Influence of precision treatment with artificial intelligence–assisted subtyping on first-line therapeutic efficacy in HR+/HER2– breast cancer.

Abstract

e13054 Background: The combination of CDK4/6 inhibitors and endocrine therapy (ET) has become the first-line standard treatment for HR+/HER-2- advanced breast cancer (ABC). A certain number of patients who are not sensitive to endocrine therapy and progress rapidly during CDK4/6 inhibitor treatment 20% to 30% of the patients experienced disease progression within 12 months. Screening of those insensitive individuals and further exploration of the optimal treatment strategies is currently the greatest challenge faced by HR+ ABC. Methods: This is a randomize, open-labeled, multi-center phase Ib/II trial (NCT05759572). AI-assisted digital pathology classified of SNF4 subtype patients had pathologically confirmed hormone receptor-positive, HER2-negative with untreated ABC were enrolled. In safety lead-in phase, 9 patients were enrolled (following the 3+3+3 protocol) to receive oral apatinib (250 mg per day) with dalpiciclib (125 mg per day for 3 weeks, followed by 1 week off) and endocrine therapy to determine the safety and dose for subsequent phase II part. In pahse II patients were randomly assigned (1:1) to receive dalpiciclib (125 mg per day for 3 weeks, followed by 1 week off) and endocrine therapy with or without oral apatinib (250 mg per day). Randomisation was stratified according to visceral metastasis, previous endocrine therapy in the adjuvant or neoadjuvant setting. Safety was analyzed in all randomly assigned patients who received at least one dose of study treatment. Results: Between March 1, 2023, and August 1, 2024, 157 patients were screened and 145 were eligible and enrolled. In safety lead-in phase 9 patients were enrolled and the recommended phase 2 dose was determined as oral apatinib (250 mg per day) and dalpiciclib (125 mg per day for 3 weeks, followed by 1 week off) with endocrine therapy. In phase II part 136 patients were randomly assigned to the apatinib+dalpiciclib with ET ( precision group, n = 68) or dalpiciclib with ET (control group, n = 68). Median progression-free survival was significantly longer in the dalpiciclib group than in the placebo group (27.8m vs 19.4m; stratified hazard ratio 0.57 [95% CI 0.36–0.91]; two-sided log-rank p = 0.017). Adverse events of grade 3 or 4 were reported in 72(93.5%) of 77 patients in the precision group and 62 (91.1%) of 68 patients in the control group. The most common adverse events of grade 3 or 4 were neutropenia (71 [92.2%] in the precision group vs 62 [91.1%] in the control group) and leukopenia (70 [91.0%] in the precision group vs 60 [88.2%]). Conclusions: AI-assisted digital pathology classification identified SNF4 patients who were resistant to ET combined with CDK4/6 inhibitor. The first-line treatment of apatinib combined with ET and CDK4/6 inhibitor, significantly improved the prognosis of these SNF4 patients with tolerated toxicity. Further prospective phase III trial has currently been conducted. Clinical trial information: NCT05759572 .