“We were we able to identify which patients would respond to CDK4/6 inhibitors. In addition, the study revealed that specific radiomic features were strongly linked to clinically relevant outcomes in these patients,” says study author Anant Madabhushi, PhD.
Approximately 50% of people diagnosed with metastatic breast cancer develop liver metastasis. The combination of endocrine therapy (ET) with cyclin-dependent kinase 4/6 inhibitors (CDK4/6) is the preferred first line treatment for metastatic breast cancer. However, CDK4/6 inhibitors, which target specific enzymes, are cost-prohibitive.
A team of researchers from Winship Cancer Institute of Emory University and Case Western Reserve University has developed a method using artificial intelligence and machine learning to analyze imaging scans and determine whether the treatment will be effective. They completed a study, with results recently published in the journal npj Breast Cancer, consisting of 73 patients whose breast cancer had spread to the liver and were treated with CDK4/6 inhibitors.
“We analyzed CT scans of the liver metastasis taken before and after the treatment to see if we could find patterns that could predict how well the treatment worked, how long the patients survived and their response to treatment,” says co-lead author Vidya Viswanathan, a first-year resident in the Department of Radiology and Imaging Sciences at Emory University School of Medicine.
The researchers examined seven radiomic features extracted from the medical images to analyze heterogeneity within the tumor and in the adjacent tumor habitat. Radiomic features consist of distinctive image and shape descriptors that can be derived from images with advanced analysis to help predict prognosis and therapeutic response. The researchers then applied a Radiomic Risk Score (RRS) derived from the features to predict early response and overall survival.
“We were we able to identify which patients would respond to CDK4/6 inhibitors. In addition, the study revealed that specific radiomic features were strongly linked to clinically relevant outcomes in these patients,” says study author Anant Madabhushi, PhD, the Robert W. Woodruff Professor of Biomedical Engineering at Emory University School of Medicine and Georgia Institute of Technology College of Engineering, a member of the Cancer Immunology research program at Winship Cancer Institute of Emory University, and research career scientist at the Atlanta Veterans Administration Medical Center.
In clinical practice, incorporating these radiomic features and the Radiomic Risk Score developed by the team offers the opportunity to enhance treatment decision-making.
“Currently, there are no robust biomarkers that can predict responses to CDK4/6, and it is not clear which patients benefit from this therapy,” says co-lead author Mohammadhadi Khorrami, a postdoctoral fellow in biomedical engineering at Emory University. “This study holds great promise in advancing our understanding of the treatment outcomes in patients with breast cancer that has metastasized to the liver.”
Information obtained from analyzing radiomic features and assigning an RRS helps physicians determine which patients are likely to benefit from this specific therapy. This enables a more targeted approach, minimizing unnecessary treatments – and the associated costs and potential side effects – for patients who may not respond well.
Radiomic features can also serve as non-invasive markers for monitoring treatment response over time. By tracking changes in the features during therapy, clinicians can evaluate the effectiveness of CDK4/6 inhibitors and make timely adjustments, if needed.
“This information enables a more proactive and dynamic approach to patient management, ensuring that treatment plans are modified to maximize therapeutic benefit,” says Viswanathan.
The next steps for advancing this research include validating the RRS in a prospective clinical trial.
Related stories:
- New AI-based biomarker can help predict immunotherapy response for patients with lung cancer
- Low-cost AI tools aim to predict Indian population cancer risk and therapeutic benefit
- AI tools aim to identify which prostate cancer patients may require adjuvant therapy
