Keynote Presentation: A Regulatory Perspective on Molecular Diagnostic Devices

00:50:57
https://www.youtube.com/watch?v=En8C5VQs32g

Résumé

TLDRDr. Eunice Lee's presentation provides an overview of the regulatory landscape for molecular diagnostic devices, particularly focusing on companion diagnostics and next-generation sequencing (NGS) oncology panels. She explains the classification of in vitro diagnostic devices (IVDs) based on risk, the evaluation of their safety and effectiveness, and the regulatory pathways for different classes of devices. The presentation highlights the importance of companion diagnostics in relation to therapeutic products, the challenges of validating NGS tests, and recent FDA authorizations that have influenced the regulatory framework. Dr. Lee emphasizes the need for collaboration between diagnostic and drug sponsors and discusses the evolving regulatory models in personalized medicine, underscoring the significance of valid scientific evidence in the review process.

A retenir

  • 🔍 Overview of IVD regulation and classification
  • 📊 Importance of analytical and clinical validation
  • 💡 Companion diagnostics are essential for therapeutic use
  • ⚖️ Regulatory pathways differ by device class
  • 🧬 NGS tests present unique validation challenges
  • 🤝 Collaboration is key between stakeholders
  • 📅 Recent FDA authorizations shape regulatory landscape
  • 📈 Parallel review program expedites coverage decisions
  • 📚 Resources available for further information
  • 🔗 Contact for follow-up questions

Chronologie

  • 00:00:00 - 00:05:00

    The live broadcast introduces Dr. Eunice Lee from the FDA, who will discuss the regulatory perspective on molecular diagnostic devices, particularly companion diagnostics and next-generation sequencing (NGS). The event encourages audience interaction through questions and provides information on continuing education credits.

  • 00:05:00 - 00:10:00

    Dr. Lee outlines the FDA's role in reviewing in vitro diagnostic devices (IVDs), which include various tests and systems for diagnosing diseases. She emphasizes that the information shared is for discussion purposes and refers to FDA resources for specific inquiries.

  • 00:10:00 - 00:15:00

    The FDA evaluates IVDs based on valid scientific evidence to ensure safety and effectiveness. Devices are classified into three categories (Class 1, 2, and 3) based on risk, with Class 1 being low risk and Class 3 being high risk, requiring more stringent controls.

  • 00:15:00 - 00:20:00

    The intended use of a device is crucial for its classification, which determines the regulatory pathway. Class 1 devices may be exempt from premarket notification, while Class 2 devices typically require a 510(k) submission, and Class 3 devices require a premarket approval (PMA).

  • 00:20:00 - 00:25:00

    The FDA review process includes assessing intended use, performance, labeling, and compliance with quality assurance regulations. Analytical and clinical performance validations are essential for demonstrating reliability and accuracy of the tests.

  • 00:25:00 - 00:30:00

    Companion diagnostics are a specific subset of IVDs that provide essential information for the safe and effective use of corresponding therapeutics. Their approval is often tied to the approval of the therapeutic product, requiring close collaboration between regulatory bodies.

  • 00:30:00 - 00:35:00

    Key questions regarding companion diagnostics include the necessity of the test, the need for an investigational device exemption (IDE) for clinical trials, and how to handle changes in the assay used during trials. Collaboration between drug and device sponsors is critical.

  • 00:35:00 - 00:40:00

    The need for companion diagnostics arises when there is adequate evidence of clinical activity in the biomarker-selected population. An IDE may be required if the test poses significant risk during trials, and bridging studies may be necessary if the clinical trial assay differs from the final diagnostic.

  • 00:40:00 - 00:45:00

    Complementary diagnostics, unlike companion diagnostics, are not essential for drug use but help identify patient subsets that may benefit from specific therapies. Their development path is similar to companion diagnostics, but they do not require the same level of regulatory scrutiny.

  • 00:45:00 - 00:50:57

    Next-generation sequencing (NGS) oncology panels present unique challenges due to their ability to screen numerous biomarkers simultaneously. Validation for these tests is complex, and the FDA recommends representative approaches for validating variants, especially for those without companion diagnostic claims.

Afficher plus

Carte mentale

Vidéo Q&R

  • What is the focus of Dr. Eunice Lee's presentation?

    The presentation focuses on the regulatory perspective of molecular diagnostic devices, particularly companion diagnostics and next-generation sequencing (NGS) oncology panels.

  • What are companion diagnostics?

    Companion diagnostics are tests that provide essential information for the safe and effective use of a corresponding therapeutic.

  • How are in vitro diagnostic devices classified?

    IVDs are classified into three classes based on risk: Class 1 (low risk), Class 2 (moderate risk), and Class 3 (high risk), with varying levels of regulatory controls.

  • What is the significance of the FDA's review process for IVDs?

    The FDA review process ensures that devices are safe and effective for their intended use, relying on valid scientific evidence.

  • What challenges are associated with validating NGS tests?

    Challenges include the inability to evaluate every variant, limited clinical specimens, and lack of reference methods for accuracy.

  • What is the role of the IDE application in clinical trials?

    An IDE application permits the lawful shipment of a device for investigation without complying with other requirements of the Food Drug and Cosmetic Act.

  • What are the three levels of biomarkers in NGS oncology panels?

    Level 1 biomarkers have companion diagnostic claims, Level 2 biomarkers are clinically significant mutations, and Level 3 biomarkers have potential clinical significance.

  • What is the purpose of the FDA CMS parallel review program?

    The program aims to reduce the time between FDA approval of a device and CMS Medicare coverage.

  • How does the FDA evaluate the performance of IVDs?

    The FDA evaluates performance through analytical and clinical validation, ensuring reliability and accuracy of test results.

  • What resources are available for more information on the topics discussed?

    Additional resources can be found on the FDA website and through the links provided in the presentation.

