Background Circulating tumor-associated cells (CTACs) are rare among peripheral blood nucleated cells (PBNCs), posing a challenge to use them as Multi Cancer Early Detection (MCED) tool. We developed an attention-gated U-Net model for pixel-level CTAC discrimination that preserves key morphological and fluorescence details. Methods Model suitability was explored in asymptomatic individuals (n = 428) and patients with advanced solid tumors (n = 354). Clinical performance was assessed in case-control cohort of therapy-naive stage I/II cancers (n = 185), benign conditions (n = 129), and asymptomatic (n = 111) individuals, followed by validation across four prospective studies on distinct patient populations: recurrent cancer with low tumor burden (n = 224); solid tumors in peri-operative setting (n = 17); suspected cancer (n = 259); and asymptomatic population (n = 7183). PBNCs were isolated using blinded peripheral blood specimens, stained with EpCAM/Hoechst33342, and imaged. Pathologists’ review established ground truth annotations. The U-Net pipeline encoded spatial features via convolutional and pooling layers to generate pixel-wise segmentation masks for CTAC identification. Sensitivity was determined by CTAC detection in cancer specimens and by their absence in healthy or benign samples. Results The model had 90.68% sensitivity and 99.53% specificity in exploratory study. In case-control cohort, sensitivity was 88.65% in benign conditions and > 99.9% in asymptomatic individuals, while specificity was 78.95% in benign conditions. Among four prospective studies, sensitivity was (a) 91.96% in pretreated low tumor burden patients; (b) 100% in pre-surgery, and 29.41% in post-surgery specimens; (c) 96.34% Positive Predictive Value (PPV) and 32.35% Negative Predictive Value (NPV) for diagnostic triaging; and (d)11% PPV and 99.97% NPV for MCED in asymptomatic individuals. Conclusion The attention-enhanced U-Net achieved robust performance in CTAC detection across case-control and prospective cohorts, supporting its clinical utility for cancer detection.
3531 Background: Colorectal cancer (CRC) is a heterogeneous disease with distinct molecular and clinical differences between right- and left-sided tumors. This study analyzes these variations to understand their impact on tumor behavior and treatment strategies. Methods: A total of 445 colonic tumor samples (132 right-sided, 313 left-sided) were profiled to assess mutations, amplifications, and fusions in key cancer-related genes along with targeted transcriptome analysis of 20,802 genes in a subset using semiconductor based next-generation sequencing (NGS) platform at Datar Cancer Genetics. Immunotherapy biomarkers (TMB, MSI, and PD-L1 22C3 TPS) were analyzed in a subset. Results: Right-sided and left-sided colon cancers exhibit substantial molecular heterogeneity, driven by distinct genetic and epigenetic alterations (Table 1). Right-sided tumors were more frequently associated with MSI and had statistically significant higher incidence of BRAF mutations. KRAS mutations were frequently observed in both right-sided and left-sided tumors at equal rates. ERBB2 amplifications were exclusive to left side tumors, whereas oncogenic ERBB2 mutations were equally distributed. Located around ERBB2, PGAP3 gene co‐amplification too was exclusive to left sided tumors. TFE3 alterations were absent from left sided tumors and common on right side. TP53 mutations, though more common in left-sided tumors, the difference was not statistically significant. Gene expression profiling of a subset, including 103 left-sided and 41 right-sided colon tumors, revealed activation of the Wnt / β-catenin signalling pathway, RAS/MAPK pathway, TGF-β signalling pathway, and immune-related pathways, though these differences were not statistically significant, suggesting that while specific drivers may differ—such as the predominance of APC mutations in left-sided tumors (56.8% vs 37.9%) leading to WNT activation and the higher incidence of RSPO2/3 fusions (7.1% vs 1.7%) in right-sided tumors -eventually some pathways are commonly implicated in colorectal cancer biology. Conclusions: Existing therapies like ICIs, HER2 inhibitors, and emerging molecules such as RSPO2/RSPO3 inhibitors could have differing impact based on tumor sidedness. Integrating these distinctions into drug development and clinical trials holds potential to optimize treatment outcomes. Molecular profiles of right- and left-sided colon tumors. Gene Right (%) Left (%) p-Value(Chi-square test) TP53 64.5% 73.2% 0.075644 APC 37.9% 56.8% 0.001354 KRAS 50.0% 43.6% 0.220194 BRAF 18.8% 3.0% 0.00001 TFE3 9.5% 0% 0.109087 ERBB2 mutation 2.3% 2.0% 0.80941 ERBB2 amplification 0% 5.9% 0.023006 PGAP3 amplification 0% 7.1% 0.673427 RSPO2/3 fusion 7.1% 1.7% 0.827207 Immunotherapy Biomarkers TMB 10-14 24.7% 29.3% 0.458066 TMB ³15 15.6% 7.6% 0.059925 MSI-High 8.1% 3.4% 0.066213 PD-L1 Positive 15% 5.6% 0.012571
