Back to Search Start Over

Development and validation of a [18F]FDG PET/CT-based radiomics nomogram to predict the prognostic risk of pretreatment diffuse large B cell lymphoma patients.

Authors :
Li, Mingshan
Yao, Hongyang
Zhang, Peng
Zhang, Lingbo
Liu, Wei
Jiang, Zhiyun
Li, Wei
Zhao, Shu
Wang, Kezheng
Source :
European Radiology; May2023, Vol. 33 Issue 5, p3354-3365, 12p, 1 Diagram, 3 Charts, 5 Graphs
Publication Year :
2023

Abstract

Objective: In this study, based on PET/CT radiomics features, we developed and validated a nomogram to predict progression-free survival (PFS) for cases with diffuse large B cell lymphoma (DLBCL) treated with immunochemotherapy. Methods: This study retrospectively recruited 129 cases with DLBCL. Among them, PET/CT scans were conducted and baseline images were collected for radiomics features along with their clinicopathological features. Radiomics features related to recurrence were screened for survival analysis using univariate Cox regression analysis with p < 0.05. Next, a weighted Radiomics-score (Rad-score) was generated and independent risk factors were obtained from univariate and multivariate Cox regressions to build the nomogram. Furthermore, the nomogram was tested for their ability to predict PFS using time-dependent receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA). Results: Blood platelet, Rad-score, and gender were included in the nomogram as independent DLBCL risk factors for PFS. We found that the training cohort areas under the curve (AUCs) were 0.79, 0.84, and 0.88, and validation cohort AUCs were 0.67, 0.83, and 0.72, respectively. Further, the DCA and calibration curves confirmed the predictive nomogram's clinical relevance. Conclusion: Using Rad-score, blood platelet, and gender of the DLBCL patients, a PET/CT radiomics-based nomogram was developed to guide cases' recurrence risk assessment prior to treatment. The developed nomogram can help provide more appropriate treatment plans to the cases. Key Points: • DLBCL cases can be classified into low- and high-risk groups using PET/CT radiomics based Rad-score. • When combined with other clinical characteristics (gender and blood platelet count), Rad-score can be used to predict the outcome of the pretreatment of DLBCL cases with a certain degree of accuracy. • A prognostic nomogram was established in this study in order to aid in assessing prognostic risk and providing more accurate treatment plans for DLBCL cases. [ABSTRACT FROM AUTHOR]

Details

Language :
English
ISSN :
09387994
Volume :
33
Issue :
5
Database :
Complementary Index
Journal :
European Radiology
Publication Type :
Academic Journal
Accession number :
163255680
Full Text :
https://doi.org/10.1007/s00330-022-09301-5