Voir plus de résumés vidéo

Accédez instantanément à des résumés vidéo gratuits sur YouTube grâce à l'IA !
Sous-titres
en
Défilement automatique:
  • 00:00:01
    hello everyone and welcome to today's
  • 00:00:03
    live broadcasts and keynote presentation
  • 00:00:05
    a regulatory perspective on molecular
  • 00:00:08
    diagnostic devices presented by dr.
  • 00:00:11
    Eunice Lee chief of the molecular
  • 00:00:13
    pathology and cytology branch in the
  • 00:00:16
    office of in vitro diagnostics and
  • 00:00:18
    radiological health at the US Food and
  • 00:00:21
    Drug Administration
  • 00:00:22
    my name is Gary Stein and I'll be your
  • 00:00:25
    moderator for today's event we're
  • 00:00:27
    delighted to bring you this educational
  • 00:00:28
    web seminar presented by lab roots lab
  • 00:00:32
    roots is the leading scientific social
  • 00:00:35
    networking website and producer of
  • 00:00:37
    educational virtual events and webinars
  • 00:00:40
    before we begin I would like to remind
  • 00:00:42
    everyone that this event is interactive
  • 00:00:45
    we encourage you to participate by
  • 00:00:47
    submitting as many questions as you want
  • 00:00:49
    at any time you want during the
  • 00:00:51
    presentation just click on the ask a
  • 00:00:54
    question box located on the far left of
  • 00:00:57
    your screen and type your questions into
  • 00:00:59
    the drop-down box that appears on the
  • 00:01:02
    screen dr. Lee will be responding to
  • 00:01:04
    your questions via email if you have
  • 00:01:07
    trouble seeing or hearing the
  • 00:01:08
    presentation please click on the help
  • 00:01:11
    desk button located in the promotional
  • 00:01:14
    board at the bottom slash center of your
  • 00:01:16
    screen or use the ask a question box to
  • 00:01:20
    let us know that you're having a problem
  • 00:01:22
    this presentation is educational and
  • 00:01:25
    thus offers continuing education credits
  • 00:01:27
    please click on the continuing education
  • 00:01:30
    credits tab located in the top right
  • 00:01:33
    corner of the presentation window and
  • 00:01:35
    follow the process to obtain your
  • 00:01:38
    credits today's webcast will be
  • 00:01:40
    available for on-demand viewing through
  • 00:01:43
    July of 2018 so please join me in
  • 00:01:47
    welcoming dr. Eunice Lee I will now turn
  • 00:01:49
    the presentation over to her
  • 00:01:56
    thank you very much for that
  • 00:01:57
    introduction Gary and I'd like to extend
  • 00:02:00
    a thank you to the organizers of this
  • 00:02:02
    event as mentioned my group is the
  • 00:02:07
    molecular pathology and cytology branch
  • 00:02:10
    and it's within the division of
  • 00:02:11
    molecular genetics and pathology at the
  • 00:02:14
    FDA we review a variety of invitro
  • 00:02:17
    diagnostic devices ranging from whole
  • 00:02:21
    slide imaging systems to
  • 00:02:22
    immunohistochemistry tests or IHC tests
  • 00:02:25
    and nucleic acid-based tests for some
  • 00:02:29
    oncology indications so today I'll
  • 00:02:32
    provide a regulatory perspective on
  • 00:02:34
    molecular diagnostic devices with
  • 00:02:37
    particular focus on companion
  • 00:02:39
    diagnostics and next-generation
  • 00:02:41
    sequencing based on ecology panels
  • 00:02:45
    before I begin I would just like to note
  • 00:02:47
    that the information presented today are
  • 00:02:51
    only intended for summary and discussion
  • 00:02:54
    purposes if there are specific questions
  • 00:02:56
    about any of the examples that I talked
  • 00:03:00
    about today about the specific devices
  • 00:03:01
    please refer to the decision summaries
  • 00:03:04
    and the summaries of safety and
  • 00:03:06
    effectiveness data or SS EDS
  • 00:03:08
    which are publicly available documents
  • 00:03:10
    on the FDA website and at the last slide
  • 00:03:14
    I'll have some a link to the medical
  • 00:03:16
    device database where this information
  • 00:03:18
    can be accessed today I'll start by
  • 00:03:24
    providing an overview of the regulation
  • 00:03:27
    of in vitro diagnostic devices or IV DS
  • 00:03:30
    then I'll touch on some assay
  • 00:03:32
    performance considerations and discuss
  • 00:03:34
    how we evaluate IV D products then I'll
  • 00:03:38
    move to companion diagnostics providing
  • 00:03:42
    an overview and discussing some key
  • 00:03:44
    issues in Co development and then I'll
  • 00:03:47
    move beyond companion diagnostics and
  • 00:03:49
    briefly talk about follow-on companion
  • 00:03:51
    diagnostics and complementary
  • 00:03:53
    diagnostics and then I'll end by
  • 00:03:56
    discussing NGS or next-generation
  • 00:03:58
    sequencing oncology panels I'll talk
  • 00:04:02
    about the validation considerations for
  • 00:04:04
    these types of
  • 00:04:04
    tests and also provide some examples of
  • 00:04:08
    recent FDA authorizations that have
  • 00:04:11
    significantly impacted the current
  • 00:04:13
    regulatory paradigm for these types of
  • 00:04:15
    devices so what is an IV D in the Code
  • 00:04:22
    of Federal Regulations or the CFR which
  • 00:04:25
    is basically the set of rules and
  • 00:04:26
    regulations by which we abide an IV D is
  • 00:04:30
    defined as reagents instruments and
  • 00:04:32
    systems intended for use in diagnosis of
  • 00:04:35
    disease or other conditions including a
  • 00:04:38
    determination of the state of health in
  • 00:04:40
    order to cure mitigate treat or prevent
  • 00:04:43
    disease or it's equally such products
  • 00:04:46
    are intended for use in the collection
  • 00:04:48
    preparation and examination of specimens
  • 00:04:51
    taken from the human body and this
  • 00:04:53
    definition can be found in 21 CFR part
  • 00:04:56
    eight oh nine point three when the
  • 00:05:01
    agency reviews an IV D we rely on valid
  • 00:05:06
    scientific evidence in order to provide
  • 00:05:08
    reasonable assurance that the device is
  • 00:05:11
    safe and effective for its intended use
  • 00:05:14
    so with regard to safety the central
  • 00:05:17
    question is really are there reasonable
  • 00:05:20
    assurances based on valid scientific
  • 00:05:22
    evidence that the probable benefits to
  • 00:05:25
    health from use of the device will
  • 00:05:27
    outweigh any probable risks with regard
  • 00:05:30
    to effectiveness similar question is are
  • 00:05:33
    there reasonable assurances again based
  • 00:05:36
    on valid scientific evidence that the
  • 00:05:38
    use of the device in the target
  • 00:05:40
    population will provide clinically
  • 00:05:42
    significant results so within the
  • 00:05:45
    regulatory framework when evaluating the
  • 00:05:49
    safety and effectiveness of the devices
  • 00:05:51
    the devices are actually classified
  • 00:05:54
    based on a risk based classification
  • 00:05:57
    system so there are three regulatory
  • 00:06:00
    classes for devices class 1 class 2 and
  • 00:06:04
    class 3 and there are differing levels
  • 00:06:07
    of controls for each of these classes
  • 00:06:09
    where the current level of control will
  • 00:06:12
    increase from class 1 to class 3
  • 00:06:15
    class 1 devices are considered to be low
  • 00:06:18
    risk these devices are subject to
  • 00:06:21
    general controls class two devices are
  • 00:06:24
    moderate risk and they are subject to
  • 00:06:27
    both general controls as well as special
  • 00:06:29
    controls and class 3 devices are the
  • 00:06:32
    highest risk category and for these both
  • 00:06:35
    general and special controls alone are
  • 00:06:37
