Imaging Differences in Metaplastic and Invasive Ductal Carcinomas of the Breast

Transcription

1 Yang et al. Metaplastic Versus Invasive Ductal reast Carcinoma Women s Imaging Original Research WOMEN S IMGING Wei Tse Yang 1 ryan Hennessy 2 Kristine roglio 3 Chadwick Mills 1,4 Nour Sneige 5 W. Grant Davis 5,6 Vicente Valero 7 Kelly K. Hunt 8 Michael Z. Gilcrease 5 Yang WT, Hennessy, roglio K, et al. Keywords: breast neoplasm, invasive ductal carcinoma, mammography, metaplastic breast carcinoma, sonography DOI: /JR Received February 16, 2007; accepted after revision June 30, Department of Diagnostic Imaging, The University of Texas M. D. nderson Cancer Center, PO ox , Unit 1350, Houston, TX ddress correspondence to W. T. Yang (wyang@di.mdacc.tmc.edu). 2 Department of Gynecology Medical Oncology, The University of Texas M. D. nderson Cancer Center, Houston, TX. 3 Quantitative Sciences Division, The University of Texas M. D. nderson Cancer Center, Houston, TX. 4 Present address: Department of Radiology, aylor College of Medicine, Houston, TX. 5 Department of Pathology, The University of Texas M. D. nderson Cancer Center, Houston, TX. 6 Present address: Physicians Reference Laboratory, Overland Park, KS. 7 Department of reast Medical Oncology, The University of Texas M. D. nderson Cancer Center, Houston, TX. 8 Department of Surgical Oncology, The University of Texas M. D. nderson Cancer Center, Houston, TX. CME This article is available for CME credit. See for more information. JR 2007; 189: X/07/ merican Roentgen Ray Society Imaging Differences in Metaplastic and Invasive Ductal Carcinomas of the reast OJECTIVE. The purpose of this study was to compare the imaging features of metaplastic breast carcinoma with those of invasive ductal carcinoma. MTERILS ND METHODS. Women diagnosed on preoperative mammography or sonography with metaplastic breast carcinoma and T-stage matched invasive ductal carcinoma of the breast from a single pathology database were included in the study. Clinical and pathologic information on all metaplastic cancers was documented. Mammography and sonography variables were recorded using the I-RDS lexicon. Groups were compared using Fisher s exact test, the chi-square test, or Wilcoxon s rank sum test, as appropriate. RESULTS. Forty-three patients diagnosed with metaplastic carcinoma were matched to 43 patients with ductal carcinoma by tumor T stage. Patients with metaplastic carcinoma were younger (median, 46 vs 53 years, p = 0.048) than those with ductal carcinoma. Mammographically, metaplastic carcinomas were less frequently irregular in shape (16% vs 74%, p < ) and less frequently showed microlobulated or spiculated margins (19% vs 56%, p = ) and calcifications (25% vs 51%, p = 0.02) when compared with ductal carcinomas. Sonographically, metaplastic carcinomas were less frequently irregular in shape (27% vs 69%, p = 0.001) and less frequently showed angular margins (9% vs 49%) and posterior acoustic shadowing (9% vs 49%, p < ). CONCLUSION. Characteristic malignant imaging features, including irregular shape, spiculated margins, segmentally distributed pleomorphic calcifications, and posterior acoustic shadowing, are uncommon in metaplastic carcinomas. These carcinomas tend to show more benign imaging features, such as round or oval shape with circumscribed margins, when compared with ductal carcinomas. nvasive breast carcinoma usually I has an epithelial morphology with ductal differentiation. Metaplastic carcinoma, on the other hand, is a heterogeneous group of neoplasms with mixed epithelial and mesenchymal differentiation. This mixed cell differentiation is seen both morphologically and immunophenotypically, as evidenced by immunohistochemical expression of markers of mesenchymal cells (vimentin), epithelial cells (pancytokeratin), and myoepithelial cells (S-100, smooth-muscle actin, and p63). Metaplastic breast carcinoma is uncommon, accounting for fewer than 5% of breast carcinomas [1 5]. Wargotz and Norris [1 3] and Wargotz et al. [4] suggested five principal variants, including matrix-producing carcinoma, spindle-cell carcinoma, squamous cell carcinoma, carcinosarcoma, and metaplastic carcinoma with osteoclastic giant cells [6]. Recently, lowgrade fibromatosis-like tumors have been described as a biologically distinct subtype [7, 8]. The differential diagnosis between invasive ductal carcinoma and metaplastic carcinoma is important for treatment planning and prognosis. recent study described significantly worse overall survival for the metaplastic subtype compared with a control group [9]. We undertook a study to compare imaging features of invasive ductal carcinomas and metaplastic carcinomas to determine whether significant differences are shown on conventional imaging methods, including mammography and sonography. Materials and Methods Clinical Data and Pathology Forty-three patients from the surgical pathology database of a single institution who had a diagnosis of metaplastic breast carcinoma from January 1995 to June 2005 and who had undergone preoperative imaging with mammography or sonography were included in this study. second study group comprised 43 patients with invasive ductal carcinoma 1288 JR:189, December 2007

2 Metaplastic Versus Invasive Ductal reast Carcinoma TLE 1: Clinical Data and Pathology Features of 43 Metaplastic and 43 Ductal Cancers Metaplastic Ductal Feature No. % No. % xillary nodes at surgery Positive Negative Unknown Estrogen receptor status Positive Negative Unknown Progesterone receptor status Positive Negative Unknown HER2/neu overexpression Positive Negative Unknown Histologic subtype Predominantly sarcomatoid Mixed sarcomatoid and epithelial Squamous Low-grade fibromatosis Miscellaneous a Note Data are numbers of tumors. Percentages may not total 100 due to rounding. a One invasive ductal carcinoma with squamous differentiation and one matrix-producing carcinoma with squamous differentiation. from the same surgical pathology database diagnosed during the same time period who were matched to the patients with metaplastic carcinoma by tumor T stage (size) and who also had breast imaging available for review. Tumors in seven patients were classified as stage T1 (< 2 cm), in 34 patients as stage T2 (> 2 cm and < 5 cm), and in two patients as stage T3 (> 5 cm) in each group. total of 21,037 patients were diagnosed with invasive ductal carcinoma during this period. Information about histopathologic features was obtained from pathology reports or, for selected cases requiring clarification of the pathology findings, through review of H and E stained slides by one of two dedicated breast pathologists and one breast pathology fellow. The clinical charts were reviewed by a senior medical oncology fellow to collect information on tumor stage, estrogen and progesterone receptor status, HER2/neu status, and clinical follow-up. The clinical and pathologic tumor stages were determined using the sixth edition of the cancer staging manual of the merican Joint Committee on Cancer [10]. Imaging Mammography Mammography was performed using a Lorad M3 (Hologic) or a DMR series mammography unit (GE Healthcare). Standard twoview diagnostic mammography was performed, with additional views as deemed necessary. Two mammographers with 3 and 10 years experience in breast imaging reviewed all available mammograms and sonograms. These reviews were done independently and without knowledge of the clinical and pathologic findings. greement as to the presence or absence of findings was by consensus. reast parenchymal density was classified according to the merican College of Radiology I- RDS classifications [11]. Mammograms were reviewed for focal masses, calcifications, asymmetric density, architectural distortion, and associated features such as skin thickening and retraction, nipple retraction, and axillary lymphadenopathy. Sonography Real-time gray-scale sonography and color Doppler sonography were performed using an Elegra unit (Siemens Medical Solutions) by an attending radiologist who was one of a group of 10 radiologists (including two of the authors) assigned to breast sonography. Sonograms were assessed for masses (solid or cystic) and their shapes, margins, echo patterns, posterior acoustic features, calcifications, vascularity determined by color Doppler imaging, and effects on surrounding tissue, according to the I-RDS sonography lexicon [12, 13] as determined