| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH |
From the 1Department of Obstetrics and Gynecology, Máxima Medical Centre, Veldhoven, the Netherlands; 2Department of Obstetrics and Gynecology, Academic Hospital Maastricht, Maastricht, the Netherlands; 3Department of Obstetrics and Gynecology, Radboud University Medical Centre, Nijmegen, the Netherlands; 4Department of Obstetrics and Gynecology, Tweesteden Hospital, Tilburg, the Netherlands; and 5Department of Obstetrics and Gynecology, Academic Medical Centre, Amsterdam, the Netherlands.
 |
ABSTRACT |
|---|
TOP
 ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS: Women scheduled for surgical treatment of an adnexal mass were included in a multicenter prospective study. All women underwent two-dimensional and three-dimensional ultrasonographic examination in the week before surgery. All parameters were compared in women with benign tumors, borderline tumors, and malignant tumors using receiver operating characteristic analysis and likelihood ratios.
RESULTS: We included 181 women; 144 had a benign mass, 26 had a malignancy, and 11 had a borderline tumor. At three-dimensional ultrasonography, the most striking difference was found in the presence of central vessels in an adnexal mass. Central vessels assessed by three-dimensional ultrasonography were present in 15% (21 of 144) of the benign masses, 69% (18 of 26) of the malignant masses, and 27% (3 of 11) of the masses of borderline malignancy. The likelihood ratios for presence of central vessels for a mass being malignant and/or borderline was 4.9 (95% confidence interval 2.1–12). Mean gray index and flow index were also significantly different between the groups, but other features were not.
CONCLUSION: The central localization of vessels in an adnexal mass, as observed by three-dimensional ultrasonography, the mean gray index, and the flow index are potentially important parameters for distinguishing benign from malignant adnexal masses.
LEVEL OF EVIDENCE: II-2
At present, several parameters are available for distinguishing benign and malignant masses. The gray-scale two-dimensional sonographic parameters that are used most frequently are tumor diameter or volume, septation and presence of papillary projections, echogenicity, and the presence of free fluid. Ovarian tumor blood flow can be evaluated by B-mode color Doppler ultrasonography and waveform analysis.1 The resistance to flow is lower in malignant tumors than in benign tumors. Frequently used Doppler parameters are resistance index, pulsatility index, and peak systolic velocity. To improve the preoperative assessment of adnexal masses, most of these parameters have been combined with patient characteristics in diagnostic models. Although initial publication reported an almost perfect performance of these models, external validation showed their diagnostic performance to be less than good.2,3
A potentially new diagnostic tool in the assessment of the adnexal mass is three-dimensional ultrasonography. This technique visualizes the adnexal mass in all three planes (coronal, sagittal, and frontal), thus facilitating spatial evaluation instead of two-dimensional ultrasonography, in which only two planes can be visualized. With three-dimensional ultrasonography, a volume of ultrasonographic data is acquired and stored, whereas with two-dimensional ultrasonography only slices can be assessed. The integration of the acquired information enables reconstruction of the organ being scanned. The addition of three-dimensional power Doppler imaging allows visualization and quantitative judgment of vascularity in all three planes.
Early publications on three-dimensional ultrasound examination of adnexal masses have suggested that the specificity of ultrasonographic imaging might improve with the addition of three-dimensional power Doppler.4,5 These studies reported that masses with central vascular flow, vascular flow within excrescences, or flow within septation, as well as masses including vessels with a chaotic architecture or complex branching pattern, are more likely to be malignant.
The objective of this prospective study was to estimate whether data acquired by three-dimensional ultrasonography and three-dimensional power Doppler investigation could potentially contribute to the differentiation between benign and malignant ovarian masses.
|
MATERIALS AND METHODS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In this prospective study all women who were scheduled for surgery because of an adnexal mass between January 2003 and March 2005 were eligible for the study. The study was approved by the institutional review board of the Máxima Medical Centre. All women received verbal and written information about the study. After informed consent had been obtained, women were scheduled for two-dimensional and three-dimensional gray-scale ultrasonography and for two-dimensional and three-dimensional power Doppler ultrasonography in the week before surgery. The two-dimensional and three-dimensional ultrasound examinations were performed using a 7.5-MHz transvaginal transducer (Voluson 730; Kretztechnik, Zipf, Austria). All ultrasound examinations were performed transvaginally. Power Doppler settings were kept constant for every patient: frequency 3–9 MHz, pulse repetition frequency 0.6 kHz, gain 4.0 dB, and wall motion filter "low 1".