    not sufficient to mitigate the risks
  • 00:06:40
    that are associated with the devices for
  • 00:06:42
    their intended uses so the risk
  • 00:06:45
    classification is wholly dependent upon
  • 00:06:49
    the intended use of the device and this
  • 00:06:51
    is basically the manner in which the
  • 00:06:53
    device is intended to be used so the
  • 00:06:56
    intended use is a statement that will
  • 00:06:58
    specify the analytes that are being
  • 00:07:00
    measured the specimen type that's being
  • 00:07:03
    tested as well as providing the context
  • 00:07:06
    in which the device should be used so
  • 00:07:08
    for example if there's a clinical claim
  • 00:07:10
    for prognosis or diagnosis or monitoring
  • 00:07:13
    that will be included in the intended
  • 00:07:15
    use statement as mentioned this is a
  • 00:07:18
    risk based classification and so for IV
  • 00:07:22
    DS it's really the risk that is
  • 00:07:24
    associated with obtaining false results
  • 00:07:26
    from that test and it's the level of
  • 00:07:30
    risk that is posed to the patients and
  • 00:07:33
    the users of that device the
  • 00:07:35
    classification is not based on
  • 00:07:38
    technology so even if you have a very
  • 00:07:40
    complex technological assay this alone
  • 00:07:43
    will not automatically place the device
  • 00:07:45
    in the high-risk category however the
  • 00:07:48
    technology or the methodology will
  • 00:07:50
    influence the types of validation
  • 00:07:52
    studies that will be needed to
  • 00:07:54
    demonstrate the safety and effectiveness
  • 00:07:56
    of the product
  • 00:08:01
    so within the regulatory framework the
  • 00:08:05
    classification of the device will
  • 00:08:08
    determine the pre-market regulatory
  • 00:08:10
    pathway for that device although all
  • 00:08:13
    devices are subject to be labeled
  • 00:08:17
    properly to be manufactured under a
  • 00:08:19
    quality assurance program and to have
  • 00:08:23
    establishment registration and device
  • 00:08:25
    listing forms on file with the FDA but
  • 00:08:29
    each class here has a different
  • 00:08:32
    regulatory path for example most class 1
  • 00:08:36
    low-risk devices are exempt from
  • 00:08:39
    premarket notification that is a 510 K
  • 00:08:42
    application is not required for those
  • 00:08:45
    devices however as I mentioned those
  • 00:08:47
    class 1 devices are still subject to
  • 00:08:50
    registration and listing class 2 devices
  • 00:08:54
    most of them will require a 510 K
  • 00:08:57
    submission in order to obtain FDA
  • 00:09:00
    clearance and they obtain clearance by
  • 00:09:04
    demonstrating that the device is
  • 00:09:06
    substantially equivalent to a legally
  • 00:09:09
    marketed predicate device in some cases
  • 00:09:13
    the class 2 devices may be exempt from
  • 00:09:18
    premarket notification and in other
  • 00:09:21
    cases there may be an absence of a
  • 00:09:24
    predicate device where that device type
  • 00:09:26
    has not previously been classified and
  • 00:09:29
    in that case then a de novo request for
  • 00:09:32
    classification can be submitted for that
  • 00:09:34
    device and once authorized that device
  • 00:09:37
    can then serve as a predicate for
  • 00:09:39
    subsequent devices of the same device
  • 00:09:41
    type and then for class 3 high-risk
  • 00:09:46
    devices these are subject to a PMA
  • 00:09:49
    application to be submitted for FDA
  • 00:09:51
    approval so what do we look at when we
  • 00:09:56
    review of these medical devices listed
  • 00:10:01
    here are the elements of FDA review the
  • 00:10:05
    intended use
  • 00:10:06
    the indications for use as discussed in
  • 00:10:08
    the previous slide are very central to
  • 00:10:11
    our review we also look at the
  • 00:10:13
    performance of the device which is
  • 00:10:15
    comprised of the analytical validation
  • 00:10:17
    and the clinical validation when
  • 00:10:21
    applicable the instrument and the
  • 00:10:23
    software validation is also reviewed we
  • 00:10:26
    also look at the labeling and the
  • 00:10:28
    labeling for the device must be
  • 00:10:30
    compliant with CFR eight or nine point
  • 00:10:33
    ten and as I mentioned previously all
  • 00:10:36
    devices must be manufactured under a
  • 00:10:38
    quality assurance program and they must
  • 00:10:41
    meet the quality system regulation
  • 00:10:43
    however for PMA applications the quality
  • 00:10:46
    system regulation information is
  • 00:10:49
    reviewed in the pre market setting so
  • 00:10:53
    I'll delve a little bit further into the
  • 00:10:55
    scientific review as I mentioned this is
  • 00:10:58
    comprised of both the analytical and the
  • 00:11:01
    clinical performance the analytical
  • 00:11:03
    performance should really demonstrate
  • 00:11:05
    the reliability and the accuracy of the
  • 00:11:08
    analyte measurements detected by that
  • 00:11:11
    test the studies can be specific to that
  • 00:11:15
    assay methodology as well as the
  • 00:11:18
    intended use and an example of this is
  • 00:11:21
    for IHC or immunohistochemistry assays
  • 00:11:25
    reader studies are needed
  • 00:11:27
    however from molecular nucleic
  • 00:11:29
    acid-based tests reader studies aren't
  • 00:11:31
    needed but accuracy studies are required
  • 00:11:35
    with regard to clinical performance the
  • 00:11:39
    test should be correlated with a
  • 00:11:41
    clinically meaningful and clinically
  • 00:11:43
    significant outcome and some measures to
  • 00:11:46
    demonstrate this can be clinical
  • 00:11:48
    sensitivity and specificity or positive
  • 00:11:52
    and negative predictive values as
  • 00:11:53
    appropriate for a few considerations for
  • 00:12:00
    analytical performance the test is
  • 00:12:02
    really considered as the whole system so
  • 00:12:05
    the validation should encompass all the
  • 00:12:07
    way from specimen collection and
  • 00:12:10
    preparation all the way through to
  • 00:12:13
    result reporting and therefore this
  • 00:12:15
    includes the pre and a little
  • 00:12:17
    steps the validation studies should also
  • 00:12:20
    be conducted with the intended specimen
  • 00:12:22
    type by using clinical specimens from
  • 00:12:25
    the intended population it's also
  • 00:12:27
    important to evaluate the performance of
  • 00:12:30
    the device across the entire reportable
  • 00:12:33
    and measuring range for that device and
  • 00:12:35
    this includes around the clinical
  • 00:12:38
    cut-offs and importantly the studies
  • 00:12:41
    should follow the protocol that will be
  • 00:12:43
    in the final device labeling for
  • 00:12:49
    clinical performance it should be
  • 00:12:51
    determined how the device will be used
  • 00:12:53
    in the clinical setting and then the
  • 00:12:56
    study design should ensure that it is
  • 00:12:58
    appropriate in order to support the
  • 00:13:00
    intended use of the device the clinical
  • 00:13:03
    study as well as the statistical
  • 00:13:06
    analysis plan should be pre specified
  • 00:13:08
    and clinical performance should be
  • 00:13:11
    established by comparing to an endpoint
  • 00:13:13
    or an appropriate surrogate and
  • 00:13:17
    typically analytical validation will
  • 00:13:20
    precede clinical validation now we'll
  • 00:13:25
    move on to companion diagnostics
  • 00:13:30
    companion diagnostics are a subset of IV
  • 00:13:33
    DS the definition of a companion
  • 00:13:36
    diagnostic is that it's a test that
  • 00:13:39
    provides information that is essential
  • 00:13:42
