Impact of Velaglucerase Alfa Therapy on Bone Marrow Burden Score in Adults with Type 1 Gaucher Disease: 7-Year Experience Deborah Elstein;1 Andrew H. Haims;2 David Zahrieh;3 Gabriel M. Cohn;3 Ari Zimran1 1Shaare Zedek Medical Center and Hebrew University-Hadassah Medical School, Jerusalem, Israel; 2Yale University School of Medicine, New Haven, CT, USA; 3Shire Human Genetic Therapies, Inc., Lexington, MA, USA 27-30 June 2012 1

Co-authors’ Disclosures The authors declare the following potential competing interests: Dr Elstein participated as an investigator in the clinical trials from which these data were taken and served as a consultant to Shire Human Genetic Therapies, Inc. (Shire HGT) during the course of the clinical trials. Dr Haims has served as a consultant to Shire HGT. Dr Cohn and Mr Zahrieh are employees of Shire HGT. Dr Zimran participated as an investigator in the clinical trials from which these data were taken and served as a consultant to Shire HGT during the course of the clinical trials. Dr Zimran receives consulting fees from Protalix Biotherapeutics; has options in Protalix Biotherapeutics and sits on its Scientific Advisory Board; receives support from Genzyme for participation in the International Collaborative Gaucher Group Gaucher Registry and receives honoraria from Actelion and Pfizer.

Introduction Type 1 Gaucher disease (GD1) is associated with significant bone pathology Velaglucerase alfa has been studied in 3 Phase III trials and an extension study to these trials, which is ongoing Since 2010 velaglucerase alfa has been approved for the long-term treatment of GD1 in 39 countries, including the USA, Israel, and the EU 3

Introduction: BMB scoring system Semi-quantitative magnetic resonance imaging (MRI) is used to evaluate bone marrow involvement and response to enzyme replacement therapy (ERT) in GD1 Appearance of marrow in MR images of lumbar spine and femur is assessed using a bone marrow burden (BMB) scoring system (Maas M, et al. Radiology 2003;229:554-61) 0–1 point is considered normal at each site (LS and FN) 2-8 points for increasing abnormality at each anatomic site (8 = most abnormal) Total possible score of LS +FN = 0 - 16 points 4

7 years of longitudinal data have been reported: TKT025: A Brief History TKT025 was a 9-month, open-label, Phase I/II trial of velaglucerase alfa in adult patients with GD1 TKT025EXT (ClinicalTrials.gov identifier NCT00391625) was an extension study to TKT025 7 years of longitudinal data have been reported: Hematologic and visceral parameters (Zimran et al. Blood 2010;115:4651-6.) Achievement and maintenance of therapeutic goals (Elstein et al. Blood Cells Mol Dis 2011;46:119-23) Bone mineral density (BMD) (Elstein et al. Blood Cells Mol Dis 2011;47:56-61.)

Eligibility Criteria for Patient Enrollment and Dosing TKT025 9-month trial TKT025EXT Extension study Enrollment Diagnosis of type 1 GD GD-related anemia (1O) Thrombocytopenia Age 18 years Intact spleen No GD treatment in past 12 months 60 U/kg EOW velaglucerase alfa, IV infusion Complete TKT025 604530 U/kg EOW velaglucerase alfa, IV infusion TKT025EXT was completed in December 2011. Maximum duration of velaglucerase alfa therapy in TKT025EXT was 79.5 months. bIn TKT025EXT, stepwise dose reduction was implemented after 3 months (1 year from start of TKT025) if patients met at least 2 of 4 therapeutic goals (the 1-year goals for hemoglobin concentration, platelet count, spleen volume, and liver volume). EOW, every other week; IV, intravenous. 6

