Metabolic vulnerabilities in ovarian cancer decoding the nexus between nutrient adaptation and therapy resistance.

Ovarian cancer cells require energy and building blocks to grow rapidly. One of the most prominent metabolic adaptations they adopt is aerobic glycolysis, commonly referred to as the Warburg effect ( Fig. 1 ). Even when oxygen is available, these cells preferentially convert glucose into lactate rather than oxidizing it through mitochondrial respiration [39] , [40] . This process is less efficient in ATP yield but supports the biosynthetic needs of proliferation. In ovarian tumors, glycolysis is markedly elevated. Glucose uptake is high, and the excess lactate produced is exported into the TME. This lactate plays a critical role beyond metabolism [41] , [42] . It acidifies the TME, creating hostile conditions for immune cells such as cytotoxic T lymphocytes and natural killer cells. The acidic environment reduces immune surveillance, allowing tumor cells to escape detection and destruction. Fig. 1 Pathways and regulation of glycolysis, the TCA cycle, and oxidative phosphorylation. The figure illustrates glucose metabolism from cellular uptake through glycolysis in the cytoplasm, entry into the tricarboxylic acid (TCA) cycle, and mitochondrial oxidative phosphorylation, highlighting key regulatory enzymes and metabolites. Abbreviations: 1,3BPG, 1,3-bisphosphoglycerate; 2PG, 2-phosphoglycerate; 3PG, 3-phosphoglycerate; 3PPyr, 3-phosphopyruvate; 3PSer, 3-phosphoserine; ECAR, extracellular acidification rate; F1,6BP, fructose-1,6-bisphosphate; F6P, fructose-6-phosphate; G3P, glyceraldehyde-3-phosphate; G6P, glucose-6-phosphate; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HIF1, hypoxia-inducible factor 1; HK2, hexokinase 2; LDH, lactate dehydrogenase; MCT4, monocarboxylate transporter 4; MPC1, mitochondrial pyruvate carrier 1; OCR, oxygen consumption rate; PDH, pyruvate dehydrogenase; PDK, pyruvate dehydrogenase kinase; PEP, phosphoenolpyruvate; PFK1, phosphofructokinase 1; PFKFB3, 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase 3; PGM1, phosphoglucomutase 1; PHGDH, phosphoglycerate dehydrogenase; PK, pyruvate kinase; ROS, reactive oxygen species; TPI, triosephosphate isomerase. Pathways and regulation of glycolysis, the TCA cycle, and oxidative phosphorylation. The figure illustrates glucose metabolism from cellular uptake through glycolysis in the cytoplasm, entry into the tricarboxylic acid (TCA) cycle, and mitochondrial oxidative phosphorylation, highlighting key regulatory enzymes and metabolites. Abbreviations: 1,3BPG, 1,3-bisphosphoglycerate; 2PG, 2-phosphoglycerate; 3PG, 3-phosphoglycerate; 3PPyr, 3-phosphopyruvate; 3PSer, 3-phosphoserine; ECAR, extracellular acidification rate; F1,6BP, fructose-1,6-bisphosphate; F6P, fructose-6-phosphate; G3P, glyceraldehyde-3-phosphate; G6P, glucose-6-phosphate; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HIF1, hypoxia-inducible factor 1; HK2, hexokinase 2; LDH, lactate dehydrogenase; MCT4, monocarboxylate transporter 4; MPC1, mitochondrial pyruvate carrier 1; OCR, oxygen consumption rate; PDH, pyruvate dehydrogenase; PDK, pyruvate dehydrogenase kinase; PEP, phosphoenolpyruvate; PFK1, phosphofructokinase 1; PFKFB3, 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase 3; PGM1, phosphoglucomutase 1; PHGDH, phosphoglycerate dehydrogenase; PK, pyruvate kinase; ROS, reactive oxygen species; TPI, triosephosphate isomerase. Several signaling pathways drive this metabolic reprogramming. Hypoxia-inducible factor 1 alpha (HIF-1α) is stabilized in low oxygen regions, which are common in growing tumors [34] , [43] ( Fig. 2 ). It activates transcription of glycolytic enzymes and glucose transporters. The PI3K/Akt pathway also promotes glycolytic gene expression and enhances glucose uptake. In parallel, the oncogene MYC amplifies glycolysis and boosts glutamine metabolism, further supporting biosynthesis and redox balance. Together, these regulators orchestrate a robust glycolytic phenotype. They make glycolysis a central axis of energy production and an important survival mechanism. Targeting this pathway is an attractive strategy. Inhibitors of key enzymes like hexokinase 2 (HK2), lactate dehydrogenase A (LDHA), and pyruvate kinase M2 (PKM2) have shown promise in preclinical studies [44] , [45] . Blocking monocarboxylate transporters (MCTs), which export lactate, can prevent acidification and restore immune activity in the TME. Glycolysis inhibition may also sensitize tumors to chemotherapy and immunotherapy. Reducing lactate levels could boost immune infiltration and decrease antioxidant capacity, making cancer cells more vulnerable. Overall, aerobic glycolysis provides a metabolic advantage to ovarian tumors but also creates a dependency that can be therapeutically exploited. Fig. 2 Regulation of lactate metabolism by oncogenes and the “lactate exchange” phenomenon between tumor cells. Hypoxic tumor cells (glycolytic type) rely on glycolysis and glutamine metabolism for energy, exporting excess lactate into the tumor microenvironment (TME) via high MCT4 expression. This accumulation lowers pH, creating an acidic niche that impairs immune cell function and metabolism. In contrast, oxygen-rich tumor cells (oxidative type) preferentially import lactate through high MCT1 expression and use it as an energy source via lactate oxidation and the TCA cycle. Regulation of lactate metabolism by oncogenes and the “lactate exchange” phenomenon between tumor cells. Hypoxic tumor cells (glycolytic type) rely on glycolysis and glutamine metabolism for energy, exporting excess lactate into the tumor microenvironment (TME) via high MCT4 expression. This accumulation lowers pH, creating an acidic niche that impairs immune cell function and metabolism. In contrast, oxygen-rich tumor cells (oxidative type) preferentially import lactate through high MCT1 expression and use it as an energy source via lactate oxidation and the TCA cycle. Collectively, work on aerobic glycolysis and lactate handling in ovarian cancer does more than recapitulate textbook biochemistry. It shows that the extent and regulation of glycolytic flux, lactate export and “lactate shuttling” between hypoxic and oxygenated tumor cells are tightly linked to immune exclusion, stemness and chemotherapy tolerance. Quantitative studies in patient samples and models consistently indicate that highly glycolytic, MCT4-high tumors display more acidic, immunosuppressive microenvironments and poorer outcomes, whereas tumors that can flexibly switch between glycolysis and mitochondrial respiration are better equipped to survive therapeutic stress. Thus, the circuits highlighted in Fig. 1 , Fig. 2 should be viewed as an integrated stress-adaptation module, rather than as isolated pathways, and they provide a mechanistic rationale for co-targeting glycolysis, lactate transport and immune checkpoints in metabolically “hot” ovarian cancers. Beyond supplying energy and biosynthetic precursors, central carbon metabolism also reshapes the ovarian cancer epigenome and signaling networks through metabolite-mediated protein modifications. A key node in this crosstalk is acetyl-CoA, which serves as the obligate donor for lysine acetylation on histones and numerous non-histone proteins. Cytosolic and nuclear acetyl-CoA pools are tightly coupled to glycolytic and TCA-cycle activity through ATP-citrate lyase and acetyl-CoA synthetase 2, which convert citrate and acetate—or, indirectly, lactate-derived pyruvate—into acetyl-CoA. When glycolytic flux and lactate production are high, as in many OC cells, acetyl-CoA availability increases and drives histone acetylation at promoters and enhancers of genes involved in cell-cycle progression, DNA damage response, epithelial–mesenchymal transition, and stemness [44] , [45] . This metabolite-epigenetic axis therefore links nutrient abundance to transcriptional programs that sustain tumor cell survival and promote resistance to platinum agents and PARP inhibitors under therapeutic stress. In parallel, acetyl-CoA-dependent acetylation of non-histone substrates—including metabolic enzymes and DNA repair proteins—adds an additional regulatory layer. Acetylation can alter enzyme activity, subcellular localization, or protein–protein interactions, thereby fine-tuning glycolysis, mitochondrial respiration, and redox homeostasis in ways that favor survival in nutrient-limited or drug-exposed conditions. These acetylation events are dynamically opposed by NAD + -dependent sirtuin deacetylases, positioning the NAD + /NADH ratio—already central to OXPHOS and PARP activity—as a critical rheostat for both metabolism and chromatin state [44] , [45] . Moreover, lactate itself can act as a signaling metabolite by fueling histone lactylation, an emerging post-translational modification that has been implicated in the activation of pro-survival and immunoregulatory gene expression programs in highly glycolytic tumors. Taken together, these findings indicate that metabolite-driven protein modifications constitute an integral component of metabolic reprogramming in ovarian cancer, coupling nutrient adaptation to stable transcriptional re-wiring and helping to consolidate therapy-resistant phenotypes. While glycolysis is a dominant feature, many ovarian cancer cells retain or even elevate mitochondrial OXPHOS. This process, centered in the mitochondria, provides efficient ATP production and supports key biosynthetic and redox processes. OXPHOS is particularly important in tumor cells that must survive under fluctuating nutrient and oxygen conditions [20] , [46] ( Fig. 3 ). Ovarian tumors exhibit striking metabolic flexibility. Some cells toggle between glycolysis and OXPHOS depending on their environment or therapeutic pressure. This capacity to switch metabolic modes allows tumors to adapt and survive under stress. Within ovarian cancer, at least two broad metabolic phenotypes exist: OXPHOS-high and glycolysis-high. OXPHOS-high tumors demonstrate elevated mitochondrial activity, increased oxygen consumption, and high expression of electron transport chain genes [21] , [26] . Glycolysis-high tumors, by contrast, show reduced mitochondrial respiration and depend heavily on cytosolic glucose metabolism. This distinction carries clinical relevance. Studies show that OXPHOS-high tumors are more sensitive to chemotherapy. One reason lies in reactive oxygen species (ROS) [47] , [48] . OXPHOS generates ROS as a byproduct. When chemotherapy adds further oxidative stress, these cells may be pushed beyond their tolerance, leading to cell death. Glycolytic tumors, on the other hand, have stronger antioxidant defenses and lower ROS, which helps them resist treatment. Fig. 3 Oxidative phosphorylation. Aerobic glycolysis converts glucose to lactate, while oxidative phosphorylation (OXPHOS) in mitochondria efficiently generates ATP. OXPHOS remains active in many tumors and cancer stem cells. Inhibitors targeting subunits of the mitochondrial electron transport chain are promising therapeutic candidates, particularly for HGSOC. Abbreviations: CytC, cytochrome C; ETC, electron transport chain; LDH, lactate dehydrogenase; Q, ubiquinone; QH 2 , ubiquinol; TCA, tricarboxylic acid cycle; FADH 2 , flavin adenine dinucleotide; NADH, nicotinamide adenine dinucleotide. Oxidative phosphorylation. Aerobic glycolysis converts glucose to lactate, while oxidative phosphorylation (OXPHOS) in mitochondria efficiently generates ATP. OXPHOS remains active in many tumors and cancer stem cells. Inhibitors targeting subunits of the mitochondrial electron transport chain are promising therapeutic candidates, particularly for HGSOC. Abbreviations: CytC, cytochrome C; ETC, electron transport chain; LDH, lactate dehydrogenase; Q, ubiquinone; QH 2 , ubiquinol; TCA, tricarboxylic acid cycle; FADH 2 , flavin adenine dinucleotide; NADH, nicotinamide adenine dinucleotide. In high-grade serous ovarian cancer (HGSOC), this metabolic classification may predict treatment response. OXPHOS-high tumors respond better to platinum agents, while glycolysis-dominant subtypes tend to be more chemoresistant. Clear cell OCCC, known for its resistance, often exhibits low OXPHOS and avoids mitochondrial stress [49] , [50] . Targeting OXPHOS may overcome resistance. Several mitochondrial inhibitors are being evaluated. Metformin, a widely used diabetes medication, inhibits mitochondrial complex I and reduces ATP production. It also increases ROS and reduces NAD + regeneration, which impairs DNA repair [51] , [52] . IACS-010759 is a more potent complex I inhibitor under clinical investigation. Devimistat (CPI-613) disrupts the TCA cycle and further impairs mitochondrial metabolism. These agents show increased efficacy when combined with chemotherapy or PARP inhibitors. They can trigger synthetic lethality by increasing oxidative stress and impairing repair pathways. OXPHOS also plays a role in NAD + homeostasis, making it relevant in resistance to DNA damage-targeting agents. In summary, mitochondrial metabolism in ovarian cancer supports energy production, redox control, and therapy resistance. Its inhibition can selectively disrupt tumor survival mechanisms, especially in resistant phenotypes. Glutamine is the most abundant amino acid in circulation, and ovarian tumors are often highly dependent on it ( Fig. 4 ). Unlike glucose, glutamine feeds the TCA cycle more directly and provides the carbon backbone for nucleotide and amino acid synthesis. It also supports redox homeostasis through the synthesis of glutathione, a key antioxidant [27] , [53] . Glutamine metabolism becomes particularly important under stress. Fig. 4 Three rate-limiting enzymes in aerobic glycolysis. HK2, PFK1, and PKM2 are the key rate-limiting enzymes in the glycolytic pathway. HK2 catalyzes the phosphorylation of glucose to glucose-6-phosphate (G-6-P) and binds to VDAC1 on the outer mitochondrial membrane, enabling efficient ATP entry to fuel this step. PFK1 converts fructose-6-phosphate (F-6-P) to fructose-1,6-bisphosphate (F-1,6-BP); its activity is enhanced by fructose-2,6-bisphosphate (F-2,6-BP), generated by PFKFB3. PKM2 catalyzes the conversion of phosphoenolpyruvate (PEP) to pyruvate and can also translocate to the nucleus, where it acts as a protein kinase to phosphorylate histone H3 and activate transcription factors such as c-MYC, STAT3, and NF-κB, thereby indirectly sustaining aerobic glycolysis and driving tumor cell proliferation. Three rate-limiting enzymes in aerobic glycolysis. HK2, PFK1, and PKM2 are the key rate-limiting enzymes in the glycolytic pathway. HK2 catalyzes the phosphorylation of glucose to glucose-6-phosphate (G-6-P) and binds to VDAC1 on the outer mitochondrial membrane, enabling efficient ATP entry to fuel this step. PFK1 converts fructose-6-phosphate (F-6-P) to fructose-1,6-bisphosphate (F-1,6-BP); its activity is enhanced by fructose-2,6-bisphosphate (F-2,6-BP), generated by PFKFB3. PKM2 catalyzes the conversion of phosphoenolpyruvate (PEP) to pyruvate and can also translocate to the nucleus, where it acts as a protein kinase to phosphorylate histone H3 and activate transcription factors such as c-MYC, STAT3, and NF-κB, thereby indirectly sustaining aerobic glycolysis and driving tumor cell proliferation. In tumors with ARID1A mutations, the dependency on glutamine is especially pronounced [54] , [55] . These mutations are common in clear cell and endometrioid ovarian cancers. Loss of ARID1A reprograms metabolism by suppressing glucose oxidation and enhancing glutamine utilization [27] , [56] . The resulting metabolic bottleneck creates a vulnerability. Blocking glutamine metabolism in ARID1A-deficient tumors can lead to synthetic lethality. These agents reduce glutamine flux, impair redox balance, and sensitize tumors to other treatments. In ARID1A-mutant models, CB-839 induces cell death and reduces tumor growth [57] . It also appears to improve immune infiltration, likely by reducing immunosuppressive metabolites and enhancing T cell viability. In addition to GLS inhibitors, other strategies target glutamine transporters or enzymes downstream in the pathway. Some also combine glutamine depletion with standard chemotherapy. This approach may be particularly effective in tumors that show metabolic plasticity under drug pressure [58] , [59] . Because normal tissues rely less on glutamine, these therapies may offer a favorable therapeutic window. Glutamine therefore occupies a central position at the interface between bioenergetics, biosynthesis and redox control in ovarian cancer. In genetically defined subsets such as ARID1A-mutant clear cell and endometrioid tumors, suppression of glucose oxidation further hard-wires glutamine dependence, creating a potential metabolic bottleneck. Taken together, Fig. 1 , Fig. 2 , Fig. 3 , Fig. 4 are therefore intended not as textbook schematics, but as a conceptual framework that synthesises what metabolic research in ovarian cancer has revealed. They highlight how canonical central-carbon pathways—glycolysis, the TCA cycle, oxidative phosphorylation and their key rate-limiting enzymes—form a highly interconnected network that cancer cells dynamically rewire to balance ATP production, biosynthesis and redox control under therapeutic pressure. By mapping where oncogenic signals, hypoxia and nutrient stress converge on this network, these figures illustrate why ovarian tumors can shift between glucose, glutamine and lipid fuels, and how this flexibility underpins resistance to platinum agents, PARP inhibitors, anti-angiogenic therapy and immunotherapy. This integrated view moves beyond pathway description and sets the stage for later sections that focus on which metabolic nodes appear most druggable, in which molecular contexts, and with which combinatorial partners. Fatty acids serve as a rich energy source and a structural component of cell membranes. In ovarian cancer, lipid metabolism plays an especially important role during metastasis and in response to treatment. Many ovarian tumors colonize the omentum, an adipose-rich environment that supplies exogenous lipids. Tumor cells exploit this niche by inducing lipolysis in nearby adipocytes, releasing fatty acids for uptake. This lipid fueling process is facilitated by several proteins. Fatty acid-binding protein 4 (FABP4) is expressed at the interface between tumor cells and adipocytes [60] , [61] . It binds and transports fatty acids into cancer cells. Once inside, fatty acids are directed into mitochondria for β-oxidation via the carnitine palmitoyltransferase 1 (CPT1) pathway. This process generates ATP and NADPH, supporting both energy and redox balance. Fatty acid oxidation (FAO) becomes more prominent in stress conditions, such as chemotherapy exposure [9] , [62] . Platinum-resistant ovarian cancer cells often reduce glucose metabolism and switch to FAO as an alternative energy source. This metabolic shift also supports survival by producing NADPH, which detoxifies ROS. Inhibiting FAO with CPT1 blockers like etomoxir can impair this adaptation and restore sensitivity to platinum agents. Beyond energy production, lipid synthesis is also upregulated in ovarian cancer. Fatty acid synthase (FASN), a key enzyme in de novo lipogenesis, is overexpressed in many tumors. It supports membrane synthesis, signaling, and cell proliferation. Targeting FASN disrupts these processes and sensitizes tumors to chemotherapy. Lipid metabolism also intersects with immune regulation. Tumors that rely on FAO may create immunosuppressive environments. Certain lipid species can alter dendritic cell function or promote T cell exhaustion [28] , [63] . Targeting FAO may therefore have dual benefits: impairing tumor survival and improving immune responses. In summary, lipid metabolism is deeply integrated into ovarian cancer biology. Tumors use both imported and synthesized fatty acids to fuel growth, survive stress, and interact with their microenvironment. These pathways offer distinct and actionable therapeutic targets, especially in metastatic and chemoresistant disease. The ovarian cancer tumor microenvironment (TME) is a metabolically coupled ecosystem in which cancer cells, cancer-associated fibroblasts (CAFs), adipocytes and immune cells constantly exchange nutrients and metabolites. CAFs frequently undergo aerobic glycolysis, upregulate monocarboxylate transporter 4 (MCT4), and export lactate, pyruvate and other high-energy metabolites that are then imported and oxidized by adjacent cancer cells—the so-called “reverse Warburg effect”. This stromal–epithelial lactate shuttle supports oxidative phosphorylation, enhances bioenergetic flexibility and promotes invasion and therapy resistance, with high stromal MCT4 expression emerging as a poor-prognosis marker in multiple solid tumors and likely relevant to desmoplastic, CAF-rich ovarian cancers as well [28] , [63] . Beyond acting as passive stromal scaffolds, CAFs in ovarian cancer function as dynamic metabolic “engineers” that continuously remodel nutrient fluxes within the TME. Ovarian cancer cell–conditioned media can drive fibroblast phenoconversion toward a glycolytic, autophagy-dependent CAF phenotype that secretes lactate, pyruvate, alanine and glutamine-rich metabolites; these are then oxidized by adjacent cancer cells to sustain OXPHOS, TCA anaplerosis and redox balance under chemotherapy- or hypoxia-induced stress [28] , [63] . In parallel, CAFs themselves undergo metabolic reprogramming orchestrated by oncogenic signaling and stromal cues—including TGF-β, IL-6 and CXCL12—that enhance glycolysis, glutamine synthesis and antioxidant capacity, thereby enabling CAFs to buffer nutrient fluctuations and continuously supply metabolic substrates to the tumor compartment [28] , [63] . Through this coupling, CAF-derived metabolites and cytokines promote epithelial–mesenchymal transition, stemness and invasion, and have been linked to platinum resistance and immune evasion in preclinical ovarian cancer models. Conceptually, these data position CAFs as metabolic gatekeepers that translate stromal stress signals into durable shifts in tumor-cell bioenergetic wiring, highlighting CAF-directed strategies (e.g., interference with CAF activation programs, lactate/glutamine shuttling or CAF-derived exosomes) as rational partners for tumor-intrinsic metabolic inhibitors in future combination therapies. Metabolic competition and crosstalk also profoundly shape antitumor immunity. Effector CD8 + T cells require robust glycolysis and mitochondrial function, whereas regulatory and exhausted T cells rely more on oxidative and fatty acid–driven metabolism; in a glucose-depleted, lactate-rich TME, effector T cells are deprived of key nutrients and exposed to high lactate, which directly impairs proliferation, cytokine production and cytotoxicity and thus limits the efficacy of immune checkpoint blockade [28] , [63] . Tumor- and stroma-derived lipids further promote immunosuppressive programs in regulatory T cells, tumor-associated macrophages (TAMs) and myeloid-derived suppressor cells, underscoring that lipid reprogramming is as central as glucose metabolism to immune escape in the ovarian TME [28] , [63] . In ovarian cancer, adipocytes represent a particularly powerful stromal driver of these lipid-centred interactions. Omental and peritoneal adipocytes undergo hormonally and cytokine-driven lipolysis, releasing free fatty acids, ketone bodies and adipokines that are avidly taken up by cancer cells via transporters such as CD36 and FABP4, as well as by lipid-scavenging immune cells in ascites and the omental niche [28] , [63] . Once imported, these fatty acids fuel mitochondrial β-oxidation to sustain ATP and NADPH production, support membrane synthesis and reactive-oxygen-species buffering, and cooperate with CAF-derived metabolites to maintain tumor growth under hypoxic, anti-angiogenic or chemotherapy-induced stress. In parallel, adipocyte-derived adipokines—including apelin, IL-17A and leptin—activate STAT3- and APJ-driven signalling that upregulates CD36/FABP4-dependent lipid uptake, reinforces epithelial–mesenchymal transition and cancer stemness, and promotes resistance to platinum agents and anti-VEGF therapy in preclinical models of ovarian cancer. Moreover, work in lipid-rich ovarian cancer ascites indicates that adipocyte-conditioned microenvironments not only feed tumor cells but also reprogram infiltrating macrophages and regulatory T cells toward FAO-dependent, immunosuppressive states, thereby blunting CD8 + T-cell fitness and dampening responses to immune checkpoint blockade [28] , [63] . Taken together, these findings argue that adipocytes in the omentum and peritoneal fat pads should be viewed as metabolically active hubs that coordinate tumor–stroma–immune crosstalk; targeting adipocyte-to-tumor lipid transfer or adipokine-driven signalling may therefore synergise with glycolysis, FAO or checkpoint inhibitors to dismantle the metabolic ecosystem that underlies therapy resistance in ovarian cancer. TAMs themselves exemplify this immunosuppressive metabolic rewiring: pro-tumor macrophages increasingly depend on fatty acid oxidation, oxidative phosphorylation and altered amino-acid metabolism, driven by hypoxia, lactate and cholesterol efflux, which reinforce an M2-like phenotype and blunt phagocytosis and T-cell activation [28] , [63] . Together, these observations highlight that therapeutic strategies aimed at ovarian cancer metabolism must consider not only tumor-cell-intrinsic pathways but also CAF, T-cell and TAM metabolism as interconnected targets; rational combinations that alleviate nutrient competition, modulate lactate or lipid fluxes, and restore immune-cell metabolic fitness are therefore particularly attractive avenues for overcoming resistance [28] , [63] . Many of the metabolic alterations discussed in this review are not unique to ovarian cancer but represent core hallmarks of malignant growth shared across multiple tumor types. Aerobic glycolysis (the Warburg effect), glutamine addiction, enhanced mitochondrial OXPHOS, redox rewiring, and increased lipid utilization have been described in diverse cancers, including colorectal, lung, and breast malignancies [64] , [65] . These adaptations support rapid ATP generation, biosynthesis of nucleotides and lipids, and maintenance of redox balance, thereby enabling proliferation and survival under therapeutic stress [28] , [63] . From this broader perspective, ovarian cancer fits into a conserved framework in which metabolic plasticity allows tumor cells to toggle between glucose, glutamine, and fatty acids as major fuels and to dynamically reconfigure NAD + , NADPH, and glutathione pools in response to chemotherapy, targeted agents, or immune pressure. At the same time, ovarian cancer exhibits several metabolic features that are either highly enriched or functionally accentuated compared with other solid tumors, reflecting its preferred metastatic niches and etiologic links to endometriosis. One prominent example is the dependency of omental metastases on adipocyte-derived lipids: adipocytes in the omentum donate fatty acids that are imported via CD36 and chaperoned by FABP4 into ovarian cancer cells, where they are oxidized to support rapid growth and contribute to carboplatin resistance [28] , [63] . While adipocyte–tumor crosstalk has been reported in other cancers, the magnitude of this FABP4/CD36-driven lipid fueling appears particularly critical in high-grade serous ovarian cancer and is tightly coupled to its characteristic pattern of peritoneal spread. A second ovarian cancer–specific theme is the strong linkage between ARID1A inactivation, glutamine dependence, and clear cell histology. ARID1A mutations occur across tumor types, but in ovarian clear cell carcinoma they are especially frequent and create a pronounced reliance on glutamine metabolism via upregulation of GLS1, rendering these tumors exquisitely sensitive to glutaminase inhibition with agents such as CB-839 [28] , [63] . This genetically defined, glutamine-addicted state is less consistently observed in ARID1A-mutant cancers at other sites, underscoring a context-specific convergence of chromatin remodeling defects, endometriosis-associated oxidative stress, and nutrient wiring in the ovary. Third, iron-rich, endometriosis-derived microenvironments give rise to a ferroptosis-related landscape that is unusually prominent in ovarian clear cell and endometrioid carcinomas. Endometriotic cyst fluid is enriched in iron, promoting chronic oxidative stress and shaping the evolution of endometriosis-associated ovarian cancers in ways not typically seen in other tumor settings [28] , [63] . Recent work suggests that dysregulated iron metabolism and ferroptosis resistance are tightly intertwined with chemoresistance in these subtypes, offering opportunities to exploit ferroptosis-inducing strategies that may be less directly translatable from non-gynecologic malignancies [28] , [63] . By explicitly distinguishing these shared metabolic hallmarks from ovarian cancer–specific adaptations, our review aims to situate ovarian cancer within the broader field of cancer metabolism while highlighting features that may not be apparent from pan-cancer studies alone. On one hand, this framing facilitates knowledge transfer from other malignancies—for example, borrowing OXPHOS- or glycolysis-targeting strategies that have been developed in colorectal or lung cancer. On the other hand, it emphasizes ovarian-specific phenomena such as adipocyte-driven lipid fueling, ARID1A-dependent glutamine addiction, and iron-mediated ferroptosis biology, which may, in turn, inspire new lines of inquiry in cancers that share similar microenvironmental or genetic constraints. Despite this increasingly refined map of ovarian cancer metabolism, several caveats warrant emphasis. Most of the pathway-level insights summarized in this section have been derived from established cell lines, bulk tumor specimens, or ex vivo models maintained under supraphysiologic nutrient conditions, which cannot fully recapitulate intra-tumoral and spatial heterogeneity in vivo. Static transcriptomic or proteomic readouts also capture metabolic capacity rather than real-time flux, making it difficult to distinguish causal, resistance-driving programs from context-dependent or compensatory states. More systematic use of stable-isotope tracing, single-cell and spatial metabolomics, and longitudinal sampling before and after therapy in patient-derived material will therefore be essential to validate which metabolic adaptations constitute truly actionable vulnerabilities versus epiphenomena of oncogenic signaling or treatment-induced stress.

