Preoperative quantitative dynamic contrast-enhanced MRI and diffusion-weighted imaging predict aggressive disease in endometrial cancer

Abstract

Background

Quantitative dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and diffusion-weighted imaging (DWI) may yield preoperative tumor biomarkers relevant for prognosis and therapy in cancer.

Purpose

To explore the value of preoperative DCE-MRI and DWI for the prediction of aggressive disease in endometrial cancer patients.

Material and Methods

Preoperative MRI (1.5-T) from 177 patients were analyzed and imaging parameters reflecting tumor microvasculature (from DCE-MRI) and tumor microstructure (from DWI) were estimated. The derived imaging parameters were explored in relation to clinico-pathological stage, histological subtype and grade, molecular markers, and patient outcome.

Results

Low tumor blood flow (F_b) and low rate constant for contrast agent intravasation (k_ep) were associated with high-risk histological subtype (P ≤ 0.04 for both) and tended to be associated with poor prognosis (P ≤ 0.09). Low tumor apparent diffusion coefficient (ADC) value and large tumor volume were both significantly associated with deep myometrial invasion (P < 0.001 for both) and were also unfavorable prognostic factors (P = 0.05 and P < 0.001, respectively).

Conclusion

DCE-MRI and DWI represent valuable supplements to conventional MRI by providing preoperative imaging biomarkers that predict aggressive disease in endometrial cancer patients.

Keywords

Magnetic resonance imaging (MRI)functional MRI MR diffusion/perfusion prognostic biomarker endometrial carcinoma

Introduction

Endometrial cancer is the most common pelvic gynecologic malignancy in industrialized countries and the incidence is increasing (1). Primary surgical treatment is currently guided by putative preoperative risk based on curettage/biopsy yielding histological subtype and grade, and on diagnostic imaging evaluating local tumor extent and distant spread. However, these preoperative assessments have reported limitations in reproducibility and accuracy when compared to the gold standard being the final surgicopathological International Federation of Gynecology and Obstetrics (FIGO) stage (2 –6). Thus, identification of robust imaging biomarkers that can aid in preoperative staging and prognostication of endometrial cancer are highly warranted.

Magnetic resonance imaging (MRI) has long been considered the preferred preoperative imaging method in endometrial cancer (7 –9). Conventional pelvic MRI provides information on tumor size, myometrial invasion, cervical stroma invasion, and suspicion of pelvic lymph node metastases, together indicative of the tumor’s aggressiveness (1,10,11). Novel functional MRI techniques have recently shown to be promising adjunct techniques in the preoperative assessment of endometrial carcinomas (7,9,12 –21).

Dynamic contrast-enhanced (DCE)-MRI is a functional imaging technique enabling characterization of tumor microvasculature and angiogenic tumor profile in vivo (14). The calculated DCE-MRI parameters are, however, affected by the employed acquisition technique as well as the choice of pharmacokinetic model (22). A few recent studies of endometrial cancer patients, using different DCE-MRI models in patient series comprising 54–80 patients, have suggested a link between specific preoperative DCE-MRI parameters and an aggressive phenotype (15,16,20,21).

Diffusion-weighted imaging (DWI) measures the microscopic mobility of water molecules in the tissue. The diffusion properties are highly influenced by tissue microstructure, microcirculation, and cellular density. The reported diagnostic performance of DWI for preoperative staging in endometrial cancer is in the range of those reported for conventional contrast-enhanced MRI (7,9,13).

The aim of this study was to explore the value of quantitative DCE-MRI and DWI tumor parameters, when based on a semi-automated approach, for the preoperative prediction of deep myometrial invasion, cervical stroma invasion, and lymph node metastases in a large patient series with endometrial cancer. We also aimed to explore whether the functional tumor parameters are related to known histologic and immunohistochemical markers of aggressive disease and to clinical outcome in endometrial cancer.

Material and Methods

Patient series, study setting, and clinical outcome

This prospective study was conducted under institutional review board-approved protocols with informed consent from all patients. From June 2009 to November 2013, preoperative pelvic MRI including DCE-MRI and DWI was performed in 177 patients in whom a uterine tumor was visible at MRI and histologically verified as endometrial carcinoma. All patients were diagnosed and treated at the same university hospital. Follow-up data regarding recurrence, progression, and survival have been collected from patient records and from correspondence with responsible physicians. Date of last follow-up was 15 December 2015 and mean follow-up for survivors was 41 months (range = 6–73 months).

