SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References

The present study retrospectively evaluated the mammographic findings of 606 Japanese women with breast cancer (median age 50 years; range 27–89 years) and correlated them with histopathological characteristics. Mammographic findings were evaluated with an emphasis on mass shape, margin, density, calcification, and the presence of architectural distortion; these findings were correlated with histopathological characteristics such as intrinsic subtype, histological grade, lymphovascular invasion, and the Ki-67 labeling index. An irregular mass shape and masses with a spiculated margin were significantly higher in the group of patients with luminal A breast cancer than in patients with masses that were lobular or round, or in tumors with an indistinct or microlobulated periphery (= 0.017, = 0.024, < 0.001, and = 0.001, respectively). Irregular mass shape and spiculated periphery were significantly lower in patients with Grade 3 cancer (< 0.001 for both). In terms of lymphovascular invasion, there were significant differences between oval and irregular or round mass shape (= 0.008 and = 0.034), between tumors with a microlobulated and indistinct periphery (= 0.014), between tumors with a punctate and amorphous or pleomorphic calcification shape (= 0.030 and 0.038), and between the presence and absence of architectural distortion (= 0.027). Equivalent or low-density masses were also higher in Grade 1 breast cancers (= 0.007). There were significant differences in the Ki-67 labeling index between irregular and lobular or round tumors (< 0.001 and = 0.014), as well as between spiculated and indistinct or microlobulated tumors (< 0.001 for both). Significant differences were noted in the mammographic features of different primary breast cancer subtypes. These proposed mammographic diagnostic criteria based on biological characteristics may contribute to a more accurate prediction of biological behavior of breast malignancies. (Cancer Sci 2011; 102: 2179–2185)

The incidence of breast cancer has increased worldwide, which is considered due, in part, to mass screening programs resulting in the discovery of clinically occult breast lesions. Mammographic screening has been demonstrated to reduce breast cancer mortality in both Western and Oriental populations.(1) This mortality may be as great as 63% in women attending for screening.(2) Therefore, million of mammographic examinations are being performed yearly worldwide, and mammography has become the gold standard for detecting breast disorders. Strict attention to high-quality interpretation is required for successful of a mammographic diagnosis. Thus, it is important to establish an accurate diagnostic system for mammography.

Traditionally, prognostic determinations are made mainly on the basis of pathological information, including histological grade and lymphovascular invasion.(3–5) In addition to histological information, the status of molecular markers that have prognostic and predictive value can contribute to the selection of an optimal treatment strategy. These markers include estrogen receptor (ER), progesterone receptor (PgR), and human epidermal growth factor receptor 2 (HER2) and determining the status of these markers has become standard practice in the management of breast cancer because ER and HER2 positivity can predict a patient’s response to endocrine therapy or targeted therapy with monoclonal antibodies directed against HER2.(6) In addition, the St Gallen international expert consensus meeting on the primary treatment of early breast cancer reported that features indicative of increased risk of recurrence, thus indirectly supporting the addition of chemotherapy to endocrine therapy, include lower expression of steroid hormone receptors, Grade 3 tumors, high proliferation (as measured by conventional or multigene assays), and extensive peritumoral vascular invasion.(7) However, these therapeutic determinations have been derived mainly from pathological information.

The appearance of tumors on mammograms has a generally good correlation with subsequent histological characteristics. For example, microcalcification is the hallmark of ductal carcinoma in situ;(8) spiculation is significantly correlated with low histologic grade; and ill-defined masses and microcalcifications are features of high-grade tumors.(8) Accurate correlation of mammographic findings with corresponding histopathologic features is considered one of the most important aspects of mammographic evaluation. Full histopathological information, including histological grades and intrinsic subtypes, is determined correctly after surgery.(9) Therefore, the purpose of the present study was to retrospectively evaluate mammographic findings and to compare the histopathological characteristics of the different tumors (i.e. intrinsic subtype, histological grade, lymphovascular invasion, and Ki-67 labeling index) in Japanese patients.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References

Patients.  The mammographic and histopathologic features of 606 Japanese breast cancer patients who had undergone surgery at Tohoku University Hospital, Sendai, between January 2005 and June 2010 were reviewed retrospectively. All patients provided informed consent and the study protocol was approved by the Ethics Committee at Tohoku University Graduate School of Medicine. The median age of the patients was 50 years (range 27–89 years).

Imaging devices and breast tissue specimens.  All mammographic examinations were performed with dedicated machines. Analog mammographic examinations were performed with one unit (MAMMOMAT 3000 Nova; Siemens, Erlangen, Germany) using a screen–film technique (Min-R 2000 Min-R EV; Kodak Health Imaging, Rochester, NY, USA). Digital mammograms were acquired by using a system with an amorphous selenium DirectRay digital detector (LOARD Selenia; Hologic, Waltham, MA, USA). The system was connected to a viewing monitor (MammoRead; TOYO, Tokyo, Japan).

