Molecular Genetics and Metabolism 111 (2014) 499–506

Contents lists available at ScienceDirect

Molecular Genetics and Metabolism journal homepage: www.elsevier.com/locate/ymgme

Outcomes of patients treated through the Canadian Fabry disease initiative S.M. Sirrs a,⁎, D.G. Bichet b, R. Casey c, J.T.R. Clarke d, K. Lemoine e, S. Doucette f, M.L. West g, On behalf of the CFDI investigators a

Department of Medicine University of British Columbia, Canada Department of Medicine University of Montreal, Canada Department of Pediatrics University of Calgary, Canada d Department of Pediatrics, Hospital for Sick Children and Centre Hospitalier Universitaire de Sherbrooke, Canada e Department of Pediatrics, Capital District Health Authority, Canada f Department of Community Health and Epidemiology, Dalhousie University, Halifax, Nova Scotia, Canada g Department of Medicine Dalhousie University, Canada b c

a r t i c l e

i n f o

Article history: Received 12 December 2013 Received in revised form 28 January 2014 Accepted 28 January 2014 Available online 2 February 2014 Keywords: Fabry disease Outcomes Enzyme replacement therapy Agalsidase

a b s t r a c t Background: The Canadian Fabry disease initiative (CFDI) tracks outcomes of subjects with Fabry disease treated enzyme replacement therapy (ERT) given to subjects who meet evidence-based treatment guidelines and cardiovascular risk factor modification. Methods: We report 5 year follow-up data on 362 subjects for a composite endpoint (death, neurologic or cardiovascular events, development of end-stage renal disease or sustained increase in serum creatinine of 50% from baseline). Results: At enrollment, 86 subjects had previously received ERT (Cohort 1a) and 67 subjects were newly started (Cohort 1b) and randomized to agalsidase alfa or agalsidase beta. 209 subjects did not initially meet ERT criteria (Cohort 1c), 25 of whom met ERT criteria in follow-up and were moved to Cohort 1b (total N = 178 ERT treated subjects). Use of supportive therapies such as aspirin (78%), renin-angiotensin blockade (59%), and statins (55%) was common in ERT treated subjects. In Cohort 1a, 32 subjects met the composite endpoint with 8 deaths. In Cohort 1b, 16 subjects met the composite endpoint with 1 death. Cohort 1b had fewer clinical events than Cohort 1a (p = 0.039) suggesting that the treatment protocol was effective in targeting subjects at an earlier stage. 19.4% of Cohort 1b subjects on agalsidase alfa and 13.3% on agalsidase beta had a clinical event (p = 0.57). 10 Cohort 1c subjects had clinical events, none of which would have been prevented by earlier use of ERT. Conclusions: Cardiovascular risk factor modification and targeted use of ERT reduce the risk of adverse outcomes related to Fabry disease. © 2014 Elsevier Inc. All rights reserved.

1. Introduction Fabry disease is an X-linked disorder causing progressive renal, cardiac, and neurologic disease due to deficiency of alpha galactosidase A [1]. Treatment for Fabry disease using enzyme replacement therapy (ERT) has been studied in some randomized trials [2–5] and observational cohorts [6,7] but, despite use of ERT for more than a decade, there remains uncertainty about its efficacy as published data emphasize surrogate endpoints, have small patient numbers, short trial duration [8–10], and lack documentation about use of supportive therapies which may impact outcomes such as angiotensin converting enzyme inhibitors (ACEI) [11,12], acetylsalicylic acid (ASA), and statin agents. The Canadian Fabry disease initiative (CFDI; clinical trials registration number NCT00455104) is a project sponsored by the Canadian ⁎ Corresponding author at: Level 4-2775 Laurel Street, Vancouver, BC V5Z 1M9, Canada. Fax: +1 6048755967. E-mail address: [email protected] (S.M. Sirrs). 1096-7192/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ymgme.2014.01.014

Institutes of Health Research and the provincial governments of Canada in collaboration with Genzyme, a Sanofi Company, and Shire Human Genetic Therapies. We present data at 5 years of follow-up to answer two questions: 1. What are the clinical outcomes of subjects with Fabry disease who receive therapy targeting cardiovascular risk factors and ERT? 2. Do the Canadian evidence-based ERT guidelines exclude from therapy subjects who may derive benefit from ERT? 2. Materials and methods 2.1. Study design The study design has been previously reported [13]. Briefly, all Canadians between the ages of 5 and 85 years with confirmed Fabry disease are eligible for enrollment and are placed in one of three cohorts:

500

S.M. Sirrs et al. / Molecular Genetics and Metabolism 111 (2014) 499–506

Cohort 1a Subjects receiving ERT prior to the start of the CFDI and are maintained on the ERT formulation (agalsidase alfa or agalsidase beta) at entry. Cohort 1b Subjects naive to ERT who meet ERT criteria and undergo 1:1 randomization, stratified by gender, to either agalsidase alfa or agalsidase beta. Cohort 1c Subjects naive to ERT who do not meet ERT criteria OR who do not consent to ERT. At any point, should a subject in Cohort 1c develop an indication for ERT, they are offered randomization into Cohort 1b. ERT is given intravenously every other week using product monograph doses: agalsidase alfa 0.2 mg/kg and agalsidase beta 1 mg/kg. 2.2. Actions during shortages of agalsidase beta Due to a world-wide shortage, subjects in Cohort 1a receiving agalsidase beta were offered a switch to agalsidase alfa (Fig. 1) and all new Cohort 1b subjects were given agalsidase alfa. Randomization was resumed when agalsidase beta supplies again became available (January 2012). 2.3. Patient assessment, cardiovascular risk factor modification, and indications for ERT Subjects were seen at one of the five treatment centers across Canada at 6- (Cohorts 1a and 1b) to 12-month (Cohort 1c) intervals and underwent cardiac, renal, and neurologic testing as described previously (13). Use of medications such as ASA, statins, and reninangiotensin blockade was at the discretion of the treating clinician with recommendations for use of these medications included in Canadian treatment guidelines for Fabry disease (14). Most subjects were considered candidates for high risk blood pressure (b 130/80 mm Hg), and lipid (LDL b 2.0 mmol/L) targets and ASA prophylaxis [14]. Smoking cessation was strongly encouraged.

ERT is recommended if subjects meet treatment guidelines [14] which are reviewed annually. Briefly, subjects are eligible for ERT if they have nephropathy (declining glomerular filtration rate (GFR) or a GFR below normal or proteinuria), cardiac disease (ventricular hypertrophy or cardiac event), stroke or transient ischemic attack (TIA), or incapacitating neuropathic pain or gastrointestinal symptoms.

2.4. Definition of outcomes The primary endpoint is a composite clinical endpoint consisting of: renal events (development of end-stage renal disease OR decline in GFR of 50% or greater, sustained for N30 days and excluding other causes), cardiovascular events (pacemaker or other intracardiac device, coronary artery bypass grafting, valve replacement surgery, coronary angioplasty or stent, cardioversion, hospitalization or emergency room visit for unstable angina/acute coronary syndrome, myocardial infarction, congestive heart failure, tachy- or brady-arrhythmia, heart block, cardiac arrest), cerebrovascular event (TIA or stroke documented by a physician or acute hearing loss), or death. Data on renal function in subjects with end stage renal disease on entry into the CFDI are not included in the renal outcomes but all other data from these subjects are included in the analysis of the other components of the composite clinical endpoint.

2.5. Study oversight and role of the study sponsors The CFDI study was approved by the institutional research board at all sites. All subjects provided informed consent. Adverse events were reviewed by an independent data safety and monitoring board. Study outcomes were reviewed annually by financial sponsors and an independent scientific oversight committee. Data analysis, interpretation, and writing of this report were done solely by the authors without input from study sponsors.

Total number of subjects N=363

Incomplete follow-up (available data

included) 1a N=2 1b N=3 1c N=13

On ERT as of Oct 1 2006 Cohort 1a N=86

Met ERT treatment guidelines Cohort 1b N=67

On agalsidase alfa at study onset

On agalsidase beta at study onset

N=49

N=37

Switched to agalsidase alfa during drug shortages N= 36

Declined enrolment N=1

Total Cohort 1b N=92

Remained on agalsidase beta during

drug shortages N=1

Randomized to agalsidase alfa N=62

Did not meet ERT treatment guidelines on enrolment Cohort 1c N=209

Met ERT criteria at some point in follow up and moved to Cohort 1b

Did not meet ERT criteria during follow up and remained in Cohort 1c

N=25

N=184

Randomized to agalsidase beta N=30

Switched to agalsidase alfa during

Remained on agalsidase beta during

drug shortages N= 4

drug shortages N=26

Fig. 1. Subject enrollment, randomization, and actions during the drug shortages. A total of 363 subjects have been approached for enrollment in the CFDI and their distribution by cohort and drug assignment is shown here. Data on subjects who were lost to follow up is included to the time of the last follow-up visit. Reasons for loss of follow-up include: declined medical follow-up (N = 10), moved out of country (N = 2), enrolled in alternate clinical trial (N = 3), alternate diagnosis for symptoms (N = 1), noncompliance (N = 1), and unknown (N = 1).

