viscosities as originally dened by the
manufacturers using the test method still valid in the USP
29 monographs (Ubbelohde capillary viscometer) as well
as the recently redened nominal viscosities according to
the current harmonised Ph. Eur., USP and JP monographs
(rotating viscometer with single-cylinder spindles and rate
of revolution specied). Note that only the specications
of the Methocel
TM
products have been redened [24].
Viscosity data are provided in the latter document justifying
the redenition of these products. A redenition of the
Metolose products was apparently not necessary as the
grades could have been historically characterised using the
Brookeld LV viscometer (see Table 1).
Satisfactory matching with the nominal viscosities is
observed when the apparent viscosity has been determined
according to the conditions specied in the current
methylcellulose Ph. Eur. and JP monographs and in the
hypromellose Ph. Eur., USP and JP monographs (i.e.
Brookeld LV), although the value found for the very high-
viscosity derivative (Methocel
TM
K100M) was borderline
considering both sets of 75-140 % limits.
Examining now the viscosities determined with the relative
or the absolute measuring system and interpolated at 10 s
-1

(former conditions) it appears that none of the values are
in line with the nominal viscosities, whatever the limits
considered. It comes out that these results justify the
change of the viscosity determination method to the test
presently described in the harmonised methylcellulose and
hypromellose monographs (see Table 2).
4. CONCLUSION
The results presented here lead to concluding remarks in
terms of both technical and compendial standpoints. Some
of them were already drawn by Iza and Grossiord [2].
Technically, the determination of the viscosity of Newtonian
solutions is not an issue and in principle both absolute and
* Nominal viscosities established using the USP 29 test method (Ubbelohde capillary viscometer)
** Nominal viscosities newly established using the Ph. Eur./USP/JP harmonised test method of the hypromellose Ph. Eur., USP and JP monographs
(see Table 2); the same conditions are valid for the present methylcellulose Ph. Eur. and JP monographs but not for the methylcellulose USP
monograph, where the Ubbelohde capillary viscometer method is still specied
*** Viscosities kindly determined by J.-L. Grossiord (University of Paris-Sud, France), using the LV Brookeld-type Haake RheoVisco L2 viscometer
Table 5 - Apparent viscosities (mPas) at 20 C of 2 % solutions of a methylcellulose grade and various hypromellose
grades, as determined using the present harmonised Ph. Eur., USP/NF and JP test method
and the former Ph. Eur. 5 test method at 10 s
-1
Grade Original
nominal
viscosity*
75-140 %
limits using
the USP 29
test method*
Redened
nominal
viscosity**
Redened
75-140 %
limits**
Rheomat 15
T-FC with
cylindrical
spindles
(relative
viscosities
at 10 s
-1
)
Rheomat
15 FC with
DINco-axial
cyliders
(absolute
viscosities
at 10 s
-1
)
Brookeld
LVDV-II+
(relative
viscosities)
Methocel A15C 1500 1125 - 2100 1730 1298 - 2422 Not measured # 25: 1615
# 3/12 rpm:
1530
Methocel
TM
E4M
(type 2910)
4000 3000 - 5600 3550 2663 - 4970
# B: 1790
# C: 1741
# 25: 1964
# 4/60 rpm:
4570
Metolose 60 SH 4000
(type 2910)
4000 3000 - 5600 - - # C: 1908 # 25: 2152
# 4/60 rpm:
3730
Methocel
TM
K4M
(type 2208)
4000 3000 - 5600 3550 2663 - 4970 # C: 2347 # 25: 2992
# 4/60 rpm:
4330***
Methocel
TM
K15M
(type 2208)
15000 11250 - 21000 17700 13275 - 24780 # C: 5495 # 14: 5920
# 4/6 rpm:
14100***
Methocel
TM
K100M
(type 2208)
100000
75000 -
140000
97000
72750 -
135800
# C: 14655
# D: 13308
# 14: 12916
# 4/3 rpm:
75500***
relative measuring systems can be used, as shown here
for a low-viscosity hydroxyethylcellulose grade. However,
certain criteria have to be fullled in order not to generate
inadequate ow conditions (offsetting or turbulence). The
situation is more complicated for non-Newtonian uids
as the apparent viscosity measured depends on the shear
rate, which can only be calculated for absolute measuring
systems (shear rate values provided for relative measuring
systems are only approximative). A consequence is that
the viscosity as specied in the carmellose sodium,
hydroxyethylcellulose, hydroxypropylcellulose and
the methylhydroxyethylcellulose/hymetellose Ph. Eur.
monographs (and formerly in the hypromellose Ph. Eur.
monograph), i.e. at a given reference shear rate (e.g.
10s
-1
), strictly cannot be determined by means of relative
measuring systems, those commonly found in user
laboratories. It is also shown that most of the extrapolated
viscosities obtained for the present hydroxyethylcellulose,
hypromellose and methylcellulose samples with the
various measuring systems do fall in the 75-140 % nominal
viscosity ranges. Similarly, the simple recommendation
of utilising a rotating viscometer equipped with a single
cylinder measuring system at a given rotational speed is not
sufcient, unless the system is explicitly described. In this
respect, the target viscosities of the methylcellulose and
hypromellose products using the Brookeld LV viscometer
and the current pharmacopoeial conditions were reached,
but it is essential to recall that these conditions have been
proposed on a trial and error basis, trying to correlate the
