Neighbourhood species richness and drought-tolerance traits modulate tree growth and δ13 C responses to drought

Mixed-species forests are promoted as a forest management strategy for climate change adaptation, but whether they are more resistant to drought than monospecific forests remains contested. In particular, the trait-based mechanisms driving the role of tree diversity under drought remain elusive. Using tree cores from a large-scale biodiversity experiment, we investigated tree growth and physiological stress responses (i.e. increase in wood carbon isotopic ratio; δ13 C) to changes in climate-induced water availability (wet to dry years) along gradients in neighbourhood tree species richness and drought-tolerance traits. We hypothesized that neighbourhood species richness increases growth and decreases δ13 C and that these relationships are modulated by the abiotic (i.e. climatic conditions) and the biotic context. We characterised the biotic context using drought-tolerance traits of focal trees and their neighbours. These traits are related to cavitation resistance versus resource acquisition and stomatal control. Tree growth increased with neighbourhood species richness. However, we did not observe a universal relief of water stress in species-rich neighbourhoods. The effects of neighbourhood species richness and climate on growth and δ13 C were modulated by the traits of focal trees and the traits of their neighbours. At either end of each drought-tolerance gradient, species responded in opposing directions during dry and wet years. We show that species' drought-tolerance traits can explain the strength and nature of biodiversity-ecosystem functioning relationships in experimental tree communities experiencing drought. Mixing tree species can increase growth but may not universally relieve drought stress.

A Radiation Therapy Contouring Atlas for Delineation of the Level I-II Axilla in the Prone Position

PURPOSE/OBJECTIVE(S)

When patients are treated prone for whole breast irradiation (WBI), the axilla typically receives less dose than when patients are treated with WBI in the supine position. There are situations, however, where the axilla is a target as in a situation where the physician intends to treat with "high tangents" to provide good coverage of both the breast and level I-II axilla. In these scenarios, ideal target delineation when the patient is in the prone position is not well-defined. While different consensus guidelines exist for delineation of the nodal areas in the supine position, to our knowledge there are no contouring guidelines for the regional nodes in the prone position based on bone, skin, vascular and muscle landmarks.

MATERIALS/METHODS

Forty-three patients treated with high tangents in the prone position from 2012-2018 were identified as representative cases. The level I and II regional nodal contours from the Radiation Therapy Oncology Group (RTOG) breast cancer atlas were adapted for prone position by a radiation oncologist and a breast radiologist based on anatomic considerations and changes observed from supine to prone positioning on diagnostic imaging. The revised nodal contours were reviewed by an expanded expert multidisciplinary panel including additional breast radiation oncologists and surgical oncologists to delineate the level I and II axilla on noncontrast computed tomography (CT) scans.

RESULTS

We adapted the RTOG breast cancer atlas, supported by detailed figures, in order to create a CT based atlas of the level I and II axillary lymph node stations in the prone position. For the level I axilla, the cranial and caudal anatomic boundaries remain unchanged. With transition to the prone position from supine, tenting of the pectoralis major occurs displacing the muscle from the chest wall and shifting the axillary space anteromedial to the lateral border of the pectoralis major. Therefore, the anterior boundary is now defined by the plane of the anterior extent of the pectoralis major to skin. The medial boundary is defined by the plane of the lateral border of the pectoralis major and pectoralis minor including to ribs and intercostal muscles. The lateral boundary is defined by the skin, cropped by 5mm. The posterior boundary is defined by the plane of the anterior surface of the latissimus dorsi and subscapularis muscle to skin. For the level II axilla, the cranial, posterior, medial and lateral boundaries remain unchanged. With transition from supine to prone, the axilla exhibits an anterolateral shift, now laterally abutting the lateral border of the pectoralis minor. The anterior boundary is now defined as the posterior aspect of the pectoralis major muscle. The caudal boundary is where the pectoralis minor inserts into ribs.

CONCLUSION

The adaptations to the RTOG breast cancer atlas for prone positioning will enable radiation oncologists to more accurately target the level I and II axilla when treating patients prone in whom the axilla is a target in addition to the breast.

Effectiveness and Toxicity of Five Fraction Prone Accelerated Partial Breast Irradiation

PURPOSE/OBJECTIVE(S)

Accelerated partial breast irradiation (APBI) after breast conserving therapy (BCT) is increasingly used to treat women with early stage breast cancer. Our institution was an early adopter of 5-fraction ABPI and delivers APBI primarily to patients in the prone position. This study reports long term oncologic and cosmetic outcome in a large cohort of women treated with 5-fraction external beam APBI.

MATERIALS/METHODS

We included patients receiving APBI 600 cGy × 5 fx delivered every other day or every day between 2010 and 2022. Late toxicities and clinician and patient rated cosmesis were evaluated for patients with > 6 month follow up. Univariate and multivariate logistic regression models were used to identify clinical and dosimetric factors associated with development of acute and late toxicities, clinician and patient rated cosmesis. All statistical tests were two-sided, and the null hypothesis was rejected for p<0.05. Kaplan Meier methodology was used to calculate overall survival (OS), disease-free survival (DFS) and locoregional recurrence-free survival (LR-RFS).

RESULTS

Four hundred and forty-two patients received APBI either daily (56%) or every other day (44%). Most of the patients (92%) were treated in the prone position. Average mean heart dose was 23 cGy for left-sided and 11 cGy for right-sided breast cancers. Ipsilateral lung V30% ≤ 30%. At a median follow up of 48 months (range: 5.96 - 155 months), 12 (2.7%) patients developed a local recurrence, 14 (3.2%) patients developed a contralateral breast primary, 10 patients (2.3%) developed a distant metastasis and one patient (0.2%) developed a local recurrence followed by a distant metastasis 1 month later. Out of 258 patients with > 6 month follow up, rates of late grade 1-2 telangiectasia, fibrosis, edema, atrophy and hyperpigmentation were 4%, 18%, 1%, 19% and 7% respectively. There was only one late grade 3 event in a patient who developed significant breast atrophy. The rate of good-excellent physician and patient rated cosmesis was 95% and 89% respectively (N = 256 at median follow up of 80 months). On multivariate logistic regression, patients who did not receive any adjuvant endocrine or chemotherapy were at increased risk of developing a local recurrence. Patients with PR negative disease were at increased risk of distant metastasis. Patient who experienced any grade of acute dermatitis during treatment were at increased risk of any high grade (grade ≥ 2) late adverse event and worse physician rated cosmesis. Daily or every other day treatment did not correlate with worse toxicity or clinical outcomes. Plastic surgery involvement, LVI, EIC, lobular histology, and ER negativity did not correlate with an increased risk of recurrence. Five-year LR-RFS, DFS and OS were 98%, 92.5% and 98.6% respectively.

CONCLUSION

Five- fraction APBI delivered primarily in the prone position either daily or every other day was effective with low rates of local recurrence, minimal toxicity and excellent cosmetic scores at long term follow up.

Abstract P4-10-11: Pregnancy-associated breast cancer in a contemporary cohort of newly diagnosed women

Introduction: Pregnancy-associated breast cancer (PABC) refers to breast cancer (BC) diagnosed during pregnancy, lactation, or in the postpartum period. There is evidence that PABC is associated with a poorer prognosis, and that the development of the disease is influenced by the unique hormonal milieu of pregnancy. The purpose of this study was to investigate the clinicopathologic characteristics associated with PABC in a contemporary cohort of women with newly diagnosed BC.

Methods: Our institutional Breast Cancer Database was queried for women diagnosed with breast cancer between 2010-17 who had at least one full term pregnancy (FTP). Variables of interest included patient demographics and clinical and tumor characteristics. PABC was defined as breast cancer diagnosed within 24 months of delivery. Statistical analyses included Pearson's chi-square and logistic regression.