Samples were stained using H&E. Histochemical and immunohistochemical analyses for ER, HER2, and Ki-67 were performed at the Department of Pathology, Tohoku University Hospital. Surgical specimens were fixed in 10% formaldehyde solution and cut into serial 5-mm slices, embedded in paraffin, cut into 4-μm sections, and placed on the glue-coated glass slides. We used the avidin–streptavidin immunoperoxidase method using the clone 6F11 antibody (Ventana, Tucson, AZ, USA) in an automated immunostainer (Benchmark System; Ventana). A standardized immunohistochemistry kit (HercepTest for Immunoenzymatic Staining; Dako, Copenhagen, Denmark) was used for HER2 staining. The Ki-67 labeling index was determined using an MIB-1 monoclonal antibody (code M7240; Dako). Both H&E and immunohistochemical staining were performed by a single experienced technician. Positive controls for ER and HER2 were breast carcinoma, whereas negative controls for immunostaining were hepatocellular carcinoma.

Imaging and histopathological analyses.  Two experienced breast surgeons independently evaluated the mammographic findings. These two investigators were blinded as to the histopathological diagnosis and the clinical outcome of the patients. If there were discrepancies in the interpretation of the mammograms, a final decision was reached using consensus evaluations from eight experienced breast surgeons and radiologists. Mammographic findings were subsequently analyzed according to the American College of Radiology Breast Imaging Reporting and Data System (BI-RADS).(10) The presence of a mass, calcifications, focal asymmetric density (FAD), and architectural distortion were each recorded. Figure 1 shows representative mammographic findings. Mass shape was tentatively classified into round, oval, lobular, and irregular. Margins were classified as microlobulated, indistinct, spiculated, and “other”. Density was classified into high, equivalent, or low. Calcification shape was tentatively classified into punctate, amorphous, pleomorphic, and linear. Finally, FAD was classified as with or without architectural distortion.

image

Figure 1.  Representative mammographic findings in breast carcinoma cases. (a) Round mass shape, microlobulated margin and intermediate density mass. (b) Lobular mass shape, indistinct margin, and high density mass. (c) Irregular mass shape, spiculated margin, and high-density mass. (d) Amorphous calcifications. (e) Pleomorphic or linear calcifications. (f) The presence of architectural distortion.

Download figure to PowerPoint

Two experienced pathologists independently evaluated surgical specimens. Histopathological evaluations were based on the World Health Organization (WHO) histological classification of tumors of breast and Rosen’s Breast Pathology.(11,12) The presence of ER was determined by nuclear staining and was graded from 0 to 8 using the Allred score, with positivity defined as a score of ≥3.(13) With regard to HER2 evaluation, membranous staining was graded as 0–1+, 2+, and 3+.(14) Samples scored as 2+ were subjected to FISH to calculate the gene copy ratio of HER2 to CEP17 (PathVysion HER2 DNA Probe kit; Abbott, Chicago, IL, USA). Positivity was defined as a HER2:CEP17 signal ratio (FISH score) >2.2.(14) Histological grades were assessed according to the criteria of Elston and Ellis.(4) The Ki-67 immunoreactivity was evaluated by examining high-power fields and counting 1000 tumor cells in the hot spots.(15) In addition, the presence or absence of lymphovascular invasion was determined according to Rosen’s Breast Pathology.(12) Intrinsic subtypes were classified according to the St Gallen international expert consensus on the primary therapy of early breast cancer 2011(16) as follows: luminal A was ER and/or PgR positive, HER2 negative, and Ki-67 low (<14%); luminal B was either ER and/or PgR positive, HER2 negative and Ki-67 high, or ER and/or PgR positive, any Ki-67, and HER2 overexpressed or amplified; the HER type was HER2 overexpressed or amplified and ER and PgR absent; and triple negative was ER, PgR and HER2 negative.

We compared mammographic findings, including mass shape, margin, density, calcification, FAD, and architectural distortion, with the histopathological characteristics of the tumors, including intrinsic subtype, histological grade, lymphovascular invasion, and the Ki-67 labeling index.

