Regressive Change in High-Grade Ductal Carcinoma In Situ of the Breast. Jason K. Wasserman, MD, PhD, and Carlos Parra-Herran, MD, FCAP

Transcription

1 AJCP / Original Article Regressive Change in High-Grade Ductal Carcinoma In Situ of the Breast Histopathologic Spectrum and Biologic Importance Jason K. Wasserman, MD, PhD, and Carlos Parra-Herran, MD, FCAP From the Department of Pathology and Laboratory Medicine, University of Ottawa, The Ottawa Hospital and Eastern Ontario Regional Laboratory Association, Ottawa, Canada. Key Words: Breast pathology; Breast cancer; Ductal carcinoma in situ; Regression Am J Clin Pathol September 2015;144: CME/SAM ABSTRACT Objectives: High-grade ductal carcinoma in situ (HG-DCIS) of the breast often shows tumor attenuation and reactive fibrosis. These changes, previously described as regressive, have been paradoxically associated with an increased risk of invasive carcinoma. We aimed to further characterize the spectrum of the so-called regressive changes (RCs) in HG-DCIS. Methods: We reviewed 52 consecutive cases of HG-DCIS on biopsy specimens followed by excision. RCs were divided into early (stage 1) and advanced (stages 2 and 3) stages according to the degree of ductal fibrosis and tumor effacement. The presence of inflammation, hormone receptor status, and diagnosis on excision were recorded. Results: RCs were seen in 51 (98%) cases: 96%, 76.4%, and 39.2% cases showed stages 1, 2, and 3, respectively. Periductal T cells with a normal CD4/CD8 ratio were constantly seen. Advanced RCs and inflammation were more frequent in estrogen and progesterone receptor negative tumors. RCs were not associated with invasion but correlated with a larger residual HG-DCIS volume on excision. Conclusions: Regression in HG-DCIS is frequent. It may reflect a targeted immune response to certain phenotypes, mainly hormone receptor negative lesions. Nonetheless, RCs do not lead to complete tumor obliteration but correlate with aggressive tumor characteristics instead. Upon completion of this activity you will be able to: list the biologic and clinical characteristics of high-grade ductal carcinoma in situ of the breast (HG-DCIS). describe the morphologic spectrum of the so-called regressive changes in HG-DCIS. discuss the associations between HG-DCIS regression and tumor characteristics, such as hormone receptor profile, extent of disease, and diagnosis on excision. The ASCP is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The ASCP designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 Credit per article. Physicians should claim only the credit commensurate with the extent of their participation in the activity. This activity qualifies as an American Board of Pathology Maintenance of Certification Part II Self-Assessment Module. The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose. Exam is located at Ductal carcinoma in situ of the breast is a heterogeneous group of lesions with a wide range of morphologic appearances and biologic behavior. Among them, high-grade ductal carcinoma in situ (HG-DCIS) carries particular biologic and clinical importance. Compared with non high-grade lesions, HG-DCIS has a more diverse genetic and phenotypic portrait. Likewise, it carries a stronger association with recurrence and progression to invasive carcinoma. Tumor regression, defined as a continuum of changes leading to obliteration and effacement of a neoplastic population, has been described in a wide spectrum of human malignancies, including breast cancer. 1 While still poorly understood, regression is believed to represent an immune response from the host attempting to contain and even eradicate the neoplastic population. Regressive changes in mammary ductal carcinoma in situ were first described in 1934 as circumferential layers of collagen and elastic tissue Am J Clin Pathol 2015;144:

