AACR 2016

Genetics may influence risk for breast cancer after chest radiotherapy to treat childhood cancer

Dr Lindsay Morton - National Cancer Institute, Bethesda, USA

I’d like to start by giving you a sense of the population of cancer survivors in the United States and why there are important both clinical and public health reasons to understand the long-term health of this population. With advances in treatment and early detection people are living longer after a cancer diagnosis today than ever before. Because of that, the population of cancer survivors has more than quadrupled from the late 1970s to reach nearly fifteen million individuals today. That number is projected to reach eighteen million individuals by the year 2022. About one in five cancers diagnosed today occurs in one of these cancer survivors and these so-called second cancers are a major cause of morbidity and mortality. So focussing in particular on childhood cancer, there are over 400,000 childhood cancer survivors in the United States and it’s well established that childhood cancer survivors, particularly those treated with chest radiotherapy, have increased risk of subsequently developing breast cancer which is the focus of the current study.

So in the next slide I’m going to give you a feel for the magnitude of this risk with data from a recent study that shows the cumulative risk of breast cancer by age in different populations. Starting with the bottom line, the light blue line, this is the risk in the general population. The grey and the red lines represent the risk of breast cancer in BRCA1 and 2 mutation carriers, so we would consider this to be a high risk population. Then the blue and yellow lines represent the risk among survivors of Hodgkin lymphoma or survivors of other childhood cancers who were treated with chest radiotherapy. So this represents a very high risk population but it’s not known whether some of these individuals might be genetically susceptible to developing breast cancer after chest radiotherapy for childhood cancer.

We therefore formed a partnership, it was led by the National Cancer Institute and St Jude Children’s Research Hospital, to launch a study to investigate this particular question. The population derives from two cohorts of childhood cancer survivors who were followed by investigators at St Jude and we have the initiating PI of one of these studies, the Childhood Cancer Survivors Study, and the co-PI of the St Jude lifetime cohort here in the audience, Dr Les Robison.  So the Childhood Cancer Survivors Study includes children diagnosed from 1970 to 1986 from 26 participating institutions and the St Jude lifetime cohort includes children diagnosed from 1963 to 2005, all of whom were treated at St Jude. These cohorts provide a very rare combination of resources: they have long term follow up for the occurrence of second cancers, they collected germline DNA and they looked at each of the survivors’ medical records in order to abstract detailed treatment data from their childhood cancers.

So in this analysis we included over 3,400 childhood cancer survivors combined from these two cohorts, of whom 207 subsequently developed breast cancer. To give you a sense of this population about two-thirds of the population that had breast cancer had first primary Hodgkin lymphoma whereas the remaining third had a mixture of other childhood cancer types. What we did was we also had a team of medical physicists at MD Anderson that reconstructed the radiation dose to the breast and then we divided our cohort into two groups, based on the level of this radiation exposure. About two-thirds of the cases had at least 10Gy of radiation to the breast during the course of their childhood cancer treatments. The median age at breast cancer in the population was 39 years old, so these are breast cancer cases that are occurring at a much younger age than you typically see in the general population.

I’m going to present today the results for two of our top results. As I mentioned, we divided our population into whether or not they had received above or below 10Gy of radiation to the breast. The first association that I’m going to present represents a marker on chromosome 1q41. This actually skipped a slide so I’ll just go ahead and say verbally. We conducted a genome-wide association study and so what this does is it uses a marker based on known genetic variation and we looked at nearly 17 million different positions in the genome to understand the genetic variation at those positions and to see whether or not the genotypes at each of those positions was associated with breast cancer risk.

So our top association was for a marker on chromosome 1q41 and the p-values that I show here demonstrate that there is a very significant difference in the genotype distribution at this particular location among the childhood cancer survivors who developed breast cancer compared with those who did not. The relative risk for developing breast cancer was nearly twofold for each copy of the risk allele and we each carry two risk alleles. Interestingly, though, this association, as highlighted in red, was restricted to the children who had received chest radiotherapy. In other words, there was no effect of this particular genotype on breast cancer risk among the childhood cancer survivors who had not received chest radiotherapy.

We found similar results for another marker on chromosome 11. Unlike the chromosome 1 SNP which was a relatively common SNP, the chromosome 11 SNP was a relatively rare marker where the genetic variant that conferred risk was present in only 2-3% of controls. Here we saw about a 2.6-fold increased risk of breast cancer, again restricted to the childhood cancer survivors who had received at least 10Gy of radiation to the breast.

So what are these SNPs? Based on previous literature there is evidence for both of these SNPs as being plausible breast cancer susceptibility loci. So the chromosome 1 marker maps near a gene called PROX1 which is a transcription factor involved in early embryonic development, cellular proliferation and migration. Very importantly the Cancer Genome Atlas data show us that PROX1 is altered most frequently in breast tumours and alterations are present in about 13% of breast tumour cells. The chromosome 11 marker maps near a gene called transgelin or TAGLN which is also a gene that is associated with cellular migration and previous studies have shown that it has been overexpressed in breast tumours compared to normal breast tissue.

So based on the previous data combined with our study results it leads to this intriguing hypothesis that the germline variants that we’ve identified could create a pro-proliferative, pro-invasive phenotype that supports the growth of malignant breast cells but only following the transformation of those cells by the ionising radiation exposure.  However, it’s really important to remember that this is a discovery study and we used SNP markers so the exact functional variant is not known and much work is needed to be done in the laboratory in order to understand specifically whether these genes are the ones that are involved in the subsequent breast cancer risk or, in fact, these markers are tagging a different effect within the genome that don’t even involve these genes.

So what does the study do? It’s the first study that provides evidence that germline genetics outside of the context of high risk syndromes can modify the risk for breast cancer in childhood cancer survivors treated with chest radiotherapy.

So what does the road ahead look like? As I’ve mentioned, this is a discovery study and I think of the road ahead as being long, winding with a little bit of an uphill climb. In addition to the laboratory work that I’ve already mentioned, our results require replication before they can be evaluated in the clinic. There are a number of important unanswered questions. One is the relationship between the genotype and the risk with radiation exposures typical of patients who are treated today. Patients treated today often receive lower treatment doses and lower treatment volumes than those who were treated in the past who are part of these two childhood cancer survivor cohorts. We also have to ask whether or not there’s an effect of the age at treatment and we could not answer that conclusively with the data from these two studies alone. We also need to ask whether or not there could be an effect that is perhaps stronger or weaker for a particular breast cancer molecular subtype, given what data in the general population have taught us about different genetic etiologies of different breast cancer molecular subtypes.

So once all of this work is done, the laboratory work and the additional replication, we think that these studies have tremendous potential for the future and we’re really excited about it. There is the possibility that understanding genetic susceptibility such as this could impact front line therapy decisions. It may alter the risk and benefit calculation for receipt of radiotherapy for a childhood cancer patient who has been diagnosed today. And perhaps even more importantly for the many survivors who are already out there these data could impact the screening recommendations. Today the Children’s Oncology Group recommends annual dual modality screening for childhood cancer survivors who were treated with chest radiotherapy beginning at age 25 or eight years after treatment, whichever one is later. If germline susceptibility confers different levels of risk then perhaps we can use this inherited susceptibility to identify those who would benefit the most from the screening.