All examinations were performed by three of the authors (P.G., K.K., E.M.). The acquired three-dimensional ultrasonographic data were stored on a hard disk to enable full evaluation at a later point. Subsequently, all recordings were reviewed systematically by one of the authors (P.G.) without knowledge of the final histopathologic diagnosis.
To perform volume measurements, the virtual organ computer-aided analysis (VOCAL; Kretz, Marl, Germany) was used.6 Volume measurements were performed by manual delineation of the outside of the adnexal mass using the 9º-rotation step (Fig. 1). In practice the data set was rotated clockwise through 180º, and the outside of the mass was delineated every 9º until a calculated volume was generated. The volume was stored and analyzed with specially designed software (3D View Function; GE Medical Systems, Kretz Ultrasound, Austria).7 Analogous to two-dimensional pixels, the stored volume ultrasound information obtained using three-dimensional sonography is defined by voxels (smallest measurable volume). Gray-scale voxels contain all three-dimensional gray-scale information grades from white to black. Voxels with a power Doppler signal are defined as color voxels. The range of color voxels can vary between 0 and 100. As described by Pairleitner et al,7 the "histogram" feature of 3D View Function calculates three vascularization indices: vascularization index, flow index, and vascularization-flow index. The vascularization index is the ratio of the number of color voxels to the number of all voxels in the defined contour and is expressed as a percentage. The flow-index (scale 0–100) is the mean value of the color voxels and is thought to express the average intensity of flow in the vessels. The vascularization-flow index measures the ratio of the mean value of the color voxels and all the voxels in the defined contour ([vascularization index x flow index]/100) and is a feature of both vascularization and flow. It identifies the extremes between low vascularization and low blood flow, on one side, and high vascularization and high blood flow, on the other. With the "histogram" feature of 3D View Function, the mean value of the gray-scale voxels can also be calculated and is expressed as the mean gray index (scale 0–100). The mean gray index is a quantification of the echogenicity of an adnexal mass. When the mean gray index is close to zero the adnexal mass is mainly sonolucent, whereas a mean gray index close to 100 refers to a high echogenic mass.
|
For each patient, the following three-dimensional gray-scale ultrasonographic and three-dimensional power Doppler data were recorded: tumor volume, mean gray index, localization of vessels, vascularization index, flow index, and vascularization-flow index. The localization of vessels was ordered into four categories (Fig. 2). The first category ("no vessels") contained masses without any sign of vascularization, neither peripheral (in the wall of the mass) nor central (within the mass). The second category ("peripheral vessels") contained cystic masses with vascularization of the wall of the mass and no vascularization within the mass. The third group ("central vessels") contained masses with central vascularization, but no vascularization of the wall of the masses. The central vascularization could be within septa, solid areas, or papillary formations. The fourth group ("peripheral and central vessels") contained masses that had both central vascularization and peripheral vascularization.
|
If it was impossible to include the adnexal mass into one volume box because of its size, the volume of the tumor could not be measured. A representative part of the adnexal mass was assessed to define the localization of the vessels and to calculate the vascularization indices.
For each patient, the following two-dimensional gray-scale ultrasonographic data were recorded: presence and thickness of septations, presence of papillary formations and free fluid, echogenicity and the mean diameter of the mass, and the following power Doppler parameters: resistance index, pulsatility index, and peak systolic velocity.
In addition to the ultrasonographic parameters, we recorded serum CA 125 level, menopausal status, and age. The CA 125 was measured with chemiluminescent immunometric assay (Immulite 2000 OM-MA; Diagnostic Products, Los Angeles, CA).
In case of suspicion of malignancy, surgery was performed by a median laparotomy and combined with frozen section diagnosis. In case frozen section diagnosis showed malignancy a complete debulking was pursued. The risk for malignancy was assessed by means of the results of a clinical examination, two-dimensional gray-scale ultrasonographic information (septation, presence of papillary formations, echogenicity, and presence of free fluid), two-dimensional power Doppler parameters (resistance index), age of the patient, and level of serum CA 125. If the mentioned parameters did not lead to any suspicion of malignancy, a cystectomy was performed if the woman was younger than 45 years and an ovariectomy was performed if the patient had was older than 45 years. All adnexal masses were sent to the laboratory of the Department of Pathology and examined by one of the experienced pathologists. A definite paraffin diagnosis was obtained.
