< Return to Video

OECD Webinar | Training and orientation on validation of methods at the OECD

  • 0:00 - 0:04
    Καλησπέρα, Καλημέρα ή Καλό βράδυ
  • 0:04 - 0:06
    αναλόγως σε ποιο σημείο του πλανήτη βρίσκετε και
  • 0:06 - 0:11
    καλωσήρθατε σε αυτό το σεμινάριο σχετικά με την εκπαίδευση
  • 0:11 - 0:16
    orientation on validation of Methods του ΟΟΣΑ.
  • 0:16 - 0:20
    Είμαστε εδώ στις 14 Νοεμβρίου 2023
  • 0:20 - 0:25
    2023 and my name is Anon uh and uh I
  • 0:25 - 0:29
    will give you um just uh one
  • 0:29 - 0:31
    introductory
  • 0:31 - 0:34
    slide uh about the objectives of the
  • 0:34 - 0:35
    this
  • 0:35 - 0:39
    webinar uh and uh then we will have our
  • 0:39 - 0:42
    speaker for today miam uh
  • 0:42 - 0:45
    Jacobs uh first the objectives of this
  • 0:45 - 0:48
    webinar of first to uh clarify terms
  • 0:48 - 0:52
    used in the context of validation for
  • 0:52 - 0:55
    test guidelines development
  • 0:55 - 0:59
    ATD uh then uh another objective is to
  • 0:59 - 1:01
    explain
  • 1:01 - 1:05
    why validation of methods is important
  • 1:05 - 1:08
    in the context of test guidelines
  • 1:08 - 1:12
    development uh also uh one objective is
  • 1:12 - 1:15
    to guide laboratories embarking on
  • 1:15 - 1:16
    methods
  • 1:16 - 1:19
    validation uh on how to approach
  • 1:19 - 1:23
    validation and evaluate uh method
  • 1:23 - 1:26
    Readiness and how uh relevance and
  • 1:26 - 1:30
    reproducibility of methods uh can be uh
  • 1:30 - 1:31
    tested
  • 1:31 - 1:34
    evaluated and then we will have uh a
  • 1:34 - 1:39
    moment in this webinar for a Q&A
  • 1:39 - 1:42
    session uh so you are most welcome to
  • 1:42 - 1:46
    use the Q&A box uh we will be posting
  • 1:46 - 1:50
    material resource in the chat uh but the
  • 1:50 - 1:52
    chat is not accessible to participants
  • 1:52 - 1:57
    only the the Q&A uh boxer so we will be
  • 1:57 - 2:00
    selecting uh questions from the Q&A and
  • 2:00 - 2:05
    raising them to uh uh our speaker and
  • 2:05 - 2:07
    anyone who would like to uh to
  • 2:07 - 2:11
    answer uh so also please note uh this
  • 2:11 - 2:13
    webinar is recorded and it will be made
  • 2:13 - 2:16
    publicly uh available as a replay on the
  • 2:16 - 2:18
    o
  • 2:18 - 2:21
    website so without further Ado I'm going
  • 2:21 - 2:26
    to hand over um the presenter wres to
  • 2:26 - 2:30
    our speaker for today uh Miriam Jacobs
  • 2:30 - 2:34
    uh who is uh uh the national coordinator
  • 2:34 - 2:36
    for the test guidelines program in the
  • 2:36 - 2:43
    United Kingdom and Miriam can you hear
  • 2:44 - 2:48
    us yes I can hear you can you hear me
  • 2:48 - 2:50
    yes very well
  • 2:50 - 2:55
    great okay I'll start sharing my screen
  • 2:59 - 3:04
    now you can see your screen y
  • 3:05 - 3:08
    okay okay welcome everybody um this is
  • 3:08 - 3:11
    really a preamble for the workshop that
  • 3:11 - 3:13
    we will be having on the 14th and 15th
  • 3:13 - 3:17
    of December uh later this year um and
  • 3:17 - 3:19
    this is just to get common understanding
  • 3:19 - 3:22
    with respect to um validation
  • 3:22 - 3:26
    terminology and um some PR principles
  • 3:26 - 3:28
    and considerations that that uh are
  • 3:28 - 3:30
    related to that
  • 3:30 - 3:35
    so I'm going to go to the next slide
  • 3:36 - 3:40
    now um so here's an overview of the talk
  • 3:40 - 3:43
    can you see my slides
  • 3:43 - 3:48
    okay yes very well yeah okay great um so
  • 3:48 - 3:50
    there are five aspects to this talk the
  • 3:50 - 3:54
    first part is to discuss why uh we want
  • 3:54 - 3:57
    to have uh validation for regulatory
  • 3:57 - 4:01
    purposes why it's important um so the
  • 4:01 - 4:03
    per first part is uh looking at the
  • 4:03 - 4:05
    crisis and reproducibility that we have
  • 4:05 - 4:07
    and the relevance of to the oecd test
  • 4:07 - 4:09
    guideline program of the mutual
  • 4:09 - 4:12
    acceptance of data principle the second
  • 4:12 - 4:14
    point we look at an overview of
  • 4:14 - 4:16
    validation terminology focusing
  • 4:16 - 4:19
    particularly on prospective validation
  • 4:19 - 4:21
    and more really the early stages of
  • 4:21 - 4:23
    validation the third aspect we'll look
  • 4:23 - 4:26
    at examples of what can go wrong and why
  • 4:26 - 4:29
    and how we can recover from that the
  • 4:29 - 4:32
    fourth one we'll look at what you can do
  • 4:32 - 4:35
    um very early on in the process and
  • 4:35 - 4:37
    finally we'll have some questions and
  • 4:37 - 4:42
    answers uh questions and discussion
  • 4:42 - 4:47
    session so the first one crisis and
  • 4:48 - 4:49
    reproducibility that there's a lack of
  • 4:49 - 4:52
    reproducibility in research is really
  • 4:52 - 4:55
    very well known um there was a survey
  • 4:55 - 4:57
    conducted by nature and published in
  • 4:57 - 5:01
    2016 where where they reported amongst
  • 5:01 - 5:04
    their respondents um More than 70% of
  • 5:04 - 5:06
    the researchers tried and failed to
  • 5:06 - 5:09
    reproduce other scientists experiments
  • 5:09 - 5:12
    and many couldn't reproduce their own
  • 5:12 - 5:15
    experiments and if you have a look here
  • 5:15 - 5:19
    um unfortunately for us as as working in
  • 5:19 - 5:22
    toxicology and and biology and
  • 5:22 - 5:26
    chemistry these were the worst um uh
  • 5:26 - 5:29
    failures really in chemistry and biology
  • 5:29 - 5:32
    so um in
  • 5:32 - 5:35
    biology uh here you can see more than
  • 5:35 - 5:38
    60% um failed to reproduce their own
  • 5:38 - 5:40
    experiments and this was from a sample
  • 5:40 - 5:42
    of over
  • 5:42 - 5:45
    1,500 um researchers and it was just on
  • 5:45 - 5:47
    the basis of a brief online
  • 5:47 - 5:50
    questionnaire but it's really quite
  • 5:50 - 5:52
    shocking and that really goes to show
  • 5:52 - 5:55
    how um
  • 5:55 - 6:00
    uh how one needs to be very careful with
  • 6:00 - 6:01
    literature
  • 6:01 - 6:03
    reports
  • 6:03 - 6:08
    furthermore um in another report uh from
  • 6:08 - 6:11
    Arrington atal in 2021 they looked at
  • 6:11 - 6:15
    193 experiments just to look at where at
  • 6:15 - 6:17
    the protocol status where they were
  • 6:17 - 6:20
    fully described and what the barriers
  • 6:20 - 6:21
    were to conducting
  • 6:21 - 6:24
    replications in the experiments
  • 6:24 - 6:28
    reported and in this
  • 6:28 - 6:33
    um uh report you can see here that very
  • 6:33 - 6:35
    often there was very little help from
  • 6:35 - 6:37
    the authors when they tried to find out
  • 6:37 - 6:41
    more about the protocols um in most
  • 6:41 - 6:44
    cases clarifications were needed often
  • 6:44 - 6:47
    reagents were offered up to 50% but
  • 6:47 - 6:51
    often not um and then codes were
  • 6:51 - 6:54
    sometimes shared but often no
  • 6:54 - 6:57
    response um and data sharing as well was
  • 6:57 - 6:58
    really
  • 6:58 - 7:01
    Limited
  • 7:02 - 7:05
    something that uh the jrc have reported
  • 7:05 - 7:08
    on and and given presentations on
  • 7:08 - 7:12
