Disease biomarkers play important roles in modern medicine， enabling early disease diagnosis， precise subtyping， prognostic evaluation， and targeted therapy. Conventional biorecognition elements， such as antibodies and aptamers， offer high specificity but suffer from inherent limitations， including high production costs， instability in complex biological matrices， and significant batch-to-batch variation. Collectively， these constraints restrict their scalability and versatility for high-throughput， cost-effective， and multi-scenario clinical applications. Molecularly imprinted polymers （MIPs）， prepared via molecular imprinting technology （MIT）， have emerged as highly robust and promising synthetic receptors. During their formation， functional monomers assemble around a target molecule， and highly specific binding sites are formed after polymerization and template removal. MIPs exhibit a unique combination of advantages， ranging from cost-effectiveness to high selectivity， as well as high affinity comparable to natural counterparts， and outstanding physicochemical stability under harsh conditions. Driven by their highly flexible and customizable design capabilities， MIP development has witnessed remarkable advancements in recent years. These polymers are suitable for analyzing different types of disease biomarkers， ranging from proteins， peptides， and saccharides to complex biological entities such as whole cells and extracellular vesicles. Moreover， MIPs can be functionally engineered to integrate signal transduction or stimuli-responsive features， facilitating the creation of intelligent biomedical platforms. This review systematically summarizes the progress in applying MIT for disease diagnostics and therapeutics. It first elaborates on diverse imprinting techniques specifically tailored for different biomarker classes， including proteins， peptides， saccharides， cells， and extracellular vesicles. Subsequently， the article provides a comprehensive overview of their applications in diagnostics， encompassing biosensing， bioimaging， and bioseparation， as well as therapeutic applications， including drug delivery， photothermal and photodynamic therapy， biotoxin removal， and cell behavior regulation. Additionally， critical challenges hindering clinical translation are discussed， such as biocompatibility， long-term toxicity， high large-scale manufacturing costs， and the lack of standardized clinical validation protocols. Finally， promising future directions are outlined， emphasizing the development of biodegradable materials， integration with artificial intelligence， and adoption of green synthesis strategies. These synergistic approaches are expected to stimulate innovation， enabling the safe， reliable， and scalable translation of MIPs into real-world biomedical and clinical applications. In summary， MIPs represent a versatile， robust， and economically viable alternative to conventional biorecognition elements. Their customizable nature， combined with functional engineering capabilities， positions MIPs as a promising technology in disease diagnosis and therapy. Future efforts focusing on clinical translation， sustainability， and intelligent platform integration are likely to accelerate their adoption across diverse biomedical fields.
