What we did is we took two of our best databases and sets of estimates statistically; we took data from our cancer registry, the SEER program, and we also had information from the Census Bureau because they have made some very useful population projections that incorporate a lot of factors like longevity and births and so on.

So we put those two sources of information together to come up with an estimate of the breast cancer population and the incidence of breast cancer in the US.

What we did was we can estimate the rate of cancer from the cancer registry, we can multiply those by population estimates from the census and come up with an estimate of burden.

So what is the news, then, here in the United States?

I think there are two points.

When we made our forecast for the total number of cases, not surprisingly our forecast is that it’s going to increase because we already knew that it was going to increase.

Our particular estimates, based on the most recent data, are for all breast cancers combined, including the in situ cancers as well as the invasive, an increase from around 283,000 cases in 2011 up to around 440,000 cases in 2030.

So that’s about a 50% rise.

There are three factors driving the increase: the first is we have more people.

More people are getting into the cancer-prone years.

We have a big baby boomer cohort in this country which is very large and so there are just a lot of women who are getting into those years.

The second is that life expectancy is improving so many women are just living longer and living long enough to develop a breast cancer.

The third factor is the rates of breast cancer in the United States are very high, as they are in many developed countries.

Not only are they high but the incidence of the most common type of breast cancer, the oestrogen receptor positive cancers, has actually been increasing slowly.

So when you put all that together in the models we come up with the estimate of about a 50% overall increase.

There were some interesting discontinuations in your graph. It seems that what women were doing in terms of hormone replacement therapy made a difference to the breast cancer incidence.

Yes.

That is data that were previously reported and widely discussed in the literature and I pointed that out, the drop circa 2002.

I wanted to point that out, to just mention that our models can account for that sort of effect in order to get good estimates of the other parameters that we need to make our forecast.

Why is oestrogen receptor positive breast cancer on the increase?

That’s a great question.

The strength and the weakness of our study combined, the strength is that we don’t have to know the reasons why to make a forecast.

So the forecast is useful for planning purposes and we can make the forecast without knowing all the reasons, that’s the strength.

The weakness is that we don’t have enough information to try to partition the increase into different sorts of risk factors, that would be a different type of study.

Presumably the increases in ER and the decreases in ER- are both reflecting all those factors, the lifestyle and so on.

It’s not clear at this time how to partition up and know exactly which causes contribute how much.

The specific information is useful in other ways but that’s not what our study was really designed to do.

We were asking how many cases were there going to be.

It is interesting that ER- tumours are on the decrease. What are the implications of that?

The implication is that we don’t really know all the reasons why and so one of the points of a descriptive study such as ours is to highlight that it’s happening, to raise awareness of that, so that investigators can go out and try to find those reasons.

One thing that is exciting, and the annual report to the nation highlighted this observation about triple negative breast cancer, which is the most common type of ER-.

They pointed out that parity without breastfeeding appears to be a very strong risk factor, so understanding the importance of this.

But our study says it’s important, it’s a population trend, that’s one association.

It would be really great to tease out all the different associations with other studies because there may be some prevention clues that have been overlooked.

You were also asked, and there were quite a lot of questions, about carcinoma in situ because you’ve categorised in situ cancer along with breast cancer. There were some worries about that, weren’t there, because we don’t know the natural history of DCIS. What did you make of that?

I think it’s interesting.

I hope our study points out the attention of understanding more about the natural history.

Again, one of the strengths of the modelling is we don’t have to know the details of the natural history to make a forecast of cases.

From a purpose of population, these are events.

Women have a diagnosis and so on and they’re reported as a cancer to the registry.

Very briefly, what do you think doctors, cancer doctors, should be doing about this at the moment?

Our data needs to get filtered through the experts who advise the cancer doctors and that’s what’s exciting.

We can provide information to help people with those expertise make good decisions and focus on important questions.

That’s the purpose of a descriptive study is to find clues.