CT-Based Peritumoral and Intratumoral Radiomics as Pretreatment Predictors of Atypical Responses to Immune Checkpoint Inhibitor Across Tumor Types: A Preliminary Multicenter Study

Abstract

ObjectiveTo develop and validate a new strategy based on radiomics features extracted from intra- and peritumoral regions on CT images for the prediction of atypical responses to the immune checkpoint inhibitor (ICI) in cancer patients.MethodsIn total, 135 patients derived from five hospitals with pathologically confirmed malignancies receiving ICI were included in this retrospective study. Atypical responses including pseudoprogression (PsP) and hyperprogression disease (HPD) were identified as their definitions. A subgroup of standard progression disease (sPD) in 2018 was also involved in this study. Based on pretreatment CT imaging, a total of 107 features were extracted from intra- and peri-tumoral regions, respectively. The least absolute shrinkage and selection operator (Lasso) algorithm was used for feature selection, and multivariate logistic analysis was used to develop radiomics signature (RS). Finally, a total of nine RSs, derived from intra-tumoral, peri-tumoral, and combination of both regions, were built respectively to distinguish PsP vs. HPD, PsP vs. sPD, and HPD vs. sPD. The performance of the RSs was evaluated with discrimination, calibration, and clinical usefulness.ResultsNo significant difference was found when compared in terms of clinical characteristics of PsP, HPD, and sPD. RS based on combined regions outperformed those from either intra-tumoral or peri-tumoral alone, yielding an AUC (accuracy) of 0.834 (0.827) for PsP vs. HPD, 0.923 (0.868) for PsP vs. sPD, and 0.959 (0.894) for HPD vs. sPD in the training datasets, and 0.835 (0.794) for PsP vs. HPD, 0.919 (0.867) for PsP vs. sPD, and 0.933 (0.842) for HPD vs. sPD in the testing datasets. The combined RS showed good fitness (Hosmer–Lemeshow test p > 0.05) and provided more net benefit than the treat-none or treat-all scheme by decision curve analysis in both training and testing datasets.ConclusionPretreatment radiomics are helpful to predict atypical responses to ICI across tumor types. The combined RS outperformed those from either intra- or peri-tumoral alone which may provide a more comprehensive characterization of atypical responses to ICI.

Multi-Omics Approaches for the Prediction of Clinical Endpoints after Immunotherapy in Non-Small Cell Lung Cancer: A Comprehensive Review

“…Directly targeting the endpoint has the advantage of fully appreciating the complexity of the response to immunotherapy. Using either baseline imagings [ 94 ] or longitudinal imagings with delta-radiomics analyses [ 95 , 96 ], several models were proposed with very high performances [ 97 ]. With the advantage of prospectively included, but retrospectively analyzed patients treated with nivolumab, a radiomics signature achieved an AUC of 0.77 in the validation dataset [ 98 ].…”

Novel Imaging Biomarkers to Assess Oncologic Treatment–Related Changes

“…In the context of assessing treatment-related changes, radiomics has been applied for several cancer types including brain, 40,[58][59][60][61] prostate, 42,62,63 lung, [64][65][66][67][68][69] esophageal, 70,71 breast, 42,72 and head and neck. 42,[73][74][75][76] In this section, we first discuss the application of radiomics to identifying and distinguishing typical and atypical responses to treatment, including RN, PsP, and HPD (Fig.…”

“…Antunes et al 96 demonstrated the role of a radiomic descriptor quantifying the spatial relationship of the textural pattern gradient in differentiating PsP from glioblastoma multiforme recurrence after chemoradiation. He et al 64 studied the role of baseline CT radiomics across patients with multiple tumor types treated with immune checkpoint inhibitors to predict the risk of PsP and to differentiate it from true progression. They analyzed textural features extracted from both intratumoral and peritumoral regions and found that a combination of features from both regions (AUC, 0.919) could predict the risk of PsP and differentiate it from true progression more accurately than peritumoral (AUC, 0.900) or intratumoral (AUC, 0.891) features alone.…”

“…Over a set of two experiments that included 228 patients from prospective clinical trials, Tunali et al 66 explored the utility of radiomics in combination with clinical data to predict rapid disease progression and HPD. He et al 64 postulated that textural features extracted from within the tumor and the surrounding region could predict the development of HPD. Their machine-learning model combining the two features could predict the risk of HPD and differentiate HPD from standard progression (AUC, 0.933).…”

PET/CT radiomics for prediction of hyperprogression in metastatic melanoma patients treated with immune checkpoint inhibitors

“…As of today, there are no reliable, clinically validated biomarkers of PD-HPD. In recent years, a few studies were published investigating suitability of image-based biomarkers for PD-HPD prediction (24)(25)(26). Most of these studies, however, considered only CT-based radiomic signatures.…”