Form 8-K AMICUS THERAPEUTICS INC For: Mar 01

180 180 B asel i n e M ont h 6 M ont h 1 2 M ont h 1 8 B asel i n e M ont h 6 M ont h 1 2 M ont h 1 8 ERT, in both males and females + m ig a la s t a t studies 50 30 40 0  - G a l A m ut a nt f or m s ( i n or de r o f a m i no a c i d s ubs t i t u t i on f r om N - t o C - t e r m i nu s ) lin ic a l S t u d ie s L ar g er S u b set C lin ic a l S t u d ie s L ar g er S u b set to migalastat were not significantly different (all doses) (150 mg QOD) M u ta ti o n s Patient samples are not required and the approach is applicable to both males and females 30 20 0 0 -1 0 0 5101520 • Includes all known types of mutations (i.e., missense, small insertions and deletions that maintain reading frame, 3 • Patients with amenable mutations showed consistent decreases in these substrate levels; larger 3 deposition, and plasma lyso-Gb concentrations studies are representative of the larger subset of amenable mutations identifying the target population for treatment with migalastat: patients with FD who have amenable • Approximately 30-50% of patients with FD are estimated to have amenable mutations; the majority to determine amenability to treatment with migalastat  - G a l A A c t i v i t y ( % W T ) N um be r o f A m e na bl e M u t a t i on s A b so l u t e I n cr ea se ( % W T ) P l asm a L yso - G b 3 1 ( M ea n C h an g e f r o m B as el i n e ; n M ) R el at i ve I n cr ease ( - f o l d ) Substrate Responses in Study 012 Amenable Non-Amenable 150150 120120 9090 6060 3030 00 -3 0-3 0 • In patients with amenable mutations, the plasma lyso-Gb3 levels were comparable to those seen with • In two male subjects with non-amenable mutations, plasma lyso-Gb3 increased following switch from ERT as compared to two (1M, 1F) who remained on ERT Phase 2/3 Amenable Mutations Compared to All • In total, 51 different amenable mutations were identified in 126 subjects from Phase 2 and 3 clinical • This represents 19% of all amenable mutations to date 80 6030 4020 2010 00 C • This set of amenable mutant forms of -Gal A (n=51) represented in clinical trials were compared to the larger FD-associated subset that met the amenable mutation criteria (n=268); the responses • The results suggest that the amenable mutant forms evaluated in Phase 2 and 3 clinical studies are representative of the larger subset of amenable mutant forms Amenable Mutations Grouped by Phenotype C l in ic a l S t u d y A m e n a b l eA l l A m e n a b l e 40125 M u ta ti o n s 100 75 50 1025 0 C l assi c N o n - cl assi cB o thC l ass i c N o n - c l ass i cB o th amenable in each phenotype category; mutant forms with unknown phenotype were excluded • A database of ~800 FD-associated GLA mutations was compiled based on literature review carboxyl-terminal nonsense mutations, complex mutations, large deletions or insertions, truncations, frameshift mutations, splice site mutations) • Includes information on whether that mutation has been associated with the classic and/or late-onset (variant) phenotype in the literature • The results show that a majority , ~65%, of all amenable mutations as well as those represented in migalastat clinical studies are associated with classic FD Conclusions • The results indicate that the Migalastat Amenability Assay and the amenable mutation criteria have high predictive value in identifying FD patients who show a pharmacodynamic response to oral administration of migalastat based on assessment of -Gal A in WBCs, kidney interstitial capillary GL-3 • The results indicate that the amenable mutations evaluated in the migalastat Phase 2 and 3 clinical • These results support the clinical validation of the Migalastat Amenability Assay and its utility in mutations of amenable mutations are associated with the classic phenotype of the disease • As new GLA mutations are identified, they can readily be tested in the Migalastat Amenability Assay Dotted line indicates the total number of amenable mutations with phenotype classification; percentages (%) indicate the % of total 6 5% 3% 3 2% 6 7% 2 5% 8% Patients with amenable GLA mutations in the Migalastat Amenability Assay 1Baseline corrected. Blue dotted line represents zero; data points represent the mean, Error bars are SD; Least Squares (LS) Mean at Month 18 showed results comparable to the mean (data not shown); Based on subjects with available samples for this analysis M i g al ast a t ER T ( n = 2 ) ( n = 2 ) ( n = 2 ) ( n = 2 ) ( n = 2 ) ( n = 2 ) ( n = 2 ) ( n = 2 ) M i g al ast a t ER T ( n= 32 ) ( n= 31 ) ( n= 32 ) ( n= 31 ) ( n= 17 ) ( n= 17 ) ( n= 17 ) ( n= 15 ) Migalastat Amenability Assay Procedure and Data Overview Binary format: 0 and 10 µM Migalastat (AT1001) Transfection control: qPCR e • The assay includes: A) a thorough and rigorous set of plasmid DNA quality control assessments and storage specifications; B) a simple binary design wherein GLA transfected HEK-293 cells are incubated in the absence or presence of a single concentration of migalastat (10 ); C) a quantitative real-time PCR (qPCR) transfection efficiency control measurement obtained from every sample; D) rigorous and consistent assay acceptance criteria - m i g a l as t at 120 110 100100% 90WT 80 70 60 40 20 103% WT • The assay data show that 600 tested -Gal A mutant forms span the entire length of the gene and show a wide range of -Gal A enzyme activities both at baseline and after incubation with migalastat Comparison to -Gal A Responses in Phase 2 and 3 • The mutant -Gal A responses to migalastat in the Migalastat Amenability Assay and in white blood cells (WBCs) of male Fabry patients orally administered migalastat in clinical studies were compared • The degree of consistency was evaluated by calculating the sensitivity, specificity, positive predictive value, and negative predictive value ValueValue 0.93751.01.00.87523 1.01.01.01.014 1.00.750.8751.022 1.01.01.01.015 (150 mg QOD)1.00.8750.9461.051 • A high degree of consistency between the Migalastat Amenability Assay results and the male subject WBC -Gal A results was obtained Comparison to Substrate Responses in Study 011 15 10 5 -5 -1 5 • Male and female kidney interstitial capillary GL-3 (IC GL-3) and plasma lyso-Gb absolute changes after six months of treatment were grouped by GLA mutation category decreases were observed with increasingly higher baseline values • In patients with non-amenable mutations, no consistent reductions in lyso-Gb3 were observed ValueValue • In Study 011, comparisons of Migalastat Amenability Assay results to patient substrate responses to migalastat showed high consistency Parameter ComparedPositiveNegativeNumber of with GLP HEK AssaySensitivity SpecificityPredictivePredictiveDifferent Patients Male Kidney IC GL-31.01.01.01.018 Male Plasma Lyso-Gb31.01.01.01.016 Male and Female Plasma 0.92860.68750.83870.846244 Lyso-Gb3 Patients with a kidney IC GL-3 or plasma lyso-Gb3 absolute change <0.0 after 6 months of treatment were categorized as showing “good” responses, and patients with 0.0 were categorized as showing “non/limited” responses; absolute change from baseline in Fabry substrate (i.e., kidney IC GL-3 or plasma lyso-Gb3) is calculated as the value after 6 months of migalastat