Abstract:BACKGROUND:Systemic therapy agents targeting immune checkpoint inhibitors have been approved for use since 2011. This type of therapy aims to trigger a patient's immune response to attack tumor cells, rather than acting against the tumor directly. Radiomics is an automated method of medical image analysis that is now being actively investigated for predictive markers of treatment response in immunotherapy.OBJECTIVE:To conduct an early systematic review determining the current status of radiomic features as potential predictive markers of immunotherapy response. Provide a detailed critical appraisal of methodological quality of models, as this informs the degree of confidence about current reports of model performance. In addition, to offer some recommendations for future studies that could establish robust evidence for radiomic features as immunotherapy response markers.METHOD:A PubMed citation search was conducted for publications up to and including April 2020, followed by full-text screening. A total of seven articles meeting the eligibility criteria were examined in detail for study characteristics, model information and methodological quality. The review was conducted in the Cochrane style but has not been prospectively registered. Results are reported following Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA) guidelines.RESULTS:A total of seven studies were examined in detail, comprising non-small cell lung cancer, metastatic melanoma and a diverse assortment of solid tumors. Methodological robustness of reviewed studies varied greatly. Principal shortcomings were lack of prospective registration, and deficiencies in feature selection and dimensionality reduction, model calibration, clinical utility and external validation. A few studies with overall moderate to good methodological quality were identified. These results suggest that current state-of-the-art performance of radiomics in regards to discrimination (area under the curve or concordance index) is in the vicinity of 0.7, but the very small number of studies to date prevents any conclusive remarks to be made. We recommended future improvements in regards to prospective study registration, clinical utility, methodological procedure and data sharing.CONCLUSIONS:Radiomics has a potentially significant role for predicting immunotherapy response. Additional multi-institutional studies with robust methodological underpinning and repeated external validations are required to establish the (added) value of radiomics within the pantheon of clinical tools for decision-making in immunotherapy.

[Advances in Imaging-Based Evaluation of Solid Tumors Treated With Immune Checkpoint Inhibitors]

Positron Emission Tomography Molecular Imaging to Monitor Anti-Tumor Systemic Response for Immune Checkpoint Inhibitor Therapy

Role of medical imaging for immune checkpoint blockade therapy: From response assessment to prognosis prediction

Advances in molecular imaging of immune checkpoint targets in malignancies: current and future prospect

Molecular Imaging for Cancer Immunotherapy: Seeing is Believing

Advances in models predicting efficacy of immune checkpoint inhibitors

Application of Molecular Imaging in Immune Checkpoints Therapy: from Response Assessment to Prognosis Prediction.

Immunotherapy and the role of imaging

Challenges of Evaluating Immunotherapy Efficacy in Solid Tumors

Progress of Molecular Imaging-guided Tumor Immunotherapy

Outcome Prediction and Evaluation by Imaging the Key Elements of Therapeutic Responses to Cancer Immunotherapies Using PET

[Advancements in Radiomics for Immunotherapy of Non-small Cell Lung Cancer]

Imaging evaluation of therapeutic response in patients with pancreatic cancer: recent advances

Fluorescence Imaging of Tumor Immune Contexture in Immune Checkpoint Blockade Therapy.

Monitoring Immunotherapy With Optical Molecular Imaging

Systematic Review of Radiomic Biomarkers for Predicting Immune Checkpoint Inhibitor Treatment Outcomes.

Emerging Biomaterials Imaging Antitumor Immune Response

Advances in artificial intelligence to predict cancer immunotherapy efficacy

Applications of CT-based radiomics for the prediction of immune checkpoint markers and immunotherapeutic outcomes in non-small cell lung cancer

Predictive Biomarkers for Cancer Immunotherapy with Immune Checkpoint Inhibitors

Immune Checkpoint Inhibitors Combined with Targeted Therapy: the Recent Advances and Future Potentials