It has been more than twenty years since the collapse of the World Trade Center on September 11th, 2001. Dr. Slavin discusses Myriad Genetics’ company-wide involvement in the effort to help identify victims of the tragedy with John Ryan, Sr. VP of Operations, Women’s Health and Oncology, and Benoît Leclair, Program Director of Quality Assurance and Regulatory Affairs.
It has been more than twenty years since the collapse of the World Trade Center on September 11th, 2001. Dr. Slavin discusses Myriad Genetics’ company-wide involvement in the effort to help identify victims of the tragedy with John Ryan, Sr. VP of Operations, Women’s Health and Oncology, and Benoît Leclair, Program Director of Quality Assurance and Regulatory Affairs.
[music]
0:00:11.7 Dr. Thomas Slavin: I am Dr. Thomas Slavin, Chief Medical Officer for Myriad Genetics.
Welcome to Inside the Genome.
0:00:17.7 DS: Hi everyone, welcome back to the podcast. Today we have two special guests. They
were individuals that worked for Myriad, and I was really touched by one of the recent stories
behind our involvement in 9/11, which I had no idea was going on. So I just wanted to find out
more, talk with them and really have them tell the story as they saw it. We have John Ryan, he is
senior vice president of operations, and we are also joined by Benoit Leclair, he is Program Director
of Quality Assurance and Regulatory Affairs. So thank you both for coming on today.
0:00:56.1 John Ryan: Thanks for having us.
0:00:56.9 Benoit Leclair: Our pleasure.
0:00:58.4 DS: Yeah. So, as I alluded to... I mean really, this was news to me, I had no idea Myriad
had any involvement in the 9/11 tragedy as a whole. Can you elaborate a little bit on what prompted
Myriad's involvement here and how it played out over the months to follow the horrible event?
0:01:22.7 JR: I can kick things off, if you'd like, Benoit.
0:01:25.3 BL: Sure.
0:01:25.8 JR: So one of the sort of interesting thing that came out of all this is, and Benoit can
certainly add details to this, is that... Well, he and I had come from a background where we had
training and experience in mass disasters. I had previously to Myriad worked at the Armed Forces
DNA identification laboratory, who actually works with the NTSB for any type of mass disaster
that is occurring on American soil. So as part of that, I'd done a lot of training and sort of planning
around the idea of if and when something dramatic like this happened. I also...
0:02:04.3 DS: What is the NTSB just for our listeners?
0:02:06.4 JR: National Traffic Safety Board.
0:02:08.6 DS: Okay.
0:02:09.0 JR: When you have some type of large accident, there was a coordination that was
occurring between AFDEL and the NTSB, lots of acronyms, in addition to which I had also been
asked to be on the scientific advisory board for the International Commission on Missing Persons,
and our efforts were focused on the identification of remains from the Balkan wars. And Benoit,
coming from his background in Canada and working with Swiss Air. And Benoit, if you wanted to
add some color there.
0:02:39.3 BL: Yes. In 1998, there was an air crash that landed in Canadian territory, it was Swissair
12/14/21 Page 2 of 11
9.11 Response Podcast
Flight 111 that was traveling from JFK to Geneva and it crashed in the ocean near Halifax. And the
RCMP, I was working at the RCMP at the time, the Royal Canadian Mounted Police for the
forensic lab division, and we were tasked with identifying the remains. So we ended up processing
something like 3000 remains, separate remains from the crash site. And at the time, we also saw in
the flight manifest that there were... The core of the passengers were related, meaning entire
families were on board, and also personal effects that could be used for identification had been
traveling with them. So we knew we would have to build pedigrees from within the set of unknown
genotypes we would produce, and that prompted me to write a piece of software to accelerate this
and make it even more robust in excluding possible mistakes. And that activity in 1998, New York
was aware that I had done this. So when in early September, 2001, New York gave me a call and
asked if I could lend a hand with adapting my piece of software to this new set of circumstances.
0:04:23.5 DS: So they contacted you. And what was your next step then? How did this all
transpire?
