For more than a century, the idea that physical activity could influence cancer has hovered at the margins of oncology. A new review published in Nature Reviews Clinical Oncology by Joshua W. Bliss of Weill Cornell Medical College and Lee W. Jones of City of Hope argues that the field of exercise oncology has now reached a decisive turning point, one in which structured exercise is being tested not merely as supportive care but as a potential anticancer therapy in its own right. Drawing together epidemiological data, landmark clinical trials and mechanistic studies, the authors lay out both the promise and the unresolved questions surrounding exercise as a therapeutic modality, and they propose a translational framework aimed at building what they call precision exercise oncology.
The epidemiological case begins with prevention. Large observational studies have consistently linked higher levels of physical activity with lower risk of several cancer types, including breast and colorectal cancer. One widely cited analysis of 1.44 million adults associated higher leisure-time physical activity with reduced risk across 26 cancer types, and more recent work using accelerometer-measured activity and step counts in cohorts such as the UK Biobank has reinforced the signal with objective measurement. Yet the review is careful to stress the limits of such data. Residual confounding, reverse causation and tumour-site-specific biases, including the healthy screenee effect in which people who volunteer for screening tend to be healthier overall, mean that observational evidence alone cannot establish that exercise causes lower cancer incidence. Mendelian randomization analyses and target trial emulation frameworks are being deployed to strengthen causal inference, but the authors argue that definitive answers will require interventional trials.
On the survival side, the observational picture is similarly consistent. Post-diagnosis physical activity has been associated with lower cancer-specific and all-cause mortality across several cancer types, with pooled analyses and large cohort studies in breast, colon and prostate cancer converging on the same direction of effect. Notably, some studies suggest the benefit may vary by tumour biology: associations between activity and survival in colorectal cancer differ according to molecular markers such as beta-catenin alterations and IRS1 expression, hinting that exercise might interact with tumour genotype. However, evidence for a benefit in disease-free survival, the endpoint most often used to judge adjuvant therapies, remains limited, and observational associations cannot exclude the possibility that healthier patients exercise more rather than exercise making patients healthier.
The decisive shift has come from randomized trials. Two adequately powered phase III studies have now directly tested structured exercise against clinical disease outcomes. The first, the CHALLENGE trial reported in the New England Journal of Medicine, enrolled patients with colon cancer who had completed surgery and adjuvant chemotherapy and found that a structured exercise programme improved both disease-free survival and overall survival. The result, described by many in the field as a watershed, provides the strongest evidence to date that a behavioural intervention can alter the natural history of a resected cancer. The second, the ECHO trial in ovarian cancer, tested exercise added to first-line systemic therapy and was negative for its primary endpoint of progression-free survival, a reminder that timing, tumour type and treatment context may all determine whether exercise can bend the survival curve.
Between these two results lies a more complicated experimental landscape. No adequately powered trial has yet demonstrated that structured exercise improves tumour response to systemic therapy, including immunotherapy. The only statistically significant signal on tumour response comes from the LEANer trial, in which a combined exercise and nutrition intervention was associated with higher rates of pathological complete response in breast cancer, but that finding emerged from an under-powered subgroup analysis and therefore requires confirmation. Trials of exercise during neoadjuvant chemotherapy for breast cancer, including Neo-train, BENEFIT and the Neo-ACT programme, have so far produced mixed results on tumour response and chemotherapy completion, underscoring how much remains unknown about the right dose, timing and patient population.
Mechanistically, the biological plausibility of exercise as an anticancer agent has grown substantially. Preclinical work shows that modifying an individual’s exposure to exercise can reshape the host macroenvironment, producing changes that propagate to distant tissues and to the tumour microenvironment itself. In mouse models, aerobic exercise normalizes tumour vasculature, reducing hypoxia and improving the delivery and efficacy of chemotherapy, and recent studies have implicated S1PR1 signalling in tumour endothelial cells as a mediator of improved vascular function in pancreatic cancer. Exercise also mobilizes and activates immune cells: epinephrine- and interleukin-6-dependent recruitment of natural killer cells, CXCR3-driven infiltration of CD8-positive T cells, and engagement of the IL-15/IL-15R alpha axis have all been shown to enhance tumour control in animal models, and in some settings to sensitize tumours to immune checkpoint blockade.
Human correlative studies are beginning to echo these findings. Early-phase trials in prevention and window-of-opportunity settings, in which exercise is delivered between diagnosis and surgery, have documented changes in blood-based and tissue-based biomarkers, including sex hormones, insulin and insulin-like growth factor axis proteins, inflammatory markers, immune cell activity and gene expression profiles in tumour tissue. Trials in postmenopausal women at elevated breast cancer risk showed that year-long aerobic exercise programmes lowered serum estrogens and altered adipokines. Exercise training in people with Lynch syndrome, an inherited predisposition to colorectal cancer, reduced inflammatory responses and promoted mucosa-associated immunity in the intestinal lining. Yet the review emphasizes a critical gap: no early-phase study has established a biologically active dose based on a validated surrogate endpoint, leaving clinicians without a rational prescription analogous to a drug dose.
To close that gap, Bliss and Jones propose an integrated translational framework modelled on drug development. Exercise therapy, they argue, should move through the same staged pipeline as molecular therapeutics: early-phase dose-finding studies, then phase II randomized trials with biomarker and clinical endpoints, and only then definitive phase III testing, all conducted in parallel with correlative science aimed at elucidating mechanisms of action, predictors of response, active dose ranges and optimal prescriptions. Emerging tools could accelerate this process, including molecular profiling of acute exercise responses, ultrasensitive plasma-based monitoring of tumour burden and molecular residual disease assays that could serve as sensitive intermediate endpoints. The framework also envisions precision exercise oncology, in which a patient’s tumour subtype, molecular features and treatment regimen would guide the specific exercise prescription, much as biomarkers now guide drug selection.
The stakes of getting this right are considerable. Leading oncology agencies, including the Clinical Oncology Society of Australia, the American Society of Clinical Oncology and the American Cancer Society, already endorse exercise as a supportive-care adjunct to mitigate the acute and late effects of treatment, and international consensus guidelines exist for cancer survivors. But supportive care and therapy are different claims, requiring different standards of evidence. If the CHALLENGE result can be replicated and extended to other tumour types and treatment settings, exercise could become one of the few interventions capable of improving survival after curative-intent treatment at relatively low cost and low toxicity. Conversely, if the field continues to rely on observational associations and under-powered endpoints, exercise risks remaining a well-intentioned recommendation without therapeutic standing. The review’s message is that the scientific infrastructure now exists to settle the question, and that the next decade of rigorously designed exercise trials will determine whether the oldest of health behaviours earns a formal place in the oncology pharmacopoeia.
Subject of Research: Exercise therapy as an anticancer intervention in cancer prevention and treatment
Article Title: Exercise therapy in cancer prevention and treatment
Article References: Bliss, J. W., & Jones, L. W. (2026). Exercise therapy in cancer prevention and treatment. Nature Reviews Clinical Oncology. https://doi.org/10.1038/s41571-026-01201-7
Image Credits: AI Generated
DOI: 10.1038/s41571-026-01201-7
Keywords: exercise oncology, cancer prevention, CHALLENGE trial, ECHO trial, disease-free survival, tumour microenvironment, immunotherapy, biomarkers, precision oncology, clinical trials, cancer survivorship, epidemiology
News Source: Nathaniel Bowman. (October 7, 2026). Exercise Emerges as a Serious Candidate Therapy in Cancer Prevention and Treatment. Scienmag.