measurements with the nominal viscosities as originally
dened by the manufacturer.
Complex issues also exist from a compendial viewpoint.
First the methods specied in the monographs most
of the time differ from those originally used by the
manufacturers to dene their viscosity product grades.
Second, the pharmacopoieal method generally varies
from one monograph to the other even though the water-
soluble cellulose ethers concerned are chemically very
similar. This originates from the fact that three different
pharmacopoeias were involved in the international
harmonisation process, without forgetting the impact of
the national product manufacturers. A third remark deals
with the status of the viscosity determination in the various
compendial monographs. From a scientic point of view
the viscosity, which directly derives from the molecular
mass and the conformation of the polymer chains, could
be considered part of the product identication. However,
as this characteristic is not a safety issue viscosity
determination appears in the pharmacopoeia monographs
either in the mandatory test section or, in the case of the
methylcellulose and hypromellose Ph. Eur. monographs,
in the so-called non-mandatory Functionality-related
characteristics section. Finally, if a mandatory test is
specied, it should be well dened and applicable whatever
the product manufacturer, at least according to the Ph.
Eur. concept. However, in the present situation, this
requirement cannot be fullled because the viscosity
grades have been characterised using different measuring
systems. This is valid for both absolute and relative methods.
The monographs of the three derivatives studied here are
illustrative (see Tables 1 and 2). In the hydroxyethylcellulose
Ph. Eur. monograph it is specied to determine the viscosity
at a given shear rate (100 s
-1
, 10 s
-1
or 1 s
-1
), simply using a
rotating viscometer method at 20 C, but the commercially
available hydroxycellulose grades were characterised by the
main manufacturers either by means of a Brookeld LV
viscometer or a Hoeppler falling ball viscometer. In the USP
monograph, it is stated to use the conditions specied in
the labeling, which are dependent on the manufacturer. As
for hypromellose, the monographs of three pharmacopoeias
specify the capillary method for derivatives of nominal
viscosity of up to 600mPas and the rotating viscometer
method with specic operating conditions for derivatives of
more than 600 mPas (the same is true for the Ph. Eur. and
JP methylcellulose monographs). This probably corresponds
to those used by one main hypromellose manufacturer,
but not by the other who dened its viscosity grades using
specically ASTM D2363 Ubbelohde viscometers. Note also
that the 600 mPas boundary relies on the practical limit
when using common capillary viscometers and does not
coincide with the transition from Newtonian and non-
Newtonian behaviour.
From this analysis, it can be concluded that the viscosity
determination test present in the various water-soluble
cellulose ether monographs should be reconsidered, having
in mind that the measurement could have three aims:
(i) unambiguous identication of different grades of a
given derivative;
(ii) quality control purposes, in particular verication
of batch-to-batch consistency in the customer-supplier
relationship; and
(iii) provision of specic information for the intended uses
of the product grade.
In practice, all water-soluble cellulose ethers are similarly
concerned with these aims. However, for a given derivative,
viscosity of an aqueous solution could not be of relevance
for some intended uses (aim iii). This is for instance the
case when this type of derivative is used as a binder for
granulation or as a lm-forming agent for solid dosage
forms. In contrast, such determination is relevant when
the product is utilised as a viscosifying agent for solutions,
hydrogels or suspensions, as well as a carrier for compressed
hydrophilic matrices.
Coming now to the necessary harmonisation of the
monographs, three different approaches could be envisaged:
either the present situation of heterogeneity in
measuring conditions is maintained; or
a single determination method is proposed for all
water-soluble cellulose ether monographs; or
reference is made to the manufacturers method.
The rst approach is not satisfactory because, as already
said, often the conditions specied in the monographs
differ from those used by the manufacturers to dene their
viscosity products. The second possibility is scientically
sound but the same hurdle as in the rst approach still
holds. Moreover, it would be necessary to choose between an
absolute measuring system (an equipment not commonly
available) when specifying a dened shear rate (e.g. 10s
-1
),
a relative measuring system (e.g. Brookeld type) that
would obviously not be universal, and a generic method
(e.g. rotational viscometry) that would not be devoid of
ambiguity. To the reviewer, because of historical reasons
the third approach of using the conditions specied in
the labeling is the most logical one, as it prevents the
various manufacturers from redening their product
viscosity grades according to a single compendial test (as
recently done by a hypromellose manufacturer [24]), even
