The diagnostic performance of ADC value for tumor grade,deep myometrial invasion and lymphovascular space invasion in endometrial cancer: a meta-analysis

Abstract

Background

Disputes exist regarding whether the apparent diffusion coefficient (ADC) can differentiate the tumor grade, deep myometrial invasion and lymphovascular space invasion (LVSI) in endometrial cancer. The aim of this review was to assess the diagnostic performance of the ADC value in endometrial cancer.

Material and Methods

The PubMed, Web of Science, Embase and Cochrane Library databases were searched for studies that used the ADC value to assess tumor grade, deep myometrial invasion and LVSI in endometrial cancer. We used forest plots to analyze the heterogeneity and generate the pooled sensitivity (SEN) and specificity (SPE). We used summary receiver operating characteristic (SROC) curves to work out the area under the SROC curve (AUC). Likelihood ratios (LRs) were also obtained.

Results

Of the 460 identified studies, 11 studies met our inclusion criteria and were included. Overall, nine studies (491 patients) aimed at differentiating high tumor grade had a pooled SEN, SPE and AUC of 77%, 73% and 81%, respectively; three studies (181 patients) for differentiating deep myometrial invasion had a pooled SEN, SPE and AUC of 71%, 67% and 77%, respectively; and two studies (106 patients) for differentiating LVSI had a pooled SEN and SPE of 66% and 74%, respectively. The positive and negative LRs were 2.77 and 0.35 for the tumor grade, 2.08 and 0.45 for deep myometrial invasion, and 2.48 and 0.45 for LVSI.

Conclusion

This meta-analysis showed that the ADC value had a moderate diagnostic performance for the tumor grade, deep myometrial invasion and LVSI in endometrial cancer.

Keywords

Endometrial cancer tumor grade myometrial invasion lymphovascular space invasion apparent diffusion coefficient meta-analysis

Introduction

In industrialized countries, endometrial cancer is one of the most common malignant gynecological tumors with an increasing morbidity (1,2). Surgical resection remains the main therapeutic and staging approach. The therapeutic decision depends on the tumor grade, depth of myometrial invasion and lymphovascular space invasion (LVSI). Arguments exist regarding whether pelvic lymphadenectomy should be performed in women with early-stage endometrial cancer. Randomized controlled trials have not provided support for routine systematic pelvic lymphadenectomy in low-risk early-stage endometrial cancer (3). Furthermore, for cervical stenosis patients, it is difficult to perform the curettage. Thus, preoperative accurate assessments of the tumor grade and the International Federation of Gynecology and Obstetrics (FIGO) stage in endometrial cancer are essential for the treatment decision, surgical planning and prognosis evaluation (4).

Magnetic resonance (MR) imaging has an excellent spatial resolution and soft-tissue contrast. Its multiparameter imaging capability makes it a good imaging tool for the preoperative evaluation of endometrial cancer. MR imaging has been used to assess tumor grade, myometrial invasion and LVSI (5). Moreover, studies have shown that quantitative apparent diffusion coefficient (ADC) measurement of diffusion-weighted imaging (DWI) provides a new approach for determining these factors in endometrial cancer (6). However, these studies included a limited number of patients and reported varied diagnostic performances (7 –10). Thus, there is a need for an updated systematic review assessing the diagnostic value of the quantitative ADC value.

The purpose of this meta-analysis is to determine the diagnostic performance of the quantitative ADC value for the tumor grade, deep myometrial invasion and LVSI in endometrial cancer, using pathological diagnosis as a reference standard.

Material and methods

This meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) recommendations (11).

Search strategy

We conducted online literature searches to identify studies up to 18 December 2018 in a wide range of electronic databases: PubMed, Web of Science, Embase and Cochrane Library. The key words were as follows: ‘endometrial cancer’, ‘ADC’, ‘tumor grade’ or ‘myometrial invasion’, ‘lymphovascular space invasion’ (see supplementary material 1). These searches were performed separately and had no limitation regarding language or time. Two investigators screened the search results to exclude irrelevant articles. The full text was examined for potentially eligible studies, and disagreements were resolved through discussion.

