Protocol AGAL-025-03: Multi-center, Open-label Study of the Safety and Efficacy of Fabrazyme^® in Patients with Fabry Disease that previously participated in the AGAL-008-00 Study

Fabrazyme^® (agalsidase beta)

Drug Name	Generic Name	Studied Indications or Disease	Approved U.S. Drug Label
Fabrazyme®	agalsidase beta	Fabry disease	Prescribing Info

NAME OF SPONSOR/COMPANY

Genzyme LLC, 500 Kendall Street, Cambridge, Massachusetts, 02142

Investigators and Study Center(s)

Twenty-five study centers in North America and Europe enrolled and treated patients in this extension study.

Publication (Reference)

Publications

Studied Period

15 January 2004 (first patient enrolled) to
12 September 2005 (last patient completed)

Phase of Development

Phase 4

Objectives

The primary objective of this study was to evaluate the stabilization of renal function with Fabrazyme by means of estimating the difference within the placebo patients’ inverse serum creatinine slope while in study AGAL-008-00 versus the inverse serum creatinine slope while in the open-label extension study (AGAL02503). (As described in more detail below, the primary efficacy analysis compared the placebo period slope with the Fabrazyme period slope; because some placebo patients were transitioned to Fabrazyme during AGAL-008-00, the Fabrazyme period includes renal data from the AGAL-008-00 study).

Methodology

This was a multi-center, multi-national, open-label extension trial designed to confirm the safety and effectiveness of Fabrazyme in patients with advanced Fabry disease who received Fabrazyme or placebo as treatment for this disease in the AGAL-008-00 trial. Only those patients who successfully completed the AGAL-008-00 trial were eligible for participation in the AGAL02503 study.

Patients in the AGAL02503 study were to receive Fabrazyme (1.0 mg/kg infusion) every 2 weeks for approximately 18 months. Patients who met specific criteria were given the option of receiving their treatment at home under the supervision of a home infusion nurse (versus at the site). At the time of each dose, the patients’ vital signs were measured. Serum creatinine was assessed every 4 weeks. Prior to each infusion, patients may have been pretreated for prophylactic management of infusion-associated reactions (IARs). Treatment-emergent adverse events (AEs) were also assessed, and concomitant medication use was reviewed. Patients went to the site every 3 months for a variety of efficacy and safety assessments. Final study evaluations were performed 2 weeks (± 7 days) after the patient’s last infusion. Patients were called approximately 1 month after the final infusion for assessment of AEs and review of concomitant medication use. Throughout the study, patients continued to receive their current standard of care.

Number of Patients (Planned and Analyzed)

PLANNED: Approximately 70 patients.

ANALYZED: A total of 67 patients entered the study. Of these, 28 patients had been randomized to placebo (Placebo/Fabrazyme) and 39 patients had been randomized to Fabrazyme (Fabrazyme/Fabrazyme) in the AGAL-008-00 study. All 67 patients were included in the Intent-to-Treat (ITT) population, and 66 were in the Per-Protocol (PP) population; 1 patient was excluded due to a major protocol violation (missed >20% of the scheduled Fabrazyme infusions). A total of 9 female patients were enrolled (4 Placebo/Fabrazyme, 5 Fabrazyme/Fabrazyme). By design, analysis of the primary efficacy endpoint only included the Placebo/Fabrazyme patients. An alternative post-hoc efficacy analysis was performed on the 67 patients in AGAL02503 comparing the 28 placebo patients with the 39 Fabrazyme patients as they were randomized in the original Phase 4 trial, AGAL-008-00. In this analysis, only the double-blind placebo data were used for the patients randomized to placebo. For the patients randomized to Fabrazyme, some analyses were conducted using only the double-blind Fabrazyme data, while other analyses were performed using all the data collected while the patient had been on Fabrazyme (i.e., double-blind and open-label data). This analysis was also performed by baseline estimated glomerular filtration rate (eGFR) subgroup (≤60 vs. >60 mL/min/1.73m²) for patients originally randomized to placebo (N=19 vs. N=9) and Fabrazyme (N=24 vs. N=15).

Diagnosis and Main Criteria for Inclusion

Patients who successfully completed the AGAL-008-00 study and provided informed consent for the extension study prior to any study-related procedures being performed were eligible to participate in this study. Patients who were unable to complete the double-blind study (AGAL-008-00), had undergone kidney transplantation or were currently on dialysis, had diabetes mellitus or presence of confounding renal disease, had a clinically significant organic disease or an unstable condition that in the opinion of the investigator precluded participation in the trial, or was unwilling to comply with the requirements of the protocol were not eligible for participation in this study.

