Impact of DWI and ADC values in Ovarian-Adnexal Reporting and Data System (O-RADS) MRI score

Patients and study setting

The study was approved by the institutional review board, and informed consent was required for data analysis.

This is a retrospective single-centre cohort study conducted between January 2015 and June 2022 in the Radiology Department of Umberto I Hospital, Sapienza University of Rome, Italy.

We initially identified 173 patients consecutively with 213 adnexal masses indeterminate on ultrasound examination.

Inclusion criteria were: age > 18 years, standardized MRI examination with DWI and DCE sequences, subsequent surgery with histological examination or stability at follow-up imaging for at least 1 year.

Exclusion criteria were: age < 18 years (n. 2), no standard MRI examination (n 25), previous hysterectomy (n. 3), acute symptoms (n. 2) and no histopathological findings or follow-up < one year (n. 9).

The final cohort included 140 patients with 172 adnexal masses.

Patient enrolment, data collection and lesion classification according to ORADS-MRI score were retrospective.

Magnetic resonance imaging

All MRI examinations were performed on a 3-T system (GE Discovery MR 750, GE Healthcare, Milwaukee, WI, USA) and on a 1.5-T system (MAGNETOM Avanto; Siemens Healthcare, Erlangen, Germany) using a 32-channel phased-array coil positioned on the lower abdomen.

Before the beginning of the examination, 20 mg of joscine N-butylbromide (Buscopan; Boehringer Ingelheim, Ingelheim, Germany) was injected intravenously to reduce motion artefacts caused by bowel peristalsis, if not contraindicated.

The standard MRI protocol included the following sequences, focusing on the lower abdomen from the pubic symphysis to the iliac crests: T2 fast spin-echo (FSE) weighted imaging (WI). On the sagittal, axial and coronal planes; axial T1 FSE WI with and without fat saturation (LAVA-Flex implementation of Dixon method), axial diffusion weighted images (DWI) with b-values of 0–1000 s/mm2 to obtain apparent diffusion coefficient (ADC) maps; dynamic T1weighted 3D gradient-echo with fat saturation in the axial plane during contrast uptake and delayed post-contrast T1-weighted 3D gradient echo with fat saturation in the axial plane.

Gadolinium chelate (gadoteric acid) was given at a dose of 0.2 mL per kilogram of body weight by using a power injector at a rate of 2 mL/sec, followed by 20 mL of normal saline to flush the tubing. Images were obtained sequentially at 2.4-s intervals beginning 10 s before the bolus injection, for a total of 320 s, Table 1.

Table 1 MR scanning parameters in detailImage analysis

All images were analysed independently by two radiologists (S.C. and V.M.) with 4 years and 1 year of experience in female pelvic imaging, respectively.

All readers, blind to clinical and histological data, retrospectively classified adnexal masses according to the six categories of the O-RADS MRI scoring system, published by Thomassin et al. in January 2020 [5].

O-RADS MRI risk stratification system has six classification categories: O-RADS MRI 0 (incomplete examination), O-RADS MRI 1 (normal ovaries), O-RADS MRI 2 (almost certainly benign), O-RADS MRI 3 (low risk), O-RADS MRI 4 (intermediate risk) and O-RADS MRI 5 (high risk).

According to previously published studies, the following MRI characteristics were analysed for each adnexal mass [12,13,14,15]: laterality, morphology (unilocular or multilocular), wall and septa (thin or irregular), content (fluid, solid, mixed), tissue characteristics (solid, adipose, fibrotic, blood), T2weighted signal intensity (SI) and DWI SI, free intraperitoneal fluid, peritoneal implants, time intensity curve (TIC) of the solid component. A TIC is created by placing a region of interest (ROI) in the most enhancing part of any solid tissue of the lesion and another on the external myometrium, trying to avoid the external myometrial vessels or fibroids, in accordance with O-RADS MRI system [16].

A gradual increase in the signal intensity of the solid tissue, without a well-defined “shoulder”, was defined as curve type 1. A moderate initial increase in the signal intensity of solid tissue relative to that of myometrium, followed by a plateau, was defined as curve type 2. An initial increase in the signal intensity of solid tissue that was steeper than that of myometrium was defined as curve type 3[17].

ADC

Two radiologists (S.C. and L.M.) with 4 and 27 years of experience in female pelvic imaging, respectively, analysed ADC maps obtained from single-exponential DWI sequences on a post-processing workstation (AW VolumeShare 7, GE Healthcare, Milwaukee, WI, USA).

A tissue is considered benign if it is hyperintense in ADC and hypointense at b 1000, while it is malignant if it is hypointense in ADC and hyperintense at b 1000 [18].

Both radiologists independently drew a two-dimensional (2D) region of interest (ROI) on the ADC map. In particular, a circular ROI was placed in the slice containing the darkest part of the lesion, corresponding to the highest signal intensity at high b-values in the DWI, and correlates with the area of contrast enhancement on the post-contrast image. Moreover, T2-weighted and DCE images were used as anatomical reference.

The ROI was positioned by excluding areas of macroscopic necrosis, surrounding structures and areas with susceptibility artefacts. When lesions with multiple solid components were found, 4 to 6 ROIs were positioned on the targets and the ROI with the lowest ADC value was recorded.

Adipose tissue and blood components show low signal in ADC and are a common pitfalls in DWI. A combination of low signal at b1000 and low signal in ADC associated with markedly hypointense tissue in T2 weighted sequence (dark T2/dark DWI) is referred to benign lesions with fibrotic content [14]. Adnexal masses, without enhancing solid tissue, with adipose, haematic or fibrotic contents, are considered benign (O-RADS MRI score 2) and were excluded from the analysis of the ADC value.

Reference standard

Histopathological diagnosis or imaging follow-up for at least 1 year was the reference standard. Histological diagnosis, which is considered the gold standard, was performed after complete surgical excision or after biopsy for inoperable lesions. Lesions were analysed by a pathologist blinded to MRI findings with more than 5 years of experience in female genital tumours and were classified as benign, borderline and malignant lesions according to the World Health Organization’s (WHO) International Classification of Diseases for Oncology (ICD-O) [19]. Malignant lesions were further classified into low grade and high grade based on the extent of cell anaplasia and the percentage of undifferentiated cells [20].

Statistical analysis

Statistical analysis was performed using SPSS 25 (IBM SPSS statistics). To define the optimal cut-off for the ADC variable in predicting O-RADS categories, two different ROC curves were made; if the AUC was significant, the optimal cut-off was identified, to maximize the sum of sensitivity and specificity. ANOVA test and Bonferroni test were performed to detect differences in the ADC variable according to histotype (borderline tumours, low-grade serous carcinomas or high-grade serous carcinomas + ovarian carcinomas G3). Chi-squared test, Gamma Index and Cohen’s Kappa were calculated to measure the agreement between two readers. Each test was considered statistically significant if the p-value was lower than 0.05.

留言 (0)

沒有登入
gif