Preoperative evaluation of endometrial carcinoma by contrast-enhanced ultrasonography

Authors


Dr J Yang, Department of Obstetrics and Gynecology, Peking Union Medical College Hospital, Beijing 100730, China. Email yangjiaxin007@hotmail.com

Abstract

Objective  To investigate the potential usefulness of contrast-enhanced ultrasonography in the preoperative evaluation of endometrial carcinoma and observe its enhancement pattern and time–intensity curve.

Design  A prospective study.

Setting  Gynaecological department of a college hospital.

Population  A total of 35 women diagnosed with endometrial carcinoma.

Methods  The patients were evaluated with real-time grey-scale contrast-enhanced ultrasonography. In 21 women, the parameters of time–intensity curve were compared between the endometrial lesion and normal myometrium. Findings about the depth of myometrial invasion on sonograms were compared with histological findings.

Main outcome measures  Enhancement pattern, parameters of time-intensity curve, depth of myometrial invasion detected on sonograms.

Results  In visual evaluation, the brightness of the power Doppler signal and the amount of recognisable vascular areas increased in each tumour after contrast agent administration. Feeding vessels of the tumour were shown in 77.1% (27/35) of women. The other 22.9% (8/35) of women revealed that the signals were first visualised in the central portion of the tumour. The arrival time and peak time of endometrial lesion tended to be shorter than normal myometrium. The tumours started to enhance earlier than or at the same time as myometrium in 90.5% (19/21) of women. The peak intensity, enhancement intensity, and rising rate were higher in endometrial lesion than normal myometrium. There was no myometrial invasion in eight women, inner half myometrial invasion in 19 women and the outer half myometrial invasion in eight women. The corresponding values for ultrasound were 9, 17, and 9. The sensitivity of contrast-enhanced ultrasonography in detecting deep invasion was 75.0%, while the specificity was 88.9%, and the accuracy was 85.7%. The overall accuracy of assessment of myometrial invasion was 68.6%.

Conclusion  There is some benefit in contrast-enhanced ultrasonography of endometrial carcinoma. It may provide better information in tumour imaging. Large studies are needed to determine the appropriate use and benefit of this new procedure.

Introduction

Endometrial cancer is the most common gynaecological malignancy. About 75–80% women present with stage I disease. Staging in endometrial cancer is surgical–pathological based on the International Federation of Gynecology and Obstetrics (FIGO) system. The depth of myometrial invasion is an important factor in staging because it is related to occurrence of lymph node metastases and, indirectly to a 5-year survival expectancy. Preoperative assessment of endometrial cancer may optimise treatment planning and the surgery to be undertaken. Contrast-enhanced ultrasonography has gained acceptance in many fields, particularly in the liver.1 In gynaecology, its usefulness was reported with tissue enhancement of the placenta2 and diagnosis of ectopic pregnancy.3 During the past 10 years, contrast-enhanced ultrasonography has been widely used in the discrimination of benign from malignant adnexal masses.4,5 To date, there are no published reports on the use of contrast ultrasonography in endometrial carcinoma.

In this report, we performed contrast-enhanced ultrasonography in 35 women with endometrial carcinoma to observe its enhancement pattern and time–intensity curve, and to evaluate the clinical value of this method in preoperative assessment of myometrial invasion.

Patients and methods

Thirty-five women with early-stage endometrial carcinoma from our unit between June 2006 and December 2007 were recruited for the study. Informed consent was obtained from all women. All women underwent total hysterectomy with bilateral salpingo-oophorectomy by either laparotomy or laparoscopy, and the final diagnoses were based on histological examination of the specimens.

