D1 Q3w Research Articles

Final results of AGITG ATTAX3 study: Randomized phase II study of weekly docetaxel (T), cisplatin, and fluoropyrimidine (F) with or without panitumumab (P) in advanced esophagogastric (OG) cancer.

4081 Background: This randomized phase II study evaluated the efficacy and safety of P, a fully human mAb against the epidermal growth factor receptor combined with T-based chemotherapy in advanced OG cancer. Methods: Eligible pts had histologically confirmed metastatic OG cancer (adeno-carcinoma and squamous cell carcinoma) were ≥18 years of age, PS 0-2, with adequate renal, haematologic and liver function with measurable disease. All pts provided informed consent. Selection was not based on kras determination. Pts received T 30mg/m2 d1,8, C 60mg/m2 d1 and F; investigator choice of 5FU infusion 160mg/m2/d or capecitabine 500mg/m2bd continuous ±P 9 mg/kg d1 q3w. Treatment was administered for 8 cycles or until PD. The primary endpoint was response rate according to RECIST (1.1) assessed q6w. Planned enrolment target was 100 pts. Stratification variables included histology, PS and choice of F. Results: From April 2010 to November 2011, 77 pts were enrolled from 15 institutions. A safety alert from the REAL3 study (also involving P in OG cancer) prompted an unplanned review of data from ATTAX3 by the IDMC. The IDMC found no evidence of adverse outcomes associated with P, but as it did not appear that P would significantly improve efficacy, they recommended cessation of the study to new enrolment. Previously enrolled pts were treated and followed according to protocol. Median follow up is 24m. Treatment arms were well balanced; median age 59, 64y, male 77%, 87%, PS0-1 95%,90%, adenocarcinoma 90%,90%, capecitabine 67%, 66% for TCF/TCF-P, respectively. Common grade 3/4 toxicities include infection 18%,24%, febrile neutropenia 10%, 5%, anorexia 10%, 24%, nausea 18%,30%, stomatitis 3%,5%, diarrhoea 15%,24% , acneiform rash 0%, 8%, fatigue 18%, 30%, hypomagnesemia 10%, 16% for TCF/TCF-P. Efficacy outcomes are summarized in Table. Conclusions: The addition of P to T-based chemotherapy in advanced OG cancer did not improve efficacy and was associated with an increase in some toxicities. Clinical trial information: ACTRN12609000109202. [Table: see text]

Phase I study of volasertib (BI 6727) combined with afatinib (BIBW 2992) in advanced solid tumors.

2521 Background: Volasertib is a potent and selective cell cycle kinase inhibitor that induces mitotic arrest and apoptosis by targeting Polo-like kinases (Plk). Volasertib and afatinib, an irreversible ErbB family blocker, have shown single agent anti-tumor activity and manageable safety profiles in patients (pts) with advanced solid tumors. This dose escalation study was designed to determine the maximum tolerated dose (MTD) of two combination schedules of volasertib and afatinib in pts with advanced solid tumors refractory to or not amenable to standard therapy. Methods: In a 3 + 3 design, cohorts of 3–6 pts received volasertib 150–300 mg IV d1 Q3W + afatinib 30–50 mg PO QD d2–21 Q3W (Schedule A) or afatinib 50–90 mg d2–6 Q3W (Schedule B). Up to 12 additional pts were enrolled at the MTD. Primary endpoint was the MTD per schedule. Secondary endpoints included pharmacokinetics (PK), safety and efficacy (RECIST). Results: 57 pts (median 58 yr; ECOG PS 0/1/2: 35%/60%/5%) were treated (n=29, Schedule A; n=28, Schedule B). MTD was volasertib 300 mg/afatinib 30 mg (Schedule A) and volasertib 300 mg/afatinib 70 mg (Schedule B). Cycle 1 dose limiting toxicities (DLTs) were experienced by 5 (Schedule A) and 7 (Schedule B) pts. Most common DLTs were diarrhea (n=5), neutropenia (n=3), fatigue (n=2) and decreased ejection fraction (n=2) in Schedule A, and thrombocytopenia (n=6), neutropenia (n=5), diarrhea (n=4) and febrile neutropenia (n=3) in Schedule B. Most common grade 3/4 adverse events were neutropenia (n=8), thrombocytopenia (n=6), diarrhea (n=3) and febrile neutropenia (n=3). Volasertib exhibited multi-exponential PK behavior with a long half-life (130 hr), moderate clearance (900 mL/min) and large volume of distribution (Vss >6000 L). Co-administration of volasertib and afatinib had no effect on the PK profile of either drug. Two pts in Schedule A (volasertib 300 mg/afatinib 30 mg) achieved partial responses (tumor types: NSCLC, head and neck). Conclusions: MTD of volasertib was 300 mg Q3W combined with afatinib 30 mg d2-21 (Schedule A) or afatinib 70 mg d2-6 (Schedule B). Both agents could be combined at previously shown active single agent doses. At the MTD, treatment was manageable and showed preliminary anti-tumor activity. Clinical trial information: NCT01206816.

Plasma VEGF-a, angiopoietin-2 (ANG2) and soluble(s)TIE2 in patients (pts) with HER2-negative locally recurrent or metastatic breast cancer (LR/MBC) treated with first-line bevacizumab (A) and paclitaxel (T) without or with capecitabine (X).

