Diagnosis of adnexal malignancies by using color Doppler energy imaging as a secondary test in persistent masses

Transcription

1 Ultrasound Obstet Gynecol 9;:277 2 Diagnosis of adnexal malignancies by using color Doppler energy imaging as a secondary test in persistent masses S. Guerriero, S. Ajossa, A. Risalvato, M. P. Lai, V. Mais, M. Angiolucci and G. B. Melis Department of Obstetrics and Gynecology, University of Cagliari, Cagliari, Italy Key words: TRANSVAGINAL ULTRASOUND, COLOR DOPPLER ENERGY, CA 25, CYSTIC OVARIAN DISEASES ABSTRACT The purpose of this prospective study was to compare the accuracy of B-mode transvaginal ultrasonography alone and in combination with color Doppler energy (or power Doppler) imaging in differentiating benign from malignant adnexal masses. A total of consecutive persistent adnexal masses (59 benign, 33 malignant) were studied before surgery by B-mode transvaginal ultrasonography with and without color Doppler energy. In addition, CA-25 plasma levels were determined and spectral Doppler analysis was performed. By color Doppler energy imaging, a mass was considered malignant when arterial flow was visualized in an echogenic portion of a mass defined as malignant by B-mode. Intratumoral arterial blood flow could be readily detected by color Doppler energy imaging in all malignant tumors and in 94% of the benign tumors. The combined use of transvaginal B-mode ultrasonography and color Doppler energy imaging has greater accuracy in the diagnosis of ovarian malignancies than transvaginal ultrasonography alone (value of kappa: 0. and 0.63, respectively), reducing the number of falsepositive results. The use of spectral Doppler analysis was of limited diagnostic value, with a kappa value of 0.7 for the pulsatility index (< ) and of 0.4 for the resistance index (< 0.4). Also, the association with CA-25 increased the number of false-negative results. In conclusion, the use of color Doppler energy imaging seems to be a useful secondary test when a mass is suspected to be malignant by B-mode ultrasonography. INTRODUCTION There have been conflicting reports about the proportion of benign ovarian tumors that have arterial blood flow detectable by color Doppler imaging 3. An analysis of data from several recent studies 4 0 gives a mean detection rate of 68%. More information about the vascularity of all ovarian masses may help to distinguish between those that are malignant and those that are benign. Recently, a variation of conventional color Doppler imaging using the amplitude (energy or power) of the Doppler signal has become available 2. This method has been termed color Doppler energy (CDE) or power Doppler imaging and it has received much attention because of its reported increase in flow sensitivity 2. Thus CDE imaging has the ability to detect areas of low blood flow currently not detected by color Doppler imaging 2. This method should be able to detect and characterize blood flow in all adnexal masses. In this study, we hypothesized that CDE imaging could be a useful secondary test when a mass is suspected to be malignant by B-mode ultrasonography. Our aim, therefore, was to evaluate whether CDE imaging increased the specificity of transvaginal B-mode ultrasonography alone, and reduced the number of false-positive results in differentiating benign from malignant adnexal masses. The associations of B-mode and CA-25 levels, CDE imaging and CA-25 and B-mode with the lowest resistance index (RI) and pulsatility index (PI) values were also assessed to see whether they increased the specificity of the B-mode evaluation alone. The kappa statistic was calculated and used to evaluate the extent of agreement between test results and actual findings 3. METHODS This study was reviewed and approved by the ethical committee of the Department of Obstetrics and Gynecology of the University of Cagliari. The subjects for this study were 240 consecutive non-pregnant women under observation between January 9 and May 7 because of the presence of an adnexal mass. All the patients were referred with a diagnosis of adnexal mass based on palpation or ultrasound imaging. After 3 months follow-up in the Correspondence: Professor G. B. Melis, Department of Obstetrics and Gynecology, University of Cagliari, Ospedale San Giovanni di Dio, Via Ospedale 46, 04, Cagliari, Italy ORIGINAL PAPER 277 Received Revised 6 2 Accepted 0 2

