U Sin Wong, Dajiang Guo
Late-stage cutaneous melanoma management has already evolved into systemic therapy thanks to the fast development of novel targeted inhibitors and immune checkpoint blockers. Therefore, a reliable prediction of systemic therapy response became crucial for making a personalised management plan for melanoma patients. Although there are multiple biomarkers already available, the overall predictive confidence remains low due to heterogeneous responses among patients. Therefore, novel solid predictive biomarkers are still greatly needed. With the development of new technologies and artificial intelligence, new predictive models are being proposed and generated. In this review, we thoroughly evaluated the established biomarkers that are already in use for clinical practice. In addition, we proposed the emerging new markers with great potential by critically reviewing the literature. Furthermore, we proposed a future framework for making a tailored clinical management plan for melanoma patients.