1072 Background: In the randomized phase II ATX trial, pts with HER2-negative LR/MBC were treated with first-line AT or ATX. We determined the prognostic value for outcome of VEGF-A, ANG2 and sTIE2 measured at baseline on cycle 1 day 1 (C1D1) and after cycle 1 (C2D1). Methods: 312 pts were randomized in 1:1 ratio to AT (T 90 mg/m2 on d1, 8, 15 and A 10 mg/kg on d1, 15 q4w x 6 cycles, followed by A 15 mg/kg on d1 q3w for next cycles) or ATX (T 90 mg/m2 on d1, 8, A 15 mg/kg on d1 and X 825 mg/m2bid on d1–14 q3w x 8 cycles, followed by the same dose of A and X q3w for next cycles). The primary endpoint was progression-free survival (PFS). Secondary endpoints were objective response rate (ORR), response duration (RD), overall survival (OS) and safety. Plasma proteins on C1D1 (N = 173) and on C2D1 (N = 142) were measured by ELISA. The association of protein levels (continuous variable) with PFS and OS was evaluated by Cox proportional hazards model and Martingale residual plot. Results: At a median follow-up of 39 months (mo), there were 292 PFS events and 242 deaths. ATX significantly improved PFS as compared to AT (median 11 vs. 8.4 mo, stratified HR = 0.52; 95% CI, 0.41 – 0.67; P < .001). The confirmed ORR in measurable disease (N = 268) was 67% in ATX vs. 50% in AT. Median RD was 6.4 mo (95% CI, 6.1 – 8.3) in ATX v 5.4 mo (95% CI, 5.1 – 6.0) in AT. Median OS was 24.1 mo in ATX vs. 23.1 mo in AT (P= .44). The aselected ‘biomarker’ cohort (N = 173) and overall trial cohort had similar baseline characteristics. ANG2 on C1D1 moderately correlated with sTIE2 on C1D1 (Pearson’s r = .44, P < .001). High ANG2 on C1D1 was significantly associated with poor OS (HR = 1.6; 95% CI, 1.1 – 2.3; P = .01), but not with poor PFS (HR = 1.3; 95% CI, 1.0 – 1.3; P = .07). ANG2 on C2D1 was not significantly associated with OS (HR = 1.55; 95% CI, 0.99 – 2.4; P = .057) or with PFS (P= .6). sTIE2 and VEGF-A were not associated with outcome. All pts had very low levels of free VEGF-A on C2D1 (median 8 pg/ml). Conclusions: In HER2-negative LR/MBC, ATX is more effective (PFS, ORR and RD) than AT. A very high plasma level of ANG2 at baseline indicates a high risk for poor survival. Clinical trial information: NTR1348.

978P - Rosia: A Single-Arm Study in More Than 1000 Patients (PTS) Receiving Front-Line Bevacizumab (BEV) + Chemotherapy (CT) for Ovarian Cancer (OC)

ABSTRACT Background BEV significantly improved the efficacy of front-line CT for OC in the ICON7 and GOG-0218 phase III trials. The global single-arm ROSiA study was designed to assess the safety of BEV-containing therapy, given until progression or for up to 36 cycles, in the context of routine oncology practice. Methods Eligible pts have FIGO stage IIb–IV or grade 3 stage I–IIa epithelial ovarian, fallopian tube or primary peritoneal carcinoma, have received no prior post-surgical therapy for OC, are aged ≥18 years and have ECOG PS 0–2. Prior neoadjuvant CT is permitted. Pts with uncontrolled hypertension, clinical signs/symptoms of GI obstruction, or a history of abdominal fistula, GI perforation or intra-abdominal abscess within the preceding 6 months are ineligible. After definitive surgery, pts receive BEV 15 mg/kg q3w (or 7.5 mg/kg at the investigator's discretion) in combination with 4–8 cycles of CT (paclitaxel [175 mg/m2 d1 q3w or 80 mg/m2 qw) + q3w carboplatin [AUC 5 or 6]). BEV is continued at the same dose as a single agent until disease progression (PD), unacceptable toxicity or for up to 36 cycles. The primary objective is to assess safety (CTCAE v4.03). Secondary endpoints include progression-free survival, overall response rate by RECIST and/or CA-125 response criteria, duration of response and overall survival. The study includes exploration of potential correlations of plasma, tumour and genetic biomarkers with BEV efficacy and toxicity. Results Between Dec 2010 and Apr 2012, ∼1000 pts were enrolled from 35 countries, predominantly Spain (n = 180), Italy (n = 110) and France (n = 101). Baseline characteristics were available from 677 pts as of 30 Jan 2012. Most pts had stage III/IV OC (stage I/II/III/IV: 7%/11%/60%/18%). The majority (93%) received paclitaxel q3w and 94% received BEV at a dose of 15 mg/kg; 14% had received prior neoadjuvant CT. Baseline characteristics from the entire population of ∼1000 pts will be reported, together with details of CT and BEV schedules selected. Conclusions ROSiA should provide valuable new information on the safety of BEV for up to 36 cycles and in pts who have received neoadjuvant CT, as well as enabling extensive translational research. Disclosure N. Colombo: NC has received honoraria for advisory boards and speaker engagements from Roche. F. Selle: FS acts as a Consultant for Roche and PharmaMar. P. Pautier: PP has served on Advisory Boards for Roche and Ipsen. D. Bollag: DB is an employee and holds stocks in F Hoffmann-La Roche Ltd. A.M. Oza: AO has received research funding from Roche. All other authors have declared no conflicts of interest.

Prospective randomized phase II trial of oral vinorelbine (NVBo) and cisplatin (P) or pemetrexed (Pem) and P in first-line metastatic or locally advanced non-small cell lung cancer (M or LA NSCLC) patients (pts) with nonsquamous (non SCC) histologic type: NAVoTRIAL01—Preliminary results.

7575 Background: NVBo+P is considered as a standard treatment (trt) in M or LA NSCLC. The recent approval of Pem+P as front line chemotherapy (CT) for non-SCC demonstrates that today, histology could become a “new guidance” to treat patients (pts). Phase III trial (Scagliotti. JCO 2008) showed efficacy of Pem+P in patients with non-SCC pts. Moreover, the higher chemosensitivity of non-SCC is recognised and reported with other chemotherapies (Ardizzoni.JNCI 2007). In GLOB 3 study, NVBo+P also showed better survival in adenocarcinoma (11.7m) than in SCC (8.9m) (Tan.Ann of Oncol 2009). This trial was set up to assess the efficacy of NVBo+P compared to Pem+P for pts with non-SCC histological type. The primary end-point was disease control rate (DCR) including overall response rate (ORR) and stable disease (SD). Methods: Pts were randomised to receive NVBo 80 mg/m² D1D8 (60 at Cycle 1) + P 80 mg/m² D1 q3w or Pem 500 mg/m² + P 75 mg/m² D1 q3w. After 4 cycles of trt, for pts showing a disease control, single agent as continuation maintenance (CM) was proposed until progression or toxicity. Given that Pem+P is now considered as a reference trt in non-SCC, pts were randomised on a 2/1 basis and stratified according to stage (IIIB - IV - relapse), non-SCC confirmed by histology or cytology, gender, smoking status and centre. Results: From 10/09 to 02/11, 153 pts were randomised to receive NVBo+P (102 pts) or Pem+P (51 pts). The efficacy between the 2 arms is similar: The ORR/DCR (Recist 1.1) after 4 cycles of trt are 26%/75% in the NVBo arm (100 eval pts in ITT population) and 24%/78% in the Pem arm (50 eval pts in ITT population). At the end of December 2011, 7 patients are still treated in CM: 5 pts in the NVBo arm and 2 pts in the Pem arm. Conclusions: The 2 CT regimen, NVBo+P or Pem+P have a similar efficacy in terms of ORR/DCR for pts with advanced non-SCC NSCLC. These results appear to be in line with the published data. As the analysis is currently in progress, final results including safety data, PFS and OS will be presented during the meeting.