The field of ovarian cancer metabolism is rapidly evolving. Research has uncovered major pathways that help tumors survive therapy. Yet, many questions remain. There is still much to learn about how to best apply this knowledge in the clinic. Understanding the when, where, and how of metabolic targeting is key to future progress. Understanding tumor metabolism in real time is difficult. Most studies rely on tissue samples. These are usually taken at diagnosis or surgery. But metabolism is dynamic. It changes during treatment. It also varies by location within the tumor and across metastases. We need better ways to monitor these changes. That’s where non-invasive metabolic imaging comes in. Several new imaging technologies are being developed. These can visualize metabolic activity without needing a biopsy. One promising tool is hyperpolarized magnetic resonance spectroscopy (MRS) [107] , [108] . This method tracks how labeled carbon compounds—like 13C-pyruvate—are metabolized in tumors. It shows whether a tumor is glycolytic or OXPHOS-active in real time. Another technique uses positron emission tomography (PET) with radiotracers that detect glucose or amino acid uptake. For example, FDG-PET measures glucose uptake and is widely used in oncology. But newer tracers can also target glutamine metabolism or lipid use. These tools help clinicians identify the dominant metabolic phenotype of a tumor. They also allow tracking changes in response to treatment. If a tumor switches from OXPHOS to glycolysis, imaging can reveal that. This information can guide therapy adjustments. Ultimately, metabolic imaging could become a standard part of precision oncology. It offers a real-time window into tumor biology. With more clinical validation, these tools could transform how we monitor and treat ovarian cancer. Not every tumor uses nutrients the same way. Some are glutamine-addicted. Others rely on fatty acids or show strong glycolysis. Knowing which pathway dominates in a given tumor is critical. It helps match the right drug to the right patient. To do this, we need reliable biomarkers. These are measurable features—genes, proteins, or metabolites—that reveal the metabolic state of a tumor. Several candidates have emerged. High expression of GLS1, SLC1A5, or MYC may signal glutamine dependence. Elevated LDHA, HIF-1α, or MCT4 suggests a glycolytic phenotype. FABP4, CPT1A, and FASN indicate fatty acid metabolism is active. But more work is needed. Biomarkers must be validated across different ovarian cancer subtypes. They should also be stable and accessible, ideally measurable in blood, ascites, or imaging. Metabolomic profiling is another approach. By analyzing blood or tumor samples, researchers can detect patterns in metabolites. These metabolic fingerprints may identify patients who would benefit from a specific metabolic therapy. Gene expression signatures also hold promise. Algorithms can classify tumors based on the activity of metabolic genes. Some models already divide high-grade serous tumors into OXPHOS-high and glycolytic groups. In future trials, integrating biomarker analysis will be crucial. It allows smarter trial design. Instead of testing a drug in all patients, we can focus on those most likely to benefit. This increases the chance of success and reduces unnecessary side effects. Tumor metabolism doesn’t happen in isolation. Cancer cells live in a complex ecosystem. They interact with stromal cells, immune cells, and adipocytes. These neighboring cells exchange nutrients, signals, and support. Ignoring this environment limits our understanding. To fully target metabolism, we must consider the whole ecosystem. Cancer-associated fibroblasts (CAFs) are one key player. They can release nutrients like glutamine or lactate [109] , [110] . In some tumors, CAFs even recycle metabolic waste from cancer cells. This helps sustain tumor growth under stress. Adipocytes are another contributor, especially in ovarian cancer. Tumors that spread to the omentum interact with fat cells. These adipocytes donate fatty acids to tumor cells. This supports energy production via β-oxidation. Blocking this interaction could limit metastasis. Immune cells are also deeply affected by tumor metabolism. High lactate levels suppress T cell function. Tumors can outcompete immune cells for glucose and glutamine. This weakens the immune response and blunts immunotherapy effects. Future therapies should aim to disrupt these metabolic relationships. For example, inhibiting lactate transport may restore immune activity. Targeting stromal glutamine recycling may cut off tumor nutrient supply. New technologies, like single-cell RNA sequencing and spatial metabolomics, can map how different cells in the tumor microenvironment interact. This will help identify which pathways to target and in whom. In summary, tumors are not isolated. They co-opt the metabolic capabilities of their surroundings. Targeting the tumor alone is not enough. We must also intervene in the metabolic crosstalk of the ecosystem. Over the past decade, rapid technological advances have fundamentally reshaped how we study metabolic reprogramming and therapy resistance in ovarian cancer. High-resolution mass-spectrometry–based metabolomics and lipidomics, together with stable isotope tracing, now allow quantitative mapping of nutrient utilization, anaplerotic routes, and redox circuits in patient samples and preclinical models. These platforms have been critical for uncovering glycolysis-versus-OXPHOS phenotypes, glutamine addiction, and fatty-acid–driven metabolic states across ovarian cancer subtypes, as well as for identifying metabolite signatures associated with platinum or PARP inhibitor resistance. Integration of metabolomics with transcriptomic, proteomic, and phosphoproteomic profiling has further enabled systems-level reconstruction of metabolic networks, revealing how oncogenic signaling, DNA damage responses, and the tumor microenvironment converge on specific metabolic nodes. At the same time, single-cell and spatial technologies are beginning to resolve the metabolic ecosystem of ovarian cancer at unprecedented resolution. Single-cell RNA-seq and ATAC-seq can stratify malignant, stromal, and immune populations into distinct metabolic states, while spatial transcriptomics, imaging mass spectrometry, and multiplex immunofluorescence capture the geographical organization of nutrient gradients, hypoxic niches, and immune exclusion zones within primary tumors and omental metastases. Complementary model systems—such as patient-derived organoids, tumor slice cultures, and microfluidic co-culture platforms that incorporate adipocytes, fibroblasts, and immune cells—provide experimentally tractable systems in which to functionally test metabolic dependencies and to track therapy-induced rewiring under controlled conditions. Despite these advances, substantial technical bottlenecks still limit the translation of metabolic insights into routine clinical practice. Many state-of-the-art methods require fresh or rapidly frozen tissue and are highly sensitive to pre-analytical variables such as ischemia time, making standardized sampling across centers challenging. Bulk metabolomics and transcriptomics average signals across heterogeneous cell populations and may obscure rare but clinically important resistant clones, whereas single-cell and spatial multi-omics remain costly, technically demanding, and restricted to specialized laboratories. Stable isotope–based flux analyses are difficult to implement at scale in patients and provide only short-term snapshots of dynamic metabolic behavior. Likewise, advanced metabolic imaging modalities, including hyperpolarized MRS and novel PET tracers, are currently limited by scanner availability, cost, regulatory hurdles, and the lack of harmonized acquisition and analysis pipelines. Data integration and interpretation represent additional bottlenecks. Multi-omics and imaging datasets are high-dimensional and noisy, and robust computational frameworks are needed to deconvolve tumor-intrinsic metabolic states from host and treatment-related confounders. Moreover, most studies to date rely on relatively small, highly selected patient cohorts, and there is a paucity of prospective, longitudinal datasets with paired pre- and post-treatment samples. Addressing these gaps will require standardized pre-analytical workflows, multi-center consortia, and closer integration of bioinformatics, systems biology, and clinical trial design. Only then will technological advances fully translate into clinically actionable metabolic phenotypes and rational, metabolism-informed strategies to overcome therapy resistance in ovarian cancer. Another key question is when to use metabolic therapies. Should we use them at diagnosis? Or reserve them for relapse? Should they be given with chemotherapy or only afterward? Some argue that metabolic therapies should be given upfront, alongside standard chemotherapy. This might prevent the emergence of resistant clones. For example, combining OXPHOS inhibitors with platinum could kill both sensitive and resistant cells from the start. Others suggest using metabolic therapy at relapse, when resistance is established. At this point, tumors may be more metabolically rigid. This could make them more vulnerable to targeted interventions. There’s also the question of sequencing. Should metabolic inhibitors be used before, during, or after chemotherapy? Preclinical data suggest that priming tumors with metabolic stress may make them more sensitive to chemo. But this needs to be tested in clinical settings. Combination strategies also raise timing issues. If we combine PARP inhibitors with GLS inhibitors, which should come first? Do we give them together or in alternating cycles? These decisions affect efficacy and toxicity. Clinical trials are starting to explore these questions. But results are still early. Adaptive trial designs may help test multiple timing strategies in parallel. In the future, real-time biomarkers and imaging may guide treatment sequencing based on how the tumor adapts. In short, timing matters. Understanding the optimal sequence for metabolic therapies is essential for maximizing benefit and minimizing harm. Not all patients are the same. Even with the same tumor subtype, individual differences in metabolism can affect therapy response. These differences stem from diet, age, genetics, comorbidities, and lifestyle. For instance, patients with diabetes may have altered glucose metabolism. This could influence how their tumors respond to glycolysis inhibitors. Patients taking metformin for diabetes may already be affecting their tumor metabolism—possibly for better or worse. Obesity is another variable. Adipose tissue alters systemic metabolism. It also provides fuel for tumors. In ovarian cancer, obese patients may have higher levels of circulating fatty acids. Their tumors may lean more heavily on lipid metabolism. This could make them more responsive to FAO inhibitors. Nutrient availability from the diet also matters. Studies have shown that calorie restriction or ketogenic diets can shift tumor metabolism. These interventions may improve the effects of certain therapies. But the data is still early, and caution is warranted. Genetic factors also shape how patients metabolize drugs and nutrients. Polymorphisms in metabolic enzymes can affect drug clearance or nutrient handling. Future treatments may need to account for these differences. In clinical trials, patient variability is often ignored. Most studies assume one-size-fits-all. But integrating metabolic profiling into trial design can change that. By accounting for host metabolism, we may improve treatment precision and avoid unnecessary toxicity. In summary, each patient brings a unique metabolic background. Understanding and accounting for this can personalize therapy and improve outcomes.