Results: Out of a total of 1934 women, 42 (2.2%) had PABC. Median follow up in the total cohort was 4.5 years. After adjusting for age at diagnosis, PABC was associated with older age at first FTP, ethnic minority status, BRCA mutation carriers, presentation with a palpable mass, higher histologic grade, and ER-negative and triple negative receptor status. Variables that were not significantly different between PABC and non-PABC cases included tumor histology, multifocality, presence of lymphovascular invasion, and family history of breast cancer.

Table:Selected Characteristics of Women with PABCVariableNon-PABC (n=1892)PABC (n=42)P-value*Age at first full term pregnancy  &lt;0.001&lt;35 years1610 (85%)28 (66.7%) ≥35 years277 (15%)14 (33.3%) Race  0.001White1397 (73.8%)23 (54.8%) Black181 (9.6%)8 (19%) Asian175 (9.2%)10 (23.8%) Hispanic131 (6.9%)1 (2.4%) Other8 (0.4%)0 (0%) BRCA 1,2 Positive56 (3%)9 (21.4%)&lt;0.001Method of Presentation  0.002Breast Exam579 (30.6%)30 (71.4%) Mammography1137 (60.1%)10 (23.8%) Ultrasound87 (1.6%)2 (4.8%) MRI67 (3.5%)0 (0%) Other22 (1.2%)0 (0%) Invasive Grade  0.014Low213 (15%)0 (0%) Intermediate763 (53.8%)12 (37.5%) High442 (31.2%)20 (62.5%) Estrogen Receptor  0.034Positive1572 (83.9%)29 (69%) Negative301 (16.1%)13 (31%) Triple Negative135 (7.1%)7 (16.7%)0.041*P-values are age-adjusted.

Conclusions: The association of PABC with ethnic minority status in our cohort is interesting and may be reflected in the increased proportion of triple negative breast cancers in the PABC group. In our contemporary cohort, PABC was associated with older age at first FTP. As more women delay childbearing, risk for PABC may increase. Our findings suggest that women who become pregnant at older ages should be followed carefully during pregnancy and in the postpartum period, especially if they are BRCA mutation carriers. The optimal approach for monitoring older women during pregnancy and the postpartum period is unclear. Clinical breast exam may play an important role, especially for those women known to be at increased risk for breast cancer.

Citation Format: Gooch JC, Chun J, Jubas T, Guth A, Schnabel F. Pregnancy-associated breast cancer in a contemporary cohort of newly diagnosed women [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 P4-10-11.

Abstract P4-10-07: Breastfeeding experience among breast cancer patients in the modern era

Introduction: In recent years, the uptake of breastfeeding has become more common as it is regarded as healthy and beneficial for both mother and infant. The process of parturition and lactation plays a role in the normal differentiation and development of breast tissue, and multiparity has been associated with a decreased risk for breast cancer. The current study sought to describe the breastfeeding experience of a modern cohort of women with newly diagnosed BC, and to examine the clinicopathologic characteristics of their disease.

Methods: A retrospective review of our institutional Breast Cancer Database from 2009-2017 was performed to identify women with at least one full term pregnancy (FTP). Clinicopathologic and demographic information was recorded, including breastfeeding experience and cumulative duration of nursing. Women were grouped by self-reported breastfeeding experience and duration of breastfeeding for analysis. Pearson's chi-square tests were performed.

Results: Of 1919 patients, 1053 (54.9%) reporting breastfeeding. Breastfeeding increased from a low of 30.4% among women with first FTP (FFTP) in the 1950's to 84.6% with FFTP in the 2010's. There were no significant differences between those who did and did not breast feed with regards to race, family history, BRCA status, pathologic stage, grade, tumor histology, lymphovascular invasion (LVI), multifocality, tumor size or receptor status. When stratified by duration of breastfeeding, the most striking finding was that women who breastfed for &gt;12 months were more likely to have tumors associated with LVI (p = 0.028).

Table– Breastfeeding Experience Among Parous Women with Breast CancerVariableNo Breastfeeding (n=866)Breastfeeding (n=1053)p-valueRace  0.432White648 (74.8%)767 (72.8%) Black73 (8.5%)112 (10.6%) Asian78 (9%)102 (9.7%) Hispanic64 (7.4%)67 (6.4%) Other3 (0.3%)5 (0.5%) Family history272 (31.4%)311 (29.6%)0.397BRCA 1,2 positive23 (2.7%)42 (4.0%)0.108Final Pathology Stage  0.2240190 (21.9%)222 (21.1%) I426 (49.2%)507 (48.1%) II197 (22.8%)229 (21.7%) III40 (4.6%)63 (6.0%) IV1 (0.1%)3 (0.3%) No residual (neoadjuvant)12 (1.4%)29 (34.1%) Invasive Grade  0.127Low92 (14.1%)120 (15.3%) Intermediate371 (56.7%)398 (50.6%) High191 (29.2%)268 (34.1%) Histology  0.130DCIS189 (21.8%)223 (21.2%) IDC531 (61.3%)688 (65.3%) ILC113 (13.1%)99 (9.4%) Other33 (3.8%)43 (4.1%) LVI127 (14.7%)174 (16.5%) Multifocality147 (17%)183 (17.4%) Median tumor size (cm; range)1.4 (0-9.5)1.3 (0-12.5)0.489Estrogen Receptor  0.206Positive726 (84.7%)861 (82.6%) Negative131 (15.3%)182 (17.4%) Progesterone Receptor  0.275Positive621 (72.5%)732 (70.2%) Negative236 (27.5%)311 (29.8%) HER2/neu Receptor  0.068Positive78 (12%)121 (15.4%) Negative571 (88%)667 (84.6%) 

Conclusions: Breastfeeding experience was not generally associated with significant differences in tumor or patient characteristics. However, breastfeeding for longer than 12 months was associated with LVI. It is possible that changes in the breast tissue that occur during the process of pregnancy and prolonged lactation may influence future tumor development. These findings are hypothesis generating and suggest that the relationship of prolonged breastfeeding and breast cancer development should be investigated further.

Citation Format: Gooch JC, Chun J, Jubas T, Guth A, Schnabel F. Breastfeeding experience among breast cancer patients in the modern era [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 P4-10-07.

Abstract P5-02-01: The relationship of breast density in mammography and magnetic resonance imaging (MRI) in women with triple negative breast cancer (TNBC)

Introduction:

TNBC represent 10%–20% of invasive breast cancers. Previous studies showed that TNBC usually present with benign features on mammography, ultrasound and MRI. However, there is a dearth of information on the relationship of mammographic breast density (MBD), background parenchymal enhancement (BPE) and fibroglandular tissue (FGT) on MRI with TNBC. The purpose of this study was to evaluate the relationship between BD, BPE, and FGT in women with TNBC compared to non-TNBC in a contemporary cohort of women with breast cancer.

Methods:

The Institutional Breast Cancer Database was queried for women who had invasive breast cancer and underwent mammography and MRI between (2010-2017). Variables of interest included clinical, pathologic, and imaging characteristics. Statistical analyses included Pearson's Chi Square and logistic regression.

Results:

Of 2224 women, 210 (9%) had TNBC. The median age was 59 years (22-95) and median follow up was 4 years. When we looked at the clinical characteristics of women with TNBC compared to non-TNBC, race, BRCA1,2 status, method of presentation, palpability, histology, grade, and Ki67 were statistically different (Table 1). When we looked at the correlation of MBD, FGT, and BPE for women with TNBC, MBD was correlated with FGT (r=0.64) but weakly correlated with BPE (r=0.22). We found a significant association of low BPE and TNBC compared to the non-TNBCs (p=0.021) (Table 1). In a short period of time, only 8 women with TNBC had a recurrence with no significant association with MBD, BPE, or FGT (Table 1).

