Dose-Dense Sequential Adjuvant Chemotherapy With Epirubicin, Paclitaxel and CMF Versus Epirubicin, Cmf and Weekly Docetaxel or Paclitaxel Followed by Trastuzumab for One Year in Patients With Early Breast Cancer

Annals of Oncology 23 (Supplement 9): ix95–ix115, 2012 doi:10.1093/annonc/mds392

246O

DOSE-DENSE SEQUENTIAL ADJUVANT CHEMOTHERAPY WITH EPIRUBICIN, PACLITAXEL AND CMF VERSUS EPIRUBICIN, CMF AND WEEKLY DOCETAXEL OR PACLITAXEL FOLLOWED BY TRASTUZUMAB FOR ONE YEAR IN PATIENTS WITH EARLY BREAST CANCER

G. Fountzilas, H. Gogas, N. Pavlidis, A. Eleftheraki, D. Skarlos, A. Koutras, E. Timotheadou, C. Papandreou, D. Pectasides, M. Dimopoulos Data Office, Hellenic Cooperative Oncology Group (HeCOG), Athens, GREECE

247O PR

THE DISCREPANCY BETWEEN HIGH PATHOLOGICAL COMPLETE RESPONSE (PCR) RATE AND LOW BREAST CONSERVING SURGERY (BCS) FOLLOWING NEOADJUVANT THERAPY: ANALYSIS FROM THE NEOALTTO TRIAL (BIG 1-06)

C. Criscitiello1, H.A. Azim, Jr2, D. Agbor-Tarh3, E. De Azambuja4, M. Piccart5, J. Baselga6, H. Eidtmann7, S. Di Cosimo8, I. Bradbury3, I.T. Rubio9 1 Department of Medicine, European Institute of Oncology, Milan, ITALY, 2Breast Cancer Translational Research Laboratory, Institute Jules Bordet, Brussels, BELGIUM, 3Frontier Science, Frontier Science, Kincraig, UNITED KINGDOM, 4 Breast Data Centre, Institute Jules Bordet, Brussels, BELGIUM, 5Institut Jules Bordet, Institut Jules Bordet, Brussels, BELGIUM, 6Hematology/Oncology, MGH Cancer Center, Massachusetts General Hospital, Boston, MA, UNITED STATES OF AMERICA, 7Klinik für Gynäkologie und Geburtshilfe, Universitäts Frauenklinik Kiel, Kiel, GERMANY, 8Medical Oncology, Istituto Nazionale Tumori, Milan, ITALY, 9 Department of Gynecology, Hospital Universitario Vall d’Hebron, Barcelona, SPAIN Background: The NeoALTTO trial showed that paclitaxel plus lapatinib and trastuzumab nearly doubles the rate of pCR compared to paclitaxel combined with either drug alone (51.3% vs 29.5% vs 24.7%). However, this high pCR rate did not translate into a higher rate of BCS, which was around 40% across the 3 arms. We investigated different factors that may have affected the choice of surgery.

248O

ANALYSIS OF CORRELATION BETWEEN WEIGHT AT DIAGNOSIS, WEIGHT GAIN AFTER BREAST CANCER TREATMENT AND RECURRENCE IN WOMEN WITH EARLY STAGE BREAST CANCER (EBC)

P. Fedele1, A. Nacci2, A.M. Lapolla1, L. Orlando1, P. Schiavone1, A. Marino1, M. Cinefra1, A. Ardizzone1, M. De Pasquale1, S. Cinieri1 1 Ospedale Perrino, Oncologia Brindisi, Brindisi, ITALY, 2U.O.C. Oncologia, Ospedale A. Perrino, Brindisi, ITALY Background: Overweight at the time of EBC diagnosis has been linked frequently to poorer survival in most studies and some evidence suggests that women who gain weight after breast cancer diagnosis are at increased risk of cancer recurrence and death. Most previous studies on this topic have relied on retrospective chart reviews. The aim of this prospective, observational, single-center study is to determine whether weight at diagnosis and weight gain after EBC treatment are predictive of BC recurrence. Methods: From August 1997 to March 2012, the study included a total of 520 EBC patients (stage I-IIIa). We assessed weight and body mass index (BMI = kg/m2) at baseline (≤ 1 month after surgery) and 24 months after completion of treatment (chemotherapy ± radiotherapy). The chi square test (X2) was conducted to determine if a significant correlation exists between BC recurrence and 3 categories of BMI at diagnosis (lean weight: BMI <25; overweight: BMI 25-30; obese: BMI >25) and BC recurrence and weight changes after EBC treatment (loss of <1 kg/m2; loss of ≥ 1 kg/ m2; gain of <2kg/m2; gain of >2 kg/m2). Results: Median age was 55 years (range 28-81); 58% of patients were postmenopausal, stage I-II in 89%; ER + /PGR+ in 69%; ER-/PGR- in 20%; HER2+ in 20%; 72% underwent conservative surgery + radiotherapy; 57% received chemotherapy (CT) and 78% received endocrine therapy alone or after CT. Median BMI at diagnosis was 26.8, after treatment 27.7. After a median follow up of 13 years 194 patients recurred. Statistical analysis is reported in table 1. Conclusions: Our findings show that EBC patients gain weight after treatment. A significant correlation was found between weight gain after EBC treatment and recurrence, suggesting that women who gain weight after breast cancer diagnosis may be at increased risk of poor outcomes. Disclosure: All authors have declared no conflicts of interest.

© European Society for Medical Oncology 2012. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected]

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

Background: Dose-dense sequential chemotherapy including anthracyclines and taxanes has been well established in the adjuvant setting of high-risk operable breast cancer. However, the preferable taxane and optimal schedule of administration have not been defined, as yet. Patients and methods: From July 2005 until November 2008, 1,001 patients (990 eligible) were randomized to receive 3 cycles of epirubicin 110 mg/m2 followed by 3 cycles of paclitaxel 200 mg/m2 followed by 3 cycles of CMF (cyclophosphamide 840 mg/m2; methotrexate 57 mg/m2; fluorouracil 840 mg/m2) with G-CSF support (Group A; 333 patients) or to 3 cycles of epirubicin followed by 3 cycles of CMF, as in Group A, followed 3 weeks later by 9 weekly cycles of docetaxel 35 mg/m2 (Group B; 331 patients) or 9 weekly cycles of paclitaxel 80 mg/m2 (Group C; 328 patients). Radiation and hormonal therapy were given after the completion of chemotherapy. Trastuzumab was administered for 1 year to all HER2-positive patients post radiation. Results: At a median follow-up of 60 months, 123 patients had documented disease relapse (51 in Group A, 37 in Group B and 35 in Group C) and 81 deaths (30 in group A, 22 in group B and 29 in group C) had been observed. The 3-year disease-free survival (DFS) rate was 86%, 91% and 89%, with overall survival (OS) rates of 96%, 97% and 96%, respectively. No differences were found in DFS or OS between the three treatment groups (log-rank, p = 0.38 and p = 0.48, respectively). The most frequently reported severe adverse events were neutropenia (29% vs 27% vs 24%, p = 0.31) and leucopenia (12% vs 13% vs 10%, p = 0.66). Febrile neutropenia occurred in fifty patients (6%, vs 5% vs 5%, p = 0.81). Severe mucositis was more frequent in Group B (3% vs 6% vs 1%, p = 0.001), while severe neuropathy was more frequent in Group A (4% vs 0% vs 1%, p < 0.001). Conclusions: No significant differences in DFS and OS between the three regimens were identified. Taxane regimen and schedule of administration preferentially influenced the type of severe toxicities. HER2-positive patients demonstrated comparable 3-year DFS and OS rates with those reported in other similar studies. Disclosure: All authors have declared no conflicts of interest.

Patients and methods: In the NeoALTTO trial, patients ( pts) with HER2+ breast cancer were randomized to either trastuzumab, lapatinib or their combination concomitantly with paclitaxel prior to surgery. The 1ry endpoint was pCR, defined as the absence of invasive cancer in the breast at the time of surgery. Here, we investigated the association between achieving pCR, and type of surgery, age, histology, grade, tumor size, ER status, multicentricity, response to therapy and the country where the treatment was given. Results: 429 pts were eligible for the analysis (26 have been excluded as they did not undergo breast surgery), of whom 160 (37%) achieved a pCR. 242 (57%) and 187 (43%) pts had mastectomy and BCS, respectively. Mastectomy was more frequent if the patient was < 50, if treated in developing country, if the tumor was multicentric, >5 cm, or ER-. All pts diagnosed with lobular cancer (n = 17) underwent mastectomy regardless of pCR. 68 pts had a radiological complete response, yet 36 of those (53%) were subjected to mastectomy (25 pts (70%) achieved a pCR). Of the 128 pts considered for BCS at screening, only 95 (74%) had a conservative surgery and rates were similar according to pCR status (79% in pCR vs. 72% in no pCR). Conversely, 30% of pts initially evaluated as inoperable or requiring mastectomy had a BCS. Conclusion: Tumor characteristics prior to neoadjuvant therapy appeared to play a main role in deciding the type of surgery irrespective of response. This may deny a large fraction of women the chance of preserving their breast. These results should be taken into account in addressing the criteria for BCS after neoadjuvant therapy. GSK distributed the study drugs and provided financial support to the NeoALTTO trial, but imposed no restriction to the current analysis. Disclosure: E. De Azambuja: advisory board from GSK; honoraria from Roche. M. Piccart: Consulting/advisory role/honoraria from SanofiAventis, Amgen, Bayer, Bristol Myers Squibb, Roche, Glaxo Smith Kline, Boehringer, PharMar. S. Di Cosimo: Speaker for GSK. All other authors have declared no conflicts of interest.

submitted abstracts

breast cancer, early stage

Annals of Oncology

Table: 248O

No of patients Recurrences p

249PD

Body Mass Index At diagnosis <25 kg/m2

25-30 kg/m2

>30 kg/m2

Changes after BC treatment Loss >1 kg/m2 Loss <1 kg/m2

Gain <2 kg/m2

Gain >2 kg/m2

179

184

157

64

32

185

145

72 0.17

73

49

20 0.34

15

63 0.0008

76

PROTEOMIC PREDICTORS OF OUTCOME AFTER ADJUVANT ANTI-HORMONAL THERAPY FOR HORMONE RECEPTOR-POSITIVE BREAST CANCER

Purpose: This study aimed to identify predictive proteomic biomarkers of outcome in women with estrogen and/or progesterone receptor-positive (ER/PR-positive) breast cancer after adjuvant tamoxifen, with sufficient power to alter patient management. Methods: Using reverse phase protein arrays (RPPA), 140 antibodies were applied to a training set of 197 ER/PR-positive breast cancers to identify predictors. An algorithm was developed that predicted patient outcomes using a subset of antibodies. Since RPPA is a useful exploratory tool but does not lend itself as a practical clinical tool to assay validated biomarkers, quantitative immunofluorescence for selected proteins was applied to 313 ER/PR-positive breast cancers (test set) for validation. Seventy-seven other ER/PR-positive cancers with transcriptional profiling data were used to compare the performance of the proteomic biomarkers and established genomic predictors. All patients were treated with adjuvant tamoxifen after loco-regional therapy. Results: Two different combinations (4-protein/3-protein models) of four proteins (CCNB1/PAI1/PR/BCL2), subdivided lymph node-negative breast cancer patients into low-, medium- and high-risk groups with significantly different 10-year recurrence-free survival. The proteomic markers predicted 10-year distant metastasis-free survival in lymph node-negative patients in the test set in the low-, medium- and high-risk groups as follows: 0.9, 0.8, 0.7 (4-protein model ( p = 0.05)), 0.91, 0.8, 0.74 (3-protein model ( p = 0.004)), with 74%/64% patients in the low-risk groups, respectively. The proteomic models outperformed clinical variables and genomic predictors in multivariate analyses. Conclusion: This study validates proteomic biomarkers that can be assayed in a practical inexpensive manner using immunofluorescence to identify lymph node-negative ER/PR-positive patients with excellent outcomes after adjuvant tamoxifen. Disclosure: All authors have declared no conflicts of interest.

250PD

CLINICAL SIGNIFICANCE OF MICRORNA EXPRESSION AS A PROGNOSTIC FACTOR IN EARLY N+ BREAST CANCER (BC)

E. Ciruelos1, G.A. De Velasco1, A. Gamez2, C. Castañeda3, J. Sepúlveda1, I. Ghanem1, H. Cortes-Funes1, J.A. Fresno4 1 Medical Oncology, University Hospital 12 de Octubre, Madrid, SPAIN, 2 Medicine, Autonoma University, Madrid, SPAIN, 3Medical Oncology, Instituto de Enfermedades Neoplasicas, Lima, PERU, 4Medical Oncology, University Hospital LaPaz, Madrid, SPAIN Background: MicroRNAs (miRNAs) are small noncoding RNA molecules that post-transcriptionally regulate gene expression. In the present study, we describe miRNA expression in early node-positive BC and identify a 8-miRNA score that could contribute to prognosis assessment. Methods: First, microarray analysis was performed using RNA samples extracted from primary breast cancer. The expression of miRNAs was analyzed by RT-qPCR using the TaqMan Arrays from Applied Biosystems. After a suitable normalization of the 677 miRNAs analyzed, 96 presented a good correlation between their expression

ix | Abstracts

251PD

EUROPEAN INTER-INSTITUTIONAL IMPACT STUDY OF MAMMAPRINT

P. Cusumano1, D.G. Generali2, E. Ciruelos3, L. Manso3, I. Ghanem3, E. Lifrange4, G. Jerusalem4, J. Klaase5, L. Stork-Sloots6, M. Lutke Holzik7 1 Surgical Oncology, CHU Sart Tilman, Liege, BELGIUM, 2Unit, Istituti Ospitalieri di Cremona, Cremona, ITALY, 3Medical Oncology, Hopital 12 de Octubre, Madrid, SPAIN, 4Hematology and Medical Oncology, University of Liege, CHU Sart Tilman, Li, BELGIUM, 5Surgery, Medisch Spectrum Twente, Enschede, NETHERLANDS, 6Medical Affairs, Agendia NV, Amsterdam, NETHERLANDS, 7 Surgical Oncology, Saxenburgh Groep Hardenberg, Hardenberg, NETHERLANDS Background: Medical board interpretation of prognostic and predictive factors and quality of IHC analysis can substantially affect the adjuvant treatment advice, resulting in substantial differences in treatment recommendations; with risk of potential over- and under treatment. Introduction of a centralized multi-gene assay could potentially lead to more formalized treatment recommendations throughout Europe. The primary aim of this study was to measure the impact of MammaPrint on chemotherapy treatment advice for patient cohorts from 3 different European institutes. Methods: Between 2008 and 2010, tumor samples from early stage breast cancer patients (T1-3, N0-1, M0, < 70 years and MammaPrint assessed) were prospectively collected in 3 European hospitals: 92 patients from Belgium, 36 patients from Italy and 66 patients from the Netherlands. Blinded for the actual treatment advise, the clinical characteristics of the patients were sent to the multidisciplinary teams of the 3 contributing hospitals and a team of a fourth hospital and treatment advise was assessed. Subsequently, MammaPrint results were disclosed and treatment advice was assessed again. Discordance in treatment advise between the 4 multidisciplinary teams was determined and the impact of MammaPrint was measured. Results: Treatment advice as assessed by the Spanish multidisciplinary team was changed in 27% of patients after disclosure of MammaPrint results, in 33% of patients as assessed by the Dutch team, in 26% of patients by the Italian team and in 22% of patients by the Belgian team. The inter-institutional agreement for the four multidisciplinary teams in treatment advice was 57% ( prior to MammaPrint disclosure) and increased to 79% after disclosing MammaPrint results. Conclusion: This study demonstrates high variability in adjuvant chemotherapy treatment advice between multidisciplinary teams from 4 different European hospitals based on traditional patient and tumor related parameters. MammaPrint would have modified adjuvant chemotherapy treatment advice in around 27% of patients and lead to a more consistent and objective treatment advice. Disclosure: L. Stork-Sloots: I am an employee of Agendia. All other authors have declared no conflicts of interest.

Volume 23 | Supplement 9 | September 2012

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

B. Hennessy1, D. Faratian2, Z. Ju3, A. Lluch-Hernandez4, S. Myhre5, A. M. Gonzalez-Angulo6, J. Overgaard7, J. Alsner7, A. Borresen-Dale5, G.B. Mills8 1 Medical Oncology, Beaumont Hospital, Dublin, IRELAND, 2Division of Pathology & Edinburgh Breakthrough Research Unit, University of Edinburgh, Edinburgh, UNITED KINGDOM, 3Biostatistics and Applied Mathematics, MD Anderson Cancer Center, Houston, TX, UNITED STATES OF AMERICA, 4 Serv. Hematologia Y Oncologia Medica, Hospital Clinico Universitario de Valencia, Valencia, SPAIN, 5Department of Genetics, Institute for Cancer Research, Norwegian Radium Hospital, Rikshospitalet University Hospital, Oslo, Norway, Oslo, NORWAY, 6Breast Medical Oncology, MD Anderson Cancer Center, Houston, TX, UNITED STATES OF AMERICA, 7Department of Experimental Clinical Oncology, Aarhus University Hospital, Aarhus, DENMARK, 8 Systems Biology, MD Anderson Cancer Center, Houston, TX, UNITED STATES OF AMERICA

in frozen and paraffin-embeded tissue. A supervised analysis was done in order to identify miR that could predict relapse. Results: 172 patients with node-positive chemotherapy-treated early BC were included in the present study. The median age at diagnosis was (53.7) years, 128 cases (74.4%) had positive estrogen receptor, and 163 patients (94,7%) received adjuvant chemotherapy. After a median follow up of 8,1 years, 71p (41.2%) relapsed. Supervised analyses identified 8 microRNAs (has miR-30e, miR-30a, miR-21, miR-210, miR-93, miR-150, miR-99b, miR-572) associated with higher risk of relapse (5 year PFS 50% vs 90%, HR 3.75, 95% CI 2.27 – 6.19) and with estimated overall survival (median not reached, HR 4.51, 95% CI 2.36 – 8.61). Interestingly, miRNAs defined 2 prognostic groups (high and low risk) regardless of type of adjuvant chemotherapy (with or without anthracyclines) and histological subtype (luminal A, B or triple negative). Conclusions: This study suggests that miRNA expression could contribute to assess molecular prognosis of breast cancer and identifies clusters of miRNAs that could improve outcome prediction. A validation study in node-negative and triple negative disease is planned. Disclosure: All authors have declared no conflicts of interest.

Annals of Oncology 252PD

META-ANALYSIS OF PROSPECTIVE EUROPEAN STUDIES ASSESSING THE IMPACT OF USING THE 21-GENE RECURRENCE SCORE ASSAY ON CLINICAL DECISION MAKING IN WOMEN WITH ER-POSITIVE, HER2-NEGATIVE EARLY STAGE BREAST CANCER

J. Albanell1, S. Holt2, J. Gligorov3, W. Eiermann4, C. Svedman5 1 Medical Oncology, University Hospital del Mar, Barcelona, SPAIN, 2Surgical Department, Prince Philip Hospital, Llanelli, UNITED KINGDOM, 3Medical Oncology Department, Tenon Hospital, Paris, FRANCE, 4Medical, Interdisciplinary Oncology Center, Munich, GERMANY, 5Medical, Genomic Health Inc, Redwood City, CA, UNITED STATES OF AMERICA

253PD

SECRAB (SEQUENCING OF CHEMOTHERAPY AND RADIOTHERAPY IN ADJUVANT BREAST CANCER) COSMESIS RESULTS

I.N. Fernando1, S.J. Bowden2, C. Brookes2, A.M. Brunt3, M. Churn4, J.H. Steven1, R.K. Agrawal5, D. Rea2 1 Cancer Centre, University Hospitals Birmingham NHS Foundation Trust, Birmingham, UNITED KINGDOM, 2CRCTU, School of Cancer Sciences, University of Birmigham, Birmingham, UNITED KINGDOM, 3The Cancer Centre, University Hospital North Staffs, Stoke-on-Trent, UNITED KINGDOM, 4Deansley Centre, New Cross Hospital, Wolverhampton, UNITED KINGDOM, 5Department of Oncology, Shrewsbury and Telford Hospital NHS Trust, Shrewsbury, UNITED KINGDOM Background: SECRAB, a randomised trial comparing sequential (Seq) to synchronous (Syn) chemo-radiation using a CMF or Anthracycline-CMF regimen, demonstrates reduces local recurrence rates after Syn treatment for early breast cancer (EBC) (Fernando 2011 EJC:47 S2). Cosmesis study results are presented.

254PD

PATHOLOGICAL COMPLETE RESPONSE TO TRASTUZUMAB SUBCUTANEOUS FIXED-DOSE FORMULATION IN THE HANNAH STUDY: SUBGROUP ANALYSIS OF PATIENT DEMOGRAPHICS AND TUMOR CHARACTERISTICS AND INFLUENCE OF BODY WEIGHT (BW) AND SERUM TROUGH CONCENTRATION (CTROUGH) OF TRASTUZUMAB

B. Melichar1, D. Stroyakovskiy2, J.S. Ahn3, M.V. Kopp4, V. Srimuninnimit5, G. Kunz6, J. Li7, T. van der Horst8, S. Muehlbauer8, C. Jackisch9 1 Oncology, University Hospital Olomouc Palacky University, Faculty of Medicine, Olomouc, CZECH REPUBLIC, 2Chemotherapy, Moscow City Oncology Hospital 62, Moscow, RUSSIAN FEDERATION, 3Haematology/Oncology, Samsung Medical Centre, Seoul, KOREA, 4Chemotherapy, Samara Regional Clinical Cancer Center, Samara, RUSSIAN FEDERATION, 5Department of Medicine, Siriraj Hospital, Bangkok, THAILAND, 6Oncology, St Johannes Hospital, Dortmund, GERMANY, 7Clinical Pharmacology, Genentech Inc, South San Francisco, CA, UNITED STATES OF AMERICA, 8Clinical Science, F. Hoffmann La-Roche Ltd, Basel, SWITZERLAND, 9Oncology, Klinikum Offenbach GmbH, Offenbach, GERMANY Background: Intravenous (IV) trastuzumab (H) is the standard of care for HER2-positive breast cancer. The Phase III, neoadjuvant-adjuvant HannaH study (Jackisch et al. EBCC 2012) demonstrated the non-inferiority of H subcutaneous (SC) vs. H-IV with respect to serum Ctrough of H and pathological complete response ( pCR), with comparable safety profiles for the two formulations. The aim of the presented analyses was to determine whether pCR rates (H-SC vs. H-IV) were consistent across subgroups, and to investigate the relationship of pCR with BW and serum Ctrough of H. Methods: Subgroup analyses of pCR were performed in the per-protocol population (PPP) based on patient demographics (race, age [< 65 and ≥ 65 years]) and disease characteristics (breast cancer type/subtype/histological grade, hormone receptor status). Further analyses, including multiple logistic regression (MLR), investigated the relationship of pCR rate with BW and serum Ctrough of H. Covariates in the MLR model included treatment arm, serum Ctrough, BW, and all interactions between covariates, in order to account for the different dosing schemes (BW-based vs. fixed dosing). Results: pCR subgroup analyses showed that the numerically higher pCR rate point estimate in H-SC vs. H-IV was reflected in the majority of subgroups, with all confidence intervals for the rate difference (SC-IV) containing ‘0‘. The findings of the PPP analysis were supported by results obtained in the ITT population. Further analyses, including MLR, showed that neither BW nor serum Ctrough of H correlated with pCR rates in either treatment arm.

Table: 253PD Category Excellent

Cosmesis, n (%) Telangiectasia, n (%)

Seq

Syn

Seq

Syn

Seq

Syn

Odds Ratio (OR) (95% CI) Excellent/good vs moderate/poor

14 (23)

15 (21)

37 (60)

37 (51)

11 (18)

21 (29)

1.87, (0.82,4.27) p = 0.1

79 (68)

99 (62)

33 (28)

54 (34)

4 (4)

7 (4)

1.28, (0.37,4.48) p = 0.7

Good

Volume 23 | Supplement 9 | September 2012

Moderate/Poor

doi:10.1093/annonc/mds392 | ix

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

Background: The Oncotype DX® Recurrence Score is a validated assay to help inform the appropriate treatment of estrogen receptor-positive (ER+), early stage breast cancer. Adjuvant treatment traditions vary significantly among different countries. Prospective studies assessing the impact of the Oncotype DX test on adjuvant treatment decisions have been performed in several European countries. This is a meta-analysis of these studies. Methods: Four prospective studies assessing the impact of using the Recurrence Score result on clinical decision making in patients with node negative and pN1 (mi) disease were identified. Node positive patients were excluded. The identified studies had a similar study design with consecutive patients and treatment recommendations before and after having the Recurrence Score results were recorded. In three of the studies, medical oncologists completed questionnaires regarding their confidence in their recommendation before and after knowing the patient’s Recurrence Score result. The final results of the studies in Germany, Spain and the UK have been presented and final data from the French study will be presented at ASCO 2012. Results: A total of 589 patients with node negative or pN1 (mi) ER pos HER2 negative early breast cancer were included in the four identified studies. Overall, 45% (range 36-52%) of patients in these studies were recommended chemo-endocrine and 55% endocrine treatment alone. After having the Recurrence Score result, 47% (range 38-60%) of patients recommended chemotherapy were changed to endocrine treatment alone and 17% (range 11-22%) of those recommended endocrine treatment alone were recommended chemo-endocrine treatment. There was a significant improvement in physician confidence in treatment recommendations when using the assay. Conclusion: Using Oncotype DX® is associated with a significant change in treatment decisions and a reduction in chemotherapy use in studied European countries despite differences in therapeutic traditions. The consistency of the results from different countries underlines the tests utility. Disclosure: J. Albanell: Participated advisory board (Genomic Health). S. Holt: Participated as a speaker, and in advisory board for Genomic Health. J. Gligorov: Has participated as speaker for Genomic Health. W. Eiermann: Participated as speaker and in advisory board for Genomic Health. C. Svedman: Employee and stocks in Genomic Health.