Study inclusion and exclusion criteria

We included studies that met the following criteria:

Types of research: both retrospective and prospective cohort studies.

Participants: regardless of the FIGO stage, only the patient whose final stage was based on the results from surgical staging.

Target condition: presurgical MR imaging diagnosis of tumor grade, deep myometrial invasion and LVSI in primary endometrial cancer.

Index test: DWI-MR imaging sequences with the ADC value of the tumor, without considering the applied protocol of acquisition (method of drawing region of interest (ROI) or using MR imaging or positron emission tomography (PET)-MR imaging). All ADC values were derived from the tumor.

Reference standard: pathological diagnosis of tumor grade, presence of deep myometrial invasion and LVSI.

Minimum data required: sufficient data to work out the true positive (TP), false positive (FP), false negative (FN) and true negative (TN).

The exclusion criteria were as follows:

Study population fewer than 10 patients.

Publication type other than original research (e.g. review articles).

Pathological diagnosis was not derived from surgical specimen, such as D & C (endometrial biopsy or dilatation and curettage).

Studies enrolled not only endometrial cancer (e.g. mixed endometrial cancer with cervical cancer).

Incomplete data to establish the 2 × 2 table.

Data extraction and quality assessment

Two reviewers independently extracted the data of the diagnostic accuracy results and additional useful information. Disagreements were resolved by a third physician. The main outcome indexes extracted were the number of patients with endometrial cancer, sensitivity (SEN) and specificity (SPE) of each original study (If a study had two or more SENs and SPEs, the result with the highest accuracy was extracted). G3 is defined as high grade, and G1 and G2 are defined as low grade. Deep invasion is defined as myometrial invasion greater than 50% of thickness. LVSI is considered positive by referring histological reports. The clinical stage refers to the FIGO classification.

We referred to the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) to assess the quality of the included studies (12). In the Index Test domain, studies that used the mean ADC value from the whole tumor volume were considered low risk. In the Flow and Timing domain, which refers to the time interval between imaging and surgical resection, 30 days was considered appropriate.

Statistical analysis

Meta-disc1.4 (Universidad Complutense, Madrid, Spain), SPSS 20 (IBM, Chicago, IL, USA), STATA 15.0 (Stata Corp, College Station, TX, USA and Open Meta Analyst) and Reveman 5.2 were used to analyze the data. We assessed publication bias using a Deeks’ funnel plot and a Deeks’ asymmetry test (13). The calculation formula of the SEN and SPE were as follows: SEN = TP/(TP + FN); SPE = TN/(FP + TN). TP + FN was equal to the histological positive patient number (patients’ number of G3, deep myometrial invasion or positive LVSI). FP + TN was equal to the negative patient number (patients’ number of G1 + G2, myometrial invasion less than 50% or negative LVSI).

The heterogeneity of the SEN and SPE of each included study was analyzed using the forest plots on the basis of a random-effects model. The summary receiver operating characteristic (SROC) curve was used to assess the pooled diagnostic performance. The presence of heterogeneity was assessed by means of a test on the Q statistic and I square index (14,15). A p value <0.05 indicated the existence of heterogeneity. I square was used to interpret the degree of heterogeneity. The heterogeneity was considered significant when I square was greater than 50%. A Spearman correlation coefficient was used to determine the threshold effect (a strong positive correlation of r > 0.6 or p < 0.05 was suggestive of the threshold effect) (16).

Clinical utility

Positive and negative likelihood ratios (LRs) for the ADC value were obtained to characterize the clinical utility of the ADC value. The posttest probability of endometrial cancer was calculated by LRs (17). The details of the calculation process are shown in supplementary material 1. The ADC value posttest probabilities for G3, deep myometrial invasion and LVSI were calculated separately.

