Abstract Background: Both the Aphinity and ExteNET trials for anti-HER2 targeted agents were challenged in showing significant benefit of added HER2-targeted treatment. The findings suggested biological heterogeneity in HER2+ cancers, not entirely identified by IHC/FISH, requiring more nuanced biomarkers to clearly identify patient subsets who may derive benefit from new HER2-targeted agents. We have previously shown that BluePrint (80-gene) molecular subtyping reclassifies nearly half of HER2+/ER+ patients to Luminal-type with differential neoadjuvant treatment response (Whitworth, Ann Surg Oncol 2014; Whitworth, ASCO 2018). Here, we evaluated the reclassification rate in the real-world diagnostic setting. Methods: Physicians regularly provide pathology reports to Agendia, Inc for samples which are processed for MammaPrint (70-gene signature) and BluePrint molecular assays as part of routine diagnostic care. For this analysis, 4986 sequentially available pathology reports (submitted between October 2016 to October 2017) were reviewed; HER2 and ER IHC results were captured. The molecular subtype was compared to IHC/FISH status. Results: HER2 IHC/FISH results were available for 1568 samples. Of those, 85% (1330/1568) were HER2-nonamplified, 10% (153/1568) were HER2-equivocal, and 5% (85/1568) were HER2-amplified by IHC/FISH. Of the HER2-nonamplified tumors, BluePrint reclassified 0.1% (2/1330) as HER2-type. Of the HER2-equivocal tumors, none were HER2-type by BluePrint; 91% (139/153) were Luminal-type and 9% (14/153) were Basal-type. Of the HER2-amplified tumors, 15% (13/85) were dominant HER2-type, 79% (67/85) were dominant Luminal-type, and 6% (5/85) were Basal-type. Conclusions: In this set of tumors identified as HER2-amplified by IHC/FISH, BluePrint reclassified 85% of tumors to non-HER2 molecular subtypes, mostly Luminal-type for ER-positive tumors and Basal-type for ER-negative tumors. Moreover, BluePrint gave clarity where IHC/FISH could not, classifying all HER2-equivocal tumors to non-HER2 subtypes. Additional therapeutic options should be explored for HER2+/ER+ BluePrint Luminal-type patients who have observed much lower pCR rates versus BluePrint HER2-type patients (12% vs. 51%, respectively; Lee, AACR 2018). HER2 IHC/FISH vs BluePrint SubtypeHER2 IHC/FISH StatusHER2-type Luminal-type Basal-type Total IHC ER-IHC ER+IHC ER-IHC ER+IHC ER-IHC ER+ Nonamplified (ER-Unknown, n=28)118120353361330Equivocal (ER-Unknown, n=6) 2131104153Amplified (ER-Unknown, n=2)112 664 85Total21310140067401568 Citation Format: Treece T, Audeh W, Navarro F, Wei J. BluePrint molecular subtyping versus HER2 assessment by immunohistochemistry and FISH in the real-world diagnostic setting [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 P3-10-22.