Methods: 2296 women were randomised and cosmesis was assessed in 382. Cosmesis and telangiectasia were assessed as excellent, good, moderate and poor for patients having wide local excision (WLE) while telangiectasia alone was assessed for those having mastectomy. Photographs were taken for 301 patients for blinded independent review using a consensus method (Haviland 2008 Clin Oncol 20:497). Patient perception was assessed using EORTC BR23 quality of life questionnaire (Q 39-42). Assessments were made at baseline, 1, 2 and 5 years post surgery. Results: The results of clinician assessed cosmesis and telangiectasia are shown below. There was no statistically significant difference between the arms. Data on independent assessment of change in breast appearance after WLE was available for 145 (98%) patients. Minimal, mild and marked changes were observed in 70, 29 and 1% of patients respectively for the Syn arm and 72, 28 and 0% for the Seq arm. There was no difference between treatment arms (OR, mild/marked vs minimal, 1.10, 95% CI 0.54-2.26, p = 0.8). There was no change in patient’s perception of their breast appearance (baseline vs. final assessment questionnaire). Conclusions: There was no significant difference in cosmesis or telangiectasia between the two arms as assessed by the clinician or by independent photographic review. There was no difference in patient perception of breast appearance. Contrary to the result of the main study, which showed a borderline significant worsening of telangiectasia in the Syn arm, no such difference was seen in the cosmesis study. Syn chemo-radiation reduces local recurrence in EBC without affecting cosmesis. Disclosure: All authors have declared no conflicts of interest.

Annals of Oncology

Conclusions: BW and serum Ctrough of H did not impact on efficacy ( pCR). The 600 mg fixed dose of H-SC is efficacious irrespective of BW. This finding supports H-SC as a treatment alternative to the registered IV formulation. Disclosure: B. Melichar: Prof. Melichar has participated in advisory board meetings and received honoraria from F. Hoffmann-La Roche Ltd. J. Li: Jing Li is an employee of F. Hoffmann La-Roche Ltd, and owns stock in Roche Holdings. T. van der Horst: Tina van der Horst is an employee of F. Hoffmann-La Roche Ltd. S. Muehlbauer: Susanne Muehlbauer is an employee of F. Hoffmann-La Roche Ltd and owns stock in Roche Holdings. C. Jackisch: Christian Jackisch has participated in advisory board meetings for F. Hoffmann-La Roche Ltd. All other authors have declared no conflicts of interest.

255PD

CORRELATION BETWEEN CIRCULATING TUMOR CELLS (CTCS), PET/CT RESPONSE AND PATHOLOGICAL COMPLETE RESPONSE (PCR) IN PRIMARY HER2-POSITIVE (HER2+) BREAST CANCER PATIENTS: A SUB-STUDY FROM THE NEOALTTO TRIAL

Background: Although CTC detection has been studied in patients receiving preoperative chemotherapy, fewer data exist for preoperative chemotherapy combined with anti-HER2 agents. Here, we report a sub-study of CTC detection within the NeoALTTO trial. Methods: NeoALTTO is a randomized phase III trial in which patients with primary HER2+ breast cancer were randomized to trastuzumab, lapatinib or their combination for 6 weeks followed by the addition of paclitaxel for 12 weeks prior to surgery. Participation in the CTC sub-study was optional. Blood samples were prospectively collected at baseline, after 2 weeks of anti-HER2 treatment alone and prior to surgery. A total of 22.5mL of blood was reduced to 7.5mL using a modified ficoll procedure that was then processed using CellSearch®. Evaluation of HER2 expression in CTCs was performed using the CellSearch® HER2 profiling kit. CTCs with HER2 staining intensity of 2 + /3+ were considered as HER2+. Associations between CTC detection and primary tumor characteristics, pCR and PET/CT response (at week 2 & 6) were assessed using the chi-square test. Results: Out of 455 patients randomized, samples for CTC analysis were available for 51 (11%) patients from 16 sites, of whom 11 (21%) had ≥ 1 CTC/ 22.5ml in at least one time point (Table 1). At baseline, week 2 and surgery, we detected ≥1 CTC/22.5ml in 5/46 (11%), 4/41 (10%), and 5/31 (16%) patients with evaluable samples respectively. HER2+ CTCs were still detectable after 18 weeks of preoperative treatment with anti-HER2 agents plus paclitaxel in 3/31 (10%) patients. No significant association was observed between CTC detection (either at baseline or week 2 or surgery), and clinicopathologic characteristics, pCR ( p = 0.36) or PET/CT response at week 2 ( p = 0.34) or week 6 ( p = 0.90). Conclusion: CTC detection does not appear to be associated with pCR in patients receiving preoperative anti-HER2 agents plus paclitaxel, acknowledging the lack of power within our study.

Patient

CTC / HER2+ CTC Treatment arm (number per 22.5mL) pCR Baseline Week 2 Prior to Surgery

1 2 3 4 5 6 7 8 9 10 11

L L L L T T L+T L+T L+T L+T L+T

0 NA 1 /1 NA 0 65 /52 0 1 /0 1 /0 1 /0 0

0 2 /2 0 4 /3 1 /0 0 0 0 0 0 1 /0

2 /2 2 /2 0 NA 0 2 /1 1 /0 NA 0 NA 1 /0

L: lapatinib; T: trastuzumab; NA: not applicable. Disclosure: All authors have declared no conflicts of interest.

ix | Abstracts

NO NO NO NO NO NO NO NO YES YES YES

FIVE YEARS FOLLOW UP OF PROSPECTIVE RANDOMIZED STUDY TO ASSESS THE EFFECT OF PROPHYLACTIC CRANIAL IRRADIATION IN HIGH RISK BREAST CANCER PATIENTS

E.A. Elkhouly, H.M. Metwally, T.A. Hashem, K.K. Abdalaziz Clinical Oncology Dep., Menoufia University Hospital, Menoufia, EGYPT Aim: To assess the effect of Prophylactic cranial irradiation (PCI) in patients with high risk early breast cancer with primary end point CNS relapse where the secondary end point is to assess the toxicity and QOL. At 5 years the aim is to assess CNS relapse, Overall Survival, NCF and QOL re-evaluation. Methods: It is 5 years re-evaluation of 62 patients with high risk invasive breast cancer. After adjuvant treatment, patients were randomized into: arm I who didn’t receive PCI, arm II who received PCI. Patients were collected from Jan. 2005 and followed up till Dec. 2007 with re-evaluationone on April 2012. After a signed informed consent, the whole brain received 240 cGy per fraction, once daily, five times per week to a total dose of 2400cGy. Brain assessment by MRI studies as a baseline before PCI, at 6 months in the first year, and then annually for at least 5 years, and Assessment of Neuro-cognitive Function (NCF): basically before PCI, 6 months, 1 year after PCI using Mini-Mental State Exam (MMSE). Health related quality of life was assessed for all patients with the Functional Assessment of Cancer Therapy-Brain (FACT-Br) Questionnaire Version 4, before PCI and 3 weeks, three months after PCI. Patients were followed up till April 2012. Results: At the end of the study, there is no incidence of CNS metastasis in PCI arm compared to 3.2% in control arm. In PCI arm, there is no statistical difference in MMSE (before and after PCI, p = 0.619 ) and QOL assessment. At 5 years, in PCI arm, the incidence of CNS metastasis is 3.2 (only one case) while in Control arm it is 12.9 (four cases). In PCI arm,there was no statistically significant change in NCF and QOL in all items of FACT-Br questionnaire except in additional concerns ( p <0.01). There are no statistical differences between Median survival of PCI arm( 51 month) and Control arm ( 52.5 month) as p- value > 0.05, Log rank of 5 years survival = 0.59 Conclusion: The rationale behind PCI is to control or eradicate undetectable micro-metastases before they become clinically significant without inducing severe adverse effects .PCI associated with decreased incidence of CNS metastases with tolerated toxicities and did not negatively affect patient survival. This may be translated into therapeutic gain at least for this short follow up period which warrants further evaluation after a longer follow up period. Disclosure: All authors have declared no conflicts of interest.

257P

EGFR AND P53 EXPRESSION IN ANDROGEN RECEPTOR (AR)-POSITIVE, TRIPLE NEGATIVE BREAST CANCER (TNBC)

A. Gucalp1, G. Gupta2, S. Patil3, Y.H. Wen4, M. Akram4, E. Brogi4, S. Powell2, A. Ho2, C.A. Hudis1, T.A. Traina1 1 Dept of Medicine, Memorial Sloan-Kettering Cancer Center, New York, NY, UNITED STATES OF AMERICA, 2Radiation Oncology, Memorial Sloan-Kettering Cancer Center, New York, NY, UNITED STATES OF AMERICA, 3Biostatistics and Epidemiology, Memorial Sloan-Kettering Cancer Center, New York, NY, UNITED STATES OF AMERICA, 4Dept of Pathology, Memorial Sloan-Kettering Cancer Center, New York, NY, UNITED STATES OF AMERICA Background: TNBCs are a heterogeneous group characterized by lack of ER/PR/ HER2 expression. Doane et al described a subgroup of TNBC associated with AR-dependent growth. While expression of EGFR and p53 are common in TNBC, their role in AR+ TNBC is less clear. We report the rates of EGFR and p53 expression and clinicopathologic features of AR+ TNBC from a retrospective cohort at MSKCC. Methods: We identified 1,032 patients ( pts) with resectable, TNBC (ER/PR < 1%; HER2 < 2 + /FISH < 2) who had surgery at MSKCC (1998-2006). Exclusions: neoadjuvant chemotherapy, prior XRT, inflammatory/metastatic BC. TMAs were constructed from 210 readily available primary tumors (> 1 cm) with each tumor represented by 3 cores. AR was tested with DAKO Clone AR441, dilution (dil) 1:500; ratio of DAB nuclear staining to hematoxylin signal >1 SD above mean was defined as AR+. EGFR and p53 were tested with Thermo-Scientific Clone EP38Y, dil 1:50 and DAKO Clone D07, dil 1:50 respectively. Scoring: 0-1+ negative, 2-3+ positive. Fisher’s exact test used to evaluate correlation between AR with EGFR and p53. RFS and OS evaluated using Kaplan-Meier methods. Clinicopathologic variables by AR status were compared using Chi-square/t-tests. Results: 166 pts had adequate cores for AR testing. 10% were AR+ (17/166). 160 pts were evaluable for EGFR and 164 pts for p53. Median (med) followup: AR + =6 years (y), AR- = 5.6y. Adjuvant chemotherapy received: AR+ 82%, AR- 87%, P = 0.40. EGFR was expressed more frequently in AR- than AR+ TNBCs (111/143 (78%) vs 9/ 17 (53%); P =0.04). p53 expression was similar in AR- and AR+ TNBC (82/147 (56%) vs 7/17 (41%); P =0.30). Clinicopathologic variables based on AR and EGFR status appeared similar. 3y RFS and OS data are below. Conclusion: Consistent with published reports, loss of AR expression correlates with EGFR expression. In contrast, no association was observed with p53 and AR. These

Volume 23 | Supplement 9 | September 2012

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

H.A. Azim, Jr1, F. Rothe1, C.M. Aura2, M. Bavington3, M. Maetens1, G. Rouas1, E. De Azambuja4, C. Sotiriou1, S. Di Cosimo5, M. Ignatiadis1 1 Breast Cancer Translational Research Laboratory, Institute Jules Bordet, Brussels, BELGIUM, 2Molecular Pathology Laboratory, VHIO, Barcelona, SPAIN, 3 Programming, Frontier Science, Kincraig, UNITED KINGDOM, 4Breast Data Centre, Institute Jules Bordet, Brussels, BELGIUM, 5Medical Oncology, Istituto Nazionale Tumori, Milan, ITALY

256P

Annals of Oncology

data are hypothesis generating regarding the mechanism of interaction of AR and EGFR in TNBC.

Disclosure: S. Haouas: Grant from Association pour la Recherche en Thérapeutiques Innovantes en Cancérologie (ARTIC). All other authors have declared no conflicts of interest.

-3 year RFS (95% CI) -3 year OS (95% CI) AR+ ARAR+ EGFR+ (n = 9) AR+ EGFR- (n = 8) AR- EGFR+ (n = 111) AR_ EGFR- (n = 32)

76% 81% 67% 88% 82% 81%

(47-90) (74-87) (39-98) (39-98) (73-88) (61-92)

88% (59-97) 93% (87-96) 78% (36-94) 100% 91% (84-95) 97% (80-100)

CLINICAL BEHAVIOR AND PROGNOSIS OF DIFFERENT IMMUNOHISTOCHEMISTRY-DETECTED SUBTYPES OF INVASIVE BREAST CANCER: A MONOINSTITUTIONAL SERIES

A. Ferro1, C. Eccher2, R. Triolo1, A. Caldara1, M.C. Di Pasquale1, S. Moroso1, L. Russo1, M. Barbareschi3, E. Galligioni1 1 Medical Oncology, Santa Chiara Hospital, Trento, ITALY, 2E-health, FBK, Trento, ITALY, 3Pathology, Santa Chiara Hospital, Trento, ITALY

Disclosure: All authors have declared no conflicts of interest.

258P

259P

PROGNOSTIC FACTORS IN EARLY-STAGE TRIPLE NEGATIVE BREAST CANCER (TNBC): THE LIMITS OF CLINICAL AND PATHOLOGICAL FEATURES

Background: Triple negative breast cancer (TNBC) patients ( pts) is an heterogeneous population regarding prognostic. Therefore, clinical and histological features were evaluated in a large monocenter cohort of patients treated for localized TNBC to identify good-prognostic TNBC. Methods: All consecutive early-stage TNBC (ER 0%, PR 0%, HER2 neg) patients treated at European Georges Pompidou Hospital, Paris, France, between 2000 and 2011 were included. Records were reviewed for demographic, clinical and pathological data. Prognostic factors were determined using univariate and multivariate stepwise log-rank analysis on disease-free survival (DFS) and overall survival (OS). Results: This analysis included 128 women with early-stage TNBC. Clinical and histological characteristics are summarized below. After a median follow-up of 37 months 36 relapses and 19 deaths were observed. The 3-years recurrence rate was 30% (95%CI 22-40) in the whole population. For DFS, bad prognostic factors in univariate analysis were: large tumor size (T3-4), node involvement (N1-3), node capsular effraction, lymphovascular invasion (LVI) and high grade (SBR 2-3). Multivariate analysis identified tumor size (T3-4) (HR 3.70, 95%CI 1.61-8.52) and LVI (HR 2.87, 95%CI 1.39-5.93) as independent factors. The 3-years recurrence rate remained high (20%, 95%CI 10-34) in patients with two good prognostic factors (T0-2 and no LVI). For OS, bad prognostic factors significant in univariate and multivariate analysis were: node capsular effraction (HR 3.57, 95%CI 1.17-10.92) and LVI (HR 3.43, 95%CI 1.18-9.92). Conclusions: In this early-stage TNBC series, LVI was a bad prognostic factor of relapse and death. However, the 3-years recurrence rate remained high in patients with good prognostic features. Therefore, new biomarkers are mandatory for a better stratification of this heterogeneous population. Clinical and histological characteristics

260P

PERITUMORAL VASCULAR INVASION AS PRINCIPAL ISTOLOGICAL PROGNOSTIC FACTOR IN TNBC

G. Rizzo, V. Fregoni, G. Daprada, L. Pavesi, G. Gallizzi, I.T. Lorenzetti, R. Tancredi, E. Ferraris, A. Pagani Oncology Division, Fondazione Salvatore Maugeri, Pavia, ITALY

N (128pts) % Median age (yrs) Histological type Invasive ductal carcinoma Invasive lobular carcinoma Tumor stage T0-T2 T3-T4 Node stage N0 N1-3 Node capsular effraction pos neg NE Tumor grade (SBR) 1 2-3 NE* Lymphovascular invasion (LVI) pos neg NE** *Due to neoaduvant chemotherapy **Due to small biopsies.

Volume 23 | Supplement 9 | September 2012

56 range 22-88 123 5

96 4

105 23

82 18

97 31

76 24

14 111 3

11 87 2

8 107 13

6 84 10

34 74 20

27 58 16

Purpose: Triple-negative breast cancers (TNBC) is a specific histological subclass of breast cancer that has gained attention in recent years for its clinical and molecular heterogeneity with still no evidence of an optimal therapeutic strategy. Aim of this study is to identify clinical and pathological features of TNBC reviewed in our center and try to figure out prognostic factors that can drive therapeutic approach. Methods: In this retrospective study we reviewed 127 cases of TNBC (median Age 59) that underwent surgical treatment from 2003 to 2008 analyzing the outcome (in terms of Disease Free Survival) in relation to clinical and pathologic features. Results: Univariate analysis performed on the entire cohort revealed that the cha Staging (P <0.0001, HR = 7.8245 [1.59-38.27]), the lympho-plasmacytic peritumoral reaction (P = 0.01, HR: 0.3160 [0.1633 to 0.6116] ), node involvment (P = 0.0241, HR: 1.9083 [1.0210 to 3.5667]) and peritumoral vascolar invasion (PVI) (P <0.0001, HR: 4.0261 [1.6608 to 9.7602]) are all prognostic factor of DFS. Multivariate analysis confirmed, in particular, PVI, observed in 24% of cases, as the main prognostic factor (P = <0.0001, HR 3.5655 [1.2-10.3]). Conclusions: In the TNBC several genes involved in angiogenesis are frequently pointed out. In our hands peritumoral vascular invasion is a major prognostic factor that would identify a subgroup of TNBC with worse prognosis and therefore this would require a more aggressive therapeutic regimen with the inhibition of angiogenesis as a possible tool for the optimal treatment. Disclosure: All authors have declared no conflicts of interest.

doi:10.1093/annonc/mds392 | ix

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

N. Pécuchet1, M. Le Frère Belda2, T. Popovski3, S. Haouas1, F. Chamming’S4, F. Lécuru5, S. Oudard1, J. Medioni1 1 Dept. Medical Oncology, Hopital European George Pompidou, Paris, FRANCE, 2 Dept. Pathology, Hopital European George Pompidou, Paris, FRANCE, 3 Gynecology, Necker University Hospital, Paris, FRANCE, 4Dept. Radiology, Hopital European George Pompidou, Paris, FRANCE, 5Dept. Gynecologic Surgery, Hopital European George Pompidou, Paris, FRANCE

Background: Invasive breast cancer (IBC) is a heterogeneous disease. Gene expression profiling has identified several biologically distinct subtypes of IBCs. As proposed by Cheang et al, immunohistochemical (IHC) markers can be used as a surrogate for the molecular classification of IBC. Purpose: To evaluate different clinical behavior, relationship with other clinical-pathological features and survival outcomes of patients ( pts) with different subtypes of IBC as classified using IHC markers. Methods: We evaluated 3461 IBC pts treated from 1995 to 2008 classified as: luminal A (ER and PR + , HER2- and Ki67< 14%), luminal B (ER and/or PR + , HER2- and Ki67 ≥ 14%), luminal C (ER and/or PR + , HER2 + , any Ki67), HER2+ (ER and PR, HER2 + , any Ki67), triple negative-TN (ER and PR-, HER2-, any Ki67). Log-rank test and Cox regression model were performed to evaluate the impact of IHC subtypes on overall survival (OS), Event Free Survival (EFS) and their correlation with other known prognostic factors. Results: We identified 909 (26.4%) luminal A, 1722 (49.9%) luminal B, 325 luminal C (9.4%), 209 (5.7%) HER2+ and 296 (8.6%) TN. Median age was 61 years. Luminal A was more frequently associated with older age, smaller size, negative axilla involvement, low grade ( p < 0.001). There were 644 (18,6%) events (local and distant relapses, contralateral and second tumors): 100 in luminal A (11%), 334 in luminal B (19.4%), 62 in luminal C (19%), 66 in HER2+ (31%) and 84 in TN (28%). Different subtypes showed preferential sites of first local or distant relapses: luminal A had more bone and loco-regional and less visceral relapses than other subtypes. Median disease free interval (DFI) was longer in luminal A (60.3 months) than in luminal B (39.1 m), C (27.8 m), HER2 (30.8 m) and TN (23.3 m). At median follow up of 77 months, EFS and OS were 94.0 and 92.1% in luminal A, 86.2 and 82.7% in luminal B, 86.2 and 86.8% in luminal C, 72.7 and 72.7% in HER2 + , 78.0% and 73.9% in TN ( p < 0.001). Luminal A presented the best prognosis among other luminal subtypes. IHC-based subtypes prognosis (EFS and OS) was independent of nodal status, grading, tumor size and age. Conclusions: In our experience IHC-based classification appared to be useful to divide IBCs in different biological entities. Its application could help the tailoring of adjuvant therapies improving patient outcomes. Disclosure: All authors have declared no conflicts of interest.

Annals of Oncology

261P

BREAST CANCER SUBTYPES AND OUTCOME IN ELDERLY BREAST CANCER PATIENTS AGED 70 YEARS AND OLDER

R. Königsberg1, G. Pfeiler2, N. Hammerschmid2, T. Klement2, A. Brunner3, C. Dittrich1 1 Ludwig Boltzmann Institute for Applied Cancer Research (lbi-acr Vienna) – LB Cluster Translational Oncology, 3rd Medical Department – Centre for Oncology and Haematology, Kaiser Franz Josef-Spital, Vienna, AUSTRIA, 2Department of Obstetrics and Gynaecology, Medical University of Vienna, Vienna, AUSTRIA, 3 Department of Obstetrics and Gynaecology, Landesklinikum Thermenregion Moedling, Mödling, AUSTRIA

262P

OUTCOMES OF ETHNIC MINORITY GROUPS WITH NODE-POSITIVE, NON-METASTATIC BREAST CANCER IN A TERTIARY REFERRAL CENTRE IN SYDNEY

S.H. Lim1, J. Descallar2, P. Sayaloune3, G. Papadatos4, P. De Souza1, G. Delaney4 1 Medical Oncology, Liverpool Hospital, Liverpool, NSW, AUSTRALIA, 2Statistics, Ingham Institute for Applied Medical Research, Liverpool, NSW, AUSTRALIA, 3 Data Management, Liverpool Hospital, Liverpool, NSW, AUSTRALIA, 4Radiation Oncology, Liverpool Hospital, Liverpool, NSW, AUSTRALIA Background: South-Asian women in Europe are significantly younger than non-Asian at diagnosis, present with larger tumours, and have higher mastectomy rates. In Australia, there is a paucity of data on early breast cancer in ethnic minority groups. The aim of this study is to examine racial differences in tumour characteristics and outcome in early node-positive breast cancer patients at Liverpool and Macarthur Cancer Therapy Centres in New South Wales (NSW), Australia. These tertiary centres service 20% of the NSW population, with 40% from a non-English speaking background. Methods: A retrospective review of patients referred to Liverpool and Macarthur with node-positive non-metastatic breast cancer diagnosed before November 2006. Variables included tumour size, number of positive nodes, histological grade, hormone receptor status, age at diagnosis, country of birth and treatment. Outcome was disease-free survival. Women of Australian/New Zealand or European backgrounds were classed as Caucasian; women from Indian sub-continent, East Asia, South-east Asia, Middle East and Polynesia were classed as Asian. The study was approved by ethics. Results: There were 644 patients, with median follow up of 6.1 years. The proportion of Australian/New Zealand, European, South/East/South-east Asian, Middle-Eastern and Polynesian populations comprised 48%, 20%, 12%, 6% and 2% respectively. Women with Asian backgrounds were significantly younger at presentation (48 vs 55 years), and more likely to undergo mastectomy (46% vs 34%) and chemotherapy (82% vs 68%) than non-Asians. Tumour grade, stage and receptor status were not statistically different between the two groups. Median age, tumour size, number of nodes involved, disease-free survival and overall survival, respectively, was 52.5 years, 22mm, 2/19, 148 months and 177 months. Multivariate analysis identified risk factors of tumour size > 2cm (hazard ratio HR = 1.51, confidence interval CI

ix | Abstracts

263P

OMISSION OF SURGERY IN SELECTED ELDERLY PATIENTS WITH EARLY BREAST CANCER DOES NOT INCREASE CANCER SPECIFIC MORTALITY

M. Hamaker1, E. Bastiaannet2, C. Smorenburg3, G. Liefers2, S. De Rooij4, J.E. A. Portielje5 1 Geriatric Medicine, Diakonessenhuis, Zeist, NETHERLANDS, 2Surgery, LUMC, Leiden, NETHERLANDS, 3Internal Medicine, Medisch Centrum Alkmaar, Alkmaar, NETHERLANDS, 4Internal Medicine, AMC, Amsterdam, NETHERLANDS, 5Interne Geneeskunde, Hagaziekenhuis van Den Haag, The Hague, NETHERLANDS Aim: to determine the clinical characteristics and cause of death for a multicentre cohort of elderly patinets with resectable breast cancer for whom surgery was omitted. Methods: Chart review of 187 patients aged 75 years or older with early stage primary breast cancer, diagnosed at five centres in the Netherlands between 1990 and 2005 who did not receive primary surgery. Clinical, social, and functional characteristics as well as reason for omission of surgery were collected; comorbidity burden was scored using the Charlson index. Cause of death was retrieved from a national database containing death certificate data. Results: Median age was 85.9 years (range 75.0-97.7 years). Median Charlson score was 1 (range 0-7). For 32% omission of surgery was at the patient’s request; other reasons were age (4%), comorbidity (6%), overall health satus (5%), or not specified (53%). Thirteen patients did not receive any treatment (8%); all others received hormonal therapy. Seve patients (4%) required secondary surgical intervention due to disease progression (median time to surgery 1 year). By December 2010, 17 patients had died (95%); 34% died of breast cancer and for 21% of patients dying of other causes, breast cancer was mentioned as a potential underlying cause of death. For the remaining patients, breast cancer was not mentioned as a potential cause of death or as an underlying disease. Median survival was 2.1 years (range 0.2-10.7 years) and did not differ between patients dying of breast cancer versus all other causes of death ( p = 0.9). Risk of dying of breast cancer decreased with increasing comorbidity ( p = 0.06) while age, stage and receptor status were not associated with cancer specific mortaliyt. Conclusion: Omission of surgery for elderly patients with early breast cancer may not increase cancer specific mortality and in a significant number of patients, breast cancer was not considered clinicaly relevant at the time of death. As in one-third of cases conservative treatment was at patient’s request, future studies should focus on providing women with the information that they consider most relevant, such as quality of ife and functional outcomes, enabling well-informed decision making. Disclosure: All authors have declared no conflicts of interest.

