Hi, my name is Paul Offit. I'm from the Children's Hospital, Philadelphia and the
Perelman School of Medicine of the University of Pennsylvania.
What I thought I would talk about in this lecture is rotavirus vaccines.
And it's the only, actually, what I'll talk just about one specific vaccine for the whole lecture.
And it's not because I think rotaviruses are especially important among the vaccines,
rather it's because I actually was fortunate enough to be part
of the team at Children's Hospital, Philadelphia that developed this vaccine.
So I really watched it go from, from early research through development and so
basically from bench to bedside and it was an educational process for me.
And so I thought it would be fun to kind of go through this story.
So rotaviruses are a virus that infects the small intestine and it causes fever and vomiting
and water loss or dehydration in young children.
This is a study done by Bill Rodriguez at the Children's Hospital in DC.
And he looked at rotavirus as compared to other viruses that cause the so-called stomach viruses,
and found that rotaviruses were particularly capable of causing vomiting and dehydration.
And that's why, when you have the stomach virus, you sort of lose water when you have dehydra-
when you have diarrhea, you lose water when you have fever,
and it's hard to rehydrate yourself when you're vomiting.
So that's sort of those three things together can rapidly lead to dehydration,
which can cause hospitalization, and death.
This is just a stained section of the small intestine showing the virus infecting the intestine.
And what you can see here is that,
if you look here at this sort of, these kind of finger-like projections, so-called villi, into the small intestine,
You can see that the virus is infecting these cells that are detected
as so-called mature epithelial cells that line the intestine.
That's what rotaviruses do. They infect those cells, and they damage those cells,
and they make it very difficult to resorb water.
And it causes diarrhea, and it causes vomiting, and it causes dehydration.
So, every year, in the United States, prior to the vaccines are being licensed and used,
which was around 2006, rotavirus accounts for about 2.7 million cases, about 500,000 doctor visits,
about 270,000 emergency department visits, 70,000 hospitalizations, and 20-60 deaths each year.
If you assume a broad cohort between 3.5 and 4 million children every year in the U.S.,
that meant that about 1 out of every 50 children born in the United States would be hospitalized
with dehydration secondary to rotavirus infection.
Those numbers are dramatically decreased since the vaccine came. And then we'll talk about that.
In the developing world, rotavirus is a killer. It accounts for about 500,000 deaths a year.
That means about 2,000 children die every day from this virus, from the dehydration caused by this virus.
Actually, it's the single agent. It's one of the most important killers of infants and young children in the world.
And for that reason there was a tremendous amount of interest
both publicly and privately to try develop a vaccine to prevent it.
Now, the original idea on how to make a vaccine, and it's sort of Edward Jenner-like approach.
And I talked about Edward Jenner in another lecture, in a history of vaccines lecture.
But the initial idea was why - since all mammals that live
on the face of the earth have their own rotavirus strain -
why don't we actually do the same thing that Edward Jenner did?
Remember, Edward Jenner used the cowpox virus to protect against human smallpox.
He - we now know--he didn't know that, but we now know the cowpox is similar enough to,
was similar enough to human smallpox, that infection with one could prevent disease caused by the other.
So, really, that was the same thinking here was
'why not use a non-human rotavirus strain to protect against human rotavirus?'
This just shows you all the [inaudible] species that could be affected by rotavirus.
But species barriers are high, meaning that as it was true, smallpox were,
cowpox could cause significant disease in cows but not people,
and human smallpox could cause significant disease in human but not cows. That's also true here.
So for example, cow rotavirus could cause severe vomiting and diarrhea in cows,
but really doesn't do that in people. And vice versa, human rotaviruses really don't cause disease in cows.
So, it's the species barriers that are high, and that was the original idea.
So what happened was, there was a group at the National Institute of Health,
headed by Al Kapikian, Rob Chanock, Taka Hoshino, Harry Greenberg, Jorge Flores and others;
who looked at a strain called RRV. Now that strain stands for Resist Rotavirus.
This was a virus, a rotavirus strain that was isolated from a monkey in Northern California in the 1980's
by H. Milerby. That virus was then purified by growing it in cell culture, and not so much weakened.