This study was supported by the National Natural Science Foundation of China ( 82574740 ); and Nankai University Eye Institute ( NKYKK202509 ).

Therapy resistance remains the most significant challenge in ovarian cancer treatment. Standard therapies—including platinum-based chemotherapy, PARP inhibitors, and anti-angiogenic agents—often induce an initial response. However, most tumors eventually adapt. Resistance develops, and tumors relapse. A key driver of this adaptation is metabolic reprogramming. Cancer cells adjust their energy sources and metabolic circuits to survive drug-induced stress. Platinum-based chemotherapy is the frontline treatment for most ovarian cancer patients. Drugs like cisplatin and carboplatin cause DNA damage. They form DNA adducts and crosslinks that disrupt replication. In sensitive cells, this leads to cell death. But in resistant cells, metabolic reprogramming plays a critical role in evading this fate. Platinum-resistant ovarian cancer cells often increase mitochondrial activity. They rely more on OXPHOS to generate ATP [46] , [66] . These cells also exhibit increased mitochondrial mass and oxygen consumption. By upregulating OXPHOS, they produce enough energy to support DNA repair and cell survival under stress. At the same time, resistant cells often shift to FAO. This allows them to bypass glucose metabolism when needed. FAO produces both ATP and NADPH, which are vital for managing oxidative stress. Since platinum drugs generate ROS, FAO helps buffer this damage. By supporting redox balance, FAO contributes directly to therapy resistance. In one study, chemoresistant ovarian cancer cells showed a marked increase in CPT1A, the enzyme that enables mitochondrial fatty acid entry [63] , [67] . Inhibition of CPT1A restored sensitivity to cisplatin. This suggests FAO is not just a bystander—it actively drives survival under chemotherapy. Glutamine plays a second critical role in platinum resistance. It feeds into the tricarboxylic acid (TCA) cycle, replenishing intermediates that support energy production. More importantly, glutamine contributes to glutathione (GSH) synthesis [68] , [69] . Beyond mitochondrial rewiring and FAO, platinum-resistant cells frequently exploit glutamine metabolism to reinforce their antioxidant capacity. By funnelling glutamine-derived carbon into the TCA cycle and nitrogen into glutathione biosynthesis, these cells maintain ATP production and buffer chemotherapy-induced ROS [70] , [71] . Platinum-resistant cells are highly adaptable. They survive even when glucose is scarce. Sensitive cells die under glucose starvation, but resistant ones switch to alternate fuels. These include glutamine, fatty acids, and branched-chain amino acids. This metabolic flexibility is a major survival advantage [72] , [73] . It means that cutting off one energy source won’t be enough. Tumor cells can simply pivot to another pathway. This is why combination strategies—targeting multiple metabolic routes—are more likely to succeed. Interestingly, the ability to resist glucose deprivation correlates with common resistance-related mutations. These include TP53 loss, PI3K/Akt activation, and others that rewire nutrient sensing and metabolism. In summary, platinum-resistant ovarian cancer cells survive by increasing mitochondrial OXPHOS, switching to fatty acid oxidation, and enhancing glutamine use. These shifts help them generate ATP, maintain redox balance, and evade cell death. Targeting these adaptations could resensitize tumors and improve chemotherapy outcomes. PARP inhibitors (PARPi) have transformed treatment for patients with homologous recombination-deficient (HRD) ovarian cancers. These include tumors with BRCA1 or BRCA2 mutations. PARP enzymes help repair single-strand DNA breaks. When PARP is inhibited, these breaks accumulate and lead to double-strand breaks. HRD cells cannot repair them, leading to cell death—a concept known as synthetic lethality [74] , [75] . But resistance to PARPi is common. Metabolic changes play a major role in this resistance. PARP enzymes use nicotinamide adenine dinucleotide (NAD + ) as a substrate. Each cycle of DNA repair consumes NAD + . When PARP is overactive, NAD + gets depleted. Some resistant cells restore NAD + pools to survive PARPi therapy. They do this through the NAD + salvage pathway [22] , [76] . Enzymes like NAMPT and NMNAT1 become upregulated. This leads to higher NAD + levels and enables cells to maintain PARP activity or use alternate repair pathways. In one study, PARPi-resistant cells had elevated NMNAT1 expression. Blocking NMNAT1 reduced NAD + levels and restored sensitivity to olaparib. This shows how metabolic adaptation—specifically NAD + recycling—helps bypass PARPi effects. OXPHOS is another factor. HRD cells that survive PARPi treatment often show higher OXPHOS activity. They rely on mitochondrial ATP to power DNA repair machinery. Mitochondrial respiration also helps regenerate NAD + from NADH, further supporting resistance. These insights suggest a new strategy: combine PARPi with OXPHOS or NAD + synthesis inhibitors. Metformin and IACS-010759 are promising candidates in this space. Some PARPi-resistant tumors restore homologous recombination. This can occur through genetic changes, such as BRCA reversion mutations. But it can also happen through metabolic remodeling. By increasing ATP and nucleotide pools, tumor cells can rebuild DNA repair functions. For example, cells increase flux through one-carbon metabolism and serine/glycine biosynthesis [77] , [78] . These pathways provide components for nucleotide synthesis and methylation reactions. They also generate NADPH, which supports redox balance. Inhibiting these metabolic programs may prevent resistance. For example, SHMT2 inhibitors, which block serine-driven one-carbon metabolism, show potential in reversing PARPi resistance. Glutamine is critical for de novo nucleotide synthesis. It donates nitrogen for purine and pyrimidine production. During PARPi stress, surviving cells may increase glutamine uptake and metabolism to restore DNA repair capacity [79] , [80] . Some single-cell studies have shown that glutamine-dependent metabolic states are enriched in PARPi-resistant clones. These cells upregulate GLS and glutamine transporters. Blocking glutamine utilization in these cells increases PARPi efficacy. This again underscores glutamine’s importance—not just in energy production, but in resistance to DNA-targeting agents. Combining PARP inhibitors with glutaminase inhibitors could prevent or delay resistance. In conclusion, resistance to PARPi involves restoring NAD + levels, boosting mitochondrial OXPHOS, and increasing nucleotide synthesis through glutamine and one-carbon metabolism. Targeting these pathways offers a way to enhance and prolong PARPi responses in ovarian cancer. Beyond chemotherapy and PARPi, ovarian cancer is also treated with anti-angiogenic agents and, increasingly, immunotherapies. Both approaches aim to modify the tumor environment or activate the immune system. However, metabolic adaptation can undermine their effectiveness. Anti-angiogenic agents like bevacizumab inhibit vascular endothelial growth factor (VEGF) [81] , [82] . This reduces blood vessel formation in tumors. As a result, tumors experience hypoxia and nutrient deprivation. In response, cancer cells stabilize HIF-1α and shift to anaerobic glycolysis. They reduce mitochondrial activity and increase glucose consumption. This adaptation helps them survive low oxygen conditions. It also increases lactate production, which acidifies the environment. The acidic TME supports invasion and suppresses immune cells. Moreover, glycolysis allows tumor cells to keep growing even with limited blood supply. In this way, anti-angiogenic therapy can inadvertently promote metabolic resistance. Targeting glycolysis or HIF-1 signaling in combination with anti-VEGF agents may improve efficacy. Inhibitors of LDHA or MCT4 can reduce lactate production and limit immune suppression [19] , [83] . Another resistance mechanism involves lipid metabolism. In hypoxic and nutrient-poor environments, tumor cells seek alternative fuels. They often turn to fatty acids supplied by adipocytes in the tumor stroma. Adipocytes release fatty acids through lipolysis. Cancer cells then import these fats using CD36 or FABP4 and metabolize them via FAO [60] . This switch supports energy production under angiogenesis blockade. Studies in ovarian cancer models show that FAO is upregulated after anti-VEGF treatment. Blocking FAO with CPT1 inhibitors like etomoxir can reduce tumor growth in this setting [84] . Lipid fueling from the TME represents a metabolic escape route. Interrupting this process may restore the effectiveness of anti-angiogenic therapies. Tumor metabolism doesn’t only affect cancer cells. It also impacts the immune landscape. Highly glycolytic tumors produce excess lactate. This acidifies the TME and suppresses immune effector cells. CD8 + T cells and NK cells are particularly vulnerable. They need glucose to function. In a glucose-deprived, lactate-rich environment, they become exhausted. Their ability to kill tumor cells decreases. Checkpoint inhibitors, like anti-PD-1 or anti-CTLA-4, aim to reinvigorate these cells. But if the metabolic environment remains hostile, these drugs may fail. This is one reason why immunotherapy has shown limited benefit in ovarian cancer to date. Modulating tumor metabolism can enhance immunotherapy. Inhibiting glycolysis in tumor cells may reduce lactate levels and improve immune function. Combining metabolic therapy with immunotherapy may thus yield better results. In ARID1A-deficient ovarian cancers, this approach has shown promise [85] , [86] . Glutaminase inhibition not only suppressed tumor growth but also enhanced immune responses. The combination of CB-839 with anti-PD-L1 antibodies produced synergistic effects in preclinical models. In summary, resistance to anti-angiogenic and immune therapies is deeply tied to metabolic adaptation. Tumors respond to hypoxia by switching to glycolysis. They scavenge lipids from adipose tissue and suppress immune cells through lactate production. Disrupting these processes may improve therapeutic outcomes. Across treatment modalities, glutamine emerges as a convergent metabolic backbone that sustains therapy-resistant states in ovarian cancer. Beyond serving as a carbon source for anaplerosis, glutamine-derived nitrogen supports de novo nucleotide synthesis and, together with cystine imported via system Xc − , fuels glutathione (GSH) biosynthesis. In this way, glutamine metabolism simultaneously maintains mitochondrial ATP production, buffers chemotherapy-induced reactive oxygen species (ROS), and underpins DNA repair capacity—three interconnected pillars of resistance to platinum agents, PARP inhibitors and anti-angiogenic therapy [87] , [88] . In platinum-based chemotherapy, glutaminase (GLS)-driven conversion of glutamine to glutamate replenishes tricarboxylic acid (TCA) intermediates while supplying the glutamate required for GSH synthesis, thereby neutralizing ROS and attenuating DNA crosslink–induced cell death [87] , [88] . PARP inhibitor resistance is likewise reinforced by glutamine-fuelled respiration and nucleotide biosynthesis, which facilitate NAD + regeneration and sustain DNA repair programs despite PARP blockade. Under anti-angiogenic pressure, where hypoxia and glucose deprivation are prominent, glutamine becomes a critical alternative fuel that preserves mitochondrial membrane potential and bioenergetic homeostasis, allowing tumor cells to tolerate vascular-targeted stress. These functional roles are further strengthened by recurrent genetic and epigenetic alterations that hard-wire glutamine dependence in specific ovarian cancer subtypes. Loss of ARID1A and overexpression of HNF1B in clear cell and endometrioid carcinomas reprogram central carbon metabolism away from glucose oxidation and toward glutamine-driven anaplerosis and GSH-centred antioxidant defenses. In high-grade serous tumors, oncogenic MYC and PI3K–Akt–mTOR signalling upregulate glutamine transporters and GLS, thereby reinforcing glutamine addiction and linking proliferative signalling to therapy-resistant metabolic states. Such lineage- and genotype-specific programs help explain why certain histologic and molecular subsets exhibit particular sensitivity to pharmacologic glutaminase inhibition. At the tumor–host interface, glutamine metabolism also integrates microenvironmental and immunologic components of resistance. Cancer-associated fibroblasts and other stromal cells can secrete glutamine or glutamine-rich metabolites, buffering fluctuations in systemic nutrient supply and sustaining tumor GLS flux under treatment. Conversely, intense glutamine consumption by ovarian cancer cells deprives infiltrating T cells of this essential amino acid, promoting T-cell dysfunction and contributing to an immune-suppressed niche. Taken together, these findings argue that glutamine should be viewed not merely as one auxiliary fuel, but as a nodal driver of chemoresistance and targeted-therapy escape. Rational combinations that pair glutaminase or glutamine-transporter inhibitors with platinum agents, PARP inhibitors or anti-angiogenic drugs—particularly in ARID1A-mutant or otherwise glutamine-addicted subtypes—may erode both metabolic flexibility and redox buffering, while strategies that rebalance glutamine availability in favour of effector lymphocytes could enhance the efficacy of immunotherapy-based regimens. A concise overview of the major metabolic adaptations that underlie resistance to current therapeutic modalities in ovarian cancer is provided in Table 1 . Table 1 Metabolic