    recently um uh this promap promoting
  • 7:12 - 7:14
    reusable open methods and protocols um
  • 7:14 - 7:17
    so there's a nice pre-print from um
  • 7:17 - 7:21
    safia Batista leet um online where they
  • 7:21 - 7:24
    give recommendations on how to report
  • 7:24 - 7:27
    protocols failure in protocols and
  • 7:27 - 7:30
    reproducibility in protocols is the
  • 7:30 - 7:33
    first step in the validation process
  • 7:33 - 7:34
    that's necessary for improving
  • 7:34 - 7:37
    reproducibility of
  • 7:37 - 7:40
    results and being able to transfer the
  • 7:40 - 7:42
    test method it Fosters confidence in
  • 7:42 - 7:44
    results that you can obtain with a
  • 7:44 - 7:46
    certain method and it really facilitates
  • 7:46 - 7:48
    the use in the weight of evidence Hazard
  • 7:48 - 7:50
    and risk
  • 7:50 - 7:54
    assessment and this is also
  • 7:54 - 7:57
    um supported by an oecd call for Urgent
  • 7:57 - 8:00
    mobilization of national and Regional
  • 8:00 - 8:01
    resources to support the valid
  • 8:01 - 8:04
    validation of new methods for the safety
  • 8:04 - 8:06
    assessment of of of
  • 8:06 - 8:09
    chemicals with the need to be able to
  • 8:09 - 8:12
    have good validation and good
  • 8:12 - 8:18
    information around first first line um
  • 8:20 - 8:23
    protocols the mutual acceptance of data
  • 8:23 - 8:26
    and the practices regards uh the OCD
  • 8:26 - 8:28
    test
  • 8:28 - 8:32
    guidelines is the next key element um to
  • 8:32 - 8:33
    begin
  • 8:33 - 8:36
    with uh that we need to consider
  • 8:36 - 8:38
    reproducible data is essential to
  • 8:38 - 8:39
    support the international Mutual
  • 8:39 - 8:41
    acceptance of data principle for all
  • 8:41 - 8:44
    oecd member country chemical regulatory
  • 8:44 - 8:46
    uh
  • 8:46 - 8:48
    systems it's critical
  • 8:48 - 8:51
    for um ensuring there are no trade
  • 8:51 - 8:55
    barriers it's critical for ensuring
  • 8:55 - 8:58
    reproducibility of data transfer of data
  • 8:58 - 9:00
    and the next few slides will'll go
  • 9:00 - 9:01
    through these in more
  • 9:01 - 9:06
    detail the reason it was created um was
  • 9:06 - 9:09
    really in response to fraudulent studies
  • 9:09 - 9:12
    submitted to Regulators uh the Mad
  • 9:12 - 9:14
    framework was constructed to ensure the
  • 9:14 - 9:17
    generation of highquality and reliable
  • 9:17 - 9:20
    non-clinical test data for regulatory
  • 9:20 - 9:23
    purposes good lab practice provides the
  • 9:23 - 9:24
    quality standards and the test
  • 9:24 - 9:27
    guidelines provide the scientific
  • 9:27 - 9:30
    standard regulatory author authorities
  • 9:30 - 9:31
    receiving the data under the mutual
  • 9:31 - 9:34
    acceptance of data agreement under the
  • 9:34 - 9:37
    orices of the oecd know that particular
  • 9:37 - 9:39
    quality and scientific standards have
  • 9:39 - 9:41
    been followed and they know then that
  • 9:41 - 9:43
    they don't have to re-evaluate the test
  • 9:43 - 9:46
    protocol to determine its robustness as
  • 9:46 - 9:48
    it has a consensus by countries by the
  • 9:48 - 9:52
    Tes guard guideline program and part of
  • 9:52 - 9:56
    that approach is really is truly by
  • 9:56 - 9:58
    consensus every country every member
  • 9:58 - 10:03
    country needs to agree so it's not by
  • 10:05 - 10:07
    majority what documents are covered by
  • 10:07 - 10:09
    the mutual acceptance of
  • 10:09 - 10:13
    data the OCD test guidelines and the OCD
  • 10:13 - 10:16
    principles for of good lab practice are
  • 10:16 - 10:18
    covered by mad they those are the only
  • 10:18 - 10:21
    documents that are covered by mad all
  • 10:21 - 10:24
    the other oecd documents are not
  • 10:24 - 10:26
    guidance documents describing certain
  • 10:26 - 10:28
    test methods in the series and testing
  • 10:28 - 10:32
    and assessment aren't covered um an
  • 10:32 - 10:35
    additional uh in and additional
  • 10:35 - 10:37
    documents such as detailed review papers
  • 10:37 - 10:40
    validation reports peer reviews they're
  • 10:40 - 10:43
    not covered by
  • 10:43 - 10:47
    mad guidance documents can often be
  • 10:47 - 10:49
    [Music]
  • 10:49 - 10:53
    um uh validated test methods um but they
  • 10:53 - 10:55
    are examples where the scientific
  • 10:55 - 10:57
    standard wasn't agreed by consensus in
  • 10:57 - 10:58
    the working group of national
  • 10:58 - 11:00
    coordinators
  • 11:00 - 11:02
    and it's considered that the scientific
  • 11:02 - 11:04
    standard for Mutual acceptance of data
  • 11:04 - 11:06
    hasn't been
  • 11:06 - 11:09
    met at that time but that doesn't mean
  • 11:09 - 11:12
    to say that the guidance documents for
  • 11:12 - 11:15
    test methods um can't be adopted as test
  • 11:15 - 11:17
    guidelines at a later date when more
  • 11:17 - 11:19
    information and more uh relevant
  • 11:19 - 11:21
    information is available to show
  • 11:21 - 11:24
    reliability and
  • 11:25 - 11:27
    reproducibility what does the oecd
  • 11:27 - 11:29
    council act on the mutual acceptance of
  • 11:29 - 11:33
    data say it says that the data generated
  • 11:33 - 11:35
    in the testing of chemicals in an oecd
  • 11:35 - 11:37
    member country in accordance with oecd
  • 11:37 - 11:40
    test guidelines and oecd principles of
  • 11:40 - 11:42
    good lab practice shall be accepted in
  • 11:42 - 11:44
    other member countries for purposes of
  • 11:44 - 11:46
    assessment and other uses relating to
  • 11:46 - 11:49
    the protection of man and the
  • 11:49 - 11:51
    environment however the data
  • 11:51 - 11:52
    requirements are government
  • 11:52 - 11:55
    prerogatives interpretation of test
  • 11:55 - 11:58
    results is a government
  • 11:58 - 12:01
    prerogative there can be there is no
  • 12:01 - 12:03
    need for any repeat testing for the same
  • 12:03 - 12:04
    data
  • 12:04 - 12:07
    requirement however acceptance doesn't
  • 12:07 - 12:10
    to automatically mean the use of the
  • 12:10 - 12:14
    data by that
  • 12:15 - 12:18
    country the regulatory use needs data
  • 12:18 - 12:21
    requirements are country prerogatives
  • 12:21 - 12:23
    the test guidelines are intended
  • 12:23 - 12:26
    intended to address and respond to the
  • 12:26 - 12:29
    existing or foreseeable regulatory needs
  • 12:29 - 12:31
    in relation to hazard
  • 12:31 - 12:33
    assessment and the kind of regulatory
  • 12:33 - 12:36
    needs um that the test guidelines
  • 12:36 - 12:38
    address um I'm sure you're all very
  • 12:38 - 12:40
    familiar with these determination of
  • 12:40 - 12:42
    physicochemical properties environmental
  • 12:42 - 12:45
    fate and behavior properties chemical
  • 12:45 - 12:47
    mechanisms of action and
  • 12:47 - 12:50
    bioactivity um absorption distribution
  • 12:50 - 12:52
    metabolism and excretion data for model
  • 12:52 - 12:55
    building toxicity data for screening and
  • 12:55 - 12:58
    priority setting toxicity data for
  • 12:58 - 13:01
    identification of a Hazard and or Target
  • 13:01 - 13:03
    organ and toxicity data for
