Assessing the number of genetic mutations in tumours such as lung cancer cells (pictured) can guide treatment decisions
Shutterstock / David A Litman
Genetic tests that predict the efficacy of certain cancer treatments aren’t as effective for people of African or Asian ancestry.
These tests measure the number of genetic mutations in solid tumours, or what’s known as tumour mutational burden (TMB). If someone has a tumour with 10 or more genetic mutations, they have a high TMB. Drugs called immune checkpoint inhibitors, which work by turning the immune system against tumours, have been shown to be quite effective against cancers in people with high TMB, but less so in people with fewer mutations.
The most accurate way to measure TMB is by doing genetic analyses of both tumour and normal tissue samples. This is called tumour-normal sequencing, and it ensures that mutations people inherent from their parents aren’t counted as tumour mutations. “If you don’t have matched normal tissue from an individual, you can actually overestimate the tumour mutational burden,” says Amin Nassar at Yale University in Connecticut.
Advertisement
Most treatment centres currently use tumour-only genetic sequencing, which estimates TMB by comparing results with genetic databases instead of a person’s own tissue. This saves time and money, but the vast majority of genetic information in these databases is from white people of European descent, which means that these comparisons may lead to more TMB misclassifications among people of Asian or African descent.
To determine how common this is, Nassar and his colleagues collected data on 3618 people from the National Cancer Institute’s Cancer Genome Atlas Program who had cancers that could be treated with immune checkpoint inhibitors. The researchers used genetic testing, rather than self-reported race, to determine that 582 people were of non-European ancestry. Since tumour and normal tissue samples were available for all participants, researchers estimated TMB with both tumour-only and tumour-normal sequencing.
They found that, on average, tumour-only sequencing more than doubled TMB in people of non-European ancestry compared with tumour-normal sequencing. For people with European ancestry, tumour-only sequencing increased TMB 1.5 times.
Using this data, the researchers developed a simple formula to recalibrate tumour-only sequencing results that accounts for ancestry, cancer type and TMB overestimates. They then tested the formula in a group of 456 patients who had undergone both tumour-only and tumour-normal sequencing and found it reduced overestimations of TMB in tumour-only sequencing results. When tested on 2800 people treated for cancer at the Dana-Farber Cancer Institute in Boston, the formula revealed that 43.6 per cent of those with African ancestry, 37 per cent of those with Asian ancestry and 21 per cent of those with European ancestry had been misclassified by tumour-only sequencing as having high TMB.
“If we’re misclassifying more Blacks and Asians as having high TMB, then we’re more likely to give them [immune checkpoint inhibitors],” says Nassar. Not only are these drugs ineffective on tumours with low TMB, but they can also cause serious side effects like diabetes, myocarditis and trouble breathing, he says.
What’s more, even accurate TMB scores may not be as useful for guiding treatment decisions for people who are Black or Asian. As part of this work, the researchers also looked at a separate cohort of 879 people treated at the Dana-Farber Cancer Institute for non-small cell lung cancer. In that group, 758 people had European ancestry, 64 had African ancestry and 57 had Asian ancestry. All had high TMB according to tumour-only sequencing and were treated with immune checkpoint inhibitors.
The algorithm reclassified 569 people as having low TMB. The researchers then analysed treatment outcomes by racial group and recalibrated TMB scores. They found that, after 36 months on checkpoint inhibitors, the overall survival rate for people of European ancestry with high TMB was 42 per cent compared with 22 per cent in those with low TMB.
“In white populations, we know that high TMB correlates with a better outcome” after treatment with checkpoint inhibitors, says Nassar. But “for Blacks and Asians, this was actually not the case”.
The researchers found that people of non-European ancestry who had high TMB scores had lower survival rates after treatment with checkpoint inhibitors than those with low TMB scores. In Black people, survival rates were 26 and 31 per cent for high and low scores, respectively. For people with Asian ancestry, they were 12 and 29 per cent.
This disparity might be due to healthcare inequities, a lack of racial diversity in clinical trials or the study’s limitations, like a small sample size, says Nassar. “Tumour mutational burden doesn’t extrapolate to non-whites,” he says.
To better treat cancers in Black and Asian people, we must first understand why this is and what other measures may need to be used instead. That means increasing diversity at all stages of the drug development process, including the number of Black and Asian participants in clinical trials, Nassar says. In the near term, though, it is crucial that doctors are aware of these biases when they make treatment decisions based off TMB, he says.
Journal reference: Cancer Cell, DOI: 10.1016/j.ccell.2022.08.022
More on these topics: