Risk-adapted modulation through de-intensification of cancer treatments: an ESMO classification

BACKGROUND

The landscape of clinical trials testing risk-adapted modulations of cancer treatments is complex. Multiple trial designs, endpoints, and thresholds for non-inferiority have been used; however, no consensus or convention has ever been agreed to categorise biomarkers useful to inform the treatment intensity modulation of cancer treatments.

METHODS

An expert subgroup under the European Society for Medical Oncology (ESMO) Precision Medicine Working Group shaped an international collaborative project to develop a classification system for biomarkers used in the cancer treatment de-intensification, based on a tiered approach. A group of disease-oriented clinical, translational, methodology and public health experts, and patients' representatives provided an analysis of the status quo, and scanned the horizon of ongoing clinical trials. The classification was developed through multiple rounds of expert revisions and inputs.

RESULTS

The working group agreed on a univocal definition of treatment de-intensification. Evidence of reduction in the dose-density, intensity, or cumulative dose, including intermittent schedules or shorter treatment duration or deletion of segment(s) of the standard regimens, compound(s), or treatment modality must be demonstrated, to define a treatment de-intensification. De-intensified regimens must also portend a positive impact on toxicity, quality of life, health system burden, or financial toxicity. ESMO classification categorises the biomarkers for treatment modulation in three tiers, based on the level of evidence. Tier A includes biomarkers validated in prospective, randomised, non-inferiority clinical trials. The working group agreed that in non-inferiority clinical trials, boundaries are highly dependent upon the disease scenario and endpoint being studied and that the absolute differences in the outcomes are the most relevant measures, rather than relative differences. Biomarkers tested in single-arm studies with a threshold of non-inferiority are classified as Tier B. Tier C is when the validation occurs in prospective-retrospective quality cohort investigations.

CONCLUSIONS

ESMO classification for the risk-guided intensity modulation of cancer treatments provides a set of evidence-based criteria to categorise biomarkers deemed to inform de-intensification of cancer treatments, in risk-defined patients. The classification aims at harmonising definitions on this matter, therefore offering a common language for all the relevant stakeholders, including clinicians, patients, decision-makers, and for clinical trials.

Abstract PD9-06: Evaluation of the predicted sensitivity to endocrine therapy (SET2,3 index) and the 21-gene Breast Recurrence Score® assay in node-positive postmenopausal breast cancer: Results from an analysis in the SWOG S8814 trial

Background. SWOG trial S8814 randomized postmenopausal patients with pathologic lymph node-positive (N+) breast cancer that was hormone receptor-positive to receive adjuvant anthracycline-based chemotherapy (cyclophosphamide, doxorubicin, fluorouracil) followed by tamoxifen endocrine therapy for 5 years (CAF-T), versus tamoxifen alone (TAM). The 21-gene Breast Recurrence Score® assay was prognostic in S8814 and predicted chemotherapy benefit in patients with higher Recurrence Score® (RS) (Albain et al, Lancet Oncol 2009). Other prognostic signatures have yet to be evaluated in this cohort. The sensitivity to endocrine therapy index (SET2,3) measures non-proliferative hormone receptor-related transcription (SETER/PR) adjusted for a baseline prognosis index derived from tumor size, nodes involved and a 4-gene molecular subtype (RNA4) (Du et al, Ann Oncol 2021). SET2,3 has been shown to provide prognostic information independent from neoadjuvant chemotherapy response. We sought to evaluate the predictive and prognostic value of SET2,3 in SWOG 8814. Methods. Independently, the SET2,3 index and cut point were calibrated from their diagnostic platform to the whole transcriptome RNA sequencing (RNAseq) platform in 85 sample pairs. Expression of the 31 transcripts used for SET2,3 were provided from RNAseq data of 283 tumors in S8814 (all previously tested for RS). Blinded calculated results of SET2,3 were then merged with outcome data. The planned analysis tested whether SET2,3 (continuous index, dichotomized high/low) provided additional prognostic information to RS (overall and in pts. with RS≤25) by treatment arm, and whether low SET2,3 was associated with chemo benefit. Cox proportional hazards models of disease-free survival (DFS) included SET2,3; RS; treatment arm; and (where relevant) interaction term and reported hazard ratios (HR) and 95% confidence intervals (95%CI). Results. There were 106 events over median follow-up of 9.1 years in 283 patients. 175 patients had RS ≤25, 108 had RS >25. Distribution of the SET2,3 low was similar in both RS high (51%) and low groups (47%), reflecting minimal correlation between the two. As proportional hazards assumptions were met during the first 5 years only the analysis was restricted to 5 years. Adjusting for treatment arm, high SET2,3 category was highly prognostic in this randomized trial (HR 0.27, 95% CI 0.15-0.49, p<0.0001). High SET2,3 was not predictive of chemotherapy response (interaction p=0.83). In multivariable Cox models (Table), continuous RS and SET2,3 were independently prognostic in the overall population for each treatment arm (p≤0.01), whereas only SET2,3 was prognostic for patients with RS≤25 (N=175, p<0.001). In patients with RS≤25, continuous SET2,3 was prognostic within the CAF-T arm (HR 0.34, p=0.006) with similar results in the TAM alone arm (HR 0.38, p=0.062). Conclusions. SET2,3 added independent prognostic information to RS results in the S8814 trial for patients with N+ disease treated with tamoxifen, though it was not predictive of benefit from adjuvant chemotherapy. When RS result ≤25, SET2,3 remained independently prognostic. Hence, SET2,3 provided independent information complementary to RS, possibly because it incorporates tumor size and number of positive nodes. SET2,3 warrants further evaluation in patients with N+ breast cancer.

Table: Multivariable Cox models in the overall population and subset with RS≤25 by treatment arm.CohortTreatment ArmContinuous Recurrence ScoreContinuous SET2,3HR (95%CI) per 10 unitsp-valueHR (95%CI)per 1 unitp-valueAll RS(N=283)CAF-TAM(N=166)1.21 (1.04-1.40)0.0120.48 (0.31-0.76)0.002TAM(N=117)1.44 (1.18-1.76)< 0.0010.48 (0.27-0.88)0.017RS≤25 (N=175)CAF-TAM(N=99)1.43 (0.58-3.49)0.440.34 (0.15-0.73)0.006TAM(N=76)1.66 (0.46-5.93)0.440.38 (0.14-1.05)0.062

Citation Format: Corey W. Speers, W. Fraser Symmans, William E. Barlow, Alex Trevarton, Stephanie The, Lili Du, James M. Rae, Steven Shak, Frederick L. Baehner, Priyanka Sharma, Lajos Pusztai, Gabriel N. Hortobagyi, Daneil F Hayes, Kathy S. Albain, Andrew Godwin, Alastair Thompson. Evaluation of the predicted sensitivity to endocrine therapy (SET2,3 index) and the 21-gene Breast Recurrence Score® assay in node-positive postmenopausal breast cancer: Results from an analysis in the SWOG S8814 trial [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr PD9-06.

Impact of homologous recombination deficiency biomarkers on outcomes in patients with triple-negative breast cancer treated with adjuvant doxorubicin and cyclophosphamide (SWOG S9313)

Background

Homologous recombination deficiency (HRD)-causing alterations have been reported in triple-negative breast cancer (TNBC). We hypothesized that TNBCs with HRD alterations might be more sensitive to anthracycline plus cyclophosphamide-based chemotherapy and report on HRD status and BRCA1 promoter methylation (PM) as prognostic markers in TNBC patients treated with adjuvant doxorubicin (A) and cyclophosphamide (C) in SWOG9313.

Patients and methods

In total, 425 TNBC patients were identified from S9313. HRD score, tumor BRCA1/2 sequencing, and BRCA1 PM were carried out on DNA isolated from formalin-fixed paraffin-embedded tissue. Positive HRD status was defined as either a deleterious tumor BRCA1/2 (tBRCA) mutation or a pre-defined HRD score ≥42. Markers were tested for prognostic value on disease-free survival (DFS) and overall survival (OS) using Cox regression models adjusted for treatment assignment and nodal status.

Results

HRD status was determined in 89% (379/425) of cases. Of these, 67% were HRD positive (27% with tBRCA mutation, 40% tBRCA-negative but HRD score ≥42). HRD-positive status was associated with a better DFS [hazard ratio (HR) 0.72; 95% confidence interval (CI) 0.51-1.00; P = 0.049] and non-significant trend toward better OS (HR = 0.71; 95% CI 0.48-1.03; P = 0.073). High HRD score (≥42) in tBRCA-negative patients (n = 274) was also associated with better DFS (HR = 0.64; 95% CI 0.43-0.94; P = 0.023) and OS (HR = 0.65; 95% CI 0.42-1.00; P = 0.049). BRCA1 PM was evaluated successfully in 82% (348/425) and detected in 32% of cases. The DFS HR for BRCA1 PM was similar to that for HRD but did not reach statistical significance (HR = 0.79; 95% CI 0.54-1.17; P = 0.25).

Conclusions

HRD positivity was observed in two-thirds of TNBC patients receiving adjuvant AC and was associated with better DFS. HRD status may identify TNBC patients who receive greater benefit from AC-based chemotherapy and should be evaluated further in prospective studies.

Clinical Trials Number

Int0137 (The trial pre-dates Clinicaltrial.Gov website establishment).

Abstract PD9-12: Integrative molecular profiling of breast cancer brain metastasis and patient-derived xenograft organoids from resected breast cancer brain metastases to interrogate and prioritize therapeutic personalized strategies

Background: Breast cancer brain metastasis (BM) is an area of unmet need in metastatic breast cancer patients. Novel therapeutic interventions to help prevent and treat BM are warranted. We conducted integrative molecular profiling of BM and matched primary tumors (PT) using next-generation DNA and RNA sequencing to examine the molecular landscape. In addition, we established patient-derived xenograft/organoid (PDX/PDO) to examine drug sensitivity according to the molecular and clinical features of the BM.

Methods: Archived, formalin fixed paraffin-embedded BM was collected retrospectively. BM were also collected prospectively at the time of clinically indicated surgical resection through the central nervous system tissue banking and the Michigan Oncology Sequencing Center (MI-ONCOSEQ) protocols. Matched archived PT tissues were collected when available. Integrative next-generation sequencing was conducted using the MI-ONCOSEQ platform. The prospectively collected BM were further used to establish PDXs/ PDOs. Successfully established PDXs/PDOs were used for ex vivo drug testing via MiDrugScreen, a novel drug sensitivity testing platform, where testing was performed in a dose-response format with drug selection prioritized by clinical scenario and molecular alterations if known a priori.

Results: 12 matched BM-PT pairs were analyzed: 6 triple negative, 5 HER2 positive, and 1 ER positive HER2 negative. All except one (11/12) had TP53 mutations. When present, TP53 mutations in BM were also found in PT (except for 1 unknwon case in PT due to low coverage). ER+HER2- was the only one without TP53 mutation but had hyper-mutation (APOBEC signature). Driver mutations and unique copy number alterations (CDKN2A loss in 1/12, mutations in PIK3CA in 1/12 and ESR1 in 1/12, CCNE1 amplification in 1/12) were noted in BMs. In 75% of cases, mutational burden was higher in BM vs. PT. 2 PDX/PDO were available for drug testing. PDO-BC9 was noted to have RB1 (splice acceptor) and LOH. As predicted by this alteration, PDO-BC9 was insensitive to CDK4/6 inhibitors (palbociclib, abemaciclib) tested on MiDrugScreen panel. PDX-BC4 was established from PIK3CA and ESR1 mutated BM from an ER+HER2- patient who had previously progressed on endocrine therapy with a CDK4/6 inhibitor. As predicted, the PDX-BC4 was resistant to CDK4/6 inhibitor but interestingly sensitive to PIK3CA, ERK, and MEK inhibitors.

Conclusions: TP53 mutation was highly prevalent and may be a biomarker for increased risk of BM. Further study is warranted to see if specific TP53 mutations are associated with a risk of BM development and can be used in risk stratification for BM specific intervention. Unique molecular alterations in BM compared to matched PT may have a therapeutic implication as a target or resistance biomarker. Conducting drug testing in addition to molecular profiling has the strong potential of being informative in tailoring or prioritizing therapeutic agents in the era of precision medicine. Additional BM PDXs/PDOs from breast and other solid tumors are being examined using this novel therapeutic tailoring approach with the combination of MIONCOSEQ and MiDrugScreen.

Citation Format: Morikawa A, Robinson DR, Soellner M, Wu Y-M, Lonigro R, Gilani R, Cheng X, Lachacz E, Thomas D, McMurray K, Smerage J, Henry NL, Heth J, Chinnaiyan A, Hayes DF, Merajver S. Integrative molecular profiling of breast cancer brain metastasis and patient-derived xenograft organoids from resected breast cancer brain metastases to interrogate and prioritize therapeutic personalized strategies [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 PD9-12.

Abstract P4-02-01: Analytical validation of an automated digital scoring protocol for Ki67: International multicenter collaboration study

Background/Goal: Ki67 expression has been a valuable prognostic marker in breast cancer, but has not seen broad adoption due to lack of standardization between institutions. Automation could represent a solution. Here we tested 3 automated digital image analysis (DIA) platforms including an open source platform to: (i) Investigate the reproducibility of Ki67 measurement across platforms with supervised classifiers performed by the same operator and by multiple operators. (ii) Compare accuracy of the 3 DIA platforms against outcome (prognostic potential). (iii) Assess inter-laboratory reproducibility of a calibrated DIA tool to evaluate Ki67 in breast cancer among 10 participating labs of the International Ki67 in Breast Cancer Working Group (IKWG).

Methods: The Mib-1 antibody (Dako) was used to detect Ki67 (dilution 1:100). HALO (H) (IndicaLabs), QuantCenter (QC) (3DHistech), QuPath (QP) (open-source software) digital image analysis (DIA) platforms were used to evaluate Ki67 expression. As a ground truth, we evaluated Ki67 LI with meticulous manual tissue segmentation using the Spectrum Webscope (SW) (Aperio). Calibration was performed using 30 ER+ breast cancer cases from phase 3 of the IKWG initiative where blocks were centrally cut and stained for Ki67. The inter-laboratory analysis was done with 10 participating laboratories divided into 2 groups where members within the same group were given the same set of images. The outcome cohort consisted of 149 breast cancer cases from the Yale Pathology archives in tissue microarray format. Intra-class correlation coefficient (ICC) was used to measure reproducibility with the pre-specified criterion for success being to exceed 0.80. Kaplan-Meier analysis supported with log-rank test was performed to assess prognostic potential.

Results: All 3 DIA platforms showed excellent inter-platform reproducibility (ICC: 0.933, CI: 0.879-0.966). Also, excellent reproducibility was found between all DIA platforms and the reference standard Ki67 values of SW (QP ICC: 0.970, CI: 0.936-0.986; H ICC: 0.968, CI: 0.933-0.985; QC ICC: 0.964, CI: 0.919-0.983). The intra-DIA reproducibility was also excellent for all platforms (QP ICC: 0.992, CI: 0.986-0.996; H ICC: 0.972, CI: 0.924-0.988; QC ICC: 0.978, CI: 0.932-0.991). Comparing each DIA against outcome, the hazard ratios were similar (QP=3.309, H=3.077, QC=3.731). The inter-operator reproducibility was particularly high (ICC: 0.962-0.995). As QP is open source software and also showed the lowest intra-DIA platform variability, we selected the QP platform to investigate inter-laboratory reproducibility among 10 IKWG labs. The different-section ICC across the 10 labs was 0.974 (CI: 0.954 - 0.986). The same-section ICC estimate was 0.984 (CI: 0.971-0.992) for group 1 and 0.978 (CI: 0.956-0.989) for group 2.

Conclusions: Our results showed outstanding reproducibility both within and between DIA platforms. We also found the platforms essentially indistinguishable with respect to prediction of breast cancer patient outcome. Automated Ki67 evaluation using a calibrated, open-source DIA platform (QuPath) met the pre-specified criterion of success in the multi-institutional setting. Assessment of clinical utility is planned.

Citation Format: Acs B, Leung SC, Pelekanou V, Bai Y, Martinez-Morilla S, Toki M, Chang MC, Gholap A, Jadhav A, Hugh JC, Bigras G, Laurinavicius A, Augulis R, Levenson R, Todd A, Piper T, Virk S, van der Vegt B, Hayes DF, Dowsett M, Nielsen TO, Rimm DL. Analytical validation of an automated digital scoring protocol for Ki67: International multicenter collaboration study [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 P4-02-01.