Table 1.Imaging Characteristics among TNBC compared to non-TNBCVariableTNBC (N=210)%Non-TNBC (N=2014)%P-valueRace    0.001White13665153376 Black35171749 Hispanic1261156 Asian23111739 Other42191 BRCA1/2    &lt;0.001Positive3025425 Negative897579195 Method of Presentation    &lt;0.001Breast exam1125475738 Mammography7737105753 Ultrasound731106 MRI126553 Palpable    &lt;0.001Yes1185783342 No9144115258 Histology    &lt;0.001DCIS with Microinvasion21382 IDC19693159079 ILC5226913 Invasive Other731176 Invasive Grade    &lt;0.001Grade 11131016 Grade 22714111858 Grade 31688650126 ER    &lt;0.001Positive00189194 Negative2101001226 PR    &lt;0.001Positive00160080 Negative21010041221 Ki67    &lt;0.001Median (range)60 (0-99) 10 (0-99)  Mammographic Density    0.165Less dense82417846 More dense11959103454 MRI BPE    0.021Low BPE707655564 High BPE222431236 MRI FGT    0.370Less dense475440449 More dense404642151 

Conclusions:

In our study population, MBD and FGT did not differ between patients with TNBC compared to non-TNBC. Interestingly, we found a higher proportion of women with lower BPE in the TNBC compared to the non-TNBC group. BPE refers to the amount of enhancing fibroglandular tissue and has been demonstrated to reflect variations in estrogen-mediated vascular permeability. Lower BPE in TNBC may reflect the fact that these tumors are not hormonally sensitive. This may also have implications for radiogenomics, which aims to correlate imaging characteristics with gene expression and genome-related characteristics in tumor biology. Further studies are warranted in looking at these imaging biomarkers and TNBC.

Citation Format: Chun J, Schnabel F, Gooch J, Lee J, Jubas T, Goodgal J, Guth A, Moy L. The relationship of breast density in mammography and magnetic resonance imaging (MRI) in women with triple negative breast cancer (TNBC) [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 P5-02-01.

Abstract P5-23-03: Gene expression profiling in male breast cancer

Background:

Tumor gene expression profiling tests are widely used to quantify risk of recurrence of breast cancer and guide systemic therapy in early stage breast cancer. These assays have not been well validated in a male cohort. The purpose of this study was to determine the current rates of genomic testing in male breast cancer (MBC), the distribution of risk of recurrence scores (RRS) in early stage MBC, and the effect of RRS on systemic treatment patterns.

Methods:

The National Cancer Database was queried and found to include 6,227 cases of pathologic T1/T2 and N0/N1 MBC from 2008-2014 with known genomic testing status. Of the 1478 (23.74%) male patients who had gene expression profiling performed, variables of interest included age at diagnosis, stage, estrogen receptor status (ER), RRS, and treatment. We compared the treatment patterns of the 1,343 patients who had RRS available and the 4,527 patients who were not tested. A similar analysis was performed in female breast cancer (FBC) patients to serve as a point of reference. Statistical analysis included multivariate logistic regression and Pearson's chi-square test.

Results:

Of the 1,478 (23.74%) cases of MBC who had gene expression profiling, the most significant variables included: younger age, non-Black race, diagnosis after 2010, tumor Grade II, Estrogen Receptor (ER) positivity, and N0 or N1mi disease. Of those who had results, the distribution of RRS was 59.3% low, 27.4% intermediate, and 13.3% high. A similar distribution was found in 154,705 women who were tested during the same study period. Risk scores in men were significantly associated with tumor grade and size, but not nodal status. 83.4% of men with a low RRS were treated with hormone therapy alone, with an adjusted odds ratio (AOR) of 7.18 (CI 5.78-8.91, p&lt;0.001). Also, 61.8% of men with a high RRS received combination chemotherapy and hormone therapy, with an AOR of 5.16 (CI 3.60-740, p&lt;0.001).

Conclusion:

Although there is limited literature supporting the use of genomic assays in MBC, our study found similar rates of testing in men and women with early stage breast cancer. Treatment patterns varied significantly based on risk stratification, implying that physicians are using gene profiling assays to help guide treatment in MBC. Understanding the role for genomic profiling in MBC is particularly important as these results will be incorporated into the new AJCC 8th edition staging system. Long term follow up is needed to determine whether these tests accurately predict prognosis and recurrence in a male cohort.

Citation Format: Dubrovsky E, Raymond S, Chun J, Schnabel F. Gene expression profiling in male breast cancer [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P5-23-03.

Abstract P5-16-26: National trends in neoadjuvant therapy for breast cancer

Purpose:Neoadjuvant therapy has been widely integrated in the treatment of locally advanced breast cancer. Over time, this strategy has been extended to include patients with earlier stage disease to allow for assessment of in vivo response to treatment. The aim of this study was to describe the national trends in neoadjuvant therapy for all invasive breast cancers with a particular focus on triple negative disease and HER2 status.

Methods: The National Cancer Database (NCDB), an oncology outcomes database that collects data from more than 1500 Commission on Cancer (CoC) accredited cancer programs, was queried for all women diagnosed with invasive breast cancer from 2006-2013. Patients with unknown systemic therapy sequence were excluded. Women were classified by whether or not they received neoadjuvant systemic, chemo and/or endocrine, therapy.

Results: We identified 1,221,976 cases that were eligible for this analysis. Of these, 29.7% were HER2 negative, 18.4% were classified as triple negative, and 8.9% received neoadjuvant systemic therapy. The percentage of patients receiving neoadjuvant therapy increased from 7.5% in 2006 to 9.8% in 2012 with a slight decrease to 9.5% in 2013. This increase in the use of neoadjuvant therapy over the time period was statistically significant (p&lt;0.0001). There was a small increase in the percent of patients with HER2 positive status who received neoadjuvant therapy, from 5.7% in 2006 to 6.5% in 2013. During this time period, there was a 9% increase in the percent of triple negative patients who received neoadjuvant therapy (13.1% in 2016 to 22.1% in 2013).

Number and percent of patients who received/did not receive neoadjuvant therapy by year2006 (N=136117)2007 (N=143033)2008 (N=148888)2009 (N=154713)2010 (N=154040)2011 (N=162333)2012 (N=163395)2013 (N=159457)No Neoadjuvant Therapy | 125908 (92.5)131559 (91.98)136593 (91.74)141364 (91.37)139459 (90.53)146500 (90.25)147401 (90.21)144306 (90.5)Neoadjuvant Therapy | 10209 (7.5)11474 (8.02)12295 (8.26)13349 (8.63)14581 (9.47)15833 (9.75)15994 (9.79)15151 (9.5)

Conclusions: Over the time period from 2006-2013, there has been an apparent increase in the percentage of patients who received neoadjuvant therapy. This trend is accompanied by increases in the percentage of TNBC patients and in Her2 positive patients who received neoadjuvant therapy. Other factors and the joint effects of these factors on the observed increase in the use of neoadjuvant therapy are under evaluation to elucidate the basis for this observation in the NCDB data.

Citation Format: Schnabel F, Schwartz S, Hochman T, Chun J, Goldberg J. National trends in neoadjuvant therapy for breast cancer [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P5-16-26.

Abstract P6-09-13: Multi-institutional evaluation of women at high-risk for developing breast cancer

Background

Well-established risk factors for breast cancer (BC) include family history, BRCA mutations and biopsies with atypical hyperplasia (AH) or lobular carcinoma in situ (LCIS). Several institutions have registries of these high-risk women but outcomes from these registries require large numbers and long follow-up. We sought to compare characteristics between high-risk populations and evaluate early outcomes.

Methods

Women enrolled in IRB-approved high risk registries at NYU Langone Medical Center (NYU) and University of Vermont (UVM) were evaluated for risk category, uptake of prevention and development of breast cancer. Descriptive statistics were used to summarize the data and Pearson's Chi-Square and Fisher's Exact Tests were performed to compare the variables of interest among the two high risk registries.

