False Discovery Rate Research Articles

Abstract 3294: Subtype-specific genetic drivers of immune evasion in breast cancer

Abstract Introduction: Immune evasion is a hallmark of cancer and a significant driver of therapeutic resistance. While immunotherapy has shown promise in highly immunogenic cancers, its efficacy in breast cancer (BC) remains limited, particularly in less immunogenic subtypes. We aimed to investigate the prognostic relevance of two immune-related gene signatures—106 genes associated with evasion of immune destruction (EID) and 182 genes linked to evasion of killing by cytotoxic T lymphocytes (ECTL)—across distinct BC subtypes. Methods: Transcriptomic and clinical data from three integrated datasets (GEO, n=1, 965; TCGA, n=1, 069; and GSE96058, n=2, 976) were analyzed. Expression of immune evasion-related genes was stratified by BC subtype using PAM50 classification. Survival analysis was conducted using Kaplan-Meier curves and Cox regression models, stratifying gene expression into high- and low-expression groups. A False Discovery Rate (FDR) correction was applied to ensure statistical robustness. Subtype-specific survival associations were further evaluated with individual gene-level analyses. Results: High mean expression of the ECTL signature was significantly associated with improved overall survival (OS) in triple-negative BC (TNBC) patients (HR = 0.25, 95% CI = 0.16-0.4, p = 8.6e−10, FDR < 1%). Similar trends were observed in HER2-enriched BC but were less pronounced after FDR correction. For the 106 EID genes, high mean expression correlated with favorable OS in TNBC (HR = 0.3, 95% CI = 0.18-0.49, p = 3.4e−7). Gene-level analysis identified essential immune evasion genes, including CXCL10, CXCL9, CXCR4, and JAK3, as robust predictors of survival in TNBC, validated across independent datasets. The new, combined four-gene signature showed significant predictive power for response to immune checkpoint inhibitors (ICI), with an AUC of 0.722 (p = 1.3e−7). Conclusions: Immune evasion-related gene signatures exhibit subtype-specific prognostic value, particularly in TNBC and HER2-enriched BC. A signature based on genes within these signatures represent potential biomarkers for predicting survival outcomes and guiding immunotherapy. Citation Format: Otilia Menyhart, Balazs Gyorffy. Subtype-specific genetic drivers of immune evasion in breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3294.

Abstract 6217: Methylation-derived inflammation phenotypes and anal cancer risk among men with HIV

Abstract Introduction: Inflammation is a hallmark of cancer. DNA methylation (DNAm) derived immune cell type proportions and circulating inflammation-related protein markers are precise measures of inflammatory phenotypes. We investigated the associations of DNAm-derived inflammation-related phenotypes with risks of anal cancer and precancer (high-grade squamous intraepithelial lesions, HSIL) among men with HIV. Method: We matched 22 anal cancer cases to 37 controls with normal anal epithelium and 54 HSIL to 90 controls using propensity score methods. DNAm data were profiled by MethylationEPIC v2.0 array on the pre-diagnostic blood biospecimens and processed by R minfi and ChAMP packages. Immune cell composition, including 4 myeloid (neutrophils, eosinophils, basophils and monocytes) and 8 lymphoid cell sub-lineages (B lymphocytes naïve (Bnv), B lymphocytes memory [Bmem], T helper lymphocytes naïve [CD4nv], T helper lymphocytes memory [CD4mem], T regulatory cells [Treg], T cytotoxic lymphocytes naïve [CD8nv], T cytotoxic lymphocytes memory [CD8mem], and natural killer lymphocytes) were estimated among normalized DNA methylation β values using a reference-based deconvolution algorithm. Proportions of these 12 major immune cell types were estimated to be 100%. Ratios of CD4nv/CD4mem, CD8nv/CD8mem, Bnv/Bmem, CD8/Treg, and neutrophil/lymphocyte were computed. Additionally, 47 DNAm-predicted circulating inflammation-related protein levels were calculated using well-established prediction models. Conditional logistic regression models were built to estimate odds ratios (OR) and 95% confidence intervals (CI) on per standard deviation increase in inflammation phenotypes adjusting for cell type heterogeneities. Results: Higher CD4nv/CD4mem (OR=2.70, 95%CI:1.09-6.70) and lower CD8/Treg (OR=0.56, 95%CI:0.31-0.99) was positively associated with anal cancer risk at nominal P < 0.05. Higher Bnv/Bmem (OR=1.47, 95%CI:1.02-2.12) was linked to increased HSIL risk at nominal P < 0.05. Elevated levels of DNAm-derived circulating protein MMP9 (OR=1.86, 95%CI: 1.20-2.88), C5 (OR=1.78, 95%CI: 1.23-2.57), B2M (OR=1.75, 95%CI:1.16-2.65), CRP (OR=2.39, 95%CI:1.32-4.33), MMP12 (OR=1.66, 95%CI:1.15-2.41), CCL11(OR=1.66, 95%CI:1.15-2.38) , TNFRSF1B (OR=1.85, 95%CI:1.20-2.87), HGF (OR=2.10, 95%CI:1.32-3.35), TGFA (OR=2.38, 95%CI:1.31-4.35), VEGFA (OR=1.82, 95%CI:1.21-2.72), CCL17 (OR=1.93, 95%CI:1.26-2.95); and decreased levels of ADAMTS13 (OR=0.47, 95%CI:0.28-0.78) were associated with increased risk of anal cancer and HSIL (combined analysis) compared to controls at false discovery rate (FDR) <5%. No significant associations were observed when comparing cancers with controls or comparing HSIL with controls, separately at FDR<5%. Discussion: Our findings suggest that inflammation plays a role in anal carcinogenesis. Future research is warranted to validate our results. Citation Format: Shuai Xu, Bryan E. Shepherd, Tebeb Gebretsadik, Anna Junkins, Jirong Long, Qiuyin Cai, Staci L. Sudenga. Methylation-derived inflammation phenotypes and anal cancer risk among men with HIV [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 6217.

