Stromal Signature Research Articles

Stromal Expression Profiling Reveals Immune-Driven Adaption to Malignancy in Canine Melanoma Subtypes.

Canine mucosal melanoma (CMM) is the most common oral malignancy in dogs and is significantly more aggressive than its cutaneous counterpart (CCM), yet the reasons for this disparity remain unclear. Cancer-associated stroma (CAS) plays a crucial role in tumour progression, but a detailed understanding of CAS in canine melanoma is missing. To assess stromal reprogramming, we analysed CAS from 21 CMM, 14 CCM and normal stroma from 10 skin and 9 oral mucosa samples by laser-capture microdissection followed by RNA sequencing. Results were assessed in relation to subtypes, prognostic factors including mitotic count (MC), ulceration, necrosis, pigmentation and immune cell infiltration (CD3, CD20 and CD68), scored using immunohistochemistry and RNA in situ hybridisation. Stromal reprogramming was evident in both subtypes but significantly more pronounced in CMM. Immune-excluded tumours exhibited higher MC than desert/cold ones. MC strongly correlated with genes associated with B-cells, T-helper cells and CTLA4 in CCM, suggesting CAS reprogramming to depend on tumour malignancy. Finally, we identify an immune-suppressive stromal signature in a subset of CMM characterised by the downregulation of key immune checkpoints and pathways. Together, these findings provide a solid foundation for understanding the role of CAS in canine melanoma, specific to cutaneous and mucosal subtypes.

Single cell sequencing delineates T-cell clonality and pathogenesis of the parapsoriasis disease group

BackgroundMycosis fungoides (MF), the most common cutaneous T-cell lymphoma, is often underdiagnosed in early stages due to similarities with benign dermatoses such as atopic dermatitis (AD). Furthermore, the delineation from so-called “parapsoriasis en plaque,” a disease that can appear either in a small- or large-plaque form, is still controversial. ObjectiveTo characterize the parapsoriasis disease spectrum. MethodsWe performed single-cell RNA sequencing of skin biopsies from patients within the parapsoriasis-to-early-stage MF spectrum, stratified for small and large plaques, and compared them to AD, psoriasis and healthy control skin. Results6 out of 8 large-plaque lesions harbored either an expanded alpha/beta or gamma/delta T-cell clone with downregulation of CD7 expression, consistent with a diagnosis of early-stage MF. By contrast, 6 out of 7 small-plaque lesions were polyclonal in nature thereby lacking a lymphomatous phenotype, and also revealed a less inflammatory microenvironment than early-stage MF or AD. Of note, polyclonal small- and large-plaque lesions characteristically harbored a population of NPY+ innate lymphoid cells and displayed a stromal signature of complement upregulation and antimicrobial hyperresponsiveness in fibroblasts and sweat gland cells, respectively. These conditions were clearly distinct from AD or psoriasis, which uniquely harbored CD3+CRTH2+ IL13-expressing “Th2A” cells or strong type 17 inflammation, respectively. ConclusionThese data position polyclonal small- and large-plaque dermatitis lesions as a separate disease entity, that characteristically harbors a so far undescribed ILC population. We thus propose the new term “polyclonal parapsoriasis en plaque” to this kind of lesions, as they can be clearly differentiated from early and advanced-stage MF, psoriasis and AD on several cellular and molecular levels.

Three-dimensional culture in a bioengineered matrix and somatic cell complementation to improve growth and survival of bovine preantral follicles.

Here we explored poly(ethylene glycol) (PEG) bioengineered hydrogels for bovine preantral follicle culture with or without ovarian cell co-culture and examined the potential for differentiation of bovine embryonic stem cells (bESCs) towards gonadal somatic cells to develop a system more similar to the ovarian microenvironment. Bovine preantral follicles were first cultured in two-dimensional (2D) control or within PEG hydrogels (3D) and then co-cultured within PEG hydrogels with bovine ovarian cells (BOCs) to determine growth and viability. Finally, we tested conditions to drive differentiation of bESCs towards the intermediate mesoderm and bipotential gonad fate. Primary follicles grew over the 10-day culture period in PEG hydrogels compared to 2D control. Early secondary follicles maintained a similar diameter within the PEG while control follicles decreased in size. Follicles lost viability after co-encapsulation with BOCs; BOCs lost stromal cell signature over the culture period within hydrogels. Induction of bESCs towards gonadal somatic fate under WNT signaling was sufficient to upregulate intermediate mesoderm ( LHX1 ) and early coelomic epithelium/bipotential gonad markers ( OSR1 , GATA4 , WT1 ). Higher BMP4 concentrations upregulated the lateral plate mesoderm marker FOXF1 . PAX3 expression was not induced, indicating absence of the paraxial mesoderm lineage. Culture of primary stage preantral follicles in PEG hydrogels promoted growth compared to controls; BOCs did not maintain identity in the PEG hydrogels. Collectively, we demonstrate that PEG hydrogels can be a potential culture system for early preantral follicles pending refinements, which could include addition of ESC-derived ovarian somatic cells using the protocol described here. We demonstrate that three-dimensional bioengineered hydrogels could aid in the survival and growth of small bovine preantral follicles. Moreover, bovine embryonic stem cells have the potential to differentiate towards precursors of somatic gonadal cell types, presenting an alternative cell source for preantral follicle co-culture.

