ROS mostly come from the electron transport chain (ETC) in the mitochondria. There is a lot of superoxide anion made by the ETC. Five protein groups, called complexes I through V, make up the ETC, and they work together to make ATP [35, 36]. During oxidative phosphorylation, superoxide is made when electrons leave complexes I and III. mtROS are primarily generated at complexes I and III of the ETC [37]. Their contributions fluctuate based on factors such as metabolic state, substrate availability, oxygen tension, mitochondrial membrane potential, and the direction of electron flux. Complex I serves as a significant source of superoxide production, particularly under conditions that promote reverse electron transport (RET), such as succinate accumulation and elevated membrane potential or depolarization. Conversely, complex III can also contribute notably to ROS generation in specific physiological and pathological contexts, primarily due to electron leakage. Consequently, the formation of mtROS cannot be solely linked to a particular respiratory complex; rather, it is a dynamic process influenced by various conditions [38–40]. When an electron from NADH is transferred to flavin mononucleotide (FMN) and subsequently overflows to O₂, or when reduced flavin mononucleotide (FMNH₂) contributes an electron to O₂, the hydrophilic region of complex I may produce superoxide anion. Complex III (ubiquinol-cytochrome c oxidoreductase) receives reducing equivalents generated by complexes I and II and employs the Q-cycle mechanism for their degradation [41]. Genetic and acquired abnormalities in the respiratory chain of the mitochondria may cause complex III to overproduce ROS. Therefore, it is yet unknown how complex III might contribute to the generation of gross mtROS in steady-state physiological settings [42].

Mitochondria also generate spatially restricted ROS nanodomains that allow local redox signaling, which can selectively modify nearby proteins without causing excessive oxidative stress. This spatial specificity is tightly associated with mitochondrial dynamics, including a constant cycle of fission and fusion, which regulates organelle morphology, bioenergetic efficiency and ROS distribution [43, 44]. Mitochondria are also important for redox homeostasis, with mitochondrial quality control mechanisms, especially mitophagy, being involved. Selective degradation of ROS producing damaged mitochondria prevents excessive oxidative stress and ensures a functional mitochondrial network. Thus, dysregulation of mitophagy can lead to accumulation of dysfunctional mitochondria, increasing ROS levels, and contributing to tumor progression or cell death in a context-dependent manner [45–47]. ROS signaling is spatially and temporally separated, allowing localized oxidative signaling without macromolecular damage. mtROS, cytosolic ROS, peroxisomal ROS and membrane-associated NOX-derived ROS have different roles depending on their site of formation and antioxidant buffering capacity. Tumor subpopulations are characterized by heterogeneous redox microenvironments in terms of oxygen tension, metabolic gradients and organelle-specific antioxidant systems, impacting therapy resistance, metastatic adaptation and ferroptosis sensitivity [48].

In addition to the ETC, emerging evidence suggests that enzymes associated with the tricarboxylic acid (TCA) cycle such as pyruvate dehydrogenase and α-ketoglutarate dehydrogenase are major sources of mtROS. Interestingly, the generation of mtROS is strongly dependent on the availability of metabolic substrates, and different fuels (e.g., glucose, glutamine, fatty acids) have differential effects on the redox balance, due to their effects on NADH/FADH2 generation and electron flux [49, 50]. Cancer cells produce increased ROS due to metabolic changes such the Warburg effect and mitochondrial issues. Mutations in mitochondrial DNA (mtDNA) increase genomic instability and ROS generation [51, 52]. Despite being encased in proteins that are just as protective as those found in nuclear chromatin, mtDNA is especially vulnerable to ROS produced by the respiratory chain. In addition to the aging process and illnesses linked to advanced age, its mutations can cause several disorders. Research points to a connection between mitochondrial genetic malfunction and aging [53, 54]. When nucleosomes that structure DNA within chromosomes are disrupted, the chromatin ceases to be compacted and coiled, resulting in the fragmentation of DNA. Chromatin is crucial for the regulation of gene transcription; thus, alterations in its functional qualities may result in errors that induce mutations [55, 56]. For example, when ROS reacts with nitrogenous bases and deoxyribose sugar in DNA, it causes a lot of oxidative damage, which may lead to mutations, cancer, cell death, necrosis, and genetic diseases [57].

The NADPH oxidase (NOX) enzyme family is crucial for the active generation of ROS as signaling mediators. NOX isoforms exhibit distinct regulatory mechanisms, tissue distributions, and biological functions in cancer. NOX1, NOX2, and NOX5 primarily produce superoxide, while NOX4 predominantly generates hydrogen peroxide and is unique in being constitutively active, not requiring stimulus-dependent assembly. Beyond serving as a general source of ROS, NOX4 plays a significant role in regulating fibrosis, epithelial-mesenchymal transition (EMT), remodeling of the extracellular matrix (ECM), and the activation of the TME stroma. Increased NOX4 expression has been strongly associated with the activation of cancer-associated fibroblasts (CAFs), transforming growth factor-β (TGF-β) signaling, tumor invasion, and metastatic progression [58–60].

In addition to ROS produced by tumor cells themselves, ROS generated by immune cells such as macrophages, neutrophils, myeloid-derived suppressor cells (MDSCs), and other stromal populations are substantial contributors to redox remodeling within the TME [61]. These ROS are involved in regulating cytokine signaling, immune suppression, angiogenesis, and matrix remodeling, thereby influencing tumor progression and therapeutic responses. In cancer, inflammatory signaling and oxidative stress are interconnected through various factors, including cytokines, growth factors, Toll-like receptor agonists, leptin, and hormones like angiotensin II, which stimulate NOX activity [62, 63].

ROS buildup is also influenced by peroxisomes, cytochrome P-450 enzymes, xanthine oxidase (XO), and endoplasmic reticulum (ER) stress responses. Exposure to environmental carcinogens and oncogenic signaling (such as RAS and MYC) increase oxidative stress in the TME [64, 65]. Numerous studies have demonstrated that ROS, which are essential for controlling tumor genesis, development, and advancement, are more prevalent in cancer cells than in normal cells. It’s interesting to note that ROS is said to have two roles in the formation and management of tumors [66, 67]. On the one hand, by controlling metabolism, proliferation, and angiogenesis, elevated ROS levels provide larger survival advantages. However, ROS are also an effective tool for inhibiting or eliminating cancer cells, which is important during radiation and chemotherapy. By controlling functional proteins such as GPXs, SLC7A11, and PARP-1, ROS-dependent apoptosis, ferroptosis, and autophagy-mediated cell death have been shown to inhibit tumor growth [23, 68, 69]. XO, which catalyzes the oxidative hydroxylation of hypoxanthine to xanthine and xanthine to uric acid, is responsible for the last two stages of purine metabolism in humans. Aging is linked to oxidative stress, immunosenescence, and inflammation, and XO is elevated throughout this time. Endothelial dysfunction, a crucial process that causes a number of age-related illnesses, including cardiovascular diseases (CVDs), appears to be indicated by XO [70–72]. The ER oxirreductin-1 (ERO1) protein is the main source of cellular ROS in eukaryotic cells, which are produced by the ER. A healthy quantity of ROS is necessary for the formation of disulfide bonds, a crucial step in the machinery that folds proteins [73]. ER Ca²⁺ releases can also generate ROS and Redox-sensing thiol groups regulate the ER’s RyR, a Ca²⁺ channel. The RyR can release calcium from the ER by oxidizing its thiol sites [74]. NOX hyperactivity and mitochondrial malfunction, two major biological sources of ROS, are intimately linked to dysregulated calcium transmission. Additionally, the ER expresses the NOX isoform NOX4, which adds to the oxidative stress vicious loop within the ER [42, 75].

Enzymatic antioxidants like SOD, catalase (CAT), GSH peroxidase (GPx), and peroxiredoxins, as well as non-enzymatic compounds like GSH and vitamins C and E, help cells prevent the buildup of ROS. To preserve redox equilibrium, cancer cells frequently upregulate antioxidant systems through transcription factors like Nrf2 [76, 77]. While it is well known that Nrf2 activation can shield cells from a variety of stresses and toxins, uncontrolled Nrf2 activation has been linked to a number of cancer forms. Cancer cell lines and tumor samples from lung, breast, esophageal, endometrial, and prostate malignancies showed constitutively high Nrf2 levels [78–80]. In fact, there was a negative correlation between Nrf2 levels in the tumor and the prognosis of cancer patients. Increased amounts of Nrf2-target antioxidant proteins, which combat oxidative stress, are responsible for the positive impact of Nrf2 overexpression on tumor growth and survival [81–83]. The main class of antioxidant enzymes that break down superoxide anions is constitutive SOD. Dismutase superoxide is quickly broken down by all types of SOD into the more stable ROS (H₂O₂), which is subsequently transformed into water and oxygen by either CAT or GPx. Although it can also be found in the cytosol or mitochondria, the highly effective enzyme CAT is mostly expressed in peroxisomes. Additionally, GPx uses GSH to catalyze the reduction of H₂O₂ to water [84, 85]. All cell compartments contain large amounts of GSH, and the primary thiol-based defense mechanism against oxidative and electrophilic stress signals in the cell is the ratio of its oxidized/reduced form (GSH/GSSG). GSSG reduces H₂O₂ to H₂O and O₂ by giving it an electron. GSH is a cofactor for a number of enzymes, including GPx, which effectively inhibits oxidative damage and, in conjunction with glutaredoxin (Grx), modifies the redox state of proteins by reversible S-glutathionylation [86, 87]. Vitamin C (ascorbic acid) is a powerful water-soluble antioxidant that primarily targets oxygen-centered reactive species, such as superoxide, hydroxyl, and peroxyl radicals. This action helps reduce oxidative damage and lipid peroxidation [88].