    for the safe and effective use of a
  • 00:13:44
    corresponding therapeutic because the
  • 00:13:48
    use of the companion diagnostic and the
  • 00:13:52
    corresponding therapeutic are
  • 00:13:53
    interdependent the use of the companion
  • 00:13:56
    along with the drug is specified in the
  • 00:14:00
    labeling of both the diagnostic as well
  • 00:14:03
    as the corresponding therapeutic it's
  • 00:14:07
    expected that there is contemporaneous
  • 00:14:10
    regulatory approvals for both the device
  • 00:14:12
    and the corresponding therapeutic
  • 00:14:16
    therefore when we review an IV D
  • 00:14:19
    application for a companion diagnostic
  • 00:14:21
    the review timeline is tied to the
  • 00:14:24
    corresponding application for the
  • 00:14:26
    therapeutic
  • 00:14:27
    product there are however exceptions
  • 00:14:30
    when contemporary knees approval may not
  • 00:14:33
    be possible and this may happen if there
  • 00:14:37
    is an unmet need or if it's in the best
  • 00:14:40
    interest of Public Health to approve a
  • 00:14:42
    drug prior to approval of the device in
  • 00:14:45
    such cases then the companion diagnostic
  • 00:14:49
    approval will be included as a post
  • 00:14:52
    marketing commitment on the therapeutic
  • 00:14:55
    side so for more information about FDA
  • 00:15:01
    expectations for companion diagnostics
  • 00:15:03
    you can refer to the final draft
  • 00:15:05
    guidance which is titled in-vitro
  • 00:15:08
    companion diagnostic devices the final
  • 00:15:11
    guidance was issued in August of 2014 in
  • 00:15:14
    there the definition of a companion is
  • 00:15:16
    provided and various scenarios for use
  • 00:15:20
    are described in addition FDA policies
  • 00:15:23
    for approval and labeling are outlined
  • 00:15:30
    to date there are more than 40 approved
  • 00:15:33
    companion diagnostic and therapeutic
  • 00:15:35
    combinations all of them except for one
  • 00:15:38
    are for oncology indications for 18
  • 00:15:42
    approved cancer therapeutics there are
  • 00:15:45
    33 approved unique IVD companion
  • 00:15:49
    diagnostics and all of them except for
  • 00:15:51
    one went through the PMA pathway and
  • 00:15:55
    these tests cover a total of 18
  • 00:15:58
    different biomarkers for a complete and
  • 00:16:01
    up-to-date listing of the cleared and
  • 00:16:04
    approved companion diagnostic devices
  • 00:16:07
    you can refer to the website shown on
  • 00:16:10
    this slide so as I mentioned the use of
  • 00:16:16
    a device
  • 00:16:16
    the use of a companion diagnostic is
  • 00:16:19
    tied to the use of a corresponding
  • 00:16:22
    therapeutic product and so there are
  • 00:16:25
    several key questions that will arise
  • 00:16:27
    regarding Co development first how do
  • 00:16:31
    you know that a companion diagnostic
  • 00:16:33
    device is needed second when you use a
  • 00:16:36
    test to identify a biomarker in
  • 00:16:40
    the clinical trial for the therapeutic
  • 00:16:42
    product will an investigational device
  • 00:16:45
    exemption or IDE application be needed
  • 00:16:48
    and third what if the final device was
  • 00:16:52
    not used in the therapeutic clinical
  • 00:16:55
    trial and finally how can a test receive
  • 00:16:59
    the same companion diagnostic claim as
  • 00:17:01
    the originally approved companion
  • 00:17:04
    diagnostic so I'll answer each one of
  • 00:17:07
    these questions in the subsequent slides
  • 00:17:10
    so first about the need for a companion
  • 00:17:13
    diagnostic typically both the
  • 00:17:16
    therapeutic and the device are studied
  • 00:17:18
    in the same clinical trial
  • 00:17:20
    therefore the clinical validation of the
  • 00:17:23
    device is supported by the results of
  • 00:17:25
    the therapeutic trial in other words the
  • 00:17:28
    results from the primary efficacy
  • 00:17:29
    population that are used to support
  • 00:17:32
    approval of the drug that is the
  • 00:17:35
    population that will also be used to
  • 00:17:37
    support device approval so when you use
  • 00:17:41
    a device in a clinical trial it's very
  • 00:17:43
    important that that assay is fully
  • 00:17:46
    specified and is locked down prior to
  • 00:17:49
    its use in that trial and what I mean by
  • 00:17:52
    that is that a specific assays should be
  • 00:17:55
    identified for detecting the marker in
  • 00:17:57
    that trial the specific protocol should
  • 00:18:00
    be identified and the clinical decision
  • 00:18:05
    points that will be used to define
  • 00:18:06
    biomarker positivity versus biomarker
  • 00:18:09
    negativity should be specified in
  • 00:18:11
    addition the specific specimen type
  • 00:18:14
    should be identified it's also important
  • 00:18:18
    that the assay is adequately validated
  • 00:18:21
    prior to its use in the trial so
  • 00:18:25
    ultimately the decision about the need
  • 00:18:28
    for a companion diagnostic is made by
  • 00:18:31
    the therapeutic review division but they
  • 00:18:35
    make this decision in consultation with
  • 00:18:37
    the device center so for these companion
  • 00:18:40
    diagnostics they really take a great
  • 00:18:42
    deal of collaboration and coordination
  • 00:18:44
    between the respective review centers as
  • 00:18:47
    well as between the respective device
  • 00:18:49
    and drug sponsors
  • 00:18:51
    so when thinking about whether companion
  • 00:18:55
    diagnostic is needed the central
  • 00:18:57
    questions are really is there adequate
  • 00:18:59
    evidence of clinical activity of the
  • 00:19:01
    therapeutic in the biomarker selected
  • 00:19:04
    population as identified by the
  • 00:19:06
    companion diagnostic and most
  • 00:19:09
    importantly is the test essential for
  • 00:19:13
    the safe and effective use of the
  • 00:19:15
    corresponding therapeutic product if the
  • 00:19:17
    answer to this is yes then undoubtedly
  • 00:19:19
    a companion diagnostic will be needed so
  • 00:19:24
    what about the need for an IDE
  • 00:19:27
    application an approved IDE permits the
  • 00:19:32
    device to be shipped lawfully in order
  • 00:19:34
    to conduct investigations of the device
  • 00:19:36
    without complying with other
  • 00:19:38
    requirements of the Food Drug and
  • 00:19:40
    Cosmetic Act that would apply to devices
  • 00:19:42
    in commercial distribution so typically
  • 00:19:45
    when a device is used in a clinical
  • 00:19:48
    trial its be it has not been approved
  • 00:19:51
    for that indication to be legally
  • 00:19:54
    marketed and so an IDE provides for an
  • 00:19:57
    exemption for use of that device and the
  • 00:19:59
    trial so when we look at all device
  • 00:20:04
    investigations some of them may be
  • 00:20:06
    subject to the IDE regulation while
  • 00:20:09
    others may be exempt from the IDE
  • 00:20:11
    regulation of those that are subject to
  • 00:20:13
    the IDE regulation some are considered
  • 00:20:17
    to be significant risk and therefore
  • 00:20:19
    subject to the full IDE requirements
  • 00:20:21
    while others may be non significant risk
  • 00:20:24
    and although an IDE application does not
  • 00:20:27
    need to be submitted to the agency for
  • 00:20:29
    these NSR studies they are subject to
  • 00:20:32
    the abbreviated IDE requirements which
  • 00:20:35
    can be found in 21 CFR part 812 so the
  • 00:20:42
    determination about the need for an IDE
  • 00:20:44
    is not a benefit risk decision it's
  • 00:20:48
    really an assessment of risk that is
  • 00:20:50
    posed to the patients who are enrolled
  • 00:20:54
