Discovery of QLS1522, a mutant-selective allosteric PI3Kalpha inhibitor for the treatment of PIK3CA-mutant solid tumors

Background: PIK3CA is a predominant oncogene in human cancers. While approved PI3Kalpha inhibitors like Alpelisib validate that PI3Kalpha blockade can suppress tumors and improve outcomes in PIK3CA-mutant ER+HER2- breast cancer, their major on-target toxicity against wild-type PI3Kalpha interferes with regulation of glucose metabolism which reduces their therapeutic index and limits clinical efficacy. Development of PIK3CA mutant selective inhibitors can significantly improve therapeutic window and anti-tumor efficacy by substantially reducing side effects. Experimental procedures: The enzymatic activity was evaluated in biochemical and biophysical assays. Viability of cell lines was measured using CellTiterGlo®. Antitumor activity was evaluated in ER+ breast cancer cell line-derived xenograft (CDX) models with either kinase domain or helical domain mutations. The impact on glucose metabolism was analyzed using an oral glucose tolerance test (OGTT). Results: QLS1522 is a potent, selective allosteric PI3Kalpha inhibitor designed to spare wild-type PI3Kalpha. It exhibited nanomolar potency against the PIK3CA H1047R mutation with high selectivity over wild-type PI3Kalpha and other isoforms (beta, gamma, delta). In a 447-kinase selectivity panel, QLS1522 showed outstanding specificity, with only Aurora-B kinase exhibiting >50% inhibition at 10 µM (IC₅₀ of 7.8 µM for QLS1522 vs. 1.4 µM for STX-478). Cell panel studies revealed that QLS1522 exerted broad inhibitory effects across cancer cell lines carrying either kinase domain or helical domain mutations. In comparative analyses with other clinical-stage candidates such as STX-478 and RLY-2608, QLS1522 showed enhanced inhibition of pAKT in the T47D breast cancer cell line harboring the PIK3CA H1047R mutation. Notably, it maintained selectivity comparable to STX-478 and better than RLY-2608 in wild-type SKBR3 cells. In vivo , QLS1522 was well tolerated and exhibited potent antitumor efficacy in multiple ER+ breast cancer CDX models, without inducing significant glucose metabolism abnormalities. The combination of QLS1522 with standard-of-care (SoC) agents resulted in enhanced efficacy compared to monotherapy or SoC alone. QLS1522 demonstrated a favorable drug-like profile, characterized by superior solubility, permeability, and excellent pharmacokinetic properties across species, with a wide therapeutic window. Conclusions: QLS1522 is a mutant-selective allosteric PI3Ka inhibitor and represents a promising therapeutic candidate for treating PIK3CA-mutant solid tumors. IND-enabling studies are currently in progress.