Cancer remains a major cause of death worldwide and current therapies are often limited by toxicity, resistance and poor tumor selectivity. Photodynamic therapy (PDT) offers spatiotemporally controlled cytotoxicity but its efficacy is constrained by suboptimal photosensitizer delivery, tumor heterogeneity and the context-dependent role of autophagy. In this Keynote Review, we discuss how nanomedicine can be engineered for targeted, stimuli-responsive PDT while modulating autophagic flux to overcome resistance. We further highlight how artificial intelligence, including machine learning and deep learning, can integrate multi-omics and imaging data to guide target selection, nanocarrier design and personalized, autophagy-informed PDT strategies.
Esmaeilpour et al. (Sun,) studied this question.