833 Background: Targeting signaling pathways in colorectal cancers is a common therapeutic strategy. In addition to targeted drugs, non- anticancer drugs approved for use in other health conditions can inhibit key pathways. De novo tumor transcriptome analysis can identify patient-specific up- and down-regulated transcripts, based on which the corresponding repurposed drugs can be used in synergistic combinations with standard of care (SoC) regimens. Methods: This retrospective study was conducted on a real-world cohort of 459 colorectal cancer cases (218 females, 241 males; median age 56 years) who had taken Exacta multi-omics tumor profiling for selection of personalized treatments. Transcriptome analysis was performed as part of the evaluations to determine up and down-regulated gene transcripts for selecting targeted, endocrine, and cytotoxic anticancer agents as well as non-anticancer (repurposed) drugs and phytochemicals that could be used for tandem targeting of the dysregulated signaling pathways, as indicated in the table below. Results: Transcriptome analysis of this cohort revealed dysregulation of ≥1 pathway in 391 patients (85.2%). MAPKs (n = 244, 53.2%) and MMPs (n = 210, 45.8%) were most frequently upregulated. Celecoxib (n = 292, 63.6%), Atorvastatin (n = 255, 55.6%), and Doxycycline (n = 208, 45.3%) were the most frequently indicated repurposed drugs for targeting the identified dysregulated pathways. Potential indications for ≥1 repurposed drugs were seen in 353 patients (76.9%). Conclusions: Several cellular signaling pathways are known or potential therapeutic targets in cancers. The present study indicates that it may be possible to use multi-omics analyses to create personalized treatment strategies using a synergistic combination of repurposed drugs to potentiate the action of SoC systemic anticancer agents in colorectal cancer. Biomarkers for non anti-cancer drugs selection. Drug Biomarker/Gene Expression Aspirin PTGS2 (COX2) Atorvastatin HMGCR, MAPK Bromelain PTGS2 (COX2) Celecoxib PTGS2 (COX2), FZD, WNT, ΜΑΡΚ Chloroquine HMGB1 Curcumin MMP, BIRC5, BCL2 Doxycycline MMP Metformin MMP, BCL2, ΜΑΡΚ Quercetin FZD, WNT Resveratrol MMP, BCL2, PCNA, PTGS2 (COX2) Vitamin C SLC2A1 (GLUT1) Artesunate FZD, WNT, MMP, BCL2 Cannabidiol MMP Epigallocatechin gallate MMP, MAPK Mebendazole XIAP, MMP, MAPK
e15170 Background: Despite significant morphological overlap between sub-groups, significant clinical heterogeneity is the hallmark of NENs. We sought to examine the molecular profiles of NENs to demonstrate the molecular heterogeneity that underpins differences in tumor biology. Methods: Tissue samples from 76 patients with NENs were analyzed, encompassing six WHO categories and various anatomic sites of origin (ASO). Molecular profiling was conducted using a semiconductor-based NGS platform at Datar Cancer Genetics. Mutation frequencies and pathway alterations were evaluated across WHO categories. Results: The cohort was distributed across WHO categories: NET grade 1 (N = 1), NET grade 2 (N = 9), NET grade 3 (N = 8), small cell NEC (N = 25), large cell NEC (N = 18), and mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) (N = 5). An additional 10 tumors were classified as poorly differentiated NEC. Lung was the most common ASO (37%), followed by the pancreas (16%), gastrointestinal tract accounted (15%), female genitourinary tract (8%) and other rare sites contributed the rest. PD-L1 22C3 analysis (N = 42) showed negative status in 90% of cases, and 10% of cases showed positivity, with positive cases being limited to NECs. MSI status was stable in all cases (N = 40). High TMB (>10 muts/mb) was observed in 27% of samples (N = 13/48).High TMB was restricted to NECs, except for two cases of pancreatic NETs showing high TMB. Molecular profiling of the entire cohort together showed TP53 