Zeyun Qiao, Lei Tang, Nana Luo, Zilin Cheng, Pingsheng Hao
Relapse after treatment withdrawal remains a major challenge in psoriasis. Despite high rates of clinical clearance with modern systemic therapies, durable off-therapy control is uncommon, and recurrence often reappears at previously affected sites. Current models centered on resolved-skin inflammatory memory explain important aspects of local recall, but may not fully account for the heterogeneity of relapse timing, pattern and durability seen across patients. Here, we review psoriasis relapse through a clinically oriented multilevel framework that integrates three interacting dimensions: an antigenic dimension, blood-accessible relapse-associated signals and programs, and tissue permissiveness. Presentation-level antigen evidence raises the possibility that selected psoriatic phenotypes are associated with distinct upstream inflammatory programs, although direct links to relapse remain unproven. Circulating skin-homing memory populations, relapse-associated blood candidates and resident-derived mobile programs suggest, but do not establish, that relapse-relevant memory may remain accessible outside the skin during remission. At the tissue level, molecular scarring, epidermal tissue-resident memory T cells and stromal-epithelial support help explain why recurrence is preferentially re-executed in previously involved skin. We propose the Antigen-Blood-Tissue model as a falsifiable structure for interpreting relapse heterogeneity rather than a closed explanation of recurrence. Clinically, this framework suggests that visible clearance and biological stability should not be considered interchangeable. It also supports the view that remission may represent a biologically variable state, and that prospective studies aligned to treatment tapering or withdrawal will be needed to determine whether biological stability can eventually be defined beyond clinical remission and related prospectively to relapse outcomes.