Tip:
Highlight text to annotate it
X
ERIC GREEN: So next up, Erin Ramos on "Clinically Relevant Variants Resource."
ERIN RAMOS: Thank you, Eric. Good afternoon, everyone. I also wanted to extend a welcome
to my colleagues from AMP, Association for Molecular Pathologists, for joining us today.
And we may also have colleagues from the College of American Pathologists listening via Webcast.
So if you are there, thank you for joining us.
This presentation will have a very similar flow to Anastasia's. In February of 2012 we
also, we brought a concept to Council. The focus was on to develop a resource of potentially
actionable genetic variants for use in the clinic. This concept, although I think there
was support in general from Council, like Anastasia said, there were significant concerns
that were raised that we took back with us and considered to bring back to you during
this Council session.
So the original proposal focused on supported identification and dissemination of consensus
information on potentially actionable genetic variants in clinical care. It had three main
goals, identifying these genetic variants with implications for clinical care and disseminating
the evidence, developing clinical decision support systems, for incorporating these variants
into clinical care. And then building upon and then unifying existing programs with the
hopes of reducing duplicative efforts across research and clinical organizations.
So, to put this into context, we had actually been thinking about such an initiative over
the past couple of years. Eric has heard from numerous colleagues as he goes to present
at various meetings that NHGRI really should take some leadership in integrating all this
information on the variants that are being identified. And we've also heard it at a variety
of population genomics meetings, such a need.
So based on that we organized this workshop in December of 2011, titled, "Characterizing
and Displaying Genetic Variants for Clinical Action." And then we took some of those ideas
and incorporated them into the Concept Clearance that we brought to you in February. Again,
the concept was deferred. One of the main concerns heard from you in February was that
this is an important effort and it touches on more than just NHGRI.
There are many professional societies that have a stake here that they have relationships
and represent important communities that would be using this information in the clinic. And
that it was very important for us to reach out to them, to learn what they are doing
and see if we could partner in moving such an effort forward.
So shortly after Council we did initiate some of these conversations. We've had great interactions
with ACMG, AMP, ASHG and, as I said, College of American Pathologists. Again we formed
a Council subgroup to further discuss some of the concerns that were raised in February.
Based on that discussion, we revised our concept document. We circulated that to our Council
subgroup. We've had addition conversations with genome staff, with our colleagues in
other organizations and around the NIH. And then slightly revised the concept a bit more
and circulated that final concept to the entire Council in May.
So what we heard from you in February, NHGRI must clearly be viewed as a leader, not just
funding another group to produce a gene list. So if we really wanted this initiative to
grow legs and be adopted by the community, it had to be seen and understood by the community
that we are really the driving force behind this.
There were concerns about the number of awards, whether a significant, a single award or multiple
awards would be the appropriate way to go. There was a concern raised for the potential
for a single awardee to produce a less optimal product. There was a concern raised that the
focus should be on generating the first list of genes and not focusing so much on the variants
because we know that that list will continue to grow.
The consensus process must include a clear hand-off to professional societies for guideline
development so that NHGRI can certainly play a role in pulling all this information together,
synthesizing what we have learned and then working with the grantee to provide the information
in a package that the clinical societies could use to develop guidelines, which then could
be used by the clinical community. And, of course, it would be best if other ICs collaborate
or at least clearly support the program.
So we heard the need for this to really be a trans-NIH effort if possible. So, again,
I wanted to thank the members of our subgroup that participated on our call and subsequent
e-mail conversation. And also for our colleagues from the professional societies that participated
in that call, it was a great discussion. We were able to really expand the discussion
around the concerns that were raised and make revisions go our concepts based on these discussions.
So to address the first concern raised in February, we did specify involvement of ACMG
and AMP as co-leads on this initiative. Actually, in the title we included NHGRI along with
these organizations. And then also specified throughout the document that it was extremely
important to collaborate with relevant professional societies and other organizations that are
stakeholders in this effort.
We emphasized, again, the need for collaborative effort. We did discuss the possibility of
having multiple awardees but it was decided that that may not be a workable approach.
So, although we are sticking with the single awardee, we've emphasized that NHGRI will
provide significant leadership through this cooperative agreement. We also clearly articulated
that there will be a governance structure that includes NHGRI and then the other professional
societies that would be included on the various committees that are formed. And, of course,
other stakeholders that we all feel are relevant.
