Search

Context.— Breast carcinoma grade, as determined by the Nottingham Grading System (NGS), is an important criterion for determining prognosis. The NGS is based on 3 parameters: tubule formation (TF), nuclear pleomorphism (NP), and mitotic count (MC). The advent of digital pathology and artificial intelligence (AI) have increased interest in virtual microscopy using digital whole slide imaging (WSI) more broadly. Objective.— To compare concordance in breast carcinoma grading between AI and a multi-institutional group of breast pathologists using digital WSI. Design.— We have developed an automated NGS framework using deep learning. Six pathologists and AI independently reviewed a digitally scanned slide from 137 invasive carcinomas and assigned a grade based on scoring of the TF, NP, and MC. Results.— Interobserver agreement for the pathologists and AI for overall grade was moderate (κ = 0.471). Agreement was good (κ = 0.681), moderate (κ = 0.442), and fair (κ = 0.368) for grades 1, 3, and 2, respectively. Observer pair concordance for AI and individual pathologists ranged from fair to good (κ = 0.313–0.606). Perfect agreement was observed in 25 cases (27.4%). Interobserver agreement for the individual components was best for TF (κ = 0.471 each) followed by NP (κ = 0.342) and was worst for MC (κ = 0.233). There were no observed differences in concordance amongst pathologists alone versus pathologists + AI. Conclusions.— Ours is the first study comparing concordance in breast carcinoma grading between a multi-institutional group of pathologists using virtual microscopy to a newly developed WSI AI methodology. Using explainable methods, AI demonstrated similar concordance to pathologists alone.

Androgen receptor (AR) inhibitor therapy is a developing treatment for AR-positive breast cancer (BC) with ongoing clinical trials. AR splice variant-7 (AR-V7) is a truncated variant of AR that leads to AR inhibitor therapy resistance in prostate cancer; recent studies have identified AR-V7 in BC and theorized that AR-V7 can have a similar impact. This study assessed the prevalence and clinicopathologic features associated with AR-V7 in a large BC cohort. BC samples were evaluated by MSK-Fusion targeted RNAseq for AR-V7 detection and MSK-IMPACT targeted DNAseq, including triple-negative tumors with no driver alteration and estrogen receptor-positive/ESR1 wildtype tumors progressing on therapy. Among 196 primary and metastatic/recurrent cases (196 RNAseq, 194DNAseq), 9.7% (19/196) were AR-V7 positive and 90.3% (177/196) AR-V7 negative. All AR-V7 positive BC were AR-positive by immunohistochemistry (19/19). The prevalence of AR-V7 by receptor subtype (N = 189) was: 18% (12/67) in ER-/PgR-/HER2-negative BC, 3.7% (4/109) in ER-positive/HER2-negative BC, and 15.4% (2/13) in HER2-positive BC; AR-V7 was detected in one ER-positive/HER2-unknown BC. Apocrine morphology was observed in 42.1% (8/19) of AR-V7 positive BC and 3.4% (6/177) AR-V7 negative BC (P

Classic adenoid cystic carcinomas (C-AdCCs) of the breast are rare, relatively indolent forms of triple negative cancers, characterized by recurrent MYB or MYBL1 genetic alterations. Solid and basaloid adenoid cystic carcinoma (SB-AdCC) is considered a rare variant of AdCC yet to be fully characterized. Here, we sought to determine the clinical behavior and repertoire of genetic alterations of SB-AdCCs. Clinicopathologic data were collected on a cohort of 104 breast AdCCs (75 C-AdCCs and 29 SB-AdCCs). MYB expression was assessed by immunohistochemistry and MYB-NFIB and MYBL1 gene rearrangements were investigated by fluorescent in-situ hybridization. AdCCs lacking MYB/MYBL1 rearrangements were subjected to RNA-sequencing. Targeted sequencing data were available for 9 cases. The invasive disease-free survival (IDFS) and overall survival (OS) were assessed in C-AdCC and SB-AdCC. SB-AdCCs have higher histologic grade, and more frequent nodal and distant metastases than C-AdCCs. MYB/MYBL1 rearrangements were significantly less frequent in SB-AdCC than C-AdCC (3/14, 21% vs 17/20, 85% P

Purpose: Lynch syndrome is defined by germline pathogenic mutations involving DNA mismatch repair (MMR) genes and linked with the development of MMR-deficient colon and endometrial cancers. Whether breast cancers developing in the context of Lynch syndrome are causally related to MMR deficiency (MMRd), remains controversial. Thus, we explored the morphologic and genomic characteristics of breast cancers occurring in Lynch syndrome individuals. Experimental Design: A retrospective analysis of 20,110 patients with cancer who underwent multigene panel genetic testing was performed to identify individuals with a likely pathogenic/pathogenic germline variant in MLH1, MSH2, MSH6, or PMS2 who developed breast cancers. The histologic characteristics and IHC assessment of breast cancers for MMR proteins and programmed death-ligand 1 (PD-L1) expression were assessed on cases with available materials. DNA samples from paired tumors and blood were sequenced with Memorial Sloan Kettering–Integrated Mutation Profiling of Actionable Cancer Targets (≥468 key cancer genes). Microsatellite instability (MSI) status was assessed utilizing MSISensor. Mutational signatures were defined using SigMA. Results: A total of 272 individuals with Lynch syndrome were identified, 13 (5%) of whom had primary breast cancers. The majority of breast cancers (92%) were hormone receptor–positive tumors. Five (42%) of 12 breast cancers displayed loss of MMR proteins by IHC. Four (36%) of 11 breast cancers subjected to tumor-normal sequencing showed dominant MSI mutational signatures, high tumor mutational burden, and indeterminate (27%) or high MSISensor scores (9%). One patient with metastatic MMRd breast cancer received anti-PD1 therapy and achieved a robust and durable response. Conclusions: A subset of breast cancers developing in individuals with Lynch syndrome are etiologically linked to MMRd and may benefit from anti-PD1/PD-L1 immunotherapy.

Supplementary Figure 1. Comparison of tumor mutational burden, dominant mutational signatures and MSISensor scores of Lynch syndrome BC cohort to control-matched BC cohort using MSK-IMPACT targeted sequencing.