Results

Included studies

A total of 460 studies were identified through the electronic search. The flow chart of the selection process is shown in Fig. 1. Among these studies, 391 studies were discarded after reviewing the title or abstract; 39 studies were discarded after reviewing the full text; and 19 studies were discarded for failing to meet the inclusion criteria. The remaining 11 studies were ultimately included, with 9 studies for the tumor grade (7 –9,18 –23), 3 studies for the myometrial invasion (9,24,25) and 2 studies for the LVSI (7,9). No additional studies were identified from the references or reviews. No discrepancies occurred between the two reviewers during the study selection.

Fig. 1.

Flowchart shows the flow of information extraction of this meta-analysis.

Methodological quality of included studies

The risk of bias and applicability concerns graph of the selected studies are shown in Fig. 2(a) and (b). We referred to the QUADAS-2 to assess the included studies. In the Patient Selection domain, the studies that did not indicate patients were enrolled consecutively were considered an unclear risk; the studies that did not indicate inclusion or exclusion criteria or clearly describe the reasons for eliminated patients were defined as an unclear risk. In the Index Test domain, one study did not implement blinding (20), while three studies did not describe the blinding (9,19,25); two studies used the quartile ADC and one study used the max ADC from the whole tumor volume (7,8,18), four studies’ mean ADC value, which was derived from the tumor on a single slice, were considered an unclear risk (19 –21,23), and two studies’ minimum ADC values derived from PET-MR imaging were considered a high risk (9,25). In the Flow and Timing domain, four studies had an interval time of more than 30 days between MR imaging and the operation (the longest interval time was 51 days) (9,18,23,24), and two studies did not describe the interval time (7,19). Details show in the supplementary material 2. The Deeks’ funnel plot shows no publication bias for the tumor grade (Fig. 3).

Fig. 2.

(a) Risk of bias and applicability concerns summary graph shows review authors’ judgements regarding each domain for each included study. (b) Risk of bias and applicability concerns. Review authors’ judgements regarding each domain present as percentages across included studies. Stacked bar charts of QUADAS-2 scores of methodologic study quality expressed as the percentage of studies that met each criterion. For each quality domain, the proportions of included studies that suggest low, high, or unclear risk of bias and/or concerns regarding applicability are displayed in green, red and yellow, respectively.

Fig. 3.

Deeks’ funnel plot for graphical exploration of publication bias for the tumor grade. Deeks’ test for asymmetry is not significant (p = 0.62). Deeks’ funnel plot shows an absence of publication bias.

Characteristics of included studies

A total of 778 patients (491 patients for tumor grade, 181 patients for myometrial invasion and 106 patients for LVSI) in academic diagnostic studies in periods that ranged from January 2005 to March 2016 were included in the final analyses. Overall, four studies were prospective, three studies were performed using 3T MR (two studies used PET-MR), and eight studies were performed using 1.5T MR. There were six studies that enrolled endometrioid and nonendometrioid tumors, while the other studies enrolled only endometrioid tumors. A total of six studies drew ROIs of the whole tumor volume, in which one study included the necrotic area in the ROI and two studies did not describe whether the ROI included the necrotic area. The main patient characteristics of the primary studies are listed in Table 1. The age was similar across studies (range 24–93 years). The menstrual status of patients was not taken into consideration.

Table 1.

Characteristics of included studies.