Molecular classifications of gastric cancer and their clinical potential

Gastric cancer if the 5th most common oncological disease in the world and one of the leading causes of death associated with this pathology. In clinical practice, the Lauren classification is widely used for gastric cancer characterization, but it does not provide accurate information on tumor progression and does not allow to select the optimal therapeutic approach. More modern tumor typologies, for example proposed by the The Cancer Genome Atlas (TCGA) and the Asian Cancer Research Group (ACRG), are based on profiling of molecular changes in the tumor genome. Currently, several new classifications exist dividing gastric cancer into groups depending on response to different treatment, for example, checkpoint inhibitors or therapy based on activity of pathological pathways associated with immunity, DNA repair, oncogenic and stromal signatures. The proposed typologies improve diagnosis and treatment of this pathology. The review describes currently available classifications of gastric tumors and considers their practical potential.

Detection and assessment of immune and stromal related risk genes to predict preeclampsia: A bioinformatics analysis with dataset.

This study aimed to investigate immune score and stromal score-related signatures associated with preeclampsia (PE) and identify key genes for diagnosing PE using bioinformatics analysis. Four microarray datasets, GSE75010, GSE25906, GSE44711, and GSE10588 were obtained from the Gene Expression Omnibus database. GSE75010 was utilized for differential expressed gene (DEGs) analysis. Subsequently, bioinformatic tools such as gene ontology, Kyoto Encyclopedia of Genes and Genomes, weighted gene correlation network analysis, and gene set enrichment analysis were employed to functionally characterize candidate target genes involved in the pathogenesis of PE. The least absolute shrinkage and selection operator regression approach was employed to identify crucial genes and develop a predictive model. This method also facilitated the creation of receiver operating characteristic (ROC) curves, enabling the evaluation of the model's precision. Furthermore, the model underwent external validation through the other three datasets. A total of 3286 DEGs were identified between normal and PE tissues. Gene ontology and Kyoto Encyclopedia of Genes and Genomes analyses revealed enrichments in functions related to cell chemotaxis, cytokine binding, and cytokine-cytokine receptor interaction. weighted gene correlation network analysis identified 2 color modules strongly correlated with immune and stromal scores. After intersecting DEGs with immune and stromal-related genes, 13 genes were selected and added to the least absolute shrinkage and selection operator regression. Ultimately, 7 genes were screened out to establish the risk model for discriminating preeclampsia from controls, with each gene having an area under the ROC curve >0.70. The constructed risk model demonstrated that the area under the ROC curves in internal and the other three external datasets were all greater than 0.80. A 7-gene risk signature was identified to build a potential diagnostic model and performed well in the external validation group for PE patients. These findings illustrated that immune and stromal cells played essential roles in PE during its progression.

SERPINE1: Role in Cholangiocarcinoma Progression and a Therapeutic Target in the Desmoplastic Microenvironment.

A heterogenous population of inflammatory elements, other immune and nonimmune cells and cancer-associated fibroblasts (CAFs) are evident in solid malignancies where they coexist with the growing tumor mass. In highly desmoplastic malignancies, CAFs are the prominent mesenchymal cell type in the tumor microenvironment (TME), where their presence and abundance signal a poor prognosis. CAFs play a major role in the progression of various cancers by remodeling the supporting stroma into a dense, fibrotic matrix while secreting factors that promote the maintenance of cancer stem-like characteristics, tumor cell survival, aggressive growth and metastasis and reduced sensitivity to chemotherapeutics. Tumors with high stromal fibrotic signatures are more likely to be associated with drug resistance and eventual relapse. Identifying the molecular underpinnings for such multidirectional crosstalk among the various normal and neoplastic cell types in the TME may provide new targets and novel opportunities for therapeutic intervention. This review highlights recent concepts regarding the complexity of CAF biology in cholangiocarcinoma, a highly desmoplastic cancer. The discussion focuses on CAF heterogeneity, functionality in drug resistance, contributions to a progressively fibrotic tumor stroma, the involved signaling pathways and the participating genes.

Identifying Diffuse Glioma Subtypes Based on Pathway Enrichment Evaluation.

