Abstract

BACKGROUND: Amivantamab, a bispecific antibody targeting EGFR and MET, is increasingly used in advanced non-small cell lung cancer. Dermatologic and mucosal adverse events are highly prevalent and often more severe and clinically complex than with conventional EGFR inhibitors, frequently challenging treatment continuation. However, structured management guidelines for amivantamab-associated toxicities are currently lacking.

METHODS: An international panel of dermatologists with expertise in oncodermatology and members of the EADV Task Force "Dermatology for Cancer Patients" conducted a structured review of clinical trial data, emerging real-world evidence, and existing toxicity management frameworks. Key clinical challenges were identified through iterative expert discussions, and consensus was achieved through repeated rounds of manuscript review and refinement. Pragmatic grading systems and stepwise management algorithms were developed.

RESULTS: We propose comprehensive strategies for prevention and monitoring and provide toxicity-specific management algorithms for acneiform eruption, erosive pustular dermatosis (EPD)-like scalp reactions, paronychia, anogenital ulcerations, and mucositis. For EPD-like scalp eruptions and anogenital ulcerations, where validated grading systems are lacking, we introduce pragmatic severity-based grading frameworks. Emphasis is placed on early recognition, individualized supportive care, appropriate use of systemic therapies, and timely treatment modification to minimize unnecessary interruptions while ensuring patient safety.

CONCLUSIONS: Amivantamab is associated with a distinctive and clinically impactful toxicity profile that frequently complicates routine management. These consensus-based recommendations provide thoracic oncologists with practical tools to support early intervention, optimized supportive care, and rational treatment adaptation, potentially improving quality of life, reducing avoidable treatment interruptions, and supporting sustained oncologic benefit.