The distribution of each of the two-dimensional and three-dimensional parameters was compared among the three categories. Distribution of continuous variables was calculated by constructing normal plots.8 Depending on normality, we calculated means or medians. In case of a normal distribution, statistical comparison was made with analysis of variance. In cases of nonnormal distributions, we used the Kruskal-Wallis test.
Subsequently, for each of the continuous variables we performed a receiver-operating-characteristic analysis, in which malignant and borderline adnexal masses were considered to be the disease, and benign masses to be non-disease. We calculated an area under the receiver-operating-characteristic curve (AUC).9 The AUC can have, in general, a value between 0.5 and 1; an AUC of 0.5 expresses no discriminatory capacity at all, whereas an AUC of 1 expresses perfect discriminatory capacity. Categorical variables were compared using 
2 analysis and further assessed by calculating likelihood ratios. A likelihood ratio of a test result was the ratio of the probability of the test result in women with malignant disease (borderline included) and the probability of the test result in women with benign disease. A likelihood ratio above 1 increases the probability of malignancy whereas a likelihood ratio below 1 decreases the probability of disease. In general, a likelihood ratio between 1 and 2 is considered as not useful, and a likelihood ratio between 2 and 5 as moderately useful. The test becomes useful at higher likelihood ratios. These values, however, should be considered together with the prevalence of disease and subsequent therapeutic consequences.
|
RESULTS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Because of initial suspicion of malignancy, surgery was performed by a median laparotomy and combined with frozen section diagnosis for 29 of the 144 women with benign pathology. Forty-five percent of the women with borderline tumors (5 of 11) were operated upon laparoscopically, and all but one of the women with malignant tumors had laparotomies.
Baseline characteristics are shown in Table 1. The mean age was 46 years (standard deviation [SD] 16.0, range 15–85 years) for women with benign masses, 61 years (SD 12.8, range 42–87 years) for women with malignant masses, and 51 years (SD 15.0, range 38–83 years) for women with a borderline malignancy (P
.001). Among the 144 women with benign masses, 54 were postmenopausal (38%), compared with 5 of 11 (45%) among women with borderline malignancy and 18 of 26 (69%) among women with malignancy (P=.005).
|
The median level of serum CA 125 in women with malignant masses was 155 kilounits/L (range 2.1–13,000), whereas the median level of serum CA 125 was 11 kilounits/L (range 1.4–450) and 29 kilounits/L (range 5.1–146) for benign masses and borderline masses, respectively (P<.001). Twenty-five of 140 benign masses (18%) showed a CA 125 level of 35 kilounits/L or greater, and 10 of 140 showed a CA 125 level of 100 kilounits/L or greater. Among these 10 benign masses with a serum CA 125 level of 100 kilounits/L or greater were seven endometriomas, two fibromas, and one mucinous cystadenoma. Twelve of 37 malignant and borderline tumors (32%) showed a serum CA 125 level less than 35 kilounits/L.
Tables 2 and 3 show the distribution of the two-dimensional and three-dimensional parameters with their 95% confidence intervals (CIs). Ten adnexal masses (six malignant tumors, one borderline tumor, and three benign tumors) could not be included in one volume box because of their sizes. Consequently their volumes could not be calculated. In eight of these 10 cases, it was possible to assess a representative part of the tumor to define the localization of the vessels and to calculate the vascularization indices.
|
|
The malignant masses also had a higher flow index compared with the benign and borderline masses (P=.04, area under the receiver-operating-characteristic curve 0.65). The likelihood ratio for malignancy increased with an increasing flow index, suggesting that a mass with a high flow index (greater than 30) is more at risk for malignancy (Table 3). The mean vascularization index and vascularization flow index did not differ significantly among the three categories, indicating that these tests do not have value.
Table 3 also shows the localization of vessels. Central vessels were present in 69% (18 of 26) of the malignant masses, 27% (3 of 11) of the masses of borderline malignancy, and 15% (21 of 144) of the benign masses. The likelihood ratios for absence of vessels or for presence of peripheral vessels only were 0.42 (95% CI 0.13–1.3) and 0.54 (95% CI 0.35–0.86), respectively. When central vessels were present, the risk of a malignancy increased, as indicated by likelihood ratios of 3.3 (95% CI 1.3–6.8) and 4.9 (95% CI 2.1–12) for central vessels only and central and peripheral vessels in the same mass. Table 4 shows which benign masses had central vascularization.
|
Among the 181 included tumors, 75 masses were sonolucent without septation or sonolucent with septa smaller than 3 mm. If we exclude these so-called simple cysts, the presence of central vessels and a flow index greater than 30 showed likelihood ratios of 4.4 (95% CI 1.8–9.9) and 1.8 (95% CI 1.2–2.7), respectively.
