Abstract P1-06-03: Validating the link between obesity and breast inflammation in women with breast cancer (BC)

Background: In post-menopausal women, obesity is a risk factor for the development of BC that expresses the estrogen and progesterone receptors (ER/PR). In mouse models of obesity, we previously described crown-like structures (CLS), consisting of macrophages surrounding dead adipocytes in white adipose tissue (WAT) of the mammary gland, which were associated with increased levels of proinflammatory mediators known to be involved in carcinogenesis. We translated these findings to women (n = 30), and provided the first evidence of CLS in the human breast (CLS-B). The presence and severity of CLS-B (CLS-B index) correlated with elevated body mass index (BMI), increased adipocyte size, activation of NF-κB, and increased levels of proinflammatory mediators (TNF-α, IL-1β, COX-2 and PGE2) and aromatase. We expanded our population to prospectively validate these preliminary findings.

Methods: We prospectively collected WAT from women undergoing breast and reconstructive surgery. WAT was subjected to immunohistochemistry for CD68, a macrophage marker, to detect CLS-B by light microscopy. Adipocyte diameter was measured on photomicrographs using the Canvas 11 Software. Endpoints were 1) CLS-B presence/absence and 2) CLS-B index (proportion of slides with CLS-B). Associations between CLS-B and clinicopathologic features were analyzed using logistic regression and Fisher's exact test.

Results: From 04/2010-02/2012, WAT (100 mastectomy and 5 abdominal reconstructions) was obtained from 101 women; median age 49 (range 26-80). CLS-B were found in 54 (53%) patients (pts). CLS-B were seen in 9/37 (24%) normal weight pts (BMI <25), 23/39 (59%) overweight pts (BMI 25-29.9), and 22/25 (88%) obese pts (BMI ≥30). Pts with CLS-B had significantly larger average adipocyte diameter (106.5 +/- 11.5 microns) compared to those without CLS-B (91.5 +/- 16.1 microns; p<0.001). Consistently, CLS-B index correlated with BMI (p<0.001) and adipocyte size (p<0.001). Breast inflammation was seen in pts with all tumor phenotypes: CLS-B were seen in 24/41 (59%) pts with ER/PR+, HER2- tumors; 7/16 (44%) pts with HER2+ tumors; and 3/10 (30%) pts with ER/PR/HER2- tumors. A higher CLS-B index was seen in WAT from ER+ tumors, but this was not statistically significant (p = 0.08). Regular use of nonsteroidal antiinflammatory drugs was protective against CLS-B (p = 0.17 for association with CLS-B, and p = 0.04 for association with CLS-B index in multivariable analyses). Among 25 pts with bilateral breast WAT, concordant CLS-B findings (+/-) were found in 20 (80%) pts. Among pts with paired breast and abdominal WAT, concordant findings were seen in 4/5 (80%) pts.

Conclusions: Findings from this prospective study, the largest reported to date, extend our previous observation that CLS-B are associated with BMI and adipocyte size. These results provide a plausible pathophysiological link between obesity and BC. Breast inflammation occurs in association with all BC phenotypes. Preliminary data suggest concordance between breasts and between abdominal and breast WAT. Hence, abdominal WAT may prove useful as a surrogate for breast WAT; biopsies of abdominal subcutaneous WAT are more easily done, which could prove useful in developing interventions to attenuate WAT inflammation.

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P1-06-03.

P5-14-13: Favorable Prognosis in Patients with T1a,b Node-Negative Triple Negative Breast Cancers Treated with Multimodality Therapy

Purpose: To evaluate the clinical characteristics, natural history and outcomes in patients with ≤1cm, node-negative triple negative breast cancer (TNBC).

Materials and Methods: After excluding patients who received neoadjuvant therapy, 1,022 TNBC patients who received definitive breast surgery from 1999 to 2006 were identified from an institutional database. Among these, 194 patients had node-negative tumors ≤1cm and comprise the study population. Clinical data was abstracted and survival outcomes were analyzed.