  • 13:03 - 13:05
    characterizing a hazard and residues
  • 13:05 - 13:08
    chemistry
  • 13:08 - 13:10
    data although data requirements are
  • 13:10 - 13:13
    country prerogatives and mad is about
  • 13:13 - 13:16
    avoiding repeat testing if data
  • 13:16 - 13:18
    requirements diverge extensively between
  • 13:18 - 13:21
    countries Mutual acceptance of data
  • 13:21 - 13:24
    principle will erode as test guidelines
  • 13:24 - 13:25
    data will not be accepted across
  • 13:25 - 13:28
    countries having different
  • 13:28 - 13:29
    requirements there therefore the more
  • 13:29 - 13:31
    compatible and similar the data
  • 13:31 - 13:33
    requirements are between countries the
  • 13:33 - 13:37
    more beneficial mad will be
  • 13:41 - 13:43
    globally and although the interpretation
  • 13:43 - 13:45
    of test results is a government
  • 13:45 - 13:47
    prerogative the test guidelines often
  • 13:47 - 13:49
    integrate the transformation of the raw
  • 13:49 - 13:51
    experimental data for example through
  • 13:51 - 13:53
    prediction models data interpretation
  • 13:53 - 13:56
    procedures to generate a retest result
  • 13:56 - 13:58
    that addresses more directly a
  • 13:58 - 13:59
    regulatory need
  • 13:59 - 14:01
    for example the identification of a
  • 14:01 - 14:04
    hazard the data transformation and
  • 14:04 - 14:06
    interpretation procedures implemented in
  • 14:06 - 14:08
    the test guideline are the outcome of
  • 14:08 - 14:11
    country's agreement to do so in order to
  • 14:11 - 14:13
    generate meaningful data and reduce the
  • 14:13 - 14:16
    room for misinterpretation and maintain
  • 14:16 - 14:18
    a common Level Playing
  • 14:18 - 14:21
    Field but it Remains the member
  • 14:21 - 14:23
    country's prerogative to use the
  • 14:23 - 14:26
    Standalone test result to to satisfy a
  • 14:26 - 14:28
    data requirement or to use the test
  • 14:28 - 14:30
    result de with other sources of
  • 14:30 - 14:32
    information in combination with other
  • 14:32 - 14:35
    levels of interpretation or criteria
  • 14:35 - 14:37
    that meet a country specific regulatory
  • 14:37 - 14:40
    need and to not use that test result if
  • 14:40 - 14:42
    their data requirement cannot be
  • 14:42 - 14:45
    satisfied with that test result alone or
  • 14:45 - 14:47
    in
  • 14:48 - 14:52
    combination so that's uh part one um
  • 14:52 - 14:55
    regarding the crisis in reproducibility
  • 14:55 - 14:58
    and uh the Mad principle the mutual
  • 14:58 - 15:02
    acceptance of of data principle um now I
  • 15:02 - 15:05
    want to move to the section two giving a
  • 15:05 - 15:08
    brief overview of validation
  • 15:10 - 15:12
    terminology here I want to look at
  • 15:12 - 15:15
    standard perspective validation um other
  • 15:15 - 15:18
    aspects to consider Al also highly
  • 15:18 - 15:20
    relevant but not not going into any
  • 15:20 - 15:22
    great detail here our retrospective
  • 15:22 - 15:25
    validation test guideline revision and
  • 15:25 - 15:27
    test guideline
  • 15:27 - 15:31
    augmentation so retrospective validation
  • 15:31 - 15:33
    for example can happen where if you have
  • 15:33 - 15:36
    uh a lot of relevant data and you want
  • 15:36 - 15:41
    to go um uh and uh provide a data
  • 15:41 - 15:45
    package to show that the test method or
  • 15:45 - 15:48
    improvements in a test method um can be
  • 15:48 - 15:51
    done on the basis of data collected
  • 15:51 - 15:53
    historically and and this is happened
  • 15:53 - 15:55
    several times for example for inv Vivo
  • 15:55 - 15:58
    test guidelines um there are a couple of
  • 15:58 - 16:01
    of examples recently where where we have
  • 16:01 - 16:02
    done some retrospective
  • 16:02 - 16:06
    validation test guideline revision also
  • 16:06 - 16:11
    um Can Happen uh with respect to um more
  • 16:11 - 16:14
    data being provided um being able to
  • 16:14 - 16:16
    expand the chemical applicability domain
  • 16:16 - 16:19
    for example or test guideline
  • 16:19 - 16:22
    augmentation um when uh there might be a
  • 16:22 - 16:24
    specific element of a test guideline
  • 16:24 - 16:27
    that can be expanded further in a
  • 16:27 - 16:29
    particular direction and the evidence
  • 16:29 - 16:31
    has been collected to be able to show
  • 16:31 - 16:34
    how it can be expanded or where a test
  • 16:34 - 16:37
    guideline may have fail to become
  • 16:37 - 16:39
    adopted and it because it needs a bit of
  • 16:39 - 16:40
    further
  • 16:40 - 16:43
    work um and I will give one example of
  • 16:43 - 16:45
    of of of that
  • 16:45 - 16:48
    later uh but mainly the next few slides
  • 16:48 - 16:52
    really focus on standard perspective um
  • 16:52 - 16:54
    prevalidation
  • 16:54 - 16:57
    and these are the main resources that
  • 16:57 - 17:00
    we're using so the OCD um guidance
  • 17:00 - 17:04
    document 34 uh published in 2005 and
  • 17:04 - 17:06
    whilst it is being
  • 17:06 - 17:11
    revised uh at the moment and work is on
  • 17:11 - 17:14
    foreseen really for the next year or so
  • 17:14 - 17:15
    um
  • 17:15 - 17:19
    it's the basic uh requirements of
  • 17:19 - 17:23
    validation um I don't imagine will uh uh
  • 17:23 - 17:26
    change that much not in relation to to
  • 17:26 - 17:28
    um uh what we're considering here and
  • 17:28 - 17:31
    then the other resource is um the paper
  • 17:31 - 17:34
    from efam published in 2004 that was
  • 17:34 - 17:36
    very much part of this guidance document
  • 17:36 - 17:40
    205 at the time a modular approach to
  • 17:40 - 17:43
    the Ean principles on test
  • 17:44 - 17:50
    validity um these are the modules um so
  • 17:50 - 17:51
    starting first of all with a test
  • 17:51 - 17:53
    definition that is what are the
  • 17:53 - 17:55
    biological process and endpoints that
  • 17:55 - 17:57
    are modeled by the test system the
  • 17:57 - 17:59
    second one is the Within lab variability
  • 17:59 - 18:02
    or reproducibility can the results be
  • 18:02 - 18:05
    reproduced or replicated in house by a
  • 18:05 - 18:07
    different operator and over different
  • 18:07 - 18:11
    part time periods transferability can
  • 18:11 - 18:13
    others run the method successfully is
  • 18:13 - 18:15
    the pro protocol sufficiently
  • 18:15 - 18:18
    detailed number four is the between lab
  • 18:18 - 18:20
    variability can the results be
  • 18:20 - 18:23
    reproduced or replicated by others and
  • 18:23 - 18:26
    um how accurate is it um what is the
  • 18:26 - 18:28
    sensitivity and specificity of the test
  • 18:28 - 18:29
    method
  • 18:29 - 18:32
    um predictive capacity do the
  • 18:32 - 18:34
    experimental results match with uh
  • 18:34 - 18:36
    usually literature-based
  • 18:36 - 18:38
    expectations sometimes in Vivo based
  • 18:38 - 18:41
    expectations it varies and broader
  • 18:41 - 18:42
    scientific
  • 18:42 - 18:44
    consensus what is the chemical
  • 18:44 - 18:47
    applicability domain what can or can't
  • 18:47 - 18:50
    be tested um what are the performance