	Author	Year	Study design	Patient number	Mean age (year)	Field (T)	b value (s/mm²)	ADC cut-off value (10⁻³ mm²/s)	ADC type	ROI selection	Necrotic area	Histopathologic subtypes	Interval time (d)
Tumor grade	Woo	2013	Retro.	33	52 (30–84)	1.5	0, 600, 1000	1.32	95% ADC	Volume	Included	Mixed	1–48
	Nougaret	2015	Retro.	70	64 (24–91)	1.5	0, 500, 1000	0.73	50% ADC	Volume	Avoided	Mixed	NA
	Kishimoto	2015	Retro.	30	57 (40–76)	1.5	0, 1000	0.725	Mean ADC	SS^#	Avoided	Mixed	NA
	Shih	2015	Pro.	36	57 (33–82)	3.0*	50, 1000	NA	Mini ADC	Volume	NA	Mixed	0–26
	Bakir	2016	Retro.	50	63 (39–60)	1.5	0, 500, 1000	0.79	Mean ADC	SS	Avoided	Mixed	7
	Yan	2017	Retro.	80	52 (24–74)	1.5	0–650	0.83	Mean ADC	LS	Avoided	EEA	14
	Chen	2017	Retro.	73	55 (34–76)	3.0	0, 700, 1400, 2000	0.936	Mean ADC	Volume	Avoided	EEA	3–13
	Takayama	2018	Pro.	32	60 (36–85)	3.0	0, 500, 1000	NA	Mean ADC	SS	NA	EEA	3–51
	Lavaud	2018	Retro.	87	65 (32–84)	1.5	0, 1000	1.73	Max ADC	Volume	Avoided	Mixed	<30
Deep myometrial invasion	Husby	2014	Pro.	100	66 (32–93)	1.5	0, 1000	<0.75	Mean ADC	Volume	Avoided	Mixed	0–34
	Shih	2015	Pro.	36	57 (33–82)	3.0*	50, 1000	NA	Mini ADC	Volume	NA	Mixed	0–26
	Budak	2017	Pro.	45	57	1.5*	50, 400, 800	0.77	Mini ADC	SS	Avoided	Mixed	<14
Lymphovascular space invasion (LVSI)	Nougaret	2015	Retro.	70	64 (24–91)	1.5	0, 500, 1000	0.73	50% ADC	Volume	Avoided	Mixed	NA
	Shih	2015	Pro.	36	57 (33–82)	3.0*	50, 1000	NA	Mini ADC	Volume	NA	Mixed	0–26

Mean of three random ROIs.

*Data derived from PET-MRI.

ADC: apparent diffusion coefficient; EEA: endometrioid endometrial adenocarcinoma; LS: largest slice (slice containing the largest tumor cross-section); mixed: endometrioid endometrial adenocarcinoma and other histological subtypes; MRI: magnetic resonance imaging; NA: not available; PET: positron emission tomography; Pro: prospective; Retro: retrospective; ROI: region of interest; SS: single slice.

Diagnostic performances of ADC value in preoperative diagnosis

The diagnostic performances of the ADC value for the tumor grade, deep myometrial invasion and LVSI are shown in the forest plots of Fig. 4 and in the SROC plots of Fig. 5.

Fig. 4.

Forest plots of diagnostic performance (sensitivity and specificity) of each study for the tumor grade (a, b); the deep myometrial invasion (c, d) and the LVSI (e, f). Horizontal lines indicate a per-node basis with 95% CIs. Diamond at bottom and vertical dotted line indicate the pooled sensitivity or specificity with 95% CIs. Heterogeneity statistics (I square value, Q value) for sensitivity and specificity are also shown.

Fig. 5.

SROC curve of the included studies using ADC values to differentiate the G3 tumor (a) and the deep myometrial invasion (b).

All analyses were performed in a random-effects model. No significant heterogeneity was identified for the SEN or SPE in the tumor grade (p = 0.47 and I square = 0.0% and p = 0.22 and I square = 24.9%, respectively). Significantly heterogeneity was identified for the SEN and SPE in deep myometrial invasion (p = 0.04 and I square = 69.7%, and p = 0.01 and I square = 77.9%, respectively). Meta-regression could not be performed to explain the significant heterogeneity of myometrial invasion because of the limited included studies. No significant heterogeneity was identified for the SEN and SPE in LVSI (p = 0.85 and I square = 0.0% and p = 0.20 and I square = 39.0%, respectively). The Spearman rank correlation analysis showed that the correlation coefficient was -0.45 (p = 0.22) for the tumor grade and 0.5 (p = 0.67) for myometrial invasion, which indicated no threshold effect in the tumor grade and deep myometrial invasion. The pooled SEN, SPE and AUC were 77% (95% confidence interval (CI): 69.3–84.0%), 73% (95% CI: 67.9–77.4%) and 81% (95% CI: 78.2–83.5%), respectively, for tumor grade; 71% (95% CI: 60.1–80.8%), 67% (95% CI: 56.6–75.8%) and 77% (95% CI: 69.9–83.1%), respectively, for deep myometrial invasion; and 66% (95% CI: 50.5–79%) and 74% (95% CI: 60.8–84.5%) for LVSI. The SROC could not be established due to only two studies being included.

