Consensus Molecular Subtypes Research Articles

Consensus molecular subtypes (CMS) as prognostic and predictive biomarkers of RAS wild-type (WT) metastatic colorectal cancer (mCRC): Pooled analysis of the randomized trials FIRE-1, FIRE-3, XELAVIRI, and PanaMa.

3529 Background: CMS were associated with prognostic relevance in RAS WT mCRC, while their role as predictive biomarker is debatable. We aimed to assess the predictive impact of CMS on treatment with anti-EGFR vs. anti-VEGF antibodies vs. cytotoxic treatment alone in a pooled analysis. Methods: Available CMS data (called by predictedCMS method) of the randomized controlled trials FIRE-1 (FUFIRI vs. mIrOx); FIRE-3 (FOLFIRI+ cetuximab vs. bevacizumab), XELAVIRI (sequential vs. initially combined FOLFIRI+bevacizumab) and PanaMa (FOLFOX+panitumumab followed by FU/FA +/- panitumumab) of RAS WT mCRC were included. Non-evaluable CMS were excluded. Kaplan-Meier estimated overall survival (OS) and progression-free survival (PFS). Uni- and multivariate Cox regression analysis (including all baseline characteristics) was used to calculate hazard ratio (HR) and 95% confidence interval (95% CI). Results: 1147 tumors had available CMS data and 746 (65.0%) were RAS wild-type (WT): CMS1, n=89; CMS2, n=346; CMS3, n=71; CMS4, n=240. CMS were significant prognostic biomarkers in terms of PFS (median months, CMS1, 6.7; CMS2, 11.1; CMS3, 8.1; CMS4, 10.0, P<0.001) and OS (median months: CMS1, 14.8; CMS2, 28.7; CMS3, 19.5; CMS4, 26.5; P<0.001). CMS remained significant in a multivariate backward elimination Cox regression analysis including all baseline characteristic variables and BRAF status with regard to PFS ( P=0.03) and OS ( P=0.04) Interaction of CMS and treatment was significant regarding OS ( P<0.001) and PFS ( P=0.001). The longest OS was observed using anti-VEGF antibodies in CMS1 and anti-EGFR antibodies in CMS2 or CMS4 RAS WT tumors (Table). Conclusions: CMS were confirmed as prognostic biomarkers and might contain predictive information for the choice of anti-EGFR or anti-VEGF antibodies in RAS WT mCRC. Prospective validation remains mandatory. [Table: see text]

Overexpression of TSPAN8 in consensus molecular subtype 3 colorectal cancer

BackgroundRecently, consensus molecular subtypes (CMSs) have been proposed as a robust transcriptome-based classification system for colorectal cancer (CRC). Tetraspanins (TSPANs) are transmembrane proteins. They have been associated with the development of numerous malignancies, including CRC, through their role as “master organizers” for multi-molecular membrane complexes. No previous study has investigated the correlation between TSPANs and CMS classification. Herein, we investigated the expression of TSPANs in patient-derived primary CRC tissues and their CMS classifications. MethodsRNA samples were derived from primary CRC tissues (n = 100 patients diagnosed with colorectal adenocarcinoma) and subjected to RNA sequencing for transcriptome-based CMS classification and TSPAN-relevant analyses. Immunohistochemistry (IHC) and immunofluorescence (IF) stains were conducted to observe the protein expression level. To evaluate the relative biological pathways, gene-set enrichment analysis was performed. ResultsOf the highly expressed TSPAN genes in CRC tissues (TSPAN8, TSPAN29, and TSPAN30), TSPAN8 was notably overexpressed in CMS3-classified primary tissues. The overexpression of TSPAN8 protein in CMS3 CRC was also observed by IHC and IF staining. As a result of gene-set enrichment analysis, TSPAN8 may potentially play a role in organizing signaling complexes for kinase-based metabolic deregulation in CMS3 CRC. ConclusionsThe present study reports the overexpression of TSPAN8 in CMS3 CRC. This study proposes TSPAN8 as a subtype-specific biomarker for CMS3 CRC. This finding provides a foundation for future CMS-based studies of CRC, a complex disease and the second leading cause of cancer mortality worldwide.

The prevalence of sickle cell phenotype and its association with clinical outcomes in patients with solid malignancies.