Abstract 1929: Metabolomic insights into androgen deprivation therapy, and saliva as a non-invasive surrogate for serum to enable community-based biomarker studies

Abstract Introduction: Androgen deprivation therapy (ADT) significantly alters metabolism in prostate cancer patients. Serum metabolomics are the gold standard for studying these effects; however, saliva offers a non-invasive alternative for biomarker assessment. This study evaluates the impact of ADT on serum and saliva metabolites and the correlation between the two matrices. By evaluating saliva as a potential surrogate for serum, this work explores its application in biomarker analysis of chronic risk factors within community health programs. Methods: Blood and saliva were collected from 20 eugonadal men and 20 men receiving hormonal therapy for prostate cancer, and were processed per standard protocols and stored at -80°C. Untargeted metabolomics using UPLC-MS analysis of polar and non-polar metabolites were performed; and glucocorticoids were quantified using targeted LC-MS/MS. Data were analyzed using MassHunter, METLIN database, and MetaboAnalyst, with log-transformation and by Mann-Whitney tests (False Discovery Rate (FDR)-adjusted P-values < 0.05). Results: In untargeted metabolomics, 1, 198 metabolites were identified across serum and saliva. Comparing ADT to eugonadal serum, 35 metabolites had fold-changes (FC)>2 and 35 had FC<0.5, with 248 metabolites reaching statistical significance (adjusted-P<0.05). Similarly, in saliva, 54 metabolites had FC>2 and 69 had FC<0.5, with 82 metabolites showing significant differences (P<0.05). Comparing serum to saliva across all participants, 248 metabolites had FC>2 and 268 had FC<0.5, with 697 reaching statistical significance (P<0.05). When integrating fold changes (FC>2 or FC<0.5) and significance thresholds (adjusted-p<0.1), we identified 14 upregulated and 20 downregulated metabolites in serum, and 14 upregulated and 28 downregulated metabolites in saliva of ADT patients compared to eugonadal patients; and 247 upregulated and 268 downregulated metabolites in saliva compared to serum among participants. Targeted metabolomics also confirmed strong correlations between serum and salivary cortisol (r=0.612, p<0.001) and cortisone (r=0.618, p<0.001). These findings underscore the potential of glucocorticoids as biomarkers for stress-related metabolic alterations, suggesting that saliva potentially reflects the systemic metabolomic profile. This highlights saliva’s promise for non-invasive biomarker studies in stress and broader community health applications. Conclusion: ADT is associated with significant changes in serum and salivary metabolome, highlighting its systemic and local impact on metabolism. Moreover, targeted metabolomic analysis revealed strong correlations between glucocorticoids in serum and saliva, suggesting saliva may serve as a non-invasive surrogate for serum in biomarker studies to facilitate research in community health programs. Citation Format: Alireza Abdshah, Mohammad Alyamani, Marijo Bilusic, Dipen J. Parekh, Benjamin Spieler, Rakesh Singal, Brandon A. Mahal, Erin Kobetz, Sanoj Punnen, Matthew C. Abramowitz, Jaime R. Merchan, Nima Sharifi. Metabolomic insights into androgen deprivation therapy, and saliva as a non-invasive surrogate for serum to enable community-based biomarker studies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 1929.

Abstract 5103: Digital spatial profling with GeoMx to identify differential protein expression in Non-Hispanic/Latino and Hispanic/Latino Patients with metastatic hormone sensitive prostate cancer

Background: Emerging evidence suggests that there is a complex interplay between ethnicity and disease biology in metastatic hormone sensitive prostate cancer (mHSPC). We examined digital spatial profiling with GeoMx to evaluate differential protein expression between non-Hispanic/non-Latino (NHL) and Hispanic/Latino (HL) patients with mHSPC. Methods: Pre-treatment prostate biopsy specimens were obtained from patients on a clinical trial as part of an IRB approved protocol. The GeoMx Digital Spatial Profiler (DSP) was used to evaluate protein expression in selected regions of interest (ROIs). Differential expression analysis was performed using R (v 4.4.2). To determine significant differences in protein expression, a false discovery rate (FDR)-adjusted q-value was applied and a Log2Fold change (Log2FC) with a q-value < 0.01 was considered significantly differentially expressed. Results: A total of 8 patients (4 NHL and 4 HL) were enrolled at City of Hope including network sites and who had pre-treatment biopsy specimens that were available for digital spatial profiling analysis. We examined the Log2FC between protein expression in NHL and HL patients and filtered for q-values < 0.01. We identified multiple proteins that were significantly differentially expressed. Proteins that were more highly expressed (Log2FC > 1) in NHL patients included CD44, fibronection, and Ki-67. Proteins that were more moderately expressed (Log2FC < 1) in NHL patients included beta-2 microglobulin and. In contrast, HL patients were found to have higher expression in BCL6, CD68, CD80, Foxp3, PARP, and STING (Log2FC < 0). Conclusions: Digital spatial profiling with GeoMx reveals significant differences in protein expression in NHL vs. HL patients with mHSPC. Notably, these include markers of cancer stem cells (CD44), DNA damage pathway proteins (PARP), and both immunoregulatory (Foxp3) and immunostimulatory (STING) proteins. This may have both prognostic and therapeutic implications. Citation Format: Peter D. Zang, Joe Bonner, Alexander Chehrazi-Raffle, Lixin Yang, Natasha Garg, Swapnil Rajurkar, James Shen, Naveen Gupta, Amartej Merla, Sumanta Pal, Raju Pillai, Stephen Gruber, Tanya B. Dorff. Digital spatial profling with GeoMx to identify differential protein expression in Non-Hispanic/Latino and Hispanic/Latino Patients with metastatic hormone sensitive prostate cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 5103.

Abstract 2267: Mapping genetic risk for Barrett's esophagus and esophageal adenocarcinoma to key tissues and transcriptional regulators

Abstract Background: Esophageal adenocarcinoma (EAC), the most common subtype of esophageal cancer in Western countries, typically arises from Barrett’s esophagus (BE), a pre-neoplastic condition marked by replacement of the normal esophageal epithelium with intestinal metaplasia. Genome-wide association studies (GWAS) have revealed a substantial heritable component of BE/EAC susceptibility, but mechanistic understanding remains limited. Identifying tissues, cell types, and transcriptional regulators mediating genetic risk is essential for elucidating etiologic pathways and defining novel molecular targets for translation. Methods: We employed a permutation-based approach to evaluate the enrichment of BE/EAC-associated variants in open chromatin regions (OCRs) across >50 diverse tissues and cell types. For a subset of implicated tissues, we further assessed enrichment in the binding regions of multiple transcription factors (TFs) using ChIP-seq profiles, with complementary analyses based on DNAse-I footprinting data and motif scanning. Multiple testing correction was implemented using the false discovery rate (FDR<0.05). Results: BE/EAC variants were significantly enriched in OCRs of gastrointestinal tissues, particularly fetal stomach (OR = 1.6, P = 6e-04) and fetal large intestine (OR = 1.5, P = 1.5e-03). Enrichments were also observed in muscle (OR = 1.5, P = 9e-04) and fibroblast (OR = 1.4, P = 8.8e-03) cells, suggesting a broader multi-lineage regulatory landscape. Key TFs implicated in subsequent testing included GATA6 (OR = 3.0, P <1e-04) and GATA4 (OR = 2.2, P = 2e-04) in GI tissues; EZH2 in muscle cells (OR = 5.7, P = 2e-04); and EZH2 (OR = 3.5, P <1e-04) and AR (OR = 1.4, P = 1.2e-03) in fibroblast cells. Conclusions: This study indicates that BE/EAC risk variants act through regulatory elements across diverse tissues, with prominent roles suggested in gastrointestinal, muscle, and stromal cells. The significant enrichment signals for transcription factors such as GATA6, GATA4, and EZH2 underscore their regulatory importance in GI-tract development and BE/EAC pathogenesis. These insights enhance our understanding of the cellular/molecular origins and genetic underpinnings of BE/EAC, providing a promising foundation for experimental and translational studies targeting specific transcriptional regulatory networks. Citation Format: Elnaz Naderi, Xu Zhang, Kathryn Graz, Carlos Maj, Johannes Schumacher, Li Yan, Matthew F. Buas. Mapping genetic risk for Barrett's esophagus and esophageal adenocarcinoma to key tissues and transcriptional regulators [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 2267.

