Abstract

Abstract Background: Acquired, somatic TP53 mutations leading to loss of p53 tumor suppressive functions are the most common genomic alteration across many cancer types. Breast cancer (BC) is the most common malignancy in females, and 50-75% of estrogen receptor (ER)-negative BC (with or without HER2 amplification) and 10-15% of ER+ BC acquire TP53 alterations. Pathogenic germline variants (PGVs) in TP53 occur in approximately 1:5000 people and lead to the Li Fraumeni Syndrome (LFS) in which individuals have an up to 90% risk of developing cancer over their lifetimes. The most common cancer in females with LFS is BC which occurs at a median age of onset of 30-35, nearly 30 years younger than sporadic BC. The distribution of BC subtypes is significantly different in LFS-BC patients compared to sporadic BC, with over half of LFS-BC being HER2+ BC and over 90% being ER+. The characteristics of LFS-BC, namely early onset and predominance of HER2+ disease, suggest different tumorigenesis mechanisms compared to sporadic breast cancer that may have implications for early detection, interception and treatment. Patients and Methods: We performed deep phenotyping on 93 LFS patients with 130 LFS-BC compared to 198 control patients with 209 nonLFS-BC. Controls were defined as having BC diagnosed age<50 and negative germline cancer susceptibility gene testing. We obtained clinical cases and/or blocks for histopathological analysis. DNA and RNA was extracted from available FFPE tumor blocks and analyzed using a targeted capture panel with copy number backbone-based DNA sequencing and RNAseq of ribosomal RNA depleted total RNA. A subset of LFS-BC underwent whole genome sequencing. Finally, slides underwent both single and multiplex immunohistochemistry (IHC) of immune markers. This series includes results of 22 LFS-BC, including seven cases of LFS ductal carcinoma in situ (LFS-DCIS) and 15 invasive ductal carcinomas (LFS-IDC). In addition, matched germline blood for all cases and surrounding (n=15) and/or contralateral (n=11) normal was analyzed. Finally, 27 non-LFS early onset breast cancer controls with matched blood normal were analyzed by the same pipeline. Results were also compared to sporadic BC from The Cancer Genome Atlas (TCGA). Results: Of the 137 adult women in our LFS cohort, 93 (68%) had developed at least one BC, with a total of 110 LFS-BC. The prevalence of BC in our cohort was nearly identical to the prevalence in the National Cancer Institute (NCI) public LFS database (69% of 698 females, p=0.82). The median (IQR) age of first BC diagnosis in our cohort was 36 (29,45) years, similar to the public cohort with a median (IQR) of 31 (27,38) years at first BC diagnosis. Of the 93 adult BC patients in our LFS cohort, 60 (64%) had developed synchronous or metachronous bilateral BC. In contrast, only 36 (18%) of the early-onset nonLFS-BC control cohort developed more than one BC (p<0.0001). LFS-BC was more commonly HER2+ compared nonLFS-BC (51% vs 20%, p<0.0001) Sixty pathological cases from LFS-BC were obtained including 40 (66%) invasive carcinomas, 19 (31%) pure DCIS, one malignant phyllodes. The majority of the invasive carcinomas were ductal (35, 88%) the other five were lobular, mixed ductal and lobular, and micropapillary types. All 40 invasive carcinomas were high grade, with associated DCIS in 27 (68%), lymphovascular invasion in 16 (40%), and sclerotic stroma in eight (20%). The majority of DCIS (either pure or with invasive carcinoma) was high grade (29, 63%) with comedonecrosis (19, 41%), and with associated calcifications (16). High-risk lesions such as radial scar, papilloma, atypical ductal hyperplasia, and lobular neoplasia were present in seven tumors. High p53 IHC nuclear expression (>50%) was seen in the majority of tumors from patients with missense mutations; whereas no protein expression (<1%, null pattern) was seen in majority of tumors from carriers of large exonic deletion, truncating, and splicing mutations in