We've clarified that the applicant would be expected to describe their rationale for focusing
on genes or variants. So whichever approach we choose, we want a description of why they
are focusing on that path.
We also drew a bright line between synthesizing the evidence and developing consensus findings,
which we felt is within our purview and then producing the clinical guidelines, which would
be outside our purview. So we just tried to clarify the language that we used there so
everyone understands our role, and that we would try to hand off this information in
a package that could be used by professional societies to develop these guidelines.
And along those lines we eliminated the aim that focused on development of clinical decision
support. We heard, although the need for clinical decision support tools is certainly an important
one, that with this effort, we wouldn't want to develop clinical decision support tools
on our resource because our resource would have the information on potentially relevant
genetic variants that wouldn't have had that sort of stamp of approval from the professional
guidelines.
We just wanted to make it clear what this information will be used foróand that clinical
decision support tools can be developed down the road, that this program is one step in
that path.
So after we circulated that revised draft to our Council subgroup, we've had, again,
follow-up conversations and made some relatively minor revisions. We did, however, remove the
organization names from the title. We heard from NIH Office of the Director that non-federal
groups can't be listed as participating organizations on any funding opportunities. And, again,
the reason that we included these societies in the title was to indicate to the community
that we're serious about working with other societies and making this a collaborative
effort.
I think in the end, by removing the institutions from the title, it probably benefits all of
us because it allows us to be more inclusive and make sure that we hear from you and others
how the relevant societies are and include them in the process. So, again, we've had
some great conversations with College of American Pathologists and also follow-up conversations
with ASHG. We've been able to incorporate those groups as well in the revision.
So we've change our name, again. For those of you, even prior to the workshop we had
been referring to this resource as ClinBase (?). And then we learned that that name was
trademarked. So then we thought of ClinAction. That sounds reasonably good because you think
about actionable variants. Although we've heard from some of you that it implies, actually,
taking and action and that we didn't want to confuse the community by thinking that
all the variants in this resource meant that you needed to take action. So that one is
out.
For a while we were calling this the Resource Formerly Known as ClinAction. [Laughter] And
you could superimpose my face there on Edvard Munch's the scream, which recently sold for
$120 million dollars. That's what I felt like. So we decided to be boring and just refer
to this as Clinically Relevant Variants Resource. And hopefully , if this moves forward, work
with the awardee and the steering committee to come up with a more appropriate name. And
we are just going to leave it at that.
So I have summarized some of the changes that we made and just to articulate what this looks
in our revised purpose and goals. So this was our original purpose and goals statement.
We've modified the purpose to read: Support a process for identification and dissemination
of consensus information on genetic variants relevant to clinical care.î So we removed
that emphasis on potentially actionable.
The first goal, identify genetic variants with implications for clinical care and disseminate
the evidence. This was a little bit broad and the disseminating the evidence part needed
to be described a bit more. So identify genetic variants with likely implications for clinical
care incorporate these variants in evidence into a resource that can serve as a substrate
for development of practice guidelines.
I mentioned we eliminated that second aim of focusing on development of clinical decision
support tools. And that was changed to focus on establishing this process for transferring
this information to appropriate organizations for development of these guidelines. So, again
hopefully, the grantee and the steering committee can work with the societies to make sure we
collect the right information, we package it in a way that can be useful for the societies
to review and make further decisions on which variants they like to recommend as moving
forward for clinical use.
And the third aim hasn't changed. This is an important one, build upon existing programs.
Unify these efforts. Reduce duplicative efforts across research and clinical organizations.
And just to drill this home the text has change. We emphasize that this needs to be multi-institutional
approach, again to bring in similar efforts and diverse perspectives.
We've highlighted, again, the close involvement with other professional societies and a jointly
appointed steering committee. We will work with our professional society colleagues to
facilitate consultation with other important stakeholders, including regulatory agencies,
clinicians, etcetera. The applicants are expected to survey the landscape of ethical, legal,
social and policy issues regarding results reporting and try to integrate these and build
these into this proposed effort.
Applicants are expected to describe their plans (The text highlighted in orange is new.)
for engagement and integration with ongoing efforts, data synthesis, creation, development
of consensus findings. Handoff to clinical organizations, again as I mentioned, for development
of these practice guidelines and dissemination. Approaches for grouping or bidding (?) genes
or variants into categories of clinical relevance. And plans for dealing with profusion of variants
of unknown significance or for focusing on particular variants at least as an initial,
manageable approach.