Histological diagnosis and immunohistochemistry

All patients were diagnosed with primary endometrial carcinoma and 99% (175/177) of the patients were surgically staged according to the FIGO staging system (2009) (23). In the remaining 1% (2/177), the diagnosis and presumed FIGO stage were based on results from preoperative specimen and clinical diagnostic work-up. In patients surgically staged, myometrial invasion and cervical stromal invasion were estimated grossly and confirmed microscopically (24). The pathologists also documented the numbers and size of metastatic lymph nodes. DNA ploidy (diploid or aneuploid) from hysterectomy specimens was reported for 59% (105/177) of the patients.

Immunohistochemical staining for expression of estrogen receptor α (ERα) and progesterone receptor (PR) was performed in 90% (160/177) of the patients using tissue microarrays sections (TMA) generated from formalin-fixed paraffin-embedded (FFPE) tumor tissue (25,26).

MRI protocol

Imaging was performed on a 1.5-T MRI scanner (Siemens Avanto running Syngo v. B17, Erlangen, Germany) using a six-channel body coil and a standardized imaging protocol (3,15,16). To reduce motion artefacts, 20 mg butyl-scopolamine bromide (Buscopan; Boehringer, Ingelheim, Germany) was administered intravenously before scanning.

Structural MRI protocol: This included pelvic sagittal and axial oblique (perpendicular to the long axis of the uterus) T2-weighted (T2W) images, together with axial oblique T1-weighted (T1W) gradient-echo images before and 2 min after i.v. administration of contrast agent (Dotarem, Guerbet (Paris, France): 0.1 mmol gadolinium per kilogram of body weight, 3 mL/s injection speed).

Physiological MRI protocol: Pelvic DCE-MRI was obtained for 12 oblique axial slices applying a 3D spoiled gradient echo (FLASH) sequence (TE/TR = 1.05/2.64 ms, flip angle [FA] = 12°, matrix = 256 × 256, field of view (FOV) = 300 × 300 mm², slice thickness = 5 mm, number of averages [NA] = 1) with a temporal resolution of 2.49 s. Sequential images were acquired from 30 s before administration of i.v. contrast to 6.3 min after contrast injection. At 2 min post contrast medium administration, a pause of 33 s was utilized to acquire the diagnostic T1W contrast-enhanced series. Pelvic DWI was obtained from an axial 2D echo planar imaging (EPI) sequence with b-values of 0 and 1000 s/mm² (matrix = 128 × 128, slice thickness = 5 mm, TE/TR = 79/3, 100 ms, FOV = 300 × 300 mm², NA = 12).

MRI analysis

All MR images were evaluated by a radiologist with > 10 years of experience in pelvic MRI and who was blinded for tumor stage, histological diagnosis, and patient outcome. Structural MRI revealed a measurable endometrial mass in all included patients and these images were used for anatomical tumor localization in the functional images (DCE-MRI and DWI). To avoid inaccuracies due to imperfect co-registration of structural and functional images, regions of interest (ROIs) were drawn directly on the DCE-MRI images at 2 min post contrast and on the DWI ADC maps (Fig. 1). In both cases placement of the ROI aimed to comprise a representative part of the tumor, avoiding inclusion of normal myometrial tissue and necrotic or hemorrhagic areas if present. For DCE-MRI, a ROI representing normal myometrial tissue was also drawn on the same slice as the tumor ROI. The volume of the analyzed DCE-MRI tumor ROIs and normal myometrial ROIs had a median of 0.20 mL (range = 0.01–3.80) mL and 0.16 mL (range = 0.05–0.83 mL, respectively; wide range being explained by variable tumor size and variable normal myometrial tissue depicted on the selected slices. The volume of the DWI tumor ROIs had a median of 0.35 mL (range = 0.12–3.37 mL. Tumor volume was estimated based on the standard anatomical images with measurements of maximum tumor diameter in three orthogonal planes (x, y, and z) using the following equation: tumor volume = x × y × z / 2.

Fig. 1.

