1. Denosumab Safety and Efficacy in Giant Cell Tumor of Bone (GCTB): Interim Results From a Phase 2 study Sant Chawla, 1 Jean-Yves Blay, 2 Javier Martin Broto, 3 Edwin Choy, 4 Martin Dominkus, 5 Jacob Engellau, 6 Robert Grimer, 7 Robert Henshaw, 8 Emanuela Palmerini, 9 Peter Reichardt, 10 Piotr Rutkowski, 11 Keith Skubitz, 12 David Thomas, 13 Yufan Zhao, 14 Yi Qian, 14 Ira Jacobs 14 1 Sarcoma Oncology Center, Santa Monica, CA, USA; 2 University Claude Bernard Lyon I, Centre Léon Bérard, Department of Medicine, Lyon, France; 3 Hospital Son Dureta, Palma de Mallorca, Spain; 4 Dana Farber/Harvard Cancer Center, Massachusetts General Hospital, Boston, MA, USA; 5 Medizinische Universitaet Wien, Vienna, Austria; 6 Skåne Universitetssjukhus, Lund, Sweden; 7 Royal Orthopaedic Hospital, Birmingham, UK; 8 Georgetown University College of Medicine, Washington, DC, USA; 9 Istituti Ortopedici Rizzoli, Bologna, Italy; 10 HELIOS Klinik Bad Saarow, Bad Saarow, Germany; 11 Maria Sklodowska-Curie Memorial Cancer Center and Institute of Oncology, Department of Soft Tissue/Bone Sarcoma and Melanoma, Warszawa, Poland; 12 Masonic Cancer Center, University of Minnesota, Minneapolis, MN, USA; 13 Peter MacCallum Cancer Centre, East Melbourne, Victoria, Australia; 14 Amgen Inc., Thousand Oaks, CA, USA

2. Giant Cell Tumor of Bone (GCTB) Aggressive, primary osteolytic tumor Causes local pain and impairs mobility and function 1 No approved or effective medical therapy Surgical intervention often associated with significant morbidity 2 1.Mendenhall WM et al. Am J Clin Oncol. 2006;29:96–9. 2.Balke M et al. J Cancer Res Clin Oncol. 2009;135:149–58.

3. RANK and RANKL in Giant Cell Tumor of Bone 1. Atkins GJ, et al. J Bone Miner Res. 2006; 21:1339–49. 2 Huang L, et al. Am J Pathol. 2000;156:761–7. 3. Kartsogiannis V, et al. Bone. 1999;25: 525–34. 4. Roux S, et al. Am J Clin Pathol. 2002; 117:210–6. 5. Burgess TL, et al. J Cell Biol. 1999;145:527–38. 6 Lacey DL, et al. Cell. 1998;93:165–76. 7. Yasuda H, et al. Proc Natl Acad Sci USA. 1998;95:3597–602. 8. Bekker PJ et al. J Bone Miner Res. 2004;19:1059–66. RANK expression in GCTB 8 RANKL expression in GCTB 8 Tumors contain osteoclast-like giant cells expressing RANK and stromal cells expressing RANK ligand (RANKL), a key mediator of osteoclast formation, activation, function, and survival. 1-4 Excessive RANKL secretion causes an imbalance in bone remodeling in favor of bone breakdown. 5-7

4. RANKL is a Central Mediator of Bone Destruction in Giant Cell Tumor of Bone Fusion Osteoclast Precursors RANK RANKL Stromal Cells of GCTB Activated Osteoclast Precursor Osteoclast Activated Giant Cell RANKL

5. Denosumab in Giant Cell Tumor of Bone Fully human monoclonal antibody that binds to RANKL 1 Inhibits osteoclast-mediated bone destruction Initial open-label, proof-of-concept, phase 2 study of denosumab in GCTB (N = 37): 2 –Tumor response in 86% of patients with GCTB –Clinical benefit in 84% of patients (reduced pain or improvement in functional status per investigator) –No serious treatment-related adverse events 1.Bekker PJ et al. J Bone Miner Res. 2004;19:1059–66. 2.Thomas D et al. Lancet Oncol. 2010;11:275–80.

6. Phase 2 Follow-on Study: Interim Analysis SC: subcutaneous 18152345 Denosumab 120 mg sc DayMonth Months 7– N Cohort 2: Salvageable GCTB, surgery planned Safety Surgery: delay, avoidance, or reduced morbidity Cohort 1: Surgically unsalvageable GCTB Safety Disease progression (investigators’ assessment) 6 Adults or skeletally mature adolescents with GCTB

7. Results Subject Demographics and Disease Characteristics Characteristic (All enrolled subjects) Cohort 1 Surgically Unsalvageable N = 112 Cohort 2 Salvageable, Surgery Planned N = 50 Female,%63 %58% Age, median (min–max)32 (13–76)34 (17–56) Location of target lesion, % Femur, tibia, patella/knee, or tarsus6%64% Lung 30%4% Sacrum 22%6% Pelvic bone 14%8% Humerus, radius, ulna, or metacarpus5%12% Vertebrae: cervical, thoracic, or lumbar10%2% Skull 6%0% Soft tissue: cervical, thoracic pelvic, or abdominal 4% N = All enrolled subjects

8. Safety Results Denosumab Exposure and Adverse Events All Subjects N = 158* Median (Q1 – Q3) number of doses received10 (6 – 15) Median (Q1 – Q3) months on study7 (3 – 12) Subjects with Adverse Events, % AEs of grade 3 or 4 considered related to denosumab4.4% Hypophosphatemia2.5% Dysmennorrhea0.6% Osteonecrosis of the jaw (ONJ)1.9% Hypocalcemia (grade 1 or 2)4.4% * N = number of subjects who received at least 1 dose of denosumab

9. Efficacy Response per Investigator Assessment No Disease Progression in the Majority of Subjects * N = the number of subjects who received denosumab, had the opportunity to be on study for ≥6 months, and had disease progression data at the time of analysis. The disease response data analysis was based on the best response reported during the assessment period. Cohort 1: Surgically Unsalvageable N = 73* Cohort 2: Salvageable, Surgery Planned N = 23 * 2 (3%) 38 (52%) 1 (1%)

10. Efficacy: Cohort 2 At 12 Months, Most Subjects in Cohort 2 Had No Surgery or a Less Morbid Surgical Procedure Than Planned Surgical Procedure, n*Planned (N = 23)Actual (N = 23) Total number of surgeries238 Major surgeries103 Hemipelvectomy10 Amputation20 Joint/prosthesis replacement51 Joint resection22 Marginal excision, en bloc excision, or en bloc resection 70 Curretage2 4 Other † 41 No surgeryN/A15 * In order from most morbid to least morbid † Other planned skeletal procedures included replacement of proximal tibia, sacral lesion/bone resection, and pelvic resection (1 each).

11. Results – Efficacy 36-year-old man with GCTB and severe pain, left lower extremity Baseline After 29 weeks of denosumab treatment

12. Pre-Treatment (August 2009) Courtesy of Alexander Fedenko, MD and Elke Ahlmann, MD (USC)

13. Post-Treatment (Limb Salvage, September 2011) Courtesy of Alexander Fedenko, MD and Elke Ahlmann, MD (USC)

14. Summary Denosumab was well tolerated by these subjects with GCTB; no new risks were observed in this population –ONJ, a known risk with denosumab, was observed at a low rate consistent with that seen in other studies Disease progression was halted in 99% of subjects Of 23 subjects for whom surgery was planned: –15 had no surgery –8 had less morbid surgeries than planned This study is ongoing; denosumab continues to be studied as a potential treatment for GCTB