Missing ScriptStart HereBlogCountry ExplorerIntegrative Care GuideEvidence ExplorerPodcastAbout

Evidence Explorer / Evidence record

CONVENTIONAL CARE COMPARATOR

Topical diclofenac to prevent capecitabine hand–foot syndrome

Mixed / uncertain

A 2024 phase III trial found a sizable preventive benefit, but an August 2026 prospective real-world study did not reproduce it. This is off-label supportive care, not cancer treatment.

MS-N016 · Literature searched October 13, 2026 · Appraised October 13, 2026 · Hand–foot syndrome

Evidence at a glance

  • Randomized trial: 264 people with breast or gastrointestinal cancer starting capecitabine were assigned to topical diclofenac or placebo; 263 were analyzed.
  • Primary result: grade 2–3 hand–foot syndrome occurred in 3.8% with diclofenac and 15.0% with placebo—an absolute difference of 11.2 percentage points (95% CI 4.3–18.1; P=.003).
  • Dose-impact result: HFS-related capecitabine dose reduction occurred in 3.8% versus 13.5%—an absolute difference of 9.7 points (95% CI 3.0–16.4).
  • Reality check: in a 151-person two-center observational study published August 20, 2026, grade 2 or worse HFS occurred in 16.9% with prophylactic diclofenac versus 13.8% with active monitoring (OR 1.28; 95% CI 0.52–3.10; P=.591).
  • Bottom line: the randomized signal is promising, but the effect may depend on product, application, population and adherence; replication is incomplete.

What was studied?

The D-TORCH phase III study was single-site, randomized and double blind. Participants used topical diclofenac or placebo gel for up to 12 weeks or until HFS developed. The later DICLO-HFS study was prospective but not randomized: 71 patients received 1% diclofenac twice daily to hands and feet and 80 underwent active monitoring, with median follow-up of 19.1 weeks.

How to read the disagreement

The randomized trial provides stronger causal evidence because allocation and blinding reduce selection and expectation bias. The observational study cannot cancel that result, but it is important because it tested routine practice in colorectal and gastric cancer at two centers and found no benefit. Differences in application sites, product formulation, adherence, capecitabine regimens and baseline risk may matter. A protocol for a Japanese 0.1% cream trial is ongoing evidence generation, not a positive result.

Safety and practical context

Topical does not mean risk-free. Diclofenac labeling includes kidney, gastrointestinal bleeding and cardiovascular warnings; systemic exposure is lower than with oral NSAIDs but not zero. NSAID allergy, anticoagulants, other NSAIDs, dehydration, kidney impairment and broken skin require review. Painful peeling, blistering or impaired function should trigger oncology-team contact because capecitabine interruption or dose adjustment may be needed. Preventing HFS and treating established HFS are different questions.

Funding and conflicts: PubMed classifies D-TORCH as non-US government-supported research. Detailed conflict and funding statements for D-TORCH and DICLO-HFS were not fully reverified in this pass and should be checked before guideline-level synthesis.

Questions for the care team

  • Would this off-label prevention strategy fit the capecitabine regimen and skin condition?
  • Which exact formulation, application sites and monitoring plan would be used?
  • Do kidney function, anticoagulants or NSAID allergy change the risk?
  • Which symptoms should prompt a call or treatment hold under the oncology team’s instructions?

Sources

Learn how to judge a study · Return to the evidence map

This evidence brief is educational, not a personal treatment recommendation. Research regimens are not prescribing instructions. Symptom benefit does not establish tumor response, recurrence prevention or survival benefit.