Quantitative Analysis of Rectal Cancer Biopsies With the Digital Pathology Segmentation Algorithm QuantCRC Associates...

Area of Science:

• Digital pathology
• Oncology
• Computational pathology

Background:

• Neoadjuvant therapy is standard for rectal cancer.
• Predicting pathologic complete response (pCR) and recurrence-free survival (RFS) is crucial.
• Digital pathology algorithms offer potential for outcome prediction.

Purpose of the Study:

• To evaluate the QuantCRC digital pathology algorithm for predicting pCR, RFS, and correlating with transcriptomic profiling in rectal cancer.
• To assess QuantCRC's association with clinical outcomes in pretherapy biopsies.

Main Methods:

• QuantCRC algorithm applied to 288 (training) and 37 (validation) pretherapy rectal cancer biopsies.
• Associations with pCR and RFS analyzed using logistic regression and Cox modeling.
• Correlation with transcriptomic digital spatial profiling performed.

Main Results:

• QuantCRC-derived tumor-infiltrating lymphocytes (TILs) associated with pCR (OR=1.05, P=.038) and were higher in pCR cases (91.3 vs 55.9/mm², P=.004).
• QuantCRC %high tumor grade independently predicted worse RFS (HR=1.27, P=.002), with reduced 5-year RFS for patients with ≥10.1% high grade (69% vs 83%, P=.007).
• High IL-6/JAK/STAT3 signaling associated with worse RFS; low IL-6/JAK/STAT3 expression linked to higher TILs.

Conclusions:

• QuantCRC-derived TILs in pretherapy biopsies predict pCR to neoadjuvant therapy in rectal cancer.
• QuantCRC %high tumor grade predicts recurrence risk, identifying patients needing closer monitoring.
• Digital pathology analysis of rectal cancer biopsies shows promise for personalized treatment strategies.