Oxidative stress and cancer: current insights and therapeutic implications Authors: María Escribano-Cebrián, Marina Bataller, Marc Biasini, Yoelsis García-Mayea, David Virós, Matilde E. LLeonart Published: 2026 ABSTRACT Oxidative stressgood or evil? Oxidative stress occurs when the balance between reactive oxygen species (ROS) and antioxidant defenses shifts toward an excess of ROS; while essential in physiological processes, it plays a context-dependent role in cancer, contributing to both the promotion and inhibition of tumorigenesis. Small to moderate amounts of ROS activate pathways supporting tumor progression and proliferation, while large amounts lead to genomic instability and cell death.ROS are generated endogenously and exogenously. In cancer, ROS activate pathways that prompt tumor development (KRAS, MYC, PI3K-Akt-mTOR) and block tumor suppressors (p53, BRCA1), allowing tumorigenesis and drug resistance. They also modulate the tumor microenvironment (TME) by altering tumor, stromal and immune cell interactions, which initiate angiogenesis, epithelial-mesenchymal transition (EMT), inflammation and metastasis. Myeloid-derived suppressor cells (MDSCs) and cancer-associated fibroblasts (CAFs) contribute to ROS-driven immunosuppression.Cancer cells mainly rely on glycolysis and oxidative phosphorylation (OXPHOS) to sustain their energetic and metabolic requirements. Generated ROS act as metabolic byproducts and signaling molecules supporting proliferation and tumorigenesis. Cancer stem cells (CSCs) produce low ROS levels by activating antioxidant pathways and mitochondria remodeling, ensuring recurrence and persistence.There is a redox duality that presents challenges and opportunities for therapies. Pro-oxidant approaches attempt to overwhelm the tumor's defenses, while antioxidants preserve healthy tissues. Advances in targeted redox modulation with immunotherapies improve therapy effectiveness. We propose a new “Adaptive Directed Redox Therapy” (ADRT), which involves a dynamic, feedback-controlled methodology that alternates pro- and antioxidant phases to selectively collapse tumor redox balance while preserving normal tissues. Glossary 2-Deoxy-d-glucose Adaptive Directed Redox Therapy Protein kinase B AMP-activated protein kinase Antigen presentation Ataxia-telangiectasia Adenosine triphosphate Breakpoint cluster region-Abelson tyrosine kinase B-Raf proto-oncogene, serine/threonine kinase Breast cancer type 1 susceptibility protein Bombesin receptor subtype-3 Cancer-associated fibroblasts Cluster of Differentiation Cancer stem cells Cytotoxic T-lymphocyte-associated protein 4 Damage associated molecular pattern molecules Dendritic cells Dynamin-related protein 1 Epidermal growth factor receptor Epithelial-mesenchymal transition Extracellular signal-regulated kinases Focal adhesion kinase Fatty acid transport protein 2 Fumarate hydratase Forkhead box O Ferroptosis suppressor protein 1 Glucose-6-phosphate dehydrogenase Glyceraldehyde-3-phosphate dehydrogenase Granulocyte-Macrophage Colony-Stimulating Factor Glutathione peroxidase Glutathione Hydrogen peroxide Hepatocellular carcinoma Hypoxia-inducible factor Hexokinase 2 Voltage-gated proton channel 1 Immunogenic cell death Isocitrate dehydrogenase Indoleamine 2,3-dioxygenase 1 Interferon-gamma Interleukin Janus kinase c-Jun N-terminal kinase Kelch-like ECH-associated protein 1 Kirsten rat sarcoma viral oncogene homolog Mitogen-activated protein kinase kinase kinase 4 Mitogen-activated protein kinase Monocarboxylate transporter 1 Myeloid-derived suppressor cells Mitogen-activated protein kinase kinase 1 Mitofusin 1 MicroRNAs Matrix metallopeptidase Mammalian sterile 20-like kinase 1/2 Mitochondrial ROS MYC proto-oncogene, bHLH transcription factor Nicotinamide adenine dinucleotide phosphate Neutrophil extracellular traps Nuclear factor kappa-light-chain-enhancer of activated B cells Nuclear factor erythroid 2-related factor 2 Natural killer cells NADPH oxidases Nanoparticles Nuclear factor erythroid 2-related factor 2 Oxidative phosphorylation SHC adaptor protein Poly(ADP-ribose) polymerase Programmed cell death protein 1 Programmed death-ligand 1 Pyruvate dehydrogenase kinase 1 Photodynamic therapy PPAR-γ coactivator-1-alpha Phosphoglycerate kinase 1 Phosphoinositide 3-kinase p53-inducible gene 3 Protein kinase C Pyruvate kinase M2 Pentose phosphate pathway Phosphatase and tensin homolog Retinoblastoma Reactive nitrogen species Reactive oxygen species RAS-selective lethal 3 Senescence-associated secretory phenotype Solute carrier family 7 member 11 Superoxide dismutase Sphingosine kinase 1 Tyrosine-protein kinase Signal transducer and activator of transcription 3 Stimulator of interferon genes Regulatory T-cells Tumor-associated macrophages Tumor-associated neutrophils T-cell receptor mitochondrial transcription factor A Transforming growth factor beta Tumor microenvironment Triple negative breast cancer Tumor necrosis factor-alpha Succinate dehydrogenase Vascular endothelial growth factor VEGF receptor von Hippel-Lindau Amino acid transport system xc Introduction Oxidative stress is among the most crucial cellular stress effectors in cancer. It represents a cellular state in which reactive species, such as ROS, reactive nitrogen species (RNS), and electrophilic metabolites, exceed the capacity of scavenging and repair pathways to avoid damage to DNA, proteins, and lipids and alter redox-sensitive cellular signal pathways [1]. In cancer, oxidative stress can occur as either a dual phenomenon in which the normal state involves low and moderate ROS levels which promote the development and function of signaling networks, driving cell proliferation, migration and defense. However, if there is uncontrolled accumulation of ROS, these processes will create mutations and affect cellular metabolism, inflammation, and cell death pathways. Alternatively, an overload of ROS beyond endogenous capabilities may lead to cell senescence or death. ROS therefore play a dual role, as both as mediators of signaling and as cytotoxic agents [[2], [3], [4], [5]]. Several cellular respiration reactions consume oxygen leading to ROS formation. Aerobic respiration in normal cells leads to the generation of adenosine triphosphate (ATP) and the conversion of metabolic substrates into ATP, CO2 and H2O. The final and most effective step in respiration is OXPHOS, which is also a primary endogenous source of ATP, as well as one of the major internal sources of ROS. While electrons are transported from the inner mitochondrial membrane, complexes I and III generate superoxide which quickly become hydrogen peroxide (H2O2) through superoxide dismutases (SOD) [6]. In normal cells, this reaction is controlled by endogenous cell signaling and it is an essential mechanism in maintenance of homeostasis. However, due to factors including oncogenic signals, loss of tumor suppressor factors, mitochondrial disorders and changes in the metabolism of cancer cells, the system loses its balance, leading to a chronic ROS increase [7]. The main cell regulator that combats oxidative stress is the Kelch-like ECH-associated protein 1 (KEAP1)- nuclear factor erythroid 2-related factor 2 (NRF2) pathway. NRF2 is an antioxidant protein which stimulates many defense proteins, such as antioxidant enzymes and various cytoprotective mechanisms which provide protection from ROS damage [8]. Although it might have a suppressing effect on early cancer progression, its constitutive activation in established tumors promotes cell survival and expansion [9,10]. This suggests that the dual action of signaling pathways may be a cause of malignancy by supporting the survival and growth processes. In this review, we focus on OXPHOS and NADPH oxidases (NOX) enzymes mediated ROS production, although there are several alternative pathways other than mitochondrial respiration to produce ROS. Cytochrome P450 enzymes are an important example, because they are often upregulated in tumor cells and during substrate metabolism, producing ROS [11]. Chronic inflammatory conditions, predominantly inside the TME, are another very significant contributor. Inflammatory cytokines produced by macrophages and other cells in the TME recruit neutrophils and tumor-associated macrophages (TAMs), with a respiratory burst via nicotinamide adenine dinucleotide phosphate (NADPH) oxidase that generates reactive oxidants, which kill tumor cells and the surrounding normal cells. It also shows that mitochondrial ROS (mtROS) can promote activation of NADPH oxidase which results in increased tissue oxidative injury [12,13]. Several exogenous sources of ROS have also been shown to directly induce oxidative stress and thus initiate cancer. These include environmental chemicals such as heavy metals, organic solvents, pesticide and quinones that either produce free radicals or inactivate antioxidant defenses, leading to increased DNA damage and tumor development [14,15]. Environmental exposure and indoor pollution is believed to induce cancer, particularly lung and skin cancers, as a result of redox-active compounds inducing DNA oxidative damage [14,16,17]. Smoking remains the most significant external agent in terms of cancer risk. Its ability to produce ROS and RNS and delivery pro-oxidants into the lung may contribute to the development of cancer through DNA, protein, and lipid damage, causing genomic instability [[18], [19], [20]]. Biological carcinogens such as Human papillomavirus, Hepatitis B and Epstein-Barr virus also promote the generation of ROS by inducing ROS-producing enzymes and by modulating antioxidant systems [21]. Furthermore, ionizing radiation and non-ionizing radiation lead to the production of hydroxyl radicals, which cause DNA damage leading to secondary malignancies and cancer [22,23]. Cumulative effects on oxidation could lead to normal cells becoming malignant cells. These changes involve not only DNA damage causing genome instability and mutations, but also affect the cell pathways that control proliferation, death, metastasis and defense to therapy [2,3]. Cancer is therefore not a purely intrinsic cellular program, but it also plays a key role in the cells' interactions with each other and with the TME. For this reason, it is important to understand how cell redox status regulates signaling pathways to modulate cell proliferation, differentiation, survival and cell death [3]. ROS and other redox signaling also influence stromal and immune cells in the TME, affecting its vascularization and metabolism, indicating that cancer is essentially a complex system-level redox phenomenon, and not simply an intra-cellular issue [24]. Although various comprehensive reviews analyze different aspects of redox regulation in cancer (ROS induced tumorigenesis, TME redox remodeling, cell adaptation, metabolic reprogramming, ferroptosis, antioxidant treatments …), they consider these areas as isolated components of redox states regulating carcinogenesis, rather than as a functionally coordinated system. An integrated model of how intracellular and intercellular redox state contributes to genomic instability, TME remodeling, metabolic adaptation, stem cell properties and therapeutic resistance is therefore lacking in the current literature. In this review, we aim to bridge this gap while discussing the relationship between oxidative stress and cancer from the molecular to the clinical level, and to update our previous study [25]. Here, we focus on how cells use redox thresholds to balance cell survival, proliferation, differentiation and plasticity, thereby contributing to cancer development and therapy resistance. We mainly focus on the two major intracellular ROS-generating sources: OXPHOS and NOX enzymes, as key regulators of redox signaling, metabolic adaptation, genomic instability, and TME remodeling. We present how the modulation of ROS levels leads to the evolution of tumors via dynamic regulation, from an intracellular point of view to the entire TME. Determining and measuring the cellular redox status is currently of vital importance in the development of effective anti-cancer therapies. Accordingly, we suggest the novel term: Adaptive Directed Redox Therapy (ADRT). We define this as a dynamic therapeutic model based on the diagnosis and targeted delivery of redox modulating molecules that facilitate the selective collapsing of malignancy specific redox balance with minimal damage to normal tissues. This constitutes an advance in tailored therapies able to exploit the redox vulnerabilities of cancer cells. Redox signaling networks driving tumor evolution ROS not only function as damaging agents, but also act as central mediators within an integrated network of genomic instability, metabolic reprogramming and TME remodeling. In various types of tumor, the continuous dynamic balance between oxidative stress and antioxidant adaptation shapes the tumor's evolution, heterogeneity and resistance. Oxidative stress as an evolutionary driver of tumor heterogeneity As mentioned above, ROS promote genomic instability not only through the formation of DNA lesions but also by integrating mutagenic processes, DNA repair deficiency and epigenetic reprograming as a single evolutionary pressure. Oxidative DNA damage, which includes modifications to bases and breaks in DNA strands, directly stimulates mutations [26]. ROS directly impact the landscape of DNA repair mechanisms. Chronic oxidative stress blocks specific DNA repair pathways, such as homologous recombination and base excision repair, by inactivating repair enzymes including 8-oxoguanine glycosylase and mutY DNA glycosylase, leading to lesion accumulation, such as 8-oxo-guanine, which further increases mutations through error-prone repair [27]. Importantly, this impaired repair cooperates with Apolipoprotein B mRNA editing catalytic polypeptide-like induced mutations, suggesting ROS integrating other sources of genomic instability [28]. Oxidative stress also actively induces a broad, genome-wide, epigenetic reprograming that further sustains genomic instability. ROS-driven alteration of DNA methylation status and histone modification leads to DNA demethylation, de novo methylation sites and gene dysregulation at sites like breast cancer type 1 susceptibility protein (BRCA1) [29] and overall genome-wide hypomethylation at specific genomic sites, particularly telomeres, contributing to tumor-associated inflammatory signals and tumor-promoting transcription programs [30]. Apart from these epigenetic alterations, oxidative stress also affects post-transcriptional regulatory mechanisms. As an example, ROS-dependent repression of Dicer during inflammatory processes results in global microRNA (miRNA) maturation defects and promotes tumorigenesis [31,32]. These findings highlight the idea that genome stability in cancer is largely mediated through redox-controlled genome