Abstract P4-08-06: Impact of molecular subtypes on long-term outcomes in triple-negative breast cancer (TNBC) patients treated with adjuvant AC chemotherapy on SWOG S9313

Introduction: TNBC is heterogeneous disease with several molecularly defined subtypes (Lehman et al), each of which may be predictive of response to chemotherapy. TNBC molecular subtypes are associated with varied pathological responses to neoadjuvant chemotherapy. However, subtype specific long-term outcomes for TNBC patients treated with uniform adjuvant chemotherapy are not known.

Aims: To characterize long-term outcomes of TNBC molecular subtypes (TNBCtypes) in patients treated with adjuvant doxorubicin (A) and cyclophosphamide (C) on S9313

Methods: SWOG 9313 accrued 3,125 women with early stage breast cancer to two alternative dose schedules of AC with no difference in outcomes between the two arms (J Clin Oncol 2007). From this trial we identified 425 (14%) patients with centrally determined TNBC for whom tissue was available. Microarray profiling was performed on genomic RNA extracted from pre-treatment FFPE tissue. A 101-gene expression model which has shown to reproduce the classification provided by the original 2188-gene algorithm (Ring et al) was applied to the microarray profiling to generate the following TNBCtypes–Basal-Like 1 (BL1), Basal-Like 2 (BL2), Mesenchymal (M), mesenchymal stem–like (MSL), and luminal androgen receptor (LAR). Immunomodulatory +/- (IM) status was assigned independent of the subtypes. Sequencing of BRCA1/2 from tumor DNA was also performed. The subtypes were tested for prognostic effect on DFS and OS using Cox regression model with adjustment for nodal status.

Results: For 425 TNBC patients, the median age was 45 years, 33% were node-positive and 10-year DFS and OS = 66.3% and 74.1%, respectively. A total of 381/424 (89.7%) cases could be classified into TNBCtypes with distribution as follows: BL1=24%, BL2=8%, M=24%, MSL=11%, LAR=9%, unclassified (UNL) =24%. No association between TNBCtypes and race or nodal status was noted. Compared to other subtypes LAR subtype was associated with older age at diagnosis (median age 53 vs 45, p<0.001). Overall 24% of samples were IM+ and 25% demonstrated deleterious tBRCA1/2 mutation. DFS, tBRCA1/2 mutation and IM+ status distribution across different subtypes are provided in the table. All subtypes except for LAR demonstrated a drop in hazard function for recurrence after 5 years.

 5 year DFS (%)10 year DFS (%)DFS HR (95% CI), p valueDeleterious tBRCA1/2 mutationIM+ statusBL184.5%77.5%141%60%BL281.3%70.5%1.59 (0.81-3.13) p = 0.1816%12%M69.2%61.2%2.06 (1.25-3.40) p = 0.00528%0%MSL54.8%50.0%2.38 (1.33-4.28) p = 0.00418%7%LAR74.3%53.8%2.24 (1.22-4.14) p = 0.0112%8%UNL76.4%71.8%1.36 (0.80-2.33) p = 0.2620%30%

Conclusions: In the presence of adjuvant AC, TNBC molecular subtypes have varied prognosis, with BL1 subtype demonstrating the best prognosis and MSL and LAR subtypes demonstrating the worst prognosis. LAR subtype is associated with older age at diagnosis and continued elevated hazard function for recurrence after year 5. tBRCA1/2 mutations are distributed across all subtypes with the highest prevalence in BL1 and M subtypes. IM+ status was infrequently noted in non-BL1 subtypes. These findings underscore TNBC heterogeneity and the need to account for this heterogeneity in prospective clinical trials.

Citation Format: Sharma P, Barlow WB, Hout DR, Seitz RS, Bailey DB, Godwin AK, Pathak H, Timms KM, Solimeno C, Linden HM, Porter P, Tripathy D, Hortobagyi GN, Thompson A, Pusztai L, Hayes DF. Impact of molecular subtypes on long-term outcomes in triple-negative breast cancer (TNBC) patients treated with adjuvant AC chemotherapy on SWOG S9313 [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 P4-08-06.

Abstract GS4-07: Race, ethnicity and clinical outcomes in hormone receptor-positive, HER2-negative, node-negative breast cancer: results from the TAILORx trial

Background: Black race is associated with worse outcomes in localized hormone receptor (HR)-positive breast cancer in population-based and in clinical trial cohorts, whether using self-identified race (Albain et al. JNCI 2009 [PMID: 19584328; Sparano et al. JNCI 2012 [PMID: 22250182) or genetically-identified race (Schneider et al. J Precision Oncol 2017 [PMID: 29333527]). This disparity persists after adjustment for treatment delivery parameters (Hershman et al. JCO 2009 [PMID:19307504]). We evaluated clinicopathologic characteristics, treatment delivered and clinical outcomes in the Trial Assigning Individualized Options for Treatment (TAILORx) by race and ethnicity (Sparano et al. NEJM 2018 [PMID: 29860917]).

Methods: The analysis included 9719 evaluable TAILORx participants. The association between clinical outcomes and race (white, black, Asian, other/unknown) and ethnicity (Hispanic vs. non-Hispanic) was examined, including invasive disease-free survival (iDFS), distant relapse-free interval (DRFI), relapse-free interval (RFI), and overall survival (OS). Proportional hazards models were fit including age (5 categories), tumor size (>2 cm vs. <=2 cm), histologic grade (high vs. medium vs. low vs. unknown), continuous recurrence score (RS), race, and ethnicity in the overall population and randomized treatment arms in the RS 11-25 cohort.

Results: The study population included 8189 (84%) whites, 693 (7%) blacks, 405 (4%) Asians, and 432 (4%) with other/unknown race. Regarding ethnicity, 7635 (79%) were non-Hispanic, 889 (9%) Hispanic, and 1195 (12%) unknown. There was no significant difference in RS distribution (p=0.22) in blacks compared with whites, or in median (17 vs. 17) or mean RS (19.1 vs. 18.2). There was likewise no difference in Hispanic vs. non-Hispanic ethnicity for RS distribution (p=0.72) or median (17 vs. 17) or mean RS (18.5 vs. 18.0). Black race (39% vs. 30%) and Hispanic ethnicity (39% vs. 30%) were both associated with younger age (</=50 years) at diagnosis. The use and type of adjuvant chemotherapy and endocrine therapy, and duration of endocrine therapy, were similar in black (vs. white) and Hispanic (vs. non-Hispanic) populations. In proportional hazards models, black race (compared with white race) was associated with worse clinical outcomes in the entire population and in those with a RS 11-25 (see table). Hispanic ethnicity was generally associated with better outcomes (compared with non-Hispanic ethnicity). For the cohort with a RS of 11-25, there was no evidence for chemotherapy benefit for any racial or ethnic group.

Race (black vs.white) and clinical outcomes in proportional hazards modelsClinical endpointEntire Population (N=693 black) Hazard ratio for eventRS 11-25 (N=471 black) Hazard ratio for eveniDFS1.33 (p=0.005)1.49 (p=0.001)DRFI1.21 (p=0.28)1.60 (p=0.02)RFI1.39 (p=0.02)1.80 (p<0.001)OS1.52 (p=0.005)1.67 (p=0.003

Conclusions: In patients eligible and selected for participation in TAILORx, black women had worse clinical outcomes despite similar 21-gene assay RS results and comparable systemic therapy. This adds to an emerging body of evidence suggesting a biologic basis or other factors contributing to racial disparities in HR-positive breast cancer that requires further evaluation.

Citation Format: Albain K, Gray RJ, Sparano JA, Makower DF, Pritchard KI, Hayes DF, Geyer, Jr. CE, Dees EC, Goetz MP, Olson, Jr. JA, Lively T, Badve SS, Saphner TJ, Wagner LI, Whelan TJ, Ellis MJ, Paik S, Wood WC, Ravdin PM, Keane MM, Gomez HL, Reddy PS, Goggins TF, Mayer IA, Brufsky AM, Toppmeyer DL, Kaklamani VG, Berenberg JL, Abrams J, Sledge, Jr. GW. Race, ethnicity and clinical outcomes in hormone receptor-positive, HER2-negative, node-negative breast cancer: results from the TAILORx trial [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 GS4-07.

Abstract PD5-07: A phase III randomized trial of anastrozole and fulvestrant versus anastrozole or sequential anastrozole and fulvestrant as first-line therapy for postmenopausal women with metastatic breast cancer: Final survival outcomes of SWOG S0226

Background: Anastrozole depletes estrogen via aromatase inhibition and fulvestrant binds and degrades estrogen receptor. In a Phase III trial we compared the concurrent use of these agents to anastrozole alone or sequential anastrozole and fulvestrant in first-line therapy of hormone receptor-positive metastatic breast cancer in postmenopausal women, and demonstrated improved progression-free (PFS) and overall survival (OS)-NEJM 2012. Now we report PFS and OS five years after the initial positive findings. Methods: A total of 707 patients were randomized to either 1 mg anastrozole P.O. daily (Arm 1) or to the combination of anastrozole and fulvestrant (Arm 2). Fulvestrant was administered as a loading dose of 500 mg on day 1, 250 mg on days 14, 28 and monthly thereafter. Randomization was stratified by adjuvant tamoxifen use. The primary endpoint was PFS with OS a secondary outcome. 40% patients not in visceral crisis crossed over to fulvestrant after progression on arm 1. Analysis of survival was by 2-sided stratified log-rank tests and Cox regression using intent-to-treat. Subset analyses include treatment effect by adjuvant tamoxifen exposure, initial sites of metastases and time from diagnosis. Results: There were 646 PFS events (328 and 318 for arms 1 and 2, respectively) among 694 eligible patients (345 and 349, respectively). Overall, median PFS was 13.5 months for arm 1 and 15.0 months for the arm 2 (log-rank p=0.007; HR=0.81 (95% CI 0.69-0.94)). This benefit extended similarly in visceral and non-visceral subgroups. In subset analysis for Arms 1 and 2, respectively, in tamoxifen-naive women (60%, n=414), median PFS was 12.7 vs. 16.7 months (log-rank p=0.002; HR=0.73 (95% CI 0.60-0.89) while in women exposed to tamoxifen, median PFS was 13.9 vs. 13.6 months (log-rank p=0.57; HR=0.93 (95% CI 0.73-1.19)). An improved OS in the combination arm was seen, median OS 42 and 50 months in arms 1 and 2, based on 261 and 247 deaths, respectively (log-rank p=0.028; HR=0.82 (95% CI 0.69-0.98)). In subset analysis in tamoxifen-naive women, median OS was 40.3 vs. 52.2 months for Arms 1 and 2, respectively (log-rank p=0.007; HR=0.73 (95% CI 0.58-0.92)) while in women exposed to tamoxifen, median OS was 43.5 vs. 48.2 months (log-rank p=0.85; HR=0.97 (95% CI 0.74-1.27). Patients with initial diagnosis >10 years benefitted most from the combination (HR=0.66 (95% CI 0.49-0.89)) regardless of tamoxifen exposure. Patients in Arm 1 who crossed over had post-progression survival similar to post-progression survival of Arm 2 patients. Conclusion: The addition of fulvestrant to anastrozole was associated with improved long-term PFS and OS compared to anastrozole alone, despite the use of fulvestrant at a dose lower than the approved, and despite the substantial cross over to fulvestrant after progression on anastrozole alone. The benefit was especially notable in those without recent exposure to adjuvant endocrine therapy. Ongoing translational medicine studies will further refine the need for up front fulvestrant. ClinicalTrials.gov:NCT00075764. Funding: NIH/NCI U10CA180888, U10CA180819 and AstraZeneca.

Citation Format: Mehta RS, Barlow WE, Albain KS, Vandenberg TA, Dakhil SR, Tirumali NL, Lew DL, Hayes DF, Gralow JR, Linden HM, Livingston RB, Hortobagyi GN. A phase III randomized trial of anastrozole and fulvestrant versus anastrozole or sequential anastrozole and fulvestrant as first-line therapy for postmenopausal women with metastatic breast cancer: Final survival outcomes of SWOG S0226 [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 PD5-07.

Abstract P1-09-05: Radiosensitization of androgen receptor (AR)-positive triple-negative breast cancer (TNBC) cells using seviteronel (INO-464), a selective CYP17 lyase and AR inhibitor

Background: Increased rates of locoregional recurrence (LR) have been observed in TNBC despite chemotherapy and radiation (RT). A novel radiosensitizer screen nominated the AR as a promising target for treatment of radioresistant breast cancer, including TNBC. We assessed the activity of seviteronel (Sevi), a selective CYP17 lyase and AR inhibitor in Phase 2 clinical development for advanced breast and prostate cancer, as a potential radiosensitizer in AR+ TNBC model.

Methods: Clonogenic survival assays were used to determine the intrinsic RT sensitivity of 21 breast cancer cell (BCC) lines. IC50 values were determined for 130 clinically available compounds and correlation coefficients were calculated using IC50 values and SF-2Gy. Gene expression was measured using RNA Seq or qRT-PCR and protein expression was measured using RPPA arrays. AR function was assessed using functional inhibition with Sevi in MDA-MB-453, ACC-422, ACC-460, SUM-185 (all four AR+ TNBC), MDA-MB-231 (AR- TNBC), and T47D (AR- ER+) BCC lines. Double-stranded DNA (dsDNA) break repair was assessed with γH2AX foci counting. In vivo tumor growth was measured with varying control and treatment groups (16-20 tumors/group). Kaplan-Meier analysis was performed to estimate local control. A Cox proportional hazards model and multi-variate analysis (MVA) were used to determine variables associated with LRF survival.

Results: Our novel radiosensitizer screen identified the activity of anti-androgen therapy as a potentially effective strategy for radiosensitization in RT-resistant BCC lines (R2 =0.46, p-value < 0.01) (Speers et al, J Clin Oncol 35, 2017 (suppl; abstr e12102). Heterogeneity in AR expression was identified in human BCC lines and TNBC samples from patients (N=2098). There was a strong correlation between AR RNA expression and protein expression across all BC intrinsic subtypes. AR inhibition using Sevi induced radiation sensitivity in vitro with an enhancement ratio (ER) of 1.24-1.69 in four different AR+ TNBC lines. No such radiosensitization was seen in AR(-) TNBC or ER+, AR(-) BCC lines. Radiosensitization was at least partially dependent on impaired dsDNA break repair with significant delays in dsDNA break repair at 16 and 24 hours in all AR+ TNBC lines examined (p-value < 0.01). AR inhibition with Sevi significantly radiosensitized AR+ TNBC xenografts in mouse models and markedly delayed tumor-volume tripling time (TTT) and tumor growth (MDA-MB-453: median TTT 16.1 days for RT alone vs. not reached after 45 days for Sevi+RT, p-value <0.001). Similar delays were seen in tumor growth, weight, and tumor doubling. Clinically, TNBC patients whose tumors had higher than median expression of AR had higher rates of LR after RT (HR for LR ˜3, p-value <0.01, 2 independent datasets). In MVA, high AR expression was the variable most significantly associated with worse LR survival after RT in TNBC patients, outperforming all other variables (HR of 3.42; p-value < 0.01).

Conclusions: Our results implicate the AR as a mediator of radioresistance in breast cancer and support the rationale for developing Sevi as a novel radiosensitizing agent in AR+ TNBC.

Citation Format: Speers CW, Chandler B, Olsen E, Wilder-Romans K, Moubadder L, Nyati S, Rae J, Hayes DF, Spratt DE, Wahl DR, Eisner J, Feng FY, Pierce LJ. Radiosensitization of androgen receptor (AR)-positive triple-negative breast cancer (TNBC) cells using seviteronel (INO-464), a selective CYP17 lyase and AR inhibitor [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 P1-09-05.

Abstract P2-03-01: Analytical validation of a standardized scoring protocol for Ki67 assessed on breast excision whole sections: An international multicenter collaboration

Aims: (i) Determine whether between-observer reproducibility for Ki67 when assessed on whole sections according to a standardized scoring protocol is adequate for clinical application. (ii) Compare between-observer reproducibility of Ki67 scores assessed on hot-spots to scores using a global method that averages across a tissue section.