Abstract Background: Response to neoadjuvant therapy (NT) predicts progression-free and overall survival in triple negative breast cancer (TNBC). Carboplatin has shown efficacy in patients with TNBC. The current phase II prospective neoadjuvant trial was designed to decrease toxicities and improve efficacy. Methods: Patients with TNBC received carboplatin (carb) and nab-paclitaxel (nab). Pre-NT biopsies were procured to evaluate for biological predictors of pathological complete response (pCR). Newly diagnosed stage II-III patients with TNBC were treated with 4 cycles of carb (AUC 6, day 1 of 28 day cycle) and weekly nab 100 mg/m2 x 16. Targeted accrual goal is 70. RNA extracted from formalin fixed paraffin embedded (FFPE) biopsies pre-NT was tested for MammaPrint/BluePrint and custom Agilent full genome microarrays for gene expression (GE, by Agendia Inc). The raw gMeanSignal was log2 transformed and normalized to the 75thpercentile for GE analysis. Association between MammaPrint/ BluePrint results and pCR was tested by Fisher exact test. The linear model from R limma package was applied. Ingenuity Pathway Analysis (IPA) was applied to assess functional pathways associated with pCR. Cellular distribution by CIBERSORT analysis was carried out to estimate the abundance of 22 different cell types in each patient sample, and test whether the distribution of cell types is different between pCR and non-responders. Results: A total of 64 patients were enrolled. Two patients were deemed ineligible (Her2+), and three were too early, resulting in 59 patients evaluable for pathological response. The pCR rate was 47% (RCB0, 28/59). Eight patients had RCB I. RCB0 plus RCBI reached 61%. Sufficient quality RNA and DNA were available from the first 43 of 55 pts with TNBC. 44/59 (75%) required dose modifications (mostly hematologic), 5 patients had grade 3 peripheral neuropathy (PN), 3 had grade 2 PN, and 3 patients had grade 2 LFTs. In the 53 pts with GE assessment, pCR was inversely associated with luminal BluePrint type (p=0.04). With fold change >1.5 and p-value < 0.05, 36 genes were differentially expressed (DE) in TNBC. CIBERSORT analysis suggested that T-cell regulatory cells (TREGS) were associated with pCR in TNBC, and 5 cell types (plasma cells, TREGS, macrophage, dendritic cells and neutrophils) presented differently between all pCR and non-pCRs with P-value <0.05. TDP analysis to assess correlation with pCR is ongoing. Conclusions: The combination of carboplatin and nab-paclitaxel given in the neoadjuvant setting reached a promising pCR rate of 47%. The MammaPrint non-luminal BluePrint subtype was predictive of pCR in TNBC. Preliminary analysis suggested that a 36-gene signature for TNBC was associated with pCR. CIBERSORT analysis revealed 5 cell types with different abundance between the pCR and non-responders, suggesting the need to target the tumor microenvironment. Citation Format: Yuan Y, Frankel P, Li M, Kruper L, Jones V, Treece T, Waisman J, Yim J, Tumyan L, Schmolze D, Hurria A, Yeon C, Mortimer J, Somlo G. Phase II trial of neoadjuvant carboplatin and nab-paclitaxel in patients with locally advanced triple negative breast cancer [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 P1-15-07.

Abstract Background: Invasive lobular carcinoma (ILC) of the breast has been shown to be more strongly associated with risk factors that modulate hormone levels, including obesity, as compared with invasive ductal carcinoma (IDC). Our previous study indicated the biology of disease engages disparately for ILC and IDC. This study further evaluated the effect of metabolic syndrome (MS) in patients with ILC histopathological tumor types using whole transcriptome gene expression (GEA) and pathway analyses. Methods: This analysis included 76 patients with ILC from the PROMIS and IMPACt studies for whom metabolic characteristics were captured with informed consent. To be classified as having MS, the patient had to exhibit any 3 of the 5 metabolic factors (obesity, hypertension, hypercholesterolemia, hypertriglyceridemia, diabetes mellitus). Risk of recurrence was established using the 70-gene signature (70-GS). Pathway enrichment analysis was performed in DAVID v 6.8 (https://david.ncifcrf.gov/). Benjamini-Hochberg corrected P< 0.05 was considered significant. Results: Thirty-four percent (26/76) of patients with ILC had MS. Significantly more ILC patients with MS than without MS were 70-GS high risk (42% vs. 12%, respectively, [P=0.007]). GEA identified 486 significant differentially expressed genes between MS and non-MS ILC patients. These differentially expressed genes indicated that the immune pathways were affected rather than growth factor pathways. The significantly enriched pathways were driven by genes upregulated on average across all MS ILC patients, and represented in the immune response (P< 0.001), adaptive immune response (P=0.001), B cell signaling (P< 0.001), inflammatory response (P<0.05), chemokine-mediated signaling (P<0.05), T cell co-stimulation (P<0.05), and intracellular signal transduction processes (P<0.05); as well as enriched in systemic lupus erythematosus (largely overlapping with nucleosome assembly, P=0.001) and osteoclast differentiation (P<0.05) pathways. While early infiltration of cytolytic T-lymphocytes may protect against tumor development, humoral-mediated immunity and a sustained state of chronic inflammation may result in polarization toward a pro-tumor microenvironment. ILC MS Status by 70-GS Risk ClassificationILC MS Status70-GS High Risk70-GS Low RiskP-valueTotalNo6 (12%)44 (88%)0.00750Yes11 (42%)15 (58%) 26 Conclusions: ILC is generally considered to have a low risk of recurrence, but with poorer long- term prognosis. We show that ILC patients with MS have a significantly higher 70-GS risk of recurrence than their non-metabolic counterparts. Our gene expression analysis suggests that ILC in the setting of metabolic syndrome, likely results in chronic immune activation and has distinct drivers of disease progression and metastasis, specifically to bone. The current study underscores the complex interplay between inflammation and immune suppression. Studies are needed to further characterize differences in enriched pathways as it relates to optimizing targeted therapeutics. Citation Format: Robinson P, Treece T, Osipo C, Uygun S, Kling H, Qamar R, Zon R, Levine E, Budway R, Mavromatis B, Untch S, Bernards R, Audeh W, Soliman H, IMPACt Investigators Group. Metabolic syndrome increases risk of recurrence and impacts immune pathways in invasive lobular carcinoma [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 PD7-04.