Sequential first-line treatment for metastatic colorectal cancer with capecitabine, irinotecan, and bevacizumab: Capecitabine plus bevacizumab (Cape-Bev) versus capecitabine, irinotecan plus bevacizumab (CAPIRI-Bev): The AIO KRK 0110/ML22011 randomized, phase III trial.

TPS3639 Background: Several randomized trials have indicated that combination chemotherapy applied in metastatic colorectal cancer (mCRC) does not significantly improve overall survival when compared to the sequential use of cytotoxic agents (CAIRO, MRC Focus, FFCD 2000-05). The present study investigates the question in bevacizumab-based first-line treatment including escalation- and de-escalation strategies. Methods: The AIO KRK 0110/ML22011 trial (ClinicalTrials.gov NCT01249638) is a two-arm, multicenter, open-label randomized phase III trial comparing the efficacy and safety of Cape-Bev versus CAPIRI-Bev in the first-line treatment of metastatic colorectal cancer. Patients with unresectable metastatic colorectal cancer, ECOG PS 0-1, will be assigned in a 1:1 ratio to receive either capecitabine 1250 mg/m2 bid for 14d (d1-14) plus bevacizumab 7.5 mg/kg (d1) q3w (Arm A) or capecitabine 800 mg/m2 BID for 14d (d1-14), irinotecan 200 mg/m2 (d1) and bevacizumab 7.5 mg/kg (d1) q3w (Arm B). Patients included into this trial are required to consent to the analysis of tumor tissue and blood for translational investigations. In Arm A, treatment escalation from Cape-Bev to CAPIRI-Bev is recommended in case of progressive disease (PD). In Arm B, de-escalation from CAPIRI-Bev to Cape-Bev is possible after 6 months of treatment or in case of irinotecan-associated toxicity. Re-escalation to CAPIRI-Bev after PD is possible. The primary endpoint is time to failure of strategy (TFS). Secondary endpoints are ORR, OS, PFS, safety and quality of life. Conclusion: The AIO KRK 0110 trial is designed for patients with disseminated, but asymptomatic mCRC who are not potential candidates for surgical resection of metastasis. Two bevacizumab-based strategies are compared: one starting as single-agent chemotherapy (Cape-Bev) allowing escalation to CAPIRI-Bev and another starting with CAPIRI-Bev and allowing de-escalation to Cape-Bev and subsequent re-escalation if necessary. By January 2012, 79 of planned 516 patients have been enrolled.

Bevacizumab (BEV) plus capecitabine as maintenance therapy after initial treatment with BEV plus XELOX in previously untreated patients (pts) with metastatic colorectal cancer (mCRC): Mature data from STOP and GO, a phase III, randomized, multicenter study.

3565 Background: Colorectal cancer is one of the most frequent malignancies, second after breast cancer in women and third after lung cancer and prostate cancer in men. The aim of this study was to evaluate and compare the progression-free survival (PFS) between two arms: Arm A is a combination of BEV + XELOX; Arm B is a combination of BEV + XELOX for 6 cycles followed by maintenance BEV + capecitabine as first-line therapy in mCRC. Methods: BEV (7.5 mg/kg) + XELOX (capecitabine 1000 mg/m2 bid d1–14 + oxaliplatin 130 mg/m2 d1 q3w) were administered until progression (Arm A) or 6 cycles of BEV + XELOX followed by BEV + capecitabine were administered until progression (Arm B). PFS was the primary endpoint; secondary endpoints included overall survival (OS), objective response rate (ORR), and safety. A sample size of 118 pts was required to detect with 80% power an increase of 1.5 months in median PFS between two arms with a standard deviation of 3.9 months and significance level of 0.05 (10% drop-out rate). Results: A total of 123 pts were randomized. Demographic characteristics were balanced between the arms. Median treatment period was 7.5 (range 0.5–13.9) and 8.1 (range 0.1–20.7) months in Arms A and B, respectively. There was a statistically significant difference in median PFS between arms, although there was no significant difference in ORR and OS (see table). Tolerability was acceptable in both arms with the following grade 3/4 adverse events (AEs): Arm A 48.4%; Arm B 34.4% (p=0.116). Grade 3/4 diarrhoea occurred in 9.7% vs. 3.3%, weakness in 8.1% vs. 8.2%, hand-foot syndrome in 3.2% vs. 1.6%, and neuropathy in 4.8% vs. 3.3% of pts in Arms A and B, respectively. Conclusions: These findings suggest that maintenance therapy with BEV + capecitabine following induction with 6 cycles of BEV + XELOX may be superior to continuous BEV + XELOX until progression inpts with previously untreated mCRC. [Table: see text]

GILT study: Oral vinorelbine (NVBo) and cisplatin (P) with concomitant radiotherapy (RT) followed by either consolidation (C) with NVBo plus P plus best supportive care (BSC) or BSC alone in stage (st) III non-small cell lung cancer (NSCLC): Final results of a phase (ph) III study.

7001 Background: Concurrent chemo-radiotherapy (CT-RT) is considered as a standard in st III NSCLC. Published trials with C after CT-RT show encouraging but discordant results. This ph III was set up to assess C in st III NSCLC. Methods: Patients (pts) received NVBo 50 mg/m² D1, D8, D15 + P 20 mg/m² D1-D4 q4w / 2 cycles (cy) + RT (66 Gy / 33 Fr). C for OR+SD pts: NVBo 60-80 mg/m² D1D8 + P 80 mg/m² D1 q3w / 2 cy + BSC (Arm A) or BSC (Arm B). PFS was the primary endpoint. Results: From 07/05 to 05/09, 279 pts received CT/RT and 201 pts (72%) were randomised to receive CT+BSC or BSC as C. Toxicity (tox) G3-4 (% pt) CT-RT/ C (Arm A/B): anaemia 3.2/3.5/1.1; thrombopenia 2.5/1.2/0.6; neutropenia (N) 11.2/11.7/5.7; febrile N 1.4/1.0/0; nausea (G3) 5.0/4.7/2.9; vomiting (G3) 3.9/3.5/2.0; anorexia 3.6/1.2/3.0; dysphagia 1.8/2.3/1.0; fatigue 3.3/ 2.3/1.0; pneumonia/ pneumonitis 2.6/0/2.0; CT-RT pain 2.2; CT-RT oesophagitis 8.6; 3 toxic deaths. Conclusions: In this ph III, NVBo+P+RT reports a high level of efficacy (OR 60.7%; DCR 86.0%) and low tox. The DCR is significantly improved in pts who received C with NVBo+ P in eval pts (p=0.0084). Lung tox was not enhanced by using NVBo+P as C. However no survival advantage for C was achieved. [Table: see text]

