| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Genitourinary Imaging |
1 From the Departments of Obstetrics and Gynecology (M.R.O.) and Clinical Physiology and Nuclear Medicine (J.S.J.), Kuopio University Hospital, Puijonlaaksontie 2, 70210 Kuopio, Finland; and Department of Obstetrics and Gynecology, Tampere University Hospital, Tampere, Finland (P.P.K.). Received August 29, 2001; revision requested October 17; final revision received May 20, 2002; accepted June 18. Supported by the Finnish Cultural Foundation and the Research Foundations of Orion Corporation and Instrumentarium. Address correspondence to M.R.O. (e-mail: maija-riitta.orden@kuh.fi).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Seventy patients who were suspected of having ovarian tumors were examined preoperatively with contrast material–enhanced US. Images obtained during a 5-minute examination were stored digitally, and the behavior of the contrast agent was evaluated objectively with measurement of the time-dependent image intensity at the region of interest with a computer program. A time-intensity curve in each case was derived and analyzed. The Mann-Whitney U test was used to compare intensity changes and tumor parameters in benign and malignant adnexal tumors.
RESULTS: Both the baseline and maximum power Doppler intensities, as well as the absolute and relative (percent) rise in intensity, were significantly higher (P < .001) in malignant as compared with benign tumors. The arrival time was shorter (17.5 vs 22.5 seconds; P = .005) and the duration of contrast agent effect was longer (190.4 vs 103.6 seconds; P < .001) in malignant tumors than they were in benign tumors. The area under the time-intensity curve was significantly greater in malignant tumors compared with that in benign tumors (P < .001).
CONCLUSION: After microbubble contrast agent injection, malignant and benign adnexal lesions behave differently in degree, onset, and duration of Doppler US enhancement.
© RSNA, 2003
Index terms: Ovary, neoplasms, 852.31, 852.32 • Ovary, US, 852.12988 • Ultrasound (US), contrast media, 852.12988 • Ultrasound (US), Doppler studies, 852.12988
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Intravascular US contrast agents enhance imaging of tumor vessels by providing a stronger Doppler signal. They have been used in studies of breast (4–6), liver (7–9), and prostate (10) lesions, and some investigators have experience with their use in gynecologic and obstetric US (11–14).
The purpose of our study was to evaluate the effects of a microbubble contrast agent on the power Doppler US examination of patients with adnexal tumors, with a special focus on the timing of the transit of the microbubble bolus.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
All except seven patients underwent surgery, and the final diagnosis was based on findings at histologic examination of the specimens removed. One 67-year-old and another 73-year-old patient had apparent ovarian cancer with bilateral large adnexal tumors. They did not undergo surgery because of their poor general condition at the time it was planned. The diagnosis in these cases was based on clinical, US, and computed tomographic findings and the further course of the disease. Five women with apparent benign adnexal tumors did not undergo surgery: In one, this was because of poor general condition; in three, the patients had functional cysts that disappeared or became smaller at repeat US examinations; and in one, the patient had an infectious process of the adnexal region from which she fully recovered. Clinical follow-up evaluation in these five patients revealed no signs of malignancy. The median interval between the contrast material–enhanced US and surgery was 1.5 days (range, 0–119 days). Thus, in 14 patients, the diagnosis was ovarian cancer; in four patients, the diagnosis was ovarian tumor with borderline malignant histologic findings; and in 52 patients, the diagnosis was benign adnexal tumor (Table 1).
|
At two-dimensional real-time US, each mass was classified as follows: a unilocular cyst (a single cyst without septa and without solid parts or papillary excrescences), a multilocular cyst (a cyst with at least one septum but no solid parts or papillary excrescences), a unilocular solid cyst (a single cyst containing solid parts or papillary excrescences but no septa), a multilocular solid cyst (a cyst with at least one septum and solid parts or papillary excrescences), or a solid tumor (a tumor with solid components in 80% or more of the tumor) (15). The maximum diameter and volume (in cubic centimeters) of the tumor were measured. Twenty-three women had bilateral adnexal masses. No woman had a benign tumor on one side and a malignant tumor on the other. All tumors were examined thoroughly with power Doppler US. However, each woman contributed only one tumor to the study. The tumor with the greatest number of vessels was chosen. A vessel was defined as a separate colored area typical of a blood vessel.
