1 University of Chicago Medical Center, Chicago IL 2 US Oncology Research, Tyler, TX

ASCO 2010

A Placebo-Controlled, Randomized Phase 2 Study of Conatumumab (C) or AMG 479 (A) or Placebo (P) + Gemcitabine (G) in Patients (pts) With Metastatic Pancreatic Cancer (mPC)HL Kindler,1 D Richards,2 J Stephenson,3 L Garbo,4 C Rocha Lima,5 H Safran,6 J Wiezorek,7 E Feigal,7 SL Bray,8 CS Fuchs9

1University of Chicago Medical Center, Chicago IL2US Oncology Research, Tyler, TX3Greenville Hospital System, Greenville, SC4New York Oncology Hematology, P.C. Albany, Albany, NY5University of Miami, Sylvester Comprehensive Cancer Center, Miami, FL6Rhode Island Hospital, Providence, RI7Amgen Inc., Thousand Oaks, CA8Amgen Ltd, Cambridge, UK9Dana-Farber Cancer Institute, Boston, MA

Abstract #4035

ASCO 2010

BACKGROUNDThe DR5/TRAIL Pathway in Pancreatic Cancer

• Pancreatic tumors express higher levels of death receptor 5 (DR5) than normal pancreas1

• Mutations in KRAS are found in > 90% of pancreatic cancer (PC)2

– Transformation by RAS sensitizes cells to TRAIL-induced apoptosis3,4

• Conatumumab (AMG 655) is an investigational, fully human, monoclonal antibody agonist of DR5– Binds DR5, activates caspases, and induces apoptosis5

• In mouse models of PC, conatumumab demonstrated single-agent activity, which was enhanced in combination with gemcitabine5

• In a phase 1b study in metastatic PC patients, conatumumab appeared safe in combination with gemcitabine6

– Disease control rate (partial response [PR] + stable disease [SD]) = 69%; 6-month overall survival (OS) = 76%

ASCO 2010

BACKGROUNDThe DR5/TRAIL Pathway in Pancreatic Cancer

ASCO 2010

BACKGROUNDThe Type 1 Insulin-Like Growth Factor Receptor (IGF-1R) Pathway and Pancreatic Cancer

• IGF1-R, IGF-1, and IGF-2 are over-expressed in PC cells and tumors• Risk of PC increases in patients with increased pathway activity

(insulin sensitivity, Akt2 amplification, and low serum IGFBP-1)• Pharmacological and genetic blockade of IGF-1R activity inhibits the

growth of multiple PC xenograft models (alone and in combination with gemcitabine)

• Tumors driven by KRAS remain sensitive to IGF-1R inhibition7-9

• AMG 479 is an investigational, fully human, monoclonal antibody antagonist of IGF-1R– Blocks binding of IGF-1 and IGF-2 to IGF-1R8

• In mouse models of PC, AMG 479 demonstrated single-agent activity; activity was enhanced in combination with gemcitabine8

• In a phase 1b study in patients with advanced solid tumors, AMG 479 appeared safe in combination with gemcitabine10

ASCO 2010

BACKGROUNDThe Type 1 Insulin-Like Growth Factor Receptor (IGF-1R) Pathway and Pancreatic Cancer

ASCO 2010

STUDY RATIONALE

• Current therapies for PC are inadequate: novel approaches are required

• Preclinical and phase I data support the evaluation of conatumumab and AMG 479 in PC patients

• Randomized phase II studies may be superior to single-arm trials in evaluating novel agents in PC

• Therefore, a 3-arm, placebo-controlled, randomized phase 2 study was conducted to evaluate conatumumab + gemcitabine and AMG 479 + gemcitabine versus placebo + gemcitabine in metastatic PC patients

ASCO 2010

PRIMARY ENDPOINT

• 6-month OS rate

ASCO 2010

METHODSStudy Schema

aThe AMG 479 + gemcitabine arm was open label due to anticipated thrombocytopenia and hyperglycemia.