264P

IS THERE ANY DIFFERENCE IN CLINICOPATHOLOGICAL FACTORS AND PROGNOSIS OF THE YOUNG BREAST CANCER?

Y. Kajiura1, Y. Ishida2, H. Yagata1, N. Hayashi1, A. Yoshida1, H. Yamauchi1 1 Breast Surgical Oncology, St. Luke’s International Hospital, Tokyo, JAPAN, 2 Center for Clinical Epidemiology, St. Luke’s Life Science Institute, Tokyo, JAPAN Background: Young breast cancer patients have been increasing in Japan over the last several decades and have been reported to have poorer prognosis than older women. However, differences in prognosis for young patients remain uncertain. The objective of this study is to compare clinicopathological factors and prognosis between breast cancer patients under 35 years versus those 35-39 years of age. Patients and methods: Sixty hundred eighty four women under 40 years were diagnosed breast cancer in our hospital from 1994 to 2008. Patients with bilateral breast cancer, without surgical treatment in our hospital, those with incomplete clinical data, and patients who received neoadjuvant chemotherapy were excluded from further analysis. The remaining 400 women were enrolled in this study. We first assessed clinicopathological factors associated with prognosis in all enrolled patients. Clinicopathological factors included family history of breast cancer, palpated tumor size and lymph nodes, pathological findings (histological type, tumor size, lymph node metastasis, nuclear grade, vessel invasion, lymphatic invasion, and ER, PgR, HER2 status), surgical procedure (mastectomy or breast conserving surgery), postoperative radiation therapy and adjuvant therapy. Next, we compared clinicopathological factors, disease-free survival (DFS) and overall survival (OS) between patients under 35 years versus those 35-39 years of age.

Volume 23 | Supplement 9 | September 2012

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

Background: Tumor characteristics and patterns of recurrence in luminal A, luminal B, Her2 pos and triple negative breast cancer (bc) patients ( pts) 70 years and older are still an area of uncertainty. In this investigation clinical pathological characteristics, disease recurrence and death were analysed according to the 4 bc subcategories and age in pts ≥70 years. Methods: Data of 274 elderly bc pts (≥70 years of age) diagnosed between 1998 and 2004 were retrospectively analyzed by computer based chart information. Baseline tumor characteristics, patient demographics and patterns of recurrence were compared between luminal A, luminal B, Her2 pos and triple negative bc pts and 3 age categories (70-75, 76 -80, ≥ 81 years). Results: Mean age was 76.95 years and mean time of follow-up was 52.23 months (range 0-144 months). Overall, recurrences were detected in 18.4% of pts. At the end of follow-up 69.0% of pts were alive, 15.0% and 16.1% died from bc and from other causes, respectively ( p = 0.745). Median overall survival (OS) was 114.1 months (95% confidence interval 100.3 -181.7 months). 41.1%, 20.4%, 16.1% and 12.5% of patients had luminal A, luminal B, Her2 pos and triple negative bc, respectively. No significant differences were found regarding recurrences according to bc subtypes ( p = 0.241). Triple negative bc had a significant shorter OS compared to lumianl A, luminal B and Her2 pos bc (72.7 months vs 124.1, 119.4 and 105.0 months, respectively ( p < 0.001)). Overall, 21.5% of pts received no therapy. Pts aged ≥81 years (15.8%) received significantly less ( p < 0.01) chemo-/antihormonal therapy compared to pts ≤80 years. Pts aged 70-75 years (42.7%) had significantly ( p = 0.032) more recurrences compared to pts 76 years or older. Pts aged 70 - 75, 76 - 80 and ≥ 81 years had median OS of 116.9, 115.4 and 73.1 months. Conclusions: Pts aged 70-75 years had significantly more recurrences; perhaps due to their lower risk of death from other causes. These data emphasize the necessity to foresee all bc subgroups of the elderly for adjuvant treatment irrespective of age but carefully considering co-morbidities, performance status and life-expectancy. Disclosure: All authors have declared no conflicts of interest.

1.06-2.41, p value 0.02), > 10 lymph nodes involved (HR = 2.64, CI 1.75-3.98, p-value <0.0001) and receptor positivity (HR = 0.40, CI 0.26-0.60, p-value <0.0001). Conclusion: Characteristics and outcome in our cohort are consistent with the literature. The Asian ethnic minority population had a younger age at diagnosis, higher mastectomy and adjuvant chemotherapy rates. Disclosure: All authors have declared no conflicts of interest.

Annals of Oncology

Results: Of the 400 patients, 139 (34.8%) were under 35 years old and 261 (65.2%) were 35-39 years old with a median follow-up of 72.5 months (range,1-211months). Postoperative radiation therapy was associated with longer DFS (HR, 0.130; p = 0.007) and OS (HR, 0.017; p = 0.001) in multivariate analysis of the 400 patients. However, there were no significant differences in clinicopathological factors, DFS and OS between under 35 years and those 35-39 years of age. Conclusion: Our results suggested that postoperative radiation therapy improved the prognosis in young breast cancer patients. However, age at diagnosis was not associated with DFS or OS in patients under 40 years. Therefore, we recommend patients under 40 years to receive unified therapy. Disclosure: All authors have declared no conflicts of interest.

265P

267P

BREAST CANCER IN YOUNG WOMEN IN INDIA

S.P. Deshmukh1, A.D. Mane1, B.P. Zade2, S.P. Sane3 1 Surgical Oncology, Ruby Hall Clinic, Pune, INDIA, 2Radiation Oncology, Ruby Hall Clinic, Pune, INDIA, 3Statistics, Ruby Hall Clinic, Pune, INDIA

TUMOR CHARACTERISTICS, TREATMENT AND OVERALL SURVIVAL (OS) IN BREAST CANCER PATIENTS (PTS) OVER 80 YEARS: A COHORT FROM A SINGLE INSTITUTION

C. Baldini1, A. Donnadieu1, A. Kramar2, V. Servent1, J. Bonneterre3 1 Breast Cancer, Centre Oscar Lambret, Lille, FRANCE, 2Biostatistics, Centre Oscar Lambret, Lille, FRANCE, 3Breast Cancer, Centre Oscar Lambret, Université Lille Nord de France, Lille, FRANCE The incidence rate of breast cancer in elderly women over 75 years old is 220 per 100000 in France and is likely to increase in the future. We studied the impact of tumor and treatment characteristics on overall survival. Patients and methods: Data were collected for 208 non metastatic breast cancer pts over 80 years (213 tumors), treated from January 2000 to December 2009. Pathological data were obtained on a biopsy or on mastectomy specimens if patients had not received presurgical treatment. Statistical methods: Survival rates were estimated by Kaplan-Meier method. Tests were performed with the logrank test. Five year relative survival rates were estimated by Ederer2 method. Results: Median age at diagnosis was 83 years (80-95). 29% tumors were T1, 58% T2 and 15% T3, 141 (66%) were ductular, 34 (16%) lobular and 38 (18%) had another histological type. SBR grade (I/II/III) was 22%/49%/29% and unspecified in 42 (20%). 167 pts (80%) were ER + , 129 (62%) PR + ; 40 (19%) ER–PgR–, and 171 (81%) ER + PgR+. HER2 status, known in 126 pts, was positive in 18 (14%), and negative in 108 (86%). Ten pts were strictly triple negative (5%)(ER = 0, PgR = 0, HER2 = 0). Ki 67 among 59 tumors was over 15% in 43 pts (73%). Forty-nine pts (23%) had conservative surgery, 80 mastectomy (38%), 72 axillary dissection (34%) and 30 sentinel lymph node biopsy (14%). Axillary lymph node metastasis were observed in 52 pts (38%). 195 pts received hormonotherapy (68 as adjuvant therapy, 89 before surgery or alone), 72 patients radiotherapy (37%), 5 chemotherapy (3%) and 1 had trastuzumab (0.5%). Median overall survival was 59 months. Median follow up was 76 months. The only significant prognostic factors on overall survival were age (48 months over 83 years, 79 before), ER and PgR status. ER positive and

Volume 23 | Supplement 9 | September 2012

DISEASE-FREE SURVIVAL ACCORDING TO PATHOLOGIC RESPONSE AND P95-HER2 IN THE CHER-LOB NEOADJUVANT STUDY OF CHEMOTHERAPY PLUS TRASTUZUMAB, LAPATINIB OR COMBINED TRASTUZUMAB AND LAPATINIB IN HER2+ OPERABLE BREAST CANCER

V. Guarneri1, G. Bisagni2, A. Bottini3, K. Cagossi4, A. Frassoldati5, F. Piacentini1, C. Holford6, J. Bruey7, R. D’Amico1, P.F. Conte1 1 Dept. of Hematology/Oncology, Modena University Hospital, Modena, ITALY, 2 Department of Oncology, Arcispedale S Maria Nuova, Reggio Emilia, ITALY, 3 Unit, Istituti Ospitalieri di Cremona, Cremona, ITALY, 4Medical Oncology, Ospedale Ramazzini, Carpi, ITALY, 5Dept of Clinical Oncology, Azienda Ospedaliera di Ferrara St. Anna, Ferrara, ITALY, 6Research & Development, GlaxoSmithKline, Stockley Park, Uxbridge, UNITED KINGDOM, 7Biomarker Research & Development, BioTheranostics, Inc, San Diego, CA, UNITED STATES OF AMERICA Introduction: This is a randomized phase II trial of preoperative sequential taxanes-anthracyclines in combination with trastuzumab, lapatinib, or combined trastuzumab and lapatinib in HER2 positve operable breast cancer. We previously reported that dual HER2 blockade with trastuzumab and lapatinib significantly increased the pathologic complete response ( pCR) rate. The aim of the present post-hoc analysis is the evaluation of disease free survival (DFS) according to pCR, p95-HER2 expression, and treatment arm. Methods: 121 patients were randomly assigned to weekly paclitaxel followed by FE75C combined with trastuzumab (arm A, n= 36), lapatinib (arm B, n= 39), or trastuzumab and lapatinib (arm C, n= 46). P95-HER2 expression was measured by IHC (bioTheranostics, Inc. San Diego, CA). Patients were classified as having a p95-HER2 positive tumor in case of strong immunostaining in ≥80% of cells. pCR was defined as disappearance of infiltrating disease in breast and axillary nodes. DFS was calculated from the date of surgery to the date of recurrence (local or distant) or death. Results: Overall, pCR was observed in 32% of patients. The pCR rates per treatment arm were 25% in arm A, 26.3% in arm B, and 46.7% in arm C (Arm C vs combined Arms A + B: p = 0.018). Thirty-five patients (29%) resulted p95-HER2 positive. At the time of the present analysis, 17 DFS events have been reported. The expression of p95-HER2 was not significantly correlated with the probability of relapse, overall and per treatment arm. The risk of relapse was significantly lower for patients achieving a pCR (hazard ratio 0.09, p = 0.019). The 3-yr DFS rate was 88% in arm C vs 72% in arms A + B. Conclusions: The achievement of a pCR in breast and axillary nodes is predictive of outcome in HER2 positive breast cancer patients treated with neoadjuvant therapy. This is consistent with other neoadjuvant studies. p95-HER2 appears to be neither predictive nor prognostic in this patient population. The combination arm resulted in a trend for a better DFS, probably reflecting the higher pCR rate observed. Sponsored by GlaxoSmithKline. Disclosure: C. Holford: Employer GlaxoSmithKline. J. Bruey: Employee at bioTheranostics, Inc. P.F. Conte: GlaxoSmithkline: research funds, speaker’s bureau, advisor. All other authors have declared no conflicts of interest.

268P

PROLIFERATION DETERMINED BY KI67 MARKER DEFINES PATHOLOGICAL COMPLETE RESPONSE IN A DOSE-DENSE NEOADJUVANT CHEMOTHERAPY SCHEDULE IN LOCALLY ADVANCED BREAST CANCER PATIENTS

M.Y. Plata Fernandez1, A. Sanchez-Muñoz2, A. Jaén-Morago1, M. Lomas-Garrido1, M. Fernández1, C. Llacer2, M. Fernández1, N. Ribelles2, E. Alba-Conejo2, P. Sánchez-Rovira1 1 Oncología Médica, Complejo Hospitalario de Jaén, Jaén, SPAIN, 2Oncology, Hospital Universitario Virgen de la Victoria, Malaga, SPAIN Background: The value of Ki67 proliferation biomarker during neoadjuvant chemotherapy (NC) is less clear than with neoadjuvant endocrine therapy. We study the role of proliferation according to Ki67 marker measured by immunohistochemistry (IHC) as a predictive factor of pathological complete response to NC. Methods: From May 2002 to June 2005, 127 pts diagnosed as stage II-III, including inflammatory tumors, breast cancer patients ( pts) received one of the 2 NC regimens every 2 weeks with prophylactic growth factor support. Study A: adriamicin 40mg/ m2 d1 plus P 150mg/m2-G 2000 mg/m2 d2 for 6 cycles (n = 54); study B: epirubicin

doi:10.1093/annonc/mds392 | ix

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

Background: The incidence of breast cancer is rising in India. It presents at a younger age in Indian population as compared to the western countries. Aims and objectives: This is a retrospective study of all breast cancer patients less than 40 years of age treated in single tertiary care center from June 2006 to June 2011. The aim was to assess the factors that may influence clinical outcome and prognosis including demographics, clinical characteristics and pathological findings and treatment. Materials and methods: Clinical data was collected from medical records. Variables like age, stage at presentation, surgery type, chemotherapy, radiation, tumour size, grade, nodal status, perinodal extension, lymphovascular emboli, ER, PR and Her2 neu status were analyzed in relation to outcome. Results: Out of 613 breast cancer patients, 91 were under 40 years of age corresponding to an incidence of 14.8%. Median tumour size was 3 cm and lymph node positivity was 56.9%. Lymphovascular emboli was positive in 42 patients (48.8%) and perinodal extension was positive in 36 patients (41.8%). Thirty patients (34.8%) were ER positive, while 39 patients (45.3%) were PR positive. Her 2 neu receptors were positive in 20 patients (23.2%). Thirty nine patients were triple negative (45.3%). The median follow up period was 27 months with the DFS being 73.2% and OS being 87.2%. In univariate analysis, factors significantly associated with survival were stage at presentation, presence of lymhovascular emboli, presence of perinodal extension and grade of the tumour. Conclusions: Breast cancer in India is seen in younger patients and most of these are triple negative breast cancers. Patients with breast cancer below 30 years of age are surviving more than the age group of 30 to 40 years. Survival of young breast cancer patients in India is comparable to western studies. Disclosure: All authors have declared no conflicts of interest.

266P

negative pts had a median survival of 60 months and 32 months respectively. Five year relative survival rate was 63%. Conclusion: Even if most pts had a ER+ tumor, the rate of poor prognosis factors is high: 29% SBR III, 14% HER2 positive, 5% triple negative and 73% had a KI67 index over 15%. Despite these poor prognostic factors, survival was rather long with a median OS of 5 years. Disclosure: All authors have declared no conflicts of interest.

Annals of Oncology

90 mg/m2-cyclophophamide 600 mg/m2 d1 for 3 cycles, followed by a second sequence with paclitaxel (P) 150 mg/m2- gemcitabine (G) 2500 mg/m2 d1 +/trastuzumab 2mg/kg/wk according to status Her2 (n= 73). In study A pts did not receive trastuzumab regardless of Her2 status. Ki67 was centrally assessed by IHC (Master Diagnostica clon SP6). Tumors were considered to have high proliferation rates (Ki67) if ≥ 20% of cell nuclei stained positive. Median age was 48 years. Her2+ was observed in 31%, HR negative 34%, grade III 46% and high Ki67 47%. Pathological complete response ( pCR) defined as absence of invasive cells in breast and lymph node was achieved in 35 pts (28%). Univariate and multivariate logistic regression models were used to study the association of each clinical-pathological variable with pCR. Results: High Ki67 was the main predictive factor of pCR to NC. In the univariate analysis histological grade ( p = 0.001), HR ( p < 0.001), Ki67 ( p = 0.001) and p53 ( p < 0.001) were statistically associated with pCR. A multivariate logistic regression showed than only high Ki67 ( p = 0.02; OR = 5.5 CI95% 1.2- 18) and HR ( p = 0.045; OR = 0.25 CI95% 0.05-0.97) were predictive for pCR. Conclusion: High proliferation determined by ki67 marker is an independent predictive factor for pCR in locally advanced breast cancer patients treated in two dose-dense NC schedules. Disclosure: All authors have declared no conflicts of interest.

269P

E. Puma1, P. Mariani1, S. Damian1, M.C. Dazzani1, E. De Benedictis1, M. Parati2, L. Sica2, A. Tessari1, F.G.M. De Braud1, A. Moliterni1 1 Medical Oncology, Istituto Nazionale dei Tumori di Milano, Milan, ITALY, 2 Oncologia Medica, Fondazione IRCCS - Istituto Nazionale dei Tumori, Milano, ITALY Background: The ECTO-1 study demonstrated the efficacy of concurrent doxorubicin and paclitaxel (AT) for 4 cycles followed by cyclophosphamyde/ methotrexate/fluorouracil (CMF) for 4 cycles in the neoadjuvant treatment of operable breast cancer (Gianni L. et al. Clin Cancer Res 2005). With the purpose of ameliorating the tolerability of the regimen, we designed the ASTER study to reduce both the duration and the total dose of neo-/adjuvant treatment with AT followed by CMF. Herein we report on the efficacy of the neoadjuvant portion of the trial. Methods: A total of 70 patients with operable breast cancer were enrolled between September 2008 and November 2011. Median age was 51 years (range 32-73); 74% of patients presented with hormone receptor positive (HR +) and 26% of patients with hormone receptor negative (HR -); 8% of patients presented HER2 overexpression/ amplification; most of the patients had cT2 (96%); half presented with node involvement cN1 (54%). Patients were treated with neoadjuvant Adriamycin (60 mg/ mq) + Paclitaxel (200 mg/mq) q21 for 3 cycles followed by CMF i.v. 1, 8q28 for 3 cycles. Results: The pathological complete response ( pCR), defined as the absence of neoplastic cells in the primary tumor was obtained in 11.5% of cases; tnpCR, defined as the absence of neoplastic cells in the primary tumor and in nodes, was obtained in 10% of cases. Overall, 29% of patients with triple negative breast cancer achieved a pCR, while only 6% of HR positive breast cancer achieved a pCR. Conclusions: Neoadjuvant treatment with three cycles of AT followed by three cycles of CMF is effective in patients with operable breast cancer. The rate of pCR matched perfectly with the data obtained with the original schema AT x 4 followed by CMF x 4 in 69 patients treated with primary chemotherapy at our institute, within the ECTO-1 study. The ASTER study sets the base for developing a less toxic chemotherapy regimen sequential and non-cross resistant containing anthracycline and taxane for the treatment of operable breast cancer. Disclosure: All authors have declared no conflicts of interest.

270P

LONG-TERM OVERALL SURVIVAL (OS) AND DISEASE-FREE SURVIVAL (DFS) ACCORDING TO PCR IN BREAST CANCERS SUBTYPES: LUMINAL A AND B, TRIPLE NEGATIVE AND HER2 + , TREATED BY NEAODJUVANT CHEMOTHERAPY (NCT)

C. Abrial1, Q. Wang-Lopez1, M. Mouret-Reynier2, P. Dubray-Longeras2, I. van Praagh-Doreau2, X. Durando2, P. Gimbergues3, J. Nabholtz2, F. Penault-Llorca4, P. Chollet2 1 Clinical Research, Centre Jean Perrin, Clermont-Ferrand, FRANCE, 2Oncology Department, Centre Jean Perrin, Clermont-Ferrand, FRANCE, 3Surgery Department, Centre Jean Perrin, Clermont-Ferrand, FRANCE, 4Dept. de Pathologie, Centre Jean Perrin, Clermont-Ferrand Cedex, FRANCE Background: In breast cancer, positive Hormone Receptors (HR) constitute a favourable prognostic factor whereas HER-2 overexpression is an adverse prognostic factor associated with a more aggressive tumor. In a retrospective database of 282

ix | Abstracts

271P

ADDITIONAL SAFETY RESULTS OF HANNAH: A PHASE III RANDOMISED, OPEN-LABEL, INTERNATIONAL STUDY OF THE SUBCUTANEOUS FORMULATION OF TRASTUZUMAB (H) IN HER2-POSITIVE EARLY BREAST CANCER PATIENTS

C. Jackisch1, M. Dank2, G. Frasci3, K.H. Park4, R.I. Lopez5, M. Johnston6, D. Heinzmann7, H. Weber7, G. Ismael8 1 Department of Obstrics and Gynecology, Klinikum Offenbach GmbH, Offenbach, GERMANY, 2Semmelweis Egyetem, Radiológiai és Onkoterápiás Klinika, Budapest, HUNGARY, 3Division of Medical Oncology A, National Tumor Institute, Naples, ITALY, 4Department of Internal Medicine, Korea University Anam Hospital, Seoul, KOREA, 5Medical Oncology Department, National Oncology Institute, Panama, PANAMA, 6Genentech Inc, South San Francisco, CA, UNITED STATES OF AMERICA, 7Clinical Science, F. Hoffmann-La Roche Ltd, Basel, SWITZERLAND, 8Medical Oncology, Fundation Amaral CarvalhoHospital of Jau, Jau, BRAZIL Background: HannaH demonstrated non-inferiority of the fixed-dose subcutaneous (SC) formulation of H compared with the intravenous (IV) formulation, based on co-primary endpoints of pharmacokinetics (Ctrough) and efficacy ( pCR) (Jackisch et al, EBCC 2012). Methods: Pts with HER2-positive, operable, locally advanced or inflammatory BC were randomised to receive 8 cycles of chemotherapy (4x docetaxel 75 mg/m2 followed by 4x FEC 600/75/600 mg/m2) concurrently with 3-wkly H, either SC (600 mg/5 mL) or IV (8 mg/kg to 6 mg/kg). After surgery, pts continued H-SC or IV to complete 1 year of treatment. Safety has been monitored until 24 mths after last dose of H, including cardiac monitoring. Results: Safety data from 298 (IV) /297 (SC) HannaH pts were analysed. At a median F/U of 12.3 months, 116 pts had completed adjuvant treatment and received a median of 17.5 (IV) and 17.0 (SC) H cycles. Pt characteristics were balanced at baseline. Overall, safety was comparable between arms. Most common AEs (>25% in either arm) were alopecia, nausea, neutropenia, diarrhoea, asthenia, and fatigue, with similar incidence in the two arms. Incidence of severe AEs (≥ grade 3) was comparable between arms (both 52%); the most common of which were haematologic (36.9 [IV] v 35.4% [SC]), GI (6.4 v 5.7%) and infections (5.0 v 6.7%). Serious AEs (SAEs) were reported in 12.4% (IV) v 20.9% (SC) of pts. The imbalance was largely driven by increased reporting of SAEs in the “infections” disease category in the SC v the IV arm (38.7 v 35.1%). Multiple logistic regression analyses did not reveal interactions between route of H treatment, body weight, and exposure (AUC) for SAEs or grade 3–5 AEs. AEs led to the withdrawal of 2.3% (IV) and 5.7% (SC) pts, with a contribution from cardiac AEs (3.0% [SC] v 1.3% [(IV]). Grade 1/2 AEs were reasons for withdrawal in 43% (IV) v 61% (SC) of pts. Grade 1/2 cardiac AEs led to withdrawal in 1.0% (IV) v 2.0% (SC) of pts. Incidence of cardiac AEs was similar in both arms: 12.1% % (IV) v 11.4% (SC). Conclusions: The safety profile of H-SC was comparable to that of H-IV. Detailed analyses of observed imbalances in SAEs and withdrawals will be provided.

