CFS 2011, Chemotherapy Foundation Symposium, 8-12 November, New York, USA

Link between obesity and breast cancer explained?

Dr Andrew Dannenberg – Weill Cornell Cancer Center, New York, USA


It’s known that obesity is a risk factor for hormone receptor positive breast cancer in post-menopausal women; it’s also known that obesity can lead to worse outcomes in women with breast cancer. So we began our studies in mice and studied two different models of obesity in mice and were able to demonstrate that obesity leads to an inflammatory process in the mouse mammary gland. That inflammatory process is manifested by so-called crown-like structures - crown-like structures represent adipocytes that are either necrotic or in the process of dying surrounded by a halo of macrophages. In association with this inflammatory process one sees high levels of inflammatory mediators, molecules such as TNF and IL-1 and COX-2. Each of these molecules was previously known to turn on the aromatase gene which is rate limiting for the synthesis of oestrogen. So, having discovered this new inflammatory process related to obesity in the mouse mammary gland, we then asked whether aromatase levels were up concordantly and found that that was in fact the case. So histologic inflammation, high levels of inflammatory mediators and an associated increase in aromatase, all of which could explain the link between obesity and hormone receptor positive breast cancer, could explain that is.

A potential flaw in the work is would it only be a mouse phenomenon or would it, in fact, translate to women. Since we studied obesity in mice in both the dietary model of obesity and a genetic model of obesity with similar effects, we thought the likelihood of this translating to women was greater than had we only used a single pre-clinical model. So, based on the mouse findings, the first discovery of inflammation in the mouse breast with a histologic lesion, crown-like structure and associated changes in expression of genes, we then moved on to design a human study. We were able to demonstrate very clearly that the same histologic lesion that we had observed in the mouse mammary gland occurs in the breasts of women and we call that CLSB or crown-like structure of the breast. It correlated with body mass index, that is women who were overweight or obese were far more likely to have these lesions in the breast than lean women; I can add that the severity of the inflammatory process in the human breast correlated with body mass index, so women who had higher body mass index had more severe inflammation. We then went on to ask whether there was a relationship between body mass index and the diameter or the size of the adipocytes in the human breast and we found that elevated body mass, that is being overweight or obese, those women had larger adipocytes and there was a very nice correlation between body mass index and the size of adipocytes in the breast. You might say why is that important? Well, our hypothesis is that it’s the larger adipocytes that are more likely to die than smaller adipocytes, not entirely sure why, it could relate to oxygen tension differences but at any rate we then looked to ask, if the idea was correct that larger adipocytes would be more likely to die leading to an influx of  macrophages and this inflammatory process, then you might predict a relationship between the size of the adipocytes and the severity of breast inflammation, which is in fact what we observed.

We then went on to look at some of the same genes that we had looked at in mice; we again looked at the expression of pro-inflammatory genes, those that are known to regulate aromatase, TNF, IL-1β, COX-2, and were able to demonstrate that each of those was elevated in association with these crown-like structures of the breast, so there was a nice correlation between the severity of the histologic lesion and the elevation of the pro-inflammatory mediators. Again, that translated to increased levels of aromatase, so a nice correlation was observed between the levels of not only the inflammatory lesions but the inflammatory mediators and aromatase. So, taken together, we think that a) this is the first time an inflammatory process has been described in the human breast with obesity or being overweight as the cause. We know that chronic inflammation of most epithelia predisposes to cancer so we could have the same discussion about ulcerative colitis and risk of colon cancer, or H. pylori gastritis and risk of stomach cancer, or reflux Barrett’s and oesophageal cancer, or chronic viral hepatitis and liver cancer, but this is the first demonstration of an inflammatory process in the human breast and it happens to be that overweight and obese is etiologically linked. So the question then becomes how do you harness this information? How do you take this knowledge and how do you affect public health? What are the strategies for moving forth, either from a prevention standpoint, given the epidemic of overweight and obesity that we face in the US but also globally, and are there ways that we can harness this knowledge to alter the natural history of breast cancer in those who are overweight or obese? So one possibility is caloric restriction, and so we’re initiating pre-clinical studies to ask whether or not this inflammatory process can be reversed, if so there may be drugs that are caloric restriction by medics that will be useful, or we may be able to harness our knowledge about the signalling pathways that are aberrant in the obese related to inflammation that we’ve discovered, and to develop targeted therapies with an appropriate safety index that will be useful at large. Nutraceuticals, another possibility, but too soon to say because we don’t have data. But what we do know is that our models in mice predicted for what we have found in women and therefore, although in oncology one commonly hears about the use of animal models of disease, in fact it’s quite rare when somebody has a model in a mouse and actually uses that data to predict what happens in a person with a very tight correlation. So we now have that privilege, we now have that finding and therefore we can go back and utilise the mouse models to try to develop evidence based interventions and, where appropriate, take it forward in people.

Are you investigating this with any other types of cancer?

We are very, very interested in the role of adipose tissue, not only in breast cancer but the role of adipose tissue in cancer biology in general. In my view, oncology will see rapid growth in the area of obesity and cancer, cancer-associated adipocytes and the interplay between cancer cells and adipocyte lineage cells and that’s a subject that we are very, very interested in. I have not discussed that at this particular meeting.