3303jum387-530online_Layout 1 2/19/14 1:38 PM Page 421

ORIGINAL RESEARCH

Nonmasslike Lesions on Breast Sonography Comparison Between Benign and Malignant Lesions Suk Jung Kim, MD, Young Mi Park, MD, Hyun Kyung Jung, MD

Objectives—To compare the imaging and clinical features of benign and malignant nonmasslike lesions in the breast. Methods—During a 2-year period at a single institution, 186 nonmasslike lesions in 158 women were pathologically confirmed through surgery or sonographically guided biopsy. The sonographic patterns (mottled, geographic, and indistinct) and distributions (focal and regional) were compared between benign and malignant lesions. The presence of sonographically visible calcifications, amount of color Doppler signals, presence of positive findings on mammography, and presence of symptoms were also compared between the two groups. Results—A total of 156 lesions (84%) were confirmed as benign and 30 (16%) as malignant. On sonography, malignant nonmasslike lesions more frequently had mottled and geographic patterns and regional distribution than benign lesions (P < .0001). Malignant lesions also more frequently had sonographically visible calcifications (40% versus 0%; P < 0.0001) and a greater amount of color Doppler signals than benign lesions (P < .0001). On mammography, malignant lesions more frequently had densities and calcifications than benign lesions (30.4% versus 7.1%; P = 0.0052; 73.9% versus 6.1%; P < .0001, respectively). Clinically, malignant lesions were more frequently palpable and accompanied by localized pain than benign lesions (50% versus 2.6%; P < .0001; 13.3% versus 0.6%; P = .0025). Received May 7, 2013, from the Department of Radiology, Haeundae Paik Hospital, Inje University College of Medicine, Busan, Korea (S.J.K., H.K.J.); and Department of Radiology, Busan Paik Hospital, Inje University College of Medicine, Busan, Korea (Y.M.P.). Revision requested June 5, 2013. Revised manuscript accepted for publication July 18, 2013. This work was supported by a 2008 grant from Inje University. Address correspondence to Young Mi Park, MD, Department of Radiology, Busan Paik Hospital, Inje University College of Medicine, 633-165 Gaegeum-dong, Busanjin-gu, Busan 614-735, Korea. E-mail: [email protected] Abbreviations

BI-RADS, Breast Imaging Reporting and Data System; DCIS, ductal carcinoma in situ doi:10.7863/ultra.33.3.421

Conclusions—The imaging and clinical features of malignant nonmasslike lesions differed significantly from those of benign nonmasslike lesions. Key Words—breast; breast ultrasound; nonmasslike lesions; sonography

W

ith advances in sonographic equipment enabling higher resolution, some lesions that are clearly discernible from adjacent normal parenchyma but do not conform to the definition of a mass have been increasingly recognized during breast sonography. Authors have suggested that some cases of ductal carcinoma in situ (DCIS) and invasive lobular carcinoma may appear as such lesions on sonography,1–14 and these observations have created the necessity for the correct identification and characterization of such lesions. The current American College of Radiology Breast Imaging Reporting and Data System (BI-RADS) lexicon only covers mass lesions, and there is no established term that describes these lesions.15 Therefore, such lesions have been described using various terms, such as indeterminate, nonmass lesion, nonmasslike lesion, nonmass image-forming lesion, ductal changes, ductlike structures, and focal shadowing.1–4,12,16

©2014 by the American Institute of Ultrasound in Medicine | J Ultrasound Med 2014; 33:421–430 | 0278-4297 | www.aium.org

3303jum387-530online_Layout 1 2/19/14 1:38 PM Page 422

Kim et al—Nonmasslike Lesions on Breast Sonography

For the first time, the Japanese Association of Breast and Thyroid Sonology systematically organized and classified nonmass image-forming lesions and presented provisional guidelines for diagnosis.17 Despite obvious evidence that nonmasslike lesions do exist, to date, no established criteria exist to differentiate benign from malignant nonmasslike lesions. Therefore, the purpose of this study was to investigate the incidence and histologic characteristics of nonmasslike lesions and to compare imaging features and clinical characteristics between benign and malignant nonmasslike lesions.