Abstract Background: The PROMIS trial (NCT01617954) previously showed that an OncotypeDx (ODx) Intermediate Recurrence Score (RS 18-30) led to uncertainty in prescribing chemotherapy (CT), especially in the middle of the intermediate range from RS 21-26 where an equal number of patients were recommended to receive and forego CT (Tsai, JAMA Oncology 2018). Forty-seven percent (3183/6711) of randomized TAILORx patients were classified as RS 18-25 and are well represented in PROMIS. These patients with RS 18-25 may still lack definitive CT recommendation following TAILORx, reflexing to age and menopausal status to make a decision. Here, we re-evaluate PROMIS using the subgroup analyses adopted by TAILORx. Methods: MammaPrint (MP) risk of recurrence was determined for ODx intermediate patients by standard diagnostic testing (Agendia, Irvine, CA). Clinical risk was assessed using the MINDACT, modified Adjuvant Online! algorithm (Cardoso, NEJM 2016). The MP high and low risk classification, and patient and tumor characteristics were re-evaluated and subdivided by RS 18-25 vs. RS 26-30. Results: The 840 eligible patients in PROMIS were classified as 61.3% (515/840) clinically low risk and 37.0% (311/840) clinically high risk (including 84 lymph node positive patients). Half (342/684) of all patients with an RS 18-25 and 20.5% (32/156) patients with RS 26-30 were MP low risk. There was no significant difference in the distribution of MP risk in women age ≤50 yrs vs. >50 years (Yates chi-square P=0.62); MP classified 46.4% (84/181) patients age ≤50 yrs and 44.0% (290/659) patients age >50 yrs as low risk. In the clinically-low risk subset of 515 patients, there was also no significant difference in the distribution of MP risk by age (Yates chi-square P=0.89); MP classified 48.3% (56/116) patients age ≤50 yrs and 49.6% (198/399) patients age >50 yrs as low risk. Conclusions: In light of TAILORx and uncertain CT benefit in women ≤50 yrs, MammaPrint provides a definitive high or low risk answer and identifies 46% of these women who may safely forego CT based on MINDACT data. An analysis of young patients in the MINDACT trial showed that MP low risk patients age <45 yrs and 45-55 yrs had very good 5-yr DMFS of 95-98%, in both clinically low and high risk groups (Alders, SABCS 2017). MammaPrint Risk by RS and AgeMammaPrint RiskRS 18-25 RS 26-30 GrandClassification≤50 yrs>50 yrsAll Ages≤50 yrs>50 yrsAll AgesTotalHigh Risk7426834223101124466Low Risk8026234242832374All15453068427129156840 Citation Format: Soliman H, Lo S, Qamar R, Budway R, Levine E, Whitworth P, Mavromatis B, Zon R, Untch S, Treece T, Blumencranz L, Audeh W, Tsai M, PROMIS Investigators Group. MammaPrint identifies 46% of patients, age ≤50 years with oncotype RS 18-30, as low risk and safe to forgo chemotherapy [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 P4-08-10.

Abstract Background: Neoadjuvant chemotherapy (NAC) is employed in patients with larger tumors to attempt to downstage locally advanced cancers to allow breast conservation and to assess in vivo tumor response. The Multi-Institutional Neoadjuvant Therapy MammaPrint Project I (MINT) study asked a secondary question of whether complete nodal downstaging could also be achieved with NAC. Methods: This analysis included 147 eligible invasive breast cancer patients with high tumor burdens, classified as cT2-4N0-3M0 (T2 greater than 3.5cm if N0). Patients who had a positive core biopsy and/or fine needle aspiration (FNA) on an axillary node prior to starting NAC were included in this analysis. Those who had a surgical sentinel lymph node biopsy were not included. Nodal involvement was established following neoadjuvant treatment by axillary lymph node dissection (ALND). Results: This population was 54% postmenopausal, average age 53 yrs (range 25 to 80 yrs). Tumor characteristics were 91% invasive ductal carcinoma; 65% T2, 29% T3, 6% T4; 87% LN1, 13% LN2-3; 3% low grade, 38% intermediate grade, 59% high grade; 65% ER-positive, 49% PR-positive, and 28% HER2-positive by immunohistochemistry; 84% High Risk (HR) and 16% Low Risk (LR) by MammaPrint (MP). After NAC, 45% (66/147) of these LN-positive patients were down-staged to ypN0 and also achieved a complete pathological response in the primary tumor. The potential for down-staging was inversely-related to tumor burden, where 47% (60/128) of N1, 35% (6/17) of N2, and 0% (0/2) of N3 patients were down-staged to ypN0. There were 3 patients who were down-staged (2 N2 to N1, and 1 N3 to N2), but not to ypN0. At surgery, 34% (44/128) of patients had no change, and 19% (24/129) progressed in LN staging. Pre vs Post NAC Nodal StagePre NAC Nodal StageypN0ypN1ypN2ypN3TotalcN16044222128cN2626317cN3 112Total6646296147 Conclusions: We confirmed that upon achieving a complete response of the primary tumor that there was also a pathologic complete response in the LN. About 53% of patients had no change or progression of LN involvement following NAC. Citation Format: Blumencranz P, Habibi M, Treece T, Blumencranz L, Yoder E, Audeh W, Carter E, McNaughton L, Roussos J, Shivers S, Acs G, Cox C, MINT Investigators Group. Neoadjuvant chemotherapy for breast cancer: Nodal downstaging is highly correlated with pathological complete response [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 PD8-04.