Abstract 3780: GEP01: Population pharmacokinetic-pharmacodynamic (PK/PD) model for neutropenia of lapatinib and iv vinorelbine in the treatment of HER2-positive locally advanced or metastatic breast cancer

Abstract Background: A novel combination of po lapatinib (LPT), a selective dual ErbB1/ErbB2 targeted drug, + iv vinorelbine (VNR), a cell-cycle inhibiting agent, could provide a high-potential treatment for locally-advanced or metastatic HER2-overexpressing breast cancer refractory to first or second-line chemotherapy associated to trastuzumab. Neutropenia is the main dose-limiting toxicity occuring in VNR treatment. LPT has been shown to be a strong inhibitor of CYP3A4 which is also mainly involved in VNR metabolism. We have demonstrated a PK interaction related to the combination of VNR and LPT. The aim of this study was to construct a semiphysiological population PK/PD model and to identify PD factors responsible for the neutropenia caused by VNR with this drug combination. Methods: Women with HER2+ locally advanced or metastatic breast cancer progressing after ≤ 2 lines of trastuzumab-based treatment, were treated with LPT starting 7 days (D) (D-7 to D0) before adding VNR on a D1 & D8 q3w IV schedule. LPT was given po continuously. Dose levels [DL, LPT (mg)/VNR (mg/m2)] ranged from 750/20 to 1,250/30. For PD analysis, blood count was measured within 7 days before beginning of treatment and up to 3 times per cycle post-treatment. The model consisted of a proliferating compartment that was sensistive to VNR, three transit compartments (transition rate constant, ktr) that represented maturation, and a compartment of circulating blood cells with a feedback mechanism from the circulating cells (Circ0/Circ)α with Circ0 representing the baseline value of circulating neutrophils and a feddback parameter α. VNR concentration-time profiles affected the proliferation of neutrophils by an inhibitory Emax model. Covariate analysis was performed using different patients characteristics and LPT dose level. Population PK/PD was modelled using a non linear mixed effect model program (Monolix version 3.2s). Results: From 29 patients, 552 neutrophil counts were obtained for PK/PD modeling. The mean estimated (SE) parameters obtained for the structural model were CIRC0= 3.7(0.19)*109/L, EC50=0.55(0.01) µg/L, ktr=0.71(0.02) h−1, α=0.24(0.025). No significant covariate effect, including LPT dose values on any PD parameters for VNR could be identified. However, a strong trend was observed between LPT dose level increase and proliferation rate decrease in the proliferating compartments. Conclusions: The model successfully described myelosuppression. When LPT dose increases, ktr value increases and ≤ value decreases indicating a PD interaction between LPT and VNR which is consistent with PK interaction observed between these two drugs. This PK/PD interaction might increase the exposure to VNR and consequently alter the hematological tolerance. A predictive PK/PD modeling of tumor growth kinetics in patients is ongoing. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 3780. doi:1538-7445.AM2012-3780

P1-17-02: A Phase 1/2 Study of SAR245408 (S08) in Combination with Trastuzumab (T) or Paclitaxel (P) and T in Patients with HER2+ Metastatic Breast Cancer (MBC) Who Progressed on a Previous T-Based Regimen.

Abstract BACKGROUND: Most HER2+ MBC patients (pts) treated with a combination of T+P progress within 1 year. Activation of downstream pathways through either deficiency in PTEN or mutations in the PI3K pathway has been implicated in the development of resistance to T. S08 is a potent, orally bioavailable, pan-PI3K inhibitor that inhibits phosphorylation of multiple downstream components of the PI3K/PTEN signaling pathway and has demonstrated activity as a single agent and in combination with other anticancer agents (Edelman G, et al., ASCO 2010; Traynor AM, et al. ASCO 2010). PATIENTS and METHODS: This ongoing, open-label multicenter phase 1/2 study (NCT01042925) was designed to assess safety and tolerability of S08 in combination with either T or T+P. Eligible female pts were ≥18 yrs, ECOG PS 0–2, with adequate organ and marrow function and had advanced or recurrent HER2+ MBC disease refractory to T. Pts have received, and progressed on at least 1 prior T-containing regimen for metastatic disease. Pts were allocated to Arm 1 or Arm 2 and received different dose levels of S08 (starting dose 200 mg PO, daily) in combination with either T 8/6 mg/kg IV on Day (D) 1 q3w (Arm 1) or T 8/6 mg/kg IV on D1 + P 80 mg/m2 on D 1, 8, and 15 q3w (Arm 2). In phase 1, a standard ascending 3 + 3 dose escalation design was used in each arm to evaluate safety/tolerability of the combinations. Following establishment of preliminary maximal tolerated doses (MTDs) for each combination, subsequent pts will be accrued to the phase 2 portion. Approximately 25 additional pts will be enrolled in each arm to further evaluate the safety and estimate the overall response rate (ORR) in each arm. RESULTS: As of June 1st 2011, 33 pts median age 55 yrs were enrolled; 18 to Arm 1 and 15 to Arm 2. Based on preliminary data of the study, in Arm 1, the most common treatment emergent adverse events (TEAEs) regardless of relationship include rash, diarrhea, fatigue, nausea, vomiting; in Arm 2 were neutropenia, diarrhea, fatigue, nausea, hyporexia, hypokalemia, peripheral neuropathy, rash and hyperglycemia. SAEs reported in Arm 1 (3 subjects) included Gr3 dehydration (2 cases), Gr3 epigastric pain and Gr2 dyspnea; those SAEs in Arm 2 (4 subjects) included Gr4 neutropenia, Gr3 anorexia, Gr3 dehydration, Gr3 epigastric pain, Gr3 thromboembolism, Gr2 nausea, Gr2 pneumonitis and Gr2 headache. A total of 4 DLTs were reported; 2 in Arm 1 [Gr3 skin rash (S08; 400 mg dose)] and 2 in Arm 2 [Gr4 neutropenia (S08; 200mg dose)] (table 1). With phase I nearly complete, the MTD for arm 1 is S08 300 mg PO daily and T 8/6 mg/kg D1 q3w. The data for Arm 2 (current dose level ongoing at the maximum allowed per protocol) will be completed in August 2011. Preliminary PK data did not show interactions between S08 and either T or T+P. DISCUSSION: S08 can be combined with either T or with T+P. Additional safety, PK and efficacy data will be presented from phase 1. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P1-17-02.