The notion was that it would be weakened in humans because it wasn't a human virus.
This is sort of the same idea Edward Jenner had.
And when the vaccine was then given by mouth at 2, 4, and 6 months of age
to children in Sweden, and in Finland, it worked.
It seemed to protect against the moderate to severe dehydration
as it's shown on the first two rows on the slide.
But when the vaccine was then tested in Rochester, NY in a trial; it didn't seem to work.
So, I think what people concluded after this series of trial was that a non-human virus - in this case, a monkey virus -
was inconsistently capable of protecting against human rotavirus disease.
Now we worked, and by we, I mean Stanley Plotkin and Fred Clark, who headed this program;
worked at Children's Hospital, Philadelphia; with a strain called WC3.
And that was actually isolated from a calf with diarrhea at the Kenneth Square facility
which is the large animal facility, the , the veterinary school here at the University of Pennsylvania.
And we took that virus and then we went back to the Wistar Institute
and we purified it by passing it in a cell culture.
And since it was from the third calf that we tested, it was called WC3 or Wistar Calf Three.
We gave this virus by mouth to children at 2, 4 and 6 months of age in Philadelphia
and found that it was excellent at protecting against moderate to severe disease
and actually very good at protecting against even mild disease caused by rotavirus.
But when we then tested the vaccine in
Cincinnati or in Bangui in the Central African Republic,
we found that it didn't work well. So basically we've found
with the calf strain exactly the same
thing that the NIH researchers had found
with a simian strain, a monkey strain.
That the protection against rotavirus could occur
with non-human strains, but it was inconsistent in it's ability to protect.
So it was really back to the drawing board. And this slide shows what
the drawing board was. We, we needed to
determine which rotavirus proteins were responsible
for evoking virus specific neutralizing antibodies. In other words,
what rotovirus proteins induce protective
immunity and which rotovirus proteins
were responsible for viral virulence. So it is
the way that you make any vaccine.
You're really trying, you're trying to separate
out the part of the, in this case, the virus
that is pathogenic, that is disease causing from the part of the virus
that induces an immune response which is
protective. Hoping in this case
that the genes that are responsible for making the proteins that cause disease are different
from the genes that are required to make the proteins
that induce a protective immune response. So
just a brief word about rotavirus structure,
the cartoon of the rotavirus particle is shown on the right and there are two proteins
to focus on. One is the protein called VP4 which stands for
viral protein four. It's the viral protein that's responsible for
binding to cells before it infects them.
It's also called a P protein,
or a P serotype and you'll see what I mean by
that in a second. P, just because it's sensitive to
proteases. Proteases are proteins that cleave proteins.
and this, in order for the virus actually
to enter the cell that VP4 needs to be
cleaved to two smaller proteins called VP5
and VP8. The other protein defunct to
focus on is called VP7, which just stands
for Viral Protein seven. And that's kind
of a coat protein. But again, on the
surface of the virus. And it it's a
glyco-protein, so it's also called a, a G
protein. I'll talk about that also in a second;
so, two surface proteins, VP4 and VP7.
And as a general rule when you're trying to protect against viral infections, you want to try and make
antibodies to the surface of a virus; so to this, surface of a bacteria.
Because that would then prevent the virus or
bacteria from binding to a cell, entering a cell
and causing disease. It's sort of a universal truth of vaccines. You're really,
for the most part, trying to prevent
virus-cell binding.
Then, the rotavirus genome consists of eleven separate
segments of double stranded RNA.
And if you look at sort of, well, column A if
you take the virus
and you disrupt it with a, just a detergent and then you
put it on top of,
so called, poly-acrylamide gel which is just a plastic
mesh and then you take the virus
and you run it down with an electric current. You
can actually separate those individual
double strained RNA segments in.. by size.
So you can see the, the so called electropherotype of
of virus A. And virus A and virus B are different.
They're different strains therefore they
have different ways in which their double-stranded RNA
segments migrate in this gel.
Now, if you take those two
viruses and you co-infect cells at the
same time,
you find that about 20 percent of the time, the progeny viruses that are generated
actually are reassorted viruses. Which is to say
that they're a combination of the two viruses.
