Cancers rewire metabolism to resist drugs, and the rewiring opens targets

Resistance as rewiring

In a preprint, researchers report that SMARCA4 mutations are enriched in relapsed T-cell acute lymphoblastic leukaemia samples compared with diagnosis, and that SMARCA4-mutant cells resist a broad range of chemotherapy agents. The mechanism the authors report is impaired remodelling of AP-1-associated regulatory regions: the mutation changes which parts of the genome the cell can open and use. Pharmacological inhibition of AP-1 mimics the chemotherapy resistance of the mutant cells, the authors report, which strengthens the case that this regulatory pathway, rather than a single drug pump or mutation, carries the resistant state. This suggests resistance here is a programme the cell has switched into, not a single broken part. m:…0747

A second preprint finds a similar logic in the amino acid stress response. Its authors report that pharmacological perturbation of that pathway has broad anti-tumour activity across multiple aggressive malignancies, and that sensitivity is associated with a chronic stress-adapted metabolic state marked by constitutive amino acid response signalling, elevated NFE2L2 expression and reduced mitochondrial abundance. One compound, GCN2-IN-6, does not reproduce the effect of genetic GCN2 suppression, the authors report, indicating context-dependent effects. This suggests the stress response is a real dependency for some tumours, but also that a drug and a gene knockout are not interchangeable ways of probing it. m:…8708

Metabolism as the lever

The clearest metabolic story comes from metastatic colorectal and pancreatic cancer. In a preprint, researchers report that HER3 activation drives a liver-metastasis-specific glycolytic program, promoting glycolysis through the AKT/RSK-PFK2 axis and reducing oxidative phosphorylation through DRP1-mediated mitochondrial remodelling. Blocking HER3 with sapitinib increases dependence on oxidative phosphorylation in liver metastases, the authors report, which sets up the combination: pairing HER3 blockade with oxidative phosphorylation inhibitors results in synergistic tumour-free survival in orthotopic and patient-derived xenograft models. This suggests a two-step logic worth noting: the first drug does not just slow the tumour, it pushes it onto a second metabolic pathway, and the second drug then blocks the escape route. m:…9661

A breast cancer preprint works the same logic through polyamines, small molecules involved in cell growth. Its authors report that the ODC-OAZ1-AZIN1 axis shows a polarised seesaw configuration across diverse tumour types including breast cancer, that the axis configuration determines breast cancer subtype sensitivity to polyamine-targeting inhibitors, and that targeted reconfiguration of the axis can shift drug sensitivity in resistant breast cancer cells. This suggests sensitivity to this drug class is not a fixed property of a subtype but a state that can, in the laboratory setting, be moved. That distinction matters for how trials might be designed: testing the axis state, not just the subtype label, could separate likely responders from likely non-responders. m:…1811

A vulnerability created by the rewiring

The fifth preprint shows the rewiring creating a vulnerability in the opposite direction: a deficiency that makes cells more sensitive, not less. Its authors report that SDHB deficiency leads to reduced ABCB1 expression in paraganglioma, renal cell carcinoma and gastrointestinal stromal tumours, that KDM4A downregulation is the mechanism for this silencing, and that SDHB-deficient cells are hypersensitive to drugs that are substrates of ABCB1. Vincristine-containing chemotherapy improves progression-free survival in SDHB-mutant paraganglioma patients, the authors report. This suggests a tumour's metabolic defect can remove one of its own defences, leaving it exposed to an existing drug class. m:…4194

Taken together, the five preprints argue against treating resistance as a wall and for treating it as a trade. Each resistant state described here buys survival at a cost: a remodelled regulatory landscape, a chronic stress adaptation, a glycolytic program, a polyamine configuration, or a silenced drug transporter. This suggests the productive question for drug development is what each resistant state now depends on, or has given up. All five findings are preprints, and most rest on laboratory or animal models rather than clinical trials, so none of them is a treatment claim. But as a research direction, they converge: map the rewiring first, then target what it costs. m:…0747 m:…9661 m:…4194