dysregulation. A second important issue is that ROS does not simply induce DNA damage, but also acts as a selecting force that drives cancer cell fitness. Overexpression of mtROS in phosphatase and tensin homolog (PTEN)-deficient tumors, which activate phosphoinositide 3-kinase/Protein kinase B (PI3K/Akt) pathway driven metabolic reprogramming, creates a pro-oxidant environment that paradoxically selects tumor cell clones with the highest antioxidant capacity to counteract that effect [33]. NRF2 mediated upregulation of antioxidant programs leads to tumor cell tolerance to the stress, increasing resistance to therapy, like platinum-based chemotherapy, but also to ferroptosis, via the glutathione (GSH) and glutathione peroxidase (GPX4) axis [34,35]. This highlights the dualistic effects of ROS, which at moderate levels promote cell proliferation via mitogen-activated protein kinase/extracellular signal-regulated kinases (MAPK/ERK) and PI3K/Akt pathway activation, whereas in high concentrations they induce genomic destabilization and cell death via cytotoxicity [2,4,5]. Oxidative stress is therefore both the driver for mutagenesis and the selective force for antioxidant adaptation and therapy-resistant tumor cell clones, with redox balance regulating tumor evolution. Redox signaling networks within the TME The TME should be considered as an interconnected redox-regulated system where immune and stromal cells cooperate with tumor cells in modulating ROS production and signaling, eventually contributing to tumor growth. This integrated redox-network comprises several cell types: MDSCs, tumor-associated neutrophils (TANs), TAMs, tumor cells themselves, and CAFs that modulate ROS through various mechanisms to promote tumor progression. One of the key characteristics of this system is metabolic and inflammatory coupling. CAFs rewire metabolism to predominantly use aerobic glycolysis, supporting tumor growth through lactate secretion [36,37], but also through the recruitment of monocytes which differentiate to MDSCs via the chemokine (C-C motif) ligand 2/C-C chemokine receptor type 2 axis. MDSCs then produce ROS through NADPH oxidases (NOX2 and NOX4) and indoleamine 2,3-dioxygenase 1 (IDO1), suppressing cluster of differentiation (CD8+) T-cells responses and inducing tumor immune evasion [38]. Thus, the tight relationship between metabolism, inflammation and tumor growth can be seen to be mediated through ROS and its associated signals. Cell-cell interaction provides another key redox driven signal within the TME. Tumor-derived extracellular vesicles and exosomes, with active NOX enzyme, mediate cell-cell communication by affecting immune cells, stimulating TAM M2-polarization and TAN-N2 response via redox driven signaling [39,40]. These redox-driven extracellular vesicle-mediated signals promote tumor growth by enhancing angiogenesis, inhibiting immune responses, reducing cytotoxic cells and promoting inflammation in the TME. Inflammation activation through redox-sensitive transcription factors such as Nuclear factor kappa-light-chain-enhancer of activated B-cells (NF-κB) drives sustained pro-inflammatory signals, a tumor-promoting environment and DNA damage [33]. These inflammatory and oncogenic signals are orchestrated by transcriptional factors such as SOX2, FOSL2 and downstream tumor necrosis factor-alpha (TNF-α)/NF-κB and IL6- Janus kinase (JAK)-signal transducer and activator of transcription 3 (STAT3) signals, which further reinforce redox driven signals to promote tumor development [41]. Senescent stromal cells also contribute to therapy resistance through the secretion of senescence-associated secretory phenotype (SASP) cytokines [42]. Factors like interleukin-8/focal adhesion kinase/tyrosine-protein kinase Src (IL-8/FAK/Src) create a self-sustaining signaling loop that supports cell proliferation and metastasis [43]. Increased extracellular matrix production by CAFs and the secretion of E-cadherin, which promote metastasis, is again supported by cell-cell interactions and redox driven signaling [44]. These findings highlight the synergistic role of redox signaling regarding cancer development and resistance through feedback mechanisms within the TME. The key concept in all these studies is that the TME produces high levels of ROS while simultaneously controlling it carefully through inflammation, hypoxia-reoxygenation cycles and tumor cell-stroma interactions. Cancer cells therefore need to maintain the ROS level in a specific and safe window to survive. Redox regulation of cancer cell fate Redox regulation refers to the dynamic balance between oxidant and antioxidant processes that controls cellular signaling, metabolism and stress adaptation. Within this framework, oncogenes, tumor suppressors, and redox regulators are not individual factors in the control of cancer cell fate, but are instead part of redox networks whose signaling coordinates the behavior of cancer cells. One of the most significant current concepts in the field is that oxidative stress is a unifying organizing principle in the regulation of many hallmarks of cancer by a few limited areas of action. Here, we classify these mechanisms into four higher-order conceptual areas that contain the major regulatory hubs, contextual regulators and therapeutic vulnerabilities involved in redox-driven cancer progression. Redox signaling networks in proliferation and survival Kirsten rat sarcoma viral oncogene homolog (KRAS) and PI3K/Akt/mTOR pathways are identified as main redox-sensitive hubs that translate ROS production into proliferation and survival signaling. KRAS oncogene is a key regulator of the redox balance in cancer cells. Mutated KRAS drives ROS production through both RAC1-mediated activation of NOX4 and mitochondrial activation involving cyclooxygenase-2 and the electron transport chain, leading to the production of superoxide and H2O2 [45]. These ROS activate KRAS-dependent signaling pathways, including the mTORC2–Akt axis, which increases mitochondrial fragmentation and mitophagy. Crucially, ROS-dependence survival in mutant KRAS tumors is a selective therapeutic liability. In addition to this role in promoting ROS production, KRAS signals also promote an antioxidant program. The activation of NRF2 then stimulates GSH production and ferroptosis suppressor protein 1 (FSP1), which reduces lipid peroxidation using NADPH [46]. Mutant KRAS tumors also exhibit dependence on the Solute carrier family 7 member 11 (SLC7A11)/GSH/GPX4 axis for suppressing ferroptosis, and particularly aggressive cancers often rely on this axis [47]. These findings indicate that many oncogenes are closely linked both to enhanced ROS production and to the activation of adaptive antioxidant programs, thereby maintaining redox balance in a pro-tumorigenic state. The PI3K/Akt/mTOR axis further corroborates this idea, as activation of this pathway increases ROS production and tumor cell survival, promotes inflammation and enhances metabolic adaptation. ROS-mediated oxidation and inhibition of PTEN activates PI3K/Akt/mTOR in an iterative forward loop, increasing tumor proliferation, migration and invasion, in part through activation of downstream effector molecules like C-X-C motif chemokine receptor 4 [48]. Additionally, PI3K/Akt/mTOR signaling induces the production of inflammatory cytokines like IL-1β, IL-6 and TNF-α, thereby bridging redox stress to pro-tumorigenic inflammation [49]. The MYC proto-oncogene, bHLH transcription factor (MYC) oncogene also activates this axis by increasing mitochondrial dysfunction, resulting in mtROS production. MYC-driven mitochondrial fragmentation involving Dynamin-related protein 1/Mitofusin 1 (Drp1/MFN1) activates PI3K/Akt/mTOR signaling, as well as NF-κB signaling [50,51]. MYC-induced ROS hyper-production can also trigger metabolic stress, RNA degradation and cell death, showing a context-dependent function that may be therapeutically exploitable [52]. Finally, other oncogenic drivers of this axis such as B-Raf proto-oncogene, serine/threonine kinase (BRAF^V600E) enhance the antioxidant capacity of cancer cells and stimulate GSH synthesis through the induction of glutamate-cysteine ligase in order to counter oxidative stress and promote metastatic dissemination [53]. After human epidermal growth factor (EGFR) transactivation mediated by EGFR2, Bombesin receptor subtype-3 (BRS-3) is found to trigger a ROS dependent ERK signaling pathway to mediate proliferation and clonogenic capacity of non-small cell lung cancer [54]. The integration of these pathways suggests a common functional principle in which the cancer cell relies on ROS levels produced as signal molecules in order to maximize the proliferation-promoting signaling output, while at the same time suppressing the detrimental effects of overwhelming oxidative stress. Redox control of metabolic plasticity Redox signaling serves as a crucial driver of metabolic plasticity, encompassing adaptive changes in glycolysis, mitochondrial respiration, lipid metabolism and biosynthetic pathways that support tumor survival under oxidative stress. This ability of cancer cells to flexibly adjust to changing environmental conditions and to adapt to oxidative stress permits tumor growth under challenging circumstances. NRF2 is a central regulator of this axis. Physiological regulation of NRF2 ensures redox balance, but its constitutive activation in cancer by the mutation of KEAP1 or nuclear factor erythroid 2-related factor 2 (NFE2L2) enhances its target gene expression (antioxidant and detoxifying genes), promotes GSH synthesis and increases NADPH production. It protects the tumor cells against oxidative stress, and promotes tumor progression and therapy resistance to radio- and chemotherapy [9,10]. However, NRF2 signaling shows context-dependent effects. It is required for antioxidant signaling under moderate oxidative stress conditions, but sustained activation of NRF2 may be required for tumor cell survival and therapy resistance in established tumors. Decreased expression of NRF2 leads to suppression of antioxidant pathways, which favors oxidative stress-induced ferroptotic cell death [55]. Increased activity of PI3K/Akt and KRAS increases SLC7A11 and GPX4 expression, which inhibits lipid ferroptosis [47]. BRCA1-deficient tumors rely on NRF2, which can be induced by NF-κB and sphingosine kinase 1 (SPHK1) to prevent ferroptosis and enhance resistance to poly(ADP-ribose) polymerase (PARP) inhibitors [56]. NOX proteins regulate ROS production in a variety of ways which are central to metabolic adaptation. For example, NOX2 and NOX4 signaling promotes hypoxia-inducible factor 1 (HIF-1) stabilization and the subsequent induction of PI3K/HIF-1α/COX-2 cascade for angiogenesis, metabolic reprogramming and tumor development [57]. In addition, NOX4 plays an important role in maintaining CSC function, genomic instability and resistance in TME [58,59]. NOX5 activation in esophagus carcinoma promotes invasion by generating membrane-associated ROS, which activate Src signaling [60,61], whereas in breast cancer protein kinase C (PKC)-induced NOX1 and NOX5 signaling increases the matrix metallopeptidase-9 (MMP9) activation, leading to the acquisition of an invasive phenotype [62] (Fig. 1).Fig. 1Role of oncogenes, tumor suppressors, and dual regulators in the control of ROS and redox signaling in cancer. Tumor suppressor genes (i.e., p53, BRCA1, p16, and RB1) impede tumor progression and maintain redox homeostasis by counteracting ROS accumulation; however, their alteration leads to high ROS levels, favoring tumorigenesis. Oncogenes (i.e., MYC, KRAS, and PI3K/Akt/mTOR pathway) promote ROS generation, facilitating tumor proliferation, aggressiveness, inflammatory cytokine secretion, and ferroptosis resistance. Dual regulators (i.e., ERK1/2, JNK, NRF2, NOX) can promote both progression and apoptosis, affecting redox equilibrium and playing a role in both cancer cells growth and death. Increased mtROS levels also favor tumorigenesis, as altered signaling affects mitochondria function and pathways, leading to mitochondrial fission and mitophagy. The (+) symbols represent activation; (−) represents inactivation. Arrows symbolize a decrease (↓), or an increase (↑) in ROS levels or gene/protein expression. Blocking arrows (⊣) represent suppression or deletion. (*) represents mutated genes and resistance. Role of oncogenes, tumor suppressors, and dual regulators in the control of ROS and redox signaling in cancer. Tumor suppressor genes (i.e., p53, BRCA1, p16, and RB1) impede tumor progression and maintain redox homeostasis by counteracting ROS accumulation; however, their alteration leads to high ROS levels, favoring tumorigenesis. Oncogenes (i.e., MYC, KRAS, and PI3K/Akt/mTOR pathway) promote ROS generation, facilitating tumor proliferation, aggressiveness, inflammatory cytokine secretion, and ferroptosis resistance. Dual regulators (i.e., ERK1/2, JNK, NRF2, NOX) can promote both progression and apoptosis, affecting redox equilibrium and playing a role in both cancer cells growth and death. Increased mtROS levels also favor tumorigenesis, as altered signaling affects mitochondria function and pathways, leading to mitochondrial fission and mitophagy. The (+) symbols represent activation; (−) represents inactivation. Arrows symbolize a decrease (↓), or an increase (↑) in ROS levels or gene/protein expression. Blocking arrows (⊣) represent suppression or deletion. (*) represents mutated genes and resistance. These collective observations demonstrate that redox signaling does not only act as a passive output of metabolism, but also contributes actively to metabolic flexibility. Redox-induced genomic instability Excess or unregulated ROS promote genomic instability that drives tumor evolution and heterogeneity. The Breakpoint cluster region-Abelson tyrosine kinase (BCR-ABL) oncoprotein, a prototypical case, significantly increases ROS production, which induces oxidative DNA damage and breaks, and subsequent mutations during S and G2/M phases that promote tumor development and drug resistance in leukemias [63,64]. Tumor suppressors play key protective roles. For example, p53, a central coordinator of cellular redox response, activates antioxidant genes under moderate oxidative stress but undergoes pro-apoptotic activity in highly stressed cells, including the transcriptional activation of p53-inducible gene 3 (PIG3) and BCL2-associated X protein, and through the SHC adaptor protein (p66Shc) pathway [65]. However, mutated p53 uniquely alters the redox responsive signaling, selectively modulating NRF2 signaling and antioxidant gene expression [66]. The loss of p53 further deregulates metabolism and ferroptosis, favoring tumor progression [67]. Ataxia-telangiectasia (ATM) mutated plays a crucial role in sensing oxidative DNA damage, amplifies redox signaling cascade, and initiates DNA repair