So you can see at the virus strain on the right has all of its genes from
virus A, and only one gene from virus B.
Well, if virus A and virus B are different within their capacity to cause disease in experimental animals,
in our case, we looked at mice, then you can
figure out the genetics of virulence.
If those two strains are different with
regard to their ability to induce
neutralizing antibodies in the serum, then,
you can say - - you can figure out the
genetics of serotype which is to say that,
that you can distinguish viruses based on
their ability to evoke antibodies which
neutralize specific strains or specific serotypes.
And so, again, by making these reassortant viruses as you see on the right,
you can figure out the genetics of virulence.
You can figure out the genetics of
serotype, or said another way,
neutralization phenotype, or said another way,
just the genes responsible for evoking protective antibodies.
So to make a long story short, to sort of summarize
ten years worth of work in one slide,
it's actually a little depressive. You can't do
that, but to summarize ten years
of work in one slide, what we found was
that each of those two surface proteins
evoked neutralizing antibodies or said
another way, each of those two different proteins
evoked or determined serotype.
So in a sense then, then rotaviruses actually
are very similar to the influenza viruses,
which also are distinguished on the basis
of two surface proteins.
And that's the way that those viruses are characterized.
H1N1, H5N1, the so-called bird flu.
That's the way that the influenza viruses are characterized and that's the way the rotaviruses are characterized also.
So, G1P1, for example.
The addition studies that were done
by researchers at the National Institutes of Health specifically headed by
Taka Hoshino as well as researchers in our lab
determined ultimately
that there were four genes that were required for virulence. So by doing studies in mice, we showed that really
there were four different genes, all of
which were required for virulence.
Well, this was good news. It meant that you
could include the genes, the human genes
that were responsible for invoking
neutralizing antibodies, but as long as
you didn't include all four that were
responsible for virulence, then the virus wouldn't be virulent,
or said another way, the virus shouldn't cause disease.
And this was advanced beyond Max Tyler.
Remember we talked about Max Tyler in another lecture.
And Max Tyler was the one working at
the Rockefeller Institute who showed
that you could weaken viruses by
serially passaging them in a cell culture.
But it was done really in a blind way, in a sense that
you would pass it in non-human cells like in his case,
you pass the Yellow Fever virus in mouse,
mouse embryo cells, or chicken embryo cells.
And then he would take the virus out
to see whether it was weak enough by testing it in people.
We, at Children's Hospital Philadelphia, at least had,
we thought we had defined virulence genes,
by doing studies in mice.
But again, you know, these were studies in
mice, you never really know
until you test it in people.
And so, you have to go very slowly, when you test in people,
starting in adults, who you know have
already been exposed to rotavirus and have antibodies
And then working your way down, ultimately to children who hadn't been exposed to the virus.
So this was reassuring, but, you really never know until
you do the right kind of studies.
So the first rotavirus vaccine
was, that was out there was called a Rota
shield. It was again the NIH reserchers
had taken their simian strain, their so called research rotavirus or RRV strain.
And they had it re-assorted it into that genes that determine
human serotype, so called G1, G2, and G4.
They assumed that the RRV strain was similar not
to the human G3, so there really there are
sort of four major human G serotypes;
G1, G2, G3, G4 and [inaudible] take this,
this simian monkey strain.
You can then reassort into those human genes that
determine those serotypes.
And so that's what was done with this original vaccine.
And, and the vaccine was on the,
was actually brought onto the market in August 1998.
It was on the market for about 10 months in the United States when this headline
was published in a CDC journal called
Morbidity and Mortality Weekly Report.
It was called intussusception among recipients of rotavirus vaccine in the United States,
1998 to 1999.
And what had happened was, intussusception
is caused when the small intestine sort of folds into itself, or invaginates into itself
and gets stuck.
And when that happens, there can be a critical loss
of blood supply to the intestine mucosal surface,
which can cause intestinal mucosal surface damage
and bleeding.
In addition, because the intestine has living on its surface,
trillions, literally, of bacteria.
