Influence of the imaging method and histopathological features on preoperative size prediction of ductal carcinoma in situ (DCIS) of the breast

Introduction

Ductal carcinoma in situ (DCIS) of the breast is a noninvasive malignant tumor located in the mammary ducts. In a DCIS neoplasm, the proliferation of malignant cells is limited only to the epithelium without invasion into the periductal stromal tissue [1, 2]. In recent studies, DCIS was diagnosed in 15% to 30% of all breast carcinomas. In between 25% and 56% of clinically occult breast cancers, DCIS was detected mammographically [1, 2]. DCIS is regarded as a potential precursor of invasive carcinoma and a risk factor for the development of cancer in the same or contralateral breast [3, 4].

Mammography is a very useful method for the detection of DCIS, however, it has poor accuracy in cases which lack microcalcifications [5, 6]. In mammography, 62-98% of DCIS are recognised by microcalcifications, whereas only 2-23% of DCIS are associated with other symptoms: focal mass, asymmetric density or architectural disorder [7]. However, a large number of DCIS (6-23%) tumors are not detectable by mammography [8, 9].

The role of ultrasonography in DCIS detection is still under discussion. Some authors have proposed a limited application of ultrasound imaging in the diagnosis of DCIS, as it can be a particularly useful tool in the evaluation of non-calcified DCIS, not only in detecting the lesion but also in evaluating its size [10]. Furthermore, ultrasound imaging enables the detection of mammographically occult DCIS in patients with dense breasts [11].

Percutaneous biopsy is the most effective technique for preoperative breast tumor diagnosis. While core needle biopsy is recommended as a more accurate ‘gold standard’ in this regard, ultrasound-guided fine needle aspiration biopsy is still in use due to its availability and for economic reasons [13, 14].

Breast-conserving surgery has become the standard method of treatment for DCIS. However, some patients require more than one surgical procedure to achieve clear histological margins because of inadequate excision [1]. Therefore, an exact description of DCIS size is essential to avoid recurrence.

The aim of the present study was to perform a retrospective evaluation of the accuracy of mammography and ultrasonography in predicting preoperative DCIS size compared to final histopathologic measurement of the extent of DCIS. The impact of other related clinical and histopathological factors was also investigated.

Material and methods

Patients



The patients’ age ranged from 48 to 68 years (average age 57.39 ±10.61 years). All patients were in the menopausal period.

The patients whose postoperative histopathological diagnosis revealed neoplasms other than DCIS cancer were excluded from the analysis. Other exclusion criteria were previous history of any other malignancy, previous breast surgery, family history of breast malignancy and premenopausal age.

The analyzed patients were divided into two groups. Group 1 consisted of 19 female patients with a palpable breast tumor, while group 2 comprised 13 female individuals with no palpable breast lesion: the breast tumors in group 2 being detected during mammography screening. All patients were preoperatively diagnosed with either core needle biopsy or ultrasound-guided fine needle biopsy.



Imaging data



The following mammographic and ultrasound features were analyzed: size, presence of calcifications, mammographic density, parenchymal pattern and BIRADS classification.



Histopathological assessment



The histopathological results were analyzed with particular attention being paid to the size, grade and surgical margin. A histological margin of 10 mm was required. Cases in which this was not achieved during the first operation proceeded to reexcision or mastectomy.



Statistical analysis



The relationships between age, mammographic and histopathological size, density and grade were evaluated using the χ2 test. All other variables were analysed using ANOVA.

Results

Both study groups were comparable according to demographic characteristics. The mean age of the patients did not differ significantly between the groups. It was found to be 61 ±10.2 and 63 ±9.5 for group 1 and 2, respectively.

The imaging features are presented in Tables I and II, and the pathological features in Tables III and IV. In no patient was the mammographical or ultrasound extent larger than histopathological extent. In group 1, mammography size was smaller than histopathological size in 14 patients (73.68%), while they were equal in 5 cases (26.31%). However, in group 2, mammography size was the same as histopathological size in 5 cases (38.46%), but was smaller in 8 individuals (61.53%).

In group 1, ultrasound measurement indicated that tumor size was smaller than the histopathological diameter in 10 patients (52.63%), but identical in 9 cases (47.36%). However, in group 2, the ultrasound diameter was smaller than the histopathological size in 8 cases (61.53%), and was identical in 5 patients (38.46%). The above mentioned percentage schedule of imaging methods size prediction between groups 1 and 2 is statistically insignificant.

The underestimation appeared in patients with the tumor size < 1.5 cm and < 1 cm in groups 1 and 2, respectively: this correlation is statistically significant (p < 0.05).