0:04:31.5 BL: Well, of course, I said yes, and I would help. The forensic community is a small one.
Everybody knows everybody. And when one of us is faced with a huge challenge, everybody comes
to help out. So the automatic response is, of course, I'm going to give you a hand. So I worked
remotely throughout the 15 months that my assistance was required, basically receiving data sets
from New York and processing them in Utah and returning identification leads back to New York
on CDROM/
DVDs. And not the victim's name, it was all coded with numbers and the software
would find the best matches and then return them to New York and they would pursue those leads,
and that the bulk of them led to identifications of the victims. So that's how it came to be for me.
0:05:42.4 DS: Yeah. How did you get involved, John?
0:05:44.2 JR: It was an organization literally on 9/11. It was that classic thing of where we all
witness the planes hitting the towers, and it was sort of the surreal moment where I literally was at
home and I stood up and I said, "I have to get to work." Because any time there had been this type
of event, my organization had been involved, and I got to work and quickly came to the realization
that I was working for a different company. And wasn't sure if we were gonna have a role or not.
And so I actually, Benoit, myself, Tom Shull and Brian Ward actually circled up and we started to
talk about whether we felt like there was something that we could do to help. If you remember at
that time, there was a belief that there was gonna be north of 30,000 people killed in this event,
were the estimates. And so at that time, what we had built at Myriad Genetics from a forensics
capacity, was the only high throughput laboratory in the world at that time that had forensic
accreditations, forensic processes. And so we felt that minimally one of the things that we could do
would be to process all of what would be ascribed as reference samples, either from relatives of
individuals or from personal effects. And so we quickly made an assessment as to we felt that we
could do this and that it was a unique offering at that time.
0:07:09.3 JR: And we proposed it to the CEO of the company, Pete Meldrum, and our assumption
was that we were gonna be turned down, it just really wasn't within kind of the central business
focus of the company. Instead, Pete came back and he said, "Go for it, this is a national tragedy. We
wanna be involved and we wanna do what we can to help." He also really focused on the idea that
we were gonna go in and help. But TJ, you mentioned that you hadn't heard about this before. That
12/14/21 Page 3 of 11
9.11 Response Podcast
was by design. We very much wanted to be involved with this, but we didn't want this to be
something that was utilized as sort of a banner of, "Hey, look what we're doing," and we wanted it
to be very much a pure effort. So by the next day, we actually reached out to the New York office
of the medical examiner, City Medical Examiner, excuse me. And because we actually had forensic
contracts with them for databasing of their convicted offender profiles, they knew our work, they
knew our capabilities, they had audited us and had a strong working relationship, and they said,
"Absolutely."
0:08:13.4 JR: So we actually were receiving reference samples, personal effects within one week.
Pretty quickly, we discovered that this wasn't gonna be on the scale that we thought it had been, but
that the demand was still gonna be very, very high. One of the things that happened was they
quickly realized that not only for reference samples and personal effects, but for the actual victim
samples themselves, there simply wasn't enough processing capacity within the United States, and
they reached out to us and asked whether we felt we had the capability to be able to do that. One of
the things to recognize with all of this is that this was actually a gigantic crime scene. And to be
able to be involved with something like that, you have to be able to meet all the requirements for
something to be within the legal arena. We worked with the New York State Department of Health
to actually get accredited to not only be a reference laboratory for personal effects and reference
samples, but actually to be able to process crime samples. And so we got our New York State
Department of Health tech lead certification very, very rapidly. And as part of that, we started
receiving actual victim samples within a month after the tragedy.
0:09:26.8 DS: Wow, that's incredible. The company was already doing a lot of this forensic work. I
mean, what was the origins of even that work in the beginning? Was it just that this was VNTR
heart genetics, these variable number of tandem repeats in the background, and we're working on
ways... That was kind of some of the early days of genetics. Is that where it just kind of naturally
came out of, or was there some interest in forensics at the time for the company?