Supplementary Legend

The glycosphingolipid disialoganglioside GD2 is a cell surface-associated antigen expressed on tumors of neuroectodermal origin that serves as a target of immunotherapy in select cancer types. Information about the expression of GD2 in breast cancer is limited. In the present study, we investigate the utility of GD2 as a potential biomarker for targeted treatment. The study cohort consists of 386 breast carcinomas of several histologic types. GD2 expression was assessed in both whole tumor sections and tissue microarrays with anti-GD2 3F8 monoclonal antibody immunohistochemistry and correlated with clinicopathologic features and survival outcomes. A total of 134 (35%) breast carcinomas were positive for GD2, with a median H-score of 100. 3F8 staining displayed granular and predominantly cytoplasmic or perinuclear patterns, which was confined to the neoplastic tissue in nearly all cases. GD2 positivity was significantly associated with tumor histologic type (P=0.0015), low grade (P

The surgical margin status of breast lumpectomy specimens for invasive carcinoma and ductal carcinoma in situ (DCIS) guides clinical decisions, as positive margins are associated with higher rates of local recurrence. The "cavity shave" method of margin assessment has the benefits of allowing the surgeon to orient shaved margins intraoperatively and the pathologist to assess one inked margin per specimen. We studied whether a deep convolutional neural network, a deep multi-magnification network (DMMN), could accurately segment carcinoma from benign tissue in whole slide images (WSIs) of shave margin slides, and therefore serve as a potential screening tool to improve the efficiency of microscopic evaluation of these specimens. Applying the pretrained DMMN model, or the initial model, to a validation set of 408 WSIs (348 benign, 60 with carcinoma) achieved an area under the curve (AUC) of 0.941. After additional manual annotations and fine-tuning of the model, the updated model achieved an AUC of 0.968 with sensitivity set at 100% and corresponding specificity of 78%. We applied the initial model and updated model to a testing set of 427 WSIs (374 benign, 53 with carcinoma) which showed AUC values of 0.900 and 0.927, respectively. Using the pixel classification threshold selected from the validation set, the model achieved a sensitivity of 92% and specificity of 78%. The four false-negative classifications resulted from two small foci of DCIS (1 mm, 0.5 mm) and two foci of well-differentiated invasive carcinoma (3 mm, 1.5 mm). This proof-of-principle study demonstrates that a DMMN machine learning model can segment invasive carcinoma and DCIS in surgical margin specimens with high accuracy and has the potential to be used as a screening tool for pathologic assessment of these specimens.

As neoadjuvant chemotherapy (NAC) for breast cancer has become more widely used, so has nipple-sparing mastectomy. A common criterion for eligibility is a 1 cm tumor-to-nipple distance (TND), but its suitability after NAC is unclear. In this study, we examined factors predictive of negative nipple pathologic status (NS−) in women undergoing total mastectomy after NAC. Women with invasive breast cancer treated with NAC and total mastectomy from August 2014 to April 2018 at our institution were retrospectively identified. Following review of pre- and post-NAC magnetic resonance imaging (MRI) and mammograms, the association of clinicopathologic and imaging variables with NS− was examined and the accuracy of 1 cm TND on imaging for predicting NS− was determined. Among 175 women undergoing 179 mastectomies, 74% of tumors were cT1-T2 and 67% were cN+ on pre-NAC staging; 10% (18/179) had invasive or in situ carcinoma in the nipple on final pathology. On multivariable analysis, after adjusting for age, grade, and tumor stage, three factors, namely number of positive nodes, pre-NAC nipple-areolar complex retraction, and decreasing TND, were significant predictors of nipple involvement (p

Breast carcinoma grading is an important prognostic feature recently incorporated into the AJCC Cancer Staging Manual. There is increased interest in applying virtual microscopy (VM) using digital whole slide imaging (WSI) more broadly. Little is known regarding concordance in grading using VM and how such variability might affect AJCC prognostic staging (PS). We evaluated interobserver variability amongst a multi-institutional group of breast pathologists using digital WSI and how discrepancies in grading would affect PS. A digitally scanned slide from 143 invasive carcinomas was independently reviewed by 6 pathologists and assigned grades based on established criteria for tubule formation (TF), nuclear pleomorphism (NP), and mitotic count (MC). Statistical analysis was performed. Interobserver agreement for grade was moderate (κ = 0.497). Agreement was fair (κ = 0.375), moderate (κ = 0.491), and good (κ = 0.705) for grades 2, 3, and 1, respectively. Observer pair concordance ranged from fair to good (κ = 0.354-0.684) Perfect agreement was observed in 43 cases (30%). Interobserver agreement for the individual components was best for TF (κ = 0.503) and worst for MC (κ = 0.281). Seventeen of 86 (19.8%) discrepant cases would have resulted in changes in PS and discrepancies most frequently resulted in a PS change from IA to IB (n = 9). For two of these nine cases, Oncotype DX results would have led to a PS of 1A regardless of grade. Using VM, a multi-institutional cohort of pathologists showed moderate concordance for breast cancer grading, similar to studies using light microscopy. Agreement was the best at the extremes of grade and for evaluation of TF. Whether the higher variability noted for MC is a consequence of VM grading warrants further investigation. Discordance in grading infrequently leads to clinically meaningful changes in the prognostic stage.

Context.— The Ventana programmed death ligand-1 (PD-L1) SP142 immunohistochemical assay (IHC) is approved by the US Food and Drug Administration as the companion diagnostic assay to identify patients with locally advanced or metastatic triple-negative breast cancer for immunotherapy with atezolizumab, a monoclonal antibody targeting PD-L1. Objective.— To determine interobserver variability in PD-L1 SP142 IHC interpretation in invasive breast carcinoma. Design.— The pathology database was interrogated for all patients diagnosed with primary invasive, locally recurrent, or metastatic breast carcinoma on which PD-L1 SP142 IHC was performed from November 2018 to June 2019 at our institution. A subset of cases was selected using a computerized random-number generator. PD-L1 IHC was evaluated in stromal tumor-infiltrating immune cells using the IMpassion130 trial criteria, with positive cases defined as immunoreactivity in immune cells in 1% or more of the tumor area. IHC was interpreted on whole slide images by staff pathologists with breast pathology expertise. Interobserver variability was calculated using unweighted κ. Results.— A total of 79 cases were assessed by 8 pathologists. Interobserver agreement was substantial (κ = 0.727). There was complete agreement among all 8 pathologists in 62% (49 of 79) of cases, 7 pathologists or more in 84% (66 of 79) of cases, and 6 pathologists or more in 92% (73 of 79) of cases. In 4% (3 of 79) of cases, all of which were small biopsies, pathologists' interpretations were evenly split between scores of positive and negative. Conclusions.— The findings show substantial agreement in PD-L1 SP142 IHC assessment of breast carcinoma cases among 8 pathologists at a single institution. Further study is warranted to define the basis for discrepant results.

