#USA #Caris Life Sciences #Irving #AI Breast Cancer #NRG Oncology

Introduction

Caris Life Sciences has made significant strides in the realm of breast cancer research with the recent publication of a study in Cancer Research Communications. This innovative research showcases a multimodal and multitask deep learning model designed to accurately predict late distant recurrence risk in hormone receptor-positive (HR+) early breast cancer patients. Given that HR+ breast cancer accounts for a substantial 70-80% of all breast cancer cases, this study's findings could greatly impact treatment approaches, particularly concerning extended endocrine therapy decisions.

Study Overview

The study, titled "Development and Validation of a Multimodal-Multitask Deep Learning Approach for Estimating Late Distant Recurrence Risk in Hormone Receptor–Positive Early Breast Cancer," was conducted in collaboration with prominent institutions, including the NSABP Foundation/NRG Oncology and the ECOG-ACRIN Cancer Research Group. The three entities pooled their expertise to explore how artificial intelligence could provide insights into the recurrence risks facing patients, thereby aiding in the decision-making process regarding prolonged therapy.

The researchers focused on the risk posed by late distant recurrence in HR+ early breast cancer cases—a scenario that often persists even after five years of conventional endocrine therapy. While extending endocrine therapy for an additional five years may lower this risk, it can lead to adherent side effects, adding a layer of complexity to treatment decision-making.

Groundbreaking Insights

Dr. George W. Sledge, Chief Medical Officer for Caris Life Sciences, emphasized the transformative potential of AI in deriving valuable clinical insights from routine data collection. By merging AI-driven analysis of standard pathology images with essential clinical data, the model is aimed at providing precise risk stratification. This, in turn, can significantly enhance individualized treatment decisions.

The AI model processes digitized hematoxylin and eosin pathology images alongside clinical information. This technology was honed using tumor specimens from 2,271 patients participating in the NSABP B-42 clinical trial and validated against a further 4,300 specimens from the esteemed TAILORx study. Notably, the research highlighted a staggering 10-year absolute distant recurrence risk discrepancy of almost 8% when comparing high- and low-risk patient groups.

Implications for Patients

These findings elucidate the model’s reliability and underscore its capability to inform extended endocrine therapy discussions. For patients identified as high-risk, there’s potential for considerable advantages from prolonged therapy, which further supports the model's practical application in clinical settings.

Despite the availability of genomic assays that provide important prognostic insights, their high costs, limited accessibility, and long turnaround times can hinder their utility in real-world settings. In contrast, the AI-based model by Caris utilizes commonly available diagnostic data, which may enhance oncologists' ability to discern patients at heightened risk and tailor treatment plans accordingly.

Product Development: Caris MI Clarity™

Following this breakthrough, Caris introduced Caris MI Clarity™, a pioneering prognostic test that delivers insights on both early and late distant recurrence risks for postmenopausal patients with HR+/HER2-negative, node-negative early-stage breast cancer. This innovative product goes beyond merely providing prognosis; it also offers decision support for chemotherapy, identifying which patients may benefit from chemotherapy. Moreover, it facilitates discussions around extended therapy beyond the initial five years, with options for reassessing recurrence risk during treatment.

Conclusion

Caris Life Sciences is at the forefront of harnessing artificial intelligence to enhance prognostic models in breast cancer treatment. By integrating advanced technologies with widely used diagnostic practices, they are paving the way for more personalized treatment options and better patient outcomes. As they continue to refine their models and explore AI's capabilities within oncology, the potential for improved patient care becomes increasingly tangible. This study marks a significant milestone in utilizing technology to revolutionize breast cancer management, emphasizing the importance of ongoing innovation in healthcare.