Background: The nuclear proliferation biomarker Ki67 has multiple potential roles in breast cancer, including aiding decisions based on prognosis, but unacceptable levels of between-laboratory variability have been observed. The International Ki67 in Breast Cancer Working Group has undertaken a systematic program to determine whether Ki67 measurement can be analytically validated and standardized across labs. In phase 1, variability in visual interpretation was identified as an important source of variability. Phases 2 and 3a showed that adherence to defined scoring methods substantially improved reproducibility in scoring tissue microarrays and core-cut biopsies. We now assess whether acceptable reproducibility can be achieved on whole sections.

Methods: Adjacent sections from 30 primary ER+ breast cancers were centrally stained for Ki67 to assemble 4 sets of 30 stained tumor sections, circulated around 23 labs in 12 countries. Ki67 was scored by 2 methods by all labs: (a) global: 4 fields of 100 tumor cells each were selected to reflect observed heterogeneity in nuclear staining (b) hot-spot: the field with highest Ki67 percentage of tumor cells with nuclear staining was selected and up to 500 cells scored. Ki67 scores were log2-transformed for statistical analyses and back-transformed for presentation. The primary objective was to assess whether either method could achieve an intraclass correlation coefficient (ICC) significantly greater than 0.8, considered substantial to almost-perfect reproducibility. Secondary objectives were to assess which method had highest observed ICC and to assess whether observers identified the same “hot-spots”.

Results: ICC for the global method was 0.87 (95%CI: 0.799-0.93), marginally meeting the prespecified success criterion. The ICC for the hot-spot method was 0.83 (95%CI: 0.74-0.90) and had a CI extending below the success criterion. Across the 23 labs, geometric mean value of the 30 scores ranged from 8.5 to 19.6 for the global method and from 12.8 to 30.3 for the hot-spot method. The overall mean (95% CI) of these values was 12.9 (11.9-14.0) and 20.9 (19.1-22.8), respectively. Visually, between-laboratory agreement in location of selected hot-spot varies between cases. The median times for scoring were 9 and 6 minutes for global and hot-spot methods respectively.

Conclusions: The global method marginally met the prespecified criterion of success; it should now be evaluated for clinical validity in appropriate cohorts of cases. The hot-spot method was observed to have slightly less reproducibility between labs. The time taken for scoring by either method is practical using counting software we are making publicly available. Establishment of external quality assessment schemes is likely to improve the reproducibility between labs further.

(Supported by a grant from the Breast Cancer Research Foundation)

Citation Format: Nielsen TO, Leung SCY, Zabaglo LA, Arun I, Badve SS, Bane AL, Bartlet JMS, Borgquist S, Chang MC, Dodson A, Ehinger A, Fineberg S, Focke CM, Gao D, Gown AM, Gutierrez C, Hugh JC, Kos Z, Lænkholm A-V, Mastropasqua MG, Moriya T, Nofech-Mozes S, Osborne CK, Penault-Llorca FM, Piper T, Sakatani T, Salgado R, Starczynski J, Sugie T, van der Vegt B, Viale G, Hayes DF, McShane LM, Dowsett M. Analytical validation of a standardized scoring protocol for Ki67 assessed on breast excision whole sections: An international multicenter collaboration [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 P2-03-01.

Abstract P2-09-26: Frequency and mechanisms of elevated somatic mutation burden in metastatic breast cancer and response to immune checkpoint blockade

Background: Immune checkpoint blockade (ICB) is effective in the treatment of various malignancies. Thus far, however, results in breast cancer have been mixed. Elevated tumor mutational load, and subsequent increased likelihood of forming immunogenic neoantigens, has been correlated with response to ICB. Mutational load observed in breast cancers varies widely. However, most studies have assessed mutational load using primary tumors. Few studies have explored the frequency of high mutational load, molecular mechanisms accounting for this phenomenon, and its potential impact on response to ICB in metastatic breast cancer (MBC).

Methods: From 2011-2016, 124 patients (pts) with MBC of varying subtypes underwent research biopsy of their metastatic disease for whole genome, exome and transcriptome sequencing of tumor and matched normal sample through the Michigan Oncology Sequencing Center (Mi-OncoSeq). Those pts with elevated somatic mutation load were defined as having greater than 10 mutations per megabase of targeted sequencing and mutational signatures accounting for high mutation load were noted. Pts treated subsequently with ICB were followed to assess response.

Results: Twelve MBC pts had high mutation load (10% of cohort). Eight pts had estrogen receptor (ER) positive MBC and 4 pts had metastatic triple negative breast cancer (TNBC). In 5 cases, a clear mutational signature accounting for high mutation load was evident. Two TNBC cases harbored an APOBEC mutational signature in addition to 1 TNBC and 2 ER positive tumors displaying a microsatellite instability signature (MSI-H). Among the tumors with MSI-H signature, 1 case was associated with a pathogenic germline alteration in MLH1. Two pts were subsequently treated with ICB on a clinical trial. One pt came off study after 3 months due to progressive brain metastases and another had partial response to therapy lasting 7 months.

Conclusions: Elevated somatic mutation burden in MBC is observed in approximately 10% of pts, and is detected in both ER positive and TNBC. Since high mutation burden has been associated with increased likelihood of response to ICB, identification of this genomic feature could have important therapeutic implications for MBC pts.

Citation Format: Cobain EF, Robinson DR, Wu Y-M, Lonigro R, Vats P, Rabban E, Kumar-Sinha C, Schott AF, Smerage JB, Morikawa A, Burness ML, Van Poznak CH, Griggs J, Wicha M, Hayes DF, Chinnaiyan AM. Frequency and mechanisms of elevated somatic mutation burden in metastatic breast cancer and response to immune checkpoint blockade [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 P2-09-26.

De-escalating and escalating treatments for early-stage breast cancer: the St. Gallen International Expert Consensus Conference on the Primary Therapy of Early Breast Cancer 2017

The 15th St. Gallen International Breast Cancer Conference 2017 in Vienna, Austria reviewed substantial new evidence on loco-regional and systemic therapies for early breast cancer. Treatments were assessed in light of their intensity, duration and side-effects, seeking where appropriate to escalate or de-escalate therapies based on likely benefits as predicted by tumor stage and tumor biology. The Panel favored several interventions that may reduce surgical morbidity, including acceptance of 2 mm margins for DCIS, the resection of residual cancer (but not baseline extent of cancer) in women undergoing neoadjuvant therapy, acceptance of sentinel node biopsy following neoadjuvant treatment of many patients, and the preference for neoadjuvant therapy in HER2 positive and triple-negative, stage II and III breast cancer. The Panel favored escalating radiation therapy with regional nodal irradiation in high-risk patients, while encouraging omission of boost in low-risk patients. The Panel endorsed gene expression signatures that permit avoidance of chemotherapy in many patients with ER positive breast cancer. For women with higher risk tumors, the Panel escalated recommendations for adjuvant endocrine treatment to include ovarian suppression in premenopausal women, and extended therapy for postmenopausal women. However, low-risk patients can avoid these treatments. Finally, the Panel recommended bisphosphonate use in postmenopausal women to prevent breast cancer recurrence. The Panel recognized that recommendations are not intended for all patients, but rather to address the clinical needs of the majority of common presentations. Individualization of adjuvant therapy means adjusting to the tumor characteristics, patient comorbidities and preferences, and managing constraints of treatment cost and access that may affect care in both the developed and developing world.

Elevated S100A8 protein expression in breast cancer cells and breast tumor stroma is prognostic of poor disease outcome

PURPOSE

Elevated S100A8 expression has been observed in cancers of the bladder, esophagus, colon, ovary, and breast. S100A8 is expressed by breast cancer cells as well as by infiltrating immune and myeloid cells. Here we investigate the association of elevated S100A8 protein expression in breast cancer cells and in breast tumor stroma with survival outcomes in a cohort of breast cancer patients.

PATIENTS AND METHODS

Tissue microarrays (TMA) were constructed from breast cancer specimens from 417 patients with stage I-III breast cancer treated at the University of Michigan Comprehensive Cancer Center between 2004 and 2006. Representative regions of non-necrotic tumor and distant normal tissue from each patient were used to construct the TMA. Automated quantitative immunofluorescence (AQUA) was used to measure S100A8 protein expression, and samples were scored for breast cancer cell and stromal S100A8 expression. S100A8 staining intensity was assessed as a continuous value and by exploratory dichotomous cutoffs. Associations between breast cancer cell and stromal S100A8 expression with disease-free survival and overall survival were determined using the Kaplan-Meier method and Cox proportional hazard models.

RESULTS

High breast cancer cell S100A8 protein expression (as indicated by AQUA scores), as a continuous measure, was a significant prognostic factor for OS [univariable hazard ratio (HR) 1.24, 95% confidence interval (CI) 1.00-1.55, p = 0.05] in this patient cohort. Exploratory analyses identified optimal S100A8 AQUA score cutoffs within the breast cancer cell and stromal compartments that significantly separated survival curves for the complete cohort. Elevated breast cancer cell and stromal S100A8 expression, indicated by higher S100A8 AQUA scores, significantly associates with poorer breast cancer outcomes, regardless of estrogen receptor status.

CONCLUSIONS

Elevated breast cancer cell and stromal S1008 protein expression are significant indicators of poorer outcomes in early stage breast cancer patients. Evaluation of S100A8 protein expression may provide additional prognostic information beyond traditional breast cancer prognostic biomarkers.

Abstract P4-12-13: High intratumoral and stromal S100A8 expression is prognostic of poor outcome in breast cancer

Background: S100A8 and S100A9 are members of a family of calcium binding proteins that regulate inflammatory response, and are biomarkers of inflammatory diseases, S100A8/A9 preferentially form heterodimers that interact with their receptor, RAGE, to activate signaling pathways (ERK1/2 MAPK, JNK, and NF-κB) and stimulate tumor cells. Elevated expression of S100A8/A9 has been observed in cancers of the bladder, esophagus, colon, ovary, and breast. S100A8/A9 are expressed intratumorally by cancer cells and in the stroma by infiltrating immune and myeloid cells as well. We investigated the associations of elevated expression of intratumoral and stromal S100A8 with survival outcomes in breast cancer.

Methods: Tissue microarrays (TMA) were constructed from breast cancer specimens from patients with stage I-III breast cancer treated at the University of Michigan Comprehensive Cancer Center between 2004-2006, ensuring a minimum of 10-year follow-up. Each patient was represented on the TMA by representative regions of non-necrotic tumor and distant normal tissue. Automative Quantitative Immunofluorescence (AQUA) was performed for S100A8 protein, and samples were scored for intratumoral and stromal S100A8 expression. S100A8 staining was assessed as a continuous value and by exploratory dichotomous cutoffs. Associations with disease-free survival (DFS) or overall survival (OS) and S100A8 expression, either as continuous value or based on the exploratory cutoffs, were determined using the Kaplan-Meier method and Cox proportional hazards models.

Results: In the entire patient cohort, high intratumoral S100A8 expression, as a continuous measure, was a significant prognostic factor for OS (univariable hazard ratio [HR] 1.26, 95% confidence interval [CI] 1.02-1.56, p=0.036), and for DFS (multivariable HR [95%CI] = 1.24 [1.01-1.53], p = 0.043). Exploratory analyses demonstrated optimal cutoffs of intratumoral and intrastromal staining that greatly separated survival curves. We evaluated whether the prognostic significance of S100A8 expression is different in breast cancer patients based on hormone receptor status and determined that neither intratumoral nor stromal S100A8 expression were significantly associated with outcomes.

Conclusions: Elevated intratumoral and stromal expression of S100A8 are significant indicators of poor outcome in breast cancer patients. These data further support a biological role for S100A8 signaling in mammary carcinogenesis and aggressive tumor behavior. Evaluation of S100A8 protein expression might provide additional prognostic information beyond traditional breast cancer prognostic biomarkers. Further validation is necessary to investigate these findings.

Citation Format: Miller P, Kidwell K, Thomas D, Sabel M, Rae J, Hayes DF, Lippman ME, El-Ashry D. High intratumoral and stromal S100A8 expression is prognostic of poor outcome in 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 P4-12-13.

Abstract P1-03-01: An international multicenter study to evaluate reproducibility of automated scoring methods for assessment of Ki67 in breast cancer

Background: The nuclear proliferation biomarker Ki67 has multiple potential roles in breast cancer, including prognosis-based decisions, but unacceptable between-laboratory variability has limited its clinical value. The International Ki67 Working Group (IKWG) has undertaken a systematic program to determine whether Ki67 immunohistochemistry can be analytically validated and standardized across laboratories. Technological advances and broader availability of devices for automated assessment of stained slides raise the possibility that these machines may improve on reproducibility of traditional pathologist-based visual Ki67 assessment.

Aims: To characterize reproducibility of automated machine-measured Ki67 expression using slides previously analyzed in the IKWG phase 3 study that evaluated reproducibility of visual Ki67 assessment.

Methods: Two sets of 30 previously stained slides containing core-cut biopsy sections of breast tumors were circulated to 14 laboratories for scanning and automated assessment of Ki67 expression. Sites were instructed to return average and maximum percentage of tumor cells positive for Ki67 for each slide, where maximum is designed to reflect “hot spot” analysis. Two laboratories returned scores from 2 operators; not all laboratories reported values for maximum Ki67 scores. Different operators were treated as distinct laboratories in analyses. Sixteen and 10 score sets were available for average and maximum Ki67 analyses, respectively, encompassing 7 unique scanner and 10 software platforms. Pre-specified analyses included evaluation of reproducibility across all laboratories as well as within a subgroup limited to those using Aperio scanners. The primary reproducibility metric was intraclass correlation coefficient between laboratories (ICC), regardless of device platform or software.

Results: Geometric means across 30 cases for 16 operators ranged from 11.06% to 38.11% with overall mean 16.75% (95% CI:14.45-19.42) for average scores. Geometric means for 10 operators ranged from 16.44% to 68.73% with overall mean 25.16% (95% CI: 18.71-33.84) for maximum scores. ICC for automated average scores across 16 operators was 0.83 (95% CI: 0.73-0.91) and ICC for maximum scores across 10 operators was 0.63 (95% CI: 0.44-0.80) although one outlier lab dramatically affected results. For the laboratories using the Aperio platform (8 score sets), ICC for automated average scores was 0.89 (95% CI; 0.81-0.96). These results are similar to ICC of 0.87 (95%CI; 0.81-0.93) reported using these same slides in the Phase 3 visual assessment reproducibility study in which observers counted 500 cells per slide (Leung et al, NPJBrCancer, in press).

Conclusions: Between-laboratory reproducibility for automated machine assessment of average Ki67 is similar to that for pathologist-based visual assessment of Ki67. However, the observed ICC was markedly numerically lower for the maximum score method compared to the average method, suggesting that the maximum score may not be useful as a reproducible measure of proliferation. Automated average scoring methods show promise for standardization of Ki67 scoring, supporting future studies to clinically validate Ki67.

Citation Format: Rimm DL, McShane LM, Leung SCY, Bai Y, Bane AL, Bartlett JMS, Bayani J, Chang MC, Dean M, Denkert C, Enwere E, Galderisi C, Gholap A, Hugh JC, Jadhav A, Kornaga E, Laurinavicius A, Levenson R, Lima J, Miller K, Pantanowitz L, Piper T, Ruan J, Srinivasan M, Virk S, Wu Y, Yang H, Hayes DF, Nielsen TO, Dowsett M. An international multicenter study to evaluate reproducibility of automated scoring methods for assessment of Ki67 in 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 P1-03-01.

Abstract P6-12-05: Targeting estrogen receptor mutations for treatment of endocrine therapy resistance in breast cancer

This abstract was withdrawn by the authors.

Abstract PD7-07: Discovery of molecular predictors of late breast cancer specific events (BCSE) in ER+, node+ breast cancer – new transcriptome expression whole gene analysis of the phase III adjuvant trial SWOG S8814

BACKGROUND: Unique genes and pathways were identified for prognosis on tamoxifen (T, 5 yrs) and prediction on CAF-T vs T in S8814 using whole transcriptome RNA-Seq from archival FFPE tissue. (Albain, et al; Cherbavaz, et al; SABCS 2015) Discovery was robust for early DFS events but sparse for late. The aims of this new analysis were to 1) utilize a new endpoint BCSE for gene discovery of late events, prognosis and prediction and 2) add intronic counts to the previous exonic results to define whole genes impacting on late BCSE.