    into that trial so the question is does
  • 00:20:57
    use of the test result pose a
  • 00:21:00
    significant risk if the answer is yes
  • 00:21:02
    and most likely that trial will be
  • 00:21:05
    use of the device in that trial will be
  • 00:21:08
    considered to be significant risk and an
  • 00:21:10
    IDE application will be needed in order
  • 00:21:14
    to determine if an IDE is needed for a
  • 00:21:16
    specific trial we recommend that a study
  • 00:21:19
    risk determination pre submission is
  • 00:21:21
    submitted to the Center for Devices so
  • 00:21:28
    the third question was about what
  • 00:21:32
    happens if the clinical trial assay or
  • 00:21:34
    CTA differs from the companion
  • 00:21:37
    diagnostic so one example of when this
  • 00:21:40
    might happen is if a prototype of an
  • 00:21:43
    assay is used to enroll patients in the
  • 00:21:45
    trial and then the assay goes through
  • 00:21:47
    some optimization or some changes and
  • 00:21:51
    then a different assay will come forward
  • 00:21:53
    for marketing authorization in such a
  • 00:21:56
    case changing the test can lead to
  • 00:21:59
    changes in the assay performance and
  • 00:22:01
    this in turn can change the patient
  • 00:22:04
    population from what was selected in the
  • 00:22:06
    therapeutic clinical trial when this
  • 00:22:09
    happens
  • 00:22:10
    bridging studies are required in order
  • 00:22:13
    to perform a bridging study patient
  • 00:22:16
    specimens that tested both CTA negative
  • 00:22:19
    and CTA positive should be retained and
  • 00:22:22
    banked so that they can be retested with
  • 00:22:24
    the final companion diagnostic device in
  • 00:22:28
    such a case there should be no
  • 00:22:30
    collection or selection bias and
  • 00:22:32
    therefore you should ensure that the
  • 00:22:35
    samples that are stored there and their
  • 00:22:38
    integrity is maintained and one way to
  • 00:22:41
    do this is through demonstration of the
  • 00:22:43
    stability of the analytes in those
  • 00:22:46
    stored specimens in order to ensure that
  • 00:22:50
    the storage duration was acceptable and
  • 00:22:52
    the quality of the assay is acceptable
  • 00:22:56
    when performing bridging study
  • 00:22:58
    statistical analysis plan should
  • 00:23:01
    consider potential discordance 'as and
  • 00:23:03
    also should account for any missing
  • 00:23:06
    samples importantly bridging studies are
  • 00:23:10
    not just a method comparison study
  • 00:23:12
    bridging studies should evaluate the
  • 00:23:15
    concordance between the CTA test results
  • 00:23:18
    and the companion diagnostic test
  • 00:23:19
    results but it should also demonstrate
  • 00:23:23
    that the clinical outcomes that were
  • 00:23:25
    correlated with the CTA are maintained
  • 00:23:28
    with the test results from the companion
  • 00:23:30
    diagnostic device the last question was
  • 00:23:37
    about the possibility for the same
  • 00:23:40
    companion diagnostic claim as the
  • 00:23:43
    original companion diagnostic claim for
  • 00:23:46
    a new test it's recognized that often
  • 00:23:51
    times samples from the original clinical
  • 00:23:54
    trial that's used to support the
  • 00:23:56
    approval of the original companion
  • 00:23:58
    diagnostic may not exist any longer or
  • 00:24:01
    they just may not be available for
  • 00:24:03
    retesting in such a case there are two
  • 00:24:08
    potential pathways that could be
  • 00:24:10
    followed in order to gain the same
  • 00:24:13
    companion diagnostic claim the first is
  • 00:24:17
    that they can follow the path of the
  • 00:24:19
    original companion diagnostic they could
  • 00:24:21
    perform a new clinical trial in order to
  • 00:24:24
    obtain clinical outcome data that's
  • 00:24:27
    correlated with the new test we
  • 00:24:30
    recognize that it can be extremely
  • 00:24:32
    burdensome and costly to perform a new
  • 00:24:35
    clinical trial and so an alternative
  • 00:24:38
    path is to get a follow-on companion
  • 00:24:41
    diagnostic claim a follow-on companion
  • 00:24:45
    diagnostic should consistently and
  • 00:24:47
    accurately select the same intended use
  • 00:24:49
    patient population as the originally
  • 00:24:52
    approved companion diagnostic device and
  • 00:24:55
    therefore the can follow on companion
  • 00:25:01
    diagnostic should demonstrate the same
  • 00:25:04
    or the comparable level of analytical
  • 00:25:07
    and clinical performance for the
  • 00:25:08
    specific mutations or the specific
  • 00:25:11
    analytes that are being tested in the
  • 00:25:14
    originally approved companion diagnostic
  • 00:25:18
    so in order to demonstrate this
  • 00:25:21
    performance it's important that a
  • 00:25:23
    clinical concordance study which is
  • 00:25:26
    essentially a method comparison
  • 00:25:28
    study is performed to compare the new
  • 00:25:32
    test with the previously approved
  • 00:25:35
    companion diagnostic so a term that has
  • 00:25:43
    emerged in the recent past is
  • 00:25:49
    complementary Diagnostics and these are
  • 00:25:53
    tests that in contrast to companion
  • 00:25:56
    diagnostics these are not essential for
  • 00:26:00
    the safe and effective use of the
  • 00:26:02
    therapeutic product complementary
  • 00:26:05
    Diagnostics identify a biomarker to find
  • 00:26:08
    subsets of patients who respond
  • 00:26:10
    particularly well to a drug and aid in
  • 00:26:14
    the benefit risk assessment for
  • 00:26:16
    individual patients therefore it is not
  • 00:26:19
    a prerequisite for receiving the drug an
  • 00:26:22
    example of a situation in which a
  • 00:26:26
    complementary diagnostic would arise is
  • 00:26:28
    when a drug is approved for an
  • 00:26:30
    all-comers however there may be
  • 00:26:34
    differential clinical outcomes based on
  • 00:26:37
    biomarker status as defined by this
  • 00:26:39
    complementary diagnostic and so in such
  • 00:26:42
    cases then a complementary diagnostic
  • 00:26:45
    may be needed because the complementary
  • 00:26:48
    Diagnostics are also studied in the
  • 00:26:51
    clinical trial that's used to evaluate
  • 00:26:53
    the performance of the therapeutic
  • 00:26:54
    product the development path for
  • 00:26:57
    complementary Diagnostics is similar to
  • 00:27:00
    that for companion diagnostics so now
  • 00:27:07
    I'll move on to NGS oncology panels
  • 00:27:11
    these next-generation sequencing based
  • 00:27:14
    on ecology panels are increasingly being
  • 00:27:17
    employed in the clinical setting part of
  • 00:27:21
    their advantage is that they have the
  • 00:27:24
    ability to screen a very large number of
  • 00:27:27
    biomarkers which can be up to thousands
  • 00:27:30
    of biomarkers that can be screened
  • 00:27:32
    simultaneously therefore the tests
  • 00:27:36
    have broad intended uses due to the
  • 00:27:40
    sheer volume of the number of sequences
  • 00:27:43
    that can be obtained from a single run
  • 00:27:46
    validation for such a test on a per
  • 00:27:49
    variant or per market basis is not
  • 00:27:51
    feasible and therefore these types of
  • 00:27:54
    tests have introduced challenges to the
  • 00:27:57
    companion diagnostic paradigm which
  • 00:28:00
    traditionally has been one test for one
  • 00:28:03
    drug as I mentioned previously the
  • 00:28:10
    validation studies for a device should
  • 00:28:13
    be designed to support the performance
  • 00:28:15
    characteristics of that device for its
  • 00:28:17
    intended use but as I mentioned there
  • 00:28:21