10563 Background: Sickle cell disease (SCD) is the most common inherited blood disorder in the United States and is caused by a mutation of β-globin ( HBB-Mt). The risk of developing solid malignancies in this population remains controversial, with research showing increased rates of some solid tumors (colon, kidney, and thyroid) and decreased rates of others (breast and male genitourinary). We report the prevalence of HBB mutations across a cohort of solid tumors and clinical outcomes between HBB-Wild type (-Wt) v Mutant (-Mt) tumors. Methods: Non-small cell lung cancer (NSCLC, N = 3307), breast (BC, N = 1960), colorectal (CRC, N = 2161), prostate (N = 899), and gynecologic cancers (N = 2,805) in persons of African descentwere tested at Caris Life Sciences (Phoenix, AZ) with NextGen Sequencing of DNA (whole exome) and RNA (whole transcriptome). Her2/Neu (+: ≥3+ and >10%) and HR (ER or PR+: ≥1+ and ≥1%) expression was tested by immunohistochemistry. Tumors were assessed for single nucleotide variations (SNVs) and insertions/deletions (indels) in HBB associated with a sickle cell phenotype. Mutation prevalence for all other genes was calculated for pathogenic SNVs/indels. Differentially regulated pathways were assessed by gene set enrichment analysis (GSEA). Fisher’s Exact/χ2 tests were applied as appropriate with p-values adjusted for multiple comparisons ( p < .05). Real-world overall survival (OS) and ethnicity data was obtained from insurance claims and log-rank estimates were calculated for molecularly defined subpopulations. Results: Amongst persons of African descent, uterine sarcoma had the highest prevalence of HBB-Mt (17.1%, 43/251) followed by uterine serous carcinoma (9.1, 70/768) and endometrial carcinoma (7.8, 55/708). Tumors with the highest absolute number of HBB-Mt tumors were NSCLC (7.0, 230/3307), CRC (8.1, 176/2161) and BC (6.7, 131/1960). In CRC, there was no significant difference in the prevalence of right-sided tumors between HBB-Wt v Mt (42% v 51, p = .06), nor in the prevalence of HBB-Mt by consensus molecular subtype ( p = .34). No significant enrichment of HBB-Mt across BC subtypes was observed (HR- HER2-: 7.8%, HR-HER2+: 9.7, HR+/HER2-:5.8, HR+/HER2+: 8.1, p = .34). No difference in the prevalence of pathogenic alterations was observed between HBB-Mt vs -Wt across NSCLC, CRC, and BC ( p > .05). HBB-Mt NSCLC were significantly enriched (GSEA) for genes related interferon /, inflammatory response and IL6/JAK/STAT3. No significant difference in OS was observed in CRC (median survival HBB-Wt: 957 days v -Mt: 1034, p = .43), NSCLC (-Wt: 591 v -Mt: 648, p = .59) or BC (-Wt: 924 v -Mt: 922, p = .61). Conclusions: No significant differences in the genomic landscape nor in OS of HBB-Mt vs HBB-Wt NSCLC, CRC or BC were observed. NSCLC HBB-Mt was enriched with inflammatory response genes. Future work should focus on potential role HBB-Mt plays in NSCLC.

A single-sample gene-expression classifier for the intrinsic consensus molecular subtypes in colorectal cancer.

e15587 Background: An analysis of colorectal cancer epithelial cells showed two intrinsic subtypes called iCMS2 and iCMS3, in addition to the bulk consensus subtypes (CMS1-4). The intrinsic subtypes are prognostically important. We present a robust method for calculating the iCMS subtypes in individual samples, as opposed to batches, applicable in a clinical setting. Methods: We selected the subset of iCMS genes (N = 201) with the strongest epithelial expression in colorectal cancer, aiming to reduce the signal from other cell types. We then calculated gene expression centroids by taking the per-gene mean expression of 8 random samples from three non-batch corrected datasets (1779 samples total) that were previously used to derive the iCMS subtypes. Thirty-two centroids were calculated per dataset and per iCMS, for a total of 192 centroids. The single sample classifier (SSC) calculates the most likely iCMS class based on the class of the nearest centroid by non-parametric correlation distance. A confident call is made when the correlation is at least 0.1 and distance from the nearest centroid of the opposite iCMS class is greater than 0.05. Results: The SSC achieved 98.3% agreement with the reference calls in training data. The SSC was applied to unseen data for validation, by classifying non-batch corrected samples one at a time. The same data were clustered with the reference published iCMS public dataset and the reference iCMS was estimated at the batch level by proximity to reference samples in the hierarchical tree structure. Concordance between reference and SSC calls was 91.4% in non-batch corrected data and only modestly improved to 93.9% when batch correction was applied, showing the robustness of this approach. Concordance remained acceptable even when non-confident SSC calls were included (86.9%). The proportion of calls that were judged confident with the SSC was 77.9%, slightly lower than the proportion of confident reference calls (87.6%). In addition, the SSC was applied to data from the VELOUR trial, for which reference iCMS calls were also available. The SSC iCMS was prognostic (HR 0.601 for OS in iCMS2 vs iCMS3, P < 0.00001) and performed at least as well as the reference iCMS (HR 0.669, P = 0.0004). Conclusions: The SSC reproduces the iCMS classification and appears robust to batch effect. The SSC calls are prognostic in clinical trial data. The SSC R package is available for download (https://github.com/CRCrepository/iCMS.SSC). [Table: see text]

A consensus molecular subtypes classification strategy for clinical colorectal cancer tissues.

Consensus Molecular Subtype (CMS) classification of colorectal cancer (CRC) tissues is complicated by RNA degradation upon formalin-fixed paraffin-embedded (FFPE) preservation. Here, we present an FFPE-curated CMS classifier. The CMSFFPE classifier was developed using genes with a high transcript integrity in FFPE-derived RNA. We evaluated the classification accuracy in two FFPE-RNA datasets with matched fresh-frozen (FF) RNA data, and an FF-derived RNA set. An FFPE-RNA application cohort of metastatic CRC patients was established, partly treated with anti-EGFR therapy. Key characteristics per CMS were assessed. Cross-referenced with matched benchmark FF CMS calls, the CMSFFPE classifier strongly improved classification accuracy in two FFPE datasets compared with the original CMSClassifier (63.6% versus 40.9% and 83.3% versus 66.7%, respectively). We recovered CMS-specific recurrence-free survival patterns (CMS4 versus CMS2: hazard ratio 1.75, 95% CI 1.24-2.46). Key molecular and clinical associations of the CMSs were confirmed. In particular, we demonstrated the predictive value of CMS2 and CMS3 for anti-EGFR therapy response (CMS2&3: odds ratio 5.48, 95% CI 1.10-27.27). The CMSFFPE classifier is an optimized FFPE-curated research tool for CMS classification of clinical CRC samples.