Cancer cells are dependent on the antioxidant systems to keep ROS levels below cytotoxic levels, while maintaining proliferative signaling that is redox-dependent. This is usually termed as “redox addiction”. This increased reliance on GSH, Trx, Nrf2-mediated transcriptional programs and NADPH regeneration pathways allows malignant cells to withstand chronic oxidative stress, hypoxia, metabolic reprogramming and therapeutic insult [89]. But this adaptive antioxidant buffering also provides therapeutic vulnerabilities. Disruption of critical redox defense pathways in many tumors leads to synthetic lethal interactions by selectively overloading the oxidative stress tolerance of cancer cells while sparing normal tissues. Antioxidant systems in cancer are thus not only protective mechanisms but also exploitable metabolic dependencies and key determinants of therapy resistance [90]. Figure 1 shows the main sources and regulation mechanisms of ROS in cancer.

ROS can cause oxidative DNA damage, including base modifications (e.g., 8-oxo-dG), strand breaks, and mutations that activate oncogenes or inactive tumor suppressors such as TP53. Chronic oxidative stress thus acts as a mutagenic driver of tumorigenesis [91, 92]. Cancer cells are continuously exposed to oxidative DNA damage and are highly dependent on DNA damage response (DDR) pathways for survival. Oxidative lesions such as 8-oxoG are primarily repaired via the base excision repair (BER) pathway through glycosylases including OGG1 and MUTYH, preventing mutagenic G:C→T:A transversions. Persistent oxidative stress induces replication stress characterized by fork stalling, nucleotide pool oxidation, and ssDNA accumulation, activating ATM/ATR signaling and checkpoint responses. Consequently, many tumors rely on PARP-mediated repair and ATR/CHK1 pathways, creating redox-associated synthetic vulnerabilities that can be therapeutically exploited by targeting BER, ATM, or ATR [93, 94]. Cells have many different ways to correct DNA damage. The BER fixes single-strand breaks (SSBs) and oxidized bases, while nucleotide excision repair (NER) resolves worse problems. But excessive ROS can overload these repair pathways or directly harm their function by oxidizing repair enzymes, cofactors, or proteins that are important for finding DNA damage [95]. ROS can impair genomic integrity through epigenetic processes. Oxidative stress can alter DNA methyltransferases and histone-modifying enzymes, causing abnormal methylation patterns and histone acetylation and methylation [96]. These modifications can turn off tumor suppressor genes, activate oncogenes, and modify chromatin structure, making certain regions more vulnerable to DNA damage or copying errors. The epigenetic and chromatin-remodeling effects of ROS contribute to increased genomic instability, resulting in mutagenesis during tumor growth [97].

Besides DNA damage, ROS can affect epigenetic remodeling by the redox-sensitive regulation of chromatin-modifying enzymes. Oxidative stress modulates TET dioxygenases, histone demethylases, DNA methyltransferases, and histone deacetylases, therefore controlling DNA methylation, histone acetylation, and chromatin accessibility. ROS-induced metabolic rewiring influences epigenetic regulation by changing the availability of cofactors such as α-ketoglutarate, acetyl-CoA, NAD⁺, and S-adenosylmethionine. Redox biology directly regulates cancer epigenetics and phenotypic heterogeneity via promotion of transcriptional plasticity, epithelial–mesenchymal transition, stemness, immune evasion, and treatment resistance [98].

ROS affects redox-sensitive signaling pathways such as MAPK/ERK, PI3K/AKT, and NF-κB, promoting cell growth and survival. Low to moderate ROS levels activate these cascades, while high ROS cause apoptosis or necroptosis. ROS modulate several cell growth and survival signaling pathways as intracellular second messengers. ROS reversibly oxidize cysteine residues on regulatory proteins, including phosphatases and kinases, altering their activity and downstream signaling [99–101]. For example, ROS-mediated inactivation of MAPK phosphatases sustains activation of MAPKs, which in turn enhance transcription of cyclins, c-Myc, and other cell cycle regulators. This signaling supports G1/S phase transition and promotes the continued proliferation of tumor cells, even under oxidative stress. Persistent ROS signaling thus provides cancer cells with a growth advantage while subtly promoting genomic instability through replication under oxidative conditions [102, 103]. Redox signaling intersects closely with metabolic reprogramming in cancer cells. ROS stabilize hypoxia-inducible factors (HIFs) under non-hypoxic conditions, driving expression of glycolytic enzymes and glucose transporters, which shifts energy metabolism toward glycolysis, which is a hallmark of rapidly proliferating tumor cells [104]. When Nrf2 is activated at the same time, it also increases the expression of antioxidant and metabolic genes, keeping ROS levels in a range that helps cells survive and providing biosynthetic precursors for cell division. These changes work together to give cancer cells both energy and a way to keep their redox levels stable [26, 105].