Axial oblique MRI and parametric maps from DCE-MRI in a 68-year-old patient with endometrial carcinoma, FIGO stage 1A, grade 3. Contrast-enhanced (CE) T1W MRI, T2W MRI, DWI with (b = 1000 s/mm²) and corresponding ADC map (upper panel). Parametric maps depicting blood flow (F_b), the transfer rate constant from blood to extravascular extracellular space (EES) (K^trans), the intravasation rate constant from EES to blood (k_ep) and the fractional volume of EES (v_e) (lower panel). At conventional diagnostic MRI the endometrial carcinomas (arrows) are typically depicted as hypo-intense relative to the surrounding myometrium on CE T1W image, hyper-intense on T2W image and DWI, and with restricted diffusion visualized by hypo-intensity on the ADC map.

The DCE-MRI time series were analyzed applying the extended Tofts pharmacokinetic model (27,28) combined with a population-based arterial input function in nordicICE (Nordic-NeuroLab AS, Bergen, Norway). From these analyses, tumor parameters representing blood flow (F_b), the transfer rate constant from blood to extravascular extracellular space (EES) (K^trans), the intravasation rate constant from EES to blood (k_ep), and the fractional volume of EES (v_e), were derived (Fig. 1). K^trans and v_e are independent model parameters from which k_ep is computed based on the equation k_ep = K^trans/v_e.

Statistical analysis

For all the derived imaging parameters of the tumor and normal myometrium, median values with 95% confidence interval (CI) of the median were calculated. The Wilcoxon signed rank test was used to analyze differences between the DCE-MRI parameters in tumor and myometrium of the same patient. The Mann–Whitney U test was used to analyze differences in imaging parameters in relation to staging parameters, histological subtype, grade, and immunohistochemical markers. Correlation between continuous imaging parameters was evaluated using Spearman’s rank-order correlation test.

Quartile limits were applied to explore prognostic value of the imaging parameters. The Mantel–Cox (log rank) linear trend test with Kaplan–Meier plot was used for univariate analyses of time to recurrence (for patients considered cured by primary treatment) or progression (for patients known to have residual disease after primary treatment). The Cox proportional hazards model was used to study the effect on survival of several variables simultaneously and to estimate hazard ratios.

The data were analyzed using SPSS 23.0 (IBM, Armonk, NY, USA). All reported P values were two-tailed and considered to indicate statistical significance when < 0.05.

Results

Patients and treatment

Clinicopathological patient characteristics are listed in Table 1. Altogether, 98% (174/177) of the patients underwent primary surgical resection. One patient (1/177) with FIGO stage 4B underwent tumor debulking without hysterectomy and two patients (2/177) were not surgically staged due to presumed FIGO stage 4.

Table 1.

Clinicopathological patient characteristics.

Patients (n)	177
Mean age (range) (years)	66 (29–93)
Postmenopausal (n, %)	162 (92)
Lymph node sampling (n, %)	149 (84)
Adjuvant therapy (n, %)	63 (36)
FIGO stage* (n, %)
IA	87 (49)
IB	43 (24)
II	20 (11)
III	23 (13)
IV	4 (2)
Histological subtype (n, %)
Endometrioid	138 (78)
Grade 1	69 (50)
Grade 2	42 (30)
Grade 3	27 (20)
Non- endometrioid	36 (20)
Missing	3 (2)

According to International Federation of Gynaecology and Obstetrics (FIGO) 2009 staging criteria.

Tumor microvasculature (DCE-MRI parameters)

The median values of the DCE-MRI perfusion parameters F_b, K^trans, and v_e were all significantly lower in endometrial tumor tissue compared to normal myometrial tissue (P < 0.001 for all, Table 2), whereas no difference was observed for k_ep.

Table 2.

Quantitative DCE-MRI perfusion parameters measured in ROIs within endometrial tumor tissue and in corresponding regions in normal myometrium (n = 177).

DCE-MRI parameters	Endometrial carcinoma Median (95% CI)	Normal myometrium Median (95% CI)	P value*
F_b (min^–1)	49 (45–54)	88 (82–95)	<0.001
K^trans (min^–1)	0.034 (0.028–0.038)	0.086 (0.079–0.089)	<0.001
k_ep (min^–1)	0.40 (0.36–0.42)	0.42 (0.39–0.43)	0.26
v_e (%)	9 (8 –11)	21 (20 –22)	<0.001

Related samples Wilcoxon signed rank test. Significant P values are given in italics.

