Gray-scale and color Doppler ultrasound characteristics of endometrial cancer in relation to stage, grade and tumor size




To describe the gray-scale and vascular characteristics of endometrial cancer in relation to stage, grade and size using two-dimensional (2D)/three-dimensional (3D) transvaginal ultrasound.


This was a prospective multicenter study including 144 women with endometrial cancer undergoing transvaginal ultrasound before surgery. The sonographic characteristics assessed were echogenicity, endometrial/myometrial border, fibroids, vascular pattern, color score and tumor/uterus anteroposterior (AP) ratio. Histological assessment of tumor stage, grade, type and growth pattern was performed.


Hyperechoic or isoechoic tumors were more often seen in Stage IA cancer, whereas mixed or hypoechoic tumors were more often found in cancers of Stage IB or greater (P = 0.003). Hyperechogenicity was more common in Grade 1–2 tumors (i.e. well or moderately differentiated) (P = 0.02) and in tumors with a tumor/uterine AP ratio of < 50% (P = 0.002), whereas a non-hyperechoic appearance was more commonly found in Grade 3 tumors (i.e. poorly differentiated) and in tumors with a tumor/uterine AP ratio of ≥ 50%. Multiple global vessels were more often seen in tumors of Stage IB or greater than in Stage IA tumors (P = 0.02), in Grade 3 tumors than in Grade 1 and 2 tumors (P = 0.02) and in tumors with a tumor/uterine AP ratio of ≥ 50% (P < 0.001). A moderate/high color score was significantly more common in tumors of higher stage (P = 0.03) and larger size (P = 0.001).


The sonographic appearance of endometrial cancer is significantly associated with tumor stage, grade and size. More advanced tumors often have a mixed/hypoechoic echogenicity, a higher color score and multiple globally entering vessels, whereas less advanced tumors are more often hyperechoic and have no or a low color score. Copyright © 2011 ISUOG. Published by John Wiley & Sons, Ltd.


Endometrial cancer is the most common malignancy of the female genital tract1. Prognosis depends on age of the patient, histological type and grade of tumor, tumor size, depth of myometrial invasion, cervical stromal invasion and lymph node metastases2. Tumor size and accurate assessment of the degree of myometrial and cervical invasion cannot be revealed by clinical examination. Preoperative knowledge of all these parameters is advantageous when planning treatment. The clinical challenge is the optimal selection of patients for more radical treatment with pelvic and para-aortic lymphadenectomy when there is a higher risk of advanced disease and relapse, whilst avoiding overtreatment in low-risk cases. This is important as endometrial cancer predominantly occurs in postmenopausal women with serious comorbidities.

Pretreatment evaluation of women with endometrial cancer using modern imaging methods is becoming more common as a result of the benefits of such techniques in tailoring treatment. However, there is little consensus to date on the optimal imaging routine in the preoperative assessment of endometrial carcinoma, and practice varies amongst many gynecologists. The accuracy of transvaginal ultrasound varies between studies3–7, and some show results comparable with that of magnetic resonance imaging (MRI)8, 9. Three-dimensional (3D) ultrasound gives new possibilities for assessing endometrial vascular morphology because a hologram of the entire vascularization can be displayed and viewed from any angle. However, few papers are available describing the sonomorphological features of endometrial cancer in relation to tumor stage, grade and size.

We hypothesized that morphological and vascular characteristics might correlate with tumor stage, grade and size, knowledge that could be of value in assessment of the preoperative tumor and perhaps be used in risk evaluation in the future. The aim of the study was to describe the sonographic features of endometrial cancer using standardized criteria, and relate the findings to tumor stage, grade and relative size.