ROS also regulate the PI3K/AKT pathway, a central axis controlling cell survival, metabolism, and proliferation. Oxidative modification of PTEN, a tumor suppressor phosphatase, at its catalytic cysteine leads to reversible inactivation, resulting in the accumulation of phosphatidylinositol (3,4,5)-trisphosphate (PIP3) and sustained AKT activation. Activated AKT phosphorylates downstream targets such as mTOR, BAD, and GSK3β, which inhibit apoptosis, promote protein synthesis, and enhance metabolic adaptation. This ROS-mediated PTEN inactivation is exploited by tumor cells to survive in high-ROS microenvironments and under nutrient or oxygen limitation, enabling continued proliferation [42, 106].

ROS stabilize hypoxia-inducible factor 1-alpha (HIF-1α), upregulating vascular endothelial growth factor (VEGF) and promoting angiogenesis. Oxidative stress enhances EMT and matrix metalloproteinase (MMP) activation, facilitating metastasis. These pro-angiogenic factors stimulate endothelial cell proliferation, migration, and tube formation, supporting neovascularization within the TME. Enhanced vascular networks provide oxygen and nutrients to proliferating tumor cells and facilitate tumor growth, while ROS-mediated modulation of VEGF signaling also primes endothelial cells for sprouting and remodeling [107]. ROS directly influence endothelial dynamics by affecting cytoskeletal organization, cell–cell junctions, and interactions with the ECM. Moderate ROS levels enhance endothelial cell motility, promoting sprouting and extension of nascent vessel networks. However, excessive ROS disrupts adherens junctions and tight junction proteins, increasing vascular permeability. This results in abnormal, leaky tumor vessels, elevated interstitial fluid pressure, and irregular blood flow, which together facilitate tumor progression and create heterogeneous oxygenation conditions that promote selection of aggressive cancer phenotypes [108, 109]. On the other hand, oxidative stress contributes to EMT in cancer cells, a critical step in metastasis. ROS regulate EMT transcription factors such as Snail, Twist, and ZEB1, and modulate the expression of cell adhesion molecules like E-cadherin and N-cadherin. This redox-mediated regulation drives the transition of epithelial tumor cells to a mesenchymal phenotype, increasing motility, invasiveness, and the ability to disseminate from the primary tumor site. By promoting EMT, ROS directly enhances the metastatic potential of tumor cells [110, 111].

In addition to directly promoting EMT, ROS actively remodel the TME to promote invasion and metastatic dissemination. Oxidative stress regulates ECM remodeling through activation of MMPs, collagen crosslinking enzymes and integrin-mediated signaling pathways which contribute to tumor cell motility and tissue invasion [112]. CAFs are major players in this redox-remodeled microenvironment, producing ROS and secreting cytokines, growth factors, and ECM components to support EMT, stromal stiffening, angiogenesis, and metastatic progression. ROS also mediates immune suppression in the TME through tumor-associated macrophages, MDSCs, regulatory T cells and checkpoint signaling pathways that facilitate immune escape during metastatic dissemination [113, 114]. Moreover, oxidative signaling participates in the formation of pre-metastatic niches by promoting the release of inflammatory cytokines, alteration of vascular permeability, metabolic adaptation, and recruitment of stromal and immune cell populations that support secondary tumor colonization. These findings emphasize that ROS-driven metastasis is not just tumor cell–intrinsic but instead reflects complex redox crosstalk between malignant cells and the surrounding microenvironment [115].

ROS are a key component of the immunological landscape of the TME. Chronic oxidative stress impacts innate and adaptive immunity and often results in an immunosuppressive environment conducive to tumor development. Tumor-associated macrophages subjected to persistent ROS are often characterized by an M2-like phenotype, including the synthesis of immunosuppressive cytokines and angiogenic and ECM remodeling signals. At the same time, ROS modulate dendritic cell maturation and antigen-presentation capacity, thereby impacting T-cell priming and antitumor immune activation [17, 116]. Oxidative stress induces T-cell dysregulation and exhaustion. Elevated levels of ROS in the TME attenuate T-cell receptor signaling, reduce cytokine secretion, and upregulate the expression of inhibitory checkpoint molecules such as PD-1, TIM-3, and LAG-3. Furthermore, ROS-mediated stabilization of HIF-1α in hypoxia promotes immune escape through up-regulation of PD-L1 expression in tumor and stromal cells. These mechanisms together suppress cytotoxic T-lymphocyte activity and limit efficient antitumor immunity. Importantly, ROS have dual immunological effects depending on concentration, localization, and duration of exposure. Moderate levels of ROS are required for immune cell activation and antimicrobial defense [117, 118].

