AACR 2021 | Racial differences in somatic mutations among patients with early-onset CRC

My team’s research program is centered upon the distinct burden of early-onset gastrointestinal cancers, or the GI cancer burden among adults younger than age 50 years. And we take a really interesting translational take on it to really assess aspects of this distinct burden from cells to society. And so to give you a little bit of background on this research space, the incidents of colorectal cancer among young adults has been on the rise over the last several decades about, in the United States, a one and a half percent annual increase. And this is mirrored with similar trends in countries worldwide.

And what’s striking about this is we don’t fully understand the etiologies that are driving this increasing incidence. And for individuals at average risk, screening would not be recommended for this population until, what is somewhat shifting to the trend of age 45. So beyond this increasing incidence in our group and others starting to demonstrate that early-onset colorectal cancer is a distinct malignancy with both clinical and molecular features compared to lat- onset cases. One of the other things we’ve demonstrated are distinct disease burden across population subgroups within the early-onset colorectal cancer cohort overall.

So these differences persist by race and ethnicity, by sex, by community health behaviors. So there’s a multitude of complex interrelated factors that yield striking variation in the early-onset colorectal cancer burden overall. And so what we sought to do was be the first study to begin exploring molecular differences, perhaps underlying early onset colorectal cancer disparities.

So the unique thing about our study is that we really wanted to assess differences within the early-onset colorectal cancer population, not just how does early-onset disease differ from that in cases aged 50 years and older, but within this population, what are some of the biological differences that may be driving the distinct disease burden we see across this population? And so we decided to take a first…

So to do this, it’s a bit challenging given the limited case numbers at individual institutions. So we took the opportunity to leverage a unique resource, the AACR project GENIE consortium data that included clinical grade targeted sequencing as well as clinical demographic data for over 6,000 pathologically confirmed colorectal cancer cases, of whom approximately a third or just shy of 30%, were diagnosed with early-onset disease based on age at cancer sequencing. And so we really had this unique opportunity to have a robust early-onset colorectal cancer cohort, to be able to explore differences by race and ethnicity within this population, which has not been able to been done before.

And so the first thing we did is predict microsatellite stability by tumor mutation burden. Giiven the approximately five out of every six young patients with colorectal cancer don’t have a genetic predisposition to disease, our interest and focus was really on that pool of cases with non-hyper mutated or microsatellite stable and colorectal cancer to really understand the molecular differences that may underlie disparities for particularly this tumor type.

And so approximately 90% of our cases in this cohort had a microsatellite stable colorectal cancer, which was about 5,500 cases whom were included for further analysis. And so to transition to our findings, I would say one of the really striking observations that we made is when we assessed tumor mutation burden by race and ethnicity within the population of early onset colorectal cancer cases. And what we observed is that compared with non-Hispanic whites, younger than age 50 at cancer sequencing, young non-Hispanic black individuals presented with a significantly higher tumor mutation burden.

Again, this is within the cohort of microsatellite stable colorectal cancers only. But in contrast, compared with whites, we saw no significant difference in tumor mutation burden with Asians and Pacific Islanders. And so that finding that blacks have the significantly higher tumor mutation burden than whites within early-onset colorectal cancer cases was quite intriguing, potentially clinically impactful and warranted further study. And so what we next sought to do was look at unique somatic mutation patterns in early onset colorectal cancer by race and ethnicity. We wanted to know which somatic gene variations had distinct patterns in early- versus late-onset. But in this case, we stratified by race and ethnicity to be able to not mask findings within some of these smaller population groups that you would see if you had just adjusted for race in the model.

And in doing so we saw several genes that had unique mutation patterns in early onset colorectal cancer, compared with late onset cases and adjusted models among whites. But what was really striking from those results were that the young Asian Pacific Islander cohort, the genes that were differentially expressed in that population, compared with late-onset cases, those genes did not overlap with the genes that were unique to early onset colorectal cancer in whites.

And similarly, you saw different genes emerged that had unique patterns in early onset colorectal cancer cases among non-Hispanic blacks. So to give you an example of that, Asian and Pacific Islander patients with an early onset microsatellite stable colorectal cancer, were about 48% and 66% less likely to have a non silent mutation in APC and PIK3CA respectively. And were about 4.7 times more likely to have a non-silent FAT1 mutation than those with late onset microsatellite stable tumors.

For the young, non-Hispanic black patients, they were more likely to have non-silent mutations and CREBBP and TGFb receptor two compared with late onset cases. These unique gene patterns in young blacks and young Asian and Pacific Islanders were not genes that had unique patterns in the non-Hispanic white cohort. And we also went on further to look at significant… We also went on to also look at mutations that had significantly different frequencies within the early-onset pool, by race and ethnicity, to confirm some of that heterogeneity and these mutation patterns within the population of early onset colorectal cancer cases.

The overall take home for our study, in it being first of its kind to demonstrate differences in the biology of early-onset colorectal cancer by race and ethnicity, that warrant further study and could potentially suggest unique molecular mechanisms driving early-onset colorectal carcinogenesis across population subgroups, somewhat aiming towards the idea of precision medicine and better understanding the disproportionate disease burden, at a biological level, for early-onset colorectal cancer across diverse population subgroups.

Our study is first of its kind in that we revealed molecular differences in early-onset colorectal cancer by race. I want to acknowledge that validation of these findings is needed in an independent cohort, but these may help us decipher if the molecular characteristics within the microsatellite stable colorectal cancer itself may contribute to these differences in disease burden that we’re seeing. I do want to also make note that race is a social construct, but unfortunately one of the considerations for this study is that GENIE does not currently have data available for genetic ancestry of these individuals.

But the hope is to really start to translate these findings from the societal level to the cellular level in leveraging preclinical models in the next step to further explore some of these findings and disentangle, hopefully reduce, and eventually mitigate disparities in the early onset colorectal cancer burden worldwide.

We were fortunate. It’s actually very timely. We just published our nature reviews cancer article on this topic, suggesting the need for the study of biology of early onset colorectal cancer disparities.

I’ll say one more thing that maybe makes it in the video. But that’s to consider that complex interrelated factors contribute to early onset colorectal cancer disparities, at the individual level and the built environment, social behaviors that could influence the biology underlying these disparities. So in the future and taking this work to the next level, it’s important to contextualize that this is really one piece of the puzzle, and better understanding this network of interrelated factors will help us disentangle the drivers, biologically, of early onset colorectal cancer disparities.

So in the long term the differences we observed in tumor mutation burden within the population of early onset colorectal cancer patients may carry clinical impact, but our findings could also help develop therapeutic or prognostic biomarkers for early-onset colorectal cancer, and really moves towards the idea of facilitating precision medicine for young patients diagnosed with colorectal and eventually, perhaps, could be applied to other gastrointestinal cancers as well.