Colorectal Cancer

By understanding how the spatial arrangement of colorectal tumors contributes to disease development, we aim to improve approaches to treating the disease.

In the United States, colorectal cancer is the second most common cause of cancer-related death [i]. Although CRC is most common in older adults, cases in adults under 50 are rising [ii]. CRC remains challenging to treat - especially when diagnosed at late stages - and so there is a significant need for improved treatment strategies. 

LSP investigators study primary CRC tumors to understand how the microenvironment of CRC tumors contributes to disease development and progression. We recently generated the first ever 3-dimensional tissue atlas (Lin et al. Cell, 2023), which revealed the complex, interconnected 3D architecture and immune interactions within colorectal tumors. We also applied machine learning models to high-dimensional imaging data to identify tissue features - a combination of molecular and spatial patterns - associated with certain patient outcomes. Although these studies were just for proof-of-concept, they highlight the potential value of integrating image data with ML/AI models.

Future work is leveraging larger cohorts to ask whether tissue features can improve the early detection of colorectal cancer or predict a response to treatment. In the long term, these efforts could help stratify patient populations, making it possible to tailor a particular treatment to the subset of patients most likely to respond.