2 case of premenopausal women or after a few weeks in the case of postmenopausal women, adnexal masses in 78 women were found to persist and were subsequently examined by laparoscopy (n = 8) or laparotomy (n = 60). The average age of the study population was 4 ± 5 years (mean ± SD), ranging from 4 to 77 years. A total of 27 patients were premenopausal (7%; average age 33 ± 9 years) and 5 were postmenopausal (29%; average age 60 ± 8 years). Within 5 days before surgery, all patients with a persistent mass underwent transvaginal ultrasonography. An Acuson 28 XP/0 ultrasound system with a MHz endovaginal probe (Acuson Inc., Mountain View, CA, USA) was used and the unit was equipped with a color pulsed Doppler system that had been upgraded with color Doppler energy imaging. The maximum transverse, anteroposterior and longitudinal diameters of each mass were measured and the mean diameter was obtained and used in the analysis. With the use of B-mode ultrasonography, a mass was considered benign when it had a typical benign pattern based on the following morphological criteria: () Endometrioma: circular homogeneous hypoechoic tissue without papillary proliferations and with a clear demarcation from the ovarian parenchyma 4,5. (2) Cystic teratoma: one of the three following echo patterns: (a) (b) (c) densely echogenic mural tubercle with posterior acoustic shadow associated with cystic echo pattern, echogenic thin band-like echoes (hyperechogenic sparkling lines and dots in a dark field), dense echo pattern associated with posterior acoustic shadow with or without a cystic component 5,6. (3) Serous cystadenoma and a serous cyst: anechoic unilocular or bilocular cystic mass with a thin regular wall and sometimes also a thin, regular septum without endocystic vegetations 9,5. (4) Hydrosalpinx: an irregular, elongated mass filled with anechoic fluid 9. (5) Benign mucinous cystoadenoma: multilocular mass, with a thin regular wall and septa, containing liquids of different echogenicities, and without endocystic vegetation 9. (6) Functional hemorrhagic ovarian cyst: echogenic mass with an increased capacity for sound transmission displaying a fishnet or a whirled appearance 9. (7) Subserous leiomyoma: mass, separate from the ovaries, with an echogenicity similar to that of the myometrium. Calcified foci within the mass may be present. The mass could be homogeneous or slightly heterogeneous, with thin, multiple acoustic shadows without a hyperechoic focus 9. (8) Paraovarian cyst: unilocular cystic adnexal mass separate from the ipsilateral ovary 9 with a regular wall and a good through-transmission. By B-mode transvaginal ultrasonography, a mass was suspected to be malignant when an echogenic structure situated adjacent to the wall of the cyst was present; when a large (> 3 mm), irregular, homogeneous or heterogeneous echogenic structure was present, suggesting an irregular solid portion; or when an irregular, thickened (> 3 mm) wall or septum was present 9,5,7. All scans were completed by transvaginal CDE imaging. A log compression (dynamic range of energy signal) of db was used for low-intensity signals. The Doppler study was performed by looking for color signals along the wall and within the septa. When these were detected, the pulsed Doppler gate was superimposed, and the PI and RI were electronically calculated. When multiple signals were obtained from the same mass, the lowest PI and RI values were used for the statistical analysis. If no color signal was detected and consequently no Doppler waveforms were obtained, the result of the Doppler study was regarded as negative. By spectral Doppler analysis, malignancy was indicated by a RI of 0.4 8, a PI of 9 or a PI of All scans were performed in the follicular phase by the same physician (S.G.). The intraobserver coefficient of variation was determined by analyzing three sets of five consecutive waveforms from the vessel with the lowest PI and RI in the first ten masses studied. The intraobserver variabilities for RI and PI