This study was a four-center collaboration between the Ultrasound Unit at the Department of Obstetrics and Gynecology, Lund University Hospital, Lund, Sweden, the Gynecologic Oncology Unit, Catholic University of Sacred Heart, Rome, Italy, the Department of Obstetrics and Gynecology, Leuven University Hospital, Leuven, Belgium and the Department of Obstetrics and Gynecology of Ziekenhuis Oost-Limburg, Genk, Belgium. Ethical approval was obtained from the Ethical Committee of Lund University (LU-412-07) and from the local Ethical Committees of the Catholic University of Sacred Heart (Rome, Italy), Leuven University (Leuven, Belgium) and Ziekenhuis Oost-Limburg (Genk, Belgium). From January 2007 until May 2009, consecutive women with histologically confirmed endometrial cancer, planned for surgery, and who were undergoing transvaginal ultrasound examination within a week before surgery, were included in the study. All women underwent total abdominal or robotic hysterectomy, both with bilateral salpingoophorectomy with systematic pelvic and para-aortic lymphadenectomy, if appropriate, according to local protocols. The International Federation of Gynecology and Obstetrics (FIGO 2009) staging criteria were used based on the surgical specimen obtained at hysterectomy10. All ultrasound examinations were performed by the principal investigators at each center (E.E. in Lund, A.C.T. in Rome and C.V.H. in Leuven/Genk). The sonomorphological and vascular pattern assessment criteria were defined at a consensus reading session, in which all examiners were included, before the start of the study. During the study, the sonomorphological assessment was made only by the senior ultrasound examiner (i.e. E.E., A.T. or C.V.H.). The ultrasound examinations were performed within 7 days before surgery.

The ultrasound equipment used in Lund was a Voluson E8 ultrasound system (GE Healthcare, Pollards Wood, Chalfont St. Giles, UK) with a RIC5-9 transducer, or a IU22 ultrasound system (Philips Healthcare, Andover, MA, USA) with a 3D9-3v transducer, and the GE Voluson E8 with a RIC5-9 transducer was employed in Rome and in Leuven/Genk. In all women we collected still images with all measurements, videoclips with and without power Doppler and 3D volumes with and without power Doppler. For each patient a predefined examination protocol form that included information on the tumor/uterine anteroposterior (AP) ratio, echogenicity, the endometrial myometrial border (regular/irregular), the presence and number of fibroids and vascularization, was filled out by the examining doctor (E.E., A.T. or C.V.H.) directly after the examination. In the presence of submucous fibroids, we did not include the fibroids in the endometrial thickness or tumor measurements, unless the fibroid was completely intracavitary.

The women were examined in the lithotomy position with an empty bladder. The uterus was scanned in the sagittal plane from cornu to cornu and in the (oblique) transverse plane from the cervix to the fundus. Having established an overview of the whole uterus, the image was magnified to contain only the uterine corpus. We defined the tumor mass in the sagittal plane, and we measured the tumor/uterine AP ratio at the point where we found the deepest myometrial invasion. We chose to use the tumor/uterine AP ratio to describe tumor size, instead of endometrial thickness or the tumor/uterine ratio of the 3D volume, because we found that this was the best objective parameter related to tumor size to predict deep myometrial invasion according to receiver–operating characteristics (ROC) curves (the area under the curve (AUC) tumor/uterine ratio, endometrial thickness and tumor/uterine ratio of the 3D volume were 0.79, 0.74 and 0.77, respectively). The tumor/uterine AP ratio was classified into < 50% or ≥ 50%, as we found that the best cut-off was around 50%.

The tumor echogenicity was assessed at the time of real-time two-dimensional (2D) examination and was defined as hyperechoic (Figure 1a), isoechoic (Figure 1b) or hypoechoic when the majority of the tumor mass showed this appearance by ultrasound in comparison with the echogenicity of the surrounding myometrial tissue. When different echogenic patterns were observed, the tumor echogenicity was classified as mixed (Figure 1c).

Figure 1.

(a) Hyperechoic endometrial cancer, Stage IA, Grade 1. (b) Isoechoic endometrial cancer, Stage IA, Grade 2. (c) Endometrial cancer with mixed echogenicity, carcinosarcoma, Stage IB, Grade 3.

Next, we assessed the endometrial/myometrial border as regular or irregular11. In addition to the sonomorphologic characteristics, the presence of uterine fibroid was also described in the protocol.

The amount of tumor vascularization was subjectively classified, during real-time 2D ultrasound examination, using a ‘color score’: absent (score = 1), minimal (score = 2), moderate (score = 3) or high (score = 4), as introduced by the International Ovarian Tumor Analysis (IOTA) group12 for the assessment of ovarian mass vascularity.