Test Product, Dose, and Mode of Administration

Fabrazyme was supplied in 20 mL vials (35 mg/vial) as a lyophilized preparation. Each vial of Fabrazyme was reconstituted with 7.2 mL of Sterile Water for Injection. The appropriate amount of reconstituted Fabrazyme was further diluted with a 0.9% sodium chloride solution to a final total volume of 100 mL or 500 mL. The total volume (100 mL or 500 mL) was chosen at the investigator’s discretion and was not to be changed during the trial.

Patients received 1.0 mg/kg Fabrazyme intravenously (IV) every 2 weeks for the duration of the study.

At the discretion of the investigator (for the 500 mL infusions only), patients may have received IV infusions at a rate of no more than 0.25 mg/min (15 mg/hr) over approximately 4 to 6 hours. After patient’s tolerance to the infusion was well established (including open-label infusions in the AGAL-008-00 study), the infusion rate could have been increased in increments of 0.05 to 0.08 mg/min (increments 3 to 5 mg/hr) with each subsequent infusion. The total infusion time could not be less than approximately 90 minutes. If the investigator opted to infuse the patient according to the ramp scheme (for 100 mL infusions only), the total IV infusion time via IV pump was to be no less than 1.9 hours. With the ramp scheme, the Fabrazyme infusion began at a rate of 0.1 mg/kg/hr and was gradually increased by 0.2 mg/kg/hr every 15 minutes (if there were no signs of a reaction) until a maximum rate of 0.7 mg/kg/hr was reached.

Duration of Treatment

Patients were treated for approximately 18 months.

Reference Therapy, Dose and Mode of Administration

Criteria for Evaluation:
EFFICACY: The primary efficacy endpoint was the within-patient change in estimated inverse serum creatinine slope for the patients randomized to placebo in the AGAL-008-00 study (placebo period) versus the estimated inverse serum creatinine slope while in the open-label extension study, AGAL02503 (Fabrazyme period). Given that some placebo patients were transitioned to open-label Fabrazyme during AGAL-008-00, the Fabrazyme period includes serum creatinine values from the AGAL-008-00 study. As described in the Statistical Methods below, alternative post-hoc efficacy analyses were also performed comparing the placebo patients with the Fabrazyme patients as they were randomized in the original Phase 4 trial, AGAL-008-00. Additional post-hoc subgroup analyses were performed by baseline eGFR subgroup (>60 and ≤60 mL/min/1.73 m²).

Secondary efficacy endpoints included within-patient change in serum creatinine area under the curve (AUC), treatment group differences in slopes of inverse serum creatinine, within-patient change in eGFR slopes, treatment group differences in slopes of eGFR, change in plasma GL-3 levels, change in proteinuria and albuminuria, and change in serum creatinine and eGFR. Additional (“other”) endpoints related to renal function included inverse serum creatinine, estimated creatinine clearance, urinary protein, and urinary albumin.

SAFETY: Safety was assessed in terms of number and duration of infusions; treatment-emergent AEs and treatment-emergent serious adverse events (SAEs), including deaths; treatment-emergent related AEs that occurred on the day of infusion (IARs); IgG antibody formation to r-hαGAL; IgE antibody, serum tryptase, complement activation, and skin testing; laboratory evaluations; vital sign parameters; physical examinations; electrocardiogram (ECG) parameters; and review of concomitant medication use.

Statistical Methods

For the purposes of analysis, a patient’s treatment group, either placebo or Fabrazyme, was the group the patient was randomized to in the AGAL-008-00 double blind study.

EFFICACY: Efficacy analyses were performed on both the ITT and PP populations, and the analyses performed on the ITT population were considered primary. The ITT population included all randomized patients from AGAL-008-00 who provided informed consent for the AGAL02503 study and had at least 1 visit in the AGAL02503 study. The PP population was the ITT population excluding any patient who missed 20% or more of the scheduled Fabrazyme infusions.

