PRKAR2A is notable because its protein product, RIIα, has a dual role in PKA regulation. Within the inactive PKA holoenzyme, RIIα restrains the catalytic subunits. It can also recruit type II PKA to A-kinase anchoring proteins, or AKAPs, which concentrate signalling complexes in defined cellular microdomains. The same subunit can therefore suppress local kinase activity while determining where that activity is positioned. The review proposes that RIIα may act as a driver, a compensatory brake, a localisation determinant or a passenger state marker. Existing evidence cannot discriminate among these models, and the metastasis-specific role of PRKAR2A remains unresolved.

One line of evidence came from an analysis comparing pathological complete response with residual disease in TNBC. PRKAR2A messenger RNA differed near the threshold for statistical significance, but the signal did not survive into the final consensus predictor. This observation suggests a possible association with treatment response or tumour state, yet it does not demonstrate that PRKAR2A causes response, persistence or metastatic progression. A separate experimental context found increased PRKAR2A expression in a lipid-adapted TNBC heterospheroid state. Because the investigators did not perturb the gene specifically, the result cannot establish whether PRKAR2A helps cells adapt to the metabolic environment or merely changes alongside other state-associated transcriptional changes.

RIIα was also detected in fractions containing mitochondria–endoplasmic-reticulum contact sites, where local PKA activity was associated with control of calcium transfer and apoptosis. The experiment linked mitochondrial–endoplasmic-reticulum signalling to a defined biological output, but it did not isolate RIIα as the responsible regulatory component. A fourth line of evidence placed PRKAR2A within a multigene prognostic signature. That inclusion indicates a possible prognostic association, although the gene lacked independent and cell-resolved validation in the evidence reviewed. Together, these observations identify RIIα as a plausible participant in spatially organised cAMP–PKA signalling, rather than a demonstrated metastasis regulator.

Related work involving AKAP2 provides a stronger mechanistic clue but not direct proof for PRKAR2A. Depleting AKAP2 altered focal-adhesion signalling, invasion, primary-tumour growth and lung signal. However, the experiments did not establish which PKA regulatory subunit mediated those effects and did not show a burden-independent role in metastasis. AKAP2 can anchor several signalling systems, so its depletion cannot be treated as a PRKAR2A-specific perturbation. This distinction is central to the review’s conclusion: evidence that an anchoring protein influences tumour behaviour is not equivalent to evidence that PRKAR2A routes cAMP–PKA signals in metastatic cells.

Testing the candidate hypothesis will require experiments designed to separate function, localisation and correlation. The review calls for inducible PRKAR2A depletion rather than reliance on constitutive knockout effects, together with rescue experiments performed at near-endogenous levels and, ideally, with spatial domain specificity. Compartment-targeted PKA reporters would help determine whether PRKAR2A changes where cAMP-dependent phosphorylation occurs. Endogenous mapping of PRKAR2A-containing complexes could establish which AKAPs and regulatory components accompany RIIα in tumour cells. Stage-resolved metastasis assays would then need to test whether manipulating PRKAR2A alters dissemination, survival or outgrowth independently of effects on primary-tumour growth.

Until those decisive tests are completed, the implications should remain narrow. PRKAR2A may help organise the subcellular distribution of PKA, and its expression may mark particular tumour states or contribute to a larger prognostic pattern. The current evidence does not show that manipulating PRKAR2A would prevent TNBC metastasis, improve treatment response or provide a clinically useful biomarker. The review therefore positions RIIα as a candidate spatial signal router and a framework for future experiments, not as a validated target. Its most important unresolved question is whether this dual-function regulatory subunit actively directs metastasis-relevant signalling in the correct subcellular compartment.

Subject of Research: The candidate role of PRKAR2A-encoded RIIα in spatial cAMP–PKA signalling associated with triple-negative breast cancer metastasis.