Vascular morphology was assessed using saved 3D volumes by scrolling through the parallel sections of the 3D volume and by looking at the vascular tree in the rendered volume, using 4D View software (GE) and QLAB software (Philips). For the description of endometrial vascular morphology we used the terminology suggested by the International Endometrial Trial Analysis (IETA) group11: no vessels, single vessel (Figure 2a), multiple focal vessels (Figure 2b), multiple global vessels (Figure 2c) and scattered vessels. Figure 3 shows sectional planes and a rendered 3D volume.

Figure 2.

Classification of endometrial tumoral vascularization as having: single branching vessel (a), multiple focal vessels (b) or multiple global vessels (c).

Figure 3.

Three-dimensional multiplanar imaging, with rendered view, showing vascular tree with multiple global vessels in a case of Grade 3 endometrial cancer.

In all centers a dedicated pathologist, with substantial experience in gynecologic oncology, assessed the pathological specimens and filled out a predetermined protocol. The histopathological variables assessed were histological type, grade of differentiation, pathological stage, tumor growth pattern (tumors mainly endocavitary-protruding or mainly infiltrating the myometrium, or a combination of both growth patterns) and the presence of fibroids. In the presence of fibroids the pathologists tried to determine the normal myometrial thickness exclusive of the fibroid or in a location where no fibroid was found and related the myometrial invasion to the estimated normal myometrial thickness. Only epithelial malignant tumors such as pure endometrioid carcinoma, serous/papillary serous carcinoma, clear cell carcinoma and carcinosarcoma (i.e. ‘malignant Müllerian tumor’) were included. Endometrioid and mucinous adenocarcinoma were classified into three grades (Grade 1 = well differentiated, Grade 2 = moderately differentiated and Grade 3 = poorly differentiated). Serous/seropapillary cancer, clear cell cancer and carcinosarcoma were classified as Grade 3. High-risk endometrial cancer was defined as Grade 3 endometrioid cancer or other histotypes (see above), myometrial invasion > 50% or cervical stroma invasion (Figure 4)2, 13.

Figure 4.

High-risk endometrial cancer characterized by deep myometrial invasion and/or cervical invasion, and/or Grade 3 or non-endometrioid histology. Sonographic characteristics include high color score, multiple global vessels and mixed or hypoechoic echogenicity.

The chi-square test or Fisher's exact test was used for categorical data and the Mantel–Haenszel chi-square test was used for variables with several ordered categories. All calculations were performed using Statistical Package for the Social Sciences (SPSS) software, version 10.0.1 (SPSS Inc., Chicago, IL, USA) and P < 0.05 was considered significant.


Lund contributed 56 patients, Leuven/Genk contributed 46 and Rome contributed 42. The patient demographics are shown in Table 1. The histological characteristics are presented in Table 2. Forty-nine per cent of the tumors were Stage IA and 51% were Stage IB or greater; 38% (27/71) of the Stage IA tumors were confined to the endometrium. Eighty-three per cent (120/144) of the tumors were pure endometrioid adenocarcinomas, 68% (98/144) were Grade 1 or Grade 2 and 32% (46/144) were Grade 3 (Table 2). A mainly infiltrating growth pattern was seen in 33% (24/73) of tumors of Stage 1B or greater, but this growth pattern was seen in only 3% (3/71) of Stage 1A tumors (P = 0.02). Multiple global vessels at ultrasound examination were seen in 63% (17/27) of tumors with infiltrating growth pattern at histology, compared with 44% (50/115) of protruding tumors (P = 0.07).

Table 1. Demographic background data of women with endometrial cancer (n = 144)
  • Data are given as mean ± SD or n (%).

  • *

    Cervical, ovarian or endometrial cancer.

  • Oral/dermal estradiol in combination with sequential or continuous progesterones.

  • Oral/vaginal estriol or vaginal estradiol. BMI, body mass index.

Age (years)66 ± 10.2
BMI (kg/m2)28.7 ± 6.5
Postmenopausal132 (92)
Family history of gynecological cancer*20 (14)
Family history of breast cancer23 (16)
Family history of breast cancer and/or gynecological cancer40 (28)
Current high- or medium-potency hormone use15 (10)
Current low-potency estrogen use52 (36)
Current tamoxifen use4 (3)
Table 2. Tumor histological characteristics
Histologyn (%)
  • *

    Among women with a Stage 1A tumor, the tumor was confined to the endometrium (i.e. in situ growth only) in 27 (38%).