Materials and Methods From March 2010 to July 2012, a total of 9528 breast sonographic examinations were performed at our institution. Among these examinations, we identified lesions designated as nonmasslike lesions on breast sonography through a search of our radiology database. An nonmasslike lesion was defined as a hypoechoic area whose character was different from that of surrounding glands or the same area in the contralateral breast. Lesions showing ductal hypoechoic areas were excluded from this study. At our institution, whole-breast sonography is usually indicated for screening purposes in women with dense breast tissue, in high-risk women younger than 40 years with a family history of breast cancer, in women who request the screening themselves, for diagnostic purposes in women with newly diagnosed breast cancer to evaluate the extent of disease, and in symptomatic women with a palpable abnormality, nipple discharge, or breast pain. The Institutional Review Board approved this retrospective study, and the requirement for informed consent from the patients was waived. A total of 505 nonmasslike lesions were detected during a total of 9528 breast sonographic examinations. Among the 505 nonmasslike lesions, 189 cases were lost to followup, and 110 cases were followed with serial sonographic examinations (mean follow-up interval, 10 months; range, 4 days–26 months). Of the lesions, 206 were pathologically confirmed by surgery (n = 60), 14-gauge core needle biopsy (n = 115), or vacuum-assisted biopsy (n = 31). In 20 lesions diagnosed by surgery, the exact pathologic outcomes were unclear; therefore, these lesions were excluded from the study population. In our data set, a minimum 2-year followup period could not be achieved. Finally, a total of 186 pathologically confirmed nonmasslike lesions in 158 women made up our final study cohort. The patients’ ages ranged from 23 to 68 years (mean age, 45 years). Whole-breast sonography was performed with an iU22 ultrasound machine (Philips Healthcare, Bothell,

422

WA) equipped with a linear 5–12-MHz transducer. The sonographic examinations were performed by 1 of 2 dedicated breast-imaging radiologists (S.J.K. and H.K.J.) with 5 and 3 years of experience in breast sonography, respectively. Grayscale sonography was first performed to evaluate the sonographic characteristics of the nonmasslike lesions. Extended field-of-view technology was frequently used to visualize the entire area of involvement and to highlight the contrast with normal tissue. Next, color Doppler sonography was obtained. The color box was enlarged to encompass the lesions as much as possible and to also include a small margin of normal breast tissue, if possible. The color Doppler gain was increased so that only background noise was suppressed and small vessels could be detected. During the color Doppler examination, pressure was applied to the lesion with a probe as lightly as possible to avoid obliterating vessels. Mammography was performed with a full-field digital mammographic system (Mammomat Inspiration; Siemens AG, Munich, Germany). Standard mediolateral oblique and craniocaudal images of the breasts were obtained from each patient. Percutaneous biopsy was performed with either a 14-gauge semiautomated core biopsy needle (Stericut coaxial; TSK Laboratory, Tochigi, Japan) or a handheld vacuum-assisted device with an 8-gauge probe (Mammotome; Devicor Medical Products; Cincinnati, OH) under sonographic guidance. Two dedicated breast imaging radiologists (Y.M.P. and S.J.K.), with 10 and 5 years of experience in breast imaging, respectively, analyzed the sonograms and classified the sonographic patterns into mottled, geographic, and indistinct groups by consensus. Classification of the sonographic patterns followed the guidelines of the Japanese Association of Breast and Thyroid Sonology.17 Indistinct patterns were defined as relatively uniform hypoechoic areas whose margins were not clearly defined (Figures 1 and 2). A mottled pattern was defined as a number of small islands of hypoechoic areas (Figures 3 and 4). A geographic pattern was defined as a confluent hypoechoic area with a cobblestone appearance that resembled geographic maps, as if mottled hypoechoic areas were fused into one area (Figures 5 and 6). The readers also classified the sonographic distributions of the nonmasslike lesions into focal and regional groups by consensus. A focal distribution was defined as a lesion involving less than 1 quadrant of the breast, and a regional distribution was defined as a lesion involving more than 1 quadrant of the breast. Diffuse involvement of the whole breast was not observed in our study population. During sonographic interpretation, the readers were blinded to mammographic findings, clinical profiles, and histologic results, as well as the proportion of cases

J Ultrasound Med 2014; 33:421–430

3303jum387-530online_Layout 1 2/19/14 1:38 PM Page 423

Kim et al—Nonmasslike Lesions on Breast Sonography

with benign and malignant histologic results. The readers also counted the number of vessels within the lesions on color Doppler sonography and categorized the amount of color Doppler signals as absent (no vessels), mild (≥1 but

Nonmasslike lesions on breast sonography: comparison between benign and malignant lesions.

To compare the imaging and clinical features of benign and malignant nonmasslike lesions in the breast...
1MB Sizes 0 Downloads 3 Views