Volume 23 | Supplement 9 | September 2012

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

NEOADJUVANT CHEMOTHERAPY IN OPERABLE BREAST CANCER: DATA FROM THE ASTER STUDY (AT FOR 3 CYCLES FOLLOWED BY CMF FOR 3 CYCLES)

breast cancer patients diagnosed from 1991 to 2006, we analysed the overall survival (OS) and the disease-free survival (DFS) of 4 phenotypes: HR-/HER-2- (triple negative, TN); HR + /Ki-67 < 20% (luminal A), HR + /Ki-67 > 20% or Her2+ (luminal B) and HER-2 overexpression (HER-2 + /HR-). Patients and methods: Median diameter of the invasive tumour was 40 mm [10-130]. 231 (82%) patients had a canalar carcinoma. 33% of the tumours were grade III SBR. Patients received either an anthracycline-based chemotherapy (47%), a taxane-based chemotherapy (21%) or a combination of both (32%). The median number of NCT courses was 6 [1-8] followed by a surgery for 97.6%, a radiotherapy for 93%, an adjuvant chemotherapy for 20% and/or an hormonotherapy for 56%. Only 5 patients received adjuvant herceptin for 1 year. In the different subgroups, we had 123 luminal A tumors (44%), 67 luminal B tumours (24%), 21 Her2+ tumours (7%) and 71 TN tumours (25% of patients). Results: According to tumour’s phenotype, the pCR (Chevallier’s classes 1 + 2) was: 3.3% for luminal A, 16.6%, for luminal B, 33.3% for Her2+ HR- and 26% for triple negative tumours. On 2 April 2012, the median follow-up was 148 months (range, 63- 252). We evaluated the OS and the DFS in the four subgroups, at 10 years. -For TN: OS and DFS were 58.5% and 56.2%. -For luminal A: OS and DFS were 76.1% and 64.7%. -For luminal B: OS and DFS were 81.7% and 72.4%. -For Her2 + HR- subgroup: OS and DFS were 64.5% and 56.4%. pCR is an objective method improvement after NCT, but it is variable among subsets, and may constitute an acquired prognostic factor. However, it was much smaller in hormonal-positive tumors subsets and may there not be an endpoint per se. We observed a significant improvement of DFS and OS for patients who reached a pCR only in TN breast cancer. However Her2 subset may be too small to reach significance here. Conclusion: OS and DFS were better in luminal A and B subgroups. Conversely OS and DFS decreased for TN and Her2+ subtypes. When TN breast cancer patients reached a pCR, OS and DFS were significantly improved. Disclosure: All authors have declared no conflicts of interest.

Annals of Oncology

Disclosure: C. Jackisch: I have participated in an advisory board for Roche. M. Johnston: I am currently an employee of Genentech Inc. D. Heinzmann: I am currently an employee of F. Hoffmann-La Roche. H. Weber: I am currently an employee of F. Hoffmann-La Roche. G. Ismael: I declare that my institute has received research funding from Roche and honoraria for attendance at conferences. All other authors have declared no conflicts of interest.

272P

SUBCUTANEOUS INJECTION OF TRASTUZUMAB – ANALYSIS OF ADMINISTRATION TIME AND INJECTION SITE REACTIONS

Background: Intravenous (IV) trastuzumab (H) is the standard of care for HER2-positive breast cancer (BC). A subcutaneous (SC) formulation of H has been developed. H SC formulation contains recombinant human hyaluronidase, which transiently hydrolyzes hyaluronan to facilitate delivery of H to the circulation. The HannaH study (Jackisch, et al. EBCC 2012) demonstrated the non-inferiority of H SC (vs. H IV) in terms of pharmacokinetics and efficacy in patients with HER2-positive early BC. Safety observations for the two routes of administration were also similar. Methods: H SC injection was via handheld syringe, and an injection time of about 5 minutes was recommended by the study protocol. The actual time taken was measured in the safety population at each administration. Injection site reactions (ISRs) associated with H SC were identified using a predefined basket of preferred terms from the medical dictionary for drug regulatory activities. Results: Duration of SC injections was generally between 1–5 minutes, with an average duration of 3.3 minutes (Table 1) compared with 60–90 minutes taken for IV administration. A low incidence of ISRs were seen with SC injections (Table 1), with all but two cases (grade 2) being grade 1; every case was reversible.

Injection duration (minutes)

Number of injections n (%) Incidence of ISRs n (%)

<2 ≥ 2 to < 3 ≥ 3 to < 4 ≥ 4 to < 5 ≥ 5 to < 6 ≥6 Missing Total

51 (1.1) 2231 (49.2) 961 (21.2) 200 (4.4) 1024 (22.6) 45 (1.0) 25 (0.5) 4537 (100)

0 (0) 27 (1.2) 15 (1.6) 6 (3.0) 27 (2.6) 9 (20.0) 0 (0) 84 (1.9)

Summary of injection duration and ISRs (injection site pain) in the H SC arm of the HannaH study (safety population) (N = 297) Incidence of ISRs was highest for injection durations of >6 minutes with 6 pts reporting 9 ISRs over the course of their treatment. In 3 pts (6 ISRs) injection time was prolonged due to injection pain. The remaining three ISRs constituted transient tissue tenderness at injection site.

Conclusion: In the HannaH study, injection of H SC was generally well tolerated with a low incidence of ISRs (grades 1 and 2). These findings support the potential of H SC to provide improved convenience for patients compared to the existing IV formulation. Disclosure: X. Pivot: X. Pivot has served as a consultant or advisor for Hoffmann-La Roche, Novartis, GlaxoSmithKline and has received honoraria from sanofi-aventis. S.D. Moodley: SD Moodly has served as a member of an advisory board for La Roche Hoffman, sanofi-aventis, Britol Meyers Squibb, Nayer Schering, and has received corporate-sponsored research funding from sanfoi-aventis. M. Johnston: M Johnston is an employee of Genentech, a member of the Roche Group. T. van der Horst: T Horst is an employee of F. Hoffman-La Roche Inc. A. Bouhlel: A Bouchel is an employee of F. Hoffmann-La Roche Inc. C. Jackisch: C. Jackisch has served on advisory boards for Hoffmann-La Roche Inc. All other authors have declared no conflicts of interest.

Volume 23 | Supplement 9 | September 2012

IMMUNOGENICITY OF TRASTUZUMAB INTRAVENOUS AND SUBCUTANEOUS FORMULATIONS IN THE PHASE III HANNAH STUDY

R. Hegg1, T. Pienkowski2, S. Chen3, E. Staroslawska4, S. Falcon5, N. Kovalenko6, N. Al-Sakaff7, D. Heinzmann7, G. Kolaitis8, G. Ismael9 1 Oncology, Hospital Pérola Byington and FMUSP, Sao Paulo, BRAZIL, 2MSC Memorial Cancer Centre, MSC Memorial Cancer Centre and Institute Maria Sklodowska-Curie, Warsaw, POLAND, 3Changhua Christian Hospital, Taipei, TAIWAN, 4St. John’s Cancer Center, Lublin, POLAND, 5Hospital Nacional Edgardo Rebagliati Martins, Lima, PERU, 6Stavropol Regional Clinical Oncology Dispensary, Stavropol, RUSSIAN FEDERATION, 7Roche, Basel, SWITZERLAND, 8 Roche, Nutley, NJ, UNITED STATES OF AMERICA, 9Hospital Amaral Carvalho, Jau, BRAZIL Background: HannaH demonstrated non-inferiority of the fixed-dose subcutaneous (SC) formulation of trastuzumab (H) to the intravenous (IV) formulation, based on pharmacokinetics (Ctrough) and efficacy ( pathological complete response [ pCR]) (Jackisch et al, EBCC 2012). We report on efficacy and safety in relation to immunogenicity (anti-drug antibodies [ADAs]). Methods: Blood samples for ADA testing were drawn at baseline (BL), then day 1 of cycles 2, 5 ( pre-surgery), 13 and 18 ( post-surgery) and at months 3, 6, 12, 18 and 24 following last H dose. Screening for ADAs to H and rHuPH20 was by bridging immunoassays. Positive ( pos) samples were re-tested, with confirmed pos patients ( pts) then tested for neutralizing antibodies (abs). Relationship between ADAs (against H and/or rHuPH20) with respect to, efficacy ( pCR) and safety (infusion-related reactions, IRRs) was investigated. Results: The median follow-up (FU) (including ADA testing) was ∼12.3 months. At BL, 5.9% of pts (17/290) in the IV arm and 4.2% of pts (12/287) in the SC arm were pos for ADAs to H. Post-BL (treatment and FU), 3.4% (10/295) and 6.8% (20/295) pts in IV and SC, respectively were pos for ADAs to H. One pt was confirmed pos twice (non-consecutive time points). At BL, 7.6% of pts (22/290) in the SC arm were pos for ADAs to rHuPH20. Post-BL, the rate was 11.5% (34/295). Titer ranges post-BL were similar to those observed pre-BL. One pt was pos for ADAs to both H and rHuPH20. No neutralizing ADAs to H or rHuPH20 were seen. The pCR rate did not differ significantly between Anti-H ab (AHA)-pos (30% [IV n = 10], 55% [SC n = 20]) and AHA-neg (36.5% [IV n = 10], 41.6% [SC n = 20]) pts. Anti-rHuPH20 status did not correlate with pCR rates (41.2% [ pos n = 34] vs 41.7% [neg n = 254]) in the SC arm. In pts given H SC or H IV, occurrences of IRRs were similar in AHA-neg (46.3% [SC n = 255] 37.6% [IV n = 263]) and -pos (40% [SC n = 20] 25% [IV n = 10]) pts. The incidence of pts with an IRR was similar in both anti-rHuPH20 status groups (SC arm only; neg 44.5% [n = 254], pos 41.2% [n = 34]). Conclusion: Using a highly sensitive assay, ADAs against both H (IV/SC) and rHuPH20 (SC only) were observed transiently and were of no relevance in terms of efficacy or safety. Immunogenicity monitoring in the study is ongoing. Disclosure: T. Pienkowski: Dr Pienkowski has carried out research sponsored by F Hoffmann La Roche. N. Al-Sakaff: I have an interest in relation of one or more organistations that could be perceived as a conflict of interest in the context of the subjuect of this abstract. Interest - other substantive relationship. Employee of F. Hoffman-La Roche Ltd. D. Heinzmann: I have an interest in relation of one or more organistations that could be perceived as a conflict of interest in the context of the subjuect of this abstract. Other substantive relationship: Employee of F Hoffman-La Roche and stockholder. G. Kolaitis: Dr Kolaitis is an employee and stockholder of F Hoffmann La Roche. G. Ismael: I have an interest that could be perceived as a conflict of interest in the context of the subjuect of this abstract. Corporate sponsored research - Roche Other substantive relationships - honoraria for conferences (Roche). All other authors have declared no conflicts of interest.

274P

A PHASE I DOSE ESCALATION AND BIOEQUIVALENCE STUDY OF A TRASTUZUMAB BIOSIMILAR (FTMB) IN HEALTHY MALE VOLUNTEERS

L. Wisman1, E. De Cock1, J. Reijers2, I. De Visser2, M. De Kam2, S. van Os1, G. Voortman1, L. Hooftman1, K. Burggraaf2 1 Clinical Operations, Synthon BV, Nijmegen, NETHERLANDS, 2Cardiovascular Research, Center for Human Drug Research, Leiden, NETHERLANDS FTMB is a biosimilar of trastuzumab (Herceptin®), a humanized monoclonal antibody targeting the human epidermal growth factor receptor 2 (HER2). Herceptin® is registered for treatment of HER2-positive breast cancer and metastatic gastric cancer. Pharmacokinetic profiles of FTMB were compared to Herceptin® in this combined dose escalation and bioequivalence study. In the dose escalation part healthy male volunteers received single doses of 0.5, 1.5, 3.0 or 6.0 mg/kg FTMB in consecutive cohorts to assess the safety profile. At each dose level, subjects were randomized to FTMB (n = 6) or placebo (n = 2), with the exception of the 6 mg/kg cohort, where subjects were randomized to FTMB (n = 9), Herceptin® (n = 9) or placebo (n = 2). Safety data were evaluated by an independent data safety monitoring board. To establish bioequivalence a total of 92 healthy male subjects, including those of the 6 mg/kg dose escalation cohort, were randomized equally to FTMB or Herceptin®. Blood samples were taken prior and at regular, predefined time points up

doi:10.1093/annonc/mds392 | ix

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

X. Pivot1, V. Semiglazov2, S. Chen3, S.D. Moodley4, A. Manihkas5, M. A. Coccia-Portugal6, M. Johnston7, T. van der Horst8, A. Bouhlel8, C. Jackisch9 1 Service Oncologie Medicale, C.H.U. Jean Minjoz, Besancon Cedex, FRANCE, 2 Breast Cancer Department, N.N.Petrov Research Inst. of Oncology, St. Petersburg, RUSSIAN FEDERATION, 3Division of Breast Surgery, Chang Gung Memorial Hospital, Taipei, TAIWAN, 4Department of Oncology, Wits Oncology Donald Gordon Medical Centre, Johannesburg, SOUTH AFRICA, 5 St Petersburg City Oncology Centre, St.Petersburg, RUSSIAN FEDERATION, 6 Eastleigh Breast Cancer Care Centre, Pretoria, SOUTH AFRICA, 7Genentech Inc, South San Francisco, CA, UNITED STATES OF AMERICA, 8F. Hoffman-La Roche Ltd, Basel, SWITZERLAND, 9Department of Obstrics and Gynecology, Klinikum Offenbach GmbH, Offenbach, GERMANY

273P

Annals of Oncology

to 9 weeks after dosing. Trastuzumab levels were determined in serum with a validated bridging ELISA method. The mean area under the concentration-time curve from time zero to infinity (AUC0-inf ), normalized to 1 mg/kg and corrected for content differences between FTMB (Test) and Herceptin® (Reference) was 221.4 µg.d/ mL for Test and 245.6 µg.d/mL for Reference. The ln-transformed Test/Reference (T/ R) ratio for AUC0-inf was 89.6% (90% confidence interval (CI): 85.1-94.4%), demonstrating bioequivalence based on the acceptance interval of 80.0-125.0%. For the secondary endpoint, maximum concentration observed (Cmax), the T/R ratio was 89.4% (90% CI: 83.4-95.9%). Interestingly, the elimination of both FTMB and Herceptin® was shown to be non-linear. Adverse events other than the documented side effects of trastuzumab (fever, influenza-like illness, and fatigue) did not occur, and signs of cardiotoxicity were not observed. In conclusion, in this first bioequivalence study with a trastuzumab biosimilar in healthy male volunteers, single administration of FTMB was considered well tolerated in doses up to 6 mg/kg, and FTMB was demonstrated to be bioequivalent to Herceptin®. Disclosure: All authors have declared no conflicts of interest.

275P

EFFICACY AND SAFETY OF TRASTUZUMAB IN SMALL HER2 POSITIVE TUMORS. A SINGLE INSTITUTION EXPERIENCE

medullary 1 (1.5%). Histological grade: G1 2 (3.0%), G2 44 (67.7%) and G3 19 (29.2%). The Ki 67 index was ≥ 13% in 45 (69.2%). Hormonal receptor status: Positive 42 (64.7%). Local treatment: breast conservative surgery and local irradiation 57 (87.7%) and mastectomy 8 (12.3%). Systemic treatment: 47 (72.3%) received T in combination and/or after adjuvant chemotherapy (mostly anthracycline/taxane-based). Cardiac Toxicity: 2 cases of asymptomatic and transient left ventricular ejection fraction decrease below 20% after T. With a median follow-up of 35.4 months, there were 5 (7.7%) metastatic recurrences, 2/18 (11.1%) in the non-T treated group, and 3/47 (6.4%) in the T-treated group. At 60 months, relapse-free survival estimates were 80.7% and 91.2% in the non-T and T-treated groups ( p > 0.05), respectively. Conclusions: Our data suggest that infra-centimetric HER2+ tumors are associated with a general poor outcome and Trastuzumab-based therapies are justified. Future evaluation of dual anti-HER2 blockade in combination with chemotherapy seems warranted in this setting. Disclosure: All authors have declared no conflicts of interest. 277P

M.V. Sano, E. Taibi, M. Alì, M. Chiarenza, S. Clementi, M. Caruso, R.A. Aiello Medical Oncology, Humanitas Centro Catanese di Oncologia, Catania, ITALY

276P

INFRA-CENTIMETRIC HER2 POSITIVE BREAST CANCER: THE SOUTHERN ITALY EXPERIENCE

G.R.R. Ricciardi1, B. Adamo1, M. Caruso2, C. Iacono3, G. Blanco4, N. Borsellino5, L. Blasi6, R. Bordonaro7, G. Tuccari8, V. Adamo1 1 Human Pathology, Unit of Integrated Therapies in Oncology, University Policlinic "G. Martino", Messina, ITALY, 2Oncologia Medica, Humanitas Centro Catanese di Oncologia, Catania, ITALY, 3Dept. Oncology, Ospedale Civile Maria Patern, Ragusa, ITALY, 4Oncologia Medica, Istituto Oncologico del Mediterraneo, Viagrande (Catania), ITALY, 5Oncology, Ospedali Buccheri la Ferla, Palermo, ITALY, 6Oncology, Hospital St Raphael Giglio, Cefalù, ITALY, 7Oncology, Azienda Ospedaliera ARNAS Garibaldi, Catania, ITALY, 8Department of Human Pathology, Section of Pathologic Anatomy, University Policlinic "G. Martino", Messina, ITALY Background: HER2 positive (HER2+) disease is relatively uncommon in women with small, early breast cancers (BC), accounting for approximately ∼10% of the cases. Several trials have shown benefit of adjuvant trastuzumab (T) for node positive or supra-centimetric HER2+ BC but there are limited data concerning infra-centimetric HER2+ disease. The aim of this study is to identify the pattern of use of T and risk of recurrence in small HER2+ tumors. Methods: This observational, multicenter, retrospective study was conducted in 9 oncology centers in Sicily during 2005-2011. Multifocal and metastatic tumors were excluded. Results: 65 cases were evaluated. Median age was 52.7 years (range 36-77). Tumor stage: T1a 21 (32.3%) and T1b 44 (67.9%). Nodal status by sentinel biopsy: 48 (74%) had node-negative disease. Invasive histological type: ductal 61 (93.9%), lobular 3 (4.6%) and

ix | Abstracts

S. Saji1, M. Takada2, N. Honma3, N. Masuda4, Y. Yamamoto5, K. Kuroi6, H. Yamashita7, S. Ohno8, T. Ueno9, M. Toi9 1 Target Therapy Oncology, Kyoto University Graduate School of Medicine, Kyoto, JAPAN, 2Breast Surgery, Kyoto University, Kyoto, JAPAN, 3Research Team For Geriatric Pathology, Tokyo Metropolitan Institute of Gerontology, Tokyo, JAPAN, 4 Surgery and Breast Oncology, NHO Osaka National Hospital, Osaka, JAPAN, 5 Breast and Endocrine Surgery, Kumamoto University, Kumamoto, JAPAN, 6 Breast Surgery, Tokyo Metropolitan Komagome Hospital, Tokyo, JAPAN, 7 Breast and Endocrine Surgery, Hokkaido University Hospital, Hokkaido, JAPAN, 8 Breast Oncology, National Kyushu Cancer Center, Fukuoka, JAPAN, 9Breast Surgery, Kyoto University Graduate School of Medicine, Kyoto, JAPAN Background: Local production of estrogen using aromatase at tumor site was thought to be the direct target of aromatase inhibitor (AI). However recent our works (Honma N, Can Sci 2011, Takada M, Breast 2012) and others (Lonning PE, Clin Can Res 2011) postulated that not local aromatase but whole body aromatase (possibly correlates to serum estrone (E1) level) would be the critical target of this drug. Patients and methods: Among the 116 postmenopausal patients who enrolled to the JFMC 34-0601 phase II clinical trial of neo-adjuvant 24 weeks (wks) treatment with AI, exemestane, serums from 60 patients at pre-treatment, at 4wks and at the end of treatment (24wks) were subjected to this study. Estradiol (E2), E1, dehydroepiandrosterone (DHEA) and androstenedione (A) were measured using LC-MS/MS analysis for precise measurement, and E1- sulfate (E1S) was by RIA. Results: Mean pre-treatment serum levels of E2, E1 and E1S were 2.29, 12.86 and 198.87 pg/ml, respectively. All these estrogens decreased to under-detection level (0.5 pg/ assay for E1 and E2, 5 pg/assay for E1S) at 4wks of treatment, and it maintained to the end of the treatment in almost all patients irrespective of clinical response. Patients with higher concentration of pre-treatment E2 and E1 (cutoff = mean value) had more chance to obtain clinical objective response ( p = 0.0245 and p = 0.0448, respectively). However in multivariate analysis, only E1 is the independent predictive factor for objective response (Odds 5.97, p = 0.0113), probably due to positive correlation of E2 to BMI, which is also predictive factor for AI response. When comparing pre-treatment level and those at 24wks (or at termination of treatment), patients with progressive disease showed significant increase of DHEA and A (p = 0.0313 for both), although patients with objective response maintained those at the same level. Conclusion: Pretreatment serum concentration of E1, not E2, is the independent predictive factor of clinical response to neo-adjuvant exemestane. Patients with disease progression showed significant increase of androgens during the treatment. Serum steroid concentrations using precise LC-MS/MS analysis may have the informative value for the AI treatment, and the increased androgens would be the important subject to be studied further. (UMIN trial ID; C000000345) Disclosure: S. Saji: S.S. has a honorarium for lecture from Pfizer. M. Toi: M.T. has a research grant and honorarium for lecture from Pfizer. All other authors have declared no conflicts of interest. 278P

THE IMPACT OF PROGESTERONE RECEPTOR ON NODE-NEGATIVE BREAST CANCER SURVIVAL

M. Solak1, F. Paksoy2, O. Keskin1, N. Kertmen1, M.K. Altundag1 1 Medical Oncology, Hacettepe University Oncology Institute, Ankara, TURKEY, 2 Medical Oncology, Ankara Onkoloji Hastanesi, Ankara, TURKEY Background: The aim of this study was to evaluate tumor characteristics and survival rates of node-negative breast cancer patients according to progesterone receptor (PR) status. Methods: Of 2,218 breast cancer cases diagnosed between 1995 and 2011 in Hacettepe University, Institute of Oncology; 864 node-negative patients with known hormone receptor status were analyzed for age, tumor size, grade, lymphovascular

Volume 23 | Supplement 9 | September 2012

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

Background: Patients affected by breast cancer with negative lymph-node are considered at high risk if at least one of the following factors is present: tumor size > 2 cm, HR negative, histological and/or nuclear grade 2-3 or age < 35 years. However some patients with HER2 positive cancers develop a recurrence despite small tumor size (< 2 cm) and none of these factors. Four large trials confirmed the potential benefit of incorporating trastuzumab in adjuvant regimens as a standard treatment. The effectiveness of adjuvant trastuzumab has been demonstrated either for node negative pT1c patients and for diseases at a higher risk. However these trial were not able to provide any information regarding the outcome of patients with node negative tumours <1cm. Purpose: To assess the efficacy and safety of trastuzumab in HER2 positive small tumors ( pT1 pN0 M0/ stage I). Patients and methods: Between January 2005 and February 2012, 88 patients with small ( pT1) Her2 positive node negative breast cancers were retrospectively reviewed at our institution. Nine patients(10,2%) were staged pT1a N0, 17(19,3%) pT1b pN0 and 65(73,8%) pT1c pN0. Among them 83(94%) patients were treated with adjuvant anthracyclines regimens for 4 or 6 cycles followed by one year of trastuzumab (52 weeks). Median age at diagnosis was 51,5 years (range 28-73) Fifty-two patients were followed up for at least 1 year, the median follow up was 35 months (range 13–63). The Left Ejection Ventricular Fraction (LEVF) was assessed at baseline and every three months. Results: Two patients relapsed; (one lung metastasis during treatment, one local recurrence 13 months after the last dose). An asymptomatic absolute decrease in LVEF by 10% was recorded in three patients. No patients discontinued the treatment. Conclusions: The use of trastuzumab in HER2 positive small tumor seems safe and effective. More data should be available for tumors smaller than 1 cm. Disclosure: All authors have declared no conflicts of interest.

SERUM CONCENTRATION OF ESTRONE (E1), NOT ESTRADIOL (E2), IS THE INDEPENDENT PREDICTIVE FACTOR OF RESPONSE TO NEO-ADJUVANT EXEMESTANE TREATMENT IN POSTMENOPAUSAL BREAST CANCER PATIENTS: JFMC 34-0601 TR

Annals of Oncology

invasion and HER-2/neu status. Estrogen receptor (ER), PR and HER-2 status were determined using immunohistochemistry. Patients were classified into four groups; ER + /PR + , ER + /PR-, ER-/PR+ and ER-/PR-. Pearson Chi-Square test and Kaplan-Meier with log-rank test were performed for statistical analysis. Results: 575 (66.6%) patients had ER + PR + , 64 (7.4%) had ER + PR-, 43 (5.0%) had ER-PR+ and 182 (21.1%) had ER-PR- tumor. The median age of patients were 48 years (20-83) and were similar in all groups (p = 0.18). Majority of patients with PR- tumors were postmenopausal at the time of diagnosis (p = 0.021). The follow-up period was 39.5 (1-273) months. Conclusions: Patients in PR- groups are more likely to have large-size and HER-2 positive breast cancer. ER- groups are associated with high grade disease. Patients with PR+ breast cancer have significantly longer overall survival, compared to patients with ER+ tumors. Our data suggest that PR status is an important prognostic factor for node-negative breast cancer survival. Disclosure: All authors have declared no conflicts of interest.