Results

Between 2003-14, 1035 women enrolled in these high risk registries. There were significant differences in age and risk characteristics but we found a 99% concordance of variables collected between both high risk registries. Among all risk groups there was a low uptake of prevention opportunities, with 8% taking chemoprevention and 7% undergoing risk-reducing surgeries. Women with AH/LCIS accounted for 66% of those choosing chemoprevention while women with BRCA mutations accounted for 76% of those undergoing risk-reducing surgeries. To date, 43 women (4%) have been diagnosed with breast cancer. 86% were diagnosed with stage 0-1 disease and 70% had moderate or poorly differentiated cancers. There was no significant difference in background risk characteristics when comparing those with breast cancer to those who have not yet developed breast cancer.

Table 1. Clinicopathologic Characteristics between UVM and NYUVARIABLESUVM (N=496, 48%)%NYU (N=539, 52%)%P-valueMEDIAN AGE (years)46 (20-75)50 (20-87)p&lt;0.001RISK FACTORSOne or more 1° relative with BC4428928653p&lt;0.001BRCA positive2559217p&lt;0.001AH631324545p&lt;0.001LCIS22411221p&lt;0.001UPTAKE OF BC PREVENTION METHODSChemoprevention2755410p&lt;0.01Prophylactic bilateral mastectomy10.2458p&lt;0.001Bilateral salpingo-oophorectomy10.2428p&lt;0.001DEVELOPED BC316122p&lt;0.01STAGE OF BC000542p&lt;0.01I2684650 IIA, IIB2618 IIIA, IIIB, IIIC31000 

Conclusions

Despite the low uptake of chemoprevention and risk reducing surgery, only 4% of patients went on to develop breast cancer in the study period. The majority of cancers involved moderate or high-grade lesions and were early stage, suggesting a benefit to participation in surveillance programs. We have demonstrated a high degree of concordance between high risk registries, suggesting no barriers to multi-institutional collaboration. High risk registries represent an important resource for studies into methods to prevent breast cancer and improve outcomes from this disease.

Citation Format: Wood M, Schnabel F, Chun J, Schwartz S, Rounds T, Cuke M. Multi-institutional evaluation of women at high-risk for developing breast cancer. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P6-09-13.

Abstract P2-12-12: MarginProbe device use and re-excision rates for breast conservation surgeries

Background: Current methods of intraoperative assessment of lumpectomy margins are limited. Previous studies have found a lower rate of re-excisions with the adjunctive use of the MarginProbe device (Dune Medical Devices Ltd, Israel). The purpose of this study was to compare the tumor characteristics and re-excision rates before and after the use of MarginProbe for patients who had breast conservation surgery (BCS) at our institution.

Methods: The Breast Cancer Database of our medical center was queried for patients who underwent BCS from 1/2010-3/2015 by three breast surgeons. 2 surgeons used the MarginProbe to direct excision of additional margins at the time of primary lumpectomy surgery and 1 surgeon performed routine 6-surface cavity shavings. We compared our historical data (1/2010-12/2014) to MarginProbe data (1/2015-4/2015). The following variables were included: age, mammographic breast density, tumor characteristics, and re-excision rates. Statistical analyses were performed using Pearson's Chi-Square and Fisher's Exact Tests.

Results: We had a total of 1201 women who had BCS among the 3 breast surgeons. The median age was 61 years. The median invasive size was 1.2 cm. Majority of cancers were early stage (stage 0, I), invasive ductal carcinoma (61%), ER-positive (86%), PR-positive (74%), and Her2Neu-negative (88%). These tumor characteristics were not statistically different in the pre- and post-MarginProbe groups. The majority of patients had dense breasts (51%) and density did not differ among the pre- and post-MarginProbe groups (p=0.86). For the surgeons who used the MarginProbe for margin assessment at the time of surgery, the re-excision rate fell from 17% to 0% and 35% to 20% during the 4-month period. In contrast, the surgeon who routinely performed 6-surface shavings had a re-excision rate that fell from 13% to 12% in the same time period. 88% of MarginProbe readings were false positive. There was one false negative reading.

Table 1. Tumor CharacteristicsVARIABLESTotal N=1201%No MarginProbe (N=1144, 95%)%MarginProbe (N=57, 5%)%P-valueTUMOR STAGE0292242782414240.95I64454614543053 IIA, IIB22319211191221 IIIA, IIIB, IIIC39338312 IV30.330.300 HISTOLOGYDuctal carcinoma in situ305252892516280.75Invasive ductal carcinoma73061695613561 Invasive lobular carcinoma1129107959 Invasive other54553512 ESTROGEN RECEPTOR STATUSNegative16314157146110.46Positive101486963865189 PROGESTERONE RECEPTOR STATUSNegative307262942613230.56Positive86774823744477 HER2-NEU STATUSNegative788887538835810.22Positive92108610614 Equivocal19217225 

Conclusions: Routine use of the MarginProbe device was associated with lower re-excision rates compared to historical data and concurrent 6-cavity shaving approach. Better intraoperative margin assessment and lower re-excision rates will decrease the burden of breast cancer on patients and the health care system and support the practice of breast conserving surgery.

Citation Format: Schnabel F, Guth A, Axelrod D, Chun J, Schwartz S, Shapiro R. MarginProbe device use and re-excision rates for breast conservation surgeries. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P2-12-12.

Abstract P2-12-08: Oncologic outcomes after nipple-sparing mastectomy: A single-institution experience

Introduction: Nipple-sparing mastectomy (NSM) is the latest advancement in the treatment of breast cancer. Long-term oncologic outcomes in nipple-sparing mastectomy (NSM) continue to be defined. Rates of locoregional recurrence for skin-sparing mastectomy (SSM) and NSM in the literature range from 0 to 14.3%. We investigated the outcomes of NSM at our institution.

Methods: Patients undergoing NSM at our institution from 2006 to 2014 were identified. Patient demographics, tumor characteristics, and outcomes were collected. Locoregional recurrence was compared to previously published NSM and SSM results compiled from 14 and 11 studies in the literature. Institutional review board approval was obtained prior to the initiation of this study.

Results: From 2006 to 2014, 319 patients (555 breasts) underwent NSM. 149 patients (248 breasts) had long-term follow-up available. Average patient age and BMI were 47.4 and 24.28. Eighty-five percent of patients underwent mastectomy primarily for a therapeutic indication. Average tumor size was 1.41 centimeters with the most common histologic type being invasive ductal carcinoma (66.7%) followed by DCIS (23.8%). Nodal disease was present in 14.8% of patients. Average patient follow-up was 30.72 months. There was one (0.7%) incidence of ipsilateral chest-wall recurrence in a 44 year-old (p&lt;0.0001, compared to aggregate NSM and SSM data). There were 0.36 complications per patient. There were 3 incidences of nipple-areola complex (NAC) necrosis: 2 partial thickness necrosis and 1 full thickness necrosis.

