Ovarian Teratomas Appearing as Solid Masses on Ultrasonography

Transcription

1 Ovarian Teratomas Appearing as Solid Masses on Ultrasonography Dong Kyung Lee, MD, Seung Hyup Kim, MD, Jeong Yeon Cho, MD, Sang Joon Shin, MD, Kyung Mo Yeon, MD The purposes of this study were to evaluate the prevalence and imaging characteristics of ovarian teratomas that appear as solid masses on ultrasonography and to compare the ultrasonographic imaging features of the tumors with their pathologic findings. The ultrasonographic images of 202 ovarian teratomas were reviewed retrospectively. Solidappearing masses were selected from among them and were evaluated in terms of internal echotexture, the presence or lack of peripheral hypoechogenicity, posterior sonic attenuation, and tumoral calcification. Seventy-six (37.6%) masses of 202 belonged to the atypical solid-appearing masses on ultrasonography. Of 76 masses, 57 (75.0%) had peripheral hypoechogenicity; 38 masses had well-defined thin hypoechoic rims, whereas 19 had poorly demarcated peripheral hypoechogenicity. Posterior sonic attenuation was evident in 18 (23.7%) masses. The presence of peripheral hypoechogenicity, which is suggestive of the fluid portion of the tumor, might be one of the characteristic findings of solid-appearing ovarian teratomas on ultrasonography. KEY WORDS: Ovary, ultrasonography; Neoplasms, ovarian; Teratoma. M ature cystic teratomas, also called dermoid cysts, are relatively common benign ovarian neoplasm and represent ABBREVIATIONS CT, Computed tomography; MR, Magnetic resonance Received July 16, 1998, from the Department of Radiology, College of Medicine, Seoul National University and the Institute of Radiation Medicine, Seoul, Republic of Korea. Revised manuscript accepted for publication October 7, Address correspondence and reprint requests to Seung Hyup Kim, MD, Department of Radiology, College of Medicine, Seoul National University, 28 Yongon-Dong, Chongno-Gu, Seoul, , Republic of Korea. about 20% of all ovarian tumors. They usually are made up of three germ cell layers with predominance of ectodermal elements. Pathologic investigation reveals that the cyst wall is lined by keratinized squamous epithelium and that the cyst contains sebaceous material in 97% of cases. 1 3 It is well known that mature ovarian teratomas show a wide spectrum of ultrasonographic imaging appearances, and teratomas sometimes appear as solid masses, making it difficult to differentiate them from other solid ovarian tumors. 4 The present study specifically focused on the imaging characteristics of solid-appearing ovarian teratomas on ultrasonography. The purposes of this study were to evaluate the prevalence and the imaging features of ovarian teratomas that appear as solid masses on ultrasonography and to compare the imaging features with their pathologic findings by the American Institute of Ultrasound in Medicine J Ultrasound Med 18: , /99/$3.50

2 142 OVARIAN TERATOMAS J Ultrasound Med 18: , 1999 METHODS The ultrasonographic images of 202 pathologically proved ovarian teratomas in 190 patients (age range, 16 to 78 years) were reviewed retrospectively. Pelvic ultrasonography was performed with various realtime sector scanners with MHz transducers. All the patients were scanned via the transabdominal route through the distended bladder. Two radiologists characterized the ultrasonographic nature of each mass with consensus and selected predominantly solid-appearing masses, which were defined as those with more than 90% echogenic area. The features of the solid-appearing teratomas that were analyzed included internal echotexture, presence or absence of posterior sonic attenuation, tumoral calcification, and halo-like peripheral hypoechogenicity. The echogenicity of each mass was compared with that of myometrium of the uterus. Bright, echogenic foci with sharply demarcated distal acoustic shadowing were considered calcifications. In cases of masses with peripheral hypoechogenicity, we classified the masses into two groups, one with well-defined peripheral hypoechogenicity and the other with poorly defined peripheral hypoechogenicity. In all the selected cases, ultrasonographic images were compared with the pathologic findings. RESULTS Among 202 pathologically proven ovarian teratomas, 76 (37.6%) masses in 74 patients belonged to the group of solid-appearing masses on ultrasonography. Internal Echotexture Thirty-three (43.4%) of 76 masses showed homogeneous echogenicity, whereas 43 (56.6%) masses showed heterogeneous echogenicity. Twenty-eight of 33 masses with homogeneous internal echoes were hyperechoic, with only five being hypoechoic. Thirty-one of 43 masses with heterogeneous echotexture were predominantly hyperechoic, and the remaining 12 masses were predominantly hypoechoic. Posterior Sonic Attenuation and Calcification (Table 1) Posterior sonic attenuation was evident in 18 (23.7%) masses. Nine of the masses had very bright echogenic foci with clear posterior sonic attenuation, suggesting calcifications (Fig 1). In the remaining nine masses, gradual fading of sound beam was noted. Eight of the nine masses had a moderately or minimally echogenic area in front of shadowing, and no demonstrable echogenic foci were seen on ultrasonography in the remaining mass. In all nine masses that showed bright echogenic foci with clear posterior sonic attenuations, actual presence of calcified material was confirmed in pathologic evaluation. In the remaining nine masses, only three masses proved to have calcified materials; six masses did not have any calcified structures, such as tooth, even at pathologic evaluation. Pathologic examination revealed that 23 of 76 tumors had evidence of calcification, including tooth or bone. However, in only nine of 23 did ultrasonography show prominent echogenic foci with posterior sonic attenuations. In 11 of 23 masses possessing calcified structures, ultrasonography did not reveal echogenic foci or posterior sonic attenuation that may lead to the suspicion of intratumoral calcified material. Presence of Halo-like Peripheral Hypoechogenicity Among 76 masses, 57 (75.0%) had peripheral hypoechogenicity: 38 masses had well-defined, thin hypoechoic rims, whereas 19 had poorly demarcated peripheral hypoechogenicity (Figs. 2, 3). In 19 (25.0%) masses, no ultrasonographically demonstrable peripheral hypoechogenicity was noted. Correlation of peripheral hypoechogenicity with gross pathologic specimens was done in some representative cases and revealed that the halo-like hypoechoic peripheral portion of the mass consisted chiefly of fluid component. Table 1: Ultrasonographic-Pathologic Correlation of Intratumoral Calcified Materials Ultrasonographic Findings Calcified Material on Pathologic Examination Positive (N = 23) Negative (N = 53) Posterior sonic attenuation without bright echogenic foci (N = 9) 3 6 Posterior sonic attenuation with bright echogenic foci (N = 9) 9 0 No evidence of posterior sonic attenuation (N = 58) 11 47