    are challenges one is that it is not
  • 00:28:25
    feasible to validate all possible
  • 00:28:28
    alterations that may be detected by an
  • 00:28:31
    NGS test a second clinical specimens for
  • 00:28:37
    analytical validation may be limiting
  • 00:28:41
    particularly for rare tumor types or for
  • 00:28:44
    tumors in which there are low prevalence
  • 00:28:46
    biomarkers and third there is a lack of
  • 00:28:51
    reference methods and standards that can
  • 00:28:53
    be used in order to demonstrate the
  • 00:28:56
    accuracy of these NGS uncle panels so in
  • 00:29:00
    the next few slides I'll address each of
  • 00:29:03
    these challenges by presenting
  • 00:29:05
    strategies that have been implemented by
  • 00:29:08
    some of the tests that we've reviewed
  • 00:29:11
    first with regard to the inability to
  • 00:29:15
    evaluate every single variant that could
  • 00:29:19
    be detected by an NGS test it's
  • 00:29:24
    recognized that this is terribly
  • 00:29:26
    burdensome and therefore we recommend
  • 00:29:28
    that all variants with clinical
  • 00:29:31
    companion diagnostic claims should be
  • 00:29:34
    included in the validation studies
  • 00:29:36
    however these NGS tests include can
  • 00:29:40
    include many more genes or biomarkers
  • 00:29:44
    that extend beyond the markers so
  • 00:29:48
    with a companion diagnostic claim so in
  • 00:29:52
    such cases we then recommend that a
  • 00:29:54
    representative approach can be taken in
  • 00:29:57
    order to support the validation of those
  • 00:30:00
    other single nucleotide variants and
  • 00:30:03
    insertions and deletions or in Dells
  • 00:30:06
    when taking a representative approach
  • 00:30:08
    it's important that a range of variant
  • 00:30:11
    types are included which represent or
  • 00:30:14
    give consideration to the sizes of the
  • 00:30:17
    different variants as well as the
  • 00:30:19
    contexts the genomic context that those
  • 00:30:22
    variants can be detected in on the other
  • 00:30:29
    hand a representative variant approach
  • 00:30:32
    is not acceptable for fusions or gene
  • 00:30:35
    rearrangements because the performance
  • 00:30:37
    of the device for one gene fusion may
  • 00:30:41
    not be able to be extrapolated to
  • 00:30:44
    fusions detected in other genes so what
  • 00:30:49
    I really mean by this is that if fusions
  • 00:30:52
    in the AL gene are used in the
  • 00:30:54
    validation studies this cannot be used
  • 00:30:57
    in order to gain claims that the test
  • 00:31:00
    can detect fusions in other genes such
  • 00:31:03
    as rosslyn however within the AL gene we
  • 00:31:09
    recommend that the most common the two
  • 00:31:11
    or three most common out confusions are
  • 00:31:14
    represented in the validation studies in
  • 00:31:18
    order to obtain the claim that a test
  • 00:31:21
    can detect alkanes and so therefore
  • 00:31:24
    every single out fusion that has been
  • 00:31:26
    identified does not need to be included
  • 00:31:29
    in the validation studies there are also
  • 00:31:34
    can be some post-market considerations
  • 00:31:37
    in order to obtain a pre market post
  • 00:31:40
    market balance and so there may be
  • 00:31:42
    circumstances under which post market
  • 00:31:45
    data can be provided for some analytical
  • 00:31:48
    studies in order to further demonstrate
  • 00:31:50
    the performance of the device
  • 00:31:57
    regarding the potential lack of clinical
  • 00:32:00
    specimens the test performance should be
  • 00:32:03
    established with the intended clinical
  • 00:32:06
    specimens however in certain analytical
  • 00:32:09
    studies it may be appropriate to use
  • 00:32:12
    contrived samples provided at an
  • 00:32:15
    appropriately designed functional
  • 00:32:17
    characterization study is performed so a
  • 00:32:20
    contrived sample functional
  • 00:32:22
    characterization study basically should
  • 00:32:25
    demonstrate that the contrived samples
  • 00:32:28
    are able to appropriately mimic the
  • 00:32:30
    clinical specimens and therefore can be
  • 00:32:33
    used as surrogates in the validation
  • 00:32:35
    studies another consideration for pan
  • 00:32:40
    tumor claims in which it is not possible
  • 00:32:42
    to test every single tumor type that has
  • 00:32:47
    been identified in these cases we
  • 00:32:50
    recommend that comparability across a
  • 00:32:53
    variety of common and challenging tumors
  • 00:32:55
    is demonstrated and we typically
  • 00:32:58
    recommend that at least 10 tumor types
  • 00:33:00
    are represented also because different
  • 00:33:04
    tumor types can have different potential
  • 00:33:06
    interference it's important to assess
  • 00:33:09
    interference effects
  • 00:33:10
    an example is melanin in melanoma and
  • 00:33:14
    then finally we also recommend that data
  • 00:33:17
    is provided regarding the invalid rates
  • 00:33:20
    and the performance metrics across tumor
  • 00:33:23
    types and this information can be based
  • 00:33:25
    on historical data or a retrospective
  • 00:33:28
    analysis of data that already exists
  • 00:33:35
    regarding the lack of reference methods
  • 00:33:38
    in order to demonstrate accuracy a
  • 00:33:41
    validated orthogonal method may be used
  • 00:33:44
    as a comparator if the method is not an
  • 00:33:48
    fda-approved
  • 00:33:49
    or an FDA cleared test then minimal
  • 00:33:52
    level of validation data is needed for
  • 00:33:55
    the comparator and we generally
  • 00:33:57
    recommend that at a minimum limit of
  • 00:34:01
    and sensitivity studies as well as DNA
  • 00:34:03
    input and precision studies are
  • 00:34:05
    performed to support the performance of
  • 00:34:08
    that other test as a comparator method
  • 00:34:12
    to demonstrate accuracy of the test
  • 00:34:17
    additional validation considerations are
  • 00:34:21
    that the validation should evaluate how
  • 00:34:24
    different source errors can affect
  • 00:34:27
    device performance
  • 00:34:30
    NGS tests are a complex process that
  • 00:34:34
    involves numerous steps in the workflow
  • 00:34:37
    this encompasses sample preparation
  • 00:34:41
    library construction sequencing and data
  • 00:34:44
    analysis and even result reporting in
  • 00:34:48
    such cases for these types of tests
  • 00:34:51
    different error profiles are associated
  • 00:34:54
    with different analytes which can range
  • 00:34:56
    from parameters including variant type
  • 00:34:59
    genomic context and allele frequency so
  • 00:35:03
    therefore we recommend that quality
  • 00:35:05
    metrics are established at both the
  • 00:35:07
    sample and the variant levels and that
  • 00:35:10
    these quality metrics should really look
  • 00:35:12
    beyond exon level coverage and so some
  • 00:35:16
    of the quality metrics that we request
  • 00:35:20
    are listed here the reportable range
  • 00:35:23
    should be clearly defined and this is
  • 00:35:25
    with respect to genomic regions variant
  • 00:35:28
    types that will be detected by the assay
  • 00:35:29
    as well as the variant allele
  • 00:35:31
    frequencies that are detected and all of
  • 00:35:35
    the quality control metrics and the
  • 00:35:37
    validations should be developed in order
  • 00:35:39
    to ensure the performance of the device
  • 00:35:42
    within the claimed reportable range in
  • 00:35:46
    addition sequencing metrics should be
  • 00:35:49
    provided such as the run in the sample
  • 00:35:51
    QC mapping metrics which will include
  • 00:35:54
    coverage and coverage should cover depth
  • 00:35:57