Patient Demographics and Tumor CharacteristicsAge (years)47.7RaceCaucasian: 127 (85.2%) Non-Caucasian: 22 (14.8%)BMI24.28Tobacco History7 (4.7%)Radiation History8 (5.4%)BRCA 1/2 Status10 (6.7%)Family History38 (25.5%)Unilateral vs. Bilateral NSMUnilateral: 76 (51.0%) Bilateral: 73 (49.0%)Indication for MastectomyTherapeutic: 126 (84.6%) Prophylactic: 23 (15.4%)Neoadjuvant Therapy6 (4.0%)Follow-Up (months) (Range)30.72 (57.6-8.28)Tumor Size (cm)1.41Histologic Type (Percent of Therapeutic NSM)IDC: 82 (66.7%) DCIS: 30 (23.8%) ILC: 7 (5.6%) Invasive Other: 6 (4.8%) Mixed Type: 1 (0.8%)Pathologic StageStage 0: 52 (34.9%) Stage I: 54 (36.2%) Stage IIA: 14 (9.4%) Stage IIB: 8 (5.4%) Stage IIIA: 3 (2.0%) Stage IIIC: 1 (0.7%)Receptor StatusER (+): 90 (60.4%) PR (+): 79 (53.0%) Her 2/neu (+): 6 (4.0%) Ki-67 (High): 35 (23.5%)Positive Nodal Status22 (14.8%)

NSM Complications per Patient22 (14.8%)Frozen Section: 6 (7 breasts) (4.0%) Permanent Section: 2 (1.3%)Mastectomy Flap Necrosis12 (8.1%)Nipple-Areola Complex NecrosisPartial-Thickness: 2 (1.3%) Full-Thickness: 1 (0.67%)Nipple-Areola Complex Excision (Patient Preference)1 (0.67%)Implant Extrusion4 (2.7%)CellulitisOral Antibiotics: 12 (8.1%) Intravenous Antibiotics: 2 (1.3%)Hematoma4 (2.7%)Seroma3 (2.0%)Wound Dehiscence1 (0.67%)Capsular Contracture2 (1.3%)Thoracodorsal Nerve Spasm1 (0.67%)Microvascular Free Flap Failure1 (0.67%)

Conclusions: We examined our institutional outcomes with NSM and found a locoregional recurrence rate of 0.7% with no nipple-areolar complex recurrence. This rate is significantly lower than aggregate published rates for both NSM and SSM.

Citation Format: Guth AA, Frey JD, Alperovich M, Kim JC, Axelrod DM, Shapiro RL, Choi M, Karp NS, Schnabel FR. Oncologic outcomes after nipple-sparing mastectomy: A single-institution experience. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P2-12-08.

Abstract P5-15-02: Beyond the learning curve: Surgical judgment in the approach to breast conserving surgery

Background: Inadequate lumpectomy margins are associated with an increased risk of ipsilateral breast tumor recurrence in patients undergoing breast conserving surgery for malignant disease. Second surgical procedures are often required to achieve acceptable margins. The purpose of this study was to examine the practice of breast conserving surgery by experienced practitioners, focusing on the approach to margins at the time of primary lumpectomy procedure.

Methods: The NYU Langone Medical Center Breast Cancer Database was queried for patients who underwent breast conserving surgery from 1/2010-1/2013 by experienced breast surgeons. Variables of interest, segregated by surgeon, included: characteristics of additional margins taken at primary lumpectomy surgery, re-excision rates, and rates of conversion to mastectomy. Statistical analyses were performed using Pearson's Chi-Square Test, Spearman's correlation, and descriptive analyses.

Results: During the study period, 988 patients opted for breast conserving surgery for invasive and intraductal breast cancer, including 260 patients who underwent at least 1 re-excision procedure. We excluded 31 patients who had initial surgery outside of NYU, yielding a study cohort of 229 patients. Stage 0 disease was associated with an increased frequency of re-excision procedures (p&lt;0.0001). Re-excision rates differed widely among surgeons (10-36%). The average number of additional margins excised did not correlate with surgeons’ re-excision rates (p = 0.18). Additional margins taken at primary lumpectomy surgery included both false positives and true positives (Table 1). Of note, the number of false positive margin excisions was double the number of true positives. Patients who went on to mastectomy after unsuccessful primary lumpectomy surgery (45 of a total of 59 who converted to mastectomy) frequently did so based on their preference (19/45 = 42%). However, patients who converted to mastectomy after multiple excisions generally did so out of concern for extent of disease.

Conclusions: In our study, patients with pure intraductal carcinoma represented a particular challenge as surgeons’ margin assessment was less accurate than in cases of invasive cancer. Re-excision rates varied by surgeon, and did not correlate with the average number of additional margins taken at the primary lumpectomy procedure. The rate of false positive margins excised exceeded the true positive rate, reflecting the limitations of surgeons’ ability to assess margins intraoperatively. We documented a significant number of patients who opted for conversion to mastectomy after a single unsuccessful lumpectomy procedure, underscoring the need for better methods of intraoperative margin assessment to support the practice of breast conserving surgery.

Table 1: Re-excision rates and approach to margins at primary lumpectomy surgerySurgeonPatients undergoing Re-excision - N (%)Avg Additional Margins Taken (per patient) in Primary LumpectomyFalse Positives - N (%)True Positives - N (%)A67 (23%)139 (10%)26 (6%)B53 (10%)4147 (45%)60 (18%)C14 (19%)217 (20%)6 (7%)D11 (31%)01 (1%)3 (5%)E8 (36%)15 (10%)3 (6%)F27 (11%)118 (11%)10 (6%)G20 (15%)335 (29%)26 (22%)H29 (32%)237 (21%)24 (14%)Total229 (16%)2299 (22%)158 (11%)

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P5-15-02.

Abstract P1-09-03: Chemoprevention in patients with newly diagnosed breast cancers

Background: Chemoprevention (including tamoxifen, raloxifene, and exemestane) is a strategy to reduce breast cancer incidence in high risk women. Studies have shown at least a 50% decrease in the incidence of breast cancer in users of these drugs. Despite this benefit, the majority of high risk, unaffected women who are offered chemoprevention decline the therapy. However, there is a growing population of women who have used these agents for primary prevention, and a larger population of survivors who have used these drugs as part of their systemic treatment. The purpose of this study was to identify a cohort of women with newly diagnosed breast cancers who had utilized chemoprevention and describe their patterns of disease.

Methods: The Breast Cancer Database of NYU Langone Medical Center was queried for patients who used chemopreventive drugs and developed breast cancer between 1/2010-1/2012. Patients were divided into primary and secondary chemoprevention groups (no previous history of breast cancer and previous history of breast cancer, respectively). Descriptive statistics were utilized.

Results: In the study period, 24 (2%) of 996 patients had used a chemopreventive agent. For 16 of the 24 (67%), the drug was part of systemic therapy for prior breast cancer, with a median of 12 years from the initial diagnosis to the diagnosis of a second breast cancer. The primary chemoprevention group included women with risk based on family history and atypical hyperplasias. The majority of patients were diagnosed with early stage disease (88% DCIS and stage I). This likely reflects their screening behaviors. In both groups, the majority of cancers were ductal in origin. Five of the 8 patients in the primary chemoprevention group were on treatment at the time of their cancer diagnosis, while 63% of patients in the secondary group were prior users. In the secondary group, the majority of cases were contralateral second primary breast cancers, with 31% ipsilateral recurrences. Interestingly, the majority of cancers in both groups were ER/PR positive.

P4-10-08: Risk Factor Profiles of Women with DCIS and Invasive Ductal Breast Cancers

Background There are well established risk factors for breast cancer, including elements of personal and family history. Ductal carcinoma in situ (DCIS) and invasive ductal carcinoma (IDC) have long been viewed as representing two points on the spectrum of one disease. There is limited and conflicting information in the literature regarding risk factor profiles for DCIS as compared with IDC. The purpose of this study was to investigate the relationship of established risk factors in a population of women newly diagnosed with pure DCIS and IDC. Methods

The Breast Cancer Database at NYU Langone Medical Center was queried for women who were diagnosed with pure DCIS and IDC from 1/2010-3/2011. Variables of interest included: age, family history of breast cancer (FHBC), BRCA1/2 status, age at menarche and menopause, parity, age at first birth, breast feeding, body mass index (BMI), history of atypical hyperplasia or lobular carcinoma in situ (ADH, ALH, LCIS), stage, ER/PR status, and method of presentation. Wilcoxon non-parametric tests, Chi-Square tests, and Fisher's Exact Tests were used to evaluate differences among DCIS and IDC patients.