Results: Median follow-up time was 71 months (range 2–143). Median age at diagnosis was 55.5 years (range 27–84). T stage was T1mic in 16 (8.2%), T1a in 49 (25.3%), T1b in 129 (66.5%). The majority of tumors were poorly differentiated (N= 142, 73%), lacked lymphovascular invasion (N= 170, 87.6%) and were screening-detected (69%, N=134). Breast-conserving surgery (BCS) was employed in 129 (66.5%) and mastectomy in 65 (33.5%) patients. 113 (58%) patients received adjuvant chemotherapy and 123 (63%) received whole breast radiation. Patients who received chemotherapy tended to have more adverse clinical and disease features (younger age,T1b, poor tumor grade; all p<0.05). For the entire group, 5 year local recurrence-free survival was 96% and distant metastasis-free survival was 95%, with no difference in distant relapse rates between T1mic/T1a vs. T1b patients (94.5% vs 95.5%, p=0.81 )or by receipt of chemotherapy (95.9% vs 94.5%, p=0.63).

Conclusion: Excellent 5-year locoregional and distant control rates were achievable in patients with TNBC tumors ≤ 1.0 cm, 58% of whom received chemotherapy. These results identify a group of TNBC patients with favorable outcomes following early detection and multimodality treatment.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P5-14-13.

Patient characteristics associated with the decision to undergo bilateral prophylactic mastectomy for lobular carcinoma in situ

148

Background: Women at increased risk for breast cancer due to a diagnosis of lobular carcinoma in situ (LCIS) have three management options: high risk surveillance +/- chemoprevention (CP) or bilateral prophylactic mastectomy (BPM). Among a large cohort of women with LCIS, we previously reported there were no differences between women choosing CP compared to those choosing surveillance alone. The purpose of this study was to identify patient factors associated with the decision to pursue BPM for LCIS.

Methods: We reviewed our prospectively maintained LCIS database to identify women choosing BPM (1995-2009). Comparisons were made between patients who chose BPM versus those who chose high risk surveillance +/- CP.

Results: Among 995 pts with LCIS, 795 (80%) chose surveillance alone, 149 (15%) chose CP and 51 (5%) chose BPM. Compared to patients electing surveillance +/- CP patients who pursued BPM were younger at age of LCIS diagnosis (48 vs. 52 yrs, p < 0.001), more likely to have bilateral biopsies with LCIS (14% vs. 3%, p = 0.002) and more likely to be premenopausal (82% vs. 57%, p < 0.001). Patients choosing BPM were also more likely to have additional risk factors for breast cancer including: stronger family histories and extremely dense breasts (Table). Occult breast cancer was found in 4 (8%) BPM patients. At a median follow-up of 54 months (range 0-190 mos), 120/944 (13%) patients in the surveillance +/- CP group developed breast cancer.

Conclusions: High-risk surveillance alone is the preferred management option for women with LCIS at our institution. Patients electing BPM are younger and more likely to have other associated risk factors for developing breast cancer. Further research to define how the increased risk imparted by LCIS is augmented by these factors may allow for better risk stratification and more informed discussions with patients.

[Table: see text]

Abstract P4-06-07: PI3K Mutations Are More Common in Low Grade DCIS

Background: DCIS is a non-obligate precursor to invasive breast cancer yet it remains difficult to predict which lesions will progress or over what time interval. Factors associated with a higher risk of progression include high histologic grade and HER2 amplification. HER2 signaling via the PI3K pathway is associated with accelerated invasiveness in laboratory models and mutations in the helical domain of PI3K have been associated with a worse prognosis in invasive cancer. We hypothesized that PI3K pathway mutations are associated with higher risk of progression in DCIS and therefore would be more common in high grade DCIS.

Methods: 195 cases of pure DCIS were identified from the MSKCC breast service database (1999-2003). All cases were reviewed by a single pathologist to assign histologic grade. Sections were obtained from archived formalin-fixed paraffin embedded (FFPE) blocks for manual microdissection to isolate pure DCIS lesions for DNA extraction. Multiplex array (Sequenom®) genotyping for PIK3CA was performed on prePCR amplified DNA. Comparisons were made between high and low grade DCIS using Fisher's exact test.

Results: Among 195 pure DCIS cases, 89 were classified as high grade and 106 as low grade. Sequenom® analysis was informative in all cases. PIK3CA mutations were identified in 4/89 (4%) cases of high grade DCIS and in 24/106 (23%) cases of low grade DCIS (p=0.000). All 4 mutations in high grade DCIS were located in the kinase domain, whereas in low grade DCIS, 9/24 (37%) mutations were in the kinase domain (H1047R) and 15/24 (67%) mutations were in the helical domain (E542K, E545K, N345K).

Conclusion: PI3K mutations were relatively uncommon in pure high grade DCIS as compared to low grade DCIS. PI3K mutations in low grade DCIS were observed in both the kinase and helical domain. These findings support the hypothesis that breast tumorigenesis differs by grade and PI3K mutations may be more prominent in low grade carcinogenesis. The significance of helical domain mutations in low grade lesions requires further investigation.