Those bacteria can then enter
the bloodstream and cause bloodstream
infection, which can be overwhelming, even resulting
in death. So intussusception is an important medical problem. It's a serious
medical problem often requiring hospitalization.
And so, the fact that,
the, the that headline appeared in the
morbidity mortality weekly report was
really worrisome. And it was based on reports to VAERS, we've talked about in
another talk, the so-called Vaccine
Adverse Events Reporting System.
This is a system that's passively, it's sort
of a passive reporting system
to the FDA and the CDC.
If you're worried that, a vaccine
could have caused a particular event, adverse event, then you can report it to this system. And then there
were fifteen cases of intussusception were reported
to various following Rotashield.
What was worrying, worrisome, was that
thirteen of those fifteen cases occurred after the first dose, which sort of lends to
the biological plausibility of the fact that these two things were causally associated.
Eleven of those thirteen cases, again, recur within seven days of vaccine administration,
which is what you would expect if the vaccine was causing it. An eighth of those thirteen cases
occurred in children 2-3 months of age.
Now, intussusception occurs anyway. Intussusception occurs in children before there was a Rotavirus vaccine.
But, it really occurred primarily in the 5-9 month-old.
So the thinking was that
now that we were seeing it in 2-3 months of age,
when these kids were getting vaccinated,
that again lent the fact at least the fact that this notion,
that the, that the vaccine was causing
the intussusception was biologically plausible.
So the way that you answer this,
the question, is to do the kinds of study
that were done by Trudy Murphy and her co-workers,
and reported in
the New England Journal of Medicine. The CDC
really took the lead on this.
Jeff Koplan, who was the head of the CDC at the time, stopped doing some other projects,
lot of money into this, to see whether or
not that association between
intussusception and vaccination was
temporal, or causal.
And, and you gotta take your hat off to this group.
That the minute that they saw something
that might have been problematic, they,
they immediately addressed it, and found that
if you received the rotavirus vaccine,
and your first dose was
you received your first dose at 1-2 months of age,
or 3-5 months of age, or 5-8 months of age,
that within a week of getting that vaccine,
you had a 25 to 30 fold increased risk
of getting the disease --
getting the intussception
than if you didn't get the vaccine.
Similarly, there was a statistically signifigant increase,
if you, within 2 weeks of getting the vaccine,
but not greater than 2 weeks after getting the vaccine.
So this was supportive of the notion that, the
rot-, the intussusception following Rotashield vaccine
was not a, just a temporal association,
but was, in fact, a causal association
that the vaccine actually was causing
this intestinal blockage.
For that reason, then the company had actually decided to
withdraw the, the their vaccine from the
market and, and that, that was done after
the vaccine had been on the market for
about ten months. Now, you could argue
that there is a difference between,
relative risk and attributable risk.
Relative risk, as we saw, was 25 to 30 to
one, following that vaccine.
But the real risk or the attributed risk was
roughly 1 per 10,000. So let me try and
give you an example of that. That if I am, if I
walk across the street in front of my
house, I have a certain risk of being hit
by a car. That would, no doubt, be much
greater than the risk than if I was just
standing. On the, the just in my doorstep
in front of my house. And so the relative
risk would be very high. It could be a
1,000 to 1. But still the attributable
risk is, is very low. I mean, I cross the
street in front of my house all the time
and don't get hit by a car. That was true here.
So while the, the relative risk was high,
the attributable risk was pretty low.
Still 1 in 10,000 recipients.
