Drug Repositioning and the Path From Candidate Identification to New Therapeutic Frontiers

Drug discovery is an expensive and time-consuming process that is associated with high failure rates and complex regulatory hurdles. Drug repurposing is a promising alternative to the conventional approach of drug discovery that can help meet the challenge of rapidly increasing demand for therapeutically active compounds. It allows reducing the risks and costs of bringing a drug to the market by repositioning an already known or withdrawn drug. Numerous computational biology approaches, real-world evidence, genomic, and experimental screening methods have contributed to the fundamental change in drug repurposing from an academic exercise to a data-driven field during the last decades. The review presents a wide range of modern drug repositioning approaches including clinical, computational, network-driven, genomics-based prioritization, and high-throughput screening. The methods have proved effective at increasing the number of therapeutically relevant drug candidates. However, the review indicates that successful long-term outcomes in clinical practice are not guaranteed even when the candidates are identified due to various factors including the limitations of biological evaluation models, insufficient mechanistic understanding, and disease heterogeneity and other issues. In order to underline the significance of the validation and the translation beyond the mere identification, Alzheimer’s disease is used as an illustrative example. By combining the data from invitro, animal, and human studies, this review highlights the methods of assessing the repurposed candidates’ safety and effectiveness with the focus on the highly diverse and context-dependent results. Thus, the drug repurposing is proved to be not only a way of discovering new indications but also a process of validating and translating the findings. Taken together, this review highlights the need to improve drug repurposing frameworks by incorporating disease-relevant validation paradigms that employ computational prediction in conjunction with experimental and clinical assessment to reduce failure rates and increase the likelihood of successful translation.