P4-20-01: Multicentric Phase II PACS 09/Beverly1 Trial: First Efficacy and Safety Results of Neoadjuvant Chemotherapy Combined with Bevacizumab in HER2−Negative Patients with Non-Metastatic Inflammatory Breast Cancer.

Abstract Background: Inflammatory breast cancer (IBC) is a relatively rare form of breast cancer. Because this disease exhibits angiogenic properties, bevacizumab (BEV) may improves the efficacy of neoadjuvant chemotherapy (CT) in IBC. Methods: Pts with HER2−negative IBC (IHC 0 or 1+ or FISH/CISH-, T4d, any N) were included to receive 4 cycles of FEC100 + BEV 15 mg/kg, d1 q3w, followed by 4 cycles of Docetaxel 100 mg/m2 + BEV 15 mg/kg q3w. Complete surgery was performed 4–6 weeks after the last dose of CT, followed by radiotherapy and endocrine therapy for hormone receptor positive (HR+) pts. BEV was re-introduced as adjuvant therapy for 10 cycles, giving a total of 18 cycles of BEV. The primary endpoint of the PACS09/Beverly1 trial was to evaluate the pathologic complete response (pCR; Sataloff classification) rate in HER2−negative non-metastatic IBC patients(pts). Results: Between Jan 2009 and Sep 2010, 101 pts were included (100 pts were evaluable). Main baseline characteristics were: median age 49 years; postmenopausal status, 39 pts; SBR 2, 32 pts SBR3 61 pts, both HR-negative, 55 pts. 85 pts completed neoadjuvant therapy according to the protocol, ***.90 pts underwent complete mastectomy, 4 had a breast-conservative surgery. After neoadjuvant treatment, investigator-assessed pCR rate was 27% (18%-36%, CI 95%). Interestingly, a significantly greater proportion of pCR rate was observed in the HR-negative group compared to HR-positive group (38% vs 13% respectively; p=0.007). Results from central review of pCR are currently in progress. Main Grade 3,4 toxicities were: neutropenia (61 pts),febrile neutropenia (36 pts),mucitis (23 pts) Neither gr3/4 HTA nor cardiac failure or proteinuria was reported. There were no treatment-related deaths. Conclusions: These data suggest that the use of BEV with neoadjuvant CT is active in HER2−negative IBC (pCR rate 27%), with an acceptable safety profile. Interestingly, the HR-negative subgroup is significantly more responsive than the HR-positive subgroup. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P4-20-01.

A randomized, phase III trial of capecitabine plus bevacizumab (Cape-Bev) versus capecitabine plus irinotecan plus bevacizumab (CAPIRI-Bev) in first-line treatment of metastatic colorectal cancer: The AIO KRK 0110 Trial/ML22011 Trial

BackgroundSeveral randomized trials have indicated that combination chemotherapy applied in metastatic colorectal cancer (mCRC) does not significantly improve overall survival when compared to the sequential use of cytotoxic agents (CAIRO, MRC Focus, FFCD 2000-05). The present study investigates the question whether this statement holds true also for bevacizumab-based first-line treatment including escalation- and de-escalation strategies.Methods/DesignThe AIO KRK 0110/ML22011 trial is a two-arm, multicenter, open-label randomized phase III trial comparing the efficacy and safety of capecitabine plus bevacizumab (Cape-Bev) versus capecitabine plus irinotecan plus bevacizumab (CAPIRI-Bev) in the first-line treatment of metastatic colorectal cancer. Patients with unresectable metastatic colorectal cancer, Eastern Cooperative Oncology Group (ECOG) performance status 0-1, will be assigned in a 1:1 ratio to receive either capecitabine 1250 mg/m2 bid for 14d (d1-14) plus bevacizumab 7.5 mg/kg (d1) q3w (Arm A) or capecitabine 800 mg/m2 BID for 14d (d1-14), irinotecan 200 mg/m2 (d1) and bevacizumab 7.5 mg/kg (d1) q3w (Arm B). Patients included into this trial are required to consent to the analysis of tumour tissue and blood for translational investigations. In Arm A, treatment escalation from Cape-Bev to CAPIRI-Bev is recommended in case of progressive disease (PD). In Arm B, de-escalation from CAPIRI-Bev to Cape-Bev is possible after 6 months of treatment or in case of irinotecan-associated toxicity. Re-escalation to CAPIRI-Bev after PD is possible. The primary endpoint is time to failure of strategy (TFS). Secondary endpoints are overall response rate (ORR), overall survival, progression-free survival, safety and quality of life.ConclusionThe AIO KRK 0110 trial is designed for patients with disseminated, but asymptomatic mCRC who are not potential candidates for surgical resection of metastasis. Two bevacizumab-based strategies are compared: one starting as single-agent chemotherapy (Cape-Bev) allowing escalation to CAPIRI-Bev and another starting with combination chemotherapy (CAPIRI-Bev) and allowing de-escalation to Cape-Bev and subsequent re-escalation if necessary.Trial RegistrationClinicalTrials.gov Identifier NCT01249638EudraCT-No.: 2009-013099-38

Bevacizumab plus capecitabine as maintenance treatment after initial treatment with bevacizumab plus XELOX in previously untreated metastatic colorectal cancer: Updated findings from a randomized, multicenter phase III trial.