were 3 and 4%, respectively. By CDE imaging (evaluation of vessel distribution), malignancy was suspected when arterial flow was visualized in an echogenic portion of a mass defined as malignant by B-mode (in agreement with Buy and colleagues 9, using conventional color Doppler imaging). A mass was considered benign by transvaginal CDE imaging when no arterial flow was visualized in an echogenic portion or when flow was seen only in the wall of a mass defined as malignant by B-mode 9. Blood samples from all the patients were collected on the same day as ultrasound examination to measure serum levels of CA-25. The CA-25 assays were performed with an immunoradiometric method using two monoclonal antibodies (CIS Bio International, Gif sur Yvette, France). The intra-assay and interassay coefficients of variation were 3.9 and 4.2%, respectively; the sensitivity was lower than 0.5 U/ml. CA-25 values of 35 and 65 U/ml were chosen as cut-off limits for the preoperative evaluation of adnexal masses, as previously reported 20. At surgery, all ovaries were carefully observed by two of the authors (V.M. and G.B.M.) and all ovarian masses were enucleated from the ovary or removed together with the ovary. The ultrasonographic diagnoses were then compared with the final histopathological diagnoses. The sensitivity, specificity and positive and negative predictive values of transvaginal ultrasonography and all combined methods were calculated for each adnexal mass 4 6. To evaluate the overall agreement between a test result and 278 Ultrasound in Obstetrics and Gynecology

3 the actual outcome, the kappa index was calculated according to Fleiss 3. RESULTS A total of 59 of the adnexal masses were benign and 33 were malignant (prevalence per mass, 7%); seven of the malignant masses were in the premenopausal group and 26 in the postmenopausal group. Nine of the 26 ovarian carcinomas were FIGO stage I, three were stage II, nine were stage III and five were stage IV. Four masses were ovarian metastases, the primary tumor being carcinoma of the colon in one case and breast carcinoma in two cases. The remaining three malignant masses were one granulosa cell tumor and two lymphomas. Of the 60 masses suspected of being malignant by B-mode transvaginal ultrasonography, 33 were confirmed by pathology. The ultrasonographic findings of the 27 Table Description of 27 false-positive cases by B-mode ultrasonography Histopathological diagnosis Endometrioma Hemorrhagic cyst Serous cystoadenoma Serous cyst Tubo-ovarian complex Ovarian fibroma Mucocele of the appendix Uterine pedunculate fibroid Mucinous cystoadenoma Number of masses *Values are medians, with ranges in parentheses 2 Diameter of adnexal masses at ultrasonography (mm) 45 (30 89)* 53 (30 70)* 93.5 (40 200)* false-positive cases resembled findings that are characteristic of malignant masses. The details of the falsepositive cases are summarized in Table. The 33 malignant masses confirmed by pathology had an ultrasonographic mean ± SD diameter of 70 ± 36 mm, ranging from 24 mm to mm. Of 32 ultrasonographic diagnoses of benign cysts, all were confirmed by pathology (endometriomas, n = 38; serous cysts, n = 32; cystic teratomas, n = 9; mucinous cystoadenomas, n = 6; serous cystoadenomas, n = 5; hemorrhagic cysts, n = 7; hydrosalpinges, n = 9; paraovarian cysts, n = 3; uterine pedunculate fibroid, n = ; tuboovarian complex, n = 2). No false-negative results were present with the use of the typical benign patterns based on the morphological B-mode criteria. The sensitivity, specificity, positive and negative predictive values and kappa index of transvaginal ultrasonography in the diagnosis of ovarian carcinoma are reported in Table 2. Power tests were performed on CA-25 plasma levels, PI and RI to determine the probability of detecting a true difference with our sample size. There was a % likelihood that a significant change (α = 0.05) would be detected in CA-25 plasma levels, PI and RI, if it existed. The median CA-25 value in malignant masses was 253 U/ml (range U/ml) and 24.2 U/ml in benign masses (range U/ml) (p < 0.05). Intratumoral arterial blood flow was readily detected by CDE imaging in all malignant tumors and in 