0:09:52.2 JR: It's a great question, TJ, and please feel free to add to this, Benoit. We had a situation
where the Myriad was in the business of sequencing clinical samples for BRCA1 and 2. And that
market, that sample volume simply hadn't developed in the way that the company had expected it. I
mean this was a true pioneering effort by Myriad, never been done before. So they built a lot of
capacity and we ended up not getting very many samples for a period of time early on. And so as a
company, one of the questions was, well, we've made this investment, are there things that we can
utilize that we can focus on that will, frankly, fall within kind of our business, and the proposal was
made to get involved with basically databasing of samples from convicted offenders. Because, as is
very common now, this idea of the FBI being able to utilize a database to process samples at a
crime scene and link it back to a prior offender simply really didn't exist, it certainly didn't exist on
any scale. And so we got into that, and over the course of seven years, we actually processed almost
900,000 database samples.
0:11:06.3 DS: Wow, that's incredible.
0:11:07.9 BL: There's a bit of history that John is probably unaware, that had preceded his arrival at
Myriad in fact. In 1997, I came to visit Myriad for troubleshooting problems that we're having with
our DNA sequencers. We were using DNA sequencers that were the same, the same make and
12/14/21 Page 4 of 11
9.11 Response Podcast
brand as Myriad, back at the RCMP. And when I visited and I saw all the robotics that were in
place, I told Tom [0:11:41.5] ____ Shull that John referred to earlier, I told him that Myriad had the
kind of setup that would be perfectly suited to do high volume testing in the forensic domain for
convicted offenders, and that maybe maybe Myriad should think about it. So I don't know how
much of that comment that I made that day actually trickled it's way to... All the way to top of
management and then it became an avenue that Myriad decided to pursue, but I remember making
that comment and then a few years later this came to be.
0:12:30.1 JR: Wow, it's a great point, I wasn't aware of that, and I actually joined in 2000 to run
this databasing effort at Myriad. And literally, I was employee one of that fledgling effort where we
had somebody dedicated to it versus people who are doing this in their spare time, and it grew from
there.
0:12:49.4 DS: And what was the sentiment around the company as all these samples are coming in,
to triage and take care of?
0:12:56.4 JR: Oh my gosh. For the World Trade Center, this was probably one of the most... The
word "exciting" probably sounds wrong, but it was an inspiring time. Literally, everyone at the
company was desperate to be involved. We had people that were coming in on their days off to
sweep the floors and take out the trash. It was literally everyone at the company had their eye on the
idea that there had been this massive tragedy and what could we do to help? So one of our biggest
struggles actually was everybody wanting to help, not necessarily having the skill set or the
accreditation associated with it, and saying, "Hey, you're a highly skilled individual, you're not
accredited for this, but we can still use you for what might be more of a clerical task or something
like that." So it was an inspiring time to be part of this company.
0:13:51.1 DS: I can only imagine. And what was the output of all these efforts? Do you have a
highlevel
recap at the end of, you said it... This went on for about a year and a half, you said?
0:14:05.4 JR: Yeah, thereabouts, and Benoit, I had, frankly, don't have a line sight to a lot of what
you were doing on the backend, so please weigh in on this. As far as our processing was concerned,
we processed over 20,000 samples. We had, personal effects, reference samples and victim samples.
I don't have the breakdown of that right now, but it was a huge effort that over time, one of the
reasons why we actually got out of it was that pretty quickly, we were working solely on victim
samples and the quality of those samples was degrading over time. The site at the World Trade
Center literally was wet and on fire, so high heat, high humidity, very, very degrative to samples,
and so they had started to take a look at that much more specialized means of being able to get a
profile from those samples, and so we made a decision to actually exit this project. But again, as I
mentioned earlier, we tried to do this in a way that was really kind of behind the scenes and as quiet
as possible.