Article Title: PRKAR2A/RIIα as a candidate spatial cAMP–PKA signal router in triple-negative breast cancer metastasis: evidence boundaries, competing models and decisive experiments

Article References: Muzaffar, T., & Saha, B. (2026). PRKAR2A/RIIα as a candidate spatial cAMP–PKA signal router in triple-negative breast cancer metastasis: evidence boundaries, competing models and decisive experiments. Molecular Biology Reports, 53(1), Article 1768. https://doi.org/10.1007/s11033-026-12975-5

Image Credits: AI Generated

DOI: 10.1007/s11033-026-12975-5

Keywords: PRKAR2A, RIIα, triple-negative breast cancer, cAMP signalling, protein kinase A, cell migration, AKAP2, metastasis, tumour metabolism, critical review, candidate, spatial

The imaging technique at the heart of the study is ¹⁸F-fluorodeoxyglucose PET/CT, a hybrid technology that pairs the functional sensitivity of positron emission tomography with the anatomical detail of computed tomography. Fluorine-18 labelled fluorodeoxyglucose is a radioactive glucose analogue that cells import through the same transporters they use for ordinary sugar. Because most cancer cells run their metabolism on glycolysis even when oxygen is plentiful, the Warburg effect, malignant tissue accumulates the tracer far more avidly than surrounding healthy structures. A detector ring records the decay of the fluorine-18 isotope, and the resulting signal is converted into standardised uptake values, or SUVs, which quantify how intensely a given lesion traps glucose relative to the injected dose and body mass. The maximum standardised uptake value, SUVmax, captures the single hottest pixel within a lesion and has long served as a convenient, if imperfect, proxy for metabolic aggressiveness.

The Zhejiang team, led by Yunqi Zhu and corresponding author Xinhui Su, assembled thirty treatment-naïve patients with histologically confirmed testicular germ cell tumours who had undergone ¹⁸F-FDG PET/CT before any antitumour therapy. Twelve carried a diagnosis of seminoma and eighteen had non-seminomatous germ cell tumours, a group that encompasses embryonal carcinoma, yolk sac tumour, choriocarcinoma and other histologies, often in mixtures. Because the patients had not yet received chemotherapy, surgery or radiotherapy, the scans offered an unfiltered view of each tumour’s native metabolism, free from the confounding effects of treatment-induced inflammation or necrosis that plague post-therapy imaging.

The headline result is hard to overstate. Non-seminomatous tumours showed a median primary-tumour SUVmax of 16.66, compared with just 8.41 for seminomas, a difference that reached statistical significance at P equals 0.007. When the researchers looked instead at the hottest lesion anywhere in each patient, the gap widened further: 21.91 versus 8.41, with P below 0.001. In other words, the non-seminomatous tumours were burning through glucose at roughly twice the rate of their seminoma counterparts, a metabolic distinction that mirrors the known biological gulf between these entities. Non-seminomatous germ cell tumours tend to be more proliferative, more genetically chaotic and more prone to early haematogenous spread, and the PET data appear to capture that ferocity in a single number.

Disease extent itself was also strikingly different between the groups. Distant metastatic disease, classified as M1 under the American Joint Committee on Cancer staging system, was present in 55.6 percent of the non-seminomatous patients but only 8.3 percent of those with seminoma. This imbalance matters for interpretation, because a patient-level measurement such as the highest-lesion SUVmax can be inflated simply by the presence of a metabolically hot metastasis rather than by any intrinsic property of the primary tumour. The investigators were alert to this pitfall and designed their analyses accordingly, performing receiver operating characteristic analysis to compare diagnostic performance, restricting analyses to patients without distant metastases, and building exploratory multivariable models to adjust for tumour size and metastatic status.

Those sensitivity analyses proved revealing. When the analysis was confined to patients with M0 disease, meaning no distant metastases, the primary-tumour SUVmax remained significantly higher in non-seminomatous tumours, at 16.96 versus 8.00 with P equal to 0.001. However, the apparent advantage of the patient-level highest-lesion SUVmax over the primary-tumour measurement shrank dramatically: the difference in area under the receiver operating characteristic curve fell from 0.148 in the full cohort to just 0.023 in the metastasis-free subgroup. This tells an important story about what each metric actually measures. The highest-lesion SUVmax looked like a superb discriminator across all comers, achieving an apparent area under the curve of 0.944, but part of that performance was borrowed from the fact that metastatic spread is itself more common in non-seminomatous disease. The primary-tumour measurement, by contrast, appears to reflect genuine histology-linked biology.

