3006^ Background: The Ras-Raf-MEK-extra-cellular signal-regulated kinase 1 and 2 (ERK1/2) pathway is frequently deregulated in human cancer. RO5126766, a first-in-class, dual Raf/MEK inhibitor was tested in a phase I (Ph1) study. Objectives were determination of maximum tolerated dose (MTD), dose-limiting toxicities (DLTs), recommended phase II dose (RP2D), safety/tolerability, PK/PD and clinical activity. Methods: Patients (pts) with advanced or metastatic solid tumors received oral RO5126766 once daily (QD) in 28 day cycles. To increase the therapeutic window, 2 intermittent regimens were also evaluated: 7 days on/7 days off (7/7) and 4 days on/3 days off (4/3). PK and PD samples (PBMCs) were collected after a single run-in dose and on cycle 1 day 15 (C1D15). Paired skin and tumor biopsies (baseline, C1D15) were analyzed for mutational status and target inhibition. FDG-PET scans (baseline, C1D15, and C3D1) were performed. Results: 52 pts (25 QD, 13 4/3 and 14 7/7) were enrolled in 13 cohorts (QD: 0.1 - 2.7 mg, 4/3: 2.7 - 4 mg and 7/7: 2.7 - 5 mg). Mean age 50yrs, ECOG 0-1, median previous chemotherapy lines 3 (0-12). Common tumors were melanoma (21), ovarian (6) and CRC (10). Ten reversible DLTs were observed: 3 blurred vision (QD, 4/3 and 7/7), 3 CK elevations (2x QD, 7/7), 1 transaminitis (QD); 1 capillary leak syndrome (7/7), 1 febrile neutropenia (7/7) and 1 serous retinal detachment (7/7). MTDs were defined as 2.25 mg QD, 2.7 mg 7/7 and 4 mg 4/3. Most common related adverse events were skin (94%), GI (71%), metabolic (67%) and eye (58%) disorders. No cases of cutaneous squamous cell carcinoma were reported. PK was linear with plasma half-life (t ½) of 40 to 60 hrs. Tumor /skin biopsies demonstrated target modulation with inhibition of pERK/pMEK in PBMCs approaching 100%. Of 40 evaluable pts, 3 melanoma pts had PR and 9 pts had SD > 16 wks associated with reduction in SUVmax (mean: -35%) at C1D15. Conclusions: RO5126766 has a manageable safety profile. MTD was defined for all regimens and RP2D is 2.7 mg 4/3. Favorable PK/PD profile with encouraging biological and antitumor activity were demonstrated. Full safety, efficacy, PK/PD profile will be presented.

3017 Background: Mutations of Ras/Raf lead to a sustained and constitutive activation of ERK pathway. MEK1/2 is the only enzyme that activates ERK1/2; consequently MEK1/2 is a potential target to inhibit in cancers with an activated ERK pathway. RO4987655, a potent, highly selective ATP non-competitive MEK1i with an excellent selectivity profile, was tested in a 3+3 Ph1 study design. Objectives were determination of maximum tolerated dose (MTD), recommended Ph2 dose (RP2D), dose limiting toxicities (DLTs), safety/tolerability, PK/ PD and clinical activity. Methods: Patients (pts) with advanced/metastatic solid tumors received oral RO4987655 administered on a continuous daily dosing (QD) from 1-2.5 mg then twice daily (BID) from 3-21 mg total daily dose in 28 days (D) cycles. Blood PK samples were collected on cycle 1 (C1) [D 1, 8, 15 and 22]. Molecular target suppression was measured by pERKi in PBMC on C1 (D1, D15). Paired skin and tumor biopsies (optional) (baseline, C1D15) and sequential FDG-PET scans (baseline, C1D15, C3D1) were taken. Efficacy was measured as per RECIST. Results: 49 pts in 12 cohorts were enrolled at 4 sites. Mean age 52 y, ECOG 0-1, previous chemotherapy lines median 3 (0-10). Common tumor types included melanoma (27), CRC (11) and NSCLC (3). Four reversible DLTs were observed: CPK elevation (17 mg (n=1), 21 mg (n=2) and blurred vision (21 mg). MTD and RP2D were defined as 17 mg. Most commonly related adverse events included skin (91%), GI (57%), eye (26%) and CPK disorders (44%). PK profiles showed dose-linearity, a half-life of 4 to 6 hrs and drug accumulation ~ 2 fold at steady-state. High (mean 75%) and sustained (90% of time >IC50) pERKi was observed in PBMC at RP2D. FDG PET showed large decrease in SUV at RP2D (-47%, -90% to +29%) (mean, range) which correlated with PBMC pERKi in melanoma pts. PRs and SDs (>=4 mo) were observed in 2 and 5 melanoma, respectively. Conclusions: RO4987655 showed acceptable and manageable safety profile. PK was linear for the tested dose range with moderate inter-pt variability. PD activity in PBMC associated with large metabolic FDG-PET response and preliminary encouraging antitumor activity were demonstrated at the RP2D.