94% of the benign ovarian tumors. No arterial flow was detected in two serous cystoadenomas, one endometrioma, one serous cyst, one paraovarian cyst, one tubo-ovarian complex and one cystic teratoma. The median PI value in malignant masses was 0.5 (range ) and 0. in benign masses (range ) (p < 0.05). The median RI value in malignant masses was 0.4 (range Table 2 Accuracy of different ultrasonographic methods in the diagnosis of adnexal malignancies B-mode B-mode and color Doppler energy Doppler analysis, PI Doppler analysis, PI 0.8 Doppler analysis, RI 0.4 PI, pulsatility index; RI, resistance index Sensitivity Specificity Positive predictive value Negative predictive value Kappa value Ultrasound in Obstetrics and Gynecology 279

4 0.4 0.) and 0.6 in benign masses (range ) (p < 0.05). Table 2 shows the diagnostic accuracy of the combined use of transvaginal ultrasonography and CDE imaging which was associated with a significantly (p < 0.05) higher specificity in comparison with transvaginal ultrasonography alone. With the use of CDE imaging, the 2 false-positive cases included three serous papillary cystoadenomas, six endometriomas which had an atypical ultrasonographic appearance due to a vascularized echogenic portion in a homogeneous hypoechoic area, a highly vascularized uterine pedunculated fibroid, a tubo-ovarian complex and a mucinous cystoadenoma. No false-negative results were obtained with the use of the typical benign pattern based on the morphological CDE imaging criteria. The use of conventional spectral Doppler analysis alone or combined with B-mode ultrasonography was of limited diagnostic value, as shown in Tables 2 and 3. Eighty-five of the benign masses had a PI value of and 2 an RI value of 0.4. Four of the malignant masses had a PI value higher than and 4 an RI value higher than 0.4. The association of CA-25 combined with B-mode ultrasonography or with CDE imaging evaluation (Table 4) had a lower accuracy than CDE combined with B-mode ultrasonography (Table 2). DISCUSSION The predictive performance of CDE imaging as a secondary test seems to validate its clinical use in the preoperative diagnosis of adnexal tumors. Using conventional color Doppler imaging, Buy and colleagues 9 used a similar approach, but they obtained a low detection rate of blood flow in benign masses which was associated with a low Table 3 Accuracy of B-mode transvaginal ultrasonography associated with the lowest pulsatility index (PI) and resistance index (RI) B-mode and PI * B-mode and RI 0.4* *, Tests positive Sensitivity 57 Specificity Positive predictive value Negative predictive value Kappa value Table 4 Accuracy of CA-25 alone or associated with ultrasonographic methods in the diagnosis of adnexal malignancies CA U/ml CA U/ml B-mode and CA U/ml* B-mode and CA U/ml* CDE and CA U/ml* CDE and CA U/ml* Sensitivity Specificity Positive predictive value Negative predictive value Kappa value *, Tests positive 280 Ultrasound in Obstetrics and Gynecology

5 confirmation rate of the presence of arterial flow; the study, therefore, had poor reproducibility. In contrast, the present study showed a high detection rate using CDE imaging, demonstrating the presence of arterial flow in all the masses in which a color signal was detected. In CDE imaging, the ultrasound system calculates the energy of the returning Doppler signal rather than its mean frequency shift or velocity 2. This mode confers several advantages over conventional color Doppler imaging. The greatest advantage of CDE imaging is its enhanced ability to convey information-containing signals relative to noise, thereby enhancing sensitivity. Because conventional color Doppler imaging displays the mean Doppler frequency shift, which is a function of blood flow velocity, random noise may be depicted as flow in any direction 2. This can make true flow and noise indistinguishable. Because noise has low energy, however, CDE imaging demonstrates noise as a uniformly colored background that is easily distinguishable from true flow. Thus, CDE has the ability to image areas of low blood flow currently undetectable by frequency-based techniques 2. Other advantages of CDE imaging over conventional color Doppler imaging