Juvenile papillomatosis (JP) of the breast is a rare benign mass‐forming lesion occurring in young women, which is histologically characterized by a constellation of proliferative changes and large cysts, giving it the gross appearance of Swiss cheese. A subset of patients with JP report a family history of breast carcinoma and/or coexisting or subsequent breast carcinoma. We performed whole‐exome sequencing of the hyperplastic epithelial component of three JPs, including one with coexisting ductal carcinoma in situ (DCIS) and invasive ductal carcinoma of no special type (IDC‐NST). JPs harbored clonal somatic PIK3CA hotspot mutations in two cases. In the JP with coexisting DCIS and IDC‐NST, these lesions were clonally related to the associated JP, sharing a clonal PIK3CA E542K somatic hotspot mutation. JP showed a paucity of copy number alterations, whereas the associated DCIS and IDC‐NST showed concurrent 1q gains/16q losses, hallmarks of estrogen receptor (ER)‐positive breast cancers. We observed JP to harbor a dominant aging‐related mutational signature, whereas coexisting DCIS and IDC‐NST showed greater exposure to an APOBEC signature. Taken together, our findings suggest that, at least in a subset of cases, JP might constitute the substrate from which DCIS and invasive breast carcinomas develop.

Tall cell carcinoma with reverse polarity is a rare subtype of breast carcinoma with solid and papillary growth and nuclear features reminiscent of those of the tall cell variant of papillary thyroid carcinoma. These tumors harbor recurrent IDH2 R172 hotspot mutations or TET2 mutations, co-occurring with mutations in PI3K pathway genes. Diagnosis of tall cell carcinomas with reverse polarity is challenging in view of their rarity and the range of differential diagnosis. We sought to determine the sensitivity and specificity of IDH2 R172 immunohistochemistry for the detection of IDH2 R172 hotspot mutations in this entity. We evaluated 14 tall cell carcinomas with reverse polarity (ten excision and five core needle biopsy specimens), 13 intraductal papillomas, 16 solid papillary carcinomas, and 5 encapsulated papillary carcinomas by Sanger sequencing of the IDH2 R172 hotspot locus and of exons 9 and 20 of PIK3CA, and by immunohistochemistry using monoclonal antibodies (11C8B1) to the IDH2 R172S mutation. The 14 tall cell carcinomas with reverse polarity studied harbored IDH2 R172 hotspot mutations, which co-occurred with PIK3CA hotspot mutations in 50% of cases. None of the other papillary neoplasms analyzed displayed IDH2 R172 mutations, however PIK3CA hotspot mutations were detected in 54% of intraductal papillomas, 6% of solid papillary carcinomas, and 20% of encapsulated papillary carcinomas tested. Immunohistochemical analysis with anti-IDH2 R172S antibodies (11C8B1) detected IDH2 R172 mutated protein in 93% (14/15) of tall cell carcinomas with reverse polarity samples including excision (n = 9/10) and core needle biopsy specimens (n = 5), whereas the remaining papillary neoplasms (n = 34) were negative. Our findings demonstrate that immunohistochemical analysis of IDH2 R172 is highly sensitive and specific for the detection of IDH2 R172 hotspot mutations, and likely suitable as a diagnostic tool in the evaluation of excision and core needle biopsy material of tall cell carcinomas with reverse polarity.

Younger women (age ≤ 40 years) with breast cancer undergoing neoadjuvant chemotherapy (NAC) have higher rates of pathologic complete response (pCR); however, it is unknown whether axillary or breast downstaging rates differ by age. In this study, we compared pCR incidence and surgical downstaging rates of the breast and axilla post NAC, between patients aged ≤ 40, 41-60, and ≥ 61 years.We identified 1383 women with stage I-III breast cancer treated with NAC and subsequent surgery from November 2013 to December 2018. pCR and breast/axillary downstaging rates were assessed and compared across age groups.Younger women were significantly more likely to have ductal histology, poorly differentiated tumors, and BRCA mutations; 35% of tumors were hormone receptor-positive/human epidermal growth factor receptor 2-negative (HR+/HER2-), 36% were HER2-positive (HER2+), and 29% were triple negative (TN), with similar subtype distribution across age groups (p = 0.6). Overall, pCR rates did not differ by age, however among patients with TN tumors (n = 394), younger women had higher pCR rates (52% vs. 35% among those aged 41-60 years and 29% among those aged ≥61 years; p = 0.007) and were more likely to have tumors with high tumor-infiltrating lymphocyte (TIL) concentrations (p 0.001). Downstaging to breast-conserving surgery (BCS) eligibility post NAC among initially BCS-ineligible patients was similar across age groups; younger women chose BCS less often (p 0.001). Among cN1 patients (n = 813), 52% of women ≤40 years of age avoided axillary lymph node dissection (ALND) with NAC, versus 39% and 37% in the older groups (p 0.001).Younger women undergoing NAC for axillary downstaging were more likely to avoid ALND across all subtypes; however, overall pCR rates did not differ by age. Despite equivalent breast downstaging and BCS eligibility rates across age groups, younger women were less likely to undergo BCS.