0:15:13.0 BL: I mean after this the... Very often on mass disasters, the... Well, I will not generalize,
but an air crash or something like this, what can be found is usually found very... Pretty quickly in a
matter of maybe 710
days. The debris field that the World Trade Center was, it was 1.6 million
tons. Those two buildings weighed that much, and it took 10 months to excavate the site, and it was
all triaged on conveyor belts to look for remains. And so the longer... The more time had passed, the
12/14/21 Page 5 of 11
9.11 Response Podcast
heat and humidity, did its work, meaning that over time, as John said, the samples that were being
recovered were more and more degraded, yielding fewer, fewer and fewer... Or shorter genotypes, if
you will. It's like a partial plate on a car, the fewer numbers you have on the plate, the more... The
less informative it is to, in the way of pointing to a particular individual, so that's why at the
backend of this effort, it became... The number of identifications that you were seeing we... Was
producing was falling down pretty rapidly because a lot of the tested samples were not producing
enough data to be able to allow for a match to be made.
0:16:47.5 DS: And what were the sample types coming in?
0:16:52.3 BL: Oh my God.
0:16:53.5 DS: Were they DNA, or were they... Were human remains?
0:16:55.5 JR: So what we did, as far as the processing was concerned, they were doing DNA
extractions really onsite,
and then putting it out to these laboratories that were involved with the
amplification and the sequence of the genotyping of these samples and creating reports. And so they
were actually coming in in plates as DNA extracts, because one of the things that we looked at was,
at Myriad we wanted to be very cautious about only applying our efforts to where it made sense and
that we felt like we had unique capabilities, and a lot of that had to do with our high throughput
processing system, whether it was the amplification, the actual sequencing, genotyping of the
samples or the developing a profile and a report coming out of that, and that's where we felt like
that's where we had kind of that unique capability to apply to this.
0:17:51.5 DS: Yeah. And did you ever get followup
on how many people were ultimately
identified or matched from all this effort?
0:18:02.8 JR: That is ongoing, so that number always change, there was a point at which we were
associated with over 1600 identifications.
0:18:11.5 DS: Wow.
0:18:11.5 JR: And frankly, I haven't kept track of that since, but it was... And one of the things that
I really wanna say about this, and I hope this conveys, was everyone that was at Myriad was
involved with this in one way or another, and what was fascinating about it was when I sent out that
email, the number of individuals who were involved with this that are still at Myriad, 20 years later,
and when you would talk to them, they would get a catch in their voice, not simply because of the
tragedy itself and the horrible, horrible implications of that, but this point where the country but
also our company came together to accomplish something really amazing.
0:18:57.0 DS: Yeah.
0:19:00.0 JR: It was an inspiring time to be at Myriad.
0:19:04.0 DS: Yeah, no and I have many questions. The matching of the actual samples on the
12/14/21 Page 6 of 11
9.11 Response Podcast
identification, is it usually that you're matching something that was found with something that was
in someone's home? Is that...
0:19:20.7 JR: Benoit can speak to that at length.
0:19:23.6 BL: Yes, in fact, I have a slide on my computer screen over here that I produced so many
years ago in 2004. And the personal effects are often thought to be the best source of material
because hopefully it's going to contain a trace of DNA that would be a perfect match to the
genotype that will be recovered from the remains. However, when we tabulate the identifications
that were done, and when they were done, there were 1560 in 2004. At that time, personal effects
had been used in 4.8% of all identifications, so it's very, very low. And in an air crash, it's the same
thing. Personal effects that are in the luggage, the most useful ones, toothbrush and so on, are in the
luggage that had been destroyed in the crash as well. So we often, in those mass disasters, have to
do parentage analysis; that's another way of talking about paternity testing. Here, it's instead of
working with two knowns, mom and dad, well, three knowns, mom, dad and the child, where you
can figure out in about a minute on paper whether or not the parentage analysis works out. In a case
of a mass disaster, you're dealing with two knowns, but an unknown within the victim's data set, or
with a mass disaster that is involving an airplane, you might have entire families within the bulk of
victims. So parentage analysis becomes critical in order to make identifications.
0:21:20.9 BL: The big difference with New York, with the World Trade Center, is that the people...