though several measuring equipments might be needed
if the end user purchases the product from different
manufacturers. This strategy is the one adopted in the
present hydroxyethylcellulose and hydroxypropylcellulose
USP/NF monographs. In the international harmonisation
process, still in progress, it could be proposed for the other
monographs of the water-soluble cellulose ethers. The
test could be included in the mandatory core-text of the
monograph if simply referring to the conditions specied in
the labeling is acceptable to the European and the Japanese
pharmacopoeias or, for the Ph. Eur., in the functionality-
related characteristics section.
5. REFERENCES
[1] Schramm G. A practical approach to rheology and
rheometry. Karlsruhe, Germany: Gebrueder Haake GmbH;
1994.
[2] Iza M, Grossiord JL. Comparative analysis of the ow
properties of hydroxypropylmethylcellulose. Pharmeuropa
1997;9(3):566-9.
[3] Viscosity determination, general chapter 2.2.10. Ph. Eur.
6th Edition. Strasbourg, France: Council of Europe; 2008.
[4] Viscosity, general chapter <911>. USP 32, NF 27.
Rockville, USA: United States Pharmacopeial Convention,
Inc; 2009.
[5] Viscosity Determination, general chapter 2.53. JP XV.
Tokyo, Japan: The Ministry of Health, Labour and Welfare;
2006.
[6] Carmellose sodium, monograph 0472. Ph. Eur. 6th
Edition. Strasbourg, France: Council of Europe; 2008(vol 2).
[7] Hydroxyethylcellulose, monograph 0337. Ph. Eur. Suppl.
6.8. Strasbourg, France: Council of Europe; 2010.
[8] Hydroxypropylcellulose, monograph 0472. Ph. Eur. 6th
Edition. Strasbourg, France: Council of Europe; 2008(vol 2).
[9] Methylhydroxyethylcellulose, monograph 0346. Ph.
Eur. 6th Edition. Strasbourg, France: Council of Europe;
2008(vol 2).
[10] Hypromellose, monograph 0348. Ph. Eur. 6th Edition.
Strasbourg, France: Concil of Europe; 2008(vol 2).
[11] Methylcellulose, monograph 0345. Ph. Eur. 6th Edition.
Strasbourg, France: Concil of Europe; 2008(vol 2).
[12] Carboxymethylcellulose sodium. USP 32, NF 27.
Rockville, USA: United States Pharmacopeial Convention,
Inc; 2009.
[13] Hydroxyethyl cellulose. USP 32, NF 27. Rockville, USA:
United States Pharmacopeial Convention, Inc; 2009.
[14] Hydroxypropyl cellulose. USP 32, NF 27. Rockville, USA:
[15] Hymetellose. USP 32, NF 27. Rockville, USA: United
States Pharmacopeial Convention, Inc; 2009.
[16] Hypromellose. USP 32, NF 27. Rockville, USA: United
[17] Methylcellulose. USP 32, NF 27. Rockville, USA: United
[18] Carmellose sodium. JP XV. Tokyo, Japan: The Ministry
of Health, Labour and Welfare; 2006.
[19] Hydroxypropylcellulose. JP XV. Tokyo, Japan: The
Ministry of Health, Labour and Welfare; 2006.
[20] Hypromellose. JP XV. Tokyo, Japan: The Ministry of
Health, Labour and Welfare; 2006.
[21] Methylcellulose. JP XV. Tokyo, Japan: The Ministry of
Health, Labour and Welfare; 2006.
[22] Opota O, Maillols H, Acquier R et al.
Comportement rhologique des solutions aqueuses
dhydroxypropylcellulose. Inuence de la concentration et de
la masse molaire. Pharm Acta Helv 1988;63(1):26-32.
[23] Stern R. Private communication. January 4, 1999.
[24] Wright J. Dowwolf Cellulosics information sheet;
2007 October 8 [available at: http://www.dow.com/
PublishedLiterature/dh_0098/0901b80380098342.
pdf?lepath=methocel/pdfs/noreg/192-01430.
pdf&fromPage=GetDoc, accessed 2010 March].
100 Pharmeuropa Bio & Scientic Notes, 2010-2
1. PHARMEUROPA AND PHARMEUROPA BIO &
SCIENTIFIC NOTES: AIMS AND SCOPE
Pharmeuropa and Pharmeuropa Bio & Scientific Notes
(Bio & SN) are aimed at providing information on the
work of the European Pharmacopoeia Commission and of
the Biological Standardisation Programme. They provide
feedback from the users of the European Pharmacopoeia,
hence it publishes papers by individual contributors on
any aspect of pharmaceutical quality control. Articles
submitted for publication in Pharmeuropa and Pharmeuropa
Bio&SN must have a clear, direct link with the work of the
pharmacopoeias, especially the European Pharmacopoeia.
2. SUBMITTING ARTICLES TO PHARMEUROPA/
PHARMEUROPA BIO & SN
It is assumed that the copyright of material submitted is
that of the contributor or that the necessary clearances have
been obtained and that, in general, material submitted has
not been offered simultaneously to another publication.
Authors are required to clearly state if material has been
submitted to another publication. In submitting an article
for publication, it is assumed that all authors of the article
have given their consent to publication.
2.1. Where to send articles
Articles should be sent (for the attention of the Editors of
Pharmeuropa/Pharmeuropa Bio & SN) by e-mail to
info@edqm.eu or by post to the following address:
European Directorate for the Quality of Medicines &
HealthCare
Department Publications & Multimedia
7 alle Kastner
CS 30026
F-67081 Strasbourg, France
2.2. Requirements
2.2.1. General requirements
If submitting an article by post, please send two complete
hard copies accompanied by an electronic version. Articles
are typewritten (Times font, font size 12) on one side only
of A4 size paper, double spaced and with at least a 2.5 cm
margin all around. Electronic versions of texts are preferably
in Word or RTF format (PDF format is unacceptable).
Authors are encouraged, wherever possible, to avoid