Stage (FIGO 2009)10 
 IA*71 (49)
 IB34 (24)
 II12 (8)
 IIIA4 (3)
 IIIB3 (2)
 IIIC19 (6)
 IIIC24 (3)
 IVA0 (0)
 IVB7 (5)
 Endometrioid adenocarcinoma, Grade 156 (39)
 Endometrioid adenocarcinoma, Grade 242 (29)
 Endometrioid adenocarcinoma, Grade 322 (15)
 Serous/seropapillary14 (10)
 Carcinosarcoma6 (4)
 Clear cell4 (3)
Growth pattern 
 No tumor left/tumor in situ5 (4)
 Mainly exophytic/intracavitary growth103 (72)
 Mainly infiltrating myometrium27 (19)
 Both infiltrating and protruding7 (5)
 No information on growth pattern2 (1)

The sonographic characteristics of endometrial cancer in relation to stage are shown in Table 3. An irregular endometrial/myometrial border was more often seen in tumors of Stage 1B or greater than in Stage 1A tumors (P = 0.005). Among the women with Stage 1A tumors with a regular endometrial/myometrial border at ultrasound examination, 68% (15/21) had tumors limited only to the endometrium (with no myometrial invasion). Hyperechoic or isoechoic echogenicity were more common in Stage 1A tumors, whereas hypoechoic or mixed echogenicity were more common in tumors of Stage 1B or greater (P = 0.003, odds ratio (OR) = 3.4, 95% CI: 1.4–7.9). Multiple global vessels, as opposed to all other vessel patterns, and moderate or high color score, as opposed to no or low color score, were more common in tumors of Stage 1B or greater than in Stage 1A tumors (P = 0.02, OR = 2.2, 95% CI: 1.4–7.9; and P = 0.03, OR = 2.2, 95% CI: 1.1–4.5, respectively) (Table 3).

Table 3. Sonographic characteristics in relation to final stage based on hysterectomy specimen
Ultrasound findingsStage 1A (n = 71)Stage 1B or greater (n = 73)P
  • Data are given as n (%).

  • *

    Chi-square test.

  • Mantel–Haenszel chi-square test.

Endometrial/myometrial border  0.005*
 Regular21 (30)8 (11) 
 Irregular50 (70)65 (89) 
Gray-scale morphology  0.003
 Hyperechoic46 (65)35 (48) 
 Isoechoic17 (24)11 (15) 
 Hypoechoic4 (6)8 (11) 
 Mixed4 (6)19 (26) 
Vessel pattern  0.059
 No vessels21 (30)8 (11) 
 Single3 (4)8 (11) 
 Multiple focal13 (18)12 (16) 
 Multiple global28 (39)42 (58) 
 Scattered6 (9)3 (4) 
Color score  0.006
 No21 (30)9 (12) 
 Low14 (20)13 (18) 
 Moderate22 (31)27 (37) 
 High14 (20)24 (33) 
Fibroids  0.400*
 No48 (68)54 (74) 
 Yes23 (32)19 (26) 

Table 4 shows the sonographic characteristics in relation to tumor grade. The echogenicity differed significantly between Grade 1 or Grade 2 tumors vs. Grade 3 tumors (P = 0.02). Hyperechoic echogenicity was more common in Grade 1 or Grade 2 tumors whereas other echogenicity patterns were more common in Grade 3 tumors (P = 0.02, OR = 2.2, 95% CI: 1.1–4.3). Multiple global vessels, as opposed to all other vessel patterns, were more common in Grade 3 tumors than in Grade 1 and Grade 2 tumors (P = 0.02, OR = 2.5, 95% CI: 1.2–5.1). Moderate or high color score was non-significantly higher in Grade 3 tumors than in Grade 1 and Grade 2 tumors (P = 0.059, OR = 2.1, 95% CI: 0.97–4.4) (Table 4).