It was a workplace, so there were only two of the victims that were related. There were two brothers
that were working at the World Trade Center in different companies, but there were no entire
families among the victims. So all the parentage analysis had to deal with two, well, two living next
of kin that we were trying to match to victims' genotypes. So the challenge, the particular challenge
with the World Trade Center was in what John described earlier, the degradation. So you would
need basically... You would be looking for full profiles within the data set and then try to match that
full profile to partial ones. You would probably find many because you can... By the level of
destruction, it was obvious in the pictures that you can still see on the web of ground zero, the
buildings were reduced to nothing. So, if steel was broken down that much, you can imagine how
bodies were broken down.
0:22:38.7 BL: So there was somebody, a victim of the World Trade Center for which in one of my
listing, there were 200 body parts matching related to that individual. So you would use the longest
profile and then attach back the partial ones until you are reaching a certain threshold for... A
certain likelihood ratio of your finding this profile someplace else in the population. And then that
way we could create a single block of profiles, the longest one being the one we would search,
therefore the next of kin to match to. So it required a lot of computation. There were also clerical
errors that found their way into to the data when paperwork was filled out for the next of kin. At
times, if you check box, the wrong checkbox would be checked and not telling us the actual
relationship to the victim. So when we noticed that, we decided to not use the reported
relationships, but use all the next of kin as a block, anonymous block, and then testing it against all
the victims and see what the algorithm would come out as to, "Okay, these three belong together."
Then New York would look at whether or not the two next of kin are in the same family, which
99.99% of the time they were, and that would lead to an identification.
12/14/21 Page 7 of 11
9.11 Response Podcast
0:24:14.4 BL: So you can imagine the computing demands that put on the system. The software I
wrote was of course 20 years ago, computers were not nearly as fast. I would load up every week
with the data download from New York, I would load it up on a Friday afternoon when I would
receive it and it would run overnight 16 hours, and the next day I would have the list of matches to
send back to New York. So what it'll do is work billions and billions of comparisons to eventually
find the three pieces that fit together for different families.
0:24:57.5 DS: Yeah. And what are you normally looking... What's the algorithm looking at? Is it
looking at tandem repeats in the DNA?
0:25:05.1 BL: STRs, short tandem repeats, and the NTRs were preceding STRs anyway in forensic
history. But you're looking at your... Our genotype with STRs is a bunch of numbers, so it's where
the genotype for a given locus, it's the number of repeats for the two alleles, that is the genotype. So
let's say you've got 14 repeats from the paternal allele, and then 17 repeats on the maternal one, then
the genotype for that target would be 14 17.
0:25:40.9 BL: So, and then you do this on multiple different targets, and we were doing 16 at a
time. So with that number of targets, you're expecting to get enough discrimination power to be able
to tell any two people apart. The odds of having the same genome type short of having an identical
twin was into the billions. So that's how we end up being able to tell two individuals apart. But with
paternity testing or parentage analysis you're going to receive one allele from mom, one allele from
dad, and then if you know what those four alleles are, or you know what your true alleles are, let's
put it that way, you can find two parents that will each have one of those two alleles for that locus.
Then you look at the next locus and they're already sharing still one allele each and then you do this
across 16 systems, and in the end the odds of not having the right set of parents is very low.
0:26:43.4 DS: Yeah, wow. And what... This is incredible. And what has been the history since the
project stopped with forensics at Myriad, and are either of you still working on forensics in either
way?
0:27:00.1 JR: That's a great question, TJ. In 2007 as Benoit had mentioned, there was this history
of this potential application of what had been built to databasing, and that was very successful.
Within 18 months we actually owned 72% of the available contracts that were out there; we just
went in and dominated it. Soon after that one of the things that happened was that this became a
kind of a feast or famine business. It was being driven by the DNA act, and so a lot of federal
money was being pushed out to state entities and local entities. The struggle with that was we would
go through a contracting basis, so there would be six, nine months at a time where no samples
would be coming in and then the contract would get awarded, and there were times where we had
over 30,000 samples coming through in a given month. And what Myriad recognized over time was
it probably wasn't gonna be a great business to be in just from the perspective of being able to have
a steady processing capacity and whatnot.