muattions as most frequent alterations (49%), followed by RB1 mutations (17%), CTNND2 amplifications (16%), RICTOR (11%), and MYC (10%) amplifications, along with NF1 mutations (9%). The PI3K/AKT/mTOR pathway was altered in 25% of tumors, and MAPK/ERK alterations in 16% of cases. PIK3CA , KRAS , and RB1 alterations were restricted to NECs and absent in NETs. Except for TP53 acting as a sole driver in 5% of cases, no two tumors shared an identical molecular profile, highlighting the pronounced molecular heterogeneity even within the same grade and organ of origin. Conclusions: We highlight the significant molecular heterogeneity among NENs, even within the same grade and ASO. These findings emphasize the necessity of integrating molecular profiling into the clinical management of NENs to enable personalized therapeutic strategies. Further studies in larger cohorts may shed added light into further characterisation of NENs. Incidence of genomic alterations across neuroendocrine tumor grades/subtypes. Gene NETGrade 1/2 NETGrade 3 Small cell NEC Large cell NEC MiNEN SNVs /InDels TP53 10.0% 37.5% 56.0% 55.6% 60.0% RB1 -- -- 36.0% 11.1% -- NF1 -- 12.5% 4.0% 16.7% -- KRAS -- -- -- 16.7% 20.0% PIK3CA -- -- -- 11.1% 20.0% CDKN2A 10.0% 12.5% 4.0% -- -- RET 10.0% 12.5% -- -- -- BRAF -- -- -- -- 20.0% Amplifications MYC -- -- 4.0% 16.7% -- CTNND2 -- -- 16.0% 11.1% 20.0% RICTOR -- -- 12.0% 11.1% 20.0% PIK3CA -- -- 20.0% -- 20.0% FGFR1 -- -- -- 11.1% -- PDGFRA -- -- -- 5.6% 20.0% ERBB2 -- -- -- 5.6% --
e18058 Background: HNCs are a diverse group of malignancies characterized by genetic, epigenetic, and proteomic alterations that drive tumor progression, treatment resistance, and poor outcomes. Methods: Molecular profiling of 273 HNC samples was conducted using a semiconductor-based next-generation sequencing (NGS) platform at Datar Cancer Genetics. A subset of 106 cases underwent targeted transcriptomic analysis of 20,802 genes. Additionally, PD-L1 (22C3) IHC, microsatellite instability (MSI), and tumor mutational burden (TMB) were analyzed in selected samples. Results: Transcriptomic analysis highlighted significant alterations in critical cellular pathways, including cell cycle control, apoptosis, DNA repair mechanisms, and transcription regulation. Among the most dysregulated genes were DLGAP5 , CASP14 , BCL2 , and JUN . Copy number profiling in a subset (N=74) revealed significant chromosomal instability, with 34% of cases showing losses on specific chromosome arms and 9% exhibiting gains, notably in regions such as 4p, 19p and 1p. Mutations were predominantly observed in tumor suppressor genes, with a lower frequency in oncogenes. Frequently altered genes include TP53 (63%), CDKN2A (19%), PIK3CA (12%), TERT (15%), HRAS (6%) and EGFR (1%). HRAS mutations offer potential for targeted therapy with tipifarnib, a farnesyltransferase inhibitor. EGFR alterations, though rare, have proven challenging to effectively target with EGFR -specific therapies such as cetuximab or tyrosine kinase inhibitors, underscoring the critical need for innovative strategies to exploit this target. Amplifications were observed in MYC (9%), CCND1 (8%), FGF19 (6%), FGF3 (4%) and FGF4 (4%). Targetable amplifications were seen in EGFR (6%), ERBB2 (2%) and MET (1%), offering potential opportunities for personalized therapies. MET amplifications are also linked to aggressive tumor behavior. Fusion was a relatively less common event, observed in 9.5% cases, including one case each of targetable BRAF-MRPS33 and FGFR1-PLAG1 fusion. High TMB (≥10 muts/mb) was seen in 21% cases (Median 12, range 0-28). None of the tumors showed MSI-high status (n=63). Higher PD-L1 positivity was observed in TMB high samples (76% vs 59%), with PD-L1 22-C3 CPS ≥1 in 66% (50/ 76) and PD-L1 28-8 TPS ≥1 in 46% (36/ 78). Conclusions: This study underscores the genetic heterogeneity of HNCs, pointing to deregulated pathways involved in cell cycle control, apoptosis, and DNA repair as promising targets for future therapeutic development and personalized treatment strategies. Also this study points towards a subset of patients who may derive a greater benefit with immunotherapy, based on TMB/PDL1. ESCAT TIER distribution of the molecular alterations detected in head and neck cohort (N=273). TIERs Total Cases % occurrence IC 6 2% IIA 21 8% IIIA 85 31% IIIA 1 0% IIIB 57 21% IVA 180 66% IVB 71 26% X 142 52%