    completeness and uniformity and
  • 00:36:00
    additional metrics can be variant
  • 00:36:03
    calling as well as reporting metrics so
  • 00:36:07
    the validation considerations that I
  • 00:36:09
    discussed were implemented
  • 00:36:12
    and stemmed from our experiences with
  • 00:36:16
    NGS tests that we've reviewed so listed
  • 00:36:19
    here are some of the NGS tests actually
  • 00:36:22
    all of the NGS tests that we have
  • 00:36:24
    reviewed and cleared or approved the
  • 00:36:28
    first NGS tests were actually not
  • 00:36:32
    companion diagnostic devices those were
  • 00:36:37
    approved in November of 2013 and those
  • 00:36:41
    were two Illumina tests that were
  • 00:36:44
    intended to detect cystic fibrosis
  • 00:36:46
    variants that are intended to aid in the
  • 00:36:49
    diagnosis of cystic fibrosis as they're
  • 00:36:52
    not companion diagnostics and they were
  • 00:36:54
    for an aid and diagnosis for cystic
  • 00:36:56
    fibrosis those went through the 510 k
  • 00:36:58
    pathway the first NGS based companion
  • 00:37:03
    diagnostic test was approved in December
  • 00:37:07
    of 2016 and that was the foundation
  • 00:37:10
    focus CDX bracha test from from
  • 00:37:14
    foundation medicine that tests detected
  • 00:37:18
    variance in the Braco one and the
  • 00:37:20
    bracket two genes and more recently in
  • 00:37:24
    June and November of last year for
  • 00:37:27
    next-generation sequencing tests were
  • 00:37:30
    approved three of these were considered
  • 00:37:33
    next-generation sequencing oncology
  • 00:37:36
    panels and they're listed here so
  • 00:37:38
    basically these tests included genes
  • 00:37:41
    beyond those that have companion
  • 00:37:44
    diagnostic claims so the first test that
  • 00:37:48
    was approved in June of 2017 went
  • 00:37:51
    through the PMA pathway and it was for
  • 00:37:54
    the angka mine DX target test from
  • 00:37:56
    thermo Fisher Scientific the next was
  • 00:37:59
    approved in note was cleared in November
  • 00:38:02
    of 2017 and that was for the msk impact
  • 00:38:06
    test from Memorial sloan-kettering and
  • 00:38:08
    this went through the Genova pathway and
  • 00:38:11
    therefore it did not have any specific
  • 00:38:13
    companion diagnostic claims associated
  • 00:38:16
    with it and then finally later in that
  • 00:38:20
    same month on November 30th the
  • 00:38:22
    foundation 1c DX test was approved the
  • 00:38:25
    p.m. a pathway and that test is from
  • 00:38:27
    Foundation medicine so these three FDA
  • 00:38:31
    authorizations have really shaped the
  • 00:38:34
    current regulatory paradigm for NGS
  • 00:38:37
    uncle panels and that's shown here so in
  • 00:38:41
    the classical companion diagnostic model
  • 00:38:44
    where it was one test for one drug
  • 00:38:47
    you only had biomarkers with companion
  • 00:38:52
    diagnostic claims but with the recent
  • 00:38:56
    clearance of the msk impact panel this
  • 00:39:00
    really created some additional levels of
  • 00:39:04
    biomarkers that are associated with
  • 00:39:06
    different levels of evidence and so you
  • 00:39:11
    can see that the while I'll refer to
  • 00:39:13
    level 1 biomarkers as those with
  • 00:39:16
    companion diagnostic claims level 2
  • 00:39:19
    biomarkers are those that are considered
  • 00:39:21
    to be cancer mutations with evidence of
  • 00:39:24
    clinical significance and level 3
  • 00:39:27
    biomarkers are the cancer mutations with
  • 00:39:30
    potential clinical significance and I'll
  • 00:39:33
    talk about the examples here the uncle
  • 00:39:36
    mine test and the foundation one CDX
  • 00:39:39
    tests all fit into the level 1 category
  • 00:39:43
    but they actually had both level 1 2 & 3
  • 00:39:49
    biomarkers while the msk impact test
  • 00:39:52
    went through a class-2 pathway and had
  • 00:39:56
    only level 2 and level 3 biomarkers but
  • 00:39:59
    it's important to note that a test
  • 00:40:01
    cannot be cleared with only level 3
  • 00:40:05
    biomarkers and so level 3 biomarkers are
  • 00:40:08
    obtained along with level 2 and/or level
  • 00:40:12
    1 biomarker claims so to talk a little
  • 00:40:16
    bit more about each of these three tests
  • 00:40:19
    the angka my index target test is a lung
  • 00:40:22
    cancer panel the tested text single
  • 00:40:26
    nucleotide variants and deletions in 23
  • 00:40:29
    dreams from DNA as well as fusions in
  • 00:40:33
    the Ross 1 gene from RNA that has been
  • 00:40:35
    isolated from FFPE tumor tissue samples
  • 00:40:38
    from non-small-cell lung cancer patients
  • 00:40:42
    it was the first approved test that
  • 00:40:45
    simultaneously assesses variants in
  • 00:40:47
    different genes in order to determine
  • 00:40:50
    patient eligibility for treatment with
  • 00:40:52
    multiple non-small cell lung cancer
  • 00:40:54
    targeted therapies and therefore it had
  • 00:40:57
    multiple companion diagnostic claims
  • 00:40:59
    which are listed here
  • 00:41:01
    the first is that the test was approved
  • 00:41:03
    to detect b-raf v600 mutations for
  • 00:41:07
    treatment with depravity and tremendum
  • 00:41:09
    it was approved for the detection of
  • 00:41:12
    Ross one fusions for treatment with
  • 00:41:14
    crizotinib and it also was approved with
  • 00:41:17
    a follow-up companion diagnostic claim
  • 00:41:20
    to detect EGFR exon 19 deletions and l85
  • 00:41:24
    8 our substitution mutations for
  • 00:41:26
    treatment which a fitting in and
  • 00:41:28
    therefore this test had both level 1
  • 00:41:32
    level 2 and level 3 claims for the msk
  • 00:41:38
    impact test the intended use is shown
  • 00:41:41
    here the assay is a qualitative in vitro
  • 00:41:46
    diagnostic test that uses targeted
  • 00:41:49
    sequencing of FFPE tumor tissue matched
  • 00:41:53
    with normal specimens from patients with
  • 00:41:55
    solid malignant neoplasms in order to
  • 00:41:59
    detect tumor gene alterations the test
  • 00:42:02
    is intended to detect mutations that
  • 00:42:05
    include point mutations and small indels
  • 00:42:07
    as well as microsatellite instability
  • 00:42:09
    and its intended for use by qualified
  • 00:42:13
    healthcare professionals in accordance
  • 00:42:15
    with professional guidelines importantly
  • 00:42:18
    the test is not conclusive or
  • 00:42:20
    prescriptive for labeled use of any
  • 00:42:23
    specific therapeutic product and
  • 00:42:25
    therefore it does not have the level 1
  • 00:42:28
    biomarker or companion diagnostic claims
  • 00:42:34
    the msk impact test is a solid tumor
  • 00:42:38
    panel it identifies mutations in 468
  • 00:42:43
    genes as well as microsatellite
  • 00:42:46
    as I mentioned this went through the de
  • 00:42:49
    novo pathway and it really established a
  • 00:42:52
    class-2 pathway for these NGS uncle
  • 00:42:55
    panels that have only level 2 and level
  • 00:42:57
    3 claims so these are considered to be
  • 00:43:02
    class 2 tumor profiling tests and they
  • 00:43:06
    are eligible for third party review
  • 00:43:10
    third party review is a program that was
  • 00:43:14
    implemented in order to approve the
  • 00:43:16
    efficiency and the timeliness of the
  • 00:43:18
    510k review process for certain device
  • 00:43:22
    types under this program there are
  • 00:43:24
    accredited third parties who can conduct
  • 00:43:27
    the 510k review and they will provide
  • 00:43:29
    the review package as well as their
  • 00:43:32
    recommendation to the FDA after they
  • 00:43:35
    have reviewed the information and so