CI, confidence interval for the median; F_b, tumor blood flow; k^ep, intravasation constant from extravascular extracellular space (EES) to blood; K^trans, transfer constant from blood to EES; v_e, fractional volume of EES.

The derived DCE-MRI tumor parameters were not significantly different in patients with a more advanced stage, i.e. deep myometrial invasion, cervical stroma invasion, or lymph node metastases (Suppl. Table 1). Low F_b and low k_ep were, however, significantly associated with high-risk histology (endometrioid grade 3 or non-endometrioid subtype) from preoperative specimen (P ≤ 0.02 for both) and from hysterectomy specimen (P ≤ 0.04 for both). For K^trans and v_e, no difference was observed between histological risk groups or subtypes (Suppl. Table 1).

When comparing tumor DCE-MRI parameters to immunohistochemical hormone receptor status, v_e was significantly higher in tumors with low ERα and low PR expression (P = 0.001 and P = 0.03, respectively; Suppl. Table 1). The remaining DCE-MRI parameters were not significantly related to hormone receptor status.

Tumor microstructure (ADC value from DWI)

The mean ADC values were significantly lower in tumors with deep myometrial invasion (P < 0.001; Suppl. Table 1), while no significant differences in ADC values were observed related to cervical stroma invasion, lymph node metastasis, or tumors with high-risk histological features (i.e. high grade, non-endometrioid subtype, aneuploidy, ERα, and PR loss) (Suppl. Table 1).

Tumor volume from structural MRI

Median tumor volume assessed by structural MRI was 8.2 mL (range = 0.1–443.8 mL). Larger tumor volume was significantly associated with deep myometrial invasion, lymph node metastases, aneuploidy, loss of ERα expression, and high-risk histology, both from preoperative specimen and hysterectomy specimen (P ≤ 0.01 for all, Suppl. Table 1).

Correlation between tumor DCE-MRI parameters, tumor ADC value, and tumor volume

No significant correlations were observed between the tumor DCE-MRI parameters and tumor ADC value, whereas F_b and v_e were negatively correlated to tumor volume (r = –0.17 and −0.16, respectively; P = 0.03 for both) (Table 3). F_b correlated significantly with K^trans, k_ep and v_e (r = 0.36, 0.24, and 0.20, respectively; P ≤ 0.008 for all), as did K^trans with k_ep and v_e (r = 0.49 and 0.48, respectively; P < 0.001 for both). Tumor ADC value was negatively correlated to tumor volume (r = –0.46, P < 0.001).

Table 3.

Correlation between the DCE-MRI tumor parameters, tumor ADC, and tumor volume (n = 177).

	F_b (min⁻¹)		K^trans (min⁻¹)		k_ep (min⁻¹)		v_e (%)		ADC value (10⁻⁶ mm²/s)*		Tumor volume (mL)
	r^†	P ^‡	r^†	P ^‡	r^†	P ^‡	r^†	P ^‡	r^†	P ^‡	r^†	P ^‡
F_b (min⁻¹)	1		0.36	<0.001	0.24	0.001	0.20	0.008	−0.06	0.43	−0.17	0.03
K^trans (min⁻¹)			1		0.49	<0.001	0.48	<0.001	−0.08	0.29	−0.14	0.07
k_ep (min⁻¹)					1		−0.41	<0.001	−0.12	0.12	0.02	0.76
v_e (%)							1		0.03	0.69	−0.16	0.03
ADC value (10⁻⁶ mm²/s)*									1		−0.46	<0.001
Tumor volume (mL)											1

Significant correlations are given in italics.

n = 175.

†

Spearman’s correlation coefficient.

‡

Spearman’s bivariate correlation test.

F_b, tumor blood flow; k_ep, intravasation constant from extravascular extracellular space (EES) to blood; K^trans, transfer constant from blood to EES; v_e, fractional volume of EES.