The exploratory log-linear model drove the point home. After adjusting for primary-tumour size and M status, non-seminomatous tumours retained a higher primary-tumour SUVmax, with an adjusted geometric mean ratio of 1.67 relative to seminoma and a 95 percent confidence interval of 1.07 to 2.60, yielding P equal to 0.024. A geometric mean ratio of 1.67 implies that, on average, the non-seminomatous primary tumours took up roughly two-thirds more glucose than seminomas of comparable size and stage. For a single-centre study of thirty patients, this is an encouraging signal, though the authors are careful to label the modelling as exploratory and to emphasise that validation in larger independent cohorts is essential before any clinical application.

The study also placed PET-derived parameters alongside the traditional serum tumour markers that clinicians already use. Among individual biomarkers, the patient-level highest-lesion SUVmax showed the highest apparent discrimination with an area under the curve of 0.944, followed by alpha-fetoprotein at 0.910 and primary-tumour SUVmax at 0.796. Alpha-fetoprotein is a well-established hallmark of non-seminomatous histology, since seminomas do not produce it, so its strong showing is unsurprising. What is intriguing is that a purely metabolic imaging metric could rival a specific biochemical marker, suggesting that glucose avidity and embryonal differentiation are intertwined at a fundamental level. The researchers also computed a panel of volumetric parameters, including metabolic tumour volume, total lesion glycolysis and total metabolic tumour volume, which quantify not just how bright a lesion is but how much metabolically active tissue it contains, though the SUVmax comparisons dominated the significant findings.

Why should two tumours arising from the same primordial germ cell lineage differ so markedly in their sugar consumption? Seminomas are histologically homogeneous, composed of sheets of uniform cells resembling primitive spermatogonia, whereas non-seminomatous tumours can contain wildly divergent components, from gland-like yolk sac structures to syncytiotrophoblasts that secrete human chorionic gonadotropin. Embryonal carcinoma elements, in particular, are highly proliferative and express patterns of gene activity reminiscent of pluripotent stem cells, states that are metabolically expensive. Rapid proliferation demands nucleotide synthesis, lipid production and continuous ATP generation, all of which push cells toward intensified glycolysis and upregulated glucose transporters. The doubled SUVmax observed in non-seminomatous tumours is thus plausibly a readout of this underlying developmental chaos, though the study was not designed to identify the molecular mechanisms responsible.

The practical implications, if confirmed, could be meaningful. A pretreatment PET/CT that hints at histology could help clinicians anticipate the tumour type in settings where biopsy material is limited or ambiguous, flag patients at higher risk of harboring metastatic disease, and potentially refine surveillance strategies after orchiectomy. The authors themselves are appropriately measured, describing their work as a retrospective exploratory study of thirty patients from a single centre and calling for validation in larger, independent and ideally prospective cohorts. Such studies would need to address selection bias, scanner variability across institutions and the heterogeneity of non-seminomatous histologies. Still, the core observation stands out for its simplicity: before a single drop of chemotherapy, the two great families of testicular germ cell tumour already glow differently on a glucose scan, and that difference survives statistical scrutiny even when tumour size and spread are taken into account. It is a vivid reminder that cancer metabolism is not merely a byproduct of malignancy but a window into what a tumour truly is.

Subject of Research: Metabolic characteristics of testicular germ cell tumours on pretreatment ¹⁸F-FDG PET/CT imaging

Article Title: Pretreatment ¹⁸F-FDG PET/CT metabolic characteristics of seminoma and non-seminomatous testicular germ cell tumours: a retrospective exploratory study

Article References: Zhu, Y., Lu, X., Lin, L., Wang, H., Yu, Q., & Su, X. (2026). Pretreatment ¹⁸F-FDG PET/CT metabolic characteristics of seminoma and non-seminomatous testicular germ cell tumours: a retrospective exploratory study. BMC Medical Imaging. https://doi.org/10.1186/s12880-026-02855-6

