This meeting took place in Pacific Grove, CA, USA, on 12-15 December 2024. This meeting convened experts to examine why certain populations experience worse cancer outcomes and how to advance equity in cancer research. Despite rapid progress in cancer research and treatments, not all groups benefit equally from new breakthroughs​. This forum focused on the causes of these disparities and ways to ensure future advances benefit all populations.

The variation in cancer incidence and outcomes across different populations was deeply discussed. For instance, endometrial cancer mortality has increased despite an overall decline in cancer-related deaths. One major challenge is the underfunding of women's health research, leading to a lack of accessible diagnostic tests and limited understanding of risk factors. Black women, in particular, face worse outcomes even when controlling for healthcare access, raising questions about potential biological and systemic influences.

The impact of ancestry on cancer susceptibility was also a key focus of this meeting. Studies have shown that somatic-germline interactions influence cancer risk; for example, certain genetic mutations (e.g., ERG fusions in prostate cancer, or EGFR mutations in lung cancer) are more common in specific populations. However, ancestry-related variations in tumor biology are still poorly understood, and the lack of diverse datasets limits our ability to explore these connections effectively. For example, current methods may assign higher tumour mutation burdens to tumours of Non-European patients, but this is artificial and due to the fact that our references are based on European populations. This was also highlighted by discussions of the Duffy null allele, which is more common in individuals of African and Middle Eastern descent and leads to lower white blood cell counts - a normal physiological condition that was previously known as ‘benign ethnic neutropenia’.

The little data available on non-European populations remains a significant challenge. Many clinical trials lack diversity, partly because researchers do not consider ancestry in study design, and informed consent forms are often unavailable in non-English languages. There is a pressing need for funding initiatives to increase diverse participation in clinical trials and facilitate international collaboration in cancer research. In Latin America, studies on breast cancer have identified ancestry-associated risk factors, such as specific genetic variants linked to HER2+ subtypes, but further research is needed to assess their clinical implications.

Beyond genetics, social determinants of health significantly impact cancer risk and progression. For example, stress has been shown to accelerate disease progression in experimental models, and socioeconomic barriers often delay access to cancer care. Studies on Aboriginal Australians have revealed that cancer disparities stem from systemic inequalities, including poor living conditions and limited access to healthcare. Underrepresentation of African people in genomic studies remains a major issue, despite Africa's immense genetic diversity. Efforts to improve electronic health record systems and increase sequencing accessibility are needed to bridge this gap.

Overall, the symposium emphasized that overcoming disparities in cancer research requires a multi-faceted approach—one that integrates genetic, environmental, and social factors into a comprehensive understanding of cancer risk and treatment. By prioritizing diversity and inclusivity in research efforts, the scientific community can work toward more equitable cancer care and improved outcomes for historically underrepresented populations.

—-------

Nyasha Chambwe
There has been a general decrease in cancer mortality over the last two decades, but mortality from endometrial cancer has increased. Women’s health conditions are chronically underfunded, so we don’t know much about this disease. It is really hard to diagnose as no easy access tests exist. Black women have worse outcomes when controlling for access, but is the system utilised and at the same rate? Cohorts are needed with genetic and non-genetic risk factors.

They assessed the prevalence of known pathogenic variants in the All of Us cohort. Most risk is multi-gene and multi-factorial. VUS can be explored as the 2nd hit hypothesis - need a 2nd hit to become pathogenic. Questions: Is obesity protective for earlier stages? Do GLP-1 inhibitors have an effect? Do hormones have an impact on onset?

Jian Carrot-Zhang
Somatic-germline interactions exist, for example, ERG fusion is enriched in prostate cancer patients of European ancestry, and EGFR mutations are enriched in lung cancer patients of East Asian and Native American ancestry compared to European ancestry. Could be due to germline modification or germline-exposure interactions. Can ancestry-associated germline variants explain somatic differences between populations? Allele frequency, effect size? HLA-type II may play a role in these interactions. Non-European samples have more driverless samples. Usually, TMB can be used as a biomarker but underrepresented groups have artificially high TMB. There is higher TMB in lung cancer in African patients, TP53 is modified in smokers across different ancestries. What is the mechanism by which ancestry modifies somatic profile? High-penetrance in cis, los penetrance in trans. Case-case GWAS. Other hidden, correlated variables with ancestry (e.g., stress?) may also be modifying somatic profile. MOuse/IPSC signatures are not reflective of what we see in humans.