2 Wasserman and Parra-Herran / Regressive Change in HG-DCIS surrounding neoplastic epithelium and were interpreted as scars or part of a healing process. 2 They were further characterized by Chivukula et al, 3 who documented a significant association between regressive changes on biopsy specimens and invasive carcinoma on subsequent excision. Such association led researchers to hypothesize that regressive changes act as a harmful rather than a protective mechanism. Further exploration of these theories, as well as the biologic and clinical significance of regressive change in HG-DCIS, is still warranted. The aim of this study is to further characterize the phenomenon of regression in HG-DCIS by determining its frequency and distribution from early to advanced stage. We also aim to describe the relationship between regressive change and other intrinsic neoplastic features such as periductal inflammation and hormone receptor status. Finally, we seek to expand on the association between the presence of regressive change on core needle biopsy specimens and in situ or invasive carcinoma outcome on surgical excision. Materials and Methods This study was approved by the Ottawa Health Science Research Ethics Board. Study Population We searched our laboratory information system for all consecutive cases with a pathologic diagnosis of HG-DCIS on biopsy specimens, followed by surgical excision, in a 2-year period. Cases with a previous diagnosis of breast cancer or therapy targeted to the breast were excluded. Patient age and clinical indication for the biopsy were recorded in each case. Histopathologic Analysis Archival biopsy H&E-stained slides from each case were retrieved. All available material was reviewed by both authors, and the diagnosis of HG-DCIS was confirmed. The presence of necrosis and microcalcifications was recorded. Regressive changes were assessed in each in situ neoplastic gland using criteria previously described by Chivukula et al. 3 Three stages were defined: Stage 1: Minimal loss/effacement of neoplastic epithelium and mild periductal fibrosis ( 1.0 the radius of the gland) Stage 2: Significant loss/effacement of neoplastic epithelium and prominent periductal fibrosis (>1.0 the radius of the gland) Stage 3: Complete loss of neoplastic epithelium with lumen obliteration and prominent periductal fibrosis (>1.0 the radius of the gland) The relative proportion of each stage of regression was quantified as a percentage of the total of involved glands (with the sum of all three stages equaling 100%). The presence of periductal chronic inflammation in each involved gland was assessed in terms of intensity and distribution. Intensity was scored as absent, mild (<0.5 the radius of the gland), moderate ( the radius of the gland), or severe (>1.0 the radius of the gland). Likewise, distribution was scored as absent, focal (<50% of glands involved), or diffuse (>50% of glands involved). Reports from surgical excisions (lumpectomy or mastectomy) after a biopsy diagnosis of HG-DCIS were reviewed, and the final diagnosis, including the presence or absence of in situ or invasive carcinoma, was recorded. In cases with residual in situ carcinoma, the extent of disease was calculated by measuring the area of residual disease in each section multiplied by the distance between involved sections in the medial to lateral plane (volume recorded in cubic millimeters). Margin status (positive or negative and distance to the closest margin in millimeters) was also recorded. In cases with invasive carcinoma, we recorded tumor size, tumor grade, tumor stage, and status of axillary lymph nodes, if sampled. Immunohistochemistry Estrogen receptor (ER) and progesterone receptor (PR) immunohistochemical studies, from either biopsy specimens or subsequent excision material, were reviewed when available. Testing was performed using primary monoclonal antibodies against ER (clone 6F11, dilution 1:300; Vector Laboratories, Burlington, Canada) and PR (clone 1.60E, ready to use; Leica Biosystems, Concord, Canada) on 4-µm thick formalin-fixed, paraffin-embedded sections. Positive expression was defined as nuclear staining in 1% or more of tumor cells. Negative expression was defined as nuclear staining in less than 1% of tumor cells with concurrent positive staining in internal nonneoplastic tissues. Results of Her2Neu testing (immunohistochemistry or fluorescence in situ hybridization) were recorded in cases with invasive carcinoma on excision (Her2; kit name: HERCEPT test; DAKO automated stainer [DAKO, Burlington, Canada]). Immunohistochemistry for lymphocytic markers was performed in cases with paraffin-embedded material available. The panel included markers of B-cell (CD20), helper T-cell (CD4), and cytotoxic T-cell (CD8) lineages (CD20, clone MJ1, ready to use [Leica Biosystems]; CD4, clone B4B12, dilution 1:25 [DAKO]; CD8, clone 4b11, ready to use [Leica Biosystems]). Statistical Analysis Statistical analyses were performed using SPSS (for Windows 12.0; SPSS, Chicago, IL). The c 2 and Fisher exact 504 Am J Clin Pathol 2015;144:

3 AJCP / Original Article tests were used to characterize the relationship between categorical variables, as appropriate. The Mann-Whitney test was used to compare regressive change to the volume of in situ or invasive carcinoma. All statistical tests were two sided, with a P value of less than.05 considered statistically significant. Results After confirmation of initial diagnosis, a total of 52 patients with HG-DCIS on biopsy specimens were included. All patients were female, and the mean age at the time of diagnosis was 58 years (range, years). A history of calcifications on imaging was documented in 40 (77%) cases, a mass or architectural distortion in five (10%) cases, a mass containing calcifications in four (8%) cases, and non mass enhancement in three (6%) cases. Upon histopathologic analysis, comedo-type necrosis was observed in all cases, whereas microcalcifications were seen in 34 (65%) cases. Regressive Changes Are Common in HG-DCIS Review of the initial biopsy material revealed evidence of periductal fibrosis and/or tumor attenuation in 51 (98%) of 52 cases. Representative examples of each regression stage are shown in Image 1. Most cases (40 cases [78.4%]) demonstrated a combination of regression stages. Eighteen (35.2%) cases showed all three stages in different proportions, 18 cases had a combination of stages 1 and 2, and two cases showed a combination of stages 2 and 3 and stages 1 and 3, respectively. Conversely, 12 cases showed only one stage of regression (11 cases for stage 1 and one case for stage 2). The relative proportion of each stage in the total spectrum of changes varied in each case, with predominance of early stage changes: 49 (96%), 39 (76.4%), and 20 (39.2%) cases showed regression stages 1, 2, and 3, respectively. Stage 1, when present, represented on average 67.1% of the regressive change in the biopsy specimen. Stage 2, when observed, comprised on average 37% of the change. Interestingly, the relative proportion of regression stage 3 was, in comparison, less (mean, 20.4%) and never exceeded 40%. Regressive Change Is Associated With a Prominent Inflammatory Response A chronic lymphoplasmacytic inflammatory infiltrate was observed in 49 (94.2%) of 52 cases. Representative examples of inflammation associated with tumor regression are shown in Image 2. There was a significant positive correlation between both the intensity and distribution of inflammation and advanced stages of regressive change (stages 2 and 3) Table 1. Immunohistochemistry for B-cell and T-cell lymphocytic markers was performed in 20 cases. In all, the periductal inflammatory population was predominantly composed of T cells. CD4-positive cells represented a mean (SD) of 57.8% (9.8%) of the infiltrate (range, 30%-70%), and CD8-positive cells comprised a mean (SD) of 24% (5.3%) (range, 15%-35%). The mean (SD) CD4 to CD8 ratio was 2.48 (0.52) (range, ). A B-cell population, while constantly present, was less prominent; CD20-positive cells represented a mean (SD) of 18.3% (12.8%) of the infiltrate (range, 5%-55%). Advanced Regression and Inflammation Are Associated With Hormone Receptor Negative Tumors The relationship between tumor regression, inflammation, and hormone receptor status is shown in Table 2. Hormone receptor studies were available on either initial biopsy specimens or subsequent resection in 49 (94%) cases. HG-DCIS was positive for ER in 30 (61%) cases and for PR in 25 (51%) cases. HG-DCIS was negative for both hormone receptors in 19 (39%) cases. Advanced stage regressive changes were more commonly seen in hormone-negative tumors (for ER, PR, or both). This association was statistically significant for regression stage 2 and approached significance for regression stage 3. Similarly, both the intensity and distribution of the inflammatory response were greater in ER- and PRnegative tumors (either or both). Advanced Regressive Changes on Biopsy Specimens Are Associated With Larger Volume of Residual HG-DCIS on Excision Residual HG-DCIS was diagnosed, on subsequent excision, in 49 (94.2%) cases. Volume of residual disease was documented in 46 cases. When present, the mean volume of residual HG-DCIS was 49 mm 3 and ranged from 1 to 260 mm 3. Cases with only stage 1 regressive change on biopsy specimens had a lower volume of disease vs cases showing stage 2 or 3 regression (19.50 ± mm 3 vs 50 ± mm 3 ; P =.047). Margin status was negative in 45 (86.5%) cases and positive for at least one margin in five (9.6%) cases. There was no significant difference in the distance of residual HG-DCIS from the closest surgical margin between cases showing only stage 1 regression and those showing stage 2 or 3 (4.80 ± 9.44 mm vs 2.45 ± 4.98 mm; P =.513). Ductal Carcinoma In Situ Regression and Invasive Carcinoma on Excision The relationship between HG-DCIS regression and inflammation and invasive carcinoma on excision is shown Am J Clin Pathol 2015;144:

4 Wasserman and Parra-Herran / Regressive Change in HG-DCIS A B C D Image 1 Regressive changes in high-grade ductal carcinoma in situ. A, In stage 1 regression, the neoplastic cell population is either spared or only minimally depleted; surrounding stroma is fibrotic (H&E, 100). B, An inflammatory response is constantly observed in different proportions (H&E, 100). C, In stage 2 regression, the neoplastic population is significantly depleted and associated with partial obliteration of the duct by mounting fibrosis (H&E, 100). D, Stage 3 regression is characterized by effacement of carcinoma cells and complete obliteration of the duct by concentric prominent fibrosis (H&E, 100). in Table 3. Invasive ductal carcinoma was diagnosed on subsequent resection in 17 (33%) patients. The degree of regressive change, reflected on the three different stages assessed, showed no significant correlation with an invasive carcinoma diagnosis. There was no difference in the presence or degree (intensity or distribution) of inflammation between patients with and without invasive carcinoma on resection. Similarly, there was no association between regressive change and invasive tumor size, lymph node status, or pathologic tumor stage. ER and PR status in the invasive component was available in 16 of 17 cases. ER was positive in 10 (62.5%) cases and PR in eight (50%) cases. Five 506 Am J Clin Pathol 2015;144: (31.2%) cases were negative for both. There was no statistically significant correlation between the regression stage and the hormone status of the invasive tumor. Her2Neu expression studies (immunohistochemistry or fluorescence in situ hybridization) were available in 12 (71%) of 17 cases with invasive carcinoma. Of these, eight were positive for Her2Neu overexpression/amplification. Regressive changes at different stages were frequently seen in this Her2Neupositive subgroup, with seven (58%), eight (67%), and five (42%) cases demonstrating regression stages 1, 2, and 3, respectively. Although there was a trend toward more advanced regressive change in Her2Neu-positive tumors, the difference was not statistically significant.

5 AJCP / Original Article A B C D Image 2 Regressive changes and inflammation in high-grade ductal carcinoma in situ. A, Early (stage 1) regression is usually associated with mild to moderate periductal inflammation (H&E, 100). B, In more advanced lesions (stage 3), the infiltrate is more pronounced (H&E, 100). C, Well-demarcated lymphoplasmacytic aggregates associated with concentric reactive fibrosis give the appearance of a completely effaced ductal unit, in keeping with late-stage regression (H&E, 100). D, Occasionally, inflammatory infiltrates involve nonneoplastic glands (H&E, 100). Discussion In this descriptive retrospective study, we expand on the knowledge of HG-DCIS changes referred to in the literature as regressive namely, periductal fibrosis and intraductal tumor attenuation/effacement. We postulate that while these changes frequently represent partial obliteration of the in situ neoplasia, they do not seem to translate into complete tumor regression and correlate with larger volumes of in situ disease instead. Spontaneous tumor regression in cancer is defined as the partial or complete disappearance of a malignant tumor in the absence of treatment or in the presence of therapy which is considered inadequate to exert a significant influence on neoplastic disease. 4 Tumor regression is a documented yet incompletely understood and intriguing phenomenon in human cancer. It is frequent in certain malignancies, such as melanoma and neuroblastoma.5,6 Nonetheless, regression has been reported to occur in virtually all other types of cancer. The causality and underlying mechanisms of tumor regression remain unclear. It has been associated with operative trauma and subsequent disruption of vascular supply, elicitation of an immune response against malignancy, and postoperative infection.4,7,8 Spontaneous regression of breast cancer appears to be infrequent; only a few case series and reports are available in the English literature.9-12 Nonetheless, some authors Am J Clin Pathol 2015;144:

6 Wasserman and Parra-Herran / Regressive Change in HG-DCIS Table 1 Relationship Between Tumor Regressive Change in High-Grade Ductal Carcinoma In Situ and Periductal Inflammation Inflammation Regressive Change Stage 1 Regressive Change Stage 2 Regressive Change Stage 3 Present, No. Absent, No. P Value Present, No. Absent, No. P Value Present, No. Absent, No. P Value Inflammation intensity Inflammation distribution Table 2 Relationship Between High-Grade Ductal Carcinoma In Situ Regressive Changes, Inflammation, and Hormone Receptor Status Regressive Changes Positive (n = 30), No. Estrogen Receptor Progesterone Receptor Hormone Receptors (Any) Negative (n = 19), No. P Value Positive (n = 25), No. Negative (n = 24), No. P Value Positive (n = 30), No. Negative (n = 19), No. P Value Stage 1 Present Absent Stage 2 Present Absent Stage 3 Present Absent Inflammation intensity Inflammation distribution have postulated that the natural course of some screendetected breast cancers is spontaneous regression based on differences in cancer incidence between patients undergoing regular screening and controls. 13,14 The spectrum of regression has also been observed in ductal carcinoma in situ of the breast. It was first described as a healing process reflecting an immune response, directed against the neoplastic population. 2 Regressive change is associated with tumor grade and appears to be significantly more prevalent in high-grade lesions. All cases but one in our series of HG-DCIS showed evidence of regression. Likewise, a separate study found that all cases of invasive breast cancer with spontaneous regression also contained HG-DCIS. 9 The morphologic spectrum of regressive change in HG-DCIS was further studied by Chivukula et al, 3 who defined three stages of tumor regression. Periductal tumor inflammation is, as shown in our study, constantly seen in HG-DCIS with regressive changes. Inflammation is predominantly composed of T cells with a normal CD4 to CD8 ratio. Indeed, a predominant helper (CD4-positive) T-cell population is seen in earlier stages of breast cancer, including carcinoma in situ; with tumor progression, there is a subsequent increase in CD8-positive population and inversion of the CD4 to CD8 ratio. 15 Cytotoxic (CD8-positive) T cells are more frequent in invasive breast carcinoma and appear to correlate with tumor aggressiveness. 16 Finally, both helper and cytotoxic T cells have been found to be capable of killing breast carcinoma cells. 17 Therefore, the inflammatory component seen in the context of HG-DCIS regression may reflect a targeted immune response against in situ neoplasia. Mammary ductal carcinoma in situ is a heterogeneous group of lesions with a spectrum of morphologic, genetic, and biologic characteristics. Among this group, HG-DCIS appears to be different at the molecular level from non high-grade carcinomas. HG-DCIS is commonly positive 508 Am J Clin Pathol 2015;144:

7 AJCP / Original Article Table 3 Relationship Between High-Grade Ductal Carcinoma In Situ Regressive Changes and Inflammation on Biopsy Specimens and Presence of Invasive Carcinoma on Subsequent Resection Regressive Changes Invasive Carcinoma on Resection No (n = 35), No. Yes (n = 17), No. P Value Stage 1 Present Absent 2 2 Stage 2 Present Absent 9 4 Stage 3 Present Absent Inflammation intensity Inflammation distribution for Her2Neu and p53 overexpression, and expression of basal cell markers such as CK5/6, CK14, and EGFR is frequent Similarly, a significant proportion of HG-DCIS is negative for estrogen and progesterone hormone receptors. 21 As surrogate markers of specific molecular portraits in breast cancer, ER and PR status correlates with the overall tumor phenotype and genotype (luminal or basal cell). 22 We found an association between stage of regression and hormone receptor status in HG-DCIS. We hypothesize that this association is due to intrinsic immunogenic characteristics of hormone-negative in situ neoplasms. In other words, it is likely that the immune response leading to regressive change targets one or more lineage-specific markers. Further exploration of the proteomic expression profile of HG-DCIS, particularly in the context of hormone-negative lesions with regressive change, is warranted. HG-DCIS, in comparison to non high-grade counterparts, has a significantly higher rate of recurrence. 