As of 2022, there were about 2.3 million women around the world newly diagnosed with breast cancer, and about 670,000 deaths caused by the disease.
There is currently no cure for breast cancer. Depending on the type of breast cancer, treatment options may include chemotherapy, radiation, hormone therapy, specially-targeted medications, and surgery.
Additionally, different types and stages of breast cancer have different rates for recurrence and metastasis, where the cancer spreads to another part of the body.
“Breast cancer is a complex disease with multiple treatment options, including chemotherapy and different types and durations of hormone therapy,” Calvin Chao, MD, vice president of medical science at digital health company Artera, told Medical News Today.
“Patients and clinicians need to understand their risks for recurrence and decide which treatments will be the most effective, thereby avoiding both undertreatment and overtreatment.”
Artera recently announced the U.S. Food and Drug Administration (FDA) clearance of its ArteraAI Breast digital pathology-based risk stratification tool for patients with early-stage, hormone receptor-positive (HR+), HER2-negative invasive breast cancer.
Read more:Perimenopause tips from a gynecologist: Best diet, other advice
ArteraAI Breast aims to help doctors predict the likelihood of metastasis in patients with early-stage HR+/HER2- breast cancer, to help determine the intensity of their treatment.
“By scanning a patient’s pathology slides of the surgical resection tissue, ArteraAI Breast inputs the digitized images and some clinical variables into an AI model,” Chao explained.
“The multimodal AI (MMAI) model was trained on data from over 8,500 breast cancer patients from clinical trials to predict the risk of metastasis, or cancer recurrence.”
“As demonstrated in the FDA clinical validation, this AI model can accurately predict the risk of metastasis, and sort patients into MMAI Low or MMAI High risk groups based on their AI risk score,” he added.
ArteraAI Breast assists oncologists when making personalized treatment decisions for breast cancer patients.
“Oncologists routinely need to determine the optimal level of therapy intensification for their patients,” Chao said.
Read more:Colorectal cancer: Novel approach could help prevent relapses
“Patients with a lower risk profile have a much lower chance of cancer recurrence; accordingly, they do not need the same treatment intensity as patients with a higher risk profile.”
“The ArteraAI Breast results, which include the personalized AI risk score and the associated MMAI risk groups can help oncologists and patients make the right treatment decisions.”
MNT had the opportunity to speak with Richard Reitherman,MD, PhD, a board certified radiologist and medical director of breast imaging at MemorialCare Breast Center at Orange Coast Medical Center in Fountain Valley, CA, about this study.
Reitherman said that as of the early 2000s, doctors have had access to a test called Oncotype DX to help them determine a person’s risk of systemic metastasis, and whether or not they would need to add chemotherapy to their endocrine therapy.
“The Oncotype DX uses a patented methodology to analyze the histopathologic (the features of the tumor as represented on the slides) and provide what is called a recurrent score that separates women into a low, moderate, and high risk of systemic disease in the future and, therefore, consideration of adding chemotherapy to reduce this risk,” he detailed.
Read more:FDA moves to remove GLP-1 medications from 503B Bulks List
“This test can be very important, but is not always available, may take several weeks to process, and is relatively costly. Insurance usually covers the testing after surgery, but not before. Some clinical situations may consider chemotherapy prior to surgery, but lack the Oncotype recurrence score.”
“The potential breakthrough in the multi–modal artificial intelligence (MMAI) model for predicting distant metastasis in hormone positive (HR+) early-stage breast cancer is that it uses the immediately available clinical and existing histopathologic features to assigned patients into low and high risk metastasis groups without the costs and time delays associated with currently available methodology,” Reitherman added.
MNT also spoke with Donna McNamara, MD, a breast medical oncologist at the John Theurer Cancer Center at Hackensack University Medical Center in New Jersey, about this study, who commented she thought that this is an incredible milestone using AI and digital pathology marking a substantial step forward in personalizing breast cancer therapy, particularly for patients with HR+ early-stage disease.
“The ability to better stratify patients who will benefit from chemotherapy versus those who can safely avoid it is critically important,” McNamara explained.
“Many patients, especially post-menopausal women with node-negative tumors, are in a ‘gray area’ where the decision to recommend chemotherapy is not always clear-cut. This technology could provide much-needed clarity in those situations.”
“The potential to spare low-risk patients from the significant toxicities of chemotherapy is a major advantage,” she continued.
Read more:Top Health News: Technology, Mental Health and More
“Chemotherapy can have debilitating short-term and long-term side effects, including neuropathy, an increased risk of infection, and impacts on fertility. If we can safely identify patients who will not derive a significant benefit from this treatment, we can save them from unnecessary physical, emotional, and financial burdens.”
To start using ArteraAI Breast with her patients, McNamara said she would require a wealth of robust, peer-reviewed evidence, starting with data from prospective, head-to-head clinical trials that directly compare its performance against current gold standards like Oncotype DX.
“This evidence must include long-term follow-up demonstrating that using ArteraAI to guide therapy leads to at least equivalent, if not superior, patient outcomes in terms of disease-free and overall survival across diverse populations,” she detailed.
“Beyond this critical clinical validation, I would need clear answers on its practical implementation, including real-world turnaround times, cost, and insurance coverage, and how it integrates into our existing workflow.”
“Furthermore, I would need a degree of transparency into the AI model itself, moving beyond the ‘black box’ to understand the key features driving its risk assessments, as I cannot base a patient’s treatment plan on a tool without thoroughly understanding its accuracy, reliability, and practical utility,” McNamara added.