pathways. Its inhibition therefore results in protein oxidation, leading to an increased accumulation of ROS and oxidative cell death [68]. Other tumor suppressors, including LKB1, play protective roles by regulating mitochondrial metabolism and ROS production. The reduced activity of LKB1 is linked to the increase in mtROS and proliferation [69], while p38 MAPK mediates cell cycle and DNA stress responses and apoptosis following oxidative damage [[70], [71], [72], [73]]. Von Hippel-Lindau (VHL) mutations can enhance lipid ROS production and ferroptosis signaling while depressing GPX4 expression [74], whereas p16INK4a deletion induces mitochondrial dysfunction and oxidative stress [75]. Retinoblastoma (RB1) loss causes DNA damage, failed repair and tumor development, as demonstrated in CRISPR/Cas9-engineered RB1-deficient cancer cell models [76]. This implies that genomic instability is not purely stochastic, but instead a programmed response to dysregulated redox homeostasis (Fig. 1). Redox adaptations in therapy resistance Therapy resistance is closely associated with redox signaling, allowing cancer cells to cope with the therapeutic ROS that are induced by radio-, chemotherapy and targeted therapy. MAPK pathways (ERK1/2, JNK and p38) can be critically involved in redox adaptation and exhibit a threshold-dependent response to ROS levels. ERK1/2 are generally responsible for cell proliferation and survival at low ROS levels, and their hyperactivation is commonly detected in tumors such as triple-negative breast cancer (TNBC) or hepatocellular carcinoma (HCC) [[77], [78], [79]]. High ROS levels will activate stress kinases p38 and c-Jun N-terminal kinase (JNK), which could lead to endoplasmic reticulum stress, growth arrest and apoptosis [[70], [71], [72]]. These pathways are known to exhibit significant functional plasticity. JNK signaling has been implicated in tumor progression, metabolism reprogramming and inflammatory signaling (JAK-STAT3 and NF-κB) through the regulation of SOD2 and redox state [80,81]. In contrast, and depending on circumstances, it can also mediate ferroptosis and cell death in nasopharyngeal and HCC models [82,83]. Similarly, although most often found to be tumor-suppressive, p38 signaling can have protumoral effects in different TMEs [73]. Apart from its involvement in tumor growth and survival, activation of ERK1/2 by H2O2 has been reported to cooperate with p53 to provoke genomic stability and to induce apoptosis [84]. Moreover, constitutive hyperactivation of RAF-Mitogen-activated protein kinase kinase 1 (MEK)-ERK maintains mitochondrial function and redox homeostasis, whereas inhibition abrogates this state and renders cells susceptible to ROS-induced cell death, especially in combination with PARP inhibition [85]. Finally, tumor-immune cell interplay reinforces redox-driven resistance mechanisms. Cytokine signaling can be involved in ROS production (IL-6), and can strengthen tumor-immune cell inflammatory circuitries, indirectly supporting therapy resistance through immune evasion [86]. Taken cancer-type, these data show that redox-regulated pathways cannot be seen as simply unilateral but as context-dependent, thereby influencing outcomes from therapeutic interventions. Increased ROS levels or blockade of antioxidant defenses such as NRF2, GPX4 and GSH proteins could overcome therapy resistance and sensitize tumor cells to treatment. Spatio-temporal dynamics of redox regulation during tumor progression Far from being passive modifiers of cellular respiration, ROS mediate a dynamic redox network which changes as tumors grow. From initiation to metastatic spread, changes in the levels and function of ROS synergistically merge the effects of oncogenic signaling, metabolic adaptation, hypoxic stress, and TME remodeling, to increasingly direct tumor cell plasticity and survival. Here, ROS function as context-dependent modulators of tumor cell plasticity, with functions that vary significantly at different stages of tumorigenesis and in varied spatial locations and environmental contexts, ultimately orchestrating the gain of invasiveness, and metastatic capability. Redox-driven cellular plasticity and EMT during early tumorigenesis Early tumor progression is dominated by redox signaling over direct DNA damage, although ROS can still promote mutagenesis through oxidative damage to genomic DNA [87]. The increased ROS levels promote signaling events essential for EMT, in a precursor process to invasion and dissemination. EMT marks the crucial turning point at which the initial primary tumor starts to develop undifferentiated metastatic properties, with a variety of signaling events, including TGF-β, NF-κB, HIF-1α and Wnt/β-catenin, promoting its occurrence through redox modulation. One characteristic example is the transforming growth factor beta (TGF-β)/ROS axis, where TGF-β induces the generation of ROS, whose amplification further stimulates TGF-β signaling to form a positive feedback loop. This loop activation initiates pro-EMT transcriptional programs that repress E-cadherin expression through Snail under the control of NF-κB, and leading to acquisition of a migratory phenotype [88]. mtROS mediate TGF-β-induced EMT in the pancreatic cancer model, with a variety of ROS sources converging on this critical transition [89]. Combined with hypoxia and inflammation, ROS orchestrate these pro-EMT changes. NF-κB and HIF-1α activity induced by oxidative stress accelerates pro-EMT transcription, and is crucial in coordinating resistance to apoptosis, with metabolic changes such as enhanced glycolysis, hypoxic adaptation, and redox homeostasis. [90]. Similarly, signaling cascades like the Wnt/β-catenin pathway controlled by ROS are involved in promoting mesenchymal transformation and invasive growth [91]. Hence, in early tumor progression, moderate amounts of ROS regulate signaling events that lead to increased cellular plasticity. Significantly, ROS levels can exercise fine control over CSC states. Low ROS levels are associated with quiescence and a highly differentiated phenotype, whereas increased ROS levels are linked to stemness and mesenchymal features via NRF2 and HIF-1α-mediated signaling pathways [92,93]. This redox-driven cellular plasticity acts as a crucial link between EMT and resistance to cancer therapy, thereby underpinning tumor heterogeneity and adaptability. Hypoxic redox adaptation in tumor maintenance and angiogenesis When tumors exceed the nutrient and oxygen supply capacity of the existing blood vessels, focal hypoxic regions form within the TME, creating a new, both spatial and temporal layer of redox control. Here, ROS become key mediators bridging hypoxic adaptation, angiogenesis, metabolic reprograming and long-term tumor survival. Angiogenesis is a crucial process that promotes sustained tumor growth and occurs as a response to hypoxia [94]. Low to moderate angiogenesis increases in ROS under hypoxic conditions stabilize HIF-1α by inhibiting prolyl hydroxylases, and enhance the transcription of genes related to angiogenesis, glycolysis, survival and invasion [94,95]. NOX-derived ROS facilitate the continuous progression of signaling along this axis to maintain HIF-1α stabilization, vascular endothelial growth factor (VEGF) expression, endothelial cell glycolytic reprogramming and subsequent neovascularization [96,97]. mtROS promote pro-angiogenic signaling via VEGF receptor 2 activation of endothelial cells to further induce blood vessel remodeling within hypoxic niches [24]. More importantly, ROS-induced PI3K/Akt/mTOR signaling can induce HIF-1α stability independent of oxygen to maintain a pro-survival signaling environment regardless of the surrounding oxygen [57]. These events demonstrate a spatial progression in ROS activity, shifting from predominantly intracellular signaling at the tumor onset to intercellular and TME signaling within hypoxic regions. Tumor-derived exosomes further transmit ROS-associated and pro-angiogenic signals to endothelial cells and pre-metastatic niches at distant sites to amplify redox signaling throughout the tumor [98]. In addition to regulating angiogenesis, ROS are also key regulators of hypoxia-induced metabolic reprograming. Oxidative stress promotes mitochondrial dysfunction and accumulation of oncometabolites such as succinate, which further stabilizes HIF-1α signaling to reinforce glycolytic reprogramming [99]. Time-dependent sustained changes in this signaling from acute HIF-1α to chronic HIF-2α activation within long-term hypoxic niches favor tumor persistence, the maintenance of stem-like phenotypes and long-term adaptation in harsh TMEs [100]. Simultaneously, ROS-mediated redox plasticity can contribute to dynamically cycling between differentiated and CSC-like cell states to maintain the tumor cell heterogeneity that underlies therapy resistance [101]. Hypoxic tumor regions should therefore be perceived not simply as areas lacking oxygen, but as complex redox-adaptive ecosystems that utilize ROS as master integrators of stress, angiogenic signals, metabolic flexibility and cell plasticity to drive long-term tumor growth. These hypoxic niches act as redox-adaptive ecosystems capable of maintaining angiogenic signaling despite therapeutic intervention. Such adaptive redox reprogramming may underline the limited clinical success of ROS-targeted anti-angiogenic therapies, including NOX inhibitors and anti-VEGF agents such as bevacizumab [102,103]. Redox adaptation during metastatic dissemination Metastasis involves the most challenging cellular adaptation due to widespread systemic stress, including oxidative stress and immune response, encountered in the circulation and upon settlement in distant organs. The high ROS levels encountered in circulation are potentially detrimental to cancer cells, making redox adaptability essential for survival [104]. In order to cope with oxidative stress, metastatic cells exhibit adapted redox metabolism, in which mitochondrial GSH plays a central role not merely as a redox buffer but also as a regulator of stress-adaptive signaling, which provides robustness to hypoxic and oxidative damage. The acquisition of an adapted redox metabolism provides the cancer cell with increased survivability when encountering stressful environments such as the bloodstream and colonizing organs [105]. ROS thus promote not only invasion and migration at a cellular level, but are also the driving force behind oxidative stress encountered by the cells in circulation, and thereby mediate selective pressure, enabling only metastatic cells that can adapt their redox metabolism to survive and establish distant metastases [104]. In this context, antioxidant systems such as NRF2 signaling and GPX4-mediated lipid peroxide detoxification become essential for metastatic fitness and survival under oxidative stress [34,35]. In overall terms, these data strongly suggest that ROS function as dynamic redox regulators which drive tumor progression across time and space. From early tumorigenesis and EMT at moderate levels of signaling through TGF-β/NF-κB and HIF-1α pathways [88,90], through angiogenesis and metabolic adaptation under hypoxia mediated by NOX-derived ROS and HIF-1α stabilization [57,96,97], to the development of sophisticated antioxidant defenses that enable metastasizing cells to survive and successful distant colonization [104], ROS continuously adapt to and modulate the cellular microenvironment. This model accounts for the importance of tumor-stage-specific and context-dependent responses, and highlights the need for targeting dynamic redox regulation in cancer, rather than absolute redox levels, with targeted therapy. Redox regulation of anti-tumor immunity The interplay between oxidative stress and immune cells in the TME is a complex and highly interconnected system; however, it is critical in defining cancer progression and the therapeutic response. Immune cell function is mediated not by independent cell populations, but by signaling circuits regulated by redox (Fig. 2).Fig. 2Immune-cancer cell interactions under oxidative stress. Immune cells modulate the redox signaling networks in the TME. ROS levels modulate their functions by either activating or repressing various signaling pathways that directly and indirectly affect cancer cell growth and the TME, suppressing (via AMPK or STING activation) or promoting (i.e., increasing GPX or SOD concentrations-tumorigenesis). ROS levels can be low (↓), moderate or high (↑). Green arrows symbolize tumor progression-tumor survival, angiogenesis, metastasis and immunosuppression, while the red blocking lines (⊣) represent tumor suppression-tumor cells damage or death, increased immune system activation and cytotoxicity. Black arrows indicate induction of ROS production or cell activation; and black blocked lines cell function inhibition. (+), activation; (−), loss. Immune-cancer cell interactions under oxidative stress. Immune cells modulate the redox signaling networks in the TME. ROS levels modulate their functions by either activating or repressing various signaling pathways that directly and indirectly affect cancer cell growth and the TME, suppressing (via AMPK or STING activation) or promoting (i.e., increasing GPX or SOD concentrations-tumorigenesis). ROS levels can be low (↓), moderate or high (↑). Green arrows symbolize tumor progression-tumor survival, angiogenesis, metastasis and immunosuppression, while the red blocking lines (⊣) represent tumor suppression-tumor cells damage or death, increased immune system activation and cytotoxicity. Black arrows indicate induction of ROS production or cell activation; and black blocked lines cell function inhibition. (+), activation; (−), loss. The first paradigm is that these interactions may be organized in four functional areas: i) ROS-mediated anti-tumor cytotoxicity, ii) redox-driven immunosuppression, iii) redox regulation of antigen presentation (AP) and iv) innate immune redox control of adaptive immunity and immune cell plasticity. These areas are interconnected through substantial crosstalk and metabolic coupling. Redox mechanisms of innate anti-tumor cytotoxicity ROS mediate a direct anti-tumor immune response with their primary site of action being in the innate immune system. Neutrophils are the most effective producers of ROS in the immune system, generating O2• −, H2O2, and hypochlorous acid via a NOX2 dependent oxidative burst [106]. In neutrophils, stimulation with C5a and leukotriene B4 further increases ROS production through the activation of xanthine oxidase pathways, which kills tumor cells by collateral tissue damage [107]. ROS are also important in promoting the formation of neutrophil extracellular traps (NETs) through chromatin decondensation, which may promote tumor cell killing but also tumor spread [107,108]. ROS production in neutrophils is directed by