Multiregional transcriptomics identifies congruent consensus subtypes with prognostic value beyond tumor heterogeneity of colorectal cancer

Intra-tumor heterogeneity compromises the clinical value of transcriptomic classifications of colorectal cancer. We investigated the prognostic effect of transcriptomic heterogeneity and the potential for classifications less vulnerable to heterogeneity in a single-hospital series of 1093 tumor samples from 692 patients, including multiregional samples from 98 primary tumors and 35 primary-metastasis sets. We show that intra-tumor heterogeneity of the consensus molecular subtypes (CMS) is frequent and has poor-prognostic associations independently of tumor microenvironment markers. Multiregional transcriptomics uncover cancer cell-intrinsic and low-heterogeneity signals that recapitulate the intrinsic CMSs proposed by single-cell sequencing. Further subclassification identifies congruent CMSs that explain a larger proportion of variation in patient survival than intra-tumor heterogeneity. Plasticity is indicated by discordant intrinsic phenotypes of matched primary and metastatic tumors. We conclude that multiregional sampling reconciles the prognostic power of tumor classifications from single-cell and bulk transcriptomics in the context of intra-tumor heterogeneity, and phenotypic plasticity challenges the reconciliation of primary and metastatic subtypes.

Image-based consensus molecular subtyping in rectal cancer biopsies and response to neoadjuvant chemoradiotherapy

The development of deep learning (DL) models to predict the consensus molecular subtypes (CMS) from histopathology images (imCMS) is a promising and cost-effective strategy to support patient stratification. Here, we investigate whether imCMS calls generated from whole slide histopathology images (WSIs) of rectal cancer (RC) pre-treatment biopsies are associated with pathological complete response (pCR) to neoadjuvant long course chemoradiotherapy (LCRT) with single agent fluoropyrimidine. DL models were trained to classify WSIs of colorectal cancers stained with hematoxylin and eosin into one of the four CMS classes using a multi-centric dataset of resection and biopsy specimens (n = 1057 WSIs) with paired transcriptional data. Classifiers were tested on a held out RC biopsy cohort (ARISTOTLE) and correlated with pCR to LCRT in an independent dataset merging two RC cohorts (ARISTOTLE, n = 114 and SALZBURG, n = 55 patients). DL models predicted CMS with high classification performance in multiple comparative analyses. In the independent cohorts (ARISTOTLE, SALZBURG), cases with WSIs classified as imCMS1 had a significantly higher likelihood of achieving pCR (OR = 2.69, 95% CI 1.01–7.17, p = 0.048). Conversely, imCMS4 was associated with lack of pCR (OR = 0.25, 95% CI 0.07–0.88, p = 0.031). Classification maps demonstrated pathologist-interpretable associations with high stromal content in imCMS4 cases, associated with poor outcome. No significant association was found in imCMS2 or imCMS3. imCMS classification of pre-treatment biopsies is a fast and inexpensive solution to identify patient groups that could benefit from neoadjuvant LCRT. The significant associations between imCMS1/imCMS4 with pCR suggest the existence of predictive morphological features that could enhance standard pathological assessment.

Microenvironment, systemic inflammatory response and tumor markers considering consensus molecular subtypes of colorectal cancer.

Introduction: Colorectal carcinomas (CRC) are one of the most frequent malignancies worldwide. Based on gene expression profile analysis, CRCs can be classified into four distinct subtypes also known as the consensus molecular subtypes (CMS), which predict biological behaviour. Besides CMS, several other aspects of tumor microenvironment (TME) and systemic inflammatory response (SIR) influence the outcome of CRC patients. TME and inflammation have important role in the immune (CMS1) and mesenchymal (CMS4) subtypes, however, the relationship between these and systemic inflammation has not been assessed yet. Our objective was to evaluate the connection between CMS, TME and SIR, and to analyze the correlation between these markers and routinely used tumor markers, such as CEA (Carcinoembryonic Antigen) and CA19-9 (Carbohydrate Antigen 19-9). Methods: FFPE (Formalin Fixed Paraffin Embedded) samples of 185 CRC patients were collected. TME was described using tumor-stroma ratio (TSR), Klintrup-Makinen (KM) grade, and Glasgow Microenvironment Score (GMS). CMS classification was performed on tissue microarray using MLH1, PMS2, MSH2 and MSH6, and pan-cytokeratin, CDX2, FRMD6, HTR2B and ZEB1 immunohistochemical stains. Pre-operative tumor marker levels and inflammatory markers [C-reactive protein - CRP, albumin, absolute neutrophil count (ANC), absolute lymphocyte count (ALC), absolute platelet count (APC)] and patient history were retrieved using MedSolution database. Results: Amongst TME-markers, TSR correlated most consistently with adverse clinicopathological features (p < 0.001) and overall survival (p < 0.001). Elevated CRP and modified Glasgow Prognostic Score (mGPS) were associated with worse outcome and aggressive phenotype, similarly to tumor markers CEA and CA19-9. Stroma-Tumor Marker score (STM score), a new combined score of CA19-9 and TSR delivered the second best prognostication after mGPS. Furthermore, CMS4 showed association with TSR and several laboratory markers (albumin and platelet derived factors), but not with other SIR descriptors. CMS did not show any association with CEA and CA19-9 tumor markers. Conclusion: More routinely available TME, SIR and tumor markers alone and in combination deliver reliable prognostic data for choosing the patients with higher risk for propagation. CMS4 is linked with high TSR and poor prognosis, but in overall, CMS-classification showed only limited effect on SIR- and tumor-markers.

Clinical Challenges of Consensus Molecular Subtype CMS4 Colon Cancer in the Era of Precision Medicine.