include an absolute lack of aliasing and a relative angle independence 2. Since malignant growths are characterized by an irregular course of the vessels 3, it is expected that angle-independent flow analysis will allow the sonographer to detect these typical vascular features more easily. CDE imaging should be considered only as a predictive secondary test. Other parameters should first be considered: for instance, the persistence of the mass, the B-mode appearance and the menopausal status. Among women of reproductive age, an important ultrasonographic criterion to be assessed is the persistence of the mass for 3 months. Follow-up could reduce the number of patients with functional cysts that undergo surgery 5. Many benign lesions such as endometriomas 4,5,2,22, serous cysts 5,6 or dermoid cysts 5,6 are easily identified because of their characteristic findings. In our opinion, the use of CDE imaging in these lesions could be avoided; this is in agreement with the results obtained by Valentin 7 using conventional color Doppler imaging. With these simple and reproducible findings, we have never misdiagnosed an ovarian cancer, in the present study, in previous studies 4 6,2,22 and in more than 500 persistent ovarian masses undergoing surgery in our department from 23. On the contrary, in atypical cases or where the findings are suspected, the use of color Doppler can increase the diagnostic accuracy. The B-mode capability allows a diagnosis to be made because the risk of ovarian malignancy in a mass with a typical benign appearance is very low 4 7 and can be further reduced by thorough inspection at laparoscopy which is characterized by high sensitivity 23. When B-mode is inconclusive, only the addition of a new approach such as CDE imaging can reduce the number of false-positive findings. The diagnostic accuracy of the tests in discriminating a benign from a malignant mass was dependent on the menopausal status. In premenopausal women, kappa values were always lower than in postmenopausal women, probably because of the higher number of false-positive results associated with a lower incidence of malignancies in the premenopausal population. The positive predictive value was always lower in the premenopausal population in comparison with the postmenopausal population (Table ). However, Strigini and colleagues 8 found a significantly lower accuracy in the postmenopausal population in comparison with the premenopausal population and no increase in accuracy using a combined approach in the latter population. The differing results shown in our study are probably due to the low incidence of benign hemorrhagic cysts such as endometriomas or inflammatory lesions that can simulate malignancy in postmenopausal women. Our study also showed a higher detection rate of Doppler signals and an increase in accuracy using CDE imaging in both populations. Strigini and colleagues 8 analyzed the use of a cut-off level for PI only. As has been shown previously 3,4,6,8,9,7, this study confirmed the presence of an overlap of values of RI and PI between benign and malignant masses. Only recent studies have attempted to explain this overlap, investigating microscopically the vascular pattern of surgically removed tumors 24 or comparing the tumoral blood flow among different histotypes of ovarian cancer 25. From a clinical point of view, the use of transvaginal ultrasonography associated with CDE imaging evaluation of vessel distribution seems to increase the specificity and positive predictive value of the use of transvaginal ultrasonography alone, reducing the number of false-positive results. In conclusion, these results suggest that, in differentiating benign from malignant adnexal masses, the use of color Doppler energy imaging could be a secondary test when a mass has been classified as malignant by B-mode ultrasonography. ACKNOWLEDGEMENT We thank Marisa Orru, DSc, from the Department of Obstetrics and Gynecology of the University of Cagliari, for performing the assays. REFERENCES. Kurjak A, Shalan S, Kupesic S, Predanic M, Zalud I, Breyer B, Jukic S. Transvaginal color Doppler sonography in the assessment of pelvic tumor vascularity. Ultrasound Obstet Gynecol 3;3: Kurjak A, Kupesic S. Transvaginal color Doppler and pelvic tumor vascularity: lessons learned and future challenges. Ultrasound Obstet Gynecol 9;6: Tekay A, Jouppila P. Controversies in assessment of ovarian tumors with transvaginal color Doppler ultrasound. Acta Obstet Gynecol Scand 9;: Rehn M, Lohmann K, Rempen A. Transvaginal ultrasonography of pelvic masses: evaluation of B-mode technique and Doppler ultrasonography. Am J Obstet Gynecol 9;: Caruso A, Caforio L, Testa AC, Ciampelli M, Benedetti Panici P, Mancuso S. Transvaginal color Doppler ultrasonography in the presurgical characterization of adnexal masses. Gynecol Oncol 9;63: 9 Ultrasound in Obstetrics and Gynecology 2

6 6. Prompeler HJ, Madjar H, Sauerbrei W. Classification of adnexal tumors by transvaginal color Doppler. Gynecol Oncol 9;6: Alcàzar JL, Ruiz-Perez ML, Errasti T. Transvaginal color Doppler sonography in adnexal masses: which parameter performs best? Ultrasound Obstet Gynecol 9;8: Strigini FAL, Gadducci A, Del Bravo B, Ferdeghini M, Genazzani AR. Differential diagnosis of adnexal masses with transvaginal sonography, color flow imaging, and serum CA 25 assay in pre- and postmenopausal women. Gynecol Oncol 9;6: Buy JN, Ghossain MA, Hugol D, Hassen K, Sciot C, Truc JB, Poitout P, Vadrot D. Characterization of adnexal masses: combination of color Doppler and conventional sonography compared with spectral Doppler analysis alone and conventional sonography alone. Am J Roentgenol 9;66: Anandakumar C, Chew S, Wong YC, Chia D, Ratnam SS. Role of transvaginal ultrasound color flow imaging and Doppler waveform analysis in differentiating between benign and malignant ovarian tumors. Ultrasound Obstet Gynecol 9; 7: Kurjak A. Conflicting Doppler data on the assessment of pelvic tumors vascularity. Eur J Obstet Gynecol Reprod Biol 9;62: 2 2. Westkott HP. Amplitude Doppler US: slow blood flow detection tested with a flow phantom. Radiology 7;202: Fleiss JL. Statistical Methods for Rates and Proportions. New York: Wiley, 9;2 4. Guerriero S, Ajossa S, Paoletti AM, Mais V, Angiolucci M, Melis GB. Tumor markers and transvaginal ultrasonography in the diagnosis of endometrioma. Obstet Gynecol 9;: Guerriero S, Mallarini G, Ajossa S, Risalvato A, Satta R, Mais V, Angiolucci M, Melis GB. Transvaginal ultrasound and computed tomography combined with clinical parameters and CA-25 determinations in the differential diagnosis of persistent ovarian cysts in premenopausal women. Ultrasound Obstet Gynecol 7;9: Mais V, Guerriero S, Ajossa S, Angiolucci M, Paoletti AM, Melis GB. Transvaginal ultrasonography in the diagnosis of cystic teratoma. Obstet Gynecol 9;85: Valentin L. Gray scale sonography, subjective evaluation of the color Doppler image and measurement of blood flow velocity for distinguishing benign and malignant tumors of suspected adnexal origin. Eur J Obstet Gynecol Reprod Biol 7;72: Kurjak A, Zalud I, Alfirevic Z. Evaluation of adnexal masses with transvaginal color ultrasound. J Ultrasound Med ; 0: Carter JR, Lau M, Fowler JM, Carlson JW, Carson LF, Twiggs LB. Blood flow characteristics of ovarian tumors: implications for ovarian cancer screening. Am J Obstet Gynecol 9;72: Gadducci A, Ferdeghini M, Prontera C, Moretti L, Mariani G, Bianchi R, Fioretti P. The concomitant determination of different tumor markers in patients with epithelial ovarian cancer and benign ovarian masses: relevance for differential diagnosis. Gynecol Oncol 2;44: Guerriero S, Mais V, Ajossa S, Paoletti AM, Angiolucci M, Melis GB. Transvaginal ultrasonography combined with CA 25 plasma levels in the diagnosis of endometrioma. Fertil Steril 9;65: Mais V, Guerriero S, Ajossa S, Angiolucci M, Paoletti AM, Melis GB. The efficiency of transvaginal ultrasonography in the diagnosis of endometrioma. Fertil Steril 3;60: Guerriero S, Ajossa S, Melis GB. The preoperative evaluation of ovarian tumors can be improved. Am J Obstet Gynecol 7;77: Kurjak A, Jukic S, Kupesic S, Babic D. A combined Doppler and morphopathological study of ovarian tumors. Eur J Obstet Gynecol Reprod Biol 7;7: Hata K, Hata T. Intratumoral blood flow analysis in ovarian cancer: what does it mean? J Ultrasound Med 9;5: Ultrasound in Obstetrics and Gynecology