cell polarization; N1 type cells exhibit significantly higher ROS production, pro-inflammatory cytokine secretion (IL-1β, TNF-α), and migration than N2 immunosuppressive neutrophils [109,110]. ROS can also mediate tumor cell killing by M1 macrophages. These cells use increased reliance on glycolysis and the pentose phosphate pathway (PPP) in conjunction with the NADPH oxidase system to kill tumor cells and mediate enhanced AP [111,112]. ROS derived from macrophages also mediate fibrogenesis, angiogenesis, and inflammatory signaling, thereby linking cytotoxic activity with tissue remodeling processes [112]. Natural killer (NK) cells are other ROS-sensitive effector cells that mediate cytotoxicity against tumor cells by releasing cytolytic granules such as perforin and granzyme, and pro-inflammatory cytokines such as TNF-α and Interferon-gamma (IFNγ) [113]. ROS-mediated damage to NK cells may be suppressed in IL-15 treated cells by enhanced thioredoxin and mTOR activity, and this increases thiol availability which in turn enhances the killing ability of NK cells [114]. Disruption of mammalian sterile 20-like kinase 1/2 (Mst1/2) in NK cells results in increased mtROS, and this leads to apoptotic cell death and tumor metastasis [115]. Inhibition of PI3K using idelalisib reduces NOX2 mediated ROS production from tumor supportive myeloid cells and restores NK cell function [116]. Overall, tightly controlled ROS production plays a critical role in mediating direct tumor cell killing from the innate immune system. Redox-mediated immunosuppression As well as killing tumors directly, ROS also play a pivotal role in mediating the immunosuppressive networks that permit tumor immune evasion. MDSCs are major effectors of ROS mediated tumor immunosuppression [117]. In general, these cells possess large quantities of mitochondria and high levels of ROS production, which acts directly to suppress T-cell, NK-cell and B-cell activity, while mediating enhanced programmed cell death protein 1 (PD-1) expression which causes immune exhaustion [118]. MDSC-mediated suppression is driven by fatty acid transport protein 2 (FATP2) activation induced by tumor derived granulocyte-macrophage colony-stimulating factor (GM-CSF) through STAT3, promoting lipid accumulation, increased ROS production and potentiation of immune suppression, and impairing responses to PD-L1 mediated therapy [119]. Other immune factors potentiate MDSC function, such as IL-6 increasing CCR5 expression and ROS production through STAT3 [105]. Moreover, microorganisms such as Candida tropicalis increase inflammasome activation, releasing IL-1β and inducing ROS through NOX2 [120]. The redox state of MDSCs is also regulated by internal systems, including the voltage-gated proton channel Hv1, which maintains ROS production, creating an immunosuppressive phenotype regardless of the surrounding TME [121], and reduced H2O2 levels which may prevent adequate dendritic cell (DC) differentiation [122]. M2 TAMs also mediate tumor suppression, showing significant metabolic plasticity and increased resistance to oxidative stress which maintains tumor promoting function [111]. Tumor-derived lactate induces an HIF-1α activation in TAMs, causing M2 polarization and tumor promoting processes including angiogenesis and EMT [123]. There is evidence of increased lipid metabolic reprograming involving an accumulation of phospholipids and sphingomyelins, which contributes to increased ROS production and TAM function [124]. GM-CSF mediated suppression of NRF2, leading to M2 functional shift and an IL-1β mediated tumor-promoting inflammatory response, involves NOX2 mediated mtROS production via TNF-α signaling [125]. Recent studies using single cell technologies identified M2 macrophages upregulating anti-oxidative mechanisms such as SOD3 at least partially maintaining an immunosuppressive phenotype in glioblastoma, whereas the removal of SOD3 limits tumor growth [126]. Moreover, long-term stimulation by tumor-exposed factors repolarizes neutrophil metabolism toward greater production of ROS, and induces a phenotypic shift toward immunosuppressive 'N2' neutrophil subsets that promote tumor immune evasion [110]. Thus, it is evident that the central immune-suppressive role of ROS mediated by a reprogrammed myeloid cell population is crucial in tumor immunity. Redox regulation of AP and innate immune sensing In addition to the cytotoxic and immunosuppressive pathways involving ROS, redox also plays a critical role in the immune system's ability to attack tumors, particularly with regard to DCs. The process of DC maturation and AP requires a precisely controlled redox environment. Mitochondrial metabolism, and particularly electron flow through the electron transport chain, is also essential for DCs to rapidly induce an immunogenic response; the disruption of complex III impairs their costimulatory capacity, migration, and ability to activate antitumor T-cells [127]. Moderate ROS production from innate immune cells supports the activation of innate receptors and signaling pathways such as stimulator of interferon genes (STING) and Toll-like receptors, promoting potent cytotoxic CD8+ T-cell responses [128,129]. However, excessive ROS results in immature DC development, decreased expression of co-stimulatory molecules (CD80/CD86), and a lack of cytokine production, leading to attenuated anti-tumor immunity. These results are supported by murine models, where tumors induced a lack of early GM-CSF-derived DCs in bone marrow cultures, leading to an absence of any anti-tumor immune responses [130]. ROS from tumor cells inhibit both DC maturation and capacity for cross AP, with NOX2 mediated controlled endosomal pH acting to preserve antigen [131]. Levels of ROS that are too low conversely prevent this process [132], while elevated ROS can also damage DCs and limit their viability, thereby preventing the formation of anti-tumor immune responses [133]. Similar ROS dependent control mechanisms operate on macrophages in AP and cytokine release [112]. These studies collectively illustrate the immense sensitivity of AP to ROS, with levels that are either too low or too high being suppressive. Redox control of adaptive immunity and immune cell plasticity Adaptive immunity and the function of T- and B-cells are heavily regulated by redox signaling. T-cells rely on moderate ROS levels for their activation. ROS produced upon activation of the T-cell receptor (TCR) signaling cascade is a modulator of MAPK and mTOR pathways, calcium signaling and activation of NFAT which result in the induction of T-cell proliferation and function [134]. ROS increase the mitochondrial potential and metabolic rate of CD8+ T-cells, augmenting their cytotoxicity [135]. Moreover, high ROS levels affect mitochondrial function, leading to calcium dysregulation and a decrease in IL-2 production, and subsequently promote the exhaustion and dysfunction of T-cells [136]. There is a differential effect of redox status on different T-cell subsets: whilst effector T-cells are susceptible to oxidative damage, are rapidly destroyed or become dysfunctional, regulatory T-cells (Tregs) upregulate GPX and SOD to maintain the reduced TME, enabling them to survive and to retain their immunosuppressive function [137]. Tumor cells release signals such as GM-CSF to stimulate ROS production by granulocytes, which will indirectly suppress T-cell proliferation and function [138]. B-cell function and proliferation is heavily regulated by redox signaling. The activation of the B-cell receptor promotes a swift increase in ROS production, either by NOX2 or mitochondrial respiration, which is a prerequisite for the initiation of the PI3K pathway and for the subsequent cell division [139]. Beyond early B-cell activation, NOX2 signaling may also influence later B-cell fate, as impaired NCF4-dependent intracellular ROS production has been shown to promote terminal differentiation toward plasma cells [140]. Unlike T-cells, high levels of ROS in B-cells lead to an increase in AMP-activated protein kinase (AMPK) pathway activity and subsequent apoptosis or autophagy [141]. Cytokines, such as IL-6, promote cell survival despite a rise in ROS production through the upregulation of BCL-2 [142]. The negative regulation of ROS signaling in B-cells is also operated by proteins including C-terminal Src kinase, which negatively regulate ROS production during the selection processes occurring in germinal centers. Misfolded protein accumulation and impaired autophagy compromise B-cell viability and function [143,144]. Intercellular redox crosstalk within the TME Redox signaling also contributes to an intensive intercellular crosstalk between immune cells and tumor cells. ROS can regulate immune checkpoint engagement. For instance, exosomal miRNAs, such as exo-miR-155-5p, which increases PD-L1 expression in macrophages and represses CD8+ T-cell activity, can mediate their immunosuppressive effect with the increase in ROS production [145]. Inhibition of ROS production, such as targeting Hv1, has immunosuppressive effects on myeloid cells, and could potentiate the effect of T-cell stimulation [121]. Redox signaling regulates the process of cross-presentation of antigens as well as their subsequent intracellular trafficking. Accum is an innovative intracellular delivery platform that improves antigen processing and immune presentation by facilitating the escape of antigenic cargo from endosomes into the cytosol. This is achieved through the combination of endosome-disruptive and nuclear localization properties, which limit lysosomal degradation and enhance antigen cross-presentation to T-cells [146]. Neutrophil-derived and macrophage-derived ROS are also important modulators of various vascular remodeling mechanisms and angiogenesis. These processes often require the orchestrated function of various cells, and are critically dependent on redox signaling [147,148]. Chronic oxidative stress in the TME coordinates immune cell communication toward either improved anti-tumor activity, or immunosuppression and tumor survival. Metabolic plasticity as a redox-adaptive program Historically, cancer metabolism was characterized mainly by the Warburg effect: the preferential and excessive use of aerobic glycolysis for lactate production, even in the presence of oxygen, this fact was interpreted as the more limited reliance of tumor cells on the mitochondrial respiratory system, and a preferential use of glycolysis rather than respiration. Nevertheless, this vision has recently been deemed too simplistic, and mitochondrial respiration is now considered crucial for many cancer types, especially for treatment-resistant, metastatic and CSC phenotypes. The Warburg effect is currently understood to be a part of a general, redox-dependent adaptive response, in which cancer cells can utilize both the mitochondrial respiratory and glycolysis system simultaneously, depending on the conditions [149]. In this redox context, mitochondria act as signaling hubs, and ROS production is closely linked to metabolic decision-making. Low levels of mtROS act as a mitogenic and pro-adaptive signaling factor, while overwhelming ROS prompts oxidative damage and cell death [150]. This coupling explains the close coordination between redox buffering systems and adaptive metabolic rewiring. The tight interaction between oxidative stress, mitochondria dysfunction, cell survival pathways and therapy resistance relies heavily on metabolic adaptability. Pathways governed by HIF-1α and AMPK activation regulate these metabolic adaptions which due to their action lead to the creation of a “hybrid metabolism” that simultaneously supports cell proliferation, biosynthesis and the redox status [149,[151], [152], [153]]. The reprogramming of cancer cell metabolism therefore cannot be defined by a simple glycolysis/OXPHOS dichotomy, but instead as an oxidative stress-constrained adaptive system of metabolic regulation which keeps the redox status within limits allowing tumor growth and survival [149]. Other pathways to OXPHOS include the redirection of metabolic intermediates from glycolysis toward the PPP for NADPH production and the restoration of redox balance [154,155]. Rather than acting as a subsidiary pathway, PPP acts as a critical redox-regulating hub which is dynamically controlled by ROS-sensitive glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in coordination with NRF2 and the PI3K/mTOR signaling pathway [[156], [157], [158]]. Tumors use other metabolic pathways to maintain their proliferation in situations of nutrient deprivation, hypoxia and mitochondrial defects in glutaminolysis. Glutaminolysis enables cancer cells to manage oxidative stress by assisting the production of GSH. It generates TCA cycle intermediates and biosynthetic precursors; meanwhile, reductive carboxylation supports glutamine-derived citrate and acetyl-CoA production under hypoxia [119,159,160]. Moreover, mitochondrial one-carbon metabolism through SHMT2 and MTHFD2 upregulation further promotes tumor growth and metabolic adaptation [161]. Importantly, oncogenic and tumor suppressor alterations further reinforce this redox-adaptive metabolic state. The loss of p53 in KRAS-mutant tumors causes a heavy dependency on glucose-6-phosphate dehydrogenase (G6PD)-dependent PPP activity in order to maintain NADPH production and promote DNA repair, showing that genetic lesions converge on redox-dependent metabolic pathways [162]. Moreover, ROS mediated stabilization of HIF-1α drives glycolytic flux through increased hexokinase 2 (HK2), phosphoglycerate kinase 1 (PGK1), and pyruvate dehydrogenase kinase 1 (PDK1) activity, and therefore forms a feedforward loop that directly dictates the metabolic profile [163,164]. Redox adaptation is also regulated at the post-transcriptional level by ROS-responsive miRNAs. Hypoxia- and ROS-responsive miRNAs such as miR-210 mediate the coordination of metabolic adaptation, DNA repair, and survival [165,166], while miR-128-3p controls ferroptosis by redox-dependent regulation of GSH metabolism and SLC7A11 expression [167,168]. Collectively, these findings support the concept of ROS as upstream regulators of metabolic adaptation rather than simple metabolic by-products. At a systemic level, metabolic reprogramming is refined through compartmentalization of the metabolic phenotype within the TME. Hypoxic niches favor glycolysis and reductive metabolism, while more oxygenated regions or therapy-exposed cells rely on mitochondrial respiration. This spatial heterogeneity is reinforced by stromal interactions, including mitochondrial transfer from stromal cells and metabolic coupling through lactate and fatty acid exchange, which collectively enhance tumor survival and redox buffering capacity [169,170]. In addition, oncometabolites are master integrators of metabolic and redox signaling. Accumulation