treatment minus the value at baseline; the GLP HEK comparison to male kidney IC GL-3 included only males with a baseline kidney IC GL-3 level 0.1. Six months of migalastat refers to the change from baseline to month 6 in subjects randomized to migalastat in Stage 1; it refers to the change from month 6 to month 12 in subjects randomized to placebo in Stage 1. Criteria for a “good” WBC -Gal A response: 2% of normal maximal net increase after oral administration of migalastat. This comparison did not include mutant forms represented in female subjects, because PBMCs derived from females are a mixture of cells that express either wild-type or mutant -Gal A. Thus the measured -Gal A activity is dominated by the wild type enzyme, which is responsive to migalastat; hence, neither the baseline activity nor the effect of migalastat on the mutant form can be accurately PositiveNegativeNumber of Different SensitivitySpecificityPredictivePredictivePatients Phase 2 Phase 2 (150 mg QOD) AT1001-011 AT1001-012 (150 mg QOD) All male patients Lysates Step 7 qPCR qPCR assay plate plification Am curv Introduction Fabry Disease (FD) • Progressive X-linked lysosomal storage disorder caused by a deficiency in -galactosidase A • Estimated FD incidence of approximately 1 in 100,000. Actual prevalence may be higher • More than 800 disease-causing mutations in GLA have been identified;Kidney GL-3 ~60% of these are missense mutations • Affects males and females; females have a mosaic of healthy and diseased cells • Globotriaosylceramide (GL-3), a natural substrate of -Gal A, accumulates and affects multiple organs and organ systems (kidney, heart, brain, gastrointestinal, skin) • Globotriaosylsphingosine (lyso-Gb3) is another substrate of -Gal A that is elevated in plasma of male and female patients with FD Migalastat for FD: From Auray-Blais et al., 2010 • Orally administered investigational pharmacological chaperone for patients with amenable mutations • Increases stability, folding, and cellular trafficking of amenable mutant forms of -Gal A to lysosomes where the breakdown of substrate can proceed • Amenable mutant forms of -Gal A are identified using a GLP-validated HEK-293 cell-based assay (Migalastat Amenability Assay) • 30-50% of patients with FD are estimated to have amenable mutations; theMigalastat majority of amenable mutations are associated with the classic phenotypeDeoxygalactonojirimycin of the disease AT1001 Objectives • To assess the clinical validation of the Migalastat Amenability Assay, the mutant -Gal A responses to migalastat in the assay were compared to Fabry patient pharmacodynamic responses to treatment with migalastat in Phase 2 and 3 clinical studies Materials & Methods Migalastat Amenability Assay (GLP HEK Assay): • A bioanalytically validated assay used to individually express FD mutations in human embryonic kidney-293 (HEK) cells and measure increases in mutant -Gal A activity in response to 10 µM migalastat • Known FD associated missense, carboxyl-terminal nonsense, small in-frame insertion, deletion, and complex mutant forms of the enzyme qualify for testing in the Migalastat Amenability Assay • Amenable mutant forms are defined as those having a 1.2-fold relative increase and 3.0% absolute increase in -Gal A activity • • To date, 600 FD mutations have been tested; 268 have met the amenable mutation criteria Data From Three Phase 2 Studies of Migalastat: • FAB-CL-201 (NCT00214500), FAB-CL-202 (NCT00283959), FAB-CL-203 (NCT00283933) • The objectives were to evaluate the safety, tolerability, pharmacokinetics, and pharmacodynamics of migalastat in patients with FD • All three studies included males only • Study 201 evaluated different dosages; Studies 202 and 203 evaluated 150 mg migalastat HCl once every other day • All three studies were open-label, and included initial 12-24-week treatment periods and optional treatment extensions Data From Phase 3 Study AT1001-011 (NCT00925301): • A double-blind, randomized, placebo-controlled study to evaluate the efficacy, safety, and pharmacodynamics of migalastat HCl in patients with FD and amenable GLA mutations • Key Inclusion Criteria – Male or female, diagnosed with FD – Amenable GLA mutation (during screening the GLA mutation was confirmed by gene sequencing; the ‘amenable’ category was determined by a preliminary HEK-293 cell-based assay) – Naïve to enzyme replacement therapy (ERT) or has not received ERT for 6 months before screening Data From Phase 3 Study AT1001-012 (NCT01218659): • A randomized, open-label study to compare the efficacy and safety of migalastat HCl and ERT in patients with FD and amenable mutations who were previously treated with ERT • Key Inclusion Criteria – Male or female, diagnosed with FD – Amenable GLA mutation (during screening the GLA mutation was confirmed by gene sequencing; the ‘amenable’ category was determined by a preliminary HEK-293 cell-based assay) – Initiated treatment with ERT at least 12 months prior to the baseline visit The Validation of Pharmacogenetics in the Identification of Target Fabry Patients for Treatment with Migalastat Benjamin ER1, Della Valle C1, Wu X1, Katz E1, Valenzano KJ1, Bichet DG2, Germain DP3, Giugliani R4, Hughes DA5, Schiffmann R6, Wilcox WR7, Yu J1, Kirk J1, Barth J1, Castelli J1 1Amicus Therapeutics, Cranbury, NJ, USA; 2Hôpital du Sacré-Coeur, Montréal, Quebéc, H4J1C5, Canada; 3Division of Medical Genetics, University of Versailles, University Paris-Saclay, Montigny, France; 4Medical Genetics Service, HCPA/UFRGS Porto Allegre, Brazil; 5Royal Free Campus, Univ College London, London, UK; 6Baylor Research Institute, Dallas, TX; 7Dept of Human Genetics, Emory Univ, Atlanta, GA, USA

Phenotype of Fabry Disease in Patients with Mutations Amenable to Migalastat Hughes, D1, Bichet, DG2, Germain DP3, Giugliani R4, Schiffmann R5, Wilcox W6, Castelli J7, Benjamin E7, Skuban N7, and Barth J7 1 University College London, London, UK; 2Hôpital du Sacré-Coeur, University of Montreal, Canada; 3Hôpital Raymond Poincaré (AP-HP), University of Versailles – St. Quentin en Yvelines (UVSQ), Garches, France; 4Medical Genetics Service, HCPA/UFRGS Porto Alegre, Brazil; 4Royal Free Campus, University College London, London, UK; 5Baylor Research Institute, Dallas, TX; 6Department of Human Genetics, Emory University, Georgia; 8Amicus Therapeutics, 1 Cedar Brook Drive, Cranbury, NJ, USA Introduction Clinical Phenotypes Among mutations characterized in the literature, a majority (64%) of patients in the Phase 3 studies had mutations associated with the classical phenotype. Patients Had Significant Baseline Disease Severity Fabry Disease • A devastating X-linked inherited disorder caused by the functional deficiency of lysosomal galactosidase A (-Gal A), with accumulation of glycosphingolipids, including globotriaosylceramide (GL-3), leading to impairment of kidney, heart, brain, and premature death. • More than 800 disease-causing mutations in GLA have been identified (~60% missense). • Affects males and females; females have mosaic of healthy and diseased cells. Migalastat for Fabry Disease • Kidney GL-3 Study 011: Amenable mutations of patients and the corresponding clinical phenotype, based on the medical literature