The sonographic contrast agent used in this study, Sono Vue (Bracco, Geneva, Switzerland), consists of sulphur hexafluoride microbubbles surrounded by a thin layer of phospholipids and palmitic acid, which allows the bubbles to withstand several passes through the pulmonary capillaries. Like all microbubbles and unlike contrast material for computed tomography and magnetic resonance imaging (MRI), Sono Vue is a blood pool agent that remains in the intravascular fluid compartment and dose not leak into the interstitium. The size of the microbubbles is less than 8 micrometres, which ensures that there is no embolisation of capillaries.6 Venous access was obtained using a 20-G cannula, before commencing the ultrasound examination. A stabilised microbubble suspension of the agent was administered intravenously at a concentration of 8 microlitres/ml (45 micrograms/ml) followed by an additional 5 ml of physiologic saline solution to flush the cannula. The typical dose was 2.4 ml; if necessary, two further injections could be administered. Real-time grey-scale contrast-enhanced ultrasound was performed on a Philips iU22 ultrasound machine, using pulse inversion harmonics. Contrast-specific imaging mode was used for postcontrast scanning. All women were scanned with a 4- to 8-MHz transvaginal transducer. All scans were performed by a single sonographer (Z.L.). All ultrasound scans were saved in the hard drive of the machine in digital imaging and communications in medicine (dicom) format. The time–intensity curves were derived in 21 women using QLAB software. We chose tumour region and normal myometrium (where there is no feeding vessels and arched vessels) with minimal artefacts as the regions of interest. The sampling frame was 5 mm2 square. The shapes and areas were identical (Figure 1). Seven parameters were recorded and calculated: arrival time (the time interval from administration of the contrast agent to its visual observation in the tumour vessels), peak time, enhancement time (peak time − arrival time), arrival intensity, peak intensity, enhancement intensity (peak intensity − arrival intensity), and rising rate (enhancement intensity/enhancement time). Values were compared between tumour region and normal myometrium. Myometrial invasion was measured from the endometrial–myometrial interface to the deepest edge of the tumour extension into the myometrium and was categorised as 0% (stage IA), <50% (stage IB), and >50% (stage IC) myometrial invasion.

Figure 1.

Time–intensity curves derived from contrast-enhanced ultrasonography of tumour and normal myometrium (red: tumour, yellow: normal myometrium).

Statistical analysis was carried out using the statistical package SPSS 10.0. The paired t test was used to compare the parameters in malignant tumour and normal myometrium. P < 0.05 were considered to be statistically significant.

Results

Contrast-enhanced characteristics

The contrast agents were well tolerated with no adverse effects. In visual evaluation, the brightness of the power Doppler signal and the amount of recognisable vascular areas increased in each tumour after contrast agent administration. The tumour base, position, and boundary were shown more clearly than traditional ultrasound. By using contrast enhancement, 77.1% (27/35) of women showed the feeding vessels of tumour (this region was enhanced earlier than normal myometrium). The contrast agent arrived at the feeding vessels firstly and then branched into the endometrial cancer. All the endometrial cancers were enhanced completely except for necrotic areas. The other 22.9% (8/35) of women revealed that the signals were first visualised in the central portion of the tumour. The enhancing signals were coarse in large tumours, while in smaller tumours the signals were punctuate. In 74.3% (26/35) of women, the tumours washed out earlier than the normal myometrium. The cancerous regions were visualised in three women after contrast enhancement but could not be seen by standard ultrasound. Images of stage IA endometrial carcinoma with bulky tumour are shown in Figure 2.

Figure 2.

Images of a stage IA endometrial carcinoma with bulky tumour. (A) Image before the injection of the bolus showed marked enlargement of endometrium. (B) Image at 14 seconds after administration showed asymmetrically enhanced tumour. (C) Image at 18 seconds showed the maximal concentration of contrast agent in the tumour. (D) Image at 27 seconds showed that tumour washed out earlier than normal myometrium, thus formed a clear peritumoral halo.

Contrast parameters

A time–intensity curve with a clear arrival time, rapid peak, and decrease with return to the baseline level was seen in 21 scanned women. Seven parameters of malignant tumour and normal myometrium are listed in Table 1. The arrival time and peak time of the endometrial cancer group tended to be shorter than normal myometrium. In 90.5% (19/21) of women, the tumours started to enhance earlier than (12 women) or at the same time as (7 women) myometrium. The enhancement time did not reach statistical significance between malignant tumour and normal myometrium. The peak intensity, enhancement intensity, and rising rate were significantly higher in malignant tumour than normal myometrium (P = 0.001). The arrival intensity did not differ between malignant tumour and normal myometrium (Table 1).