There hasn't been any change to anticipated funding. So this is what you saw presented
in February. The only other comment that I would make is that we have reached out to
numerous, other IC colleagues to find out if they are doing similar effortsóbecause
we heard from you we didn't want to be duplicating any efforts across the NIH. And besides, if
it was a work we already were aware of at NIDMS with the farm(?) DKB, there hasn't been
anything presented to us. In fact, a couple of our colleagues at other institutes said
that they really support NHGRI taking a leadership position in such an initiative.
So I will just end there and thank our Council subgroup and other colleagues for helping
move this forward. And I guess, actually, Eric, if we could ask any of our colleagues
from the professional societies that have been working with us to add any comments before
we open it up to Council.
MARY WILLIAMS: Thanks. It is Mary Williams for the Association for Molecular Pathology.
Thanks so much for the proposal. I can tell you that our members are, they are mostly
in translational research and clinical practice. Genomics is moving so fast into clinical practice
that it is now being described as the Wild West out there. And that doesn't make our
members feel very good because that gives sort of the picture that decisions, clinical
decisions are being made around a poker table in a saloon or that surgery is happening on
the bar.
So there are guidelines that need to be written to help guide our colleagues in clinical practice.
And these guidelines are very expensive to produce. They are for evidence-based guidelines,
not just consensus-based guidelines. And they need to be coming out, not just once every
three years. They need to be coming out with multiple guidelines per year. So we want to
thank the NHGRI for proposing this.
ERIN RAMOS: Thanks, Mary. Jo?
JO BAUFMAN: Jo Baufman (?) from ASHG. I would just like to thank and congratulate the staff
of NHGRI for engaging the community in different ways, I think, thanóat least I've not seen
this in the last decade in the process of a Concept Clearance. ASHG is pleased to be
named in this. I think that we have members who can contribute a great deal to this.
Our leadership did react pretty strongly to the Concept Clearance and that is for two
reasons. First, because this is so important to all of us. Secondly, because we think it
is so broad in this process of development and engagement and then, eventually, dissemination
that we wanted to make sure that it is done right. Devil is in the detail, of course,
but I think that doing this together we can get it right. Thank you.
ERIN RAMOS: Thanks, Jo.
MIKE: [off microphone] ñCollege of Medical Genetics. This has been the rate limiting
step in getting genetics and genomics into clinical practice has been the existence of
guidelines that get it out of the hands of the rare disease experts. And I think Mary
hit one of the clear problems. And I think it is one you are going to have to think about.
Guidelines themselves are very expensive and a lot of the expense is in building the evidence
base. But the evidence base can be built in a way that makes it a much shorter trip to
the guidelines. And there are a lot of steps along the way.
If you look at Blue Cross ñ Blue Shield, one gene test assessment is $100,000 to do
the evidence base and then the guideline that comes off of that. But depending on where
you draw your line between informing the guideline and allowing the professional groups, it can
be much less expensive. And I think I would give a fair bit of attention to that.
We are extremely interested in this moving forward. But it is not just a laboratory thing.
This is going to be a very strong clinical and multi-specialty, multi-disciplinary because
genetics is multi-disciplinary and multi-organ. And many of our conditions spread across multiple
systems. So I think it is going to be a very interesting involvement of the medical specialty
world in a way that has been very different for the NIH.
ERIN RAMOS: Thanks, Mike. And I think that's a great point, that we want to try to bring
the professional societies together at an early stage so we can hear from them what
information we need to work with the grantee to collect, how we can package it and make
it the most efficient pipeline to get information from the beginning, packaged (?) in a way
that can be used to develop these guidelines. Any other questions or comments?
ERIC GREEN: I open it up for Council discussion now. Tony.
TONY: I would just like to reiterate that final point. Although NHGRI wants to take
the leadership on providing the database and the evidence behind those variants being clinically
relevant, we need to have a plan for how the professional societies are going to develop
the guidelines because it can't be done by volunteers. It is going to require funding.
And I think that point about, it's where you draw the line about, it's where you draw the
line about what you are funding to gather evidence and what's clinically relevant will
lessen the burden to the professional groups to actually develop the guidelines. So I think,
in a way, you'll be taking part of the lead on that side of it as well. It depends on
where you draw that line.
ERIC GREEN: Mike.
