Pathological image profiling identifies onco-microbial, tumor immune microenvironment, and prognostic subtypes of...

Area of Science:

• Oncology
• Microbiology
• Genomics
• Pathology

Background:

• Colorectal cancer (CRC) prognosis is influenced by histology, tissue microbes, and host gene expression.
• The complex interplay between these multimodal omics in CRC remains poorly understood.

Purpose of the Study:

• To investigate the associations between pathological images, intratumoral microbes, and host gene expression in CRC.
• To propose a novel CRC classification based on histology slide features.
• To elucidate the biological significance of microbe-host interactions within the tumor immune microenvironment (TIME).

Main Methods:

• Comprehensive profiling of pathological images, intratumoral microbes, and host gene expression in 527 CRC patients.
• Clustering patients based on histology slide features to define histology slide subtypes (HSS).
• Analysis of onco-microbial communities and tumor immune microenvironment (TIME) differences between HSS.

Main Results:

• Two distinct histology slide subtypes (HSS1 and HSS2) were identified in CRC patients.
• Significant differences in onco-microbial communities and TIME were observed between HSS1 and HSS2.
• Intratumoral microbe-host interactions were associated with prognostic heterogeneity between the subtypes.

Conclusions:

• Pathological image features can define distinct CRC subtypes with unique microbial and immune profiles.
• Tumor microbe-host interactions significantly impact CRC prognosis, mediated through the TIME.
• This study offers a new framework for classifying CRC based on histology and understanding its underlying biological mechanisms.