S54 Poster Session II. Locally advanced and metastatic disease Friday, 16 March 2007 P142 Rechallenge of patients previously treated with adjuvant anthracyclines using pegylated liposomal doxorubicin (PLD) with cyclophosphamide (C) as first-line chemotherapy for metastatic breast cancer (MBC) M. Trudeau 1 , M. Clemons 1 , L. Provencher 2 , L. Panasci 3 , L. Yelle 4 , D. Rayson 5 , J. Latreille 6 , T. Vandenberg 7 , J. Pouliot 8 . 1 Sunnybrook Health Sciences Centre, Medicine, Toronto, Canada, 2 Hopital St-Sacrement, Quebec City, Canada, 3 Jewish General Hospital, Montreal, Canada, 4 Hopital Notre-Dame, Montreal, Canada, 5 QE II Health Sciences Centre, Halifax, Canada, 6 Hopital Charles Lemoyne, Longueuil, Canada, 7 London Regional Cancer Centre, London, Canada, 8 Schering Canada Inc., Pointe-Claire, Canada Background: Anthracyclines (A) are components of most breast cancer chemotherapy regimens in the adjuvant and metastatic settings. Adjuvant pre-exposure limits the utilization of A in advanced disease due to cu- mulative cardiotoxicity. PLD (Caelyx/Doxil) has clinical activity equivalent to conventional doxorubicin in MBC with significantly less cardiotoxicity. Combinations of A + cyclophosphamide (C) are the backbone of many adjuvant therapeutic regimens making the combination of PLD + C a rational regimen for clinical investigation. Methods: Patients with measurable MBC having completed an anthra- cycline containing adjuvant regimen >12 months prior to diagnosis of metastatic disease were entered in this multi-center single arm phase II trial. Treatment was PLD 35 mg/m 2 + C 600 mg/m 2 every 3 weeks as first- line therapy for metastatic disease. Response assessment was by standard RECIST criteria. Results: Seventy-three patients were enrolled. Prior adjuvant therapy included: AC (40%), CEF/FEC (24%), AC-T (13%), FAC (8%), EC (6%), AT (5%). Median dose of prior A was 240 mg/m 2 and 576 mg/m 2 for doxoru- bicin and epirubicin, respectively. Median time since completion of adjuvant chemotherapy was 4.6 years (1−14). Patients received a median of 6 cycles (2−19) of PLD + C. Major toxicities were: grade 3/4 neutropenia (7.5%), asymptomatic >10% declines in LVEF (9%) (reversible upon discontinuation of PLD), grade 3/4 hand foot syndrome (6%). No grade 3/4 cardiac toxicities were observed, despite overall cumulative doses of anthracycline reaching 450 mg/m 2 (130–720) and 700 mg/m 2 (287–1279) for doxorubicin and epirubicin, respectively. Objective response rate (ORR) was 38% (3% CR, 35% PR), with an additional 30% having stable disease >6 months resulting in a clinical benefit rate of 69% (CB). ORR was similar for patients who had received adjuvant taxanes. Kaplan-Meier estimated median time to progression was 6.6 months. Conclusions: This combination of PLD + C every 3 weeks for patients previously exposed to adjuvant anthracyclines is well tolerated and demon- strates cardiac safety. The ORR and clinical benefit rate observed is similar to other commonly employed chemotherapeutic regimens for MBC and suggests persistence of anthracycline-sensitive disease. Re-challenge with a less cardiotoxic anthracycline may be a therapeutic alternative for patients with advanced breast cancer. P143 Retrospective pooled analysis to assess correlation between time to progression and overall survival in patients with breast cancer B. Sherrill 1 , C. Hirst 2 , Y. Wu 1 , M. Amonkar 3 , S.H. Stein 3 . 1 RTI-Health Solutions, Biometrics Department, Reaearch Triangle Park, NC, USA, 2 RTI-Health Solutions, Epidemiology Department, Manchester Science Park, UK, 3 GlaxoSmithKline, Global Health Outcomes, Collegeville, PA, USA Background: For cancer therapies, although overall survival (OS) is the gold standard for clinical benefit, the surrogate endpoints of time to tumor progression (TTP) and/or progression free survival (PFS), in advanced dis- ease, and disease-free survival (DFS) in the adjuvant setting are clinically valid endpoints and may be acceptable for drug approval. The relationship between disease progression endpoints and overall survival (OS) has been demonstrated in colorectal and non-small lung cancers in the metastatic (Johnson et al. Lancet 2006) and in colon cancer in the adjuvant settings (Sargent et al. JCO 2005). We performed a meta-analysis to explore the association between time to tumor progression (TTP) and progression-free survival (PFS) and OS in randomized controlled trials (RCTs) for metastatic breast cancer (MBC). Methods: Summary data, trial and patient characteristics were abstracted from RCTs of MBC (first line and refractory), since 1994, retrieved via a literature