A Radiomics-Clinical Model Predicts Overall Survival of Non-Small Cell Lung Cancer Patients Treated with Immunotherapy: A Multicenter Study

Abstract

Background: Immune checkpoint inhibitors (ICIs) are a great breakthrough in cancer treatments and provide improved long-term survival in a subset of non-small cell lung cancer (NSCLC) patients. However, prognostic and predictive biomarkers of immunotherapy still remain an unmet clinical need. In this work, we aim to leverage imaging data and clinical variables to develop survival risk models among advanced NSCLC patients treated with immunotherapy. Methods: This retrospective study includes a total of 385 patients from two institutions who were treated with ICIs. Radiomics features extracted from pretreatment CT scans were used to build predictive models. The objectives were to predict overall survival (OS) along with building a classifier for short- and long-term survival groups. We employed the XGBoost learning method to build radiomics and integrated clinical-radiomics predictive models. Feature selection and model building were developed and validated on a multicenter cohort. Results: We developed parsimonious models that were associated with OS and a classifier for short- and long-term survivor groups. The concordance indices (C-index) of the radiomics model were 0.61 and 0.57 to predict OS in the discovery and validation cohorts, respectively. While the area under the curve (AUC) values of the radiomic models for short- and long-term groups were found to be 0.65 and 0.58 in the discovery and validation cohorts. The accuracy of the combined radiomics-clinical model resulted in 0.63 and 0.62 to predict OS and in 0.77 and 0.62 to classify the survival groups in the discovery and validation cohorts, respectively. Conclusions: We developed and validated novel radiomics and integrated radiomics-clinical survival models among NSCLC patients treated with ICIs. This model has important translational implications, which can be used to identify a subset of patients who are not likely to benefit from immunotherapy. The developed imaging biomarkers may allow early prediction of low-group survivors, though additional validation of these radiomics models is warranted.

Artificial Intelligence and Machine Learning in Predicting the Response to Immunotherapy in Non-small Cell Lung Carcinoma: A Systematic Review

“…The studies utilized diverse AI/ML techniques, including deep learning (DL), artificial neural networks (ANNs), support vector machines (SVMs), random forests (RFs), and gradient boosting methods (e.g., XGBoost). These algorithms were applied to various data modalities, such as medical imaging (computed tomography (CT), positron emission tomography (PET)), genomic data (TMB, gene expression), clinical variables (performance status, blood counts), and immunohistochemical markers (PD-L1, TILs) [8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26]. The studies employed various performance metrics to evaluate the predictive accuracy of their AI/ML models.…”

“…. Similarly,Yolchuyeva et al (2023Yolchuyeva et al ( , 2024 developed ML models incorporating cytokine profiles and clinical variables that effectively predicted OS and PFS in patients receiving anti-PD-1/PD-L1 immunotherapy[9,10].Vanguri et al (2022) andHe et al (2022) utilized multimodal data, including CT images, genomic features, and clinical variables, to build ML models that outperformed single biomarkers like TMB in predicting immunotherapy response[12,13]. employed eXplainable AI (XAI) and ML techniques to develop a model that accurately predicted disease control rate, OS, and time to treatment failure, while also identifying influential features such as neutrophil-to-lymphocyte ratio and PD-L1 expression[14,17].…”

Imaging-Based Biomarkers Predict Programmed Death-Ligand 1 and Survival Outcomes in Advanced NSCLC Treated With Nivolumab and Pembrolizumab: A Multi-Institutional Study

“…The recent advent of radiomics through quantitative image analysis has been gaining interest in oncology as a novel strategy for predicting treatment response. 33 , 34 , 35 , 36 Nevertheless, the development of imaging-based signatures that can be robust and generalizable across academic centers has been a bottleneck to adopt radiomics in clinical practice.…”

The Cross-Scale Association between Pathomics and Radiomics Features in Immunotherapy-Treated NSCLC Patients: A Preliminary Study

“…Radiomics can be quantified using well-defined mathematical equations to obtain the underlying characteristics of the disease through image intensity, shape, or texture, thereby overcoming the subjective nature of image interpretation [10]. Many studies, including our recent works, have investigated and reported on the added clinical value of radiomics features to predict various clinical and biological outcomes in immunotherapy-treated patients, such as overall survival, progression-free survival, tumor histology, and genetic profiling, among other endpoints [9,[11][12][13]. On the other hand, the rapidly expanding field of pathomics aims to investigate the micro-scale patterns from the digitized histopathology slides, or whole slide imaging (WSI), in a high-throughput manner [4,14].…”