by the two mammographers. Disease was also assessed as unifocal, multifocal, or multicentric in all patients who underwent sonography. Sonographic assessment of the regional lymph node basins, including the axillary, infraclavicular, internal mammary, and supraclavicular regions, was documented according to previously published criteria [14, 15]. Whole-breast and nodal basin sonography are routinely performed at our institution because most patients present for staging of a known malignancy. Statistical Methods Mammography variables, including size, shape, margin, and density of masses as well as morphology and distribution of calcifications, were recorded according to I-RDS criteria. Sonography variables recorded included mass size, shape, margin, posterior acoustic phenomena, vascularity, and axillary lymph node status according to I-RDS criteria. Imaging features were tabulated and compared between groups using Fisher s exact test, the chisquare test, or Wilcoxon s rank sum test, as appropriate. Values for p are not presented when there were fewer than five patients in a category; p values of less than 0.05 were considered statistically significant. Results Patients with metaplastic breast carcinoma were slightly younger than patients with ductal carcinoma (p = 0.048). The median age for the patients with metaplastic carcinoma was 46 years (range, years) compared with 53 years (range, years) for those with ductal carcinoma. The median size of metaplastic carcinomas was 3 cm (range, 1 10 cm) compared with 2.4 cm (range, 1 5 cm) for ductal carcinomas. Thirty-seven (86%) of 43 patients with metaplastic carcinoma presented with a palpable mass, compared with 34 (79%) of 43 with ductal carcinoma. Table 1 describes the clinical and tumor characteristics of the 43 patients with metaplastic breast carcinomas. Most metaplastic breast carcinomas were hormone receptor negative (estrogen receptor and progesterone receptor) (39/40, 97.5% and 33/38, 86.8%, respectively) and HER2/neu-negative (22/23, 95.7%; unknown in 20/43, 46.5%) compared with estrogen receptor and progesterone receptor negativity rates of 25% (10/40) and 47.5% (19/40), re- JR:189, December

3 Yang et al. TLE 2: Mammographic Features Ductal (n = 43) Metaplastic (n = 43) Feature No. % No. % p Mass Yes No Increased density rchitectural distortion Mammography not performed Shape Round Oval Lobular Irregular < Margin Circumscribed Indistinct Microlobulated or spiculated Density High Isodense Calcifications Yes No Morphology Coarse or amorphous Pleomorphic, heterogeneous, or linear Distribution Grouped Segmental Diffuse Note Dash ( ) indicates p values not presented when there are fewer than five patients in a category. spectively, for ductal carcinoma, and HER2/neu negativity of 73.7% (28/38; unknown in 11.6% [5/43]). xillary lymph node disease was noted in 23.7% (9/38) of patients with metaplastic carcinoma compared with 47.5% (19/40) of patients with ductal carcinoma, and axillary nodal status was unknown in 11.6% (5/43) of patients with metaplastic carcinoma compared with 7.0% (3/43) of patients with ductal carcinoma. The median follow-up was 47 months (range, months). Four (9%) of 43 patients were lost to follow-up, 27 (62.7%) are currently alive with no evidence of disease, and 12 (27.9%) died. The cause of death was breast carcinoma in 11 patients and concurrent lymphoma in one. total of 17 recurrences were seen in 17 patients. Four of the recurrences were local, 12 were distant, and one was combined local and distant. The most common site of metastatic disease was the lung (n = 10), followed by liver and bone. The median time to recurrence was 12.5 months. Table 2 shows the mammography characteristics of metaplastic carcinomas. Mammography was not performed in three patients with metaplastic carcinoma. Mammographically, metaplastic carcinomas were more frequently round, oval, or lobular (Figs. 1 5) (84% vs 26%, p < ) and more frequently showed circumscribed margins (Figs. 1 and 3) (15% vs 1%, p = 0.001) than ductal carcinomas. Calcifications were more frequently noted in ductal (51% vs 25%, p = 0.02) than in metaplastic carcinomas (Fig. 1) and were more frequently linear, pleomorphic, or