Artificial Intelligence Can Cut Costs While Maintaining Accuracy in Colorectal Cancer Genotyping

Area of Science:

• Oncology
• Health Economics
• Medical Diagnostics

Background:

• Rising cancer care costs present a significant financial burden on healthcare systems.
• Artificial intelligence (AI) offers potential for cost reduction in diagnostic testing and therapy expenditures.
• Mismatch repair deficiency/microsatellite instability (dMMR/MSI) status is critical for guiding first-line metastatic colorectal cancer (mCRC) treatment.

Purpose of the Study:

• To evaluate the financial and clinical impact of AI-based dMMR/MSI testing in first-line mCRC.
• To compare eight different diagnostic testing strategies, including AI-driven approaches versus traditional methods.

Main Methods:

• A deterministic model was developed to simulate eight distinct testing strategies for dMMR/MSI status.
• A hypothetical, nationally representative US population sample (N=32,549) of newly diagnosed mCRC patients was used.
• Model inputs were derived from peer-reviewed literature and Medicare data to estimate costs and clinical outcomes.

Main Results:

• The strategy of high-sensitivity AI followed by high-specificity PCR/IHC for AI-negative cases yielded the greatest cost savings ($400 million, 12.9%) compared to NGS alone.
• The high-specificity AI-only strategy demonstrated the best clinical impact, achieving 97% diagnostic accuracy and reducing time to treatment initiation to under one day.
• AI integration has the potential to decrease diagnostic costs and time to treatment in mCRC without compromising accuracy.

Conclusions:

• AI-based diagnostic tools show promise in reducing both financial costs and treatment initiation times for mCRC.
• The value proposition of AI in cancer diagnostics is expected to grow with improved accuracy and reduced processing costs.
• Health systems should consider integrating AI-driven histopathology into diagnostic protocols for mCRC management.