A new study is exploring a new method for classifying mesothelioma subtypes. It offers hope for more accurate diagnoses and tailored treatments.
Improving Mesothelioma Prognosis
Mesothelioma is classified into three subtypes based on how the cancer cells look. The three types are: epithelioid, biphasic, and sarcomatoid. Each subtype has unique characteristics and can mean different things to a patient’s outlook. Accurately identifying the subtype is crucial for determining treatment options for patients.
Currently, doctors will identify a patient’s subtype just by looking at their cancer cells under a microscope. There is no standardized process, which leads to inconsistent results. Additionally, epithelioid and sarcomatoid cells can look very similar. This can make it even harder for doctors to tell what kind of mesothelioma their patient has.
The Role of Graph Neural Networks in Subtype Classification
To address these challenges, researchers developed a new approach. They used a Graph Neural Network (GNN) architecture. This advanced model uses machine learning to identify features to analyze cancer tissue samples. The GNN approach helps doctors to see cancer cells more clearly, leading to more accurate diagnoses.
Accurate subtype classification has a big impact on treatment decisions. For example, surgical treatment is more beneficial for epithelioid mesothelioma. Sarcomatoid and biphasic subtypes don’t often improve with surgery. This breakthrough offers hope for improved patient outcomes and a more personalized approach to treatment. With more research, it could lead to improved survival rates and overall quality of life for those affected by mesothelioma.
Eastwood M, Sailem H, Marc ST, et al. MesoGraph: Automatic profiling of mesothelioma subtypes from histological images [published online ahead of print, 2023 Oct 5]. Cell Rep Med. 2023;101226. doi:10.1016/j.xcrm.2023.101226. https://www.cell.com/cell-reports-medicine/fulltext/S2666-3791(23)00403-2