Table 4. Sonographic characteristics in relation to tumor differentiation grade based on hysterectomy specimen
 Histological grade
Ultrasound findingsGrade 1: well differentiated (n = 56)Grade 2: moderately differentiated (n = 42)Grade 3: poorly differentiated (n = 46)P
  • Data are given as n (%). All statistical calculations were made for well differentiated and moderately differentiated vs. poorly differentiated.

  • *

    Chi-square test.

  • Mantel–Haenszel chi-square test.

Endometrial/myometrial border   0.057*
 Regular21 (38)3 (7)5 (11) 
 Irregular35 (63)39 (93)41 (89) 
Gray-scale morphology   0.020
 Hyperechoic38 (68)25 (60)20 (44) 
 Isoechoic10 (18)9 (21)9 (20) 
 Hypoechoic1 (2)1 (2)5 (11) 
 Mixed7 (13)7 (17)12 (26) 
Vessel pattern   0.003
 No vessels21 (38)5 (12)3 (7) 
 Single5 (9)4 (10)2 (4) 
 Multiple focal9 (16)7 (17)9 (20) 
 Multiple global16 (29)25 (59)29 (63) 
 Scattered5 (9)1 (2)3 (7) 
Color score   0.020
 No22 (39)5 (12)3 (7) 
 Low10 (18)7 (17)10 (22) 
 Moderate12 (21)19 (45)18 (39) 
 High12 (21)11 (26)15 (33) 
Fibroids   0.870*
 No37 (66)32 (76)33 (72) 
 Yes19 (34)10 (24)13 (28) 

Table 5 shows the sonographic characteristics in relation to tumor/uterine AP ratio, as assessed by ultrasound. Fibroids (P = 0.047), a regular endometrial/myometrial border (P < 0.001) and hyperechoic echogenicity were significantly more common in tumors with a tumor/uterine AP ratio of < 50% of the uterus (P = 0.002, OR = 2.9, 95% CI: 1.5–5.9). Multiple global vessels, as opposed to all other vessel patterns, and moderate or high color score, as opposed to no or low color score, were significantly more common in tumors with a tumor/uterine AP ratio of ≥ 50% (P < 0.001, OR = 4.8, 95% CI: 2.4–9.8; and P < 0.001, OR = 5.5, 95% CI: 2.6–11.7, respectively) (Table 5).

Table 5. Tumor/uterine anteroposterior (AP) ratio, as measured by ultrasound in relation to sonomorphology
 Tumor/uterine AP ratio
Ultrasound findings< 50% (n = 74)≥ 50% (n = 70)P
  • *

    Chi-square test.

  • Mantel–Haenszel chi-square test.

Endometrial/myometrial border  < 0.001*
 Regular24 (32)5 (7) 
 Irregular50 (68)65 (92) 
Gray-scale morphology  0.007
 Hyperechoic52 (70)31 (44) 
 Isoechoic13 (18)15 (21) 
 Hypoechoic2 (3)5 (7) 
 Mixed7 (10)19 (27) 
Vessel pattern  < 0.001
 No vessels25 (34)4 (6) 
 Single8 (11)3 (4) 
 Multiple focal13 (18)12 (17) 
 Multiple global23 (31)47 (67) 
 Scattered5 (7)4 (6) 
Color score  < 0.001
 No26 (35)4 (6) 
 Low17 (23)10 (15) 
 Moderate18 (24)31 (46) 
 High13 (18)25 (36) 
Fibroids  0.047*
 No47 (64)55 (79) 
 Yes27 (36)15 (21) 


In this study we describe the sonographic appearance of endometrial cancer. We have shown that gray-scale and vascular morphology characteristics of endometrial cancer are significantly related to tumor stage, grade and size. Tumors of Stage IB or greater, of Grade 3 or with a tumor/uterine AP ratio of ≥ 50% are more often hypoechoic or have a mixed echogenicity, have a higher color score and more often display multiple globally entering vessels, whereas tumors of Stage IA, of Grade 1 or Grade 2, or with a tumor/uterine AP ratio of < 50% are more often hyperechoic, and have more frequently low or no vascularization. These findings indicate that low-risk endometrial cancer (Figures 1a and 2a) has a different sonomorphological appearance from that of high-risk endometrial cancer (Figures 1c and 2c).