3 J Ultrasound Med 18: , 1999 LEE ET AL 143 DISCUSSION UItrasonography plays a major role in diagnostic imaging of various pathologic ovarian processes. However, with regard to ovarian teratomas, the diagnostic value of ultrasonography in comparison to other imaging modalities, such as CT or MR imaging, is less promising owing to relatively nonspecific imaging features that reflect the complex internal structure and variable tissue components of the tumors. It is well known that a mainly cystic adnexal mass containing an echogenic focus with distal acoustic shadowing or having prominent fat-fluid or hair-fluid level is a characteristic feature of ovarian teratomas. Nevertheless, no more than half of all mature cystic teratomas showed such typical findings, 4 and certain masses are found to be mostly cystic or totally solid masses. It is true that easy differentiation of cystic from solid lesions is a distinctive merit of ultrasonography. However, it is not so easy to characterize the internal composition of mass in many cases of the ovarian teratomas by means of ultrasonography. Fatty lesions show a broad spectrum of sonographic appearances. The echogenicity of such lesions is the function of the acoustic impedance mismatch between the fat itself and the supportive connective tissues. 5,6 Consequently, a fatty lesion may appear entirely anechoic, have cystic internal echotexture, or even be markedly echogenic. In the case of teratoma, which consists predominantly of soft tissue other than fat, the mass would be hypoechoic just like other ovarian solid neoplasms. This Figure 1 Mature cystic teratoma of right ovary in a 43 year old woman. A hypoechoic area is present at the peripheral area of solid-appearing mass in right adnexa (large arrows). Note an echogenic focus in the mass with prominent posterior sonic attenuation (small arrow). varied ultrasonographic appearance often prevents radiologists from making a proper imaging diagnosis, and further imaging modalities are needed to determine the exact nature of the tumors. According to Laing and coworkers, % of dermoid cysts showed predominantly solid ultrasonographic pattern. In some reports ultrasonographically solid ovarian teratomas often were mistaken for uterus or pedunculated leiomyomas. 7 Ovarian teratomas usually are detected owing to their marked posterior shadowing, which is known to be due to the predominance of hair, to hair floating on top of sebaceous material, or to particulate matter such as tooth or bone in the masses. 4 In our series, 30% (23 of 76) of masses had actual calcified structures on pathologic evaluation, and 23.7% (18 out of 76) showed posterior sonic attenuation on ultrasonography. Consequently, in a large number of masses the proper preoperative diagnosis would not have been made when typical known ultrasonographic findings are used as diagnostic criteria in solid-appearing ovarian teratomas. Moreover, when the ultrasonographic criteria of calcified materials were applied to our series, the detection rate of intratumoral calcified materials was less than 40%. The failure to demonstrate calcified tissue in our series is thought to be partly the result of the sonographer s experience and partly the result of masking of the tissue by proximally located highly echogenic tissues, such as hair strands (the so-called tip of the iceberg sign). 8 Nine masses showed no demonstrable ultrasonographic evidence of calcified structures despite prominent posterior sonic attenuation, which is consistent with the tip of the iceberg sign. Some reports have described entirely solidappearing ovarian teratomas as being totally echogenic with no discernible anechoic component on ultrasonography. 2,4 Our study showed that 75% of solid-appearing ovarian teratomas had peripheral hypoechogenicity. The correlation of ultrasonograms with CT scans and MR images in limited cases showed that the echogenic, solid-appearing portion on ultrasonography was made up of mixture of hair, sebum, or fatty tissue intermixed with multiple soft tissue strands and that the hypoechogenicity in the peripheral portion of the mass consisted of fluid component. Pathologic specimens showed that the fluid portion consisted of pure sebum (known to be liquid at body temperature 4,5 ), serous fluid, or both. The cause of differences in the sharpness of the margins between the main echogenic portion and the peripheral hypoechogenic portion is not clear, but one possible explanation may be advanced. In the