Results Of the 593 women identified in this study, 140 (24%) had pure DCIS and 453 (76%) had IDC. The median age at diagnosis was 59 years. There were 9 (1.5%) BRCA1/2 mutation carriers. Of these, 7 had IDC and 2 had DCIS. The majority of patients with IDC were stage I (67%) and ER/PR+ (73%). The majority of patients with DCIS were also ER/PR+ (71%). In our cohort, the majority of DCIS (83%) was mammographically detected versus 47% of IDC cases. There were no statistically significant differences in breast cancer risk factors between the two groups.

Abstract P2-01-02: The Incidence of Mammographically-Occult Breast Cancer in Women Older than Seventy Years

Background: Mammographic density generally decreases over time, which increases the sensitivity of screening mammography. However a significant proportion of older women have persistently dense breast tissue. This raises the question of how best to screen older women, particularly those with mammographically dense tissue or other risk factors. Little information exists about the accuracy of screening mammography in women older than seventy years. Additionally, the frequency of older women with mammographically-occult breast cancer is also unknown. The purpose of this study was to report the incidence of mammographicallyoccult breast cancer in women older than seventy and to describe the clinical factors that may be associated with this presentation.

Methods:

A retrospective chart review was conducted of women older than seventy diagnosed with breast cancer (intraductal carcinoma, invasive ductal carcinoma and invasive lobular carcinoma) at NYU Langone Medical Center from 2002-2009. Data collected included age, presentation, stage, mammographic density, BI-RADS results, breast ultrasound (US) and MRI results, and risk factors. Breast density was categorized according to BI-RADS definitions: 1. predominantly fatty, 2. scattered fibroglandular elements, 3. heterogeneously dense, or 4. extremely dense. Descriptive analyses were applied.

Results: A total of 401 women older than seventy years were diagnosed with breast cancer at our institution. A cohort of 20 (5%) had mammographically occult disease. The median age was 77 years (range 71 to 89 yrs). Of the 20 mammographically occult cases, the majority (75%) were detected as a palpable mass, one was detected by screening US and four were detected by screening MRI. The majority of the cohort (65%) had been screened regularly with mammography prior to diagnosis. When we examined mammographic density, 60% had heterogeneously dense or extremely dense tissue. Nineteen out of twenty were diagnosed with early stage disease. There were 4 cases of DCIS, 10 (50%) were stage 1, and 5 (25%) were stage 2. Seven (35%) had invasive lobular carcinoma on final histopathology. Nine (45%) patients also had an antecedent history of breast cancer; 4 had ipsilateral recurrences and 5 had contralateral new primaries. Only 4 (20%) had a history of HRT use and 7 (35%) had a family history of breast cancer.

Discussion: A meaningful proportion of women older than age seventy diagnosed with breast cancer at our institution had mammographicallyoccult disease. A majority (60%) of these patients had heterogeneously dense or extremely dense breast tissue, potentially limiting the sensitivity of mammographic screening in this cohort. In previous work, we evaluated a large number of screening mammograms and reported an incidence of 37% dense breast tissue in women of the same age group. This suggests that increased mammographic density, even in older women, may increase the potential for mammographically occult breast cancer. Particularly for women with increased risk based on a personal or family history of breast cancer, the addition of other imaging modalities may be of value in the presence of mammographically dense tissue regardless of patient age.

Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P2-01-02.

Sucess of Brochure/One Page Universal Consent for Biospecimen Donation

Background: Accrual of biospecimens to biorepositories is a challenge and has been recognized as a critical barrier for translational research. HIPAA mandates that patients sign a separate Informed ConsentForm (ICF) for donating biospecimens for research. This poses a significant challenge to the medical staff and may negatively impact accrual to biorepositoriesMaterials and Methods: At NYU Cancer Institute, the ten page ICF was reformatted as a tri-fold brochure and one page ICF in 2005. The brochure is given to patients during the office visit or at time of registration and the triplicate one-page ICF on the day of surgery. The brochure contains information regarding biospecimen banking.The triplicate, one page consent is signed by the patient on the day of surgery.Results: In a 3-year period (2003-2005), the 10-page ICF was used to consent 433 patients from Breast, Gynecologic Oncology and GI Oncology, with an average of 2.77 patients per week. After the implementation of the novel brochure and single page ICF format, 1,426 patients were consented in the next 3- year period (2006-2008), with an average of 9.14 patients per week (10)Patients in the Breast Surgical Oncology Clinic at NYUCI were initially targeted. This was done in close cooperation with the breast surgical oncologists, nurses and operating room staff. The brochure/1-page ICF method doubled the accrual of patients from 141 to 304 patients in the first year (2005-6). However, the number of patients who consented did not increase in the next two years in spite of a significant increase in the total number of patients in the breast surgical oncology service.A root cause analysis was performed and it was determined that either the patients were not approached in the physician's office or the consenting process was not followed. It was concluded that a dedicated consenting professional was required to overcome these hurdles, who will primarily works in the operating room, where patients with diagnosis of cancer are targeted. This has yielded a significant increase in number of patients consenting for biospecimen donation. shows a significant increase in accrual since her recruitment in February 2009.Conclusions: A novel consenting mechanism with a brochure and one page consent has been successfully implemented at NYU Cancer Institute and has resulted in a significant increase in patients consenting for biospecimen donation. The process is being replicated institute wide for accrual to all biorepositories at NYULMC. This novel method can be widely employed to accrue specimens for clinical trials.

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 6081.

Mammographic Density and Lobular Involution in Older Women with Abnormal Breast Imaging

Mammographic density has been established as an independent risk factor for breast cancer, and there is data to suggest that the degree of lobular involution in the breast tissue may also function as an independent risk factor for the disease. The present study was designed to investigate the relationship between mammographic density and lobular involution in a population of mature women undergoing open biopsy for non-palpable breast lesions.A total of 199 women over the age of 60 who underwent breast biopsy with image-guided localization in 2008 at NYU Langone Medical Center formed the study population. Variables of interest included age, breast density, degree of involution, use of hormone replacement therapy (HRT) and biopsy histology (invasive ductal and lobular carcinoma and intraductal carcinoma vs. benign). Breast density was categorized according to BI-RADS classification as less dense (predominantly fatty and scattered fibroglandular elements) vs more dense (heterogeneously dense and extrememely dense). Degree of involution was classified as none (0%), partial (1-74%) and complete (&gt;75%). All specimens were reviewed by a single pathologist who used the criteria of Hartmann, et al in evaluating degree of involution. Statistical analysis was performed using Pearson's Chi-square test and logistic regression analyses.In agreement with our previously presented data, we found an inverse relationship between breast density and age (p=0.02). In our cohort of women over 60 with abnormal breast imaging, we found a trend toward an inverse relationship between age and degree of lobular involution, but this did not achieve statistical significance. 129 of our patients had biopsies yielding malignant histology. There was no significant relationship between degree of lobular involution and biopsy histology. Use of HRT did not correlate with breast density in our cohort. There was no discernable relationship between degree of involution and history of prior or current HRT use.Association of lobular involution with age, density, histology, and HRT  INVOLUTION TOTALP-VALUE NONEPARTIALCOMPLETEN (%) AGE (YRS)      60-696337617 (9%)0.18270-797202959 (29%) 80-894618123 (62%) DENSITY     Less Dense143477125 (63%)0.177More Dense3254674 (37%) HISTOLOGY     Benign5194670 (35%)0.690Malignant124077129 (65%) HRT     Never11246398 (70%)0.774Ever3112741 (30%)  Our study reinforces the general observation that breast density decreases with age. Our data do not permit us to comment on the validity of lobular involution as a risk factor for breast cancer. However, the degree of lobular involution did not predict malignant histology in our cohort. There was a trend toward an inverse relationship between mammographic density and lobular involution. This observation suggests that other factors such as stromal features may be responsible for the increased mammographic density in these older patients. Further study is warranted to better understand the significance of lobular involution and its relationship with mammographic density in all age groups of women undergoing breast cancer screening.