The strengths of the study were the size of the study population, the prospective design and the standardized ultrasound examination with predefined evaluation criteria. All examinations were performed by experienced ultrasound examiners, using high-end ultrasound equipment. However, it remains to be shown if and how our findings can be used in the clinical management of women with endometrial cancer.

It has previously been described that an irregular endometrial–myometrial border is a feature of endometrial carcinoma14–17. In our study, 20% (29/144) of the endometrial malignancies were found to have a regular endometrial/myometrial border; however, it was uncommon in tumors of Stage 1B or greater (11%, 8/73) or those filling ≥ 50% (7%, 5/70) of the uterus. Tumor growth limited only to the endometrium can explain the presence of a regular border in 52% (15/29) of the cases.

Hyperechoic endometrium is a common feature of endometrial hyperplasia18. We found that hyperechoic endometrium in endometrial cancer patients was related to smaller size of tumor and to well-differentiated tumors invading < 50% of the myometrium. In a previous study we found that it was difficult to differentiate between benign and malignant endometrial lesions in women with postmenopausal bleeding if the endometrium measured < 15 mm14. This can be explained by our findings that small endometrial cancer tumors often resemble endometrial hyperplasia, by their hyperechogenicity and regular endometrial–myometrial border.

The vascular pattern obtained using 2D ultrasound16, 19 or the color content obtained using either 2D14 or 3D ultrasound20 can be used for the prediction of endometrial cancer in women with postmenopausal bleeding. In addition, it has previously been shown that the high color density21 or high color score22 in endometrial cancer patients is related to the stage21 or to the presence of nodal metastasis22. To our knowledge this is the first study describing the vascular morphology patterns in a large series of endometrial tumors using predefined criteria11. Multiple global vessels were correlated to high-risk endometrial cancer, whereas non-vascularized tumors were mainly seen in low-risk endometrial cancers. There was also a non-significant correlation between the histological growth pattern and the vascular pattern at ultrasound examination: indeed, with an infiltrating growth pattern (rather than endocavitary protruding pattern) the tumor more commonly displays a multiple global vessel pattern (P = 0.07).

We found no association between tumor stage, grade and the presence of fibroids, but we observed that fibroids were more common in women with a relative tumor size of < 50% of the uterus. De Smet and co-workers showed that the number of fibroids was a useful variable in a mathematical model, together with grade, endometrial thickness and volume, for the prediction of myometrial invasion23. They speculated whether fibroids could restrict the deep growth of the tumor, limiting the risk of deep invasion.

Although endometrial cancer is often (75%) confined to the uterine body, data show that the mortality rate for endometrial cancer has increased rapidly over the last decades, which may be related to an increased rate of advanced-stage cancer and high-risk histologies24. To improve on outcome for patients, physicians need to identify high-risk patients and tailor treatment appropriately to provide the best long-term survival. Currently, the first step in the management of endometrial cancer is to assess the risk for metastatic disease according to a preoperative histopathological diagnosis and preoperative evaluation of the tumor extent, to select patients for appropriate surgical treatment. However, there are known limitations of misclassification of the preoperative biopsy specimen as well as of imaging methods to assess myometrial invasion and cervical stroma involvement. Up to 20% of tumors may have a worse histologic grade and occasionally a different tumor type based on analysis of the hysterectomy specimen2, 25. Similarly, the accuracy of myometrial invasion, assessed by ultrasound or MRI scans, is only 80–85%9, 26. Therefore, there is a need for additional preoperative parameters to identify women at high risk and with a poor prognosis.

To our knowledge, this is the first paper describing the sonomorphological gray-scale and power Doppler ultrasound characteristics of endometrial cancer in a large series of women using both 2D and 3D ultrasound. We have shown a correlation between the gray-scale and vascular morphology features of endometrial cancer and the stage, grade and size of the tumors. However, it remains to be determined in future studies whether sonomorphological characteristics can be used in risk-calculation models to predict low-risk and high-risk endometrial cancer and thus be used to tailor treatment. The reproducibility of the morphology assessment should also be evaluated, and the most optimal way of assessing myometrial invasion investigated.