Over the past decade, our understanding of the diversity of colorectal cancer has expanded significantly, raising hopes of tailoring treatments more precisely for individual patients. A key achievement in this direction was the establishment of the consensus molecular classification, particularly identifying the challenging consensus molecular subtype (CMS) CMS4 associated with poor prognosis. Because of its aggressive nature, extensive research is dedicated to the CMS4 subgroup. Recent years have unveiled molecular and microenvironmental features at the tissue level specific to CMS4 colorectal cancer. This has paved the way for mechanistic studies and the development of preclinical models. Simultaneously, efforts have been made to easily identify patients with CMS4 colorectal cancer. Reassessing clinical trial results through the CMS classification lens has improved our understanding of the therapeutic challenges linked to this subtype. Exploration of the biology of CMS4 colorectal cancer is yielding potential biomarkers and novel treatment approaches. This overview aims to provide insights into the clinico-biological characteristics of the CMS4 subgroup, the molecular pathways driving this subtype, and available diagnostic options. We also emphasize the therapeutic challenges associated with this subtype, offering potential explanations. Finally, we summarize the current tailored treatments for CMS4 colorectal cancer emerging from fundamental and preclinical studies.

Multi-omics cluster defines the subtypes of CRC with distinct prognosis and tumor microenvironment

BackgroundColorectal cancer (CRC) is a complex malignancy characterized by diverse molecular profiles, clinical outcomes, and limited precision in prognostic markers. Addressing these challenges, this study utilized multi-omics data to define consensus molecular subtypes in CRC and elucidate their association with clinical outcomes and underlying biological processes.MethodsConsensus molecular subtypes were obtained by applying ten integrated multi-omics clustering algorithms to analyze TCGA-CRC multi-omics data, including mRNA, lncRNA, miRNA, DNA methylation CpG sites, and somatic mutation data. The association of subtypes with prognoses, enrichment functions, immune status, and genomic alterations were further analyzed. Next, we conducted univariate Cox and Lasso regression analyses to investigate the potential prognostic application of biomarkers associated with multi-omics subtypes derived from weighted gene co-expression network analysis (WGCNA). The function of one of the biomarkers MID2 was validated in CRC cell lines.ResultsTwo CRC subtypes linked to distinct clinical outcomes were identified in TCGA-CRC cohort and validated with three external datasets. The CS1 subtype exhibited a poor prognosis and was characterized by higher tumor-related Hallmark pathway activity and lower metabolism pathway activity. In addition, the CS1 was predicted to have less immunotherapy responder and exhibited more genomic alteration compared to CS2. Then a prognostic model comprising five genes was established, with patients in the high-risk group showing substantial concordance with the CS1 subtype, and those in the low-risk group with the CS2 subtype. The gene MID2, included in the prognostic model, was found to be correlated with epithelial–mesenchymal transition (EMT) pathway and distinct DNA methylation patterns. Knockdown of MID2 in CRC cells resulted in reduced colony formation, migration, and invasion capacities.ConclusionThe integrative multi-omics subtypes proposed potential biomarkers for CRC and provided valuable knowledge for precision oncology.

The significance of m6A RNA methylation regulators in diagnosis and subtype classification of HBV-related hepatocellular carcinoma.

In recent years, abnormal m6A alteration in hepatocellular carcinoma (HCC) has been a focus on investigating the biological implications. In this study, our objective is to determine whether m6A modification contributes to the progression of HBV-related HCC. To achieve this, we employed a random forest model to screen top 8 characteristic m6A regulators from 19 candidate genes. Subsequently, we developed a nomogram model that utilizes these 8 characteristic m6A regulators to predict the prevalence of HBV-related HCC. According to decision curve analysis, patients may benefit from the nomogram model. The clinical impact curves exhibited a robust predictive capability of the nomogram models. Additionally, consensus molecular subtyping was employed to identify m6A modification patterns and m6A-related gene signature. The quantification of immune cell subsets was accomplished through the implementation of ssGSEA algorithms. PCA algorithms were developed to compute the m6A score for individual tumors. Two distinct m6A modification patterns, namely cluster A and cluster B, exhibited significant correlations with distinct immune infiltration patterns and biological pathways. Notably, patients belonging to cluster B demonstrated higher m6A scores compared to those in cluster A, as determined by the m6A score metric. Furthermore, the expression of IGFBP3 proteins was validated through immunofluorescence, revealing their pronounced lower expression in tumor tissues. In summary, our study underscores the importance of m6A modification in the advancement of HBV-related HCC. This research has the potential to yield novel prognostic biomarkers and therapeutic targets for the identification of HBV-related HCC.

Identification of unique rectal cancer-specific subtypes.

Existing colorectal cancer subtyping methods were generated without much consideration of potential differences in expression profiles between colon and rectal tissues. Moreover, locally advanced rectal cancers at resection often have received neoadjuvant chemoradiotherapy which likely has a significant impact on gene expression. We collected mRNA expression profiles for rectal and colon cancer samples (n = 2121). We observed that (i) Consensus Molecular Subtyping (CMS) had a different prognosis in treatment-naïve rectal vs. colon cancers, and (ii) that neoadjuvant chemoradiotherapy exposure produced a strong shift in CMS subtypes in rectal cancers. We therefore clustered 182 untreated rectal cancers to find rectal cancer-specific subtypes (RSSs). We identified three robust subtypes. We observed that RSS1 had better, and RSS2 had worse disease-free survival. RSS1 showed high expression of MYC target genes and low activity of angiogenesis genes. RSS2 exhibited low regulatory Tcell abundance, strong EMT and angiogenesis signalling, and high activation of TGF-β, NF-κB, and TNF-α signalling. RSS3 was characterised by the deactivation of EGFR, MAPK and WNT pathways. We conclude that RSS subtyping allows for more accurate prognosis predictions in rectal cancers than CMS subtyping and provides new insight into targetable disease pathways within these subtypes.

Abstract 409: Deciphering chromosomal instability in consensus molecular subtypes (CMS) in CRC: Insights from an integrative multi-omics approach

Abstract CRC is a heterogeneous disease with different molecular subtypes, which can have a profound impact on treatment response and patient outcomes. CMS classification based on gene expression profiles has been shown to differentiate patients into clinically relevant sub-groups. However, deep multi-omics characterization of the CMS classes based on high quality patient data is lacking. Through the integration of multi-omics data, we aim to enhance our understanding of CRC’s molecular heterogeneity and thereby facilitate the development of personalized therapeutic strategies. We performed a multi-omics analysis of CRC patients to identify potential new features of CMS groups. 890 fresh-frozen, surgically resected CRC tumor and adjacent normal samples were used to analyze whole genome and RNA sequencing data. DNA and RNA was prepared using Qiagen AllPrep Universal and KAPA Hyper Prep kit (DNA), as well as TruSeq Stranded Total mRNA kits and sequenced on a NovaSeq6000 system. We identified CMS subtypes based on RNA-Seq data and performed all statistical analyses using R. Our analysis revealed distinct molecular characteristics across the four CMS subtypes, including unique driver mutations, signaling cascades, and immune cell infiltration profiles. The poor prognosis associated with CMS4 patients was corroborated by our clinical data. Notably, we observed high global chromosomal instability (CIN) profiles for both CMS2 and CMS4 and identified recurrent global aneuploidy patterns within these subtypes that link alterations of known cancer genes to these large-scale structural events. We found higher levels of CIN in CMS2 compared to CMS4, which appears to be predominantly driven by numerical, rather than structural CIN. Specifically, in CMS2, SMAD2/4 is commonly impacted by deletion on chromosome arm 18q, while PLCG1 is more influenced by amplifications on 20q respectively, suggesting potential values as CRC CMS biomarkers. In conclusion, our integrative multi-omics analysis offers a comprehensive understanding of CRC molecular subtypes, which can inform personalized treatment strategies. Our findings underscore the importance of considering not only specific driver mutations but also large-scale structural rearrangements in the context of CMS subtypes. Citation Format: Julia Bischof, Jonathan Woodsmith, David Church. Deciphering chromosomal instability in consensus molecular subtypes (CMS) in CRC: Insights from an integrative multi-omics approach [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 409.

Abstract 5059: Robust single sample consensus molecular subtype classification for primary and metastatic colorectal cancer

Abstract Introduction Consensus Molecular Subtypes (CMS) represent a well-established molecular stratification framework for colorectal cancer (CRC). Existing methods for CMS classification rely on a representative input cohort as a preprocessing step (CMScaller) or have difficulty generalizing to metastatic samples (CMSclassifier). We developed Tempus CMS to overcome these limitations. Our method normalizes gene expression data from input samples to a static reference cohort to enable single sample classification of both primary and metastatic tumors. We evaluated the performance of this classifier on a large, de-identified CRC cohort comprising 8,489 samples from primary and metastatic sites. Methods To normalize input data, our method shifts and scales each gene expression value based on the mean and standard deviation of the expression of that gene in the reference cohort (n=2,787 primary and metastatic CRC). CMS calls are then generated via nearest template prediction similar to CMScaller (Eide et al., 2017). The analysis cohort was selected from clinical biopsies within 30 days of primary or metastatic diagnosis excluding reference cohort samples. CMS calls were assessed by comparing subtype prevalence to reported rates and by testing for known enrichments of pathways and genomic markers. Results Subtype prevalences for lower gastrointestinal (GI) biopsies and resections (n=5,090) were comparable to reported rates for primary CRC tumors: CMS1 - 12%, CMS2 - 28%, CMS3 - 19%, and CMS4 - 33% with 7.6% no calls. Biopsies from metastatic tissue showed site-specific changes in prevalence: CMS2 increased to 43% in liver samples (n=1,715) and CMS4 was identified in 49% of samples from metastatic sites excluding the liver (n=1,684). In addition, we found that purity had an effect on prevalence: in lower GI tumors with ≥ 50% tumor purity rates of CMS2 and CMS4 were 34% and 26%, respectively, while rates in tumors with purity between 20% and 50% were 20% CMS2 and 41% CMS4. Across all tissue groups, pathway enrichment showed significant associations, including TGF-ꞵ signaling in CMS4 (adj P ≤ 4.4e-15) and cell differentiation in CMS3 (adj P ≤ 3.0e-27). Expected enrichments of CMS markers were also observed: BRAF mutations and microsatellite instability in CMS1 and KRAS mutations in CMS3. In liver tissue, higher KRAS mutation rates were noted in CMS1 (57%) and CMS3 (83%) compared to GI tissue (35% and 67%, respectively). Conclusions In this study, we developed an enhanced, single sample CMS-calling algorithm optimized for primary and metastatic sites. Application of this algorithm in a large cohort recapitulated known biology, which suggests that the tool can be used to support clinical studies requiring robust molecular stratification. Citation Format: Andrew J. Sedgewick, Tushar Chandra, Timothy Taxter, Justin Guinney. Robust single sample consensus molecular subtype classification for primary and metastatic colorectal cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 5059.

