The Broken Genetic Landscape of BRCA Tumors

By Avantika Lal, PhD

The BRCA Journal is pleased to announce the launch of its "Innovative Guest" series. We will feature stories from our Young Investigator Awardees, genetic counselors, and leaders in the field of BRCA research and outreach. Our awardees will discuss their very own cutting-edge research that aims to prevent, diagnose, and cure BRCA cancers as early as possible. Other experts will tackle topics related to the implications and advances of genetic counselling and the importance of epidemiology in big data. You will read their original views and ideas here first, straight from the source.

Precision medicine is an often-used term in cancer. Every cancer is genetically unique, and one of the goals of precision medicine is to develop a molecular understanding of individual tumors. This is done by sequencing a patient's tumor and using its individual genetic sequence to identify the tumor's causes and, consequently, choose the best treatment options.

Another aim of precision medicine is to sequence the DNA of healthy people, and identify mutations— such as those in the BRCA1 and BRCA2 genes— which may place carriers at a high risk of developing cancer in the future. A more recent, but increasingly realistic, vision is to sequence DNA from people's blood in order to detect signs of a growing tumor at a relatively early stage when it can still be easily treated.

While these are easily stated goals, the complexity of precision medicine, however, is often understated. Cancer is caused by changes in the activity of genes. Consider a gene that controls cell growth; if it's more active than it should be, you get uncontrolled growth, causing the cell to develop into a tumor.

That said, there are many, many ways in which genes can be made more or less active and ultimately lead to cancer. Just to list a few: mutations in the gene sequence; deletions and duplications of pieces of DNA; rearrangement of the genome by moving pieces of DNA into different positions; and 'epigenetic' changes that control the activity of genes without changing the DNA sequence. Moreover, a large part of the human genome does not contain genes; nevertheless, changes in this 'non-coding' genome affect cells in ways that are difficult to pin down.

A typical tumor has many events of all these types— and then some. So if we examine any one aspect of a tumor in isolation— say, mutations in genes— we put ourselves in the position of the blind man touching an elephant's tail and concluding elephants must be like snakes. In other words, we fail to observe the bulk of changes affecting the tumor.

A complete picture of a tumor therefore requires a scientist to carry out not one, but several different experiments, resulting in many types of genetic data. Collecting data like this is a huge endeavor. In recent years, the Cancer Genome Atlas has made huge strides by collecting such data for over 7000 tumors. Once collected, however, integrating and analyzing such large, multidimensional datasets remains a computational challenge.

Over the last year, I have been working on methods to integrate multidimensional tumor data. In a recent paper, my co-authors and I combined data on DNA copy number, mutation, gene expression and DNA methylation from several hundred tumors and used this information to separate breast cancer into 13 molecular subtypes.

Unfortunately, the publicly available dataset we used has very few BRCA mutation carriers. In fact, there is a general dearth of multidimensional data for tumors from BRCA carriers. This is a great pity, because several lines of evidence indicate the tumors these patients develop are genetically different from other kinds of breast tumors. This is not really surprising— after all, tumors in BRCA carriers, unlike other breast tumors, develop because of the BRCA mutation. It's possible their pathway of evolution from a normal cell to a cancer is very different.

In the past, analyses of DNA sequences from breast tumors have shown that tumors from BRCA mutation carriers have more deletions and duplications of their DNA than other kinds of breast cancer. Some studies have also suggested these tumors have a characteristic pattern of point mutations (single-base changes in DNA). And finally, a few, very interesting, recent studies indicate that BRCA1 and BRCA2 tumors are genetically quite different from each other as well as from normal breast tumors—tumors, from BRCA1 carriers have many more duplications of large segments of DNA, scattered throughout the genome.

So far, we have a picture of a much more damaged genome in tumors from BRCA mutation carriers, broken at many points and then stitched together— with mistakes in the stitching. It seems likely there may also be differences in other kinds of genomic alterations, including epigenetic changes such as DNA methylation, but this remains to be seen.

Why does this matter for patients? First, understanding the biological pathways through which cancers evolve in BRCA carriers can help us develop treatments and perhaps even prevention for these cancers. Second, it helps us ensure these high-risk patients receive the best possible monitoring. If the genetic changes that cause cancer are different in BRCA mutation carriers, this suggests we need to develop specialized monitoring tests for these carriers, as a means of recognizing these changes and catching cancer earlier.

My current work, funded by the BRCA Foundation, involves sequencing tumors from BRCA mutation carriers and comparing them to public datasets on breast cancer. I examine data on gene expression, epigenetic changes, mutation and DNA rearrangements, and combine all this information in order to build an unprecedented picture of how BRCA tumors are unique at the genetic level.

The truth remains, however, that even this will just scratch the surface of the complex layers of molecular interactions that define these cancers. A tumor isn't simply a uniform mass of cells; different parts of a tumor can actually have very different genetic features. In the future, a deep, personalized understanding of cancer may focus on understanding different parts of the tumor, how they interact and compete with each other, and with surrounding tissues and the immune system. This precise and detailed approach based on differing individual information is what defines precision medicine.