footnotes and to include the information in the main body
of the text. Greek letters are written in full in the text in
capitals (for capital Greek letters) or in lower case (for lower
case Greek letters) and have a full stop before and after the
letter. This is illustrated in the following examples: .beta.
.alpha. .SIGMA. .OMEGA. The Greek symbol will be printed
in the final version of the text.
2.2.2. Requirements for tables and figures
Authors are required to include all original files for figures
and tables. Numerical data and tables are acceptable in Word
or Excel format. All figures are to be saved in individual files,
preferably in EPS files, or in TIFF format (with a resolution
of 600 dpi). PDF format is only acceptable with vectorised
graphic data. Files are to be submitted with high quality
graphics. Please note that bitmap images embedded in PDF
or Microsoft Word files and formats intended for the internet
(i.e. GIF, JPEG, PNG, etc.) will not be accepted.
Text in figures must be in Arial, Times New Roman or
Helvetica font (font size 8). If text is to be added to a figure
(e.g. key to symbols) it should be supplied on a separate page
or included as part of the figure legend. In no case should
there be handwriting on the original figure. Black and white
figures are preferred, grey scale is only acceptable at an
intensity of 25 %. Authors are asked to avoid reverse type
where possible (i.e. white lettering on a dark background)
and a shaded background (especially if a suitable electronic
version cannot be supplied).
2.2.3. Figure dimensions
Authors are asked to submit figures of an appropriate
quality and size for publication. Figures have a width of
either 230pt (corresponding to about 8 cm) if intended for
publication across 1 column of Pharmeuropa/Pharmeuropa
Bio & SN or 480 pt (corresponding to about 17 cm) if
intended for publication across 2 columns of Pharmeuropa/
Pharmeuropa Bio & SN.
3. PRESENTATION OF ARTICLES
Two types of manuscript will be considered for publication:
Articles in Pharmeuropa Bio & SN: articles describing
one or more methods of analysis applicable to
pharmaceutical quality control in the broadest sense or
the result of analysis of pharmaceutical products;
Readers Tribune: discussion articles and letters
commenting on the policy or decisions of the European
Pharmacopoeia Commission or on previously published
papers.
Further instructions pertaining to these two types of
manuscript are given below.
3.1. Articles in Pharmeuropa Bio & SN
Articles are accepted in either English or French, however
if the text is written in French it must be accompanied by
a comprehensive summary in English. Articles written in
French will not be translated. As a general rule, articles
should not exceed 3500 words and should contain a
reasonable number of figures and tables. Articles are written
in a clear, concise and coherent manner and are well
structured. Poorly written articles will not be considered for
publication and will be returned to the author.
Articles are structured as follows:
Title
Sub-title (if necessary)
Authors (including correspondence details)
Information provided in this document is intended solely as a guide for authors wishing to submit articles for publication
in Pharmeuropa and Pharmeuropa Bio & Scientic Notes.
October 2009
Pharmeuropa Bio & Scientic Notes, 2010-2 101
Abstract
Keywords
1. Aim (if necessary)
2. Introduction
3. Participants (where applicable)
4. Experimental (Materials and methods for articles on
the work of the Biological Standardisation Programme)
5. Results and discussion (may be split into 2 sections if
necessary)
6. Conclusion
7. Acknowledgements (where applicable)
8. Participants (listed by country in alphabetical order)
(where applicable)
9. Abbreviations (if necessary)
10. References
Appendices (if necessary)
Sub-sections are numbered accordingly, e.g. 1.1, 1.1.1,
1.1.2 etc; and should not exceed 5 hierarchic levels.
Indented paragraphs are preferably listed using bullet
points or dashes. If necessary, e.g. if reference is made to the
paragraph elsewhere in the text, indented paragraphs are
listed using a), b) etc.
Title. The title should be as clear and concise as possible,
contain important keywords and not exceed 80 characters
(including spaces). A sub-title may be included if necessary.
Authors. Authors names and initials are given with
their current addresses and, if different, their addresses
at the time the article was drafted. The author to whom
all correspondence is to be addressed is indicated
(corresponding author). Please supply a full postal address
and an e-mail address for the corresponding author and
ensure that these addresses are correct and sufficiently clear.
Abstract. A concise, self-contained summary of the problem
and the principal results, without experimental details, and
which does not contain references.
Keywords. General words are to be avoided. Not more than 6
keywords or phrases will usually be necessary.
Introduction. A concise statement of the background of the
topic, including relevant earlier work. The introduction
should indicate why this topic is considered important and
outline its relevance to the European Pharmacopoeia.
Participants. A general overview of the participants involved
