Markers for the early detection of colon cell proliferative disorders
Inventors
WARSINSKE, Hayley • Drake, Adam • PALANIAPPAN, Krishnan Kanna • O'DONOVAN, Brian D. • Hawkins, John
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12228574-B2
Publication Date
2025-02-18
Expiration Date
Abstract
Systems, media, compositions, methods, and kits disclosed herein relate to a panel of protein biomarkers for the early detection of colon cell proliferative disorders, including colorectal cancer. The presence or levels of the proteins in a biological sample for the protein panels described herein may be used for classifier generation, and as inputs in machine learning models useful to classify subjects in a population for the detection of colon cell proliferative disorders.
Core Innovation
The invention provides a non-invasive method of detecting colorectal cancer in a subject using a computer specifically programmed to detect the cancer. The method is based on obtaining a protein profile of the subject from a pre-determined protein panel, wherein the protein profile comprises a measured amount of a protein selected from a group of predefined proteins and is obtained or derived from a biological sample.
The computer detects an increased risk of the colorectal cancer by processing the protein profile with a trained machine learning model. The trained machine learning model is trained with training data comprising biological samples from subjects with advanced adenoma, control subjects without advanced adenoma, subjects with benign polyp, and control subjects without benign polyp, and is trained to differentiate between a presence or an absence of advanced adenoma and a presence or an absence of benign polyp.
Responsive to detecting the increased risk of the colorectal cancer, the method treats the subject with surgery, chemotherapy, immunotherapy, or radiotherapy. Dependent aspects further refine the method for specified colorectal cancer types and for specific biological sample selections, and may stratify colorectal cancer by sub-type or stage.
Claims Coverage
The document provides two independent claims that cover computer-implemented detection of increased risk for colorectal cancer using a pre-determined protein panel and a trained machine learning model trained on advanced adenoma versus control without advanced adenoma, and benign polyp versus control without benign polyp. The independent claims include an output value indicative of increased risk and subsequent treatment selection.
Pre-determined protein panel and measured protein profile
obtaining a protein profile of the subject comprising a measured amount of a protein from a pre-determined protein panel comprising at least six proteins selected from the group consisting of [named proteins], in a biological sample obtained or derived from the subject
Trained machine learning model for increased-risk output
processing the protein profile using a trained machine learning model, wherein the trained machine learning model is trained with training data comprising biological samples obtained or derived from subjects with advanced adenoma, control subjects without advanced adenoma, subjects with benign polyp, and control subjects without benign polyp, to provide an output value indicative of the increased risk of the colorectal cancer, wherein the trained machine learning model is trained to differentiate between a presence or an absence of advanced adenoma and a presence or an absence of benign polyp
Risk-responsive treatment for colorectal cancer
responsive to detecting the increased risk of the colorectal cancer, treating the subject with surgery, chemotherapy, immunotherapy, or radiotherapy for the increased risk of the colorectal cancer
Pre-determined protein panel with minimum size and risk detection workflow
obtaining a protein profile of the subject comprising a measured amount of a protein from a pre-determined protein panel comprising at least three proteins selected from the group consisting of [named proteins], in a biological sample obtained or derived from the subject
Increased-risk output from advanced-adenoma vs benign-polyp training sets
processing the protein profile using a trained machine learning model, wherein the trained machine learning model is trained with training data comprising biological samples obtained or derived from subjects with advanced adenoma, control subjects without advanced adenoma, subjects with benign polyp, and control subjects without benign polyp, to provide an output value indicative of the increased risk of the colorectal cancer, wherein the trained machine learning model is trained to differentiate between a presence or an absence of advanced adenoma and a presence or an absence of benign polyp
Across the two independent claims, the coverage centers on obtaining a protein profile from a pre-determined protein panel, detecting increased risk using a trained machine learning model trained on advanced adenoma versus control and benign polyp versus control to provide an output value, and treating the subject based on the detected increased risk. Independent claim 1 requires at least six proteins in the panel, while independent claim 14 requires at least three proteins in the panel.
Stated Advantages
Documented Applications
No documented applications found
Interested in licensing this patent?