search. Studies were selected if both progression endpoints and OS were reported. Data was analyzed using Spearman correlation coefficients and simple regression equations (Louvet et al. Cancer 2001). Results: Data from a pooled sample size of ~19,000 patients was avail- able and retrieved from 68 of 582 studies that met all eligibility criteria. TTP across treatment groups averaged 6.9 months and OS averaged 20.5 months, among 51 studies. Correlation between the 2 measures was 0.33 (p = 0.0007). PFS averaged 7.3 months in an additional 17 studies that reported this endpoint. However, since definitions provided for TTP and PFS were not consistent across studies, and often overlapped, all studies were included and OS was compared to TTP or PFS; correlation was 0.31 (p = 0.0002). Linear regression coefficients suggest that a 1-month increase in TTP or PFS corresponds to a commensurate increase in OS (b = 0.9). Conclusions: These preliminary results confirm the findings from other tumor types that increases in time to progression endpoints (TTP/PFS) observed in MBC trials ultimately correspond to survival benefits. This finding is particularly important since survival data from RCTs is often obscured by subsequent treatments or patient cross-over. Further analyses using a multivariate meta-regression approach are planned. As seen in other tumor types, one could expect similar findings when DFS is compared to OS in the adjuvant breast cancer setting. P144 CELER: an international randomized phase III trial of the CESAR Group – or: Should chemotherapy be added to hormonal therapy in advanced breast cancer? D. Sirbu 1 , C. Oprean 1 , S. Negru 1 , M. Draganescu 2 , V. Valeanu 3 , D. Reichert 4 , S. Eder 5 , B. Moritz 6 , L. Edler 6 , R. Morant 8 . 1 Oncomed, Timisoara, Romania, 2 Trestioreanu Oncology Institute Bucarest, Bucarest, Romania, 3 Oncology Clinic Bd Corneliu Coposu, Sbiu, Romania, 4 St¨ adtische Kliniken Oldenburg, Oldenburg, Deutschland, 5 Assign Clinical Research, Wien, Oesterreich, 6 Cesar Central Office, Wien, Oesterreich, 7 Deutsches Krebsforschungsinstitut Heidelberg, Heidelberg, Deutschland, 8 Tumorzentrum ZeTuP St. Gallen und Chur, St. Gallen, Switzerland There is still considerable uncertainty, whether patients with metastatic breast cancer should be treated with chemotherapy, hormonal therapy or both at the same time. The concomitant administration of chemotherapy and an aromatase inhibitor has been insufficiently studied, whereas com- binations with tamoxifen do not seem to be beneficial. Possible higher response rates of a combination therapy may be offset by decreased quality of life without increased survival. CESAR (Central European Society for Anticancer Drug Research) thus initiated the randomized CELER trial (CEsar LEtrozole Randomized study) and encouraged participation of Eastern European Countries. The trial randomizes postmenopausal women with advanced hormone-responsive breast carcinoma to treatment with either hormonal therapy with letrozole alone until progression or with letrozole in combination with a defined number of chemotherapy cycles. The two-arm trial examines the question whether hormonal therapy alone may be sufficient or proves to be inferior to a more toxic combination with cytotoxic drugs. Primary endpoint of the study is the time to tumor progression, assumed to rise from 9.4 to 12 months with the combination; secondary endpoints involve overall survival, quality of life and translational research questions on paraffin- embedded tissues. With a recruitment period of 2 years, and a follow-up period of a further 2 years, 548 patients will have to be recruited. The study has been initiated in 2006 and has enrolled 24 patients in November 2006. 46 institutions from 12 countries will participate. The study compares different approaches to patient care with advanced breast cancer, and brings into contact researchers from Switzerland to Omsk in Siberia. Investigators’ meetings combined with educational events, such as this one in St. Gallen, will help the participating researchers to build international contacts and their institutions to be at the forefront of clinical knowledge. The set-up of such a trial is an organisational challenge for the Cesar Central Office in Vienna and involves cooperation with 6 companies, the setup of an internet based electronic data capture (EDC) system and the use of biostatistical services of the DKFZ in Heidelberg. At the meeting we describe some of these challenges and solutions for an academic institution in an era of increasing regulations, bureaucratic hurdles and financial constraints in Eastern and Western Europe.