The LRs of the ADC value were derived from the included nine studies (491 patients) for the prediction of tumor grade; from the three studies (181 patients) for the prediction of deep myometrial invasion; and from two studies (106 patients) for the prediction of LVSI. The positive LRs of the ADC value for the tumor grade, deep myometrial invasion and LVSI were 2.77 (95% CI: 2.22–3.45), 2.08 (95% CI: 1.27–3.39) and 2.48 (95% CI: 1.47–4.20), respectively, and the negative LRs were 0.35 (95% CI: 0.26–0.48), 0.45 (95% CI: 0.25–0.82) and 0.45 (95% CI: 0.30–0.69), respectively. The pretest probabilities were 28% for G3 tumor grade, 44% for deep myometrial invasion and 40% for LVSI. The posttest probabilities of ADC value used to diagnose G3 tumor, deep myometrial invasion and LVSI, were 52% (95% CI: 46–57%), 62% (95% CI: 50–73%) and 62% (95% CI: 54–77%), respectively. The posttest probabilities of ADC value used to exclude G3 tumor, deep myometrial invasion and LVSI, were 12% (95% CI: 9–16%), 26% (95% CI: 16–39%) and 23% (95% CI: 16–31%), respectively.

Discussion

Histological grade, depth of myometrial invasion and LVSI are important prognostic factors strongly associated with the management strategies and survival of individuals with endometrial cancer (26). In this meta-analysis, the preoperative diagnostic ability of the ADC value in these three factors was investigated.

Previous studies have shown that the ADC value was correlated with the tumor grade and could be used to differentiate low grade from high grade endometrial cancers. It has been suggested that high risk tumors tend to have a low ADC value due to the tumor’s high cellularity, proliferation and perfusion and less extracellular space (27,28). Previous studies have reported that the SEN and SPE for the tumor grade ranged from 55–100% and 70.4–89%, respectively (7 –9,18 –23). This meta-analysis showed a moderate diagnostic performance, with the pooled SEN and SPE of 77% and 73%, respectively. However, controversy exits regarding whether the ADC value can be used to differentiate the tumor grade (29,30). A considerable overlap of the ADC value has been found between low grade and high grade endometrial cancer in previous studies, which indicates other factors, such as nuclear atypia and the different pathological proteins, may contribute to the tumor grade that cannot be displayed by the ADC value (31,32). Furthermore, ADC values are different in various histological subtypes of endometrial cancer. According to the study of Takeuchi et al., the ADC value of carcinosarcomas is higher than that of G3 endometrioid adenocarcinomas (33). Carcinosarcomas have a higher frequency of necrosis and epithelial cystic components, which can increase the tumor’s ADC value. Moreover, this meta-analysis included not only endometrioid adenocarcinoma but also nonendometrioid adenocarcinomas. Furthermore, the difference in the genomic characterization of endometrial cancer may be attributed to the different biological behavior of the same traditional histology tumor grade (34). Thus, the performance of the ADC value may be affected.

Previous studies have shown that the ADC value could be used to assess deep myometrial invasion, with the SEN and SPE ranging from 47.6–78% and 57–88%, respectively (9,24,25). In this meta-analysis, moderate SEN and SPE were identified in evaluating the deep myometrial invasion. In contrast, DWI could show the boundary of the tumor and normal myometrium more clearly than T2-weighted imaging, and a high b-value-DWI had a high contrast-to-noise ratio.(35) The SEN and SPE of DWI were 86% and 86%, respectively, for deep myometrial invasion, which indicates a higher diagnosis performance achieved than the ADC value (36). Furthermore, the ADC value measurement seriously depended on drawing the ROI on the ADC map which has a low signal-to-noise ratio. Moreover, it is difficult to draw an accurate ROI along with an unclear tumor boundary because many patients had small tumors and an atrophic uterus, and it is also affected by the movement or position of the uterus. Thus, substantial useful information would be missed.

