When it comes to colorectal cancer, we’re all at risk at least a little, says Ulrike “Riki” Peters, PhD, MPH, a molecular and genetic epidemiologist with Fred Hutchinson Cancer Center.
“Many people have average risk, some people have much lower risk and some are at really high risk. But people generally wouldn’t know if they’re at high risk,” said Peters, who has focused on both the genetic variants and environmental risk factors that can drive these cancers for over a decade.
Peters and her team are trying to help those people — and their health care providers — better understand germline genetic risk, the kind you inherit from your parents, by identifying all the gene variants that can drive colorectal cancer.
The investigators’ immediate goal is to use this information to create tests based on germline DNA that help identify who’s at high risk and who’s not.
The team’s future goal: fold together germline genetic risk data with information about the genetic damage wrought by environmental exposures — think pollution, processed meat, smoking — and create polygenic, or multi-gene, risk scores that assess a person’s colorectal cancer risk with much more precision.
With the publication of a new paper today in Nature Genetics, they’ve gotten much closer to that first mark.
Peters and her team at Fred Hutch — including researchers Li Hsu, PhD, Minta Thomas, PhD, Jeroen Huyghe, PhD and Tabitha Harrison, MPH — along with over 200 other scientific collaborators from around the world, announced their discovery of another 100 genetic risk variants for colorectal cancer.
“This is the largest and most comprehensive investigation of common genetic risk factors for colorectal cancer,” Peters said. “We’re excited about the large amount of discovery we’ve made. Previously, we had about 140 variants. Now we’re adding another 100.”
Nearly two million people are diagnosed with colorectal cancer each year worldwide, and it’s currently the world’s second most deadly cancer, especially for African American and Alaska Native people.
Actor Chadwick Boseman died of this cancer in 2020 at the age of 43, two years shy of the American Cancer Society’s newly revised recommended age for average-risk people to begin screening. Most national guidelines boards followed suit in lowering the screening age.
What’s average risk? It means you have no family history of colorectal cancer nor any personal history of inflammatory bowel disease or ulcerative colitis or Crohn’s disease, among other risk factors.
Those with higher risk — people who do have a family history making them more susceptible — may need to get screened before age 45. However, over 80% of all colorectal cancer cases do not fall in this higher-risk group.
Estimates vary, but scientists believe about a third of colorectal cancers are driven by inherited genetic variants. Peters and her team are working to create tools, like a simple blood test, that can help clinicians identify additional people at high risk who can get earlier screening.
But first they need to identify all the genetic variants that can lead to a colorectal cancer.
To do that, the scientists first used genome-wide association studies, or GWAS, which has been used for years to identify cancer risk genes. GWAS have already netted 140 independent risk variants for colorectal study.
But the scientists didn’t just use this tried-and-true method to hunt down colorectal cancer genes.
“We’re trying to understand the mechanism,” said Hsu, lead biostatistician on this and a slew of other colorectal cancer projects in collaboration within the Peters Studies lab. “We start from the DNA and ask, how does genetic variance associated with colorectal cancer risk come about? Through what particular genes? What gene expression? What particular methylation? We’re also looking at biological pathways or mechanisms.”
For their new comprehensive analysis, the scientists brought together almost all currently available GWAS data, then used new statistical methods developed by Hsu to fold in studies that look at the associations between health and the way genetic instructions are “read,” or expressed, in cells. These are called transcriptome- and methylome-wide association studies.
The transcriptome looks at RNA transcripts, or the DNA’s various marching orders, and is the first step in gene expression. The methylome recognizes genome-wide DNA methylation patterns; methylation refers to small molecules called methyl groups that are attached to DNA in a way that regulates patterns of gene expression and can lead to disease.
By amassing all of this information, Peters, Hsu and others were able to investigate the genes and the mechanisms underlying both established and new colorectal cancer risk loci (loci are the specific physical locations of a gene or variant, like the street address on the chromosome).
“To go from DNA to disease risk — from germline DNA to disease — is a complicated biological process,” Hsu said. “What sets this study apart is that we’re looking at the intermediate states — gene expression or methylation — trying to understand how the genetic variants impact the gene expression and further impact the disease risk.”
The investigators also identified credible effector genes — genes that activate a cellular process — and the tissues in which they act, furthering science’s understanding of how colorectal cancer gets started in the first place.
Thomas is a research associate tasked with creating colorectal cancer risk prediction models using machine learning and statistical approaches. She said incorporating the different genomic approaches made for a “very good study.”
“If you look at the results, we identified many genes functionally leading to colorectal cancer,” she said. “It offered us a biological glimpse inside these cancers.”
Peters, who holds the Fred Hutch 40th Anniversary Endowed Chair, looks forward to using the information to cut back on cancer cases.