Analysis Population at Baseline: n=12 Demographics Median age, years (range) 39.3 (18.8, 69.8) Female gender, n (%) 7 (58) Clinical parameters Median (range) Bone marrow burden score Lumbar spine 6 (3, 8) Femura – Bone mineral density z-score Lumbar spineb −1.5 (−2.9, −0.4) Femoral neckb −1.6 (−2.9, 0.1) Hemoglobin concentrationc, g/dL 11.1 (9.8, 13.3) Platelet countc, ×109/L 60.3 (37.0, 98.5) Spleen volume, % of body weight 3.6 (2.2, 6.5) Liver volume, % of body weight 4.4 (2.6, 5.8) Plasma chemokine (C-C motif) ligand 18 levelc, ng/mL 1978 (1563, 5247) Normal spleen volume 0.2% of body weight. Normal liver volume 2.5% of body weight. an=6 patients with Baseline femoral BMB score; descriptive statistics not calculated. bn=11; 1 patient did not have a Baseline BMD assessment of the lumbar spine and 1 patient did not have a Baseline BMD assessment of the femoral neck. cBaseline and Week 1 (prior to first dose) values averaged if both were available. 7

TKT 025 and TKT 025EXT Clinical Trial Results -10 -20 -30 -40 -50 -60 -70 -80 20 40 60 80 100 120 160 140 180 Change from Baseline statistically significant at 9 months (P<0.003) and 48 months (P<0.004) for each parameter Mean change per key clinical parameter (%) 025 (n=11) 025EXT (n=9 had data to 45 months) Week 13 25 37 41 55 67 79 94 107 119 131 143 159 171 183 195 1 2 3 Year Platelet count Hemoglobin concentration Liver volume Spleen volume Zimran A, et al. Blood 2010;115:4651-6. 8

BONE DENSITY RESULTS: Z-Scores ; n = 10, ITT Time Z-Scores Lumbar Spine Mean [95% CI] Femur Baseline -1.59 [-2.17, -1.01] -1.46 [-2.11, -0.81] Month 9 -1.50 [-2.05, -0.95] -1.39 [-2.06, -0.72] Month 24 -1.20 [-1.80, -0.60] -1.23 [-2.04, -0.42] Month 33 -1.16 [-1.80, -0.52] -1.07 [-1.82, -0.32] Month 45 -1.12 [-1.72, -0.52] -1.14 [-1.86, -0.42] Month 57 -1.07 [-1.68, -0.46] -1.10 [-1.89, -0.31] Month 69 -0.91 [-1.55, -0.27] -1.06 [-1.83, -0.29] 9 9

Lumbar Spine BMD Z-scores: Individual Patients’ Data 13 patients were screened and all consented to participate in TKT025; patient 0004 was excluded before study start because of imiglucerase antibodies. Patients 0006 and 0012 did not enroll in the long-term extension study. Patients 0002, 0003, 0007, and 0010 received concomitant bisphosphonates.   Elstein D, et al. Blood Cells Mol Dis 2011;47:56-61. 10 10

Femoral Neck BMD Z-scores: Individual Patients’ Data 13 patients were screened and all consented to participate in TKT025; patient 0004 was excluded before study start because of imiglucerase antibodies. Patients 0006 and 0012 did not enroll in the long-term extension study. Patients 0002, 0003, 0007, and 0010 received concomitant bisphosphonates.   Elstein D, et al. Blood Cells Mol Dis 2011;47:56-61. 11 11

WHO Classification (T-scores) at Baseline and 69 Months Velaglucerase alfa without concomitant bisphosphonates (All 4 patients on bisphosphonates had no change in WHO Category) Lumbar Spine (n=6) Femoral Neck (n=6) Normal = T-score ≥ –1 Osteopenia = T-score > –2.5 and < –1 Osteoporosis = T-score ≤ –2.5 Elstein D, et al. Blood Cells Mol Dis 2011;47:56-61. 12

Changes in WHO Category; n = 6 [No Bisphosphonates] Patient Lumbar Spine Femur Baseline 69 Months 025-071-0001 OPN OPO 025-071-0005 025-071-0008* NORMAL 025-071-0009 025-071-0011 025-071-1003 Normal = T-score ≥ -1; osteopenia (OPN) =T-score > -2.5 and < -1; osteoporosis (OPO) = T-score ≤ -2.5 * Patient 23-years old at Baseline; one of two patients with osteonecrosis at Baseline 13 13