METHODS: Charts of patients (pts) on CAF-T (212) vs T (142) were reviewed to define the BCSE endpoint (local/regional, contralateral, distant). Deaths without BC were treated as competing risks. BCSE models (including metagenes) of late prognosis and prediction used cumulative incidence functions. Consolidated intronic regions counts within genes were added to exonic regions counts. Using these “whole gene” (WG) counts, association of gene expression with time to BCSE was assessed by Cox regression. A multiple WG score (MWGS) for BCSE prognosis beyond 5 yrs (to 12.5 yrs) was constructed and evaluated for 1-3 and 4+ node (N) groups. False discovery rate was controlled at 10%.

RESULTS: More exons and WG were discovered for prognosis on T alone over 12.5 yrs with the BCSE endpoint than DFS. For prognosis of late BCSE after 5 yrs, more genes were discovered using WG (n=111) than by exons (n=9). There were significantly fewer genes for late BCSE on CAF-T (8, WG; 0, exons). The functions of WG prognostic for late BCSE were: cell cycle/proliferation-26 genes, chromosome segregation/mitotic spindle-22, DNA repair/maintenance-10, transcriptional/translational control-5, cell adhesion/migration-4, immune-3, diverse/unknown-32 and growth factor/hormone receptor signaling-9 (this group was only found by WGs, not exons). Of these 111 WG, a MWGS prognostic for late BCSE on T used 57 previously discovered genes pre-specified for this analysis. Probability of BCSE beyond 5 yrs for low vs high MWGS was 8% vs 21% in N1-3+ and 17% vs 42% in N4+. Late prognosis on T differed by low vs high risk defined in a metagene model: cumulative BCSE at year 10 was 0% vs 47% (low vs high risk, p=0.001). Prediction of 10-yr incidence of BCSE varied by risk level by treatment in a metagene model: low risk- CAF-T=47%, T=0% (p=0.045); high risk- CAF-T=35%, T=45% (p=0.027).

CONCLUSIONS: Gene discovery for prognosis of late BCSE is enhanced with a novel WG transcriptome expression approach. Use of chemotherapy (CT) before T significantly attenuated gene discovery, so that molecular tools for decisions on extending endocrine therapy (ET) may not be reliable in a setting of prior CT. Some pts on ET for 5 yrs may not require either longer ET or CT, given a N+ cohort was defined with no BCSE observed over 12.5 yrs. For prediction of CT benefit, CAF-T appeared to be inferior to T in a low risk metagene model for BCSE. In sum, these results add more evidence that ET alone may be sufficient (perhaps better) in select N+ settings. Validation in SWOG S1007 (RxPONDER) is planned.

SUPPORT: NCI CA180888, 180819, 180821, 180820, 180863; in part, Genomic Health, Inc.

Citation Format: Albain KS, Crager MR, Barlow WE, Baehner FL, Bergamaschi A, Rae JM, Ravdin PM, Tripathy D, Gralow JR, Livingston RB, Osborne CK, Ingle JN, Pritchard KI, Davidson NE, Carey LA, Cherbavaz DB, Sing AP, Shak S, Hortobagyi GN, Hayes DF. Discovery of molecular predictors of late breast cancer specific events (BCSE) in ER+, node+ breast cancer – new transcriptome expression whole gene analysis of the phase III adjuvant trial SWOG S8814 [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 PD7-07.

Abstract P1-01-01: Circulating tumor cell number and CTC-endocrine therapy index predict clinical outcomes in ER positive metastatic breast cancer patients: Results of the COMETI Phase 2 trial

Introduction: Only half of hormone receptor positive (HR+) metastatic breast cancer (MBC) patients (pts) benefit from endocrine therapy (ET). Circulating tumor cells (CTC) are prognostic in pts with MBC using CellSearch® technology. The CTC-endocrine therapy index (CTC-ETI) provides semi-quantitative analyses of CTC-ER (estrogen receptor), BCL2, HER2, and Ki67 expression. We hypothesized that CTC-ETI high (elevated CTC number and/or low expression of ER and BCL2, and high expression of HER2 and Ki-67) might predict resistance to ET in a prospective, multi-institutional clinical trial: COMETI-P2-2012.0 (NCT01701050).

Methods: 121 pts with ER+, HER2 negative (-), and progressive MBC after one or more lines of ET or within 12 months (mos) of completing adjuvant ET, who were initiating a new ET, were enrolled after informed consent. CTC and CTC-ETI were determined as previously reported (Paoletti C et al, CCR 2015) at baseline (BL), 1, 2, 3, and 12 mos, and/or at the time of progression. Imaging was performed every 3 mos. Association of CTC levels and CTC-ETI with patient outcomes (progression free survival (PFS); rapid progression (RP) defined as progression within 3 mos) was assessed using logrank and Fisher's exact tests. Trial design estimated 85 PFS and 51 RP events, providing >90% power (2-sided a=0.05); pts with unsuccessful BL CTC-ETI or ineligible were unevaluable. Only baseline (BL) data are reported in this abstract.

Results: 32% of enrolled pts had progression within 12 mos of completing adjuvant ET, whereas 40%, 20%, and 8% had 1, 2, ≥3 lines of ET for MBC. CTC-ETI was successfully determined in 93% of pts (90% CI, 88% to 97%). CTC were ≥5 CTC/7.5 ml whole blood in 37/108 (34%) pts evaluable for clinical validity. Elevated CTC was associated with worse PFS (median (m) PFS: 3.3 vs. 5.9 mos; P<0.01). Low, intermediate, and high CTC-ETI were observed in 75 (69%), 6 (6%), and 27 (25%) pts, respectively. CTC-ETI was associated with PFS (logrank P<0.01): pts with low, intermediate, and high CTC-ETI had mPFS of 5.7, 8.5, and 2.8 mos, respectively. In the 96 pts eligible for determination, elevated CTC was associated with RP, (65.6% vs. 42.2%; P=0.05) as was CTC-ETI (P=0.003): 79.2% (95% CI, 57.8% to 92.9%) of pts with high CTC-ETI had RP versus 41.2% (95% CI, 29.4% to 53.8%) with low CTC-ETI; in the small group with intermediate CTC-ETI 1 of 4 pts (25%) had RP.

Conclusions: In this multi-institutional, prospective study, CTC-ETI was accurately determined, confirming the previously established analytical validity of the assay, meeting the primary objective of the trial. Elevated CTC and CTC-ETI high compared to low were associated with poor outcomes to ET. CTC-ETI distribution resulted in a small number of patients assigned to the intermediate group, restricting our ability to associate this group with outcomes. These results suggest that CTC-biomarker phenotype and enumeration have clinical validity. CTC-ETI may identify ER+ HER2– MBC pts who are unlikely to benefit from ET and might be better treated with ET in combination with other therapies or proceed to chemotherapy. Further analyses including CTC-ETI at serial time points during ET are planned.

Citation Format: Paoletti C, Regan MM, Liu MC, Marcom PK, Hart LL, Smith II JW, Tedesco KL, Amir E, Krop IE, DeMichele AM, Goodwin PJ, Block M, Aung K, Cannell EM, Darga EP, Baratta PJ, Brown ME, McCormack RT, Hayes DF. Circulating tumor cell number and CTC-endocrine therapy index predict clinical outcomes in ER positive metastatic breast cancer patients: Results of the COMETI Phase 2 trial [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 P1-01-01.

Abstract P6-09-11: Genetic variation in CYP3A affects steady-state exemestane concentrations but does not explain inter-race difference

Background: Exemestane is a third generation steroidal aromatase inhibitor (AI) used for the treatment of estrogen receptor (ER) positive breast cancer in postmenopausal women. Differences in AI treatment efficacy and side effects may be due, in part, to variability in drug exposure. We previously reported that patients who self-report as white and those who carry the low-activity CYP3A4*22 single nucleotide polymorphism (SNP) have increased exemestane steady-state concentrations. Additional SNPs in CYP3A may contribute to pharmacokinetic variability and explain this inter-race difference. CYP3A5*3 (rs776746) is a non-expresser genotype that is far more common in European (minor allele frequency (MAF)∼0.94) than African (MAF∼0.18) individuals. CYP3A7*1C (rs45446698) is believed to tag adult expression of the fetal CYP3A7 enzyme and is relatively uncommon in tested cohorts (European MAF=0.04, African MAF<0.01). The objective of this secondary analysis was to determine whether these additional CYP3A SNPs contribute to variability in steady state exemestane concentrations and explain the inter-race difference.

Methods: 500 patients were randomly assigned to either drug on the Exemestane and Letrozole Pharmacogenetics (ELPh) Study. Clinical data and DNA were collected at baseline and blood samples were collected after 1 or 3 months of treatment to measure steady-state exemestane concentration via HPLC/MS. Genotyping for CYP3A5*3 and CYP3A7*1C was performed via Taqman Allelic Discrimination. Pharmacogenetic association with log-transformed concentrations were tested for each variant by inclusion in a multivariable model with CYP3A4*22 and self-reported race, assuming additive genetic effect, using Tobit regression to censor concentrations below the lower limit of quantification. SNPs with suggestive p-values <0.10 were included in a multivariable model with relevant covariates (AST or ALT>40, body mass index (BMI), and prior chemotherapy) to assess their independent contribution.

Results: In 231 evaluable patients there was a suggestive trend toward lower steady-state exemestane concentrations for CYP3A7*1C carriers (6.3 vs. 8.0 ng/mL) in the model including CYP3A4*22 and race (p=0.083). In the final multivariable model each CYP3A7*1C allele decreased exemestane concentration 31.5% (p=0.035, Table 1). CYP3A5*3 was not associated with exemestane concentration (p>0.2).

Multivariable Model of Exemestane Concentration % change in concentration (95% CI)p-valueCYP3A4*22 (rs35599367)64.5% (23%, 120%)0.0008CYP3A7*1C (rs45446698)-31.5% (-52%, -2.6%)0.035Self-Reported White47.2% (9.0%, 99%)0.012AST or ALT>4041.3% (1.0%, 98%)0.044BMI-0.9% (-2.4%, 0.55%)0.22Prior Chemotherapy-23.5% (-37%, -7.6%)0.006CI: Confidence Interval

Conclusions: Patients with breast cancer who carry CYP3A7*1C have lower steady-state exemestane concentrations but this association does not explain the greater concentrations in self-reported white patients. Ongoing analyses will determine whether exemestane concentration predicts treatment efficacy or toxicity, and if so, whether genetic and clinical factors can be useful for individualizing dosing to optimize outcomes. CYP3A7*1C should be prioritized for analyses of pharmacokinetic variability of other CYP3A substrates.

Citation Format: Hertz DL, Kidwell KM, Gersch CL, Desta Z, Storniolo AM, Stearns V, Skaar TC, Hayes DF, Henry NL, Rae JM. Genetic variation in CYP3A affects steady-state exemestane concentrations but does not explain inter-race difference [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 P6-09-11.

Epithelial Cell Adhesion Molecule Expression in Canine Tumours

Epithelial cell adhesion molecule (EpCAM) is expressed in most human normal and neoplastic tissues of epithelial derivation and may have an association with tumour cell aggressiveness, a stem cell-like phenotype and clinical outcome. Antibody-based strategies for the targeting and capture of EpCAM-expressing tumour cells are showing promise, both as diagnostic tools and potential therapies. The aim of this study was to assess EpCAM expression in canine tumours. EpCAM expression was assessed in tumour cell lines via gene expression profiling and in formalin-fixed and paraffin wax-embedded tissues from canine carcinomas representing various anatomical sites by immunohistochemistry. EpCAM mRNA expression was higher in cell lines from carcinomas than those derived from sarcomas or haemopoietic tumours. EpCAM was expressed by >2/3 of tumour cells in 71% of canine carcinomas evaluated, irrespective of histotype, with the exception of carcinomas of the adrenal gland. Canine sarcomas and haemopoietic tumours were uniformly negative. Most canine carcinomas express EpCAM and so could be suitable for the study of EpCAM-directed diagnostics and therapeutics.

SWOG S0800 (NCI CDR0000636131): addition of bevacizumab to neoadjuvant nab-paclitaxel with dose-dense doxorubicin and cyclophosphamide improves pathologic complete response (pCR) rates in inflammatory or locally advanced breast cancer

SWOG S0800, a randomized open-label Phase II clinical trial, compared the combination of weekly nab-paclitaxel and bevacizumab followed by dose-dense doxorubicin and cyclophosphamide (AC) with nab-paclitaxel followed or preceded by AC as neoadjuvant treatment for HER2-negative locally advanced breast cancer (LABC) or inflammatory breast cancer (IBC). Patients were randomly allocated (2:1:1) to three neoadjuvant chemotherapy arms: (1) nab-paclitaxel with concurrent bevacizumab followed by AC; (2) nab-paclitaxel followed by AC; or (3) AC followed by nab-paclitaxel. The primary endpoint was pathologic complete response (pCR) with stratification by disease type (non-IBC LABC vs. IBC) and hormone receptor status (positive vs. negative). Overall survival (OS), event-free survival (EFS), and toxicity were secondary endpoints. Analyses were intent-to-treat comparing bevacizumab to the combined control arms. A total of 215 patients were accrued including 11 % with IBC and 32 % with triple-negative breast cancer (TNBC). The addition of bevacizumab significantly increased the pCR rate overall (36 vs. 21 %; p = 0.019) and in TNBC (59 vs. 29 %; p = 0.014), but not in hormone receptor-positive disease (24 vs. 18 %; p = 0.41). Sequence of administration of nab-paclitaxel and AC did not affect the pCR rate. While no significant differences in OS or EFS were seen, a trend favored the addition of bevacizumab for EFS (p = 0.06) in TNBC. Overall, Grade 3-4 adverse events did not differ substantially by treatment arm. The addition of bevacizumab to nab-paclitaxel prior to dose-dense AC neoadjuvant chemotherapy significantly improved the pCR rate compared to chemotherapy alone in patients with triple-negative LABC/IBC and was accompanied by a trend for improved EFS. This suggests reconsideration of the role of bevacizumab in high-risk triple-negative locally advanced breast cancer.

Abstract P5-12-05: CYP3A4*22 polymorphism is associated with increased exemestane concentrations in postmenopausal breast cancer patients

Background: Exemestane is a second generation steroidal aromatase inhibitor (AI) used for the treatment of estrogen receptor (ER) positive breast cancer in postmenopausal women. Variability in AI treatment efficacy and side effects seen across patients may be due, in part, to inter-patient differences in drug exposure. This exposure variability is likely caused by patient genetics factors, such as single nucleotide polymorphisms (SNPs) in drug metabolizing enzymes, or clinical factors such as patient body size, organ function, and comorbidities. The objective of this secondary correlative analysis was to identify genetic and clinical characteristics that affect steady state exemestane concentration, with a specific focus on the influence of inherited genetic variants and baseline hepatic function.

Methods: 500 patients were enrolled on the Exemestane and Letrozole Pharmacogenetics (ELPh) Study and randomized to either drug. Clinical data and DNA were collected at baseline and blood samples were collected after 1 or 3 months of treatment to measure steady-state exemestane concentration via HPLC/MS. Genotyping was performed on a custom Sequenom MassARRAY iPLEX that included the recently discovered low activity CYP3A4*22 (rs35599367) SNP and several other SNPs with putative functional consequence in enzymes thought to be involved in exemestane metabolism (CYP1A1/2, CYP1B1, CYP3A4, CYP4A11, AKR1C3/4, AKR7A2). Our primary hypothesis was that patients carrying CYP3A4*22 variants would have higher serum exemestane concentrations. Other SNPs and clinical characteristics (hepatic and renal function, age, body mass index (BMI), time of sample collection, prior chemotherapy) were assessed for independent association, and then adjusted for in a multivariable tobit regression model for CYP3A4*22 on log-transformed censored exemestane concentration.