Imaging parameters and survival

Patients with the lowest values for F_b and k_ep tended to have reduced recurrence-/progression-free survival (P = 0.07 [Fig. 2a] and P = 0.09 [Fig. 2b], respectively), whereas tumor K^trans and v_e did not have an impact on survival. Patients with low tumor ADC value and large tumor volume had significantly reduced recurrence-/progression-free survival (P = 0.05, Fig. 2c; and P < 0.001, Fig. 2d, respectively). When including the imaging parameters F_b (25% lowest vs. 75% highest values), ADC (50% lowest vs. 50% highest values), and tumor volume (25% highest vs. 75% lowest volumes) in a multivariate analysis, only large tumor volume independently predicted unfavorable outcome (hazard ratio [HR] of 2.8; P = 0.002) whereas low F_b and low ADC only tended to the same (HRs of 1.7; P = 0.13 and 1.7; P = 0.16, respectively). When including k_ep (25% lowest vs. 75% highest values), ADC, and tumor volume in the multivariate analysis, low k_ep independently predicted dismal outcome (HR of 2.1; P = 0.03) as did large tumor volume (HR of 3.2; P = 0.001), whereas low ADC value only tended to the same (HR of 1.7; P = 0.12).

Fig. 2.

Kaplan–Meier curves depicting recurrence-/progression-free survival according to: (a) tumor blood flow (F_b), (b) tumor intravasation rate constant from extravascular extracellular space to blood (k_ep), (c) tumor ADC, and (d) tumor volume. Low tumor F_b and k_ep tended to be associated with reduces survival (a, b), whereas low tumor ADC (c) and large tumor volumes (d) were significantly associated with reduced recurrence-/progression-free survival.

Discussion

In this study of a large endometrial cancer patient cohort, we have shown that preoperative functional DCE-MRI and DWI yield imaging parameters which could be markers of aggressive disease. These functional imaging techniques may thus represent valuable supplements to conventional pelvic MRI, enabling improved preoperative risk stratification facilitating better tailored surgical and adjuvant treatment in endometrial cancer patients.

To our knowledge, this is to date the largest published endometrial cancer series (n = 177), in which preoperative imaging parameters have been derived both from DCE-MRI, DWI and conventional MRI. The DCE-MRI analyses have been conducted in a semi-automated approach using a commercially available software tool (Nordic-NeuroLab AS, Bergen, Norway). Previous studies including DCE-MRI in endometrial cancer comprise smaller patient series of 54–80 patients (15,16,20,21) and employ either less advanced pharmacokinetic models or in-house implementations for calculation of the tumor DCE-MRI parameters (29 –31). However, similar to our findings that low tumor blood flow (F_b) is associated with high tumor grade, non-endometrioid subtype, and adverse outcome, a previous study (employing a different pharmacokinetic model) found that low F_b is a marker of non-endometrioid subtype and associated with poor prognosis (15,16). Also in line with this, low maximum enhancement (based on semi-quantitative modeling of the time-signal intensity curves) has been reported to be more frequent in high-grade endometrial cancers (20,21), supporting the notion that reduced blood flow which is putatively linked to tissue hypoxia, is associated with tumor progression and aggressive disease.

The endometrial tumor values for the DCE-MRI parameters F_b, K^trans, and v_e were all significantly lower than for normal myometrial tissue (Table 2). The present study confirms previous findings that endometrial cancers are hypo-vascular relative to the surrounding highly vascularized myometrium (16,20,21), although the absolute values of the DCE-MRI parameters are unequal to previously reported values due to the different pharmacokinetic models employed. Interestingly, we found that both tumor F_b and k_ep, were significantly lower in tumors with high-risk histology, non-endometrioid subtype or endometrioid grade 3 (Suppl. Table 1), suggesting that both parameters represent potential imaging biomarkers characterizing the endometrial cancer phenotype. Furthermore, low F_b and low k_ep both tended to be associated with reduced survival, and when adjusting for the prognostic imaging markers low ADC value and large tumor volume, low k_ep had an independent negative prognostic impact while low F_b tended to the same.

Loss of ERα expression was both associated with high v_e (volume of EES) and with large tumor volume (from conventional MRI). Higher v_e was also associated with loss of PR receptor status (Suppl. Table 1). The higher v_e is possibly explained by microscopic tumor necrosis which is reportedly more common in ERα and PR receptor negative endometrial cancers (32).

The semi-automated pharmacokinetic modeling of the DCE-MRI (27) and the population-based arterial input function used in this study are advantageous in terms of time consumption (time of analysis: approximately 5 min/patient). The utilization of commercially available software may also render better standardization of the DCE-MRI analyses. Between centers comparisons of tumor DCE-MRI parameters are currently precluded by variation in image acquisition techniques and in the pharmacokinetic models employed. Thus, better harmonization of imaging protocols and standardization of subsequent pharmacokinetic modeling between centers is highly warranted. Furthermore, the derived tumor parameters also need to be validated for reproducibility before potential routine implementation as imaging biomarkers in the clinic.

