Survival Signature Research Articles

Metabolite signatures of chronological age, aging, survival, and longevity

Metabolites that mark aging are not fully known. We analyze 408 plasma metabolites in Long Life Family Study participants to characterize markers of age, aging, extreme longevity, and mortality. We identify 308 metabolites associated with age, 258 metabolites that change over time, 230 metabolites associated with extreme longevity, and 152 metabolites associated with mortality risk. We replicate many associations in independent studies. By summarizing the results into 19 signatures, we differentiate between metabolites that may mark aging-associated compensatory mechanisms from metabolites that mark cumulative damage of aging and from metabolites that characterize extreme longevity. We generate and validate a metabolomic clock that predicts biological age. Network analysis of the age-associated metabolites reveals a critical role of essential fatty acids to connect lipids with other metabolic processes. These results characterize many metabolites involved in aging and point to nutrition as a source of intervention for healthy aging therapeutics.

Prognostic Impact of Acute and Chronic Inflammatory Interleukin Signatures in the Tumor Microenvironment of Early Breast Cancer

Interleukins play dual roles in breast cancer, acting as both promoters and inhibitors of tumorigenesis within the tumor microenvironment, shaped by their inflammatory functions. This study analyzed the subtype-specific prognostic significance of an acute inflammatory versus a chronic inflammatory interleukin signature using microarray-based gene expression analysis. Correlations between these interleukin signatures and immune cell markers (CD8, IgKC, and CD20) and immune checkpoints (PD-1) were also evaluated. This study investigated the prognostic significance of an acute inflammatory IL signature (IL-12, IL-21, and IFN-γ) and a chronic inflammatory IL signature (IL-4, IL-5, IL-10, IL-13, IL-17, and CXCL1) for metastasis-free survival (MFS) using Kaplan–Meier curves and Cox regression analyses in a cohort of 461 patients with early breast cancer. Correlations were analyzed using the Spearman–Rho correlation coefficient. Kaplan–Meier curves revealed that the prognostic significance of the acute inflammatory IL signature was specifically pronounced in the basal-like subtype (p = 0.004, Log Rank). This signature retained independent prognostic significance in multivariate Cox regression analysis (HR 0.463, 95% CI 0.290–0.741; p = 0.001). A higher expression of the acute inflammatory IL signature was associated with longer MFS. The chronic inflammatory IL signature showed a significant prognostic effect in the whole cohort, with higher expression associated with shorter MFS (p = 0.034). Strong correlations were found between the acute inflammatory IL signature and CD8 expression (ρ = 0.391; p < 0.001) and between the chronic inflammatory IL signature and PD-1 expression (ρ = 0.627; p < 0.001). This study highlights the complex interaction between acute and chronic inflammatory interleukins in breast cancer progression and prognosis. These findings provide insight into the prognostic relevance of interleukin expression patterns in breast cancer and may inform future therapeutic strategies targeting the immune–inflammatory axis.

An Order Statistics Perspective for System Reliability

ABSTRACTStructural reliability integrates the design variables over the safety region characterized by a positive limit state function when the reliability of the entire system is of concern. Calculating the structure function can be challenging for high‐dimensional systems or intricate system architectures. In order to enhance the efficiency of time‐dependent system reliability assessment, we emulate the integral formulation in structural reliability. To elaborate further, we treat each individual unit's lifetime variable as a design variable and subsequently perform calculations involving multiple integrals. Given the ordered nature of unit failure times, we leverage the order statistics distribution to simplify the multiple integrals into a double integral, and then multiply this result by the survival signature to obtain reliability. A two‐terminal nine‐unit network system configuration is illustrated to assess the performance and effectiveness of the proposed method.