  • 18:50 - 18:52
    standards what
  • 18:52 - 18:54
    parameters can be developed that can be
  • 18:54 - 18:56
    expected from this test method when it's
  • 18:56 - 18:59
    working as expected based upon on the
  • 18:59 - 19:01
    historical data and this includes
  • 19:01 - 19:03
    reference chemicals to demonstrate
  • 19:03 - 19:06
    demonstrate equivalence in performance
  • 19:06 - 19:08
    for new me2 test systems compared to the
  • 19:08 - 19:11
    those that have been previously
  • 19:11 - 19:14
    validated so these are different modules
  • 19:14 - 19:16
    and one can come in and out of the
  • 19:16 - 19:20
    different modules it's not rigid um and
  • 19:20 - 19:24
    this is the uh from the Harang AEL paper
  • 19:24 - 19:27
    showing how they're constructed um and
  • 19:27 - 19:30
    how the mod modes lead when they're
  • 19:30 - 19:33
    successful they get um discussed by the
  • 19:33 - 19:35
    validation Management Group lead for and
  • 19:35 - 19:37
    then they can be taken forward for
  • 19:37 - 19:40
    peerreview by ex various expert panels
  • 19:40 - 19:43
    wherever they might be um whether that's
  • 19:43 - 19:47
    uh esac the European um well efam
  • 19:47 - 19:50
    scientific advisory committee or efam or
  • 19:50 - 19:52
    jfam or
  • 19:52 - 19:55
    oecd um at this
  • 19:55 - 19:58
    level so really I'm just going to talk
  • 19:58 - 20:00
    talk a little bit more about these four
  • 20:00 - 20:03
    modules
  • 20:03 - 20:06
    reproducibility uh a typical validation
  • 20:06 - 20:09
    management structure um would be having
  • 20:09 - 20:11
    a study management team here at the top
  • 20:11 - 20:13
    independent chemical selection and
  • 20:13 - 20:16
    expert review um to the side independent
  • 20:16 - 20:19
    statistical analysis um and then a
  • 20:19 - 20:23
    number of Labs at least two um sometimes
  • 20:23 - 20:27
    three and often more it really
  • 20:27 - 20:30
    depends
  • 20:30 - 20:33
    for the test definition the objective is
  • 20:33 - 20:35
    to define the scientific purpose
  • 20:35 - 20:37
    describe the mechanistic basis
  • 20:37 - 20:40
    characterize the cell line um there have
  • 20:40 - 20:45
    been quite a few um uh issues arising
  • 20:45 - 20:47
    out of cell line characterization or
  • 20:47 - 20:49
    lack of characterization that I will
  • 20:49 - 20:52
    mention later the limitations of the
  • 20:52 - 20:55
    test method the experimental design and
  • 20:55 - 20:58
    the optimized protocol which includes
  • 20:58 - 20:59
    the necessary
  • 20:59 - 21:02
    sop the specifications of endpoints and
  • 21:02 - 21:04
    endpoint measurements and the derivation
  • 21:04 - 21:06
    and expression of results and
  • 21:06 - 21:08
    interpretation um of a preliminary
  • 21:08 - 21:12
    prediction model uh the prediction Mo
  • 21:12 - 21:16
    model often needs to be revised um uh on
  • 21:16 - 21:19
    the subsequent steps that uh uh follow
  • 21:19 - 21:22
    in the prevalidation aspect and then
  • 21:22 - 21:24
    there need to be adequate controls
  • 21:24 - 21:26
    vehicle controls positive controls
  • 21:26 - 21:30
    negative controls
  • 21:31 - 21:33
    to begin with also one has requirements
  • 21:33 - 21:35
    for the initial training or proficiency
  • 21:35 - 21:37
    set of chemicals and these need to be
  • 21:37 - 21:39
    well-known chemicals to cover the
  • 21:39 - 21:41
    relevant mechanism and mode of action of
  • 21:41 - 21:45
    the test method and with
  • 21:46 - 21:49
    negatives with the second module one
  • 21:49 - 21:51
    starts to look at the variability over
  • 21:51 - 21:53
    time and addressing variability for
  • 21:53 - 21:55
    different operators but using the same
  • 21:55 - 21:57
    lab laboratory
  • 21:57 - 22:01
    setup um and the proficiency set of
  • 22:01 - 22:05
    chemicals um generally expands uh these
  • 22:05 - 22:07
    are the most obvious chemicals due to
  • 22:07 - 22:09
    Quality and quantity of reference
  • 22:09 - 22:12
    data um addressing the mode of action
  • 22:12 - 22:15
    the relative potency if possible
  • 22:15 - 22:16
    physicochemical properties and
  • 22:16 - 22:18
    appropriate negative controls
  • 22:18 - 22:21
    availability and representatives of
  • 22:21 - 22:24
    relevant uh chemical classes for that
  • 22:24 - 22:26
    particular test
  • 22:26 - 22:30
    method the s um includes all of these
  • 22:30 - 22:34
    aspects listed here um and it needs to
  • 22:34 - 22:37
    be extremely well
  • 22:37 - 22:39
    detailed
  • 22:39 - 22:44
    um the first step uh with the protocol
  • 22:44 - 22:44
    with the
  • 22:44 - 22:48
    transfer um is often the step where
  • 22:48 - 22:53
    hiccups are found um Everybody every
  • 22:53 - 22:55
    Everybody tends to think that that
  • 22:55 - 22:58
    everybody else does the same sort of pro
  • 22:58 - 23:00
    process and sop and lab practice that
  • 23:00 - 23:04
    they do but it really isn't the case so
  • 23:04 - 23:07
    basic assumptions also need to be
  • 23:07 - 23:09
    included and these are the sorts of
  • 23:09 - 23:11
    things that we've really learned over
  • 23:11 - 23:13
    the years it's much better to include a
  • 23:13 - 23:16
    lot more detail than less to be able to
  • 23:16 - 23:19
    ensure that the lab uh that the test
  • 23:19 - 23:22
    method transfers to the lab
  • 23:22 - 23:25
    successfully for the module three with
  • 23:25 - 23:28
    transferability um the robustness of a
  • 23:28 - 23:30
    test assessment of reproducibility of
  • 23:30 - 23:33
    data in a second lab is assessed and the
  • 23:33 - 23:37
    prail practicability the successful
  • 23:37 - 23:40
    transfer and this is an iterative
  • 23:40 - 23:44
    process um and with amendments on on the
  • 23:44 - 23:47
    test definition and the protocol as
  • 23:47 - 23:50
    needed as the data is generated problems
  • 23:50 - 23:53
    arise and as they resolved um the
  • 23:53 - 23:55
    clarifications need to be
  • 23:55 - 23:58
    made and for the proficiency T of
  • 23:58 - 24:00
    chemicals now as we move to this they
  • 24:00 - 24:01
    need to be well understood to cover the
  • 24:01 - 24:04
    relevant modes of
  • 24:04 - 24:08
    action for the between lab variability
  • 24:08 - 24:10
    the objectives are the assessment of
  • 24:10 - 24:12
    reproducibility of the experimental data
  • 24:12 - 24:16
    in two to four labs and sometimes more
  • 24:16 - 24:19
    um the training and proficiency set of
  • 24:19 - 24:20
    chemicals need to be the most obvious
  • 24:20 - 24:23
    chemicals again um mode of action
  • 24:23 - 24:25
    potency physicochemical properties
  • 24:25 - 24:27
    negative chemicals and availability
  • 24:27 - 24:29
    considerations there's a slide coming up
  • 24:29 - 24:33
    that gives a a bit more of an overview
  • 24:33 - 24:35
    um more detailed overview shortly here
  • 24:35 - 24:36
    it
  • 24:36 - 24:39
    is so the Essential Elements for
  • 24:39 - 24:41
    chemical selection are to address the
  • 24:41 - 24:42
    sector
  • 24:42 - 24:45
