An innovative model predicts the effectiveness of immunotherapy in lung cancer

Lung cancer remains the leading cause of cancer-related deaths worldwide. Among available treatments, immunotherapy has shown significant benefits for patients with non-small cell lung cancer, a common and aggressive form of the disease. However, not all patients respond equally to this type of treatment, and identifying those who will benefit the most is a major challenge.

A recent breakthrough combines the analysis of medical images and the study of tumor tissues to accurately predict the expression of a key protein, PD-L1, which plays a central role in the response to immunotherapy. This protein, present on the surface of cancer cells, allows the immune system to recognize and attack the tumor. Its expression level is often used to determine if a patient is eligible for immunotherapy.

Until now, the detection of PD-L1 mainly relied on laboratory tests requiring tissue samples, a method sometimes limited by the variability of results and the availability of samples. To overcome these obstacles, a team of researchers has developed an innovative model that merges data from scans and microscopic analyses of stained tissues. This model uses advanced techniques to extract both macroscopic information, such as the shape and texture of the tumor visible in images, and microscopic information, such as cellular characteristics observed under the microscope.

The study involved 328 patients with non-small cell lung cancer. The researchers divided these patients into three groups to train, validate, and test the model. The results showed that this combined approach was more effective than models based solely on imaging or tissue analysis. It accurately predicted whether a tumor expressed the PD-L1 protein, with success rates exceeding 90% in some cases.

To go further, the model was also tested on a group of 47 patients undergoing immunotherapy. The results revealed that patients for whom the model predicted high PD-L1 expression had significantly better progression-free survival and overall survival. This suggests that the model could not only help identify patients eligible for immunotherapy but also predict their response to treatment.

One of the major advantages of this method is that it uses data already available as part of standard diagnosis, such as scans and histological slides. It therefore does not require additional samples, reducing costs and the burden on patients. Additionally, it could prove useful in situations where tissue samples are insufficient or when traditional test results are uncertain.

This model thus paves the way for more precise medicine, where treatment decisions could be tailored based on the specific characteristics of each tumor. It represents an important step toward personalized management of lung cancer, providing doctors with an additional tool to optimize therapeutic strategies and improve patient outcomes.

Sources and Credits

Source Study

DOI: https://doi.org/10.1186/s13244-026-02322-4

Title: A radio-pathological fusion model for predicting PD-L1 expression and immunotherapy response in non-small cell lung cancer

Journal: Insights into Imaging

Publisher: Springer Science and Business Media LLC

Authors: Dingpin Huang; Fangyi Xu; Yi Gan; Liya Ding; Kaihua Lou; Yongcen Li; Dong Xie; Haiping Zhang; Lei Shi; Rui Xu; Hongjie Hu