The presence of random echogenicity (two-dimensional) and central vessels (three-dimensional) were the most sensitive predictors because they were present in 22 and 21, respectively, of the 37 women with borderline malignancy or invasive malignancy. Sensitivity and specificity of random echogenicity were 59% (95% CI 42–75) and 80% (95% CI 73–86), and sensitivity and specificity of presence of central vessels were 57% (95% CI 40–73) and 85% (95% CI 80–91). Presence of septa and their thickness, presence of papillary formations, the mean diameter and the volume of the mass, pulsatility index, peak systolic velocity, vascularization index, and vascularization-flow index were not predictive. Premenopausal status, absence of free fluid, an anechoic performance of the mass, and the presence of only peripheral (and not central) vessels reduced the risk for malignancy.
|
DISCUSSION |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
By means of three-dimensional power Doppler ultrasonography, it is possible to visualize vessels in all three planes (coronal, sagittal, and frontal), thus facilitating spatial evaluation. Assessing the localization of vessels by three-dimensional power Doppler ultrasonography is a simple procedure.
Since the introduction of three-dimensional transvaginal ultrasonography, several studies have been published assessing the potential contribution of three-dimensional sonography in the differentiation between benign and malignant ovarian masses (Table 5).4,5,12,14–16 The authors use morphologic criteria5,14–16 or the presence of central or chaotic vessels4,5,12 to define malignancy. Most of the studies show improvement of the specificity of ultrasonographic imaging by addition of three-dimensional ultrasonography.
|
In this prospective study, 181 adnexal masses were evaluated, and we assessed the localization of vessels by means of three-dimensional power Doppler ultrasonography. Central vessels were seen more often in malignancies (69%) than in benign masses (15%) and masses with borderline malignancy (27%). The presence of central and peripheral vessels in an adnexal mass had a likelihood ratio of 4.9 (95% CI 2.1–11). Our results are in accordance with the results of Alcazar and Castillo.12 The criteria for suspicion of malignancy used in the their study are comparable with our criteria. We suspected malignancy if blood flow was central anywhere in an adnexal mass and also if blood flow was detected in septations. However, Alcazar and Castillo12 assessed localization of blood flow by two-dimensional and three-dimensional power Doppler examination, and three-dimensional power Doppler imaging did not have a better performance than two-dimensional power Doppler imaging. We did not assess blood flow localization by two-dimensional power Doppler examination, so we cannot confirm these results.
Malignant neoplasms are highly vascularized. The neovascularization is of poor quality, with low compliance and low resistance. Because of the low resistance of the vasculature, the flow in malignant masses will be increased. In this study, we found an increased flow index in the malignant masses (32.6%) compared with the borderline masses (30.8%) and benign masses (28.5%). The flow index represents the intensity of flow in an adnexal mass. However, the low resistance of a flow index of more than 30% is close to one, indicating that it has no value in the differentiation between benign and malignant masses.
The mean gray index was significantly higher in the malignant adnexal masses than in benign and borderline masses. Adnexal masses that contain papillary formations and masses that are mixed or highly echogenic are suspicious for malignancy. These features result in a high mean gray index. However, benign teratomas or hemorrhagic masses also often show mixed echogenicity. This probably explains the low likelihood ratio of the mean gray index.
In this study we estimated whether data acquired by three-dimensional ultrasonography and three-dimensional power Doppler investigation could potentially contribute to the differentiation between benign and malignant ovarian masses. The central localization of vessels in an adnexal mass, the mean gray index, and the flow index, as observed by three-dimensional ultrasonography, are potentially important parameters for distinguishing benign from malignant adnexal masses. As stated before, assessing the localization of vessels by three-dimensional imaging is a simple procedure. Another advantage of three-dimensional ultrasonography compared with two-dimensional ultrasonography is the objective aspect of the mean gray index and flow index. The mean gray index and flow index are objective quantitative parameters based on the number of gray-scale and color voxels, whereas in two-dimensional ultrasonography the echogenicity of an adnexal mass and the amount of flow are always subjective descriptions of the examiner.