adaptations underlying therapy resistance in ovarian cancer. Therapy modality Principal metabolic adaptations Key molecular drivers / pathways Functional consequences for resistance Potential metabolic interventions Platinum-based chemotherapy Increased mitochondrial OXPHOS and biogenesis; shift toward fatty acid β-oxidation (FAO); enhanced glutamine anaplerosis and glutathione (GSH) synthesis; greater flexibility in fuel usage under stress Upregulation of CPT1A and other FAO enzymes; increased GLS expression and glutamine transporters; induction of SLC7A11 and cystine import; activation of antioxidant pathways Sufficient ATP supply to support DNA damage repair; buffering of chemotherapy-induced ROS; maintenance of redox homeostasis; survival in glucose-limited conditions Combination of platinum agents with OXPHOS inhibitors (e.g., metformin, IACS-010759) or TCA cycle disruptors; CPT1/FAO inhibition; targeting GLS or SLC7A11/GSH synthesis to weaken antioxidant defenses PARP inhibitors (PARPi) Restoration of NAD + pools through the salvage pathway; increased mitochondrial respiration; upregulated one-carbon metabolism and de novo nucleotide synthesis; enhanced glutamine utilization NAMPT–NMNAT1–mediated NAD + salvage; elevated ETC activity; activation of serine/glycine and one-carbon metabolism (e.g., SHMT2, MTHFD2); GLS-dependent glutamine flux Maintenance of PARP activity and alternative DNA repair; restoration of homologous recombination capacity; increased dNTP availability; improved tolerance to PARPi-induced DNA damage Rational combinations of PARPi with NAD + salvage inhibitors, OXPHOS inhibitors, or one-carbon pathway blockers; co-targeting glutamine metabolism (GLS inhibitors) to restrict nucleotide synthesis and enhance PARPi efficacy Anti-angiogenic therapy (e.g., anti-VEGF) HIF-1α–driven switch to anaerobic glycolysis; increased glucose uptake and lactate production; greater reliance on FAO using adipocyte-derived lipids in hypoxic, nutrient-poor niches Stabilization of HIF-1α; induction of glycolytic enzymes (HK2, PFK1, LDHA) and MCT4; upregulation of lipid transporters (CD36, FABP4) and CPT1 Survival under hypoxia and limited perfusion; acidification of the TME and promotion of invasion; metabolic escape via adipocyte-to-tumor lipid fueling; reduced effectiveness of VEGF blockade Combined targeting of glycolysis (LDH or MCT inhibitors) with anti-angiogenic agents; inhibition of FAO (CPT1 inhibitors) or adipocyte–tumor lipid transfer (CD36/FABP4 blockade) to prevent metabolic escape Immunotherapies (e.g., immune checkpoint blockade) Highly glycolytic tumor phenotype with excessive lactate secretion; nutrient competition for glucose and glutamine between tumor and immune cells; lipid-driven metabolic reprogramming of suppressive myeloid and regulatory T cells Tumor cell LDHA and MCT1/4 activity; high expression of GLUT1 and glycolytic enzymes; GLS-dependent glutamine consumption; FAO-related pathways in immunosuppressive myeloid cells and Tregs Acidic, lactate-rich TME that impairs CD8 + T-cell and NK-cell function; metabolic exhaustion of effector lymphocytes; blunted responses to checkpoint inhibition Metabolic co-targeting strategies that combine checkpoint inhibitors with glycolysis or lactate-transport inhibitors; GLS inhibition to reduce tumor glutamine consumption and improve immune cell fitness; modulation of FAO in suppressive immune subsets Metabolic adaptations underlying therapy resistance in ovarian cancer. Collectively, these observations strongly implicate metabolic rewiring as a central layer of therapy resistance; however, the strength of evidence is not uniform across mechanisms. Many resistance-associated metabolic signatures have been defined retrospectively in small, heterogeneous patient cohorts or in heavily passaged cell lines, in which prior treatment selection and clonal evolution may confound causal interpretation. Commonly used in vitro models rarely capture the full spectrum of prior therapies, the contribution of stromal and immune compartments, or host-specific factors such as comorbidities and diet that shape tumor metabolism in the clinic. Prospective translational studies and interventional trials that integrate serial metabolic profiling, pharmacodynamic biomarkers, and functional imaging will therefore be required to determine which metabolic adaptations are bona fide drivers of resistance and which are merely downstream correlates of other resistance programs.

Jingyi C, Bingzheng Z, and Xiaoman L (Corresponding authors): the main article conception, data collection and processing, funding acquisition. Xiaoyu G and Zongang L: writing, reviewing, and editing.

Ovarian cancer remains one of the most challenging malignancies to treat. Despite advances in surgery, chemotherapy, and targeted therapies, most patients eventually experience disease recurrence. At the heart of this problem lies therapy resistance. Understanding how tumors adapt and survive under treatment pressure is essential for overcoming this obstacle. One key mechanism of resistance is metabolic adaptation ( Fig. 5 ). This process allows cancer cells to change how they generate energy, synthesize building blocks, and maintain redox balance. When therapies attempt to disrupt cell growth or induce stress, ovarian cancer cells do not passively succumb. Instead, they respond by rewiring their metabolism. They shift from one nutrient source to another. They reconfigure energy production systems. They activate antioxidant pathways. All of these changes help them escape treatment effects and continue growing. Fig. 5 Conceptual summary of metabolic vulnerabilities and therapeutic targeting opportunities in ovarian cancer. This schematic provides an integrative, conceptual overview of key metabolic adaptations—encompassing glycolysis, mitochondrial oxidative phosphorylation (OXPHOS), glutamine metabolism, lipid utilization, and redox regulation—that support tumor survival and therapy resistance in ovarian cancer. The figure highlights potential metabolic nodes and intervention points that have been implicated across different therapeutic contexts, including platinum-based chemotherapy, PARP inhibition, anti-angiogenic therapy, and immunotherapy. Conceptual summary of metabolic vulnerabilities and therapeutic targeting opportunities in ovarian cancer. This schematic provides an integrative, conceptual overview of key metabolic adaptations—encompassing glycolysis, mitochondrial oxidative phosphorylation (OXPHOS), glutamine metabolism, lipid utilization, and redox regulation—that support tumor survival and therapy resistance in ovarian cancer. The figure highlights potential metabolic nodes and intervention points that have been implicated across different therapeutic contexts, including platinum-based chemotherapy, PARP inhibition, anti-angiogenic therapy, and immunotherapy. Metabolic adaptation is not random. It follows specific patterns depending on tumor type, genetic alterations, and microenvironmental conditions. Some cells increase glycolysis. Others lean heavily on mitochondrial OXPHOS [21] , [111] . Many rely on glutamine or fatty acids to meet their energy and biosynthetic needs. Importantly, these shifts are not just side effects. They are central to how resistance evolves. HGSOC, for example, can show either a glycolytic or OXPHOS-dominant phenotype. The balance between these two influences how a tumor responds to platinum chemotherapy or PARP inhibitors. Meanwhile, OCCC often exhibits glutamine addiction and antioxidant upregulation, driven by ARID1A loss and HNF1β activation [112] , [113] . These subtypes underscore that metabolic plasticity is not a uniform property of ovarian cancer but reflects distinct evolutionary strategies adopted by different tumors. Recognizing metabolic adaptation as a core resistance mechanism shifts the therapeutic paradigm beyond a traditional focus on DNA damage or apoptotic pathways, and towards energy flux, nutrient signalling and intracellular redox balance as key therapeutic battlegrounds. This perspective naturally motivates metabolic interventions: agents targeting glycolysis (e.g. 2-deoxyglucose, LDH inhibitors), OXPHOS (metformin, IACS-010759), glutamine metabolism (GLS1 inhibitors such as CB-839), and lipid metabolism, particularly in metastatic disease that exploits omental fat as an energy reservoir. Increasingly, combination approaches seek to block multiple escape routes simultaneously—for example, co-inhibiting glutamine metabolism and fatty acid oxidation, or pairing metabolic inhibitors with chemotherapy or PARP inhibitors to create context-specific synthetic lethality and drive energetic collapse or intolerable oxidative stress. However, these strategies confront fundamental challenges. The same metabolic pathways that sustain tumor cells are also essential for normal tissues, making selectivity and therapeutic window key concerns. tumors differ not only at the genomic level but also in their metabolic phenotypes; therefore, metabolic profiling—encompassing gene-expression patterns, enzyme activity and metabolite levels—is critical to identify pathway dependencies and enable rational patient stratification. A tumor with high GLS1 expression and ARID1A loss, for example, may be particularly susceptible to glutaminase inhibition, whereas a tumor with robust mitochondrial respiration may be more vulnerable to OXPHOS inhibitors. To support such decisions, robust biomarkers are needed, including molecular markers (e.g. metabolic enzyme or transporter expression), circulating metabolites in blood or ascites, and imaging-based readouts such as hyperpolarized MRI or PET tracers that report on metabolic flux. These tools will not only guide initial therapy selection but also help monitor how tumors adapt under treatment pressure. Equally important is the design of clinical trials capable of capturing these dynamics. Many metabolic inhibitors remain in preclinical or early-phase development, and their evaluation should extend beyond conventional tumor shrinkage to include metabolic endpoints, biomarker-driven stratification and, where feasible, adaptive trial designs that allow treatment modification in response to evolving metabolic profiles. Realising the potential of metabolism-informed therapy will require effective translation into clinical practice: integrating metabolic testing into oncology workflows, educating clinicians about tumor energetics, and building multidisciplinary collaborations among basic scientists, clinicians and bioinformaticians. Support from funding agencies, journals and professional societies to prioritise functional metabolic biology alongside genomics will be essential to move these concepts from theory to tangible benefit for patients. Looking ahead, a major priority is to translate descriptive metabolic knowledge into clinically actionable patient stratification. This will likely depend on prospectively validating composite biomarker panels that integrate metabolism-related gene-expression and proteomic signatures, expression or activity of key enzymes and transporters, functional metabolic imaging (e.g. glucose, lipid and glutamine tracers or hyperpolarized MRI) and circulating metabolite or lipid profiles. A concrete, testable hypothesis is that multi-parametric “metabolic scores” combining these layers will outperform single markers in predicting platinum responsiveness, PARP inhibitor benefit and susceptibility to specific metabolic combinations. Such scores could then be embedded into biomarker-enriched early-phase trials that co-target, for example, PARP and OXPHOS in OXPHOS-high, DNA-repair–deficient tumors, glutamine metabolism in ARID1A-deficient or platinum-resistant disease, or fatty acid oxidation in omental metastases with strong adipocyte–tumor lipid coupling. At the same time, future research must better account for the spatial and temporal heterogeneity of the ovarian cancer metabolic ecosystem and the constraints of systemic physiology. Single-cell and spatial multi-omics, coupled with advanced imaging and physiologically relevant co-culture or organoid models, will be essential to map metabolic crosstalk between tumor cells, immune cells, fibroblasts, endothelial cells and adipocytes, and to design regimens that reprogram the entire microenvironment rather than cancer cells alone. Parallel efforts are needed to define therapeutic windows and long-term safety, particularly in patients with metabolic comorbidities. Ultimately, the field should aim to move from enumeration of metabolic pathways to implementation of a truly metabolism-informed precision oncology framework—one in which dynamic metabolic profiling guides patient selection, treatment combinations and sequencing. Achieving this vision will require close collaboration between basic scientists, systems biologists and clinicians, but holds real promise for reducing relapse and improving survival for women with ovarian cancer. Ultimately, the goal is clear. We want to turn metabolic vulnerabilities into therapeutic opportunities. We want to use the tumor’s adaptability against itself. By depriving cancer cells of their preferred fuels—or by blocking their ability to switch—we may finally overcome therapy resistance. Ovarian cancer is a formidable enemy. But it is not invincible. With smarter, metabolism-informed strategies, we have a chance to outmaneuver it. We are now entering an era where targeting how cancer cells eat, breathe, and adapt is just as important as targeting how they grow or mutate. The challenge is large. But so is the potential. If we get this right, metabolic therapies could reshape the future of ovarian cancer treatment—and offer new hope for patients facing resistant disease.

