Abstract PS7-03: A first-in-human phase 1a/b trial of LOXO-783, a potent, highly mutant-selective, brain-penetrant, allosteric PI3Kα H1047R inhibitor in PIK3CA H1047R-mutant advanced breast cancer and other solid tumors: Results from the PIKASSO-01 study

Abstract Background: Phosphoinositide 3-kinase alpha (PI3Kα) H1047R mutations occur in ∼15% of breast cancers (BC). Approved treatments for PI3Kα mutant-driven BC include inhibitors of the PI3K/AKT pathway; these block signaling from mutant PI3Kα but also inhibit wild-type PI3Kα leading to dose-limiting toxicities (DLTs) that include hyperglycemia, rash, and GI side effects. Thus, there is a need for novel PI3Kα inhibitors with improved therapeutic indices. LOXO-783 is an oral, brain-penetrant allosteric PI3Kα inhibitor, highly selective for H1047R that induces tumor regressions in ER+, HER2- PI3Kα H1047R-mutant BC models without increasing plasma insulin or C-peptide. Here we present clinical data from PIKASSO-01, a global, first-in-human phase 1a/b trial of LOXO-783 as monotherapy and in combination with other anticancer therapies in patients (pts) with PIK3CA H1047R-mutant advanced BC (aBC) and other solid tumors (NCT05307705). Methods: Phase 1a (ph1a) evaluated LOXO-783 dose escalation (mTPI-2 design) in pts with PIK3CA H1047R-mutant solid tumors (excluding colorectal cancers). Phase 1b (ph1b) included the evaluation of LOXO-783 plus (+) endocrine therapy (ET: fulvestrant [F] / aromatase inhibitor [AI]), or LOXO-783 + paclitaxel (P) in PI3Ki-naïve/intolerant pts with PIK3CA H1047R aBC. Key objectives included safety, tolerability, PK, and antitumor activity (objective response rate, ORR and clinical benefit rate, CBR per RECIST v1.1). Serial plasma samples were collected for ctDNA analysis and glucose metabolism markers. Results: As of 14 June 2024, 149 pts were enrolled: 43 pts (35 aBC [30 HR+ HER2-, 5 TNBC], 8 other solid tumors) into ph1a and 106 pts with aBC (90 HR+ HER2-, 16 TNBC) into ph1b. Median age was 58 years. In ph1a, pts received monotherapy doses from 200 – 600 mg BID or 500 – 600 mg QD. Median prior regimens was 3 (1-5) including prior PI3K/AKT/mTORi (28%). DLTs occurred in 4 pts, 1 each at: 500 mg QD (grade [G] 3 fatigue), 600 mg QD (G3 ALT, AST, and G1 bilirubin elevation), 400 mg BID (G3 sepsis and enterocolitis), and 600 mg BID (G2 photosensitivity). In ph1b, pts received LOXO-783 (200 – 400 mg BID or 300 – 600 mg QD) in doublet therapy: + ET (n=86; 6 + AI; 80 + F) or + P (n=20). Median prior regimens for aBC was 2 (1-7). Across all treated pts in the monotherapy or doublet therapy cohorts, diarrhea was frequently observed (84% any grade [G], 7% G ≥3) and was more common/severe (G ≥3) at higher doses (≥400 mg QD and BID dosing). Overall, 37% of pts who experienced diarrhea required a dose modification. Other frequent (any G/G ≥3) treatment-emergent AEs with LOXO-783 +/- ET included fatigue (32%/2%) and nausea (31%/2%); + P included neutrophil decreased (65%/45%), anemia (40%/20%), fatigue (35%/0%), and WBC decreased (30%/25%). Treatment-related AEs led to discontinuation of any study treatment in 4% and 10% of pts who received LOXO-783 +/- ET and + P respectively. No treatment-related hyperglycemia was observed. Preliminary PK analyses showed dose-dependent and time-dependent nonlinear increases in LOXO-783 exposure across all doses (reaching H1047R IC90 from 200 mg BID + ET); tmax ∼2 h and half-life ∼19 h; and no combination drug-drug interactions. Minimal activity was observed with LOXO-783 monotherapy (ORR/CBR in evaluable pts: 3%/17% [n=29]). In doublet therapy cohorts, ORR/CBR was 5%/19% + ET (n=77) and 19%/25% + P (n=16). Most (80%; 85/106) evaluable pts had a decrease (any) in PIK3CA H1047R variant allele frequency at C1D15 compared to baseline. Conclusions: LOXO-783 demonstrated proof of concept of mutant selectivity by entirely sparing hyperglycemia in the clinic, but exhibited high rates of diarrhea, limiting ability to achieve the optimal preclinical dose of LOXO-783. Citation Format: Komal Jhaveri, Takako Eguchi Nakajima, Antonio Giordano, Dejan Juric, Cynthia Ma, Nisha Unni, Matthew P. Goetz, Shigehisa Kitano, Sherene Loi, Hope Rugo, François-Clément Bidard, Cristina Saura, Elgene Lim, Patrick Neven, Joyce O’Shaughnessy, Manali Bhave, Tira J. Tan, Philippe L. Bedard, Hélène Vanacker, Ruth O’Regan, Nick C. Turner, Javier Cortes, Stephen Chia, Mario Campone, Vincent Chau, Matthew P. Hanley, Lin Du, Aurelie Lombard, Monica Ramstetter, Shawn Estrem, Funda Meric-Bernstam. A first-in-human phase 1a/b trial of LOXO-783, a potent, highly mutant-selective, brain-penetrant, allosteric PI3Kα H1047R inhibitor in PIK3CA H1047R-mutant advanced breast cancer and other solid tumors: Results from the PIKASSO-01 study [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr PS7-03.