Raman spectroscopy (RS) is a label-free, non-destructive optical modality that provides a detailed profile of the molecular composition of a sample. There is growing interest in the clinical application of RS to characterize biomolecular signatures associated with radiotherapy response in tumor cells and tissues. A critical step before analyzing Raman data consists of performing spectral pre-processing to increase the quality of the measurements. Spectral pre-processing comprises baseline subtraction, signal smoothing, cosmic ray (CR) correction, and removal of poor-quality measurements. Herein, we present a convolutional autoencoder (AE) for single-step, automated pre-processing of Raman spectra obtained from tumor cells and tumor tissue. We trained two separate models using the same proposed architecture, one for eliminating spectral artifacts from preclinical single-cell line and xenografted tissue spectra exposed to single-fraction radiation, and the other for correcting clinical prostate tumor biopsy spectra collected from patients receiving high-dose-rate brachytherapy (HDR-BT). The autoencoder demonstrated fast, excellent performance in removing baseline, noise, and CRs. For the preclinical data, the model obtained a root mean squared error (RMSE), and a percentage root mean squared difference (PRD) of 7.1 × 10-5 and 3.1%, respectively, between the AE-corrected spectra and their corresponding target data (pre-processed by our current baseline-removal algorithm). Also, the autoencoder successfully removed 94.0% of CRs from the spectra. For the clinical biopsy data, the AE achieved an RMSE and a PRD of 8.1 × 10-5 and 3.7%, respectively, and a CR removal rate of 90.2%. Overall, the AE corrected approximately 11 000 spectra within 2.4  s without the need of a GPU. Furthermore, comparative supervised learning-based post-processing data analyses were performed separately on the spectra pre-processed by the autoencoder versus the target data, and we show consistency in the biochemical radiation response profiles extracted. Finally, the AE architecture was leveraged to train a reconstruction AE to facilitate semi-automated identification of poor-quality prostate biopsy spectra, and we demonstrate 96.4% agreement between AE and manually removed outliers. These results support the development of a deep learning framework for efficient, automated pre-processing of tumor cell and tissue Raman spectra collected for radiation response monitoring studies.

Peritoneal carcinomatosis (PC) is an aggressive manifestation of advanced gynecological and gastrointestinal malignancies with high recurrence rates despite cytoreductive surgery and chemotherapy. We developed a cationic liposomal nanoparticle (DSTAP, ∼159 nm) to deliver and retain the toll-like receptor 7/8 agonist Resiquimod (R848) within the peritoneal cavity, aiming to modulate the tumor immune microenvironment (TIME) and improve therapeutic efficacy. DSTAP-R848 was evaluated in murine colorectal and ovarian PC models, alone or combined with oxaliplatin (Oxa). Survival, cure rates, and durable immunity following tumor rechallenge were assessed. Immune polarization was examined by incubating ascites and peritoneal fluid with naïve splenocytes, while flow cytometry and metabolomic profiling characterized intratumoral immune populations and metabolic changes. Combination therapy achieved cure rates of 80% in the colorectal model and 30% in the ovarian model, with no recurrence after rechallenge; Oxa monotherapy yielded no cures. Oxa+DSTAP-R848 increased CD8+ T cells and M1 macrophages, while reducing regulatory T cells and myeloid-derived suppressor cells. Immunosuppressive and glycolytic metabolites progressively declined, correlating with reduced tumor burden. Overall, DSTAP-R848 enhances chemotherapy efficacy by reshaping immune and metabolic pathways and promoting durable anti-tumor immunity in PC.

High-grade serous ovarian carcinoma (HGSOC) frequently recurs after platinum-taxane therapy, yet recurrence-associated changes within ascites remain unclear. We performed cell-type-resolved quantitative proteomics on paired ascites samples collected at primary diagnosis and at disease recurrence from three HGSOC patients, profiling unsorted cells alongside matched CD45- tumor-enriched and CD45+ immune-enriched fractions. Conventional proteomic label-free quantitative (LFQ) analysis yielded limited significant proteins due to inter-patient heterogeneity. To address this, we applied Gene Set Enrichment Analysis (GSEA) to identify coordinated pathway-level alterations associated with recurrence. Across all cellular compartments, recurrent samples exhibited consistent downregulation of interferon-α/γ-associated pathways alongside enrichment of oxidative phosphorylation and stress-adaptive metabolic programs. In the CD45+ immune compartment, these changes were accompanied by enhanced IL-2/STAT5 signaling and reduced antigen-presentation pathways. Together, these findings suggest that recurrent ascites is characterized by a shift toward oxidative metabolism and a more immunosuppressive microenvironment.

Mass spectrometry imaging (MSI) is emerging as a powerful tool for uncovering the distribution of metabolites in the tumor microenvironment and studying tumor metabolism in vivo. To date, MSI of biobanked tissues contextualized by patient data has been limited to peptides, proteins, and glycans-with few examples for metabolites. This is because most biobanked fresh-frozen tissue required for spatial metabolomics is embedded in optimal cutting temperature (OCT) compound to preserve structural features and mitigate thermal decay. However, OCT introduces abundant polyethylene glycol and polyvinyl alcohol interferents. Herein, we use nanospray desorption electrospray ionization (nano-DESI) to demonstrate MSI of metabolites in OCT-embedded tissue. Metabolite coverage and sensitivity for tissue mimetic homogenates embedded in OCT and an MSI-compatible material, carboxymethylcellulose (CMC), exhibited excellent agreement. We apply our ambient MSI workflow to study the impact of methionine-restriction in a preclinical mouse model undergoing adoptive T-cell therapy. After tumor incubation (8 days), lymphoma-bearing mice were maintained on a complete or methionine-restricted diet for 2 days. Nano-DESI MSI revealed a heterogeneous tumor microenvironment, with multiple methionine-cycle intermediates (S-adenosylmethionine, S-adenosylhomocysteine) and related metabolites, including known T-cell modulators (1-methylnicotinamide, polyamines) localizing to tumor subregions. Methionine-restricted tumors exhibited reduced methionine and elevated S-adenosylmethionine, relative to the control group. Overall, this work establishes the potential for spatial metabolomics of fresh-frozen OCT-embedded tumors, unlocking the wealth of information stored in primary tissue biobanks and consequently accelerating our understanding of cancer metabolism and treatment.

Covariate measurement error in regression analysis is an important issue that has been studied extensively under the classical additive and the Berkson error models. Here, we consider cases where covariates are derived from tumor tissue histology, and in particular tissue microarrays. In such settings, biomarkers are evaluated from tissue cores that are subsampled from a larger tissue section so that these biomarkers are only estimates of the true cell densities. The resulting measurement error is non-negligible but is seldom accounted for in the analysis of cancer studies involving tissue microarrays. To adjust for this type of measurement error, we assume that these discrete-valued biomarkers are conditionally Poisson distributed, based on a Poisson process model governing the spatial locations of marker-positive cells. Existing methods for addressing conditional Poisson surrogates, particularly in the absence of internal validation data, are limited. We extend the simulation extrapolation (SIMEX) algorithm to accommodate the conditional Poisson case (POI-SIMEX), where measurement errors are non-Gaussian with heteroscedastic variance. The proposed estimator is shown to be strongly consistent in a linear regression model under the assumption of a conditional Poisson distribution for the observed biomarker. Simulation studies evaluate the performance of POI-SIMEX, comparing it with the naive method and an alternative corrected likelihood approach in linear regression and survival analysis contexts. POI-SIMEX is then applied to a study of high-grade serous cancer, examining the association between survival and the presence of triple-positive biomarker (CD3+CD8+FOXP3+ cells)

In this issue of Cell Chemical Biology, Pang et al.1 address the question of how effector CD8⁺ T cells acquire stem-like durability. They uncover a redox-driven metabolic program in which NQO1-mediated cycling of lawsone enhances pentose phosphate pathway, remodels mitochondrial function, and connects effector differentiation to sustained antitumor immunity.

Diet is a key modulator of human metabolism, influencing disease prevention and progression. In patients with cancer, nutritional interventions are common but often lack standardized guidelines and are mainly used for symptom relief rather than as adjuvants to immunotherapies. A major limitation is the poor mechanistic understanding of how diet influences immunity and cancer. Emerging evidence shows that diet profoundly shapes immune function and the gut microbiome, both of which affect responses to immunotherapy, disease progression, and survival. Despite this, the immune–gut axis is infrequently assessed. In this review, we focus on dietary intervention studies that include immune or microbiome assessments, highlighting mechanistic insights and clinical relevance. Diet should be used as a strategic tool to enhance cancer therapies and improve patient outcomes.

T cell–based immunotherapies have transformed cancer treatment, yet their efficacy in solid tumors is constrained by the nutrient-poor and oxidative tumor microenvironment (TME). Accumulating evidence indicates that reactive oxygen species (ROS), methionine metabolism, and the amino acid stress sensor general control nonderepressible 2 (GCN2) are tightly interconnected regulators of T cell activation, differentiation, and effector function. In this review, we detail how these pathways form an integrated redox–metabolic circuit that dynamically tunes T cell responses to environmental stress. Physiological ROS are essential for T cell receptor signaling, glycolytic reprogramming, and cytotoxicity, whereas excessive or prolonged oxidative stress drives exhaustion and apoptosis. GCN2 links amino acid availability, particularly methionine and cysteine, to adaptive transcriptional and metabolic programs that regulate glutathione synthesis and redox homeostasis. We highlight how therapeutic manipulation of methionine availability, GCN2 signaling and ROS produces highly context-dependent outcomes across immune checkpoint blockade and adoptive cell therapy settings in solid tumors. Finally, we discuss emerging strategies to interrogate and modulate this circuit using integrated omics, CRISPR-based screening, and pharmacological approaches, emphasizing the need for context-aware and temporally controlled metabolic interventions to enhance T cell–based immunotherapies in solid tumors.

Methionine is an essential amino acid critical for T cell activation. While methionine restriction (MR) combined with immune checkpoint blockade has been shown to enhance T cell function, the impact of methionine on adoptive T cell therapies is largely unexplored. Here, we examined the functionality of T cells under MR and pharmaceutical inhibition of the methionine cycle (MAT2Ai), using primary T cells and a murine adoptive T cell therapy model. In vitro, transient MR or MAT2Ai treatment increased interferon gamma (IFNγ) expression in CD8+ T cells, whereas sustained MR led to the upregulation of T cell exhaustion-associated markers. Mechanistically, transient MR suppressed the polyamine synthesis pathway, and supplementation with polyamines reversed MR-induced IFNγ expression. Genetic ablation of s-adenosylmethionine decarboxylase, an enzyme in the polyamine synthesis pathway, recapitulated the effect of MR, indicating that transient MR enhances T cell function by inhibiting polyamine synthesis. Despite this, transient MR treatment of ovalbumin (OVA)-specific (OT-I) CD8+ T cells prior to adoptive transfer did not improve antitumor efficacy against EG7-OVA tumors in vivo. In contrast, sustained dietary MR accelerated EG7-OVA tumor growth in mice treated with OT-I T cells, demonstrating that methionine availability is essential for the activity of adoptively transferred T cells. These findings suggest that enhancing methionine availability in the tumor microenvironment may improve the efficacy of adoptive T cell therapies.

Abstract T cell-based immunotherapies have remained ineffective against high-grade serous ovarian carcinoma (HGSOC). The metabolic environment of HGSOC suppresses the activity of cellular therapies, however, the metabolites that enhance or suppress T cell antitumor activity are not fully understood. Here, a pooled CRISPR-Cas9 knockout screen in primary human T cells cultured with patient-derived ascites was used to identify metabolic enzymes that inhibit effector cytokine production and cytolytic function. The screen identified PRDX1 as a negative regulator of T cell effector function. Targeted deletion of PRDX1 increased the frequency of IFN-γ-producing T cells, enhanced glucose uptake, increased mitochondrial mass, and improved T cell viability under suppressive ascites conditions. Mechanistically, PRDX1 deficiency increased intracellular reactive oxygen species (ROS) and impaired autophagic flux. The effects of PRDX1 deletion enhanced aspects of T cell function, while its effects on chimeric antigen receptor (CAR)-T cell cytotoxicity were donor dependent. Collectively, this study identifies PRDX1 as a regulator of T cell activation, metabolism, and effector function in the inhibitory physiological suppressive environment of HGSOC ascites.

CONTEXT:In the face of the growing global burden of cancer, there is increasing interest in dietary interventions to mitigate its impacts. Pre-clinical evidence suggests that time-restricted eating (TRE), a type of intermittent fasting, induces metabolic effects and alterations in the gut microbiome that may impede carcinogenesis. Research on TRE in cancer has progressed to human studies, but the evidence has yet to be synthesized. OBJECTIVE:The objective of this study was to systematically evaluate the clinical and/or metabolomic effects of TRE compared with ad libitum eating or alternative diets in people with cancer. DATA SOURCES:Ovid MEDLINE, Ovid Embase, CINAHL, Ovid Cochrane Central Register of Control Trials (CENTRAL), Web of Science Core Collection (ESCI, CPCI-SSH, CPCI-S), and SCOPUS were searched up to January 4, 2023, using the core concepts of "intermittent fasting" and "cancer." Original study designs, protocols, and clinical trial registries were included. DATA EXTRACTION:After evaluating 13 900 results, 24 entries were included, consisting of 8 full articles, 2 abstracts, 1 published protocol and 13 trial registries. All data were extracted, compared, and critically analyzed. DATA ANALYSIS:There was heterogeneity in the patient population (eg, in tumor sites), TRE regimens (eg, degree of restriction, duration), and clinical end points. A high rate (67-98%) of TRE adherence was observed, alongside improvements in quality of life. Four articles assessed cancer markers and found a reduction in tumor marker carcinoembryonic antigen, reduced rates of recurrence, and a sustained major molecular response, following TRE. Five articles demonstrated modified cancer risk factors, including beneficial effects on body mass index, adiposity, glucoregulation, and inflammation in as short a period as 8 weeks. None of the completed studies assessed the effect of TRE on the microbiome, but analysis of the microbiome is a planned outcome in 2 clinical trials. CONCLUSIONS:Preliminary findings suggest that TRE is feasible and acceptable by people with cancer, may have oncological benefits, and improves quality of life. REGISTRATION:PROSPERO registration No. CRD42023386885.

Discrete biomarkers derived as cell densities or counts from tissue microarrays and immunostaining are widely used to study immune signatures in relation to survival outcomes in cancer. Although routinely collected, these signatures are not measured with exact precision because the sampling mechanism involves examination of small tissue cores from a larger section of interest. We model these error-prone biomarkers as Poisson processes with latent rates, inducing heteroscedasticity in their conditional variance. While critical for tumor histology, such measurement error frameworks remain understudied for conditionally Poisson-distributed covariates. To address this, we propose a Bayesian joint model that incorporates a Dirichlet process (DP) mixture to flexibly characterize the latent covariate distribution. The proposed approach is evaluated using simulation studies which demonstrate a superior bias reduction and robustness to the underlying model in realistic settings when compared to existing methods. We further incorporate Bayes factors for hypothesis testing in the Bayesian semiparametric joint model. The methodology is applied to a survival study of high-grade serous carcinoma where comparisons are made between the proposed and existing approaches. Accompanying R software is available at the GitHub repository listed in the Web Appendices.

Chemoresistance poses a significant clinical challenge in managing glioblastoma (GBM), limiting the long-term success of traditional treatments. Here, a 3D tumoroid model is used to investigate the metabolic sensitivity of temozolomide (TMZ)-resistant GBM cells to iron chelation by deferoxamine (DFO) and deferiprone (DFP). This work shows that TMZ-resistant GBM cells acquire stem-like characteristics, higher intracellular iron levels, higher expression of aconitase, and elevated reliance on oxidative phosphorylation and proteins associated with iron metabolism. Using a microphysiological model of GBM-on-a-chip consisting of extracellular matrix (ECM)-incorporated tumoroids, this work demonstrates that the combination of iron chelators with TMZ induces a synergistic effect on an in vitro tumoroid model of newly diagnosed and recurrent chemo-resistant patient-derived GBM and reduced their size and invasion. Investigating downstream metabolic variations reveal reduced intracellular iron, increased reactive oxygen species (ROS), upregulated hypoxia-inducible factor-1α, reduced viability, increased autophagy, upregulated ribonucleotide reductase (RRM2), arrested proliferation, and induced cell death in normoxic TMZ-resistant cells. Hypoxic cells, while showing similar results, display reduced responses to iron deficiency, less blebbing, and an induced autophagic flux, suggesting an adaptive mechanism associated with hypoxia. These findings show that co-treatment with iron chelators and TMZ induces a synergistic effect, making this combination a promising GBM therapy.

T-cell based immunotherapies such as chimeric antigen receptor T (CAR-T) cell therapy face substantial hurdles when confronting solid tumors such as ovarian cancer, where metabolic constraints in the tumor microenvironment limit T cell infiltration and function. In particular, T cells exposed to nutrient deprivation and hypoxia upregulate autophagy, a lysosomal degradation pathway that negatively regulates effector responses. Here, we used CRISPR-Cas9 to target a folate receptor alpha (αFR) CAR expression cassette into the locus of the essential autophagy gene ATG5, thereby generating autophagy-deficient CAR-T cells in a single editing step. Targeted metabolite profiling revealed that deletion of ATG5 induced widespread metabolic reprogramming characterized by increased glucose and amino acid uptake. Functionally, ATG5-knockout CAR-T cells maintained high cytolytic activity when assayed in patient-derived ascites in vitro, and exhibited superior and long-lasting tumor control against ovarian tumors in vivo. Taken together, our results suggest that deletion of ATG5 metabolically primes CAR-T cells for enhanced cytotoxicity in immune-suppressive conditions, thereby improving the therapeutic potential of αFR CAR-T cells for ovarian cancer immunotherapy.

Prostate cancer is characterized by an immunosuppressive tumour environment. This work combines Raman spectroscopy with group-and-bases-restricted non-negative matrix factorization (GBR-NMF) and machine learning to assemble models of immune cell densities within the needle-core biopsies of patients undergoing high-dose-rate brachytherapy (HDR-BT). Raman spectral acquisition, as well as immunohistochemistry staining of CD68[Formula: see text], CD3[Formula: see text], and [Formula: see text] cells, was completed for biopsies collected before and 2 weeks following the first fraction of HDR-BT. Regression techniques, constructed using GBR-NMF scores, that produced the most accurate predictions of immune cell density by metrics of root mean-squared error (RMSE) and R[Formula: see text] were the gradient-boosted trees model of [Formula: see text] density (RMSE: 163 counts[Formula: see text], [Formula: see text]: 0.65) and the elastic net model of [Formula: see text]/ [Formula: see text] (RMSE: 0.25, [Formula: see text]: 0.82). The accuracy of these models, herein defined as the fraction of patient predictions within [Formula: see text] standard deviation of their measured values was 11/16 and 12/16, for CD68[Formula: see text] CD3[Formula: see text] and CD68[Formula: see text]/ CD8[Formula: see text] models, respectively. To further delineate which metabolites were most important in the CD68[Formula: see text]/ CD8[Formula: see text] model, this ratio was further predicted in stromal and epithelial tissues within the biopsies, and resulting models utilized the GBR-NMF scores of glutathione, collagen, palmitic acid, and the pre- or post-HDR-BT label to produce an optimal performance level according to RMSE and R[Formula: see text]. In summary, this study illustrates a novel methodology in which supervised machine learning techniques are used to model immune cells, which are prognostic indicators of disease progression.

The fate of CD8+ T cells is sculpted not only by antigenic stimulation and cytokine milieu but, increasingly, by metabolic context. In their recent Nature Immunology study, Sharma and colleagues report a previously underappreciated and temporally constrained nutrient-sensing mechanism in which methionine (Met) availability during the earliest minutes of T-cell receptor engagement exerts durable control over T-cell function, exhaustion, and antitumor efficacy. Their findings expose a critical metabolic window, within just 30 minutes of activation, during which extracellular Met shapes intracellular signaling and transcriptional fate decisions through a posttranslational mechanism involving arginine methylation of the calcium-activated potassium channel KCa3.1. These findings open the door to timed interventions that modulate Met and potentially enhance T-cell responses.

Mass spectrometry imaging (MSI) is emerging as a powerful tool for uncovering the distribution of metabolites in the tumor microenvironment and studying tumor metabolism in vivo . However, to date, MSI of primary patient biobanked tissues contextualized by patient data has been limited to peptides, proteins, and glycans – with few examples for metabolites. This is because most biobanked fresh-frozen tissue required for spatial metabolomics is embedded in optimal cutting temperature compound (OCT), which introduces high-abundance polymeric interferents. Herein, we use nanospray desorption electrospray ionization (nano-DESI) to demonstrate the MSI of metabolites in OCT-embedded tissue. Metabolite coverage and sensitivity for prepared tissue mimetic homogenates embedded in OCT and an MSI-compatible material, carboxymethylcellulose (CMC), showed excellent agreement. We apply our ambient MSI workflow to detect changes in intratumoral methionine using a preclinical cancer mouse model undergoing adoptive T-cell therapy. Eight days after tumor incubation, lymphoma-bearing mice were maintained on a complete or methionine-restricted diet for 2 days. Nano-DESI MSI revealed a heterogeneous tumor microenvironment, with multiple methionine-cycle intermediates (S-adenosylmethionine, S-adenosylhomocysteine) and related metabolites, including known T-cell modulators (1-methylnicotinamide, polyamines) localizing to tumor subregions. Methionine-restricted tumors exhibited reduced methionine levels and elevated S-adenosylmethionine, relative to the control group. Overall, this work demonstrates spatial metabolomics on fresh-frozen OCT-embedded tissue, unlocking the wealth of information stored in primary tissue biobanks and consequently accelerating our understanding of cancer metabolism and treatment. ### Competing Interest Statement The authors have declared no competing interest. Natural Sciences and Engineering Research Council, https://ror.org/01h531d29, RGPIN-2022-03696 Terry Fox Research Institute, Program Project Grant #1125 Mitacs, IT39215 Michael Smith Health Research BC, RT-2024-03771 Michael Smith Health Research BC, SCH-2025-04631 Canada Foundation for Innovation, https://ror.org/000az4664, CFI JELF 43810 Lotte and John Hecht Memorial Foundation, Program Project Grant #1125

A CNN was developed for classifying Raman spectra of radiosensitive and radioresistant tumour cells. Furthermore, a CNN explainability method was proposed to identify biomolecular Raman signatures associated with the observed radiation responses.