In recent advancements in cancer treatment, a new pathology tool utilizing artificial intelligence is redefining how patients with extensive-stage small cell lung cancer (SCLC) are assessed before undergoing chemotherapy. The implications of this innovation are significant, as it allows for more informed and personalized treatment plans for patients.
Highlights
- Guides decision-making to minimize unnecessary treatments
- Predicts overall survival rates
- Requires no additional biopsies or procedures
BUFFALO, N.Y. — Findings from a new study indicate that an AI-powered pathology tool can accurately predict whether patients with extensive-stage small cell lung cancer will respond to platinum-based chemotherapy before treatment begins, and without the need for additional biopsies. This innovation allows patients to avoid ineffective treatments, enroll earlier in clinical trials for emerging therapies, and gain a clearer understanding of their prognosis.
The novel tool, named PhenopyCell, has undergone verification by three collaborating institutions and is detailed in the journal npc Precision Oncology. It was developed by a research team co-led by thoracic oncologist Prantesh Jain, MD, FACP, from Roswell Park Comprehensive Center, alongside Anant Madabhushi, PhD, from Winship Cancer Institute at Emory University in Atlanta.
This study offers new hope for the 70% of small cell lung cancer patients diagnosed with extensive-stage disease, where the cancer has already metastasized. Diagnosis at this stage often results in a bleak survival prognosis of just 12 to 13 months, making the quick identification of effective treatments crucial.
Currently, SCLC has various subtypes, but distinguishing among them is challenging, leading to a uniform treatment approach involving platinum-based chemotherapy paired with immunotherapy. Unfortunately, by the time the effectiveness of this treatment is evident, it can be too late to switch therapies. Though new treatments for SCLC have recently gained FDA approval or are showing potential in clinical trials, they yield positive results in only a limited number of patients. Unlike many other cancer types, SCLC lacks clearly defined biomarkers—biological indicators such as proteins or genetic mutations found in blood or tissue samples—to guide treatment decisions.
“As we move into a new era of treatment options for small cell lung cancer, selecting the right approach for each patient is vital,” asserts Dr. Jain. “However, the absence of biological markers makes this difficult. PhenopyCell addresses this gap by leveraging data from various sources, including pathology slides and patient medical records, to correlate this data with patient outcomes.”
In their retrospective analysis, the researchers utilized PhenopyCell to examine standard pathology slides from 281 SCLC patients treated at Roswell Park, Winship, and University Hospitals Cleveland Medical Center. The tool utilized information regarding immune cell presence in tissue samples from diagnostic biopsies to predict patient responses to chemotherapy before treatment commenced. When compared with actual patient outcomes, the tool demonstrated greater accuracy than traditional manual analysis.
PhenopyCell indicated that tumors of patients who achieved better outcomes had a higher concentration of immune cells arranged in organized clusters surrounding tumor masses, signifying a robust immune response. Conversely, tumors linked with poorer outcomes displayed fewer immune cells and a disorganized arrangement positioned farther from the tumor itself. These patterns were only discernible through the AI-based pathology tool.
“Every patient with small cell lung cancer already has a pathology slide from their diagnostic biopsy,” emphasizes Dr. Jain. “This system operates using that existing slide, eliminating the need for further procedures or tissue collection, which adds no extra costs. In a disease where survival duration is often measured in months, this tool holds great potential to become an invaluable resource.”
The research team included experts from University Hospitals Cleveland Medical Center/Case Western Reserve University, City of Hope, and Penn State Cancer Institute.
Original release: https://www.roswellpark.org//newsroom/202603-new-ai-tool-predicts-whether-aggressive-small-cell-lung-cancer-will-respond
