Sensitivity and Specificity of Whole-body MRI for the Detection of Pediatric Malignancy.

Document Type

Article

Publication Date

1-1-2023

Identifier

DOI: 10.1097/MPH.0000000000002575

Abstract

Children with cancer often present with general and nonspecific symptoms leading to initial diagnostic workup inclusive of clinical imaging. Various sequences of magnetic resonance imaging (MRI) are becoming more available for diagnostic imaging. However, there is currently a dearth of literature quantifying the sensitivity and specificity of whole-body MRI in identifying pediatric malignancy. In this study, a retrospective analysis was performed of pediatric whole-body MRI inclusive of short tau inversion recovery sequence conducted at an academic pediatric medical center from 2013 to 2018. Kappa statistics were used to evaluate the diagnostic agreement between MRI results and the gold standard diagnostic study of the respective final diagnosis. Sensitivity, specificity, false-positive, and false-negative estimates were provided with joint 90% confidence regions. One hundred forty-two patients received a whole-body MRI during the study period. The sensitivity of whole-body MRI in detecting malignancy was found to be 93.8% with a specificity of 93.4%. The positive and negative predictive values were determined to be 65.2% and 99.1%, respectively. Our findings suggest that whole-body MRI may be of value as an initial diagnostic tool for pediatric malignancy. Larger multicenter collaboration will be needed to further support these data.

Journal Title

Journal of pediatric hematology/oncology : official journal of the American Society of Pediatric Hematology/Oncology

Volume

45

Issue

1

First Page

26

Last Page

30

MeSH Keywords

Humans; Child; Retrospective Studies; Magnetic Resonance Imaging; Sensitivity and Specificity; Neoplasms; Predictive Value of Tests; Whole Body Imaging

Keywords

Retrospective Studies; Magnetic Resonance Imaging; Sensitivity and Specificity; Neoplasms; Predictive Value of Tests; Whole Body Imaging

Library Record

Share

COinS