Background: Tumor microenvironment of metastasis (TMEM) is a microanatomical structure composed by 3 cells in direct contact, including a tumor cell expressing the actin-regulatory protein Mammalian-enabled (Mena), a perivascular Tie2hi/Vegfhi-expressing macrophage, and an endothelial cell. TMEM are intravasation sites that function as doorways for hematogenous tumor cell dissemination and metastases (Harney et al. Cancer Discovery 2015). TMEM may be identified and enumerated by triple immunohistochemistry in mouse and human mammary carcinomas. High TMEM score is associated with increased risk of distant metastasis in early stage breast cancer, and provides complementary prognostic information to IHC4 (Rohan et al. JNCI 2014) and Oncotype DX Recurrence Score in ER+, HER2-negative breast cancer (Sparano et al. NPJ Breast Cancer, 2017). Neoadjuvant chemotherapy (NAC) increases TMEM score in breast carcinoma in animal models and humans, indicating a previously unrecognized mechanism of resistance to cytotoxic therapy (Karagiannis et al. Science Trans Med 2017). Intravasation at TMEM sites may be inhibited using agents that block release of VEGF from TMEM-associated TIE2-hi, VEGF-hi macrophages (Harney et al. Mol Cancer Ther, 2017). Here we investigated whether TMEM score in post-NAC treated breast carcinoma is prognostic of distant recurrence in localized breast cancer after NAC, and thus provides a foundation for testing agents that block TMEM function in combination with NAC. Methods: We determined TMEM score in 80 evaluable patients' post-NAC specimens with residual invasive ductal carcinomas of at least 0.5 cm. Approximately 60% of patients had ER+/HER2-negative, 28% had triple negative and 12% had HER2+ disease. Most of the patients received doxorubicin/cyclophosphamide + taxane and an anti-HER2 therapy if applicable. Tissue sections from residual tumors were stained for TMEM using triple immunohistochemistry for Mena-expressing cancer cells, CD31-expressing endothelial cells and CD68-expressing macrophages. The stained slides were scanned, and the images were analyzed by three pathologists, blinded to outcome, who independently determined the tissue areas appropriate for TMEM scoring. TMEM was scored within these areas using an automated algorithm. Results: TMEM score was significantly higher in patients with distant recurrence (average TMEM=106), compared to patients without distant recurrence (average TMEM=71) (p50 yrs. vs. Conclusion: TMEM score may provide independent prognostic information for distant recurrence in patients with residual invasive carcinoma after NAC. These results support the use of agents that block TMEM function in combination with NAC, as planned in the I-SPY2 trial. Citation Format: Oktay MH, D'Alfonso T, Ginter P, Lanjewar S, Entenberg D, Pastoriza JM, Wang Y, Lin Y, Karagiannnis GS, Lin J, Ye X, Anampa J, Xue X, Rohan TE, Sparano JA, Condeelis JS. Tumor microenvironment of metastasis (TMEM) score in residual breast carcinoma post-neoadjuvant chemotherapy as an independent prognosticator of distant recurrence [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P2-08-18.

For primary triple-negative breast cancer (TNBC) treated with neoadjuvant chemotherapy (NAC), higher pretreatment tumor-infiltrating lymphocytes (TILs) correlates with increased pathologic complete response (pCR) rates, and improved survival. We evaluated the added prognostic value of residual disease (RD) TILs to residual cancer burden (RCB) in predicting survival post-NAC.We combined four TNBC NAC patient cohorts who did not achieve pCR. RD TILs were investigated for associations with recurrence-free survival (RFS), and overall survival (OS) using Cox models with stromal TILs as a continuous variable (per 10% increment). The likelihood ratio test was used to evaluate added prognostic value of RD TILs.A total of 375 RD TNBC samples were evaluable for TILs and RCB. The median age was 50 years, with 62% receiving anthracycline/taxane chemotherapy. The RCB class after NAC was 11%, 50%, and 39% for I, II, and III, respectively. The median RD TIL level was 20% (IQR 10-40). There was a positive correlation between RD TIL levels and CD8+ T-cell density (ρ = 0.41). TIL levels were significantly lower with increasing post-NAC tumor (P = 0.005), nodal stage (P = 0.032), but did not differ by RCB class (P = 0.84). Higher RD TILs were significantly associated with improved RFS (HR: 0.86; 95% CI 0.79-0.92; P 0.001), and improved OS (HR: 0.87; 95% CI 0.80-0.94; P 0.001), and remained significant predictors in multivariate analysis (RFS P = 0.032; OS P = 0.038 for OS). RD TILs added significant prognostic value to multivariate models including RCB class (P 0.001 for RFS; P = 0.021 for OS). The positive prognostic effect of RD TILs significantly differed by RCB class for RFS (PInt=0.003) and OS (PInt=0.008) with a greater magnitude of positive effect observed for RCB class II than class III.TIL levels in TNBC RD are significantly associated with improved RFS and OS and add further prognostic information to RCB class, particularly in RCB class II.

Androgen receptor (AR) inhibitor therapy is a developing treatment for AR-positive breast cancer (BC) with ongoing clinical trials. AR splice variant-7 (AR-V7) is a truncated variant of AR that leads to AR inhibitor therapy resistance in prostate cancer; recent studies have identified AR-V7 in BC and theorized that AR-V7 can have a similar impact. This study assessed the prevalence and clinicopathologic features associated with AR-V7 in a large BC cohort. BC samples were evaluated by MSK-Fusion targeted RNAseq for AR-V7 detection and MSK-IMPACT targeted DNAseq, including triple-negative tumors with no driver alteration and estrogen receptor-positive/ESR1 wildtype tumors progressing on therapy. Among 196 primary and metastatic/recurrent cases (196 RNAseq, 194DNAseq), 9.7% (19/196) were AR-V7 positive and 90.3% (177/196) AR-V7 negative. All AR-V7 positive BC were AR-positive by immunohistochemistry (19/19). The prevalence of AR-V7 by receptor subtype (N = 189) was: 18% (12/67) in ER-/PgR-/HER2-negative BC, 3.7% (4/109) in ER-positive/HER2-negative BC, and 15.4% (2/13) in HER2-positive BC; AR-V7 was detected in one ER-positive/HER2-unknown BC. Apocrine morphology was observed in 42.1% (8/19) of AR-V7 positive BC and 3.4% (6/177) AR-V7 negative BC (P 0.00001). Notably, AR-V7 was detected in 2 primary BC and 7 metastatic/recurrent BC patients with no prior endocrine therapy. We conclude that positive AR IHC and apocrine morphology are pathologic features that may indicate testing for AR-V7 is warranted in both primary and metastatic BC in the appropriate clinical context. The study findings further encourage the assessment of AR-V7 as a predictive biomarker for AR antagonist benefit in ongoing clinical BC trials.

The surgical margin status of breast lumpectomy specimens for invasive carcinoma and ductal carcinoma in situ (DCIS) guides clinical decisions, as positive margins are associated with higher rates of local recurrence. The "cavity shave" method of margin assessment has the benefits of allowing the surgeon to orient shaved margins intraoperatively and the pathologist to assess one inked margin per specimen. We studied whether a deep convolutional neural network, a deep multi-magnification network (DMMN), could accurately segment carcinoma from benign tissue in whole slide images (WSIs) of shave margin slides, and therefore serve as a potential screening tool to improve the efficiency of microscopic evaluation of these specimens. Applying the pretrained DMMN model, or the initial model, to a validation set of 408 WSIs (348 benign, 60 with carcinoma) achieved an area under the curve (AUC) of 0.941. After additional manual annotations and fine-tuning of the model, the updated model achieved an AUC of 0.968 with sensitivity set at 100% and corresponding specificity of 78%. We applied the initial model and updated model to a testing set of 427 WSIs (374 benign, 53 with carcinoma) which showed AUC values of 0.900 and 0.927, respectively. Using the pixel classification threshold selected from the validation set, the model achieved a sensitivity of 92% and specificity of 78%. The four false-negative classifications resulted from two small foci of DCIS (1 mm, 0.5 mm) and two foci of well-differentiated invasive carcinoma (3 mm, 1.5 mm). This proof-of-principle study demonstrates that a DMMN machine learning model can segment invasive carcinoma and DCIS in surgical margin specimens with high accuracy and has the potential to be used as a screening tool for pathologic assessment of these specimens.

