@article {Colene001752, author = {Rivka R Colen and Christian Rolfo and Murat Ak and Mira Ayoub and Sara Ahmed and Nabil Elshafeey and Priyadarshini Mamindla and Pascal O Zinn and Chaan Ng and Raghu Vikram and Spyridon Bakas and Christine B Peterson and Jordi Rodon Ahnert and Vivek Subbiah and Daniel D Karp and Bettzy Stephen and Joud Hajjar and Aung Naing}, title = {Radiomics analysis for predicting pembrolizumab response in patients with advanced rare cancers}, volume = {9}, number = {4}, elocation-id = {e001752}, year = {2021}, doi = {10.1136/jitc-2020-001752}, publisher = {BMJ Specialist Journals}, abstract = {Background We present a radiomics-based model for predicting response to pembrolizumab in patients with advanced rare cancers.Methods The study included 57 patients with advanced rare cancers who were enrolled in our phase II clinical trial of pembrolizumab. Tumor response was evaluated using Response Evaluation Criteria in Solid Tumors (RECIST) 1.1 and immune-related RECIST (irRECIST). Patients were categorized as 20 {\textquotedblleft}controlled disease{\textquotedblright} (stable disease, partial response, or complete response) or 37 progressive disease). We used 3D-slicer to segment target lesions on standard-of-care, pretreatment contrast enhanced CT scans. We extracted 610 features (10 histogram-based features and 600 second-order texture features) from each volume of interest. Least absolute shrinkage and selection operator logistic regression was used to detect the most discriminatory features. Selected features were used to create a classification model, using XGBoost, for the prediction of tumor response to pembrolizumab. Leave-one-out cross-validation was performed to assess model performance.Findings The 10 most relevant radiomics features were selected; XGBoost-based classification successfully differentiated between controlled disease (complete response, partial response, stable disease) and progressive disease with high accuracy, sensitivity, and specificity in patients assessed by RECIST (94.7\%, 97.3\%, and 90\%, respectively; p\<0.001) and in patients assessed by irRECIST (94.7\%, 93.9\%, and 95.8\%, respectively; p\<0.001). Additionally, the common features of the RECIST and irRECIST groups also highly predicted pembrolizumab response with accuracy, sensitivity, specificity, and p value of 94.7\%, 97\%, 90\%, p\<0.001\% and 96\%, 96\%, 95\%, p\<0.001, respectively.Conclusion Our radiomics-based signature identified imaging differences that predicted pembrolizumab response in patients with advanced rare cancer.Interpretation Our radiomics-based signature identified imaging differences that predicted pembrolizumab response in patients with advanced rare cancer.Data are available upon reasonable request. The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request and approval from study sponsor according to available guidelines at time of request.}, URL = {https://jitc.bmj.com/content/9/4/e001752}, eprint = {https://jitc.bmj.com/content/9/4/e001752.full.pdf}, journal = {Journal for ImmunoTherapy of Cancer} }