Abstract
Malignancy is defined by a set of abnormal biological capacities, termed the hallmarks of cancer. Decades of histopathologic assessment and molecular profiling of human tumors have demonstrated there are multiple ways cells can acquire each hallmark. As a result, tumors with the same clinical characteristics can vary dramatically across individuals and these distinct molecular vulnerabilities can have important prognostic and therapeutic implications. It is unclear when these differences originate.
Oncogenic aberrations are acquired within the context of germline genomes which differ across individuals at millions of polymorphic sites, but the role of germline variants in somatic evolution remains poorly understood. The most compelling example is that deleterious germline variants in BRCA1 and, to a lesser extent, BRCA2 are preferentially associated with the development of TNBC, implying germline variants sculpt specific subtypes of disease. The mechanistic basis for this preference is incompletely characterized. Additionally, germline variants that upregulate the mTOR pathway are associated with further deregulation of mTOR via somatic PTEN loss-of-function. Moreover, pathogenic germline variants in cancer predisposition genes promote somatic bi-allelic inactivation in a lineage-dependent manner and rare variants across a diverse list of genes can modulate somatic mutational processes. In prostate cancer, germline variants can modulate genomic stability, tumor-specific DNA methylation and gene regulation at the transcriptional and translational levels. Pathogenic germline variants in genes related to telomere and mitotic function increase the risk of sarcoma. Differences in breast cancer subtype frequencies across ancestral populations further suggest germline contributions. These data point to an underappreciated role of the germline genome in somatic tumor evolution.
Various lines of evidence suggest that avoidance of the adaptive immune system is another strong determinant of which somatic mutations persist within a tumor. It remains less clear how germline differences influence immunoediting. Levels of interferon signaling and cytotoxic T-cell infiltration are estimated to be 15-20% heritable. Generally, germline variants have not been considered a good source of immunogenic epitopes as cytotoxic response should be dampened by central tolerance. However, non-mutated immunogenic epitopes have been identified in genes such as ERBB2 in breast and ovarian cancer and H4 histone in prostate cancer, amongst others. Antigens with weak binding affinity for MHC receptors can escape central tolerance and elicit an immune response. Tissue-restricted post-translational modifications can also circumvent central tolerance. Peripheral self-reactive T cells are present at similar frequencies to T cells specific to foreign antigens but are held in an anergic state by regulatory T cells (Treg). However, Treg depletion in healthy mouse models led to the natural occurrence of self-reactive CD4+ T cells. Further, there is mounting evidence for innate-like T-cell populations within mouse and tumor malignancies that have increased propensity for self-reactivity. Altogether, these data suggest that under specific circumstances and disruptions to immune homeostasis, a subset of T-cells may respond to germline-derived epitopes during tumorigenesis.
Building on these observations, we sought to investigate whether germline variants sculpt somatic evolution by mediating immunoediting. Specifically, we hypothesize that the burden of germline-derived epitopes in recurrently amplified driver genes may select against gene amplification. This is because amplification of a gene with a high burden of germline-derived epitopes would increase epitope availability, likelihood of epitope presentation and immune-mediated cell death. Instead, immune pressures may select for amplification of an alternate driver gene with a lower germline-mediated epitope burden. Motivation for this hypothesis draws upon insights from the immuno-oncology and population genetics fields. First, in a subset of tumors, somatic tumor mutational burden (TMB), specifically clonal TMB, is positively associated with immunotherapy response. Second, recent successes with cancer vaccines developed from non-somatically mutated, tumor-specific overexpressed peptides, i.e. HER2, TERT and WT1, suggests germline epitopes contribute to immune response. Finally, genome-wide association studies have demonstrated the power of aggregating germline-variants with weak effects into a unified risk score, i.e. polygenic risk score (PRS). Thus, we hypothesized aggregating germline-derived epitopes, with weak individual effects, akin to PRS, may provide a powerful metric to assess germline-mediated immunoediting.
We addressed this question in breast cancer for three reasons. First, the well-characterized link between BRCA1 and TNBC susceptibility, along with high heritability estimates (~31%), suggests the germline genome plays a role in shaping breast cancer evolution. Second, breast cancer is one of the most extensively sequenced cancer types with sizeable cohorts spanning the full continuum of disease, from pre-invasive lesions to primary tumors and metastatic disease. Finally, oncogenic amplifications define five prognostic breast cancer subtypes (HER2+ and four ER+/HER2-) which are established early, evidenced by their identification in premalignant ductal carcinoma in situ (DCIS). Thus, breast cancer provides an optimal proof-of-concept for studying this phenomenon.
We leveraged paired tumor and normal sequencing data from 4,918 primary and 702 metastatic breast cancer patients as well as somatic genomic profiles from 341 patients with DCIS and evaluated the relationship between germline-derived epitope burden (henceforth referred to as GEB) and subtype commitment, defined by the acquisition of focal oncogenic amplifications. As proof of concept, identified two immunogenic peptides derived from the germline sequence of HER2 and hypothesized that the ability to present either peptide, i.e. possess MHC class I alleles that can bind and present GP2 or E75, would be negatively associated with developing HER2+ breast cancer. Indeed, we discovered that individuals that possess MHC class I alleles that can bind and present GP2 or E75 are significantly less likely to develop HER2+ breast cancer compared to other breast cancer subtypes. Beyond GP2 and E75, we found that individuals with a high GEB in ERBB2, encoding HER2, are significantly less likely to develop HER2+ breast cancer. The same negative association between epitope burden and somatic amplification is observed for four recurrent amplicons observed in high risk of relapse ER+ breast cancers. No association was observed between somatic amplifications and GEB in metastatic breast cancer suggesting a subset of tumors are able to overcome immune-mediated negative selection. Tumors that do are more aggressive and exhibited microenvironments depleted of lymphocytes, consistent with “immune cold” tumors. In DCIS, prior to immune escape, high GEB was negatively associated with progression to invasive breast cancer suggesting GEB is protective in the pre-invasive setting.
These data indicate that supposedly “benign” germline variants with little to no functional genic effect, may, in aggregate, sculpt breast cancer subtypes and disease aggression via immunoediting. Exploiting germline-mediated immunoediting may inform the development of biomarkers that predict risk of progression to invasive breast cancer and refine risk stratification within invasive breast cancer subtypes. Further, germline-mediated immunoediting points to a broad source of currently underappreciated immunogenic antigens, motivating a potential new avenue for developing cancer vaccines with desired properties, including being clonal and present in entire subgroups of disease.
Citation Format: Kathleen Houlahan, Aziz Khan, Noah Greenwald, Cristina Sotomayor Vivas, Robert B. West, Michael Angelo, Christina Curtis. Germline-mediated immunoediting sculpts breast cancer subtypes and metastatic proclivity [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr NG01.