Variant Expression Patterns in Breast Tumors

Evidence

This observational genetic association and tumor-transcriptomic study examined whether rs140068132, a germline variant in the 6q25 region upstream of estrogen receptor 1 (ESR1), was associated with gene expression in breast cancer tissue. The study focused on a subset of women with invasive breast cancer recruited through the Peruvian Genetics and Genomics of Breast Cancer Study at the Instituto Nacional de Enfermedades Neoplásicas in Lima, Peru. Women were eligible for the parent cohort if they had been diagnosed with invasive breast cancer in 2010 or later and were 21 to 79 years old at diagnosis. As of October 2025, the broader cohort included 2,156 recruited participants. (pmid:42777143)

RNA sequencing was performed on 270 breast tumors from the Peruvian study. Investigators classified tumor-intrinsic subtypes using an in-house adaptation of the PAM50 method. The adaptation accounted for the cohort's estrogen receptor-status distribution relative to the dataset used in the original PAM50 publication; before classification, the investigators sampled a training subgroup with a clinicopathologic distribution similar to that of the study cohort. Thus, the reported expression analyses were based on sequenced tumor tissue and an internally adapted intrinsic-subtype assignment approach rather than on a single pooled tumor category. (pmid:42777143)

The analytic aim was to assess whether rs140068132 acted as a potential cis expression quantitative trait locus across intrinsic subtypes. Models evaluating genotype-expression relationships were adjusted for age at diagnosis, Indigenous American genetic ancestry, and PAM50 subtype, and included an interaction term between tumor subtype and single-nucleotide polymorphism genotype. That interaction was central to the design because it tested whether the genotype-expression association differed across subtypes rather than treating subtype as adjustment alone. The analysis therefore compared expression patterns by rs140068132 genotype within the modeled subtype framework. (pmid:42777143)

The principal reported result was subtype-specific: the protective rs140068132-G allele was associated with increased expression of ARMT1, RMND1, and CCDC170 in HER2-enriched tumors. The source characterizes these genes as cis-genes in relation to the variant and reports that the expression pattern was limited to HER2-enriched tumors in the study analysis. No effect estimate, confidence interval, or p-value for these gene-expression associations was located in the evidence provided, so the magnitude and precision of each association cannot be stated here. (pmid:42777143)

The study situates this tumor-expression analysis in prior work on breast cancer susceptibility. It reports that rs140068132 had previously been associated with decreased breast cancer risk and with greater protection against hormone receptor-negative and HER2-positive subtypes. It also reports that the association was independent of other named breast cancer-associated variants in the 6q25 locus in prior conditional analyses. Those earlier risk observations provide context for studying tumor expression, but the present tissue analysis measured genotype-associated expression rather than a clinical endpoint. (pmid:42777143)

The variant was described as being 56 kb upstream of ESR1 within a known regulatory region. The authors interpreted the observed expression pattern as consistent with a cellular-context-dependent regulatory element at 6q25 that may modulate nearby gene expression. This is a biological interpretation of an association in tumor tissue, not direct evidence that the variant changes transcription-factor binding, enhancer activity, or gene regulation in the sequenced tumors. (pmid:42777143)

Analysis — Subtype-specific tumor expression

The observed pattern is consistent with the reported design and selection process: among sequenced breast tumors from Peruvian patients, rs140068132 genotype was associated with expression of ARMT1, RMND1, and CCDC170 in the HER2-enriched subtype after modeling age at diagnosis, Indigenous American genetic ancestry, subtype, and a genotype-by-subtype interaction. Including ancestry and subtype in the models addresses measured differences relevant to this cohort, while the interaction term makes the subtype-specific framing explicit. However, adjustment for these recorded variables does not convert the analysis into a causal test of how the allele affects gene expression. (pmid:42777143)

The distinction between tumor subtype and germline genotype is especially important. The analysis used tumor RNA and assigned intrinsic subtypes from tumor expression data, so its result concerns a molecular pattern in collected tumors, not a direct measurement of gene regulation in normal breast tissue or a clinical outcome. The report's proposed cis-regulatory explanation is biologically bounded: it identifies a hypothesis compatible with the location of rs140068132 and the observed subtype-stratified associations, without demonstrating the intervening molecular mechanism. (pmid:42777143)

Limitations

The retrospective design cannot establish causation. This was an observational association study of available breast tumors from a Peruvian breast cancer cohort, and selection into the sequenced tumor set, availability and characteristics of tumor tissue, subtype composition, and genotype distribution may affect which associations were detectable. The measured outcomes were tumor-gene expression levels and intrinsic subtype classifications, not direct functional perturbations of the variant or clinical outcomes. (pmid:42777143)

The authors describe the cross-subtype genotype-expression analysis as explorative because the sample size and genotype distribution limited statistical power for subtype-specific effects. They specifically characterize the CCDC170 association as borderline and subtype-specific, warranting cautious interpretation. The source also notes that rs140068132 is underrepresented in publicly available population-level and phenotype-based databases, which restricts broader cross-database analyses. These constraints limit certainty about reproducibility and the generalizability of the reported associations beyond the studied tumors and cohort. (pmid:42777143)