Precision treatment with artificial intelligence assisted subtyping enhances therapeutic efficacy in HR+/HER2- breast...

Area of Science:

• Oncology
• Genomics
• Artificial Intelligence in Medicine

Background:

• Hormone receptor-positive/human epidermal growth factor receptor 2-negative (HR+/HER2-) metastatic breast cancer often develops resistance to cyclin-dependent kinase 4/6 inhibitors (CDK4/6i).
• Identifying effective precision treatments for this patient population remains a clinical challenge.

Purpose of the Study:

• To evaluate the efficacy of subtyping-based precision therapy versus treatment of physician's choice in HR+/HER2- metastatic breast cancer patients with acquired CDK4/6i resistance.
• To identify distinct patient subtypes using artificial intelligence (AI) for targeted treatment selection.

Main Methods:

• The LINUX trial (NCT05594095) was a multicenter, randomized, controlled phase II platform trial.
• 105 patients were classified into four similarity network fusion (SNF) subtypes using AI.
• Patients were randomized to receive either subtyping-based precision therapy (N=70) or treatment of physician's choice (N=35).

Main Results:

• Objective response rates (ORR) were superior in subtyping-based groups compared to controls across SNF subtypes: SNF1 (10% vs 0%), SNF2 (65% vs 30%), SNF3 (40% vs 30%), and SNF4 (70% vs 20%).
• Subtyping-based precision therapy demonstrated significant benefits, especially for SNF2 and SNF4 subtypes.
• Grade 3-4 treatment-related adverse events were similar between groups (37%).

Conclusions:

• AI-driven subtyping can guide effective precision therapy selection for HR+/HER2- metastatic breast cancer post-CDK4/6i resistance.
• Subtyping-based precision therapy offers a superior treatment strategy compared to physician's choice for specific patient subgroups.
• Further validation in phase III trials is warranted to confirm these findings.