Phenotypic subtypes predict outcomes in colorectal cancer

Area of Science:

• Oncology
• Molecular Pathology
• Cancer Genomics

Background:

• Colorectal cancer (CRC) is a leading global cause of cancer mortality.
• Transcriptome-based classification identified four molecular subtypes (CMS1-4) with distinct genomic profiles and prognoses.
• There is a need for simpler, phenotype-based methods for clinical CRC subtyping.

Purpose of the Study:

• To develop an immunohistochemistry-based method for classifying colorectal cancer into four phenotypic subtypes.
• To analyze disease-specific survival (DSS) among these phenotypic subtypes.
• To investigate associations between phenotypic subtypes and clinicopathological variables.

Main Methods:

• Categorized 480 surgically treated CRC patients into four phenotypic subtypes (immune, canonical, metabolic, mesenchymal) using immunohistochemistry.
• Utilized CD3-CD8 tumor-stroma index, proliferation index, and tumor-stroma percentage for subtyping.
• Analyzed survival rates using Kaplan-Meier and Cox regression; examined associations with clinicopathological variables via chi-square test.

Main Results:

• Immune subtype tumors demonstrated the best 5-year DSS; mesenchymal subtype tumors had the worst prognosis.
• Canonical subtype prognostic value varied significantly across clinical subgroups.
• Immune subtype associated with female sex, stage I, and right-sided tumors; metabolic with pT3/pT4 and male sex; mesenchymal with stage IV, mucinous histology, and rectal location.

Conclusions:

• Phenotypic subtype classification effectively predicts colorectal cancer patient outcomes.
• Subtype associations and prognostic values mirror transcriptome-based consensus molecular subtypes (CMS).
• The immune subtype offers an exceptionally good prognosis; further research should explore transcriptome-phenotype concordance.