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 6072.

Does the recurrence of pregnancy-associated breast cancer during subsequent pregnancy 10 years after mastectomy support the unique hormonal milieu of pregnancy as a cause of tumorigenesis? A case and literature review

Abstract #3058

We report a case of pregnancy-associated chest wall recurrence of breast adenocarcinoma 10 years following mastectomy and reconstruction for initial pregnancy-associated breast cancer (PABC). This represents the first reported case in the literature of a pregnancy-associated recurrence of pregnancy-associated breast cancer. The patient had no episodes of recurrent breast cancer during the interim 10 years, and only experienced a recurrence with a rapidly growing left chest mass during the subsequent pregnancy. Thus, the possibility that the malignancy may have been triggered by the pregnancy cannot be ignored.&#x2028; Although prior studies following breast cancer patients and their subsequent pregnancy have not shown an increase risk of recurrence, these studies were performed in “all comers” with breast cancer, and would not have had the statistical power to distinguish a small subset of patients with an increased risk, such as PABC patients. As a result, these patients have been assumed to be at no greater risk for recurrence with pregnancy and have been advised as such.&#x2028; It is reasonable to assume that PABC patients may, in fact, represent a unique subset of breast cancer patients with the unique hormonal milieu of pregnancy playing a mechanistic role in the propagation and proliferation of the disease. In this case a patient with prior PABC had a rapidly growing progesterone receptor positive breast cancer during a subsequent pregnancy. Further, studies have indicated that there is a possibility that the hormone status of women during pregnancy may in fact lead to a change in tumor characteristics, and therefore their risks. This instance would support the concern and represent the sentinel report describing a new subset of breast cancer with a unique spectrum of risk factors.

Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 3058.

Assessment of breast cancer risk prediction models in a high-risk population

Abstract #4074

Background: A variety of risk assessment models have been developed to quantify a woman's risk for developing breast cancer. Although the Gail model (GM) is the most widely utilized model, there are limitations. The Tyrer-Cuzick model (TCM), which has been validated in the United Kingdom (UK), addresses these limitations but its performance in other populations is unclear. The purpose of this study was to evaluate the predictive ability of the TCM in a cohort of high-risk women from the New York metropolitan area and compare the results to the GM.&#x2028; Methods: The Women At Risk (WAR) Registry provided the study population. Due to the age limitation of the Gail model, we excluded women who were under 35 and over 80 years of age. Calculation of lifetime Gail scores included the following: age, race, age at first menses, age at first live birth, number of first degree relatives with breast cancer, number of previous breast biopsies, and atypical hyperplasia. These variables were also included in the lifetime Tyrer-Cuzick scores as well as information on menopausal status, use of hormone replacement therapy, LCIS, and age at diagnosis for relatives with ovarian and breast cancer. The Receiver Operating Characteristic (ROC) and estimated area under the ROC curve (AUC) with 95% confidence intervals (CI) were used to assess prediction accuracy of both models.&#x2028; Results: Out of a total study population of 1523 women, 82 (5%) developed breast cancer during a median follow-up of 56 months. The median age of women without breast cancer was 49 years, and the median age of women who developed breast cancer was 54 years. The AUC and 95% CI for the Gail model was 0.547 (0.479 to 0.615) and 0.501 (0.433 to 0.569) for the Tyrer-Cuzick model. Our results indicate poor discriminatory accuracy for both models, which performed no better than pure chance (0.5).&#x2028; &#x2028; Discussion: Quantitative breast cancer risk assessment models are critically important in developing effective risk management strategies. Although the TCM addresses some limitations of the GM, it is unclear how well it performs in populations outside the UK due to the wide variation in international breast cancer rates. In addition, the TCM excludes more recently identified risk factors such as breast density. Bayesian techniques may ultimately lead to a better risk assessment tool as they allow continual modification of the model as new risk factors are identified. Accurate breast cancer risk prediction is a complex undertaking and further studies are warranted in order to achieve truly comprehensive and widely applicable models.

Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 4074.

Factors correlating with lymph node metastases in patients with T1 breast cancer

BACKGROUND

Identification of reliable predictors of axillary metastases (ALNM) may be useful in selecting appropriate management for patients with T1-size breast cancer. This study was undertaken to determine the degree of correlation between ALNM and several variables, including age, race, menopausal status, palpability, tumor size, positive margin on initial excision, histology, grade, lymphatic invasion (LI), estrogen receptor status (ER), progesterone receptor status, S-phase, and ploidy.

METHODS

Data from 1416 patients with T1 breast cancers treated at Columbia-Presbyterian Medical Center between 1989 and 1998 was reviewed. Patients with multifocal tumors were excluded.

RESULTS

Mean patient age was 57.5 years (SD = 12.0); 65% of the patients were postmenopausal. One hundred thirty-one patients with Tla (< or =0.5 cm), 435 with T1b (0.6-1.0 cm), and 850 patients with T1c (1.1-2.0 cm) lesions were studied. The overall rate of ALNM was 23%. AM was identified in 11% of T1a, 15% of T1b, and 29% of T1c patients. Statistically significant factors from univariate analysis were age, palpability, skin changes, tumor size, LI, histology, grade, ER status, and positive margin on initial excision.

CONCLUSIONS

Axillary staging by either sentinel lymph node biopsy or level I/II axillary dissection is indicated for most T1 breast cancer patients. Omission of axillary staging can be considered for highly selected patients with T1a cancers.

Estimating the effect of an intensive surveillance program on stage of breast carcinoma at diagnosis: a propensity score analysis

BACKGROUND

The authors used propensity score adjustment to investigate the impact of intensive screening on stage of breast carcinoma at diagnosis in women who were at elevated risk for breast carcinoma.

METHODS

The authors compared 58 women participating in a surveillance program at the Columbia Presbyterian Medical Center of New York Presbyterian Hospital who developed breast carcinoma with 3022 nonparticipating breast carcinoma patients. A propensity score was constructed for each woman by using important background covariates, and multivariable regression modeling was used to estimate the association of program membership with disease stage after adjusting for the propensity score.

RESULTS

Before propensity score adjustment, nine baseline covariates significantly differed between the two groups (number of pregnancies, number of births, age at first delivery, race, how the tumor was discovered, history of prior breast disease, breast carcinoma in mother, breast carcinoma in maternal aunt, and breast carcinoma in sister), and there was a significant difference in stage at diagnosis. After adjustment, no significant differences remained. Program participants were more likely to have lower stage tumors at diagnosis than nonparticipants, but this association did not reach statistical significance (odds ratio, 1.52; 95% confidence interval, 0.94--2.46).

CONCLUSIONS

Propensity score methods can remove bias in treatment comparisons in observational studies. An intensive surveillance program at a major cancer center may have had some effect on improving stage at diagnosis, but this effect was not statistically significant.

The relationship between genetic damage from polycyclic aromatic hydrocarbons in breast tissue and breast cancer

A number of polycyclic aromatic hydrocarbons (PAH) are widespread environmental contaminants that cause mammary cancer experimentally. We investigated whether exposure and susceptibility to PAH, as measured by PAH-DNA adducts in breast tissue, are associated with human breast cancer. We carried out a hospital-based case-control study using immunohistochemical methods to analyze PAH-DNA adducts in tumor and nontumor breast tissue from cases and benign breast tissue from controls. The subjects were white, African-American and Latina women without prior cancer or treatment, including 119 women with breast cancer and 108 with benign breast disease without atypia. PAH-DNA adducts measured in breast tumor tissue of 100 cases and in normal tissue from 105 controls were significantly associated with breast cancer (OR=4.43, 96% CI 1.09-18.01) after controlling for known breast cancer risk factors and current active and passive smoking, and dietary PAH. There was substantial interindividual (17-fold) variability in adducts overall, with 27% of cases and 13% of controls having elevated adducts. The odds ratio for elevated adducts in tumor tissue compared with control tissue was 2.56 (1. 05-6.24), after controlling for potential confounders. Adduct levels in tumor tissue did not vary by stage or tumor size. Among 86 cases with paired tumor and nontumor tissue, adducts levels in these two tissues were highly correlated (r=0.56, P<0.001). However, the corresponding associations between case-control status and adducts measured in nontumor tissue from 90 cases and in normal tissue from 105 controls were positive but not statistically significant. Overall, neither active nor passive smoking, or dietary PAH were significantly associated with PAH-DNA adducts or breast cancer case-control status. These results suggest that genetic damage reflecting individual exposure and susceptibility to PAH may play a role in breast cancer; but more research is needed to determine whether the findings are relevant to causation or progression of breast cancer.