ACOSOG Z0010: A multicenter prognostic study of sentinel node (SN) and bone marrow (BM) micrometastases in women with clinical T1/T2 N0 M0 breast cancer

CRA504

Background: SN biopsy (SNB) with immunohistochemistry (IHC) of histologically negative SN identifies metastases (mets) not seen by standard histology. The impact of IHC-detected BM mets has been reported in several large single-institution studies. 5,539 patients (pts) were entered into this prospective multicenter observational study to determine the clinical significance of SN and BM mets.

Methods: Patients underwent lumpectomy and SNB with bilateral iliac crest BM aspiration. BM and histologically negative SN were evaluated with IHC in a central laboratory (results not clinically reported). Overall survival (OS), disease-free survival, and locoregional recurrence were determined. Results with OS (the primary endpoint) are reported here.

Results: SN were successfully identified in 5,184 of 5,485 pts (94.5%). Histologic SN mets were found in 1,239 pts (23.9%). IHC detected an additional 350 pts (10.5%) with SN mets. BM mets were identified by IHC in 105 of 3491 examined (3.0%). 5-yr overall survival is shown in the Table . BM IHC positivity significantly predicted decreased OS (p=0.015). A multivariable analysis that included SN and BM status, ER, PR, grade, size, and age showed that neither IHC detected mets in SN (p=0.66) or BM (p=0.08) were independent predictors of OS, although BM status showed a strong trend.

Conclusions: The detection of BM mets by IHC in pts with clinical T1/2 N0M0 breast cancer identifies those pts at significantly increased risk for death; the impact of BM mets on outcome supports and confirms prior studies. In this study, SN IHC-detected mets appear to have no significant impact on OS. The routine examination of SN by IHC is not supported in this patient population by this study.

[Table: see text]

[Table: see text]

ACOSOG Z0011: A randomized trial of axillary node dissection in women with clinical T1-2 N0 M0 breast cancer who have a positive sentinel node

CRA506

Background: Sentinel node biopsy (SNB) eliminates the need for axillary dissection (ALND) in patients whose sentinel node (SN) is tumor-free. However, completion ALND remains the gold standard for patients with a tumor-involved sentinel node. ALND achieves regional control, but its effect on survival remains controversial. The main objective of ACOSOG Z0011 was to compare outcomes of patients with hematoxylin and eosin (H&E) detected metastasis in SN managed with or without ALND and no axillary irradiation.

Methods: Clinically node-negative patients who underwent SN biopsy and had 1 or 2 SN with metastases detected by H&E were randomized to ALND or no further axillary specific treatment. All patients were treated with lumpectomy and opposing tangential field irradiation. Adjuvant systemic therapy was at the discretion of their physicians. Overall survival (OS), disease-free survival (DFS), and locoregional control were evaluated.

Results: 446 patients were randomized to SNB alone and 445 to SNB plus ALND. Patients treated with SNB alone were similar to those treated with SNB + ALND with respect to age, tumor size, Bloom-Richardson score, estrogen receptor status, adjuvant systemic therapy, tumor type, and T stage. Patients randomized to SNB alone had a median of two lymph nodes removed whereas patients randomized to ALND had a median of 17 lymph nodes removed. 17.6% of ALND patients had 3 or more involved nodes compared to 5.0% of SNB patients (p < 0.001). Median follow-up is 6.2 years. 5-year in breast recurrence after ALND was 3.7% compared to 2.1% for SNB (p = 0.16) while 5-year nodal recurrence was 0.6% compared to 1.3% (p = 0.44) respectively. The five-year OS for patients undergoing SNB + ALND is 91.9% compared to 92.5% for SNB alone (p = 0.24), and DFS is 82.2% compared to 83.8% respectively (p = 0.13).

Conclusions: Despite the widely held belief that ALND improves survival, no significant difference was recognized by this study of SN node-positive women. Although the study closed early because of low accrual/event rate, it is the largest phase III study of ALND for node-positive women, and it demonstrates no trend toward clinical benefit of ALND for patients with limited nodal disease.