Abstract This abstract was not presented at the symposium.

Abstract This abstract was not presented at the symposium.

Abstract Background: Cyclin-dependent kinase (CDK) 4/6 inhibitors are being evaluated in the adjuvant setting for patients with resected early stage hormone receptor positive (HR+) and HER2 negative (HER2-) breast cancer (BC). However, biomarkers that predict benefit from this class of agents are unknown. We have recently reported results from the phase II neoadjuvant NeoPalAna trial which demonstrated that palbociclib (Pal) enhanced the anti-proliferative activity when added upon anastrozole (Ana) monotherapy in estrogen receptor (ER) positive and HER2 negative breast cancers. Interestingly, a small group of patients was resistant to Pal, exhibiting persistent tumor cell proliferation (Ki67 >2.7%) on the combination of Ana and Pal. In this study, we evaluated the utility of a research algorithm for the 70-gene signature (70-GS) in identifying Pal resistant versus sensitive patients. Methods: Serial biopsies were collected from patients at four treatment timepoints: baseline (BL), cycle 1 day 1 (C1D1) following 28 days of Ana monotherapy, cycle 1 day 15 (C1D15) at 2 weeks post the addition of Pal, and at surgery (Surg). RNA was extracted from frozen tumor biopsies at each timepoint and run on Agilent full genome microarrays (GSE93204) at Washington University. As an exploratory analysis, genes from the GPL8253 array that match the 70-GS were used to calculate a research approximation of the 70-GS index (r-GS). The distribution of the r-GS across Ki67 response groups was evaluated. Results: Ki67 had previously been measured at each timepoint, and used to classify patients as being either Ana-sensitive (C1D1 Ki67 ≤2.7%), Pal-sensitive (C1D1 Ki67 >2.7%, C1D15 Ki67 ≤2.7%), or Pal-resistant (C1D15 Ki67 >2.7%). The r-GS was differentially regulated between sensitive (AI or Pal) and Pal-resistant groups at BL (p=0.012), C1D1 (p=0.039), and C1D15 (p=0.022). The r-GS values varied widely across patients at BL, and generally became more positive (more low risk) with treatment. There was no correlation between Ki67 levels and r-GS. Furthermore, gene expression analysis was performed to elucidate the difference between Pal-sensitive vs. Pal-resistant patients, and Ana-sensitive vs. Pal-sensitive patients. Conclusions: While on-treatment Ki67 indicated drug responsiveness, baseline r-GS significantly stratified patients into sensitive (Ana or Pal) versus Pal-resistant groups in the neoadjuvant setting. This preliminary finding suggests that the 70-GS may have clinical utility in identifying patients resistant to Pal for future studies. Additionally, results of the gene expression analysis may help to further develop genomic biomarkers for Pal and Ana sensitivity and resistance. Citation Format: Hoog JW, Treece T, Blumencranz L, Audeh W, Sanati S, Ellis MJ, Ma CX. Genomic biomarker for resistance to palbociclib in the NeoPalAna trial [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 P2-09-19.