Coronary GL-3 • Binds to Gal A, increasing its physical stability, lysosomal trafficking, and cellular activity. First-in-class orally administered (QOD) pharmacological chaperone being developed as a targeted medicine for the treatment of Fabry disease in patients with amenable GLA mutations. Between 30-50% of people with Fabry disease express mutant forms of -Gal A that are amenable to migalastat, based on an in vitro GLP-validated Migalastat Amenability Assay. • • AT1001; Migalastat HCl; Deoxygalactonojirimycin • Overall, 91% of patients had Fabry disease involvement in 2 organ systems, indicating significant disease burden. In Study 011, all patients had clinical manifestations, and 90% of patients had renal involvement, 52% had cardiac involvement, and 54% had CNS involvement. In Study 012, all patients had clinical disease manifestations, and 75% of patients had renal involvement, 72% had cardiac involvement, and 53% had CNS involvement. DESIGN of AT1001-011 (FACETS, NCT00925301) • Key Inclusion/Exclusion Criteria: • • Males and females, 16 to 74 years, diagnosed with Fabry disease. Amenable GLA mutation. Naïve to ERT or have not received ERT for 6 months before screening. eGFR (MDRD) at screening 30 ml/min/1.73 m2. Urine GL-3 at screening 4 times the upper limit of normal (24-hour collection). Subjects taking ACE inhibitors, ARBs, or renin inhibitors on a stable dose for at 4 weeks before screening. • • Patients Enrolled in the Migalastat Phase 3 Studies Are Comparable With Fabry Disease Patients Currently Receiving ERT • • BASELINE Study 012: Amenable mutations of patients and the corresponding clinical phenotype, based on the medical literature • DESIGN of AT1001-012 (ATTRACT, NCT01218659) Key Inclusion/Exclusion Criteria: • Males and females, 16 to 74 years, diagnosed with Fabry disease. Amenable GLA mutation Initiated treatment with ERT at least 12 months prior to baseline visit. Stable ERT dose for 3 months prior to baseline visit and 80% of labeled dose. eGFR (MDRD) at screening 30 ml/min/1.73 m2. Subjects taking ACE inhibitors, ARBs, or renin inhibitors on a stable dose for 4 weeks before screening. • • • Summary and Conclusions • • •The very high proportion of patients with multi-organ system involvement in the Phase 3 studies of migalastat (Studies 011 and 012), associated elevated plasma lyso-Gb3 and low -Gal A activity in patients not receiving ERT, indicate substantial disease burden in this population. Methods TESTING OF GENOTYPES FOR AMENABILITY: Patients had Low -Gal A Activity and Elevated Plasma Lyso-Gb3 Levels Low residual -Gal A activity in male patients and elevated levels of plasma lyso-Gb3 in males and females have been associated with the classical Fabry phenotype (Desnick, Brady et al., 2003; Wilcox, Oliveira et al., 2008; Rombach, Dekker et al., 2010). • A majority of patients in the Phase 3 migalastat studies had a phenotype associated with the classical phenotype. • 600 Fabry disease-causing mutations were expressed in transfected HEK-293 cells and -Gal A activity was measured in the presence and absence of 10 µM migalastat. Amenable mutant forms were defined by a 1.20-fold relative increase and a 3.0% wild-type absolute increase in the presence of 10 µM migalastat. 268 amenable mutations were identified. • • Patients enrolled in the migalastat Phase 3 studies are comparable with the current Fabry disease population being treated with ERT, as reflected in both the Fabry Outcome Survey (Mehta, Beck et al., 2009) and the Fabry Registry (Eng, Fletcher et al., 2007; 2014). • • • In Study 011, >90% of patients had plasma lyso-Gb levels comparable 3 to patients with a classical phenotype (Rombach, Dekker et al., 2010); 91% of males had plasma lyso-Gb3 >51 nM; 94% of females had plasma PHENOTYPE : Acknowledgments • Proportions of patients enrolled in Studies 011 and 012 with disease-related involvement of 2 organ systems were determined. Patient’s phenotypes (classical/non-classical) were assessed based on the medical literature definition of genotypes. The classical Fabry phenotype has been used to described patients with early onset, low residual -Gal A activity (in male patients), elevated plasma lyso-Gb3, and multiple organ-system disease. lyso-Gb >1.19 nM. • Patients and their families Patrick Deegan David Dimmock Fatih Ezgu Francois Eyskens Ulla Feldt-Rasmussen Claudio Feliciani Ana Jovanovic David Koeller Nicola Longo Robin Lachmann Charles Lourenco Joel Charrow Takashi Hamazaki Ichiei Narita Khan Nedd Kathleen Nicholls Toya Ohashi Iacopo Olivotto Seymour Packman Norio Sakai Raphael Schiffmann C. Ronald Scott Suma Shankar Katherine Sims Gere Sunder-Plassmann Akemi Tanaka Mark Thomas Roser Torra Ahmad Tuffaha Stephen Waldek William Wilcox 3 • • • • • • • Hernan Amartino Daniel Bichet Drago Bratkovic David Finegold Gerard Vockley Dominique Germain Pilar Giraldo Majed Dasouki Ozlem Goker-Alpan Roberto Giugliani Eric Hachulla Derralynn Hughes Charles Jeanette Robert Colvin Usama Sharaf El Din Maryam Banikazemi • • 44% of males had baseline Gal A Activity <1% of normal, and 87% had baseline activity <3% of normal. • (Due to previous ERT treatment in patients entering Study 012, enzyme activity and plasma lyso-Gb3 levels were confounded.) Study AT1001-012: A Randomized, Open-Label Study To Compare The Efficacy and Safety Of Migalastat and Enzyme Replacement Therapy (ERT) in Patients With Fabry Disease and Migalastat-Responsive GLA Mutations, Who Were Previously Treated With ERT Baseline Characteristics in Phase 3 Migalastat Studies Versus ERT Registries Migalastat HCl 150 mg Every-Other-Day (QOD) Placebo QOD MONTH MONTH 612 Stage 1Stage 2Optional Extension 6 Months 6 Months 12 Months Double-Blind Open-Label Kidney biopsy Screening 1:1 Randomization Stratified by Sex Double-Blind Open-Label Migalastat HCl 150 mg QOD 2 Months Study AT1001-011: A Double-Blind, Randomized, Placebo-Controlled Study to Evaluate the Efficacy, Safety and Pharmacodynamics of Migalastat HCl in Patients With Fabry Disease and Amenable GLA Mutations

Comparison of Integrated White Blood Cell -Galactosidase A Activity Exposure Between Every-Other-Day Orally Administered Migalastat and Biweekly Infusions F. K. Johnson1, K. J. Valenzano1, and J. Castelli1 1Amicus Therapeutics, Cranbury, NJ USA of Agalsidase Beta or Agalsidase Alfa randomized double-blind, placebo-controlled study, and AT1001-012, a randomized, open-label, comparator study with ERT and migalastat. Both Phase 3 studies periodically for up to 24 months. AT1001-013 was an open-label, single dose study in a fixed sequence with ERT alone first, and then co-administered with either in Fabry patients, the outcome of which was used in the current analyses. -Gal A activity in WBCs was measured pre-infusion, and at 2, 4, 24, 168, and 336 rate of turn-over of an artificial substrate, 4-MUG to 4-MU. Circulating WBCs were selected as an example of a surrogate for tissue uptake because of ease of similar changes in skin and kidney -Gal A activity levels.e The data analysis methods were comprised of modeling and simulations to predict exposure of WBC dose 1 mg/kg Agalsidase beta or 0.2 mg/kg Agalsidase alfa 1 mg/kg agalsidase beta and is approximately 6-fold