23,24 Recurrence tends to occur earlier and more often as invasive carcinoma. 25,26 Likewise, high nuclear grade in DCIS on biopsy specimens is, among other factors, a significant risk factor for invasive carcinoma on excision. 24,27 The role of changes associated with tumor regression in predicting HG-DCIS recurrence has been explored, with mixed results. Some authors identified that periductal inflammation and fibrosis are significantly associated with tumor recurrence and invasive carcinoma on excision. 3,23,28 Conversely, other studies found no such association. 29 Histopathologic evaluation of periductal fibrosis and inflammation was not standardized among references, which may account for the discrepancy in their results. Most cases in our study demonstrated some degree of regressive change and chronic inflammation. Therefore, we could not reliably assess a relationship between the presence or absence of regression on biopsy specimens and the diagnosis of invasive carcinoma on excision. Nonetheless, when dividing cases by stage of regressive change and inflammation, we did not find a significant correlation with upstage to invasive cancer. Although regression did not predict the presence of invasive carcinoma on excision in our study, we found that the degree of regressive change significantly correlated with volume of residual HG-DCIS. Several factors may account for this association. First, the development of regressive change may occur in parallel with HG-DCIS growth. On the other hand, regression may be a phenomenon preferentially elicited by types of HG-DCIS with a more rapid growth rate; as discussed earlier, regression appears to be characteristic of particular phenotypic and genotypic classes of in situ carcinoma. Whether the underlying mechanisms of regression in HG-DCIS eventually lead to stromal invasion remains unclear. The documented correlation between periductal fibrosis, inflammation and tumor effacement, and adverse prognostic features (such as hormone-negative status and extent of in situ disease) suggests that cancers displaying such changes are biologically more aggressive. These regressive changes, while reflecting a host response to the neoplasm, do not translate into complete tumor regression. To explain this paradoxical association, it has been suggested that immune responses leading to tumor regression produce a harmful effect by altering the myoepithelial cell layer and promoting stromal invasion. 3 Alternatively, we raise the possibility that the immune response against the neoplasm eradicates clones with high immunogenicity, thus selecting less immunogenic clones for survival and eventual progression to invasive carcinoma. Our study has several limitations. First, our sample is relatively small. Second, patient follow-up after surgical excision was too short to assess a possible relationship between regressive changes and patient outcomes (tumor recurrence and survival rates). Therefore, the significance of regression in the long-term biologic behavior of in situ and invasive mammary carcinoma remains to be explored and will be the focus of our future investigations. In summary, regressive change in HG-DCIS is frequent. The spectrum of regressive change is wide, ranging from mild to extensive periductal fibrosis and progressive loss of neoplastic cells. Advanced regressive changes and inflammation are more frequent in hormone-negative lesions, suggesting a relationship between molecular portraits and spontaneous regression. Advanced stage of HG-DCIS regression Am J Clin Pathol 2015;144:

8 Wasserman and Parra-Herran / Regressive Change in HG-DCIS is associated with a greater extent of in situ disease. Thus, regression in this context is only partial and does not translate into complete tumor effacement. Further study of regressive change in HG-DCIS, following unified definitions and inclusion criteria, is warranted to validate our findings and expand on the knowledge of this intriguing biologic phenomenon. Corresponding author: Carlos Parra-Herran MD, FCAP, University of Ottawa, The Ottawa Hospital General Campus, 501 Smyth Rd, CCW Room W4221, Ottawa, ON, Canada K1H 8L6; cparra@toh.on.ca. References 1. Everson TC, Cole WH. Spontaneous Regression of Cancer. Philadelphia, PA: Saunders; Muir R, Aitkenhead A. The healing of intra-duct carcinoma of mamma. J Pathol. 