279P

PROGESTERON RECEPTOR STATUS IN DETERMINING THE PROGNOSIS OF ESTROGEN RECEPTOR POSITIVE / HER2-NEGATIVE BREAST CARCINOMA

U. Yalcintas Arslan, O. Bal, T. Ozatli, K. Helvaci, O. Esbah, B. Budakoglu, U. Uyeturk, O.U. Sonmez, I. Turker, O.B. Oksuzoglu Department of Medical Oncology, Ankara Dr. A.Y.Oncology Research and Education Hospital, Ankara, TURKEY

280P

ADJUVANT AROMATHASE INHIBITORS (AIS) HORMONE THERAPY (HT): WHICH REASONS LEAD PATIENTS (PTS) TO DISCONTINUE TREATMENT? A MONO INSTITUTIONAL ANALYSIS

L. Moscetti1, M.A. Fabbri1, I. Sperduti2, P. Frittelli3, F. Nelli1, A. Massari4, G. D’Auria1, L. Pompei5, E.M. Ruggeri1 1 Divisione di Oncologia, Ospedale Belcolle, ASL di Viterbo Oncology Unit, Viterbo, ITALY, 2Biostatistics, Regina Elena Institute, Roma, ITALY, 3Surgery, Breast Surgery, Ospedale Belcolle, ASL di Viterbo Oncology Unit, Viterbo, ITALY, 4 Oncologia ed Ematologia, Anatomia Patologica, Ospedale Belcolle, ASL di Viterbo Oncology Unit, Viterbo, ITALY, 5Oncologia ed Ematologia, Radioterapia, Ospedale Belcolle, ASL di Viterbo Oncology Unit, Viterbo, ITALY Background: AIs represents the standard HT for the adjuvant treatment of post menopausal endocrine sensitive early breast cancer. A percentage of pts interrupt treatment because of toxicity. In this setting the switch to other AIs or tamoxifen (TAM) may represent an option to allow the prosecution of HT. We reviewed the main reasons for interruption of the AIs in our institution from 2006 to the present. Methods: 236 pts with a minimum treatment time up of 6 months were considered eligible for analysis. Pts characteristics: median age 64 yrs (35-89), median follow up 24 months (6-28). Prior adjuvant CT: taxanes based: 47 pts, anthracyclines based: 43 pts. 118 pts (49%) had received letrozole (L), 101 (43%) anastrozole (A), 18 (8%) exemestane (E). An alternative HT (AIs or tamoxifen [T]) was offered to pts who wanted or needed to interrupt permanently the ongoing drug.

Volume 23 | Supplement 9 | September 2012

281P

FOR WHICH PT1A, BN0M0 HORMONE RESPONSIVE INVASIVE BREAST CARCINOMAS COULD ENDOCRINE THERAPY BE AVOIDED?

C. Perrin1, J. Edeline1, J. Leseur2, V. Lavoué3, P. Tas4, H. Mesbah5, C. Lefeuvre-Plesse6, D. Gedouin6, F. Penault-Llorca7, P. Kerbrat1 1 Medical Oncology Department, Eugène Marquis Center, Rennes, FRANCE, 2 Radiation Department, Eugène Marquis Center, Rennes, FRANCE, 3Surgical Oncology, University Hospital, Rennes, FRANCE, 4Pathology Department, Eugène Marquis Center, Rennes, FRANCE, 5Informations and Statistics, Eugène Marquis Center, Rennes, FRANCE, 6Medical Oncology, Eugène Marquis Center, Rennes, FRANCE, 7Dept. de Pathologie, Centre Jean Perrin, Clermont-Ferrand Cedex, FRANCE Background: Overtreatment is a daily concern in adjuvant setting for invasive breast carcinoma. Although chemotherapy is the most controversial issue, endocrine therapies must also be discussed. Methods: In this monocentric retrospective study, we analysed results of patients treated for hormone responsive invasive pT1a,bN0M0 breast cancer, focusing on the population without adjuvant endocrine treatment. Women with previous history of invasive carcinoma were excluded. Results: In our institution, 382 patients with hormone responsive breast invasive pT1a,bN0M0 carcinoma were treated between 1997 and 2007. Local treatment involved either mastectomy or partial breast surgery followed by breast radiotherapy. After multidisciplinary discussion, among the 382 patients, 162 patients (42 %) did not receive any adjuvant endocrine treatment. Comparatively to the group treated with endocrine therapy, these patients had significantly more grade I tumors (81.4 % versus 47.2 %, p< 0.0001) and Ki67 < 14% (89.6% versus 72.4%, p < 0.0001). In patients not treated with endocrine therapy, after a mean follow-up of 7.1 years, 5 years-disease free survival (DFS) was 93% and 5 years-distant disease free survival was 98.7%. Remarkably, 5 years-DFS was particularly favourable for patients older than 50 years old (97.1% versus 82.1% for age < 50 years old, p = 0.0002) and for grade I carcinomas (97.3% versus 76.7% for grade II-III, p= 0.0005). Only one metastatic recurrence occurred after 10 years in the grade I group. Conclusion: This series of patients treated without adjuvant endocrine therapies raises the issue of avoiding anti tumor endocrine therapy in patients with hormone responsive pT1a,bN0M0, age> 50 years and grade I tumors. Disclosure: All authors have declared no conflicts of interest.

282P

EVALUATION OF BONE MINERAL DENSITY (BMD) IN EARLY BREAST CANCER (EBC) PATIENTS TREATED WITH UP-FRONT OR SWITCH SCHEDULES OF AROMATASE INHIBITORS (AIS): RESULTS OF A SINGLE CENTRE RETROSPECTIVE STUDY

P. Seminara1, A. Iannace2, G. Manna2, T. Losanno2, A. Emiliani2, E. Franzese1, L. Frati2 1 Internal Medicine, Sapienza University, Rome, ITALY, 2Department of Experimental Medicine, Sapienza University, Rome, ITALY Background: Aromatase inhibitors (AIs) are the standard adjuvant treatment of hormone responsive early breast cancer (EBC) in postmenopausal women. AIs are well tolerated but are associated with specific toxicities, including effect on bone (accelerated bone loss, bone fracture, osteoporosis) and musculoskeletal symptoms (ostealgia, arthralgia, myalgia) Aim: The present observational study investigated the effect of BMD in postmenopausal women with EBC scheduled to receive AIs up-front or switch treatment after standard 2-3 years of adjuvant Tamoxifen (TAM-AIs).

doi:10.1093/annonc/mds392 | ix

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

Background: The aim of this retrospective study was to determine whether progesteron receptor (PgR) status have an influence on the prognosis of estrogen receptor positive (ER+)/HER2-negative breast carcinoma (BC). Methods: We retrospectively reviewed the medical files of 1680 operable BC patients ( pts) diagnosed between 1996 and 2011 and 456 of whom ER,PgR and HER2 status known were included in this study. Patients were categorized into 2 groups; as group A (ER + /PgR-/HER2-negative) and group B (ER + /PgR + /HER2-negative). Twenty one percent (97 pts) of the pts were in group A. Results: Median follow up was 33.5 (0-177) months. Median age was 54 (21-90) years. Sixty-one percent (278) of the pts had node-positive BC. Sixty percent (276) of the pts were postmenopausal. Eighty percent (365) of the pts received adjuvant chemotherapy (ACT). Adjuvant hormonotherapy (AHT) was recommended to nearly all patients (mostly tamoxifen). Pts in group A had significantly higher lymph node positive disease as compared to group B (%70 vs 59%, p = .046). Although there was no statistically significant difference between two groups for the first site of recurrence (mostly bone and soft tissue, p = .51), the number of recurrence and mortality events in group B were proportionally less than group A [27.8% and 21.2% for group B and 42.7% and 32.2% for group A;p = .005 and p = .004, respectively). In the node-positive subgroup, an important difference for relative risk RR of BC recurrence between groups A and B was found (34/68 vs 75/207, RR 1.3, CI:1.02-1.86, p = .034). But, the mortality risk was similar for lymph node-positive pts in groups A and B (16/63 vs 50/201,RR 1.0, CI:0.62-1.66,p = 0.93). Although DFS for group B was shorter, it was not statistically significant (5-year DFS rate 45% vs 35% p = .24). However, 5-year overall survival (OS) was significantly longer for pts in group B than group A (78% vs 67%,p = .043). Conclusions: PgR status has a great influence on the prognosis of BC pts especially negative PgR may have a significant negative influence on the prognosis of node-positive ER + /HER2- BC pts. Disclosure: All authors have declared no conflicts of interest.

Results: According to the CTC NCI, arthralgia was the main toxicity observed (G1 19.4%, G2/3 5.5%/1.7%. Overall 24 out of 236 pts (10%) needed discontinuation of AIs as a result of toxicity. Grade 2 and 3 arthralgia was the main reason for discontinuation in 13/24 pts (54%). No differences in the incidence of arthralgia were noted in pts who had received taxanes or anthracyclines. Headache (n = 2), alopecia (n = 2), G3 itching (n = 2), diffuse skin reaction (n = 1) allergic reaction with hypertensive crisis (n = 1), xerostomia and xerophthalmia (n = 1), insomnia (n = 1) and somnolence (n = 1) were the other reasons for discontinuance. In the multivariate logistic regression analysis, age (65 yrs) and HT represent independent factors associated with the onset of arthralgia (respectively p = 0.006 and 0.008, OR 2.65, CI95: 1.32-5.31). 17/24 pts were switched to E, 5 to T and 1 each to L and A. In pts who switched to another AI we observed a reduction in the severity of the arthralgia but not its complete resolution. A complete resolution of symptoms was observed for the other toxicities except for allergic reactions that recurred after switching from L to A. Moreover, the patient with the diffuse skin reaction, chose discontinuation of HT. Conclusions: In our analysis, 10% of pts discontinued AIs due to toxicity. Switching to alternative HT, and monitoring the toxicity, represent an option to offer to pts in order to avoid a premature and permanent interruption of an effective treatment. Disclosure: All authors have declared no conflicts of interest.

Annals of Oncology

Patients and methods: We reviewed data for 89 patients with hormone receptor-positive EBC: 64 postmenopausal women received up-front AIs for 5 years; 25 patients received TAM-AIs. BMD were assessed at baseline and during hormonal treatment once a year. Osteopenic patients received Calcium and vitamin D; osteoporotic patients received also oral biphosphonates. Results: No patient with a normal baseline BMD (T-score >-1) became osteoporotic during 5 years of follow-up. However, it was noted that more women receiving AIs up-front became osteopenic, as compared with TAMàAI treated women (AIs = 58% vs. TAM-AIs = 26%, p= <0.001). In addiction, women with osteoporosis at baseline, treated with biphosphonates, became osteopenic. Using the linear regression model approach we found that only patient with a baseline T-score less than -1.5 became osteoporotic on treatment (Fig.1). Supplementary comedications have improved reduction of musculoskeletal symptoms. Conclusion: The data suggest that only those women with a T-score of less than -1.5 are at risk of developing osteoporosis during the AIs adjuvant therapy. This cut-off value might reduce the costs both for BMD monitoring and supplementary comedications. Finally, the correct management of treatment-related symptoms might improve patient’s compliance and the adherence to treatment. Disclosure: All authors have declared no conflicts of interest.

283P

MODEST VALUE OF STAGING INVESTIGATIONS IN HEAVILY NODE POSITIVE EARLY BREAST CANCER PATIENTS

Background: Staging investigations for metastatic disease are not recommended for asymptomatic patients with early breast cancer. However, it is relatively common practice to perform staging in patients who are heavily node positive (4 or more nodes involved). This study explored the benefit of routine investigations in this patient group. Methodology: All patients operated in Castle Hill Hospital, Hull, UK between January 2005 and December 2011 for early breast cancer and found to have 4 or more nodes involved on pathological assessment were identified. Staging investigations, consisting of CT chest, abdomen and pelvis and isotope bone scan, were reviewed. The need for additional investigations was recorded. Results: A total of 231 patients were identified. 59 patients were excluded as no staging investigations were performed. In the remaining 172 patients staging investigations were positive in 12% (21/172). An additional 17% (29/172) required further investigations for equivocal scans and were not found to have metastatic disease. On univariate analysis with chi-square T stage (T1 3%, T2 9%, T3 14%, T4 38%; p < 0.001), number of nodes involved (4-6 4%, 7-9 9%, 10+ 21%; p = 0.009) and ratio of involved/uninvolved nodes (<0.5 2%, >0.5 16%; p= 0.037) were significant for predicting positive staging investigations (see table). On multivariate analysis using logistic regression only T4 breast cancers were significantly associated with positive staging investigations ( p = 0.008). Conclusions: Routine staging investigations have a modest pick up rate of metastatic disease in heavily node positive patients. Importantly 1 in 6 patients will require further investigations to exclude metastatic disease. Staging is recommended for all patients with locally advanced/inflammatory breast cancers but is probably not necessary in every pN2 patient. Univariate analysis of factors predictive of positive staging investigations Positive Staging scans T Stage

Grade ER HER2 LVI Age Nodes + NAC Node Ratio

1 2 3 4 1 2 3 Pos Neg Pos Neg Pos Neg <50 >50 4-6 nodes 7-9 nodes 10+ nodes Y N <0.5

2.7% (1/37) 8.9% (7/79) 14.3% (5/35) 38.1% (8/21) 0% (0/4) 12.7% (9/71) 12.3% (12/97) 11.8% (14/118) 13.0% (7/54) 16.7% (7/42) 10.5% (13/124) 13.7% (14/102) 10% (7/70) 10.7% (6/56) 12.9% (15/116) 4.4% (3/68) 9.1% (3/33) 21.1% (15/71) 10.3% (7/68) 13.5 % (14/104) 2.1% (1/48)

>0.5

16.3% (20/123)

Disclosure: All authors have declared no conflicts of interest.

ix | Abstracts

p= 0.001

p = 0.751 p = 0.838 p = 0.284 p= 0.636 p= 0.806 p= 0.009 p= 0.637 p= 0.037

ADVANTAGES OF ONE STEP NUCLEIC ACID AMPLIFICATION (OSNA) WHOLE NODE ASSAY IN SENTINEL LYMPH NODE (SLN) ANALYSIS IN BREAST CANCER

A. Santaballa1, H. De La Cueva1, C. Salvador1, A. Garcia Martinez2, M. J. Guarin2, D. Lorente1, L. Palomar1, I. Aznar3, F. Dobon3, P. Bello4 1 Medical Oncology, Hospital Universitari i Polit, Valencia, Spain, 2Pathology, Hospital Universitari i Politecnic La Fe, Valencia, SPAIN, 3Breast Unit, Hospital Universitari i Politecnic La Fe, Valencia, SPAIN, 4Nuclear Medicine, Hospital Universitari i Politecnic, Valencia, SPAIN The aim of this study is to present our first results with OSNA assays performed in a routine clinical setting in 168 patients with invasive and in situ breast cancer and compare them with conventional histopathology results for SLN biopsies in a historical cohort in our institution. Methods: 412 patients (total study population) with clinically node negative early stage breast cancer underwent an axillary SLN in our institution. The SLN was assesed with OSNA in 168 patients. The OSNA results were compared with conventional histopathology results from 244 patients who has undergone a SLN biopsy few years earlier. Patients characteristics were evaluated in both groups and the rate of metastases detected by both methods and the surgical procedures were compared.

Characteristics Age, median (range) Tumor size, mean (range) mm Tumor histology

Histological grade

Oestrogen receptor Progesteron receptor HER2 status Ki 67

DCIS Ductal invasive Lobular invasive Other subtypes invasive 1 2 3 Unknown Positive Negative Positive Negative Positive Negative Unknown ≤ 10% > 10% Unknown

OSNA Cohort (n = 168)

Historical cohort (n = 244)

58 (33-86) 16.63 (2-60)

57 (30-93) 14.16 (2-55)

18 130 12 8

8 192 25 19

56 77 35 2 220 24 126 42 16 146 6 72 86 10

100 88 17 39 143 25 202 42 4 26 214 95 124 25

Results: SLN metastasis were found in 45 patientes in the OSNA group and in 49 in the historical group. There were no differences in rates of macrometastases (27 by OSNA, 41 by HE) and we found differences in micrometastasis rate (18 by OSNA and 8 by HE p =0.007). Axillary node dissection (ALND) was performed in 45 patients in the OSNA group and in 49 in the historical group. All the patients diagnosed by OSNA had a complete ALND during the initial surgical procedure. In the historical cohort the ALND was performed during the initial surgical procedure in 41 patients, and ALND was performed in a second surgical procedure in 8 patients. Conclusions: The OSNA assay can detect SLN metastasis as accurately as conventional pathology with increased detection of micrometastasis. The second surgery can be reduced with the OSNA assay. Disclosure: All authors have declared no conflicts of interest.

285P

WHAT DO CLINICIANS DO WITH THE RESULTS OF THE SYSTEMATIC STAGING IMAGING AT THE TIME OF BREAST CANCER DIAGNOSIS?

A. Turpin1, A. Mailliez1, P. Vennin1, J. Bonneterre2 1 Breast Cancer, Centre Oscar Lambret, Lille, FRANCE, 2Breast Cancer, Centre Oscar Lambret, Université Lille Nord de France, Lille, FRANCE Introduction: Asymptomatic distant metastases are often looked for at the time of initial diagnosis of early breast cancer. However, there is no consensus on when to perform it and on the consequences on the treatment. Patients and methods: 125 asymptomatic women receiving systemic neoadjuvant (33 patients ( pts)) or adjuvant treatment (92 pts) for breast cancer at the Oscar Lambret Center in 09/2011 were considered. The staging imaging was a PET scan: 59 pts, a CT scan + a bone scan: 61 pts and both: 5 pts. Results for each procedure were considered normal, abnormal but typically benign or potentially malignant. In this

Volume 23 | Supplement 9 | September 2012

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

A.S. Dhadda1, B. Sethugavalar1, K. Grover2, P. Kneeshaw2, P. McManus2 1 Oncology, Castle Hill Hospital, Hull, UNITED KINGDOM, 2Breast Surgery, Castle Hill Hospital, Hull, UNITED KINGDOM

284P

Annals of Oncology

case, another imaging technique was carried out to confirm the suspected diagnosis. The patient was considered metastatic if the results of two different procedures were concordant and eventually in case of response to chemotherapy. 1 patient had a biopsy (liver). 22 pts were Stage 1, 58 Stage 2, 44 Stage 3. 1 other patient had an excision of a recurrence on mastectomy scar.

Pts Potentially malignant lesions Malignant lesions Benign lesions Confirmatory Procedures

Bone scan + CT scan

PET scan

Both Total

61 13

59 8

5 2

5 33 19

2 15 19

2 1 2

125 23 (including confirmed malignant) 9 49 40

286P

PROGNOSTIC AND PREDICTIVE VALUES OF BI-RADS CLASSIFICATION IN BREAST CANCER PATIENTS

Y. Kuo1, L. Cheng2, T. Chang1 1 Surgery, National Cheng Kung University, Tainan, TAIWAN, 2Radiology, National Cheng Kung University, Tainan, TAIWAN Objective: The goal of this study is to determine the prognostic and predictive values of the Breast Imaging Reporting and Data System (BI-RADS) classification in breast cancer patients. Patients and methods: 1045 patients with breast cancer were disgnosed between January 1, 1999, and December 31, 2007, and 512 (48.9%) of them were classified as BI-RADS 5. Overall survival and disease-free survival were estimated with the Kaplan-Meier method and compared across the two groups (BI-RADS 5 versus BI-RADS 0-4) using the log-rank test. Univariate and multivariate analyses were used to identify the prognostic factors. Results: The median follow-up time was 87.8 months. Laplan-Meier analysis showed a significant difference between the two subgroups in five-year overall survival ( p < 0.0001) and five-year disease-free survival ( p < 0.0001). On multivariate analysis, the BI-RADS mammographic findings, lymph node status, HER-2 status, and tumor grade were significant factors related to five-year overall survival and disease-free survival. Conclusion: The BI-RADS classification is a reliable prognostic and predictive factor. Taiwanese breast cancer patients with BI-RADS 5 mammographic finding showed a higher relapse rate than patients with BI-RADS 0-4 mammographic finding. Disclosure: All authors have declared no conflicts of interest.

287P

STUDY OF STROMAL CD10 OVEREXPRESSION IN INVASIVE BRAEST CANCER AND ITS RELATIONSHIP WITH CLINCOPHATHOLOGIC FACTORS

A. Taghizadeh Kermani1, A. Jafarian2, R. Ashabe Yamin2, M. Seilanian Toosi1, L. Pourali3, A. Omidi Ashrafi2 1 Oncology, MUMS, Mashhad, IRAN, 2Pathology, MUMS, Mashhad, IRAN, 3 Gynecology, MUMS, Mashhad, IRAN Background: Carcinoma of the breast is the most common non-skin malignancy in women. Invasion and metastasis are considered as the major causes of cancer-related morbidity and mortality. It has long been proposed that secreted proteinases, including the matrix metalloproteinases, play an important role in tumor progression and mediating extracellular matrix remodeling. More recently, it has been suggested that extracellular proteinases also regulate growth factors and cytokines that may contribute to tumor progression. CD10 is a 90-110kd cell surface zinc-dependent metalloprotease. Since CD10 is structurally similar to matrix metalloproteinase and

Volume 23 | Supplement 9 | September 2012

288P

PROGNOSTIC TOOLS IN EARLY BREAST CANCER: PREDICTING BENEFIT OF ADJUVANT CHEMOTHERAPY

E.E. Parkes1, C. Davidson1, A. Hussain1, C.R. James1, G.G. Hanna2 1 Medical Oncology, Northern Ireland Cancer Centre, Belfast, UNITED KINGDOM, 2Centre for Cancer Research and Cell Biology, Queens University Belfast, Belfast, UNITED KINGDOM Introduction: Adjuvant chemotherapy (CT) in early breast cancer reduces the risk of mortality. However, absolute reductions in mortality can be small. For patient with low risk disease prognostic tools such as ‘Adjuvant! Online’ (AO) and ‘Predict’ (PD) can be used to estimate the benefit of adjuvant chemotherapy. We compare the survival gains estimated using AO and PD in routine clinical practice, assessing the characteristics of patients in which AO and PD disagree. Methods: In a retrospective study using the hospital electronic database, the clinical and pathological details of all patients with early breast cancer referred for adjuvant therapy at the Northern Ireland Cancer Centre in a 3 month period in 2011 were collected and were entered in to AO and PD to assess percentage benefit (absolute reduction in mortality at 10 years) from CT. We categorised patients into three prognostic groups: those where risk from CT outweighs benefit (<2% predicted benefit), marginal benefit (2 to 5%) and significant benefit from CT (>5%) We excluded patients with metastatic disease at presentation, DCIS, a second primary breast cancer or receiving neo-adjuvant treatment. Results: Of the 200 patients identified, 43 (21.5%) fell in to different prognostic groups depending on whether AO or PD was used to calculate benefit from CT. In total, AO suggested marginal or significant benefit in 69.8% of patients, compared to 60.4% using PD. Eight patients had “major” comorbidities, which is weighted only in AO, and were excluded in subsequent analysis. Of those without major comorbidities, AO offered at least 2% benefit in 80% of cases, and PD in only 57.1%. The majority (91.4%) of cases were ER positive, and node negative (82.9%). This difference was notable in women aged 65 or less, with 83.3% with >2% benefit using AO, and 61.1% using PD. AO estimates of benefit were on average 3.7% higher for this age group. HER2 status had little impact, with similar recommendations using either AO or PD. Conclusions: This study highlights lack of concordance between two available online prognostic tools, notably in ER positive, node negative patients. For patients with a marginal benefit from CT, care must be used when making adjuvant treatment decisions. Disclosure: All authors have declared no conflicts of interest.

289P

A STUDY OF THE IMPACT OF THE 21-GENE BREAST CANCER ASSAY ON THE USE OF ADJUVANT CHEMOTHERAPY IN WOMEN WITH BREAST CANCER IN A MEXICAN PUBLIC HOSPITAL

J.E. Bargallo-Rocha1, F. Lara1, R.J. Shaw Dulin1, V. Perez-Sánchez1, C. Villarreal-Garza1, H. Maldonado-Martinez1, A. Mohar-Betancourt2, C. Yoshizawa3, E. Burke4, C. Chao4 1 Breast Cancer Clinic, Instituto Nacional de Cancerología (INCan), Mexico City, MEXICO, 2Biomedical Research, Instituto Nacional de Cancerología (INCan), Mexico City, MEXICO, 3Biostatistics, Genomic Health, Inc., Redwood City, CA, UNITED STATES OF AMERICA, 4Medical Affairs, Genomic Health, Inc., Redwood City, CA, UNITED STATES OF AMERICA Background: The majority of early stage breast cancer patients in Mexico are treated through the public health system, and >80% are treated with chemotherapy (CT). The 21-gene assay (Oncotype DX®) is available in Mexico but has been utilized primarily within the private health system. There are no data that describe the potential impact of the assay on adjuvant treatment decision-making for public

doi:10.1093/annonc/mds392 | ix

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

Results: 9 pts were considered metastatic (7%): 5 bone, 2 liver, 1 lung and 1 spleen. There was 1 Stage 1, 4 Stage 2, and 4 Stage 3. 3 of the 5 pts in the neoadjuvant setting were operated on (2 after the planned chemotherapy: 1 complete response of lung metastasis and 1 sternal lesion treated by boost and 1 after modified protocol for unique liver lesion which will be treated by Cyberknife®). The 2 others had bone lesions, received modified protocol and did not have surgery. For the pts in the adjuvant setting, 2 received the full treatment (1 continued Trastuzumab beyond the year planned), 1 received modified protocol of chemotherapy then radiotherapy and hormonotherapy and the last one received chemotherapy alone. Conclusion: In this population of 125 pts selected for (neo) adjuvant chemotherapy, 40 imaging procedures were performed to check the lesions. 9 pts were considered as metastatic and only 5 had had a significant modification of their treatment. A systematic imaging staging in pts selected for adjuvant chemotherapy should not be recommended. Disclosure: All authors have declared no conflicts of interest.

stromelysin it may facilitates cancer cell invasion and/or metastasis. The aim of this study was to investigat the rate of CD10 expression in the stromal cells of invasive ductal carcinomas of breast immunohistochemically and clarify its corelation with other clinicopathological factors of the disease. Methods: 100 patients with histopathologic diagnosis of IDC and 50 patients with fibroadenoma of breast (as the control group) were selected and 150 parafin blocks were obtained. The stained slides by IHC method for CD10 marker were examined separately by two pathologists and discrepancies were reviewed in a common session to get the final result. Results: Stromal CD10 was detected in 28% of the IDC. No immunoreactivity was identified in the stromal cells of normal breast. Stromal CD10 expression in IDC was significantly correlated with tumor size (P < 0.001), histologic grade (P < 0.001), the presence of nodal metastases (P < 0.001) and estrogen receptor negative status (P = 0.003). Conclusion: Stromal CD10 expression in IDC is closely correlated with invasion and metastasis and it may play an important role in the pathogenesis of IDC. Disclosure: All authors have declared no conflicts of interest.

