Combating Underrepresentation in Most cancers Well being with CRISPR

The drivers of breast cancer in underrepresented populations is poorly understood.

Scientists design therapeutic approaches for cancer around mutations and variations that affect disease onset and progression. However, these determinants can differ across populations, and what drives cancer in underrepresented populations is poorly understood. Melissa B. Davis, a geneticist at the Morehouse School of Medicine, wants to rectify this knowledge gap.

In a study recently published in Cancer Research, Davis and a multinational team developed a CRISPR screening pipeline for identifying potential breast cancer therapeutic targets for patients with West African ancestry.¹ They found several previously unidentified targets, particularly noting that a combination of FGFR1 and EGFR inhibition blocked cancer cell proliferation.

How are technologies such as CRISPR and high-throughput sequencing helping to fill knowledge gaps for underrepresented populations?

There is a knowledge gap around disease plasticity in different populations. Everyone is represented in the clinic, but not everyone is represented in the data, and so we struggle to interpret or manage diseases that run unique courses in certain populations. Right now, we focus on the West African genetic diaspora because it bears the heaviest burden for breast cancer.

I think establishing the genetic background for populations outside of the European diaspora is important because it helps unveil potential drivers of unique cancer phenotypes that could possibly be more aggressive or drug resistant. Techniques such as CRISPR are also opportunities to identify the functional consequences of these newly unveiled variants. In this study, we used CRISPR to systematically identify important interactions in cells derived from underrepresented populations, such as how a combination of FGFR1 and EGFR inhibitors might benefit a population with these genetic backgrounds. This suggests ancestry-informed combination therapies could be repurposed or developed to improve outcomes in populations historically left out of drug development.

This study sourced primary cells from both Ghanaian and African American patients. How genetically similar are West African diaspora populations in different geographic regions?

There will always be variation between populations. For example, Ghanaian populations tend to have very little admixture from other sources, while African Americans can trace genetic ancestry from just about every region of Africa and additionally possess a degree of admixture from European ancestry. Despite this almost stochastic ancestry admixture, a large proportion of these individuals still drive towards a certain phenotype when they get cancer. As such, we measure genetic background not to stratify people into smaller groups, but to identify common regions—there is probably a signal that would allow us to identify the genomic region that harbors a unique allele or set of alleles driving this unique phenotype. In my mind, these admixed populations are ideal because everything else about the genetic background is different except for this one thing that we are looking for. This brings statistical power in a way a typical genome-wide association study would not be able to do.

The kinases of interest identified in this study are well-characterized. Was it surprising to find yet undiscovered roles?

We have a canonical understanding of what these genes do in this context, especially in the way we conceptualize gene loss-of-function in cancer. However, if we think about non-disease contexts such as development, there is more pathway variation. For instance, differential regulation of the FGFR pathway clearly results in variable phenotypes when it comes to things like height and bone density.

Personally, I was not that surprised, because we see this quite frequently. For example, oncogenic p53 mutations in our cohort tend not to be loss-of-function but rather gain-of-function. Jill Bargonetti, who I collaborate with, has studied this her whole career and has applied this to identify alternative eligibility criteria for PARP inhibitor therapy.²

The whole point is to fill knowledge gaps—like we have only been searching around the light radius of the lamp post when there is a whole field that we have yet to take advantage of. With tools like CRISPR, I think we have a great opportunity to try to model and simulate combinations that we would not necessarily have seen before. We can also correct historical inequities in who benefits from cancer research. This is how we move from representation in the clinic to representation in the data.

This interview has been condensed and edited for clarity.