Among the topics discussed at this year’s annual meeting of the American Association for Cancer Research (AACR) was disparities in cancer care. Several MSK researchers presented research that directly addresses these problems.
In a prerecorded presentation, MSK physician-scientist Sana Raoof explained how clinical trials of liquid biopsies — a type of blood test for cancer — can include concrete measures to reduce disparities related to access and inclusion.
Dr. Raoof said such trials should aim to include at least 25% of participants who are non-white — a figure that more accurately represents the percentage of non-white individuals in the U.S. population. Previous liquid biopsy trials have included 4% to 13% non-white participants.
“The lack of representation of minorities is a huge problem in cancer clinical trials,” Dr. Raoof says. “It’s not just a problem because it’s unfair and inequitable; it’s also a problem because it affects scientific conclusions.”
To improve access to liquid biopsy trials, academic medical centers should seek to partner with community hospitals that see a more diverse patient population, she says.
Liquid biopsies may aid in early detection because they could potentially be administered in a primary care physician’s office, rather than at a cancer center, where most imaging scans and other tests are performed.
“In the context of the COVID, we’re seeing later and later stages at diagnosis across multiple types of cancer,” Dr. Raoof says. “People are missing their cancer screening tests and are therefore being diagnosed when they have symptoms at later stages, when it’s harder to cure the disease. That is particularly affecting minorities who, at baseline, have worse access to healthcare.”
“It would be very useful,” Dr. Raoof continues, “if we had a point-of-care test that was convenient, quick, and cheap that could be used by everyone but might be of particular benefit to minorities to supplement cancer screening.”
Dr. Raoof wrote an editorial about the potential of liquid biopsies to address disparities in access in the Boston Globe last year.
Addressing Disparities in Colorectal Cancer Outcomes
MSK computational oncologist Francisco Sanchez-Vega is part of a team that has for a number of years been studying the genomics of colorectal cancer, trying to identify genomic features that correlate with worse outcomes. They are mining the wealth of genomic data obtained from MSK-IMPACT®, a gene-sequencing panel that looks for cancer-associated mutations in 500+ genes.
While doing this work, Dr. Sanchez-Vega noticed that, on average, African American patients in the MSK-IMPACT cohort had worse outcomes compared with other patients. The question that he and others want to understand is why.
“The general belief is that this disparity is probably driven by socioeconomic factors, differences in risk factors, differences in access to healthcare, maybe differences in environmental exposure, lifestyle, diet, etc.,” Dr. Sanchez-Vega says. “But we had a very good opportunity to try to understand if there were any genomic features that could at least partially explain these differences or give some insight into these differences.”
In his presentation at AACR, Dr. Sanchez-Vega discussed the various genomic factors that they measured, including tumor mutation burden, location of the tumor, and frequency of various oncogenes. The study population included 3,963 patients with colorectal cancer treated at Memorial Sloan Kettering Cancer Center (MSK), including 336 (8%) self-reported Black patients and 3,627 (92%) self-reported white patients.
They found some differences. For example, right-sided colon tumors were more common among Black patients (37% versus 25% in white patients). Hypermutated tumors, including tumors with a condition called microsatellite instability, were more frequent among white patients (12% versus 8% in Black patients).
Tumors in Black patients were enriched in KRAS mutations (60% versus 45% in white patients), but high-risk KRAS-G12C mutations accounted only for 3% of all driver KRAS mutations in Black patients versus 8% of all driver mutations in white patients.
When the investigators factored in all these genomic and clinical differences, however, they still did not explain the difference in outcomes between the two groups — namely, that Black patients had shorter overall survival from the time of sequencing (median 28 months versus 50 months for white patients). This was true even though all patients were treated at the same single institution (MSK) during the study.
Dr. Sanchez-Vega cautions that these results may not apply behind the specific cohort of people who came to MSK, since this population may not be fully representative of the wider community. Nevertheless, the results would seem to point to other factors — beyond genomic ones — to explain these disparities in outcomes.
Using Ancestry Testing To Improve Cancer Care
For certain types of cancer, knowing one’s precise genetic ancestry can be important for properly assessing genetic risk factors for the disease and weighing possible treatments. Perhaps the best example of this phenomenon is people with Ashkenazi Jewish European ancestry, who have a much higher risk of carrying a high-risk BRCA1 or BRCA2 mutation than the general population. These mutations put one at risk for hereditary breast, ovarian, prostate, and pancreatic cancers.
Traditionally, establishing someone’s genetic ancestry required looking at genetic markers from across the entire genome. At AACR, computational biologists Kanika Arora, Michael Berger, and their colleagues in the Marie-Josée and Henry R. Kravis Center for Molecular Oncology presented their work showing that data from MSK-IMPACT® — which sequences only a subset of genes in the human genome — can be used to reliably establish someone’s genetic ancestry.
The finding raises the possibility that genetic ancestry testing could be incorporated into routine clinical care whenever a test such as MSK-IMPACT is administered. The potential benefits include a more accurate genetic risk assessment than self-reported race.
“A lot of times in medicine, self-reported race is used as a substitute for genetic ancestry, but it’s not always a good indication of someone’s actual genetics,” explains Arora. “For example, someone who identifies as Black could have close to 100% African ancestry, or they could have 50% or less African ancestry. So, while both are Black, self-reported race does not always capture the genetic component well, which is why for certain questions about risk factors, it’s important to look in detail at measures of ancestry.”
According to the investigators, one particularly noteworthy finding to emerge from their analysis is a discrepancy in the rates of clinically actionable alterations in patients of African ancestry versus patients of European ancestry as determined by MSK’s OncoKB® database. While the overall proportion of biologically relevant driver mutations was similar between the two groups, patients of African ancestry had, on average, fewer options for drugs matching those mutations.
There are likely several reasons for that, Dr. Berger says, including the fact that genomic-sequencing panels themselves are skewed toward the genomic data collected from largely European populations.
Regardless of the reason, the data would seem to call for more equitable drug development programs, he notes. “Patients with African ancestry are not as well served as people with European ancestry by our current precision oncology therapies,” Dr. Berger says.