It is controversial whether the ADC value is able to identify the LVSI (7,9). Our meta-analysis included only two studies, which limited us to come to a conclusion. We thought that the different stages of endometrial cancers included may affect the ADC value to identify the LVSI.

Several potential explanations for the moderate diagnostic performance of the ADC value in these three predictive factors are as follows: first, the tumor’s ADC value may be different from the tumor invasion area’s ADC value, and it also cannot represent the invasion area. Second, the acquisitions of the ROIs were different in the included studies. Tumor heterogeneity was represented more accurately using the tumor volumetric ADC value than the selected ROI ADC value (37). Third, a different ROI drawing approach had a significant difference in the ADC value (38). Fourth, the different menstrual status and phase influence the ADC value (39).

The clinical usefulness of an examination is well expressed by the LRs. Clinically useful tests should have a high positive LR (>5.0, ‘good for confirming disease’) and a low negative LR (<0.2, ‘good for ruling out disease’) (17). The positive LR of 2.77 and the negative LR of 0.35 indicated limited clinical usefulness of the ADC value in diagnosing a G3 tumor. In patients with G3 tumors, a positive ADC value would increase the probability from a pretest of 28% to a posttest of 52%. This finding indicated that only 52% of the patients diagnosed as G3 tumors by the ADC value were actually G3 tumors. Insufficient studies were available for drawing a powerful conclusion in the deep myometrial invasion and LVSI. Therefore, for endometrial cancer, a single quantitative ADC value had a limited diagnostic performance for these three factors. Combined with other techniques, such as endometrial cancer radiomics, a quantitative ADC value might be more helpful to assess the tumor invasiveness (40).

There were several limitations in this meta-analysis. First, a small number of studies, particularly for deep myometrial invasion and LVSI, were included in this study, which did not enable investigation of all potential causes of heterogeneity. Second, we did not separately analyze the three predictive factors to reduce the interaction, and comparisons of the ADC value between superficial and deep myometrial invasions were not performed in different tumor grades. Third, we did not analyze the diagnostic performance of the ADC value of different ROIs.

In conclusion, this meta-analysis showed that the ADC value had a moderate diagnostic performance for the tumor grade, deep myometrial invasion and LVSI in endometrial cancers. Further studies that combine the ADC value with other techniques of analysis and imaging are warranted.

Supplemental Material

Supplemental Material1 - Supplemental material for The diagnostic performance of ADC value for tumor grade, deep myometrial invasion and lymphovascular space invasion in endometrial cancer: a meta-analysis

Supplemental material, Supplemental Material1 for The diagnostic performance of ADC value for tumor grade, deep myometrial invasion and lymphovascular space invasion in endometrial cancer: a meta-analysis by Bi Cong Yan, Mei Ling Xiao, Ying Li and Jin Wei Qiang in Acta Radiologica

Supplemental Material

Supplemental Material2 - Supplemental material for The diagnostic performance of ADC value for tumor grade, deep myometrial invasion and lymphovascular space invasion in endometrial cancer: a meta-analysis

Supplemental material, Supplemental Material2 for The diagnostic performance of ADC value for tumor grade, deep myometrial invasion and lymphovascular space invasion in endometrial cancer: a meta-analysis by Bi Cong Yan, Mei Ling Xiao, Ying Li and Jin Wei Qiang in Acta Radiologica

Footnotes

Acknowledgements

Bi Cong Yan and Mei Ling Xiao contributed equally to the writing of this article. The author contributions were as follows:

Bi Cong Yan: literature research, acquisition of data, quality assessment, drafting of the manuscript.

Mei Ling Xiao: literature research, acquisition of data.

Ying Li: quality assessment, acquisition of data, statistical analysis.

Jin Wei Qiang: study concept and design, critical revision of the manuscript for important intellectual content, study supervision.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by Shanghai Municipal Commission of Health and Family Planning (No. ZK2015A05)

References

Amant

Moerman

Neven

, et al. Endometrial cancer. Lancet 2005; 366:491–505.

