Breastfeeding and breast cancer risk: what are the facts?

A very happy new year to you all! I’ve been writing this blog over Christmas for the new GP Infant Feeding Network’s website, and wanted to share it here too. The GPIFN is a fantastic organisation, founded last year by Dr Louise Santharam, which aims to enhance medical education around breastfeeding issues so that all doctors, and GPs in particular, have more expertise to deal with lactation problems in the community.

One of the reasons why breastfeeding is important for maternal health is in the reduction in risk of breast and ovarian cancer (World Cancer Research Fund, 2014; Lancet breastfeeding series, Jan 2016).

My PhD at Imperial was in breast cancer risk and focused for the last couple of years on looking for biomarkers of risk in breast milk – could we find markers in the genetic material found in cells from breast milk that could confirm the changes we were seeing in blood cells from women at high risk of breast cancer? If we could, then a new tool to enhance screening women could be developed.

Doing a PhD means a lot of reading around the main subject you are working on, and allowing yourself to be led down some interesting paths as a consequence. I rapidly became fascinated with the papers I was reading about the possibility of reducing the risk of cancer by breastfeeding, as it just didn’t match up with my medical training. When I was at medical school, we were taught about breast cancer by breast surgeons, who would write up a list of risk factors on the board – high risks were age, family history, etc. Moderate risk factors included hormonal factors (early menarche, late menstruation, parity, age at first child). Breastfeeding was classed as a low risk modifier – the published estimates back then were a reduction in risk of only about 4-5% for every year a woman breastfeeds over a lifetime (Lancet, 2002).

Biology moves on though, and since 2002 molecular medicine has been able to look at cancers from thousands of different women. These studies have shown that ductal breast carcinomas, the most common form of breast cancer (>95%), cluster into five main biological types: luminal A, luminal B, HER2 positive, basal and the claudin-low/normal type (Perou, 2000; Sorlie, 2003 [figure of dendrogram showing clustering below]).

Many, many more subtypes are and will be discovered as genetic and pathological checks get even better, but these five subtypes are important, as they likely indicate tumours that originate from different cells of origin within the breast ducts (Visvader, 2014; see figure below).

It became apparent last year that these subtypes were also important for the impact of breastfeeding. A meta-analysis of case-control and cohort studies that divided studies of risk according to breast cancer subtype showed no effect of breastfeeding on the hormone receptor subtypes (luminal A, B and HER2), which make up 70-75% of breast cancers (Islami, 2015). However, it did show a 16-24% risk reduction for triple negative breast cancers (TNBC, tumour cells do not express oestrogen and progesterone receptors or HER2/ERBB2 receptors; shown in the Forest plot below reproduced from Islami’s paper).

This systematic review looked at ‘ever breastfeeding’, rather than the duration of breastfeeding, as there were not enough studies that had recorded that to be analysed. Given that most studies have shown breastfeeding to have a dosage effect (the longer a woman breastfeeds, the lower their breast cancer risk), future studies that look specifically at these tumours may show a greater effect with prolonged breastfeeding duration.

TNBCs are one of the most aggressive forms of breast cancer. These tumours tend to occur in younger women and those with genetic mutations. Pregnancy-associated breast cancers also tend to be TNBCs. These are cancers that are diagnosed during pregnancy and, although the cut off is unclear, for up to 5 years after birth.

The mechanism, or mechanisms, that explain how breastfeeding reduces TNBC risk are not clear, but are the subject for a future blog and a great deal of future research, some of which we are hoping to facilitate in the future at the Hearts Milk Bank. However, a small number of women (approximately 100-120 each year in the UK) who breastfeed will still develop aggressive breast cancers. As with everything in medicine and public health, breastfeeding protects some, but in others the mechanisms will a) be inadequate and the cancer would develop anyway, or b) the molecular and cellular mechanisms that occur in preparing for and performing lactation may trigger some women with a specific genetic make up to develop a cancer.


So, is breastfeeding protective from breast cancer? For the majority of tumours, perhaps not, but more research is needed to look at the effect of long-term breastfeeding on the other subtypes of breast cancer. However, for the 20-30% of TNBCs with the worst prognosis, which affect premenopausal women, breastfeeding is probably a very powerful way to reduce risk. At the Hearts Milk Bank, we will be working with scientists across a range of fields to facilitate research that can determine how this effect happens.