Rewiring the Tumor: How Cancer Metabolism Is Becoming Medicine’s Next Target

…

The concept of metabolic reprogramming as a defining feature of malignancy has evolved considerably since its early description, when proliferating tumor cells were first observed to favor glycolysis even in the presence of abundant oxygen. What was once regarded as a curious biochemical quirk is now understood as a coordinated rewiring of cellular metabolism that supports every stage of tumor evolution, from initiation and growth through invasion, dissemination, and resistance to treatment. The contributions gathered in this Research Topic illustrate how this rewiring manifests across distinct cancer types and metabolic axes, and how each axis can, in principle, be converted into a therapeutic liability.

One recurring theme across the collected articles is the interdependence between metabolic pathways and canonical signaling networks. The review by Yang and colleagues on HMG-CoA reductase exemplifies this principle. Rather than functioning solely as a housekeeping enzyme for cholesterol production, HMGCR sits at a junction where lipid biosynthesis feeds into isoprenoid-dependent post-translational modification of signaling proteins, thereby influencing pathways such as Hippo, Hedgehog, and MAPK that are frequently dysregulated in tumors. This positioning explains why statins, which deplete mevalonate pathway intermediates, have shown preclinical activity across multiple models, and why the mevalonate axis has attracted sustained interest as an adjunctive target. The bidirectional relationship between HMGCR activity and TNF-α signaling adds an immunological dimension, suggesting that lipid metabolism can shape the inflammatory milieu of tumors in ways that promote immune escape. The clinical translation of statin-based strategies, however, remains complicated by dose-limiting toxicities and the difficulty of achieving sufficient pathway inhibition in tumor tissue without harming normal tissues that also rely on the mevalonate pathway.

The contribution from Zhan and colleagues on clear cell renal cell carcinoma highlights how a defining genetic lesion can dictate metabolic dependencies. The near-universal loss of VHL function in ccRCC stabilizes HIF transcription factors even under normoxic conditions, driving a pseudohypoxic program that includes glycolytic activation, lipid droplet accumulation, and enhanced glutamine utilization. The approval and clinical study of HIF-2α antagonists such as belzutifan represent a landmark demonstration that a metabolic transcriptional driver can be targeted directly in patients. The authors’ emphasis on combining such agents with ferroptosis inducers, glutamine antagonists, and immune checkpoint inhibitors reflects a broader realization that single-agent metabolic therapies are unlikely to be curative on their own, given the redundancy and plasticity of metabolic networks. Their proposal to use radiomics and lipid metabolic phenotypes to stratify patients also points toward imaging-based biomarkers that could guide metabolic therapy selection without requiring repeated invasive biopsies.

Lactate dehydrogenase isoenzymes, examined in depth by Huang and colleagues, occupy a central position in the Warburg phenotype. The predominance of LDHA in gynecologic malignancies of the ovary, endometrium, and cervix allows tumor cells to sustain glycolytic flux under hypoxic conditions while regenerating NAD+ needed for continued ATP production. The consequences extend beyond energy generation: accumulated lactate acidifies the tumor microenvironment, suppressing cytotoxic T cell function and promoting invasion. Clinically, elevated serum and tissue LDH levels correlate with advanced stage, metastatic burden, and chemoresistance, which has long positioned LDH as a prognostic indicator. The preclinical success of LDHA inhibitors in curbing tumor growth is encouraging, yet the observation that LDHA blockade may trigger context-dependent ferroptotic death in cervical cancer models introduces both an opportunity and a caution, since ferroptosis induction could be harnessed therapeutically but may also produce unpredictable toxicity in settings where lipid peroxide defenses vary between patients.

Amino acid metabolism emerges from the multi-omic analysis of He and colleagues as a rich source of prognostic and predictive information in bladder cancer. By interrogating 144 dysregulated amino acid-related genes, the authors identified a molecular subtype characterized by heightened glutamine and tryptophan flux, proliferative signaling, and an immune microenvironment that is inflamed yet functionally suppressed. This combination is particularly relevant to immunotherapy, because tumors with high tryptophan catabolism often deplete local tryptophan and generate catabolites that impair T cell effector function, creating resistance to checkpoint blockade. The 16-gene risk model developed in the study outperformed existing approaches and predicted responsiveness to PI3K/mTOR pathway inhibition as well as immunotherapy, illustrating how metabolic gene signatures can serve dual roles as prognostic tools and treatment-selection aids. The localization of the signature to malignant epithelial cells, and the demonstration that PSPH knockdown curtails invasion, provide mechanistic validation that the signature reflects functional biology rather than mere correlation.