Abstract 2004: Integrative bioinformatics approach reveals potential liquid biopsy-based biomarkers for breast cancer metastasis monitoring

Abstract Background: Despite recent advances in screening and therapeutic strategies, breast cancer remains the first leading cause of cancer death among females. The 5-year survival rate is high as 95% at localized stages but drops significantly to 31% once the cancer metastasizes. Therefore, developing predictive tools of metastatic recurrence is an unmet clinical need to design preventive clinical trials. Methods: This study used a public dataset (GEO ID: GSE102484), which is a microarray dataset that contains primary tumor samples (n = 683) sourced from subjects with (n = 95) and without (n = 582) distant recurrence (metastasis) of breast cancer (met and non-met, respectively). Our approach began with two sets of differentially expressed genes (DEGs). The first set of DEGs are from all the samples in the dataset regardless of breast cancer subtypes, while the second set of DEGs are from triple negative breast cancer (TNBC) samples only. To identify subjects with TNBC, basal-like breast cancer samples were selected using PAM50 algorithm with genefu R package. To select potential blood biomarker candidates, we developed an integrative bioinformatics pipeline. Briefly, 1) Differential expression analysis between met and non-met were performed using limma R package. We then selected DEGs with log2 fold change ≥ 0.58 and false discovery rate (FDR) < 0.1, 2) Genes with housekeeping/hematopoietic markers were removed, and 3) The genes that can be detectable in human bloods were selected using the human secretome database from the Human Protein Atlas (HPA). Results: The differential expression analysis of all breast cancer samples identified 134 significantly upregulated genes in met and the selection process with the pipeline resulted in 13 marker candidates including COL10A1, CD93, HPRT1, MMP1, EPHB4, HSPD1, MAP4, CCDC80, MMP11, SAFB2, CALD1, CRISP3, and GAPDH. Subsequent GO enrichment analysis identified functions associated with these markers including angiogenesis, calcium ion binding, and extracellular matrix and structure organizations. Nucleotide metabolism was observed significance in pathway. We then focused on the TNBC samples from the dataset and identified 162 significantly upregulated genes in met from TNBC. Out of the 13 marker candidates from the 134 upregulated genes, 8 marker candidates including CALD1, CD93, EPHB4, SAFB2, HSPD1, MAP4, CCDC80, and HPRT1 were also found in TNBC. Conclusion: Through an integrative computational analysis, we identified 13 markers associated with breast cancer metastasis, of which 8 were specifically linked to metastasis in TNBC. Future studies will focus on validating these biomarkers using liquid biopsy samples from breast cancer patients, potentially offering a less invasive diagnostic approach. Citation Format: Jina Kim, Cynthia Jinno, Hyoyoung Kim, Sunao Tanaka, Takeo Fujii, Sungyong You, Hideki Furuya. Integrative bioinformatics approach reveals potential liquid biopsy-based biomarkers for breast cancer metastasis monitoring [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 2004.

Abstract 1890: Proteomic analysis of colorectal cancer risk across two prospective cohorts

Abstract High-throughput proteomic profiling in prospective cohorts has the potential to uncover biomarkers of cancer risk and biology of cancer development. We investigated prospective associations of up-to 7, 335 aptamers with overall and site-specific colorectal cancer in a case-cohort study within the European Prospective Investigation into Cancer and Nutrition (EPIC) and then evaluated associations in the Atherosclerosis Risk in Communities (ARIC) study. Hazard ratios and 95% confidence intervals were estimated using Prentice weighted Cox-proportional hazard models (stratified by recruitment center, sex, and 5 year age group and adjusted for fasting status, day of blood draw, body mass index, smoking status, daily alcohol consumption, physical activity level, and highest level of education) in 977 incident colorectal (colon, 658; rectum, 319) cancer cases and 5, 057 non-cases under a case-cohort design. The identified associations were then evaluated in ARIC in a cohort design including 271 colorectal (colon, 235; rectum, 36) cancers in 9, 495 individuals using multivariate Cox-proportional hazard models. Overall, 37 aptamers were associated (false discovery rate[FDR] p-value < 0.05) with colorectal cancer in EPIC; data on 27 of these were available in ARIC, 7 of which had consistent directions of association in both studies and met a nominal p-value (0.05) threshold in ARIC (positive associations: ACAA1, ASL, GDF15, IQCF1, TFF3; inverse associations: CLEC3B, COMP). Associations did not change after 5-year follow-up exclusion. In site-specific analyses in EPIC, 66 and 6 aptamers were specifically associated with colon and rectal cancer, respectively at FDR<0.05. Four of these aptamers were nominally (P<0.05) associated with colon cancer in ARIC (inverse association: CLEC3B; positive association: ACAA1, GDF15, TFF3) while a further 29 showed associations directionally consistent with those in EPIC. These large-scale proteomic analyses identified a set of proteins associated with colorectal cancer risk across two independent cohorts. These proteins reflect pathways involved in cell migration and adhesion, colorectal mucosal integrity and growth and differentiation, among others, and may represent novel biomarkers of colorectal cancer risk. Citation Format: Matthew A. Lee, Vivian Viallon, David C. Muller, Ziqiao Wang, Paula Jakszyn, Pietro Ferrari, Giovanna Masala, Domenico Pali, Salvatore Panico, Carlotta Sacerdote, Karl Smith-Byrne, Ruth C. Travis, 8 Rosario Tumino, P. Martijn Kolijn, Roel Vermeulen, Monique Verschuren, Nicholas Wareham, Elizabeth A. Platz, Nilanjan Chatterjee, Elio Riboli, Marc J. Gunter. Proteomic analysis of colorectal cancer risk across two prospective cohorts [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 1890.