Chronic oxidative stress leads to chronic inflammation, immune suppression, and tumor tolerance. Such threshold-dependent immunological behavior highlights the need for tightly controlled redox modulation strategies in cancer therapy. Moreover, recent evidence also suggests that ROS-mediated immunogenic cell death (ICD) may enhance therapeutic responsiveness via stimulation of dendritic cell activation, antigen cross-presentation, and recruitment of cytotoxic T cells. Therefore, combining ROS-modulating therapies with immune checkpoint inhibitors is a new approach to turn immunologically “cold” tumors into “hot” tumors that are more amenable to immunotherapy [119]. Figure 2 illustrates that cancer cells depend on a tightly controlled ROS “redox window,” where moderate ROS promote tumor progression, but excessive ROS trigger cell death, creating a therapeutic vulnerability.

Pro-oxidant therapeutic approaches exploit the increased ROS susceptibility inherent to malignant cells by further elevating intracellular oxidative stress beyond cytotoxic thresholds. One of the reasons why this approach has been chosen is the relative insensitivity of cancer cells compared to normal cells to an oxidative insult: Due to the already existing elevated levels of basal ROS and reduced antioxidant reserve capacity, exogenous generation of ROS is more easily triggered to cause apoptosis, ferroptosis, or necroptosis in tumor cells but not in normal tissues [121, 126]. A variety of clinically approved chemotherapeutic agents achieve their cytotoxic effects by generating ROS. The semiquinone and quinone forms of doxorubicin are a common anthracycline antibiotic applied in breast, ovarian, and hematologic malignancies, in which redox cycling of the drug gives rise to the formation of superoxide anion radicals and then hydrogen peroxide (H₂O₂). Nonetheless, doxorubicin-induced cardiotoxicity, explained by excessive ROS accumulation in cardiomyocytes with insufficient antioxidant protection, is an important clinical problem, which is why the development of tumour-selective delivery methods is necessary [127]. Arsenic trioxide (ATO, As₂O₃) which was approved against acute promyelocytic leukemia (APL) acts by dual redox process: On the one hand, it directly inhibits GPx and Trx reductase (TrxR), therefore, preventing the removal of ROS, and on the other, it causes the mitochondrial production of ROS due to interference with the ETC. In APL, ATO is a tumour-killing biological agent that has been shown to achieve complete remission in monotherapy, providing clinical validity for the concept of ROS-mediated tumour killing [120, 128]. Elesclomol (STA-4783) is a small-molecule copper chelator that delivers copper ions to the mitochondrial matrix, where they induce hydroxyl radical generation by a Fenton-like reaction. Preclinical models have demonstrated broad antitumor activity, and a Phase III trial has indicated that it enhances progression-free survival in stage IV melanoma and significantly improves it in patients with normal serum lactate dehydrogenase (LDH), which is strongly associated with greater mitochondrial oxidative metabolism and greater sensitivity to ROS [129]. One of the attractive approaches to drug repurposing is that of disulfiram-copper (DSF-Cu) complexes. The anti-alcohol dependence drug disulfiram forms bis (dithiocarbamate) complexes with copper ions in vivo, thereby inhibiting the 26S proteasome, triggering the p97-NPL4 pathway, and producing intracellular ROS. Several retrospective studies and preliminary clinical trials have reported improved survival rates in cancer patients who used disulfiram throughout chemotherapy. DF-Cu with carboplatin/paclitaxel Phase I/II trial (NCT02671890) is currently ongoing in advanced non-small-cell lung cancer (NSCLC) [130].

New photodynamic therapy (PDT) technologies are a spatially selective form of pro-oxidant modality. PDT includes the use of a tumour-targeting photosensitizer (e.g., photofrin, 5-aminolevulinic acid, temoporfin) and light activation, which delivers energy to molecular oxygen to produce singlet oxygen (¹O₂)—A highly reactive non-radical ROS that causes lipid peroxidation, mitochondrial membrane rupture, and immunogenic tumour cell death. PDT has now been approved for the treatment of oesophageal, lung, and skin malignancies. Current studies are directed toward near-infrared-activable photosensitizers with improved tissue penetration to expand their use to solid tumours [99, 131].