of fumarate, succinate and 2-hydroxyglutarate due to fumarate hydratase (FH), succinate dehydrogenase (SDH) and isocitrate dehydrogenase (IDH) mutations result in ROS generation and altered redox states of the cell, while reshaping the tumor's transcriptional profile and immunological response [171,172]. The interplay between them and the downstream signaling network clearly illustrates the complexity of the changes initiated by their presence. A critical translational implication of this integrated model is that metabolic plasticity constitutes a major barrier to therapy, as redox adaptive rewiring enables the cancer cell to survive the oxidative stress generated by the treatment itself. Enhanced mitochondrial biogenesis, increased mitophagy, and the shift to fatty acid metabolism improve the cell's capacity for redox balance under treatment [170,173]. Conversely, targeting key pathways that boost redox adaptive metabolic switches can also provide therapeutic opportunities [162,[174], [175], [176]]. On the other hand, mitochondrial replication promotes cancer cell growth through mechanisms that do not depend on OXPHOS [177]. Finally, novel approaches targeting metabolism suggest the use of specific markers such as mitochondrial ribosomal gene expression and OXPHOS signatures as predictors of treatment response [178]. Circadian chronotherapy approach can increase treatment efficacy [179]. In overall terms, this reflects a transition to targeting the complexity of the redox metabolic signaling network rather than individual pathways. Redox regulation of CSCs CSCs can be divided into two cell types: a slow-cycling, quiescent population responsible for therapy resistance and recurrence, and a rapidly cycling population associated with tumor growth [180]. Quiescent CSCs are not merely unique tumor subsets, but rather redox-defined states with high adaptability, self-renewal, and resistance to environmental and therapeutic stresses. Instead of intrinsic cell fates, CSCs exist as a state formed by state-transition, dictated by redox thresholds that bridge ROS, stemness, and differentiation. The most critical organizing principle is that proliferative CSCs possess lower intracellular redox states than differentiated tumor cells. However, this condition does not remain passive, and it is instead a result of regulated metabolic reprogramming and upregulated antioxidant systems. Similar to normal stem cells, the optimal metabolic status for CSCs involves limited production of ROS with preserved signaling capacity. This is demonstrated by their enrichment with glycolytic pathways and attenuated mitochondria activity [181,182]. However, the statement is not absolute, and there is increasing evidence to suggest that CSCs exploit mitochondrial metabolism when necessary, displaying metabolic and redox plasticity [183]. This plasticity comes from a general concept whereby CSCs exhibit redox-metabolic landscape with high flexibility. In certain contexts, CSCs increase mitochondrial mass, fusion dynamics and OXPHOS to maintain bioenergetic homeostasis and maintain their stem-like state, as demonstrated in pancreas cancer models [183]. Meanwhile, some events, such as mitochondrial uncoupling and substrate selection of specific metabolic pathways, reduce mtROS production and protect CSCs from oxidative damages [182]. This further implies that CSCs have no unique metabolic program, but instead choose their metabolic status according to dynamic requirements of the redox state and biological functional integrity. An important characteristic for the control of CSCs is that non-canonical antioxidant pathways, such as KEAP1-independent NRF2 signaling and forkhead box O (FoxO)-mediated stress adaptation, contribute to their regulation beyond conventional redox effectors. NRF2 activation independently from KEAP1 often induces a potent antioxidant response, preventing excessive ROS accumulation and conferring chemotherapy resistance [184]. This is tightly modulated by a transcriptional network involving FoxO transcription factors that regulate metabolism and redox signaling under stress. mTOR inhibition reverses all these effects inhibiting FoxO signaling and the consequent stemness properties [185]. All these pathways provide CSCs with strong redox protection mechanisms, so that they can survive under adverse cytotoxic conditions. Furthermore, the redox state of CSCs does not arise from intrinsic cell biology, but instead relies on bidirectional regulation of the TME. Hypoxic conditions within the TME maintain HIF1-α stabilization, and induce redox adaptive pathway, including miRNA mediated effects like miR-210, which drives CSCs stemness and EMT [186]. Moreover, stromal signals and lipid peroxidation products directly modulate NRF2 and GSH pathways, increasing antioxidant ability and ensuring CSC survival [93]. This suggests that CSCs indeed survive in special redox niches in the TME which sustain their stem-like states. Another important concept is the dynamic role of ROS in CSC state transitions. Instead of passively exhibiting a reduced oxidative state, CSCs can use dynamic changes in intracellular ROS levels to control differentiation and state transition. Low intracellular ROS are related to cell quiescence and long-term stem cell maintenance, while a transient moderate increase in intracellular ROS can drive cell plasticity, generating several stemness stages among the CSC population, including EMT transition via upregulation of NRF2 and HIF-1α pathways [92,93]. These phenomena create a model in which CSCs have redox-gradients as molecular determinants rather than static CSC markers. miRNA-ROS interaction forms a positive feedback loop within cells to regulate the redox state and CSC stemness. mir-27a, among others, decreases ROS production via targeting antioxidant regulators, consequently compromising CSC survival and chemotherapeutic sensitivity [187]. Conversely, miRNAs such as miR-223 and miR-3065-5p enhance antioxidant capacity and reinforce stem-like phenotypes by regulating NRF2-dependent pathways and redox homeostasis [184]. These interactions show that CSC behavior emerges from integrated regulatory circuits rather than isolated pathways. In terms of translational studies, CSCs rely on their enhanced antioxidant function and metabolic plasticity, which may hinder standard therapy as its effect is mainly based on increased ROS production in tumor cells. However, these traits are precisely what create targets for therapy; interference of the redox homeostasis could selectively kill CSCs by overwhelming antioxidant defenses or blocking their metabolic flexibility. For example, blocking NRF1 mediated antioxidant pathway via CD13 inhibition greatly sensitizes CSCs to chemotherapy by ROS accumulation and inhibition of stem-like traits [188]. CSCs can be defined as a redox adaptive state coupled with a TME-specific metabolic network. The identity of CSCs is regulated by their ability to survive in a wide range of redox conditions; therefore, a treatment that alters redox networks sustaining CSCs might therefore constitute a more effective therapeutic strategy, and prevent tumor recurrence. Targeting redox adaptation in cancer therapy Redox thresholds as a therapeutic vulnerability Therapeutic targeting of oxidative stress in cancer should not be viewed as simply a pro-oxidant versus an antioxidant framework. Cancer cells generally operate at their limit of redox tolerance where they rely on elevated ROS to drive oncogenic signaling but, at the same time use it to place constraints on cell growth [189]. In the context of such a tightly regulated redox balance, overwhelming tumor specific redox thresholds could lead to targeted cell death. Classical anti-cancer therapies like radio- and chemotherapy promote cancer cell death through DNA damage and the activation of p53, PI3K/Akt/mTOR, and/or MAPK cascades allowing ultimately to apoptosis cell death [190]. However, resistance to therapy occurs through the same pathways; upregulation of antioxidants such as SOD, catalase, and GPX are crucial for restoring redox balance [191], while activation of the PI3K/Akt, NF-κB, MAPK, and HIF-1α signaling cascades are pro-survival pathways that enhance cancer cell growth and metastatic capability by stimulating angiogenesis, metabolic reprogramming, and the development of therapy-induced phenotypes, respectively [192,193]. Therapy-induced ROS can exert both anti-tumor and pro-survival effects depending on the redox-adaptive response of cancer cells; effective anti-cancer strategies must therefore consider the complexity and plasticity of redox balance (Table 1).Table 1Overview of therapeutic strategies targeting oxidative stress and redox homeostasis in cancer, including pro-oxidant and antioxidant strategies, their molecular mechanisms, and clinical relevance.Table 1TreatmentEvidence LevelCancer TypeMechanismRedox RoleRelevanceReference2-DG + GemcitabinePreclinical evidence (combination includes approved therapy)Pancreatic cancerInhibits glycolysis while gemcitabine induces DNA damage, resulting in increased reactive oxygen species production and oxidative stress-mediated cytotoxicityPro-oxidantEnhances chemosensitivity, suppresses CSC and EMT phenotypes, and reduces tumor growth and metastasis[194]3-BromopyruvateClinical testingTNBCAlkylates glycolytic enzymes including GAPDH and depletes glutathione, causing disruption of redox balance and accumulation of reactive oxygen speciesPro-oxidantEnhances radiosensitivity and selectively targets metabolically active cancer cells, offering potential to overcome therapeutic resistance[195]Accum-OVA-pulsed mature DC vaccinationPreclinical evidenceT-cell lymphomaDisrupts endosomal membranes to enhance antigen escape and processing, promoting oxidative stress-associated immune activationPro-oxidantInduces strong T-cell activation and immune memory, delays tumor growth, and synergizes with immune checkpoint blockade to improve survival[146]AuranofinPreclinical evidenceTNBCInhibits thioredoxin reductase, impairing antioxidant defense systems and promoting accumulation of reactive oxygen speciesPro-oxidantInduces oxidative imbalance in cancer cells and represents a promising redox-targeting therapeutic strategy[196]BAY 87-2243Preclinical evidenceBRAF-mutant melanomaInhibits mitochondrial complex I, increasing mitochondrial and cytosolic reactive oxygen species and inducing oxidative stress-mediated cell deathPro-oxidantReduces tumor growth and enhances efficacy when combined with targeted therapies such as BRAF inhibitors[174]CD13 inhibitor (Ubenimex)Approved therapy and clinical testingHCCInhibits CD13 activity, blocking ROS-scavenging pathways and increasing intracellular ROS accumulation, leading to enhanced DNA damagePro-oxidantReverses chemotherapy resistance and enhances cytotoxicity of anticancer agents through redox imbalance[188]Chloroquine + CisplatinPreclinical evidence (includes approved therapies)Drug-resistant epithelial ovarian cancerInhibits autophagy and increases mtROS accumulation, leading to enhanced DNA damage through p21-mediated pathwaysPro-oxidantEnhances cytotoxicity in resistant cancer cells and overcomes chemoresistance by promoting oxidative damage[176]Compound 7b (triazolothiadiazine derivative)Preclinical evidenceHCCInduces ROS accumulation and activates oxidative stress-dependent JNK signaling, leading to apoptosis and cell cycle arrestPro-oxidantInhibits tumor growth, induces apoptosis and G2/M arrest, and reduces CSC enrichment[83]CurcuminPreclinical evidenceChemoresistant lung cancerInduces ROS accumulation, leading to activation of p38 MAPK signaling and ER stress-associated pathways, ultimately triggering caspase-dependent apoptosisPro-oxidantOvercomes chemoresistance by promoting ROS-mediated apoptosis in drug-resistant lung cancer cells, supporting its potential as an adjuvant strategy to sensitize resistant tumors[72]CWB-20145 (phenylboronic acid nitrogen mustard)Preclinical evidenceTNBCUndergoes ROS-activated DNA interstrand cross-linking, selectively inducing cytotoxicity in oxidative environmentsPro-oxidantPromotes selective tumor cell death with improved safety and efficacy compared to conventional alkylating agents[197]Elesclomol-Cu + cinnamaldehyde nanoplatformPreclinical evidenceBreast cancerGenerates ROS through copper-mediated redox cycling and cinnamaldehyde-induced oxidative stress, forming a positive feedback loop that amplifies ROS productionPro-oxidantEnhances tumor accumulation, induces immunogenic cell death, and activates antitumor immune responses[198]EmpagliflozinClinical testing (approved for other diseases)Breast cancerInduces ferroptosis by modulating miR-128-3p signaling, reducing GPX4 and GSH levels, and increasing intracellular ROS and iron-dependent lipid peroxidationPro-oxidantReduces metastatic burden and sensitizes resistant cancer cells to ferroptosis, supporting its role as an adjunct therapy[168]hsa-miR-27a overexpressionPreclinical evidenceBreast cancerDownregulates ROS defense genes and impairs GSH synthesis, leading to increased intracellular ROS levelsPro-oxidantSensitizes cancer cells to chemotherapy and reduces CSC-associated resistance mechanisms[187]Ir2 (Iridium(III) complex with α-tocopheryl succinate)Preclinical evidenceCervical, lung, and breast cancerInduces excessive ROS production, causing oxidative stress-mediated DNA damage, mitochondrial depolarization, and dual inhibition of glycolysis and OXPHOSPro-oxidantExhibits potent and selective anticancer activity through mitochondrial targeting, inducing apoptosis and autophagy[199]LysionotinPreclinical evidenceColorectal cancerInduces ferroptosis by promoting NRF2 degradation, leading to ROS accumulation and lipid peroxidationPro-oxidantSuppresses tumor growth and metastatic behavior without significant toxicity[55]Trametinib + OlaparibClinical testing (both approved therapies)Malignant pleural mesotheliomaDisrupts redox homeostasis by combining MAPK pathway inhibition with impaired DNA repair, leading to metabolic and oxidative stress-induced apoptosisPro-oxidantDemonstrates strong synergistic antitumor activity and offers a promising strategy for difficult-to-treat cancers[85]Methionine sulfoxide reductase A modulationPreclinical evidencePancreatic ductal adenocarcinomaRegulates methionine oxidation on PKM2, altering redox signaling and metabolic pathwaysPro-oxidantSuppresses metastasis through redox-dependent metabolic reprogramming[140]miR-210 overexpressionPreclinical evidenceColorectal cancerInduces ROS generation and activates oxidative stress-dependent apoptotic pathwaysPro-oxidantPromotes apoptosis and inhibits proliferation, representing a potential therapeutic target[166]miR-21-3p-loaded gold NPs + Anti-PD-1Preclinical evidence (includes approved therapy)MelanomaPromotes ferroptosis by increasing lipid peroxidation and ROS