Results: 246 (225 randomized to exemestane arm, 21 crossed-over from letrozole arm) patients had exemestane steady state levels and were evaluable in this analysis. As hypothesized, the CYP3A4*22 polymorphism (minor allele frequency=0.06) was associated with a 54% increase in exemestane concentration (95% CI: 14% - 109%, p<0.01). Exemestane concentration was 44% greater in patients who had evidence of hepatic impairment (AST or ALT>40) at baseline (95% CI: 2% - 104%, p=0.02), 1% lower per unit increase in BMI (95% CI: 0% - 3%, p=0.05), and 20% lower in patients who received prior chemotherapy (95% CI: 4% - 34%, p=0.03). Age, renal impairment, and other SNPs were not associated with exemestane concentration. After adjustment for significant clinical covariates the CYP3A4*22 SNP remained significant (p<0.01).

Conclusions: Genetic and clinical predictors of exemestane concentration were discovered in a large cohort of prospectively enrolled estrogen responsive breast cancer patients. Ongoing analyses will determine whether the variability in exemestane concentration was associated with downstream effects on estrogen depletion or treatment-related toxicity. If so, these genetic and clinical characteristics could be useful for individualizing dosing of exemestane to ensure that all patients are receiving maximal benefit with minimal toxicity.

Citation Format: Hertz DL, Kidwell KM, Seewald NJ, Gersch CL, Desta Z, Flockhart DA, Storniolo AM, Stearns V, Skaar TC, Hayes DF, Henry NL, Rae JM. CYP3A4*22 polymorphism is associated with increased exemestane concentrations in postmenopausal breast cancer patients. [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 P5-12-05.

Abstract S3-02: Molecular predictors of outcome on adjuvant CAF plus tamoxifen (T) vs T in postmenopausal patients (pts) with ER+, node+ breast cancer – Transcriptome expression analysis of the phase III trial SWOG-8814

BACKGROUND: In SWOG-8814A, pts with ER+ node+ breast cancer and low 21 gene recurrence scores (RS) had good prognosis and no CAF benefit, but high RS predicted longer survival from CAF followed by T (CAF-T) vs T (Albain, Lancet Oncol 2010). The aims of SWOG-8814B were to identify novel genes and networks for 1) prognosis of early and late relapse and 2) prediction of CAF benefit, using whole transcriptome expression analysis with next generation RNA sequencing (NGS).

METHODS: Stored RNA previously extracted for SWOG-8814A (T, CAF-T arms; T, 5 yrs) was analyzed for RNA/library yield (see companion abstract Cherbavaz et al. for methods). Genes were sequenced and expression of mRNA species was related to disease-free survival (DFS) using Cox proportional hazards. Discovery analyses controlled false discovery rate (FDR) at 10%. Genes were identified for prognosis on T and prediction on CAF-T vs T. Networks of genes/pathways were explored. Early (0-5 yrs) and late (5-13+ yrs) time periods were studied. Gene Ontology, Cytoscape, pathway and hierarchical clustering were used for functional gene and metagene analyses.

RESULTS: Of 367 samples, 354 (96%; 142 T, 212 CAF-T; 141 DFS events) had sufficient RNA/library yield, with 20,101 genes sequenced. For prognosis on T, there were 2327 and 568 genes discovered in early and all-yrs follow-up, with only 9 genes prognostic after 5 yrs. Prognosis analyses for residual risk after CAF-T were uninformative. Functional mapping found that genes prognostic for worse DFS were enriched for proliferation (G2M, M-phase), cellular metabolism, DNA repair, stress response and EMT; whereas, those with better DFS involved transcription regulation/repression via zinc finger proteins. Hierarchical clustering (T arm) found significant DFS prognostic metagene signatures for ER-related genes, immune response, ECM/stroma, chromatin remodeling-transcription factor activity and TGFb pathway. All varied for early vs late DFS events. For example, low ER/high stroma expression signatures correlated with high proliferation gene expression and were strongly associated with early events (standardized [st] HR 2.94, p<0.001). Late recurrence was associated with high proliferation, both individually (stHR 1.51, p=.035) and in combination with higher ER expression (stHR 1.51, p=0.09). Fifteen genes predicted CAF benefit (9 better DFS, 6 worse), or 129 genes if FDR relaxed to 20%. Cluster analysis for CAF prediction is ongoing.

CONCLUSIONS: Unique genes, clusters and pathways were identified by NGS of archival material in ER+ N+ breast cancer, including previously unreported signatures. While ER, stroma and proliferation-related signatures were associated with early prognosis, proliferation best predicted worse DFS after 5 yrs. NGS of the primary tumor is most informative for early events in pts with only 5 years of T, with few genes selecting only for late relapse. If validated, these signatures may identify pts with excellent DFS despite positive nodes for endocrine therapy alone as well as others for whom chemotherapy and/or biologics are also required

.

SUPPORT: NCI CA 180888, 180819, 180821, 180820, 180863; in part, Genomic Health, Inc.

Citation Format: Albain KS, Crager MR, Barlow WE, Baehner FL, Bergamaschi A, Rae JM, Ravdin PM, Tripathy D, Gralow JR, Livingston RB, Osborne CK, Ingle JN, Pritchard KI, Davidson NE, Carey LA, Cherbavaz DB, Sing AP, Shak S, Hortobagyi GN, Hayes DF. Molecular predictors of outcome on adjuvant CAF plus tamoxifen (T) vs T in postmenopausal patients (pts) with ER+, node+ breast cancer – Transcriptome expression analysis of the phase III trial SWOG-8814. [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 S3-02.

Abstract ES5-1: Predictive gene expression profiles in breast cancer, where are we?

Successful sequencing of the entire human genome offered an enormous potential for improving human health, especially in oncology. Nearly 20 years after successful completion of the human genome project, the Cancer Genome Atlas (TCGA) has catalogued the most common mutations in most cancers, including breast. Further, the technical advances that facilitate the ability to fully analyze both germline and somatic DNA and RNA has resulted in rapid (<4 weeks) and relatively inexpensive (<$3000) assays to emerge. Indeed, the NCI-sponsored Lung MAP and Cancer MATCH, and the ASCO-sponsored TAPUR, trials are now opening in the U.S., testing whether knowledge of somatic genetic changes improves patient outcomes.

However, arguably, the most clinically useful product of the Human Genome Project to date has been the emergence of multiparameter gene expression assays that can be used to direct cancer care, particularly in breast cancer. The 21-gene assay was the first of these to be shown to have sufficiently high analytical validity and clinical utility to be used to identify patients with node negative, ER positive early stage breast cancer who, assuming adequate adjuvant endocrine therapy, could be spared the toxicity of adjuvant chemotherapy based on prognosis alone. Subsequently, several other assays have now been reported to be equally satisfactory for this same use context. The Oxford Overview has suggested that adjuvant chemotherapy reduces the odds of distant recurrence and death by approximately 1/3, regardless of initial prognosis. However, many retrospective correlative studies over the last 4 decades have suggested that chemotherapy may have differing relative activities within distinct biological, or more recently designated “intrinsic,” subtypes, based mostly on expression of estrogen receptor and proliferation genes. Preliminary, prospective retrospective studies by the NSABP in B20 and SWOG in S8814 have, indeed, suggested that patients with ER positive breast cancer with low recurrence scores, as determined by the 21-gene assay, not only have better prognosis than those with intermediate or higher RS, but that chemotherapy may simply not work in this category. However, these studies were both confounded by a number of biases. Therefore, before potentially life-saving chemotherapy is with-held from node positive breast cancer patients based on prediction alone, a prospective clinical trial, SWOG S1007 (the RxPonder Trial), was initiated. The primary objective of the RxPonder trial is to determine if women with node positive, ER positive breast cancer with RS <25 may not benefit from adjuvant chemotherapy even if their prognosis is worse than those with node negative disease. This trial, which is now fully accrued, will not only tell us if we can spare yet another group of women from receiving unnecessary or inactive chemotherapy, it also offers the promise of identification of targets that could be further interrogated with newer biological treatments in the future.

Citation Format: Hayes DF. Predictive gene expression profiles in breast cancer, where are we?. [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 ES5-1.

Abstract S4-05: Interrogating the landscape of long noncoding RNAs in breast cancer to identify predictors of tamoxifen resistance

Background: We previously performed an informatics-based analysis on RNA sequencing libraries from 7,256 tumor and normal tissue specimens to delineate the landscape of long noncoding RNAs (lncRNAs) in the human transcriptome. This analysis identified 58,648 lncRNAs, including over 45,000 novel transcripts (Iyer MK et al, Nature Genetics, 2015). We now interrogate this lncRNA compendium to identify top candidate estrogen receptor (ER)-associated lncRNAs in breast cancer and characterize their association with disease progression.

Methods: To prioritize differentially expressed lncRNAs in cancer vs normal tissue, and in ER+ vs ER- disease, we performed Sample Set Enrichment Analysis (SSEA) on >1000 RNA Seq libraries, from breast cancer and normal tissue samples from The Cancer Genome Atlas project. The effect of the top prioritized lncRNA on cancer phenotypes was studied via in vitro proliferation, colony formation, invasion and tamoxifen resistance assays in MCF7 and T47D cells, and via in vivo mouse xenograft studies and chick chorioallantoic membrane (CAM) assays. To study the mechanism by which this lncRNA promotes tumor progression, we identified its top protein interactors and subdomains responsible for function, and then studied the effects of disrupting function of this lncRNA on cancer phenotypes. Finally, in a "guilt-by-association" study, we developed a signature of 150 protein coding genes most strongly associated with our lncRNA of interest, and investigated the association of this signature with clinical outcomes using Oncomine analyses.

Results: SSEA analysis on over 1000 TCGA samples nominated Breast Cancer Associated Transcript (BRCAT 431) as the top overexpressed ER-regulated lncRNA in breast cancer. In vitro experiments demonstrate that siRNA-mediated knockdown of BRCAT431 resulted in significantly decreased proliferation, colony formation, and invasion (by >50% in most assays). Tamoxifen resistance was associated with significantly increased BRCAT431 levels in both MCF7 and T47D cells, and knockdown of BRCAT431 reversed tamoxifen resistance. In vivo xenograft and CAM studies demonstrate that knockdown of BRCAT431 also significantly decreased xenograft growth and tumor invasion by >50%. RNA pulldown followed by mass spectrometry identified the RNA binding protein hnRNPL as a key protein interacting with BRCAT431. Deletion studies identified a 27 base region of BRCAT431 necessary for its interaction with hnRNPL, and loss of this region abrogated BRCAT431- induced invasion. Finally, guilt-by-association studies demonstrate a strong association between BRCAT431 overexpression and tumor grade, recurrence, and metastases.

Conclusion: In this study, we develop the largest reported compendia of breast cancer lncRNAs. We prioritize BRCAT431 as the top lncRNA upregulated in ER-positive breast cancers, and demonstrate that it confers aggressive oncogenic phenotypes in vitro and in vivo. We identify a novel mechanism by which this lncRNA functions. Our results suggest that by promoting tamoxifen resistance, BRCAT431 increases the clinical risk of recurrence and metastases in breast cancer. Overall, this study supports the rationale for investigating lncRNAs as novel biomarkers and therapeutic targets in breast cancer.

Citation Format: Feng FY, Niknafs Y, Han S, Ma T, Speers C, Malik R, Evans J, Zhang C, Pierce LJ, Hayes DF, Rae JM, Chinnaiyan AM. Interrogating the landscape of long noncoding RNAs in breast cancer to identify predictors of tamoxifen resistance. [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 S4-05.

Abstract P2-02-19: Somatic genetic profiling of circulating tumor cells (CTC) in metastatic breast cancer (MBC) patients

Introduction: Somatic mutations, including those in TP53, PIK3CA, and estrogen receptor alpha (ESR1), are key to the biology of cancer and response to therapy. Recently, somatic cancer-associated mutations have been identified in circulating cell free plasma tumor DNA (ptDNA). Less is known about the mutation profile of DNA extracted from CTC (CTC-DNA). Since CTC-DNA provides mutational information of single cells, we hypothesize CTC-DNA will complement ptDNA to give greater insight into tumor heterogeneity.

Methods: Patients with ER positive MBC who were enrolled in the Mi CTC-ONCOSEQ, a companion trial to Mi-ONCOSEQ (the Michigan Oncology Sequencing Program), and who had ≥5CTC/7.5 ml whole blood were included. CTC were enriched from white blood cells (WBC) with CellSearch© (CXC kit). CTC and WBC were then purified using DEPArrayTM. DNA from individual CTC and WBC was isolated and subjected to whole genomic amplification (Ampli 1TM WGA). Genetic analysis was performed on individual CTC, pooled CTC and pooled WBC DNA by multiplexed PCR based targeted next generation sequencing (NGS) using the Oncomine Comprehensive Panel (targeting ∼130 onco- and tumor suppressor genes) and the Ion Torrent Proton. All patients had exome sequencing performed on research biopsies of metastases using an Illumina HiSeq 2500 platform.

Results: This pilot study was conducted using high quality DNA from two patients assessed to date. Both patients had lobular carcinoma and as expected harbored somatic, deleterious CDH1 (E-cadherin) mutations (frameshift and non-sense) in both research biopsy and CTC-DNA. These data supported our approach. Patient #1 was TP53 wild type in her research biopsy, but multiple CTC harbored somatic TP53 frame-shift mutations (Table). Patient #2 harbored an ESR1 Y537S mutation in her research biopsy. However, only 4 of 7 CTC also harbored this somatic, heterozygous mutation.

Prioritized mutations in CTCPt#Cell Type (CTC vs WBC), numberGeneMutationVariant fraction (expected 1=homozygous; 0.5=heterozygous)Found in research biopsy?1CTC_A2CDH1p.I584fs1YES CTC_A4  1  CTC_A7  0.54  CTC_pool*  0.74  WBC_pool  0  CTC_A2TP53p.152_156del1NO CTC_A4  1  CTC_A7  0.51  CTC_pool*  0.88  WBC_pool  0 2CTC_A9ESR1p.Y537S0.52YES CTC_D1  0.34  CTC_D2  0.46  CTC_D6  0.65  CTC_pool*  0.35  WBC_pool  0  CTC_A12  0  CTC_D3  0  CTC_D7  0  CTC_A12CDH1p.Q641X1YES CTC_A9  1  CTC_D1  1  CTC_D3  1  CTC_D6  1  CTC_pool*  1  WBC_pool  0 * pool of all CTC

Conclusions: We demonstrate the ability to purify CTC, isolate, and amplify DNA of suitable quality for genetic analysis using a comprehensive targeted sequencing panel. Both known and novel alterations were identified in comparison to research biopsy specimens. This approach allows single cell analysis demonstrating heterogeneity of mutational status in different single cells. Studies of CTC-ESR1 and other genetic abnormalities in patients with known tissue mutations who participated in Mi CTC-ONCOSEQ are now underway.

Citation Format: Paoletti C, Cani AK, Aung K, Darga EP, Cannell EM, Hovelson DH, Yazdani M, Blevins AR, Tokudome N, Larios JM, Thomas DG, Brown ME, Gersch C, Schott AF, Robinson DR, Chinnaiyan AM, Bischoff F, Hayes DF, Rae JM, Tomlins SA. Somatic genetic profiling of circulating tumor cells (CTC) in metastatic breast cancer (MBC) patients. [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 P2-02-19.

Abstract P3-05-01: Molecular analysis of cancer tissue, circulating tumor cells (CTC) and cell-free plasma tumor DNA (ptDNA) suggests variable mechanisms of resistance to endocrine therapy (ET) in estrogen receptor (ER) positive metastatic breast cancer (MBC)

Introduction: Fifty percent of ER positive MBC patients do not benefit from ET. Potential mechanisms of resistance to ET in this patient population include absence of ER expression by deletion or suppression, alteration in ER signaling pathway genes, or upregulation of multiple growth factor receptor pathways. We hypothesized that genotyping and phenotyping of CTC combined with genomic analysis of ptDNA will provide important insights into the multiple mechanisms of ET resistance and that a set of blood tests might serve as a "liquid biopsy" abrogating the need for tissue specimens.

Methods: Twenty-four patients providing informed consent were enrolled into the Mi CTC-ONCOSEQ study, a companion trial to Mi-ONCOSEQ (the Michigan Oncology Sequencing Program). Seven of these patients (5 with ER immunohistochemistry (IHC) positive and 2 with ER negative cancers) who had available archived primary and metastatic cancer tissue, a research metastatic biopsy for genomic analysis, and who had ≥5CTC/7.5 ml whole blood (WB) characterized for ER protein (CTC-ER) are the focus of this report. All the patients were ET refractory. None of them was progressing on fulvestrant at the time of study entry. CTC enumeration and phenotyping was performed with CellSearch©. Circulating ptDNA was analyzed by droplet digital polymerase chain reaction (ddPCR). ER status from archived tissue was obtained from chart review. ER mRNA expression was determined in the research biopsy of metastatic tissue by using quantitative RNA sequencing. Mutational status of ER gene, ESR1, was determined by Next-gen Sequencing using the Illumina HiSeq 2500 platform.