The study by Li and colleagues on CELF1 in luminal A breast cancer demonstrates how post-transcriptional regulation can orchestrate metabolic phenotype. CELF1, an RNA-binding protein overexpressed in ER-positive tumors, promotes aerobic glycolysis by stabilizing or modulating transcripts encoding GLUT1 and glycolytic enzymes including HK2 and G6PD. The association of high CELF1 expression with poor survival and with ESR1 expression links metabolic rewiring to hormone receptor biology in a subtype where endocrine therapy remains the backbone of treatment. The synergistic tumor suppression observed when CELF1 knockout is combined with the GLUT1 inhibitor BAY-876 suggests that targeting the regulatory node and its downstream metabolic effector together may be more effective than either approach alone, a strategy conceptually similar to combining pathway inhibition at multiple points to overcome compensatory feedback.

Computational approaches feature prominently in the contribution of Wang and colleagues, who applied weighted gene co-expression network analysis and machine learning to the TCGA hepatocellular carcinoma cohort. Their identification of a glycolysis and lipid metabolism module, refined through LASSO-Cox modeling into a hub gene signature including PKM and ACAT2, exemplifies how large public datasets can be mined to extract metabolism-based prognostic classifiers. The capacity of the high-risk profile to predict recurrence and sorafenib response, together with in vitro validation that hub gene knockdown suppresses proliferation, addresses a frequent criticism of computational signatures, namely the lack of experimental follow-through. The integration of these hubs into a multi-omics nomogram further supports the idea that metabolic panels could become practical tools for stratifying patients with hepatocellular carcinoma, a disease in which systemic therapy options have historically been limited and response prediction remains imprecise.

Mitochondrial quality control, addressed by Han and colleagues in osteosarcoma, adds an organelle-level dimension to cancer metabolism. GABARAP-mediated mitophagy sustains pyruvate metabolism by clearing damaged mitochondria and maintaining a healthy mitochondrial pool capable of supporting the tricarboxylic acid cycle and anabolic reactions. The finding that GABARAP connects mitophagy-driven metabolic adaptation with immune evasion suggests that the metabolic and immunological phenotypes of tumors are not merely parallel processes but are mechanistically coupled. Therapeutic strategies that disrupt this coupling, whether through autophagy modulators or pyruvate dehydrogenase kinase blockers, could in principle simultaneously starve tumor cells of metabolic flexibility and relieve immunosuppression, though the essential role of mitophagy in normal tissues demands careful therapeutic window assessment.

Collectively, these eight contributions reinforce several principles that are shaping the field of cancer metabolism research. First, metabolic vulnerabilities are highly context-dependent, varying by tumor type, genetic background, tissue of origin, and microenvironmental conditions such as hypoxia and nutrient availability. A target that is critical in one malignancy may be dispensable in another, underscoring the need for the kind of molecular stratification advocated across multiple papers in this collection. Second, metabolic pathways rarely operate in isolation; they intersect with signaling cascades, epigenetic regulation, RNA processing, and immune function, meaning that metabolic interventions are likely to have effects that extend well beyond the enzyme or transporter they directly inhibit. Third, the convergence of high-throughput omics, systems biology modeling, and machine learning is accelerating the identification of metabolic dependencies at a pace that experimental validation must now strive to match.

The obstacles that remain are substantial. Metabolic plasticity allows tumor cells to reroute flux through alternative pathways when a preferred route is blocked, a phenomenon that has repeatedly undermined single-agent metabolic therapies in clinical trials. Intratumoral heterogeneity means that a biopsy sampled at one site and time point may not represent the metabolic state of the entire tumor, particularly after treatment pressure has selected for resistant subclones. Toxicity to normal tissues is a persistent concern because many metabolic enzymes targeted by cancer therapies, such as those in the mevalonate pathway or one-carbon metabolism, are also essential in proliferating bone marrow, intestinal epithelium, and immune cells. Nanotechnology-based delivery approaches, referenced within this Research Topic as part of the experimental toolkit, may help address this challenge by concentrating metabolic inhibitors in tumor tissue while sparing healthy compartments.

Looking forward, the most promising direction identified across these articles is the rational combination of metabolic interventions with immunotherapy and precision medicine frameworks. Because metabolic reprogramming directly shapes the tumor immune microenvironment through lactate accumulation, nutrient depletion, and catabolite-mediated T cell suppression, metabolic drugs may prime tumors for immune checkpoint blockade in ways that conventional cytotoxic agents cannot. Similarly, pairing metabolic targets with pathway-specific agents, as illustrated by the CELF1-GLUT1 combination strategy, offers a template for overcoming compensatory resistance. The continued integration of mechanistic biochemistry with computational discovery, biomarker development, and careful clinical trial design will determine whether the vulnerabilities described in this Research Topic can be converted into durable benefits for patients with cancer.

Subject of Research: …

Article Title: Editorial: Targeted cancer therapy through metabolic pathways

Article References: Editorial: Targeted cancer therapy through metabolic pathways. (n.d.). https://doi.org/10.3389/fmmed.2026.1842647

Image Credits: AI Generated

DOI: 10.3389/fmmed.2026.1842647

Keywords: Editorial, Targeted, cancer, therapy, metabolic, pathways, scientific research