Statistical analysis.  To compare mammographic findings with histopathological findings, multivariate analysis was used. All analyses were performed using SPSS version 10.0 (SPSS Inc., Chicago, IL, USA), with < 0.05 taken to indicate significant differences.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References

Comparison of mammographic findings with intrinsic subtype. Figure 2 summarizes the results of the numbers and ratios of each mammographic finding according to intrinsic subtype. In the luminal A group, significant differences were identified between masses that were irregular and lobular or round (= 0.017 and = 0.024), between those that had speculated and indistinct or microlobulated margins (< 0.001 and = 0.001), between those showing amorphous and pleomorphic calcification (= 0.044), and between the presence and absence of architectural distortion (= 0.002). In the HER group, significant differences were identified between masses that were irregular and oval or round (= 0.009 and < 0.001), between masses that were lobular and round (= 0.021), and between those that had spiculated and microlobulated margins (= 0.005). In the triple negative group, significant differences were identified between masses that had spiculated and indistinct margins (= 0.027), as well as between those identified as having high and equivalent or low density (= 0.027).

image

Figure 2.  Correlation between mammographic findings and intrinsic subtype: (a) mass shape, (b) margin, (c) density, (d) calcification shape, and (e) focal asymmetric density (FAD) and architectural distortion. HER, human epidermal growth factor receptor; TN, triple negative; Lum A, luminal A; Lum B, luminal B.

Download figure to PowerPoint

Comparison of mammographic findings with histological grade. Figure 3 summarizes the results of the numbers and ratios of each mammographic finding according to histological grade. There were significant differences between irregular and lobular or oval mass shape in Grade 3 (< 0.001 for all). Furthermore, in Grade 1 tumors, significant differences were found between with an indistinct and microlobulated or spiculated periphery (= 0.030 and = 0.003), between those with spiculated and indistinct or microlobulated margins (< 0.001, respectively), between those identified as high and equivalent or low density (= 0.047), and between those with a linear and amorphous calcification shape (= 0.027).

image

Figure 3.  Correlation between mammographic findings and histological grade: (a) mass shape, (b) margin, (c) density, (d) calcification shape, and (e) focal asymmetric density (FAD) and architectural distortion.

Download figure to PowerPoint

Comparison of mammographic findings with lymphovascular invasion. Figure 4 summarizes the results for the numbers and ratios of each mammographic finding according to lymphovascular invasion. There were significant differences between oval and irregular or round mass shape (= 0.008 and = 0.034), between microlobulated and indistinct periphery (= 0.014), between punctate and amorphous or pleomorphic calcification shape (= 0.030 and 0.038), and between presence and absence of architectural distortion (= 0.027).

image

Figure 4.  Correlation between mammographic findings and lymphovascular invasion: (a) mass shape, (b) margin, (c) density, (d) calcification shape, and (e) focal asymmetric density (FAD) and architectural distortion. ly−, no lymphovascular invasion; ly+, lymphovascular invasion.

Download figure to PowerPoint

Comparison of mammographic findings with the Ki-67 labeling index. Figure 5 summarizes the results of correlations between mammographic findings and the Ki-67 labeling index. The Ki-67 labeling index according to mass shape was 15.74 ± 6.21 for irregular masses, 38.82 ± 13.10 for lobular masses, 36.22 ± 15.75 for oval masses, and 37.85 ± 14.95 for round masses. According to mass periphery, the Ki-67 labeling index was 35.80 ± 28.51, 34.56 ± 29.76, 11.73 ± 10.86, and 27.50 ± 24.75 for tumors with indistinct, microlobulated, spiculated, and “other” margins, respectively. For tumors with a high and equivalent or low mass density, Ki-67 labeling index was 27.68 ± 26.75 and 13.14 ± 14.10, respectively. Tumors that showed amorphous, punctate, pleomorphic, and linear calcification had a Ki-67 labeling index of 24.55 ± 7.58, 26.00 ± 18.27, 24.68 ± 9.43, and 16.00 ± 17.23, respectively. In tumors without and with architectural distortion, the Ki-67 labeling index was 22.27 ± 8.64 and 25.02 ± 7.43, respectively. There were significant differences between irregular and lobular or round (< 0.001 and = 0.014), spiculated and indistinct or microlobulated (< 0.001 for all), and high and equivalent or low density (= 0.018) groups. A trend for a positive correlation was detected between irregular and oval mass shape, but the difference did not reach statistical significance (= 0.062). There were no significant differences according to calcification shape and the presence of architectural distortion.

image

Figure 5.  Correlation between mammographic findings and Ki-67 labeling index: (a) mass shape, (b) margin, (c) density, (d) calcification shape, and (e) focal asymmetric density (FAD) and architectural distortion.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References

Histological grade is well known to have a strong correlation with clinical outcome in patients with breast cancer.(4) Accumulating clinical evidence suggests that prognostic factors influencing breast cancer extend beyond the traditional tumor histological grade.(17) Several factors, including ER expression, HER2 status, and lymphovascular invasion, have been clearly demonstrated in recent years to contribute significantly to the management and subsequent prognosis of patients with breast cancer.(7,18) Therefore, an accurate correlation between mammographic findings and their corresponding histopathological features is considered most important in mammographic evaluation. Mammographic findings may provide insights into pathological and biological features, including tumor cell characteristics, histological grade, and cell proliferation. We attempted to determine which finding is more relevant with regard to the newly defined subtype of breast carcinoma cells. Therefore, the purpose of the present study was to evaluate the correlation between mammographic findings (e.g. mass shape, margin, density, calcification shape, FAD, and the presence of architectural distortion) with intrinsic subtype, histological grade, lymphovascular invasion, and the Ki-67 labeling index in breast cancer patients.