greater than that seen following a 0.2 mg/kg rate constants for a multiple-dose simulation. Multiple-7 every-other-day oral administrations of 150 mg noncompartmental analysis on the 14-day multiple-dose indicated by the arrows and followed by an inhibitory phase lasting another 24 hours. based on in vitro data and animal models. The following constant or zero-order rate of increase in -Gal A the characterization of -Gal A activity in plasma and patients is represented as the blue line plotted on the • Therefore, some individuals may have greater or lesser responses C ha nge f r o m B a s e l i ne i n  - G a l A A c t i v i t y i n W B C s ( n m o l / h r / m g ) P l asm a M i g al ast a t C o n c ( n M ) C ha nge f r o m B a s e l ine in  - G a l A A c t i v i t y i n W B C s ( n m o l / h r / m g ) * C ha nge f r o m B a s e l i ne i n  - G a l A A c t i v i t y i n W B C s ( n m o l / h r / m g ) * Introduction Fabry disease is an x-linked -galactosidase (-Gal A) deficiency. It involves progressive globotriaosylcerimide (GL-3) accumulation, which affects multiple organs and organ systems including the kidney and heart. Currently approved treatments include once-every-other-week infusions with enzyme replacement therapies 1 mg/kg agalsidase beta or 0.2 mg/kg agalsidase alfa. Misfolded or unstable -Gal A is degraded in the endoplasmic reticulum. Migalastat HCl is a low molecular weight iminosugar and is an analogue of the terminal galactose of GL-3 that binds to the active site of -Gal A. Pre-clinical in vitro and in vivo studies have demonstrated that migalastat acts as a pharmacological chaperone for -Gal A, selectively and reversibly binding, with high affinity, to the active site of both wild-type and specific mutant forms of -Gal A, the genotypes of which are referred to as amenable mutations.a In in vitro and in vivo models bound migalastat stabilizes -Gal A, slowing its denaturation at neutral pH and body temperature.b Migalastat binding stabilizes these mutant forms of -Gal A in the endoplasmic reticulum facilitating their proper trafficking to lysosomes where dissociation of migalastat allows -Gal A to reduce GL-3 storage material. In contrast, misfolded and/or unstable -Gal A is recognized by the endoplasmic reticulum quality control system as aberrant and targeted for degradation, never reaching the lysosome.c The PK of migalastat has been well-characterized. Migalastat is dose proportional from 50 to 1250 mg, well absorbed in 3 hours, and has a terminal half-life of approximately 4 hours.d Data Analysis Methods The studies included in this data analysis were two Phase 3 studies, AT1001-011 and AT1001-012, and a Phase 2a study, AT1001-013. AT1001-011 was a were conducted in Fabry patients with amenable mutations, patients were dosed with migalastat every-other-day, and -Gal A activity in WBCs were measured 150 mg or 450 mg migalastat in male Fabry patients with any mutation. An additional arm with 150 mg migalastat alone was used to characterize the plasma PK hours post-start of infusion of agalsidase. The bioanalytical method for determination of -Gal A activity in WBCs was a fluorescence assay, which measured the sampling, as well as excellent exposure to both agalsidase and migalastat. Importantly, migalastat-mediated changes in WBC -Gal A levels were associated with -Gal A activity following oral administration of 150 mg migalastat every-other-day for 7 doses, 14 days total, and noncompartmental analyses to estimate mean WBC -Gal A activity exposure following single infusions of 1 or 0.2 mg/kg agalsidase beta or alfa, respectively. Migalastat administration results in more consistent levels of WBC -Gal A Activity Figure 1. WBC -Gal A Activity Following 150 mg Migalastat QOD X 7 Doses vs. Single-Table 1. WBC -Gal A Activity PK Summary 1 . 0 m g / kg A g a l s i d ase B e t a 150 m g M i g a l a s t at Q O D S i m u l a t i o n 0 . 2 m g / kg A g a l s i d ase A l f a 100 10 As shown in Figure 1, every-other-day dosing with migalastat suggests more consistent levels of -Gal A activity compared to a 14-day ERT dosing interval. As shown in Table 1, simulated migalastat AUC for WBC -Gal A activity is comparable to that seen following a single dose of 1 agalsidase alfa. Single doses of agalsidase beta resulted in higher Cmax values (mean of 105.9) with rapidly declining -Gal A activity. The simulated AUC for migalastat represents an attenuated Cmax with more consistent levels of activity as a result of every-other-day dosing over the same 14-day 0 .1 interval. All baseline endogenous levels of activity were subtracted from each time point for 0 48 96 144 192 240 288 336 estimation of PK parameters. The exposure values presented in the abstract were not baseline-T i m e ( h r ) corrected. Proposed Relationship of Plasma Migalastat to WBC -Gal A Activity Figure 2. Plasma Migalastat and WBC -Gal Activity vs. Time (14 Days) Before simulating -Gal A activity over a 14-day interval, a hypothetical single-dose model was established. Data 100000100from 79 Fabry patients were pooled from the 011 and 012 Phase 3 studies. The time of blood sampling for -Gal A activity determinations were not recorded in either Phase 3 study. Therefore, a hypothetical activity-time curve was constructed using actual WBC -Gal A 1000010activity data from Phase 3 studies in Fabry patients. Each value was baseline-corrected and were used to create the red line plotted on the right Y-axis. The shape of this curve was based upon the pharmacokinetic characterization of  Gal A activity in plasma and 10001WBCs, and the duration of inhibition by migalastat assumptions were made for assigning time points to WBC activity values: duration of inhibition by migalastat from Time 0 to approximately 24 hours post-dose, a 1000 .1activity with maximum activity attained by roughly 3 081624324048hours, and a biphasic elimination rate consistent with T i m e ( h r ) WBCs. Migalastat in plasma from the 013 study in Fabry left Y-axis. Multiple-dose Simulation Figure 3. WBC -Gal A Activity Simulation Following 7 Doses with 150 mg Migalastat QOD The next step was to fit a 2-compartment model to the 100 activity-time curve to obtain volume of distribution and dose simulation was performed for a total of 14 days or migalastat. An overall AUC was estimated by a simulation. The time of migalastat dose administration is phase lasting approximately 24 hours, and an activation 1 e ( h r ) Conclusions Based on Modeling and Simulation and Limitations of the Analysis Conclusions: •Following Q14d single-dose infusions with agalsidase beta or alfa to Fabry patients, or 7 QOD oral administrations of 150 mg migalastat HCl to Fabry patients with amenable mutations over 14 days, -Gal A in WBCs were: •Comparable between agalsidase beta and migalastat, but were 6-fold greater for migalastat compared to agalsidase alfa. •More consistent following QOD administration of migalastat, which provided lower Cmax values and higher Ctrough values than single infusions of agalsidase which ultimately suggests more consistent cellular -Gal A activity levels. Limitations: •Time of sampling relative to dosing for WBC activity data from Phase 3 studies AT1001-011 and -012 was not recorded •Therefore, a hypothetical activity level-time curve was constructed from actual data •Selected activity levels from the combined 011 and 12 data sets were assigned to fit an assumed constant and rapid rate of increase and biphasic elimination rate based upon the characterization of -Gal A in plasma •WBCs are not a disease-relevant tissue for Fabry •However, circulating WBCs were selected as an example of tissue uptake because of ease of sampling, ample exposure to both -Gal A ERT and migalastat, and association with similar migalastat-mediated changes in -Gal A activity levels that were observed in skin and kidney tissue •WBC -Gal A activity may be overestimated following agalsidase administration to IgG positive patients who have greater uptake of -Gal A into WBCsf •The hypothetical model and simulation presented here represents a mosaic of different amenable mutant forms Presented at The World Symposium for Lysosomal Disorders Annual Meeting March, 2016 San Diego, CA References aIshii, Chang, et al. J Pharmacol Exp Ther. 2009. bBenjamin, Khana, et al. Mol Ther 2012. cYam, Zuber, et al. Am J Physiol, Cell Physiol. 2006. dJohnson, Mudd, et al. Clin Pharmacol Drug Develop. 2013. eGermain, Giugliani, et al. Orphanet J Rare Dis. 2012. fLinthorst, Hollak, et al. Kidney Int’l. 2004. 10 0 48 96 144 T i m 192 240 288336 Time of migalastat dose Treatment (N) AUC [hr*(nmol/hr/mg)] Cmax (nmol/hr/mg) Study 150 mg migalastat HCl (79) 2969 39.9 Combined 011/012 1.0 mg/kg agalsidase  (9) 3091 105.9 AT1001-013 0.2 mg/kg agalsidase  (8) 485 4.83 AT1001-013

Persistence of Positive Renal and Cardiac Effects of Migalastat in Fabry Patients with Amenable Mutations Following 30 Months of Treatment in the ATTRACT Study Bichet DG1, Germain DP2, Giugliani R3, Hughes D4, Schiffmann R5, Wilcox W6, Castelli J7, Cantor E7, Kirk J7, Skuban N7, and Barth J7 on behalf of the ATTRACT investigators 1Hôpital du Sacré-Coeur, University of Montreal, Canada; 2Hôpital Raymond Poincaré (AP-HP), University of Versailles – St. Quentin en Yvelines (UVSQ), Garches, France; 3Medical Genetics Service, HCPA/UFRGS Porto Alegre, Brazil; 4Royal Free Campus, University College London, London, UK; 5Baylor Research Institute, Dallas, TX; 6Department of Human Genetics, Emory University, Georgia; 7Amicus Therapeutics, 1 Cedar Brook Drive, Cranbury, NJ, USA Introduction Baseline Characteristics Baseline Disease Severity Fabry Disease • A devastating X-linked inherited disorder caused by the functional deficiency of lysosomal -galactosidase A, with accumulation of glycosphingolipids, including globotriaosylceramide (GL-3), leading to impairment of kidney, heart, brain, and premature death. • More than 800 disease-causing mutations in GLA have been identified (~60% missense). • Affects males and females; females have mosaic of healthy and diseased cells. • The stabilization or slowing of renal dysfunction and reduction of cardiac complications remain critical medical needs for individuals living with Fabry disease. Kidney GL-3 Abbreviations: CNS = Central Nervous System; eGFR = estimated glomerular filtration rate; LVH = left ventricular hypertrophy; LVMi = left ventricular mass index; TIA = transient ischaemic attack | Notes: Angiokeratoma or Corneal Whorling based on medical history finding. Cardiac Involvement includes previous cardiac event (based on medical history), LVH, or conduction abnormality (eg, tachycardia, ST-T segment abnormality) based on medical history finding or baseline assessment of LVMi. CNS involvement was based on medical history findings (stroke/TIA, tinnitus/hearing loss). Renal Involvement based on medical history finding or baseline eGFR <90 mL/min/1.73m2, 24-hr Protein 150 mg. | Coronary GL-3 Migalastat for Fabry Disease • First-in-class orally administered (QOD) pharmacological chaperone being developed as a targeted medicine for the treatment of Fabry disease in patients with amenable GLA mutations. Between 30-50% of people with Fabry disease express mutant forms of -galactosidase A that are amenable to migalastat, based on an in vitro GLP-validated Migalastat Amenability Assay. In patients with Fabry disease, migalastat binds and stabilizes the amenable mutant forms of the enzyme in the endoplasmic reticulum throughout the body and restores trafficking to lysosomes where the enzyme can catabolize accumulated glycosphingolipids. As an oral small molecule treatment, migalastat therapy is unlikely to exhibit the limitations of ERT, which include infusion-associated reactions, formation of antibodies to the exogenous protein, and the significant burden that biweekly infusions place on patients and their families. • • All Study 012 patients with amenable mutations had clinical manifestations of Fabry and were eligible for treatment based on current guidelines. The age at enrollment/start of ERT treatment and the percentages of patients with involvement of different organ systems in Study 012 were comparable with those for patients reported in the Fabry Outcomes Survey (Mehta, Ricci et al. 2004) and the Fabry Registry (Eng, Fletcher et al. 2007). These findings indicate that Study 012 patients are comparable with the current Fabry population being treated with ERT, as reflected in the ERT registries. • AT1001; Migalastat HCl; Deoxygalactonojirimycin • • 30-Month Renal Results 30-Month LVMi Results • DESIGN of AT1001-012 (ATTRACT, NCT01218659) • The 30-month analyses include patients with amenable mutations (based on the Migalastat Amenability Assay) and baseline/post-baseline measures of eGFR and mGFR (renal analyses) or LVMi (ECHO analyses). Summary of 30-Month Study Renal and LVMi Findings 31 male/female patients with amenable mutations who were randomized to the migalastat group completed the 18-month randomized period and entered the 12-month open-label extension . 49 patients with amenable mutations received 1 dose of migalastat during the combined 30 months. The annualized rates of change in eGFRCKD-EPI and mGFRiohexol for migalastat (see Table above) are comparable to those previously reported in patients receiving ERT for 18 months: -1.0 (-3.6, 1.6) and -3.2 (-7.8, 1.3), respectively. For patients receiving ERT, previously reported 18-month changes in LVMi were -2.0 (-11.0, 7.0) for all patients and +4.5 (-20.9, 29.9) for patients with baseline LVH. For renal function, in patients switched from ERT, the effect of migalastat is persistent, with similar results observed over 18 and 30 months of treatment. For LVMi, the reduction in patients switched from ERT to migalastat is also • Safety (ITT Patients) • • The 51 patients in the safety population – amenable/non-amenable mutations– had a mean duration of migalastat exposure of 896 days. Only 1 SAE was assessed as possibly related to migalastat by the investigator: proteinuria. This occurred during pregnancy in a patient with history of proteinuria during pregnancy. • • • Migalastat was generally safe and well tolerated based on adverse event, laboratory, and physical exam data. Acknowledgments Patients and their families Study 012 PIs: Patrick Deegan (UK) David Dimmock (US) Francois Eyskens (France) David Feingold (US) Ulla Feldt-Rasmussen (Denmark) Dominique Germain (France) Ozlem Goker-Alpan (US) Tashaski Hamazaki (Japan) Eric Hachulla (France) Derralynn Hughes (UK) Ana Javanovic (UK) David Koehler (US) Robin Lachmann (UK) Charles Lourenco (Brazil) Ichiei Narita (Japan) Khan Nedd (US) Kathy Nicholls (Australia) Toya Ohashi (Japan) Iacopo Olivotto (Italy) Norio Sakai (Japan) Suma Shankar (US) Gere Sunder-Plassmann (Austria) Akemi Tanaka (Japan) Mark Thomas (Australia) Roser Torra (Spain) Ahmad Tuffaha (US) Gerald Vockley (US) Randomized patients were 16-74 years of age and had: • • • A genetically confirmed diagnosis of Fabry disease. Initiated ERT 12 months before the baseline visit and a stable dose (>80% of the labeled does) for 3 months prior to the baseline visit. A responsive GLA mutation based on a preliminary cell-based assay. Estimated glomerular filtration rate (eGFRMDRD) 30ml/min/1.73m2. Patients taking angiotensin converting enzyme inhibitors, angiotensin receptor blockers had to be on a stable dose for 4-weeks before the screening visit. • persistent with similar results observed over 18 and treatment. 