Results: The 2 control patients with triple negative breast cancer had negative CTC-ER. Discordance between CTC-ER and tissue ER by IHC was observed (Table). Two of the 5 ER positive patients retained CTC-ER positivity (39% and 19% of the CTC). One of them (#7) harbored an ESR1 mutation in the research biopsy tissue and in ptDNA, whereas the other (#14) had wild type (WT) ESR1. CTC-ER protein levels in patients #12, 17 and 24 were negative. All had WT ESR1 in the research biopsy tissue. Of note, patient #12 had WT ESR1 in the research biopsy, but an ESR1 mutation was detected in her ptDNA.

Pt#CTC-ER Tissue-ER  ESR1 status in research biopsyESR1 status in ptDNA N[deg]CTC/7.5ml WB% CTC-ER +Primary by IHCMet by IHCMet research biopsy by mRNA  71839%+++Y537SY537S141619%+NA+WTWT12130%+++WTD538G17160%++weakly+WTWT242750%+weakly+weakly+WTWT

Conclusions: These exploratory data suggest heterogeneous mechanisms of resistance to ET in patients with previously determined ER-positive MBC, including ESR1 mutations in ER positive cases (seen in 2 patients) and loss of ER expression (seen in CTC of 3 patients). In contrast, other cancers continue to express WT ESR1, and therefore must have developed alternative mechanisms of resistance. At least 2 of these mechanisms can be detected and monitored with complementary circulating assays: CTC and ptDNA. Further investigations are needed to understand the heterogeneous mechanisms of resistance to ET.

Citation Format: Paoletti C, Aung K, Cannell EM, Darga EP, Chu D, Kidwell KM, Thomas DG, Tokudome N, Brown ME, McNutt LM, Gersch C, Schott AF, Park BH, Robinson DR, Chinnaiyan AM, Rae JM, Hayes DF. Molecular analysis of cancer tissue, circulating tumor cells (CTC) and cell-free plasma tumor DNA (ptDNA) suggests variable mechanisms of resistance to endocrine therapy (ET) in estrogen receptor (ER) positive metastatic breast cancer (MBC). [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 P3-05-01.

Abstract P3-07-64: Association between gene variants in SULT1A1 and UGT1A4 and disease outcomes in patients enrolled in SWOG S0226 and treated with anastrozole alone or in combination with fulvestrant for metastatic breast cancer

Background: Anastrozole (A) blocks estrogen production by inhibiting the activity of CYP19 aromatase. Fulvestrant (F) blocks estrogen receptor (ER) signaling by competitive binding, leading to ER degradation by ubiquitination. SWOG S0226 ("Phase III Randomized Trial of Anastrozole versus Anastrozole and Fulvestrant (250mg LD) as First Line Therapy for Post Menopausal Women with Metastatic Breast Cancer," ClinicalTrials.gov Identifier:NCT00075764) demonstrated that combination of A+F is superior to A alone as first-line therapy for patients with ER positive metastatic breast cancer (Mehta et al, NEJM, 2012). Our functional preclinical studies have shown that single nucleotide polymorphisms (SNPs) in SULT1A1 and UGT1A4, drug conjugation enzymes that inactivate A and F, result in decreased enzyme activity toward these drugs (Edavana et al, DMD, 2013; Edavana et al Pharmgenomics Pers Med 2013). We therefore hypothesized that these SNPs will be associated with disease outcomes in S0226 patients due to altered drug levels.

Methods: Germline DNA was available for 295 (43.5%) patients enrolled in S0226 overall (157 on A and 138 on A+F). SNPs in SULT1A1 and UGT1A4 were determined either by direct sequencing or allele-specific PCR (TaqMan) assays.

Results: There was no difference in progression-free survival (PFS) or overall survival (OS) comparing patients with or without available germline DNA (p = 0.86 and 0.36, respectively). The SULT1A1 G902A allele (rs6839), which confers decreased mRNA and enzymatic activity, was associated with improved PFS (GG/GA vs. AA; HR 0.74, 95% CI 0.56-0.98, p=0.033) and OS (HR 0.70, 95% 0.50-0.98, p=0.039). In exploratory subset analyses of PFS, the SULT1A1 G902A association was similar across both treatment arms (A HR=0.75; 95% CI 0.51-1.10; A+F HR=0.73; 95% CI 0.48-1.11). For OS there was some evidence of a difference by treatment (A HR=0.60; 95% CI 0.38-0.96; A+F HR=0.82; 95% CI 0.50-1.32), though no significant interaction was evident (p=0.30).

The UGT1A4 G-163A promoter variant, which leads to decreased protein expression, was not associated with PFS (AA/AG vs. GG HR 0.88, 95% CI 0.68-1.14, p=0.33); however, this variant was associated with OS (HR 0.71, 95% CI 0.52-0.96, p=0.027). In subset analyses with OS, the difference was marginally stronger in the A arm (HR 0.63, 95% CI 0.42-0.97, p=0.035) compared to the A+F arm (HR 0.77, 95% CI 0.49-1.21, p=0.25), though the interaction was not significant (p=0.40).

Conclusion: SULT1A1 and UGT1A4 gene variants resulting in decreased enzyme activity were associated with better PFS, OS or both in patients enrolled in SWOG S0226. Planned validation studies correlating these SNPs with drug levels and disease outcomes in additional patient cohorts will establish their clinical utility in identifying patients who benefit from A and F alone or in combination.

Funding: Supported by NIH/NCI CA118981; NIH/NCI/NCTN grants CA180888, CA180819, and CA180863; and in part by AstraZeneca.

Citation Format: Kadlubar SA, Barlow WE, Mehta RS, Daniels JR, Albain KS, Vandengerg TA, Dakhil SR, Tirumali NR, Lew DL, Gralow JR, Livingston RB, Hortobagiyi GN, Hayes DF, Rae JM. Association between gene variants in SULT1A1 and UGT1A4 and disease outcomes in patients enrolled in SWOG S0226 and treated with anastrozole alone or in combination with fulvestrant for metastatic 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 P3-07-64.

Abstract P1-04-01: Significance of circulating tumor cells in metastatic triple negative breast cancer: Results of an open label, randomized, phase II trial of nanoparticle albumin-bound paclitaxel with or without the anti-death receptor 5 tigatuzumab (TBCRC 019)

Background: Circulating Tumor cells (CTCs) are prognostic at baseline and first follow-up in patients with metastatic breast cancer (MBC). Using the most commonly used assay (CellSearch®), we have previously reported the ability to detect apoptotic vs. non-apoptotic CTCs in patients with MBC. However, there has been concern regarding the performance of the CellSearch® assay in patients with triple negative (TN) MBC. We hypothesized that CellSearch® is an effective assay in patients with TN MBC, and that CTC-apoptosis might further separate prognosis. Therefore, we studied CTCs in patients with TN MBC who participated in a prospective randomized phase II trial testing for activity of tigatuzumab (TIG) in combination with nanoparticle albumin-bound paclitaxel (nab-PAC) conducted by the Translational Breast Cancer Research Consortium (overall results reported by Forero A., et al, ASCO 2013).

Methods: Whole blood (WB) was drawn into a CellSave tube at baseline, day 15, and day 29 and CTC counts were determined using the CXC CellSearch® kit. Apoptosis was characterized by staining with a monoclonal antibody that detects a neo-epitope on fragmented cytokeratin (M-30) and independently by visual inspection (nucleic condensation and/or fragmentation, as well as granular cytokeratin). Association between levels of CTCs and CTC-apoptosis with the overall response rate (ORR) and progression free survival (PFS) at baseline, day 15, and day 29 was assessed using logistic regression, Kaplan Meier curves, and Cox proportional hazards modeling.

Results: Of the 60 patients entered into the trial, 52 were evaluable for CTCs. Of these, 19/52 (36.5%), 14/52 (26.9%), and 13/49 (26.5%) had elevated CTCs (≥5CTC/7.5 ml WB) at baseline, day 15, and day 29, respectively. Patients with elevated CTCs at each time point had worse PFS than patients with low or no CTCs. Hazard rates (HR) at baseline, day 15, and day 29 were 2.38 (95% CI: 1.27-4.45, p = 0.007), 2.87 (95% CI: 1.46-5.66, p = 0.002), and 3.40 (95% CI: 1.68-6.89, p = 0.001), respectively. The odds of overall response for those who had elevated CTCs compared to those who did not at baseline, day 15, and day 29, were 0.25 (95% CI: 0.073-0.81, p = 0.024), 0.18 (95% CI: 0.04-0.67, p = 0.014), and 0.06 (95% CI: 0.01-0.28, p = 0.001), respectively. There was no apparent prognostic effect comparing the degree of CTC-apoptosis vs. non-apoptosis.

Conclusions: Similar to observations in other intrinsic subgroups, CTCs were detected in a large fraction of TN MBC patients at baseline using CellSearch® assay, and reductions in CTC levels reflected response. In these homogenously prospectively enrolled TN MBC patients, regardless of treatment, CTCs are prognostic at baseline, day 15, and day 29. It does not appear that analysis of CTC-apoptosis is prognostic.

Supported by Susan G. Komen for the Cure, Veridex, LLC, Fashion Footwear Charitable Foundation of New York/QVC Presents Shoes on Sale™ (DFH), Associazione Sandro Pitigliani and by a studentship from FIRC (CP), Triple Negative Breast Cancer Foundation, The AVON Foundation, and The Breast Cancer Research Foundation.

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P1-04-01.

Abstract S3-02: Associations between baseline patient-reported symptoms and discontinuation of adjuvant aromatase inhibitor (AI) therapy

Background: Non-adherence and non-persistence with AI therapy are common and have been associated with increased mortality. A major reason for premature discontinuation of therapy is toxicity of AI therapy. We performed an exploratory analysis to investigate associations between patient-reported symptoms at the time of AI initiation and discontinuation of AI therapy due to toxicity.

Methods: 503 postmenopausal women with early stage ER positive breast cancer were enrolled into a randomized open label clinical trial of exemestane versus letrozole for 2 years. Questionnaires about sleep quality (PSQI), mood disorders (CESD), and general symptoms (NSABP symptom questionnaire) were completed prior to AI initiation. For each item on the symptom questionnaire, responses were dichotomized as not at all vs any. To analyze total number of symptoms the following 4 symptoms were included: PSQI>5, CESD≥16, any degree of joint pain, and any degree of difficulty concentrating. Subjects were evaluated 1, 3, 6, 12, and 24 months after AI initiation. Reasons for treatment discontinuation were prospectively recorded. The primary endpoint of this exploratory analysis was treatment discontinuation due to toxicity ≤12 months following AI initiation. Linear and logistic regression analyses were performed to analyze associations between clinical factors and baseline symptoms and treatment discontinuation by 12 months.

Results: One hundred forty-two subjects (28.8%) discontinued therapy because of toxicity by or at 12 months. On univariate analysis, poor sleep quality (PSQI>5), depression (CESD≥16), forgetfulness, and total number of symptoms were statistically significantly associated with treatment discontinuation. On multivariable logistic regression, treatment discontinuation due to toxicity was statistically significantly associated with AI medication, and there was a trend towards an association with both younger age and increasing total number of symptoms present at baseline.

Univariate and multivariate analysis of baseline predictors of AI treatment discontinuation   Univariate analysisMultivariate analysisSymptom/characteristic% subjectsOR (95% CI)P valueOR (95% CI)P valueAge (yrs) mean 58.0 (9.0)0.98 (0.96-1.00)0.060.98 (0.95-1.00)0.076Drug (vs letrozole) 49.5%1.47 (0.99-2.18)0.0541.57 (1.04-2.38)0.033Estradiol (BL, pg/ml) mean 10.6 (37.2)1.01 (0.997-1.02)0.151.01 (1.00-1.02)0.091Prior taxane (vs no) 32.7%1.23 (0.82-1.85)0.331.14 (0.73-1.78)0.57Total # symptoms1 vs 028.2%1.56 (0.89-2.73)0.0241.27 (0.70-2.31)0.092 2 vs 021.5%1.52 (0.84-2.77) 1.54 (0.83-2.88)  ≥3 vs 023.7%2.41 (1.37-4.23) 2.10 (1.15-3.84) Depression 15.2%1.03 (1.01-1.06)0.012  Forgetfulness (vs none) 45.5%1.64 (1.11-2.43)0.014  Difficulty concentrating (vs none) 23.7%1.35 (0.86-2.11)0.19  Joint pain (vs none) 57.6%1.42 (0.95-2.12)0.09  Poor sleep quality (vs good) 47.4%1.84 (1.23-2.75)0.003  Vaginal dryness (vs none) 31.5%1.37 (0.91-2.08)0.14  CI: confidence interval; OR: odds ratio

Conclusions: Patient-reported symptoms present prior to initiation of AI therapy may predispose patients to early discontinuation of treatment. Pre-emptive management of these symptoms, rather than treatment of AI toxicity after its development, may improve adherence to and persistence with therapy.

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr S3-02.

Abstract P1-08-11: Association of variants in candidate genes on lipid profiles in women with early breast cancer on adjuvant aromatase inhibitor therapy

Background

Aromatase inhibitors (AI) can exert unfavorable effects on lipid profiles, but previous studies have reported inconsistent results. Given the intricate biological relationship between estrogen and lipid profiles, these mixed results may be explained in part by variation in genes encoding proteins involved in the drug's target and in estrogen metabolism and signaling. The purpose of this study was to investigate associations of single-nucleotide polymorphisms (SNP) in candidate genes with AI-mediated changes in lipid profiles.

Methods

We completed a prospective multicenter randomized observational open-label study to test the association of SNPs in candidate genes on biomarkers of estrogenic and anti-estrogenic activity in post-menopausal women with early breast cancer who were recommended adjuvant AI therapy. Eligible women were randomly assigned to exemestane or letrozole, and were followed for 2 years. We genotyped 137 SNPs from germ line DNA in the following candidate genes: ARVCF, COMT, CYP19A1, ESR1, ESR2, PGR, EP300, EZH2, NCOA1-3, NCOR1-2, NRIP, and PELP1. Lipid profiles including total cholesterol (TC), low-density lipoprotein (LDL), high-density lipoprotein (HDL), and triglycerides (TG) were measured at baseline and 3 months after initiating AI. We conducted genetic association data analysis and multivariate linear regressions to analyze the genetic effects using dominant, recessive, and additive models. Multivariate analysis included age, body mass index, prior hormone replacement therapy, and prior tamoxifen. To adjust for multiple comparisons, only SNPs with a p<0.0003 were considered significant.

Results

We enrolled 502 women in to the study, but for this analysis we excluded women who did not have genetic data (n = 33), had incomplete data (n = 23), discontinued or crossed over AI therapy (n = 48), women not fasting at both time points (n = 89), or those on lipid-lowering medications (n = 162). A total of 200 women were evaluable (letrozole 107, exemestane 93). Lipid profiles in all patients (n = 200) at baseline and 3 months after initiating AI, respectively, were as follows: TC 204.9 and 203.3 (unchanged, p = 0.43); HDL 61.3 and 56.8 (decreased, p = 6.3E-10); LDL 122.2 and 124.6 (unchanged, p = 0.22); and TG 107.1 and 103.6 (unchanged, p = 0.26). Genetic association and multivariate analysis revealed that SNPs in ESR1 and NCOR1 are significantly associated with additional changes in lipid parameters as summarized in Table 1.

Table 1.Significant findings of multivariate linear regressions analyzing genetic associations between candidate gene SNPs and lipid profiles of AI-treated women.CohortNumberSNP (gene)Minor Allele FrequencyLipid ParameterModel UsedMean Absolute Change (mg/dL)P-valueAll patients184rs9340958 (ESR1)0.07TCRecessive-2.250.0003Letrozole96rs9340958 (ESR1)0.07TCRecessive5.280.00009 101rs3020368 (ESR1)0.09TCRecessive6.350.00007Exemestane93rs3798758 (ESR1)0.03HDLDominant, additive-7.970.00001 88rs926848 (ESR1)0.03HDLDominant, additive-7.970.00002 93rs61753150 (NCOR1)0.01TGDominant, additive-11.630.00003

Conclusions

Variants in genes involved in estrogen metabolism and signaling are associated with changes in lipid profiles in AI-treated women and should be validated in other studies.