Table 1.  Comparison of time–intensity parameters for malignant tumour and normal myometrium
ParametersTumourNormal myometriumP value
Arrival time (seconds)16.13 ± 5.3417.50 ± 6.390.023
Peak time (seconds)25.10 ± 7.8827.90 ± 8.340.002
Enhancement time (seconds)8.97 ± 3.6110.40 ± 4.700.223
Arrival intensity (dB)5.64 ± 6.855.69 ± 5.790.949
Peak intensity (dB)24.77 ± 5.7020.48 ± 5.390.000
Enhancement intensity (dB)19.13 ± 9.5314.79 ± 7.960.001
Rising rate (dB/seconds)2.23 ± 1.081.60 ± 0.920.001

Myometrial invasion

All 35 women were histologically staged as FIGO stage I. In 23% (8/35) of women, the tumour was confined to the endometrium (stage IA); in 54% (19/35), the inner half of the myometrium was involved (stage IB), and in 23% (8/35), the outer half of the myometrium was involved (stage IC). The results of assessment of myometrial invasion and histological examination are shown in Table 2. Stage IC endometrial cancer was diagnosed by contrast-enhanced ultrasonography in nine women. Six women were confirmed, while three women were overestimated. Seventeen women with stage IB endometrial cancer were diagnosed by contrast-enhanced ultrasonography, 14 of them were confirmed and 3 were overestimated (all stage IA cancers). Finally, no invasion was diagnosed in nine women by contrast-enhanced ultrasonography. Within this group, histological examination revealed five women with and four without myometrial invasion. The sensitivity, specificity, positive and negative predictive values and accuracy in each stage are shown in Table 3. The overall accuracy was 68.6%.

Table 2.  Comparison of staging based on contrast-enhanced sonography and histological findings
Contrast-enhanced sonographyHistological stageTotal
Ia (n = 8)Ib (n=19)Ic (n = 8)
Ia4329
Ib314017
Ic1269
Table 3.  Sensitivity, specificity, positive predictive value, negative predictive value and accuracy for correct determination of myometrial invasion
StageSensitivitySpecificityPositive predictive valueNegative predictive valueAccuracy
Ia0.500.820.440.850.74
Ib0.740.810.820.720.77
Ic0.750.890.670.920.86

Discussion

Intravascular contrast agents increase the intensity of Doppler signals. Thus, they are expected to permit better detection of blood flow in small, deep vessels, and this in turn improves the ability to differentiate between areas of normal and abnormal perfusion.5 We first used an ultrasonographic contrast agent to improve the visualisation of tumour vasculature and to evaluate its effects in endometrial cancer. The characteristics of contrast-enhanced ultrasonography in endometrial cancer are closely correlated with the pathophysiological basis. Malignant tumours are characterised by neovascularisation. Vessels in malignant tumours are numerous and complex. They penetrate into the tumour in a radial fashion and often develop arteriovenous shunts, which result in high blood flow velocity. Muscularisation of the vessel wall is incomplete, resulting in low resistance to flow.7 Thus, the enhancement of malignant tumour is earlier and faster than normal myometrium. Using these characteristics, small superficial tumour regions can be easily identified by contrast-enhanced ultrasonography.

In 1988, FIGO adopted the surgical staging of endometrial carcinoma with adjunctive treatment based on surgicopathological findings. An important part of the staging procedure is to assess the depth of myometrial invasion. This may determine if women should have retroperitoneal lymph node dissection. This assessment can be carried out by gross visual inspection, sonography and/or MRI. Gross visual inspection has been reported to be only 65 and 31% accurate in grade 2 and grade 3 lesions, respectively,8 and 71% accurate in cases of deep (>50%) myometrial invasion.9 MRI is considered the most accurate imaging technique for preoperative assessment of endometrial cancer because of its excellent soft-tissue contrast resolution. Overall staging accuracies have been reported at 83–92%.10 MRI is time-consuming and expensive and is not widely available, especially in developing countries. The cost of MRI is nearly ten times more than standard ultrasonography and five times that of contrast-enhanced ultrasonography. Ultrasonography has proved to be helpful in evaluating endometrial carcinoma. Some authors have reported rates ranging from 66 to 98%.11–13 To our knowledge, various studies in the literature report that accuracy rates are lower if the results are subdivided within stage I (IA, IB or IC) rather than when myometrial invasion involves half or more of its thickness. Our results are consistent with these findings.