The degree of the underestimations for both imaging methods compared to pathological size were identical, with a value of 1.1 ±0.9 cm and 1.2 ±1.1 cm for groups 1 and 2, respectively.

Furthermore, in both study groups, as final histopathologic size decreased, the degree of underestimation increased (p < 0.05).

Necrosis, calcification, density, BIRADS scale, histological grade, surgical margins or biopsy result were not found to have any association with the differences in tumor size evaluated by the imaging method (mammography, ultrasound) and pathologic measurement.

Insufficient surgical margins were noted only in cases where the true histopathological diameter of the tumor was underestimated by mammography and ultrasonography.

Discussion

Little data exist regarding the accuracy of imaging methods in predicting DCIS size. Chakrabarti et al. report mammography to be 73% accurate in predicting histological extent within 10 mm [15]. However, in contrast to the present study, the authors found the pathological size to be underestimated by > 10 mm in 17.2% of their patients and overestimated by > 10 mm in 10.2%. The overestimation was explained by collecting the tumor sample during core needle biopsy and the presence of benign calcifications.

The mammographic size is currently the standard method for estimating pathological size preoperatively, but underestimation is believed to occur in approximately 15-20% of cases [16, 17]. However, our study showed the tumor to be underestimated in 73.68%. This difference may result from the limited number of patients in our study and the use of retrospective analysis.

Microcalcification is a very characteristic radiological feature of DCIS, present in 85-90% of cases. Calcifications occur in high grade invasive carcinoma as well as in benign lesions [16]. Due to the potential inaccuracies caused by calcification, and the fact that some cases of DCIS are mammographically occult, ultrasonography may be considered for the detection of DCIS not associated with microcalcifications [17]. Typical DCIS ultrasound features are mass without microcalcifications, mass with microcalcifications, isolated microcalcifications, ductal changes and dimensions over 20 mm [18]. DCIS ultrasonography is the most accurate in the event of the co-occurrence of the above mentioned features [18].

In our study, all patients in whom DCIS extent was underestimated required a further surgical procedure. The most important factor ensuring a complete excision is the size of the excision margin taken. Some studies have shown that the percentage of patients with a complete excision is 66% when a 10-mm excision margin is taken, compared to 83% and 89% for a 20-30 mm excision margin [15, 19].

Other studies have demonstrated that the use of a large margin can reduce the effect of underestimating the pathological size by mammography or ultrasonography. Several studies have also confirmed that larger margin widths are associated with a reduced risk of local recurrence [20-22].

Although most authors report that larger mammography size is associated with underestimation [15, 21], the present study does not confirm this. In our series, the disparity between the results of the imaging and pathological tests was correlated with final pathological diameter.

Holland et al. demonstrate a greater disparity between mammographic and pathological size for cribriform and micropapillary histological type DCIS (47% underestimation > 20 mm) vs. comedo DCIS (16% overestimation by > 20 mm) [19].

Magnetic resonance mammography seems to be a method with higher accuracy in DCIS detection compared to mammography and ultrasonography. The prospective study on women with an increased risk of breast cancer showed that magnetic resonance mammography detects DCIS in 15% of patients without any lesions in mammography and ultrasonography [23].

However, final pathological size depends on the histopathological method used in the laboratory. Faverly et al. showed that the choice of the procedure used in taking sections has an influence on the diagnosis of multifocal or continuous intraductal growth, as does the differentiation of DCIS: well-differentiated lesions are usually multifocal [16]. Moreover, the histopathological diagnosis is impeded by coexisting DCIS with other breast neoplasms, especially Paget disease. Marczyk et al. described similar proliferative activity of malignant cells, measured as Ki-67 antigen expression, in Paget disease and DCIS. Additionally, DCIS presented overexpression of c-erbB2 as well as Paget disease [24].

Except for medical reasons, the high accuracy and quality of the imaging method in breast cancer diagnosis is important due to encouraging women to participate in the breast cancer screening programme [25].

Conclusions

The present study demonstrates that smaller DCIS pathological size is associated with greater discrepancies in imaging method size prediction. No other preoperative clinical and pathological factors were identified.

Acknowledgements

The authors thank Ed Lowczowski, language consultant.



The authors declare no conflict of interest.

References

1. Alvarado R, Lari SA, Roses RE, et al. Biology, treatment, and outcome in very young and older women with DCIS. Ann Surg Oncol 2012; 19: 3777-84.

2. Nekhlyudov L, Habel LA, Achacoso N, et al. Ten-year risk of diagnostic mammograms and invasive breast procedures after breast-conserving surgery for DCIS. J Natl Cancer Inst 2012; 104: 614-21.