    diversity um to have a structural
  • 24:45 - 24:47
    diversity to have a good number of
  • 24:47 - 24:51
    negatives at least 25% and ideally 50%
  • 24:51 - 24:53
    and it's actually this data that's often
  • 24:53 - 24:57
    the hardest to find the negative data
  • 24:57 - 24:59
    because it's it's often not
  • 24:59 - 25:03
    published um ideally as we try to use uh
  • 25:03 - 25:06
    nams and vitro test methods um more in
  • 25:06 - 25:09
    iarta and defined approaches uh the more
  • 25:09 - 25:12
    we have information on the range of weak
  • 25:12 - 25:15
    moderate and strong activity rather than
  • 25:15 - 25:18
    uh yes no response um the more the
  • 25:18 - 25:20
    greater applications that we we can use
  • 25:20 - 25:23
    that test method for um but it's not
  • 25:23 - 25:25
    always uh it really depends on the
  • 25:25 - 25:27
    quality of the data that's generated in
  • 25:27 - 25:28
    the in the study
  • 25:28 - 25:31
    um sometimes it's just not possible to
  • 25:31 - 25:35
    do that not at the early stages anyway
  • 25:35 - 25:37
    ideally also you have uh human invivo
  • 25:37 - 25:41
    data but um obviously for some classes
  • 25:41 - 25:44
    of chemicals such as uh pesticides um
  • 25:44 - 25:46
    it's generally quite unlikely to have
  • 25:46 - 25:49
    this sort of uh um concentration dose
  • 25:49 - 25:52
    response information um to a sufficient
  • 25:52 - 25:54
    degree whereas one might have it for
  • 25:54 - 25:56
    pharmaceuticals the Pharmaceuticals may
  • 25:56 - 26:00
    not be um holy relevant to the wider um
  • 26:00 - 26:04
    chemical um uh world that the test
  • 26:04 - 26:07
    method is intended to to be looking
  • 26:07 - 26:10
    at and then there are limitations and
  • 26:10 - 26:12
    restrictions on International um
  • 26:12 - 26:14
    transport and use that we need to
  • 26:14 - 26:18
    consider um uh chemicals that are on the
  • 26:18 - 26:20
    Stockholm pops convention that are
  • 26:20 - 26:22
    banned that shouldn't be really used
  • 26:22 - 26:25
    anymore um they ought to be considered
  • 26:25 - 26:27
    chemicals with abuse potential it's very
  • 26:27 - 26:31
    hard to import or export them in a
  • 26:31 - 26:34
    validation study so for example um well
  • 26:34 - 26:36
    drugs phob barbital cannabinoids
  • 26:36 - 26:39
    anabolic steroids and so on and then the
  • 26:39 - 26:42
    availability it can be very problematic
  • 26:42 - 26:44
    to have a chemical that it's really rare
  • 26:44 - 26:46
    and hard to find and there's no
  • 26:46 - 26:48
    long-term manufacturer or if it's really
  • 26:48 - 26:51
    expensive um it it becomes really
  • 26:51 - 26:53
    impracticable impr practical so these
  • 26:53 - 26:55
    considerations need to be uh included in
  • 26:55 - 26:57
    the chemical
  • 26:57 - 27:00
    selection
  • 27:00 - 27:04
    there is no fixed number of chemicals it
  • 27:04 - 27:07
    really needs to be on a Case specific
  • 27:07 - 27:09
    basis depending on the um literature and
  • 27:09 - 27:12
    data that you have on which you can base
  • 27:12 - 27:15
    your uh proposals for the the chemical
  • 27:15 - 27:18
    selection list and it's often a very
  • 27:18 - 27:22
    good idea to um get that list that you
  • 27:22 - 27:25
    derive get it peer-reviewed
  • 27:25 - 27:29
    independently um to and get Buy in um um
  • 27:29 - 27:31
    as you go forward and that's certainly
  • 27:31 - 27:33
    something at the oecd level with the
  • 27:33 - 27:34
    validation management groups and the
  • 27:34 - 27:36
    expert groups it's really good to have a
  • 27:36 - 27:39
    good discussion with them um if possible
  • 27:39 - 27:41
    at the outset to resolve any issues that
  • 27:41 - 27:44
    might come
  • 27:46 - 27:48
    later
  • 27:48 - 27:52
    whoops um designs of a validation study
  • 27:52 - 27:57
    so um what is not a good idea is to have
  • 27:57 - 28:00
    loads of Laboratories and just a handful
  • 28:00 - 28:02
    of chemicals that are being tested um
  • 28:02 - 28:03
    because then you're not really
  • 28:03 - 28:05
    demonstrating the applicability the
  • 28:05 - 28:09
    chemical applicability of the of the
  • 28:09 - 28:12
    assay however um sometimes you don't
  • 28:12 - 28:15
    really have a choice about uh the
  • 28:15 - 28:20
    validation Lab Partners um if if it
  • 28:20 - 28:22
    depends really on the funding stream but
  • 28:22 - 28:25
    certainly uh for example if a cro is
  • 28:25 - 28:27
    interested in participating in a test
  • 28:27 - 28:30
    method because they want to demonstrate
  • 28:30 - 28:34
    um well they want to grain gain skills
  • 28:34 - 28:36
    in in running that assay and they will
  • 28:36 - 28:38
    have a centel um that will want to use
  • 28:38 - 28:42
    it um it's not really
  • 28:42 - 28:45
    uh I mean if they want to participate
  • 28:45 - 28:49
    it's uh really to be encouraged so
  • 28:49 - 28:51
    sometimes one cannot Define the number
  • 28:51 - 28:55
    of uh or or number of Laboratories uh
  • 28:55 - 28:58
    fully upfront but as long as you have
  • 28:58 - 29:05
    um a minimum of of three um it's it's uh
  • 29:05 - 29:08
    encouraged in the expert
  • 29:09 - 29:12
    groups um for past validation studies
  • 29:12 - 29:15
    for example the situation on the left
  • 29:15 - 29:18
    has often been the case where a number
  • 29:18 - 29:21
    of different chemicals are run in all
  • 29:21 - 29:24
    three Labs but it was Al it's also been
  • 29:24 - 29:28
    commonly done to stagger um or have
  • 29:28 - 29:31
    phases in the validation study so that
  • 29:31 - 29:35
    the you have your first core uh uh test
  • 29:35 - 29:37
    chemical or proficiency chemicals in the
  • 29:37 - 29:39
    labs that are run by everybody and they
  • 29:39 - 29:41
    all demonstrate that they can reproduce
  • 29:41 - 29:44
    them well um and then you can expand
  • 29:44 - 29:46
    your chemical chemical applicability
  • 29:46 - 29:48
    doain by testing in a number of
  • 29:48 - 29:50
    different Labs or very often the lead
  • 29:50 - 29:52
    lab will often test a lot more of the
  • 29:52 - 29:54
    chemicals um to be able to demonstrate
  • 29:54 - 29:59
    the chemical applicability domain range
  • 30:03 - 30:05
    independent statistical support is
  • 30:05 - 30:07
    needed for within lab variability and
  • 30:07 - 30:11
    between lab variability and predictive
  • 30:11 - 30:13
    capacity
  • 30:13 - 30:16
    um and the TR traditional
  • 30:16 - 30:17
    process
  • 30:17 - 30:22
    um uh can take uh well this slide has
  • 30:22 - 30:24
    been used this interpretations of these
  • 30:24 - 30:29
    slides have been used um a lot um to
  • 30:29 - 30:32
    show the length of time that can it can
  • 30:32 - 30:36
    take um but sometimes it's been used to
  • 30:36 - 30:38
    show that it can take a very long time
  • 30:38 - 30:40
    and that's really not necessarily the
  • 30:40 - 30:43
    case um research and
  • 30:43 - 30:46
    development can take a great many years
  • 30:46 - 30:49
    um the prevalidation process um
  • 30:49 - 30:52
    can include an awful lot of
  • 30:52 - 30:54
    optimization of that research and
  • 30:54 - 30:57
    development uh process and sometimes it
  • 30:57 - 30:58
    sort of goes backwards and you're really
  • 30:58 - 30:59
    doing
  • 30:59 - 31:03
    R&D um but on average it can take two
  • 31:03 - 31:05
    years perhaps sometimes it will take a
  • 31:05 - 31:06
    bit more especially if there are
  • 31:06 - 31:07
    problems
  • 31:07 - 31:10
    arising the validation stage um the
  • 31:10 - 31:12
    getting to full validation where you
  • 31:12 - 31:16
    start to have uh blind chemicals um
  • 31:16 - 31:18
    which I'm not discussing any detail in
  • 31:18 - 31:21
    this one um that can take a year or it
  • 31:21 - 31:25
    can take a lot more it really um it it
  • 31:25 - 31:27
    will really depend on a case-by casee
  • 31:27 - 31:30
    basis the peer review uh independent
  • 31:30 - 31:33
    peer review can take a year and then the
  • 31:33 - 31:36
    oecd regulatory acceptance usually takes
  • 31:36 - 31:38
    around two years but it can take more
  • 31:38 - 31:40
    for a number of different reasons which
  • 31:40 - 31:43
    we will come to in the in the next
  • 31:43 - 31:48
    section and then um uh acceptance and
  • 31:48 - 31:50
    writing in to different regulatory
  • 31:50 - 31:53
    jurisdictions um often happens following
  • 31:53 - 31:56
    the oecd regulatory
  • 31:56 - 31:59
    acceptance so so the first stages here
  • 31:59 - 32:04
    um with the R&D um can happen at the
  • 32:04 - 32:11
    academic uh uh industry um uh um side um
  • 32:11 - 32:16
    or agency side as well um and then
  • 32:16 - 32:22
    really ideally uh these validation steps
  • 32:22 - 32:24
    um sometimes they are started
  • 32:24 - 32:27
    independently or by industry groups or
  • 32:27 - 32:29
    academ MC groups or so but they can
  • 32:29 - 32:33
    start to enter the um OC OCD test
  • 32:33 - 32:37
    guideline program at the pre valid valid
  • 32:37 - 32:42
    prevalidation stage and um it's really
  • 32:42 - 32:46
    good if if projects do enter um at that
  • 32:46 - 32:49
    stage to get support from the
  • 32:49 - 32:52
    international experts and nomin expert
  • 32:52 - 32:55
    nominations from the working group of
  • 32:55 - 32:57
    national coordinators that at that sort
  • 32:57 - 32:58
    of State
  • 32:58 - 33:01
    agage um because then you know you're
  • 33:01 - 33:03
    going to be resolving problems that
  • 33:03 - 33:05
    might come up later with respect to
  • 33:05 - 33:07
    regulatory acceptance you have a chance
  • 33:07 - 33:10
    to to be able to address those um and
  • 33:10 - 33:12
    get
  • 33:15 - 33:19
    support intellectual property um has
  • 33:19 - 33:23
    become a bit of a uh an issue also um in
  • 33:23 - 33:25
    the oecd test guideline program as we
  • 33:25 - 33:29
    have moved more to inv vitro test
  • 33:29 - 33:32
    methods um there often are intellectual
  • 33:32 - 33:36
    property um and patent issues around
  • 33:36 - 33:38
    some of the test methods it's not a
  • 33:38 - 33:43
    problem at all now um and uh specific
  • 33:43 - 33:45
    guidance has been developed regarding
  • 33:45 - 33:48
    these elements in oecd desk guidelines
  • 33:48 - 33:51
    it's important to declare whether the
  • 33:51 - 33:54
    test method contains predicted elements
  • 33:54 - 33:56
    right at the outset when the project is
  • 33:56 - 33:59
    submitted to the ACD test guideline
  • 33:59 - 34:01
    program it needs to be explicitly
  • 34:01 - 34:04
    indicated and the means foreseen to make
  • 34:04 - 34:05
    these protected elements available to
  • 34:05 - 34:09
    users if their method becomes a test
  • 34:09 - 34:12
    guideline so it's not a handicap it just
  • 34:12 - 34:15
    needs to be
  • 34:18 - 34:21
    transparent
  • 34:21 - 34:24
    um the test method needs to ideally it
  • 34:24 - 34:26
    will address high priority needs for the
  • 34:26 - 34:31
    test guideline Pro program um and uh a
  • 34:31 - 34:33
    guidance document for the development of
  • 34:33 - 34:34
    oecd guidelines for the testing of
  • 34:34 - 34:38
    chemicals um uh from
  • 34:38 - 34:41
    2009 states that
  • 34:41 - 34:45
    um only that the development or update
  • 34:45 - 34:47
    of a test guideline is Possible only
  • 34:47 - 34:48
    when the proposed test method or
  • 34:48 - 34:50
    modifications to the existing test
  • 34:50 - 34:52
    guideline has undergone a critical
  • 34:52 - 34:54
    appraisal concerning its validation and
  • 34:54 - 34:56
    Regulatory
  • 34:56 - 34:58
    acceptance including where feasible and
  • 34:58 - 35:01
    relevant an interlaboratory comparative
  • 35:01 - 35:03
    study I a ring test or something like
  • 35:03 - 35:05
    that describing the performance of the
  • 35:05 - 35:08
    test method and then the current
  • 35:08 - 35:12
    guidance document um um on validation
  • 35:12 - 35:13
    from
  • 35:13 - 35:16
    2005 also states that test method
  • 35:16 - 35:18
    validation is a process based upon
  • 35:18 - 35:21
    scientifically sound principles by which
  • 35:21 - 35:23
    the reliab reliability and relevance of
  • 35:23 - 35:26
    a particular test approach method or
  • 35:26 - 35:28
    process
  • 35:28 - 35:31
    is established for a specific
  • 35:32 - 35:35
    purpose so now this is the one slide
  • 35:35 - 35:39
    coming up um giving examples of what can
  • 35:39 - 35:42
    grow go wrong and why um but this is
  • 35:42 - 35:44
    what I will spend most of the time
  • 35:44 - 35:48
    on so some common stumbling blocks or
  • 35:48 - 35:50
    hurdles in the oecd test guideline
  • 35:50 - 35:53
    process first of all they can arise with
  • 35:53 - 35:56
    reproducibility and transfer of the test
  • 35:56 - 35:59
    method um and experimental Corrections
  • 35:59 - 36:01
    need to be conducted and there there may
  • 36:01 - 36:03
    be missing characterization of the test
  • 36:03 - 36:06
    method and its cell line these have
  • 36:06 - 36:09
    happened in quite a few um examples of
  • 36:09 - 36:12
    adopted test guidelines as well as those
  • 36:12 - 36:15
    uh test methods that have failed uh to
  • 36:15 - 36:17
    become
  • 36:17 - 36:20
    adopted so for example one of one of the
  • 36:20 - 36:23
    um endocrine active uh um test
  • 36:23 - 36:27
    guidelines early on um uh the test
  • 36:27 - 36:31
    method um failed to be reproducible in
  • 36:31 - 36:34
    another lab um even by the test method
  • 36:34 - 36:37
    developers and the cell line and the
  • 36:37 - 36:39
    constract had to be changed but then
  • 36:39 - 36:41
    ultimately it ended up being a
  • 36:41 - 36:43
    successful uh successfully adopted as a
  • 36:43 - 36:47
    test guideline it it worked it it was
  • 36:47 - 36:50
    fine um in another case we've had a um
  • 36:50 - 36:53
    cases a case of
  • 36:53 - 36:56
    uh uh well to very early on we didn't do
  • 36:56 - 36:58
    suff efficient characterization of the
  • 36:58 - 37:00
    invitra test methods or the cell line
  • 37:00 - 37:05
    and we had an instance um where another