For analysis of the predefined primary efficacy endpoint, 2 linear slopes were estimated for each placebo patient from his or her inverse serum creatinine values, a pre-slope and a post-slope. The pre-slope was based on values obtained during the placebo period (AGAL-008-00 double-blind study from baseline up to final visit/open-label date). The post-slope was based on values obtained during the Fabrazyme period (after the patient transitioned to Fabrazyme, either from the open-label date in AGAL-008-00, if applicable, or at entry of AGAL02503 to 18 months post-entry/dialysis date/transplant date). The within-patient difference between the placebo period slope and the Fabrazyme period slope was the primary endpoint of the study. If during the course of the AGAL-008-00 study, a patient began taking commercially available therapy or switched to open-label therapy, subsequent serum creatinine values were used to estimate the Fabrazyme period inverse serum creatinine slope as opposed to the placebo period inverse serum creatinine slope. The primary analysis was the within-placebo group comparison of the primary endpoint using a mixed effects model with a population level (fixed effect) intercept and slope and a subject level (random effect) intercept and slope, whereby patients were treated as random effects, the REML algorithm was used, and the numerator degrees of freedom were estimated using Kenward and Roger’s approximation.

The following longitudinal analyses of inverse serum creatinine were performed using a mixed model for the patients in AGAL02503 who had been randomized to placebo in AGAL-008-00: incorporated a quadratic time term for each treatment period; stratified patients by renal event/no renal event group; excluded patients with <6 observations in either the placebo or Fabrazyme treatment period; and weighted analysis wherein the weights were based on the number and logarithm of the number of serum creatinine assessments in each treatment period. Other mixed model analyses performed using the patients randomized to placebo were as follows: stratified by increase/no increase in serum creatinine group; stratified by various covariates at baseline and pre-Fabrazyme, including eGFR subgroups (≤60 and >60 mL/min/1.73m²); excluded patients with a confounding event; and adjusted for covariates at baseline and pre-Fabrazyme.

Alternative post-hoc mixed model analyses of inverse serum creatinine and eGFR were conducted to compare the randomized placebo patients versus the randomized Fabrazyme patients overall and by eGFR subgroups (≤60 and >60 mL/min/1.73m²). Likewise, mixed model analyses were conducted to compare the double-blind and open-label inverse serum creatinine and eGFR slopes for the patients randomized to Fabrazyme in AGAL-008-00.

The following secondary efficacy analyses were conducted using t-tests: AUC analysis of serum creatinine; change in plasma GL-3 summarized at 6-month time intervals; change in proteinuria and albuminuria summarized at 6-month time intervals; and change in serum creatinine and eGFR summarized at 6-month time intervals for all patients and by eGFR subgroups. These analyses evaluated within-patient changes from pre-Fabrazyme values at 6-month intervals. The primary and secondary efficacy parameters are presented graphically as appropriate.

Other renal assessments, including inverse serum creatinine, estimated creatinine clearance, urinary protein, and urinary albumin, were also performed but were not secondary efficacy endpoints. These analyses also evaluated within-patient changes from pre-Fabrazyme values at 6-month intervals.

SAFETY: Since the As-Treated and ITT populations were identical, the safety analyses were performed on the ITT population. A summary of the number of patients who received treatment (infusion) and summary statistics for duration of infusion are presented. Treatment-emergent AEs were defined as AEs that started during or after the first Fabrazyme infusion. Frequency counts of treatment-emergent AEs and the unique number of patients who had treatment-emergent AEs are presented by WHOART Body System and Preferred Term for both the AGAL-008-00 and AGAL02503 studies. A similar approach was used for treatment-emergent SAEs, except that only treatment-emergent events that occurred during the AGAL02503 study are summarized. Frequency and percentage of patients who had a treatment-emergent AE are also presented, stratified by relationship to treatment, severity, and IgG seroconversion status. By-patient listings are provided for all treatment-emergent AEs (including SAEs and deaths). Summaries of treatment-emergent IARs are presented by WHOART Body System and Preferred Term and stratified by titer group, 6-month time intervals, and prescribed infusion volume.

Changes in vital signs, physical examination findings, laboratory values, and ECGs are summarized at 6-month time periods. Listings of abnormal laboratory values and ECG findings reported as treatment-emergent AEs are provided. Frequency and percentage of patients who seroconverted and time to seroconversion in days from the first infusion of Fabrazyme are presented. Antibody titers for the IgG positive patients and a summary of time to peak titer are also provided. If a patient had IgE, serum tryptase, or complement activation testing performed, listings were prepared. Concomitant medications (including pain medications) are summarized .