Specimens submitted for intraoperative frozen sections from breast cancer patients most often include axillary sentinel lymph nodes and subareolar margins. Sentinel lymph node frozen section use has declined following the publication of the American College of Surgeons Oncology Group (ACOSOG) Z0011 trial. It is currently indicated for clinically node-negative patients undergoing mastectomy and for those that have received neoadjuvant chemotherapy. Nipple-sparing mastectomy has become increasingly popular as a procedure associated with superior aesthetic outcomes while being oncologically safe. Subareolar tissue biopsies from patients undergoing nipple-sparing mastectomy are often submitted for frozen section to exclude occult carcinoma. This chapter will provide a practical approach for evaluating frozen sections of sentinel lymph node and subareolar margin specimens. Pertinent data from clinical trials will provide context for the current use of frozen sections for breast carcinoma patients. Instructive examples of both routine and difficult cases will highlight challenges and pitfalls associated with frozen section interpretation.

Aims Here we explore the presence of mediator complex subunit 12 (MED12) exon 2 and telomerase reverse transcriptase (TERT) promoter hotspot mutations in complex fibroadenomas (CFAs) of the breast. Methods The stromal components from 18 CFAs were subjected to Sanger sequencing of MED12 exon 2 and the TERT promoter hotspot loci. The epithelial and stromal components of two MED12 mutated CFAs were subjected to laser capture microdissection, and Sanger sequencing of MED12 exon 2, TERT promoter and PIK3CA exons 9 and 20, separately. Results MED12 exon 2 mutations were identified in the stroma of 17% of CFAs. The analyses of epithelial and stromal components, microdissected separately, revealed that MED12 mutations were restricted to the stroma. No TERT promoter or PIK3CA mutations in exons 9 and 20 were detected in analysed CFAs. Conclusions Like conventional fibroadenomas, MED12 exon 2 mutations appear to be restricted to the stromal component of CFAs, supporting the notion that CFAs are stromal neoplasms.

Context.— It is unclear whether HER2+ tumors expressing both estrogen receptor (ER) and progesterone receptor (PR), that is, triple-positive breast carcinomas (TPBCs), show unique morphologic and clinical features and response to neoadjuvant chemotherapy (NAC). Objective.— To study the morphologic and immunohistochemical features of TPBCs from patients who underwent NAC. Design.— We retrospectively reviewed core biopsy and post-NAC slides of 85 TPBCs. H-scores were calculated for ER and PR. HER2 slides and fluorescence in situ hybridization (FISH) reports were reviewed. Residual cancer burden was calculated for post-NAC specimens. Results.— Eighty-one of the 85 tumors (95.3%) showed ductal histology, 3 (3.5%) were invasive lobular carcinomas, and 1 (1.2%) showed mixed ductal and lobular features. A subset showed mucinous (n = 7, 8.2%), apocrine (n = 5, 5.9%), and/or micropapillary (n = 4, 4.7%) differentiation. Fifty-four TPBCs (63.5%) showed high ER expression (H-score >200), including 27 (31.8%) with high expression of ER and PR. Fifty-two tumors (61.1%) showed HER2 3+ staining. Mean HER2/CEP17 ratio by FISH was 3.6 (range, 2–12.2) and mean HER2 signals per cell was 8 (range, 3.7–30.4). Pathologic complete response (pCR) rate was 35.3% (30 of 85). HER2 3+ staining was the only significant predictor of pCR on multivariate analysis (odds ratio = 9.215; 95% CI, 2.401–35.371; P < .001). The ER/PR expression did not correlate with response to therapy. Conclusions.— TPBCs are heterogeneous with some showing mucinous, lobular, or micropapillary differentiation. The pCR rate of TPBCs is similar to that reported for ER+/PR−/HER2+ tumors. HER2 overexpression by IHC was associated with significantly better response to therapy and may help select patients for treatment in the neoadjuvant setting.

E-cadherin (ECAD) immunohistochemical (IHC) expression is lost in ∼90% of invasive lobular carcinomas (ILCs) owing to genomic alterations of CDH1. We examined morphologic features and ECAD IHC expression in invasive breast carcinomas (BCs) with known CDH1 alterations. Between January 2014 and May 2018, 202 cases of BC with a CDH1 somatic alteration were identified. ECAD expression was lost in 77% (155/202) of cases and was retained in 23% (47/202) cases. Most (90%, 139/155) ECAD-negative cases were morphologically classified as ILC, while the remaining (10%, 16/155) were invasive mammary carcinoma with mixed ductal and lobular features (IMC). Of 47 cases with ECAD staining, 62% (29/47) were classified as ILC, 23% (11/47) were classified as IMC, and 15% (7/47) were classified as invasive ductal carcinoma (IDC). Of note, 51% (24/47) of ECAD-positive cases were initially diagnosed as IDC or IMC based on ECAD expression alone. For ECAD-negative BCs, 98% (152/155) of CDH1 alterations were truncating, and 2% (3/155) were variants of unknown significance (VUS). Truncating CDH1 alterations were identified in the majority of ECAD-positive BCs (72%, 34/47); however, VUS-type CDH1 alterations were more prevalent (28%, 13/47) in ECAD-positive BCs than in ECAD-negative BCs. Although 90% of ECAD-negative tumors were compatible with ILC in this study, 17% (29/168) of ILC cases were ECAD positive. In addition, CDH1 truncating alterations were seen in ECAD-positive ILC, supporting the notion of aberrant ECAD staining. Therefore, ECAD IHC expression must be interpreted in conjunction with morphology, and BC with classic histologic features of ILC should not be reclassified as IDC/IMC based solely on the status of ECAD IHC expression.

