Early Prediction and Monitoring of Treatment Response in Gastrointestinal Stromal Tumors by Means of Imaging: A Systematic Review

Ylva A Weeda; Gijsbert M Kalisvaart; Floris H P van Velden; Hans Gelderblom; Aart J van der Molen; Judith V M G Bovee; Jos A van der Hage; Willem Grootjans; Lioe-Fee de Geus-Oei

doi:10.3390/diagnostics12112722

Early Prediction and Monitoring of Treatment Response in Gastrointestinal Stromal Tumors by Means of Imaging: A Systematic Review

Diagnostics (Basel). 2022 Nov 7;12(11):2722. doi: 10.3390/diagnostics12112722.

Affiliations

¹ Department of Radiology, Leiden University Medical Center, 2333 ZA Leiden, The Netherlands.
² Department of Medical Oncology, Leiden University Medical Center, 2333 ZA Leiden, The Netherlands.
³ Department of Radiology, Section of Abdominal Radiology, Leiden University Medical Center, 2333 ZA Leiden, The Netherlands.
⁴ Department of Pathology, Leiden University Medical Center, 2333 ZA Leiden, The Netherlands.
⁵ Department of Surgical Oncology, Leiden University Medical Center, 2333 ZA Leiden, The Netherlands.
⁶ Biomedical Photonic Imaging Group, University of Twente, 7522 NB Enschede, The Netherlands.
⁷ Department of Radiation Science & Technology, Technical University of Delft, 2629 JB Delft, The Netherlands.

Abstract

Gastrointestinal stromal tumors (GISTs) are rare mesenchymal neoplasms. Tyrosine kinase inhibitor (TKI) therapy is currently part of routine clinical practice for unresectable and metastatic disease. It is important to assess the efficacy of TKI treatment at an early stage to optimize therapy strategies and eliminate futile ineffective treatment, side effects and unnecessary costs. This systematic review provides an overview of the imaging features obtained from contrast-enhanced (CE)-CT and 2-deoxy-2-[¹⁸F]fluoro-D-glucose ([¹⁸F]FDG) PET/CT to predict and monitor TKI treatment response in GIST patients. PubMed, Web of Science, the Cochrane Library and Embase were systematically screened. Articles were considered eligible if quantitative outcome measures (area under the curve (AUC), correlations, sensitivity, specificity, accuracy) were used to evaluate the efficacy of imaging features for predicting and monitoring treatment response to various TKI treatments. The methodological quality of all articles was assessed using the Quality Assessment of Diagnostic Accuracy Studies, v2 (QUADAS-2) tool and modified versions of the Radiomics Quality Score (RQS). A total of 90 articles were included, of which 66 articles used baseline [¹⁸F]FDG-PET and CE-CT imaging features for response prediction. Generally, the presence of heterogeneous enhancement on baseline CE-CT imaging was considered predictive for high-risk GISTs, related to underlying neovascularization and necrosis of the tumor. The remaining articles discussed therapy monitoring. Clinically established imaging features, including changes in tumor size and density, were considered unfavorable monitoring criteria, leading to under- and overestimation of response. Furthermore, changes in glucose metabolism, as reflected by [¹⁸F]FDG-PET imaging features, preceded changes in tumor size and were more strongly correlated with tumor response. Although CE-CT and [¹⁸F]FDG-PET can aid in the prediction and monitoring in GIST patients, further research on cost-effectiveness is recommended.

Keywords: FDG-PET; X-ray computed; gastrointestinal stromal tumor; personalized medicine; prediction; radiomics; response monitoring; tomography.

Publication types

Review

Grants and funding

We declare the following financial interests/personal relationships which may be considered as potential competing interests: G.M. Kalisvaart was the recipient of an educational grant (LEI-05) from Philips Electronics Nederland B.V, Eindhoven and supported by a public grant (LSHM18089) from Health Holland TKI Life Sciences & Health. The funders had no role in the design of the study, in the collection, analyses, or interpretation of data, in the writing of the manuscript or in the decision to publish the results.