An entirely different and complementary mechanism is the conceptual degradation of the antioxidant defence system, whereby cancer cells rely heavily on it to survive under sustained, constitutive oxidative stress. Adaptively, cancer cells overexpress the GSH system, the Trx system, and the Nrf2-Keap1 transcriptional axis, thereby forming targetable redox vulnerabilities. Interference with such protective mechanisms disrupts redox homeostasis, leading to oxidative cell death with selective targeting of cancer cells [132, 133]. The most common intracellular antioxidant is GSH, which is found in millimolar amounts in cancer cells and catalyses the neutralisation of ROS by GPx. The prototypic GSH depletion agent buthionine sulfoximine (BSO) competitively inhibits the rate-limiting enzyme in GSH production, gamma-glutamylcysteine synthetase (gamma-GCS). Preclinical research further showed that BSO and melphalan exhibited striking synergy in childhood neuroblastoma, prompting a Phase I clinical study that revealed the potential of BSO-mediated GSH depletion in pediatric malignancies. More recently, erastin and other system Xc-inhibitors deprive cancer cells of cysteine, the rate-limiting amino acid for GSH synthesis, and inactivate GPX4, resulting in ferroptosis and lipid peroxidation, types of iron-dependent oxidative cell death with therapeutic potential across various tumour types [132, 133]. The main secondary antioxidant pathway is the Trx system, which includes Trx (Trx1/2), TrxR1/2, and NADPH, and has critical functions in DNA synthesis, protein folding, and redox signalling. TrxR1 overexpression in a variety of tumours is associated with resistance to chemotherapy and a negative outcome. Auranofin (Ridaura) is an antirheumatic medication that is a potent, irreversible TrxR inhibitor, achieved by covalent selenocysteine formation at the enzyme’s active site upon gold(I) addition. Auranofin induces apoptosis by increasing mtROS and is in Phase I/II clinical trials for the treatment of chronic lymphocytic leukaemia (CLL), lung, and ovarian cancers. The reduced CAT level in cancer cells makes them more dependent on TrxR than normal cells, thereby providing therapeutic selectivity [133, 134].

Precision redox oncology targets the unique oxidative dependencies of cancer cells. Cancer cells with oncogenic KRAS, MYC amplification, KEAP1 mutations, mitochondrial malfunction, or high basal ROS need GSH, Trx, GPX4, FSP1, and NADPH-regenerating pathways to survive. Damage to compensatory antioxidant pathways may reduce cancer cell tolerance to oxidative stress while protecting normal tissues. GPX4 and FSP1 inhibition depletes GSH, targeting TrxR, while PARP inhibition boosts ROS in DNA repair-deficient malignancies. Context-specific synthetic lethal interactions may improve adaptive resistance and therapeutic selectivity [135–137].

In contrast to apoptosis, ferroptosis is characterized by the accumulation of oxidized PUFA-containing phospholipids that are produced by iron-catalyzed Fenton reactions and ROS-mediated lipid oxidation. Ferroptosis represents an emerging therapeutic target in redox oncology, since cancer cells with an increased oxidative stress and a dysregulated lipid metabolism are often highly sensitive to ferroptotic cell death [138]. GPX4 is a central regulator of ferroptosis sensitivity by detoxifying phospholipid hydroperoxides and preventing lethal accumulation of lipid ROS. Inhibition of system Xc–cystine transport, depletion of GSH, or direct GPX4 inhibition disrupts this protective pathway and induces ferroptosis. Another important player is ACSL4 (acyl-CoA synthetase long-chain family member 4), which enriches cellular membranes with PUFA-containing phospholipids that are highly susceptible to peroxidation and consequently increases ferroptosis sensitivity [139]. Multiple parallel ferroptosis defense systems beyond GPX4 have been recently identified by studies. Ferroptosis suppressor protein 1 (FSP1) reduces coenzyme Q10 (CoQ10) and suppresses lipid peroxidation at the plasma membrane independently of GPX4, whereas dihydroorotate dehydrogenase (DHODH) provides an additional mitochondrial ferroptosis defense mechanism through mitochondrial CoQ reduction. These compensatory antioxidant pathways are important contributors to therapy-refractory tumor ferroptosis resistance [140, 141].

Increased transferrin receptor expression, ferritin degradation, labile iron accumulation, and mitochondrial iron mishandling can potentiate ROS-induced lipid peroxidation and vulnerability to ferroptosis [142]. Conversely, cancer cells can evade ferroptosis through increased antioxidant buffering, Nrf2 activation, lipid remodeling alterations, iron sequestration, or upregulation of FSP1- and DHODH-mediated defense pathways. Ferroptosis is closely related to ROS generation, metabolic reprogramming, and therapy resistance; thus, ferroptosis-inducing strategies are increasingly studied in combination with chemotherapy, radiotherapy, immunotherapy, and nanoparticle-based ROS amplification systems [135].

The combination of ROS modulation and the currently used cancer treatment methods, such as chemotherapy, radiotherapy, and immunotherapy, is one of the most rapidly developing fields in translational oncology. The combination rationalisation behind the use of combinatorics is based on mechanistic synergy: ROS-modulating agents can sensitise tumour cells to traditional cytotoxic therapy, circumvent resistance, and augment immune-mediated tumour elimination [143, 144]. The cytotoxic action of ionising radiation (IR) is due mainly to the formation of hydroxyl radicals through radiolysis of intracellular water that causes irreparable DNA breaks in the form of double-stranded breaks (DSBs) in the replicating tumour cells. Intratumoral hypoxia restricts radiation-generated ROS, leading to radioresistance. Cancer stem cells (CSCs), which are proportionately the major culprits of radioresistance and tumour recurrence, exhibit lower ROS levels due to higher antioxidant expression; targeting this radioresistance pathway could be achieved by targeting CSC antioxidant systems [99, 143]. The most significant transformation of a fundamental model between redox biology and cancer treatment is perhaps the convergence of ROS modulation and cancer immunotherapy. One type of regulated cell death, called ICD, occurs in response to specific pro-oxidant stimuli and is characterised by the surface exposure of calreticulin (ecto-CRT), ATP secretion, a find-me signal, and HMGB1 release, a danger signal. The combined effects of these damage-associated molecular patterns (DAMPs) include stimulation of dendritic cells, cross-presentation of tumour antigens, and activation of cytotoxic T lymphocyte (CTL) responses. The main inducers of ICD, such as anthracyclines, oxaliplatin, cyclophosphamide, and PDT, share a common mechanism: They cause ER stress and ROS production in mitochondria [144, 145].