Rapid computation of survival signature for dynamic fault tree based on sequential binary decision diagram and multidimensional array

Many practical safety-critical systems typically exhibit sequence-dependent failure behaviors, limiting the efficiency of analyzing these systems. Although the survival signature-based method can address this problem to a certain extent, the dependence on Boolean states constrains its application to large systems. In this study, we present a novel method that leverages the sequential binary decision diagram (SBDD) and multidimensional array to rapidly compute survival signatures for dynamic fault trees (DFTs) of these systems. These dynamic nodes in the SBDD are represented through multidimensional arrays, which are then utilized as inputs for the subsequent computations. Ultimately, survival signatures are obtained by iteratively computing the multidimensional arrays. Additionally, two practical engineering cases are examined to highlight the superiority of the proposed methods over other methods. Compared with Boolean state vector-based methods, the proposed method achieves a 689-fold and 209-fold increase in efficiency for calculating survival signatures in their respective cases. Compared with the Monte Carlo (MC) simulation, the simulation efficiency for the reliability results improve by 60-fold and 201-fold in their respective cases.

Time-based redeployment of multi-class nodes for reliable wireless sensor network coverage

We consider the problem of redeploying nodes into a wireless sensor network (WSN) to maintain reliable area coverage over time as nodes fail. Specifically, we consider the class of time-based node redeployment policies in which the WSN is inspected after a fixed amount of time, after which new sensor nodes are redeployed to bring the number of functioning nodes to a desired level. Whereas previous research on time-based node redeployment assumes nodes are identical with respect to time to failure, we use multiple classes of sensor nodes to represent a scenario where nodes’ times to failure are dependent on positioning in the network. We propose a partial survival signature (PSS) approach for estimating area coverage reliability under a given time-based redeployment policy, where the PSS is estimated by Monte Carlo simulation. This PSS representation enables efficient re-evaluation of coverage reliability under different redeployment policies, thus allowing the use of metaheuristics to obtain a set of redeployment policies that are near-efficient with respect to cost and coverage reliability. We present a numerical example to demonstrate that the PSS approach yields accurate estimates of coverage reliability within a reasonable amount of computation time. Furthermore, we apply non-dominated sorting genetic algorithm II (NSGA-II) to optimize the numerical example with respect to cost and coverage reliability.

Construction of a Cancer Stem Cell Marker Genes-Related 22-Gene Signature for Overall Survival Prediction in High-Risk Wilms’ Tumor

Background: This study aimed to investigate the comprehensive expression profile of cancer stem cell (CSC)-related genes and construct a prognostic signature for overall survival (OS) prediction in high-risk Wilms’ tumor (WT). Materials and methods: Gene expression and survival data from 120 high-risk WT cases in the Therapeutically Applicable Research to Generate Effective Treatments (TARGET)-WT were used. Results: In total, 229 CSC-related genes were found to be significantly dysregulated in WT compared to tumor-adjacent normal tissues, among which 34 were associated with OS. Using LASSO regression, a 22-gene signature was developed, which exhibited excellent performance in 3-, 5-, and 10-year OS predictions (AUC > 0.86). The high-risk score group showed markedly poorer OS compared to the low-risk score group (median separation, HR = 6.41, 95% CI: 3.18–12.92, p = 3.2e − 9). The 22-gene signature was an independent prognostic factor for OS (HR = 5.086, 95% CI: 3.019–8.568, p < 0.001). Conclusion: This study identified a robust prognostic signature that can effectively support OS prediction.

Estimation of multicomponent system reliability for inverse Weibull distribution using survival signature

Estimation of multicomponent system reliability for inverse Weibull distribution using survival signature

A T2-weighted MRI-based radiomic signature for disease-free survival in locally advanced cervical cancer following chemoradiation: An international, multicentre study