Summary – Conclusions

The AGAL02503 study was designed as a supportive study to the double-blind, placebo-controlled AGAL-008-00 study. The AGAL-008-00 study showed a strong trend favoring Fabrazyme in the ITT population (Hazard Ratio=0.57, 95% CI = 0.27, 1.22; p=0.145) in the reduction of the occurrence of the primary endpoint (progression of renal, cardiac or cerebrovascular disease or death) when analyzed as planned in the protocol. When an appropriate adjustment for proteinuria was made, Fabrazyme-treated patients were 53% less likely to experience any clinical event in the ITT population (p=0.058) and 61% less likely in the PP population (p=0.034). Patients who completed the AGAL-008-00 study were allowed to enroll into the open-label extension study AGAL02503 for an additional 18 months of Fabrazyme treatment.

In the AGAL02503 study, most patients in the ITT population were Caucasian (n=59; 88%) and male (n=58; 87%). A total of 9 female patients were enrolled (4 Placebo/Fabrazyme, 5 Fabrazyme/Fabrazyme). Mean age for all patients at the time of enrollment in study AGAL02503 was 45.3 years. Overall, mean patient ages for all patients at onset of Fabry disease symptoms and diagnosis of Fabry disease were 13.6 and 33.6 years, respectively, and the mean ages at these time points were similar between groups.

It is important to note that placebo patients continued to have further progression of their chronic kidney disease during the AGAL-008-00 study and did not return to their baseline condition at the start of the Fabrazyme treatment period. Serum creatinine increased significantly by 0.2 mg/dL (p=0.0375), on average, from 1.6 mg/dL to 1.8 mg/dL and eGFR declined 2.7 mL/min/1.73 m², on average, from 53.5 to 50.8 mL/min/1.73 m², thereby indicating a further progression of their chronic kidney disease. For placebo/Fabrazyme patients, there was also a significant difference in age between baseline in AGAL-008-00 (43.5 years) and start of Fabrazyme treatment (45.1 years; difference 1.6 years; p<0.0001).

Efficacy

Primary Endpoint:
The predefined primary efficacy endpoint in this study was the difference in inverse serum creatinine slopes for the AGAL02503 Placebo/Fabrazyme patients (N=28) between the placebo and Fabrazyme treatment periods. In the ITT population, there was a statistically significant mean change (worsening) in slope in inverse serum creatinine over time between the placebo and Fabrazyme treatment periods (–0.029 dL/mg/year, p=0.0130 [mixed model]). However, this analysis is not statistically appropriate for determining the Fabrazyme treatment effect because the placebo treatment period preceded the Fabrazyme treatment period, thereby confounding the period effect (i.e., the fact that patients were treated with placebo earlier in their disease and then treated with Fabrazyme later in their disease) with the treatment effect. This analysis is essentially half of a cross-over design, wherein the other treatment sequence (Fabrazyme to Placebo) is missing. According to Levey, et al., “the rate of [renal] progression should be compared with the rate in a control group not receiving the intervention” (Levey et al., 1991, J Am Soc Nephrol), i.e., patients should not be used as their own control when assessing renal disease progression. Therefore, it is more statistically appropriate to compare placebo patients to Fabrazyme patients as they were randomized in the original Phase 4 study (AGAL-008-00).

Two analyses were therefore performed for the 67 patients participating in AGAL02503: a) comparing the double-blind data for placebo patients (i.e., only the data while they were on placebo) to double-blind data for patients treated with Fabrazyme, and b) comparing the double-blind data for placebo patients to all of the available data for patients treated with Fabrazyme. In the analysis based on randomized treatment in AGAL-008-00, no difference was observed in the slopes of inverse serum creatinine or eGFR between the placebo and Fabrazyme patients. However, a beneficial effect of Fabrazyme treatment over time was observed in the randomized analyses of inverse serum creatinine and eGFR slopes in the subgroup of patients with less severe kidney disease (baseline eGFR >60 mL/min/1.73 m²). In these patients, using all available data for patients treated with Fabrazyme, there was a statistically significant difference (in favor of Fabrazyme) between the placebo and Fabrazyme slopes for inverse serum creatinine (–0.069 dL/mg/year, p=0.0068) and eGFR (–6.787 mL/min/1.73 m²/year, p=0.0027). For the patients with baseline eGFR ≤60 ml/min/1.73m², there was no statistically significant difference in inverse serum creatinine and eGFR slopes between placebo- and Fabrazyme-treated patients. These results suggest that treatment with Fabrazyme early in the disease delays further progression of Fabry disease.

In Fabrazyme-treated patients with baseline eGFR >60 mL/min/1.73 m², the mean inverse serum creatinine slope was –0.005 dL/mg/year. In comparison, the mean inverse serum creatinine slope for the placebo patients was -0.074 dL/mg/year. Based on a serum creatinine value of 1.11 mg/dL, serum creatinine changed at an average rate of –0.10 mg/dL/year for the placebo group and at an average rate of –0.01 mg/dL/year for the Fabrazyme group, demonstrating stabilization of renal function.

In patients with less severe kidney disease at baseline, significantly less decline in kidney function over time (as measured by inverse serum creatinine and eGFR slopes) was observed in Fabrazyme-treated patients relative to placebo-treated patients despite the fact that the Fabrazyme/Fabrazyme patients had higher median proteinuria (0.6 vs. 0.2), a known risk factor for disease progression, at baseline than the Placebo/Fabrazyme patients.

Secondary Endpoints:
Inverse serum creatinine and eGFR slopes over time were also evaluated in Fabrazyme/Fabrazyme patients, comparing the double-blind versus open-label treatment periods. No significant change in mean slope over time was observed for either inverse serum creatinine (–0.006 dL/mg/year, p=0.5051) or eGFR
(–0.509 mL/min/1.73m²/year, p=0.5070). Since these patients were on Fabrazyme during both treatment periods, these findings were not unexpected.

In Placebo/Fabrazyme patients, the serum creatinine AUC (analysis comparing the weighted average change in serum creatinine over time in the placebo period versus the Fabrazyme period) increased significantly during the Fabrazyme period relative to the placebo period (mean AUCs of 0.30 vs. 0.02 mg/dL times (x) days, representing a mean increase of 0.28 mg/dL days; p=0.0003, 1-sample t-test). However, this within-patient analysis shares the same design and analysis flaw as the predefined primary efficacy analysis and are not appropriate.

Other analyses of secondary efficacy endpoints were performed using all patients and evaluated combined exposure to Fabrazyme during the double-blind and open-label extension studies. In all patients, significant mean (± SD) increases in serum creatinine were observed from pre-Fabrazyme levels at 6 months (0.1 ± 0.30 mg/dL, p=0.0094); 12 months (0.2 ± 0.50 mg/dL, p=0.0003); and 18 months (0.4 ± 0.78 mg/dL, p<0.0001) (1-sample t-test). In patients exposed to Fabrazyme for >18 months in the combined AGAL-008-00 and AGAL02503 studies, mean increases (range of 0.4 to 0.8 mg/dL) from pre-Fabrazyme serum creatinine values were observed at every 6-month evaluation from Month 24 (N=46) through Month 54 (N=3).

Small median increases from pre-Fabrazyme serum creatinine were also observed at every 6-month evaluation; however, the median increases were generally smaller than the mean increases. The median increase was 0.1 mg/dL at each 6-month evaluation through Month 30; 0.2 mg/dL at Months 36 and 42; and 0.5 and 0.9 mg/dL at Months 48 (N=10) and 54 (N=3), respectively, when the number of patients with data was smaller. The median serum creatinine curve remained fairly stable through 36 months in Fabrazyme/Fabrazyme patients, while rising in Placebo/Fabrazyme patients.

In all patients, significant mean (± SD) decreases in eGFR were observed from pre-Fabrazyme levels at 12 months (–2.7 ± 7.38 mL/min/1.73 m², p=0.0035) and 18 months (–5.3 ± 9.76 mL/min/1.73 m², p<0.0001) but not at 6 months (–0.5 ± 7.53 mL/min/1.73 m², p=0.5571) (1-sample t-test). In patients with >18 months of exposure to Fabrazyme in the combined AGAL-008-00 and AGAL02503 studies, mean reductions (range of –5.7 to –11.4 mL/min/1.73 m²) from pre-Fabrazyme eGFR were observed at every 6-month evaluation from Month 24 (N=46) through Month 54 (N=3). Median decreases from pre-Fabrazyme eGFR were also observed at every 6-month evaluation and were generally comparable with the mean decreases.