MIKE: I'd characterize the leadership just slightly differently. To me the issues was
not so much that NHGRI be the leader, although that is fine but it be clear that this is
the place. That this isn't, again, just one of many gene lists. I did wonder as we were
going through this whether this is a place where other institutes could be engaged in
terms of helping to support this. Because again, just like the last one we talked about,
this is one that multiple institutes surely are interested in. Don't know if that is possible
or not.
ERIN RAMOS: We've actuallyóin my conversations with other institutes trying to get a sense
of if there are duplicative efforts, I've circulated our plans and have heard back from,
you know, of course preliminarily, at least one or two institutes that are interested
in co-funding. So that is something certainly we are interested in. We recognize that this
could be a shared effort.
MIKE: I think that could be really helpful.
ERIC GREEN: Rob (?).
ROB(?): I infer that the aim is to make this a comprehensive list, a complete list. And
the path to how you get there, is it going to be a prioritized in any way?
ERIN RAMOS: Yes. So I think in the language of the concept we would ask the grantee prioritize
how they would sort of bin (?) the variants that are compiled into those that are more
ready for clinical action. There might be a more intermediate bin where there is more
information that we need and what is that information then. Other bin(?) clearly these
variants aren't ready for incorporation into the clinic. And there are other ways to bin
it but certainly that is something that we want the grantee to come up with a proposal
for prioritization and they would work with the steering committee to figure out the best
way to prioritize the variants that we have.
JIM: What Mike just mentioned is important and it might be that something along these
lines would be included in the RFA. And what I'm getting at is, in addition to it being
highly desirable for other institutes to help sponsor an effort like this, I think that
the right applicants will also, perhaps, bring in non-geneticists. And I'm thinking, for
example, of a specific situation. When we were recently going through lists of things,
many of us had assumed that long QT associated genes, known deleterious mutations in these
would be in what we call bin 1. These are highly actionable.
It looks like as you learn more about this things that there are certain long QT-associated
genes where we know enough to say that's true but others where we don't . And it took a
cardiologist to tell us that. So that is a long-winded way to say that it might be useful
in the RFA to encourage the participation of having expertise in the conditions and
not completely just the genetics of those conditions.
ERIN RAMOS: Yeah. I agree with you completely. In fact, we thought oneóand this, of course
would be up to the grantee. But one way to organize sort of sifting through these piles
of variants is to organize into sort of domain specialties of disease expertise. So we could
bring in experts from cardiology or, you know, cancer related expertise, diabetes. And that
would not only have the genetics expertise but those very familiar with the condition
to be able to make real decisions.
PARTICIPANT: One other question. Jim, I think I asked you this question four months ago.
And that is, how many genes and variants are there in bin 1?
JIM: It depends on what day you catch me at. Let's just say that I think most of us feel
very comfortable that it would be fewer than 100. And many of us would say far fewer than
100, depending on how you set the evidence bar.
PARTICIPANT: As a follow up to that, then, is this a five-year project?
JIM: I would think it is for two reasons. One is that list, those lists are going to
change as we find out more. The other thing that I think clinicians need really that bin
1, what do I need to tell people and what's life threatening in the short term that I
can do something about. That's the, although difficult, easiest question. There are many
other questions about how to characterize these results that vary on all kinds of parameters,
including the impact on people on learning this kind of information, etcetera. So I think
there is plenty to be done here for that.
ERIN RAMOS: And I would say that the first year, particularly, we need to work with the
steering committee and the grantee to get sort of the framework in place. We could test
it out on these first sets of variants. And then we to take that framework and apply it
to other variants along the way. So it could take a couple of years to work through that.
And then we could also, towards the end, come back and revisit some of the variants that
now may have more information that we can make different decisions on.
ERIC GREEN: We will do Mike and then Carlos.
MIKE: Only would only encourage you to sort of recognize how important informatics is
going to be in this. Physicians don't follow practice guidelines. You know, it is rarely
over 50 percent in reality when you really go out and survey them. Clinical decision
support tools integrated into electronic medical records is where this is going to get disseminated
so that it is well vetted and it give people guidance in the clinical scenario they are
sitting in.
So the EMR that are ongoing through NCBI and other, or NLM, to develop those standards
and things are going to get attached to this pretty closely I think to be successful.
ERIN RAMOS: That's a good point.