IntroductionTo develop and validate a T2-weighted magnetic resonance imaging (MRI)-based radiomic signature associated with disease-free survival (DFS) in locally advanced cervical cancer. Materials and MethodsThe study comprised a training dataset of 132 patients (93 Norwegian; 39 The Cancer Imaging Archive (TCIA) and an independent validation Canadian dataset of 199 patients with FIGO stage IB-IVA cervical cancer treated with chemoradiation. Radiomic features were extracted using PyRadiomics. A radiomic signature was developed based on a multivariable radiomic prognostic model for DFS built using the training dataset, with minimal redundancy maximum relevancy feature selection method. Univariate and multivariable Cox regression analyses were then conducted to examine the association of the derived radiomic signature with DFS. ResultsA radiomic signature was prognostic for DFS in the training cohort (Norwegian hazard ratio [HR] 5.54, p = 0.002; TCIA HR 3.59, p = 0.04). The radiomic signature remained independently associated with DFS (HR 3.70, p = 0.004) when adjusted for stage and tumor volume. The radiomic signature was also prognostic for DFS in the validation cohort, both on univariate analysis (HR 2.22, p = 0.003), and multivariable analysis adjusted for stage and tumor volume (HR 1.84, p = 0.04). The 4-year DFS rates of patients with radiomic signature score > 0 vs ≤ 0 were 48.2 % vs 87.9 %, and 56.4 % vs 80.8 % for training and validation cohorts respectively. ConclusionAn MRI-based radiomic signature can be used as a prognostic biomarker for DFS in patients with locally advanced cervical cancer undergoing chemoradiation.

Sampling and active learning methods for network reliability estimation using K-terminal spanning tree

Network reliability analysis remains a challenge due to the increasing size and complexity of networks. This paper presents a novel sampling method and an active learning method for efficient and accurate network reliability estimation under node failure and edge failure scenarios. The proposed sampling method adopts Monte Carlo technique to sample component lifetimes and the K-terminal spanning tree algorithm to accelerate structure function computation. Unlike existing methods that compute only one structure function value per sample, our method generates multiple component state vectors and corresponding structure function values from each sample. Network reliability is estimated based on survival signatures derived from these values. A transformation technique extends this method to handle both node failure and edge failure. To enhance efficiency of proposed sampling method and achieve adaptability to network topology changes, we introduce an active learning method utilizing a random forest (RF) classifier. This classifier directly predicts structure function values, integrates network behaviors across diverse topologies, and undergoes iterative refinement to enhance predictive accuracy. Importantly, the trained RF classifier can directly predict reliability for variant networks, a capability beyond the sampling method alone. Through investigating several network examples and two practical applications, the effectiveness of both proposed methods is demonstrated.

Predicting prostate adenocarcinoma patients’ survival and immune signature: a novel risk model based on telomere-related genes

Alterations in telomeres constitute some of the earliest occurrences in the tumourigenesis of prostate adenocarcinoma (PRAD) and persist throughout the progression of the tumour. While the activity of telomerase and the length of telomeres have been demonstrated to correlate with the prognosis of PRAD, the prognostic potential of telomere-related genes (TRGs) in this disease remains unexplored. Utilising mRNA expression data from the Cancer Genome Atlas (TCGA), we devised a risk model and a nomogram to predict the survival outcomes of patients with PRAD. Subsequently, our investigations extended to the relationship between the risk model and immune cell infiltration, sensitivity to chemotherapeutic drugs, and specific signalling pathways. The risk model we developed is predicated on seven key TRGs, and immunohistochemistry results revealed significant differential expression of three TRGs in tumours and paracancerous tissues. Based on the risk scores, PRAD patients were stratified into high-risk and low-risk cohorts. The Receiver operating characteristics (ROC) and Kaplan–Meier survival analyses corroborated the exceptional predictive performance of our novel risk model. Multivariate Cox regression analysis indicated that the risk score was an independent risk factor associated with Overall Survival (OS) and was significantly associated with T and N stages of PRAD patients. Notably, the high-risk group exhibited a greater response to chemotherapy and immunosuppression compared to the low-risk group, offering potential guidance for treatment strategies for high-risk patients. In conclusion, our new risk model, based on TRGs, serves as a reliable prognostic indicator for PRAD. The model holds significant value in guiding the selection of immunotherapy and chemotherapy in the clinical management of PRAD patients.

