Role of proton MR spectroscopy in the differentiation of borderline from malignant epithelial ovarian tumors: A preliminary study

Feng Hua Ma; Yong Ai Li; Jia Liu; Hai Ming Li; Guo Fu Zhang; Jin Wei Qiang

doi:10.1002/jmri.26541

Role of proton MR spectroscopy in the differentiation of borderline from malignant epithelial ovarian tumors: A preliminary study

J Magn Reson Imaging. 2019 Jun;49(6):1684-1693. doi: 10.1002/jmri.26541. Epub 2018 Oct 24.

Authors

Feng Hua Ma^{1

2}, Yong Ai Li¹, Jia Liu², Hai Ming Li¹, Guo Fu Zhang², Jin Wei Qiang¹

Affiliations

¹ Department of Radiology, Jinshan Hospital, Shanghai Medical College, Fudan University, Shanghai, P.R. China.
² Department of Radiology, Obstetrics & Gynecology Hospital, Shanghai Medical College, Fudan University, Shanghai, P.R. China.

PMID: 30353967
DOI: 10.1002/jmri.26541

Abstract

Background: Due to the overlapping imaging appearances between borderline and malignant epithelial ovarian tumors (EOTs), borderline EOTs often represent a diagnostic challenge on conventional MRI. Proton magnetic resonance spectroscopy (¹ H-MRS) might have potential to differentiate borderline from malignant tumors.

Purpose: To investigate the ability of ¹ H-MRS to differentiate borderline from malignant EOTs.

Study type: Prospective.

Population: In all, 278 patients with adnexal masses.

Field strength/sequence: 1.5 T Siemens Avanto MRI system and ¹ H-MRS using a point-resolved spectroscopy sequence (PRESS).

Assessment: Resonance peak integrals of the most common metabolites were analyzed and compared between the two groups.

Statistical tests: The ratios of metabolites between borderline and malignant EOTs were compared with the Mann-Whitney U-test. A receiver operating characteristic (ROC) curve was used to determine their differential diagnosis performances.

Results: In the solid components of borderline and malignant EOTs, the mean Cho/Cr, NAA/Cr, and NAA/Cho ratios were 4.4 ± 1.1 and 9.9 ± 2.8; 10.4 ± 3.0 and 2.2 ± 1.0; and 2.4 ± 0.7 and 0.3 ± 0.1, respectively (all P < 0.001). The sensitivity, specificity, and area under the curve (AUC) were 91%, 100%, and 0.98 for the Cho/Cr ratio; 100%, 98%, and 0.99 for the NAA/Cr ratio; and 100%, 100%, and 1.00 for the NAA/Cho ratio, respectively. In the cystic components, the mean Cho/Cr, NAA/Cr, and NAA/Cho ratios were 3.2 ± 0.8 and 5.1 ± 1.2; 9.1 ± 3.4 and 2.3 ± 1.4; and 2.9 ± 1.2 and 0.5 ± 0.4, respectively (all P < 0.001). The sensitivity, specificity, and AUC were 84%, 82%, and 0.89 for the Cho/Cr ratio; 94%, 97%, and 0.99 for the NAA/Cr ratio; and 94%, 97%, and 0.99 for the NAA/Cho ratio, respectively.

Data conclusion: The NAA/Cho ratio is a reliable biomarker for differentiating borderline from malignant EOTs.

Level of evidence: 2 Technical Efficacy: Stage 3 J. Magn. Reson. Imaging 2019;49:1684-1693.

Keywords: magnetic resonance imaging; magnetic resonance spectroscopy; ovarian neoplasms.

MeSH terms

Adult
Area Under Curve
Diagnosis, Differential
Diagnostic Imaging
Female
Humans
Middle Aged
Neoplasms / diagnostic imaging*
Neoplasms, Glandular and Epithelial / diagnostic imaging*
Observer Variation
Ovarian Neoplasms / diagnostic imaging*
Prospective Studies
Proton Magnetic Resonance Spectroscopy*
ROC Curve