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P1-08-11.

Abstract PD6-4: Heterogeneity of expression of estrogen receptor by circulating tumor cells suggests diverse mechanisms of resistance to fulvestrant in metastatic breast cancer patients

Introduction: Fulvestrant is a selective estrogen receptor down-regulator (SERD). Recent studies have shown that the efficacy of fulvestrant is dose-related. However, at the higher dose (500 mg/month) most cancers develop resistance and progress. We previously reported expression of several markers, including estrogen receptor (ER) and BCL-2, on breast cancer circulating tumor cells (CTC) using CellSearch®. We now report pilot data showing inter-patient heterogeneity of these markers on CTC in patients with known ER positive breast cancer whose disease is progressing on fulvestrant.

Methods: We conducted a pilot trial to determine the analytical validity of measuring expression of markers of endocrine sensitivity (ER, BCL-2) or resistance (HER-2, Ki-67) with fluorescent-labeled antibodies using the CellSearch® system. Patients with ER positive metastatic breast cancer (MBC) whose disease was progressing on any type of therapy were eligible after signed informed consent. This report is limited to the subjects who were progressing on fulvestrant. Whole blood (WB) was characterized for CTC counts and each of the four molecular markers using the CXC CellSearch® kit.

Results: Of 50 enrolled patients, seven were progressing on fulvestrant. Two patients had no detectable CTC, while five patients had an average of ≥5 CTC/7.5 mL WB. Results are shown in a table below:

   CTC-ERCTC-BCL-2Patient #Fulvestrant dose (mg/month)Days since last doseN CTC/7.5 mL of WB% of CTC-ER+N CTC/7.5 mL of WB% of CTC-BCL-2+295002880%110%4550028170%170%2250341010%714%850031812%1735%172507728%367%

These exploratory data suggest widely different mechanisms of resistance to fulvestrant in different patients with ER positive MBC. In two of the patients (29, 45) treated with 500 mg/month, both CTC-ER and CTC-BCL-2 expression were absent, suggesting no signaling through the ER pathway. We hypothesize either that fulvestrant was actively down-regulating ER, but the cancers had adopted other growth and survival pathways, or that ER negative, hormone-independent clones had evolved. In the other three cases, ER was clearly present with evidence of signaling, based on BCL-2 expression. Two of these patients (2, 17) were on the lower dose of fulvestrant, now considered to be less effective. However, the third (8) was on the higher dose and yet still had evidence of ER signaling. This observation suggests that some patients may benefit from even higher doses of SERD therapy.

Conclusions: These pilot results suggest heterogeneous biological or pharmacological mechanisms of resistance to SERD therapy. These data suggest that CTC-ER and CTC-BCL-2 expression could serve as pharmacodynamic monitoring tools for dose escalation of fulvestrant or other SERDs. Further molecular analysis might provide biological bases for resistance to fulvestrant.

Supported by Veridex, LLC, Fashion Footwear Charitable Foundation of New York/QVC Presents Shoes on Sale™ (DFH), Associazione Sandro Pitigliani and by a studentship from FIRC (CP).

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr PD6-4.

Germline genetic variants in ABCB1, ABCC1 and ALDH1A1, and risk of hematological and gastrointestinal toxicities in a SWOG Phase III trial S0221 for breast cancer

Hematological and gastrointestinal toxicities are common among patients treated with cyclophosphamide and doxorubicin for breast cancer. To examine whether single-nucleotide polymorphisms (SNPs) in key pharmacokinetic genes were associated with risk of hematological or gastrointestinal toxicity, we analyzed 78 SNPs in ABCB1, ABCC1 and ALDH1A1 in 882 breast cancer patients enrolled in the SWOG trial S0221 and treated with cyclophosphamide and doxorubicin. A two-SNP haplotype in ALDH1A1 was associated with an increased risk of grade 3 and 4 hematological toxicity (odds ratio=1.44, 95% confidence interval=1.16-1.78), which remained significant after correction for multiple comparisons. In addition, four SNPs in ABCC1 were associated with gastrointestinal toxicity. Our findings provide evidence that SNPs in pharmacokinetic genes may have an impact on the development of chemotherapy-related toxicities. This is a necessary first step toward building a clinical tool that will help assess risk of adverse outcomes before undergoing chemotherapy.

Circulating tumour cells: insights into tumour heterogeneity

Tumour heterogeneity is a major barrier to cure breast cancer. It can exist between patients with different intrinsic subtypes of breast cancer or within an individual patient with breast cancer. In the latter case, heterogeneity has been observed between different metastatic sites, between metastatic sites and the original primary tumour, and even within a single tumour at either a metastatic or a primary site. Tumour heterogeneity is a function of two separate, although linked, processes. First, genetic instability is a hallmark of malignancy, and results in 'fixed' genetic changes that are almost certainly carried forward through progression of the cancer over time, with increasingly complex additional genetic changes in new metastases as they arise. The second type of heterogeneity is due to differential but 'plastic' expression of various genes important in the biology and response to various therapies. Together, these processes result in highly variable cancers with differential response, and resistance, to both targeted (e.g. endocrine or anti-human epithelial growth receptor type 2 (HER2) agents) and nontargeted therapies (e.g. chemotherapy). Ideally, tumour heterogeneity would be monitored over time, especially in relation to therapeutic strategies. However, biopsies of metastases require invasive and costly procedures, and biopsies of multiple metastases, or serially over time, are impractical. Circulating tumour cells (CTCs) represent a potential surrogate for tissue-based cancer and therefore might provide the opportunity to monitor serial changes in tumour biology. Recent advances have enabled accurate and reliable quantification and molecular characterization of CTCs with regard to a number of important biomarkers including oestrogen receptor alpha and HER2. Preliminary data have demonstrated that expression of these markers between CTCs in individual patients with metastatic breast cancer reflects the heterogeneity of the underlying tumours. Future studies are designed to determine the clinical utility of these novel technologies in either research or routine clinical settings.

Predictors of recovery of ovarian function during aromatase inhibitor therapy

BACKGROUND

Aromatase inhibitors (AIs) may cause a rise in estrogen levels due to ovarian function recovery in women with clinical chemotherapy-induced ovarian failure (CIOF). We carried out a prospective registry trial to identify predictors of ovarian function recovery during AI therapy.

PATIENTS AND METHODS

Women with hormone receptor (HR)-positive breast cancer who remained amenorrheic and had hormonal levels consistent with ovarian failure after adjuvant chemotherapy were enrolled in a multi-institutional clinical trial of anastrozole. Subjects underwent frequent assessment using an ultrasensitive estradiol assay. Multivariable analysis was used to evaluate clinical and biochemical predictors of ovarian function recovery within 48 weeks.

RESULTS

Recovery of ovarian function during AI therapy was observed in 13 of 45 (28.9%) assessable subjects after a median 2.1 months (range 0.6-11.9). Median age at chemotherapy initiation was statistically significantly different between those who regained ovarian function (43 years, range 40-51) and those who remained postmenopausal (49 years, range 44-52; P < 0.0001).

CONCLUSIONS

A significant proportion of women with CIOF recover ovarian function during AI therapy, including a woman over age 50 at initiation of chemotherapy. Tamoxifen remains the standard of care for women with CIOF. If an AI is used, patients should be monitored frequently with high-quality estradiol assays. CLINICALTRIALS.GOV: NCT00555477.

Abstract S6-3: Neurocognitive impact in adjuvant chemotherapy for breast cancer linked to fatigue: A Prospective functional MRI study

Background: Our previous research showed evidence of compromised cognitive function prior to adjuvant chemotherapy for breast cancer, with fatigue as a contributory factor. Fatigue is a common symptom reported by women treated for breast cancer, yet its association with neurocognitive function has not been systematically examined. In this prospective study, we examined possible alterations in neurocognitive responses, namely, working memory, from pre- to post- adjuvant treatment during functional magnetic resonance imaging (fMRI) and further investigated whether early fatigue might be linked to cognitive alterations over time.

Methods: Women treated with either adjuvant chemotherapy (anthracyline-based combination regimen, n=29) or radiotherapy (n = 37) for localized breast cancer (Stages 0-IIIa) and age-matched healthy controls (n = 32) were enrolled. Participants performed a verbal working memory task (VWMT) with varying levels of demand for cognitive control during fMRI scanning and provided self-reports of fatigue (FACT-F) at two time points coincident with pre- and one-month post chemotherapy assessments. Imaging data were analyzed with general linear models using SPM5; comparative statistics were used to determine group differences, and correlational analyses addressed relationships of fatigue and neurocognitive measures.

Findings: The chemotherapy group reported significantly greater severity of fatigue (p &lt; .05) and performed less accurately on the VWMT both pre- and one-month post-treatment than the other groups. Greater fatigue was correlated with poorer performance on the VWMT at both time points across groups, with stronger correlation post-treatment (r = −.22, p = .03). A 2 time-point (pre- vs. post-treatment) × 2 group (chemotherapy vs. controls) × 2 demand-level contrasts (high minus low vs. medium minus low) analytic model showed a significant group × time interaction (p &lt; .05), mainly due to lower pre-treatment activation in an area of the prefrontal cortex supporting working memory, the anatomical left inferior frontal gyrus (LiFG), at higher task demand in the chemotherapy group. The radiotherapy group scored between the other two groups with intermediate activation of those contrasts. Of interest, lower pre-treatment activation in the LiFG in the high-low demand contrast predicted severity of fatigue across all participants at the post-treatment assessment (r = −.27, p &lt; .01), linking early compromise in neurocognitive performance with greater fatigue over time.

Discussion: Neurocognitive alterations during a working memory task and greater fatigue were evident before any adjuvant chemotherapy for breast cancer. Notably, functional alterations in working memory processes were evident with fMRI before adjuvant chemotherapy and predicted severity of post-treatment fatigue. Importantly, across all participants, greater fatigue over time was correlated with reduced cognitive performance. Taken together, these findings indicate that pre-treatment neurocognitive compromise and fatigue are key contributors to the cognitive impact often attributed solely to chemotherapy. Early therapeutic interventions targeting fatigue may improve cognitive function and reduce the distress of “chemo brain” throughout the course of adjuvant treatment.

Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr S6-3.

A SNP in steroid receptor coactivator-1 disrupts a GSK3β phosphorylation site and is associated with altered tamoxifen response in bone

The coregulator steroid receptor coactivator (SRC)-1 increases transcriptional activity of the estrogen receptor (ER) in a number of tissues including bone. Mice deficient in SRC-1 are osteopenic and display skeletal resistance to estrogen treatment. SRC-1 is also known to modulate effects of selective ER modulators like tamoxifen. We hypothesized that single nucleotide polymorphisms (SNP) in SRC-1 may impact estrogen and/or tamoxifen action. Because the only nonsynonymous SNP in SRC-1 (rs1804645; P1272S) is located in an activation domain, it was examined for effects on estrogen and tamoxifen action. SRC-1 P1272S showed a decreased ability to coactivate ER compared with wild-type SRC-1 in multiple cell lines. Paradoxically, SRC-1 P1272S had an increased protein half-life. The Pro to Ser change disrupts a putative glycogen synthase 3 (GSK3)β phosphorylation site that was confirmed by in vitro kinase assays. Finally, knockdown of GSK3β increased SRC-1 protein levels, mimicking the loss of phosphorylation at P1272S. These findings are similar to the GSK3β-mediated phospho-ubiquitin clock previously described for the related coregulator SRC-3. To assess the potential clinical significance of this SNP, we examined whether there was an association between SRC-1 P1272S and selective ER modulators response in bone. SRC-1 P1272S was associated with a decrease in hip and lumbar bone mineral density in women receiving tamoxifen treatment, supporting our in vitro findings for decreased ER coactivation. In summary, we have identified a functional genetic variant of SRC-1 with decreased activity, resulting, at least in part, from the loss of a GSK3β phosphorylation site, which was also associated with decreased bone mineral density in tamoxifen-treated women.

PD04-01: Predictors of Recovery of Ovarian Function during Aromatase Inhibitor (AI) Therapy

Background: AIs may cause a paradoxical rise in estrogen levels due to re-activation of ovarian function in women with chemotherapy-induced ovarian failure (CIOF). Therefore, identification of residual ovarian estradiol production is critical if such women are treated with adjuvant AI therapy rather than tamoxifen. We performed a prospective registry trial to identify predictors of recovery of ovarian function during AI therapy.

Methods: Women with hormone receptor (HR) positive breast cancer who were pre- or peri-menopausal at diagnosis and who remained amenorrheic for ≥8 weeks after cyclophosphamide-containing adjuvant chemotherapy were enrolled in a multi-institutional, open-label clinical trial of anastrozole (1 mg/day). Following confirmation that serum estradiol (E2) levels were &lt;20 pg/ml using an ultrasensitive E2 assay (Quest Diagnostics), subjects initiated anastrozole. Serum E2 was assessed biweekly for 12 wks, then less frequently, for 72 wks. Multivariable logistic regression was used to evaluate clinical predictors (age at AI initiation OR chemotherapy, menopausal status at chemotherapy, body mass index (BMI), baseline E2) of recovery of ovarian function defined as elevated serum E2 levels or return of menses.

Results: Sixty-nine women were enrolled; current status is given in Table 1.

P4-07-16: Development of Circulating Tumor Cell-Endocrine Therapy Index in Metastatic Breast Cancer Patients

Introduction: Only ∼ 50% of patients (pts) with estrogen receptor (ER) positive metastatic breast cancer (MBC) benefit from endocrine therapy (ET). Currently only clinical judgment can be used to identify pts with endocrine-refractory MBC, who are better palliated with chemotherapy. Circulating Tumor Cells (CTC) are reliably enumerated using an automated immunomagnetic system (CellSearch®; Veridex LLC). High CTC levels predict rapid progression in pts with MBC. We have developed a multi-parameter assay, the CTC-Endocrine Therapy Index (CTC-ETI) using CellSearch® that may identify pts with ER positive MBC who are unlikely to benefit from ET and may be better served with chemotherapy. CTC-ETI scores are assigned based on CTC levels coupled with the relative percent and degree of marker positivity on the CTC. We report preliminary results from a pilot single institutional study.

Methods: CellSearch® has 4 fluorescence channels. Three distinguish CTC from WBC (DAPI, anti-cytokeratin, anti-CD45). The 4th “empty” channel was used to measure ER, BCL-2, HER-2, and Ki-67 expression with fluorescent-labeled antibodies. These 4 markers reflect sensitivity (ER, BCL-2) or resistance (HER-2, Ki-67) to ET. Forty ml of blood was drawn into 4 CellSave® tubes from pts with progressive MBC. Whole blood from 4 tubes was pooled and divided into 4 different 7.5 ml aliquots of blood, which were processed and characterized for CTC counts and each of the four molecular markers using the CXC CellSearch® kit.

Results: 21 pts have been accrued to the feasibility study. One patient was ineligible. Five of 20 pts had low CTC counts (&lt;5 CTC/7.5ml whole blood), and are expected to have a relatively favorable prognosis. CTC-ETI was successfully determined in 10 pts (50%): 2 pts had low, while 3 had intermediate, and 5 had high CTC-ETI. Technical difficulties precluded accurate CTC-ETI in the remaining 5 patients. Of note, expression of the biomarkers among CTC in single patients was heterogeneous, suggesting that future iterations of the CTC-ETI will have to consider expression variability. Further exploratory results regarding associations between CTC-ETI and outcomes will be presented.

Conclusions: ER, BCL-2, HER-2, and Ki-67 can be accurately determined on CTC using the 4th channel in the CellSearch® system to calculate CTC-ETI. We predict that lower CTC-ETI scores (low or no CTC, or CTC with high CTC ER and BCL-2 and low CTC HER-2 and Ki-67) could be associated with favorable response to ET. Successful completion of the feasibility study will lead to a prospective trial to determine if high CTC-ETI at baseline predicts resistance and rapid progression on ET in women starting a new endocrine therapy for MBC.

