An Optimal Radiomics Nomogram Based on 18F-FDG PET/CT for Identifying Event-Free Survival in Pediatric Neuroblastoma

Lijuan Feng; Shuxin Zhang; Xia Lu; Xu Yang; Ying Kan; Chao Wang; Hui Zhang; Wei Wang; Jigang Yang

doi:10.1016/j.acra.2023.06.004

An Optimal Radiomics Nomogram Based on ¹⁸F-FDG PET/CT for Identifying Event-Free Survival in Pediatric Neuroblastoma

Acad Radiol. 2023 Oct;30(10):2309-2320. doi: 10.1016/j.acra.2023.06.004. Epub 2023 Jun 29.

Authors

Lijuan Feng¹, Shuxin Zhang¹, Xia Lu¹, Xu Yang¹, Ying Kan¹, Chao Wang², Hui Zhang³, Wei Wang¹, Jigang Yang⁴

Affiliations

¹ Department of Nuclear Medicine, Beijing Friendship Hospital, Capital Medical University, 95 Yong An Road, Xi Cheng District, Beijing 100050, China (L.F., S.Z., X.L., X.Y., Y.K., W.W., J.Y.).
² SinoUnion Healthcare Inc., Beijing, China (C.W.).
³ Department of Biomedical Engineering, School of Medicine, Tsinghua University, Beijing, China (H,Z,).
⁴ Department of Nuclear Medicine, Beijing Friendship Hospital, Capital Medical University, 95 Yong An Road, Xi Cheng District, Beijing 100050, China (L.F., S.Z., X.L., X.Y., Y.K., W.W., J.Y.). Electronic address: yangjigang@ccmu.edu.cn.

PMID: 37393177
DOI: 10.1016/j.acra.2023.06.004

Abstract

Rationale and objectives: To investigate whether the ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) positron emission tomography/computed tomography (PET/CT) radiomics features that combine tumor and bone marrow can more accurately identify event-free survival (EFS) in pediatric neuroblastoma.

Materials and methods: A total of 126 patients with neuroblastoma were retrospectively included and randomly divided into the training and validation cohorts (7:3 ratio). Radiomics features were extracted to develop a tumor- and bone marrow-based radiomics risk score (RRS). The Kaplan-Meier method was used to evaluate the effectiveness of RRS in EFS risk stratification. Univariate and multivariate Cox regression analyses were used to determine independent clinical risk factors and construct the clinical models. The conventional PET model was constructed based on conventional PET parameters, and the noninvasive combined model integrated the RRS and the noninvasive independent clinical risk factors. The performance of the models was evaluated using C-index, calibration curves, and decision curve analysis (DCA).

Results: A total of 15 radiomics features were selected to build the RRS. According to Kaplan-Meier analysis, there was a significant difference in EFS between the low-risk and high-risk groups as defined by the value of RRS (P < .05). The noninvasive combined model combining RRS and the International Neuroblastoma Risk Group stage achieved the best prognostic prediction of EFS, with a C-index of 0.810 and 0.783 in the training and validation cohorts, respectively. The calibration curves and DCA indicated that the noninvasive combined model had good consistency and clinical utility.

Conclusion: The ¹⁸F-FDG PET/CT-based radiomics of neuroblastoma allows a reliable evaluation of EFS. The performance of the noninvasive combined model was superior to the clinical and conventional PET models.

Keywords: Event-free survival; Neuroblastoma; Positron emission tomography/computed tomography; Radiomics.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Child
Fluorodeoxyglucose F18
Humans
Neuroblastoma* / diagnostic imaging
Nomograms
Positron Emission Tomography Computed Tomography* / methods
Progression-Free Survival
Retrospective Studies

Substances

Fluorodeoxyglucose F18