in the work (concerns only articles on the work of the
Biological Standardisation Programme).
Experimental/Materials and methods. Reasonably detailed
accounts of experimental procedures, sufficient to allow
readers to repeat the work.
Results. A clear presentation of the important results,
clarified by tables and figures (see below for further
requirements). When a new analytical procedure is being
described this section may include experimental detail.
Discussion. The principal conclusions to be drawn from
the results and their implications. This section may include
comparisons with other relevant studies.
Acknowledgements. It is general policy to include
acknowledgements only in articles on the work of the
Biological Standardisation Programme.
Participants (listed by country in alphabetical order). A
list of the individual participants involved in the work, in
order of country (concerns only articles on the work of the
Biological Standardisation Programme).
Abbreviations and units. Abbreviations are either defined on
first use in the text or, if they are numerous are compiled in
a list containing all abbreviations used in the text. SI units
are used.
References. These are cited in the text according to the
EDQMs guide for citing references (see Appendix 2).
Appendices. Authors are encouraged, wherever possible, to
avoid appendices and to include the information in the main
body of the text. However, if the text contains, for example,
numerous tables that have to be grouped together, authors
should present this information in the form of an appendix.
Figures and tables. Figures and tables are kept to a
minimum and are as clear and simple as possible. Before
including a figure or table, consider if it is really necessary
(for example, avoid including a figure and a table that
represent the same information). Axes are labelled using
SI units. Wherever possible, the use of figures rather than
tables to represent numerical data is encouraged.
Photographs. Only black and white, well-contrasted
photographs will be accepted.
3.2. Readers Tribune
There is no fixed format, but as a rule articles should
not exceed 800 words in length (about one page of
Pharmeuropa). The requirements for figures and tables
described in Section 3.1 apply.
4. EVALUATION PROCEDURE
Articles are evaluated by the Editorial Boards of
Pharmeuropa and Pharmeuropa Bio & SN, and are subject
to editorial modifications before publication. If the article is
considered suitable for publication, any comments received
from the Editorial Board may be sent to the corresponding
author and a revised version of the article may be requested.
The revised version is checked by the Editors to ensure
that the comments have been taken into account. Proofs
of the article with the finalised layout are sent to the
corresponding author along with a release form. Authors
may comment on the proposed layout although the Editors
of Pharmeuropa and Pharmeuropa Bio & SN reserve the
right to make the final decision regarding matters of style
(e.g. location of figures/tables).
5. PROOFS AND RE-PRINTS
A PDF version of the published article is sent to the
corresponding author, except for certain Readers Tribune
articles (e.g. letters). Re-prints are not available.
6. RETURN OF MATERIAL
The original manuscript and illustrations will not be
returned to the author.
102 Pharmeuropa Bio & Scientic Notes, 2010-2
1. For articles submitted by post, are the hard copies and
electronic copies identical?
2. For articles written in French, is a summary in English
included?
3. Is the corresponding authors postal address and e-mail
address correct and clearly indicated?
4. Is the title concise (less than 80 characters, including
spaces) and does it contain important keywords?
5. Is a list of keywords included?
6. Does the text include an abstract?
7. Do all gures and tables have a legend with gure axes
correctly labelled using SI units?
8. Are all gures and tables referenced in the text?
9. Are all references cited in the text listed in the
references section using the EDQMs guide for citing
references?
10. Is the text correctly structured using numbered sections
and sub-sections?
APPENDIX 1
CHECKLIST FOR AUTHORS OF SCIENTIFIC NOTES
APPENDIX 2
UNIFORM REQUIREMENTS FOR REFERENCE CITING IN
PHARMEUROPA AND PHARMEUROPA BIO & SN
INTRODUCTION
1. CITATION OF MATERIAL IN JOURNALS
1.1 Monograph in Pharmeuropa
1.2 Scientific note in Pharmeuropa
1.3 Scientific note in Pharmeuropa SN
1.4 Scientific article in Pharmeuropa Bio
1.5 Scientific article in Pharmeuropa Bio & SN
1.6 Standard journal article
1.7 Article with more than 3 authors
1.8 Article with no author given
1.9 Material accepted for publication but not yet
published
1.10 Volume with a supplement
1.11 Issue with a supplement
1.12 Volume with a part
1.13 Issue with a part
1.14 Issue with no volume
1.15 No issue or volume
1.16 Type of article/source indicated as required (I)
1.17 Type of article/source indicated as required (II)
2. CITATION OF BOOKS AND MONOGRAPHS
2.1 Monograph in the European Pharmacopoeia (I)
2.2 Monograph in the European Pharmacopoeia (II)
2.3 General chapter in the European
Pharmacopoeia
2.4 Monograph in the USP
2.5 General chapter in the USP
2.6 Standard book with personal author(s)
2.7 Editor(s)/Compiler(s) as author