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02855-6

Keywords: testicular cancer, testicular germ cell tumour, seminoma, non-seminomatous germ cell tumour, ¹⁸F-FDG PET/CT, SUVmax, tumour metabolism, metabolic tumour volume, alpha-fetoprotein, cancer imaging, BMC Medical Imaging, diagnostic imaging

The research, led by investigators at Tongji Hospital of Tongji Medical College, Huazhong University of Science and Technology, centres on kinesin family member C1, or KIFC1, a motor protein best known for hauling cargo along microtubules and for its role in centrosome clustering during cell division. By mining six publicly available transcriptomic datasets of cervical cancer, the team found that KIFC1 was among the most significantly dysregulated kinesin genes in tumour tissue compared with healthy cervical tissue. Pan-cancer analysis across more than twenty tumour types confirmed widespread overexpression, and examination of clinical databases linked elevated KIFC1 to lymph node metastasis. Crucially, the findings were not confined to computational correlations: immunohistochemical staining of a tissue microarray containing 115 cervical carcinoma specimens and 39 adjacent non-tumour tissues showed markedly higher KIFC1 protein levels in tumours, and patients whose tumours expressed high levels of the protein experienced significantly poorer disease-free and overall survival.

To probe what KIFC1 actually does inside cervical cancer cells, the researchers engineered HeLa and SiHa cell lines with stable KIFC1 knockdown or overexpression using lentiviral vectors. The functional consequences were striking. Depleting KIFC1 curtailed cell proliferation in cell counting and colony formation assays and impaired migration in wound healing and Transwell experiments, while forcing KIFC1 expression produced the opposite effects, enhancing growth and motility. But the most revealing clue came from RNA sequencing of KIFC1-silenced cells, which identified 2,714 differentially expressed genes. Pathway enrichment analysis pointed strongly toward metabolic reprogramming, with glycolysis and gluconeogenesis among the most significantly affected processes. Direct biochemical measurements confirmed the shift: glucose consumption, lactate production and intracellular ATP levels all fell when KIFC1 was silenced and rose when it was overexpressed, hallmarks of the Warburg effect that tumour cells exploit to fuel rapid proliferation.

The mechanistic thread connecting KIFC1 to metabolism runs through pyruvate kinase M, or PKM, the enzyme that catalyses the final, rate-limiting step of glycolysis. The PKM gene undergoes alternative splicing to yield two isoforms with opposing metabolic roles: PKM1 supports oxidative metabolism, whereas PKM2 favours aerobic glycolysis and is the dominant form in most tumours. When the researchers examined isoform expression after KIFC1 knockdown, they found PKM2 levels dropped while PKM1 rose, indicating that KIFC1 helps maintain the pro-glycolytic PKM2 balance that cervical cancer cells depend upon. The question was how a microtubule motor protein could influence a splicing decision buried in the cell’s gene-expression machinery.

The answer emerged from an unbiased proteomic approach. Using co-immunoprecipitation coupled with mass spectrometry, the team catalogued 1,067 proteins that physically associate with KIFC1, and among the enriched functional categories were spliceosome components and mRNA splicing processes. One candidate stood out: SRSF3, a serine/arginine-rich splicing factor previously shown to promote inclusion of PKM exon 10 and thereby tilt splicing toward PKM2. Co-immunoprecipitation experiments confirmed that KIFC1 and SRSF3 interact in cervical cancer cells, and database analysis revealed a robust correlation between the two genes in patient tumours. Intriguingly, KIFC1 knockdown did not change total SRSF3 abundance. Instead, nuclear and cytoplasmic fractionation combined with confocal immunofluorescence microscopy showed that depleting KIFC1 reduced the amount of SRSF3 in the nucleus and increased its cytoplasmic accumulation, suggesting that the motor protein acts as a shuttle that delivers the splicing factor to its site of action rather than as a regulator of its production.

