This video features Susan M. Domchek, MD presenting Genetics and Breast Cancer Risk. This presentation was given at our January 19th 21st Annual Highlights of the 2022 San Antonio Breast Cancer Symposium in 2023.
Hello. I'm dr Susan down check from the University of pennsylvania. I'd like to thank the organizers for the opportunity to provide an update on genetics from the CN Antonio breast cancer symposium I'm going to be presenting the following abstract which range from population estimates of control or breast cancer related to pathogenic variants. Radio Mick phenotype sis in mammography. The baby tam follow up study and the positive study looking at pregnancy outcomes after interruption of endocrine therapy. The carrier study is a population-based piece control trial. We were lucky enough at penn to be part of this. This included 32,000 women with breast cancer and 32,000 controls and sequencing performed to evaluate the association between pathogenic variants and various genes with breast cancer risk. This provided the following outcome a odds ratio with a. T. M. Pathogenic variants of 1.87 point six for BRC 15.2 BRC 22.5 for check. Two and 3.8 from county to these results were published in the new England Journal of Medicine in 2021. Here we use these cohorts to investigate contra lateral breast cancer risk in the same study population. This involved 15,000 women with unilateral invasive breast cancer. From 10 prospective epidemiologic studies in the United States. The inclusion was preserved contra lateral breast and at least one year follow up and excluded D. C. I asset initial diagnosis and looked at five specific genes. The analysis was a time to event analysis. Looking at contra lateral breast cancer risk between carriers and non carriers in each gene, a multi variant proportional hazards. Regression analysis accounted for the competing risk of death censoring at last fall or contra lateral mastectomy. And adjusted for those items listed here. The patient characteristics, the median age is 62 which is very similar to the general population with a median age. Uh About 63% were non hispanic white, 70 approximately 75% postmenopausal. And your deposit most doctoral histology and about half had used endocrine therapy. Here's the control auto breast cancer risk data. These were just published in Jae SEO last week. The risk of control out of breast cancer with A. T. M. Mutation carriers was not statistically significant. With the house ratio of 1.2. The risk was elevated in both the RC one and PRC two cares with hazard ratios approximately three for check to hazard ratio of two and for probably 21.3. These are the cumulative risk estimates again for A. T. M. We don't see an elevation of cumulative risk of breast cancer. Whereas we do for BRC one whereby 20 years the cumulative risk of breast cancer approaches 35% and for BRC two approaches 30%. For check to the 20 year cumulative risk is approximately 15%. Both for all comers and for check to 1100 L. C. Please note the common little penetrates. Check to findings of I 1 50 70 and S for 20 af were excluded from this analysis. I'll discuss that more but these are clearly pathogenic not low penetrates. Check to mutations help me to the numbers were quite small and here's the cumulative risk by estrogen receptor status. Both five year and 10 year accumulative risk. I'm going to focus on A. T. M. First which as you can see for er positive disease. The five year cumulative risk estimates was 2% for both. And for the 10% cumulative risk estimates for approximately 4% for both for BRC one BRC to the er positive risks. Um uh or elevate year. It's that there's the there's an elevated risk of er positive er negative disease. And you can see particularly for BRC one the risk of er negative disease by 10 years is as high as 30%. Check to the risk. The five year risk for cumulative risk for er positive disease is 4.5% and for 10 years 8.2%. So to compare that the 10 year cumulative risk estimate of a control our breast cancer. 3.8 for non carers 8.2% for check to carers. So while clearly elevated this is a relatively modest risk that needs to be counseled appropriately. We know that age gene and phenotype matter. So within beer say one in two for instance. And here's data for BRC. Two from this. A study from embrace in Jama in 2017. If you're a BRC two carrier who had a first breast cancer less than age 40 your residual lifetime risk up until you're 80 is as high as 68%. But if your first breast cancer was above age 50, really the highest risk you get is about 20%. So it does matter and it needs to be counseled separately. How old is the patient? Do they have an era positive or negative disease? And what is the gene that they have? A pathogenic variant? So pathogenic variants and a team do not appear to be associated with an increased risk contra lateral breast cancer. In the general population, family history does matter. A. T. M. Is really considered as only one piece of the puzzle if you will of breast cancer risk. So in individuals who have very strong family histories that should be taken into account. But increasingly we're seeing individuals who are tested but no family history who have an A. T. M. Mutation at the age of 55 those individuals do not necessarily there there is no particular need for them to have a control l mastectomy and the usual shared decision making and risk benefit and conversations apply. It is extremely unlikely that the common check to will penetrate experience. I won 50 70 and S. 4 28 