1934;18: Chivukula M, Domfeh A, Carter G, et al. Characterization of high-grade ductal carcinoma in situ with and without regressive changes: diagnostic and biologic implications. Appl Immunohistochem Mol Morphol. 2009;17: Cole WH. Spontaneous regression of cancer: the metabolic triumph of the host? Ann N Y Acad Sci. 1974;230: Diede SJ. Spontaneous regression of metastatic cancer: learning from neuroblastoma. Nat Rev Cancer. 2014;14: Kalialis LV, Drzewiecki KT, Klyver H. Spontaneous regression of metastases from melanoma: review of the literature. Melanoma Res. 2009;19: Challis GB, Stam HJ. The spontaneous regression of cancer: a review of cases from 1900 to Acta Oncol. 1990;29: Papac RJ. Spontaneous regression of cancer: possible mechanisms. In Vivo. 1998;12: Horii R, Akiyama F, Kasumi F, et al. Spontaneous healing of breast cancer. Breast Cancer. 2005;12: Larsen SU, Rose C. Spontaneous remission of breast cancer: a literature review [in Danish]. Ugeskr Laeger. 1999;161: Tokunaga E, Okano S, Nakashima Y, et al. Spontaneous regression of breast cancer with axillary lymph node metastasis: a case report and review of literature. Int J Clin Exp Pathol. 2014;7: Dussan C, Zubor P, Fernandez M, et al. Spontaneous regression of a breast carcinoma: a case report. Gynecol Obstet Invest. 2008;65: Zahl P-H, Maehlen J, Welch HG. The natural history of invasive breast cancers detected by screening mammography. Arch Intern Med. 2008;168: Zahl P-H, Andersen JM, Maehlen J. Spontaneous regression of cancerous tumors detected by mammography screening. JAMA. 2004;292: Sheu B-C, Kuo W-H, Chen R-J, et al. Clinical significance of tumor-infiltrating lymphocytes in neoplastic progression and lymph node metastasis of human breast cancer. Breast. 2008;17: Bilik R, Mor C, Hazaz B, et al. Characterization of T-lymphocyte subpopulations infiltrating primary breast cancer. Cancer Immunol Immunother. 1989;28: Müller MR, Grünebach F, Nencioni A, et al. Transfection of dendritic cells with RNA induces CD4- and CD8-mediated T cell immunity against breast carcinomas and reveals the immunodominance of presented T cell epitopes. J Immunol. 2003;170: Bobrow LG, Happerfield LC, Gregory WM, et al. The classification of ductal carcinoma in situ and its association with biological markers. Semin Diagn Pathol. 1994;11: Tang P, Wang X, Schiffhauer L, et al. Relationship between nuclear grade of ductal carcinoma in situ and cell origin markers. Ann Clin Lab Sci. 2006;36: Dabbs DJ, Chivukula M, Carter G, et al. Basal phenotype of ductal carcinoma in situ: recognition and immunohistologic profile. Mod Pathol. 2006;19: Tang P, Wang X, Schiffhauer L, et al. Expression patterns of ER-alpha, PR, HER-2/neu, and EGFR in different cell origin subtypes of high grade and non-high grade ductal carcinoma in situ. Ann Clin Lab Sci. 2006;36: Gruvberger S, Ringnér M, Chen Y, et al. Estrogen receptor status in breast cancer is associated with remarkably distinct gene expression patterns. Cancer Res. 2001;61: Pinder SE, Duggan C, Ellis IO, et al. A new pathological system for grading DCIS with improved prediction of local recurrence: results from the UKCCCR/ANZ DCIS trial. Br J Cancer. 2010;103: Kim J, Han W, Lee JW, et al. Factors associated with upstaging from ductal carcinoma in situ following core needle biopsy to invasive cancer in subsequent surgical excision. Breast. 2012;21: Wallis MG, Clements K, Kearins O, et al. The effect of DCIS grade on rate, type and time to recurrence after 15 years of follow-up of screen-detected DCIS. Br J Cancer. 2012;106: Collins LC, Tamimi RM, Baer HJ, et al. Outcome of patients with ductal carcinoma in situ untreated after diagnostic biopsy: results from the Nurses Health Study. Cancer. 2005;103: Yen TWF, Hunt KK, Ross MI, et al. Predictors of invasive breast cancer in patients with an initial diagnosis of ductal carcinoma in situ: a guide to selective use of sentinel lymph node biopsy in management of ductal carcinoma in situ. J Am Coll Surg. 2005;200: Hoorntje LE, Schipper MEI, Peeters PHM, et al. The finding of invasive cancer after a preoperative diagnosis of ductal carcinoma-in-situ: causes of ductal carcinoma-in-situ underestimates with stereotactic 14-gauge needle biopsy. Ann Surg Oncol. 2003;10: Huo L, Sneige N, Hunt KK, et al. Predictors of invasion in patients with core-needle biopsy-diagnosed ductal carcinoma in situ and recommendations for a selective approach to sentinel lymph node biopsy in ductal carcinoma in situ. Cancer. 2006;107: Am J Clin Pathol 2015;144: