CONVERSATIONS AT CHR: Breast Cancer Risk After Recent Childbirth


Hello, I’m Norbert Gleicher MD and I’m the
medical director and the Chief Scientist at the Center for Human Reproduction. And it
is my special pleasure to be here today with David Barad, who is the Clinical
Director of our IVF Program here at CHR and the Senior
Scientist at CHR. And this is a new series we are starting today where we
have conversation about a topic which we here at CHR feel is very important and may
have not gotten all the necessary attention. And a very good example
is a recent paper that was published in the Annals of Internal Medicine, with the
first author being Hazel B. Nichols and the senior author Dale Sandler. And this
paper really revised our thinking over many decades about what pregnancy does
in regards to breast cancer risks. Really for decades it was almost a
dark mark that women who have had children demonstrate
reduced breast cancer risks. And in converse, women who don’t have children
have increased cancer risk (apropos the same thing applies also for ovarian cancer). This
study, which involved a large number of investigators and a huge number of
patients’ data, came to very different conclusions. And this is exactly why this is
my real pleasure to have David here with me today because he is not only an
excellent reproductive endocrinologist, but he’s also an epidemiologist and has a
deep understanding about the analysis of these kinds of population studies. And this
kind of understanding is very very important to really interpret the data
of this paper. And that’s, David, why I would like you to take over here and
maybe give to our listeners a quick summary of what the
paper shows. And maybe even more importantly, put it into a certain
context. Yeah, so first of all the thing to understand is, as Dr. Gleicher said,
for many years we’ve understood that pregnancy’s protective against breast
cancer. So, that was surprising that this paper had come up with a different understanding, a different kind
of conclusion. To understand this better, this particular
paper is a combination of twenty previous studies. Now, these were
prospective studies that were designed to try to see what are the factors that
contribute to the onset of breast cancer and they’re collected over many years
under different settings by different groups, and individually none of those
studies showed a positive effect. But these days, what people do is they
take together these separate studies and put them together into one very large
population. So, in this particular study, there were over 9.6
million person years of follow-up. What that means is, if you had ten people all
for one year, those be ten person years. If you had twenty people followed for two
years, those would be forty person years. And so by looking at the number of years
that they were following people and the number of people that were followed, they
come up with this 9.6 million (more than 9.6 million person years) obviously there were 9.6 million people in
the trial, just some of them were followed for longer periods of time.
But what this allows is that you then can establish consequences not only at
one time point 25 or 30 years down the road, after pregnancy, but
you have enough data to establish time points.
Not only that, but these were prospectively followed groups. So each of these groups, they
picked a select population, and they followed them consecutively over time
with serial observations. So whereas, in some studies, where you just sample whole
population they’d be ask them, “Were you ever pregnant?” that would be
a cross sectional observation (a slice at a particular time). But in this, they’re able to
put together the effect of an exposure (namely pregnancy in this case)
over a period of time. So, they can calculate these person years of exposure.
And they can tell us what’s your risk a year after pregnancy, what’s your risk
in five years, ten years, twenty years in a way that, in the past, we would look
at women who were in the primary age group for breast cancer (mostly
postmenopausal people), and we would retrospectively say, “Were you
ever pregnant?” and “What was the effect there?” And now you’re able to get that
follow up over the individual years following a pregnancy. So, if I understand
you correctly, what that then means is that the new data (which obviously
suggests that pregnancy increases risk for breast cancer, at least for certain
time period) does not necessarily contradict the old data. Because
the old data was based on a long term follow-up and this is really the first
study that has the patient volume to tell us what’s
happening between immediately and 30 years old. So, most of the old data, well
some like the Nurse’s Health Study, would be following people over time, but
a lot of the older studies would simply be cross-sectional. They’re not really
follow-up. They’re looking at a bunch of people who are 55-60 years old who have
breast cancer and saying, “What were you previously
exposed to? Did you ever have children?” And, of course, in that group, women who never
had children work out to be at an increased the risk of breast cancer. In
this study, in contrast, you have a bunch of these groups put together into
this large group followed year after year to see what happens in their life.
Did they have a child? Did they have a disease? And then looking at that
longitudinal effect. So you have a little bit more detail and you can truly track what the effect of an exposure (in this case pregnancy) is over time. Now,
at the end of the day, the increases that they talk about– they say
there’s this increased risk, they estimated it to be what they call a “hazard
ratio” of 1.8 times (that means if there were no risk it would just be one), so
this is an 80% increased risk, of those people that were exposed. And
the estimate could be some variance between 63 and 99 percent
increase, but let’s go into 80 percent. That sounds really scary because it
sounds like you’re almost at 2 times greater risk of getting breast cancer if
you’re pregnant than not. But this is a relative risk, it’s not an absolute risk.
So, for instance, the people who are 45 years old in this
case, there were about 620 cases among a 100,000 women if they did not
have breast cancer and if they did not have a
pregnancy. And if they did have a pregnancy,
it increased from 620 to 661
cases per 100,000. So, the actual number of people affected aged 45
because of pregnancy for a 100,000 was 41. So, we’re not
trying to tell people, “Don’t ever get pregnant because you’re gonna increase
your risk of breast cancer.” It’s not an absolute risk, but it’s a relative risk
and that makes it interesting because we want to try to understand what is the
mechanism that leads to that relative risk. That’s exactly the point that we want to get
to here because our understanding, our belief that pregnancy
protects from
breast cancer risk was really based on the fact that
males have something to do with lactation and with breastfeeding, etc.. We
never really thought about a deep underlying physiological explanation where
that connection actually could be. And, I think, it’s exactly the time studies, that
we see here in this paper, that allow really for the development
of some very interesting hypotheses. Because, what the study showed is that