30 months of migalastat • • • • In patients with LVH at baseline, the reduction to month 30 for migalastat was statistically significant based on the 95% CIs. Conclusions Methods • • • The GFR stabilization and reduction in LVMi demonstrated with migalastat treatment are clinically relevant. Based on the literature, annualized rates of decline in GFR in ERT-treated Fabry patients are -2.2 to -2.9 ml/min/m2. RENAL • • • eGFR was assessed at intervals of 2-3 weeks until month 24 and again at month 30. CKD-Epi In patients switched from ERT to migalastat, the annualized rates 30 were: -1.0 (-3.6, 1.6) and -3.2 (-7.8, 1.3), respectively. of change (95% CI) in eGFRCKD-EPI and mGFRiohexol at month mGFRIohexol was assessed at baseline and months 6, 12, 18, and 30. The long-term effect was assessed by calculating the annualized rates of change for each patient using the slope of the linear regression between the observed values and the assessment times. • LVH is the greatest risk factor for cardiac events in Fabry disease (Patel, Cecchi et al. 2011), and any reduction in LVH has been shown to have a positive impact on cardiovascular morbidity and mortality in hypertensive heart disease (Pokharel and Bella 2013). Migalastat reduced cardiac mass in all 012 patients following 30-month treatment and, most importantly, in patients with LVH (cardiac hypertrophy), the reduction was statistically significant. The effects of migalastat on GFR and LVMi observed following 18 months persist over 30 months. The results suggest that migalastat is a promising first-in-class oral chaperone treatment for male and female patients with amenable mutations. ECHOCARDIOLOGY • Left ventricular mass index (LVMi) collected by Echo using 2D or M-mode every 6-12 months through blinded, centralized evaluation (Cardiocore, Rockville, MD). • The long-term effect was assessed by calculating the change from baseline to the last available timepoint and the 95% confidence interval for each patient. • • • A Randomized, Open-Label Study to Compare the Efficacy and Safety of AT1001 and Enzyme Replacement Therapy (ERT) in Patients with Fabry Disease and AT1001-Responsive GLA Mutations, who were Previously Treated with ERT Parameter Parameter Statistic eGFRCKD-EPI mGFRiohexol Annualized Rate of Change (mL/min/1.73 m2) Baseline – Month 30 n 31 30 Mean -1.718 -2.746 SD 2.5501 5.5318 95% CI (-2.653, -0.782) (-4.812, -0.681) Median -1.934 -3.190 Group Parameter Statistic Overall LVH at Baseline LVMi (g/m2) Baseline n 30 11 Median 89.780 109.780 Mean 94.649 116.440 SD 22.4222 20.9471 OLE Period Month 30 Actual n 29 10 Median 87.140 101.075 Mean 89.266 105.583 SD 20.2636 18.5857 Change from Baseline n 28 10 Median -4.580 -11.335 Mean -3.772 -9.959 SD 13.1540 9.3260 95% CI (-8.873, 1.328) (-16.630, -3.288) • HCl Sex Fabry Disease in 2 Organ Systems Angio-keratoma or Corneal Whorling Cardiac CNS Neuro-pathic Pain Renal Gastro-intestinal Males n (%) 21/24 (88%) 13/24 (54%) 16/24 (67%) 18/24 (75%) 14/24 (58%) 18/24 (75%) 14/24 (58%) Females n (%) 29/33 (88%) 16/33 (48%) 25/33 (75%) 12/33 (36%) 22/33 (67%) 25/33 (76%) 20/33 (61%) Intent-to-Treat Population Migalastat Arm (n=36) ERT Arm (n=21) Sex Female n (%) Male n (%) 20 (56) 16 (44) 12 (57) 9 (43) Median Age (range) 54 (18, 70) 48 (18, 72) Years since diagnosis Mean (SD) 10 (12) 13 (12) eGFRCKD-EPI, (mL/min/1.73 m2) Mean (SD) 89.6 (22) 95.8 (19) mGFRiohexol (mL/min/1.73 m2) Mean (SD) 82.4 (18) 83.6 (24) 24-hr Urine Protein (mg) Mean (SD) 260 (532) 417 (735) ACEi/ARB /RI Use: n (%) 16 (44) 11 (52) Amenable based on Migalastat Amenability Assay: n (%) 34 (94) 19 (90)

Co-Administration of the Pharmacological Chaperone AT2221 with A Proprietary Recombinant Human Acid -Glucosidase Leads to Greater Plasma Exposure and Substrate Reduction Compared to Alglucosidase Alfa Khanna R, Xu S, Hilliard D, Lun Y, Schilling A, Soska R, Nair A, Chang K, Feng J, Frascella M, Garcia A, Pendino K, Johnson FK, Benjamin ER, Gotschall R, Do H, and Valenzano KJ Amicus Therapeutics, 1 Cedar Brook Drive, Cranbury, NJ 08512, USA Introduction Pompe disease is an inherited lysosomal storage disorder that results from a deficiency in acid alpha-glucosidase (GAA) activity, and is characterized by progressive accumulation of lysosomal glycogen in cardiac and skeletal muscles. Enzyme replacement therapy (ERT) using recombinant human GAA (rhGAA) is the only approved treatment available for Pompe disease. While rhGAA provides some clinical benefits, the infused enzyme shows insufficient uptake into key disease-relevant muscles, which is likely due to sub-optimal levels of mannose-6-phosphate (M6P), a carbohydrate that binds cation-independent M6P receptors (CI-MPR) at the cell surface resulting in enzyme internalization and lysosomal targeting. In order to increase the targeting efficiency of ERT, we have developed a proprietary mammalian cell line and purification process that yields a novel form of rhGAA (designated as ATB200) with a significantly higher M6P content compared to the alglucosidase alfa. In this study, we have examined the effects of ATB200 on tissue exposure and substrate reduction with and without the addition of a small molecule pharmacological chaperone (PC) AT2221. 