F. Have a significantly elevated control or breast cancer risk. As the risk of developing a first breast cancer is only 1.5 we need to factor all this into discussions regarding contra lateral mastectomy and it's a reminder that even beer C one and two mutation carriers are not required to undergo contra lateral mastectomy and age. And your status mattered here as well. In the second abstract is looking at radio megafauna types of breast texture and associated with both breast cancer risk and the risk of if you will a false negative. We all know that there's variability by reds density categories ranging from almost entirely fatty scattered fiber, glandular heterogeneous lee, dense and extremely dense. And these different features both in prank mall features and breast density independently predict breast cancer risk and can be associated with false negatives. So in this study they looked at various characteristics of related to the the mammograms. They looked at 30,000 women including three breast cancer screening cohorts including Penn. Um in my colleague Marie McCarthy and Hispanic contests and looked at the association with breast cancer risk. And there was a discovery set and a validation set. So, you know, there's lots of things. They looked at grade level hissed a g and co occurrence run length said as if I know exactly what these mean. But they did both an unser supervised learning with higher clustering and principal component analysis. And then once they did this training set they went on to the testing set of 10,000 patients stratified by race and density. In this principal component analysis, they identified six principal components that explain to 85% of the variability and they were able to then look at that in the training set and take it to the testing set. And what you end up with at the end of the day is these R. O. C. Curves and these include in red age plus B. M. I plus bi rads density and then in blue H plus B. M. I, bi rads density and these principal components. And what you can see is that overall When you add it, it helps a little bit in terms of overall breast cancer risk and a little bit in white individuals. But there seemed to be a bigger impact in black patients. Although please note that these are overlapping numbers. So the confidence intervals overlap with each other. So there does appear to be some mild benefit here. And when you look at sort of false negatives or symptomatic interval cancers. Likewise, there seems to be an improvement and particularly for these symptomatic interval of cancers adding these principal component analysis seems to be helpful. So this adds to growing data that radiographic features. Feel intensity may add to risk assessment. The numbers are small, there's a modest improvement and discrimination and there is this reminder that there's a difference between discrimination and calibration, discrimination separates into two classes but calibration says how many patients get the the if you will cancer that you thought were going to get the cancer. So whether they agree with the different proportions and I just wanna you know point out there's multiple papers emery McCarthy has been looking at apologetic risk scores and combining these different features. There's a I. Based strategies that regina parsley at M. I. T. Has been using and uh and other risk models that incorporate many different features so more to come in this area nothing actionable at this time. The next abstract is by andre and decency and this is a 10 year update of their interesting phase three trial of low dose tamoxifen and non invasive breast cancer. Lovingly called Baby Damn. The study design looked at women under 75 with high risk lesions either A. D. H. L. C. I. S. Or er positive or unknown D. C. I. S. Randomized to tamoxifen five mg a day versus placebo treated for three years with seven years of follow up 500 patients in world from 14 centers in Italy with the primary endpoint of invasive breast cancer or D. C. I. S. Mhm. They are the authors presented their initial work at san Antonio and published it in G. C. O. In 2019 demonstrating a significant improvement in the development of breast cancer events. Either outcome or control lateral with hazard ratios of 20190.5 and 0.24 accordingly with the use of Baby Damn. After san Antonio asked the U. S. Service Representative task force included baby Tm as an option and the N. C. C. N. Recommends BB. Tm. After D. C. I. S. If patient is symptomatic or unwilling or unable to take the full dose. And this and several studies have shown that baby Tm. Is actually the most popular choice in movement with fibrous lesions in the US. With lower discontinuation rates in one year versus 20 million. So tamoxifen or relaxation or ai. So here's their updated results. This is the end of treatment in three years. Here's the follow up and see the Kurds remain separated. The hazard ratio .48. This remains statistically significant both for its bilateral and contra lateral breast events. So here's the consideration. So baby tam at five mg daily decreases breast cancer risk. Long term follow up reminder that there is no direct comparison to to 20 mg. So this is not five mg versus 20 mg. Five million. five mg is not a commercially available dose. So you can use 10 mg every other day. Um Other studies have shown that tamoxifen at 2.5 mg decreases nam a graphic breast density as much as 20 mg. This is published in J. C. Last year But we don't know if that translates to a breast cancer risk. So again no direct