Phosphoinositide 3-kinase-based dual-target inhibitors for cancer therapy: rational design, structure-activity relationships and biological evaluation

Background and purpose: The phosphoinositide 3-kinase (PI3K)/AKT/mTOR pathway is dysregulated in 30-50 % of human cancers. PI3K inhibition is clinically validated, yet of the seven PI3K pathway inhibitors approved by the US Food and Drug Administration, four have had indications withdrawn and two are no longer marketed, principally because narrow therapeutic windows and overall-survival detriments offset progression-free survival gains. Pathway reactivation and adaptive resistance further limit single-agent activity, motivating single-molecule dual-target inhibitors. Experimental approach: A structured search (January 2019 - July 2026) retained original medicinal-chemistry reports of single chemical entities engaging PI3K with a second oncology target, for which biochemical potency against both targets and an interpretable structure-activity relationship were reported. Five combinations dominated retrieval - PI3K/mTOR, PI3K/HDAC, PI3K/PARP, EGFR/PI3K and PI3K/BRD4 - and twenty representative leads are analysed. Key results: Each compound is assigned an explicit level of evidence for dual-target engagement, distinguishing measured biochemical and cellular target engagement from structural hypotheses generated by docking; nine of twenty leads reach the highest tier. In vivo pharmacokinetics are available for only seven leads and plasma protein binding for none, so a uniform in silico ADMET profile is reported for all compounds. Predictions reproduced the solubility-permeability trade-off and indicated near-complete plasma protein binding for the most lipophilic hybrids, but under-predicted clearance by up to two orders of magnitude and returned false-positive hERG calls for every compound in which the liability was measured. Solubilizing groups on the solvent-exposed vector improved solubility more than 100-fold without loss of dual potency, whereas carboxylic acids, hydroxamate zinc-binding groups and N-acylhydrazone linkers each improved one property at a cost in potency, permeability, oral bioavailability or chemical stability. Conclusion: Dual-target PI3K inhibition is now clinically validated: the pan-PI3K/mTORC1/2 inhibitor gedatolisib was approved in July 2026. The principal obstacles are not biochemical potency, which is routinely sub-nanomolar, but the 80- to 340-fold potency imbalance between the two targets of most hybrids, the lack of experimental DMPK data, low oral bioavailability of benzamide-based hybrids, and a regulatory environment now requiring randomized evidence of survival benefit.