Oxidative modification of PD-L1 by ROS at cysteine residues may affect its stability, glycosylation, and interaction with PD-1, thereby altering the response to checkpoint inhibitors. The synergistic antitumor activity of pro-oxidant agents, when combined with anti-PD-1 antibodies, has transformed immunologically cold tumours into immunologically hot tumours, thereby enabling checkpoint blockade. Several Phase I/II clinical trials are ongoing to test this combination strategy in a variety of solid tumour histologies [134, 144]. The existing guidelines on consensus have suggested no regular high-dose antioxidant supplementation during radiotherapy and platinum-based chemotherapy until conclusive results from prospective studies are available [99, 145]. However, the immunomodulatory effects of ROS are highly context-dependent and can also drive immune suppression in the TME. Chronic oxidative stress in T cells contributes to T-cell exhaustion by impairing TCR signaling and enhancing the expression of inhibitory receptors (e.g., PD-1), thereby limiting cytotoxic activity. At the same time, ROS supports the proliferation and suppressive activity of MDSCs and induces macrophage polarization to the pro-tumorigenic M2 phenotype. These effects together highlight a dual role of ROS and emphasize the need for precise modulation to improve the efficacy of immunotherapy [146–148].

Nanotechnology platforms can be used to address the selectivity challenges of systemic ROS modulation. Fenton and Fenton-like reactions catalysed by iron oxide nanoparticles and copper-based metal-organic frameworks (MOFs) are deactivated in the acidic, H₂O₂-rich TME at interstitial levels, producing highly toxic hydroxyl radicals selectively at the tumour site: a strategy known as chemodynamic therapy (CDT). Nanoplatforms loaded with glucose oxidase (GOx) use intratumoral glucose to catalyse the production of H₂O₂ in the tumour, which in turn starves cancer cells by serving as a metabolic substrate by producing hydroxyl radicals through the Fenton reaction. These methods constitute an excellent amalgamation of tumour microenvironment hijacking and targeted ROS amplification, with several nanoformulations undergoing Phase I trials [134, 149]. ROS-responsive drug delivery and prodrug strategies are under investigation to enhance tumor selectivity and reduce systemic toxicity. The high oxidative stress in many cancers makes it advantageous to release therapeutic payloads in ROS-rich tumor locations using ROS-sensitive linkers such as thioketal bonds, boronic esters, or oxidation-sensitive polymers. ROS-activatable prodrugs are oxidized in the TME to exert cytotoxicity. Nanoparticle-based systems for endogenous ROS generation and delivery of chemotherapeutic medicines, ferroptosis inducers, or immunomodulators are prospective precision redox oncology multimodal therapies [150–152].

Despite promise, key translational hurdles remain for CDT and GOx-based nanoplatforms. Their efficacy depends on heterogeneous factors of the TME (e.g., H₂O₂ levels and glucose availability), which can limit consistent ROS generation. Therapeutic accumulation is further hampered by inefficient delivery, rapid clearance, and poor tumor penetration. Furthermore, the lack of clinically predictive models and the difficulties of large-scale reproducibility are still barriers to successful clinical translation [153–155]. A summary of redox-targeted therapeutic strategies in cancer is presented in Table 1.

The main obstacle that ROS-based cancer treatments face is the need to establish a specific therapeutic redox window that activates programmed cell death in cancer cells while protecting normal tissue from permanent oxidative damage [122]. Elevated basal ROS levels in cancer cells result from their metabolic and signaling defects. In contrast, normal cells maintain their normal ROS levels because both cancer and normal cells need ROS to support vital biological functions, including signal transduction, immune system activation, and tissue repair [158]. The process of systemic ROS modulation poses a significant risk because any increase beyond the toxic threshold will cause complete damage to DNA, proteins, and lipids in healthy cells, resulting in significant side effects and a complete breakdown of cellular balance [23]. ROS function through a complex mechanism that requires specific dosage levels, as low to moderate doses activate pro-tumorigenic pathways that lead to cellular changes and cell survival. In contrast, high-dose levels exceed a cell’s capacity to handle antioxidants, leading to cell death via apoptosis, ferroptosis, and other forms of controlled cell death [159]. The clinical progress of several prooxidant agents has been severely hindered by an exceptionally narrow therapeutic window, which requires high drug concentrations to achieve therapeutic effects but results in dangerous toxicity to patients [122].