In agreement with results obtained by standard ultrasonography, the source of our diagnostic error in this study was the difficulty of distinguishing bulky intraluminal tumour with an intact but thinned myometrium from diffusely infiltrating tumour. This factor led to overestimation of depth of invasion. Minimal or microscopic invasion was the most common reason for underestimation. In this study, the accuracy rate of contrast-enhanced ultrasonography was not superior compared with standard transvaginal ultrasonography in the evaluation of myometrial invasion. The reasons and also the limitations of this study are because of the relatively small number of women and the operator’s experience. There is some benefit in contrast-enhanced ultrasonography of endometrial carcinoma. It may provide better information in tumour imaging. However, further studies are indicated to evaluate the applicability, cost-effectiveness, and reproducibility of this modality. This examination will be reserved for proven endometrial cancers preoperatively and then compared with standard sonography in the future.

Disclosure of interests

The authors declare no conflict of interest.

Contribution to authorship

Y.S. conceived and designed the study, analysed the data, wrote the paper. J.Y. conceived and designed the study and contributed the subjects. Z.L. performed the ultrasound and analysed the data. K.S. conceived and designed the study, and contributed the subjects.

Details of ethics approval

Subjects were informed of the aims, methods, the anticipated benefits and potential risks of the study and the discomfort it may entail. Informed consent was obtained from all subjects.

Funding

The study was financially supported by Peking Union Medical College Hospital Research Fund.

Acknowledgements

The authors thank Department of Imaging and Ultrasound for their valuable and encouraging help. We also thank Dr Huifang Huang and Dr Lingya Pan for their contributions of subjects.

Editor’s Commentary

The clinical role of contrast-enhanced ultrasound is rapidly expanding, adding entirely new capabilities to real-time imaging. The introduction of microbubble contrast technology has redefined the use of ultrasound in resolving vascular questions that until now have been left to computed tomography (CT) and magnetic resonance imaging (MRI).

While conventional ultrasound provides morphological information and the addition of Doppler evaluates direction and velocity of blood flow, contrast-enhanced ultrasound allows the dynamic detection of blood flow at the capillary level. This new generation of contrast agents consists of encapsulated bubbles of gas, 1–7 micrometres in diameter, are smaller than red blood cells and are purely intravascular, therefore behaving as blood-pool markers with no leakage into the extravascular space. When injected intravenously, there is enhancement of blood flow in the microvasculature (as well as the macrovasculature), a feature not obtained from traditional Doppler that is essentially limited to major blood vessels. The unique interaction between ultrasound and microbubbles therefore leads to improved resolution and lesion characterisation and offers a means of differentiating between areas of normal and abnormal perfusion. This is especially useful in malignant tumours that are characterised by neovascularisation.

The lack of radiation exposure, minimal adverse effects and known safety in renal compromise has contributed to the increasing use of contrast-enhanced sonography. Clinical applications have focused particularly in the detection and characterisation of focal liver lesions, and there are recent reports that document the potential for similar use in the kidneys, spleen, pancreas, breast and prostate. This technique, however, has not been widely explored in gynaecology with only few studies describing the use of microbubble contrast in gynaecological malignancies.

Yingna et al. report the use of Sono Vue in women with endometrial cancer. By demonstrating the feeding vessels of the tumour and employing various enhancement parameters for both normal and abnormal tissue, the overall accuracy of myometrial invasion was noted to be 68.6%. Although not as accurate as MRI, the authors highlight the potential of this modality in centres where MRI is not available. More research is urgently required in this area as ultrasound remains the fundamental imaging investigation in most gynaecological malignancies. It is probable that with further advances, this approach may in time parallel the findings of CT or MRI, with the added advantage of high spatial and temporal resolution. Future applications may also include targeted and molecular imaging, particularly in quantifying angiogenesis in cancer, with major implications for patient treatment and response.

V Harry
Department of Gynaecological Oncology, Aberdeen Royal Infirmary, Aberdeen, UK

Ancillary