And you could argue that if you took a
theoretical million children who either
did or didn't get the vaccine. Far more
who didn't get the vaccine would have been
hospitalized. And, and then frankly, even
five to ten fold more would have died from
not getting the vaccine. Because Rotavirus
kills children even in this, in the United
States. But I think at the time, the
disease wasn't perceived to be terribly
dangerous. Certainly no one wanted to get
something that you knew could have caused
intus susception and there was one child
who died of intussusception following this
vaccine even though six to twelve will die
from not having received the vaccine among
a million children. So you could have made
the argument that even with that adverse
event that still the benefits outweigh the
risk, but that's not the way that the
played out at the time and so the vaccine
was taken off the market. Now, I think the
saddest part of this, and the biggest
tragedy, is that there was a World Health
Organization meeting held in February of
2000. So it was about four months after
the withdrawal of Rotashield from the
United States, where the company was
great. I mean, wyeth made this vaccine,
stood up and said, we, we, we'll give you
the virus, the vaccine strains. We will
give you, the you know, the technology on
how to make it. We will give you the cell
substrate on which we grew this,
thisvaccine. We'll help you build the
buildings. Because the Wyeth knew that
they had a technology that now they were
not giving to American children. They
could've saved. As many as 2000 lives a
day in the developing world. But, But
count-, country after country stood up and
said, you know, if it's not safe for our
children, it not - if it's not safe for
America's children, then it's not safe for
our children, therefore, we're not going
to, to use it. Even though the risk
benefit ratio was obviously very different
in the developing world. So, I think the
tragedy was, if in this whole story, is
that for seven years, a technology that
could have saved a lot of lives in the
developing world, and certainly even have
prevented a lot of suffering in the United
States, sat on the shelf. And so, two
other, companies actually stepped forward
to make vaccines. One of them was Merkin
Company, who, who, in collaboration with,
researchers at Children's Hospital of
Philadelphia, developed this, this next
vaccine. And again, it used that bovine
strain that I talked to you about, that's,
that's called iii.'Cause we certainly knew
that, that virus was safe in children. It
just didn't seem to be, consistently
protective enough to make it a vaccine.
And so what was done, was the same thing
really that was done with the RotaShield
vaccine, which was to make a series of
reassortant vaccines. And again, we talked
about reassortants to some extent in the
in the history of vaccines talk. That
included those human genes that, that
represented those, the, the proteins that
were responsible for, for a serotypes G1,
G2, G3 and G4, as well as one of the, the
P serotypes, so called P1. So that's,
that's what the RotaTeq vaccine was. It
was given. As three doses by mouth at two,
four and six months of age to, to
children, It's This. Sorry, the, the, the
dose was 1.6 *ten^6 plaque forming units*
per strain. A plaque form unit is, is just
when you test a virus. In cell culture,
the virus can reproduce themselves, and
destroy cells, as well as cells in the
immediate area, which causes this sort of
plaque in the, in the or hole in the
culture, and that's how you determine
that. But the important to know here is
that it was three doses given by mouth at
two, four, and six months of age. There
were some advantages of the bovine strain,
because the cows are not as close
phylogenetically to humans as primates
are. The there actually was a. To provide,
less, for instance of sort of these common
less adverse events because of our
assisted reproduced self as well. So I
think the important thing here is, is that
the vaccine was tested in, in more than
70,000 children prospectively, in eleven
countries It took four years to do this
study, it probably cost about 350 million
dollars to do the definitive phase three
trial of this vaccine. And what was found
was that the vaccine was capable of
preventing any roto virus disease being
mild, moderate or severe disease at about
74%, which is roughly the equivalent of
what natural infection capacity is to
protect against roto virus disea se.
Disease, Efficacy against severe disease
was 98%, against rotavirus
hospitalizations 94%, against rotavirus
doctor visits 86%, and there was no
clinically significant increase in fever.
Or vomiting or diarrhea. Listlessness,
lethargy, or poor feeding versus placebo.
So, the vaccine appeared to be safe and
appeared to be effective. Importantly,
regarding [inaudible], within fourteen
days of any dose, there was only one case
of in the vaccine group a one in the
placebo group. Within 42 days of any dose,
[inaudible] six, six weeks of any dose,
there was six in the vaccine group, five
in the placebo group. And within one year
again, twelve in the vaccine and fifteen
in the placebo group. So, again, the
vaccine didn't appear to, to cause
[inaudible] or prevent [inaudible]. And so
those were, And, and we also have post
licensure data that I'll talk about a
little later. [inaudible] has been
introduced, this vaccine has been
introduced into the developing world.
Specifically in Nicaragua, Bangladesh,
Vietnam, Ghana, Mali. And just within this
past week Rwanda. And that's a tribute to
the Bill and Melinda Gates Foundation, who
have made that possible. you know, the,
although the World Health Organization
certainly considers all the [inaudible].