heterogeneous in morphology (40% vs 77%, p = 0.06) in ductal than in metaplastic carcinomas. Consistent with the size distribution of tumors, overlying skin thickening (13%) was relatively infrequent in this study because only three tumors were larger than 5 cm. Table 3 shows the sonographic characteristics of ductal and metaplastic carcinomas. Sonography was performed in 39 and 33 patients with invasive ductal and metaplastic carcinoma, respectively. typical malignant irregular shape was noted in 69% of ductal versus 27% of metaplastic carcinomas (p = 0.001). ngular margins were present in 49% of ductal versus 9% of metaplastic carcinomas. Circumscribed and indistinct margins were more frequent in metaplastic than in ductal carcinomas (55% vs 18%) (Fig. 4). Posterior acoustic shadowing was noted in 49% of ductal versus 9% of metaplastic carcinomas, and posterior acoustic enhancement (Figs. 2, 3, and 5) was noted in 18% of ductal versus 67% of metaplastic carcinomas (p < ). Discussion Metaplastic carcinomas displayed more benign mammographic features in our series and typically showed a round or oval mass with circumscribed margins compared with invasive ductal carcinomas, which typically showed an irregular shape and spiculated margins. significantly lower frequency of malignant-type pleomorphic, linear microcalcifications was observed in metaplastic than in ductal carcinomas. These findings concur with previous publications involving fewer patients [16 20]. Sonographically, metaplastic carcinomas showed more benign features characterized by an oval, round, or lobular solid hypoechoic mass with circumscribed or indistinct margins in our series when compared with ductal carcinomas. Metaplastic carcinomas frequently showed posterior acoustic enhancement, whereas posterior acoustic shadowing was more frequently associated with invasive ductal carcinomas. These combined mammographic and sonographic features may lead to misinterpretation of metaplastic carcinoma as a probably benign or a I-RDS category 3 lesion [11, 12], which could potentially result in delayed diagnosis of carcinoma even if appropriately categorized and evaluated. The differential diagnosis of an oval, round, or lobular solid mass with circumscribed margins includes the triad of circumscribed carcinomas: medullary, mucinous, and papillary carcinomas [21 25]. Recent publications have de JR:189, December 2007

4 Metaplastic Versus Invasive Ductal reast Carcinoma Fig. 1 Right mediolateral oblique mammogram shows lobular mass with internal coarse calcifications (arrows) in 66-year-old woman who presented with palpable mass. Final pathology showed invasive metaplastic carcinoma with malignant chondroosseous neoplasm. Fig year-old woman with palpable mass., Craniocaudad right mammogram shows high-density round mass with partially indistinct margins (long arrows) and partially circumscribed margins (short arrows)., Transverse sonogram shows oval mixed solid and cystic mass (long arrow) with partially indistinct margins showing posterior acoustic enhancement (short arrows). Final pathology revealed spindle cell carcinoma most consistent with sarcomatoid metaplastic carcinoma. Fig year-old woman with palpable mass., Left mediolateral oblique mammogram shows oval high-density mass with circumscribed margins (arrow)., Transverse sonogram of same breast shows solid oval mass with microlobulated margins (long arrow) and posterior acoustic enhancement (short arrows). Final pathology showed high-grade invasive sarcomatoid carcinoma. Fig year-old woman with palpable mass in left breast., Left mediolateral oblique mammogram shows highdensity lobular mass (long arrows) with associated left axillary adenopathy (short arrows)., Color Doppler transverse sonogram shows lobular circumscribed solid hypoechoic and hypervascular mass (arrows). Final pathology revealed high-grade invasive squamous cell carcinoma. JR:189, December