Targeting the altered metabolism of OC is emerging as a promising therapeutic approach. As we’ve seen across subtypes, cancer cells rely on key metabolic processes—glycolysis, OXPHOS, glutamine metabolism, lipid utilization, and redox balance—to fuel survival and resist treatment. These metabolic programs are not just side effects of transformation. They are central to the tumor’s ability to adapt and escape therapy. By disrupting these adaptations, we can uncover therapeutic vulnerabilities that are otherwise concealed beneath genetic and phenotypic plasticity. This section reviews six major categories of metabolic targeting strategies and their clinical implications. Glycolysis is often upregulated in OC, even under normoxic conditions. This metabolic behavior, known as the Warburg effect, supports rapid ATP production and contributes to an acidified TME. Acidification suppresses immune cell activity and facilitates tumor growth and therapy resistance. Several inhibitors have been developed to target this pathway. These include LDH inhibitors, which block the conversion of pyruvate to lactate; MCT inhibitors, which prevent lactate export and intracellular pH regulation; and 2-deoxyglucose (2-DG), a glucose analog that disrupts glycolysis at its early steps [9] , [102] . In preclinical models, these agents impair tumor cell growth and sensitize cells to other treatments. Importantly, glycolysis inhibition may also reprogram the TME. High lactate levels in tumors suppress T and NK cell function, while acidic pH dampens immune checkpoint therapy. Reducing lactate can restore immune activity and enhance responses to immunotherapy. In addition, lactate drives angiogenesis, so inhibiting its production may disrupt tumor vasculature. Despite promising preclinical data, glycolysis inhibitors often face limited efficacy in monotherapy due to metabolic plasticity. Their optimal use likely lies in combination therapies that exploit glycolytic dependence and immune modulation. While many tumors rely on glycolysis, some, especially those that are chemoresistant, exhibit a dependency on OXPHOS. These tumors tend to have high mitochondrial respiration, robust ATP generation, and active DNA repair mechanisms. OXPHOS provides not only energy but also supports NAD + regeneration and redox stability—factors critical during chemotherapy-induced stress. Inhibitors targeting OXPHOS include metformin, which impairs mitochondrial complex I, and IACS-010759, a more potent inhibitor with enhanced tumor selectivity [21] , [103] . Other compounds like devimistat disrupt mitochondrial metabolism at the level of the TCA cycle. By reducing ATP and increasing ROS, these agents create cellular stress that is especially harmful to resistant cells. When combined with platinum-based agents or PARP inhibitors, OXPHOS inhibitors can induce synthetic lethality. This is particularly effective in HRD tumors, where the repair of therapy-induced DNA damage is already impaired. In these cases, mitochondrial inhibition may close the final escape route, pushing the tumor past its tolerance threshold. Importantly, OXPHOS targeting may also prevent resistance emergence if used in early combination regimens. However, systemic effects must be carefully managed, given the universal reliance on mitochondria for cellular energy. Glutamine addiction is particularly pronounced in ARID1A-mutant and platinum-resistant ovarian cancers, in which GLS-driven glutamine catabolism sustains TCA anaplerosis and glutathione-based redox buffering. These features render the glutamine pathway an attractive therapeutic target. Glutaminase (GLS1) inhibitors such as CB-839 (telaglenastat) have demonstrated robust preclinical activity, reducing tumor burden, resensitizing ARID1A-deficient and platinum-resistant models to cytotoxic agents, and enhancing anti-tumor immune responses. By collapsing glutamine-fuelled antioxidant defenses and metabolic flexibility, these agents create a synthetic lethal context in genetically defined subsets. Early-phase clinical trials are currently evaluating the safety and combinatorial potential of CB-839 in ovarian and other solid tumors, positioning glutamine pathway inhibition as a promising strategy for biomarker-guided, stratified therapy [88] , [104] . Lipid metabolism is another key area of adaptation in OC, especially during metastasis to lipid-rich environments like the omentum. Tumors in these niches shift toward FAO to sustain energy production under nutrient stress. This switch also supports resistance to chemotherapy. Key regulators include CPT1, which facilitates fatty acid entry into mitochondria, and FASN, which drives lipid synthesis. Inhibiting CPT1 with agents such as etomoxir disrupts FAO and energy generation, while targeting FASN impairs membrane biosynthesis and signaling [29] , [60] . These interventions reduce metastatic spread and sensitize tumors to platinum agents. Metastatic OC cells frequently upregulate lipid transporters like FABP4 and CD36 to scavenge fatty acids from surrounding adipocytes. This lipid uptake feeds mitochondrial oxidation, helping tumors resist stress. Inhibiting these pathways reduces peritoneal colonization and disrupts mitochondrial energy flux. FAO is also critical for redox balance, generating NADPH to detoxify ROS. Thus, FAO inhibition can both starve the tumor and increase its oxidative burden. Lipid metabolism inhibitors, especially in the metastatic setting, offer dual benefits—disrupting growth and enhancing chemotherapy response. Oxidative stress is a double-edged sword in cancer. While low ROS levels support signaling and growth, excessive ROS can trigger cell death. Many OC cells elevate their antioxidant capacity to neutralize therapy-induced ROS and maintain redox homeostasis. Targeting this buffer system creates a therapeutic opportunity. Drugs like APR-246, which reactivates p53 and depletes GSH, and buthionine sulfoximine (BSO), which inhibits GSH synthesis, disrupt this balance [105] , [106] . Additionally, xCT inhibitors like sulfasalazine block cystine uptake, preventing GSH production. These agents sensitize tumors to chemotherapy by weakening their antioxidant shield. A related strategy involves ferroptosis induction. Ferroptosis is a form of cell death driven by iron-dependent lipid peroxidation. It is particularly effective in OCCC, which often has high iron levels and redox instability. Drugs such as erastin and artesunate initiate this process, overwhelming the cell’s capacity to prevent lipid damage. Targeting redox homeostasis can thus reintroduce vulnerability in resistant tumors. It also complements DNA-damaging therapies by increasing cellular stress beyond tolerable levels. Given the complexity and redundancy of cancer metabolism, single-agent metabolic therapies are rarely sufficient. The future lies in combination approaches that target multiple pathways or rationally pair metabolic inhibitors with existing cytotoxic, targeted or immune therapies. However, the success of such strategies critically depends on precise patient stratification. Rather than assigning fixed labels such as “glycolytic” or “glutamine-addicted” to a given histotype, it is more useful to recognize that ovarian cancer subtypes exhibit characteristic but heterogeneous metabolic propensities: high-grade serous ovarian cancer (HGSOC) can shift between glycolysis- and OXPHOS-biased states depending on its genomic background and treatment history, ovarian clear cell carcinoma (OCCC) frequently displays enhanced glutamine and redox metabolism driven by lineage- and ARID1A/HNF1B-associated programs, and endometrioid ovarian carcinoma (EnOC) often shows intermediate, PI3K–Akt–mTOR–influenced nutrient usage patterns. Matching therapies to these dynamic phenotypes is essential. For example, glycolysis inhibitors such as 2-deoxy-D-glucose may be most effective in highly glycolytic tumors, whereas biguanides like metformin are better suited to OXPHOS-dependent disease. Glutaminase inhibitors (e.g., CB-839) are particularly attractive in ARID1A-mutant or platinum-resistant tumors that rely on glutamine, and CPT1 inhibitors may limit lipid-driven metastasis and help restore platinum sensitivity. Stratification in this context relies on biomarkers—gene-expression signatures, metabolite levels and protein activity states—that can guide the selection and sequencing of metabolic drugs and their partners. Advanced tools such as hyperpolarized MRI and serum metabolomics panels are being developed to monitor metabolic activity in real time. These technologies not only support baseline classification but also reveal therapy-induced shifts in tumor metabolism, enabling adaptive treatment planning. As these approaches mature, personalized metabolic therapy will become increasingly feasible, and the most promising path forward is likely to be a flexible, stratified strategy that aligns combination regimens with each tumor’s evolving metabolic fingerprint. To provide a practical, subtype-oriented overview that links metabolic susceptibilities to candidate therapeutic strategies, we summarize potential targeted combination regimens in Table 3 . Table 3 Ovarian cancer subtypes, metabolic susceptibilities, and candidate targeted combination therapies. Ovarian cancer subtype / context Dominant metabolic susceptibility Representative metabolic inhibitor(s) or strategy Rational combination partner(s) Brief comments / translational notes HGSOC, glycolysis-dominant (high LDHA/HK2, lactate-rich TME) Dependence on aerobic glycolysis and lactate export LDH inhibitors; MCT1/4 inhibitors; 2-deoxyglucose Platinum-based chemotherapy; anti-VEGF therapy; immune checkpoint inhibitors Glycolysis/lactate blockade may resensitize chemoresistant tumors, reduce acidosis, and improve T/NK-cell function; clinical data in OC still emerging. HGSOC, OXPHOS-high (mitochondria-active phenotype) Reliance on mitochondrial complex I and TCA cycle for ATP and NAD + regeneration Metformin; IACS-010759; devimistat (CPI-613) Platinum-based chemotherapy; PARP inhibitors in HRD/BRCA-mutated disease OXPHOS inhibition increases ROS and impairs NAD + -dependent repair, potentially enhancing platinum and PARPi responses in OXPHOS-addicted tumors. HGSOC with HRD / BRCA1/2 mutation treated with PARPi NAD + salvage and OXPHOS-supported PARPi resistance; enhanced one-carbon and nucleotide metabolism NAMPT/NMNAT1 inhibition; OXPHOS inhibitors; one-carbon metabolism inhibitors PARP inhibitors (olaparib, niraparib, etc.); platinum agents Dual targeting of PARP and NAD + / OXPHOS / one-carbon pathways may prevent or reverse PARPi resistance; requires careful toxicity monitoring. Ovarian clear cell carcinoma (OCCC) with ARID1A loss Glutamine addiction via GLS1 upregulation GLS1 inhibitor (e.g., telaglenastat/CB-839) Platinum-based chemotherapy; immune checkpoint inhibitors; anti-angiogenic agents ARID1A-deficient OCCC shows synthetic lethality to GLS inhibition; combining with chemo or ICI may overcome intrinsic chemoresistance. OCCC / EnOC in iron-rich, endometriosis-derived microenvironment Ferroptosis sensitivity and redox stress vulnerability Ferroptosis inducers (e.g., erastin analogs, artesunate); xCT/GSH pathway inhibitors Platinum-based chemotherapy; radiotherapy; redox-modulating agents Iron overload and chronic oxidative stress create a ferroptosis-prone state; rational to combine ferroptosis induction with DNA-damaging therapies. Endometrioid ovarian carcinoma (EnOC) with PI3K/Akt/mTOR activation and/or ARID1A mutation PI3K/Akt-driven glycolysis and lipogenesis; context-dependent glutamine dependence PI3K/mTOR inhibitors; GLS inhibitors (in glutamine-addicted subset) Platinum-based chemotherapy; hormonal therapy; PARP inhibitors in HR-altered cases Dual targeting of PI3K/Akt/mTOR and metabolic pathways may counteract acquired resistance; requires biomarker-guided patient selection. Mucinous ovarian carcinoma (MOC) with KRAS mutation Hexosamine biosynthetic and glycosylation pathways supporting mucin production Inhibitors of hexosamine pathway or protein glycosylation (experimental); MAPK-pathway inhibitors MAPK-targeted therapy (e.g., MEK inhibitors); GI-type chemotherapy regimens Metabolic targeting remains exploratory; combination of MAPK blockade with glycosylation/hexosamine modulation is a hypothesis-generating avenue. Low-grade serous carcinoma (LGSOC), hormone- and MAPK-driven MAPK-linked metabolic rewiring (moderate glycolysis/OXPHOS); hormone-responsive metabolism MEK inhibitors; potential adjunct OXPHOS/glycolysis modulators Endocrine therapy (aromatase inhibitors, etc.); platinum-based chemotherapy MEK inhibition is an established strategy; adding metabolic agents may enhance efficacy, but OC-specific evidence is currently sparse. Ovarian cancer subtypes, metabolic susceptibilities, and candidate targeted combination therapies. Despite the compelling rationale for targeting metabolic vulnerabilities in ovarian cancer, several overarching challenges temper enthusiasm and must be acknowledged to provide a balanced view of the field. Foremost among these is the narrow therapeutic window created by the ubiquity of core metabolic pathways in normal tissues. Glycolysis, mitochondrial oxidative phosphorylation, glutamine metabolism and fatty acid oxidation are essential not only for tumor cells but also for highly energy-demanding organs such as the heart, brain and skeletal muscle, as well as for proliferating immune and hematopoietic cells. On-target, off-tumor toxicities—including fatigue, myopathy, neurocognitive changes, gastrointestinal intolerance, hepatotoxicity and myelosuppression—have limited dose intensity for several first-generation metabolic inhibitors, and may be exacerbated in patients with pre-existing metabolic comorbidities such as obesity, diabetes or cardiovascular disease. A critical future task will therefore be to define dosing schedules and combination strategies that modulate, rather than completely shut down, specific metabolic circuits in the tumor while preserving systemic homeostasis. A second major hurdle is the intrinsic plasticity and redundancy of cancer metabolism, which can undermine the durability of responses to metabolic therapies. As discussed throughout this review, ovarian tumors are capable of switching between glycolysis and OXPHOS, rerouting carbon from glucose to glutamine or fatty acids, and dynamically reconfiguring redox systems in response to therapeutic pressure. In this context, pharmacological inhibition of a single enzyme or pathway often leads to compensatory upregulation of alternative routes, re-establishing biosynthetic and energetic fluxes and ultimately fostering resistance to the metabolic inhibitor itself. Early-phase clinical trials of glycolytic, OXPHOS, glutaminase and fatty-acid-metabolism inhibitors have, in several cases, yielded mixed results—showing acceptable safety but modest single-agent activity or transient responses—highlighting the limitations of preclinical models that do not fully capture human tumor heterogeneity, stromal crosstalk and host physiology. More predictive experimental systems that integrate patient-derived organoids, co-culture with immune and stromal cells, and in vivo models with longitudinal metabolic imaging will be essential to de-risk clinical development. Finally, the influence of host physiology and the tumor microenvironment poses additional layers of complexity that complicate both efficacy and safety. Nutrient availability, hormonal milieu, inflammatory status and organ-specific niches such as the omentum or endometriotic lesions all shape tumor metabolism in ways that are difficult to recapitulate in vitro. Systemic manipulation of metabolism may inadvertently impair antitumor immunity—for example, by depriving effector T cells of glucose or glutamine—or exacerbate cachexia and frailty in already vulnerable patients. These considerations argue for a shift from empiric, one-size-fits-all use of metabolic drugs toward biomarker-guided designs in which pharmacokinetic and pharmacodynamic profiling, dynamic metabolic imaging and composite biomarker panels are used to select patients, tailor dosing, and monitor on-target engagement. Acknowledging these challenges does not diminish the promise of metabolic targeting; rather, it provides a realistic framework within which rational combinations, careful trial design and cross-disciplinary collaboration will be required to translate metabolic insights into durable clinical benefit for women with ovarian cancer.