Tumorigenesis is a multistep process in which oncogenic mutations arising in a single somatic cell confer clonal advantage as the initiating event; yet mutations alone are insufficient for full malignant transformation, as additional epigenetic alterations and extrinsic environmental pressures including aging, chronic inflammation, and genotoxic insults must converge to drive progression toward an irreversible, highly heterogeneous, and invasive lesion [160]. Tumor heterogeneity is the term for the different kinds of cancer cells that can be found in a tumor (intratumoral) or between patients (intertumoral). These differences can be in genetics, gene expression, and metabolism. This complexity, resulting from clonal evolution influenced by Darwinian selective pressures from both intrinsic cellular factors and extrinsic microenvironmental stimuli such as hypoxia, altered pH, and nutrient gradients, produces phenotypically distinct and drug-resistant subpopulations capable of immune evasion, therapeutic resistance, and sustained uncontrolled growth, rendering treatment, diagnosis, and disease monitoring exceedingly difficult [160, 161]. This tumorigenic cascade is driven by ROS dysregulation, highly reactive oxygen-derived intermediates that balance intracellular levels for normal homeostasis. DNA damage, genomic instability, and oncogenic signaling pathway activation result from oxidative stress, an imbalance between ROS generation and antioxidant defenses. It also provides a pro-tumorigenic milieu that promotes cancer cell proliferation, survival, and metastasis [162]. Redox complexity arises not merely from elevated ROS levels per se, but from the dynamic tension between oxidant production and the adaptive antioxidant machinery that cancer cells co-opt to survive; the Nrf2–Keap1 signaling axis serves as the master regulator of this response, wherein oxidative stress releases Nrf2 from its repressor Keap1, enabling nuclear translocation and transcriptional activation of over 250 cytoprotective genes governing GSH synthesis, Trx activity, and NADPH regeneration thereby establishing an adaptive redox buffer that, while protective in normal tissue, becomes a pro-oncogenic mechanism when constitutively activated in tumor cells [22, 162].

Sustained NrF2 activation in cancer, frequently mediated by loss-of-function mutations in Keap1, enhances GSH synthesis, upregulates detoxifying enzymes, and promotes drug efflux transporter expression, collectively producing a multidrug-resistant phenotype that confounds therapeutic strategies [162]. The resulting redox complexity is further amplified by the heterogeneity of ROS thresholds across distinct tumor subclones, the compartment-specific nature of ROS generation from NOX and mitochondrial ETC sources and the redundancy of parallel antioxidant systems, including both the TrxR (TXNRD1) and GSH reductase (GSR) pathways regulated by NrF2, all of which collectively obscure the precise oxidative switch governing the transition between tumor-promoting and cytotoxic ROS signaling, underscoring the urgent need for real-time redox profiling, subtype-specific intervention strategies, and biomarker-guided combination therapies [162, 163].

Malignant cells exhibit a profound redox adaptation that enables them to survive chronic oxidative stress by developing an enhanced endogenous antioxidant capacity. This metabolic flexibility renders cancer cells relatively insensitive to additional stress inducers, effectively blunting the therapeutic efficacy of conventional treatments such as chemotherapy and radiation [164]. A cornerstone of this adaptive mechanism is the persistent activation of the Nrf2–Keap1 signaling axis; while Nrf2 is typically targeted for degradation by the Keap1–Cul3 complex under resting conditions, oxidative stress stabilizes the protein, allowing it to translocate to the nucleus [158]. Upregulated GSH biosynthesis and Trx system activity have been directly correlated with cancer aggressiveness and resistance to platinum-based chemotherapy and radiation therapy, as these systems enable tumor cells to neutralize apoptotic signals and maintain redox homeostasis. This targeted disruption of redox adaptation is increasingly viewed as a means to overcome drug resistance in advanced tumors [122, 165].

Redox-based oncology translation encounters a significant challenge because it requires clinically verified biomarkers that can measure intracellular ROS levels and metabolic fluxes in real time [158]. Current research often relies on static surrogate markers, including measurement of antioxidant enzyme expression levels (e.g., SOD and CAT) and assessment of oxidative damage product accumulation (e.g., malondialdehyde and 8-oxo-G) [23]. While these are useful for indicating past oxidative stress, they fail to capture the rapid, dynamic redox states and interconversions of specific oxidant species that characterize living tumors [158, 166].

This technical limitation leads to several clinical challenges. For example, Poor Patient Stratification, as it is difficult to identify which patients possess the specific “redox phenotypes” most likely to respond to prooxidant or antioxidant interventions [167]. Researchers face difficulties evaluating tumor redox alterations because they lack access to real-time redox-monitoring technology that would enable more precise measurements. The study shows that nonspecific ROS dyes mislead data interpretation, underscoring the need for better scientific methods, including deep-tissue imaging and mass spectrometry [158, 168].