Caused by rotavirus important and, and
worthy of prevention. The World Health
Organization doesn't really have the money
to introduce this vaccine in the
developing world and, and frankly or the
countries, the countries also don't have
the money. So, that Bill and Melinda Gates
choose to spend their money by preventing
a disease which causes a lot of suffering
and death in the developing world is a
tribute to them. Now there's, there's
another rotavirus vaccine that, that's,
that's developed, that was developed by
researchers at Children's Hospital in
Cincinnati specifically Dick Ward and
David Bernstein, using the more classic
way of making a vaccine, a live weakened.
Vaccine, which was to take the vi rus, and
serially [inaudible] in non human cells as
a way of attenuating it. This vaccine is
given as two doses by mouth, which is an
advantage. It's one fewer dose. At two,
two to three, and then four to five months
of age. It's given at a somewhat lesser,
dose. Because the virus is, is a human
virus. So therefore, somewhat better at
reproducing itself in the intestine than
the bovine strain was. The, the
researchers at children's hospital in
Cincinatti, in collaboration with
GlaxoSmithKline, did a study of, 60, more
than 63,000 infants who were given this
vaccine at two to four months of age. And
the studies were performed in Latin
American countries as well as Finland and
the vaccine worked very well, preventing a
severe disease about a, [inaudible] an
efficacy rate of 85 percent and preventing
80, hospitalizations at a rate of about
85%. Rotateq. Was introduced in the United
States in 2006. Rotarix in the United
States in 2008. Although Rotarix was
introduced in the developing world also in
2-, 2006. And, what we found was that.
there was a 50, even in the US, 50 percent
immunization rate causing 80 to 90 percent
reduction in hospitalization. So it's a
there is her immunity that is being
induced by these two vaccines. So, so what
about an [inaudible]. Why this, the Rota
Virus vaccine, why did Rota shield cause
an [inaudible] interest. There were data
presented from Brazil, Mexico, Australia
and the United States on October
twenty-eighth, 2000 [inaudible] that show
the [inaudible] was actually was a rare
consequence Rota risk in Mexico and
Australia. Not at the one in 10,000 level
that was seen for Rota, Rota Shield. But,
but more an attributable risk of one per
60,000 or 90,000 doses. And also, for
RotaTeq as well, intussusception was a
rare consequence in Australia, with an
attributable risk again between sort of,
60,000 and 90,000 doses. So it now looks
like all three Rotavirus vaccines,
including Rotarix, which is simply an
attenuated human Rotavirus strain, causes
intussusceptio N. So it sort of causes us
to re-evaluate why. Rotashield caused
intussusception, I think the most likely
reason is that natural rotavirus infection
is also a rare cause of intussusception.
So then the question becomes which is
rarer, Intussusception caused by. By the
vaccine, or intussusception caused by the
disease. Because remember the vaccine
prevents the disease. So if, if
intussusception caused by the, the vaccine
is more common than intussusception caused
by the. The, the natural virus, then, the
rates of interception should go up in
countries that use the vaccine. Conversely
if, if interception caused by the. The
natural virus is more common, then
[inaudible] caused by the vaccine, given
that the vaccine prevents natural
infection. Then as you introduce the
vaccine, rates of [inaudible] should,
should go down. And sort of all of the
evidence to date, is it certainly raises
the [inaudible], the [inaudible] seem to
be going up, and if anything they seem to
be going down. A little. So I think that
what we've learned is that in the long
run, that [inaudible] is probably is a
consequence of the natural infection. It's
probably to some extent prevented by, by
vaccination. I mean it's interesting to
know what, what, what had happened if we
had found out that [inaudible] shield had
caused [inaudible]. This is after four
years after it had already been
introduced. Because now, we know that
rotarix and rota, rotatech are two more
recent vaccines for saving lives in the
developing world are certainly preventing
hospitalizations and suffering in
developed countries like the United
States. I wondered whether we would have
still made the same decision rotashield
that we found out that it was a rare cause
of [inaudible] four years later. So I'll
stop right there and thank you for your atte