S.M. Sirrs et al. / Molecular Genetics and Metabolism 111 (2014) 499–506

2.6. Statistical analysis The CFDI is not an intention-to-treat study as subjects who do not meet criteria for ERT at entry (Cohort 1c) may be switched if they later meet these criteria. One of the goals of the study is to see if subjects who do not receive ERT under CFDI guidelines are at higher risk of events and therefore clinical endpoints are described by cohort assignment at the time of an event. For example, if a subject in Cohort 1c had a clinical event that caused that person to be moved into Cohort 1b, that first clinical event is included within Cohort 1c. For subjects who switched cohorts, the length of time the subjects spent in each cohort is used to calculate median length of follow-up for that cohort. The data for clinical endpoints are presented as of January 31, 2013 and all clinical events are included. Data on subjects who were enrolled in the CFDI but did not complete follow-up are included to the time of loss to follow up (Fig. 1). Demographic data are summarized as means and standard deviation for continuous data and frequency with percentage for categorical data. Mixed models with random intercepts and slopes for each patient were used to calculate yearly change in eGFR and LV mass. Comparisons were made between Cohorts 1a and 1b using Fisher's exact and Student's t-tests where appropriate and

501

event-free survival was compared with a Kaplan–Meier curve using the log rank test. Cohort 1b currently has insufficient power to detect a difference between the two drugs (294 subjects within each arm of Cohort 1b would be required to detect a 10% difference in the rate of the composite clinical outcome), so data are combined for the comparison with Cohort 1a. A sensitivity analysis was performed in subjects who had 36-month outcome information. Unadjusted and adjusted Hazard Ratios (HR) were generated using Cox Proportional Hazards models. All model assumptions were met. Hazard ratios were adjusted for characteristics known to be related to the composite endpoint: age, gender, and proteinuria (b 0.5 and N0.5 g/day). All analyses were performed using SAS Version 9.3 (Cary, NC, USA). 3. Results 3.1. Comparison of ERT-treated cohorts We report data on 362 subjects 18 years of age and over, representing 91% of known adult cases of Fabry disease in Canada. A single subject under the age of 18 met a clinical endpoint (acute hearing loss) and is not included. Fig. 1 shows the treatment allocation of

Table 1 Baseline characteristics of ERT treated subjects enrolled in the CFDI. Parameter

N Age at baseline (years; mean (SD)) Time in CFDI (months) — median (IQR) Gender Females Males Patients with Nova Scotia mutation Sitting systolic blood pressure (mm Hg; mean (SD)) Sitting diastolic blood pressure (mmHg; mean (SD)) Serum creatinine (mg/dL; mean (SD))c MDRD GFR (ml/min/1.73 m2) Proteinuria (g/day) Left ventricular mass index (g/m2) LDL-cholesterol (mg/dL) Alpha galactosidase levels (leukocytes; nmol/h/mg protein) Chronic kidney disease stage 1 2 3 4 5 Health conditions present at baseline High blood pressure Pacemaker Dialysis Transplant TIA Stroke Concomitant medication use at baseline AceI/ARBd at baseline Statin at baseline ASA at baseline Indications for enrollment Renal Cardiac Neurologic Gastrointestinal Pain a b c d

Cohort 1ba

Cohort 1a Females

Males

1a combined

Females

Males

1b combined

23 52.6 (13) 67.6 (62.7– 69.5)

63 39.9 (13.2) 64.4 (61.6– 69)

86 43.3 (14.3) 64.4 (62.7– 69)

55 52.4 (14.7) 62.4 (46.7– 66.2)

37 40.5 (14.3) 51.8 (31.8– 65.3)

92 47.6 (15.6) 58.9 (43.5– 65.7)

23 (100%) 0 (0%) 5/19 (26.3%)b 121.4 (19.8) 77.4 (10.4) 0.87 (0.21) 78.8 (25.4) 0.5 (0.8) 112.3 (42.7) 92.6 (23.1) 9 (24.2)

0 (0%) 63 (100%) 15/56 (26.8%) 118.6 (14.1) 73.8 (11.4) 1.21 (0.7) 90.2 (44.6) 0.5 (0.7) 141.6 (58.9) 92.6 (30.9) 3.2 (5.1)

23 (26.7) 63 (73.3) 20/75 (26.7%) 119.3 (15.7) 74.7 (11.2) 1.12 (0.62) 87 (40.2) 0.5 (0.7) 134.1 (56.5) 92.6 (27.0) 7.9 (15.3)