Siegel

Naishadham

Jemal

Cancer statistics, 2013.

CA Cancer J Clin 2013; 63:11–30.

Kitchener

Swart

Qian

, et al. Efficacy of systematic pelvic lymphadenectomy in endometrial cancer (MRC ASTEC trial): a randomised study. Lancet 2009; 373:125–136.

Beddy

O'Neill

Yamamoto

, et al. FIGO staging system for endometrial cancer: added benefits of MR imaging. Radiographics 2012; 32:241–254.

Boronow

Morrow

Creasman

, et al. Surgical staging in endometrial cancer: clinical-pathologic findings of a prospective study. Obstet Gynecol 1984; 63:825–832.

Harry

Deans

Ramage

, et al. Magnetic resonance imaging in gynecological oncology. Int J Gynecol Cancer 2009; 19:186–193.

Nougaret

Reinhold

Alsharif

, et al. Endometrial cancer: combined MR volumetry and diffusion-weighted imaging for assessment of myometrial and lymphovascular invasion and tumor grade. Radiology 2015; 276:797–808.

Lavaud

Fedida

Canlorbe

, et al. Preoperative MR imaging for ESMO-ESGO-ESTRO classification of endometrial cancer. Diagn Interv Imaging 2018; 99:387–396.

Shih

Yen

Chen

, et al. Standardized uptake value and apparent diffusion coefficient of endometrial cancer evaluated with integrated whole-body PET/MR: correlation with pathological prognostic factors. J Magn Reson Imaging 2015; 42:1723–1732.

10.

Cao

Gao

Sun

, et al. Apparent diffusion coefficient of diffusion weighted MRI in endometrial carcinoma-relationship with local invasiveness. Eur J Radiol 2012; 81:1926–1930.

11.

Liberati

Altman

Tetzlaff

, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol 2009; 62:e1–e34.

12.

Schuetz

Tackmann

Hamm

, et al. [ Quality of diagnostic accuracy studies: QUADAS (quality assessment of diagnostic accuracy studies included in systematic reviews)]. Rofo 2010; 182:939–942.

13.

Deeks

Macaskill

Irwig

The performance of tests of publication bias and other sample size effects in systematic reviews of diagnostic test accuracy was assessed.

J Clin Epidemiol 2005; 58:882–893.

14.

Dinnes

Deeks

Kirby

, et al. A methodological review of how heterogeneity has been examined in systematic reviews of diagnostic test accuracy. Health Technol Assess 2005; 9:1–113.

15.

Higgins

Thompson

Deeks

, et al. Measuring inconsistency in meta-analyses. BMJ 2003; 327:557–560.

16.

Shan P, Rong C, JinWei Q, et al. Utility of DWI with quantitative ADC values in ovarian tumors: a meta-analysis of diagnostic test performance. Acta Radiol 2018;59:1386–1394.

17.

Jaeschke

Guyatt

Sackett

DL.

Users' guides to the medical literature. III. How to use an article about a diagnostic test. B. What are the results and will they help me in caring for my patients? The Evidence-Based Medicine Working Group. JAMA 1994; 271:703–707.

18.

Woo

Cho

Kim

, et al. Histogram analysis of apparent diffusion coefficient map of diffusion-weighted MRI in endometrial cancer: a preliminary correlation study with histological grade. Acta Radiol 2014; 55:1270–1277.

19.

Kishimoto

Tajima

Maeda

, et al. Endometrial cancer: correlation of apparent diffusion coefficient (ADC) with tumor cellularity and tumor grade. Acta Radiol 2016; 57:1021–1028.

20.

Bakir

Sanli

, et al. Role of diffusion-weighted MRI in the differential diagnosis of endometrioid and non-endometrioid cancer of the uterus. Acta Radiol 2017; 58:758–767.

21.

Yan

Zhao

Liang

, et al. Can the apparent diffusion coefficient differentiate the grade of endometrioid adenocarcinoma and the histological subtype of endometrial cancer? Acta Radiol 2018; 59:363–370.