Abstract 541: PARP and autophagy inhibition synergy in small cell lung cancer

Abstract Purpose: Small cell lung cancer (SCLC) is a high-grade neuroendocrine carcinoma comprising 15% of lung cancers. First-line treatment with platinum and etoposide chemotherapy—and radiotherapy for limited stage disease—produces good initial response, but most patients suffer treatment-resistant relapse within 2 years. The median survival is under 10 months, with immunotherapy increasing this by about 2 months. This regimen has changed minimally in 3 decades, fueling a need for more effective therapies. Poly (ADP-ribose) polymerase (PARP) inhibitors (PARPi) effectively induce DNA damage in SCLC and have shown potential in this setting, but response is variable, so we aimed to identify mechanisms through which SCLC may be sensitized to PARPi therapy. Methods: CRISPR dropout screens were conducted using the Toronto KnockOut v1 (TKOv1) CRISPR library in the SBC5 and H82 SCLC cell lines, with the PARPi, olaparib, as the selection pressure. DNA sequencing was performed at days 25 and 35 for SBC5 and days 28 and 39 for H82. Top hits were identified by gene dropout in the olaparib condition versus the control, with a false discovery rate (FDR) cutoff of 0.05. Gene ontology analysis was used to identify critical pathways. Stable shRNA knockdown cell lines were generated using lentiviral transduction and validated by Western blot. Cells were treated with olaparib and assayed for viability with CellTiter-Glo 2.0. In wild-type cell lines, therapeutic mTOR activation with MHY1485 was validated by Western blot and therapeutic autophagy inhibition with chloroquine (CQ) or GNS561 was validated by Western blot after 3 hours starvation in EBSS medium. Efficacy of autophagy inhibitors alone and in combination with PARPi was assayed by cell viability, and SynergyFinder+ was used to quantify synergism of the combination. Results: CRISPR screening identified mTOR pathway regulators, including components of the TSC and GATOR1 complexes, and gene ontology analysis indicated downregulation of TOR signaling and upregulation of autophagy as key pathways which confer PARPi sensitivity when lost. TSC1/2 knockdown cell lines exhibit reduced viability after PARPi treatment. MHY1485 treatment blocked autophagy as indicated by LC3B-II accumulation, but combination therapy with MHY1485 and olaparib was antagonistic due to proliferative effects of MHY1485. CQ or GNS561 treatment also blocked autophagy, and SCLC cell lines had varying sensitivity to these agents as monotherapies. Autophagy inhibition synergized with PARPi, sensitizing SCLC cell lines to PARPi therapy. Conclusions: TSC1/2 knockdown sensitizes SCLC cell lines to PARPi in concordance with our model that mTOR downregulation promotes autophagy and cell survival after PARPi therapy. While mTOR upregulation was antagonistic with PARPi and promoted cell growth, autophagy inhibitors were a superior therapeutic approach, synergizing with PARPi in vitro. Citation Format: Tony Yu, Ranya Barayan, Lifang Song, Vidhyasagar Venkatasubramanian, Sree N. Nair, Vivek Philip, Benjamin H. Lok. PARP and autophagy inhibition synergy in small cell lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 541.

Abstract 3678: Liquid biopsy fragmentomics approach for the diagnosis of uterine tumors

Abstract Background: Leiomyomas (LM), also known as fibroids, are common benign tumors of the smooth muscle of the uterus that can cause pain, infertility, and abnormal menstrual bleeding. Diagnosing LM is challenging using clinical indications alone, as they share symptoms with leiomyosarcoma (LMS), a rare, aggressive uterine malignancy with a ∼50% disease-specific survival. Pre-operative distinction between LM and LMS is difficult because these tumors are rarely biopsied before surgery. We hypothesize that a non-invasive circulating tumor DNA (ctDNA) test based on LMS- and LM-specific molecular markers will provide accurate pre-operative diagnosis and guide appropriate surgical treatment. Methods: We previously analyzed point mutations and copy number alterations (CNAs) in ctDNA using deep targeted sequencing and shallow whole-genome sequencing (WGS) from plasma specimens of 7 LMS and 12 LM patients (PMIDs: 29463554, 32232185). To build on this, we performed a new analysis of the first 4 nucleotide sequences (4-mer motifs) on the 5’ end of the cell-free DNA fragments using shallow WGS data. The frequency of 256 possible 4-mer motifs was calculated using R programming and normalized to total reads in each specimen. Two-class differential analysis identified motifs enriched in LM and LMS (false discovery rate < 0.05). Results: Our previous studies demonstrated the feasibility of ctDNA detection in LMS and LM patients. In LMS patients, ctDNA was detected in 6 of 7, with >98% specificity; in LM patients, ctDNA was detected in 6 of 12, with 76% specificity. To increase ctDNA detection sensitivity, we incorporated fragmentomics as a new ctDNA marker. Our new analysis identified 66 significantly enriched and 21 significantly decreased motifs in cell-free DNA from LMS patients compared to LM patients. Motifs associated with DNASE1L3 nuclease activity (e.g., CCCA) were significantly decreased in LM compared to LMS patients. Conclusion: We identified new tumor-specific fragmentomic markers in cell-free DNA from LMS and LM patients. Combining detection of point mutations, CNAs, and fragmentomic patterns could enable a highly sensitive ctDNA assay for accurate pre-operative distinction between LM and LMS. Next, we will validate DNASE1L3 and other nucleases in tissue microarrays of 100+ LM and 200+ LMS specimens by immunohistochemistry, and validate distinct fragmentomic patterns using expanded plasma collections. This research addresses an unmet clinical need for accurate pre-operative diagnosis of uterine tumors. Citation Format: Meagan S. Cobb, Philippe Jolivet, Kristen N. Ganjoo, Deirdre A. Lum, Matt van de Rijn, Joanna Przybyl. Liquid biopsy fragmentomics approach for the diagnosis of uterine tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3678.

Abstract 3599: Proteomic risk factors for prostate cancer: A case-cohort study in EPIC

Abstract Background: Prostate cancer (PCa) is the most prevalent cancer among men in Europe, yet its aetiology is poorly understood. Proteins contribute to the development of carcinogenesis and are also the target of most pharmacological interventions. The examination of associations between plasma proteins and PCa may enhance understanding of the aetiology of the PCa. Aim: Identify proteins associated with PCa risk within a large European prospective cohort, with emphasis on associations with aggressive subtypes of the disease. Methods: We conducted a case-cohort study in the European Prospective Investigation into Cancer and Nutrition (EPIC) study, with data for a sub-cohort of 1, 573 individuals and for 982 incident PCa cases. The SomaLogic® 7K panel was used to measure 7, 363 aptamers (representing 6, 412 unique proteins) in plasma samples drawn at recruitment. PCa diagnosis was ascertained by linkage to cancer and death registry data, and was further stratified into aggressive subtypes based on histological grade, tumour stage and mortality information. The Prentice-weight Cox regression model was applied to estimate associations between individual proteins and PCa risk, with adjustment for body mass index, smoking, alcohol intake and education level. Subgroup analyses examined associations with high-grade, advanced-stage, aggressive and extended lag-time (diagnosis more than 15 years after blood draw) PCa risk. We sought external replication for our significant findings in an independent multi-ancestry cohort, the Atherosclerosis Risk in Communities (ARIC) study (SomaLogic® 5K). Results: Following a median follow-up period of 16.2 years, the sub-cohort developed 79 cases of PCa. ACP3, FLT4, and KLK3 [HRs (95% CI): 1.20 (1.10, 1.31), 1.26 (1.13, 1.40) and 2.28 (1.96, 2.65), respectively] were associated with overall PCa risk in EPIC at FDR (false discovery rate) <0.05. In the subgroup analyses, 12 proteins were associated with high-grade, 9 with advanced stage, and 7 with aggressive PCa risk. Among these, ANKRD1 was associated with approximately a 25% higher risk of all three of these clinically relevant subtypes per SD increase. Twelve proteins were associated with PCa risk diagnosed more than 15 years after blood draw, including ANXA2, OSCAR, and SAT2. Two well-established biomarkers of PCa (ACP3 and KLK3) were replicated in ARIC, together with two novel proteins (ANXA2 and APOC2). Conclusion: Multiple novel circulating proteins were associated with PCa risk, particularly with risk for aggressive subtypes. While some proteins may serve as early biomarkers of PCa, those with evidence from long-lag time highlight underlying aetiological mechanisms in the development of PCa. Citation Format: Zhe Huang, Mahboubeh Parsaeian, Vivian Viallon, Ziqiao Wang, Keren Papier, Trishna Desai, Stephanie Chan, Antonio Agudo, Carlotta Sacerdote, David C. Muller, Domenico Pali, Giovanna Masala, Nicholas Wareham, Raul Zamora-Ros, Roel C. Vermeulen, Rosario Tumino, Ian G. Mills, Nilanjan Chatterjee, Elizabeth A. Platz, Pietro Ferrari, Marc Gunter, Elio Riboli, Tim J. Key, Joshua R. Atkins, Karl Smith-Byrne, Ruth C. Travis, The EPIC SomaLogic Working Group. Proteomic risk factors for prostate cancer: A case-cohort study in EPIC [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3599.