Preventive maintenance policy optimization for a weighted k-out-of-n: G system using the survival signature

This paper presents a novel preventive maintenance (PM) policy for a weighted k-out-of-n: G system. The system consists of critical components with high weight and other low-weight components, while the critical ones significantly impact the system's total weight. The lifetime distribution of critical and other components follows the Weibull distribution with different shape and scale parameters. Different maintenance actions are performed on the component according to the number of failed critical components and the total weight of the working components. The expected cost per renewal cycle and system availability are analyzed based on survival signature. The optimal PM cycle and the inspection period are derived by minimizing the expected cost per renewal cycle or maximizing availability. Finally, the effectiveness of the presented strategy is illustrated by a numerical case.

Survival Signature Approach for Reliability Evaluation of Linear Consecutive \(k\)-out-of-\(n\):\(G\) System

Survival Signature Approach for Reliability Evaluation of Linear Consecutive \(k\)-out-of-\(n\):\(G\) System

A Comparative Analysis of Secreted miRNAs Reveals Candidate Biomarkers for Pleural Mesothelioma Detection

Abstract Background MicroRNAs secreted by tumour cells through exosomes or in non-encapsulated form, were shown to promote tumour growth and treatment resistance, underscoring their potential as cancer biomarkers. Aims Here, we compare miRNAs secreted by pleural mesothelioma (PM) primary cells with miRNAs secreted by non-PM cells, in order to identify diagnostic biomarkers of this devastating disease. Methods We established primary cell cultures from pleural effusion of 12 PM and 7 non-PM patients. Secreted miRNAs were profiled as: (1) total secreted miRNA in cell culture supernatant (Sup), and (2) exosomal (Exo) miRNAs. Exosomes were extracted using iZON qEV columns. RNA was extracted with mirVana PARIS kit. Sequencing libraries were prepared using the QIAseq miRNA Library Kit (Qiagen). Reads were mapped on the mature miRNA sequences (MirBaseDB), followed by DESeq2 differential expression analysis. MiRNA target genes were selected using MirDB and functionally annotated using DAVID. Results We identified 309 and 84 deregulated miRNAs in PM-Exo and PM-Sup samples, respectively (p-value &amp;lt;0.05, Fig. 1A). Among the up-regulated miRNAs, we found let-7c-3p and miR-16-5p, which are known tumour suppressors, and miR-23a-3p and miR-30a-5p (Fig. 1B). Interestingly, miR-30a-5p belongs to the same family as miR-30e-5p, which together with miR-23a-3p was previously identified as part of the long survival signature in PM. Additionally, we identified 11 miRNAs that showed up-regulation in both PM-Exo and PM-Sup. Further analysis of these candidate miRNAs revealed that exosomal expression of miR-30a-5p was a significant predictor of patient survival (Fig. 1C). Finally, we identified potential target genes of the 11 up-regulated miRNAs - the target genes were significantly associated with transcriptional regulation and cell division (p-value &amp;lt;0.05). Conclusion Our comparative analysis of secreted miRNAs identified 11 candidate biomarkers for PM. Our findings emphasise the importance of the PM secretome in advancing our understanding of mesothelioma biology and discovering novel cell-free biomarkers. Further in-depth analyses are currently underway.

Identification of a novel matrix metalloproteinases-related prognostic signature in hepatocellular carcinoma.

Hepatocellular carcinoma (HCC) is the most common primary liver cancer worldwide. Cancer cells' local infiltration, proliferation, and spread are mainly influenced by the protein hydrolyzing function of different matrix metalloproteinases (MMPs). However, no study has determined the relationship between MMPs and prognostic prediction in HCC. Expression profiles of mRNA and MMPs-related genes were obtained from publicly available databases. Cox regression and LASSO Cox regression analysis were used to identify and predict MMPs-related prognostic signature and construct predictive models for overall survival (OS). A nomogram was used to validate the accuracy of the prediction model. Drug prediction was performed using the Genomics of Drug Sensitivity in Cancer (GDSC) dataset, and single-cell clustering analysis was performed to further understand the significance of the MMPs-related signature. A MMPs-related prognostic signature (including RNPEPL1, ADAM15, ADAM18, ADAMTS5, CAD, YME1L1, AMZ2, PSMD14, and COPS6) was identified. Using the median value, HCC patients in the high-risk group showed worse OS than those in the low-risk group. Immune microenvironment analysis showed that patients in the high-risk group had higher levels of M0 and M2 macrophages. Drug sensitivity analysis revealed that the IC50 values of sorafenib, cisplatin, and cytarabine were higher in the high-risk group. Finally, the single-cell cluster analysis results showed that YME1L1 and COPS6 were the major genes expressed in the monocyte cluster. A novel MMPs-related signature can be used to predict the prognosis of HCC. The findings of this research could potentially impact the predictability of the prognosis and treatment of HCC.