Neuroendocrine breast neoplasms are uncommon invasive carcinomas that have historically been poorly defined due to various definitions of what constitutes a neuroendocrine carcinoma. The 5th Edition of the World Health Organization (WHO) Classification of Breast Tumors has moved to a dichotomous classification of neuroendocrine neoplasms in the breast in order to become standardized with classifications of other organ systems. Neuroendocrine breast neoplasms in the new edition are classified as "neuroendocrine tumor" and "neuroendocrine carcinoma." Key changes are exclusion of special histologic types (solid papillary carcinoma and hypercellular variant of mucinous carcinoma) and the inclusion of large cell neuroendocrine carcinoma. Neuroendocrine tumors are genetically heterogenous and harbor molecular alterations that differ from invasive carcinoma, no special type. Neuroendocrine carcinomas (high-grade) show some overlapping molecular alterations with their counterparts in other organ systems. Data regarding the prognostic significance of neuroendocrine differentiation are conflicting, and histologic grade and tumor stage remain the main prognostic parameters. Current management of neuroendocrine neoplasms is not different from other types of breast carcinoma. This review will provide an update to the current WHO classification of neuroendocrine breast neoplasms and describe pertinent clinical, histologic, and molecular features of these uncommon tumors.

Metastasis causes ~90% of breast cancer mortality. However, standard prognostic tests based mostly on proliferation genes do not measure metastatic potential. Tumor MicroEnvironment of Metastasis (TMEM), an immunohistochemical biomarker for doorways on blood vessels that support tumor cell dissemination is prognostic for metastatic outcome in breast cancer patients. Studies quantifying TMEM doorways have involved manual scoring by pathologists utilizing static digital microscopy: a labor-intensive process unsuitable for use in clinical practice. We report here a validation study evaluating a new quantitative digital pathology (QDP) tool (TMEM-DP) for identification and quantification of TMEM doorways that closely mimics pathologists&rsquo, workflow and reduces pathologists&rsquo, variability to levels suitable for use in a clinical setting. Blinded to outcome, QDP was applied to a nested case-control study consisting of 259 matched case-control pairs. Sixty subjects of these were manually scored by five pathologists, digitally recorded using whole slide imaging (WSI), and then used for algorithm development and optimization. Validation was performed on the remainder of the cohort. TMEM-DP shows excellent reproducibility and concordance and reduces pathologist time from ~60 min to ~5 min per case. Concordance between manual scoring and TMEM-DP was found to be &gt, 0.79. These results show that TMEM-DP is capable of accurately identifying and scoring TMEM doorways (also known as MetaSite score) equivalent to pathologists.

© Author(s) (or their employer(s)) 2020. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ. http://creativecommons.org/licenses/by-nc/4.0/This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/., Aims: Here we explore the presence of mediator complex subunit 12 (MED12) exon 2 and telomerase reverse transcriptase (TERT) promoter hotspot mutations in complex fibroadenomas (CFAs) of the breast. Methods: The stromal components from 18 CFAs were subjected to Sanger sequencing of MED12 exon 2 and the TERT promoter hotspot loci. The epithelial and stromal components of two MED12 mutated CFAs were subjected to laser capture microdissection, and Sanger sequencing of MED12 exon 2, TERT promoter and PIK3CA exons 9 and 20, separately. Results: MED12 exon 2 mutations were identified in the stroma of 17% of CFAs. The analyses of epithelial and stromal components, microdissected separately, revealed that MED12 mutations were restricted to the stroma. No TERT promoter or PIK3CA mutations in exons 9 and 20 were detected in analysed CFAs. Conclusions: Like conventional fibroadenomas, MED12 exon 2 mutations appear to be restricted to the stromal component of CFAs, supporting the notion that CFAs are stromal neoplasms., This study was funded by the Breast Cancer Research Foundation. BW is funded by a Cycle for Survival grant, CS by a Fundação para a Ciência e Tecnologia grant (SFRH/BDE/110544/2015). FP is partially funded by a K12 CA184746 grant. The research reported in this paper was supported in part by a Cancer Centre Support Grant of the National Institutes of Health/National Cancer Institute (grant No P30CA008748).

Granulomatous mastitis (GM) is an uncommon entity and challenging to distinguish from breast carcinoma, both clinically and upon imaging. Cystic neutrophilic granulomatous mastitis (CNGM) is a histologically distinct GM with clear "cystic" spaces (CS) surrounded by neutrophils and may contain Gram-positive bacilli (such as Corynebacterium species). Cytologic features of GM, in particular CNGM, on ThinPrep (TP), are limited. We reviewed the cytological features of GM with a special focus on CNGM on TP. Ten cases with a concurrent/recent FNA and histologic diagnosis of GM or CNGM, or positive culture for Corynebacterium kroppenstedtii, were identified over an 11-y period (2007-2017). A double-blinded review of the TP slides was performed by two independent cytopathologists. Data were analysed using Fisher's exact test. Six cases were histologically proven CNGM and 1 FNA only case had a positive culture for C. kroppenstedtii. Three had a histological diagnosis of GM. TP from the 7 CNGM cases demonstrated CS, including 1 case with a cellblock (CB) study showing well-formed clear CS, while none of the non-CNGM cases had CS (P = .01). Bacteria were identified within CS in 5 of 7 cases of CNGM. Multinucleated giant cells, granulomas, histiocytes, and neutrophils were variably observed. All histologically confirmed cases of CNGM showed distinct CS on TP. Cellblock study is a valuable adjunct to evaluate CNGM and for ancillary studies. Although our study was limited by a small sample size, our findings necessitate the need for studies with a larger cohort.

A 34-year-old G5P4014 (5 pregnancies, 4 children born at term, 0 preterm, 1 abortion, and 4 living children) with stage IV triple-positive (ER+/PR+/Her2+) breast cancer was admitted to the medical oncology service for recurrent shortness of breath and right-sided pleuritic chest pain secondary to disease progression. The patient had widespread metastases to the bone, liver, and lungs and was being treated with docetaxel, trastuzumab, pertuzumab, and zoledronic acid. A gynecology consult was requested after the patient was incidentally noted to have a positive urine human chorionic gonadotropin (hCG)4 (Beckman Coulter Icon 25) and serum βhCG of 102 mIU/mL (Beckman UniCel® DxI). There was no evidence of impaired renal function (estimated glomerular filtration rate levels all within normal limits), and no medications that can affect hCG concentration were noted in the patient's chart. The patient's last menstrual period was approximately 1 year ago. She experienced premature ovarian failure because of chemotoxic effects on the ovary. The patient last had unprotected sexual intercourse about 6 months before her presentation. She complained of chronic lower back pain and nausea, which were attributed to chemotherapy, but denied having any vaginal bleeding, spotting, lower abdominal pain, or cramping. There was no evidence of an intrauterine or ectopic pregnancy seen on transvaginal ultrasound. The βhCG …