Several previous studies evaluated the correlation between mammographic findings and histopathological characteristics in individual patients.(8,19–21) A number of independent groups demonstrated that masses with a spiculated periphery were associated with a good outcome in patients.(19,20) Conversely, well-defined masses were associated with triple-negative breast cancer.(8,21) The results of the present study demonstrate that is a higher incidence of lower histological grade in masses with an irregular shape and/or spiculated margins, although a higher histological grade is not necessarily associated with irregular mass shape or spiculated margins. In addition, correlation of mammographic findings with the intrinsic subtype demonstrated that irregular mass shape and/or spiculated margin masses were significantly more frequently detected in luminal A breast cancers than in the other subtypes in this cohort of Japanese patients. However, oval and round mass shape and/or indistinct and microlobulated margin masses were significantly more frequently detected in triple-negative breast cancers or HER breast cancers. As for architectural distortion, the ratio of architectural distortion was significantly higher in luminal A cases and also tended to be associated with histological Grade 1. Together, these results suggest that poorly differentiated breast carcinoma cells are associated with good histological grade and luminal A subclassification. However, well-differentiated carcinoma cells are associated with adverse clinical grading and negative ER status.

Previous studies have demonstrated that these differentiations were related somewhat with adhesion factors.(22,23) Loss of adhesion factors in carcinoma cell is considered to play a role in the characteristic histological appearance of invasive carcinoma as loosely dispersed linear columns of cells and a typical discrete mass.(22) This more diffuse infiltrative pattern may explain some of the typical imaging appearances of tumors, such as spiculation and distortion.(22) In addition, adhesion factors are correlated with high histologic grade.(23) Therefore, adhesion factors may be considered to be correlated with the results of the present study in that spiculated breast cancers have a good clinical outcome and histological Grade 1. However, it is also true that numerous biological mechanisms underlying the association between the process of infiltration and histopathological characteristics remain unknown and that further investigations are required to confirm interpretation of mammography in terms of the biological and histopathologic characteristics of tumors.

To the best of our knowledge, this is the first study to compare mammographic findings with the Ki-67 labeling index and histopathological lymphovascular invasion. The results of the present study demonstrated that there was a higher incidence of a lower Ki-67 labeling index in tumors with an irregular mass shape, spiculated periphery, and equivalent or low mass density. Irregular mass shape and a spiculated periphery are well-known predictors of malignancy, but the results of the present study seem to suggest that findings of irregular shape and a spiculated periphery are relatively good prognostic predictors in terms of the Ki-67 labeling index. In addition, the results of the present study demonstrate that lymphovascular invasion was significantly greater in cases in which there was architectural distortion; however, the incidence of lymphovascular invasion was not significantly higher in spiculated masses. These results all suggest that the correlation between findings of radiological distortion and the mechanisms of lymphovascular invasion remain unknown and further investigations are required.

We also examined the correlation between mammographic calcification shape and histopathological characteristics. Previous studies have reported that triple-negative breast cancers are more likely to exhibit comedo calcifications.(8) In addition, the high frequency of comedo calcification in triple-negative breast cancers may represent a consequence of high histologic grade.(8) The presence of mammographic comedo calcification has also been reported to be associated with a poor prognosis in small screening-detected invasive cancers.(19) The results of the present study also demonstrate that non-necrotic calcifications, including amorphous and punctate calcification, are associated with a higher ratio of luminal A cases, whereas necrotic calcifications, including pleomorphic and linear calcification, were associated with a higher ratio of HER breast cancers. In addition, necrotic calcifications tended to be associated with a higher histological grade than non-necrotic calcifications. Therefore, the results suggest that the type of calcification may become a prognostic factor for breast malignancies.

We noted significant differences in the mammographic features of different primary breast cancer immunophenotypes in the present study. Stratifying the mammographic features according to immunophenotypes reveals distinct differences among cancer subtypes. However, the limitations of the present study include that fact that the study was retrospective in nature and was performed in a single institute, namely Tohoku University Hospital. Therefore, further investigations are needed, including analysis in several different institutions to further refine the new mammographic criteria. Biological and histopathological differences may result in imaging differences that may help us better understand the development of breast cancer. These proposed mammographic diagnostic criteria based on biological characteristics may contribute to a more accurate prediction of the biological behavior of breast malignancies.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References

The authors thank medical technologist Mr Masahiro Sai for his excellent technical assistance for mammography. The authors also thank medical technologist Ms Yayoi Takahashi for excellent technical assistance with the immunohistochemical staining. This work was supported, in part, by a Grant-in Aid from Kurokawa Cancer Research Foundation.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References