2.8 Organisation as author and publisher
2.9 Book with series title and individual volume
2.10 Chapter in a book
2.11 Chapter as part of an edited/translated work
3. CITATION OF MISCELLANEOUS SOURCES
3.1 Conference proceedings
3.2 Conference paper
3.3 Poster
3.4 Dissertation
3.5 EMA guideline
3.6 ICH guideline
3.7 WHO guideline
3.8 WHO report
3.9 WHO technical report series
3.10 International standards
3.11 ISO standards
4. CITATION OF ELECTRONIC MATERIAL
4.1 Electronic journal
4.2 CD-ROM
4.3 Article retrieved from a database
4.4 No author identified, no date
Instructions for authors
INTRODUCTION
The following real and fictitious examples are given as a
guide to the citation style to be used in Pharmeuropa and
Pharmeuropa Bio & SN. The style has been adapted from
the Vancouver citation style to meet the requirements
of the EDQM. References are indicated in the text using
square brackets, and are numbered in the order of citation.
References are listed in the reference list at the end of the
article. Punctuation forms an integral part of the reference;
the style used in the examples below must be carefully
copied.
1. CITATION OF MATERIAL IN JOURNALS
Cite the names and initials of all authors unless there are
more than three, when only the first three should be given
followed by et al. The authors names are followed by
the title of the article, the title of the journal abbreviated
according to the style of Index Medicus, the year of
publication, the volume number, and the first and last page
numbers.
1.1 Monograph in Pharmeuropa
[1] Methyldopa, monograph 0045. Pharmeuropa
2002;14(3):54-62.
1.2 Scientic note in Pharmeuropa
[2] Purewel TS. Test methods for dry powder inhalers to
check performance under normal use and unintentional
misuse conditions. Pharmeuropa 2002;14(3):14-18.
1.3 Scientic note in Pharmeuropa SN
[3] Weber C, Matingen S, Holzgrabe U. Batch variability of
bacitatrin: HPLC versus MEKC. Pharmeuropa Scientific
Notes 2005(1):47-51.
1.4 Scientic article in Pharmeuropa Bio
[4] Frommer I, Klaus R, Dias A et al. Feasibility study
to evaluate the correlation between results of in
vitro assay and established in vivo assays for potency
determination of various disease vaccines. Pharmeuropa
Bio 2001(1):52-66.
1.5 Scientic article in Pharmeuropa Bio & SN
[5] Ekman L, Fransson D, Claeson P et al. Development of
an Alternative Method for Determination of Terpene
Lactones in Ginkgo Dry Extract. Pharmeuropa Bio &
Scientific Notes 2009(1):67-71.
1.6 Standard journal article
[6] Turner AK, Smith S. Liver transplantation is associated
with a decreased risk for pancreatobiliary disease. Ann
Intern Med 1996;124(11):980-3.
1.7 Article with more than 3 authors
[7] Ozben T, Nacitarhan S, Tuncer N et al.Plasma and
glycosylated haemoglobin in non-insulin dependent
diabetes mellitus. Ann Clin Biochem 1985;38(3):303-6.
1.8 Article with no author given
[8] Immunologic status of the leukaemia patient and the
effects of blood transfusion on antitumor responses.
Curr Opin Gen Surg 1983;7:325-65.
1.9 Material accepted for publication but not yet published
[9] Browell DA, Orwell QR. Physiological status of the
cancer patient and the effects weight gain on staging of
solid tumours. Curr Opin Gen Surg [in press].
1.10 Volume with a supplement
[10] Brenner M, Rogerson J. SSRIs in the treatment
of dipolar depression. Environ Health Perspect
2004;102(Suppl 8):375-82.
1.11 Issue with a supplement
[11] Moithier DK, Sanchez K, Massie MJ. Womens reactions
to ovarian cancer. Semin Oncol 1994;31(19 Suppl 2):89-
97.
1.12 Volume with a part
[12] Knaup T, Naciturn S, Hendreson PR. Potency testing of
diptheria toxoid vaccines. Ann Clin Biochem 1995;32(Pt
4):203-6.
1.13 Issue with a part
[13] Sesardic GH, Millson SM. Two hundred consecutive
cases of leg ulceration in ageing patients. Pharm J
1994;107(986 Pt 3):367-8.
1.14 Issue with no volume
[14] Clawson I, Dobby T, Fillener-Tsai L. Hepatitis B (rDNA)
vaccine BRP. Clin Orthop 1997;(428):129-34.
1.15 No issue or volume
[15] Brunwell DA, Lennard TW. Collaborative study for the
establishment of a help-line for patients suffering from
cancer. Curr Opin Gen Surg 1993:325-33.
1.16 Type of article/source indicated as required (I)
[16] Schuzberger W, Fischer PA. Perindopril in Parkinsons
disease. Lancet 1986;327:1337 [letter].
1.17 Type of article/source indicated as required (II)
[17] Clements R, Jefferies R. Haematological complications
of nephropathy HVN. Kidney Int 1992;38:1285
[abstract].
2. CITATION OF BOOKS AND MONOGRAPHS
Cite authors name and initials, full title, edition,
place of publication, publisher, year of publication,
and page numbers if necessary. Do not repeat digits
unnecessarily. Page numbers are not cited for the European
Pharmacopoeia because it exists in both English and
French versions, with different pagination. Furthermore,
the monograph title and number are sufficient to lead the
reader to the specified monograph/chapter.
2.1 Monograph in the European Pharmacopoeia (I)
[18] Ferrous sulphate heptahydrate, monograph 0083. Ph.
Eur. Suppl. 4.3. Strasbourg, France: Council of Europe;
2002.
2.2 Monograph in the European Pharmacopoeia (II)
[19] Arnica flower, monograph 1391. Ph. Eur. 5th Edition.
Strasbourg, France: Council of Europe; 2004(vol 2).
2.3 General chapter in the European Pharmacopoeia
