A 3-year structured exercise program begun soon after adjuvant chemotherapy improved disease-free survival and suggested better overall survival versus health education alone, with a modest increase in musculoskeletal events.

Background

Recurrence after curative-intent surgery and adjuvant chemotherapy for stage III or high-risk stage II colon cancer remains common and treatment side effects impair function and quality of life. Preclinical and observational data suggest exercise may improve cancer outcomes, but definitive randomized evidence has been lacking.

Patients

• N = 889 adults with resected stage III (90%) or high-risk stage II (10%) colon adenocarcinoma.
• Completed adjuvant chemotherapy 2–6 months prior; ECOG 0–1; exercising less than 150 min/week at baseline; able to perform submaximal walk test.
• Median age 61 years; 51% women; BMI median 28.5; 61% Canada, 33% Australia.
• Adjuvant regimens: FOLFOX 61%, CAPOX 14.5%, capecitabine 16.9%, other 7.6%.

Intervention

Structured, behaviorally supported exercise program for 3 years targeting ≥10 MET-hours/week increase in recreational moderate-to-vigorous activity (e.g., brisk walking), with individualized modality, frequency, and duration:

• Phase 1 (0–6 months): 12 biweekly behavioral-support sessions (in-person) + 12 mandatory and 12 recommended supervised exercise sessions (alternating weeks).
• Phase 2 (6–12 months): 12 biweekly behavioral-support sessions (in-person or remote), supervised session if in-person.
• Phase 3 (12–36 months): 24 monthly behavioral-support sessions (in-person or remote), supervised session if in-person.

Control

• Health-education materials only (general physical activity and nutrition guidance) plus standard oncologic surveillance.

Outcome

• Primary: Disease-free survival (DFS): time to colon cancer recurrence (local/distant), new primary colorectal cancer, second primary cancer, or death (any cause).
• Key secondary: Overall survival (OS); patient-reported physical functioning (SF-36 physical-functioning subscale); objective fitness (predicted VO2max), 6-minute walk distance; recreational moderate-to-vigorous physical activity; safety (with focus on musculoskeletal adverse events).

Study Design

• Phase 3, multicenter, randomized controlled trial at 55 centers; 1:1 allocation via dynamic minimization stratified by center, stage, BMI, and ECOG status.
• Intention-to-treat efficacy analyses; as-treated safety analyses.
• Accrual 2009–2024; database lock January 24, 2025; ClinicalTrials.gov NCT00819208.

Level of Evidence

Level 1 (phase 3 randomized controlled trial).

Follow up period

• Median follow-up: 7.9 years.
• Reported timepoints: 5-year DFS; 8-year OS.

Results

• Primary outcome (DFS):
  • Hazard ratio (HR) for disease recurrence, new primary cancer, or death: 0.72 (95% CI, 0.55–0.94).
  • 5-year DFS: 80.3% (exercise) vs 73.9% (health-education); absolute difference 6.4%.
  • NNT: 16 over 5 years to prevent one DFS event (using 5-year absolute difference).
  • Annual incidence of DFS events: 3.7% vs 5.4%.
  • Fewer liver recurrences (3.6% vs 6.5%) and fewer new primary cancers overall (5.2% vs 9.7%).
• Secondary outcomes:
  • Overall survival: HR for death 0.63 (95% CI, 0.43–0.94); 8-year OS 90.3% vs 83.2% (absolute difference 7.1%); NNT 15 over 8 years to prevent one death.
  • Patient-reported physical functioning (SF-36): Greater improvements across 3 years in the exercise group (e.g., +6–7 point gains vs ~+1–3 in control).
  • Objective fitness and function: Higher predicted VO2max (+1.3 to +2.7 ml/kg/min) and longer 6-minute walk distance (+13 to +30 m) over 3 years in exercise group.
  • Physical activity behavior: Sustained between-group increase of ~5.2–7.4 MET-hours/week favoring exercise.
  • Safety: Any adverse event during intervention: 82.0% (exercise) vs 76.4% (control). Musculoskeletal events: 18.5% vs 11.5%; NNH ~14. Grade ≥3 AEs: 15.4% vs 9.1%. Only 10% of musculoskeletal events were judged related to the intervention.

Limitations

• Slow accrual over 15 years; fewer events than planned (224/380), potentially reducing power for some analyses.
• Enrollment 2–6 months post-chemotherapy may exclude early recurrences and favor higher-functioning participants (selection bias).
• More contact time/social interaction in the intervention arm could confound some benefits.
• Physical activity self-reported (recall bias), though corroborated by objective fitness gains.
• Lynch syndrome status not tracked; generalizability to such subgroups uncertain.

Funding

Canadian Cancer Society, Australian National Health and Medical Research Council, and Cancer Research UK; coordinated by the Canadian Cancer Trials Group. Independent data and safety monitoring committee oversight.

Citation

Courneya KS, Vardy JL, O’Callaghan CJ, et al.; for the CHALLENGE Investigators. Structured Exercise after Adjuvant Chemotherapy for Colon Cancer. New England Journal of Medicine. Published online June 1, 2025. DOI: 10.1056/NEJMoa2502760.