Ovarian cancer continues to pose one of the most pressing challenges in gynecologic oncology. It ranks among the top causes of cancer-related deaths in women worldwide [1] , [2] , [3] . Despite progress in diagnosis and treatment, outcomes remain poor. Most patients are diagnosed at advanced stages due to the disease’s insidious onset. Standard therapy includes cytoreductive surgery followed by platinum-based chemotherapy [4] , [5] . While many patients initially respond to treatment, relapse is common. More than 70% of women with advanced disease eventually develop recurrent tumors that are less responsive to subsequent therapy. This cycle of remission and relapse highlights the urgent need for new therapeutic strategies. One emerging area of focus is cancer metabolism. Tumor cells do not follow the same rules as normal cells. They rewire their metabolic processes to meet the demands of rapid growth, survival under stress, and resistance to therapy. This process is called metabolic reprogramming [6] , [7] . It allows cells to generate energy quickly, produce the molecules needed for replication, and neutralize cellular stress. In ovarian cancer, metabolic adaptation plays a central role in disease progression and therapy failure [8] , [9] . Chemotherapy, radiation, and targeted therapies apply significant stress on cancer cells. These treatments damage DNA, impair cell division, and induce oxidative stress [10] , [11] . However, many ovarian tumors adapt to these conditions. They shift their metabolism to survive. Some switch from glucose to fatty acids. Others increase glutamine usage or activate antioxidant pathways. This flexibility is called metabolic plasticity. It is one of the most important mechanisms by which cancer cells escape treatment. At the core of this plasticity are several interconnected metabolic pathways. These include glycolysis, mitochondrial oxidative phosphorylation (OXPHOS), glutamine metabolism, and lipid metabolism [12] , [13] . Each of these pathways provides energy and building blocks. They also support redox control and immune evasion. When one pathway is blocked or stressed, cancer cells often activate another. This redundancy makes therapy challenging [14] , [15] . Yet it also presents opportunities. If we understand how these switches happen, we can identify vulnerabilities. One key pathway is aerobic glycolysis. In this process, cells convert glucose into lactate even in the presence of oxygen [16] , [17] . This behavior, known as the Warburg effect, is common in many tumors. It allows rapid ATP generation and supports biosynthesis. It also creates an acidic tumor microenvironment (TME) that suppresses immune cells. Ovarian tumors often exhibit high glycolytic activity, driven by hypoxia-inducible factor 1 alpha (HIF-1α), the PI3K/Akt pathway, and MYC [18] , [19] . These tumors consume large amounts of glucose and produce lactate, contributing to therapy resistance and immune escape. Another important pathway is mitochondrial oxidative phosphorylation. Unlike glycolysis, OXPHOS occurs in the mitochondria and is highly efficient at producing ATP [20] , [21] . Some ovarian tumors maintain active mitochondria and rely on OXPHOS for energy. Others switch to this pathway when glycolysis is impaired. High-OXPHOS tumors often exhibit increased reactive oxygen species (ROS), which makes them more sensitive to chemotherapy. However, they also use OXPHOS to regenerate NAD + , a cofactor needed for DNA repair [22] , [23] . This links mitochondrial metabolism to resistance against DNA-damaging agents and PARP inhibitors. Glutamine metabolism also plays a central role in ovarian cancer. Glutamine is used to feed the tricarboxylic acid (TCA) cycle, produce nucleotides, and maintain redox balance through glutathione synthesis [24] , [25] . Many ovarian tumors are glutamine-addicted, particularly those with ARID1A mutations or platinum resistance. These tumors rely on glutaminase (GLS) to convert glutamine into glutamate. Blocking glutamine metabolism has shown promise in reversing resistance and enhancing immune responses [26] , [27] . Because normal cells are less dependent on glutamine, this strategy may offer a therapeutic window. Fatty acid metabolism is another critical adaptation in ovarian cancer. Tumor cells often scavenge lipids from surrounding adipocytes, particularly in the omentum where metastases frequently occur [28] , [29] . They use fatty acid oxidation (FAO) to generate energy and produce NADPH for antioxidant defense. Enzymes like carnitine palmitoyltransferase 1 (CPT1) and fatty acid synthase (FASN) play central roles in this process [9] , [30] . In cisplatin-resistant tumors, lipid metabolism becomes even more important. These cells reduce glucose uptake and rely on fatty acids for survival. Inhibiting FAO or lipogenesis can resensitize tumors to chemotherapy. These core metabolic pathways do not function in isolation. They interact with one another and with the tumor microenvironment. For example, lactate produced during glycolysis suppresses immune cell function. Fatty acids released from adipocytes fuel tumor growth. Glutamine supports not only energy production but also nucleotide synthesis and immune modulation. This metabolic network allows tumors to adapt dynamically to therapy. Importantly, metabolic behavior varies across ovarian cancer subtypes. High-grade serous ovarian carcinoma (HGSOC), which is the most common subtype, can be divided into OXPHOS-dominant and glycolysis-dominant groups [31] , [32] . These differences influence treatment response. Clear cell carcinoma (OCCC), on the other hand, often carries ARID1A mutations and is highly glutamine-dependent [33] , [34] . It also exhibits elevated antioxidant defenses. Endometrioid ovarian cancers share features with both clear cell and serous subtypes. They may rely on glutamine and show moderate glycolysis. Mucinous and low-grade serous tumors have their own unique traits. These include hormone responsiveness and reliance on mucin synthesis or MAPK-driven pathways. This subtype-specific metabolic heterogeneity has major therapeutic implications. A glycolysis inhibitor may work well in one subtype but fail in another. Similarly, glutamine-targeting drugs may be effective only in ARID1A-deficient tumors [35] , [36] . Understanding these differences is crucial. Precision medicine in ovarian cancer must go beyond genomics. It must incorporate metabolic profiling. There is growing interest in targeting these metabolic vulnerabilities. Several drugs are in development. These include inhibitors of glycolysis, OXPHOS, glutaminase, and lipid metabolism. Some have entered clinical trials. Others are being tested in preclinical models. Many shows enhanced efficacy when combined with existing therapies. For example, OXPHOS inhibitors may sensitize tumors to PARP inhibitors [37] , [38] . Glycolysis blockers may improve immune cell infiltration. Glutamine deprivation may reverse resistance to platinum chemotherapy. However, challenges remain. Metabolic inhibitors can affect healthy tissues. Tumor metabolism is dynamic and context-dependent. Cells may switch fuels under pressure. The tumor microenvironment also plays a role. Immune cells, fibroblasts, and stromal components influence nutrient availability. Adipocytes provide lipids. Hypoxia alters enzyme expression. All of these factors must be considered. This review seeks to map how metabolic rewiring supports therapy resistance in ovarian cancer. We explore how core nutrient pathways are altered in different contexts. We examine how these changes support survival, immune evasion, and treatment failure. We also highlight potential vulnerabilities that may be exploited. Our goal is to bridge basic research with clinical application. By understanding how ovarian tumors adapt their metabolism, we can identify points of intervention. These may include enzyme inhibitors, nutrient depletion strategies, or combination regimens with immunotherapy or chemotherapy. We also emphasize the need for translational tools. Biomarkers, metabolic imaging, and gene expression signatures can help stratify patients and guide treatment. In conclusion, therapy resistance in ovarian cancer is deeply linked to metabolic plasticity. Tumor cells alter their energy pathways to survive treatment. These adaptations create new targets. By disrupting these metabolic lifelines, we may develop more durable therapies. This approach holds promise for improving outcomes in one of the most treatment-resistant cancers.

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.

Ovarian cancer is not a uniform disease. It includes multiple subtypes, each with its own biology, genetic profile, and clinical behavior. These subtypes also differ in their metabolic wiring. Understanding these differences is essential. It helps us tailor treatments and identify subtype-specific vulnerabilities. Beyond histology and stage, subtype-specific metabolic wiring in OC is strongly imprinted by recurrent genomic and epigenetic alterations. Integrated genomic and epigenomic analyses from The Cancer Genome Atlas (TCGA-OV) first defined four major transcriptional subtypes of high-grade serous ovarian cancer (HGSOC)—proliferative, immunoreactive, mesenchymal and differentiated—that differ in the enrichment of glycolysis, hypoxia and lipid-metabolism pathways, and in associated copy-number events such as MYC or CCNE1 amplification [87] , [88] . Subsequent CPTAC proteogenomic profiling and other large-scale proteomic datasets confirmed that OXPHOS and glycolytic protein expression covaries with DNA-repair proficiency, chromosomal instability and platinum responsiveness, thereby mechanistically linking metabolic state to genotype-driven signalling programs [87] , [88] . In particular, a PML–PGC-1α axis has been implicated in driving an OXPHOS-high HGSOC phenotype with enhanced mitochondrial biogenesis and respiratory-chain activity, whereas OXPHOS-low tumors rely more heavily on glycolysis and show relatively greater chemoresistance, underscoring how stress-responsive transcriptional co-activators partition patients into metabolically and clinically distinct subgroups [87] , [88] . In ovarian clear cell carcinoma (OCCC), lineage transcription factor HNF1B and ARID1A-deficient SWI/SNF complexes remodel chromatin at metabolic-gene loci, driving a shift towards glycolysis, glycogen accumulation, glutathione synthesis and enhanced antioxidant capacity that supports survival in the iron-rich, inflammatory milieu of endometriosis-associated lesions [87] , [88] . Similar immune–metabolic coupling has been described in endometrioid carcinomas, where frequent alterations in ARID1A, PTEN and PIK3CA activate PI3K–Akt–mTOR signalling and reprogram glucose and lipid metabolism, and in mucinous or low-grade serous tumors, in which KRAS/BRAF/MAPK activation and hormone-receptor signalling promote glycosylation and lipid-remodelling demands. Collectively, these transcriptional and epigenetic programs provide a mechanistic framework to explain why ostensibly related OC subtypes develop divergent dependencies on glycolysis, OXPHOS, glutamine metabolism or redox homeostasis, rather than merely representing descriptive metabolic “labels”. HGSOC is the most common subtype. It accounts for roughly 70% of all ovarian cancer cases. It is aggressive and often diagnosed at an advanced stage. Patients typically receive platinum-based chemotherapy and, in some cases, PARP inhibitors. HGSOC is nearly always associated with TP53 mutations. This tumor suppressor gene is mutated in more than 95% of cases. Loss of p53 function disrupts cell cycle control and DNA damage response. Many HGSOC tumors also exhibit HRD. This is due to mutations in BRCA1, BRCA2, or other repair genes [86] , [89] . HRD tumors cannot repair double-strand DNA breaks effectively. This makes them vulnerable to PARP inhibitors, which block single-strand repair. However, not all HGSOCs are the same. Differences in their metabolic behavior create distinct subgroups. Some HGSOC tumors are glycolysis-dominant. They rely heavily on glucose for energy. These tumors show high expression of glycolytic enzymes like LDHA and HK2. They also produce high levels of lactate. This leads to an acidic microenvironment, which promotes invasion and suppresses immune responses. Other HGSOC tumors are OXPHOS-high. These tumors have high mitochondrial activity. They consume more oxygen and produce more ATP through oxidative phosphorylation. They also show high expression of ETC (electron transport chain) genes. This metabolic classification matters. OXPHOS-high tumors are often more sensitive to chemotherapy. They accumulate more ROS and are less able to manage oxidative stress. In contrast, glycolytic tumors are often more resistant. Their antioxidant systems are stronger, and they can survive harsh conditions. HGSOC tumors are not locked into one state. They show metabolic plasticity. Under therapy or environmental stress, they can switch between glycolysis and OXPHOS. This flexibility helps them survive treatment and contributes to relapse. Metabolic imaging tools are being developed to assess this flexibility. Techniques like hyperpolarized 13C-pyruvate MRI can track glycolysis in real time. These tools may help stratify patients and guide metabolic therapies. Targeted treatments are also emerging. For glycolytic tumors, glycolysis inhibitors (like 2-deoxyglucose or LDH blockers) may improve response. For OXPHOS-high tumors, mitochondrial inhibitors like metformin or IACS-010759 show promise. In summary, HGSOC is metabolically diverse. Stratifying tumors by their dominant metabolic phenotype can inform therapy. This approach moves us closer to personalized treatment in ovarian cancer. OCCC is a distinct and less common subtype. It accounts for about 