Conclusions: BMD improvement with velaglucerase alfa In patients with type 1 Gaucher disease and baseline osteopenia/osteoporosis who were treated with velaglucerase alfa, BMD improved in both lumbar spine (Month 24) and femoral neck (Month 33). Since the velaglucerase alfa dose was reduced from 60 to 30 unit/kg/infusion between 15-18 months, the improvement in bone pathology was not dependent upon continuous high-dose therapy. 14 14

Bone Marrow Burden Score Analysis: Objectives To evaluate the impact of up to 7 years of velaglucerase alfa therapy on BMB score at lumbar spine and femur To explore possible correlations before treatment was initiated (at Baseline) between BMB score and other clinical parameters DeMayo et al (Am J Roentgenol 2008;191:115-23) found positive correlation between spleen status (surgically absent or enlarged) and BMB scores at both sites, but not with duration of ERT with imiglucerase 15 15

Lumbar Spine BMB Scores at each assessment Change from Baseline in BMB score Patients with available data, n Median BMB score (range) Mean change (95% CI) Decrease of 2 points or more, n (%) Decrease of 3 points or more, Decrease of 4 points or more, Baseline 12 6 (3, 8) 9 months 11 4 (1, 8) –1.8 (–2.9, –0.7) 6 (55) 3 (27) 2 (18) 24 months 9 3 (1, 5) –2.8 (–4.2, –1.4) 6 (67) 4 (44) 3 (33) 33 months 1 (1, 4) –3.8 (–5.0, –2.5) 8 (89) 45 months 7 1 (1, 5) –3.3 (–4.9, –1.7) 6 (86) 4 (57) 3 (43) 57 months 8 –4.3 (–5.7, –2.8) 8 (100) 6 (75) 5 (63) 69 months –4.1 (–5.8, –2.5) 7 (88) 81 months CI, confidence interval. 16

Individual Lumbar Spine BMB Scores over 81 months (7 Years) of velaglucerase alfa therapy Patient number Baseline 9 months 24 months 33 months 45 months 57 months 69 months 81 months 1 8 5 4 3 2 6 4a 7 – 6b 9 10 11c 12d Dash indicates missing BMB score. aDiscontinued from study before 24 months. bDiscontinued from study after 48 months. cCompleted TKT025 but did not enroll in TKT025EXT. dDiscontinued from TKT025 after 3 infusions. 17

Mean temporal changes from Baseline in BMB Score at the Lumbar Spine a Clinically meaningful improvement at each anatomic site defined as a decrease of 2 points or more. CI, confidence interval. 18

all decreased by 2 points Individual Femoral BMB Scores over 81 months (7 Years) of velaglucerase alfa therapy: all decreased by 2 points Patients with MR images available at both Baseline and during TKT025EXT BMB score Patient number Baseline 9 months 24 months 33 months 45 months 57 months 69 months 81 months 2 5 – 4 3 9 1 10 Dash indicates missing BMB score. To score the femur, an image of the whole femur is needed. The imaging protocols for the studies did not specify that the whole femur needed to be imaged; consequently, many MR images only contained a part of it (eg, femoral neck) and could not be scored. 19

Clinical Correlations with Lumbar Spine BMB Score Baseline values for the hematologic and visceral parameters, chemokine (C-C motif) ligand 18 level, and lumbar spine BMD z-score were assessed for correlation with lumbar spine BMB score at Baseline The parameter with the strongest correlation with the lumbar spine BMB score at Baseline was with increased spleen volume, but this was not statistically significant (ρ=0.38; P=0.22) 20

Conclusions Bone marrow involvement improved among adults with GD1 treated with velaglucerase alfa At the lumbar spine, the mean change from Baseline in BMB score was statistically significant at 9 months and continued to improve through 5 years, despite dose reduction after the first year At 5 years, all patients with available data had experienced a clinically meaningful decrease in BMB score and the 5-year scores were the same at 7 years 21

Thank you for your attention! 22