3. Burstein HJ, Polyak K, Wong JS et al. Ductal carcinoma in situ of the breast. N Engl J Med 2004; 350: 1430-41.

4. Viehweg P, Lampe D, Buchmann J, Heywang-Köbrunner SH. In situ and minimally invasive breast cancer: morphologic and kinetic features on contrast-enhanced MR imaging. MAGMA 2000; 11: 129-37.

5. Dershaw DD, Abramson A, Kinne DW. Ductal carcinoma in situ: mammographic findings and clinical implications. Radiology 1989; 170: 411-5.

6. Lee YS, Mathew J, Dogan BE, et al. Imaging features of micropapillary DCIS: correlation with clinical and histopathological findings. Acad Radiol 2011; 18: 797-803.

7. Saito M, Matsuzaki M, Sakuma T, et al. Clinicopathological study of non-palpable familial breast cancer detected by screening mammography and diagnosed as DCIS. Breast Cancer (in press).

8. Holland R, Peterse JL, Millis RR, et al. Ductal carcinoma in situ: a proposal for a new classification. Semin Diagn Pathol 1994; 11: 167-80.

9. Moon WK, Im JG, Koh YH, et al. US of mammographically detected clustered microcalcifications. Radiology 2000; 217: 849-54.

10. Moon WK, Myung JS, Lee YJ, et al. US of ductal carcinoma in situ. Radiographics 2002; 22: 269-81.

11. Kolb TM, Lichy J, Newhouse JH. Occult cancer in women with dense breasts: detection with screening US diagnostic yield and tumor characteristics. Radiology 1998; 207: 191-9.

12. Londero V, Zuiani C, Furlan A, et al. Role of ultrasound and sonographically guided core biopsy in the diagnostic evaluation of ductal carcinoma in situ (DCIS) of the breast. Radiol Med 2007; 112: 863-76.

13. Parker SH, Jobe WE, Dennis MA, et al. US-guided automated large-core breast biopsy. Radiology 1993; 187: 507-11.

14. Philpotts LE, Hooley RJ, Lee CH. Comparison of automated versus vacuum-assisted biopsy methods for sonographically guided core biopsy of the breast. AJR Am J Roentgenol 2003; 180: 347-51.

15. Chakrabarti J, Evans AJ, James J, et al. Accuracy of mammography in predicting histological extent of ductal carcinoma in situ (DCIS). Eur J Surg Oncol 2006; 32: 1089-92.

16. Ikeda O, Nishimura R, Miyayama H, et al. Magnetic resonance evaluation of the presence of an extensive intraductal component in breast cancer. Acta Radiol 2004; 45: 721-5.

17. Evans AJ, Wilson AR, Burrell HC, et al. Mammographic features of ductal carcinoma in situ (DCIS) present on previous mammography. Clin Radiol 1999; 54: 644-6.

18. Zuiani C, Francescutti GE, Londero V, et al. Ductal carcinoma in situ: is there a role for MRI? J Exp Clin Cancer Res 2002; 21: 89-95.

19. Holland R, Hendriks JH. Microcalcifications associated with ductal carcinoma in situ: mammographic-pathologic correlation. Semin Diagn Pathol 1994; 11: 181-92.

20. Coombs JH, Hubbard E, Hudson K, et al. Ductal carcinoma in situ of the breast: correlation of pathologic and mammographic features with extent of disease. Am Surg 1997; 63: 1079-83.

21. De Roos MA, Pijnappel RM, Post WJ, et al. Correlation between imaging and pathology in ductal carcinoma in situ of the breast. World J Surg Oncol 2004; 12: 2-4.

22. Faverly DR, Burgers L, Bult P, Holland R. Three dimensional imaging of mammary ductal carcinoma in situ: clinical implications. Semin Diagn Pathol 1994; 11: 193-8.

23. Popiela TJ, Kibil W, Herman-Sucharska I, Urbanik A. The use of magnetic resonance mammography in women at increased risk for developing breast cancer. Wideochir Inne Tech Malo Inwazyjne 2013; 8: 55-62.

24. Marczyk E, Kruczak A, Ambicka A, et al. The routine immunohistochemical evaluation in paget disease of the nipple. Pol J Pathol 2011; 62: 229-35.

25. Pietrzak P, Godlewski D, Adamczak M. Breast cancer screening program attitudes among women from Wielkopolska. The report from Focus Group Interview (FGI) studies. Wspolczesna Onkol 2011; 15: 415-6.