Award Details (original) (raw)

Grant ID

Project Title

Personalized Recurrence Risk Prediction for Patients with Colorectal Cancer Precursors: A Digital Pathomics and Spatial Proteomics Approach

Award Amount

$960,000.00

Primary Organization

President and Fellows of Harvard College, Harvard Medical School

Award Start Date - Award End Date

07/01/2024 - 06/30/2028

Program Name

Research Scholar Grants

Colorectal cancer precursor lesions affect 40% of people aged 50-75 years in the U.S. Approximately 60% of patients with these lesions will experience a second lesion (commonly referred to as “recurrence”), and some will develop colorectal cancer if untreated. To prevent the recurrence and development of cancer, the current clinical guidelines use the number of lesions and visual features of the lesions under the microscope to determine the clinical follow-up schedules. However, patients with the same number of lesions and the same microscopic visual features have different risks of recurrence. Unnecessary treatments in low-risk patients cause serious complications (including massive bleeding, infection, and broken bowel). Thus, we need an accurate recurrence prediction model to reduce the health risks and healthcare costs related to managing colorectal cancer precursors. To address this unmet clinical need, we will establish an artificial intelligence (AI) system to predict the recurrence risk of precursor lesions by connecting microscopic image patterns, results of molecular analyses, and patients’ clinical profiles. We have assembled a team of experts in artificial intelligence, pathology, and colorectal cancer screening and treatment. We will employ large pathology datasets collected at the Brigham and Women's Hospital and the Massachusetts General Hospital to (i) develop an AI-based prediction model for colorectal cancer precursor recurrence risks using microscopic imaging features, (ii) identify the molecular indicators of recurrence risk using high-resolution protein profiling methods, and (iii) integrate microscopic, molecular, and clinical data to predict the recurrence risks. Our approach is innovative because it connects complementary signals from microscopic analyses, molecular, and clinical profiles and leverages new technologies in AI, high-resolution digital microscopy imaging, and detailed protein analyses of colorectal tissue samples. The proposed research is significant because it will expand our knowledge of the connections among microscopic imaging patterns, molecular variations, and disease recurrence risks. The knowledge obtained from this study will improve the current clinical guidelines for colorectal cancer screening. Thus, our study will have a positive impact on patients by providing evidence-based clinical decision-making, minimizing unnecessary treatments, and enabling personalized patient care.