5 Yang et al. Fig year-old woman with palpable mass in left breast., Left craniocaudad mammogram shows high-density round mass with partially spiculated margins (arrows). Overlying marker denotes palpable mass., Transverse sonogram shows irregular mixed solid and cystic mass (long arrow) with mild posterior acoustic enhancement (short arrows). Final pathology revealed sarcomatoid carcinoma with osseous metaplasia. TLE 3: Sonographic Features Ductal (n = 39) Metaplastic (n =33) Feature No. % No. % p Shape Oval or round Lobular Irregular Margin Circumscribed Indistinct Microlobulated ngular Cystic and solid Solid Solid and cystic Posterior acoustic phenomena Enhancement None Shadowing < Calcifications No Yes Vascularity No Yes Not available xillary lymph nodes Negative Positive Note Dash ( ) indicates p values not presented when there are fewer than five patients in a category. scribed significantly more circumscribed malignant noncalcified mammographic masses representing high-grade invasive carcinomas in patients who have gene mutations than in the control group [26, 27]. Metaplastic breast carcinomas presented at a younger age than ductal carcinomas. However, this comparison of age between the two groups (metaplastic and ductal carcinoma) is limited by selection bias because the 43 matched invasive ductal carcinomas represent a small fraction ( 0.2%) of the total number of ductal carcinomas ( 21,037) diagnosed during this period. Some studies have found tumor size to be the most important determinant of patient outcome, and tumors greater than 5 cm have the poorest prognosis [2, 28]. The presenting symptom is usually a palpable mass, as was the case in our series, in which 86% of patients presented with a palpable breast mass. Metaplastic breast carcinoma displays a biologic behavior different from that of typical invasive ductal carcinoma. Metaplastic tumors tend to be large at presentation, to be hormone receptor and HER2/neu-negative, and to have a low incidence of regional lymph node involvement [5, 29]. The number of patients with documented axillary nodal metastasis at the time of axillary nodal dissection was 23.7% in our series. This figure is within the reported range for metaplastic breast carcinoma [30, 31] but is still lower than the expected yield for invasive ductal carcinoma [32] and the percentage of axillary nodal involvement of 48.8% in the control group of ductal carcinomas in this study. Metaplastic carcinomas have a greater tendency for early hematogenous dissemination. These tumors have uncertain prognostic significance. It has been suggested that metaplastic carcinomas have a worse prognosis than typical ductal carcinomas [9, 33], and that duration of symptoms, TNM stage, tumor size, and axillary nodal status are significant prognostic factors of survival [34]. We postulate that the poorer outlook associated with metaplastic carcinomas is the result of a difference in biology that reflects more the sarcomatoid than the epithelial spectrum of tumor behavior; the sarcomatoid spectrum is associated with higher hematogenous and lower lymphatic spread and higher hormone and HER2/neu receptor negativity. It is therefore incumbent on the breast imager to be aware of more-benign-appearing mammographic and sonographic lesions that can represent aggressive malignancies with poorer prognosis and overall survival [9, 33, 34]. wareness of this overlap in the imaging findings between metaplastic carcinomas, in JR:189, December 2007

6 Metaplastic Versus Invasive Ductal reast Carcinoma vasive carcinomas in patients who have oncologic genes, and probably benign lesions should prompt timely biopsy and ancillary testing of the histologic specimens. References 1. Wargotz ES, Norris HJ. Metaplastic carcinoma of the breast. I. Matrix-producing carcinoma. Hum Pathol 1989; 20: Wargotz ES, Norris HJ. Metaplastic carcinoma of the breast. IV. Squamous cell carcinoma of ductal origin. Cancer 1990; 65: Wargotz ES, Norris HJ. Metaplastic carcinoma of the breast. III. Carcinosarcoma. Cancer 1989; 64: Wargotz ES, Deos PH, Norris HJ. Metaplastic carcinoma of the breast. II. Spindle cell carcinoma. Hum Pathol 1989; 20: Oberman H. Metaplastic carcinoma of the breast: a clinicopathologic study of 29 patients. m J Surg Pathol 1987; 11: Wargotz ES, Norris HJ. Metaplastic carcinomas of the breast. V. Metaplastic carcinoma with osteoclastic giant cells. Hum Pathol 1990; 21: Gobbi H, Simpson JF, orowsky, Jensen R, Page DL. Metaplastic breast tumors