Precision Oncology Database Highlights Racial/Ethnic Disparities Across Cancer Types

A precision oncology database used by researchers to define novel genomic variations between racial categories in different cancer types has proven to underrepresent racial/ethnic minorities across cancers and should be used cautiously in studies that directly compare racial/ethnic groups, according to researchers from Massachusetts General Hospital (MGH).¹

Findings were published in the NPJ Precision Oncology and showed that this omission may impact the validity of studies that attempt to directly compare mutational profiles between ethnic and racial groups for the most common types of cancer, thus limiting any generalization of biomarker discoveries to all patient populations.

“We sought to evaluate a real-world cancer data registry designed to accelerate precision oncology discovery. Our analysis indicates that minorities do not have sufficient representation both in general, and by various cancer types, which may impact the validity of studies directly comparing mutational profiles between racial/ethnic groups and limit generalizability of biomarker discoveries to all populations,” Sophia Kamran, MD, MGH told Targeted Oncology™.

In the study, researchers at MGH looked at the American Association of Cancer Research (AACR) Project Genomics Evidence Neoplasia Information Exchange (GENIE). GENIE is an international cancer registry which consists of clinically relevant genomic sequencing information that is publicly available.

Previously, Project GENIE has been used by research organizations to assess and define novel genomic variations between racial categories of different tumors. However, these findings have conflicted with other data in the past and have not been amenable to replication or biological explanations.

As a result, experts evaluated the GENIE tumor landscape by looking at the broader cancer population using the Centers for Disease Control and Prevention’s Wide-Ranging Online Data for Epidemiologic Research (CDC-WONDER). In the CDC-Wonder, experts have defined the number of patients with cancer expected from each racial or ethnic group for a given cancer type.²

A total of 43,472 samples were seen in GENIE in only sites across the United States, and 52,773 samples were observed among all sites, including international sites, at the time of the analysis. Within GENIE, most of the patients were White for all cancers analyzed from the US, including a range from 67.8% for gastric cancers to 91.2% for patients with melanoma. Using observed-to-expected ratios >1, White patients were adequately or over-represented for all cancer types except for lung cancer and melanoma.

Patients who were Asian or Pacific Islander had a significant over-representation across colorectal cancer (ratio 1.47; 95% CI, 1.20-1.79; P < .001), lung cancer (2.43; 1.80-3.29; P < .001), and breast cancer (1.32; 1.10-1.59; P < .001). Among these patients, there was an overall ratio of 1.46 across all cancer types (95% CI, 1.27-1.67; P < .001).

However, excluding melanoma and cervical cancer, the findings showed that Black patients were significantly under-represented across all cancers and has a ratio of 0.50 (95% CI, 0.40-0.62; P < .001). Hispanic patients also were consistently under-represented and showed a ratio of 0.57 (95% CI, 0.50-0.65; P < .001) among all cancers.

Additionally, Native American patients had no samples for 5 cancer types, including anal, prostate, endometrial, cervical, and vaginal cancers, showing this demographic to be significantly under-represented in many cancer types.

According to the researchers, some of the reasons behind the underrepresentation of minorities can be attributed to institutional exclusion that has been ingrained as well as there being mistrust by research agendas when evaluating minority groups themselves.

Overall, these findings suggest that data from GENIE do not reflect the actual landscape of patients with cancer in the United States and may misrepresent the disease burden in various racial/ethnic minority populations. To summarize, GENIE shows White patients to be adequately or over-represented across nearly all cancer types, while most minorities are underrepresented, excluding the Asian and Pacific Islander category. Currently, GENIE is not able to detect small yet potentially clinically meaningful differences between patients who are White vs those who are non-White in even the most common cancer types.

Experts at MGH are hopeful that these results will increase recruitment efforts and represent more patients in minority populations to adequately evaluate future studies.¹

“Our results reflect the historical status quo for precision medicine research. Multiple other biorepositories have found low representation of diverse individuals, from TCGA to GWAS. In addition, a recent analysis of precision oncology clinical trials revealed consistent under-representation of non-white participants. The reasons for this are likely multi-factorial. However, the oncology field is recognizing this limitation of these rich databases, and many stakeholders are making strides to correct it. For example, AACR Project GENIE is implementing various solutions to increase diversity and representation within their database, which we applaud, said Kamran, assistant professor of Radiation Oncology, and director of Diversity, Equity, and Inclusion in the department of Radiation Oncology at Harvard Medical School.

“These biorepositories democratize precision oncology research and are critically important to accelerating oncologic discoveries, therefore we merely note caution when using these for studies that directly compare racial/ethic groups currently. However, we are encouraged by the ongoing changes to improve representation of these precision oncology registries to better reflect our cancer populations,” Kamran added.

REFERENCES:

Researchers uncover racial and ethnic disparities in a widely used precision oncology data registry. News release. Massachusetts General Hospital. January 31, 2023. Accessed February 6, 2023. https://bit.ly/3Xa4GUB

Cheung ATM, Palapattu EL, Pompa IR, et al. Racial and ethnic disparities in a real-world precision oncology data registry. NPJ Precis Oncol. 2023;7(1):7. Published 2023 Jan 19. doi:10.1038/s41698-023-00351-6