4 144 OVARIAN TERATOMAS J Ultrasound Med 18: , 1999 A Figure 2 Mature cystic teratoma of right ovary in a 33 year old woman. A, Ultrasonography shows poorly defined peripheral hypoechoic area in the right adnexal mass (arrow). B, Axial T1-weighted (TR/TE = 500/30) MR image shows a wellmarginated mass in the right adnexa. Note a crescent-shaped area of high signal intensity at the periphery of the mass (arrow). C, On axial T2-weighted (TR/TE = 2000/85) MR image, the mass shows homogeneous signal intensity with no demonstrable crescent-shaped area corresponding to the one on the T1-weighted image, suggesting that its nature is fat. B C lesions with poorly defined margins between the solid-appearing lesion and the peripheral hypoechogenicity on ultrasonography, the acoustic impedance of the tissue interface may be less prominent than that of lesions with well-defined margins between the two tissues. Greasy mixtures composed of sebum and keratin, which are usually liquid at body temperature, 9,10 may result in a less prominent acoustic interface to the solid portion of the mass than would occur with pure serous fluid or pure sebum. The ultrasonographic finding of a solid-appearing ovarian mass with peripheral hypoechogenicity in ovarian teratoma can be explained as a predominatly solid mass surrounded with a small amount of peripheral fluid. This implies that when such ultrasonographic findings are seen, the proportion of the Rokitansky protuberance may be much larger than that of surrounding cystic area in the mature cystic teratoma. Thus, this ultrasonographic finding might be one of the specific findings of ovarian teratoma. It is well known that CT and MR imaging are superior to ultrasonography in evaluating and diagnosing the mature ovarian teratomas. However, considering that ultrasonography is the major screening imaging modality for various pathologic pelvic conditions, it may be noteworthy to be familiar with the varied atypical ultrasonographic appearances of ovarian teratomas. Our study has some limitations. First, it was performed retrospectively. Ultrasonography was the only preoperative imaging performed in many cases in our series, and ultrasonographic-pathologic correlation to find out the nature of peripheral hypoechogenicity also was done in a limited number of cases. We do not have any data on the prevalence of peripheral hypoechogenicity of solid-appearing ovarian tumors other than ovarian teratomas as yet. Further studies must be undertaken on whether the peripheral hypoechogenicity of solid-appearing ovarian mass on ultrasonography is one of major and specific findings of solid-appearing atypical ovarian teratomas.

5 J Ultrasound Med 18: , 1999 LEE ET AL 145 A B C Figure 3 Mature cystic teratoma of left ovary in a 40 year old woman. A, Pelvic ultrasonography shows a well-defined, crescentshaped hypoechoic area at the periphery of the left adnexal mass (arrows). B, Axial T1-weighted (TR/TE = 600/15) MR image shows a well-marginated mass in the left adnexa. A well-defined area of low signal intensity surrounds the left adnexal mass (arrow). C, On axial T2-weighted (TR/TE = 2000/80) MR image, the mass shows homogeneous high signal intensity. The peripheral area surrounding the mass on T1-weighted axial image (Fig. 3A) was not evident. Pathologic evaluation of the tumor specimen showed the persence of an area of clear serous fluid at the periphery between the central solid portion and the wall of tumor. D, Gross specimen of another patient with mature cystic teratoma. A cleft is present between the solid portion and the capsule of the tumor, which was filled with serous fluid in vivo (arrows). D REFERENCES 1. Buy JN, Ghossain MA, Moss AA, et al: Cystic teratoma of the ovary: CT detection. Radiology 171:697, Sheth S, Fishman EK, Buck JL, et al: The variable sonographic appearances of ovarian teratomas: Correlation with CT. AJR 151:331, Scanne P, Huebener KH: Computed tomography of cystic ovarian teratomas with gravity-dependent layering. J Comput Assist Tomogr 7:837, Laing FC, van Dalsan VF, Marks VM, et al: Dermoid cysts of the ovary: Their ultrasonographic appearances. Obstet Gynecol 57:99, Sanders RC: Sonography of fat. In Sanders RC, Hill MC (Eds): Ultrasound Annual. New York, Raven, 1984, p Behan M, Kazam E: The echographic characteristics of fatty tissues and tumors. Radiology 129:143, Sandler MA, Silver TM, Karo JJ: Gray-scale ultrasonic features of ovarian teratomas. Radiology 131:705, Guttman PH Jr: In search of the elusive benign cystic ovarian teratoma: Application of the ultrasound tip of the iceberg sign. J Clin Ultrasound 5:403, Rosai J: Female reproductive system/ovary. In Ackerman s Surgical Pathology. Vol. 2. St. Louis, CV Mosby, 1989, p Talerman A: Germ cell tumors of the ovary. In Kurman RJ (Ed): Pathology of the Female Genital Tract. 3rd Ed. New York, Springer-Verlag, 1987, p 687