with a dominant fibromatosis-like phenotype have a high risk of local recurrence. Cancer 1999; 85: Sneige N, Yaziji H, Mandavilli SR, et al. Low-grade (fibromatosis-like) spindle cell carcinoma of the breast. m J Surg Pathol 2001; 25: Luini, guilar M, Gatti G, et al. Metaplastic carcinoma of the breast, an unusual disease with worse prognosis: the experience of the European Institute of Oncology and review of the literature. reast Cancer Res Treat 2007; 101: Epub 2006 Sep Greene FL, Page DL, Fleming ID, et al. JCC cancer staging manual, 6th ed. New York, NY: Springer-Verlag, 2002: merican College of Radiology. CR I-RDS: mammography, 4th ed. In: CR reast Imaging Reporting and Data System: breast imaging atlas. FOR YOUR INFORMTION This article is available for CME credit. See for more information. Reston, V: merican College of Radiology, merican College of Radiology. CR I-RDS: ultrasound. In: CR reast Imaging Reporting and Data System: breast imaging atlas. Reston, V: merican College of Radiology, Mendelson E, erg W, Merritt CR. Toward a standardized breast ultrasound lexicon, I-RDS: ultrasound. Semin Roentgenol 2001; 36: Yang WT, huja, Tang, Suen M, King W, Metreweli C. High-resolution sonographic detection of axillary lymph node metastases in breast cancer. J Ultrasound Med 1996; 16: Vlastos G, Fornage D, Mirza NQ, et al. The correlation of axillary ultrasonography with histologic breast cancer downstaging after induction chemotherapy. m J Surg 2000; 179: Patterson SK, Tworek J, Roubidoux M, Helvie M, Oberman H. Metaplastic carcinoma of the breast: mammographic appearances with pathologic correlation. JR 1997; 169: Samuels TH, Miller N, Manchul L, DeFreitas G, Panzarella T. Squamous cell carcinoma of the breast. Can ssoc Radiol J 1996; 47: renner RJ, Turner RR, Schiller V, rndt RD, Giuliano. Metaplastic carcinoma of the breast: report of three cases. Cancer 1998; 82: Park JM, Han K, Moon WK, et al. Metaplastic carcinoma of the breast: mammographic and sonographic findings. J Clin Ultrasound 2000; 28: Günhan-ilgen I, Memis, Ustün EE, Zekioglu O, Ozdemir N. Metaplastic carcinoma of the breast: clinical, mammographic, and sonographic findings with histopathologic correlation. JR 2002; 178: Cardenosa G, Doudna C, Eklund GW. Mucinous (colloid) breast cancer: clinical and mammographic findings in 10 patients. JR 1994; 162: Conant EF, Dillon RL, Palazzo J, Ehrlich CM, Feig S. Imaging findings in mucin-containing carcinomas of the breast: correlation with pathologic features. JR 1994; 163: Chopra S, Evans J, Pinder SE, et al. Pure mucinous breast cancer: mammographic and ultrasound findings. Clin Radiol 1996; 51: Meyer JE, min E, Lindfors KK, Lipman JC, Stomper PC, Genest D. Medullary carcinoma of the breast: mammographic and US appearance. Radiology 1989; 170: Schneider J. Invasive papillary breast carcinoma: mammographic and sonographic appearance. Radiology 1989; 171: Kaas R, Kroger R, Hendriks JHCL, et al. The significance of circumscribed malignant mammographic masses in the surveillance of RC1/2 gene mutation carriers. Eur Radiol 2004; 14: Rhiem K, Flucke U, Schmutzler RK. RC1-associated breast carcinomas frequently present with benign sonographic features. (letter) JR 2006; 186:E11 E Gibson GR, Qian D, Ku JK, Lai LL. Metaplastic breast cancer: clinical features and outcomes. m Surg 2005; 71: Tse GM, Tan PH, Putti TC, Lui PC, Chaiwun, Law K. Metaplastic carcinoma of the breast: a clinicopathological review. J Clin Pathol 2006; 59: Epub 2006 Feb eatty JD, twood M, Tickman R, Reiner M. Metaplastic breast cancer: clinical significance. m J Surg 2006; 191: Gutman H, Pollock RE, Janjan N, Johnston D. iologic distinctions and therapeutic implications of sarcomatoid metaplasia of epithelial carcinoma of the breast. J m Coll Surg 1995; 180: Gloeckler Ries L, Reichman ME, Lewis DR, Hankey F, Edwards K. Cancer survival and incidence from the Surveillance, Epidemiology, and End Results (SEER) program. Oncologist 2003; 8: Rayson D, djei, Suman VJ, Wold LE, Ingle JN. Metaplastic breast cancer: prognosis and response to systemic therapy. nn Oncol 1999; 10: Chao TC, Wang CS, Chen SC, Chen MF. Metaplastic carcinoma of the breast. J Surg Oncol 1999; 71: JR:189, December