The main aim of our study was to assess the underlying genetic predisposition in terms of germline pathogenic variants in genes predisposing to breast cancer in women with very early onset breast cancer, that is under the age of 31, so 30 and under.
What was the methodology used in the study?
What we did initially was to screen all 380 women with genetic testing initially for BRCA1 and BRCA2, TP53 and a common genetic variant in the moderate risk gene CHEK2. Then we went on in those testing negative, where we had sufficient DNA, we carried out testing of a panel of at least ten additional genes known to link to breast cancer risk.
What was the likelihood of actionable pathogenic variant detection in breast cancer genes?
Essentially the detection rate was around 20% for BRCA1, 9% for BRCA2 and 6% for TP53. So adding those together you’ve got about a 35% probability of a fault in one of the three important breast cancer genes. Adding in the remaining ten genes only really added in about an extra 5% probability. So it didn’t really substantially increase, from basically 35% up to about 40% by adding in those additional genes.
What were the key study findings?
A number of key results. One is that BRCA1 and BRCA2, the frequency of genetic changes generally gets higher the younger you get. But we’ve noticed this previously in ovarian cancer, that once you get down below a certain age actually the frequency drops off very dramatically. For ovarian cancer that comes at around 30 years of age so it’s very rare to find pathogenic variants in those genes as a cause of ovarian cancer under 30. The same thing seems to be happening with BRCA1 and 2 is that once you get below 26, in other words 25 and under, the likelihood drops considerably. So, for instance, we found a frequency of over 30% in women aged 26 to 30 but that dropped down to actually only about 13-14% when we went below 25 and below. That’s in contrast to p53 where the frequency of mutations actually increased in that very young age group. So that was the first finding.
The second is that there is this very low additional detection rate from screening an extended gene panel. Given that low detection rate there would be concern if, for instance, you reduced the sensitivity of the analysis of BRCA1, BRCA2 and p53 by putting lots of genes into a panel. There is definitely some reduction in sensitivity the more genes you put into these panels. In particular, whether the tests on copy number variations are accurate, given that about 20% in the UK of BRCA1 genetic faults are copy number variants.
Then there’s also a link with the pathology. So we already know that triple negative breast cancer is strongly linked to BRCA1 and we again showed that with about 40% of triple negative breast cancer having BRCA1 gene faults. And also HER2 positivity, we confirmed that those were more likely to have TP53 germline variants. But what isn’t well described in the literature is that those with precancerous breast cancer, DCIS in this age group, particularly those with something called comedo-DCIS, had a very high rate of TP53. So that’s, again, a strong correlation.
So I would say those were the three main findings of the study.
What are the future implications of the results obtained from this study?
I think the main finding of this study is that it should be near compulsory if you’re doing genetic germline testing to test for faults in BRCA1, BRCA2 and TP53 in anyone diagnosed aged 30 and under. Particularly p53 should definitely be included if there is HER2 positive breast cancer or DCIS.
There would be a case, given that in the UK we currently have a panel of BRCA1, BRCA2 and PALB2, to add p53 into that panel in that age group. But what we would probably say is that one would need to be careful about testing women with breast cancer 30 and under with an extended panel if there was any loss of sensitivity whatsoever in detection of mutations in BRCA1 and BRCA2 and p53. In particular, that we would advise that obviously testing, the more genes you test the more the likelihood is that you will create uncertainty, finding a variant of uncertain significance, and that those women could be disadvantaged in ongoing follow-up because they may be worried about a risk that doesn’t really exist. Because the great majority of variants of uncertain significance turn out to be harmless.