Annals of Oncology

sector patients. The aim of this study is to characterize how assay results impact the decision-making process and confidence of public sector physicians in Mexico to determine adjuvant therapy. This is the first decision impact study of the Oncotype DX assay in Latin America. Methods: At total of 98 consecutive patients with ER + , HER2-, stage I-IIIa, N0/ N1-3 breast cancer from the Instituto Nacional de Cancerologia in Mexico City, Mexico, were enrolled in the study. Via consensus discussion in multidisciplinary team meetings, physicians completed pre- and post-assay questionnaires regarding adjuvant treatment recommendations for each enrolled patient. The primary endpoint was the overall change in physician treatment recommendations resulting from the addition of the Recurrence Score® result in the decision-making process. Results: Pre- and post-assay results were available for 96 patients. Treatment decisions changed for 31/96 (32%: 95% CI 23%-43%) patients; 17/62 (27%; 95% CI 17%-40%) N0 and 14/34 (41%; 95% CI 25%-59%) N1-3 patients. Post-assay, 8/50 (16%) of patients initially recommended hormonal therapy (HT) were recommended chemohormonal therapy (CHT) or CT, and 21/46 (46%) of patients initially recommended CHT/CT were recommended HT alone. The proportion of patients recommended CT decreased from 48% pre- to 34% post-assay ( p = 0.024), a decrease of 14% overall, 6% in N0, and 26% in N1-3 patients. Conclusions: These results suggest that use of the 21-gene assay in the Mexican public health system has an impact on adjuvant treatment recommendations and may reduce the use of chemotherapy.

Adjuvant Therapy HT alone CT + HT CT alone Total HT alone 42 (44%) 7 (7%) 1 (1%) 50 (52%) CT + HT 21 (22%) 23 (24%) 2 (2%) 46 (48%) Total 63 (66%) 30 (31%) 3 (3%) 96

Disclosure: C. Yoshizawa: Genomic Health- Employment and Stockholder. E. Burke: Genomic Health- Employment and Stockholder. C. Chao: Genomic HealthEmployment and Stockholder. All other authors have declared no conflicts of interest.

290P

BROWN FAT SEEN ON FDG PET/CT IS INCREASED IN BREAST CANCER PATIENTS COMPARED TO THEIR AGEAND WEIGHT-MATCHED CONTROLS WITH OTHER CANCERS

K.H.R. Tkaczuk1, Q. Cao2, L. Jones1, M. Smith2, S. Chumsri1, J. Jenkins2, V. Dilsizian2, W. Chen2 1 Medicine, University of Maryland Greenebaum Cancer Center, Baltimore, MD, UNITED STATES OF AMERICA, 2Radiology, University of Maryland, Baltimore, MD, UNITED STATES OF AMERICA Purpose: We previously found increased brown fat deposits in mammary tissue of Rrca1 mutant breast cancer mouse models suggesting a potential role of brown fat environment in the early breast tumorigenesis. The goal of the current human study is to test the hypothesis that the prevalence of brown fat activity seen on FDG PET/ CT is increased in breast cancer (BC) patients. Methods: We conducted a retrospective study to assess the distribution and intensity of brown fat activity on FDG PET/CT in female BC patients compared to age- and weight-matched control subjects with other cancers mostly colon cancer. We analyzed 124 FDG PET/CT scans of BC patients done at the University of Maryland and 124 age- and weight-matched control subjects who had FDG PET/ CT scan on the same day for staging of other cancers (the majority were colon cancer). Results: The prevalence of brown fat was higher in BC (12.9% or 16/124) than in their age- and weight-matched control subjects (5.6% or 7/124) ( p < 0.05). When the data was stratified by age, among those who were ≤ 50 years old, the prevalence of brown fat was 35.5% (11/31) in BC patients versus 9.1% (3/33) in the controls ( p<0.05). In contrast, among those who were > 50 years of age, there was no difference in brown fat prevalence between BC patients and controls (5.4% or 5/93 vs 4.4% or 4/91; p = NS), respectively. Brown fat was more commonly identified in the bilateral supraclavicular regions in BC patients than controls (22 vs 6, p = 0.049). There was no difference in the intensity of brown fat between the 2 groups (mean SUV max = 3.5 ± 1.5 in BC vs 3.4 ± 0.7 in controls, p = NS). Conclusion: The prevalence of brown fat seen on FDG PET/CT is increased in BC patients compared to their age- and weight-matched controls with other cancers, particularly in patients aged <= 50, and in the supraclavicular region. These clinical data provide further support to the experimental studies that brown fat may play a role in breast cancer tumorigenesis. Disclosure: All authors have declared no conflicts of interest.

ix | Abstracts

P53 CAN PREDICT OUTCOME AFTER TAXENE-BASED ADJUVANT CHEMOTHERAPY IN PATIENTS WITH EARLY BREAST CANCER

H.Jung An1, H-J. Lee2, G. Gong2, K-P. Kim1, J-H. Ahn1, K.H. Jung1, J.E. Kim1, B.H. Son3, S-H. Ahn3, S-B. Kim1 1 Department of Oncology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, KOREA, 2Department of Pathology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, KOREA, 3Department of Surgery, Asan Medical Center, University of Ulsan College of Medicine, Seoul, KOREA Background: Taxene-based adjuvant chemotherapy is widely accepted in lymph node (LN)-positive breast cancer, but few biomarkers exist to predict clinical outcomes. We prospectively investigate the role of potential biomarkers in LN-positive breast cancer population homogenously treated with docetaxel-containing adjuvant chemotherapy. Methods: From 2007 to 2008, 218 patients with LN-positive breast cancer who received adjuvant chemotherapy (4 cycles of adriamycin 60 mg/m2 and cyclophosphamide 600 mg/m2 followed by 4 cycles of docetaxel 100 mg/m2) were enrolled in this analysis. Immunohistochemical stainings for ER, PR, p53, topoisomerase-IIα, β-III tubulin, Tau, CYP3A4 and Her2/neu were performed in formalin-fixed, paraffin-embedded tissue section and their relationships with prognosis were investigated. Results: With a median follow-up of 58.3 months (range, 10.9-62.9), p53 and topoisomerase-IIα over-expression were associated with poor relapse-free survival (RFS) and overall survival (OS) along with tumor size >2 cm, number of positive LN (≥4), high tumor grade (G3), and negativity of hormone receptor in univariate analyses. However, in multivariate analyses, number of positive LN ≥4 (HR = 2.47; 95% CI, 1.07-5.74; p = 0.035), high tumor grade (HR = 2.76; 95% CI, 1.08-7.09, p = 0.034), and over-expression of p53 (HR = 6.55; 95% CI, 2.40-17.85, p < 0.001) were independent poor prognostic factors for RFS. And p53 over-expression was also related to poor OS (HR = 3.93; 95% CI, 1.04-14.81, p = 0.044). Patients with high p53 expression (≥10% of positive cells) tended to have tumors with large size, higher grade and lower hormone receptor positivity compared to those of low p53 expression. Conclusions: p53 over-expression along with number of positive LN and tumor grade was found to be useful biomarker to predict poor outcomes in patients with LN-positive breast cancer receiving docetaxel-based adjuvant chemotherapy. Disclosure: All authors have declared no conflicts of interest.

292P

SERUM LEVELS OF VEGF PRE- AND POST-TREATMENT WITH BEVACIZUMAB (BEV) FOR EARLY STAGE BREAST CANCER (ESBC): ICORG 08-10

A.M. Canonici1, I. Parker2, T. O’Shea2, B. Moulton2, G. Gullo3, M.J. Kennedy4, D. Tryfonopoulos5, N. Walsh6, N. O’Donovan6, J.P. Crown7 1 NICB, Dublin City University, Dublin, IRELAND, 2ICORG, Dublin, IRELAND, 3 St Vincent’s University Hospital, Department of Medical Oncology, Dublin, IRELAND, 4Dept Medical Oncology, St James’s Hospital, Dublin, IRELAND, 5 Medical Oncology Unit, St Vincents University Hospital, Dublin, IRELAND, 6 National Institute for Cellular Biotechnology, Molecular Therapeutics for Cancer Ireland, Dublin City University, Dublin, IRELAND, 7Department of Medical Oncology, St Vincents University Hospital, Dublin, IRELAND Background: Angiogenesis, partly mediated by vascular endothelial growth factor (VEGF), promotes metastases. BEV, an anti-VEGF monoclonal antibody which has some efficacy in metastatic BC is being studied as an adjuvant treatment for ESBC. However, there are no validated biomarkers, and the effects of BEV on VEGF levels in pts with ESBC is unknown. We studied VEGF serum concentration in ESBC receiving BEV. Methods: 106 pts with HER-2 negative ESBC were included in this study. Patients received 4 cycles of docetaxel (75 mg/m2) + cyclophosphamide (600 mg/m2) with BEV (15 mg/kg) once every three weeks for one year. Serum samples were collected prior to commencement of treatment, at 6 months and 12 months. The VEGF levels in serum samples were determined, at each time point, using a VEGF enzyme-linked immunosorbent assay (ELISA). Results: VEGF concentration was determined in serum samples from 65 patients. Serum VEGF was detectable in 62 of 65 patients and the median level in untreated patients was 325.4 pg/ml (range 0.1 - 924.8 pg/ml). All of the 62 patients showed a significant decrease in VEGF concentration after 6 and 12 months of treatment (median 9 ± 42.1 pg/ml, p < 0.001 and 9 ± 43.9 pg/ml, p < 0.001 respectively). No significant change in median VEGF levels observed at 12 months compared to 6 months (median 0 ± 60.3 pg/ml, p = 0.704). However, in 14 patients the levels of VEGF detected at 12 months were higher than at 6 months following treatment ( p = 0.045). Conclusion: Adjuvant therapy with chemotherapy and BEV is associated with a significant reduction in VEGF levels at 6 months. Disclosure: All authors have declared no conflicts of interest.

Volume 23 | Supplement 9 | September 2012

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

Post-Assay Pre-Assay

291P

Annals of Oncology 293P

QUANTIFICATION OF TUMOR-SPECIFIC DNA IN BLOOD OF HEALTHY WOMEN AND BREAST CANCER PATIENTS

S.N. Tamkovich1, V.A. Myleiko1, A.V. Starikov2, V.I. Permyakova3, V.V. Vlassov1, P.P. Laktionov1 1 Cellular Biology Group, Institute of Chemical Biology and Fundamental Medicine, Novosibirsk, RUSSIAN FEDERATION, 2Mammology Department, National Novosibirsk Regional Oncologic Dispensary, Novisibirsk, RUSSIAN FEDERATION, 3Hematology Department, Central Clinical Hospital, Novisibirsk, RUSSIAN FEDERATION

CONCOMITANT USE OF ANTIDEPRESSANT DRUGS AND TAMOXIFEN

L. Binkhorst1, R.H.J. Mathijssen1, M.P.P. van Herk-Sukel2, M. Bannink3, E.A. C. Wiemer1, T. van Gelder4 1 Medical Oncology, Erasmus MC-Daniel den Hoed Cancer Center, Rotterdam, NETHERLANDS, 2Oncology, PHARMO Institute, Utrecht, NETHERLANDS, 3 Psychiatry, Erasmus MC, Rotterdam, NETHERLANDS, 4Hospital Pharmacy, Erasmus MC, Rotterdam, NETHERLANDS The CYP2D6 enzyme is primarily involved in the metabolism of tamoxifen into its most important metabolite endoxifen. The efficacy of tamoxifen therapy may be influenced by genetic polymorphisms of this enzyme and CYP2D6-inhibiting co-medication (e.g. antidepressant drugs). It is therefore generally recommended to avoid potent CYP2D6 inhibitors (e.g. paroxetine, fluoxetine) in tamoxifen patients. In this study, we examined whether women using tamoxifen still received strong CYP2D6-inhibiting antidepressants concomitantly during the period 2005-2010. Drug dispensing data for tamoxifen and seven antidepressants associated with CYP2D6 inhibition ( paroxetine, fluoxetine, sertraline, fluvoxamine, (es)citalopram and venlafaxine), were obtained from the community pharmacy database of the PHARMO Institute. This database contains dispensing data of > 3 million people in the Netherlands. Concomitant use of CYP2D6-inhibiting antidepressants in women using tamoxifen, as well as the use of antidepressants in the total population, were determined per year using PASW Statistics 17.0 (SPSS Inc., Chicago, IL). Prevalence of the use of the four most common antidepressants among tamoxifen users, expressed as the number of women receiving an antidepressant concurrently with tamoxifen per 1000 tamoxifen users, is shown in the table. Although there is a downwards trend for paroxetine, this drug is still one of the most frequently used antidepressants in tamoxifen patients. In addition, a similar trend was observed in the total population of the database. Venlafaxine and citalopram, associated with weak CYP2D6-inhibiting properties, are increasingly co-prescribed.

2005 2006 2007 2008 2009 2010 294P

HIGH FREQUENCY OF BRCA 1/2 MUTATIONS AMONG ISRAELI NON ASHKENAZI BREAST CANCER PATIENTS

F. Elyan1, O. Maimon2, A.F. Salah3, M. Sagi1, L. Kaduri2, I. Lerer1, Y. Goldberg1, T. Hamburger2, D. Bercovich4, T. Peretz-Yablonski5 1 Department of Human Genetics, Hadassah-Hebrew University Medical Center, Jerusalem, ISRAEL, 2Sharett Institute of Oncology, Hadassah-Hebrew University Medical Center, Jerusalem, ISRAEL, 3Oncology and Radiotherapy, Hadassah University Hospital Hebrew University Medical Center, Jerusalem, ISRAEL, 4CgaGalil Genetic Analysis and Migal- Galilee Bio- Technology Center, CGA- Galil Genetic Analysis and Migal- Galilee Bio- Technology Center, Galil, ISRAEL, 5 Sharett Institute of Oncology, Hadassah University Hospital Hebrew University Medical Center, Jerusalem, ISRAEL Background: Inherited mutations in the breast cancer susceptibility genes (BRCA1 and BRCA2) are associated with a high risk of developing breast (BC) and ovarian cancers (OC) in females of different age and ethnic groups. The spectrum of mutations in these genes varies between populations, with some showing a high frequency of unique mutations. Ashkenazi Jews have a high rate of founder mutations in BRCA1/2, in some other Jewish communities in Israel (Jews who immigrated to Israel from Iraq, Iran and Yemen), other founder mutations had been identified. Still high proportions of Israeli BC cases with strong family history have none of the known mutations at the BRCA1/2 genes. Methods and patients: Over 3000 breast cancer patients were evaluated in the oncogentic clinic during 1997-2011. One hundred thirty seven of them underwent full sequencing of the BRCA1/2 genes due to strong family history of breast and/or ovarian cancer or young age at presentation. Clinical and pathological characteristics of these patients were evaluated. Result: Sixty seven percent were non Ashkenazi Jews, Mean age at BC onset was 44 (22-77). In 20 patient (15%) BRCA 1(N = 8) or 2 (N = 12) mutations were identified. Three of the carriers had bilateral BC and 5 had OC as second primary .17 were of non Ashkenazi origin. Ninety five percent of the carriers had first degree family history of breast or ovarian cancer. The pathological information was available in half of the carriers. All had high grade (2-3) tumor, 90% of them were HER2 negative and 60% ER positive. Age at presentation had no effect on BRCA1/2 status. BRCAPRO is a predictive model to assess BRCA status and has assessed in 9 carriers, in 6 of them, the probability was >80% and in one - 5%. Conclusions: A full sequence of the BRCA1/2 genes was performed in a selected group pf BC patient, who were negative for the common founder mutation in Israel. Fifteen percent were found to carry other BRCA1/2 mutations. We recommend that, patients with the following clinical features: Sephardic origin, first degree family history of BC or OC, high grade tumor, HER-2 negative, should be offered full BRCA1/2 testing. The role of BRCAPRO and other predictive model should be further evaluated. Disclosure: All authors have declared no conflicts of interest.

Volume 23 | Supplement 9 | September 2012

Paroxetine Fluoxetine Citalopram Venlafaxine

47.1 9.5 17.0 18.3

44.3 10.6 23.5 23.8

46.6 9.8 22.4 28.2

41.7 10.7 26.2 27.5

36.9 10.8 25.2 29.1

28.8 9.7 28.5 30.8

Despite the strong biological rationale to avoid potent CYP2D6-inhibiting co-medication in tamoxifen treated patients, paroxetine is still often used along with tamoxifen. In clinical practice, one should be extremely alert to co-medication in these patients. It is advised that strong CYP2D6 inhibitors are switched to weaker CYP2D6-inhibiting alternatives, if possible. Disclosure: All authors have declared no conflicts of interest.

296P

ITEM RESPONSE THEORY AND FACTOR ANALYSIS AS MEAN TO CHARACTERIZE OCCURRENCE OF RESPONSE SHIFT FOR LONGITUDINAL QUALITY OF LIFE STUDY IN BREAST CANCER PATIENTS

A. Anota1, C. Bascoul-Mollevi2, F. Guillemin3, T. Conroy4, M. Velten5, D. Jolly6, M. Mercier7, S. Causeret8, T.S. Dabakuyo9, F. Bonnetain9 1 Plateform of Clinical Research "Quality of Life and Cancer", University hospital of Besançon, Besançon, FRANCE, 2Biostatistic Unit, Centre Val-d’Aurelle - Paul Lamarque, Montpellier, FRANCE, 3Clinical Epidemiology and Evaluation Department, CIC-EC, Nancy, FRANCE, 4Oncology, Centre Alexis Vautrin, Vandoeuvre les Nancy Cedex, FRANCE, 5Epidemiology and Public Health Laboratory, College of Medicine, Strasbourg, FRANCE, 6Création Du Pôle Recherche Et Innovations, University Hospital, Reims, FRANCE, 7Clinical Research Unit In Quality of Life, CHU Jean Minjoz, Besancon, FRANCE, 8 Surgery, Centre Georges François Leclerc, Dijon, FRANCE, 9Biostatistic and Epidemiological Unit(ea 4184), Centre Georges François Leclerc, Dijon, FRANCE Aims: Health-related quality of life (HRQoL) is a dynamic process which depends on the adaptation of the patient and reflected by a Response Shift (RS) effect. RS results in a recalibration, a reprioritization and a reconceptualization of key HRQoL domains. Longitudinal analyses of HRQoL have to take into account the possible occurrence of RS. However, there is no standard of statistical analysis to characterize RS. Two complementary methods are investigated to characterize RS. Methods: This work builds on data of a prospective multicenter study including all primitive breast cancer patients or suspicion. HRQoL was evaluated using the EORTC QLQ-C30 and QLQ-BR23 at baseline, after surgery, at 3 months and 6 months, according to the « then-test/post-test » design: the retrospective assessments done after surgery and at 3 months refer to baseline HRQoL; the retrospective measurement done at 6 months refers to HRQoL at 3 months. The order then-test and post-test of HRQoL questionnaires was randomized. Recalibration was explored by Multiple Correspondence Analyses (MCA) and the

doi:10.1093/annonc/mds392 | ix

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

Despite of tumor location DNA from cancer cells was shown to got in circulating blood, thus DNA circulating in blood represent the valuable source of diagnostic material for non-invasive blood based tests for cancer. Material and methods: Blood from healthy female (n = 50) and breast cancer patients (n = 23, T1-2, N0-1, M0) were fractionated into plasma and cellular fractions, cell-surface-bound cirDNA (csbDNA) were eluted from cell surface with PBS/EDTA and trypsin solutions [1]. Total DNA was quantified with TaqMan PCR for LINE-1 repeats [2], methylated DNA of RARβ2 gene promoter was quantified with methylation-specific SYBR Green Real Time PCR [3]. Results: Breast cancer patients had significantly higher mean concentration of methylated RARβ2 gene promoter in circulating DNA (cirDNA) and csbDNA fractions as compared to healthy controls: 192 vs. 132 pg/ml and 619 ng/ml vs. 413 pg/ml, respectively; P < 0.005. No correlation was found between cirDNA levels and stage of tumor. ROC-curve analysis demonstrated sensitivity and specificity of cirDNA-based breast cancer detection as 52% and 65% respectively for cut-off value of 0.61 for plasma tumor cirDNA and 70% and 61% respectively for cut-off value of 0.75 when cirDNA and csbDNA were analysed simultaneously. The data obtained demonstrate that along with cirDNA csbDNA provides a valuable source of material for breast cancer diagnostics. 1. Clin. Chem., 2005. V. 51. P. 1317-1319. 2. Anal. Biochem., 2011. V. 408. P. 354-356. 3. Eur. J. Cancer Prev., 2011. V. 20. P. 453-455. Disclosure: All authors have declared no conflicts of interest.

295P

Annals of Oncology

Linear Logistic Model with Relaxed Assumptions (LLRA) of Item Response Theory (IRT). LLRA gives trend of item easiness parameters. Reprioritization and reconceptualization were explored by Principal Component Analyses (PCA). Results: Between February 2006 and February 2008, 381 patients were included, 90% had a confirmed breast cancer. PCA show a secondary reprioritization of the QLQ-C30’s dimensions. Fatigue and pain remain priority symptoms. Secondary symptoms are insomnia at baseline, diarrhea after surgery, nausea and vomiting at 3 months and 6 months. A stronger and stronger link between functional scales reflects a reconceptualization. Main recalibration’s profiles reflected by MCA are from one modality to an adjacent one. A lower or upward recalibration of each dimension is reflected by the IRT model. Based on retrospective assessment at six months of HRQoL at three months, arm and breast symptoms were overestimated with trend parameters equal to -0.59 and -1.05 ( p < 0.001). Conclusions: IRT models have mainly been used to validate HRQoL questionnaires. This work shows their interest to characterize occurrence of RS. Further analyses should be lead to validate their abilities to characterize all RS components. Disclosure: All authors have declared no conflicts of interest. 297P

PROFESSIONAL ATTITUDES TOWARDS TREATMENT OF RARE HISTOLOGICAL SUBTYPES OF CARCINOMAS OF THE BREAST: AN INTERNATIONAL PRACTICE SURVEY (PARIS STUDY)

Background: WHO classification has identified a dozen rare subtypes that account for less than 10% of all breast cancers (BC). They are generally not taken into account in international/national treatment (trt) guidelines. We evaluated professionals’ attitudes towards decision making regarding rare BC subtypes and their need for consensus guidelines. Methods: This international survey was conducted among 236 international BC experts including surgeons, pathologists, medical oncologists and radiation oncologists. Experts were chosen to be representative of all regions of the world. They were selected if part of regional, national or international BC guidelines committee(s), or active member of a BC-directed scientific society, or author > 50 publications in the field of BC. They were contacted through email up to three times and were asked to fill an online questionnaire. Results: 86 experts from 32 countries responded (36%). A total of 50% were medical oncologists, 21% surgeons, 17% pathologists and 12% radiation oncologists. 77% of the experts declared they based their general BC care decisions on regional, national or international guidelines. Up to 76% declared individual routine trt decisions for BC were taken by multi-disciplinary boards in their practice. However, only 10% strongly considered rare BC should be treated following national or international guidelines without specific individualisation. Interestingly, 50-80% described individualising treatment for rare breast cancers according to pathological subtype of the tumour in their practice. Three specific clinical situations of rare BC presented were associated with > 50% heterogeneity in trt decision among the 86 experts. Answers were associated with both region of residency and medical subspecialty. However, more than 90% of experts answered they would welcome international recommendations for rare BC subtypes. Conclusion: Rare BC subtypes are diagnosed and managed very heterogeneously depending on geographical location, habits and specialist training of medical professionals. There is need for international consensus guidelines regarding their diagnosis and trt. Disclosure: All authors have declared no conflicts of interest. 298P

PROFILE OF MALE BREAST CANCER: SINGLE INSTITUTION STUDY OF 76 PATIENTS FROM NORTHERN INDIA

A. Gogia1, V. Raina2, S.V.S. Deo3, B.K. Mohanti4, N.K. Shukla3 1 Medical Oncology, All India Institute of Medical Sciences (AIIMS) Institute Rotary Cancer Hospital, New Delhi, INDIA, 2Dept. of Medical Oncology, All India Institute of Medical Sciences (AIIMS) Institute Rotary Cancer Hospital, New Delhi, INDIA, 3Surgical Oncology, AIIMS, New Delhi, INDIA, 4Radiation Oncology, All India Institute of Medical Sciences (AIIMS) Institute Rotary Cancer Hospital, New Delhi, INDIA Background: Male breast cancer (MBC) is a rare disease and accounts for 1 % of all breast cancer. The aim of our study was to assess clinical, pathological parameters and outcome in MBC. Methods: This analysis was carried out in 76 male breast cancer patients with confirmed case of MBC who were registered in our breast cancer clinic between

ix | Abstracts

299

ASSOCIATION OF ALPHA2A AND BETA2 ADRENOCEPTORS EXPRESSION WITH CLINICAL OUTCOME IN BREAST CANCER

Y. Du, J. Lu Breast Surgery, Shanghai Cancer Center Fudan University, Shanghai, CHINA Previous studies have shown that alpha2-adrenoceptor antagonists can inhibit the growth of breast cancer cells. The action of beta2 adrenoceptors on cell proliferation is still controversial. In this study, we investigated the association of alpha2a and beta2 adrenoceptors expression with tumor-relevant biological markers and clinical outcome. Immunohistochemistry assay was performed on paraffin-embedded tumors to detect alpha2a and beta2 adrenoceptor expression in 220 operable breast tumors. We initially demonstrated that alpha2a expression was associated with Her-2 status (P = 0.048) and a marginal significance was observed between alpha2a expression and ER (P = 0.061). We also find that beta2 expression was not associated with any tumor-relevant biological markers. Kaplan-Meier models trended to show associations between different beta2 expression and DFS (P = 0.092). Further analysis was made and we found in hormone receptor positive breast cancer patients, strong beta2 expression correlated with better DFS than weak beta2 expression (P = 0.031). Multivariate analysis demonstrated that beta2 adrenoceptor (HR = 0.31, P = 0.039) and lymph node (HR = 2.85, P = 0.031) were independent predictors of DFS in hormone receptor positive breast cancer patients. In conclusion, this study showed that strong alpha2a expression occurred in ER positive and Her-2 negative breast cancers. In hormone receptor positive breast cancer patients, strong beta2 expression correlates with better DFS than weak beta2 expression. In hormone receptor positive breast cancer patients, beta2 ADR could be an independent predictors of DFS. Our results also suggest that the expression of beta2 adrenoceptor may be regulated by estrogen and progesterone or intercross may exist between the two pathways of beta2 adrenoceptor and hormone receptor. Disclosure: All authors have declared no conflicts of interest.