Abstract 6410: Blood DNA methylation in association with immune checkpoint inhibitor related outcomes among patients with head and neck squamous cell carcinoma

Abstract Background: Tumor tissue DNA methylation has been proposed as a potential biomarker for immune checkpoint inhibitor (ICI) response. However, less is known about blood DNA methylation as a potential biomarker for ICI treatment related outcomes in recurrent/metastatic (R/M) head and neck squamous cell carcinomas (HNSCC). Objective: Investigate associations of pre-treatment blood-derived DNA methylation profiles with survival and response outcomes among patients with R/M HNSCC who receive ICI therapy. Methods: Patients with R/M HNSCC were recruited from the Dana-Farber Cancer Institute, Dartmouth Cancer Center, and Rhode Island Hospital in an ongoing prospective study in the U.S. (data cutoff: September 2024). Peripheral blood was collected from each patient prior to start of ICI based therapy. DNA methylation was measured using the Illumina EPIC array V1 and V2. After strict quality control, the top 10% most variable CpG sites (n=67,713) were included in the analysis. Epigenome-wide association study analysis was conducted for overall survival (OS), progression-free survival (PFS), and treatment response (durable benefit, non-durable benefit, and non-benefit) in association with normalized CpG beta-values. Models were adjusted for age, sex, smoking status, and DNA methylation-derived estimated blood cell type proportions. Results: Among the 113 subjects enrolled so far, 59 (52.2%) died during follow-up and 18 (15.9%) experienced cancer progression. Of 36 patients (31.9%) who have remained progression-free survivors, 11 patients showed a complete response to the ICI therapy. The median OS and PFS were 428 and 279 days, respectively. Approximately half of the patients were categorized in the durable benefit group (n=57, 50.4 %), followed by the non-benefit group (n=30, 26.5 %) and the non-durable benefit group (n=26, 23.0 %). We identified two CpG sites (cg04997743 and cg25104727) that were associated with improved OS; and three (cg03090869, cg19041076, and cg22088248) that were associated with worse OS (false discovery rate [FDR] < 0.05). Similarly, six CpG sites (cg04079758, cg10654272, cg14132225, cg14812628, cg20071413, and cg25104727) were associated with improved PFS while a single CpG (cg24138433) was correlated with worse PFS (FDR < 0.05). The CpG sites associated with OS or PFS were annotated to CpG islands chr11:69924338-69925197, chr18:72916107-72917233, and chr19:51505914-51506363. Additionally, these CpG sites were located in genes that are broadly characterized by metabolism-related (ZADH2 and PPT1) or inflammation-related (KLK9 and ANO1) pathways. No CpG site was associated with the treatment response categories (FDR > 0.85). Conclusions: Distinct blood DNA methylation signatures may be associated with ICI therapy-related outcomes in patients with R/M HNSCC. Citation Format: Gyeyoon Yim, Lucas Salas, Kartik Sehgal, Keisuke Shirai, Rondi Butler, Julianna Miller, Hannah Stolrow, Md Saiful Islam Saif, Devin Koestler, John Wiencke, Robert Haddad, Karl Kelsey, Brock Christensen. Blood DNA methylation in association with immune checkpoint inhibitor related outcomes among patients with head and neck squamous cell carcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 6410.

Abstract 2757: Hypermethylation of FAT4 linked to tumor progression in colorectal adenocarcinoma

Abstract Introduction: The gene FAT4 encodes a cell adhesion molecule of the protocadherin family. FAT4 can be linked to cancer progression at multiple levels, including metastasis, autophagy, epithelial-to-mesenchymal-transition, and as a tumor suppressor. Aberrant methylation, a common hallmark of colorectal adenocarcinoma, disrupts normal gene regulation. Purpose: We explored the methylation changes of FAT4 during colorectal cancer progression, using a large patient cohort with methylation microarray data. Methods: Raw data files of a collection of 2,356 colorectal tumor, adenoma, and normal samples from two microarray platforms (Illumina HM450K and EPIC) were processed in R using the minfi and the watermelon libraries. Based on a matrix of β-values, methylated regions and CpG loci were identified using Kruskal-Wallis-test complemented with ROC analysis. For data visualization we employed www.epigenplot.com, a web platform established based on our database. False discovery rate was set at 10%. Results: Based on UCSC RefGene data, 18 of the HM450K and 27 of the EPIC array probes were assigned to the FAT4 gene. All the probes hypermethylated in adenocarcinoma tissues compared to normal ones belonged to the CpG island chr4:126235895-126238930 of the proximal promoter region. When comparing to normal tissues, the TSS200 region had the most prominent hypermethylation (HM450K Δβ =0.28, AUC 0.9; EPIC Δβ = 0.22, AUC 0.85, p <0.0001), while the probes of the open sea CpGs in the protein coding region have shown a slight hypomethylation (HM450K Δβ = 0.1; EPIC Δβ = 0.14, p <0.0001). These changes were also present in adenoma tissues compared to normal ones. Based on KEGG classification, three other genes associated to the mTOR signaling pathway were similarly hypermethylated in the TSS200 region (Δβ >= 0.2), including WNT2, FZD10, and WNT7A. Conclusions: FAT4, a tumor suppressor gene with previously reported downregulation in CRC, shows significant hypermethylation in colon cancer, including the precursor lesions. Our results suggest that detection of these methylation changes in the promoter region of FAT4 might be utilized as a prognostic biomarker and in the early detection of CRC. Citation Format: Dalma Muller, Balazs Gyorffy. Hypermethylation of FAT4 linked to tumor progression in colorectal adenocarcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 2757.