Laura Fejerman
Points to think about - solutions to sparse data for non-European populations and data sharing restrictions. Think about correlations between genetic ancestry, geography, identity, SES and its impact on analytical outcomes and results. Began scoring all clinical trials for increasing diversity - in many cases, people running the trials do not have inclusion/exclusion criteria relating to ancestry because investigators only expect European patients. Informed consents are not translated into Spanish (not a rare language in the US). Question - can IARC help with finding translators internationally? Funding sources are needed to increase diversity in clinical trials.  

Breast cancer in Latina women has lower incidence, there may be a genetic component. Ancestry was associated with breast cancer risk in paper from 2006, with a signal at the ESR1 locus. Then, with more data, a variant was identified. Laura is interested in HER2+ cancers. Uruguay and Argentina have a high incidence of these, and low Native American ancestry, whereas Peru shows an association of ancestry with HER2+ subtype. Genetics may be useful - People in the top 1% of PRS have high enough risk that it becomes informative for clinical management.

Alejandro Ruiz-Patiño
The CANDELA consortium studied genetic diversity in Latin America, and found that Peru has the highest recent Asian immigration. His interest started with lung cancer studies that identified a relationship between ancestry and EGFR mutation rate, but then focused on lentigo maligna melanomas, which was the most represented subtype of melanoma in the registry he was studying (however, acral melanoma is the most common type of melanoma in Colombia). There are different proportions of BRAF-mutated tumours in different departments in Colombia, but it is currently unknown whether there is a relationship with ancestry of other variables in this cohort. It was mentioned that Jewish patients have a higher proportion of BRAF^V600E mutations, while this proportion is lower in African patients. It was discussed that perhaps an initiative to make funding (and other resources) available to the Colombian team would help them tease out this relationship.

Carino Dias-Gurjao
‍Colorectal cancer risk and progression are associated with genetics, immunity, microbiome and lifestyle, all of which can be studied through DNA sequencing. African Americans have higher incidence and mortality from colorectal cancer, there is for sure an environmental component but that is not fully explained yet. When they undertook metagenomic profiling in African Americans and Europeans, the bacteria that came up as most overrepresented in AAs was Flavonifractor plautii, whereas in Europeans these were lumen-associated bacteria. There are other associations of colorectal cancer, including with diet, antibiotic regulation in countries across the world.

Sarah Moody
Working on mutational epidemiology as part of the Mutographs project. Fifty percent of cancers in Africa are expected to be infection-related. There are mutational signatures that show regional and country-specific variability, for example, SBS12 is much more common in Japanese samples (both kidney and liver cancers, but has not been seen in other samples). Why could this be? It could be due to environmental exposure or DNA repair efficiency. For example, SBS16 has also been seen only in Japanese drinkers - but this was then found to be associated with a specific germline variant. So these mutational signatures could also be due to germline variants that show distinct allele frequencies around the world. As an example of this interaction, BRCA deficiency is associated with SBS3 and ID6.

Scott Coven
Started by enlisting the questions that patients in his care commonly ask. These include ‘Why?’, ‘What can I do about it?’, ‘Are there any side effects?’, ‘Is it something that I did?’. There are many unanswered questions when studying underrepresented populations and rare cancers. In fact, there are disparities also at the diagnostic level, “epidemiologically these cancers do not exist”. There are many social determinants of health that affect disease progression, for example, stress. Stress has been shown to be associated with disease progression in mouse experiments: Pre-leukaemic mice have been made to share a space with aggressive males, which stresses them, and this has been shown to fuel clone growth. So, how to reach patients? Can they be seen in their own home? Scott’s team has started asking patients questions regarding their food security and access to healthcare and other social aspects. It is important to consider who asks the questions - patients probably feel more comfortable with social workers. Also, lots of breakups happen after treatments, as well as with childcare - also arguing about patient treatment. Communication with patients is key, for example, explain, why are you prescribing this treatment? There is perhaps also a correlation between sociodemographics and spirituality. Only a small proportion of patients answered that problems at home impacted their healthcare (about 27%).

Kian Hong Kock
Studying human diversity is scientifically important, and is a pressing equity issue. But, how do we classify individuals? There are categorical and continuous measures of diversity. But there is much confusion, for example, perhaps African Americans should be classified in the Admixed group instead of in the African group. Language also tends to be a good indicator of where someone is from. When studying the composition of blood in samples from different countries, and taking into account age, BMI, ethnicity, sex and smoking status, there are differences in immune cell composition in samples from different ethnicities. Also, a decline of CD19 cells with age is seen.

