NEW YORK — Some women at high genetic risk of breast cancer may further increase their risk if they take corticosteroids, a new analysis has found.
While a person's genetic profile can influence their risk of developing breast cancer, environmental and lifestyle factors also contribute to disease risk and could additionally interact with genetic risk factors. To explore this, researchers from Weill Cornell Medicine used UK Biobank data to examine interactions between medication use and genetic susceptibility to breast cancer. As they reported on Friday in the American Journal of Human Genetics, they found that using corticosteroids increased breast cancer risk among some women who were already at increased genetic risk for the disease.
"What is particularly surprising is that in our analysis, corticosteroid use is associated with increased breast cancer risk only in patients with a high genetic risk of breast cancer as measured using a polygenic risk score," senior author Olivier Elemento, director of the Englander Institute for Precision Medicine at Weill Cornell, said in an email. He noted that previous work had suggested corticosteroid use might influence breast cancer progression and metastasis, but that there was no known link to disease risk.
For their analysis, the researchers combined two polygenic risk scores to gauge the genetic breast cancer risk of more than 212,000 women from the UK Biobank. At baseline, none of these women had breast cancer or a personal history of the disease, but nearly 12,000 later developed breast cancer.
At recruitment to the UKB, the women also reported what medications they took, and after removing birth control medications from the mix, the researchers examined the pairwise interactions between 96 types of medication and PRS-based risk of breast cancer. This analysis uncovered five medication groups with significant interactions, all of which were related to corticosteroids. The researchers further calculated that the PRS explained about three times the variation in breast cancer risk among women who took corticosteroids versus those who did not.
The investigators also traced the top 40 SNPs from the PRS that contribute to the interaction to 35 genes. Many of these genes are targets of NFE2L2 and its protein NRF2, which mediates cellular response to stress. Its overexpression has been linked to breast cancer development and tumorigenesis.
Further, the expression of NFE2L2 appears to be influenced by common corticosteroids like dexamethasone, hydrocortisone, and betamethasone. This suggested that these SNP polymorphisms might increase breast cancer risk through effects on NRF2 activity, and that NRF2 is then further influenced by corticosteroid use to raise risk even higher. "It's just speculation at this stage, but this is what the data is pointing to," Elemento noted.
One variant in particular, rs62119267, could stratify breast cancer risk according to corticosteroid use. Of the women carrying the risk allele who took corticosteroids, 18.2 percent developed breast cancer, compared to 5.1 percent of those who did not take the medication. There was no difference in breast cancer risk among women homozygous for the reference allele.
According to Elemento, there is great interest in using polygenic risk scores to predict who is at risk of eventually developing disease. His group is working on making such scores actionable, he said, to help decide what to do with a patient with high polygenic risk of a certain disease.
The new findings, he added, suggest that patients with high polygenic risk of breast cancer might want to avoid corticosteroids so as to not exacerbate their risk, and that inhibiting corticoid signaling could potentially mitigate breast cancer risk. He noted, though, that this first needs to be validated as well as further studied.