  • 00:43:37
    really devices under this class can go
  • 00:43:41
    through the class to pathway either
  • 00:43:43
    through the third party review pathway
  • 00:43:46
    or they can directly submit a 510k
  • 00:43:48
    submission to the FDA and I'd also like
  • 00:43:52
    to note that one of the accredited third
  • 00:43:54
    party reviewers for this type of test is
  • 00:43:57
    New York State Department of Health so
  • 00:44:03
    going back to the regulatory paradigm we
  • 00:44:07
    have the three levels of review and we
  • 00:44:10
    saw based on the uncle mind test that it
  • 00:44:13
    had claims for both level for all of the
  • 00:44:16
    levels level 1 level 2 and level 3
  • 00:44:19
    importantly the validation for level 1
  • 00:44:22
    biomarkers which have companion
  • 00:44:26
    diagnostic claims require that
  • 00:44:29
    analytical validation is performed for
  • 00:44:32
    each marker associated with such a
  • 00:44:35
    companion diagnostic acclaim the
  • 00:44:38
    clinical claims are established through
  • 00:44:42
    a clinical study or for follow-on
  • 00:44:46
    companion diagnostics through clinical
  • 00:44:48
    concordance results
  • 00:44:52
    for level two claims analytical
  • 00:44:56
    validation can be conducted per marker
  • 00:44:58
    or through representative approach for
  • 00:45:01
    SMBs and indels in contrast to the level
  • 00:45:07
    one the level of evidence for clinical
  • 00:45:10
    validation is that it can be established
  • 00:45:13
    in professional guidelines but it does
  • 00:45:16
    not have to be demonstrated directly
  • 00:45:18
    with the test of interest and then for
  • 00:45:21
    level three biomarkers and a little cook
  • 00:45:24
    validation is supported through the
  • 00:45:26
    representative approach while clinical
  • 00:45:29
    validation is not demonstrated either in
  • 00:45:32
    professional guidelines or with the test
  • 00:45:34
    but rather it's suggestive based on
  • 00:45:37
    available evidence and that available
  • 00:45:39
    available evidence can include mecca's
  • 00:45:42
    mechanistic rationale as well as
  • 00:45:45
    published literature so for the last NGS
  • 00:45:51
    uncle panel that has recently been
  • 00:45:53
    approved this was the foundation once
  • 00:45:55
    EDX it's a solid tumor panel and the
  • 00:45:59
    test was approved to detect
  • 00:46:01
    substitutions in der indels which are
  • 00:46:04
    insertions and deletions as well as copy
  • 00:46:07
    number alterations in 324 genes as well
  • 00:46:10
    as select gene rearrangements it was
  • 00:46:14
    also approved to detect two genomic
  • 00:46:16
    signatures microsatellite instability
  • 00:46:18
    and tumor mutation burden the test was
  • 00:46:23
    approved for use with 15 targeted
  • 00:46:26
    therapies in five different cancer types
  • 00:46:28
    and therefore it had companion
  • 00:46:30
    diagnostic claims and it's also intended
  • 00:46:33
    to provide tumor mutation profiling in
  • 00:46:36
    accordance with professional guidelines
  • 00:46:38
    and these are level 2 biomarker claims
  • 00:46:41
    importantly this test was the first
  • 00:46:44
    breakthrough device to complete the PMA
  • 00:46:46
    process and it was the second device to
  • 00:46:49
    be reviewed and approved under the fda
  • 00:46:52
    CMS parallel review program here's an
  • 00:46:58
    excerpt a quick snapshot of the intended
  • 00:47:02
    use from the device shown here is table
  • 00:47:05
    with all of the companion diagnostic
  • 00:47:08
    claims for which the device is approved
  • 00:47:10
    you can see here the five different
  • 00:47:13
    tumor types are not small cell lung
  • 00:47:16
    cancer melanoma breast cancer colorectal
  • 00:47:19
    cancer and ovarian cancer along with the
  • 00:47:22
    associated biomarkers and therapeutics
  • 00:47:29
    so as I mentioned this was the second
  • 00:47:32
    IVD that was reviewed and approved under
  • 00:47:35
    the fda CMS parallel review program this
  • 00:47:39
    parallel review program was piloted in
  • 00:47:42
    2011 and fully implemented in 2016 and
  • 00:47:46
    it's aimed at reducing the time between
  • 00:47:50
    FDA approval of a device and CMS
  • 00:47:53
    Medicare coverage so on November 30th
  • 00:47:57
    when the FDA approved this device on the
  • 00:48:00
    same day CMS issued a proposed national
  • 00:48:03
    coverage determination or an NC d for
  • 00:48:06
    NGS cancer diagnostics for Medicare
  • 00:48:09
    beneficiaries after the public comment
  • 00:48:12
    period
  • 00:48:14
    CMS finalized the NCD on March 16th of
  • 00:48:18
    this year for NGS for Medicare
  • 00:48:21
    beneficiaries with advanced cancer the
  • 00:48:26
    coverage is for diagnostic laboratories
  • 00:48:29
    using NGS tests for patients with
  • 00:48:32
    advanced cancer and that's defined as
  • 00:48:35
    being recurrent metastatic
  • 00:48:37
    relapsed/refractory
  • 00:48:39
    or advanced cancer as stages 3 & 4 in
  • 00:48:44
    order to get coverage the test must have
  • 00:48:47
    FDA approval or clearance as a companion
  • 00:48:49
    diagnostic it also must have an
  • 00:48:52
    fda-approved or cleared indication for
  • 00:48:55
    use in the patient's cancer and the
  • 00:48:58
    results should be provided to the
  • 00:49:00
    treating physician using a report
  • 00:49:02
    template to specify the treatment
  • 00:49:05
    options the coverage decision also noted
  • 00:49:08
    that coverage for other NGS lab tests
  • 00:49:11
    for Medicare patients with cancer are
  • 00:49:14
    made by local Medicare administrative
  • 00:49:17
    ders and for more information about the
  • 00:49:19
    decision memorandum on CMS's NCD the
  • 00:49:23
    link is provided at the bottom of this
  • 00:49:25
    slide so in conclusion I think the
  • 00:49:31
    regulatory models for companion
  • 00:49:34
    diagnostics and NGS oncology panels have
  • 00:49:38
    really evolved over time due to
  • 00:49:40
    increasing efforts and personalized
  • 00:49:42
    medicine and a lot of these efforts have
  • 00:49:45
    really been driven by device innovation
  • 00:49:48
    and device advancements and a couple of
  • 00:49:52
    takeaways are that these type to bring
  • 00:49:57
    these products to market it takes a lot
  • 00:49:59
    of collaboration and coordination
  • 00:50:00
    between all of the stakeholders for
  • 00:50:04
    example for companion diagnostics it
  • 00:50:06
    takes coordination communication between
  • 00:50:09
    the diagnostic and the drug sponsors as
  • 00:50:11
    well as within the agency between the
  • 00:50:15
    review centers the respective review
  • 00:50:17
    centers also through programs like third
  • 00:50:20
    party review and parallel review it
  • 00:50:23
    takes a lot of coordination between
  • 00:50:25
    agencies as well as with the device
  • 00:50:28
    sponsors as well so with that I'd like
  • 00:50:32
    to end here and listed here are a few
  • 00:50:35
    additional resources to get some
  • 00:50:37
    additional information on some of the
  • 00:50:39
    topics that I've discussed today and
  • 00:50:41
    also if after this session any questions
  • 00:50:45
    arise about specific devices or anything
  • 00:50:47
    I've talked about today you can please
  • 00:50:49
    feel free to send me any questions at
  • 00:50:52
    the email address here thank you very
  • 00:50:54
    much
Tags
  • Molecular Diagnostics
  • Companion Diagnostics
  • Next-Generation Sequencing
  • FDA
  • In Vitro Diagnostics
  • Regulatory Framework
  • Clinical Trials
  • Biomarkers
  • Personalized Medicine
  • Device Innovation