Gliomas are highly heterogeneous in molecular, histology, and microenvironment. However, a classification of gliomas by integrating different tumor microenvironment (TME) components remains unexplored.Based on the enrichment scores of 17 pathways involved in immune, stromal, DNA repair, and nervous system signatures in diffuse gliomas, we performed consensus clustering to uncover novel subtypes of gliomas. Consistently in three glioma datasets (TCGA-glioma, CGGA325, and CGGA301), we identified three subtypes: Stromal-enriched (Str-G), Nerve-enriched (Ner-G), and mixed (Mix-G). Ner-G was charactered by low immune infiltration levels, stromal contents, tumor mutation burden, copy number alterations, DNA repair activity, cell proliferation, epithelial-mesenchymal transformation, stemness, intratumor heterogeneity, androgen receptorexpression and EGFR, PTEN, NF1 and MUC16 mutation rates, while high enrichment of neurons and nervous system pathways, and high tumor purity, estrogen receptor expression, IDH1 and CIC mutation rates, temozolomide response rate and overall and disease-free survival rates. In contrast, Str-G displayed contrastive characteristics to Ner-G. Our analysis indicates that the heterogeneity between glioma cells and neurons is lower than that between glioma cells and immune and stromal cells. Furthermore, the abundance of neurons is positively associated with clinical outcomes in gliomas, while the enrichment of immune and stromal cells has a negative association with them. Our classification method provides new insights into the tumor biology of gliomas, as well as clinical implications for the precise management of this disease.

Abstract CT235: RNA-seq molecular subtypes and gene expression signatures in the phase 3 IMpower151 study of 1L bevacizumab + pemetrexed/paclitaxel ± atezolizumab in metastatic NSCLC

Abstract Introduction IMpower151 (NCT04194203) was a phase 3 study to evaluate the efficacy and safety of atezolizumab, bevacizumab, carboplatin, plus paclitaxel or pemetrexed (ABCP) vs BCP, as 1L treatment for metastatic nsqNSCLC. IMpower151 did not meet its primary endpoint of INV-PFS. The objectives of this biomarker analysis were to characterize the molecular subtypes of metastatic nsqNSCLC and to explore the relationship between molecular subtype, predefined gene expression signatures, PD-L1 status and clinical outcomes with ABCP and BCP. Methods RNA sequencing of 250 nsqNSCLC tumor samples from IMpower151 was analyzed in the ACBP (n=123) and BCP (n=127) treatment arms. Non-negative matrix factorization (NMF) was applied to variably expressed genes to define molecular subtypes. Gene expression signatures were defined based on previously published associations with their respective biology. Tumor PD-L1 levels were determined by SP263 immunohistochemistry (Ventana). Results No PD-L1 dependent difference of survival benefit of ABCP compared to BCP was observed at the cutoff of TC 1% or TC 50%. Four molecular subtypes (NMF1-NMF4) were identified based on RNA sequencing. NMF1 (n=78) were immune-depleted and had low tumor proliferation. NMF2 (n=73) were characterized by low level of anti-tumor immune cells, increased expression of stromal signatures and high tumor proliferation. NMF3 (n=65) had the most inflamed transcriptomic profile with high enrichment of immune cells, and increased expression of stromal signatures as well. NMF4 (n=34) had the enrichment of granulocytes, a less inflamed immunophenotype and decreased expression of stromal signatures. The positive rate of PD-L1 expression with TC>=1% was significantly lower in NMF1 compared to other subtypes. None of patients with the four NMF subtypes showed significantly longer PFS or OS with ABCP than BCP. Overall, patients with NMF2 showed numerically shorter PFS and patients with NMF3 showed numerically longer PFS than other subtypes. Gene expression signatures of antitumor cytokines are correlated with survival benefit of ABCP compared to BCP. Patients with high expression level of antitumor cytokines showed substantially improved PFS and OS benefit from ABCP than BCP in both ITT population and EGFR mutation population. Conclusion Four molecular subtypes of nsqNSCLC with different immune-cell and tumor microenvironmental features were characterized. NMF2 may identify a fibrotic and immune-depleted subgroup correlated with worse prognosis. NMF3 may identify an immune-inflamed subgroup correlated with better prognosis. Neither NMF subtypes nor PD-L1 IHC level showed a correlation with survival benefit of ABCP controlled by BCP. The gene expression signature of antitumor cytokines could serve as a potential predictive biomarker for survival benefit of ABCP in metastatic nsqNSCLC including EGFR mutant patients. This finding is based on the exploratory analysis of a single clinical trial with limited sample size. It needs further investigation in other cohorts of tumor tissues and with larger clinical potential in blood samples as well. Citation Format: Caicun Zhou, Charlie Sun, Stacey Huang, Xiaorong Dong, Gongyan Chen, Zhehai Wang, Xianghua Wu, Yu Yao, Yiping Zhang, Ying Cheng, Hongming Pan, Xiaodong Zhang, Jiuwei Cui, Lifeng Wang, Xi Chen, Xiaoling Li, Ziping Wang, Qiming Wang, Jianxing He, Mengzhao Wang, Li Qian, Habib Hamidi, Shanshan Li, Marcus Ballinger, David S. Shames, Minu Srivastava. RNA-seq molecular subtypes and gene expression signatures in the phase 3 IMpower151 study of 1L bevacizumab + pemetrexed/paclitaxel ± atezolizumab in metastatic NSCLC [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr CT235.

CAF-associated genes putatively representing distinct prognosis by in silico landscape of stromal components of colon cancer.

Comprehensive understanding prognostic relevance of distinct tumor microenvironment (TME) remained elusive in colon cancer. In this study, we performed in silico analysis of the stromal components of primary colon cancer, with a focus on the markers of cancer-associated fibroblasts (CAF) and tumor-associated endothelia (TAE), as well as immunological infiltrates like tumor-associated myeloid cells (TAMC) and cytotoxic T lymphocytes (CTL). The relevant CAF-associated genes (CAFG)(representing R index = 0.9 or beyond with SPARC) were selected based on stroma specificity (cancer stroma/epithelia, cS/E = 10 or beyond) and expression amounts, which were largely exhibited negative prognostic impacts. CAFG were partially shared with TAE-associated genes (TAEG)(PLAT, ANXA1, and PTRF) and TAMC-associated genes (TAMCG)(NNMT), but not with CTL-associated genes (CTLG). Intriguingly, CAFG were prognostically subclassified in order of fibrosis (representing COL5A2, COL5A1, and COL12A1) followed by exclusive TAEG and TAMCG. Prognosis was independently stratified by CD8A, a CTL marker, in the context of low expression of the strongest negative prognostic CAFG, COL8A1. CTLG were comprehensively identified as IFNG, B2M, and TLR4, in the group of low S/E, representing good prognosis. Our current in silico analysis of the micro-dissected stromal gene signatures with prognostic relevance clarified comprehensive understanding of clinical features of the TME and provides deep insights of the landscape.

Abstract 6849: Characterizing a discernible stromal fibroblastic gene signature in the shift from high-grade prostatic intraepithelial neoplasia to aggressive prostate cancer

Abstract Background: High-grade prostatic intraepithelial neoplasia (HGPIN) is a preneoplastic lesion that precedes the development of both indolent and aggressive variants of prostate cancer (PCa). However, the specific stromal molecular mechanisms underlying the progression of HGPIN and their contribution to epithelial cancer cell motility and invasiveness remain poorly understood. This study aims to elucidate the driver genomic alterations involved in shaping the perineoplastic stroma of HGPIN, with the goal of unraveling their role in promoting cancer invasion, and metastasis. Methods: We performed gene expression profiling on laser capture microdissected perineoplastic stroma of HGPIN and PCa in tissue samples from 25 PCa cases, including 12 low-grade (Gleason 6) and 13 high-grade (Gleason >/= 8) cases. We used the standard moderated t-test from the Bioconductor limma package to assess the statistical significance of differential gene expression between perineoplastic stroma high grade PCa and perineoplastic stroma HGPIN, as well as between perineoplastic stroma low grade PCa and HGPIN. The -log10 p-values represents the level of statistical significance for each gene, while the log2 fold change represents the magnitude of gene expression difference between the two comparison groups. Additionally, we computed Gene Set Enrichment Analysis (GSEA) scores using the Bioconductor package and discerned the leading-edge genes among the pathways. This study is supported by the NIH-NCI-T32 program for next-generation pathologists at our institution. Results: In the gene expression analysis comparing High Grade PCa and HGPIN, several genes, including POSTN, BGN, GREM1, C1QA, THBS4, COL1A1, COL1A2, SFRP4, RNU6-847P, FCGR2P, and TOMM22, exhibited the highest level of both biological and statistical significance. GSEA revealed that the multi-invasive pathway had the highest enrichment score (>2). Notably, the top overexpressed genes in this pathway exhibited a fibroblastic signature (in bold), including POSTN, BGN, GREM1, FBN1, COL1A2, COL3A1, COL5A2, ASPN, SPARC, VCAN, COMP, INHBA, THBS2, and SFRP4. The presence of this signature indicates the invasiveness of the disease. Importantly, these genes were not associated with the progression to low grade PCa, indicating their exclusive association with the aggressive variant. Conclusion: To our knowledge, this study is the first to explore the stroma's role in HGPIN-to-aggressive PCa progression. It highlights a key stromal pathway enriched with fibroblastic gene signature. These findings offer potential biomarkers for targeted therapies and improved care through HGPIN patient stratification. Citation Format: Mohammad K. Alexanderani, Lucio Queiroz, Mohamed Omar, Claudio Zanettini, Hubert Pakula, Luigi Marchionni, Massimo Loda. Characterizing a discernible stromal fibroblastic gene signature in the shift from high-grade prostatic intraepithelial neoplasia to aggressive prostate 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 6849.

Abstract 1141: Spatial-specific gene signatures to predict immunotherapy outcomes in lung cancer

Abstract The widespread use of immunotherapy in lung cancer, and its more recent approval for early stages, underscores the need for biomarkers that can identify the most responsive patients. Spatial transcriptomics, which maps gene expression in its spatial tissue context, offers a unique approach over bulk transcriptomics. By incorporating spatial information, the predictive power of the signature can be enhanced. Here, we aim to develop spatially informed gene signatures that could be translated into clinical RNA in situ assays, distinguishing patients unlikely to benefit from immunotherapy and thereby sparing them unnecessary side effects. We utilized the NanoString GeoMX Whole Transcriptome Atlas for spatially resolved transcriptomic profiling of retrospectively collected lung cancer tissue samples from patients treated with immunotherapy in an advanced-stage setting (N=60). By targeting 18,190 genes within distinct areas of interest (AOIs)—including stromal (macrophages/CD68+ and leukocytes/CD45+) and tumor (cytokeratin, CK+) cells—we developed AOI-specific gene signatures to predict objective responses. These were derived from a robust computational framework employing LASSO logistic regression on a split-sample approach, yielding predictive models for treatment outcome.We achieved high predictive accuracy on the training set, with the area under the curve (AUC) exceeding 0.86 for all AOI-specific spatial signatures, indicating strong potential for clinical application. Validation against an independent cohort (N=42) corroborated the efficacy of these signatures. Our 6-gene tumor signature was validated with an AUC of 0.73 (95% CI: 0.67-0.89, p = 0.009**), while the CD45 5-gene signature showed an AUC of 0.75 (95% CI: 0.53-0.97, p = 0.022*). Our 18-gene CD68 signature trended towards validation but lacked statistical significance. A combined CD68 and CD45 signature predicted outcomes with greater accuracy, achieving an AUC of 0.79 (95% CI: 0.62-0.98, p = 0.0088*). Following gene set enrichment analysis on our differentially expressed and signature genes, we identified genes with positive coefficients in both tumor and stroma signatures that were associated with glucocorticoid response and glycolytic processes linked to T-cell homeostasis, while genes with negative coefficients were associated with epithelial cell differentiation and cytokine production. These associations are concordant with the observation that genes with positive coefficients are predictors of treatment response, while those with negative coefficients indicate resistance.Our findings indicate that AOI-specific signatures predict the immunotherapy outcome in lung cancer with high accuracy, suggesting that spatial assessment can provide substantial predictive information. The high performance of these signatures indicates their potential for prospective clinical applications. Citation Format: Thazin N. Aung, Myrto Moutafi, Ioannis Trontzas, Arutha Kulasinghe, James Monkman, Niki Gavrielatou, Ioannis Vathiotis, Jonathan H. Warrell, David L. Rimm. Spatial-specific gene signatures to predict immunotherapy outcomes in lung 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 1141.

Stroma-specific gene expression signature identifies prostate cancer subtype with high recurrence risk

Current prognostic tools cannot clearly distinguish indolent and aggressive prostate cancer (PC). We hypothesized that analyzing individual contributions of epithelial and stromal components in localized PC (LPC) could improve risk stratification, as stromal subtypes may have been overlooked due to the emphasis on malignant epithelial cells. Hence, we derived molecular subtypes of PC using gene expression analysis of LPC samples from prostatectomy patients (cohort 1, n = 127) and validated these subtypes in two independent prostatectomy cohorts (cohort 2, n = 406, cohort 3, n = 126). Stroma and epithelium-specific signatures were established from laser-capture microdissection data and non-negative matrix factorization was used to identify subtypes based on these signatures. Subtypes were functionally characterized by gene set and cell type enrichment analyses, and survival analysis was conducted. Three epithelial (E1-E3) and three stromal (S1-S3) PC subtypes were identified. While subtyping based on epithelial signatures showed inconsistent associations to biochemical recurrence (BCR), subtyping by stromal signatures was significantly associated with BCR in all three cohorts, with subtype S3 indicating high BCR risk. Subtype S3 exhibited distinct features, including significantly decreased cell-polarity and myogenesis, significantly increased infiltration of M2-polarized macrophages and CD8 + T-cells compared to subtype S1. For patients clinically classified as CAPRA-S intermediate risk, S3 improved prediction of BCR. This study demonstrates the potential of stromal signatures in identification of clinically relevant PC subtypes, and further indicated that stromal characterization may enhance risk stratification in LPC and may be particularly promising in cases with high prognostic ambiguity based on clinical parameters.

Abstract A032: Low non-immune stromal cell levels and high hypoxic tumor microenvironments are associated with pathologic complete response in the neoadjuvant I-SPY2 TRIAL

Abstract Background: Non-immune components of the tumor microenvironment (TME) play important roles in cancer. We explore two important components of the TME: non-immune stromal cell populations including fibroblasts, adipocytes, and endothelial cells; and hypoxic conditions. We leverage published gene expression-based biomarkers of stromal cell types and hypoxia, and molecular and clinical data from the I-SPY2 TRIAL (PMID: 35623341) – an adaptive neoadjuvant trial for locally advanced breast cancers. We assess the contribution of stromal cell populations and oxygen-poor environments at pre-treatment in shaping the landscape of immune enrichment and therapeutic response in breast cancers. Methods: We curated 14 stromal and 11 hypoxia gene sets, in addition to 35 immune gene sets from literature. Signatures scores of stromal, hypoxic, and immune TME were calculated from gene expression profiles of pre-treatment tumor samples from I-SPY2 patients (n=987) in the first 9 treatment arms of the trial plus control arms. Correlation of stromal and hypoxia signatures with immune signatures were assessed, and association of TME features with HER2/HR receptor status, I-SPY response-predictive subtype (RPS), and pathological complete response (pCR) were evaluated using non-parametric tests. All p-values were adjusted for multiple testing. Significance levels were defined at adj. p<0.05. Results: Both stromal and hypoxia signatures showed significant differential enrichment across HR/HER2 and RPS subtypes, and association with pCR. Several stromal signatures including adipocytes, endothelial cells, mesenchymal stem cells, and overall stromal score had the lowest enrichment in HR-/HER2- breast cancers, and in the RPS HER2-/immune+ subtype. Moreover, most stromal signatures showed anti-correlation with CD8+ and CD4+ T cell enrichment. Stromal signatures including endothelial cells, smooth muscle cells, and overall score were significantly lower in patients who achieved pCR. Two hypoxia gene sets demonstrated highly significant enrichment in HR- versus HR+ samples regardless of HER2 status. All hypoxia signatures varied significantly across RPS subtypes, with the highest scores in HER2-/Immune-/DRD+, especially relative to HER2-/Immune-/DRD- and HER2+/BP-Luminal subtypes. Hypoxia signatures strongly linked to hypoxia inducible factor (HIF) and angiogenesis via VEGF were positively associated with pCR. These hypoxia signatures showed treatment-specific associations with pCR, specifically in agents targeting mechanisms with known interactions with hypoxia-induced biology: trebananib (angiogenesis inhibitor), ganetespib (HSP90 inhibitor) and veliparib/carboplatin (DNA damage repair inhibitor). Conclusions: Hypoxia and stromal cell gene signatures show breast cancer subtype-specific biology. Relative absence of non-immune stromal cell types is associated with higher immune infiltration and higher pCR rates, while specific hypoxia pathways were associated with response to therapeutic agents targeting hypoxia-related pathways. Citation Format: Kailey Dubinsky, Elene Tsopurashvili, Denise M. Wolf, Zheyun Xu, Silver Al Khafaji Alkhafaji, Anna van Montfort, Christina Yau, Diane Heditsian, Allison Fiscalini, Laura J. Esserman, Laura van ‘t Veer, Rosalyn W. Sayaman. Low non-immune stromal cell levels and high hypoxic tumor microenvironments are associated with pathologic complete response in the neoadjuvant I-SPY2 TRIAL [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Breast Cancer Research; 2023 Oct 19-22; San Diego, California. Philadelphia (PA): AACR; Cancer Res 2024;84(3 Suppl_1):Abstract nr A032.

Single-cell transcriptome analysis of epithelial, immune, and stromal signatures and interactions in human ovarian cancer

A comprehensive investigation of ovarian cancer (OC) progression at the single-cell level is crucial for enhancing our understanding of the disease, as well as for the development of better diagnoses and treatments. Here, over half a million single-cell transcriptome data were collected from 84 OC patients across all clinical stages. Through integrative analysis, we identified heterogeneous epithelial-immune-stromal cellular compartments and their interactions in the OC microenvironment. The epithelial cells displayed clinical subtype features with functional variance. A significant increase in distinct T cell subtypes was identified including Tregs and CD8+ exhausted T cells from stage IC2. Additionally, we discovered antigen-presenting cancer-associated fibroblasts (CAFs), with myofibroblastic CAFs (myCAFs) exhibiting enriched extracellular matrix (ECM) functionality linked to tumor progression at stage IC2. Furthermore, the NECTIN2-TIGIT ligand-receptor pair was identified to mediate T cells communicating with epithelial, fibroblast, endothelial, and other cell types. Knock-out of NECTIN2 using CRISPR/Cas9 inhibited ovarian cancer cell (SKOV3) proliferation, and increased T cell proliferation when co-cultured. These findings shed light on the cellular compartments and functional aspects of OC, providing insights into the molecular mechanisms underlying stage IC2 and potential therapeutic strategies for OC.

Distinct mesenchymal cell states mediate prostate cancer progression

In the complex tumor microenvironment (TME), mesenchymal cells are key players, yet their specific roles in prostate cancer (PCa) progression remain to be fully deciphered. This study employs single-cell RNA sequencing to delineate molecular changes in tumor stroma that influence PCa progression and metastasis. Analyzing mesenchymal cells from four genetically engineered mouse models (GEMMs) and correlating these findings with human tumors, we identify eight stromal cell populations with distinct transcriptional identities consistent across both species. Notably, stromal signatures in advanced mouse disease reflect those in human bone metastases, highlighting periostin’s role in invasion and differentiation. From these insights, we derive a gene signature that predicts metastatic progression in localized disease beyond traditional Gleason scores. Our results illuminate the critical influence of stromal dynamics on PCa progression, suggesting new prognostic tools and therapeutic targets.

Immune and stromal transcriptional patterns that influence the outcome of classic Hodgkin lymphoma

Classic Hodgkin lymphoma (cHL) is characterized by a rich immune microenvironment as the main tumor component. It involves a broad range of cell populations, which are largely unexplored, even though they are known to be essential for growth and survival of Hodgkin and Reed–Sternberg cells. We profiled the gene expression of 25 FFPE cHL samples using NanoString technology and resolved their microenvironment compositions using cell-deconvolution tools, thereby generating patient-specific signatures. The results confirm individual immune fingerprints and recognize multiple clusters enriched in refractory patients, highlighting the relevance of: (1) the composition of immune cells and their functional status, including myeloid cell populations (M1-like, M2-like, plasmacytoid dendritic cells, myeloid-derived suppressor cells, etc.), CD4-positive T cells (exhausted, regulatory, Th17, etc.), cytotoxic CD8 T and natural killer cells; (2) the balance between inflammatory signatures (such as IL6, TNF, IFN-γ/TGF-β) and MHC-I/MHC-II molecules; and (3) several cells, pathways and genes related to the stroma and extracellular matrix remodeling. A validation model combining relevant immune and stromal signatures identifies patients with unfavorable outcomes, producing the same results in an independent cHL series. Our results reveal the heterogeneity of immune responses among patients, confirm previous findings, and identify new functional phenotypes of prognostic and predictive utility.

Spatial proteomic profiling of tumor and stromal compartments in non-small-cell lung cancer identifies signatures associated with overall survival.

Non-small-cell lung carcinoma (NSCLC) is the most prevalent and lethal form of lung cancer. The need for biomarker-informed stratification of targeted therapies has underpinned the need to uncover the underlying properties of the tumor microenvironment (TME) through high-plex quantitative assays. In this study, we profiled resected NSCLC tissues from 102 patients by targeted spatial proteomics of 78 proteins across tumor, immune activation, immune cell typing, immune-oncology, drug targets, cell death and PI3K/AKT modules to identify the tumor and stromal signatures associated with overall survival (OS). Survival analysis revealed that stromal CD56 (HR = 0.384, P = 0.06) and tumoral TIM3 (HR = 0.703, P = 0.05) were associated with better survival in univariate Cox models. In contrast, after adjusting for stage, BCLXL (HR = 2.093, P = 0.02) and cleaved caspase 9 (HR = 1.575, P = 0.1) negatively influenced survival. Delta testing indicated the protective effect of TIM-3 (HR = 0.614, P = 0.04) on OS. In multivariate analysis, CD56 (HR = 0.172, P = 0.001) was associated with better survival in the stroma, while B7.H3 (HR = 1.72, P = 0.008) was linked to poorer survival in the tumor. Deciphering the TME using high-plex spatially resolved methods is giving us new insights into compartmentalised tumor and stromal protein signatures associated with clinical endpoints in NSCLC.

Whole stromal fibroblast signature is linked to specific chemokine and immune infiltration patterns and to improved survival in NSCLC

ABSTRACT Cancer associated fibroblasts (CAF) are known to orchestrate multiple components of the tumor microenvironment, whereas the influence of the whole stromal-fibroblast compartment is less understood. Here, an extended stromal fibroblast signature was investigated to define its impact on immune cell infiltration. The lung cancer adenocarcinoma (LUAD) data set of the cancer genome atlas (TCGA) was used to test whole stroma signatures and cancer-associated fibroblast signatures for their impact on prognosis. 3D cell cultures of the NSCLC cancer cell line A549 together with the fibroblast cell line SV80 were used in combination with infiltrating peripheral blood mononuclear cells (PBMC) for in-vitro investigations. Immune cell infiltration was assessed via flow cytometry, chemokines were analyzed by immunoassays and RNA microarrays. Results were confirmed in specimens from NSCLC patients by flow cytometry or immunohistochemistry as well as in the TCGA data set. The TCGA analyses correlated the whole stromal-fibroblast signature with an improved outcome, whereas no effect was found for the CAF signatures. In 3D microtumors, the presence of fibroblasts induced infiltration of B cells and CD69+CD4+ T cells, which was linked to an increased expression of CCL13 and CXCL16. The stroma/lymphocyte interaction was confirmed in NSCLC patients, as stroma-rich tumors displayed an elevated B cell count and survival in the local cohort and the TCGA data set. A whole stromal fibroblast signature was associated with an improved clinical outcome in lung adenocarcinoma and in vitro and in vivo experiments suggest that this signature increases B and T cell recruitment via induction of chemokines.

Establishment of a 7-gene expression panel to improve the prognosis classification of gastric cancer patients.