Samantha Jumbe
There is a need to increase studies of cancer genomes in under- and unrepresented populations in Africa, there is a need to conduct genomic research in understudied cancers in Africa. There is sometimes a problematic attitude of researchers toward generating data in Africa, assuming that it is not feasible. But, how to challenge these attitudes but at the same time be practical and get things done? Some companies (such as LabCorp) maybe would be able to help as they have labs in multiple countries and multiple forms forms of education - virtual options have become available. Medicines and sequencing are inaccessible, and African individuals represent fewer than 1% of those included in cancer GWAS, yet they represent 90% of human genomic diversity. There is a need for electronic health records. We need to consider ancestry for variant pathogenicity interpretation.

Sandy Thompson
Has been studying disease incidence in Aboriginal people and Torres Strait Islanders in Australia. This represents one of the world’s oldest continuous cultures (65,000 years), and aboot 167 indigenous languages are still actively spoken. The terms “First Nations” and “indigenous” are used interchangeably. 3.8% of the Australian population is Aboriginal, and they have 2.3 times the burden of disease. They are segregated - they tend to live inland, which is “less nice”. She has observed that was causes cancer is poor living conditions, and smoking-related cancers are more frequent in Aboriginal people. The population tree comparin Aboriginal and non-Aboriginal people shows that Aboriginal people die younger and have a higher birth rate, which looks like the population tree of a developing country. In contrast, the population tree of non-Aboriginal people matches those of developed countries. The definition of aboriginal person consists of three ‘components’: descent, self-identification and recognition by other indigenous people. Much of cancer may be preventable, but not all. Why do people not access cancer care earlier? It is multifactorial: could be finance issues, or that people have families and other priorities in their life. It depends on ability to perceive, to seek, to reach, to pay, and to engage.

Cassie Kline
Works in pediatric neuro-oncology. Has been studying pediatric brain cancers and how these relate to ancestry - if classifying patients in the five 1000G superpopulations, these show distinct prevalence and outcomes across pediatric central nervous system tumours from the PBTA and PNOC. There is some enrichment of germ-cell tumours in Asian populations, and others - overall, there is an enrichment of different diseases across different ancestry groups. However, there is also a need to assess stress levels - perhaps through questions such as where they love but also long-term (transgenerational) stress levels may be relevant.

Patricia Possik
Has been studying acral melanoma in the Brazilian population. In Brazil, mixed populations are classified as ‘admixed’ regardless of the combining groups (e.g., African with European, African with Indigenous, etc). The majority of patients though are African/European admixed. It was suggested that perhaps admixture mapping would be a good technique in order to identify causal variants. Or - look for the haplotype distribution and BRAF^V600E mutation and see in which ones it is more likely to arise. Perhaps, discrete categories could be compared (e.g., AFR vs EUR). Different studies show different proportions of somatic BRAF mutation, they are composed of patients of different populations, and they have been assessed by different methods.  

Lucy Godley
Introduced the Duffy null project. Complete blood counts (CBCs) are composed of whote blood count + hemoglobin + platelets + MCV. In the past, a condition recognised as ‘benign ethnic neutropenia’ was recognised, in which some people had lower WBCs, but this was later recognised to be due to genetic variation (known as Duffy null) that is more common in people of African and Middle Eastern descent (⅔ of African Americans and 96% of Saudi Arabians carry it). This allele was selected in Western Africans because it is protective of malaria. This example illustrates that if only certain populations are studied (e.g., white-skinned from European descent), the range of normal physiological conditions and genetic variants influencing distinct phenotypes may be missed.

Monica Cappetta
Focused on breast cancer in Uruguayan women: 1 in 11 women in Uruguay will be diagnosed with breast cancer, which is similar to other developed countries. There is a negative correlation between global DNA methylation and ancestry; African ancestry is associated with lower DNA methylation in the blood of breast cancer patients. Latin American patients seem to have different methylation clocks, which are marks that are used to estimate a person’s age. The methylation analysis in these Uruguayan breast cancer patients and controls find candidate epigenetic biomarkers that are not found in European patients - is this environmental or genetic? (Most Uruguayan patients will probably be of European descent).  

Daniela Robles
Focused on her lab’s studies in acral melanomagenesis in Latin American patients. There is a detected correlation between ancestry and BRAF mutation rate (European-descent patients have a higher BRAF mutation frequency). These differences are also potentially linked to acral melanoma cell of origin, which may underscore differences in the biology of the disease depending on ancestry. Some questions raised are: Why is there a relationship between ancestry and somatic/copy number profile? How important is the germline contribution? How important are these differences for treatment? (e.g., metabolism of certain drugs)? Do we need more diverse experimental models? (e.g., saturation genome editing experiments)