5–10% of ovarian cancer cases. It often presents at an earlier stage than HGSOC but is typically more resistant to chemotherapy. OCCC is defined by unique molecular and metabolic traits. One of the hallmarks of OCCC is high expression of HNF1β (hepatocyte nuclear factor 1 beta) [90] . This transcription factor drives extensive metabolic changes. It upregulates glycolysis and glycogen synthesis. It also reduces mitochondrial respiration. As a result, OCCC cells shift toward a Warburg-like phenotype. HNF1β expression helps OCCC cells manage oxidative stress. It reduces ROS production and enhances survival under harsh conditions, such as hypoxia or chemotherapy. This makes OCCC intrinsically chemoresistant. Another key feature of OCCC is frequent ARID1A mutations. ARID1A is a chromatin remodeler. Its loss leads to altered gene expression, including genes involved in metabolism. ARID1A-deficient OCCC cells develop a strong glutamine dependence. They overexpress GLS and rely on glutamine to fuel the TCA cycle. Glutamine also supports GSH synthesis. This defends against oxidative damage and chemotherapy-induced ROS. This dependency creates a metabolic vulnerability. Targeting glutamine metabolism with GLS inhibitors like CB-839 selectively kills ARID1A-mutant OCCC cells [27] , [91] . It also enhances sensitivity to standard drugs. OCCC also shows high expression of antioxidant genes. Some tumors activate the NRF2 pathway, which boosts the expression of enzymes involved in ROS detoxification. This further supports their redox resilience. Despite their strong antioxidant systems, OCCC cells can be vulnerable to ferroptosis. Ferroptosis is an iron-dependent form of cell death triggered by lipid peroxidation. OCCC cells often exist in iron-rich environments due to their origin in endometriotic cysts. These cysts contain free iron and generate oxidative stress. Although OCCC adapts to this stress, it also becomes sensitive to further disruption. Ferroptosis inducers like erastin or artesunate may be effective in OCCC. They tip the redox balance and induce lethal lipid damage [23] , [70] . In summary, OCCC has a distinctive metabolic profile. It is glycolysis-driven, glutamine-addicted, and redox-optimized. However, it shows specific weaknesses. Targeting glutamine metabolism or inducing ferroptosis are promising strategies in this hard-to-treat subtype. Endometrioid ovarian carcinoma (EnOC) accounts for about 10% of ovarian cancer cases. It shares some features with OCCC but is a distinct entity. Many EnOC tumors arise from endometriosis and show intermediate behavior in terms of aggression and treatment response. EnOC exhibits moderate glycolytic activity [92] . It does not show the same strong Warburg phenotype as OCCC but still upregulates glycolytic enzymes. This supports growth in hypoxic or nutrient-limited environments. EnOC often shows overlap with clear cell features, particularly when it arises from endometriotic lesions [93] , [94] . Like OCCC, it may have faced oxidative stress during early development. This selects for cells with enhanced antioxidant capacity and metabolic flexibility. Unlike HGSOC, which is usually highly proliferative, EnOC tends to grow more slowly. As a result, its metabolic demands may be lower, but its ability to adapt under stress remains important. Some EnOC tumors have ARID1A mutations, like OCCC. These tumors may show similar glutamine dependence. However, this relationship is not always consistent. Studies show that GLS1 expression does not always correlate with ARID1A status in EnOC. This suggests a heterogeneous metabolic landscape. Some tumors may rely on glutamine for TCA cycling and antioxidant defense. Others may not. Stratifying EnOC tumors based on glutamine metabolism could help identify those that would benefit from GLS inhibition. EnOC tumors also exhibit oxidative stress adaptation [95] , [96] . Mutations in PIK3CA and PTEN, which are common in EnOC, can activate the mTOR pathway. This supports cell growth and enhances glycolysis and lipid synthesis. EnOC’s ability to survive in stressful environments reflects its origin. Like OCCC, it often arises in inflammatory and oxidative microenvironments. This may shape its metabolic profile and impact therapy response. In summary, EnOC shows intermediate metabolism. It combines moderate glycolysis, occasional glutamine reliance, and redox adaptation. It is generally more chemo-sensitive than OCCC but may develop resistance in recurrent settings. Targeting its metabolic traits could be useful, especially in ARID1A-mutant or glutamine-addicted cases. Two additional subtypes—mucinous ovarian carcinoma (MOC) and low-grade serous ovarian carcinoma (LGSOC)—are rarer but important. Each has unique biology and metabolic characteristics. MOC resembles gastrointestinal tumors in many ways. It is characterized by KRAS mutations and mucin production. These tumors often express genes related to the hexosamine biosynthetic pathway. This pathway converts glucose into building blocks for mucins and glycoproteins. MOC cells consume glucose not only for energy but also for mucin synthesis [96] , [97] . This creates a distinct metabolic demand. The production of mucins uses substrates like UDP-N-acetylglucosamine, which are derived from glucose and glutamine. This unique metabolism suggests that MOC may respond to inhibitors of glycosylation or hexosamine pathway enzymes. Such therapies are still experimental but represent potential strategies for a tumor type that is often chemoresistant. LGSOC is a slow-growing tumor subtype. It is less responsive to chemotherapy and more dependent on hormonal signaling. Estrogen and progesterone receptors are often expressed. This makes hormonal therapies, like aromatase inhibitors, an option. At the metabolic level, LGSOC is less studied. However, it shows lower glycolytic and OXPHOS activity compared to HGSOC. This may be due to its lower proliferation rate [98] , [99] . Instead, LGSOC appears to rely on MAPK pathway activation, driven by mutations in KRAS, BRAF, or NRAS. This pathway can influence nutrient sensing and mitochondrial metabolism. Therapies targeting the MEK/ERK axis may alter metabolism in LGSOC. These interventions may also sensitize tumors to standard drugs [100] , [101] . Combining MEK inhibitors with metabolic agents is an area of active research. Both MOC and LGSOC show distinct features. MOC is glycosylation-intensive. LGSOC is hormone-driven and MAPK-dependent. These characteristics open unique therapeutic avenues. While their metabolic profiles are not as well-defined as the other subtypes, their distinct biology suggests that customized metabolic strategies may offer benefits. Research into these areas is ongoing and needed. These mechanistic insights have increasingly been translated into quantitative classifiers that can be leveraged for patient stratification. Several groups have used TCGA-OV and external cohorts to perform non-negative matrix factorization or consensus clustering on metabolism-related transcripts, identifying robust metabolic subtypes with distinct patterns of glycolysis, amino-acid and lipid metabolism that independently predict survival, immune infiltration and putative drug response [100] , [101] . For example, immune- and metabolism-associated classifications stratify patients into subgroups that differ in T-cell–inflamed gene signatures, checkpoint expression and predicted benefit from immune checkpoint blockade, while lipid/glycolysis-based subtypes display divergent sensitivity to platinum therapy and targeted agents. Complementing these transcriptomic approaches, high-resolution proteomics across epithelial ovarian cancer histotypes (serous, clear cell, endometrioid) has delineated subtype-specific activation of fatty-acid oxidation, TCA-cycle and redox pathways, and pinpointed metabolic enzymes as potential biomarkers or therapeutic targets. Finally, emerging tissue and serum metabolomics studies reveal that OCCC and other histotypes exhibit non-overlapping metabolite fingerprints—including differences in glutamine, lipid and antioxidant metabolites—that can distinguish subtypes and correlate with prognosis, underscoring the feasibility of integrating metabolomic readouts into future risk-stratification algorithms [100] , [101] . To synthesize the subtype-specific metabolic features and therapeutic implications discussed above, we summarize the key characteristics of major ovarian cancer subtypes in Table 2 . Table 2 Subtype-specific metabolic phenotypes and vulnerabilities in ovarian cancer. Histologic subtype Dominant genetic / signaling features Key metabolic traits Therapy response / resistance patterns Potential metabolic vulnerabilities High-grade serous ovarian cancer (HGSOC) Nearly universal TP53 mutation; frequent HRD due to BRCA1/2 or other DNA repair gene alterations; activation of PI3K/Akt and other growth pathways Co-existence of glycolysis-dominant and OXPHOS-dominant phenotypes; high metabolic plasticity with the ability to switch between glycolysis and mitochondrial respiration; variable glutamine and lipid dependence OXPHOS-high tumors often more sensitive to platinum chemotherapy but may adapt by increasing antioxidant defenses; glycolysis-high subtypes exhibit stronger resistance and immunosuppressive, lactate-rich TMEs; subset responds to PARPi but may acquire metabolic resistance Glycolytic HGSOC: targeting glycolysis and lactate export (LDH/MCT inhibitors); OXPHOS-high HGSOC: mitochondrial complex I or TCA-cycle inhibitors; HRD tumors: combination of PARPi with OXPHOS or NAD + -modulating agents; biomarker-guided stratification based on metabolic gene signatures and imaging Ovarian clear cell carcinoma (OCCC) High frequency of ARID1A mutations; strong HNF1β expression; frequent activation of NRF2 and antioxidant programs; association with endometriosis Warburg-like metabolic profile with enhanced glycolysis and glycogen storage; reduced mitochondrial respiration; strong glutamine dependence and robust antioxidant capacity; adaptation to iron-rich, oxidative environments Intrinsically chemoresistant to platinum agents; capable of managing ROS and hypoxia; poor response to conventional regimens in advanced or recurrent disease Synthetic lethal vulnerability to GLS inhibition in ARID1A-deficient tumors; sensitivity to ferroptosis inducers due to iron load and redox stress; potential benefit from strategies that disable antioxidant systems (e.g., xCT/GSH targeting) combined with standard therapy Endometrioid ovarian carcinoma (EnOC) Frequently associated with endometriosis; recurrent mutations in ARID1A, PIK3CA, PTEN; activation of PI3K/Akt/mTOR signaling Moderate glycolytic activity; context-dependent glutamine utilization; evidence of oxidative stress adaptation and metabolic flexibility shaped by an inflammatory, endometriosis-derived microenvironment Generally more chemo-sensitive than OCCC but can acquire resistance at recurrence; metabolic behavior heterogeneous across cases, with subsets resembling clear-cell-like or serous-like profiles In ARID1A-mutant or glutamine-addicted EnOC, GLS inhibitors may be effective; PI3K/Akt/mTOR-driven metabolic programs suggest sensitivity to combined pathway and metabolic inhibition; redox-targeting approaches for tumors with high antioxidant tone Mucinous ovarian carcinoma (MOC) Frequent KRAS mutations and gastrointestinal-like molecular features; strong mucin production and intestinal differentiation signatures Heavy reliance on hexosamine biosynthetic and glycosylation pathways to sustain mucin and glycoprotein synthesis; utilization of glucose and glutamine as substrates for UDP-sugar production Often chemoresistant and poorly responsive to standard platinum-based regimens; behavior parallels mucinous gastrointestinal tumors more than other ovarian subtypes Potential susceptibility to inhibitors of glycosylation or hexosamine pathway enzymes; combination of MAPK-pathway inhibition with metabolic modulation to reduce mucin-driven survival; exploration of GI-type metabolic strategies adapted to ovarian context Low-grade serous carcinoma (LGSOC) Slow-growing; frequent activating mutations in KRAS, BRAF, or NRAS; strong dependence on hormone receptor and MAPK signaling Overall lower glycolytic and OXPHOS rates compared with HGSOC; metabolic program shaped by MAPK activation and hormonal cues rather than extreme proliferative demand Poor response to conventional chemotherapy; better sensitivity to hormonal therapies and MEK/ERK pathway inhibitors; metabolic features less well characterized but likely intertwined with MAPK signaling Targeting MEK/ERK signaling with or without adjunct metabolic agents (e.g., OXPHOS or glycolysis modulators); metabolism-informed combinations with endocrine therapy; need for further profiling of metabolic states to identify specific vulnerabilities Subtype-specific metabolic phenotypes and vulnerabilities in ovarian cancer. At the same time, a cautious interpretation of subtype-specific metabolic traits is warranted. Many of the proposed “glutamine-addicted” or “OXPHOS-high” phenotypes are based on relatively small numbers of tumors, rely on differing histopathologic or molecular subtype definitions across studies, and may shift substantially between primary, metastatic, and recurrent lesions. Existing multi-omics resources also tend to over-represent advanced, heavily treated high-grade serous disease, with comparatively sparse sampling of rarer histologies such as mucinous and low-grade serous carcinomas. As such, current subtype–metabolism relationships should be viewed as testable hypotheses rather than rigid rules, and will require validation in larger, prospective cohorts using harmonized classification criteria and standardized metabolic assays before they can reliably guide patient stratification.