No significant mean change in serum creatinine was observed at 6, 12, and 18 months post-treatment in patients with baseline eGFR >60 mL/min/1.73 m² (mean changes of –0.0 to 0.0 mg/dL), indicating stability of renal function in this group; however, a statistically significant mean increase in serum creatinine was observed at these timepoints in patients with baseline eGFR ≤60 mL/min/1.73 m² (range of 0.2 to 0.6 mg/dL, p≤0.0043; 1-sample t-test). Similar findings were observed in the analyses of eGFR, whereby no significant mean change was observed at 6, 12, and 18 months post-treatment in patients with baseline eGFR >60 mL/min/1.73 m² (range of –1.3 to 1.9 mL/min/1.73 m²), while a statistically significant mean decrease was observed in patients with baseline eGFR ≤60 mL/min/1.73 m² (mean changes of –2.0 to –7.6 mL/min/1.73 m², p≤0.0476, 1-sample t-test). Similar results were observed in analyses of median changes in serum creatinine and eGFR by baseline eGFR subgroup.

In all patients, mean proteinuria generally remained slightly lower than pre-Fabrazyme levels over time; however, the mean (± SD) decrease from pre-Fabrazyme levels was statistically significant only at 6 months after the first Fabrazyme infusion (–0.2 ± 0.67, p=0.0254) (1-sample t-test). There was no clinically significant change in albuminuria over time.

In all patients, significant reductions from pre-Fabrazyme plasma GL-3 levels were observed at 6 months, 12 months, and 18 months after the first Fabrazyme infusion (p<0.0001 at all 3 time points; 1-sample t-test). The mean plasma GL-3 value decreased from 9.0 μg/mL at the pre-Fabrazyme assessment to 4.8 μg/mL at 6 months, 4.7 μg/mL at 12 months, and 4.6 μg/mL at 18 months after the first infusion. In patients with >18 months of exposure in the combined studies, mean plasma GL-3 levels also remained normal (≤7.03 μg/mL) from Month 24 (N=47) through Month 54 (N=2), ranging from 4.3 to 4.8 μg/mL

Safety Results:
The median time on Fabrazyme was 35.5 months (range of 14.3 to 51.6 months). Over the course of these studies, 28 of 67 (42%) patients received at least 1 full infusion of Fabrazyme in a 100 mL infusion volume in ≤2 hours. Sixty-one of the 67 (91%) patients received at least 1 full infusion in a 500 mL infusion volume; 48 of the 61 (79%) patients received at least 1 full infusion in ≤2.5 hours, and 41 of the 61 (67%) patients received at least 1 full infusion in ≤2 hours. Among all patients treated, 58 of 67 (87%) patients received the majority of their infusions in ≤2.5 hours, and 53 (79%) patients received the majority of their infusions in ≤2.0 hours. For all patients, the median duration of infusion decreased during the study.

Two (3%) patients died during the AGAL02503 study. One patient was hospitalized with neoplasm NOS, pneumonia, and chronic obstructive airways disease and died shortly thereafter. The immediate cause of death was unknown, although primary lung cancer or lymphoma was suspected. The other patient, who had a medical history significant for stroke and cardiovascular disease, died of an acute myocardial infarction. Both deaths were assessed by the investigator as unrelated to Fabrazyme treatment.
Twenty-four (36%) patients experienced a total of 50 treatment-emergent SAEs during the AGAL02503 study. The most common treatment-emergent SAE was uremia (3 patients, 4%). Most of the treatment-emergent SAEs were unrelated to Fabrazyme as assessed by the investigator. One patient had Fabrazyme-related treatment-emergent SAEs. This patient experienced 2 treatment-emergent SAEs considered to be definitely related to Fabrazyme (severe angioedema and a positive skin test that was mild in severity).

Two (3%) patients discontinued their participation in the study due to a treatment-emergent AE. One patient was withdrawn from the study due to a positive skin test (as reported above). Another patient was discontinued from the study following hospitalization due to a finger fracture and joint dislocation related to a fall. Per the investigator’s comments, the patient was not able to return to the site for subsequent visits and discontinued participation in the study following this event.

The most common treatment-emergent AEs that occurred during the AGAL-008-00 and AGAL02503 studies without regard to causality were rhinitis (55% of patients), coughing (49%), rigors (48%), fever (45%), headache (42%), upper respiratory tract infection (42%), nausea (39%), oedema dependent (39%), vomiting (39%), abdominal pain (34%), pain (34%), back pain (33%), myalgia (31%), and dizziness (30%). A similar profile was observed for Fabrazyme-related treatment-emergent AEs. Most of the treatment-emergent AEs were mild or moderate in severity. The incidence of treatment-emergent AEs was generally comparable among IgG positive and IgG negative patients.