e21040 Background: SPARC is an albumin-binding protein associated with poor prognosis that may mediate intratumoral accumulation of nab-P. SPARC is associated with both tumor cells and stromal components. We examined SMS in paired primary (P) and metastatic (M) TNBC biopsies (bxs) to 1) determine changes in expression, and 2) examine potential correlations with response to therapy with nab-P/C/B. Methods: First- and second-line TNMBC pts (n = 27) received nabP (100 mg/m2) and C (AUC = 2) on d1,8,15 of a 28 d cycle. B (10 mg/kg) was given on d1, 15. Paraffin- embedded P and M bxs were required (first line). PFS, ORR, and CBR were determined. Tumor SMS was determined using a validated SPARC IHC method. Scoring included: % cells stained in each field, intensity of staining, and overall score. Multivariable IHC data were analyzed using GeneSpring and Nexus analysis programs. Results: SMS data were available for 20 pts. SMS of M bx distinguished a low risk (n= 5, PFS = 16.0 mos) from a high risk group (n = 15, PFS = 4.9 mos), p = 0.03. SMS of P bx did not correlate with outcome. Of 17 first-line paired P and M bx (from 15 pts), 8 M SPARC signatures were similar to the P (p < 0.05) and 9 were dissimilar (p ≥ 0.05). For the dissimilar group, 3 microenvironment variables were higher in the M bx compared to P bx: % Blood vessel, % Inflammatory cells, and % Fibroblasts, p = 0.006, 0.06, and 0.06. In 14 pts with paired bx and ORR data, M bx that were similar to their P had better response than those with dissimilar M bx: CR 3/7 (43%), PR 4/7 (57%), and SD 0/7 (0%)vs CR 0/7 (0%), PR 5/7 (71%), and SD 2/7 (29%), p = 0.03. PFS for similar and dissimilar pairs was not different. Conclusions: In this small sample of TNMBC, the SMS of M bxs discriminated tumors with a short PFS from those with a longer PFS. Archival P bx SMS did not correlate with outcomes and M SMS were significantly different from the corresponding P SMS in over half of the pts. The change in the SMS from the P to M tumors was associated with reduced response to nab-P/C/B chemotherapy. SMS should be explored further in both primary and metastatic tumors in order to further define it prognostic and predictive significance. Author Disclosure Employment or Leadership Position Consultant or Advisory Role Stock Ownership Honoraria Research Funding Expert Testimony Other Remuneration Abraxis BioScience Abraxis BioScience Abraxis BioScience

8578 Background: Tumor overexpression of the albumin-binding protein SPARC (secreted protein acidic and rich in cysteine) is associated with poor prognosis in multiple cancers. SPARC in the tumor microenvironment may be associated with tumor cells and stromal components. While previous studies examined SPARC only in tumor cells and/or fibroblasts, it remains unclear which tumor SPARC components may be associated with poor prognosis in different tumors. Recently we have found that plasma SPARC may contribute to poor outcomes in tumor models. Herein, we examined all tumor microenvironment components as well as plasma SPARC levels to develop a SPARC signature that may be correlated to outcomes. Methods: 76 chemo naïve and pretreated patients (pts) with unresectable stage IV melanoma were treated with nab-paclitaxel (nab-P, 100 mg/m2) and carboplatin (C, AUC 2) on d 1, 8, and 15 of a 28 day cycle until disease progression. SPARC in tumor biopsies was measured using a validated SPARC IHC. IHC data were analyzed using array analysis programs. SPARC plasma level in pts (baseline and posttreatment [Tx]) and normal controls were quantitated using a validated SPARC ELISA (blinded to patient outcome). Results: Tumor SPARC IHC data was available for 40 pts. A unique SPARC signature was developed for melanoma pts to distinguish early progressers (LR, n = 9) from later progressers (HR, n = 31). Median PFS increased from 3.8 to 6.7 months (logrank, p = 0.02) and median OS increased from 9.6 to 18.0 months (p = 0.07) for the HR vs. LR groups, respectively. Plasma SPARC levels were higher in melanoma pts than normal controls (median 259 ng/ml [95% CI = 273-327], n = 321 from 76 pts; vs. median 153 ng/mL [95% CI = 99-207], n = 50 from 50 pts; t-test, p < 0.0001). Conclusions: The SPARC microenvironment signatures (SMS) was designed to discriminate between low-risk and high-risk groups with respect to PFS/OS. These data suggest that SPARC in the tumor microenvironment may play an important role in the outcome of melanoma pts. Plasma SPARC levels in melanoma pts were higher than normal controls. Correlation of plasma SPARC with outcomes will be presented at ASCO. Author Disclosure Employment or Leadership Position Consultant or Advisory Role Stock Ownership Honoraria Research Funding Expert Testimony Other Remuneration Abraxis BioScience Abraxis BioScience Abraxis BioScience Abraxis BioScience