AI Unveils Immune Patterns to Predict Rectal Cancer Treatment Outcomes

A recent study conducted by researchers at University College London (UCL) and UCL Hospitals (UCLH) demonstrates that artificial intelligence (AI) can effectively predict treatment responses in rectal cancer patients. By analyzing standard tissue samples, the research highlights how the immune landscape around tumors influences patient outcomes.

The study, published in eBioMedicine, reveals that the immune microenvironment plays a critical role in cancer progression and treatment efficacy. Researchers explored routine pathology images and employed AI to identify and quantify key immune cells around rectal cancer tumors. This innovative approach aims to understand better how these immune characteristics affect patient survival and disease recurrence.

AI Enhances Traditional Pathology Analysis

Traditionally, pathologists examine biopsy slides under a microscope, a process that can be time-consuming and subjective. The research team aimed to leverage AI’s capabilities to recognize immune cell “signatures” in these images, potentially streamlining the analysis. Dr. Charles-Antoine Collins-Fekete, a senior author of the study, stated, “Pathology slides are already part of routine care, so they’re an abundant source of data.” He emphasized that AI can analyze these slides significantly faster and more cost-effectively than conventional methods, such as whole-genome sequencing.

The study analyzed samples from three patient groups, including participants from the ARISTOTLE clinical trial. Results indicated that patients with higher levels of lymphocytes—immune cells that combat infections and cancer—tended to have improved survival rates and lower chances of cancer recurrence. Conversely, those with elevated macrophage levels—immune cells that can inadvertently support tumor growth—exhibited poorer outcomes.

Linking Immune Features and Genetic Factors

The AI system was trained on millions of pathology images and evaluated 900 patient samples. It effectively measured immune cell levels before and after treatment. Findings revealed that patients experiencing an increase in tumor-infiltrating lymphocytes often had better outcomes, as chemotherapy can stimulate an active immune response. In contrast, those whose tumors remained “cold” immunologically faced a higher likelihood of early recurrence.

The researchers also investigated the impact of genetic changes on immune responses. For instance, patients with a normal KRAS gene and elevated lymphocyte counts had better survival rates than those with KRAS mutations and fewer lymphocytes. Additionally, high macrophage levels were particularly detrimental in patients with mutations in the TP53 gene.

Dr. Zhuoyan Shen, the first author of the study, noted that while experienced pathologists can recognize some immune features, these insights are not typically incorporated into treatment decisions. The AI method identifies hidden immune signatures, providing biological insights similar to those obtained through costly techniques like whole-genome sequencing.

The combined analysis of immune and genetic data allows for a clearer understanding of cancer behavior, aiding in the classification of patients into high- and low-risk groups. This stratification can inform treatment decisions, such as administering more aggressive therapies to high-risk patients while reducing exposure for those at lower risk.

The research team discovered that tumors with high mitotic activity, which indicates rapid cell division, tend to suppress immune responses, leading to poorer patient outcomes. This finding suggests that fast-growing cancers may pose greater challenges for the immune system.

To facilitate access to their findings, the researchers developed a free online tool called Octopath, allowing clinicians to upload pathology slides for automated immune analysis. While the tool shows promise, further research is necessary to validate the results across larger and more diverse patient populations. The team aims to explore additional immune cell types and advanced methods to deepen the understanding of cancer-immune interactions.

Professor Maria Hawkins, a senior author and UCLH consultant clinical oncologist, expressed optimism about the potential of AI in cancer classification. She stated, “This is an early step towards the use of AI to aid the further classification of cancer. In the future, clinicians and patients will discuss personalized treatment based on timely information provided by AI.”

The study signifies a significant advancement in using AI to enhance cancer treatment approaches, potentially leading to improved patient outcomes through tailored therapies. Further research will be crucial in integrating these findings into everyday clinical practice.