3579 Background: Colorectal cancer is one of the most frequent malignancies, second to breast cancer in women and third to lung cancer and prostate cancer in men. The aim of this study in first-line metastatic colorectal cancer (mCRC) was to achieve a better progression-free survival (PFS) and less risk of toxicity by administrating bevacizumab (BEV) + capecitabine + oxaliplatin (XELOX) for 6 cycles, stop oxaliplatin and go with maintenance therapy (BEV + capecitabine) until progression. Methods: BEV (7.5 mg/kg) + XELOX (capecitabine 1000 mg/m2 bid d1–14 + oxaliplatin 130 mg/m2 d1 q3w) were administered until progression (Arm A) or 6 cycles of BEV + XELOX followed by BEV + capecitabine were administered until progression (Arm B). PFS was the primary endpoint; secondary endpoints included overall survival (OS), objective response rate (ORR), and safety. A sample size of 118 patients (pts) was calculated to achieve 80% power to detect an increase of 1.5 months in median PFS between Arm A (9.5 months) and Arm B (11.0 months) with a standard deviation of 3.9 months and significance level of 0.05 using a 10% drop-out rate. Results: A total of 122 pts were randomized. No significant differences were found in demographic characteristics between the two arms. Median treatment period was 6.1 (range 0.7–13.4) and 6.8 (range 0.7–12.4) months in Arms A and B, respectively. Interim analysis showed no statistically significant differences in median PFS and ORR between arms (Table). Tolerability was also acceptable in both arms with grade 3/4 diarrhoea in 7.7% vs. 8.2%, weakness in 15.2% vs. 8.4%, hand-foot syndrome in 6.3% vs. 9.4%, and neuropathy in 2.8% vs. 4.6% of pts in Arms A and B, respectively. Conclusions: BEV + capecitabine as maintenance therapy following induction BEV + XELOX is noninferior to continuous BEV + XELOX until progression. These interim findings suggest that maintenance therapy with BEV + capecitabine is an appropriate option following induction BEV + XELOX in pts with mCRC. Updated data will be presented. Efficacy Arm A (n=61) Arm B (n=61) P value Median PFS, months 8.3 9.9 0.064 ORR, % 57.4 69.2 0.207

Abstract 5468: GEP01: A phase I pharmacokinetic study of lapatinib and iv vinorelbine in the treatment of HER2-positive locally advanced or metastatic breast cancer

Abstract Background: A novel combination of oral lapatinib (LPT), a selective dual ErbB1/ErbB2 targeted drug, + iv vinorelbine (VNR), a cell-cycle inhibiting agent, could provide a high-potential treatment for locally-advanced or metastatic HER2-overexpressing breast cancer refractory to first or second-line chemotherapy associated to trastuzumab. LPT has been shown to be a strong inhibitor of CYP3A4 which is also mainly involved in VNR metabolism. In this study we investigated the potential pharmacokinetic (PK) interactions related to the association of VNR and LPT. Methods: Women with HER2+ locally advanced or metastatic breast cancer progressing after ≤ 2 lines of trastuzumab-based treatment, were treated with LPT starting 7 days (D) (D-7 to D0) before adding VNR on a D1 & D8 q3w IV schedule. LPT was given orally continuously. Dose levels [DL, LPT (mg)/VNR (mg/m2)] ranged from 750/20 to 1,250/30. A total of 29 patients, 36 to 75 years old, were treated by the association of LPT + VNR. For PK analysis, 7 time point samples were collected on D1 of cycle 1 for LPT and VNR assays. For VNR and LPT respectively, whole blood and plasma concentrations were measured using UPLC coupled with tandem mass spectrometry validated methods. Population PK was modelled using a non linear mixed effect model program (Monolix version 3.1s) by computing the maximum likelihood estimator of the parameters without any approximation of the model (no linearization). Results: A three-compartment open model adequately described VNR time versus concentration courses. The inter-individual variabilities (ISV) could be well estimated for all stuctural parameters (clearance : CL, volume of distribution : V, inter-compartmental clearances : Q) except for Q3 and V3. The population PK parameters obtained for the structural model were: CL=40.2 L/h, V1=8.49 L, Q2=46.9 L/h, V2=1290 L, Q3=61.9 L/h and V3=60.9 L. Influence of platelet counts and LPT dose values on blood VNR CL were significant according to the Bayesian Information Criteria (BIC). A one-compartment model adequately fitted the LPT plasma concentration-time data. The population PK parameters were CL=27.9 L/h, V=61.5 L and the absorption constant, ka=0.071 h−1. The ISV were 55.7% and 81.7% for CL and V respectively. No covariate effect, including body surface area and VNR dose values on any PK parameters for LPT could be identified. Conclusions: A PK interaction occured between VNR and LPT. When the LPT dose level increased, the VNR CL decreased significantly, probably due to CYP450-3A4 interaction. The LPT dose which should decrease VNR CL by 50% was defined as 1790 mg. No covariate effects on any LPT PK parameters had been identified. A PK/PD modelling using tumour evaluation and absolute neutrophil count (ANC) is ongoing. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 5468. doi:10.1158/1538-7445.AM2011-5468

Effective management of patients receiving XELOX: Evaluation of impact of dose modifications on outcome in patients from the NO16966, NO16967, and NO16968 trials.

497 Background: In patients (pts) with metastatic colorectal cancer (MCRC), XELOX is non-inferior to FOLFOX4 in terms of PFS and OS as either first-line (NO16966, Cassidy et al. JCO 2008) or second-line therapy (NO16967, Rothenberg et al. Ann Oncol 2008). In pts with stage III colon cancer (NO16968, Haller et al. ECCO-ESMO 2009), adjuvant XELOX is superior to 5-FU/LV in terms of DFS. While XELOX is generally well tolerated, regional differences in fluoropyrimidine tolerability were noted in a pooled analysis [Rothenberg et al. ASCO GI 2008], leading to use of lower than recommended doses of capecitabine (i.e. 1,000mg/m2 bid d1–15 with oxaliplatin 130mg/m2 d1 q3w) in some countries. Methods: NO16966, NO16967 and NO16968 protocols included standard dose/schedule modifications for capecitabine, 5-FU and oxaliplatin for treatment-related adverse events (AEs). Safety parameters included AEs, deaths, laboratory parameters, exposure to trial medication, and withdrawals. In NO16966 and NO16967, Kaplan-Meier curves for PFS were developed for pts with no treatment modifications, and pts with dose reductions, treatment interruptions or cycle delays to assess the effect of treatment modifications on efficacy. NO16968 also included a planned analysis of the effect of dose modifications on DFS. Results: NO16966 included 1335 pts with previously untreated MCRC receiving XELOX or FOLFOX4. NO16967 included 627 pts with previously treated MCRC receiving XELOX or FOLFOX4. NO16968 included 1886 pts with resected stage III disease receiving XELOX or 5-FU/LV; dose modifications were required for capecitabine in 65% and oxaliplatin in 62% of pts in NO16968. Kaplan-Meier curves of PFS (NO16966/NO16967) or DFS (NO16968) for pts who did and did not require dose modifications indicated that the efficacy of XELOX did not appear to be compromised by dose modifications. Indeed, pts who required dose modifications seemed to have a favourable outcome compared with those who did not. Conclusions: From these data we cannot make a recommendation that initial dosing should be lower than the labeled dose. However, it is clear that dose modification does not seem to impact patient outcome. [Table: see text]

A randomized, multicenter phase III trial of bevacizumab plus capecitabine as maintenance treatment after initial treatment with bevacizumab plus XELOX in previously untreated metastatic colorectal cancer.