The US contrast agent (Levovist; Schering, Berlin, Germany) consisted of galactose microparticles and a small (0.1%) admixture of palmitic acid. When the contrast agent is mixed with water, it provides microbubbles of air covered by a thin stabilizing layer of palmitic acid. The contrast agent (dose, 8.5 mL) was administered manually at a concentration of 300 mg/mL through an antebrachial vein with an 18-gauge cannula at an injection speed of 1 mL/sec and was flushed with 5 mL of saline. A US scanner (Sequoia 512; Acuson, Mountain View, Calif) with transvaginal (EC-10C5; Acuson) and transabdominal (6C2; Acuson) transducers were used.
Tumor vascularization was visualized by means of the power Doppler technique, and blood flow velocity waveforms were obtained from several different sites of the tumor. The final Doppler examination was registered from that part of the tumor in which the lowest pulsatility index (PI) and resistive index (RI) values were obtained. Correction of the insonation angle was performed in every pulsed Doppler waveform obtained, and the angle was kept at less than 60°. The lowest PI and RI and the highest peak systolic flow and time-averaged maximum velocity values obtained were registered.
Power Doppler US was performed first without contrast agent enhancement and then during and after administration of the contrast agent in exactly the same part of the tumor. Identical power Doppler settings (ie, sensitivity, penetration, gain, filters) were used before and after administration of the contrast agent. Sensitivity was set to the optimal level to visualize small vessels but to avoid noise artifacts. The US equipment was programmed to store the whole examination of 5 minutes as 3-second clips in the hard disk of the US device. With 10 patients, the contrast-enhanced examination was performed twice in about 15 minutes to find out the reproducibility of the contrast agent effect. All US examinations were performed by one of the authors (M.R.O.). The contrast agent was well tolerated, and no harmful effects were noticed.
The effect of the contrast agent was evaluated with visual observation and measurement of the power Doppler intensity with a computerized data program (Data Pro, version 2.1; Noesis, Courtboeuf, France). The digital clips were stored on magneto-optical disks and exported to a personal computer to process the data with the computerized data program. The region of interest (ROI) included one-sixth to one-fourth of the total cross-sectional area of the tumor. The ROI was placed on an area of the tumor that was richest in solid tissue and vasculature. With visual evaluation, the number of blood vessels (ie, separate colored areas typical of blood vessels) in the ROI was calculated before and after administration of the contrast agent.
A new vessel (ie, a recognizable vascular area) in our study, therefore, could be either a new vessel that was previously not imaged or a branch of a vessel already recognized before administration of the contrast agent. The computerized data program analysis produced a number of active points, which means a number of pixels with nonzero values. The total intensity, namely, the sum of the pixel values of the ROI, was measured just before contrast agent administration and at every 5 seconds after the administration up to 120 seconds and then at every 10 seconds up to 300 seconds (ie, 5 minutes). Each measurement of 3 seconds with an average heart rate was the mean of three to four heart cycles. The number of heart cycles could be calculated, with the computerized data program showing the intensity curve of each 3-second clip. The data were transferred to a computerized spreadsheet program (Excel 97; Microsoft, Redmond, Wash), and a time-intensity curve was derived in each patient.
The time-intensity curves were analyzed for the following indices: the baseline intensity (in decibels); the arrival time (in seconds), defined as the first point of the curve clearly above the baseline intensity, followed by a further rise; the time to peak (in seconds), defined as the time from the start of the injection to the maximum intensity of the curve; the rising time (in seconds), defined as the time from the start of the rising curve to its maximum; maximum intensity (in decibels); absolute (in decibels) and relative (percentage, from change in decibels) rise in intensity; and the rise rate (in decibels per second), defined as the increase of the intensity divided by the rising time. The time-intensity curves were also analyzed according to the method described by Schwarz et al (16).
The analysis was based on the finding that the washout phase is biphasic, that is, a fast linear phase followed by a slow linear phase. The fast and slow portions of the washout phase could be identified in time-intensity curves. By using linear fitting procedures for time-intensity (measured in decibels) curves, the rates of fast (slope 1, measured in decibels per second) and slow (slope 2, measured in decibels per second) washout phases were derived by using the computerized spreadsheet program. The time required for the Doppler intensity to return to the precontrast value, or duration of contrast agent effect (DCE 1 and 2), and area under the time-intensity curve (area 1 and 2) were calculated based on the 
Ymax and slope values (Fig 1). In the reproducibility study (n = 10), the coefficient of variation (CV) to measured indices was calculated as suggested by Glüer et al (17). Standardized CV (SCV) also was calculated as follows: SCV = CV/(4 SD/mean) of the whole study group (n = 70) (18).
