Early detection of skin cancer

At EADO 2025, Prof Dr Lieve Brochez from the University Hospital Ghent, Belgium, presented on behalf of the Skin Cancer Research Institute Ghent, work on improving early detection of melanoma. Given the rising incidence of skin cancer, now accounting for 1 in 3 cancers, the team emphasised the need for effective, targeted detection strategies. While dermatologists best monitor high-risk individuals (e.g. with prior skin cancer or genetic predispositions), this group is small. The broader low-risk population, though individually less at risk, contributes significantly to the total number of cases.

To address this, Ghent developed a “one-spot check” consultation for patients with an alarming skin lesion (e.g. new, changing, non-healing, or referred lesions). Among over 4,000 such consultations, they achieved a 14% skin cancer detection rate, with 2% melanomas, higher than general screening rates. Importantly, dermatology nurses led most of these consultations under supervision, helping meet demand and reduce waiting times (90% seen within 4 weeks).

Efforts are underway to improve nurse training in dermoscopy and explore AI support for diagnosis. The team is also testing this model outside hospitals and among general practitioners to maintain high detection rates.

In a separate oral presentation, they discussed a prospective study on smartphone apps for skin lesion diagnosis. Despite stricter EU regulations, current apps lack sufficient accuracy. Patients expect ≥95% sensitivity and accept lower specificity, but real-world results show lower performance, including missed skin cancers. This raises concerns about app reliability and the adequacy of current regulatory standards.

Overall, the Ghent team promotes nurse-led triage, AI integration, and evidence-based app evaluation to improve early skin cancer detection for the broader population.