Patient willingness to de-escalate adjuvant therapy based on predictive biomarker results in stage IIB-IIID melanoma: A cross-sectional survey

INTRODUCTION: Predictive biomarkers such as circulating tumour DNA (ctDNA) and gene expression profiling (GEP) may guide adjuvant melanoma therapy de-escalation by identifying low-recurrence-risk patients. However, their implementation depends on patient willingness to act on test results. This study investigated patient preferences regarding predictive biomarkers. METHODS: A cross-sectional survey was conducted among consecutive patients with AJCC 8th-edition stages IIB-IIID melanoma at a tertiary centre (N = 159, response rate 80.7%). The self-constructed questionnaire assessed biomarker-related attitudes, including willingness to forgo adjuvant therapy, accuracy thresholds, and undertreatment-overtreatment trade-offs, alongside sociodemographic characteristics. Validated instruments assessed decision-making preferences (Control Preferences Scale), health literacy (HLS-EU-Q16), fear of progression (PA-F-KF), and anxiety (GAD-7). Alongside descriptive analyses, logistic regression examined factors associated with willingness to forgo therapy. RESULTS: Despite limited prior awareness (34.2%), 97.4% considered biomarker research important. Most patients expected personalised therapy (82.3%) as primary benefit and wished to receive prognostic information on adverse events (91.1%) and disease progression (83.5%). Given a favourable five-year test result, 38.0% would forgo adjuvant therapy; 73.3% required ≥ 95% test accuracy. Two-thirds (66.9%) considered undertreatment worse than overtreatment; only 18.3% would accept de-escalation given a 10% false-negative rate. Lower health literacy (OR 0.61, p = 0.045) and an academic degree (OR 2.31, p = 0.037) were associated with willingness to forgo therapy. The latter emerged only after adjustment for health literacy, potentially reflecting a suppression effect. CONCLUSIONS: Melanoma patients demonstrate strong interest in biomarkers but demand high accuracy with low tolerance for false-negative predictions. Implementation of biomarker-guided decisions requires targeted patient education addressing predictive uncertainty.