[Table: see text]

Negative Estrogen Receptor and HER2 Assays at Core Biopsy of Invasive Cancers Should Be Confirmed in the Surgical Specimens

Background: Management of patients with invasive breast carcinoma is determined by the expression of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) in the tumor cells. It is therefore critical to accurately assess ER, PR, and HER2 (ERPRH2) and avoid false-negative results that could lead to withholding of potentially beneficial therapy. Most laboratories, including ours, evaluate ERPRH2 status of invasive carcinoma on core biopsy (CB) material. We also routinely retest the cancer in the surgical specimen (SS) for any marker negative on CB. Our study evaluates discrepancies in ER and HER2 results in CB and SS to determine if repeat testing is necessary.Methods: Retrospective review of the pathology database found 186 invasive carcinomas from 181 patients in which ERPRH2 had been performed on the CB, and negative stains repeated on the SS. For the purpose of this study, immunoperoxidase stains (IHC) were repeated simultaneously on the CB and SS for the marker with discordant result, using the same antibody. Two study pathologists reviewed all IHC. HER2 FISH was performed on the SS for all cases with a discordant HER2 IHC result.Results: The concordance rate between CB and SS results was 97.3% (181/186) and discordant results were found in 5/186 (2.7%) cases (Table 1). Three cases were discordant in ER (including 2 cases that were PR(-) on CB and SS) and two cases were discordant for HER2. Discrepancy was due to intratumoral heterogeneity in 2 cases. The CB sampled the HER2(-) area of the tumor in one case, but staining of a larger section unveiled the positive focus; the positive result was confirmed by amplification of HER2 detected by FISH. The second case was an invasive lobular carcinoma (IL) with mixed classical and histiocytoid morphology; only the classical IL was ER(+), but the CB had sampled the histiocytoid IL. In two other cases the discrepancy resulted from technical error. Both cases had been prospectively interpreted as ER(-) on CB and ER(+) on SS, but repeat IHC for ER showed positive staining in both the CB and SS. The fifth case was HER2 (0/1+) on CB but equivocal (1-2+) in the SS, where a larger portion of tumor was evaluated; the equivocal result led to reflex HER2 FISH with detection of low level HER2 amplification. Relying solely on the CB would have resulted in the misclassification of 2 tumors as triple negative; repeat stains on the SS showed that one was HER2(+) and the other ER(+).Conclusions: Concordance in the ER and HER2 results between CB and SS was high (97.3%), but 2.7% of cases showed discordant findings. Factors associated with discordance included intratumoral heterogeneity tumor, technical error, and equivocal findings. A triple negative profile on CB converted to either ER(+) or HER2(+) after staining on SS in 2 cases (40% of discordant cases and 1.1% of all cases), impacting patient management. Our findings suggest that any ER and/or HER2 negative result obtained at CB should be confirmed on the SS to ensure appropriate patient management.Findings of discordant casesCaseER (%)HER2HER2 FISH CBSSCBSSSS1OOO3+4.42O951+1+N/P3O1001+1+N/P45501+1+N/P590N/P0-1+1-2+2.6N/P- not performed

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 6008.

Clinical Management Factors Contribute to the Decision for Contralateral Prophylactic Mastectomy (CPM)