comparison but nobody takes 20 mg of tamoxifen prevention or at least I'm not very successful in getting people to do it. So the real question is should we really start at 20 and then these ask would more patients. Uh more women be willing to take it if we offered it from the get go at five mg. And I will admit that this this presentation made me think to myself, I really should be doing more of that. I should be offering it at five mg. Given the data that is uh it's well tolerated and patients preferring. And finally we come to the positive trial which is this pregnancy outcome and safety of interrupting therapy for our women with endocrine response to breast cancer presented by Ian Partridge really kind of a tour de force study um getting out a very practical and important which is many of our breast cancer survivors desire pregnancy, retrospective evidence suggests pregnancy after breast cancer does not worsen disease outcome regardless of former receptor status. But I think we all recognize that the risk the lion's stopping internet therapy early and standard 5 to 10 years of endocrine therapy complicates pregnancies because you know, people are you know, this is a time when their fertility may be declining. So there's a lot of issues that are uh you know, the confounding this. So the goal of positive was prospectively studied pregnancy after breast cancer when after an interruption of endocrine therapy, Women had to be premenopausal wanting to get pregnant unless we're able to 42 years at least 18 months and no more than 30 months prior endocrine therapy for stage one through three breast cancer prior chemotherapy was allowed fertility preservation was allowed no clinical evidence of recurrence with the primary outcome of breast cancer free interval, which was time from enrollment to the breast cancer event. With secondary outcomes of pregnancy. Here are the trial procedures. Once they enrolled, they had a planned undergoing therapy interruption within one month up to two years to attempt pregnancy, conceived deliver breastfeed including a three month washout period. And if no pregnancy by a year, fertility assessment was recommended. Honestly, I usually do that earlier. An endocrine therapy was strongly recommended after pregnancy to complete 5 to 10 years of therapy with this long term follow up. So here's the study schema. 500 women plan simple size um Uh and again, primary analysis acker 1600 patient years of follow up a lesson or equal to 46 breast cancer events considered safe. Uh and different various interim analyses which is you know that they would stop at the annual breast cancer rate was 4%. Um for instance, they used a cohort of about 1500 soft text patients as an external control. Using this bootstrap matching method to compare the positive patients to the soft tech uh soft and tax patients. Again, 116 centre 20 countries, four continents. Um 516 patients for the primary advocacy analysis 497 for the outcomes of pregnancy. And in terms of patient characteristics, 20 about 20% were over 40 the others I've seen here, 75%. No prior births, about half stage one half stage to um 42% terms alone, 36% with O. F. S. And 62% prior chemotherapy, 45% with prior mastectomy. You can see at 48 months 8.9% with an event. And here is the stopping rule at 15%. And in terms of distant were at 4.5%. Um with this would be 22 events here if accordingly um overlaying the soft and text that's the positive is in solid, the soft and text is in dotted and you can see that there's no difference. And if anything soft and text, the curves are separating for soft and text to look a little worse. But of course we don't want to read too much into this at this time. When you look at the three year breast cancer, you know, cumulative incidents, it did vary according to pathologic features. Not surprisingly, those with a higher risk. In terms of nodal status, tumor size or tumor grade um had a little bit more had a higher cumulative risk. And when you looked at those with who were not pregnant versus those who were pregnant. Um There there really were no differences in terms of the risks. Um In a time dependent cox model analysis, Pregnancy outcomes, 74% of the patients did have at least one pregnancy, 70% within two years. Um 64% of all women had had one, at least one live birth, which was 86% of those who became pregnant. Um there was, you know, 11 stillbirth and at least one elective abortion. There were 16. So this just gives you an idea, 34% needed, went on to C sections 11% had pregnancy complications most commonly hyper Clancy, hypertension, preeclampsia and diabetes. Um So the conclusion here at the temporary interruption of endocrine therapy to attempt pregnancy. Among women who desire pregnancy did not seem to affect short term disease outcome will have to see about longer term outcome, although again the occurs sort of potentially worse for soft, but you know, we all know that we didn't put, you know, we didn't refer just anyone to positive. People would tend positive, knows, I'm not sure that we can feel sort of okay sending them to to interrupting their endocrine therapy early because I think that again is really where the risk is. So this this additional information to add to the already challenging and nuanced conversations and again, really has to matter about the absolute risk for the patient and their their overall feeling about risk avoidance and their desire for pregnancy. So with that I thank you for your attention and I would be happy to take questions