0:28:03.2 JR: At the same time, really the clinical diagnostic business had been taking off. We
ended up at times kind of competing for resources, and so we were growing rapidly on both fronts.
And Myriad made a decision in 2007 to get out of forensics. And when I looked at it from a
12/14/21 Page 8 of 11
9.11 Response Podcast
business perspective, it made sense. Honestly, looking at it kind of personally and also
professionally, it was very, very hard to not be involved with that so directly anymore.
0:28:37.3 DS: Yeah, especially because it was such a part of your life. And Benoit, are you still
working in this field at all?
0:28:47.0 BL: No. I started working at Myriad in 2001 and I was involved in many forensic
projects at the time, but when the forensic division was closed and then all of my time was basically
reallocated
to clinical testing. I've not done any... I've done two events, for Swissair and the World
Trade Center. And now there have been other disasters and there's always a large number of people
that are willing to help out with those events. And it's quite demanding to work on mass disasters
because of pressures of time. It's very high pressure, next of kin are looking for the retrieval and
identification of their loved one, political pressure and so on. So those were long weeks for 15
months that I was working on it, upward, north of 7080
hours a week, so it's quite taxing to do that
kind of work. And after two events I said, "Well, if somebody rings me up one day for help, I won't
say no." But it was a very...
0:30:32.9 BL: As John said earlier, "exciting" is not the right word, but certainly inspiring time to
have been involved in this. For me, it was a challenge as to how we're going to do this and to
identify as many victims as possible as quickly as possible. The intellectual challenge was worth it.
And then it advanced the field of mass disaster identification initiatives and now there are
commercial products that are involved in doing this and also missing person databasing has taken
off as well. I still have friends in the community that I talk to on a regular basis, but I'm not active
on the forensic front anymore.
0:31:26.4 DS: Yeah.
0:31:27.5 JR: Similarly...
0:31:27.9 DS: Sorry.
0:31:29.4 JR: Similarly, I also, at that time I was on the Department of Justice Office of Inspector
General team that was taking a look at some of the practices that the FBI laboratory, and I was on
the scientific advisory board for the ICMP, for both of those I took a step back. It's one of those
classic things where if you're not actively involved with it, pretty quickly your accreditations go
away, you fail on what's happening in the field. Lots of friends in the field and it's exciting to see
where it's gone. So we sort of had our moment where we were able to have an impact on this, and
then we've been continued to be engaged by all the challenges that Myriad is throwing at us in the
clinical arena.
0:32:12.9 DS: Yeah, plenty to do there. No, really appreciate personally all the work that both of
you have done and shared here today. This is incredible, and hopefully the listeners have a... I didn't
know any of this, honestly didn't really think about the background of how to identify these
individuals and get information back to loved ones, so this was quite revealing on a lot of fronts. So
thank you both, incredible commitment, long hours, I'm sure during that time. We're still, as you
12/14/21 Page 9 of 11
9.11 Response Podcast
brought up John, still doing good now just helping out on the clinical front, which is a huge need as
well. But it's interesting to hear some of the origins of the company and the decisions that were
made to get us to where we are today, so thank you both for your time.
0:33:12.4 JR: No, and thank you for the opportunity. Just first up, I'll say that Benoit's estimate of
his hours is very conservative.
0:33:19.8 DS: I bet.
0:33:21.1 JR: He would [0:33:21.2] ____ night and day and all weekends. A lot of the people of
the company were. This was truly a companywide
effort, so many people were involved with this
and they take deep pride in having been involved and been successful at it. So it's an honor to come
here and talk about it, but frankly, when I come here and talk about it, it is representing the
hundreds of people from Myriad Genetics that were involved with this.
0:33:46.7 DS: Well, thank you both.
[music]