1. ATB200 Has a Higher M6P Content and Results in Better Tissue Uptake and Greater Glycogen Reduction in vivo Compared with Alglucosidase Alfa 3.AT2221 Co-administration Leads to Greater ATB200-mediated Glycogen Reduction and Reduces Lysosome Proliferation in Disease-relevant Muscles of Gaa KO Mice (A) AT2221 Dose Range Determination (A) CI-MPR Affinity Chromatography (B) Quadriceps of Gaa KO Mice 300 Uptake 6 5 Quadriceps 125 250 4 100 * 4 200 3 * 75 2 2 50 150 1 0 25 0 0 Substrate 400 Alglucosidase alfa (20 mg/kg) ATB200 (20 mg/kg) ATB200 + AT2221 (20 + 3 mg/kg) ATB200 + AT2221 (20 + 10 mg/kg) ATB200 + AT2221 (20 + 30 mg/kg) Untreated 100 5 300 4 75 (A) Twelve-week-old male Gaa KO mice were administered a total of 2 bi-weekly bolus IV injections of 20 mg/kg alglucosidase alfa or ATB200. In addition, ATB200 was co-administered with various doses of AT2221 (3-30 mg/kg). Glycogen levels were determined in quadriceps collected 14 days post-last dose. (Left) While ATB200 alone resulted in greater glycogen reduction compared to alglucosidase alfa, its efficacy is further improved by co-administration, mostly significantly with 10 mg/kg AT2221 (green arrow). Bars represent Mean ± SEM of 7-21 mice/group. * p<0.05 vs. alglucosidase alfa; # p<0.05 vs. ATB200 alone in 2-sided t-test. (Right) PAS staining also showed the lowest glycogen level in quadriceps with the co-administration of 10 mg/kg AT2221 with 20 mg/kg ATB200 (upper panels). Co-administration of 10 mg/kg AT22221 also led to marked further reduction in glycogen compared to ATB200 alone in heart (bottom panels) and additional tissues (data not shown). Magnification = 200x. 200 3 50 100 2 0 25 1 0 0 (B) Subsequently, the effect of co-administration of 20 mg/kg ATB200 + 10 mg/kg AT2221 was compared with 20 mg/kg alglucosidase alfa or ATB200 alone in another 2-biweekly-administration study in male Gaa KO mice of (B) IHC Examination of LAMP1 in Disease-Relevant Muscles of Gaa KO Mice rhGAA lacking M6P rhGAA containing M6P (C) Glycogen by PAS Stain twelve-weeks of age. IHC examination of lysosome marker LAMP1 in quadriceps revealed a substantial up-regulation of LAMP1 in fibers of untreated animals (top panel), which is indicative of lysosomal proliferation, a hallmark of Pompe disease. Unlike alglucosidase alfa, ATB200 alone leads to a marked decrease in LAMP1 signal, whose level was lowered further still with the co-administration of AT2221, approaching that seen in WT tissues. The change in LAMP1 level closely follows the change in glycogen level in quadriceps, and is repeated in additional tissues, such as heart, diaphragm, and soleus. Magnification = 400x. (C) Moreover, the fiber type response to ATB200 was investigated by IHC with LAMP1 antibody (top) and a type I (slow twitch) fiber-specific antibody NOQ7.5.4D (bottom) on adjacent sections of soleus, which has a relative equal representation of both type I and type II (fast twitch) fibers. ATB200 alone is much more effective than alglucosidase alfa, as indicated by the normalization of LAMP1 levels in most type I fibers and, significantly, a fraction of type II fibers as well, contrary to their reported resistance to alglucosidase alfa. With co-administration, a reversal of lysosomal proliferation was achieved in the majority of muscle fibers, regardless of fiber type. This Alglucosidase alfa (20 mg/kg) Untreated ATB200 (5 mg/kg) ATB200 (10 mg/kg) ATB200 (20 mg/kg) (A) Alglucosidase alfa or ATB200 was loaded onto a CI-MPR column. Only enzyme that contained M6P was retained and then eluted from the column using free M6P of increasing concentration (dotted line in red). Both unbound (flow-thru) and bound/eluted fractions were collected and assayed for GAA activity. The majority of ATB200 (91%) was bound compared to alglucosidase alfa (27%), suggesting that ATB200 has a higher M6P content, which is key to the efficient endocytosis and lysosomal targeting of rhGAA. (B) Twelve-week-old male Gaa KO were administered 2 bi-weekly bolus intravenous (IV) injections of alglucosidase alfa (20 mg/kg) or ATB200 (5-20 mg/kg) via tail vein (n=6-7 per group). Quadriceps were collected 14 days post the last dose and measured for GAA activity and glycogen levels. ATB200 shows dose-dependent increases in uptake and substrate reduction. Importantly, 5 mg/kg ATB200 is comparable to 20 mg/kg alglucosidase alfa, whereas 20 mg/kg ATB200 is significantly better than alglucosidase alfa, indicating improved potency of ATB200. Bars represent mean ± SEM. * p<0.05 vs. alglucosidase alfa in 2-sided t-test. (C) Paraffin sections of quadriceps and diaphragm from study described in panel B were also examined for glycogen accumulation by Periodic acid-Schiff’s reagent (PAS), which stains glycogen magenta. Consistent with the biochemical measurements, 20 mg/kg ATB200 appeared more effective in glycogen reduction compared to 20 mg/kg alglucosidase alfa in both tissues. Images of age-matched wild-type (WT) animal are shown in the insets. Each image is representative of 6-7 animals per group. Magnification is 200x. Untreated Alglucosidase alfa ATB200 ATB200 + AT2221 WT (C) IHC Examination of Fiber Type Response to ATB200 in Soleus * * result is consistent with the ATB200 + AT2221 compared quadricepsanddiaphragm observed superiority to alglucosidase alfa (B),tissueswith of in a * * * * * * * predominant type II fiber content. Asterisks mark all the type I fibers in a section, while the red triangles highlight * * * the type II fibers with significantly signals. Magnification = 400x. reduced LAMP1 * * * * * * * * * * * 2. The Pharmacological Chaperone AT2221 Increases the Stability and Exposure of ATB200 * ** * Untreated Alglucosidase alfa ATB200 ATB200 + AT2221 WT (A)Thermostability (A) The stability of ATB200 in acidic or neutral pH buffers was evaluated in a thermostability assay using SYPRO Orange. AT2221 stabilizes ATB200 at pH 7.4 in a concentration-dependent manner, to approaching the level seen at pH 5.2, a condition that mimics the acidic environment of the lysosome, as demonstrated by a nearly 10oC increase in the melting temperature (Tm) of ATB200. (B) Cynomolgus monkeys (2-3 years of age) were administered a single 2-hour IV infusion of 100 mg/kg ATB200 alone or with oral administration of 175 mg/kg AT2221 30 minutes earlier. Plasma samples were collected over the following 24 hours 100 80 60 Summary and Conclusions 7. 4 ( pH of bl ood ) 7. 4 + 10uM A T 222 1 7. 4 + 30uM A T 222 1 7. 4 + 100uM A T 222 1 5 . 2 ( p H o f l yso so m e s) 40 20 We have developed a novel rhGAA, ATB200, with a significantly higher M6P content compared to alglucosidase alfa, which resulted in greater enzyme uptake and glycogen reduction in disease-relevant tissues of Gaa KO mice, likely due to the improved endocytosis and lysosome targeting of the exogenous recombinant enzyme mediated by the binding of M6P to its receptor CI-MPR. 0 40 50 60 70 80 T em p ( ° C ) (B) Plasma Exposure in NHPs  More importantly, we showed that co-administration with the optimized pharmacological chaperone AT2221 leads to further improvement of the efficacy of ATB200, possibly via binding and stabilizing ATB200 in the blood, keeping the enzyme in a properly folded, active form that is more accessible for tissue uptake and lysosomal delivery. As a result, AT2221 improves the exposures of ATB200, broadens its bio-distribution, and achieves significantly greater glycogen reduction in disease-relevant cell types/tissues that have responded poorly to alglucosidase alfa, such as type II skeletal muscle fibers and skeletal muscles with a higher content of type II fibers. 1500 1250 m g / kg A T 222 1 and GAA activity was determined. Co-1000 administration resulted in an approximate 2-fold increase in ATB200 exposure (AUC) and half-life (T½), compared to administration of ATB200 alone. Each time point represents the mean ± SEM of 8 NHPs (4 males and 4 females)/group. 750 500 250  Taken together, these preclinical data highlight the efficacy of our proprietary rhGAA, ATB200, in mice when combined with a pharmacological chaperone using our proprietary CHART platform, and thus warrant further investigation. 