CARLOS: First I want to say I think this is a really important thing to do. And I really
support the idea that NHGRI is going to take this on as a kind of flagship to put their
imprimatur on or our imprimatur. The thing that concerns me a little bit is distinguishing
it from other existing effort from OMIM to HMGD to all the other databases that folks
have tried to put together on what are important variants that you need to flag and know about.
And secondly, how it will overlap with, and I think it should go hand in hand with the
Mendelian centers and the work that they are doing. So I imagine as we are talking about
both having a really great database of known and really critically important and bona fide
highly penetrant alleles is one thing. And another is that, you know, things that you
need to screen and know about because it would be really hazardous not to return that information
to participants once you know it.
Figuring that fine line, I think, is really the whole challenge that this faces. And sometimes
it is not an informatics problem, right? I mean the informaticians can build databases
and they can curate them and they can do it really carefully so that it is not polluted
and so on. But where you draw that line is a totally different issue. And so I think
you really want to think hard about what it is you are going to ask of this resource,
particularly if you are not going to want to fund this in perpetuity but rather create
it in such a way so that it a really good systematic way of building that resource so
that you can continue to build it and make it available.
ERIN RAMOS: That's a great point. And part of the reason we are trying to engage the
societies and other stakeholders early through the steering committee and other committees
is to hear from them exactly what data points that they needóso that we can make these
decisions, so we have the critical information in place. We've alsoówe are certainly talking
with NCBI and the various resources that they have to see how we can coordinate efforts
to take information that we develop into their systems and vice versaóso that we are just
more easily synchronizing the information that we have. Yes, so I think those are great
points.
ERIC GREEN: [off microphone]
PARTICIPANT: I think to your point, though, one of the things this has over OMIM and several
of the other resources is exactly that process, to engage stakeholders. There is a little
bit of experience with this in Pharmacogenetics Research Network with the CPIC (?) guidelines.
I think that sort of provides a little bit of a model for how to go forward.
PARTICIPANT: Was there any thought given in terms of reach through. I appreciate the way
you kind of drew the semi-permeable line there between creating the evidence, etcetera, and
then handing it off for the guidelines. What if the guidelines are bad? What if there are
no takers? Nobodyóyou know, we've already spent our money on creating the guideline
for the last eight you gave us. Can't handle another eight. What becomes the responsibility
of identifying these is there is no receptor node?
ERIN RAMOS: Well, I think at that point we would have to revisit our role. I mean we'll
do our best to put the information in a way that can be used. Again, working with the
societies, we do hear it is expensive. But if we can take on some of that by curating
the information and moving it forwardówe certainly want to evaluate this resource along
the way and make sure that the information that we are presenting is being used. Hopefully
it will be and I expect it will be. But, I mean it is a fair point. We can come back
here and revisit how we want to move forward and helping facilitate getting these guidelines
to come out of the resource that we put together. But that will have to be collaborative, again,
with societies and other folks.
PARTICIPANT: We must have learned a lot from the copy number variant guidelines and sort
of establishment of that, David Ledbetter's (?) talk at the AIRLE (?) and since then.
Is this the same kind of problem trying to figure outóI know it is different types of
data that you are dealing with here. But is that what the issue is? Is that what you are
trying to learn by doing this RFA? Or is it--? Do you know what I'm talking about, Erin?
ERIN RAMOS: The ISCA consortium?
PARTICIPANT: Yeah.
ERIN RAMOS: International Standards for Cytogenomics Arrays.
PARTICIPANT: Right.
ERIN RAMOS: Actually, I'm going to goóthe ISCA is having their national meeting today
and tomorrow. So I'm going to go to that meeting. I mean that is one example of an approach
for pulling this information together. So they have a database that can be used for
the cytogenomics community. I'm not exactly clear on the approach they took to coming
up with the data that is included in that data base. But we certainlyóone, we can learn
from their approach. And, two, any information that we can integrate from the evidence they
already have into our resource.
PARTICIPANT: I was thinking about the buy-in that Pearl just mentioned. Somebody, I think
at least a lot of people accepted that. I don't know what that process was, who was
involved and whether it was the groups that are talking about this one as well.
ERIN RAMOS: We certainly can learn more from David and ISCA along the way. That's a good
point
ERIC GREEN: Okay. Seeing no more discussion, I'm looking for a motion on this Concept Clearance.
Motion. Second. Okay. Any other discussion before we vote? Okay. All in favor? Opposed?
Abstained. Okay. It is approved. Thank you, Erin. Bettie, you ready?