Synthetic lethality in cancer drug discovery: challenges and opportunities

Synthetic lethality, first proposed more than two decades ago, has long held immense promise for targeted cancer therapy. Although the clinical success of PARP inhibition in BRCA-mutant cancers stands as proof of concept, few other synthetic lethal interactions have been translated from preclinical findings into effective therapies. This slow pace of translation stems in part from the difficulty of developing drugs against genetic dependencies, but also reflects the cell- and tissue-specific nature of these interactions. In this Review, we outline recent advances in the discovery and validation of synthetic lethal pairs, from their discovery in large-scale genetic screens to the development of drugs for the clinic. We discuss how alternative CRISPR-based approaches — including combinatorial screens, base editing and saturation mutagenesis — are now being used to discover new tractable interactions. We also examine how machine learning models can enable prioritization of candidate pairs and the identification of biomarkers for patient stratification. Finally, we highlight alternative phenotypic readouts, such as high-content imaging and single-cell profiling, which enable the dissection of phenotypes beyond simple cell growth or fitness. Together, these developments are refining the synthetic lethality paradigm and advancing its potential for cancer therapy.