accumulationPro-oxidantEnhances immunotherapy efficacy by sensitizing tumor cells to ferroptotic cell death[200]Mitochondrial complex I inhibitor (IACS-010759) + LDH inhibitor (NCI-006)Preclinical evidence (includes clinical agent)Colorectal and gastric cancerBlocks metabolic flexibility by inhibiting both glycolysis and OXPHOS, leading to energy depletion and increased ROS levelsPro-oxidantProduces synergistic antitumor effects through metabolic disruption and redox imbalance[201]NF-κB inhibitors (QNZ and BMS-345541)Preclinical evidenceLung cancerReduce expression of antioxidant transporters, decreasing GSH synthesis and increasing ROS levelsPro-oxidantSensitizes cancer cells to ferroptosis and enhances vulnerability to oxidative stress-based therapies[202]NIR-715 PDTPreclinical evidenceEsophageal and breast cancer, melanomaGenerates ROS upon light activation, inducing endoplasmic reticulum stress and ICDPro-oxidantEnhances antitumor immunity and suppresses tumor growth with low toxicity[203]NOX1 inhibitor (GKT771)Preclinical evidenceColorectal cancerInhibits NOX1-mediated ROS production associated with tumor proliferation and angiogenesisPro-oxidantReduces tumor growth and improves response to immune checkpoint blockade[102]Methyl 3-(4-nitrophenyl) propiolatePreclinical evidenceMultiple cancer typesInduces ROS production via cytochrome P450 activity, leading to oxidative stress and mitochondrial apoptosisPro-oxidantSelectively targets cancer cells with high basal ROS and disrupts oncogenic signaling pathways[11]Olaparib + SPHK1 inhibitor (PF-543)Preclinical evidence (includes approved therapy)Ovarian cancerInhibits SPHK1 signaling, reducing NRF2 activity and promoting ferroptosis through increased oxidative stressPro-oxidantEnhances treatment sensitivity and provides a strategy to overcome resistance[56]PEGylated mesoporous cerium oxide-coated upconversion NPsPreclinical evidenceCervical carcinomaDecomposes H2O2 to generate O2 and enhances ROS-mediated cytotoxicity during photodynamic therapy, thereby alleviating hypoxiaPro-oxidantImproves PDT efficacy, supports synergistic chemotherapy, and inhibits tumor growth with low systemic toxicity[204]Perfluorohexane-Ce6 liposome O2 nanodroplets + sonodynamic therapy + Anti-PD-1Preclinical evidence (includes approved therapy)Colorectal cancerGenerates ROS through sonosensitizer activation, inducing oxidative stress, apoptosis, and ICDPro-oxidantEnhances immunotherapy by improving immune activation, reducing regulatory T-cells, and promoting long-term immune memory[205]Poly(lactic-co-glycolic acid)-encapsulated MnO2 NPsPreclinical evidenceBone metastasisReact with H2O2 to generate oxygen, reducing hypoxia-associated signaling and remodeling the oxidative TMEPro-oxidantReverses hypoxia, improves antitumor immunity, decreases Tregs, and prolongs survival[206]RSL3 + ErastinPreclinical evidenceOsteosarcomaBlocks cystine import and GPX4 activity, promoting accumulation of lipid radicals and ferroptosisPro-oxidantSelectively eliminates osteosarcoma cells and undifferentiated stem cells while sparing normal bone cells[207]SR18292Preclinical evidenceMultiple myelomaInhibits PPAR-γ coactivator-1-alpha (PGC-1α)-driven OXPHOS, inducing ROS accumulation, energy depletion, and apoptosisPro-oxidantSuppresses tumor growth and promotes apoptosis with minimal toxicity[175]Wireless implantable electrochemical drug delivery system with polypyrrole NPsPreclinical evidenceMultiple cancer typesEnables electrochemical redox-controlled drug release through oxidation-reduction reactions in polypyrrole NPsPro-oxidantAllows precise and programmable drug release, supporting targeted therapy with reduced systemic toxicity[208]AcriflavinePreclinical evidenceGlioblastoma stem-like cellsInhibits HIF-1 dimerization, reducing RAGE expression and hypoxia-associated inflammatory signalingAntioxidantInterferes with hypoxia- and inflammation-driven glioblastoma progression and supports targeting of stem-like tumor cells[209]Berberine, anastrozole, or pranoprofenPreclinical evidence (includes approved therapy)Colitis-associated colon cancerPromote Dicer expression, reducing oxidative stress-induced DNA damage and inflammatory IL-6 signalingAntioxidantAlleviate inflammation and suppress inflammation-driven carcinogenesis, supporting preventive and therapeutic potential[32]DaphnetinPreclinical evidenceLung adenocarcinomaInhibits MAPK, STAT3, and NF-κB signaling while reducing oxidative stress-associated signaling activityAntioxidantSuppresses proliferation, migration, invasion, and clonogenic growth, supporting its potential as a therapeutic agent[81]FluorofenidoneClinical testingPulmonary fibrosisReduces oxidative stress by increasing SOD activity and decreasing lipid peroxidation, while inhibiting PI3K/Akt/mTOR signaling and promoting autophagyAntioxidantAlleviates fibrosis, inflammation, EMT, and oxidative damage, supporting therapeutic potential in fibrotic disease[49]IDO1 inhibitor (Epacadostat)Clinical testingLung squamous cell carcinomaBlocks IDO1 signaling, reducing kynurenine production and associated ROS-mediated immunosuppressionAntioxidantPartially restores T-cell function by limiting metabolic and oxidative immune evasion[38]lncRNA HITTPreclinical evidenceColon, cervical, and lung cancerInhibits PKM2 oligomerization, reducing glycolysis and lactate production and modulating macrophage polarization within the TMEAntioxidantSuppresses tumor growth, sensitizes cancer cells to nutrient stress, and promotes antitumor immune responses by shifting macrophages toward an M1 phenotype[210]Lonidamine + 5-(tetradecyloxy)-2-furancarboxylic acidPreclinical evidence (includes approved therapy)Ovarian and colon cancer CSCsSimultaneously inhibits glycolysis and lipolysis, disrupting metabolic pathways associated with oxidative stress adaptation and CSC survivalAntioxidantReduces CSC viability more effectively than either agent alone and supports metabolic targeting of resistant cell populations[211]MetforminApproved therapy and clinical testingTNBCInhibits mitochondrial respiratory complex I and activates AMPK, reducing ATP production and imposing energetic and redox stress on tumor cellsAntioxidantEnhances chemotherapy efficacy, may reduce required drug doses, and exerts additional immunostimulatory effects[212]MitoQClinical testing (approved as a dietary supplement)Breast cancerReduces mitochondrial superoxide production and limits oxidative damage associated with metastatic progressionAntioxidantShows preventive potential against cancer recurrence and metastasis through mitochondrial redox control[213]MnTE-2-PyP5+ or DPIPreclinical evidenceBreast cancerScavenge ROS or inhibit NOX activity, reducing ROS-dependent STAT3 activation required for M2 macrophage polarizationAntioxidantSuppress protumoral macrophage function and may enhance antitumor immunity when combined with immunotherapy[214]mTOR inhibitor (Rapamycin)Approved therapy and clinical testingLung adenocarcinomaInhibits mTOR signaling, reversing Warburg-like metabolic reprogramming and helping normalize redox balanceAntioxidantReduces stemness, invasiveness, viability, and colony formation, supporting therapeutic relevance in metabolically active tumors[185]NOX2 inhibitor (GSK-2795039)Preclinical evidenceLung squamous cell carcinoma and other solid tumorsInhibits NOX2-mediated ROS production in MDSCs, reducing oxidative immunosuppression in the TMEAntioxidantRestores CD8+ T-cell proliferation and supports immune reactivation in solid tumors[38] Table 1Overview of therapeutic strategies targeting oxidative stress and redox homeostasis in cancer, including pro-oxidant and antioxidant strategies, their molecular mechanisms, and clinical relevance.Table 1TreatmentEvidence LevelCancer TypeMechanismRedox RoleRelevanceReference2-DG + GemcitabinePreclinical evidence (combination includes approved therapy)Pancreatic cancerInhibits glycolysis while gemcitabine induces DNA damage, resulting in increased reactive oxygen species production and oxidative stress-mediated cytotoxicityPro-oxidantEnhances chemosensitivity, suppresses CSC and EMT phenotypes, and reduces tumor growth and metastasis[194]3-BromopyruvateClinical testingTNBCAlkylates glycolytic enzymes including GAPDH and depletes glutathione, causing disruption of redox balance and accumulation of reactive oxygen speciesPro-oxidantEnhances radiosensitivity and selectively targets metabolically active cancer cells, offering potential to overcome therapeutic resistance[195]Accum-OVA-pulsed mature DC vaccinationPreclinical evidenceT-cell lymphomaDisrupts endosomal membranes to enhance antigen escape and processing, promoting oxidative stress-associated immune activationPro-oxidantInduces strong T-cell activation and immune memory, delays tumor growth, and synergizes with immune checkpoint blockade to improve survival[146]AuranofinPreclinical evidenceTNBCInhibits thioredoxin reductase, impairing antioxidant defense systems and promoting accumulation of reactive oxygen speciesPro-oxidantInduces oxidative imbalance in cancer cells and represents a promising redox-targeting therapeutic strategy[196]BAY 87-2243Preclinical evidenceBRAF-mutant melanomaInhibits mitochondrial complex I, increasing mitochondrial and cytosolic reactive oxygen species and inducing oxidative stress-mediated cell deathPro-oxidantReduces tumor growth and enhances efficacy when combined with targeted therapies such as BRAF inhibitors[174]CD13 inhibitor (Ubenimex)Approved therapy and clinical testingHCCInhibits CD13 activity, blocking ROS-scavenging pathways and increasing intracellular ROS accumulation, leading to enhanced DNA damagePro-oxidantReverses chemotherapy resistance and enhances cytotoxicity of anticancer agents through redox imbalance[188]Chloroquine + CisplatinPreclinical evidence (includes approved therapies)Drug-resistant epithelial ovarian cancerInhibits autophagy and increases mtROS accumulation, leading to enhanced DNA damage through p21-mediated pathwaysPro-oxidantEnhances cytotoxicity in resistant cancer cells and overcomes chemoresistance by promoting oxidative damage[176]Compound 7b (triazolothiadiazine derivative)Preclinical evidenceHCCInduces ROS accumulation and activates oxidative stress-dependent JNK signaling, leading to apoptosis and cell cycle arrestPro-oxidantInhibits tumor growth, induces apoptosis and G2/M arrest, and reduces CSC enrichment[83]CurcuminPreclinical evidenceChemoresistant lung cancerInduces ROS accumulation, leading to activation of p38 MAPK signaling and ER stress-associated pathways, ultimately triggering caspase-dependent apoptosisPro-oxidantOvercomes chemoresistance by promoting ROS-mediated apoptosis in drug-resistant lung cancer cells, supporting its potential as an adjuvant strategy to sensitize resistant tumors[72]CWB-20145 (phenylboronic acid nitrogen mustard)Preclinical evidenceTNBCUndergoes ROS-activated DNA interstrand cross-linking, selectively inducing cytotoxicity in oxidative environmentsPro-oxidantPromotes selective tumor cell death with improved safety and efficacy compared to conventional alkylating agents[197]Elesclomol-Cu + cinnamaldehyde nanoplatformPreclinical evidenceBreast cancerGenerates ROS through copper-mediated redox cycling and cinnamaldehyde-induced oxidative stress, forming a positive feedback loop that amplifies ROS productionPro-oxidantEnhances tumor accumulation, induces immunogenic cell death, and activates antitumor immune responses[198]EmpagliflozinClinical testing (approved for other diseases)Breast cancerInduces ferroptosis by modulating miR-128-3p signaling, reducing GPX4 and GSH levels, and increasing intracellular ROS and iron-dependent lipid peroxidationPro-oxidantReduces metastatic burden and sensitizes resistant cancer cells to ferroptosis, supporting its role as an adjunct therapy[168]hsa-miR-27a overexpressionPreclinical evidenceBreast cancerDownregulates ROS defense genes and impairs GSH synthesis, leading to increased intracellular ROS levelsPro-oxidantSensitizes cancer cells to chemotherapy and reduces CSC-associated resistance mechanisms[187]Ir2 (Iridium(III) complex with α-tocopheryl succinate)Preclinical evidenceCervical, lung, and breast cancerInduces excessive ROS production, causing oxidative stress-mediated DNA damage, mitochondrial depolarization, and dual inhibition of glycolysis and OXPHOSPro-oxidantExhibits potent and selective anticancer activity through mitochondrial targeting, inducing apoptosis and autophagy[199]LysionotinPreclinical evidenceColorectal cancerInduces ferroptosis by promoting NRF2 degradation, leading to ROS accumulation and lipid peroxidationPro-oxidantSuppresses tumor growth and metastatic behavior without significant toxicity[55]Trametinib + OlaparibClinical testing (both approved therapies)Malignant pleural mesotheliomaDisrupts redox homeostasis by combining MAPK pathway inhibition with impaired DNA repair, leading to metabolic and oxidative stress-induced apoptosisPro-oxidantDemonstrates strong synergistic antitumor activity and offers a promising strategy for difficult-to-treat cancers[85]Methionine sulfoxide reductase A modulationPreclinical evidencePancreatic ductal adenocarcinomaRegulates methionine oxidation on PKM2, altering redox signaling and metabolic pathwaysPro-oxidantSuppresses metastasis through redox-dependent metabolic reprogramming[140]miR-210 overexpressionPreclinical evidenceColorectal cancerInduces ROS generation and activates oxidative stress-dependent apoptotic pathwaysPro-oxidantPromotes apoptosis and inhibits proliferation, representing a potential therapeutic target[166]miR-21-3p-loaded gold NPs + Anti-PD-1Preclinical evidence (includes approved therapy)MelanomaPromotes ferroptosis by increasing lipid peroxidation and ROS accumulationPro-oxidantEnhances immunotherapy efficacy by sensitizing tumor cells to ferroptotic cell death[200]Mitochondrial complex I inhibitor (IACS-010759) + LDH inhibitor (NCI-006)Preclinical evidence (includes clinical agent)Colorectal and gastric cancerBlocks metabolic flexibility by inhibiting both glycolysis and OXPHOS, leading to energy depletion and increased ROS levelsPro-oxidantProduces synergistic antitumor effects through metabolic disruption and redox imbalance[201]NF-κB inhibitors (QNZ and BMS-345541)Preclinical evidenceLung cancerReduce expression of antioxidant transporters, decreasing GSH synthesis and increasing ROS levelsPro-oxidantSensitizes cancer cells to ferroptosis and enhances vulnerability to oxidative stress-based therapies[202]NIR-715 PDTPreclinical evidenceEsophageal and breast cancer, melanomaGenerates ROS upon light activation, inducing endoplasmic reticulum stress and ICDPro-oxidantEnhances antitumor immunity and suppresses tumor growth with low toxicity[203]NOX1 inhibitor (GKT771)Preclinical evidenceColorectal cancerInhibits NOX1-mediated ROS production associated with tumor proliferation and angiogenesisPro-oxidantReduces