[20] Microbiological quality of pharmaceutical preparations,
general chapter 5.1.4. Ph. Eur. 4th Edition. Strasbourg,
France: Council of Europe; 2001.
2.4 Monograph in the USP
[21] Calcium gluceptate. USP 25, NF 20. Rockville, USA:
2.5 General chapter in the USP
[22] Electrophoresis, general chapter <726>. USP 25,
NF 20. Rockville, USA: United States Pharmacopeial
Convention, Inc; 2002.
Instructions for authors
2.6 Standard book with personal author(s)
[23] Ringsven MK, Bond D. Medicines management skills
for pharmacists. 2nd Edition. Albany, USA: Delmar
Publishers; 1996:134-48.
2.7 Editor(s), compiler(s) as author
[24] Norman IJ, Croft SJ, editors. Mental health care for
retired people. Dundee, Scotland: Churchill Livingstone;
1996.
2.8 Organisation as author and publisher
[25] Royal Cheshire Hospital, University of Cheshire,
Department of Clinical Pharmacy. Compendium of
pharmacy research and practice development, 1999-
2000. Cheshire, UK: Cheshire University; 2001.
2.9 Book with series title and individual volume
[26] Bennett GL, Hammlet R. Chemokines; iodination
of and use in receptor binding analysis. Methods in
Enzymology. Brighton, UK: Academic Press; 2002(vol
288):134-48.
2.10 Chapter in a book
[27] Porter RJ, Meldrum BS. Antiepileptic drugs. In: Basic
and clinical pharmacology. 6th Edition. Paris, France:
Appleton and Lange; 1995:361-80.
2.11 Chapter as part of an edited/or translated work
[28] Phillips SJ, Summers JP. Hypertension is dangerous:
stroke and myocardial infarction events in the over
50 age group. In: Laragh JH, Brenner BM, editors.
Hypertension, the pathophysiology, diagnosis, and
management. 2nd Edition. London, UK: Raven Press;
1995:465-78.
3. CITING MISCELLANEOUS SOURCES
3.1 Conference proceedings
[29] Nicouma J, Sissons H, editors. Recent advances in
clinical cardiology. Proceedings of 11th international
congress of EMG and clinical cardiology: Cape Town,
South Africa; 1995 Oct 15-19. Brighton, UK: Dingthton
publishers Int; 1996.
3.2 Conference paper
[30] Walker AR, McKellar RB. Preliminary observations on
the density of mitochondria of human kidney cells.
In: Irvin AD, Cunningham MP, Dustoff AS, editors.
Advances in the control of theileriosis. Proceedings
of the 5th international conference for research on
animal diseases; 1981 Feb 9-13. Oxford, UK: Delaware
Publishers; 1981:125-6.
3.3 Poster
[31] Gerard M, Moussard N, Whitehew LW. Comparative
study of reversed phase and hydrophobic interaction
chromatography for the analysis of human growth
hormone formulations. Poster 32. The 8th European
meeting of chromatography and molecular biology: La
Grande Motte, France; 2001 June.
3.4 Dissertation
[32] Kappin SJ. Post-hospital home health care in the elderly
population: up-take and use. St. Louis, MO: Washington
Univ; 1995 [dissertation].
3.5 EMA guideline
[33] Note for guidance on dry powder inhaler(DPI) for use
in-chronic asthma: chemistry, manufacturing, and in-
house controls: chemistry, manufacturing, and controls
documentation. Ref: CPMP/QWP/2845/00. EMEA; 2000
Nov [guideline].
3.6 ICH guideline
[34] Stability testing: photostability testing of drug
substances for use in animals. Ref: Q1B. ICH; 1998 Nov
6 [guideline].
3.7 WHO guideline
[35] Horowitz B, Minor P, Morgenthaler JJ. Guidelines on
viral activation and addition procedures intended to
assure the viral safety of animal blood plasma products.
WHO Expert Committee on Biological Standardization;
2003 Dec 26-30.
3.8 WHO report
[36] Swahele S, Lilley N,Yu MW et al. An international
collaborative study to establish a WHO international
standard for B20 rDNA assays. Ref: BS/00.1928. WHO
Expert Committee on Biological Standardization; 2000.
3.9 WHO technical report series
[37] Technical Report Series 96, 47th report. WHO Expert
Committee on Biological Standardization; 2004 Jul.
3.10 International standards
[38] European Committee for Standardization (CEN).
Respiratory therapy equipment Part 1: Nebulizing
systems and their components. Ref: EN 13544:2001.
3.11 ISO standards
[39] International Organization for Standardization. Particle
size analysis -laser diffraction methods Part 1: General
principles. Ref: ISO 13320-1:1999.
4. CITATION OF ELECTRONIC MATERIAL
The basic format for electronic sources follows the same
principles as for printed references. If deemed necessary, the
source of the material (CD-ROM, computer file, etc.) should
be identified. The URL address and the date the information
was accessed are included at the end of the reference.
4.1 Electronic journal
[40] VandenBos G, Knapp C, Dowler S. Role of reference
elements in the selection of resources by pharmacology
students. J of Bibliographic Research 1992;5:118-23
[available at: http://jbr.org/articles.html, accessed 2001
Oct 13].
4.2 CD ROM
[41] Ley J-T, McBairn H. Clinical cardiology, illustrated 2nd
Edition. San Fransisco: CDEA Multimedia Group; 1995
[CD-ROM].
4.3 Article retrieved from a database
[42] Anderson C. Role of the pharmacist in the community
pharmacy setting. J Appl Psychology 1998 Sep
19;78:443-9 [available at: Ovid database, accessed 2000
Oct 13].
4.4 No author identied, no date
[43] GVUs 9th WWW survey. [available at: http://www.
dc.gatech.edu/gvu/usersurveys/, accessed 2000 Aug 8].
2
0
1
0
-2
O
ctober
2010
P
H
A
R
M
E
U
R
O
P
A