“Most people don’t have a positive family history of colorectal cancer,” she said. “But this allows us to identify those at particularly high risk, who are not aware of their risk, and who could benefit from earlier, more frequent screenings or from dietary interventions. Some of these are modifiable risk factors.”
— Fred Hutch research associate Dr. Minta Thomas
Finding these genetic variants is exacting work. But the more data researchers have (in this case, people’s DNA and tissue samples), the better they can distinguish between actual risk — even if it’s quite small for any one of the genetic variants — and what scientists call “noise”: background information that’s unrelated to what they’re investigating.
In this study, researchers pulled together data on a whopping 100,204 colorectal cancer cases and 154,587 healthy controls. People of European descent made up 73% of the cases; 27% were of Asian origin.
“It’s really a numbers game,” Peters said. “When you look at genetic risk factors, there are many, many risk factors that are contributing. But there are millions and millions of variants in the genome that we have to weed through to understand which one is significantly associated with this cancer.”
The problem, Peters said, is many of the variants they find have weak effects.
“They’re not easily found,” she said. “The signal-to-noise ratio is not in our favor. That’s why we need such large sample sizes. We’re buffering against false positives by being stringent with our threshold for true risk.”
And though they’ve essentially brought the world’s resources together — “there’s almost no more data,” Peters said — they still haven’t identified all the genetic risk factors for this cancer.
“There are still more to be discovered,” she said. “But building large, good data sets cost time and money.”
It also requires equity.
Current risk scores are problematic because they’re highly Eurocentric; the vast amount of DNA used to map the human genome has primarily come from white people. As a result, most risk scores are accurate for that particular population but not so much for others.
But Peters and her colorectal cancer collaborators have been diligently gathering large pools of DNA from different ethnic/minority groups to identify as many genetic risk variants as possible. The scientists are currently analyzing genetic risk by comparing DNA from two racial and ethnic groups: African American and Hispanic individuals with and without colorectal cancer.
“We were close to bringing that in for this paper, but the sample size for African American and Hispanic people was too small and that’s a problem for us,” Peters said. “At this point, the risks we’ve identified are applicable to people of Asian and European descent.”
And that’s not good enough, Peters stressed.
“We have ongoing work to fix it so the risks we’ve identified apply to everybody,” she said. “We want to have precision medicine tools that are not increasing health disparities and inequality and we’re not there yet. We need to specifically focus resources to improve that effort.”
Thomas said universally accurate polygenic risk scores will eventually be created from the Hutch research. But they’re not ready yet.
“This particular study was more about discovery to identify the colorectal cancer loci or the genes associated with it,” she said.
And it’s not just about getting high-risk people in for early screening.
As Peters mentioned, these new findings may also provide opportunities for interventions that could potentially reduce a person’s risk of colorectal cancer.
“These genetic effector genes we’re identifying can really link to potential targets for chemoprevention or for targets for treatment not currently being used,” Peters said. “We think it’s a missed opportunity.”
Chemoprevention, or using drugs or chemicals to fend off disease, has been used in other diseases to great effect, she said.
“Much has happened in the chemoprevention of stroke,” she said. “They’ve identified genetic risk factors and they have drugs that can target the genes. Statins are heavily used and they are very good at preventing people from getting major cardiovascular disease effects.”
But chemoprevention strategies aren’t as common in cancer, she said, which is “very unfortunate,” especially since in addition to identifying new genetic risk factors, Peters and her team found potential candidates for interventions. And they identified currently FDA-approved drugs that might be put to use on these targets, either directly or indirectly.
“Industry is doing a lot in terms of drugs and they fund large-scale sequencing efforts, but they’re not as interested in chemoprevention for cancer,” Peters said. “Everything is happening in cancer treatment. That’s where industry is making the money.”
Peters and her team remain undaunted by these challenges, continuing to home in on colorectal cancer’s genetic roots with an eye toward putting their hard work to use as quickly as possible in the general public.
“It’s on the pathway to the clinic,” she said.
Funding for this study came from multiple public and private sources, including the National Institutes of Health, the National Cancer Institute and Cancer Research UK. View the full list of funders in the acknowledgements section here.
Diane Mapes is a staff writer at Fred Hutchinson Cancer Center. She has written extensively about health issues for NBC News, TODAY, CNN, MSN, Seattle Magazine and other publications. A breast cancer survivor, she blogs at doublewhammied.com and tweets @double_whammied. Email her at email@example.com. Just diagnosed and need information and resources? Visit our Patient Care page.
Are you interested in reprinting or republishing this story? Be our guest! We want to help connect people with the information they need. We just ask that you link back to the original article, preserve the author’s byline and refrain from making edits that alter the original context. Questions? Email us at firstname.lastname@example.org