RICTOR/mTORC2 downregulation in BRAFV600E melanoma cells promotes resistance to BRAF/MEK inhibition.

The main drawback of BRAF/MEK inhibitors (BRAF/MEKi)-based targeted therapy in the management of BRAF-mutated cutaneous metastatic melanoma (MM) is the development of therapeutic resistance. We aimed to assess in this context the role of mTORC2, a signaling complex defined by the presence of the essential RICTOR subunit, regarded as an oncogenic driver in several tumor types, including MM. After analyzing The Cancer Genome Atlas MM patients' database to explore both overall survival and molecular signatures as a function of intra-tumor RICTOR levels, we investigated the effects of RICTOR downregulation in BRAFV600E MM cell lines on their response to BRAF/MEKi. We performed proteomic screening to identify proteins modulated by changes in RICTOR expression, and Seahorse analysis to evaluate the effects of RICTOR depletion on mitochondrial respiration. The combination of BRAFi with drugs targeting proteins and processes emerged in the proteomic screening was carried out on RICTOR-deficient cells in vitro and in a xenograft setting in vivo. Low RICTOR levels in BRAF-mutated MM correlate with a worse clinical outcome. Gene Set Enrichment Analysis of low-RICTOR tumors display gene signatures suggestive of activation of the mitochondrial Electron Transport Chain (ETC) energy production. RICTOR-deficient BRAFV600E cells are intrinsically tolerant to BRAF/MEKi and anticipate the onset of resistance to BRAFi upon prolonged drug exposure. Moreover, in drug-naïve cells we observed a decline in RICTOR expression shortly after BRAFi exposure. In RICTOR-depleted cells, both mitochondrial respiration and expression of nicotinamide phosphoribosyltransferase (NAMPT) are enhanced, and their pharmacological inhibition restores sensitivity to BRAFi. Our work unveils an unforeseen tumor-suppressing role for mTORC2 in the early adaptation phase of BRAFV600E melanoma cells to targeted therapy and identifies the NAMPT-ETC axis as a potential therapeutic vulnerability of low RICTOR tumors. Importantly, our findings indicate that the evaluation of intra-tumor RICTOR levels has a prognostic value in metastatic melanoma and may help to guide therapeutic strategies in a personalized manner.

Abstract 5061: Improved survival and unique mutational signatures in small cell lung cancer arising in patients with limited tobacco use