Short-term outcome of chronic immunosuppression on the development of breast lesions in premenopausal heart and lung transplant patients

The risk of development of breast lesions in patients on chronic immunosuppression is unknown. In order to assess this risk, a retrospective review was performed of the records of 87 women between the ages of 12 and 47 years who received thoracic organ transplant from 1987 to 1996 at our institution. Inclusion criteria consisted of patients who were premenopausal, had no previous history of breast disease, and survived for at least 1 year posttransplantation. All patients were on a triple immunosuppressive regimen consisting of cyclosporine, steroids, and azathioprine. Mean follow-up was 4 +/- 1.2 years with a range of 1-6 years. During this period, 21 patients (24%) with a mean age of 38 +/- 10 years had screening or diagnostic mammography. The remainder of patients with a mean age of 24 +/- 9 years were followed clinically. Overall, 10 patients (11%) developed a total of 17 palpable, solid lesions at 33 to 72 months posttransplantation. Fifteen of these lesions were surgically excised. Five of the patients had multiple lesions. Pathological examination of the specimens revealed fibroadenoma in nine, fibrocystic disease in four, low grade phylloides tumor in one, and T-cell lymphoma in one case. None of the patients have developed primary breast cancer during follow-up. In conclusion, short-term immunosuppression does not increase the risk of the development of benign breast lesions in young women after thoracic organ transplantation, but rather the distribution of benign lesions is similar in an age-matched population. There were several cases of multiple fibroadenomas in the transplant population, but mammography revealed no malignant disease in this age group and does not need to be utilized in this population beyond what is considered standard for immunocompetent patients. The long-term effect ofimmunosuppressive therapy on the developmentof breast cancer in this group remains to be defined.

Organochlorine compounds (DDE and PCB) in plasma and breast cyst fluid of women with benign breast disease

The organochlorines, dichloro-diphenyl-trichloroethane and polychlorinated biphenyl (PCB) are pervasive environmental contaminants. Results from previous studies have been conflicting regarding the relationship between the internal dose of these organochlorine residues and breast cancer risk. To determine whether these compounds are present in breast cyst fluids and whether cyst fluid and plasma concentrations are correlated, we analyzed organochlorines in paired cyst fluid and plasma samples from 24 subjects using gas chromatography and electron capture detection. All but one of the women had a history of multiple cysts, suggesting that they were at elevated risk for future breast cancer. DDE (a metabolite of dichloro-diphenyl-trichloroethane) was present in 22 of the cyst samples and PCB was detected in 19 of the cyst samples. Organochlorine levels were more concentrated in the plasma than in breast cyst fluids. Levels of DDE in plasma were significantly correlated with those in cyst fluid (r = 0.73; P < 0.001); in contrast to PCB levels in cyst and plasma (r = 0.37; P = 0.12). Congener specific analysis of the PCBs showed that some individual congeners were preferentially excluded from or concentrated in the cyst fluid. To our knowledge, this study is the first to demonstrate that PCB and DDE are present in cyst fluids and thus in contact with the ductal epithelium of the breast. These results support the use of plasma DDE as a proxy for DDE in the target tissue in research on the role of environmental factors in breast cancer.

Immediate breast reconstruction in patients with locally advanced disease

Immediate breast reconstruction for patients with early-stage disease is well established. This study evaluates a consecutive series of 22 patients with locally advanced disease (stage IIB or III) who underwent mastectomy and immediate breast reconstruction. All patients received several cycles of neoadjuvant chemotherapy (average, 3.5 cycles) followed by completion of chemotherapy beginning approximately 3 weeks following surgery. The perioperative morbidity was 14% and no patient suffered a delay in the resumption of chemotherapy. Patients have been particularly grateful about being offered reconstruction in this setting. Our preliminary results with this technique have been encouraging and further study is warranted.

The relationship of urinary thromboxane excretion to cyclosporine nephrotoxicity

Alterations in renal prostaglandin production have recently been postulated to modulate the decrease in renal blood flow associated with cyclosporine nephrotoxicity. In particular, increases in renal production of the potent vasoconstrictor thromboxane A2 have been implicated in the pathogenesis of this disorder. The present study was undertaken to explore the relationship between alterations in urinary thromboxane B2 excretion (UTxB2V) and CsA nephrotoxicity in two rat models. Male Sprague-Dawley (SD) rats were treated for 14 days with CsA 50 mg/kg/day (n = 8) or olive oil (C) (n = 9) by gavage. Creatinine clearance (Ccr), urine flow (V), and urinary excretion rates of sodium, N-acetyl-beta-D-glucosaminidase (NAG), glucose, and TxB2 were determined before and after treatment. A similar study was conducted using Fischer rats (CsA: n = 10, C: n = 13). In Fischer rats, CsA caused a 35% decrease in Ccr (P = 0.01), a 33% decrease in sodium excretion (P = 0.02), and a greater than 2-fold increase in NAG excretion (P = 0.03), while V, glucose excretion, and UTxB2V did not change. Although similar changes in sodium and NAG excretion were seen after CsA administration in SD rats, Ccr was not affected. Additional findings in SD rats included a 3-fold increase in V (P less than 0.01), a 24-fold increase in glucose excretion (P = 0.03), and a 5-fold increase in UTxB2V (P = 0.04). Thus, Fischer rats developed CsA nephrotoxicity in the absence of increased UTxB2V. In contrast, SD rats failed to develop nephrotoxicity despite a marked increase in UTxB2V. We conclude that changes in renal TxA2 production are unrelated to the development of CsA nephrotoxicity.

Renal vascular reactivity in the bile duct-ligated rat

The renal dysfunction associated with liver disease appears to be caused by decreased renal perfusion. Altered renal vascular reactivity (RVR) may contribute to the development of decreased renal blood flow. This study was undertaken to assess the effect of bile duct ligation (BDL) on RVR to norepinephrine and angiotensin II in the rat and to evaluate the role of prostaglandins in the modulation of altered RVR. Male Sprague-Dawley rats underwent either sham operation (SO) or BDL. Four days later, isolated kidney perfusions were established. Dose-response curves to norepinephrine (SO, n = 9; BDL, n = 10) and angiotensin II (SO, n = 7; BDL, n = 9) were obtained. The perfusate was modified to contain indomethacin, 10(-6) mol/L, and dose-response curves to norepinephrine (SO, n = 10; BDL, n = 8) and angiotensin II (SO, n = 4; BDL, n = 8) were obtained as well. Baseline resistances did not differ between any of the groups. RVR to norepinephrine did not appear to be altered by BDL; however, increased RVR to norepinephrine was unmasked by inhibition of prostaglandin synthesis. BDL did not affect the RVR to angiotensin II with or without indomethacin. BDL is associated with an increase in renal vascular sensitivity to norepinephrine, which is balanced by a compensatory increase in renal prostaglandin activity in response to norepinephrine stimulation. These findings are consistent with theories that decreased renal perfusion associated with liver disease may be mediated by catecholamines.