The study, led by researchers from the University of Chicago Medicine Comprehensive Cancer Center, in collaboration with scientists from InSilico Medicine and Prelude Therapeutics, found that blocking the activity of protein arginine methyl transferase 5 (PRMT5) is a potential treatment strategy against ACC. ACC accounts for only 1-5% of head and neck cancers and 25-35% of salivary gland neoplasms.

The disease itself is very, very, very rare, which makes it very difficult to study, explained Evgeny Izumchenko, PhD, Assistant Professor of Medicine at UChicago. He clarified that, in addition to being rare, it is also hard to detect early because patients tend to not show symptoms until it has progressed significantly. Not much is known about this disease not much is known about how to treat it; you don't have the rich history of treated patients that you can look back to and define what would be the best approach, Izumchenko said.

Given the lack of targeted therapies for ACC, the team turned to artificial intelligence (AI), which has been gaining traction in the discovery of novel therapeutic targets. Using an AI-based predictive discovery tool, the team analyzed gene expression data from 87 ACC tumor samples and 35 matched normal controls to identify potential drug targets. Among the top-ranked candidates was PRMT5, an enzyme implicated in epigenetic regulation (altering gene expression and protein activity without changing the DNA sequence itself) and known to play a role in cancer development.

Once they identified PRMT5 as a promising target, the researchers collaborated with Prelude Therapeutics, a company that had developed a highly selective PRMT5 inhibitor called PRT543. The researchers evaluated PRT543 in cellular and animal models, including ACC cell lines (cells derived from cancer tissues), organoids (3D tumor models derived from patient samples), and patient-derived xenografts (PDXs), which are human tumors implanted in mice.

Source: https://www.uchicagomedicine.org/forefront/cancer-articles/ai-helps-find-promising-therapeutic-target