that risk, as small as it may be, continues increasing after pregnancy for
about four years. Yeah, it peaked at about 4.6
years and then slowly decreased until about
almost 20 years. Which is the usual time when studies were
evaluated and only after around 20 years, indeed, then do we see a decline because those
women– who were destined or may have genetic predispositions or other
predispositions– they had their breast cancers already. And so, the remaining ones
have actually lower risk. So when we did the previous studies, that was on that time
period when it appears to be protected. But, now, you have to remember the incidence
of breast cancer increases with age. And so, although very few women
who are young– say a thirty-year-old has a baby, there may be an increase
5 years after she’s had the baby as a relative risk. But, a percentage
increase risk in the incidence of breast cancer is going to affect many more
people who are 55 than who are 35. So, if I can summarize the first part of our
discussion, there, indeed, is news here in the sense that it seems that pregnancy
(at least on the temporary basis), increases the risk of breast cancer.
Not altogether temporary, up to 20 years with a peak
at around 4-5 years. That increase, I think it’s fair to
say, is very small. So, as you correctly
stated, nobody should understand that as, “Don’t get pregnant because you will get
breast cancer!” Nothing like that at all. What it really means, and I think here I
would like to come into the second (maybe a more important) part of our discussion.
What this really means is that there are some patients, who for some reason, with
pregnancy, develop a high risk of breast cancer. And therefore, what are
those reasons? What is the underlying pathophysiology that exposes
that small group of women to an increased risk of breast cancer in
association with pregnancy? The first thing that most
endocrinologists would go to would be the exposure to hormones. Estrogen, to be more
specific. Well, except, you have to remember, for instance, in the Women’s
Health Initiative, the group that was in the estrogen-only group didn’t see an
increase in relative risk. Whereas, those who were taking estrogen plus progesterone
did see an increased risk. And interestingly, in earlier studies,
there had been a similar increased short-term risk and after exposure to birth
control pills. Some other similar prospectively followed things. So, I mean,
you could make an argument that has to do with exposure to actually
progesterone or progestins. Some other people have suggested that maybe
a certain subgroup of breast cancer patients that already had breast
cancer and that visually small breast cancer is fed by those hormones,
that it doesn’t create new cancer. That was one of the explanations for the old birth
control studies, which are 20 years old now. But, we have some other ideas. Yeah,
and I think that is what makes the study, in our opinion, so interesting.
Because when one thinks about hormone effects,
they’re usually relatively short-lived. And here we are seeing a relatively
long term effect (up to 8 to 20 years). Does it seem likely that the hormone
effect may last for 20 years? I mean, there are different kinds of hormonal
effects. So, there’s the hormone effect if you give something, endometrium
grows, becomes hyper-plastic, some tissue grows under the influence
of a steroid. But, there are other modulating effects on the genome
that changes the way it’s methylated, changes the way it’s expressed as an
epigenetic. That possibly could have a longer-term effect and I think that’s
part of the understanding of some of the carcinogenic effect of certain
exposures. Even estrogenic exposures in the environment. In the environment they call those epigenetic estrogen effects. But that
could also have effects, the same kind of modulation could have effects on other systems. And I think one very attractive hypothesis that all of this
really raises is whether this isn’t an effect on the immune system.
The rationale behind that is that we all know that the maternal immune
system has to reprogram itself in order to tolerate the pregnancy. In other words,
the maternal immune system reprograms itself to its tolerance towards that
very rapidly growing fetus, so this is not even comparable to an organ
transplant. Because there, the mass of foreign material remains the
same. If you get one kidney, the kidney doesn’t grow tenfold, hundredfold,
thousandfold, after your transplant. Here, you have a “transplant” that
grows exponentially and yet, is not being rejected by the maternal immune system. And when we
do a transplant, we’re giving immunosuppression which totally
suppresses things. But here, you’re actually reprogramming the immune system to
accept this very different individual growing zone. And, we know that there
are some suggestions that what happens during that reprogramming, the
induction of so-called “tolerance pathways,” is something that pregnancy has
in common–potentially– with successful organ transplantation, but also with
carcinogenesis and particularly metastatic disease. And we now know from
recent studies in oncology that tumors really have various tricky
ways of disarming the immune system on top of everything else. The immune system has mechanisms, some of which may be borrowed
from the whole maternal embryonic adjustment of things. Of course, yes, that is increasingly being recognized that the micro
environments, for example, in the early pregnancy, and in organ transplantation,
and in cancer, have a lot of similarity. So, the hypothesis that we are
talking about is the idea that the maternal immune system, in going
through pregnancy, at least partially disarms itself. And that disarmament
allows either already existing tumors or even fresh tumors to grow. While, with a
completely normally functioning immune system, those tumors would be destroyed
by the immune system but yet, they do not get destroyed. This is
obviously a very interesting concept if it is what is happening because that may
have significance then for many other tumors. I mean, starting with ovarian
cancer, where the same association has been believed to exist now for decades.
Meaning, pregnancy protects against ovarian cancer. So, that raises the question
whether that is really true, but it would even apply to other tumors
possibly. So, what this study really suggests is that we need to look at
cancer following pregnancy in a much broader way. All right, so if we can
summarize, a very important paper that for reasons which we here do not
understand, has gotten a surprisingly little attention in the broad media, but
even within the medical community. [This is] a paper that may have significant
consequences if what we discussed here has, indeed, relevance. [This is] not a paper that
should alarm anybody, by any means, because the observed increase in risk is
really quite small. But, [this is] a paper that should excite mostly basic
scientists who want to understand how the immune system interplays with pregnancy
and also with cancer. David, I thank you very much. It was very interesting,
thank you.

One Reply to “CONVERSATIONS AT CHR: Breast Cancer Risk After Recent Childbirth”

Leave a Reply

Your email address will not be published. Required fields are marked *