B
I
O
&

S
C
I
E
N
T
I
F
I
C

N
O
T
E
S
This issue is included in
the subscription to Pharmeuropa
2010 Council of Europe

ISSN 2075-2164
ISSN-L 2075-2164
Director of the Publication
Dr S. Keitel
Editorial Board
Prof. H. Altorfer
Dr S.R. Andersen
Dr T. Arvidsson
Dr F. Bailleul
Prof. Dr K.H.C. Baser
Dr A.F. Bristow
Dr L. Bruckner
Dr K.H. Buchheit
Prof. D.H. Calam
A. Costanzo
Dr C. Coune
Dr M. Ek
F. Ezrati
Prof. Dr G. Franz
Prof. S.H. Hansen
Dr C. Jarvis
Prof. H.J. de Jong
P. Jongen
K. Kreft
Prof. H.G. Kristensen
Dr G. Lee
M. Marre
Prof. B. Meier
Dr C. Milne
Dr J.M. Morris
Prof. A. Nicolas
C. Nicolas
Dr V. ppling
Prof. P.-P. Pastoret
Dr J.C. Rouby
S. Schaeffer
Prof. G. Scriba
Prof. Dr R. Seitz
Dr G. Silvester
J.-M. Spieser
Dr T. van der Stappen
Dr B. Strukelj
Dr E. Terao
Dr M. Terberger
L. Toumasson
Dr D. Wood
Page layout
I. Bylinski
I. Iankova
C. Knaup
A. Roberts
I. Vernay
Copyright
All rights conferred by virtue
of the International Copyright Convention are
specifically reserved to the Council of Europe
and any reproduction or translation requires
the written consent of the Publisher.
Contributions
The Editor is pleased to consider contributions
for publication in Pharmeuropa Scientific Notes
provided they have some relationship with the
work of the European Pharmacopoeia.
It is assumed that the copyright of material
submitted is that of the contributor or that the
necessary clearances have been obtained and
that material submitted has not been offered
simultaneously to another publication.
Authors alone are responsible for the views
expressed in signed contributions. Please see
Instructions for Authors (www.edqm.eu).
Mention of specific companies or
of certain manufacturers products is made only
for the convenience of the reader. This does not
imply that they are endorsed or recommended
by the European Pharmacopoeia Commission or
by the Council of Europe in preference to others
of a similar nature that are not mentioned.
European Directorate for the Quality of
Medicines & HealthCare (EDQM)
7 alle Kastner
CS 30026
F-67081 STRASBOURG
FRANCE
Tel: +33 (0)3 88 41 30 30
Fax: +33 (0)3 88 41 27 71
E-mail: info@edqm.eu
Internet: http://www.edqm.eu
Bank: Socit Gnrale
Account N: FR76
30003 02360 00550034256/76
Strasbourg
PH
A
R
M
EU
R
O
PA
BIO
&
SC
IEN
TIFIC
N
O
TES 2
0
1
0
-2
C
o
n
te
n
ts
Calibration of Hum
an Coagulation Factor VIII Concentrate
Ph. Eur. BRP Batch 4 for Use in Potency Assays
ent of Replacem
ent Batches
of Heparin Low
-M
olecular-M
ass for Assay Biological Reference
Preparations
Collaborative Study for Validation of a Serological Potency Assay
for Rabies Vaccine (inactivated) for Veterinary Use
ent of the Second
International Standard for Gram
icidin
ent of the Third International
Standard for Nystatin
Q
uality Control of M
etoprolol Extended-Release Form
ulations in
the Presence of Ethanol
Possible Am
biguities w
hen Testing Viscosity in Com
pendial
M
onographs - Characterisation of Grades of Cellulose Ethers
O
c
t
o
b
e
r

2
0
1
0
P
H
A
R
M
E
U
R
O
P
A

B
I
O

&

S
C
I
E
N
T
I
F
I
C

N
O
T
E
S

2
0
1
0
-
2

HTTP 1

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

HTTP 1

Transféré par

Droits d'auteur :

Formats disponibles

2

2010 Council of Europe

software[3]. Each run

for calculations and

forte, Metoprolol Sandoz

and Seloken ZOC

in the presence of 10 to 40% (v/v%) ethanol

and Seloken ZOC

were released fast in

Retard and to some

forte decreased in the ethanol containing media.

and Seloken ZOC

forte, Metoprolol Sandoz

containing 100 mg metoprolol per dosage form

Retard contains metoprolol tartrate while

Retard was analysed and those containing metoprolol

forte, Metoprolol Sandoz

were analyzed. The concentration of the

Retard Metoprolol tartrate Extended-release tablet (non-

forte Metoprolol succinate

Metoprolol succinate Extended-release tablet (dividable

Metoprolol succinate Extended-release tablet (dividable

forte, Metoprolol Sandoz

and Seloken ZOC

Retard formulation is a non-

Retard tablets in the presence of 0, 10, 20

forte tablets in the presence of 0, 10, 20 and

forte formulation is a non-dividable ER

delivering metoprolol succinate is a

Retard. The high surface area contributes to a higher

Retard. This illustrates the substantial

in the presence of 0, 10, 20 and 40% ethanol

tablets in the presence of 0, 10, 20 and

Retard and Logimax

forte are non-

forte. Both of these

forte, Metoprolol GEA

and Seloken ZOC

Retard dosage form. The similarity

and Seloken ZOC

forte was not disintegrated in contact with

and Seloken ZOC

Retard formulation. This

and Seloken ZOC

and Seloken ZOC

will leave the

forte ER-tablets do. This may, to a

and Seloken ZOC

Retard and Logimax

forte dosage forms

formulations were more sensitive to

Retard in the absence of

2010 Council of Europe

Vous aimerez peut-être aussi