TP53. Using p53 IHC and sequencing results, we found evidence of biallelic loss of TP53 (fully mutant p53) in 21 of 22 (95%) LFS-BC, two of 15 surrounding normal (13%) and no contralateral normal breast tissue. Nearly all LFS-BC underwent biallelic loss of TP53 via focal copy neutral or deletion loss of heterozygosity (89%) or a somatically acquired TP53 mutation (9%). No recurrent oncogenic variants were identified except ERBB2 (HER2) copy number amplification. In contrast, recurrent oncogenic alterations in PIK3CA, GATA3 and MAP3K1 were seen in early onset nonLFS-BC, similar to sporadic BC in TCGA. Compared to nonLFS-BC, LFS-BC had lower or similar tumor mutational burden, copy number burden and aneuploid chromosome arm number. LFS-BC were instead marked by elevated focal segments of loss of heterozygosity (LOH) and allelic imbalance. Accordingly, copy number signature analysis from whole genome sequencing showed nearly all LFS-BC had undergone chromothripsis (CN5) with focal LOH on a diploid background (CN9). We identified 5115 significantly differentially expressed genes comparing LFS-BC to normal breast tissue from LFS patients. Using Gene Set Enrichment Analysis (GSEA), we found most significantly up and down regulated pathways in LFS-BC were similarly regulated in non-LFSBC compared to normal breast tissue. Notably, the apoptosis pathway was significantly down-regulated in LFS-BC but up-regulated in HER2+ early onset nonLFS-BC, likely reflecting a loss of p53 regulation of genes such as BAX and TP53I3. In addition, cholesterol homeostasis and androgen response pathways were significantly up-regulated in LFS-BC; whereas they were significantly down-regulated in both ER+Her2- and HER2+ non-LFSBC. We then performed multiplex IHC on ER+ TP53WT nonLFS-BC (n=7), nonLFS-TNBC (n=16), and LFS-BC (n=33). We found that CD8+ T cell infiltration in LFS-BC was comparable to non-LFS triple-negative BC (TNBC), a tumor type which responds well to immune checkpoint blockade therapy, and higher than in ER+ nonLFS-BC. CD20+ infiltration in LFS-BC was lower than both TNBC and ER+ nonLFS-BC. Using single stain IHC and comparing an LFS breast tumor to its surrounding normal, we found that FoxP3+ Tregs, CD68+ and CD20+ cells were enriched in the tumors compared to surrounding normal breast tissue. Finally, we compared genomic, transcriptomic and immune cell differences between DCIS and IDC in LFS. Interestingly, all genomic instability measures and gene expression profiles were similar, including focal LOH and allelic imbalances. However, CD4+ and CD8+ T-cells and CD20+ B-cells became more excluded from the tumor tissue with increasing stage of tumor. Using the ACT activity assay, we found that there was significantly higher Ki67+ and Granzyme B in LFS-DCIS as compared to LFS-IDC indicating active T-cell killing. Finally, FoxP3+ T-regulatory cell and CD68+ macrophages were lower in LFS-IDC compared to LFS-DCIS. Conclusions: Overall, our data suggest that TP53 related breast tumor formation, at least in the germline state, requires biallelic loss of TP53 as a first step. Loss of p53 tumor suppressive function in the female breast leads to focal genomic instability events, possibly with selection for ERBB2 amplification. p53 deficiency in LFS-BC leads to failure to induce apoptosis but maintained cell cycle arrest. LFS-BC resemble sporadic TNBC with high CD8+ T-cells but low CD20+ B-cells. Genomic instability and p53 transcriptional dysfunction occur very early in tumor formation; whereas the immune microenvironment evolves to be more tumor supportive during the progression from DCIS to IDC in LFS-BC. This work provides the first insights into the mechanism of germline TP53 mutation driven breast tumor formation in humans. Citation Format: Kara N. Maxwell. Distinct genomic and immunologic tumor evolution in germline TP53-driven breast cancers [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 NG02.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call