AIMS Neoadjuvant chemotherapy (NAC) is often used to treat localised invasive breast cancer. Invasive mucinous carcinoma (IMC) is considered to be an indolent form of invasive breast cancer, and is rarely treated with NAC. We report the largest series of IMCs treated with NAC, and report a characteristic, but not well recognised, pattern of pathological response. METHODS AND RESULTS Our series included seven patients with IMC treated with NAC. Three patients presented with locally advanced disease, three patients had tumours that were HER-2/neu-positive, and four patients had tumours with admixed mucinous and micropapillary features. Clinical and imaging assessment of response showed persistent and, in some cases, progressive disease, despite evidence of significant pathological response in these cases. Pathological assessment after NAC demonstrated marked reduction in tumour cellularity, but persistent space-occupying mucin pools, showing acellular mucin in one case

The assessment of sentinel lymph nodes (SLN) on hematoxylin and eosin (HE)-stained sections in cases of classic type of invasive lobular carcinoma (cILC) is considered unreliable, particularly in cases with minimal involvement, that is by either isolated tumor cells (pN0i+) or micrometastases (pN1mi). Although the impact of minimal SLN involvement has been shown to be insignificant in most clinical trials (even though cILC was either under-represented or not separated in the respective cohorts), the results of MIRROR trial did emphasize the need for additional therapy in cases with minimally involved SLN to ensure improved disease-free survival. We sought to study the role of cytokeratin immunohistochemistry (CK-IHC) in evaluating SLN in cILC. A total of 582 cILC cases with SLN diagnosed over a 12-year period (2005 to 2016) were reviewed. In all, 394/582 (68%) cases had HE(-)/CK(-) SLN. In total, 188/582 (32%) cases showed some degree of SLN involvement of which 143/582 (25%) cases had readily identifiable SLN involvement on HE slides. Overall, 45/582 (7.7%) cases had HE(-)/CK(+) SLN. The following data relate to the latter subset of 45 cases. Mean age of patients: 61 y (range: 32 to 86 y); right: 24 (53%), left: 21 (47%); multifocal and/or multicentric: 22 (49%); mean size: 2.0 cm (range: 0.25 to 4.4 cm); mean number of SLN: 2.5; mean number of involved SLN: 1.2; and cases with prior needle core or excisional biopsy: 45 (100%). CK(+) cells were identified in isolation or in loose clusters, either in subcapsular sinuses or nodal cortex or both. Overall, 30/45 (67%) showed ≤200 CK(+) cells (ie, pN0i+), and 15/45 (33%) showed200 CK(+) cells (ie, pN1mi). In total, 15/45 (33%) cases underwent axillary lymph node dissection, of which 4/45 (9%) cases were positive. cILC recurred in 3/45 (7%) cases. On statistical analyses, the number of CK(+) cells (≤/200) did not correlate with either axillary lymph node-positivity or with recurrence. Number of CK(+) cells (≤/200) readily distinguished pN0i+ from pN1mi based on AJCC's numerical criteria. CK(+) cells could be quantified in linear terms (ie, AJCC's size criteria of pN0i+ and pN1mi was applicable) in only 2 cases. On the basis of these findings, the use of CK-IHC staining should be considered for SLN in cases of cILC to ensure detection, and precise determination of extent, of involvement; however, the prognostic significance of this procedure would have to await results of additional studies with long-term follow-up.

Mammary myofibroblastoma is a benign spindle cell tumor that can show variable morphologic patterns and lines of differentiation. Diffuse and strong immunohistochemical expression of CD34 is a characteristic of myofibroblastoma and greatly aids in confirming a diagnosis. Myofibroblastoma has been shown to belong to a family of CD34-positive tumors with similar morphology that show a deletion of 13q14, which includes RB1 and FOXO1A genes. The purpose of this study was to better understand a subset of myofibroblastomas that is characteristically CD34-deficient by immunohistochemistry. Six myofibroblastomas were studied by immunohistochemistry and fluorescence in situ hybridization (FISH) for RB1. Patients included five women and one man, aged 41-85 years (median, 52.5). Tumor size ranged from 0.4 to 1.5 cm (mean, 0.95). Tumors showed spindle cell morphology in five cases and epithelioid features in one case. Two tumors showed complete lack of CD34 staining. The remaining showed weak focal or weak patchy CD34 staining. Dichotomous staining was seen in one case with CD34-positive spindle cell areas and CD34-negative myxoid areas. All six tumors showed ER expression, five of six showed desmin expression, and four of six showed bcl-2 positivity. Two of six (33.3%) tumors showed deletion of RB1 by FISH, including one that showed loss of Rb immunohistochemical staining. Myofibroblastomas uncommonly show absent/focal expression of CD34, a potential diagnostic pitfall, particularly in small samples. Characteristic staining with other immunohistochemical markers is seen which can aid in confirming the diagnosis. These tumors may harbor deletion of RB1, similar to CD34-positive myofibroblastomas, and this deletion may not correlate with loss of Rb by immunohistochemistry.

Pathologic analysis of surgical excision specimens for breast carcinoma is important to evaluate the completeness of surgical excision and has implications for future treatment. This analysis is performed manually by pathologists reviewing histologic slides prepared from formalin-fixed tissue. In this paper, we present Deep Multi-Magnification Network trained by partial annotation for automated multi-class tissue segmentation by a set of patches from multiple magnifications in digitized whole slide images. Our proposed architecture with multi-encoder, multi-decoder, and multi-concatenation outperforms other single and multi-magnification-based architectures by achieving the highest mean intersection-over-union, and can be used to facilitate pathologists' assessments of breast cancer., Accepted at Computerized Medical Imaging and Graphics

Intraoperative radiotherapy (IORT) is a novel and increasingly utilized radiation technique in the treatment of breast carcinoma. There are few reports on the histologic changes seen in breast tissue from patients who have undergone IORT. We sought to evaluate the histologic changes observed in specimens received following IORT, as well as report an unusual case which prompted our study. A retrospective review of patients who received IORT and subsequently had breast tissue histologically evaluated at our institution was performed. Fifteen post-IORT specimens from 12 patients, including the patient from the reported case, were studied. We report a case of a 77-year-old woman found to have mammographic microcalcifications at the lumpectomy site 6 months following lumpectomy and IORT for ductal carcinoma in situ (DCIS). A stereotactic biopsy showed abundant desquamated anucleate squamous cells with calcification and keratin material associated with squamous metaplasia of ducts. Carcinoma was not present. The predominant findings in the post-IORT specimens were fat necrosis and scar (n = 5), recurrent invasive carcinoma (n = 5), surgical site changes (n = 3), abscess (n = 1), and exuberant squamous metaplasia with calcification (n = 1). Five of fifteen (33%) post-IORT specimens showed squamous metaplasia, all of which were collected within 6 months of IORT delivery. The morphologic changes observed after IORT are similar to those seen after external beam radiotherapy. Exuberant squamous metaplasia is an uncommon consequence of IORT; however, pathologists should be aware of this phenomenon and review a history of prior intraoperative radiation before raising concern for malignancy.