Abstract 2336: Discovering plasma proteins associated with common solid cancer incidence in ARIC

Background: Plasma proteins, reflecting both exogenous and endogenous factors, could serve as the basis for risk stratification to guide decision-making for primary and secondary prevention of cancer in adults. Published studies have investigated the prospective risk of cancers and other chronic diseases associated with Olink-based proteomic data available in the UK Biobank Study. There is a need for further investigations based on alternative measurement platforms, longer follow-up data, and diverse populations. Aim: Identify proteins measured by SomaScan assay® 5K that are associated with risk of solid cancers in the Atherosclerosis Risk in Communities (ARIC) study, a prospective cohort of middle-aged and older Black and White men and women. Methods: The study was conducted based on a cohort of 9,495 individuals for which a total of 590, 416, 22, 97, 271, 88, 136 incident cancers of prostate, lung, liver, kidney, colorectal, pancreatic, and bladder were observed over a maximum follow-up period of 25.9 years. Log2-transformed relative fluorescence unit of 4,955 SOMAmers measured in plasma collected at Visit 2 were adjusted in a linear regression model including proteomic PEER factors, study sites, and ten genetic principal components (PCs). Corrected protein quantifications based on rank-inverse normalization of the residuals were used in the analysis. Incident cancers were ascertained primarily by cancer registry linkage. The Cox regression model and time-dependent coefficient Cox model for age at Visit 2 were applied to estimate associations between individual proteins and each solid cancer risk, adjusting for potential environmental risk factors, race, genetic PCs, and PEER factors. Sensitivity analyses examined associations with extended lag-time (diagnosis more than 5 years after Visit 2) and minimally adjusted Cox model for cancer risk. Results: We identified 32 proteins (19 positively, 13 inversely) for liver cancer risk, 5 proteins (4 positively, 1 inversely) for lung cancer risk in both main and sensitivity analyses at false discovery rate (FDR) <0.05. 2 known proteins (MMP7 and HAVCR1) were identified to be positively associated with kidney cancer, and 2 (ACP3 and KLK3) were identified for prostate cancer. We further identified GSTM1 associated with bladder cancer risk [HR (95% CI): 0.68 (0.56, 0.82)] and found evidence of enrichment of signals for proteomics association with colorectal cancer at nominal threshold P < 0.05 (371 proteins for colorectal cancer, 401 proteins for colon cancer only, and 425 proteins for rectal cancer). Conclusion: We discovered multiple circulating proteins associated with lung and liver cancer risks, and confirmed previously established circulating proteins for kidney and prostate cancer risks. Additional evaluations in independent and large cohort studies and laboratory testing will be necessary to validate our results. Support: NHGRI, NHLBI, NCI, NPCR. Citation Format: Ziqiao Wang, Vernon A. Burk, Nilanjan Chatterjee, Elizabeth Platz. Discovering plasma proteins associated with common solid cancer incidence in ARIC [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 2336.

Abstract 4990: A discovery of novel methylation profiles in head and neck squamous cell carcinoma in relation to pretreatment pain by cancer subtype

Abstract Pain is the initial symptom reported by nearly 75% of individuals diagnosed with head and neck squamous cell carcinoma (HNSCC), yet effective management remains elusive, with up to 80% of patients reporting inadequate pain management despite opioid use. Pretreatment pain is a key predictor of survival and recurrence, but the molecular mechanisms underlying pain in HNSCC are not well understood. This pilot study aims to explore the epigenetic mechanisms underlying pain perception in HNSCC patients by cancer subtype and identify potential biomarkers to improve pain management strategies. DNA methylation levels were generated for individuals newly diagnosed with HNSCC of oral cavity (OCSCC; n=18) or oropharynx (OPSCC; n=8) prior to definitive treatment using the Illumina Infinium MethylationEPIC v2.0 array. Perceived worst pain score was measured using the Brief Pain Inventory-Worst Pain item prior to treatment, categorized according to NCCN guidelines, with patients reporting no pain serving as the reference group. After quality control, M-values for 26 samples and 913, 962 methylation probes were evaluated. Differentially methylated positions (DMPs) associated with pain severity were identified combining subtypes, with hypo- (fold change (FC) < -2) and hypermethylated (FC > 2) sites determined after adjusting for age. Stratified analyses were performed by cancer type. All analyses set a significance level with a false discovery rate < 5%. Pain phenotype and blood collection occurred on the same day, after diagnosis but prior to definitive treatment. The cohort had a mean age of 65.1 ± 11.3 years, 42.3% of those reporting any level of pain were male. Pre-treatment pain was reported by 94.4% (N=17) of OCSCC and 25% (N=2) of OPSCC patients. DMP analysis revealed probe cg08573204_TC21 in LMO1 gene was hypermethylated in OCSCC patients with mild pain versus no pain (p=0.02, Log2FC=1.11). While LMO1 has not been previously linked to pain, its role in adrenergic cell identity suggests potential involvement in pain pathways. In patients with moderate/severe pain, cg18566479_BC21 located in CHDH gene was hypermethylated in OCSCC (p=1.09E-04, Log2FC=1.98). CHDH, associated with metabolic signatures in HNSCC and neuroinflammatory activation in neuropathic pain, showed significant hypermethylation, suggesting its potential as a biomarker for pain management in HNSCC. Differential DNA methylation may be associated and predictive of the development and severity of pain in HNSCC. With future work in larger samples, this research may provide insight into the biological underpinnings of pain and guide cancer symptom management by using DNA methylation profiles to identify patients at risk for pain who may benefit from preventative measures. Citation Format: Monica A. Wagner, Yanning Wu, Naji Ayyash, Fredrick R. Schumacher, Quintin Pan. A discovery of novel methylation profiles in head and neck squamous cell carcinoma in relation to pretreatment pain by cancer subtype [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 4990.

Abstract 2286: Integrating multi-ancestry genomic and proteomic data to identify blood risk biomarkers and target proteins for breast cancer genetic risk loci

Abstract Genome-wide association studies (GWAS) have identified more than 200 risk loci for breast cancer. However, target genes and their encoded proteins in these loci remain largely unknown. In this study, we utilized genetic prediction models for 1, 350 circulating proteins derived from individuals of African (n= 1, 871) and European (n= 7, 213) ancestry to investigate genetically predicted protein levels in association with breast cancer risk among females of African (n=40, 138), Asian (n= 137, 677), and European (n=247, 173) ancestry. We identified 70 blood protein biomarkers associated with breast cancer risk, overall or by subtypes, at a false discovery rate (FDR) <0.05, including 37 proteins encoded by genes located at least 500kb away from any of the known risk loci identified in GWAS. Of the 33 proteins located at GWAS-identified risk loci, associations for 14 proteins were significantly attenuated after adjustment for the index risk variant of each respective locus, suggesting that these proteins may be target proteins for the risk loci. Encoding gene expression levels in normal breast tissue could be genetically predicted for 35 of the 70 identified proteins, and 17 encoding genes were associated with breast cancer risk in the same direction (P <0.05). Our study identified potential protein targets of GWAS risk loci and biomarkers for breast cancer risk and provided additional insights into breast cancer genetics and etiology. Citation Format: Guochong Jia, Jie Ping, Ran Tao, Jirong Long, Lili Liu, Shuai Xu, Stefan Ambs, Mollie E. Barnard, Yu Chen, Ji-Yeob Choi, Yu-Tang Gao, Montserrat Garcia-Closas, Jian Gu, Jennifer J. Hu, Motoki Iwasaki, Esther M. John, Sun-Seog Kweon, Koichi Matsuda, Keitaro Matsuo, Katherine L. Nathanson, Barbara Nemesure, Olufunmilayo I. Olopade, Tuya Pal, Sue K. Park, Boyoung Park, Michael F. Press, Maureen Sanderson, Dale P. Sandler, Song Yao, Ying Zheng, Prisca O. Adejumo, Thomas Ahearn, Abenaa M. Brewster, Anselm J. Hennis, Hidemi Ito, Michiaki Kubo, Eun-Sook Lee, Siew-Kee Low, Timothy Makumbi, Paul Ndom, Dong-Young Noh, Katie M. O'Brien, Andrew F. Olshan, Mojisola M. Oluwasanu, Min-Ho Park, Sonya Reid, Taiki Yamaji, Gary Zirpoli, Ebonee N. Butler, Wei Zheng. Integrating multi-ancestry genomic and proteomic data to identify blood risk biomarkers and target proteins for breast cancer genetic risk loci [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 2286.

Abstract 4629: Plasma exosomal miRNA and disease progression in metastatic breast cancer on CDK4/6 inhibitors: A feasibility study

Abstract Background: Treatment resistance is common among metastatic breast cancer patients receiving CDK4/6 inhibitors (CDKi). Exosomal miRNA play a key role in intercellular communication and gene regulation that may predict treatment outcomes. We hypothesize that patient’s blood exosomal miRNA can uncover biological pathways critical for CDKi response and resistance. Methods: Plasma samples from metastatic breast cancer patients were prospectively collected every 12 weeks during CDKi treatment until discontinuation. Next generation sequencing was performed on isolated plasma exosomes for miRNA profiling. MiRNA independently associated with treatment response (hazard ratio [HR] < 1) and resistance (HR > 1) after Bonferroni correction were identified with mixed-effects Cox regression with time-to-disease progression as the outcome. We generated a miRNA-gene network using miRNet and performed pathway enrichment analysis of target genes using hypergeometric test. Results: Eleven patients provided 26 samples during CDKi treatment. Five (45%) progressed; they were more likely to be hormone receptor negative (40% vs 0%) and a race-ethnic minority (80% vs 17%). After adjustment, 54 and 18 miRNAs were associated with treatment response and resistance, respectively. A total of 48 pathways (false discovery rate [FDR] < 0.05) were identified in relation to treatment outcomes (Table 1). Of these, 25 pathways (52%, e.g. TGF-β, insulin, Jak-STAT) were associated exclusively with response, 2 (4%, long-term depression and oocyte meiosis) were associated with resistance, and 21 (44%, e.g. cell cycle, p53, neurotrophin) were shared by both. The study’s feasibility is demonstrated by the identification of numerous cell-cycle and cancer-specific pathways. Conclusion: This study highlighted the potential of exosomal miRNA as gene expression modulators to uncover key pathways underlying CDKi treatment outcomes. Findings warrant validation in independent studies. Citation Format: Ding Quan Ng, Julia Trudeau, Casey Hudson, Khai Kober, Kord M. Kober, Munjal M. Acharya, Alexandre Chan, Ritesh Parajuli. Plasma exosomal miRNA and disease progression in metastatic breast cancer on CDK4/6 inhibitors: A feasibility study [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 4629.

Effects of Transcranial Magnetic Stimulation on Cognitive-Affective Task-Based Functional Connectivity.

Objective: Repetitive transcranial magnetic stimulation (rTMS) uses electromagnetic fields to induce electrical currents in the superficial cortex, and this electric signal is believed to propagate to functionally connected distal brain regions. We previously reported that rTMS targeting the postcentral gyrus affected resting-state functional connectivity with the posterior insula. The current study investigated whether rTMS targeting the postcentral gyrus would affect task-based functional connectivity (TBFC) with the posterior insula during a cognitive-affective distress task. Methods: Twenty-five healthy participants were assigned to 10 Hertz (Hz) (n = 13) or 1 Hz (n = 12) rTMS groups. Participants received five consecutive days of once-daily rTMS and underwent pre- and post-rTMS magnetic resonance imaging (MRI) scans while completing a cognitive-affective distress task with negative auditory feedback. rTMS coil placement over the right postcentral gyrus was guided with neuronavigation, and TBFC analysis of the MRI data was performed using the bilateral auditory cortex as a seed region-of-interest. Results: There was an false discovery rate (FDR)-corrected significant group-by-session-by-condition interaction in a right putamen/posterior insula cluster: in the distress condition, the 1 Hz rTMS group had significantly weaker (i.e., smaller absolute value) negative TBFC following rTMS (p = 0.005), while the 10 Hz group had no significant effect. Conclusion: This preliminary, proof-of-concept study suggests that rTMS can modulate TBFC in distal brain regions implicated in the neural response to cognitive-affective negative feedback. Future research should investigate whether rTMS can both modulate insula-associated TBFC and improve cognitive-affective task performance or mood outcomes, potentially by increasing the number of rTMS sessions or using different rTMS pulse sequences.

Abstract 3243: Fragmentomic features of plasma cell-free DNA in patients with osteosarcoma

The rapid advancement of liquid biopsy technologies represents a major breakthrough in cancer management, allowing for sensitive detection of tumor-derived molecules and providing real-time insights into disease progression. Osteosarcoma (OS), the most common primary bone malignancy, currently lacks robust biomarkers for therapy response. We hypothesize that a multi-omic and multi-modal approach to identifying OS-specific markers in circulating tumor DNA (ctDNA) will facilitate precise patient stratification and improve treatment outcomes. It has been demonstrated by others that plasma ctDNA in OS patients carry specific genomic markers, such as point mutations and copy number alterations. However, the fragmentomic features of cell-free DNA (e.g., fragment length and nucleotide composition) remain poorly characterized in OS patients. To provide the rationale for studying fragmentomic features of plasma DNA in OS, we sought to demonstrate the feasibility of detecting distinct fragmentomic patterns in the publicly available whole genome sequencing data from cfDNA of 3 OS patients (NCBI SRA accession number PRJNA917431) together with our previously published shallow whole genome sequencing data from cell-free DNA of 7 patients with leiomyosarcoma, 12 patients with leiomyoma, 3 patients with dedifferentiated liposarcoma and 3 patients with well-differentiated liposarcoma (PMIDs: 29463554, 32232185, 34986184). We determined the first 4 nucleotide sequence (i.e., a 4-mer motif) on each 5′ fragment end of plasma DNA molecules. We calculated the frequency of each of the 256 possible motifs and normalized it to the total number of reads in each specimen. We applied multiclass differential analysis of the frequency of these motifs in 5 different types of tumors, and identified 64 motifs significantly enriched in OS (mean frequency difference between groups > 2.5-fold, false discovery rate < 0.05). In OS cell-free DNA, we observed decrease of motifs associated with DNASE1L3 nuclease (e.g., CCCA, CCAG, CCTG) and increase of motifs such as TAAA, AAAA, and TTTT. In the next steps, we will examine the expression of different DNA nucleases in patient tissue specimens by immunohistochemistry and will analyze plasma DNA from a new cohort of 50 OS patients to validate our preliminary findings. We will investigate whether these fragmentomic patterns of cell-free DNA have the predictive and prognostic value in OS patients. Our preliminary results warrant further studies of the frequency and diversity of the 4-mer motifs and expression of different DNA nucleases in OS. Citation Format: Anastasia E. Mackeracher, Joanna Przybyl. Fragmentomic features of plasma cell-free DNA in patients with osteosarcoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3243.

Abstract 3311: Uncovering methylation signatures and pathways associated with resistance to anti-EGFR therapy in colorectal cancer

Abstract Background: Although anti-epidermal growth factor receptor (EGFR) therapy has established efficacy in RAS/BRAF wild type (WT) metastatic colorectal cancer (mCRC), a significant proportion (30-50%) of WT patients exhibit primary resistance to this treatment. Existing biomarker discovery has predominantly focused on tumor-specific DNA mutations, which suffer from variable mutation frequencies among patients. In contrast, DNA methylation represents a common and early event in carcinogenesis. For this reason, methylation markers hold great promise for detecting resistance mechanisms in primary tumors. Several studies have shown an association between DNA methylation patterns and resistance to anti-EGFR therapy in CRC. To validate and build upon these findings, we investigated the methylomes of primary tumors with resistant and responsive phenotypes. Material and Methods: We isolated DNA from FFPE tissue of 40 primary CRC tumors. DNA was subjected to bisulfite conversion and Infinium MethylationEPIC v2.0 BeadChip (Illumina). Subsequently, ChAMP was used to determine the methylation values (β values) and perform differential methylation analysis. The CpG sites were ranked based on the largest Δβ values and the significance of p-values across the patient groups. To ensure the reliability of our findings, we performed a false discovery rate (FDR) evaluation using the Benjamini-Hochberg method, which led to the selection of a subset of differentially methylated CpG sites (DMCs) with the highest likelihood of being genuine associations. Additionally, a gene set enrichment analysis (GSEA) was performed with methylGSA. Results: 48 DMCs were identified with adjusted p-values < 0.05. The affected genes are involved in cancer progression and invasiveness, suggesting a role in a yet-to-be discovered resistance mechanism. The GSEA revealed significant enrichment in 53 KEGG and 33 Reactome pathways, including MAPK-related and Akt signaling pathways. There were 470 significant gene ontology (GO) pathways: 347 in biological processes, 71 in cellular components and 52 in molecular functions. Conclusion and future work: We have found methylation sites associated with resistance to anti-EGFR therapy in CRC patients. In addition to identifying critical DMCs, the pathways uncovered hold potential as a rich source of predictive biomarkers. Next, we will include 24 additional patients. When increasing the sample size, we expect a higher statistical power and an enrichment of CpGs with significant p-values. Afterwards, we will independently validate the most significant CpG sites in a follow-up study using a highly sensitive multiplexed droplet digital PCR assay on additional tissue biopsies. Citation Format: Ana Regina de Abreu, Yann Heylen, Joe Ibrahim, Marc Peeters, Guy Van Camp, Ken Op de Beeck. Uncovering methylation signatures and pathways associated with resistance to anti-EGFR therapy in colorectal cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3311.

Abstract 6053: Down-regulation of interferon gamma response and other immune-related pathways in PIK3CA mutant colorectal cancer

Abstract Background: The PIK3CA gene, encoding the p110α protein, is commonly mutated across different cancer types, including colorectal cancer (CRC). As 18% of CRCs have a PIK3CA mutation, a more comprehensive understanding of its influence on immune processes is needed. Here we sought to isolate and characterize the RNA expression signature of PIK3CA mutant CRC at both the tumor and patient-derived cancer organoid (PDCO) level. Methods: Patient/Originator Specimen (tumor) and Organoid Culture (PDCO) RNA sequencing data of CRC patients with and without PIK3CA mutations were sampled from the National Cancer Institute’s Patient-Derived Models Repository (NCI PDMR) Database. After filtering out patients with coexisting BRAF, KRAS, and NRAS mutations, patients were separated into two groups: wild-type (WT) and PIK3CA mutant (PK). The PDCO dataset contained 15 patients (nine WT vs six PK), and the tumor dataset contained eight patients (four WT vs four PK). Gene Set Enrichment Analysis (GSEA) was performed on the differential expression results comparing these two groups. A cutoff of |log10 false discovery rate (FDR)| ≥ 1.3 was used to determine significance for pathways. Data analysis was performed in R with the following packages: DESeq2, tximport, tximportData, dplyr, readr, magrittr, and ggplot. Results: In the PDCO data set, two immune pathways were significantly up-regulated in WT compared to PK: interferon alpha (IFNα, FDR q-value < 0.001) and interferon gamma (IFN-γ, 0.018). In the tumor data set, more immune pathways were significantly down-regulated in PK compared to WT. PK had a down-regulation in pathways involved in T cell activation and function like IL6 JAK STAT3 signaling (q = 0.009) and IL2 STAT5 signaling (q = 0.003). Interestingly, in the tumor data set, PK had multiple inflammatory pathways down-regulated compared to WT: TNFA signaling via NFKB, allograft rejection, inflammatory response, complement (all q < 0.001), IFN-γ response (q = 0.004). IFN-γ response was investigated further due to it being significantly down-regulated in PK in both PDCO and tumor data sets. Significant enrichment was seen in common genes for both groups as follows: PDE4B (organoid log2FC = -1.24 and tumor log2FC = -3.97), IRF5 (-1.11 and -3.05), BTG1 (-0.753 and -2.72), TNFAIP2 (-0.5 and -2.69), and CD40 (-0.405 and -2.09). IRF5 has been implicated in M1 macrophage polarization, CD40 in CD8 T cell activation, PDE4B and TNFAIP2 in inflammation, and BTG1 in B cell regulation. Downregulation of these genes hints at PK possibly inhibiting important immune cell markers and the possible mechanisms through which PK is acting. Conclusions: These analyses indicate mutated PIK3CA may play a role in suppressing the immune response in CRC, which can be seen even in PDCOs. Further investigation is needed to determine the mechanisms by which PIK3CA is suppressing the IFN-γ response and other immune-related pathways. Citation Format: Alec B. Cornelio, Alexa E. Schmitz, Xingqi Shen, Shirsa Udgata, Cheri A. Pasch, Dustin A. Deming. Down-regulation of interferon gamma response and other immune-related pathways in PIK3CA mutant colorectal cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 6053.