  • 37:05 - 37:08
    um test method um was found not to be
  • 37:08 - 37:11
    the cell line that it was stated to be
  • 37:11 - 37:15
    um during the validation and at the upon
  • 37:15 - 37:17
    adoption and it was later found out by
  • 37:17 - 37:19
    another group um that the
  • 37:19 - 37:21
    characterization was wrong and the cell
  • 37:21 - 37:22
    line had been
  • 37:22 - 37:25
    contaminated the validation had been
  • 37:25 - 37:28
    successful the cell the test method had
  • 37:28 - 37:31
    been adopted at a test guideline and um
  • 37:31 - 37:34
    a clarification sentence was uh provided
  • 37:34 - 37:36
    within that test
  • 37:36 - 37:38
    guideline sometimes the material
  • 37:38 - 37:40
    supporting the performance or validity
  • 37:40 - 37:43
    of the test method is not
  • 37:43 - 37:47
    adequate um and so the greater detail
  • 37:47 - 37:50
    given the better for all of the
  • 37:50 - 37:52
    evaluations the references the
  • 37:52 - 37:55
    validation reports and making sure that
  • 37:55 - 37:57
    they're all accessible some times there
  • 37:57 - 38:00
    is a lack of adherence to validation
  • 38:00 - 38:03
    criteria um sometimes there can be
  • 38:03 - 38:04
    issues regarding reference or
  • 38:04 - 38:07
    proficiency chemicals so here an example
  • 38:07 - 38:10
    is the draft sip induction test
  • 38:10 - 38:14
    guideline where um although successfully
  • 38:14 - 38:20
    validated uh by the uh the jrc the wnt
  • 38:20 - 38:24
    wanted to see an expansion of the um
  • 38:24 - 38:27
    reference uh chemicals um so work on
  • 38:27 - 38:31
    that is ongoing at the moment but um and
  • 38:31 - 38:35
    that's it's resolvable um and there were
  • 38:35 - 38:37
    there was a very strong justification
  • 38:37 - 38:40
    for the uh chemicals that were used in
  • 38:40 - 38:43
    the original validation because they
  • 38:43 - 38:45
    were the only chemicals that were human
  • 38:45 - 38:46
    data was
  • 38:46 - 38:48
    available
  • 38:48 - 38:51
    um so this has slowed down the adoption
  • 38:51 - 38:53
    unfortunately of this particular test
  • 38:53 - 38:55
    guideline but it's
  • 38:55 - 38:57
    resolvable other
  • 38:57 - 39:00
    um cons considerations might be a poorly
  • 39:00 - 39:02
    written draft test guideline or
  • 39:02 - 39:04
    sometimes to be perfectly honest the
  • 39:04 - 39:05
    adopted test
  • 39:05 - 39:08
    guideline we find out later it's not
  • 39:08 - 39:12
    sufficiently clear um for the end users
  • 39:12 - 39:14
    and we need to go back and revise the
  • 39:14 - 39:16
    data interpretation procedure we need to
  • 39:16 - 39:19
    clarify the language but it's really
  • 39:19 - 39:23
    only after trying it and having it out
  • 39:23 - 39:27
    there um that we are able to
  • 39:27 - 39:30
    do those Corrections and we've had to do
  • 39:30 - 39:31
    that a couple of times recently with the
  • 39:31 - 39:35
    TG uh 456 for
  • 39:35 - 39:39
    example sometimes the there's a changed
  • 39:39 - 39:41
    regulatory purpose of a test guideline
  • 39:41 - 39:44
    or it's not clearly defined and
  • 39:44 - 39:47
    increasingly the WT really does want to
  • 39:47 - 39:52
    see um a clear indication of the
  • 39:52 - 39:55
    regulatory purpose of the test guideline
  • 39:55 - 39:57
    and how it can be used how it can fit it
  • 39:57 - 40:00
    um particularly now for inv vitro test
  • 40:00 - 40:02
    methods for
  • 40:03 - 40:05
    nams sometimes there's a low level of
  • 40:05 - 40:08
    commitment by the lead country um there
  • 40:08 - 40:10
    can be a staff change or reduction in
  • 40:10 - 40:12
    res resources and the project needs to
  • 40:12 - 40:16
    be taken over uh by another country or
  • 40:16 - 40:19
    or it just gets put to one side for a
  • 40:19 - 40:24
    year or two until someone is willing or
  • 40:24 - 40:27
    able um to take the project over or it
  • 40:27 - 40:30
    will get ultimately dropped if no one is
  • 40:30 - 40:32
    willing to take it
  • 40:32 - 40:37
    over um and you know we all move jobs um
  • 40:37 - 40:39
    uh resources change
  • 40:39 - 40:45
    so it's it's uh it's it's a realistic um
  • 40:45 - 40:48
    uh problem that that
  • 40:48 - 40:51
    occurs sometimes a project proposal is
  • 40:51 - 40:54
    discussed at the uh working group of
  • 40:54 - 40:55
    national
  • 40:55 - 40:57
    coordinators and um further work is
  • 40:57 - 41:01
    really needed to explore issues within
  • 41:01 - 41:03
    that work that proposal there might be a
  • 41:03 - 41:05
    need for an expert consultation then
  • 41:05 - 41:07
    might be a need for a stronger evidence
  • 41:07 - 41:10
    base for the proposal and a detailed
  • 41:10 - 41:13
    review paper might be requested first
  • 41:13 - 41:15
    and that has happened several times as
  • 41:15 - 41:17
    well with the test uh test method
  • 41:17 - 41:20
    proposals going forward to the
  • 41:20 - 41:24
    w&t so we have that um at the moment for
  • 41:24 - 41:28
    um uh just thinking of some
  • 41:28 - 41:32
    um uh with a Maps test for example uh
  • 41:32 - 41:34
    we've been doing a more detailed report
  • 41:34 - 41:36
    and having an expert cons independent
  • 41:36 - 41:39
    expert uh consultation for a project on
  • 41:39 - 41:42
    the on the work
  • 41:42 - 41:48
    plan sometimes things go arai because um
  • 41:48 - 41:50
    of how the expert groups are working
  • 41:50 - 41:54
    there can be um insufficient briefing or
  • 41:54 - 41:57
    management of or listening
  • 41:57 - 42:00
    uh to new review experts or national
  • 42:00 - 42:03
    coordinators and the interactions there
  • 42:03 - 42:06
    or the recordkeeping not might not be
  • 42:06 - 42:09
    sufficiently detailed enough and so
  • 42:09 - 42:11
    keeping really good records of all of
  • 42:11 - 42:15
    the record of the discussions and
  • 42:15 - 42:18
    ensuring uh agreement on those records
  • 42:18 - 42:20
    as one progresses really makes a
  • 42:20 - 42:23
    difference uh in achieving
  • 42:23 - 42:27
    efficiencies and briefing uh new new
  • 42:27 - 42:29
    reviewers that may may join um because
  • 42:29 - 42:32
    often there can be circular discussions
  • 42:32 - 42:37
    that start to occur um as uh uh uh ideas
  • 42:37 - 42:40
    or concerns expressed earlier and dealt
  • 42:40 - 42:44
    with um get reiterated um by people who
  • 42:44 - 42:46
    were not in experts who were not
  • 42:46 - 42:49
    involved in the work
  • 42:49 - 42:52
    previously sometimes
  • 42:52 - 42:58
    um uh it's uh the National
  • 42:58 - 42:59
    coordinators
  • 42:59 - 43:02
    don't consider that the test method
  • 43:02 - 43:06
    addresses a key endpoint Gap or there
  • 43:06 - 43:07
    are a lack of lead countries coming
  • 43:07 - 43:10
    forward to address an oecd identified
  • 43:10 - 43:12
    cap
  • 43:12 - 43:15
    so for example with a thyroid um
  • 43:15 - 43:19
    validation effort at the moment on going
  • 43:19 - 43:23
    um um being led by the jrc but with
  • 43:23 - 43:25