This study has some limitations. First, it was not possible to exactly co-localize the histological sections and the MRI tumor planes for ROI analyses, implying that the tumor tissue examined for the histologic and immunohistochemical parameters does not necessarily overlap with the tumor volume studied for functional imaging. Care was, however, taken to place the ROIs in what appeared to be a representative part of the tumor, in terms of the tumors vascular profile. Second, the analyzed ROIs did not comprise the entire tumor volume as the ROIs were drawn on a single slice avoiding both the periphery of the tumor as well as necrotic areas of the tumor. This was done to avoid including normal myometrial tissue (due to temporal changes in uterine position caused by arterial pulsation and bowel peristalsis) and necrotic tumor tissue. Third, one of the principal challenges in all pharmacokinetic DCE-MRI analysis is a reliable definition of the arterial input function, and the use of a population-based arterial input function may have influenced our results if the cardiac output or the vascular resistance differs greatly between patients (14). However, population-based arterial input function was chosen due to variable quality of the arterial vessel signals in the scanned DCE-MRI volumes.

In conclusion, endometrial carcinomas with high-risk histologic subtype and poor prognosis are characterized by low tumor blood flow (F_b) and low rate constant for contrast agent intravasation (k_ep) when assessed preoperatively by DCE-MRI. Low tumor ADC value and large tumor volume are also bad prognostic factors. DCE-MRI and DWI may represent valuable supplements to conventional pelvic MRI and provide novel imaging biomarkers enabling improved preoperative risk stratification in endometrial cancer patients.

Footnotes

Acknowledgments

The authors thank the Department of Radiology, the Department of Gynecology and Obstetrics, and the Department of Pathology (all at Haukeland University Hospital). They also thank Jenny A Husby, Inger J Magnussen, Anna Berg, Kathrine Woie, Ingunn Stefansson, Helga B Salvesen, Eva Øksnes, Ellen Valen, and Britt Edvardsen for invaluable radiological/clinical/pathological input and/or laboratory work.

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 The Western Norway Regional Health Authority, Bergen Research Foundation, Research Funds at the Department of Radiology (Haukeland University Hospital), Norwegian Research Council, The University of Bergen, The Meltzer Foundation and The Norwegian Cancer Society (The Harald Andersen’s legacy).

References

Amant

Moerman

Neven

et al.

Endometrial cancer. Lancet 2005; 366: 491–505.

Salvesen

Haldorsen

Trovik

. Markers for individualised therapy in endometrial carcinoma. Lancet Oncol 2012; 13: e353–e361.

Haldorsen

Husby

Werner

HMJ

et al.

Standard 1.5-T MRI of endometrial carcinomas: modest agreement between radiologists. Eur Radiol 2012; 22: 1601–1611.

Eltabbakh

Shamonki

Mount

. Surgical stage, final grade, and survival of women with endometrial carcinoma whose preoperative endometrial biopsy shows well-differentiated tumors. Gynecol Oncol 2005; 99: 309–312.

Frumovitz

Singh

Meyer

et al.

Predictors of final histology in patients with endometrial cancer. Gynecol Oncol 2004; 95: 463–468.

Leitao

Jr. Kehoe

Barakat

et al.

Accuracy of preoperative endometrial sampling diagnosis of FIGO grade 1 endometrial adenocarcinoma. Gynecol Oncol 2008; 111: 244–248.

Haldorsen

Salvesen

. Staging of endometrial carcinomas with MRI using traditional and novel MRI techniques. Clin Radiol 2012; 67: 2–12.

Kinkel

Kaji

et al.

Radiologic staging in patients with endometrial cancer: a meta-analysis. Radiology 1999; 212: 711–718.

Haldorsen

Salvesen

. What is the best preoperative imaging for endometrial cancer? Curr Oncol Rep 2016; 18: 1–11.

10.

Todo

Choi

H-J

Kang

et al.

Clinical significance of tumor volume in endometrial cancer: a Japan–Korea cooperative study. Gynecol Oncol 2013; 131: 294–298.

11.

Ytre-Hauge

Husby

Magnussen

et al.