|
2 test was used to investigate the association between benign and malignant tumors and patient or tumor characteristics (categorical variables). P < .05 was considered to indicate a statistically significant difference. Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of most sensitive indices to show the risk of malignancy were calculated. |
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
If the DCE 1 index limit of greater than 48 seconds was used to determine the risk of malignancy, 13 of 14 (sensitivity, 93%) malignant lesions would be identified as having a long DCE 1. Forty-eight of 52 (specificity, 92%; accuracy, 92%) benign lesions had a DCE 1 shorter than 48 seconds. With a DCE 2 index greater than 150 seconds, sensitivity was 86% (12 of 14), specificity was 77% (40 of 52), and accuracy was 79% (52 of 66) for the diagnosis of ovarian malignancy. Both areas 1 and 2 were significantly larger in malignant tumors than they were in benign tumors. With a cutoff limit of 25 dB x sec in area 1, sensitivity was 100% (14 of 14), specificity was 83% (43 of 52), and accuracy was 86% (57 of 66) for the diagnosis of ovarian malignancy. With area 2 and a cutoff limit of 90 dB x sec, sensitivity was 79% (11 of 14), specificity was 85% (44 of 52), and accuracy was 83% (55 of 66) for malignancy (Table 5).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Findings in a meta-analysis (19) revealed that optimal ovarian lesion characterization with conventional methods can be obtained through the combination of gray-scale US morphology and color Doppler flow imaging information. In this study, we used a US contrast agent to enhance the visualization of tumor vasculature and to evaluate its suitability for the differentiation of benign from malignant adnexal tumors. To the authors’ knowledge, this study is the first examination of transit time of a US contrast agent in imaging of benign and malignant ovarian tumors, as well as the first study in which digital analysis of power Doppler intensity was used in adnexal tumors with the used new program.
The addition of a vascular contrast agent, such as the US contrast agent used in this study, allows a more complete delineation of the vascular anatomy through enhancement of the signal strength from small vessels and provides an entirely new opportunity to time the transit of an injected bolus. Doppler US examinations of breast tumors have revealed that carcinomas show a higher number of vessels and color pixel densities than do benign lesions (20). Breast tumor studies with contrast agents also indicate that contrast reaches a peak more rapidly in carcinomas than it does in benign lesions and that the signal intensity enhancement is greater and longer in the malignancies (5,6). In our study, the brightness of the power Doppler signal and the amount of recognizable vascular areas increased in each tumor after contrast agent administration. The number of recognizable vessels was significantly higher in malignant lesions than it was in benign lesions both before and after addition of the contrast agent.
The shape of the time-intensity curves appeared to be as reported (16,21) previously with the US contrast agent used in this study. After a quick rise in intensity, the washout phase was biphasic, with first a fast linear decrease (distribution phase) and then followed by a slow linear decrease (elimination phase). Both the baseline and maximum power Doppler intensities, and also the absolute and relative rise in intensity, were significantly higher in malignant tumors than they were in benign tumors. The arrival time of the contrast agent was significantly shorter in malignant tumors than it was in benign tumors. This could be caused by the high-velocity flow through the arteriovenous shunts that are typically found in malignant neovascularization. The initial arrival of contrast agent induces an excessive broadening of color areas, an effect known as blooming. This phenomenon is likely to be more pronounced in carcinomas with excessive vasculature and, thus, contributes to the earlier arrival time.
In our study, in addition to intensity changes, the DCE and the area under the time-intensity curve appeared to be the best discriminating factors between benign and malignant tumors. Mean DCE 2 in malignant tumors was 190.4 seconds, and in benign tumors it was 103.6 seconds (P < .001). The longer persistence of enhancement in malignant lesions could be explained by the pooling of the contrast agent in dilated and blind-ending vessels. Once in the bloodstream, the US contrast agent used in this study has an average duration of blood pool enhancement of 2–5 minutes (22), although a later liver-specific phase has been described (23). The stabilized air microbubbles gradually dissolve in plasma, and each passage of a capillary bed accelerates this activity (24). Therefore, bypass of capillary beds through shunts will reduce the rate of bubble destruction and, thus, result in higher bubble concentrations in diverging vasculature of malignant tumors.
As Table 2 shows, the PI and RI values were significantly lower in malignant adnexal tumors than they were in benign adnexal tumors. However, findings in numerous studies (25–27) have indicated that the overlapping in these indices reduces their value for diagnostic purposes. Similar overlapping could be seen in indices measured from our time-intensity curves. It may, thus, be that the diagnostic value of one index is limited, whereas the overall shape of the curve appeared to be reproducible. On the basis of the findings of our study, we can say that the examination is more reliable with tumors of prominent vasculature, because in these cases the possible accidental movement of the probe or tumor does not seriously disturb the analysis.
Many of our measured indices showed high variation in measured values, a variation that was also demonstrated in high CV values. SCV values were much lower and suggest that the contrast agent effect may be recorded in such a reproducible way that the variations in tumoral vasculature may be discerned. Obviously, the vasculature under the ROI in duplicate examinations was not exactly the same, and this led to variation of absolute values of indices, although the qualitative shape of the time-intensity curves was more constant.
Development of color Doppler techniques and improved sensitivity of modern US devices have made it possible to image smaller vessels even without contrast agents. However, the dynamic features noted here and in previous studies of the breast and the liver (4–6,24) are important to echo-enhancement studies because they measure the transit of a bolus of microbubbles through the vasculature. Benign and malignant ovarian tumors behave differently at US in both degree and dynamics of enhancement after the injection of a contrast agent. This phenomenon could be exploited in the treatment of tumors through the use of microbubbles as drug delivery vehicles and as a means of increasing capillary permeability, which improves local access of the released therapeutic agent, such as antiangiogenetic factors in cancer treatment (28–31).
In conclusion, our study presents an objective digital analysis of kinetics of a US contrast agent in imaging benign and malignant adnexal tumors, with a special focus on the timing of the transit of the microbubble bolus. Our study findings indicate that the signal enhancement is quicker, stronger, and longer in malignant tumors than it is in benign tumors. On the basis of our duplicate measurements, reproducibility of the technique appears to be reasonable, especially when related to biologic variation of the recorded parameters. However, a larger study is needed to confirm this finding.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, M.R.O., P.P.K.; study concepts and design, M.R.O., P.P.K.; literature research, M.R.O.; clinical studies, M.R.O.; data acquisition, M.R.O.; data analysis/interpretation, M.R.O., J.S.J.; statistical analysis, M.R.O., J.S.J.; manuscript preparation, M.R.O.; manuscript editing, J.S.J., P.P.K.; manuscript definition of intellectual content, revision/review, and final version approval, M.R.O., J.S.J., P.P.K.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. C. Fleischer, A. Lyshchik, H. W. Jones Jr, M. Crispens, M. Loveless, R. F. Andreotti, P. K. Williams, and D. A. Fishman Contrast-Enhanced Transvaginal Sonography of Benign Versus Malignant Ovarian Masses: Preliminary Findings J. Ultrasound Med., July 1, 2008; 27(7): 1011 - 1018. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Barua, J. S. Abramowicz, J. M. Bahr, P. Bitterman, A. Dirks, K. A. Holub, E. Sheiner, M. J. Bradaric, S. L. Edassery, and J. L. Luborsky Detection of Ovarian Tumors in Chicken by Sonography: A Step Toward Early Diagnosis in Humans? J. Ultrasound Med., July 1, 2007; 26(7): 909 - 919. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. T. Rissanen, P. Korpisalo, J. E. Markkanen, T. Liimatainen, M.-R. Orden, I. Kholova, A. de Goede, T. Heikura, O. H. Grohn, and S. Yla-Herttuala Blood Flow Remodels Growing Vasculature During Vascular Endothelial Growth Factor Gene Therapy and Determines Between Capillary Arterialization and Sprouting Angiogenesis Circulation, December 20, 2005; 112(25): 3937 - 3946. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. C. Testa, G. Ferrandina, E. Fruscella, C. Van Holsbeke, E. Ferrazzi, F. P. G. Leone, D. Arduini, C. Exacoustos, D. Bokor, G. Scambia, et al. The Use of Contrasted Transvaginal Sonography in the Diagnosis of Gynecologic Diseases: A Preliminary Study J. Ultrasound Med., September 1, 2005; 24(9): 1267 - 1278. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. S. Abramowicz Ultrasonographic Contrast Media: Has the Time Come in Obstetrics and Gynecology? J. Ultrasound Med., April 1, 2005; 24(4): 517 - 531. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Marret, S. Sauget, B. Giraudeau, M. Brewer, J. Ranger-Moore, G. Body, and F. Tranquart Contrast-Enhanced Sonography Helps in Discrimination of Benign From Malignant Adnexal Masses J. Ultrasound Med., December 1, 2004; 23(12): 1629 - 1639. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. C. Fleischer, K. J. Niermann, E. F. Donnelly, T. E. Yankeelov, K. M. Canniff, D. E. Hallahan, and M. E. Rothenberg Sonographic Depiction of Microvessel Perfusion: Principles and Potential J. Ultrasound Med., November 1, 2004; 23(11): 1499 - 1506. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D Cosgrove Angiogenesis imaging - ultrasound Br. J. Radiol., December 1, 2003; 76(suppl_1): S43 - S49. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
