Genetic mutations predict future breast cancer biology

The study challenges the conventional wisdom that most cancers develop as a result of random mutations that accumulate over a lifetime, pointing out instead that genetic sequences inherited from parents (the so-called germline genome) play an active role in determining whether cells with potentially cancer-causing mutations are recognized and eliminated by the immune system or lie dormant unnoticed and develop into early cancers.

“Apart from highly penetrant genes that confer significant cancer risk, the role of genetic factors remains poorly understood, and most malignant tumors are thought to arise from random errors or bad luck during cell division,” said Christina Curtis, PhD, the RZ Kao Professor of Medicine and professor of Genetics and Biomedical Data Science. “This would mean that tumor development is random, but that's not what we're observing. Rather, we find that the path to tumor development is constrained by genetic factors and immunity. These new results reveal a new class of biomarkers for predicting tumor progression and an entirely new way of understanding the origins of breast cancer.”

Curtis is the lead author of the study, which was published May 31. SciencePostdoctoral researcher Dr. Kathleen Houlahan is the lead author of the study.

“In 2015, we hypothesized that some tumors are 'born malignant,' meaning that their grade and potential to metastasize are determined early in the disease's progression,” Curtis said. “Since then, we and others have confirmed this finding in multiple tumors, but our current findings shed entirely new light on just how early this occurs.”

New insight into the origins of cancer

This study provides a nuanced yet powerful new understanding of the interplay between emerging cancer cells and the immune system that may help researchers and clinicians more accurately predict and treat breast cancer.

Currently, only a few high-profile genetic mutations associated with cancer are regularly used to predict cancer, including BRCA1 and BRCA2, which occur in about 1 in 500 women and increase the risk of breast and ovarian cancer, and rare mutations in a gene called TP53 that cause a condition called Li-Fraumeni syndrome that predisposes to childhood and adult-onset tumors.

The findings suggest that dozens to hundreds of additional genetic mutations, also found in healthy people, may play a role in determining why some people remain cancer-free throughout their lives.

“Our findings not only explain which type of breast cancer an individual is likely to develop, but also suggest how aggressive and likely that type of breast cancer is to spread,” Houlahan said. “Furthermore, we anticipate that these genetic variants may influence a person's risk of developing breast cancer.”

The genes we inherit from our parents are known as the germline genome. The germline genome is a mirror of our parents' genetic makeup and varies subtly from person to person, which is why some people have blue eyes, brown hair, and type O blood. Some inherited genes contain mutations that increase cancer risk from the start, such as BRCA1, BRCA2, and TP53. However, identifying other germline genomic mutations that are strongly associated with future cancer has proven difficult.

In contrast, most cancer-associated genes are part of what's called the somatic genome. During our lives, tens of millions of cells divide and die. Every time the DNA in a cell is copied, mistakes can happen and mutations can accumulate. DNA in a tumor is often compared to the germline genome in an individual's blood or normal tissues to identify which changes likely led to the cells becoming cancerous.

Classification of Breast Cancer

In 2012, Curtis began a deep dive into the types of somatic mutations that occur in thousands of cases of breast cancer with the help of machine learning. She was eventually able to classify breast cancer into 11 subtypes with different prognoses and risks of recurrence, and found that four of the 11 groups were significantly more likely to recur even 10 or 20 years after diagnosis. This is important information for clinicians making treatment decisions and discussing long-term prognosis with patients.

Previous studies have shown that people who inherit BRCA1 or BRCA2 gene mutations are more likely to develop a subtype of breast cancer called triple-negative breast cancer. This correlation suggests that something behind the scenes in the germline genome influences which subtype of breast cancer a person will develop.

“We wanted to understand how inherited DNA influences tumor development,” says Houlahan, and to do so, they took a closer look at the immune system.

It's a quirk of biology that even healthy cells regularly decorate their outer membranes with tiny globs of protein that float in the cytoplasm — an outward display that reflects the cell's internal style.

This display is based on so-called HLA proteins, which vary widely from person to person. Like fashion police, immune cells called T cells roam the body looking for suspicious or overly flashy ornaments (called epitopes) that might indicate something is wrong in the cell. Virus-infected cells display pieces of viral proteins; diseased or cancerous cells adorn themselves with abnormal proteins. These imperfections trigger T cells to destroy the culprits.

Houlahan and Curtis decided to focus on cancer genes: normal genes that, when mutated, can free cells from the regulatory pathways that keep them in a normal state. Often, these mutations take the form of multiple copies of the normal gene, lined up head to tail along the DNA, the result of a type of genomic perturbation called amplification. Amplification of specific cancer genes drives different cancer pathways and, in Curtis' original work, was used to distinguish between subtypes of breast cancer.

The importance of decoration

The researchers wondered whether highly conspicuous epitopes might catch T cells' attention more easily than other, more modest displays (think golf-ball-sized dangling turquoise earrings versus simple silver studs). If so, cells that inherited a flashy version of an oncogene might have a harder time successfully amplifying without alerting the immune system than cells with a more modest version of the same gene. (One pair of overly flashy turquoise earrings might be forgivable, but five pairs might cause the patrolling fashionista T cells to switch from clicking to quitting.)

The researchers studied nearly 6,000 breast cancers across a range of disease stages to see whether each tumor subtype correlated with a patient's germline cancer gene sequence. They found that people who inherited cancer genes with a high germline epitope load (i.e., lots of ornamentation) and an HLA type that allows those epitopes to be prominently expressed were significantly less likely to develop breast cancer subtypes in which those cancer genes were amplified.

But there was also a surprise: The researchers found that cancers with a high germline epitope load, which allows them to escape roaming immune cells early in development, tend to be more aggressive and have a worse prognosis than less aggressive cancers.

“In the early, pre-invasive stages, high germline epitope load helps protect against cancer,” Houlahan says, “but once they start fighting the immune system and are forced to develop mechanisms to overcome it, tumors with high germline epitope load become more aggressive and more likely to metastasize. As the tumor progresses, this pattern is reversed.”

“Essentially, there is a tug of war between tumors and immune cells,” Curtis says. “In the pre-invasive setting, nascent tumors may initially be susceptible to immune surveillance and destruction. Indeed, many tumors are likely eliminated this way and remain unnoticed. But the immune system does not always win: some tumor cells are not eliminated, and those that survive develop ways to evade immune recognition and destruction. Our findings may shed light on this opaque process and inform optimal timing of therapeutic interventions, as well as how to heat up immunologically cold tumors to make them more susceptible to treatment.”

The researchers envision a future where germline genomics is used to further classify the 11 breast cancer subtypes identified by Dr. Curtis, to guide treatment decisions and improve prognosis and monitoring of recurrence. The findings of this study may provide further clues in the quest for personalized cancer immunotherapy and may one day enable clinicians to predict cancer risk in healthy people from a simple blood sample.

“We started with a bold hypothesis,” Curtis says, “The field had never thought about tumor origin and evolution in this way. We are looking at other cancers from a new perspective: genetic and acquired factors, and tumor-immune coevolution.”

This research was funded by the National Institutes of Health (grants DP1-CA238296 and U54CA261719), the Canadian Institutes of Health Research, and the Chan Zuckerberg Biohub.