300

THE PERCENTAGE OF CD44-/CD24-/LIN- CELLS IN PRIMARY BREAST TUMORS AND CORRELATION BETWEEN PROGNOSTIC FACTORS

E. Yetimoglu1, D. Cabuk2, E. Karaoz3, S. Kaya2, S. Temiz2, O. Acikgoz2, K. Uygun2, Z. Canturk4, E. Yirmibesoglu5, G. Aksu5 1 Internal Medicine, Kocaeli University Medical School Hospital, Kocaeli, TURKEY, 2 Medical Oncology, Kocaeli University Medical School Hospital, Kocaeli, TURKEY, 3Kocaeli University Center for Stem Cell and Gene Therapies Research and Practice, Kocaeli University Medical School Hospital, Kocaeli, TURKEY, 4 General Surgery, Kocaeli University Medical School Hospital, Kocaeli, TURKEY, 5 Radiation Oncology, Kocaeli University Medical School Hospital, Kocaeli, TURKEY Introduction: Breast cancer (BC) is associated with different molecular profiles and histological types. Different histological types contain cells with different phenotype and genotype. Many hypotheses have been taken for understanding the causes of breast cancer’s heterogenecity. One of these is breast cancer stem cell hypothesis. The hypothesis was proposed that cancers contain a subpopulation of cells similar to epithelial stem cells. There are many methods used for identification of BC stem cells. One of these methods is using cell surface markers. BC stem cells are defined by the phenotype CD44 + /CD24-/low/Lin-. The aim of this study is to investigate the percentages of CD44 + /CD24-/low/Lin- cells and the correlation between stem cells and prognostic factors in primary BC. Material and method: Twenty three women who underwent surgery for BC between May 2010 and January 2011 were enrolled in this study. Fresh tumor tissues were

Volume 23 | Supplement 9 | September 2012

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

Z. Fadoukhair1, D. Lefeuvre2, K. Hofert3, E. Lanoy2, R. Rahhali4, M. Mathieu5, C. Mazouni6, S. Delaloge1 1 Breast Cancer Unit, Department of Medical Oncology, Institut Gustave Roussy, Villejuif, FRANCE, 2Biosatitistics, Gustave Roussy, villejuif, FRANCE, 3Medical Oncology, University of liverpool, liverpool, UNITED KINGDOM, 4Medical Oncology, National Institute of Oncology, Rabat, MOROCCO, 5Department of Pathology, Institut Gustave Roussy, Villejuif, FRANCE, 6Breast Cancer Unit, Department of Surgery, Institut Gustave Roussy, Villejuif, FRANCE

1996-2011 at Institute Rotary Cancer Hospital, All India Institute of Medical Sciences (AIIMS). Analysis was performed with descriptive statistics, the log rank test was used to compare outcome. Results: The median age was 56 years (range 28-80). The median duration of symptoms was 10 months (range 0.5-120). Breast lump was the commonest (93%) presenting symptom (right > left side). TNM Stage distribution was stage I -3 %, stage II- 20%, stage III- 55%, and stage IV- 22%. Positive family history was elicited in 8 (10.5%) patients. The median clinical tumour size was 3.9 cm. Modified Radical mastectomy was the commonest surgical procedure. IDC was the most common histology. Sixty percent of tumours were high grade and 55% had pathological node positive disease. ER/PR, her2neu status was available in 75% and 65% respectively. ER/PR and her2neu positivity was 90% and 30% respectively. Triple negative breast cancer (TNBC) constituted 20%. Median follow up was 36 months. Ten patients had relapsed of which 70% were distant, bone being the most common site followed by lung and liver. Five years disease free (DFS) and overall survival (OS) was 40% and 60%. Higher nodal stage, tumour size (>5 c.m.), negative ER/PR status, positive her2neu status, and visceral metastasis at baseline predicted poor outcome. Conclusion: Our population had more advanced disease at presentation, higher her2neu positivity and higher triple negative status as compared to western population results in poorer outcome. Disclosure: All authors have declared no conflicts of interest.

Annals of Oncology

broken up enzymatically and CD44 + /CD24-/low/Lin- cell phenotype was idendified by using surface marker antibodies. The percentage of this phenotype was determined by surface marker expression of the cells by using flowcytometry. Results: The mean age of the patients ( pts) was 47 and 52% had early stage BC. CD44 + /CD24-/low/Lin- cells were present in all tumor tissues and the mean percentage of these cells was 1.43 ± 1.16%. The percentage of CD44 + /CD24-/low/ Lin- cells was higher in postmenopausal women, early stage, Her-2 negative and low grade tumors, but it was not statistically significant. There was inverse correlation between involved lymph node numbers and the percentage of CD44 + /CD24-/low/ Lin- cells but it was not statistically significant. There was no statistically significant correlation between the percentage of CD44 + /CD24-/low/Lin- cells and menopausal status, stage, grade, number of lymph nodes, hormone reseptors and HER-2 status. Conclusion: In our study CD44 + /CD24-/low/Lin- cells were present in all tumor tissues and the mean percentage of these cells was 1.43 ± 1.16% as shown in previous studies. But there was no statistically significant correlation between prognostic factors and the percentage of the CD44 + /CD24-/low/Lin- cells. Disclosure: All authors have declared no conflicts of interest.

89 patients (38.3%) and HER2 tissue positive expression was observed in 77 patients (33.2%) with a moderate concordance of 76.7%. Elevated serum HER2 ECD correlated with postmenopausal ( p < 0.001), high tumor grade ( p < 0.001), negativity of both estrogen ( p = 0.007) and progesterone receptors ( p < 0.001), high level of carbohydrate antigen 153 (CA153) ( p = 0.039) and tissue polypeptide specific antigen (TPS) ( p = 0.018). Conclusion: HER2 ECD, which is associated with poor prognosis, may provide more additional information compared with HER2 tissue alone. We support that it is necessary to decrease the cut-off value in evaluating serum HER2 ECD level for early stage breast cancer. Disclosure: All authors have declared no conflicts of interest.

EARLY BREAST CANCER AGRESSIVENESS DOES NOT DIFFER BETWEEN INSULIN-SENSITIVE AND INSULIN-RESISTANT POSTMENOPAUSAL NON-DIABETIC WOMEN

A. Emiliani1, A. Iannace1, G. Nigita2, T. Losanno1, G. Manna1, E. Franzese3, L. Frati1 1 Department Experimental Medicine, Sapienza University, Rome, ITALY, 2Division of Surgery, Policlinico Casilino, Rome, ITALY, 3Internal Medicine, Sapienza University, Rome, ITALY

301

Background: The metabolic syndrome is known to negatively influence breast cancer outcome. One of the syndromès components is hyperinsulinism, which can exert tumor-promoting effects directly or through increased hepatic IGF1-production. A fasting insulin concentration > 10 µIU/ml and/or a HOMA score > 2.6 are considered diagnostic for the metabolic syndrome. We performed a pilot study to evaluate the relation between insulin resistance and tumor characteristics. Study design: Prior to surgery we collected blood samples of 33 consecutive non-diabetic postmenopausal women with early breast cancer to measure fasting insulin and glucose concentration. Correlation analyses were performed for both fasting insulin and HOMA-score (insulin resistance index) and parameters tested were body mass index, mitotic activity index (MAI), Bloom Richardson score (BNR) and tumor diameter. We also compared nodal status, ER-status and Her2-status of insulin-resistant and insulin-sensitive breast cancer patients. Results: Median age was 66 (49- 86) years and median BMI was 25.8 (18.6- 45.4) kg/ m2. Median MAI was 7 (1-52), median tumor diameter was 12 (3- 70) mm and 9 patients had node-positive disease ( 2 N0itc, 5 N1, 2 N2). Insulin resistance was found in 5 (15 %) patients. BMI showed a strong positive correlation with insulin concentration and HOMA score (2-sided Pearson test; P= 0.000). Insulin resistant and insulin sensitive patients however did not differ in terms of MAI, BNR, tumor diameter, nodal status, ER-status and Her2-status. Conclusion: In postmenopausal non-diabetic women with early breast cancer insulin resistance is encountered quite often. It does however not appear to alter cancer biology. Disclosure: All authors have declared no conflicts of interest. 302 Ma1,

STUDY ON SERUM HER2 EXTRACELLULAR DOMAIN EXPRESSION IN EARLY STAGE BREAST CANCER PATIENTS Yang2,

Li3,

Wang2,

Han3,

Shi1

L. H. J. F. X. Y. 1 Medical Oncology, Cancer Institute/Hospital, Beijing, CHINA, 2Pathology, 3 Cancer, Beijing, CHINA, Clinical Laboratory, Cancer Institute/Hospital, Beijing, CHINA Background: The measurement of the human epidermal growth factor receptor 2 (HER2) protein in the serum of metastatic breast cancer patients has now been reported, but there are no consistent data to support the clinical utility of serum HER2 extracellular domain (ECD) for patients with early breast cancer. We aimed to evaluate the correlation between serum HER2 ECD levels and tumor HER2 status, and analyze their relationship with clinicopathological parameters in patients with early stage disease. Methods: A prospective study was conducted on 232 breast cancer patients with stage I-II diseases before treatment. Preoperative serum samples were measured by enzyme-linked immunosorbent assay (ELISA). Tissue HER2 status was analyzed by immunohistochemistry and fluorescence in situ hybridization assays. Results: The median serum HER2 ECD concentration was 6.8 ng/ml (range 1.3 42.1 ng/ml). The best diagnostic cut-off value was 7.4 ng/ml (with 72.9% sensitivity and 85.3% specificity), which was lower than HER2 ECD cut-off value with metastasis breast cancer (15 ng/ml). High serum HER2 ECD levels were reported in

Volume 23 | Supplement 9 | September 2012

CLINICAL RELEVANCE OF HORMONE RECEPTORS, KI67 AND HER-2 STATUS AS BIOMARKERS SURROGATE FOR BREAST CANCER MOLECULAR SUBTYPES: A RETROSPECTIVE ANALYSIS OF DISEASE FREE SURVIVAL IN PATIENTS WITH T1 N0/N+ BREAST CANCER

Background: Early breast cancer (EBC) is a heterogeneous disease with distinct clinical, pathological, and molecular features and different treatment responsiveness and outcomes. Aim of study was to evaluate outcomes according to molecular subtypes breast cancer classified by four biomarkers using immunohistochemistry (IHC). Patients and methods: We retrospectively reviewed 264 women with T1 N0/N+ breast cancer, referred to a single centre (1986 - 2009) and treated according to European guidelines. The relationships between classical prognostic factors, molecular subtypes classified by IHC and Disease free survivals (DFS) were analyzed. Results: Univariate survival analysis showed a significantly different DFS at 5- and 7-ys for N0 and N+ patient populations, with a better outcome for patients with node-negative tumors (5ys: N0 = 95.9% vs. N + =88.8, p = 0.03; 7ys: N0 = 94.8% vs. N + =86.0%, p = 0.02). Nevertheless, long-term outcome (15-ys) displayed an inversion of the survival curves with a lower DFS rate of 65% in N0 vs. 86% in N+ patients ( p < 0.0001). Regarding to Ki-67 tumor expression, patients with low Ki67 values (Ki-67 < 14%) had a better 5-ys DFS compared to patients with high Ki-67 (Ki-67 ≥ 14%). A significant difference in DFS has been observed also considering the tumor grading (G1 = 100%, G2 = 88%, G3 = 94.4%, p = 0.03). Based on molecular subtypes breast cancer classification by IHC, the 5-ys DFS rate was 98.4% for Luminal A (HR + , HER2-, Ki-67 < 14%), 96.3% for Luminal B HER2 negative (HR + , HER2-, Ki-67 ≥ 14%), 83.3% for Luminal B HER2 positive (HR + , HER + , any Ki-67), 83.6% for HER2-like (HR-, HER2+) and 74.5% for Basal-like tumors (HR-, HER2-). Conclusions: At long-term follow-up, patients with T1N0 and patients with G2 EBC showed worst outcomes, probably because they are considered to have a low recurrence risk and not received adjuvant chemotherapy. Ki-67 expression is an important prognostic factor in hormone-receptors positive disease, allowing better definition of Luminal A and B molecular subtypes. Classic prognostic factors evaluated by IHC could be used as biomarkers surrogate for breast cancer molecular subtypes and might improve therapeutic decision in EBC patients. Disclosure: All authors have declared no conflicts of interest.

304

PROGNOSTIC VALUE OF CIRCULATING TUMOR CELLS IN EARLY BREAST CANCER PATIENTS DETECTED BY RT-PCR OF MAMAGLOBIN

A.M. Hilal1, H.M. Elzawahrey1, A.A. Abd Elwahab2, M.N. Abdelhafez1, M.M. Moneer3, A.D. Darwish1 1 Medical Oncology, National Cancer Institute, Cairo, EGYPT, 2Molecular Biology, National Cancer Institute, Cairo, EGYPT, 3Epidemiology & Biostatistics, National Cancer Institute, Cairo, EGYPT Background: The identification of circulating tumor cells (CTC) could potentially become an important prognostic factor in breast cancer patient. The aim of this prospective study is to detect CTC in the blood of breast cancer patients and to correlate between detection of CTC and other prognostic factors, disease free survival and overall survival. Material and methods: The study was conducted at medical oncology department NCI, Cairo University during the period from August 2008 to August 2011. Forty two consecutive consenting female patients with non metastatic breast cancer who ended their adjuvant chemotherapy and radiotherapy at least 2 years were recruited. Detection of CTC in the blood of our patients was done by measuring the gene expression for mammaglobin by RT-PCR, and then the relative fold change was calculated relatively to normal samples.

doi:10.1093/annonc/mds392 | ix

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

H.K. van Halteren1, S.B. Bins2, W. De Roos3, A. Bosch3, J. Klein Gunnewiek4, E. Ruijter5, J. Enserink3 1 Department of Internal Medicine, Ziekenhuis Gelderse Vallei, Ede, NETHERLANDS, 2Internal Medicine, Gelderse Vallei Hospital, Ede, NETHERLANDS, 3Surgery, Gelderse Vallei Hospital, Ede, NETHERLANDS, 4 Clinical Chemistry, Gelderse Vallei Hospital, Ede, NETHERLANDS, 5Pathology, Alysis Hospital, Arnhem, NETHERLANDS

303

Annals of Oncology

Results: The median CTC fold changes were 9.3, ranging from 0 to 20.8 in the whole studied group while zero for the control group. There was a highly statistically significant difference ( p = 0.001) between CTC fold changes for stage IIIA compared to stage II tumor, and a statistical significance ( p = 0.070) in favor of higher CTC fold changes in those with Her2-neu receptor positivity. There was no significant correlation between higher CTCs and other factors related to the patients or disease characteristics. There was also no relation between CTC fold changes and overall survival or disease free survival. Conclusion: In this small study, mammaglobin is considered a sensitive marker for detection of CTC in breast cancer. CTC fold changes are correlated with tumor stage and Her2 status. Further studies including larger number of patients and followed for longer period are recommended to evaluate this protocol more thoroughly. Disclosure: All authors have declared no conflicts of interest.

305

IMMUNOHISTOCHEMICAL KI67 LABELING INDEX HAS A SIMILAR PROLIFERATION PREDICTIVE POWER AS VARIOUS FIRST-GENERATION GENE SIGNATURES IN BREAST CANCER

Background: Immunohistochemical assessment of Ki67 labeling index (IHC Ki67) has been described as a prognostic and predictive marker for breast cancer. On the other hand, several genetic prognostic markers have been described for breast cancer However, few clinical data sets provide information on correlations between genomic markers, mRNA Ki67, and IHC Ki67 in the same cohort of patients with survival data. The objective of this study is to examine the association between immunohistochemical Ki67 labeling index, Ki67 mRNA expression level, and first-generation gene signatures in a cohort of breast cancer patients. Patients and methods: We assessed associations between IHC Ki67 and first-generation gene signatures in a panel of 40 tumor samples, using an oligonucleotide microarray. Gene expression analyses included Ki67 alone (MKi67), 21-gene signature, Mitosis Kinome score signature (MKS), and Genomic grade index (GGI). Correlation coefficients were calculated by Spearman’s rank correlation test. To assess the relationship between IHC Ki67 and survival outcomes, survival curves were calculated by the Kaplan–Meier method and compared with the log-rank test according to IHC Ki67. Results: Median age at diagnosis was 51 years. Treatments included adjuvant chemotherapy (32 patients) and endocrine therapy (21 patients). IHC Ki67, MKi67, and 3 genetic markers were highly correlated in all cases (Rho: 0.71–0.79). Estrogen receptor (ER)-positive cases exhibited strong correlations between IHC Ki67 and other signatures (Rho: 0.79–0.83). ER-negative cases displayed slightly lower correlations (Rho: 0.61–0.74). In ER-positive cases, the low IHC Ki67 group exhibited significantly longer relapse-free survival (RFS) than the high IHC Ki67 group ( p = 0.007). This difference was confirmed by multivariate analysis. Conclusions: Our data indicate that IHC Ki67 exhibits similar predictive power for proliferation in ER-positive cancers as genomic markers. Further study of IHC Ki67 is needed to define prognostic factors and predictive factors for chemotherapy using central laboratory assessment. Disclosure: All authors have declared no conflicts of interest.

306

ONCOTYPE DX® - THE SíRIO-LIBANêS HOSPITAL CANCER CENTER EXPERIENCE

A.A.L. Pereira1, A.K. Shimada1, F.C. Santini1, E.C. Nascimento2, K.B. Ribeiro3, R.J. Marques1, M.S. Mano1, A. Katz4 1 Centro de Oncologia - 2° Andar, Sirio Libanes Hospital, Sao Paulo, BRAZIL, 2 Imunohistoquímica e Biologia Molecular, Diagnostika, Sao Paulo, BRAZIL, 3 Instituto de Ensino e Pesquisa, Sirio Libanes Hospital, Sao Paulo, BRAZIL, 4 Centro de Oncologia, Sirio Libanes Hospital, Sao Paulo, BRAZIL Background: The Oncotype Dx® recurrence score (RS) assay quantifies the risk of distant recurrence (rDR) and its use has increased despite the lack of prospective studies. Methods: This is a cross-sectional study of consecutive patients (PTS) from our Institution with histologically confirmed invasive breast cancer (IBC) who underwent surgery with curative intent and in whom Oncotype Dx (ODx) was performed. The main objectives were to compare (1) the predicted rDR by RS and Adjuvant! (2) Risk allocation by RS and St Gallen Criteria and (3) the agreement between histological grading (HG) and RS.

ix | Abstracts

307

STUDY ON MICROMETASTASIS IN SENTINEL LYMPH NODE OF PATIENTS WITH BREAST CANCER

D. Pérez Callejo1, E. Almagro Casado1, B. Cantos1, C. González-Lois2, A. Ruiz-Valdepeñas1, M. Palka1, Á. Ortiz2, C. Maximiano1, M. Méndez García1, M. Provencio Pulla1 1 Medical Oncology, Hospital Universitario Puerta de Hierro Majadahonda, Madrid, SPAIN, 2Anatomical Pathology, Hospital Universitario Puerta de Hierro Majadahonda, Madrid, SPAIN Introduction: The nodal ratio it’s the most important factor when delivering a prognosis in breast cancer, and survival is clearly related with the number of affected nodes. During the last few years the technique of the sentinel lymph node (SLN) has demonstrated to be a useful tool to predict the nodal status, being necessary until today to evaluate if a lymphadenectomy is needed when this node is affected. Due to the doubtful prognosis in these particular cases relating micrometastasis (NMIC) with a needed axillary lymph node dissection (ALND), in our Hospital we analyze the SLNs evaluated. Methods: We perform a descriptive retrospective study in 216 patients of the Hospital Universitario Puerta de Hierro who have gone through a SLN biopsy (SLNB) of breast carcinoma between January 2010 and March 2012, being introduced the OSNA technique in May 2011. We analyze the following factors: histology, processing method of the SLN, number of SLNs, intraoperative and definitive diagnosis, performance of ALND, number of nodes dissected along the procedure and total number of affected nodes. Results: Among 216 patients treated with SLNB, 26.8% (58) showed a positive SLN, subdivided in Nmic 8.8% (19 patients) and metastasis 18%, (39 patients). Out of the total number of nodes, 57% of them were processed with the OSNA method and the other 43% with histopathological (HP) examination. The specificity of the HP technique in these series was of a 100%, with a sensitivity of 73.33%. Out of the 58 SLN that presented SLN positive, 96.6% (56) received an ALND. Out of the 19 SLN with Nmic, 17 cases had an ALND, and 3 cases were affected again later. The other two SLNs, processed by histopathology, presented a negative intraoperative diagnosis for infiltration, and therefore it was decided not to proceed with a delayed lymphadenectomy. Conclusions: Although we still don’t know the impact of the persistence of tumoral disease in nodes with disease free survival and overall survival, with the data gathered in our study and given that more than the 15% of the nodes removed were positive, we conclude that lymphadenectomy should not be dismissed after detecting Nmic in the SLN. It is necessary to carry out studies with a larger number of patients, under a long-term monitoring, to determine the real impact on survival. Disclosure: All authors have declared no conflicts of interest.

308

POPULATION BASED OUTCOME OF HER2 POSITIVE EARLY BREAST CANCER IN TRASTUZUMAB ERA

E. Matos, C. Grasic Kuhar, B. Zakotnik Dept. of Medical Oncology, Institute of Oncology Ljubljana, Ljubljana, SLOVENIA Introduction: About 15% of breast cancer (BC) patients are HER2 positive (HER2+) with high risk of cancer recurrence and death. With adjuvant trastuzumab (T) treatment the risk of reccurence almost halved and the risk of death decreased for one third. Adjuvant treatment with T has been available in Slovenia since August 2005, immediatelly after the first positive trial results.

Volume 23 | Supplement 9 | September 2012

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

N. Niikura1, T. Iwamoto2, S. Masuda3, N. Kumaki4, M. Shirane5, K. Mori5, T. Okamura1, Y. Saito1, Y. Suzuki1, Y. Tokuda1 1 Breast and Endocrine Surgery, Tokai University School of Medicine, Kanagawa, JAPAN, 2Department of Breast and Endocrine Surgery, Okayama University Hospital, Okayama, JAPAN, 3Department of Pathology, Nihon University School of Medicine, Tokyo, JAPAN, 4Department of Pathology, Tokai University School of Medicine, Kanagawa, JAPAN, 5Product Research, Chugai Pharmaceuticals Co. Ltd, Kanagawa, JAPAN

Results: From October/2006 to April/2012, 105 PTS with ER positive IBC were evaluated. Median age: 55y. Sixty-eight (64.8%) were EC IA; axillary lymph node involvement was seen in 25 PTS (14 micro and 11 macrometastasis). The rDR by RS was low in 64 PTS (60.9%), intermediate in 34 (32.4%) and high in 7 (6.7%). According to Saint Gallen, 12 (12%), 68 (68%) and 20 PTS (20%) were classified as low, intermediate and high risk, respectively, among 100 classifiable PTS. There was no statistically significant agreement between risk allocation by RS and Saint Gallen criteria (Kappa coefficient = -0.043; p = 0.401). Ki-67 data was available for 89 PTS: <15% in 46; 15-19% in 9 and ≥20% in 34 PTS. There was correlation between RS and Ki-67 (Pearson chi-Square = 6.472; p= 0.167). Among the 80 node-negative PTS, there was no significant correlation between the predicted average rDR, using ODx, and Adjuvant!OnLine (Pearson correlation coefficient = 0.082, p = 0.792) with median risk of 10% vs 15.6%, respectively. HG was available in 103 PTS: 28 (27.2%) were grade 1, 69 (67%) grade 2 and 6 (5.8%) grade 3. A linear trend association between HG and RS was observed, ie, for increasing levels of the RS, we observed lower frequencies of grade 1 and higher frequencies of PTS with grade 2 (Linear trend chi-square = 4.62; p = 0.032). Conclusions: We found no statistically significant agreement between Oncotype Dx®, the Saint Gallen criteria, Adjuvant!Online, or Ki-67. However, a potential association between HG and RS was noted. The rDR may be overestimated by clinicopathological-based classifications Disclosure: All authors have declared no conflicts of interest.