Subareolar sclerosing duct hyperplasia (SSDH) remains to be fully characterized nearly 20 years after initial description. Thirty-five SSDH cases diagnosed over a 16-year period (January 2000 to December 2015) were reviewed. All patients were female (mean age = 59 years, range = 18-80) who had presented with a unilateral solitary lesion (left 22, right 13) with a mean size of 1.3 cm (range = 0.4-3.0 cm), and showed florid and papillary epithelial hyperplasia with dense sclerosis without involvement of nipple or areolar epidermis. Significant lesions concurrent within SSDH included low-grade adenosquamous carcinoma (n = 1), ductal carcinoma in situ (DCIS; n = 1), lobular carcinoma in situ (LCIS; n = 1), and atypical ductal hyperplasia (ADH; n = 13). No case of SSDH recurred in a mean follow-up of 44 months (range = 6-189). Subsequent significant lesions occurred in 6 patients: DCIS (n = 3; ipsilateral 2, contralateral 1), ipsilateral ADH (n = 2), and ipsilateral atypical lobular hyperplasia (n = 1). Long-term follow-up for patients with SSDH is indicated as DCIS can occur subsequently in either breast.

Aims The recent American Society of Clinical Oncology/College of American Pathologists guidelines for human epidermal growth factor receptor 2 (HER2) testing in breast cancer recommend repeat testing based on tumour grade, tumour type, and hormone receptor status. The aim of this study was to test the value of these criteria. Methods and results HER2 status was concordant in the core biopsies and excision specimens in 392 of 400 invasive carcinomas. The major reasons for discordance were amplification around the cut-off for positivity and tumour heterogeneity. Of 116 grade 3 carcinomas that were HER2-negative in the core biopsy, four were HER2-positive in the excision specimen. Three of these four either showed borderline negative amplification in the core biopsy or were heterogeneous. None of the 55 grade 1 carcinomas were HER2-positive. Review of repeat testing of HER2 in routine practice suggested that it may also be of value for multifocal tumours and if recommended by the person assessing the in-situ hybridization. Conclusions Mandatory repeat HER2 testing of grade 3 HER2-negative carcinomas is not appropriate. This is particularly true if repeat testing is performed after borderline negative amplification in the core biopsy or in HER2-negative heterogeneous carcinomas.

Recent publications have brought attention to the histopathological, immunohistochemical, and molecular aspects of the rare breast tumor resembling the tall cell variant of papillary thyroid carcinoma (BrTC). Nine archived cases of this entity were retrieved, reviewed, and compared with randomly selected tall cell variants of papillary thyroid carcinoma (ThTC). Seven of the BrTC cases as well as 5 cases of solid papillary carcinoma of breast were analyzed by Oncomine next-generation sequencing. BrTC and ThTC were histologically distinguishable by the presence of solid architecture, luminal histiocytes, and reverse polarity in the former, and psammoma bodies, giant cells, and optically clear nuclei in the latter. Sequencing revealed IDH2 R172 single-nucleotide variants in all 7 BrTCs, 6 of which had concurrent PIK3CA mutations. None of the conventional solid papillary carcinomas demonstrated IDH2 mutation. BrTC bears superficial resemblance to other papillary tumors but is unique in terms of histology and molecular profile.

Summary Tumor-infiltrating lymphocytes (TILs) have emerged as prognostic in triple-negative breast cancer (TNBC). We aimed to assess the consistency of hotspot placement and TIL enumeration among multiple pathologists. Additionally, we assessed hotspot TIL count consistency by comparing hotspot counts in 3 separate locations within a single whole-tissue section. Anti-CD8 immunohistochemistry was performed on a representative section from 66 cases of primary TNBC, which were then scanned as whole-slide images. Quantification of the tissue area and combined stromal and intratumoral CD8+ TILs was performed using digital image analysis (DIA) within 2.2 mm–diameter circle hotspots. TIL counts were quantified as absolute counts and densities (absolute count/tissue area in micrometers2). For each case, 6 pathologists placed a single hotspot, defined as an area with the subjectively highest CD8+ immunoreactivity, within the tumor bed. Separately for each case, a single pathologist placed hotspots in 3 different locations within a single tumor section. Intraclass correlation coefficients (ICCs) were generated following TIL enumeration via DIA. ICCs for single hotspot placement by 6 pathologists were 0.96 for density and 0.97 for absolute counts, respectively. In 32% of cases (21/66), all the hotspots placed by the 6 pathologists were in the same location. When evaluating hotspots in 3 different locations within a tumor, the ICC was 0.95 for both density and absolute counts. Hotspot evaluation by DIA is a reproducible method for CD8+ TIL quantification, and the use of hotspots may reduce TIL count variation caused by intratumoral TIL heterogeneity.

Granulomatous lobular mastitis (GLM) is an uncommon condition that typically occurs in parous, reproductive-aged women and can simulate malignancy on the basis of clinical and imaging features. A distinctive histologic pattern termed cystic neutrophilic granulomatous mastitis (CNGM) is seen in some cases of GLM and has been associated with Corynebacterium infection. We sought to further characterize the clinical, imaging, and histopathologic features of CNGM by studying 12 cases and attempted to establish the relationship of this disease with Corynebacterium infection. Patients were women ranging in age from 25 to 49 years (median: 34 y), and all presented with a palpable mass that was painful in half of the cases. In 2 of 9 cases, imaging was highly suspicious for malignancy (BI-RADS 5). CNGM was characterized by lobulocentric granulomas with mixed inflammation and clear vacuoles lined by neutrophils within granulomas. Gram-positive bacilli were identified in 5/12 cases. In 4 patients, the disease process worsened after the diagnostic core biopsy, with the development of a draining sinus in 2 cases. No growth of bacteria was seen in any microbial cultures. No bacterial DNA was identified by 16S rDNA polymerase chain reaction for 1 case that showed gram-positive bacilli on histology. Patients were treated with variable combinations of surgery, antibiotics, and steroids. The time to significant resolution of symptoms ranged from 2 weeks to 6 months. Similar to other forms of GLM, CNGM can mimic malignancy clinically and on imaging. When encountered in a needle core biopsy sample, recognition of the characteristic histologic pattern and its possible association with Corynebacterium infection can help guide treatment.