Gastric cancer (GC) ranks fifth in incidence and fourth in mortality worldwide. The high death rate in patients with GC requires new biomarkers for improving survival estimation. In this study, we performed a transcriptome-based analysis of five publicly available cohorts to identify genes consistently associated with prognosis in GC. Based on the ROC curve, patients were categorized into high and low-expression groups for each gene using the best cutoff point. Genes associated with survival (AUC > 0.5; univariate and multivariate Cox regressions, p < 0.05) were used to model gene expression-based scores by weighted sum using the pooled Cox β regression coefficients. Cox regression (p < 0.05), AUC > 0.5, sensitivity > 0.5, and specificity > 0.5 were considered to identify the best scores. Gene set enrichment analysis (KEGG, REACTOME, and Gene Ontology databases), as well as microenvironment composition and stromal cell signatures prediction (CIBERSORT, EPIC, xCell, MCP-counter, and quanTIseq web tools) were performed. We found 11 genes related to GC survival in the five independent cohorts. Then, we modeled scores by calculating all possible combinations between these genes. Among the 2,047 scores, we identified a panel based on the expression of seven genes. It was named GES7 and is composed of CCDC91, DYNC1I1, FAM83D, LBH, SLITRK5, WTIP, and NAP1L3 genes. GES7 features were validated in two independent external cohorts. Next, GES7 was found to recategorize patients from AJCC TNM stages into a best-fitted prognostic group. The GES7 was associated with activation of the TGF-β pathway and repression of anticancer immune cells. Finally, we compared the GES7 with 30 previous proposed scores, finding that GES7 is one of the most robust scores. As a result, the GES7 is a reliable gene-expression-based signature to improve the prognosis estimation in GC.

Biology and performance of pre- and post-pembrolizumab (pembro) vesical imaging–reporting and data system (VI-RADS) to predict the pathological response in muscle-invasive urothelial bladder cancer (MIBC): Full data analysis from a clinical trials pipeline.

4591 Background: The VI-RADS is a standardized reporting system that uses multiparametric magnetic resonance imaging (mpMRI) parameters to predict the probability of MIBC. It has been also used in studies using standard neoadjuvant chemotherapy. No studies have analyzed the performance of VI-RADS to predict the pT0 or pT≤1 response post-immunotherapy (IO) in MIBC. Methods: In PURE-01 study patients (pts) were staged with mpMRI before and after treatment (3 cycles of pembro) prior to radical cystectomy (RC). All mpMRI scans were centrally reviewed. Logistic regression models analyzed pre- and post-pembro VI-RADS against pT≤1 (primary endpoint) and pT0 (secondary endpoint). VI-RADS scores were dichotomized between 0-3 (0=no evidence of disease) and 4-5. Covariates included cT-stage, age, gender, PD-L1 combined positive score (CPS) and tumor mutational burden (TMB). Model performance was also tested in the ongoing NureCombo (nivolumab+nab-paclitaxel; NCT04876313) and SURE-01 (sacituzumab govitecan; NCT05226117) studies. Transcriptome-wide expression profiling was matched with VI-RADS scores. Results: In PURE-01, 115 pts had centrally-reviewed MRI scans (N=230 mpMRI), treated between 02/17 and 08/18. In total, 32 (27.8%) had pure/predominant non-urothelial carcinoma (UC) histology, 54 (47%) had a cT3-4N0 MIBC. Pre-pembro: 21 pts (18.3%) had no measurable disease (VI-RADS=0), 37 (32.2%) a VI-RADS 1-3 score, and 57 (49.5%) had a VI-RADS 4-5 score. 20 pts (17.4%) had a downstage from VI-RADS 4-5 to VIRADS 0-3 post-pembro. Pre-pembro VIRADS 4-5 were associated with higher angiogenesis and epithelial-mesenchymal transition (EMT) activity vs VI-RADS 0-3 (p=0.011 and 0.033). Both pre-pembro and post-pembro VI-RADS 0-3 scores were the only significant covariates against pT≤1 endpoint on multivariable analyses (MVA): the strongest effect was seen with post-pembro VI-RADS 0-3 against pT≤1 response (OR: 23.4, 95%CI: 7-95.3, p&lt;0.0001). The AUC of this model was 0.90. Model-based decisions were supported in the range of threshold probabilities of 10-80%. Post-pembro VI-RADS 4-5 were characterized by higher stromal signature scores vs VI-RADS 0-3. When post-therapy VI-RADS were tested in evaluable pts from the combined NureCombo+SURE-01 cohort (N=17pts, 34 mpMRI): 90% of post-VI-RADS 0-3 revealed a pT≤1 stage at radical cystectomy. Conclusions: VI-RADS scores mirrored distinct biological features in MIBC, recapitulating tissue biomarkers of IO response. VI-RADS scores post-IO revealed a strong association with pathological downstaging and represented a robust feature based on which selecting pts for bladder-sparing strategies. Results seemed to be also applicable to other novel neoadjuvant therapies beyond IO monotherapy. Clinical trial information: NCT02736266 .