Abstract 3807: CXCL9:SPP1 macrophage polarity as a prognostic biomarker for localized colorectal cancer after surgical resection

Abstract Background: There is an unmet need for additive predictive biomarkers for the recurrence of stage II/III colorectal cancer (CRC) after surgery. As recent studies showing the potential prognostic role of macrophage polarity defined by CXCL9 and SPP1, instead of conventional M1 and M2, we tried to evaluate the prognostic role of this biomarker in real-world practice setting, delineating biologic mechanisms. Method: The RNA sequencing data of 273 tumors derived from patietns with stage II/III colorectal cancer who underwent surgical resection in 5 Korean cancer-centers were available for analysis. We initially divided two subgroups based on the ratio of expression level of CXCL9 per SPP1 expressions (CS ratio) and evaluated prognostic outcomes and compared immune cell proportions by deconvolutional method between subgroups. Result: The CS ratio was higher in consensus molecular subtype (CMS)2 or CMS3 than others, and no significant differences were observed according to sidedness, stage, microsatellite instability status and medical centers. Total 69 (25,3%) patients were classified as low level of CS ratio group, and there were no significant differences in baseline characteristics, including age, sex, CMS, sidedness, stage, microsatellite instability status and medical centers. With a median follow-up of 58.2 months, the low CS ratio group showed poor 5-year recurrence-free survival (50.6 vs. 71.1%, HR=2.27, P = 0.002) and 5-year overall survival (69.7% vs. 85.9%, HR=1.75, P = 0.033). These trends were consistent across the subgroups. In the multivariate analyses including the potential prognostic factors (P &amp;lt; 0.1 in the univariate analyses), low CS ratio was significantly associated with poorer for RFS (adjusted HR = 1.96, P = 0.011) and OS (adjusted HR = 1.81, P = 0.002). These trends were consistent in the TCGA dataset (5 years OS: 69.7% vs. 85.9%, HR=1.98, P = 0.011). The low CS ratio group showed upregulated WNT signaling, and extracellular matrix organization pathway. Immune cell deconvolutional method exhibited the low CS ratio group showed more proportion of neutrophils, monocytes, whereas CD8+T cells, activated memory CD4+T cells were decreased, compared to high CS ratio group. Conclusions: These results suggest that CS ratio could be an applicable prognostic biomarker in locally advanced colorectal cancer after surgical resection. Based on this classification, more advanced treatment strategy may be required for low CS ratio group to improve surgical outcomes. Citation Format: Yeong Hak Bang, Ji Hye Choi, Hyunsu Kim, Kyunghee Park, Dae Hee Pyo, Yong Beom Cho, Woong-Yang Park. CXCL9:SPP1 macrophage polarity as a prognostic biomarker for localized colorectal cancer after surgical resection [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3807.

Abstract 4894: Multi-omic approach for drug repurposing in the poor prognosis CMS4 subtype of colon cancer

Abstract Introduction: Colon cancer (CC) is one of the deadliest malignancy in developing countries. Key drivers considered for the clinical management are RAS, BRAF alterations together with TNM staging and microsatellite status (MS). With the advancement of sequencing technologies and bioinformatic approaches, various molecular classifications of CC have been proposed. Guinney et al. proposed a transcriptomic-based molecular subtyping, the so-called four Consensus Molecular Subtypes (CMS). CMS4 patients have the worst prognosis. Biologically, CMS4, also known as mesenchymal, shows high expression of genes related to EMT, matrix remodelling, TGFb signaling, and inflammatory-related system and a peculiar enrichment in stromal cells. Thus, in the present study we are focusing on CMS4 subgroup in order to identify drugs to be repurposed in such a clinical setting through a multi-omic approach. Methods: Using TCGAbiolinks R package, we retrieved STAR Counts transcriptome profiling data, for COAD and READ. CMSclassifier was used to determine the CMS of the samples. We separated the coding genes from the long noncoding RNAs (lncRNAs) using annotation databases generated from Ensembl. We estimated the fraction of cell populations throughout the CIBERSORTx. Features selection approaches was performed both for coding genes and lncRNA. Data normalization has been carried out through variance stabilizing transformation for read count data. To identify CMS and potential biomarkers, we adopted mixOmics N-integration method. Our independent validation cohort consists of 100 patients, locally enrolled, upon local Ethical Committee approval. Features contributing to CMS4 subtype have been extracted and use to construct drug-gene interaction network. Results: To set-up the classification model the DIABLO approach was used, reaching a mean AUCROC of 0.9022. Contributing features related CMS4 subtypes included nine coding genes, six lncRNAs and, regarding CibersortX deconvolution, Macrophage M0 and M2.The drug-gene interaction network includes six subnetworks centered to ALOX5, KCNMA1, AQP9, TGFB3, THBS4 and DPYSL3. The role of TGFb pathway was confirmed, considering the contribution to classification. Interestingly, we found interactions with Non-Steroid Anti-Inflammatory Drugs (NSAIDs), such as Mesalazine. Moreover, THBSA gene, involved in cell-to-cell, cell-to-matrix and stromal response, could also be targeted. Conclusions: The results of our drug repositioning approach have been biologically corroborated considering, as mentioned above, the peculiar enrichment of CMS4 subtype for TGFb pathway and stromal cells. The chance to modulate the macrophagic activity in the tumor microenvironment through NSAIDs could be evaluated. Experimental validation on patient-derived organoids (PDOs), related to samples included to local validation cohort, is ongoing. Citation Format: Tommaso M. Marvulli, Debora Traversa, Roberta Di Fonte, Leonarda Maurmo, Amalia Azzariti, Letizia Porcelli, Claudio A. Coppola, Concetta Saponaro, Eliseo Mattioli, Francesco A. Zito, Rossella Fasano, Simona Serratì, Davide Quaresmini, Oronzo Brunetti, Stefania Tommasi, Raffaella Massafra, Simona De Summa. Multi-omic approach for drug repurposing in the poor prognosis CMS4 subtype of colon cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4894.