tumor growth and improves response to immune checkpoint blockade[102]Methyl 3-(4-nitrophenyl) propiolatePreclinical evidenceMultiple cancer typesInduces ROS production via cytochrome P450 activity, leading to oxidative stress and mitochondrial apoptosisPro-oxidantSelectively targets cancer cells with high basal ROS and disrupts oncogenic signaling pathways[11]Olaparib + SPHK1 inhibitor (PF-543)Preclinical evidence (includes approved therapy)Ovarian cancerInhibits SPHK1 signaling, reducing NRF2 activity and promoting ferroptosis through increased oxidative stressPro-oxidantEnhances treatment sensitivity and provides a strategy to overcome resistance[56]PEGylated mesoporous cerium oxide-coated upconversion NPsPreclinical evidenceCervical carcinomaDecomposes H2O2 to generate O2 and enhances ROS-mediated cytotoxicity during photodynamic therapy, thereby alleviating hypoxiaPro-oxidantImproves PDT efficacy, supports synergistic chemotherapy, and inhibits tumor growth with low systemic toxicity[204]Perfluorohexane-Ce6 liposome O2 nanodroplets + sonodynamic therapy + Anti-PD-1Preclinical evidence (includes approved therapy)Colorectal cancerGenerates ROS through sonosensitizer activation, inducing oxidative stress, apoptosis, and ICDPro-oxidantEnhances immunotherapy by improving immune activation, reducing regulatory T-cells, and promoting long-term immune memory[205]Poly(lactic-co-glycolic acid)-encapsulated MnO2 NPsPreclinical evidenceBone metastasisReact with H2O2 to generate oxygen, reducing hypoxia-associated signaling and remodeling the oxidative TMEPro-oxidantReverses hypoxia, improves antitumor immunity, decreases Tregs, and prolongs survival[206]RSL3 + ErastinPreclinical evidenceOsteosarcomaBlocks cystine import and GPX4 activity, promoting accumulation of lipid radicals and ferroptosisPro-oxidantSelectively eliminates osteosarcoma cells and undifferentiated stem cells while sparing normal bone cells[207]SR18292Preclinical evidenceMultiple myelomaInhibits PPAR-γ coactivator-1-alpha (PGC-1α)-driven OXPHOS, inducing ROS accumulation, energy depletion, and apoptosisPro-oxidantSuppresses tumor growth and promotes apoptosis with minimal toxicity[175]Wireless implantable electrochemical drug delivery system with polypyrrole NPsPreclinical evidenceMultiple cancer typesEnables electrochemical redox-controlled drug release through oxidation-reduction reactions in polypyrrole NPsPro-oxidantAllows precise and programmable drug release, supporting targeted therapy with reduced systemic toxicity[208]AcriflavinePreclinical evidenceGlioblastoma stem-like cellsInhibits HIF-1 dimerization, reducing RAGE expression and hypoxia-associated inflammatory signalingAntioxidantInterferes with hypoxia- and inflammation-driven glioblastoma progression and supports targeting of stem-like tumor cells[209]Berberine, anastrozole, or pranoprofenPreclinical evidence (includes approved therapy)Colitis-associated colon cancerPromote Dicer expression, reducing oxidative stress-induced DNA damage and inflammatory IL-6 signalingAntioxidantAlleviate inflammation and suppress inflammation-driven carcinogenesis, supporting preventive and therapeutic potential[32]DaphnetinPreclinical evidenceLung adenocarcinomaInhibits MAPK, STAT3, and NF-κB signaling while reducing oxidative stress-associated signaling activityAntioxidantSuppresses proliferation, migration, invasion, and clonogenic growth, supporting its potential as a therapeutic agent[81]FluorofenidoneClinical testingPulmonary fibrosisReduces oxidative stress by increasing SOD activity and decreasing lipid peroxidation, while inhibiting PI3K/Akt/mTOR signaling and promoting autophagyAntioxidantAlleviates fibrosis, inflammation, EMT, and oxidative damage, supporting therapeutic potential in fibrotic disease[49]IDO1 inhibitor (Epacadostat)Clinical testingLung squamous cell carcinomaBlocks IDO1 signaling, reducing kynurenine production and associated ROS-mediated immunosuppressionAntioxidantPartially restores T-cell function by limiting metabolic and oxidative immune evasion[38]lncRNA HITTPreclinical evidenceColon, cervical, and lung cancerInhibits PKM2 oligomerization, reducing glycolysis and lactate production and modulating macrophage polarization within the TMEAntioxidantSuppresses tumor growth, sensitizes cancer cells to nutrient stress, and promotes antitumor immune responses by shifting macrophages toward an M1 phenotype[210]Lonidamine + 5-(tetradecyloxy)-2-furancarboxylic acidPreclinical evidence (includes approved therapy)Ovarian and colon cancer CSCsSimultaneously inhibits glycolysis and lipolysis, disrupting metabolic pathways associated with oxidative stress adaptation and CSC survivalAntioxidantReduces CSC viability more effectively than either agent alone and supports metabolic targeting of resistant cell populations[211]MetforminApproved therapy and clinical testingTNBCInhibits mitochondrial respiratory complex I and activates AMPK, reducing ATP production and imposing energetic and redox stress on tumor cellsAntioxidantEnhances chemotherapy efficacy, may reduce required drug doses, and exerts additional immunostimulatory effects[212]MitoQClinical testing (approved as a dietary supplement)Breast cancerReduces mitochondrial superoxide production and limits oxidative damage associated with metastatic progressionAntioxidantShows preventive potential against cancer recurrence and metastasis through mitochondrial redox control[213]MnTE-2-PyP5+ or DPIPreclinical evidenceBreast cancerScavenge ROS or inhibit NOX activity, reducing ROS-dependent STAT3 activation required for M2 macrophage polarizationAntioxidantSuppress protumoral macrophage function and may enhance antitumor immunity when combined with immunotherapy[214]mTOR inhibitor (Rapamycin)Approved therapy and clinical testingLung adenocarcinomaInhibits mTOR signaling, reversing Warburg-like metabolic reprogramming and helping normalize redox balanceAntioxidantReduces stemness, invasiveness, viability, and colony formation, supporting therapeutic relevance in metabolically active tumors[185]NOX2 inhibitor (GSK-2795039)Preclinical evidenceLung squamous cell carcinoma and other solid tumorsInhibits NOX2-mediated ROS production in MDSCs, reducing oxidative immunosuppression in the TMEAntioxidantRestores CD8+ T-cell proliferation and supports immune reactivation in solid tumors[38] Overview of therapeutic strategies targeting oxidative stress and redox homeostasis in cancer, including pro-oxidant and antioxidant strategies, their molecular mechanisms, and clinical relevance. Pro-oxidant approaches: exploiting redox vulnerabilities The ultimate objective of pro-oxidant therapy is to overwhelm the redox capacity of cancer cells, leading to an irreversible accumulation of ROS. While numerous chemotherapeutic drugs and radiotherapy increase ROS and trigger apoptosis through mechanisms such as mitochondrial dysfunction and DNA damage [190,196];, targeted approaches are beginning to effectively deplete cellular antioxidants and increase ROS production. Auranofin-induced ROS-dependent death in TNBC [196] and the inhibition of GPX4 by RAS-selective lethal 3 (RSL3) inhibitor have led to ferroptotic death in osteosarcoma by promoting lipid peroxidation [207,215]. Combination therapy approaches are particularly promising for overcoming resistance and increasing efficacy; the depletion of intracellular GSH or activation of thioredoxin inhibitors alongside a pro-oxidant agent (pharmacological ascorbate, menadione or arsenic trioxide) can synergistically increase ROS to levels incompatible with life, causing the death of cancer cells in breast cancer and B-cell malignancies [216]. Furthermore, the enhancement of olaparib-induced cell death through suppression of the NF-κB/NRF2 pathway by inhibition of SPHK1 promotes ROS-dependent ferroptosis [56]. Advanced nanotechnology based approaches are further utilizing this redox imbalance of cancer cells for tumor-targeted delivery and the enhancement of therapeutic effects. Nanoparticle (NP)-mediated delivery of targeted pro-oxidant therapies, such as mitochondrial-targeted nanoagents and ROS-producing NPs, promote enhanced ROS generation leading to an “ROS overload” intercellular context that primarily affects the higher baseline ROS of malignant cells [2,3,217]. Combination therapy approaches are also being used in cancer therapeutics, and particularly those combining ROS-generating NPs within the TME with standard treatments such as photodynamic therapy (PDT). For example, Ce6-lipid and Fe3O4 NPs promote Fenton reactions and lipid peroxidation, significantly enhancing PDT efficacy while overcoming limitations associated with oxygen availability and light penetration [218]. Furthermore, antibiotics such as linezolid can be used as mitochondrial-targeting strategies to induce mitochondrial dysfunction in CSCs and chemoresistant populations by disrupting metabolism, increasing ROS, and reducing tumor growth in breast cancer [219]. Similarly, antimycin A inhibits mitochondrial complex III, resulting in mtROS production, DNA damage, and PARP activation, ultimately leading to apoptosis in a redox-dependent manner, as these effects can be reversed by antioxidants [220]. Taken together, these findings support mitochondrial-targeting antibiotic-like approaches as promising strategies to exploit redox vulnerabilities in cancer cells [212]. The main limitations in the use of pro-oxidant agents are the heterogeneity of tumor cells within the tumor, increased capacity for antioxidant buffering or development of adaptation within redox states of malignant cells and metabolic plasticity which influences oxygen delivery. These findings highlight that pro-oxidant therapy is inherently context-dependent and requires biomarker-guided application [189,191,193]. Antioxidant approaches: the therapeutic paradox While not initially intended for anti-cancer treatments, antioxidants are commonly used to reduce the toxicity of chemotherapies and radiation, which is also beneficial in tumor cells due to mitigating damage from ROS [202]. However, this presents a contradiction, as it allows the tumor to survive at the expense of the therapeutic effects, since it is not eliminated by oxidative cell death pathways -such as those involving DNA damage- and also due to the reinforcement of antioxidant defenses, such as the NRF2/KEAP1 pathway, which is responsible for enhancing tumor survival, tumor growth, and metastasis by upregulating various oncogenic factors [189,221]. This is the antioxidant paradox, which runs counter to the concept of antioxidant supplementation in cancer therapy [222]. Mitochondria-targeted antioxidants such as MitoQ have been shown to promote survival and inhibit tumor growth, recurrence and metastasis by limiting the activation of CSCs by reducing mtROS [213]. When delivered appropriately at certain times in treatment, redox modulation may indeed operate to suppress toxicity, or by the precise administration of targeted compounds, promote survival within normal tissues without detriment to anti-cancer therapeutics [213,222,223]. These findings show that antioxidant approaches must be temporally and contextually optimized, to avoid protecting tumor cells while simultaneously preserving normal tissue function. Redox modulation of the TME and immune response Therapeutic approaches targeting redox also have important consequences for the TME, where ROS play key roles in regulating the immune system. Elevated ROS have a potent immunosuppressive effect by negatively impacting T-cell activation and by suppressing immune regulatory effects mediated by macrophages and MDSCs [38,112,214]. The ability to negatively impact ROS in these cells can activate an anti-tumor immune response. The inhibition of ROS in MDSCs and tumor-associated macrophages promotes the proliferation of CD8+ T-cells and reduces immunosuppression, thereby potentiating anti-tumor effects with the administration of immune checkpoint therapy [38,214]. The control of redox by other agents such as NOX inhibitors or by the administration of antioxidant agents may reverse the immunosuppressive profile of macrophages and downregulate the expression of PD-L2, another crucial molecule involved in immune suppression [214]. NP technology designed to utilize the redox balance in the TME to positively affect immune activation and the anti-tumor immune response has now been introduced. The use of perfluorocarbon-based nanodrugs improves cancer immunotherapy by promoting ROS generation, relieving tumor hypoxia, inducing tumor apoptosis and activating DCs and T-cells after insufficient radiofrequency ablation [205]. Similarly, manganese dioxide NPs with tunable O2 released for sustained reoxygenation of bone metastases have been shown to decrease hypoxia and immunosuppression, enhance NK cytotoxicity, lower Treg cells, and prolong survival in mouse models [206]. Such therapies represent a shift toward TME-specific interventions, in which the manipulation of redox can target both the tumor itself and modulate an anti-tumor immune response. Metabolic redox sensitization strategies Metabolic pathways provide prime therapeutic targets given their intrinsic relationship to redox homeostasis. Inhibition of either glycolysis or OXPHOS leads to an energy deficit and redox imbalance that sensitizes cancer cells to redox stress. The glycolytic inhibitors 2-deoxy-d-glucose (2-DG), lonidamine and 3-bromopyruvate inhibit ATP production and enhance ROS production, resulting in chemosensitization and targeting of CSCs [194,195,211]. In pancreatic cancer, 2-DG enhances chemosensitivity to gemcitabine and targets the CSC population [194], while in ovarian cancer lonidamine inhibits stemness in models of tumor recurrence [211]. Other than direct glycolytic inhibition, some therapy-resistant tumors rely on monocarboxylate transporter 1 (MCT1)-dependent lactate uptake to sustain energy production under metabolic stress, suggesting that lactate transport may be an additional target for disrupting metabolic adaptation [224]. The antibiotic antimycin A inhibits mitochondrial complex III, resulting in mtROS production, which induces DNA damage and activates PARP, leading to DNA damage-induced apoptosis (these effects can be reversed by antioxidants, and thus confirm the redox dependency) [220]. Metformin has also been shown to target OXPHOS-dependent tumors due to its ability to inhibit ATP production [212]. A combined targeting of glycolysis and OXPHOS will limit metabolic compensation and therefore exacerbate redox stress [199,201]. Further targeting of the metabolic regulators pyruvate kinase M2 (PKM2) or HIF-1α enhances susceptibility to oxidative stress [209,210]. This shows that metabolic targeting is fundamentally a redox intervention. Translational challenges Cancer's adaptive capability can compromise redox-targeted therapy efficacy as it allows