Thirty-seven (55%) patients experienced a total of 287 treatment-emergent IARs during the AGAL-008-00 and AGAL02503 studies. The most common treatment-emergent IARs were rigors (37% of patients), fever (22%), hypertension (15%), temperature changed sensation (12%), and vomiting (10%). Two patients had a severe treatment-emergent IAR (angioedema or Fabry pain). The percentage of patients who experienced a treatment-emergent IAR was highest during the first 6 months of Fabrazyme therapy (42%); the number of reported treatment-emergent IARs was also highest during this timeframe (136/287 IARs). Treatment-emergent IAR incidence tended to be higher with the ramp infusion scheme, although caution should be used in interpreting these results due to the relatively small number of patients who received the 100 mL infusion.  A total of 10 Placebo/Fabrazyme patients received at least one 100 mL infusion.  In 8 of these patients, the date of their first 100 mL infusion was the same as the date of their first Fabrazyme infusion.  Six of these 8 patients experienced at least 1 treatment-emergent IAR which required an interruption or decrease of the infusion rate and/or corrective treatment to ameliorate the patients' symptoms.  Five of these 6 patients were able to tolerate an infusion rate of 0.3 mg/kg/hour. 

Most (79%) of the patients developed IgG antibodies (seroconverted) during the study. Most of these patients converted within 3 months of their first Fabrazyme infusion, and almost all seroconverted by Month 6. The majority of treatment-emergent IARs were reported within the first 6 months of exposure, showing a peak at Month 6. The percentage of patients who experienced a treatment-emergent IAR decreased thereafter.

Of the 53 patients who seroconverted, 6 (11%) were low responders, and 12 (23%) patients demonstrated a downward trend in antibody titer. Twenty-four (45%) patients demonstrated a plateau, and 11 (21%) patients tolerized. A higher percentage of males than females seroconverted during the study (83% versus 56%, respectively). Among the males who seroconverted, the majority were categorized as downward trend (12/48, 25%) or plateau (24/48, 50%). Of the 5 females who seroconverted, 1 was a low responder and 4 tolerized. For all patients, the mean time to seroconversion was 88.8 days. Similar results were observed for male patients (78.2 days), whereas the mean time to seroconversion was longer in females (189.8 days).

In the AGAL02503 study, 9 patients underwent IgE testing, 2 of whom tested positive. Skin testing was only performed on 1 patient, which yielded a positive result. This patient, who initially tested IgE positive (with a serum tryptase of 17.8 μg/mL), subsequently tested IgE negative (serum tryptase of 10.3 μg/mL).

Seven of the 9 patients who underwent complement activation testing secondary to the occurrence of treatment-emergent IARs had a positive test result. All 7 patients whose plasma was positive for complement activation were also IgG positive.

No new safety concerns were identified based on review of the laboratory, vital sign, physical examination, and ECG findings, as well as concomitant medication use.

The safety profile of Fabrazyme remains favorable.

Conclusion:

In summary, due to the sequential design of the study (active treatment period following the placebo period, in a single arm study for a progressive disease), the primary endpoint as initially chosen is not considered adequate to compare the effects of the two treatments. The sponsor believes that it is more statistically appropriate to compare placebo patients with Fabrazyme patients as they were randomized in the original Phase 4 trial, AGAL-008-00. Results of this analysis indicate there is no difference in the slopes of inverse serum creatinine between the Fabrazyme and placebo patients; however, after stratification based on kidney function, randomized analyses showed a beneficial effect of Fabrazyme for both inverse serum creatinine and eGFR slopes for patients with less severe kidney disease (baseline eGFR >60 mL/min/1.73 m²). Results of the secondary efficacy endpoint analyses also generally showed a slowing of Fabry disease progression over time in Fabrazyme-treated patients with less severe kidney disease. This suggests the need to treat patients early in their disease before severe kidney damage occurs. After the start of Fabrazyme treatment, mean plasma GL-3 decreased and then remained normal throughout the study. While this study did not show a benefit in some patients with moderately severe kidney disease in terms of kidney function, these patients may still benefit from Fabrazyme treatment in other organ systems, based upon the observed GL-3 clearance and the results of the original AGAL-008-00 study. Therefore, given the rarity of this disease and the small sample size, the beneficial effects of Fabrazyme seen in this trial are encouraging. This extension study demonstrates that Fabrazyme can be safely administered as an IV dose of 1.0 mg/kg every 2 weeks in patients with advanced Fabry disease.