0 0 2 4 6 8 10 12 16 20 24 28 T i m e ( h r ) WORLD LDN 2016 GAA Activity nmol/mL/hr ATB200 (µg/mL) Diaphragm Quadriceps GAA Activity nmol/mL/hr GAA Activity nmol/mL/hr + M6P + M6P Glycogen µg/mg protein GAA Activity nmol/mg protein/hr Untreated Alglucosidase alfa (20 mg/kg) ATB200 (5 mg/kg) ATB200 (10 mg/kg) ATB200 (20 mg/kg) Glycogen µg/mg protein Untreated Alglucosidase alfa (20 mg/kg) ATB200 (20 mg/kg) ATB200 (20 mg/kg) + AT221 (3 mg/kg) ATB200 (20 mg/kg) + AT221 (10 mg/kg) ATB200 (20 mg/kg) + AT2221 (30 mg/kg) Heart Quadriceps NOQ7.5.4D LAMP1 Diaphragm Heart Quadriceps 100 m g / kg A T B 2 00 + 175 1 0 0 m g/ k g A TB 2 0 0 a l one ATB200 + PK MeasurementATB200AT2221 AUC (µg . hr/mL)36476801 T½ (hr)1.23.2 Cmax (µg/mL)11551105 Tmax (hr)22.5 pH pH pH pH p H Test ConditionTm (oC) pH 7.456.2 pH 7.4 + 10 µm AT222161.6 pH 7.4 + 30 µm AT222162.9 pH 7.4 + 100 µm AT222166.0 pH 5.267.3 **** * ** *** ** * ** * * * ** * *** *** * * * ** * ** * ** ** ** ** * * * ** ** ** **** **** ** ** ** **** * **** *** * *** **** * * ** ** ** ** * * ** * ** *** * ** *** * **** * ** ** * * * * * ATB200 91 % 9% * * * * * # * 73 %Alglucosidase Alfa 27 %

SixmonthsofMigalastat Contentin Treatment AdultMale Reduces Patients PodocyteGlobotriaosylceramide withFabryDisease 1, 1, 2, Castelli J.2, Williams H.2, Mauer M.3 Najafian B. Sokolovskiy A. Barth J. 1.University of Washington, Seattle, WA; 2. Amicus Therapeutics; 3. University of Minnesota, Minneapolis Background Deficiency of -galactosidase-A in Fabry disease leads to accumulation of globotriaosylceramide (GL-3) inclusions in cells, causing organ damage. Progressive kidney failure is a major complication of Fabry disease. Results 240 220 200 180 160 140 120 100 80 60 40 20 0 Volume of GL-3 inclusions per podocyte Figure 4. (A) Plasma lyso-6000 Gb was reduced after 6 months 3 treatment. (B and C) The decrease in plasma lyso-Gb3 correlated with %change in GL-3 inclusion content of podocytes from baseline to 6 months. Podocytes are terminally differentiated cells with limited regeneration capacity. Recent studies suggest a key role for podocytes in Fabry nephropathy. In young Fabry patients, we showed that podocyte GL-3 accumulation occurs early, is progressive with age, and is associated with podocyte injury and proteinuria (Najafian et al. Kidney Int 2011). Reducing podocyte GL-3 burden may reduce progression of Fabry nephropathy. However, podocyte are far more resistant than other kidney cells to clear GL-3 following enzyme replacement therapy. 5000 4000 3000 BL 6M 2000 10% 0% -10% -20% -30% -40% -50% -60% -70% -80% 15% 10% 5% 0% -5% -10% -15% -20% -25% B C 1000 A 0 Baseline Migalastat (MIG) is an investigational pharmacologocal chaperone that stabilizes “amenable” mutant -gal-A and enhances its trafficking to lysosome. MIG reduced peritubular capillary endothelial cell GL-3 in 6 months (study 011). 6 Months BL 6M Figure 1. (A) Volume of GL-3 inclusions per podocyte was reduced from baseline (BL) to 6 months (6M) post-treatment. (B) A glomerulus from a Fabry patient at baseline; and (C) 6 months after migalastat. -30% -90% -80 -70 -60 -50 -40 -30 -20 -10 -80 -70 -60 -50 -40 -30 -20 -10 A B  Plasma Lyso-Gb3 (nmol/ml) Month 6 - Baseline  Plasma Lyso-Gb3 (nmol/ml) Month 6 - Baseline Podocyte Vol. vs. Podocyte Inclusion Vol. Changes After 6 Months of Migalastat Podocyte Inclusion Podocyte Volume Volume Density 15% 10% 5% 0% -5% -10% -15% -20% -25% -30% A 10% 0% -10% -20% -30% -40% -50% -60% -70% -80% -90% B 12000 0.50 0.45 10000 Materials and Methods 8 paired biopsies (baseline and 6 months post-migalastat) from male patients with “amenable” GLA mutations 4 paired biopsies R=0.98 P=0.00003 0.40 8000 0.35 6000 0.30 0.25 4000  Urine Protein/day (mg) Month 6 - Baseline  Urine Protein/day (mg) Month 6 - Baseline 0.20 2000 6 m 12 m 0.15 Figure 5. (A and B) There were statistical trends towards associations between 24-hr urine protein and % change in GL-3 inclusion content of podocytes. A B C Placebo Migalastat Open label extension 0 0.10 BL 6M BL 6M Figure 2. (A and B) GL-3 reduction in podocytes was closely paralleled by a reduction in podocyte volume. (C) Volume fraction of GL-3 inclusions in podocytes did not change during the 6 months migalastat treatment. Migalastat Migalastat Open label extension No statistically significant changes in eGFR, albuminuria or proteinuria over 6 months treatment. 4 paired biopsies Conclusions In patients with Fabry disease and “amenable” mutations, migalastat treatment led to a reduction in podocyte GL-3 within 6 months. This reduction correlated with proportional reduction in podocyte volume, leading to no significant change in GL-3 volume fraction in podocytes. 500 0 -500 -1000 -1500 -2000 -2500 -3000 -3500 800 P=0.01 700 P=0.002 600 500 R=0.82 p=0.02 The observed direct relationship between reduction in foot process width and GL-3 content in podocytes following 6 months of migalastat treatment is suggestive of reduced podocyte injury. 400 -4000 A B -120 -80 -40 0 40 80 120160 300  Foot Process Width (nm) Normal BL 6M * Asterisk indicates patients with no paired biopsies after 6 months migalastat (e.g. only BL and M6 on placebo) **Data includes all amenable male patients with ICF and paired assessable biopsies It will be crucial to confirm these findings in larger cohorts and examine if with longer treatment duration, podocytes further benefit from migalastat treatment. Figure 3. (A) Average foot process width in Fabry patients before or after 6 months treatment with migalastat was greater than values from 9 healthy control subjects . Foot process width was reduced in 5/7 and increased in 2/7 cases after 6 months MIG, but the change was not statistically significant. (B) The magnitude of foot process width reduction correlated with the magnitude of reduction in GL-3 inclusion volume in podocytes. Likewise, the magnitude of foot process width reduction correlated with the magnitude of reduction in GL-3 inclusion volume density in podocytes (R=0.82, p=0.02) and reduction in podocyte size (R=0.089, p=0.007). Biopsy structural parameters by electron microscopic stereology Vv(Inc/PC): Fraction of podocyte cytoplasm occupied by GL-3 inclusions V(PC): Average podocyte volume V(Inc/PC): Average volume of GL-3 inclusions per podocyte Future studies are needed to confirm if migalastat can prevent or ameliorate podocyte loss. This study shows that sensitive quantitative stereological methods can assess treatment efficacy in much shorter time periods (e.g. 6 months) than scoring methods. Other Parameters Age, eGFR, 24 hr urine protein, albumin/creatinine ratio, protein/creatinine ratio, plasma lyso Gb3, and peritubular capillary inclusion score (BLISS) This study was sponsored by a research contract from Amicus Therapeutics Study 011 Migalastat Baseline V(PC), µm3 V(Inc/PC), µm3 Foot Process Width (nm) V (Inc/PC), µm3 Vv(Inc/PC)  Vv(Inc/PC) Month 6 - Baseline % Vv(Inc/PC) Month 6 - Baseline % V(Inc/PC) Month 6 - Baseline % V(Inc/PC) Month 6 - Baseline Plasma Lyso-Gb3 (nmol/ml) P=NS P=NS P=0.004 R=0.69 P=0.06 R=0.64 P=0.09 Hypothesis MIG reduces GL-3 inclusion content in podocytes in patients with Fabry disease with amenable mutations. R=0.79 P=0.02 R=0.83 P=0.01 P=0.02 P=0.0004