Abstract 2496: Molecular characterization of stage III colon cancer patients with recurrence after adjuvant chemotherapy

Abstract Introduction: Patients with stage III colon cancer are routinely treated with resection followed by adjuvant chemotherapy (ACT) with a fluoropyrimidine and oxaliplatin. This one-size-fits-all approach does not account for heterogeneity in tumor biology and fails to cure ~30% of patients, highlighting the urgency for biomarkers to better understand and predict resistance to ACT or to offer alternative targeted treatment. Aim: To identify molecular characteristics predictive of resistance to standard ACT in stage III colon cancer. Methods: Patients with resected stage III colon cancer who received ACT (93% CAPOX, 7% CAP) were selected from the Prospective Dutch Colorectal Cancer cohort (PLCRC). RNA exome capture sequencing was performed on resected FFPE tumor tissue to compare differential gene expression, hallmark pathway enrichment, immune cell deconvolution and consensus molecular subtypes (CMS) between patients with versus without recurrence after surgery and ACT. Postsurgery circulating tumor DNA (ctDNA) detected in PLCRC-PROVENC3 (separate abstract) was used to select a subgroup with minimal residual disease (MRD) postsurgery, as sensitivity analysis of patients with recurrence despite ACT (chemoresistant) versus curation by ACT (chemosensitive). Results: For 264 out of 278 patients (95%), tumor-derived RNA sequencing libraries were of sufficient quality. 67 patients (25%) with recurrence (median after 13 months [IQR 7-23]) were compared to 178 patients with at least 2 year recurrence-free follow-up (median 42 months [IQR 30-53]). A total of 320 genes were differentially expressed between patients with versus without recurrence. Genesets defining epithelial-mesenchymal transition (p = 0.002) and angiogenesis (p &amp;lt; 0.001) were enriched in the primary tumor of patients with recurrence, in part explained by the estimated larger fraction of stroma (Wilcoxon effect size r 0.18, p &amp;lt; 0.001) and cancer-associated fibroblasts (CAFs) (r 0.19, p = 0.003). In contrast, the fraction of CD4+ T cells was lower (r 0.14, p = 0.001). Consistently, CMS4 was found more frequently in tumors from patients with versus without recurrence (57% versus 34%, p &amp;lt; 0.001). Of the subgroup with ctDNA-based MRD postsurgery (26 out of 171 assessed patients), 19 patients (73%) with recurrence (median after 12 months [IQR 7-17]) were compared to 7 patients without recurrence (minimum follow-up 32 months). This sensitivity analysis showed similar trends in pathways, immune cells and CMS as the main analysis, with a more pronounced effect size for CAFs (r 0.27). Conclusion: Stage III colon tumors of patients who experience recurrence after resection and ACT are enriched for stroma and CAFs, while displaying less CD4+ T cells. The observed trends are sustained in the patients with MRD postsurgery. These findings may help to better predict resistance to standard ACT and elucidate targets for alternative personalized treatment. Citation Format: Ingrid Franken, Steven Ketelaars, Carmen Rubio-Alarcón, Sietske van Nassau, Suzanna Schraa, Gerrit Meijer, Mark Sausen, Miriam Koopman, Geraldine Vink, Sanne Abeln, Remond Fijneman, Jeanine Roodhart, on behalf of the PLCRC-MEDOCC group. Molecular characterization of stage III colon cancer patients with recurrence after adjuvant chemotherapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2496.

Abstract 5397: Characteristics of cancer associated fibroblasts isolated from refractory colorectal cancer using RNA sequence

Abstract Primary tumor location of colorectal cancers (CRCs) is emerging as an important prognostic factor owing to distinct biological features. In fact, the patients with right-side CRCs (R-CRCs), in which primary tumors were resected in our department and followed by systemic chemotherapy, have a poorer prognosis in 2-year overall survival and progression-free survival than those with left-side CRCs (L-CRCs). Furthermore, cancer-associated fibroblasts (CAFs) promote tumor malignancy, but the precise transcriptional mechanisms regulating the acquisition of the CAF phenotype are not well understood. Here, we isolated the CAFs and normal counterparts from 10 CRCs which were performed radical resection, including 5 L-CRCs and 5 R-CRCs, and analyzed transcriptomic profile of CAFs by RNA-sequencing. Gene Set Enrichment Analysis (GSEA) was performed and inflammatory signatures, such as "HALLMARK IL2 STAT5 SIGNALING"(NOM p-value&amp;lt;0.001, FDR q-value=0.146), "HALLMARK IL6 JAK STAT3 SIGNALING"(NOM p-value=0.014, FDR q-value=0.162), "HALLMARK INTERFERON GAMMA RESPONSE"(NOM p-value=0.032, FDR q-value=0.1726), were enriched in the CAFs isolated from R-CRCs (R-CAFs). In addition to sidedness of CRCs, based on gene expression profiles from bulk tumors, CRCs were classified into four consensus molecular subtypes (CMS1-4). Among them, The fibrotic CMS4 subtype portends poor relapse-free survival. We identified 4 CMS4-CRCs, which have the characteristics of stromal activation and the worst prognosis among other CRCs. GSEA showed that the signature of "HALLMARK TNFA_SIGNALING_VIA_NFKB" (NOM p-value=0.023 FDR q-value=0.207), "HALLMARK GLYCOLYSIS" (NOM p-value=0.036 FDR q-value=0.239) were upregulated in the CAFs isolated from CMS4-CRCs. These results indicated that CAFs isolated from refractory CRCs have unique phenotypes and the elucidation of these mechanisms might lead to a promising treatment for refractory CRCs. Citation Format: Hiroaki Kasashima, Yasuhiro Fukui, Ken Yonemitsu, Kisyu Kitayama, Yuichiro Miki, Mami Yoshii, Tatsunari Fukuoka, Tatsuro Tamura, Masatsune Shibutani, Takahiro Toyokawa, Shigeru Ree, Masakazu Yashiro, Kiyoshi Maeda. Characteristics of cancer associated fibroblasts isolated from refractory colorectal cancer using RNA sequence [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 5397.