Preoperative tumor size at MRI predicts deep myometrial invasion, lymph node metastases, and patient outcome in endometrial carcinomas. Int J Gynecol Cancer 2015; 25: 459–466.

12.

Whittaker

Coady

Culver

et al.

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

13.

Husby

Salvesen

Magnussen

et al.

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

14.

Leach

Morgan

Tofts

et al.

Imaging vascular function for early stage clinical trials using dynamic contrast-enhanced magnetic resonance imaging. Eur Radiol 2012; 22: 1451–1464.

15.

Haldorsen

Stefansson

Gruner

et al.

Increased microvascular proliferation is negatively correlated to tumor blood flow and is associated with unfavourable outcome in endometrial carcinomas. Br J Cancer 2014; 110: 107–114.

16.

Haldorsen

Grüner

Husby

et al.

Dynamic contrast-enhanced MRI in endometrial carcinoma identifies patients at increased risk of recurrence. Eur Radiol 2013; 23: 2916–2925.

17.

Lin

K-K

Chang

C-J

et al.

Myometrial invasion in endometrial cancer: diagnostic accuracy of diffusion-weighted 3.0-T MR Imaging—Initial Experience. Radiology 2009; 250: 784–792.

18.

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.

19.

Nakamura

Imafuku

Nishida

et al.

Measurement of the minimum apparent diffusion coefficient (ADCmin) of the primary tumor and CA125 are predictive of disease recurrence for patients with endometrial cancer. Gynecol Oncol 2012; 124: 335–339.

20.

Ippolito

Cadonici

Bonaffini

et al.

Semiquantitative perfusion combined with diffusion-weighted MR imaging in pre-operative evaluation of endometrial carcinoma: results in a group of 57 patients. Magn Reson Imaging 2014; 32: 464–472.

21.

Ippolito

Minutolo

Cadonici

et al.

Endometrial cancer: diagnostic value of quantitative measurements of microvascular changes with DCE-MR imaging. MAGMA 2014; 27: 531–538.

22.

Khalifa

Soliman

El-Baz

et al.

Models and methods for analyzing DCE-MRI: a review. Med Phys 2014; 41: 124301–124301.

23.

Pecorelli

. Revised FIGO staging for carcinoma of the vulva, cervix, and endometrium. Int J Gynaecol Obstet 2009; 105: 103–104.

24.

Silverberg SG, Kurman RJ, Nogales F. Tumors of the uterine corpus. In: Tumors of the Breast and Female Genital Organs. World Health Organization Classification of Tumors. Pathology and Genetics. Lyon: IACR Press Inc., 2003.

25.

Engelsen

Stefansson

Akslen

et al.

GATA3 expression in estrogen receptor α-negative endometrial carcinomas identifies aggressive tumors with high proliferation and poor patient survival. Am J Obstet Gynecol 2008; 199: 543–543.

26.

Trovik

Wik

Werner

HMJ

et al.

Hormone receptor loss in endometrial carcinoma curettage predicts lymph node metastasis and poor outcome in prospective multicentre trial. Eur J Cancer 2013; 49: 3431–3441.

27.

Tofts

. Modeling tracer kinetics in dynamic Gd-DTPA MR imaging. J Magn Reson Imaging 1997; 7: 91–101.

28.

Tofts

Brix

Buckley

et al.

Estimating kinetic parameters from dynamic contrast-enhanced t1-weighted MRI of a diffusable tracer: standardized quantities and symbols. J Magn Reson Imaging 1999; 10: 223–232.

29.

St Lawrence

Lee

. An adiabatic approximation to the tissue homogeneity model for water exchange in the brain: II. Experimental validation. J Cereb Blood Flow Metab 1998; 18: 1378–1385.

30.

Taxt

Jirik

Rygh

et al.

Single-channel blind estimation of arterial input function and tissue impulse response in DCE-MRI. IEEE trans Biomed Eng 2012; 59: 1012–1021.

31.

Thng

Koh

Collins

et al.

Perfusion magnetic resonance imaging of the liver. World J Gastroenterol 2010; 16: 1598–1609.

32.

Bredholt

Mannelqvist

Stefansson

et al.

Tumor necrosis is an important hallmark of aggressive endometrial cancer and associates with hypoxia, angiogenesis and inflammation responses. Oncotarget 2015; 6: 39676–39691.

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

0.55 MB

0.57 MB