INTRODUCTION:Rates of CPM are reported to be increasing, yet factors driving this trend are unclear. We performed this analysis to determine if increasing rates of CPM are being driven by increased recognition of risk factors for contralateral breast cancer (CBC) or by treatment factors related to the index lesion.METHODS:From 1997-2005, 2967 patients with Stage 0-III primary unilateral breast cancer underwent mastectomy at MSKCC and were entered into a prospectively maintained database. Demographic, tumor and treatment factors were abstracted and comparisons made between patients who did and did not undergo CPM within one year of treatment for their index cancer. Generalized estimating equations were used to fit logistic regression models to identify independent predictors of CPM.RESULTS:The overall rate of CPM was 13.8%(n=408), increasing from 6.7% in 1997 to 24.4% in 2005 (p&lt;0.0001). The median age of CPM pts was 44.9 vs 53.2 in the non-CPM group (p&lt;.001) and only 7% of CPM pts were non-white compared to 25% of non-CPM pts (p&lt;.001). Although 69% of CPM pts had a family history (FH) of breast cancer (vs 40% non-CPM; p&lt;.001) only 8% had 2 or more first degree relatives affected. Genetic testing was performed in 29% of CPM pts; of those 37 (31%) were positive. The use of MRI increased from 1.3% to 36.3%of cases over the study period. MRI at diagnosis (43% vs 16%) and MRI generated biopsy in the contralateral or bilateral breasts (29% vs 4%) were strongly associated with CPM (p&lt;.0001). Prior attempts at breast conservation (BCT) (28% vs 16%; p&lt;.001) were more common in the CPM group, but number of attempts did not differ between groups. Patients undergoing CPM were more likely to have DCIS versus an invasive histology (p&lt;0.0001), and of those with invasive disease, CPM patients had smaller tumors (1.2cm vs. 1.8cm, p&lt;0.0001) and were more likely to be node negative (53% vs 43%, p&lt;0.0001). ER, PR, and HER2 status were not associated with CPM. CPM rates among surgeons ranged from 9.8% to 26%. Multivariate analysis of predictors of CPM for 2387 patients with complete data is shown in the Table.VariableORp95%CIWhite race3.6&lt;0.00012.4-5.4Age&lt;502.3&lt;0.00011.8-3.1FH breast cancer2.9&lt;0.00012.3-3.7MRI at Dx2.2&lt;0.00011.7-2.8BCT attempted1.60.000081.2-2.1Reconstruction3.2&lt;0.00012.3-4.5DCIS histology1.40.021.1-1.9*adjusted for surgeonCONCLUSIONS: Factors associated with clinical management of the index cancer such as preoperative MRI with the potential for additional biopsy, failed attempt at BCT, and breast reconstruction were strongly associated with CPM. Age and FH were also independent predictors; however the FH profiles of CPM pts in this series do not support increased recognition of mutation carriers and truly high risk FH as a major cause of increasing rates of CPM. The lack of association with ER status, which results in treatment that decreases the risk of CBC, provides additional support that patients may be choosing CPM for reasons other than future risk. These data suggest that the need for additional procedures during management of the index cancer may be contributing to the increasing use of CPM among a relatively moderate risk patient population. Efforts to optimize BCT and minimize unnecessary tests may help to curb this trend.

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 38.

Impact of HER2 Status on Risk of Sentinel Node Metastasis: An Independently Validated Multivariable Model

Purpose The presence of axillary lymph node metastases is dictated by tumor biology and is a strong prognostic indicator in breast cancer. We have previously published a user-friendly nomogram that provides a risk estimate for sentinel lymph node (SLN) metastasis in women with breast cancer (Bevilacqua et al. 2007, J Clin Oncol 25: 3670). At that time, HER2 testing was not uniformly performed, but is now standard of care. The purpose of this study was to determine if the addition of HER2 status or grouping of patients by molecular subtype improves the prediction of SLN metastasis.Patients and Methods: The ability of clinical and pathologic features to predict the presence of SLN metastasis in patients presenting with clinically node negative invasive breast cancer, was assessed with multivariable logistic regression (MVA) for 4723 sequential SLN biopsy procedures with known ER, PR, and HER2 from 1996 to 2004. HER2 status was defined as positive if IHC=3+ and/or FISH≥2. The modeling (n=3297) and validation (n=1426) groups were identified by simple random sampling. Two models were created using the modeling population: one included ER, PR, and HER2 as separate variables and one combined these markers into 4 subtypes defined as: Luminal A-like=ER or PR +, HER2 −; Luminal B-like=ER or PR +, HER2 +; HER2-like=ER and PR −, HER2 +; Basal-like=ER and PR −, HER2 −. The validation group was used to assess the calibration (intercept, slope, Emax, Eavg) and discrimination (AUC, area under the receiver operating curve) of the models.Results: In addition to age, tumor size, tumor type, lymphovascular invasion, tumor location, and multifocality, subtype was found to be an independent predictor of SLN metastasis on MVA (p=0.003). The Basal-like subtype was associated with a significantly lower risk of SLN metastasis (compared with referent Luminal A-like subtype, OR=0.58). HER2 alone was not found to be an independent predictor of SLN metastasis (p=0.56), while ER and PR remained significant (p=.05, p=.02). Compared to our previous model without HER2, the AUC was slightly decreased by the addition of either HER2 or subtype into the model, but the calibration was slightly improved (Table 1). DiscriminationCalibrationModelAUCInterceptSlopeEmaxEavgPrevious Model0.735- 0.0440.8690.0390.020Mew model with HER20.733- 0.0030.8890.0270.019New model with subtype0.731- 0.0010.8870.0270.020Table 1: Discrimination and calibration measures comparing the new model (with HER2 alone or subtype) with the previously published model (without HER2 or subtype).Conclusion: The addition of subtype, defined by combining HER2 status with ER and PR, slightly increases the calibration of the new model but does not increase its discrimination compared to the previous model. Breast tumor subtype is a significant independent predictor of risk of SLN metastasis, with basal-like subtype having a lower risk of SLN metastasis.

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 1004.