tumors to reprogram their metabolic and antioxidant pathways, and to promote survival through alternative energy pathways, autophagic activation, increased fatty acid oxidation and induction of other survival pathways [264,265]. Furthermore, TME heterogeneity complicates treatment, as hypoxic niches alter redox signaling and metabolic states through HIF-1α-dependent pathways [193]. These challenges highlight the need for personalized therapeutic approaches that account for tumor-specific redox states, metabolic configuration, and adaptive capacity. Meanwhile, the development of appropriate redox biomarkers is crucial for efficient clinical translation. Baseline ROS levels can influence sensitivity to ROS-based therapies, and may serve as indicators of therapeutic response [217]. Chronotherapeutic strategies based on circadian oscillations in metabolism might further enhance therapeutic efficacy [179]. Taken together, more dynamic and adaptive treatment strategies whereby redox balance can be adjusted in response to changing tumor status should be considered. Conceptualization of Adaptive Directed Redox Therapy (ADRT): a dynamic framework for precision redox oncology There are two fundamental limitations in each approach; prolonged pro-oxidative stress may induce healthy tissue damage, whereas an antioxidant strategy may make the tumor resistant to adaptation [225,226]. Recent studies have shown that resistance emerges through a series of changes in cellular states, accompanied by a gradual improvement in the cells’ ability to survive and adapt, in a process known as the “resistance continuum” [227,228]. To address this paradox, we propose a dynamically controlled cellular redox regulation framework, or an “Adaptive Directed Redox Therapy” (ADRT) (Fig. 3). While conventional redox modulation strategies aim to impose a statically higher or lower level of ROS, ADRT is based on the principle that tumor vulnerability lies in the instability of redox balance rather than its overall level. The model introduces a temporally controlled and context-dependent modulation of oxidative stress, aiming to exploit the limited adaptive capacity of tumor cells compared to normal tissues.Fig. 3ADRT: a closed-loop redox detection and response system. ADRT is proposed as a dynamic, feedback-controlled therapeutic platform that continuously monitors and adapts to the oxidative state of both the TME and the surrounding healthy tissues. While healthy cells maintain antioxidant defenses and homeostasis of ROS, cancer cells operate near their oxidative threshold and exhibit persistently elevated ROS levels. The NP-based platform incorporates redox-sensitive sensor elements capable of detecting local oxidative levels and translating them into controlled therapeutic responses. Upon detection of ROS, an integrated decision module regulates the NPs' activity, activating either pro-oxidant domains (which induce apoptosis or ferroptosis in cancer cells) or antioxidant domains (which protect normal cells from oxidative damage) depending on ROS concentrations, thereby maintaining oxidative balance. A biosensor microchip provides real-time monitoring of the redox environment and establishes a closed-loop feedback system that continuously adjusts nanoparticle activity based on local ROS levels. Beyond direct redox modulation, ADRT reshapes the immune landscape, promoting antitumor immunity while reducing tumor-promoting inflammation, and can be synchronized with standard therapies to maximize cancer cell death while preserving the integrity and function of normal tissues. ADRT: a closed-loop redox detection and response system. ADRT is proposed as a dynamic, feedback-controlled therapeutic platform that continuously monitors and adapts to the oxidative state of both the TME and the surrounding healthy tissues. While healthy cells maintain antioxidant defenses and homeostasis of ROS, cancer cells operate near their oxidative threshold and exhibit persistently elevated ROS levels. The NP-based platform incorporates redox-sensitive sensor elements capable of detecting local oxidative levels and translating them into controlled therapeutic responses. Upon detection of ROS, an integrated decision module regulates the NPs' activity, activating either pro-oxidant domains (which induce apoptosis or ferroptosis in cancer cells) or antioxidant domains (which protect normal cells from oxidative damage) depending on ROS concentrations, thereby maintaining oxidative balance. A biosensor microchip provides real-time monitoring of the redox environment and establishes a closed-loop feedback system that continuously adjusts nanoparticle activity based on local ROS levels. Beyond direct redox modulation, ADRT reshapes the immune landscape, promoting antitumor immunity while reducing tumor-promoting inflammation, and can be synchronized with standard therapies to maximize cancer cell death while preserving the integrity and function of normal tissues. At its conceptual core, the ADRT framework conceptualizes the process of redox modulation as a rhythmic dynamic process in which phases of oxidative stress are alternated with phases of recovery, continuously de-stabilizing the tumor cells. Since tumor cells exist at the threshold of oxidative damage and unlike healthy cells have a limited capacity for long-term restoration (which have a sufficient antioxidant capacity in terms of SOD, catalase and GPX [229,230]), repetitive dynamic oxidative bursts might deplete crucial antioxidant reserves such as GSH, ATP or NADH, and lead to terminal oxidative collapse [231,232]. This theoretical proposed model for ADRT brings together various redox-dependent processes which are individually confirmed by numerous experiments. Oxidative stress has been shown to inhibit antioxidant systems such as GPX4 and amino acid transport system xc (xCT)/SLC7A11, which sensitizes cells to lipid peroxidation and oxidative damage that ultimately can cause cell death [233]. The long-term repeated oscillation of oxidative stress may perturb the NRF2/KEAP1-mediated antioxidant response and affect tumor cells' adaptive mechanism [234]. Furthermore, a cell's redox regulation ability is heavily dependent on its metabolism, the cell's bioenergetics and redox potential, which influence its adaptive responses to oxidative stress [235]. Oxidative stress also regulates gene transcription via epigenetic modification (such as histone deacetylase activity), causes mutations, and modulates tumor cell survival and apoptotic programs [236]. Moreover, a temporary increase in oxidative stress may result in immunogenic cell death (ICD) by releasing damage-associated molecular pattern molecules (DAMPs) such as ATP, high-mobility group box-1 and calreticulin, which enhance the immune cell response, including activating DCs and eliciting T-cell responses [237]. Taken together, these factors show that temporal modulation of oxidative stress can act on the metabolism, survival and immunogenic ability of the tumor cell simultaneously, while this combined effect with an integrated approach has never been reported. The ADRT model should involve a closed-loop therapeutic system, and use nanocarriers with a capacity for real time monitoring of ROS levels, followed by a feedback-controlled release of therapeutic agents. While some components of this system are currently available, the technologies for the practical implementation of the entire system have yet to be developed. Nanocarriers which are responsive to local ROS levels or other tumor specific signals such as pH and hypoxia are the focus of extensive investigation, and are one feasible approach for the proposed ADRT model [238]. ROS-responsive nanoenzymes and drug carriers can modulate the oxidative potential in a context-specific manner. There have been many developments in redox sensing technologies, which include electrochemical and optical biosensors that are able to detect species including H2O2 [239]. Wireless communication and feedback systems are also being considered [198,208,240]. However, continuous, reliable and spatial resolved redox status monitoring is not yet possible. No approved redox sensor has yet been developed for clinical use. The ADRT concept could potentially increase the effectiveness of current therapeutic methods by regulating and synchronizing the redox states of tumor cells with existing therapeutic schedules. For instance, coordinating oxidative peaks with the administrations of chemotherapeutics or radiation would potentially enhance their cytotoxic effect on tumors, while mitigating damage to normal tissue. A dynamic increase in ROS levels has also shown to improve the immunogenicity of tumor cells, enhance AP, thereby facilitating T-cell responses and potentially potentiating the efficacy of immune checkpoint inhibitors including anti-PD-1 and anti-cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) therapies [197,203,241]. These multi-drug and multi-mode combination treatment strategies are supported by pre-clinical evidence for the effect of redox regulation on tumor immunity, metabolism and therapy response. Precise optimization of the timing, extent and sequence of different treatment options are necessary for this multi-modality treatment. The idea of a “redox thermostat”, i.e., a device that could dynamically switch between pro-oxidative and anti-oxidative drugs based on real time feedback, should only be regarded as a promising prospect in technological development before it is considered a clinical application that need to be fully investigated at the experimental level. Another issue regarding the model is the multicellular complexity of the TME, which underscores that ADRT should be regarded as a response-guided approach, not as an attempt to define a universal ROS set point for cancer cells. Depending on the nature, intensity, duration and source of the oxidative stress, tumor cells, CD8+ T-cells, Tregs, NK cells, DCs, MDSCs and TAMs vary in their basal redox state and oxidative stress responses [200,204]. Thus, the appropriate therapeutic parameter in ADRT is a patient- and tumor-specific, multicellular therapeutic window. This window should be defined as the degree and duration of redox disruption sufficient to induce immunogenic cell death, while maintaining the normal redox-dependent functions of antitumor immune cells. By design, ADRT must preserve the functions of CD8+ T cells, NK cells, and dendritic cells (DCs), and block the immunosuppressive mechanisms of MDSCs and tumor-associated macrophages (TAMs). Consequently, while transient oxidative stress can induce apoptosis (AP) and lymphocyte activation, sustained oxidative stress is associated with T-cell exhaustion, NK-cell dysfunction, and myeloid-cell-dependent immunosuppression [[242], [243], [244]]. Due to dynamic changes in the tumor microenvironment (TME) during treatment, ADRT can make adaptive adjustments to the intensity, duration, and spatial distribution of treatment, thereby offering a verifiable approach to exploit the redox susceptibility of cancer cells without compromising antitumor immunity [204,[243], [244], [245], [246]]. ADRT will undeniably face significant translational barriers that need to be specifically adjusted before it is implemented in clinical practice. Firstly, tumor heterogeneity, both spatial and temporal, makes it difficult to establish consistent redox targets because ROS levels are heterogeneous and variable throughout tumor tissue due to variations in hypoxia and metabolism. The lack of suitable redox biomarkers also makes it difficult to determine the optimal therapeutic dose and monitor the response. The in vivo detection and quantification of ROS levels and the relative changes although possible, still remain technically challenging. Furthermore, efficient and precise drug delivery is always a critical challenge in cancer therapy; further progress on the intelligent nanocarrier design is essential for dynamic delivery. Systematic redox modulation could carry the risk of side effects and damage to normal tissues with high metabolic activity, such as normal cells and liver. Overall, reliable evidence for the sensitivity, stability, specificity and biocompatibility of ROS sensors under in vivo conditions is limited and challenging, hindering the precise real time monitoring of tumor redox environment. In overall terms, the ADRT concept could be seen as a prospect for the future which can integrate our current understanding of redox biology with prospective advanced technologies. It reformulates cancer treatment as a dynamic control of tumor redox environment, rather than a static alteration of ROS levels. The proposed model represents a hypothetical extension to experimentally validated redox fluctuations in cancer biology. By emphasizing the transient nature of the manipulation system level analysis and the adaptive treatment approach, ADRT can provide a unified perspective for directing future research in redox oncology. Conclusion Oxidative stress plays a pivotal role in cancer biology, as it could be considered a continuous driver of tumorigenesis and genomic changes. The disruption of the equilibrium between ROS levels and antioxidants promotes tumor initiation, progression and may affect response to therapies. ROS can be defined as signaling regulators that favor cell death or survival, depending on their concentrations and tumor context. Oncogenic pathways such as KRAS, MYC, and PI3K/Akt/mTOR increase ROS levels, whereas cancer cells respond with antioxidants like GSH, GPX4, and the NRF2 pathway activation. This helps tumors survive and resist therapy. Cancer cells can also switch between different metabolic states (OXPHOS and glycolysis) to survive in adverse TME. Emerging combinatorial therapies, such as pro-oxidant agents (RSL3, PDT), antioxidants (GPX4, NOX inhibitors) and glycolysis (2-DG) inhibitors, as well NP delivery, aim to take advantage of these weaknesses in the redox system, and advance toward more personalized and effective cancer treatments. Controlling oxidative stress dynamics would represent a crucial achievement in precision oncology. The proposed ADRT feedback-guided strategy, which alternates oxidative and protective phases to disrupt tumor redox homeostasis while protecting healthy cells, may represent a challenge and opportunity for future cancer therapies. CRediT authorship contribution statement María Escribano-Cebrián: Writing – original draft. Marina Bataller: Writing – review & editing. Marc Biasini: Writing – review & editing. Yoelsis García-Mayea: Writing – review & editing. David Virós: Writing – review & editing. Matilde E. LLeonart: Conceptualization, Investigation, Supervision, Validation, Writing – review & editing. Additional information No additional information. Declaration of competing interests The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.