Further research steps to be undertaken are analyses in which the additional value of three-dimensional parameters is estimated. A larger study population with a larger percentage of malignant masses would be helpful.
|
Footnotes |
|---|
doi:10.1097/01.AOG.0000240138.24546.37
|
REFERENCES |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2. Timmerman D, Verrelst H, Bourne TH, De Moor B, Collins WP, Vergote I, et al. Artificial neural network models for the preoperative discrimination between malignant and benign adnexal masses. Ultrasound Obstet Gynecol 1999;13:17–25.[Medline]
3. Mol BW, Boll D, De Kanter M, Heintz AP, Sijmons EA, Oei SG, et al. Distinguishing the benign and malignant adnexal mass: an external validation of prognostic models. Gynecol Oncol 2001;80:162–7.[Medline]
4. Cohen LS, Escobar PF, Scharm C, Glimco B, Fishman DA. Three-dimensional power Doppler ultrasound improves the diagnostic accuracy for ovarian cancer prediction. Gynecol Oncol 2001;82: 40–8.[Medline]
5. Kurjak A, Kupesic S, Sparac V, Kosuta D. Three-dimensional ultrasonographic and power Doppler characterization of ovarian lesions. Ultrasound Obstet Gynecol 2000;16:365–71.[Medline]
6. Raine-Fenning NJ, Campbell BK, Clewes JS, Kendall NR, Johnson IR. The reliability of virtual organ computer-aided analysis (VOCAL) for the semiquantification of ovarian, endometrial and subendometrial perfusion. Ultrasound Obstet Gynecol 2003;22:633–9.[Medline]
7. Pairleitner H, Steiner H, Hasenoehrl G, Staudach A. Three-dimensional power Doppler sonography: imaging and quantifying blood flow and vascularization. Ultrasound Obstet Gynecol 1999;14:139–43.[Medline]
8. Altman DG. Preparing to analyze data. In: Practical statistics for medical research. 1st ed. New York (NY): Chapman and Hall; 1991. p. 133–4.
9. Altman DG. Some common problems in medical research. In: Practical statistics for medical research. 1st ed. New York (NY): Chapman and Hall; 1991. p. 417–8.
10. Folkman J. The role of angiogenesis in tumor growth. Semin Cancer Biol 1992;3:65–71.[Medline]
11. Tailor A, Jurkovic D, Bourne TH, Collins WP, Campbell S. Sonographic prediction of malignancy in adnexal masses using multivariate logistic regression analysis. Ultrasound Obstet Gynecol 1997;10:41–7.[Medline]
12. Alcazar JL, Castillo G. Comparison of 2-dimensional and 3-dimensional power-Doppler imaging in complex adnexal masses for the prediction of ovarian cancer. Am J Obstet Gynecol 2005;192:807–12.[Medline]
13. Guerriero S, Ajossa S, Risalvato A, Lai MP, Mais V, Angiolucci M, et al. Diagnosis of adnexal malignancies by using color Doppler energy imaging as a secondary test in persistent masses. Ultrasound Obstet Gynecol 1998;11:277–82.[Medline]
14. Bonilla-Musoles F, Raga F, Osborne NG. Three-dimensional ultrasound evaluation of ovarian masses. Gynecol Oncol 1995;59:129–35.[Medline]
15. Hata T, Yanagihara T, Hayashi K, Yamashiro C, Ohnishi Y, Akiyama M, et al. Three-dimensional ultrasonographic evaluation of ovarian tumours: a preliminary study. Human Reprod 1999;14:858–61.
16. Alcazar JL, Galan MJ, Garcia-Manero M, Guerriero S. Three-dimensional sonographic morphologic assessment in complex adnexal masses: preliminary experience. J Ultrasound Med 2003;22:249–54.
This article has been cited by other articles:
|
|
 |
|
  M. J. Kudla, I. E. Timor-Tritsch, J. M. Hope, A. Monteagudo, D. Popiolek, S. Monda, C. J. Lee, and A. A. Arslan Spherical Tissue Sampling in 3-Dimensional Power Doppler Angiography: A New Approach for Evaluation of Ovarian Tumors J. Ultrasound Med., March 1, 2008; 27(3): 425 - 433. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. Alcazar, D. Rodriguez, P. Royo, R. Galvan, S. Ajossa, and S. Guerriero Intraobserver and Interobserver Reproducibility of 3-Dimensional Power Doppler Vascular Indices in Assessment of Solid and Cystic-Solid Adnexal Masses J. Ultrasound Med., January 1, 2008; 27(1): 1 - 6. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