Annals of Oncology

Aim: Retrospective analysis of relapse free (RFS) and overall survival (OS) in HER2+ BC patients treated with adjuvant trastuzumab in years 2005-09. Results: Adjuvant T received 317 patients, median age 53 (min 23; max 78) years. Ninety five percent were invasive ductal carcinoma and 97% grade 2/3, 55% were ER positive. Tumor stage was T1 (29%), T2 (48%), T3 (11%), T4 (10%), unknown (2%). N stage was N0 (26%), N1 (50%), N2 (16%), N3 (7%), unknown (1%). All patients recieved adjuvant chemotherapy, 41% anthracycline-based, 48% anthracycline and taxane- based. The median follow-up time was 4.5 (min 0.7; max 6.9) years. There were 66 relapses and 28 deaths. RFS was 98.1, 92.3, 84.2 and 80.8% at 1, 2, 3, 4 years, resp. OS was 99.7, 96.6, 93.4 and 92.5%, at 1, 2, 3, 4 years, resp. The independent prognostic factor for RFS were T stage (HR 1.2 95% CI1.0-1.4) and N stage (HR1.3 (1.1-1.5). Chemotherapy regimen was not found as a prognostic factor for RFS. Conclusions: This is one of the first reports of the population based results of the adjuvant T treatment. Our results, based on adjuvant T treatment of real-life HER2 + BC patients, are fully comparable to the results obtained in HERA and Joint American study with 4-years RFS around 80%. Disclosure: All authors have declared no conflicts of interest.

309

DELAYED CARDIAC TOXICITY IN BREAST CANCER PATIENTS TREATED WITH ADJUVANT OR NEOADJUVANT CHEMOTHERAPY (INCLUDING ANTHRACYCLINES) AND TRASTUZUMAB

311 New treatments in the (neo) adjuvant settings for HER2 positive breast cancer have recently changed their prognosis due to the systematic use of Trastuzumab (T). Progress however has been associated with a risk of cardiac toxicity. Since the incidence of cardiotoxicity is low and inconsistent across the 5 major adjuvant trials, we undertook a study to determine the cardiotoxicity rate of adjuvant or neoadjuvant sequential chemotherapy (3 FEC 100-3 Docetaxel) and T in non selected breast cancer pts treated at the Oscar Lambret Center between 2005-2007 and who did not relapse. Patients ( pts) and methods: 121 pts met the inclusion criteria and were invited for a cardiac evaluation. 79 accepted and had a clinical examination, an Electrocardiogram (ECG) and an isotopic or echocardiographic measurement of Left Ventricular Ejection Fraction (LVEF). We also retrospectively collected, LVEF values at chemotherapy initiation, at T initiation, and after 3, 6, 9 and 12 months of T. Cardiotoxicity was defined as a decrease in LVEF of more than 10% or a level of LVEF less than 55% or clinical or ECG symptoms. The 33 pts not responding to the convocation had similar characteristics to those analysed. Results: 51 pts received 1 year T, 7 pts had 6 months T and 21 pts stopped T before 1 year because of a decrease in LVEF. Median follow up was 5.1 years (3.2 to- 6.5). 66 pts had at least 4 evaluations and were considered evaluable. No pts experienced clinical or ECG symptoms. 30 pts always had a normal LVEF. 36 pts (55%) had a normal initial LVEF but a decrease during follow up; 20% appeared more than 6 months after initiation. 11 pts (16%) had previously experienced a decrease in LVEF with anthracyclines, eight of whom also had cardiotoxicity during T. No risk factor of cardiac disease was observed in pts with cardiotoxicity. No correlation was found with age, side of the tumor, previous radiotherapy. Cardiotoxicity was more frequently observed in pts with ER negative tumors (74% vs 39%). At the time of the last evaluation, no abnormal LVEF was observed. Conclusion: The cardiotoxiciy rate was unexpectedly high in these non selected pts. Nevertheless, LVEF remained normal after a median follow up of 5.1 year. Disclosure: All authors have declared no conflicts of interest.

310

ENDOMETRIAL EFFECTS OF TAMOXIFEN AND EXEMESTANE IN EARLY BREAST CANCER: FINAL RESULTS OF A RANDOMIZED PHASE III TRIAL

O. Garrone1, E. Principe2, M. Monteverde3, M. Occelli1, G. Bertelli4, B. Favilla2, T. Catzeddu1, P. Vanella1, E. Miraglio1, M. Merlano1 1 Clinical Oncology, Azienda Ospedaliera St. Croce e Carle, Cuneo, ITALY, 2 Gynecology, Azienda Ospedaliera St. Croce e Carle, Cuneo, ITALY, 3Oncology, S. Croce General Hospital, Cuneo, ITALY, 4Oncology, South West Wales Cancer Instiute, Swansea, UNITED KINGDOM Background: Tamoxifen (T) has been for a long time the treatment of choice in hormonal receptor positive early breast cancer patients (EBCP). It is known that T has partial estrogen agonistic activity on uterus determining an increase in gynecological symptoms. Long-term T treatment can lead to an increased incidence of endometrial cancer. The availability of aromatase inhibitors (AI) has modified the prescription options in postmenopausal breast cancer patients while remaining the standard of care in premenopausal setting. AI do not mediate their effects through the estrogen receptor potentially conferring tolerability advantages to women particularly in relation to their effects on the endometrium. So far some studies including all the available AI have been published which confirmed the lack of estrogen effects of AI on endometrium compared to T. In 2001 we started a phase III study comparing T and Exemestane (E) in EBCP in order to evaluate the differences in the endometrial changes recorded by transvaginal ultrasound (TVUS). Material and methods: Postmenopausal hormonal receptor positive EBCP were randomly assigned to receive T or E. A baseline TVUS was performed and repeated after 6 and 12 months. The primary end-point, endometrial thickness (ET) was evaluated with T-test. Results: From 2001 to 2008 220 patients were enrolled in our department. Patients’ characteristics were well balanced in each arm. Definitive results are summarized in the table:

EVALUATION OF AROMATASE INHIBITOR-RELATED ARTHRALGIA IN EARLY BREAST CANCER PATIENTS

N. Romero Laorden, C. Olier, X. Mielgo, S. Falagán, A. Velástegui, S. Hernando, J.C. Cámara, A. Hurtado, J. Garcia-Donas, C. Jara Medical Oncology, Hospital Universitario Fundacion Alcorcon, Alcorcon, Madrid, SPAIN Introduction: Aromatase inhibitors (AI) have demonstrated a disease-free survival benefit compared with tamoxifen as adjuvant therapy in the treatment of hormone-dependent early breast cancer in postmenopausal women. Although the AI are associated with fewer thromboembolic events and endometrial abnormalities than tamoxifen, an increase in arthralgia, skeletal, and muscle pain has been reported and cited as the primary reason for discontinuation of AI therapy. The aim of our study was to evaluate the incidence and the impact of the arthralgia syndrome in our population.

Volume 23 | Supplement 9 | September 2012

ET (median) Baseline E T

ET (median) After 6 m

p (ET baseline vs 6 ET (median) months) After 12 m

3 (1.4-5) 3 (1.17-9.3) NS 3 (0.14-5) 6.64 (1.9-19.7) <0.001

p (ET baseline vs 12 months)

2.8 (0.02-13.5) NS 6 (1.93-25.2) <0.001

Conclusions: Our results confirmed the data obtained in similar studies. In particular E was associated with significantly less endometrial thickening than T during the projected period of therapy in postmenopausal hormone receptor positive EBCP. Disclosure: All authors have declared no conflicts of interest.

doi:10.1093/annonc/mds392 | ix

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

A. Mailliez1, N. Kotecki2, C. Kouakam3, C. Fournier4, J. Bonneterre5 1 Breast Department, Centre Oscar Lambret, Lille, FRANCE, 2Breast Cancer Departement, Centre Oscar Lambret, Lille, FRANCE, 3Soins Externes, Centre Oscar Lambret, Lille, FRANCE, 4Biostatistics, Centre Oscar Lambret, Lille, FRANCE, 5Breast Cancer Department, Centre Oscar Lambret, Université Lille Nord de France, Lille Cedex, FRANCE

Methods: Postmenopausal patients with the diagnosis of early-stage hormone-sensitive breast cancer receiving AI therapy at our institution were included prospectively between 2010-2012. We performed a specific survey to evaluate articular toxicity collecting presence of arthralgias, EVA scale, use of analgesic, start time and localization of the pain. Clinical data was collected from our department database. Univariate analysis was used to compare those with AI-related arthralgia and different demographic and clinical features. Results: A total of 83 women were included, with a mean age 63.2, BMI 28.4, menopause mean age 48.2. Regarding to the features of the primary tumor, 81.9% had ductal carcinoma histology, 28.9% stage I, 54.2% stage II, 16.8% stage III. At the moment of the evaluation, 68.7% (57/83) were receiving treatment with anastrozole, 15.7% (13/83) letrozole, 15.7% (13/83) exemestane. 86.7% (72/ 83) had joint pain, of them 68.1% (49/72) related their pain to the hormone therapy, 31.9% (23/72) to previous diseases or another treatments. The articulations more frequently affected were located in the back, knees and hands. Regarding to the pain intensity, mean maximum EVA was 6.03, minimum 1.06. Only 2 of the patients affected stopped therapy, and it was changed for another aromatase inhibitor. No differences were seen between the different aromatase inhibitors. Conclusions: The incidence of arthralgia associated with aromatase inhibitors was superior in our series compared with results previously published in clinical trials. Further research in this field should be developed to clarify the prevalence and severity of joint symptoms related to AI and to define factors that may be associated with an increased risk of this treatment-related adverse effect. Disclosure: All authors have declared no conflicts of interest.

Annals of Oncology

312: Table: Chemotherapy uptake within MWRH

Decision-making tools (DMT) Nottingham Prognostic Index (NPI) · Excellent (≤ 2.40) · Good (2.41 – 3.40) · Moderate (3.41 – 5.40) · Poor (≥ 5.41) Adjuvant! Online (AO) – additional 10-year survival benefit of chemotherapy and hormonal therapy · 0.0 – 2.5% · 2.6 – 5.0% · 5.1 – 7.5% · ≥ 7.5% Oncotype Dx Recurrence Score (RS) Assay · Low (1–17) · Intermediate (18 – 30) · High (>31)

312

Chemotherapy Uptake Offered

Undertaken

Not offered

2/7 16/41 21/29 0/0

1/7 13/41 21/29 0/0

5/7 25/41 8/29 0/0

8/21 12/33 9/13 10/ 10

6/21 10/33 9/13 10/10

13/21 21/33 4/13 0/10

13/38 20/33 6/6

11/38 18/33 6/6

25/38 13/33 0/6

CORRELATION BETWEEN DECISION-MAKING TOOLS AND CHEMOTHERAPY UPTAKE IN NODE-NEGATIVE ENDOCRINE RECEPTOR EARLY-STAGE BREAST CANCER IN MID-WESTERN IRELAND

K.I. Quintyne1, B. Woulfe1, J.C. Coffey2, R.K. Gupta1 1 Mid-Western Cancer Centre (MWCC), Mid-Western Regional Hospital (MWRH), Limerick, IRELAND, 2Department of Surgery, Mid-Western Regional Hospital (MWRH), Limerick, IRELAND

313

DOCETAXEL IN THE TREATMENT OF BREAST CANCER: A MULTICENTER RETROSPECTIVE STUDY FROM CHINA

B. Xu Medical Department, Cancer Hospital, Chinese Academy of Medical Sciences, Beijing, CHINA Background: Docetaxel is considered as a fundamental drug in the treatment of breast cancer. It has been extensively incorporated into the neoadjuvant, adjuvant and metastatic treatment. The aim of this study was to investigate how breast cancer patients are treated with docetaxel in China. Methods: A retrospective review of chemotherapy with docetaxel performed from 2009 to 2011 was carried out in China. Study included all patients diagnosed with invasive breast cancer and treated with docetaxel-containing regimens in forty-two cancer centers from 12 provinces in China. Regimens were compared in different subgroups based on stage, subtype, and lymph node (LN) status. Patterns of chemotherapy were also compared to published guidelines. Results: Among 2188 breast cancer patients treated with docetaxel, 1881 (86.0%) were in adjuvant or/and neoadjuvant setting (including 91 in both settings). The mean age was 48.7 (range, 14-82). Compared with 288 patients using docetaxel as single agent therapy, more patients (86.8%) used docetaxel-containing combination regimens. The mean cycle administered and dose for every cycle was 4.8 and 73.0 mg/m2, respectively. Dose reduction and delay occurred in 409 (19.7%) patients mainly due to non-medical factors (10.9%) and hematologic toxicity (5.9%). TAC, TA, TC, TX, and AC-T regimens were given in 34.8%, 19.7%, 17.4%, 5.3%, and 2.2% of patients respectively. TAC was used more frequently in triple-negative breast cancer (TNBC) than other patients (43.0% vs. 32.7%, P = 0.004). In the (neo) adjuvant setting, TAC regimen was used more frequently in LN-positive patients

ix | Abstracts

314

MOVING FROM 2D TO 3D-CRT PLANNING OF CHEST WALL FOR POSTMASTECTOMY BREAST CANCER PATIENTS: MANSOURA UNIVERSITY EXPERIENCE

I. Awad, D. Zayed, N. Abotouk, T. Dawood Clinical Oncology, Mansoura University Hospital School of Medicine, Mansoura, EGYPT Background: This study evaluates the dose distribution of the wedged tangential beam three-dimensionally planned conformal radiotherapy (3D-CRT) compared to the previously used (5 years ago) two-dimensionally(2D) planned radiotherapy of the chest wall for postmastectomy breast cancer patients in Clinical Oncology and Nuclear Medicine Department - Mansoura University. Patients and methods: Thirty six breast cancer patients were randomized for planning by both the standard 3D-CRT and 2D-RT techniques for radiotherapy of the chest wall. Dose-volume histograms were carried out by the 3D treatment planning system. They were assessed for the PTV and organs at risk. The total dose was 50 Gy in 25 fractions. Results: The three–dimensionally planned conformal radiotherapy showed a significantly better homogeneity index of the PTV (chest wall). The ipsilateral mean lung dose was significantly reduced with the tangential beam 3D-CRT plans with an average of 24.6% (1217 cGy versus 1614 cGy). For the left sided breast cancer patients, the mean heart dose was also reduced by an average of 48.6% (718 cGy versus 1398 cGy). Conclusions: The tangential beam 3D-CRT planning demonstrated a significantly better homogeneity index for the PTV of the postmastectomy breast cancer patients with a statistically significant reduction in the mean doses of the ipsilateral lung and the heart for the left –sided breast cancer patients. Disclosure: All authors have declared no conflicts of interest.

315TiP

SAFEHER: A STUDY OF ASSISTED- AND SELF-ADMINISTERED SUBCUTANEOUS TRASTUZUMAB (H-SC) AS ADJUVANT THERAPY IN PATIENTS WITH EARLY HER2-POSITIVE BREAST CANCER (EBC)

J. Gligorov1, H.A. Azim2, B. Ataseven3, M. Delaurentiis4, K.H. Jung5, F. Herbst6, A. Llombart7, A. Manikhas8, S. Osborne9, X. Pivot10 1 Medical Oncology Department, Tenon Hospital, Paris, FRANCE, 2Clinical Oncology Department, Cairo University, Cairo, EGYPT, 3Department of Obstetrics and Gynaecology, Rot-Kreuz-Klinikum, Munich, GERMANY, 4Breast Oncology Department, National Cancer Institute “Fondazione Pascale”, Napoli, ITALY, 5Department of Oncology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, KOREA, 6Oncology, F. Hoffmann-La Roche Ltd., Basel, SWITZERLAND, 7Medical Oncology Service, Hospital Arnau de Vilanova, Valencia, SPAIN, 8Oncology, Oncology Hospital of St. Petersburg, St. Petersburg, RUSSIAN FEDERATION, 9Statistics, F. Hoffmann-La Roche Ltd., Basel, SWITZERLAND, 10Chimiothérapie – Oncologie, CHU Jean Minjoz, Besançon, FRANCE Purpose: One year of H-based therapy, consisting of 18 q3w cycles, is standard of care for the adjuvant treatment of HER2-positive EBC. H is administered

Volume 23 | Supplement 9 | September 2012

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

Background: Good clinical judgement remains the best tool in determining a therapeutic plan; however this process has become easier, and more reproducible with the ongoing progression of decision-making adjuncts such as: Nottingham Prognostic Index (NPI), which incorporates pathological features; Adjuvant! Online (AO), which incorporates clinic-pathological features; and Oncotype Dx ® Recurrence Score (RS) assay, which incorporates genetic features. We herein report our experience with the use of these tools, and their potential impact on chemotherapy uptake for patients in MWRH, Limerick, Ireland. Methods: Study period: 01/09/2008 – 31/12/2012. Data was derived from (1) MWCC Oncology database (2) pathology reports (3) patient files (4) RS assay reports. Data was collated and entered into an Access database and exported into SPSS (v.18) for analysis. Results Seventy-seven (77) patients with early stage endocrine positive breast cancer were analysed. Median age: 55.9 years (range: 30.6 – 72.3 years). Median follow-up: 20.9 months. Findings are shown below in tabular form. Conclusions: Chemotherapy uptake from the decision-making tools was varied within the cohort, but it was uniformly noted that patients were more likely to undertake chemotherapy if they had higher indices derived from these DMT, except in cases where the patients declined therapy, or if physicians’ preference differed on clinical grounds. This highlights that none of the available DMT are perfect in isolation, but when used in concert, they can elucidate the best potential treatment pathway following consultation between patients and their healthcare providers; this is given by the fact that they all rely on different subsets of information to predict potential outcome. Further work is needed to find an algorithm that can appropriately incorporate these DMT any potential patient’s best interest. Disclosure: All authors have declared no conflicts of interest.

than LN-negative group (44.2% vs. 30.0%, P < 0.001). Of 1682 patients in adjuvant setting, 729 (43.3%) were treated with triplet (TAC or AC-T). In 290 patients receiving neoadjuvant chemotherapy, only 94 (32.4%) used TAC and none used AC-T (P = 0.013). Conclusions: Docetaxel was widely used in the treatment of breast cancer in China, especially in the (neo)adjuvant setting. TAC regimen was the most frequent option, especially in patients with TNBC or LN-positive BC. Although AC-T was also recommended in adjuvant setting by most guidelines, it was less commonly used compared to TAC regimen in China. Disclosure: All authors have declared no conflicts of interest.

Annals of Oncology

316TiP

every 3 months for a total of 5 years of treatment. Supplemental vitamin D (≥ 400 IU) and calcium (≥ 500 mg) will be administered. The trial, sponsored by Amgen Inc. and registered with the ClinicalTrials.gov identifier NCT01077154, began enrolling patients in June 2010 and is expected to complete enrolment in late 2012. Disclosure: R.E. Coleman: Other substantive relationships: Expert testimony Novartis. R. Bell: Advisory Board Member : Amgen. M. Martin: Advisory Board Member amgen, roche, bayer, novartis, gsk, sanofi. A. Braun: Employee : Amgen Inc Stock : Amgen Inc. C. Ke: Employer : Amgen Inc. Stock ownership : Amgen Inc. T. Maniar: Employee :: Amgen Inc. Stock ownership :: Amgen Inc. P.E. Goss: Honoraria for Pfizer and Novartis. All other authors have declared no conflicts of interest.

Downloaded from http://annonc.oxfordjournals.org/ by guest on March 17, 2016

intravenously (IV) over 30–90 mins. An SC formulation of H has been developed, which is rapidly administered (up to 5 mins), potentially improving convenience for patients and clinical staff, and reducing administration costs. The Phase III HannaH study (NCT00950300) demonstrated that the pharmacokinetics and efficacy of H-SC were non-inferior to that of H-IV, meeting the co-primary endpoints. The safety profile of H-SC was comparable and consistent with the known safety profile of H-IV. SafeHer is designed to further evaluate the safety and tolerability of H-SC in a broader patient population; to allow greater understanding of a range of safety data. H-SC will be administered via one of two different routes (handheld syringe [vial formulation] or single-use injection device [SID]; which allows self-administration). Supportive data on SID safety and patient satisfaction with self-administration will be collected. Methods: SafeHer is a multicentre, international, Phase III open-label trial (NCT01566721). The primary objective is the safety and tolerability of H-SC. Secondary objectives include disease-free survival, overall survival and patient satisfaction with SID administration. Immunogenicity of H and recombinant human hyaluronidase in a subset of patients using the SID at select sites is an exploratory objective. Planned enrolment is 2500 patients, assigned to one of two cohorts at the investigators’ discretion ± concurrent/sequential chemotherapy. All patients will receive H-SC at a fixed dose of 600mg regardless of body weight, for a total of 18 cycles (q3w) via injection into the thigh over a period of up to 5 minutes. Patients in cohort A (n = 1800) will receive H-SC using handheld syringes. Patients in cohort B (n = 700) will receive H-SC using an SID, first via assisted administration and then self-administered (in select patients). Enrolment began in May 2012 and parallel substudies may be performed at selected centres to evaluate medical resource utilisation (“time and motion study”). Disclosure: J. Gligorov: I disclose potential conflict of interest : advisory boards and speaker honoraria for Roche laboratories. H.A. Azim: I serve on the advisory board of Roche in the middle east. B. Ataseven: I am a Roche International Steering Committee member for the SafeHer trial. I have received speaker honoraria from Roche for giveing lectures. I participated in and received a travel grant from Roche for SABCS 2011. M. Delaurentiis: I am a member of a Roche Advisory Board. F. Herbst: I am an employee of, and am a stockholder in, F. Hoffmann-La Roche. A. Llombart: I am a member of a Roche Advisory Board. S. Osborne: I am an employee of Roche. X. Pivot: I am a member of Roche and GSK Advisory Boards. All other authors have declared no conflicts of interest. DENOSUMAB VERSUS PLACEBO AS ADJUVANT TREATMENT FOR WOMEN WITH EARLY-STAGE BREAST CANCER AT HIGH RISK OF DISEASE RECURRENCE (D-CARE): AN IN PROGRESS, PHASE 3 CLINICAL TRIAL

R.E. Coleman1, C. Barrios2, R. Bell3, D.M. Finkelstein4, H. Iwata5, M. Martin6, A. Braun7, C. Ke8, T. Maniar7, P.E. Goss9 1 Academic Unit of Clinical Oncology, University of Sheffield, Sheffield Cancer Research Centre, Sheffield, UNITED KINGDOM, 2Department of Medicine, PUCRS School of Medicine, Porto Alegre, BRAZIL, 3The Andrew Love Cancer Centre, Geelong Hospital:Deakin University, Geelong, AUSTRALIA, 4Biostatistics, Massachusetts General Hospital, Boston, MA, UNITED STATES OF AMERICA, 5 Aichi Cancer Center, Breast Oncology, Nagoya University School of Medicine, Nagoya, JAPAN, 6Medical Oncology Service, Hospital General Universitario Gregorio Marañón, Madrid, SPAIN, 7Hematology/oncology, Amgen Inc., Thousand Oaks, CA, UNITED STATES OF AMERICA, 8Biostatistics, Amgen Inc., Thousand Oaks, CA, UNITED STATES OF AMERICA, 9Cancer Center, Massachusetts General Hospital, Boston, MA, UNITED STATES OF AMERICA Background: Bone represents approximately 40% of all first recurrences in women with early-stage breast cancer and is a common site of distant recurrence. In preclinical studies, RANKL inhibition significantly delays skeletal tumour formation, reduces skeletal tumour burden, and prolongs survival of tumour-bearing mice. Denosumab is approved for the prevention of skeletal-related events (SREs) in patients with established bone metastases from solid tumours. Purpose: The D-CARE trial is designed to assess if denosumab treatment prolongs bone metastasis-free survival (BMFS) and disease-free survival (DFS) in the adjuvant breast cancer setting. The primary endpoint of this event-driven trial is BMFS. Secondary endpoints include DFS and overall survival. Additional endpoints are safety, breast density, incidence of SREs (following the development of bone metastasis), patient reported outcomes, and biomarkers. Methods: In this international, randomized, double-blind, and placebo controlled phase 3 trial, approximately 4,500 women with stage II or III breast cancer, at high risk for recurrence and with known hormone and HER-2 receptor status, are eligible. High risk is defined as biopsy evidence of breast cancer in regional lymph nodes, tumour size > 5 cm, (T3), or locally advanced disease (T4). Standard-of-care adjuvant or neoadjuvant chemo-, endocrine, or HER-2 targeted therapy, alone or in combination, must be planned. Patients with a prior history of breast cancer (except DCIS or LCIS) or distant metastasis, oral bisphosphonate (BP) use within 1 year of randomization, or any intravenous BP use, are not eligible. Patients are randomized 1:1 to receive denosumab 120 mg or placebo subcutaneously monthly for 6 months, then

Volume 23 | Supplement 9 | September 2012

doi:10.1093/annonc/mds392 | ix