    there the discussions at the oecd expert
  • 43:25 - 43:28
    group um countries do need to take a
  • 43:28 - 43:32
    lead and come forward uh to address to
  • 43:32 - 43:35
    to to address these assays and these are
  • 43:35 - 43:37
    assays that were first identified um
  • 43:37 - 43:41
    over 10 years ago in an oecd thyroid um
  • 43:41 - 43:44
    scoping document so the delay there has
  • 43:44 - 43:46
    really been a lack of countries coming
  • 43:46 - 43:49
    forward to do the work um the jrc came
  • 43:49 - 43:53
    forward and have done uh um the
  • 43:53 - 43:56
    preliminary or the prevalidation work
  • 43:56 - 43:57
    for
  • 43:57 - 43:59
    a large number of the assays I think 17
  • 43:59 - 44:03
    or 18 or more um but still uh more
  • 44:03 - 44:05
    commitment is needed from from lead
  • 44:05 - 44:07
    countries to take forward some of the
  • 44:07 - 44:12
    these work these these pieces of work um
  • 44:12 - 44:14
    and some of them are so for example
  • 44:14 - 44:17
    Peppa are taking forward at least two I
  • 44:17 - 44:20
    think of the thyroid assays now but more
  • 44:20 - 44:22
    countries are needed to take these
  • 44:22 - 44:24
    forward sometimes there are simp it's
  • 44:24 - 44:26
    simply conflict of regulatory needs
  • 44:26 - 44:28
    across R different
  • 44:28 - 44:32
    countries um and that means that the the
  • 44:32 - 44:33
    project won't be taken
  • 44:33 - 44:36
    up sometimes also there can be cultural
  • 44:36 - 44:38
    safety differences or there can be
  • 44:38 - 44:41
    issues um around things that you think
  • 44:41 - 44:44
    are easily available so for example
  • 44:44 - 44:46
    human serum use in invitro test
  • 44:46 - 44:49
    guidelines um a UK project proposal that
  • 44:49 - 44:54
    was made several years ago um uh a cro
  • 44:54 - 44:58
    came forward and um um uh was able to
  • 44:58 - 45:01
    adapt some of the skin sensitization
  • 45:01 - 45:02
    assays
  • 45:02 - 45:07
    442 um for human serum use and they had
  • 45:07 - 45:11
    obtained the serum from
  • 45:11 - 45:13
    Sigma it was openly available it
  • 45:13 - 45:15
    appeared to be uh
  • 45:15 - 45:19
    regulated um but actually it turned out
  • 45:19 - 45:22
    not to be um and the issues that arose
  • 45:22 - 45:25
    around the discussion at the w&t table
  • 45:25 - 45:29
    really start started to unpe unpeel or
  • 45:29 - 45:32
    expose a lot of different issues that we
  • 45:32 - 45:34
    needed to consider when using human
  • 45:34 - 45:36
    tissue including competition with
  • 45:36 - 45:40
    medical use um as well as exploitation
  • 45:40 - 45:45
    of of donors for example so that led to
  • 45:45 - 45:47
    a much deeper piece of work and in
  • 45:47 - 45:51
    including um um workshops to to address
  • 45:51 - 45:55
    those those problems so and these things
  • 45:55 - 45:56
    are not obvious when they they first
  • 45:56 - 46:00
    come so and they need to be resolved um
  • 46:00 - 46:02
    as we go through the process to be able
  • 46:02 - 46:06
    to achieve uh full adoption as a test
  • 46:06 - 46:10
    guideline under mad so there can be this
  • 46:10 - 46:12
    and here's just this is just a start of
  • 46:12 - 46:15
    a list really of really common um or
  • 46:15 - 46:17
    they not I wouldn't say they're really
  • 46:17 - 46:21
    common but they really concrete examples
  • 46:21 - 46:25
    of where there have been um
  • 46:25 - 46:28
    stumblings uh with respect to test
  • 46:28 - 46:31
    method adoption at the
  • 46:32 - 46:35
    oecd now let's look at practical
  • 46:35 - 46:38
    purposes what can be done right to begin
  • 46:38 - 46:40
    with
  • 46:40 - 46:42
    um
  • 46:42 - 46:46
    really it's it's really if validation
  • 46:46 - 46:48
    needs can be incorporated right at the
  • 46:48 - 46:51
    outset into research outputs the more we
  • 46:51 - 46:54
    do that as we develop our work in the
  • 46:54 - 46:57
    R&D side as it comes as the test methods
  • 46:57 - 47:00
    come through to um optimization the more
  • 47:00 - 47:02
    that we have a detailed description of
  • 47:02 - 47:02
    the
  • 47:02 - 47:05
    SOP the more we publish the data
  • 47:05 - 47:09
    associated with the result um the more
  • 47:09 - 47:12
    that the DAT data is findable accessible
  • 47:12 - 47:15
    interoperable reusable not as a PDF but
  • 47:15 - 47:17
    in a machine readable
  • 47:17 - 47:20
    format the more efficient the whole
  • 47:20 - 47:22
    process will become and the more
  • 47:22 - 47:24
    transparent it will become which will
  • 47:24 - 47:27
    increase Trust there is insufficient
  • 47:27 - 47:28
    space in scientific literature
  • 47:28 - 47:30
    manuscripts for all of this sort of
  • 47:30 - 47:34
    information but there are um websites
  • 47:34 - 47:37
    where one can um work on protocols
  • 47:37 - 47:39
    together um there is uh I mentioned
  • 47:39 - 47:42
    right at the outset this um prom map um
  • 47:42 - 47:46
    initiative and guidance um developed uh
  • 47:46 - 47:49
    by staff at the um
  • 47:49 - 47:51
    efam
  • 47:51 - 47:55
    um this is most critical step to really
  • 47:55 - 47:56
    uh
  • 47:56 - 48:00
    get on top of um right to begin with
  • 48:00 - 48:03
    with respect to the smooth operation of
  • 48:03 - 48:03
    of
  • 48:03 - 48:07
    validation um and the metha transfer is
  • 48:07 - 48:09
    often delayed at this
  • 48:09 - 48:11
    step and here are just a list of
  • 48:11 - 48:16
    resources that one can go to to um
  • 48:16 - 48:19
    improve the quality of this the SOP and
  • 48:19 - 48:23
    the data generated so I've just taken
  • 48:23 - 48:27
    I've taken mainly European examples
  • 48:27 - 48:32
    um um and oecd examples here and there
  • 48:32 - 48:34
    are there are a lot more internationally
  • 48:34 - 48:38
    but here here is just a
  • 48:38 - 48:41
    selection um the pepper platform has
  • 48:41 - 48:44
    particularly uh provided some very nice
  • 48:44 - 48:47
    uh guidance on how to um uh report test
  • 48:47 - 48:50
    method Readiness for example and there
  • 48:50 - 48:52
    are a number of different guidance
  • 48:52 - 48:54
    documents at the oecd the is
  • 48:54 - 48:56
    critical for test methods
  • 48:56 - 49:00
    developers um GD 211 for describing
  • 49:00 - 49:02
    non-guideline inv vitro tests and that's
  • 49:02 - 49:05
    the basis of uh of these uh test method
  • 49:05 - 49:07
    Readiness um
  • 49:07 - 49:09
    um uh
  • 49:09 - 49:13
    spreadsheets um reusable and open
  • 49:13 - 49:16
    protocols
  • 49:16 - 49:18
    um I'm so so on so there's a whole a
  • 49:18 - 49:22
    whole uh uh list of resources here to to
  • 49:22 - 49:24
    make use
  • 49:24 - 49:27
    of and with that thank you for your
  • 49:27 - 49:29
    attention and um we can have some
  • 49:29 - 49:31
    discussion and questions
  • 49:31 - 49:34
    now
Title:
OECD Webinar | Training and orientation on validation of methods at the OECD
Description:

more » « less
Video Language:
Greek
Duration:
49:33

English subtitles

Incomplete

Revisions Compare revisions