Abstract Small-cell lung cancer (SCLC) remains the deadliest histologic lung cancer subtype with a median survival of only 12.3 months among trial-eligible patients with extensive-stage (ES-SCLC) disease, as demonstrated in the IMPower133 trial. Heavy tobacco use, &amp;gt;40 pack-years, is strongly associated with developing SCLC, which near universally harbors RB1 and TP53 co-mutations. However, there is a paucity of information regarding SCLC tumors that occur in patients with minimal tobacco use. This represents an unmet need given an increasing appreciation for low pack-year and never-smoker SCLC with some studies, like the CAPSTONE-1 trial, demonstrating an incidence of never-smoking SCLC &amp;gt;20%. To address this, we clinically characterized the treatments and outcomes of a 164 SCLC patient cohort with ≤10 pack year smoking histories treated at the University of Texas MD Anderson Cancer Center; 22 of which underwent whole-exome sequencing (WES). The overall survival of limited-stage (LS-SCLC) and extensive-stage (ES-SCLC) disease in the low pack-year cohort were 33.6 and 16.6 months, respectively (P=0.003); superior to historic medians of roughly 17.0 and 12.3 months, respectively. Among those with ES-SCLC, those who reported 1-10 pack-years had significantly improved median overall survival as compared to never-smokers; 19.5 vs 15.8 months, respectively (P=0.02). This trend was observed in LS-SCLC as well with 1-10 pack-year and never-smokers demonstrating median overall survivals of 88.6 vs 26.2 months, respectively (P=0.2). Demographic differences including age were similar across these cohorts. WES of tumors from 22 patients demonstrated significantly fewer mutations in RB1 and TP53 than was observed in previously reported cohorts of SCLC from smokers. Among the low-pack year cohort, 7/22 (32% vs 98.2%:P&amp;lt;0.001) harbored an RB1 alteration, 7/22 (32% vs 90.9%: P&amp;lt;0.001) harbored a TP53 mutation, and 4/22 (18% vs 90.9%: P&amp;lt;0.001) had co-occurring RB1/TP53 mutations. No patients harbored mutations in EGFR, HER2, or RET. Mutations in potentially targetable pathways were observed throughout the cohort including the DNA-damage repair pathways, RTK/RAS, and PI3K. Taken together, these data demonstrate that patients with low pack-year SCLC have improved overall survival compared to historic averages for SCLC patients with a history of heavy tobacco use. Paradoxically, however, light-smokers with SCLC live longer than never-smokers, and these patients may harbor a higher frequency of targetable alterations than typical SCLC patients, highlighting the importance of molecular profiling in SCLC patients with minimal tobacco use. Citation Format: Kyle Concannon, Moushumi Sahu, Whitney Lewis, Hai Tran, Lixia Diao, Yuanxin Xi, Benjamin Morris, Simon Heeke, Ximing Tang, John Stewart, Ivan Valiev, Gabriela Raso, Jing Wang, Koji Sasaki, Carl Gay, John Heymach, Lauren Byers. Improved survival and unique mutational signatures in small cell lung cancer arising in patients with limited tobacco use [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 5061.

Abstract 3519: Reconstructing a latent representation of gene expression from genomic alterations to improve clinical utility of real-world clinicogenomics data

Abstract Background: Molecularly and clinically well-annotated patient datasets are ideal for studying tumor biology and developing robust machine learning (ML) models for predicting outcome and treatment response. These data however rarely exist in real-world settings or in sufficient quantities within research contexts. Large publicly available datasets like The Cancer Genome Atlas (TCGA) which provid multi-omic profiles for diverse cancer types, have profoundly advanced cancer research and facilitated development of novel therapies and personalized medicines. However, the absence of patient outcome data tied to treatment limits the applicability of these data for understanding and modeling treatment response. Real-world clinicogenomics cohorts, such as the AACR Project GENIE, on the other hand are typically very rich in clinical annotations, including treatment regimens and outcomes measures. These data are however sparsely annotated for patient tumor molecular profiles, rarely exceeding ~100’s of genes profiled. We hypothesized that it would be possible to reconstruct latent tumor mRNA representations from limited genomic and clinical data available in real-world clinicogenomic cohorts, and that these reconstructed expression profiles would be useful for a variety of clinically meaningful downstream applications. Methods: We developed an ML model, called Mut2Ex, to reconstruct tumor gene expression profiles using genetic information available on commercial next generation sequencing panels using a Principle Label Space Transformation (PLST) we adapted to regression problem, along with embeddings from clinical information (OncoTree code, sex and stage) generated by a language model. Mut2Ex was trained on ~1200 cell lines from DepMap representing 26 cancer types, to generate ~2000 reconstructed mRNA gene profiles that were applied to a variety of clinical tasks. We used Mut2Ex to reconstruct mRNA profiles for ~10,000 tumors from TCGA and ~180,000 tumors from AACR Project GENIE. Results: Reconstructed mRNA expression by Mut2Ex was highly correlated with true expression in cell lines (rho = 0.926, [0.924-0.928 95% CI, N = 1184]). Compared to true expression, reconstructed profiles recapitulate sub-clusters within cancer types, PAM50 subtyping in breast tumors, survival signatures in colorectal tumors and multiple oncogenic signatures in a pan-cancer manner. Analysis of reconstructed expression for AACR Project GENIE tumors revealed expected enrichment of known driver genes within expression subtypes and enrichment of oncogenic signatures associated with distinct clinical outcomes in a cancer type specific manner. Conclusions: Our flexible analytic framework for reconstructing gene expression profiles from clinicogenomics data substantially augments the clinical utility and value of data acquired in real-world settings. Citation Format: Maayan Baron, Sunil Kumar, Felicia Kuperwaser, Dillon Tracy, Emily Vucic, Jeff Sherman. Reconstructing a latent representation of gene expression from genomic alterations to improve clinical utility of real-world clinicogenomics data [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 3519.

Abstract 5437: Chemoproteomics-enabled discovery of covalent inhibitors targeted at AGPAT4: Unravelling tumor lineage plasticity to overcome drug resistance in hepatocellular carcinoma

Abstract The liver is a unique organ that is responsible for many metabolic functions. During hepatocellular carcinoma (HCC) development, these metabolic machineries are extensively reprogrammed to support the insatiable nutrient requirements of HCC. Tumor lineage plasticity, a recognized hallmark of cancer, is a phenomenon in which tumor cells co-opt developmental pathways to attain cellular plasticity, enabling them to evade targeted therapeutic interventions. Cancer cells can reprogram their metabolic pathways to match their increased metabolic needs for cancer cell survival under adverse conditions. Identifying novel metabolic targets related to stemness can offer promising strategies for targeting cancer stemness roots and hence to kill and control their growth. Through pathway enrichment analysis of genes that linked metabolism and stemness, we identified an aberrant glycerophospholipid metabolism signature, with AGPAT4 ranking as the top-hit. AGPAT4 upregulation in HCC is strongly correlated with aggressive clinical features, including survival, metastasis, and stemness signatures. AGPAT4 expression peaked during early liver progenitor development, decreased during hepatocyte maturation and progressively increased from well-differentiated to poorly differentiated HCCs. Enrichment of AGPAT4 in HCC is mediated by promoter binding of SOX9 to drive AGPAT4 transcriptional activity. AGPAT4 inhibition can mitigate tumor initiation, self-renewal, metastasis, and sorafenib resistance. Mechanistic studies revealed an AGPAT4-mediated phosphatidic acid production axis that promotes HCC through the regulation of mTOR signaling. Inhibition of Agpat4 by AAV8 shRNA reduced tumorigenicity and stemness, and sensitized HCC tumors to sorafenib. AGPAT4 overexpression can predict sorafenib response in clinical settings. Through chemoproteomics screening of a cysteine-reacting compound library using activity-based protein profiling, a cysteine-reacting compound with high binding affinity and selectivity towards AGPAT4 was identified and found to work synergistically with sorafenib to suppress HCC, as demonstrated in HCC patient-derived tumor xenograft (PDTX) models. Toxicity analysis through histological examination of organs, body weight measurements, and biochemical tests revealed minimal toxicity associated with the covalent inhibitor. In conclusion, AGPAT4 is a novel metabolic driver of oncogenic stemness, dedifferentiation, and metastasis in HCC. AGPAT4-induced tumor lineage plasticity may represent an Achilles heel for HCC treatment, and inhibition of AGPAT4 may widen the therapeutic window for sorafenib treatment in the clinic. Citation Format: Kai-Yu Ng, Tin-Yan Koo, Tsz-Lok Fong, Ya Gao, Tin-Lok Wong, Yuan Gao, Jing-Ping Yun, Xin-Yuan Guan, Ming Liu, Clive YS Chung, Stephanie Ma. Chemoproteomics-enabled discovery of covalent inhibitors targeted at AGPAT4: Unravelling tumor lineage plasticity to overcome drug resistance in hepatocellular carcinoma [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 5437.

Abstract 5447: Microtubule-associated protein MAP2 promotes drug resistance and cancer stemness in hepatocellular carcinoma through integrin dysregulation

Abstract Integrins are adhesion molecules that mediate mechanical and chemical signal transduction pathways to support cell survival and proliferation. Dysregulated integrin signaling empowers tumor cells to drive oncogenic stemness functions, including tumor initiation, epithelial plasticity, metastatic reactivation, and resistance to therapies. However, the interplay between cancer stemness and integrin signaling in hepatocellular carcinoma (HCC) remains poorly understood. HCC cells marked by CD133 represent an important functional subset of HCC tumors, displaying a dedifferentiated status with stem cell traits. In this study, transcriptome profiling reveals specific downregulation of integrin α family genes and integrin signaling in ‘HCC’ CD133+ cells isolated from NRAS+AKT-driven HCC, but not epithelial-specific ‘normal’ CD133+ cells isolated from regenerating liver. Of note, one of the most differentially upregulated genes identified in CD133+ HCC cell profiling, microtubule-associated protein MAP2, demonstrates the ability to suppress integrin expression. MAP2 overexpression is frequently observed in HCC and correlates with aggressive clinical and stemness features, including survival, tumor stage, and stemness signatures. Epigenetic modifications by H3K27Ac contribute to MAP2 upregulation. Functionally, MAP2 promotes cancer stemness and cell invasion, and confers resistance to the targeted therapy sorafenib. Mechanistically, the inhibition of collagen-binding integrin α subunits by MAP2 subsequently ablates integrin β1-mediated cell adhesion and FAK signaling. Estramustine Phosphate (EMP), previously reported to inhibit the interaction of MAP2 with actin filaments, attains a synergistic effect in suppressing tumor initiation and growth of HCC cell lines, HCC patient-derived organoids and NRAS+AKT HCC mouse model when used in combination with sorafenib. In summary, MAP2 inhibition may represent a potential novel therapy for HCC by targeting its cancer stemness roots and altering integrin signaling. Ongoing work is focused on the study of MAP2 regulation of integrin signaling and cell behaviours in the maintenance of a more stemness state in HCC. Citation Format: Ut Kei Lou, Ka-Hei Lam, Huajian Yu, Jia Jian Loh, Ki-Fong Man, Lei Zhou, Yuan Gao, Tin-Lok Wong, Cheng-Han Yu, Stephanie Ma. Microtubule-associated protein MAP2 promotes drug resistance and cancer stemness in hepatocellular carcinoma through integrin dysregulation [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 5447.

Deep learning-based multi-model prediction for disease-free survival status of patients with clear cell renal cell carcinoma after surgery: a multicenter cohort study.

Although separate analysis of individual factor can somewhat improve the prognostic performance, integration of multimodal information into a single signature is necessary to stratify patients with clear cell renal cell carcinoma (ccRCC) for adjuvant therapy after surgery. A total of 414 patients with whole slide images, computed tomography images, and clinical data from three patient cohorts were retrospectively analyzed. The authors performed deep learning and machine learning algorithm to construct three single-modality prediction models for disease-free survival of ccRCC based on whole slide images, cell segmentation, and computed tomography images, respectively. A multimodel prediction signature (MMPS) for disease-free survival were further developed by combining three single-modality prediction models and tumor stage/grade system. Prognostic performance of the prognostic model was also verified in two independent validation cohorts. Single-modality prediction models performed well in predicting the disease-free survival status of ccRCC. The MMPS achieved higher area under the curve value of 0.742, 0.917, and 0.900 in three independent patient cohorts, respectively. MMPS could distinguish patients with worse disease-free survival, with HR of 12.90 (95% CI: 2.443-68.120, P <0.0001), 11.10 (95% CI: 5.467-22.520, P <0.0001), and 8.27 (95% CI: 1.482-46.130, P <0.0001) in three different patient cohorts. In addition, MMPS outperformed single-modality prediction models and current clinical prognostic factors, which could also provide complements to current risk stratification for adjuvant therapy of ccRCC. Our novel multimodel prediction analysis for disease-free survival exhibited significant improvements in prognostic prediction for patients with ccRCC. After further validation in multiple centers and regions, the multimodal system could be a potential practical tool for clinicians in the treatment for ccRCC patients.