Abstract 6430: Gene expression signature quantifies the cellular contribution of the tumor vs. the microenvironment in determining colorectal cancer prognosis and therapeutic outcome

Abstract Accumulating evidence has suggested that cancer progression and therapeutic response may depend on not only tumor cells but also their tumor microenvironment (TME). However, their contributory roles are complex and not well understood. Here we report an analysis of 2373 human colorectal cancer (CRC) tumors representing all stages, leading to identification of a 20-gene signature capable of quantifying the cellular contribution of the tumor vs. the TME in determining colorectal cancer prognosis and therapeutic outcome. We classified these tumors into the consensus molecular subtypes (CMS1-4) with distinct, variable TME cellular features defined by CIBERSORT deconvolution analysis of bulk gene expression data. Surprisingly, all the 10 positive signature genes (portending worse survival) were strongly correlated with the TME-rich CMS1 and CMS4 tumors. By contrast, all the 10 negative signature genes (portending better survival) were highly related to the TME-poor CMS2 and CMS3 tumors. A single cell expression analysis from an independent database further revealed that strikingly, all the 10 positive genes were principally expressed by the immune/stromal TME whereas all the 10 negative genes were predominantly expressed in epithelial tumor cells. These data suggest that the TME has an equally important role as do tumor cells in determining CRC prognosis and support targeting the TME to improve survival. Moreover, distinct gene expression from tumor cells versus the TME also showed a differential impact on therapeutic outcome. Retrospective analyses on two independent clinical trial datasets show that the 20-gene signature score significantly predicted progression free survival in metastatic CRC patients treated with cetuximab, an FDA-approved EGFR inhibitor therapy. These data suggest a potential for the 20-gene signature as a predictive biomarker to identify “sensitive” versus “resistant” subpopulations of CRC patients to improve outcome of EGFR-targeted therapy. Citation Format: Mingli Yang, Timothy J. Yeatman, Michael V. Nebozhyn, Michael J. Schell, Lance Pflieger, Andrey Loboda, Warren J. Pledger, Ramani Soundararajan, Michelle Maurin, Heiman Wang, Jetsen Rodriguez Silva, Ashley Alden. Gene expression signature quantifies the cellular contribution of the tumor vs. the microenvironment in determining colorectal cancer prognosis and therapeutic outcome [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6430.

Abstract 881: Cell-type-specific gene programs suggest immune and stromal drivers of relapse in colorectal cancer

Abstract Background: Molecular profiling of colorectal cancer (CRC) has demonstrated marked heterogeneity between tumors that appear histologically similar. Although existing classification systems such as the Consensus Molecular Subtypes (CMS) have shown clinical utility, these bulk sequencing-based methods fail to adequately address the intra-tumoral heterogeneity of CRC. To disentangle the complex landscape and identify clinically-relevant properties of CRC we generated a single cell sequencing (scRNASeq) database from 225 tumor samples derived from resection of primary colon or liver metastases. Our analysis defines cell-type-specific gene programs significantly associated with relapse and survival and suggests interventions which may abrogate their effect. Methods: Utilizing a bio-repository of CRC specimens with an average of 5 years of clinical follow-up, we performed scRNASeq on 225 patient samples comprising ~ 1,000,000 total cells and 2.4 billion gene expression measurements. Following traditional clustering approaches, we then inferred gene expression programs through consensus non-negative matrix factorization (cNMF) to characterize cell activities. Gene program activity was explored on a cell-type-specific level and associated with clinical outcomes such as relapse. Treatment effects were additionally explored. Results: We identified 56 cell-type specific gene programs significantly associated with relapse in adult CRC patients. Interestingly these programs spanned all cellular compartments with the majority found in stromal or immune, as opposed to tumor cells. Program activity such as the non-canonical Granulocyte Chemotaxis program in B Cells was protective of relapse (relapse median expression 0% activity vs non-relapse activity 40% pAdj &amp;lt;.05 × 1050) whereas the Epithelial Mesenchymal Transition (EMT) program in Cancer Associated Fibroblasts (CAF) was strongly associated with relapse (relapse median expression 16% activity vs non-relapse activity 0% pAdj &amp;lt;.05 × 1010). Cumulative program usage enabled stratification on a patient level to provide an estimate of comparative disease free survival (DFS) and overall survival (OS); gene-programs were significantly associated with outcomes such as the CAF-EMT program which was associated with a DFS HR of 3 (pAdj &amp;lt;.0005). Responses to chemotherapy and treatment effect were also explored; prior treatment with Bevacizumab was associated with a significant reduction in the CAF-EMT program (11% activity in non-exposed patients versus 0% activity in exposed patients pAdj &amp;lt;.05 × 1010). Conclusions: Analysis of the largest CRC scRNASeq cohort to date identifies a number of previously undescribed cell-type specific gene programs significantly associate with relapse after surgery. These programs provide insights into the underlying mechanisms and suggest that the TME plays an important role in relapse. Citation Format: Nicholas J. Hornstein, Mingshuang Wang, Abdelrahman Yousef, Saikat Chowdhury, Fadl Zeineddine, Mohammad Zeineddine, Mahmoud Yousef, Shuangjie You, Betul Gunes, David Menter, Scott Kopetz, John Paul Shen. Cell-type-specific gene programs suggest immune and stromal drivers of relapse in colorectal cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 881.