Supported by Veridex, LLC, Fashion Footwear Charitable Foundation of New York/QVC Presents Shoes on Sale ™ (DFH), Associazione Sandro Pitigliani and by a studentship from FIRC (CP).

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P4-07-16.

Plasma letrozole concentrations in postmenopausal women with breast cancer are associated with CYP2A6 genetic variants, body mass index, and age

The associations between plasma letrozole concentrations and CYP2A6 and CYP3A5 genetic variants were tested in the Exemestane and Letrozole Pharmacogenomics (ELPH) trial. ELPH is a multicenter, open-label prospective clinical trial in women randomly assigned (n≈250 in each arm) to receive 2 years of treatment with either oral letrozole (2.5 mg/day) or oral exemestane (25 mg/day). CYP2A6 and CYP3A showed effects on letrozole metabolism in vitro. DNA samples were genotyped for variants in the CYP2A6 and CYP3A5 genes. Plasma letrozole concentrations showed high interpatient variability (>10-fold) and were associated significantly with CYP2A6 genotypes (P<0.0001), body mass index (BMI) (P<0.0001), and age (P=0.0035). However, CYP3A5 genotypes showed no association with plasma letrozole concentrations. These data suggest that CYP2A6 is the principal clearance mechanism for letrozole in vivo. CYP2A6 metabolic status, along with BMI and age, may serve as a biomarker of the efficacy of letrozole treatment or a predictor of adverse effects.

Abstract P3-02-06: Correlation of BCL-2 and Apoptosis on Circulating Tumor Cells and Breast Cancer Tissue

Background: Circulating Tumor Cell (CTC) levels are prognostic markers in metastatic breast cancer (MBC). Further phenotypic characterization of CTC may provide an opportunity for non-invasive evaluation of predictive and prognostic markers. Apoptosis is a common form of chemotherapy-induced cell death. Monoclonal antibody (MAb) M30 recognizes a neo-epitope on fragmented cytokeratin and is a marker of apoptosis. BCL-2 is an anti-apoptotic marker and may predict for resistance to anti-neoplastic therapy. We have previously reported the results of a pilot clinical trial to estimate M30 and BCL-2 expression in CTC from patients with MBC. The current study was performed to correlate the expression of both BCL-2 and M30 in CTC with breast cancer tissue samples. Methods: Of the 85 patients in the original pilot study, 52 (61%) had evaluable tissue available (39 primary and 12 metastatic lesions; one unknown) and were included in this analysis. CTC were collected at baseline and were isolated, enumerated, and phenotypically characterized for M30 and BCL-2 using the CellSearch® System. CTC phenotype is reported as percentage of cells with positive staining. Tissue Microarrays (TMA) were constructed and immunohistochemically stained for M30 (anti-Cytodeath™ M30, Roche) and Bcl-2 (M-0887, DAKO) and scored using the Allred method.

Results: Tissue staining for M30 and Bcl-2 were inversely correlated. Of the 52 patients with TMA available for investigation, 22 (42%) had &gt; 5 CTC/7.5 ml whole blood. A positive, but non-significant, correlation was observed between increasing numbers of CTC levels and tissue BCL-2 Allred Score (Spearman r=0.36; p=0.1310). M30 staining was detected in ≥10% of CTC in 20 of the 22 (91%) patients with elevated CTC, while BCL-2 staining was detected in ≥10% of CTC in 19 of the 22 (86%) patients. Little if any correlation was observed between CTC M30 and tissue M30 expression, however CTC BCL-2 was positively correlated with tissue Bcl-2 expression (Spearman r=0.47; p=0.0440). Conclusions: BCL-2 and M30 can be characterized in both CTC and breast cancer tissue. The likelihood of having elevated CTC may be associated with higher BCL-2 expression, and CTC BCL-2 was associated with tissue BCL-2 expression. Although the clinical implications of these findings are unknown, they offer the opportunity for future trials to investigate whether CTC BCL-2 and M30 might be useful to identify patients with cancers that are resistant to standard therapies. Furthermore, CTC BCL-2 and M30 might be monitored in clinical trials of novel, apoptosis-inducing therapeutic agents.

Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P3-02-06.

Abstract P2-14-09: Prospective Evaluation of Change in 2-Point Discrimination of Index Finger as a Potential Early Predictive Marker for Carpal Tunnel Syndrome among Women Receiving Adjuvant Aromatase Inhibitor Therapy for Postmenospausal Breast Cancer in the Exemestane and Letrozole Pharmacogenomics (ELPh) Trial

Introduction: Third generation aromatase inhibitors (AIs) represent an integral part of hormonal therapy in postmenopausal women with hormone receptor (HR)-positive breast cancer. AIs are associated with musculoskeletal symptoms in up to 50% of women. Post-hoc analyses of adjuvant AI trials (ATAC and IES) have suggested that AIs might be associated with carpal tunnel syndrome (CTS), a pressure-induced neuropathy disorder caused by compression on the median nerve. The clinical diagnosis of CTS is made with typical symptoms of pain, weakness, and paresthesias in affected arm. A variety of tests, including change in 2-point discrimination (2PD) can be used to aid in diagnosis. However, the actual incidence of CTS and clinical utility of diagnostic tests such as 2PD have not been prospectively examined among women receiving AIs. Methods: Postmenopausal women with stage 0-III HR-positive breast cancer, who had completed local therapy and, if indicated, adjuvant chemotherapy, and who were enrolled in the multi-center Exemestane and Letrozole Pharmacogenetics (ELPh) trial underwent prospective evaluation of 2PD with the Disc-criminator™ (sliding aesthesiometer) at baseline, and 3 months, following initiation of the AI. The end of the Disc-criminator™ was applied at the two points at same time to the skin on the volar tip pulp of the index fingers, and the threshold value (in mm) was determined as the shortest distance between the two points a woman was able to differentiate. The exercise was repeated thrice at each point. Abnormal 2PD thresholds were defined using standard criteria (outside 95 percentile for age). The differences in mean 2PD from baseline to 3 months were analyzed using a multivariate mixed effects model where the correlations from repeated measures were accounted for by assuming an unstructured covariance structure. A p value &lt; 0.05 was considered statistically significant.

Results: A total of 104 women underwent baseline 2PD testing. The mean age was 59 years, 55.8% had stage I disease, and 42.3% received adjuvant chemotherapy. We observed abnormal 2PD thresholds in 1.9% and 3.5% of women at baseline and 3 months respectively. There was a significant worsening in the adjusted mean 2PD from baseline (3.4 mm) to 3 months (4 mm, p=0.01). The increase in mean 2PD following 3 months of AI therapy was higher among women with age &gt; 55 (p=0.02), BMI &gt; 25 (p=0.002), African Americans (p=0.02), and those who received adjuvant chemotherapy (p=0.05), as compared to their counterparts. Conclusion: Adjuvant AI therapy was associated with a significant worsening of 2PD at 3 months, particularly among older women, overweight women, and those receiving adjuvant chemotherapy. Correlation with CTS symptoms and need for surgical release will be presented at the meeting. Our results suggest that 2PD is a non-invasive method that may potentially allow for early detection of CTS. If confirmed, change in 2PD could serve as an objective early predictor for subsequent CTS in postmenopausal women with breast cancer initiating AI therapy.

Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P2-14-09.

Abstract P6-15-03: Phase Ib Trial of the Gamma Secretase Inhibitor (GSI), MK-0752 Followed by Docetaxel in Locally Advanced or Metastatic Breast Cancer

Background: The cancer stem cell hypothesis asserts that there is a small population of cells within a tumor that has the ability to self renew and differentiate, and that these cells drive tumor growth and metastasis but are resistant to conventional cytotoxic chemotherapy. Pathways involved in stem cell growth and differentiation are viable targets for new anticancer therapies. One such pathway, Notch, is inhibited by GSIs which prevent translocation of Notch intracellular domain to the nucleus. Inhibition of GS concurrent with chemotherapy might improve disease control by targeting both stem cells and differentiated cells within the tumor. This Phase Ib clinical trial was designed to determine the MTD of the GSI, MK-0752, in combination with docetaxel, and to evaluate an effect on stem cell markers in serial tumor biopsies.

Methods: Eligible subjects had metastatic breast cancer or locally advanced breast cancer that did not respond to anthracycline therapy. Patients with disease that progressed on a taxane, or who had received a taxane within 6 months were excluded. MK-0752 was administered orally on days 1-3 of each 21-day cycle of therapy, in escalating doses. Dose levels (mg/day) 1=300; 2=450; 3=600; and 4=800. Docetaxel 80 mg/m2 IV was administered day 8, with pegfilgrastim day 9 each cycle. Treatment was continued until disease progression, unacceptable toxicity, or symptomatic deterioration. The trial was monitored using the Time to Event Continual Reassessment Method, targeting a 20% toxicity rate. Tumor biopsies were performed at baseline, after 1 cycle, and at treatment completion in a subset of patients.

Results: 30 patients were enrolled between Mar 2008 and Jan 2010. Dose limiting toxicities of the combination included diarrhea, hand-foot syndrome, and LFT elevation. 20/30 patients experienced Grade 1 or 2

fatigue. The final estimates and confidence intervals for the probability of dose limiting toxicity at each dose level are summarized in the table:

Probability = probability of dose-limiting toxicity 20 enrolled patients had measurable disease by RECIST criteria. Of these, 9 had PR, 8 SD, and 3 PD, for an estimated RR of 45% to the combination. 2 patients have been maintained on therapy in excess of 22 cycles. Conclusions: Dose level 3 was identified for further study in a Phase II randomized trial. Efficacy of docetaxel was not inhibited by MK-0752, as a 45% RR in patients with measurable disease was observed. There is intriguing long term disease stabilization in 2 patients. Evidence of an effect of the combination on the stem cell population was apparent on serial biopsies as presented at SABCS Dec 2009 (Abstract # 48); additional biopsy data will be presented.

Cytochrome P450 2D6 activity predicts discontinuation of tamoxifen therapy in breast cancer patients

The selective estrogen receptor modulator tamoxifen is routinely used for treatment and prevention of estrogen-receptor-positive breast cancer. Studies of tamoxifen adherence suggest that over half of patients discontinue treatment before the recommended 5 years. We hypothesized that polymorphisms in CYP2D6, the enzyme responsible for tamoxifen activation, predict for tamoxifen discontinuation. Tamoxifen-treated women (n=297) were genotyped for CYP2D6 variants and assigned a 'score' based on predicted allele activities from 0 (no activity) to 2 (high activity). Correlation between CYP2D6 score and discontinuation rates at 4 months was tested. We observed a strong nonlinear correlation between higher CYP2D6 score and increased rates of discontinuation (r(2)=0.935, P=0.018). These data suggest that presence of active CYP2D6 alleles may predict for higher likelihood of tamoxifen discontinuation. Therefore, patients who may be most likely to benefit from tamoxifen may paradoxically be most likely to discontinue treatment prematurely.

Multiplexed AQUA-based assessment of SWOG 9313 shows prognostic value of continuous ER, PR and HER2 assessment

Abstract #704

Introduction: HER2 is expressed at high levels due to gene amplification in about 15% of breast cancer cases and it has been shown to be a poor prognostic marker. However, HER2 is also expressed in normal breast duct tissue, albeit a much lower levels. We hypothesized that continuous analysis of expression using AQUA will provide prognostic information beyond that attainable with conventional methods.&#x2028; Methods: A tissue microarray was made from 2123 cases of the 3122 patients accrued to SWOG 9313, in which sequential doxorubicin and cyclophosphamide (A-C) was compared to combination AC. A multiplexed assessment of HER2 and estrogen receptor (ER) was performed on the same slide using the immunofluoresence-based AQUA® method of automated quantitative analysis. Reproducibility and fidelity of multiplexing were determined for each marker by regression analysis.&#x2028; Results: As expected, both ER and PR were significantly predictive of disease-free survival (DFS) when both are tested as continuous variables, both adjusted for node status, tumor size, treatment and menopausal status (p-values 0.005 and &lt;0.001, respectively). HER2, measured as a continuous variable showed a bi-phasic effect. It has been previously reported (Camp et al, Cancer Research 2003, 63;1445) that both the high and low expressers of HER2 have worse outcome (low levels are equivalent to that seen in normal breast ducts). Splitting the SWOG cohort by deciles shows that both the top and bottom decile have worse DFS than the middle 80% (log rank p=0.012). Also, modeling the hazard ratio as a function of concentration shows a U-shape relationship showing both high and low HER2 expression is associated with poorer DFS.&#x2028; Conclusions: The AQUA method provides a reproducible method of continuous measurement of ER, and HER2 on the same slide. In this cohort both ER and PR as continuous variable are highly prognostic, as expected, but multiplexing with HER2 did not affect outcome. Quantitative analysis demonstrated that both low and high levels of HER2 expression were associated with poor outcome. Studies are ongoing to determine the significance of this observation with respect to biological classifications of breast cancer and relationships with breast cancer therapies.

Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 704.

Associations between estrogen receptor (ER) Alpha expression levels and ER genotypes

Abstract #3049

Background: Estrogen receptors (ER) and associated factors, including progesterone receptors (PgR), modulate estrogenic effects both in normal tissue and in hormone receptor (HR)-positive breast cancers. ERα expression occurs in the majority of breast cancers, although the level of overexpression varies between tumors. Multiple single nucleotide polymorphisms (SNPs) in the genes that encode ER, termed ESR1 (ERα) and ESR2 (ERβ), have been shown to affect phenotypic outcomes, including breast cancer risk, bone mineral density, hot flashes, and lipid levels. We hypothesized that SNPs in genes that encode ER may affect level of ER overexpression.&#x2028; Methods: DNA was extracted from white blood cells from 100 subjects with HR-positive breast cancer enrolled in a tamoxifen observational study. SNPs in ESR1 (XbaI: rs#9340799, PvuII: rs#2234693) and ESR2 (01: rs#1256049; 02: rs#4986938) were determined from germline DNA using TaqMan. ER overexpression in paraffin-embedded formalin-fixed tumor specimens was quantified with immunohistochemistry (IHC) of tumor specimens on slides (Ventana Image Analysis System) and with RT-PCR using the Oncotype DX breast cancer assay (Genomic Health, Inc). Gene expression of other ER-associated genes (PgR, BCL2, and SCUBE2) was determined using RT-PCR (Oncotype DX). Expression of ER and associated genes across ER genotype was performed using the Kruskal-Wallis test. For all analyses, a p value of &lt;0.05 was considered statistically significant.&#x2028; Results: ER expression assessed using IHC and RT-PCR was statistically significantly correlated, with a Pearson correlation coefficient of 0.43 (p=0.0006). However, there was no association between the ESR1 or ESR2 genotypes and ER expression assessed by either method. A trend between ESR1 XbaI genotype and PgR expression by RT-PCR was noted (p=0.08 for gene-dose effect). No other associations between ER polymorphisms and expression of ER-related genes were identified.&#x2028; Conclusions: SNPs in genes encoding for ER do not appear to be associated with level of ER expression in breast cancer. Differences in ER expression between individual tumors are likely due to mechanisms other than these inherited mutations in the ER genes.

Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 3049.

Estrogen receptor genotypes, menopausal status, and the lipid effects of tamoxifen

Tamoxifen induces important changes in serum lipid profiles in some women; however, little information is available to predict which women will experience improved lipid profiles during tamoxifen therapy. As part of a multicenter prospective observational trial in 176 breast cancer patients, we tested the hypothesis that tamoxifen-induced lipid changes were associated with genetic variants in candidate target genes (CYP2D6, ESR1, and ESR2). Tamoxifen lowered low-density lipoprotein cholesterol (P<0.0001) by 23.5 mg/dl (13.5-33.5 mg/dl) and increased triglycerides (P=0.006). In postmenopausal women, the ESR1-XbaI and ESR2-02 genotypes were associated with tamoxifen-induced changes in total cholesterol (P=0.03; GG vs GA/AA) and triglycerides (P=0.01; gene-dose effect), respectively. In premenopausal women, the ESR1-XbaI genotypes were associated with tamoxifen-induced changes in triglycerides (P=0.002; gene-dose effect) and high-density lipoprotein (P=0.004; gene-dose effect). Our results suggest that estrogen receptor genotyping may be useful in predicting which women would benefit more from tamoxifen.