474 Background: Colorectal cancer is one of the most frequent malignancies, second to breast cancer in women and third to lung cancer and prostate cancer in men. The aim of this study in first-line metastatic colorectal cancer (mCRC) was to achieve a better progression-free survival (PFS) and less risk of toxicity by administrating bevacizumab (BEV) + capecitabine + oxaliplatin (XELOX) for 6 cycles, stop oxaliplatin and go with maintenance therapy (BEV + capecitabine) until progression. Methods: BEV (7.5 mg/kg) + XELOX (capecitabine 1,000 mg/m2 bid d1–14 + oxaliplatin 130 mg/m2 d1 q3w) were administered until progression (Arm A) or 6 cycles of BEV + XELOX followed by BEV + capecitabine were administered until progression (Arm B). PFS was the primary endpoint; secondary endpoints included overall survival (OS), objective response rate (ORR), and safety. A sample size of 118 patients (pts) was calculated to achieve 80% power to detect an increase of 1.5 months in median PFS between Arm A (9.5 months) and Arm B (11.0 months) with a standard deviation of 3.9 months and significance level of 0.05 using a 10% drop-out rate. Results: A total of 122 pts were randomized. No significant differences were found in demographic characteristics between the two arms. Median treatment period was 6.1 (range 0.7–13.4) and 6.8 (range 0.7–12.4) months in Arms A and B, respectively. Interim analysis showed no statistically significant differences in median PFS and ORR between arms (see table). Tolerability was also acceptable in both arms with grade 3/4 diarrhoea in 7.7% vs. 8.2%, weakness in 15.2% vs. 8.4%, hand-foot syndrome in 6.3% vs. 9.4%, and neuropathy in 2.8% vs. 4.6% of pts in Arms A and B, respectively. Conclusions: BEV + capecitabine as maintenance therapy following induction BEV + XELOX is non-inferior to continuous BEV + XELOX until progression. While this study is ongoing, these interim findings suggest that maintenance therapy with BEV + capecitabine is an appropriate option following induction BEV + XELOX in pts with mCRC. [Table: see text] [Table: see text]

Efficacy of capecitabine (C)-based therapy in patients with first-line metastatic breast cancer (MBC) previously treated with adjuvant anthracyclines and taxanes.

665 Background: Anthracyclines and taxanes are used increasingly in neoadjuvant and adjuvant chemotherapy (CT) in breast cancer. Consequently, MBC is frequently diagnosed in anthracycline- and/or taxane-pretreated pts. Retreatment with these agents is commonly compromised by resistance and/or cumulative toxicity. C is effective in the management of first-line MBC; the current analysis was undertaken to determine the first-line clinical efficacy of C in pts pretreated with adjuvant CT. Methods: 3 studies of C 1000 mg/m2, bid, d1-14 q3w, were included: a retrospective study of single-agent C (n=226; Debled et al 2009); a phase II study of single-agent C (n=161; Kaufmann et al 2009); a phase III, randomized study (RIBBON-1) of C ± bevacizumab (A; 15mg/kg, d1 q3w; n=615). Results: Prior exposure to neoadjuvant and adjuvant CT was reported in around 50-75% of pts. Of these, around 60-75% and 40-45% received anthracyclines or taxanes, respectively. Overall response rate (ORR) and CBR (CR+PR+SD) were around 25% and 60%, respectively (Table). In Debled et al, univariate analysis demonstrated similar time to treatment failure (TTF) in chemotherapy-naive and -pretreated pts (8.9 vs. 8.3 months; p=0.46). A similar analysis in RIBBON-1 supports a substantial benefit with C in chemotherapy-naïve and -pretreated pts (median PFS: 6.7 and 4.4 months, respectively), with consistent effects in taxane- (6.1 and 4.2 months) and anthracycline- pretreated pts (6.4 and 4.4 months). Notably, the addition of A appeared to overcome the effect of prior taxane exposure: CA median PFS 8.7 and 8.3 months in taxane-pretreated and -naïve pts, respectively. Conclusions: These data demonstrate that C consistently achieves improved clinical outcome in pts who have received prior neoadjuvant or adjuvant CT. Debled et al 2009 Kaufmann et al 2009 RIBBON-1 Study C (n=226) C (n=161) C (n=206) CA (n=409) Neoadjuvant or adjuvant CT, % 48 54 75.7 70.4 Anthracycline, % 39 35.8 69.4 60.4 Taxane, % 5 24.5 40.8 39.4 ORR, % 56 26.1 23.6 35.4 CBR, % 76 64 60.2 72.9 TTF, months 8.8 NR NR NR PFS, months NR 7.3* 5.7 8.6 * TTP Author Disclosure Employment or Leadership Position Consultant or Advisory Role Stock Ownership Honoraria Research Funding Expert Testimony Other Remuneration Roche Roche Roche

Pharmacokinetic (PK) Interactions between Lapatinib (L) and Vinorelbine (VNR) in a Phase I Study in Locally Advanced or Metastatic Breast Cancer (LAMBC) Patients(Pts) Overexpressing HER2. GEP-01 – A Study of the FNCLCC-Group of Early Phase Trials.

