What practicing pathologists and oncologists should know about the new computational pathology-based companion...

Area of Science:

• Digital pathology
• Computational pathology
• Artificial intelligence in medicine

Background:

• AI integration in pathology transforms histological and immunohistochemical (IHC) slide assessment.
• Current AI tools detect, grade, and classify diseases, moving towards predicting biomarkers and patient prognosis.
• Emerging AI applications include companion diagnostic assays for predicting drug response.

Purpose of the Study:

• To provide an overview of novel computational pathology companion diagnostic tools.
• To inform pathologists and oncologists about transitioning AI tools from research to clinical practice.
• To highlight the potential of AI in improving cancer patient care through personalized medicine.

Main Methods:

• Review of current AI applications in pathology, including supervised machine learning models.
• Analysis of studies demonstrating AI's superiority in assessing IHC markers like TROP2, HER2, and PD-L1.
• Discussion of the development and clinical translation of computational pathology tools.

Main Results:

• Supervised machine learning models show superiority in assessing IHC markers compared to human scoring (e.g., TROP2 in TROPION-PanTumor01 trial).
• AI approaches demonstrate promise in refining evaluations of HER2 and PD-L1, identifying patient subgroups for targeted therapies.
• Novel computational tools are being developed as companion diagnostics, linking biomarker evaluation to drug response prediction.

Conclusions:

• AI is revolutionizing pathology by enabling more accurate and efficient diagnostics.
• Computational pathology companion diagnostics represent a significant advancement in personalized cancer treatment.
• Understanding and adopting these AI tools is crucial for clinicians to improve patient outcomes.