Abstract
Vitiligo-like depigmentation (VLD) arising under immune checkpoint blockade in melanoma is associated with objective tumour response and substantially longer survival. Its prognostic value is not coincidental: depigmentation and tumour regression share a CD8⁺ effector programme directed at melanocyte differentiation antigens and recruited through the IFN-γ-CXCL9/CXCL10-CXCR3 axis, an axis independently required for the efficacy of PD-1 blockade. The depigmenting lesion is therefore an accessible compartment in which the antitumour effector response could, in principle, be measured directly. This opportunity is currently unexploited: VLD is graded by inspection, and its molecular correlates are sought almost exclusively in plasma, where a locally generated chemokine signal is diluted into a whole-body average confounded by tumour burden, concurrent immune-related adverse events and corticosteroid exposure. Paired-matrix data from spontaneous vitiligo show that the same chemokines discriminate active from stable depigmentation several-fold in lesional skin interstitial fluid whilst discriminating poorly in serum from the same patients, a finding not yet tested in the ICI setting. Building on this observation as an explicit hypothesis, we reframe the problem as one of matrix selection, pre-analytical control and analytical validation: we compare skin-proximal and systemic matrices against tumour biopsy and circulating tumour DNA, define the pre-analytical variables that are peculiar to oncology, and outline a translational roadmap by which a validated cutaneous assay could convert a visible prognostic sign into a quantitative, early and repeatable readout of antitumour immunity.