22.

Chen

, et al. Quantitative evaluation of diffusion-kurtosis imaging for grading endometrial carcinoma: a comparative study with diffusion-weighted imaging. Clin Radiol 2017; 72:911–995.

23.

Takayama

Nishie

Togao

, et al. Amide proton transfer MR imaging of endometrioid endometrial adenocarcinoma: association with histologic grade. Radiology 2018; 286:909–917.

24.

Husby

Salvesen

Magnussen

, et al. Tumour apparent diffusion coefficient is associated with depth of myometrial invasion and is negatively correlated to tumour volume in endometrial carcinomas. Clin Radiol 2015; 70:487–494.

25.

Surer

Toptas

Aydin

, et al. Correlation of minimum apparent diffusion coefficient and maximum standardized uptake value of the primary tumor with clinicopathologic characteristics in endometrial cancer. Mol Imaging Radionucl Ther 2017; 26:24–32.

26.

Prat

Prognostic parameters of endometrial carcinoma.

Hum Pathol 2004; 35:649–662.

27.

Henzler

Schmid-Bindert

Schoenberg

, et al. Diffusion and perfusion MRI of the lung and mediastinum. Eur J Radiol 2010; 76:329–336.

28.

Yoshikawa

Ohsumi

Sugata

, et al. Relation between cancer cellularity and apparent diffusion coefficient values using diffusion-weighted magnetic resonance imaging in breast cancer. Radiat Med 2008; 26:222–226.

29.

Bharwani

Miquel

Sahdev

, et al. Diffusion-weighted imaging in the assessment of tumour grade in endometrial cancer. Br J Radiol 2011; 84:997–1004.

30.

Shen

Chiou

Wang

, et al. Diffusion-weighted single-shot echo-planar imaging with parallel technique in assessment of endometrial cancer. AJR Am J Roentgenol 2008; 190:481–488.

31.

Whittaker

Coady

Culver

, et al. Diffusion-weighted MR imaging of female pelvic tumors: a pictorial review. Radiographics 2009; 29:759–774, 774–778.

32.

Tang

Dundamadappa

Thangasamy

, et al. Correlation of apparent diffusion coefficient with Ki-67 proliferation index in grading meningioma. AJR Am J Roentgenol 2014; 202:1303–1308.

33.

Takeuchi

Matsuzaki

Uehara

, et al. Pathologies of the uterine endometrial cavity: usual and unusual manifestations and pitfalls on magnetic resonance imaging. Eur Radiol 2005; 15:2244–2255.

34.

Burk

R D

Chen

Saller

, et al. Integrated genomic and molecular characterization of cervical cancer. Nature 2017; 543:378–384.

35.

Kuhl

Textor

Gieseke

, et al. Acute and subacute ischemic stroke at high-field-strength (3.0-T) diffusion-weighted MR imaging: intraindividual comparative study. Radiology 2005; 234:509–516.

36.

Andreano

Rechichi

Rebora

, et al. MR diffusion imaging for preoperative staging of myometrial invasion in patients with endometrial cancer: a systematic review and meta-analysis. Eur Radiol 2014; 24:1327–1338.

37.

Lambregts

Beets

Maas

, et al. Tumour ADC measurements in rectal cancer: effect of ROI methods on ADC values and interobserver variability. Eur Radiol 2011; 21:2567–2574.

38.

Inoue

Fujii

Kaneda

, et al. Apparent diffusion coefficient (ADC) measurement in endometrial carcinoma: effect of region of interest methods on ADC values. J Magn Reson Imaging 2014; 40:157–161.

39.

Tamai

Koyama

Saga

, et al. Diffusion-weighted MR imaging of uterine endometrial cancer. J Magn Reson Imaging 2007; 26:682–687.

40.

Ytre-Hauge

Dybvik

Lundervold

, et al. Preoperative tumor texture analysis on MRI predicts high-risk disease and reduced survival in endometrial cancer. J Magn Reson Imaging 2018; 48:1637–1647.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.10 MB

0.47 MB