Abstract Background: Lapatinib is effective in overexpressing HER2 LAMBC pts. Following anthracyclines, taxanes and capecitabine, VNR is an active agent in metastatic setting. Both drugs use the same main metabolic pathway: Cytochrome P450 3A4. We investigated the combination of L + VNR, seeking the recommended dose for further phase II studies, and looking for the PK interactions between the 2 drugs.Methods: Women with HER2+ LAMBC, in progression after ≤ 2 lines of trastuzumab-based treatment were treated with a 7 day (D-7 to D0) loading dose of L before starting VNR on a D1 and D8 q3w iv regimen. L was given po continuously. G-CSF as primary prophylaxis of febrile neutropenia (FN) was not permitted. Dose levels (DL) (L mg/VNR mg/m²) ranged from 750/20 to 1250/30 with 3 patients per DL. PK samples were collected on 7 points on D1 of cycle 1 for L and VNR dosages. PK time points analysis was performed on plasma for L and whole blood for VNR. For both drugs the analytical method was Ultra Performance Liquid Chromatography (UPLC) coupled with tandem mass spectrometry (MS-MS) as detection. VNR PK parameters were estimated by nonlinear mixed effect modeling (NONMEM) and L PK parameters were estimated by non compartmental software (MicroPharm).Results: Seventeen pts and 68 PK time points were available for VNR analysis. VNR was ascribed to a tri-compartment model. When considering all VNR data together, VNR clearance (CL) was 37.8±16.7 L/h, while it showed a gradual decrease according to specific dose of L: 47.5±20.4 L/h for L 750 mg (3 patients) vs 35.6±4.3 L/h for pooled data for L 1000 and 1250 (14 patients). This trend was not significant likely due to low number of patients treated at lower dose of L, while no significant difference was observed between L 1000 and L 1250. The mean elimination half life time (t1/2) of VNR was 0.12 h. Six patients and 42 PK time points were analyzed for L. L CL were respectively 83.8 ± 41.3 and 57.1 ± 5.5 L/h for patients who received 750 mg (3 patients) and 1000 mg (3 patients) of L with 20 mg of VNR. The mean elimination half life time (t1/2) of L was 2.89 h.Conclusions: A potential PK interference occurred between VNR and L. When the dose of L increased, both VNR and L CL decreased, likely due to CYP450-3A4 interactions. This trend should be confirmed by the PK analysis of L and VNR in more patients, including an intermediate DL (1250/22.5) still ongoing. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 5091.

Association of the fibroblast growth factor receptor-4 (FGFR4) Gly388Arg polymorphism with clinical and pathological response to primary systemic therapy (PST) for early breast cancer (EBC)

525 Background: The FGFR4 Arg388 genotype has recently been shown to be associated with decreased disease free survival after adjuvant systemic therapy for EBC and thus therapy resistance (Harbeck et al, J Clin Oncol. 2006). Primary systemic therapy (PST) for EBC allows in-vivo chemosensitivity testing. Pathological complete response (pCR) in the breast following PST is associated with improved survival. We evaluated the FGFR4 Gly388Arg polymorphism as a predictive factor for response to PST for EBC in the setting of a prospective, randomized phase 2 trial. Methods: Patients (pts, n = 257) with T2–4/N0–2/M0 EBC were randomized to 4 cycles (cy) A 60 mg/m2 + C 600 mg/m2 day (d) 1 every three weeks (q3w) followed by 4 cy D 100 mg/m2 d1 q3w, or 4 cy A 60 mg/m2 + P 500 mg/m2 on d1 q3w followed by 4 cy D 100 mg/m2 d1 q3w, stratified by hormone receptor (HR) status. Surgery was performed after last chemotherapy. Genotyping was performed using a Taqman Genotyping assay. The genotype was confirmed by RFLP-PCR for 50% of samples with 100% congruence. Results: Genotype was available for 252 of 257 pts enrolled into this study. Using Fisher´s Exact Test there was no correlation between genotype and clinical tumour size (cT), HR status, HER2neu status, grading, menopausal status or triple negative phenotype. The G/G genotype correlates significantly with clinically positive axillary lymph nodes before chemotherapy (p = 0.03). Considering all pts the G/G genotype (vs G/A + A/A) is associated with a significantly lower overall response rate (ORR)(OR 0,47; 65% vs 80%; p = 0.03). Separate analysis of treatment arms showed a significant correlation between the G/G genotype (vs other genotypes) and a lower pCR rate (OR 0.33; 12.5% vs 30.4%; p = 0.04), clinical complete response (p = 0.02) and ORR (OR 0.25; p = 0.01) in the AC-D arm, whereas there was no correlation for pts treated with AP-D. Conclusions: This study provides important evidence that the FGFR4 genotype might serve as a valuable predicitve marker for response to PST with AC-D for EBC. No significant financial relationships to disclose.

GEP01: A phase I study of lapatinib (L) and vinorelbine (VNR) in HER2 overexpressing (HER2+) locally advanced or metastatic breast cancer (LAMBC) patients (pts): A FNCLCC Group of early phase trials study

1051 Background: Lapatinib is an effective anti-HER-2 therapy in LAMBC pts, currently investigated in the adjuvant setting. Following anthracyclines, taxanes and capecitabine, VNR is an active agent in metastatic setting. Its main toxicity consists of neutropenia and may challenge the standard weekly, day D1 and D8 regimen. We investigated the combination of L + VNR, seeking the recommended dose for further phase II studies and the potential pharmacokinetic (PK) interactions. Methods: Women with a HER-2+ LAMBC, in progression after ≤ 2 lines of trastuzumab-based treatment were treated with a 7D (D-7 to D0) loading dose of L before starting VNR on a D1 and D8 q3w IV regimen. L was given orally continuously. Primary prophylaxis of febrile neutropenia (FN) with G-CSF was not permitted. Dose levels (DL, L [mg]/VNR [mg/m2]) ranged from 750/20 to 1250/30 with 3 pts/DL (6 in case of dose-limiting toxicity [DLT]). DLT was defined on tolerance at cycle 1 and included grade (gr) 4 neutropenia (PN) ≥ 7D, FN, thrombocytopenia (gr4 or symptomatic gr3), omission of D8 for haematological toxicity, and any drug-related gr3–4 non-haematological toxicity. PK samples were collected on 7 points on D1 of cycle 1 for L and VNR dosages. Results: From August 2007 to December 2008, 15 evaluable pts were enrolled (median age 58 [46–75], 43% PS 0, 8 pts previously exposed to only 1 line of trastuzumab) and 65 cycles were administered. Toxicity is available for 13 pts. Of 3 pts treated at DL4 (1000/25), 2 developed a DLT: 1 FN and 1 gr4 PN > 7D. Other significant toxicities (% pts) included gr2 anaemia 8%, gr4 PN 46%, gr1 diarrhoea 62%, gr2 nausea/vomiting 8%, gr1 skin rash 23%, gr2 transaminases 23%; no decrease of cardiac function occurred. From DL1 to DL3 (750/20, 1000/22.5), total body clearance of VNR decreased by 50% (32 ± 19 L/h vs 17 ± 7 L/h). Conclusions: Maximal tolerated dose has been reached at 1000/22.5 for the combination of L with VNR given on a D1 and D8 q3w schedule. Given a potential PK interference which would yield to higher exposure to VNR, an intermediate DL is going to be explored (1250/22.5) to allow an accurate definition of the recommanded dose of the combination for future phase II comparison studies. [Table: see text]