Follow-up experiments cemented SRSF3 as the functional intermediary. Silencing SRSF3 in HeLa and SiHa cells suppressed proliferation, migration and all three glycolytic readouts, mirroring the effects of KIFC1 depletion. RNA immunoprecipitation demonstrated that SRSF3 binds directly to PKM messenger RNA, and its loss reproduced the isoform shift, raising PKM1 and lowering PKM2. Most persuasively, rescue experiments showed that restoring SRSF3 in KIFC1-depleted cells partially reversed the suppression of proliferation, migration, glucose consumption, lactate production and ATP generation, and re-established the PKM2-dominant profile. Taken together, these results delineate a coherent KIFC1–SRSF3–PKM axis in which a transport motor governs the subcellular localisation of a splicing factor, which in turn dictates a metabolic enzyme isoform switch that sustains the glycolytic appetite of the tumour.

What elevates this finding from an interesting piece of cell biology to a potentially transformative therapeutic insight is its connection to radiotherapy. Aerobic glycolysis is increasingly recognised as a driver of radioresistance because it supplies the energy and biosynthetic intermediates needed for DNA repair and helps tumour cells buffer the oxidative stress inflicted by radiation. When the researchers exposed KIFC1-silenced cells to graded doses of irradiation, clonogenic survival dropped significantly, particularly at 4 Gy and above, and radiation-induced apoptosis increased. Pharmacological activation of PKM2 with the selective activator TEPP-46 partially restored proliferative capacity and blunted the apoptotic response, demonstrating that the radiosensitising effect of KIFC1 loss depends on PKM2-driven glycolysis. Restoring SRSF3 produced a similar partial rescue, confirming the pathway’s role in the radiation response.

DNA damage assays provided a mechanistic window into why glycolytic disruption sensitises cells to radiation. After 4 Gy irradiation, KIFC1-depleted cells displayed elevated γ-H2AX fluorescence, a marker of DNA double-strand breaks, along with increased 53BP1 recruitment and reduced RAD51 foci, indicating impaired homologous recombination repair. Both SRSF3 restoration and TEPP-46 treatment partially reversed these changes, reinforcing the idea that a glycolysis-competent state is required for tumour cells to mend radiation-induced damage efficiently. The team then translated these findings into vivo using subcutaneous xenografts in nude mice. KIFC1 knockdown alone suppressed tumour growth, and combining it with a 6 Gy focal dose of irradiation produced significantly slower tumour growth than either intervention alone, accompanied by reduced Ki-67 staining and the same PKM isoform shift observed in culture. The authors candidly note that the in vivo effect was one of tumour growth delay rather than complete regression, a discrepancy they attribute to the complexity of the tumour microenvironment, including hypoxia, stromal interactions and the metabolic plasticity that allows tumours to compensate by activating alternative energy pathways.

The study does carry limitations that the authors themselves acknowledge. Validation in larger clinical cohorts will be needed to firmly establish KIFC1 as a prognostic and predictive biomarker, the precise splicing sites mediating SRSF3’s control of PKM in cervical cancer were not directly mapped, and the molecular determinants of the KIFC1–SRSF3 interaction, including possible adaptor proteins, remain to be identified. Nevertheless, the work expands the functional repertoire of KIFC1 well beyond mitosis, revealing a post-transcriptional mechanism by which intracellular transport machinery shapes metabolic adaptation and therapeutic resistance. If pharmacological inhibitors of this axis can be developed or repurposed, clinicians may one day combine them with radiotherapy to strip cervical tumours of the metabolic armour that currently lets them survive treatment, offering new hope for patients whose disease has stopped responding to conventional care.

Subject of Research: The role of the KIFC1–SRSF3–PKM splicing–metabolic axis in cervical cancer glycolysis and radiosensitivity

Article Title: Targeting the KIFC1‐SRSF3‐PKM Axis Suppresses Cervical Cancer Glycolysis and Enhances Radiosensitivity

Article References: Liu, J., Li, J., Zhou, H., Ha, W., Wu, X., Liu, X., Jiang, Y., Gong, C., Cheng, Y., Liu, Q., Chao, T., & Xiong, H. (2026). Targeting the KIFC 1‐ SRSF 3‐ PKM Axis Suppresses Cervical Cancer Glycolysis and Enhances Radiosensitivity. Journal of Cellular and Molecular Medicine, 30(18), Article e71373. https://doi.org/10.1111/jcmm.71373

Image Credits: AI Generated

DOI: 10.1111/jcmm.71373

Keywords: cervical cancer, KIFC1, SRSF3, PKM2, glycolysis, radioresistance, alternative splicing, Warburg effect, DNA damage repair, kinesin motor protein, radiosensitivity, tumour metabolism

The single most transformative technical development of the period is widely agreed to be the sequencing revolution. The landmark 2005 demonstration of genome sequencing in microfabricated high-density picolitre reactors opened the door to reading tumour genomes at scale, an advance that would have been unimaginable when the journal launched. From that technological foundation flowed the systematic discovery of cancer genes, the classification of tumours by their mutational signatures and the rise of precision oncology as a clinical discipline. The identification of recurrent mutations of the BRAF gene in human cancer in 2002, before the sequencing revolution fully matured, had already signalled what was to come: single genetic lesions, once found, could define entire treatment strategies.

Clinical actionability has expanded at a remarkable pace. Analyses quantifying the expanding landscape of clinical actionability for patients with cancer show that a steadily growing fraction of tumours now harbour alterations for which targeted drugs or guided treatment decisions exist. The poster child of this era remains BRAF V600E inhibition in melanoma, but the paradigm has matured in sophistication. When colon cancers proved unexpectedly unresponsive to BRAF inhibition, researchers discovered that feedback activation of EGFR was protecting the cells—a finding that led directly to the FDA-approved combination of BRAF and EGFR inhibitors for BRAF-mutant colorectal cancer, a therapy developed in René Bernards’ laboratory at the Netherlands Cancer Institute. The lesson embedded in that story reshaped the field: cancers are wired for resilience, and single-agent thinking is rarely sufficient.

That lesson now drives the logic of synthetic lethality, one of the most productive conceptual frameworks of the past decade. Rather than attacking oncogenes directly, synthetic lethal strategies exploit vulnerabilities created by a tumour’s alterations. A striking recent example is the demonstration that amplification of the cyclin E1 gene CCNE1, common in aggressive ovarian and other cancers, creates a dependence on the PKMYT1 kinase that can be pharmacologically exploited. This approach extends druggability into territory long considered untouchable, and drug discovery scientist Paul Workman of the Institute of Cancer Research has championed the broader project of drugging the cancer genome—developing chemical inhibitors and small-molecule research tools against targets once dismissed as intractable.

Perhaps no target symbolises both the promise and the frustration of the era better than RAS. Mutant RAS genes were identified in the early 1980s and were long considered undruggable, a verdict repeated in reviews for decades. Yet the past few years have delivered direct RAS inhibitors, and a 2026 phase three trial reported that the RAS inhibitor daraxonrasib outperformed chemotherapy in previously treated metastatic pancreatic cancer—one of the most difficult malignancies to treat. Allan Balmain, whose laboratory uses mouse models to trace the cells that acquire initiating oncogenic mutations, notes that interpreting such advances requires a deeper understanding of tumour evolution itself: when a mutation arises, in which cell it arises, and what environmental forces promote or suppress its expansion.

That evolutionary perspective has been genuinely subversive. Sequencing of normal human skin revealed a high burden of somatic mutations and pervasive positive selection operating in histologically normal tissue—mutant clones competing for space in organs that look entirely healthy. Even more provocative, work in mosaic mouse skin has shown that injury can prevent the expansion of Ras-mutant cells, overturning the intuitive assumption that wounding uniformly promotes carcinogenesis. Complementing this, chemically induced skin tumours have been traced to long-lived stem cells of the upper hair follicle, underscoring that the identity of the cell of origin is a decisive variable in tumour outcome. Cancer, in this view, is not simply a cell-autonomous genetic disease but an ecological process embedded in tissue dynamics, environmental exposure and time.

The cancer stem cell concept illustrates how an important idea can be discovered, oversimplified and then properly understood only over decades. The founding observation—that human acute myeloid leukemia is organized as a hierarchy originating from a primitive hematopoietic cell—was published in 1997, before the review period began. The intervening years saw the concept generalized, contested and eventually refined. As Hans Clevers and Eduard Batlle argued in revisiting the field, functional stemness depends on cellular context and state rather than immutable identity, and epithelial cell plasticity allows cells to cross lineage boundaries that once seemed fixed. The modern synthesis holds that many tumour cells can transiently adopt stem-like behaviour, which has important consequences for therapy resistance and relapse, even if the simplest hierarchical models proved too rigid.

Technology has also reshaped how tumours are studied. Clevers’ pioneering work on organoids—self-organizing, three-dimensional cultures derived from adult stem cells—created patient-derived models that preserve the genetic and functional features of the original tissue, bridging the long-standing gap between cell lines and patients. Organoid biobanks now support drug screening, gene-function studies and personalised treatment predictions, and the technology has been commercialised widely, with Clevers holding patents and co-founding companies in the field. Alongside organoids, powerful tools such as CRISPR-based functional genomics, exemplified by the synthetic lethality screens of the Bernards laboratory, allow systematic interrogation of cancer vulnerabilities directly in human cells.

Beyond genes, the anniversary reflections give substantial weight to metabolism and the tumour microenvironment, areas that were marginal when the journal began but are now central. Tumour-derived lactic acid was shown to functionally polarise tumour-associated macrophages, demonstrating that a metabolic by-product can reprogram immune cells within the microenvironment. Cancer-associated cachexia, the devastating wasting syndrome that contributes to a large fraction of cancer deaths, has recently been connected to mechanistic insight in human studies, replacing decades of descriptive work. Karen H. Vousden, whose research spans p53 biology and metabolism, has explored how dietary composition interacts with tumour metabolism, and precision nutrition companies are now testing whether manipulating non-essential amino acids can enhance therapy in patients. The message is that cancer cannot be understood as a tumour in isolation; it is a systemic disease reshaping the entire body.

The newest force transforming the field is artificial intelligence. The transformer architecture described in ‘Attention is all you need’ underpins models that now predict biomolecular interactions with remarkable accuracy, as demonstrated by AlphaFold 3, and AI agents are being designed to empower biomedical discovery end to end—from hypothesis generation to experimental execution. Marinka Zitnik of Harvard Medical School leads work on multimodal and agentic AI systems paired directly with laboratory experiments, and is team lead of AURORA within Cancer Grand Challenges. The authors’ cautious optimism is tempered by experience: past waves of overhyped technologies promised more than they delivered, and AI will need to demonstrate genuine predictive power in clinics, not only impressive benchmarks, to transform outcomes.

Woven through the reflections is an honest accounting of unmet expectations. Not every large-scale endeavour—from indiscriminate high-throughput screening to simplistic gene-expression signatures—delivered what early enthusiasts forecast. Combination therapy remains difficult to develop because toxicity, not efficacy, is usually the limiting factor. Drug resistance evolves faster than new drugs arrive, and late-stage solid tumours continue to claim most of the lives that cancer kills. Yet the trajectory is unmistakable: five-year survival has risen for many malignancies, targeted combinations have turned once-fatal diseases into chronic or curable conditions, and the mechanistic depth of the field is unrecognisable compared with 2001. The authors’ shared conclusion is that the next 25 years will be judged by how well the field integrates these hard-won insights—evolutionary, ecological, metabolic and computational—into treatments that reach every patient, not just those with actionable mutations in privileged tumour types.

Subject of Research: A 25-year retrospective on transformative advances and unmet expectations in cancer research

Article Title: Reflecting on 25 years of cancer research: transformative advances and unmet expectations

Article References: Balmain, A., Bernards, R., Clevers, H., Vousden, K. H., Workman, P., & Zitnik, M. (2026). Reflecting on 25 years of cancer research: transformative advances and unmet expectations. Nature Reviews Cancer. https://doi.org/10.1038/s41568-026-00974-4

Image Credits: AI Generated

DOI: 10.1038/s41568-026-00974-4

Keywords: cancer research, Nature Reviews Cancer, tumour evolution, precision oncology, synthetic lethality, RAS inhibitors, cancer stem cells, organoids, tumour metabolism, cachexia, artificial intelligence, drug discovery