ASCO 2010

METHODSSecondary Endpoints

• Response rate (by RECIST)– Investigator defined, no central review

• Progression-free survival (PFS, investigator defined)• OS• Incidence of adverse events (AE) and laboratory

abnormalities• Incidence of anti-conatumumab or anti-AMG 479

antibodies• Pharmacokinetics of conatumumab and AMG 479• Dose intensity of gemcitabine when combined with

conatumumab, AMG 479, or placebo

ASCO 2010

METHODSKey Inclusion Criteria

• Histologically or cytologically documented metastatic adenocarcinoma of the pancreas

• ≥ 18 years of age• ECOG PS 0 or 1• Adequate hematologic, hepatic, renal, and coagulation

function• Amylase and lipase ≤ 2.0 x ULN • Adequately controlled type 1 or 2 diabetes (amended during study to

FBS <160 mg/dL only) – Fasting blood sugar ≤ 160 mg/dL – HbA1c < 8%

• Written informed consent

ASCO 2010

METHODSKey Exclusion Criteria

• Uncontrolled cardiac disease• Prior chemotherapy or radiation in the adjuvant or

metastatic setting

ASCO 2010

METHODSStatistics

• The study was designed as an estimation study– Planned sample size of 120 (40/arm)– As designed (estimation, not hypothesis-testing), the trial has 80%

power to detect a difference between a 6-month OS rate of:• 45% for placebo• 69% for conatumumab or AMG 479 (target improvement, 10%)

ASCO 2010

RESULTSBaseline Demographics

ASCO 2010

RESULTSBaseline Lesion Sites

ASCO 2010

RESULTSEfficacy

• The PFS and OS results were robust after adjusting for baseline covariates (age, gender, ECOG PS status, liver metastases, and tumor sum of longest diameter)

aPrimary endpoint. HR, hazard ratio; CI, confidence interval.

ASCO 2010

RESULTSProgression-Free Survival

Full analysis set. aStratified hazard ratio (estimates relative to placebo + gemcitabine arm). bStratified log-rank test.HR, hazard ratio; KM, Kaplan-Meier; CI, confidence interval.

ASCO 2010

RESULTSOverall Survival

Full analysis set. aStratified hazard ratio (estimates relative to placebo + gemcitabine arm). bStratified log-rank test.

ASCO 2010

RESULTSBest Overall Tumor Response

CR, complete response; PR, partial response; SD, stable disease; PD, progressive disease.

ASCO 2010

RESULTSGrade 3-5 Adverse Eventsa

aIn ≥ 5% of patients. bOne GI perforation, 1 GI hemorrhage,1 hepatic failure in a patient with liver metastases, 4 PD. cOne acute renal failure, 1 hemorrhage, 2 PD. dOne GI hemorrhage.

ASCO 2010

RESULTSDrug Exposure

aRatio of cumulative dose of drug : protocol-specified cumulative dose over the specified period. IP, investigational product.

ASCO 2010

CONCLUSIONS

• This randomized phase 2 study shows evidence of activity for AMG 479 with trends across several efficacy parameters in metastatic PC– Improved 6-month OS rate (primary endpoint) (57% vs 50%) – Improved 12-month OS (39% vs 23%)– Longer PFS (5.1 vs 2.1 months, HR 0.65)– Longer OS (8.7 vs 5.9 months, HR 0.67)– Higher rates of SD + PR (51% vs 41%)

ASCO 2010

CONCLUSIONS (continued)

• The study also shows evidence of activity for conatumumab– Improved 6-month OS rate (primary endpoint) (59% vs. 50%) – Longer PFS (4.0 vs 2.1 months, HR 0.65)– Higher rates of SD + PR (61% vs 41%)

• Both drug combinations were well-tolerated• These data suggest that further study of AMG 479 +

gemcitabine is warranted in advanced PC

ASCO 2010

REFERENCES1. Data on file. Amgen Inc.

2. Koorstra JB, et al. Pancreatology. 2008;8:110-125.

3. Nesterov AM, et al. Cancer Res. 2004;64:3922-3927.

4. Drosopoulos KG, et al. J Biol Chem. 2005;280:22856-22867.

5. Kaplan-Lefko PJ, et al. Cancer Biol Ther. 2010;9:618-631.

6. Kindler HL, et al. J Clin Oncol. 2009;27:abstr 4501.

7. Beltran PJ, et al. Mol Cancer Ther. 2009;8:1095-1105.

8. Sell C, et al. Mol Cell Biol. 1994;14:3604-3612.

9. Yeh AH, et al. Oncogene. 2006;25:6574-6581.

10. Sarantopoulos J, et al. J Clin Oncol. 2008;26:abstr 3583.

ASCO 2010

ACKNOWLEDGMENTS

• This study was sponsored by Amgen Inc. (ClinicalTrials.gov: NCT00630552)