55 (100%) 0 (0%) 16/52 (30.8%) 127.2 (17) 77.6 (11.6) 0.85 (0.2) 79.5 (23.2) 0.5 (0.8) 113.5 (34.5) 111.9 (34.7) 22.8 (21)

0 (0%) 37 (100%) 5/34 (14.7%) 127.2 (12.5) 79.8 (9.4) 1.26 (0.49) 78.3 (29.7) 1.1 (1.3) 137.6 (63.3) 84.9 (27.0) 3.3 (5.3)

55 (59.8) 37 (40.2) 21/86 (24.4%) 127.2 (15.4) 78.4 (10.8) 0.99 (0.38) 79.1 (25.4) 0.7 (1) 123.2 (49.3) 104.2 (34.7) 13.8 (18.5)

p-Value (1a vs. 1b) 0.055 b0.0001

b0.0001 0.86 0.0014 0.031 0.12 0.15 0.18 0.19 0.036 0.043 0.008 (for the trend)

6 (27.3%) 8 (36.4%) 6 (27.3%) 0 (0%) 2 (9.1%)

23 (37.1%) 14 (22.6%) 10 (16.1%) 2 (3.2%) 13 (21%)

29 (34.5) 22 (26.2) 16 (19.1) 2 (2.4) 15 (17.9)

16 (30.2%) 26 (49.1%) 11 (20.8%) 0 (0%) 0 (0%)

11 (32.4%) 8 (23.5%) 8 (23.5%) 0 (0%) 7 (20.6%)

27 (31) 34 (39.1) 19 (21.8) 0 (0) 7 (8.1)

11 (50%) 3 (13.6%) 1 (4.4%) 1 (4.4%) 9 (39.1%) 6 (26.1%)

22 (34.9%) 12 (19.1%) 9 (14.3%) 10 (15.9%) 12 (19.1%) 9 (14.3%)

33 (38.8) 15 (17.7) 10 (11.6) 11 (12.8) 21 (24.4) 15 (17.4)

32 (58.2%) 6 (10.9%) 1 (1.8%) 0 (0%) 9 (16.4%) 9 (16.4%)

22 (59.5%) 0 (0%) 5 (13.5%) 2 (5.4%) 3 (8.1%) 4 (10.8%)

54 (58.7) 6 (6.5) 6 (6.5) 2 (2.2) 12 (13) 13 (14.1)

0.01 0.034 0.30 0.008 0.056 0.68

15 (68.2%) 13 (59.1%) 17 (77.3%)

35 (56.5%) 22 (35.5%) 40 (64.5%)

50 (59.5) 35 (41.7) 57 (67.9)

31 (56.4%) 22 (40%) 39 (70.9%)

19 (54.3%) 13 (37.1%) 22 (62.9%)

50 (55.6) 35 (38.9) 61 (67.8)

0.65 0.76 1

10 (43.5%) 13 (56.5%) 2 (8.7%) 6 (26.1%) 5 (21.7%)

29 (46%) 18 (28.6%) 8 (12.7%) 26 (41.3%) 33 (52.4%)

39 (45.4) 31 (36.1) 10 (11.6) 32 (37.2) 38 (44.2)

10 (18.2%) 43 (78.2%) 9 (16.4%) 4 (7.3%) 2 (3.6%)

21 (56.8%) 18 (48.7%) 5 (13.5%) 3 (8.1%) 2 (5.4%)

31 (33.7) 61 (66.3) 14 (15.2) 7 (7.6) 4 (4.4)

0.13 b0.0001 0.52 b0.0001 b0.0001

Cohort assignment reflects the cohort at the time of data analysis January 31, 2013; some subjects crossed over from Cohort 1c to 1b and are included in Cohort 1b (see text). Percentage values where presented are based on total number of subjects for whom information on a given datapoint is available. Excludes subjects with end stage renal disease on study entry. ACEI angiotensin converting enzyme inhibitor ARB angiotensin receptor blocker.

502

Table 2 Clinical endpoints and risk factor modification of subjects in the ERT cohorts of the CFDI. Cohort 1a

a b c

Males

1a combined

Females

Males

1b combined

23

63

86

55

37

92

3 (13%) 7 (30.4%) 2 (8.7%) 0/21 (0%) 8 (34.8%) 1484.63 15.5 (7.8–35.8)

5 (7.9%) 16 (25.4%) 6 (9.5%) 3/51 (5.9%) 24 (38.1%) 4013.9 13.9 (9.4–21.8)

8 (9.3%) 23 (26.7%) 8 (9.3%) 3/72 (4.2%) 32 (37.2%) 5498.53 14.3 (10.1–20.9)

1 (1.8%) 5 (9.1%) 2 (3.6%) 0/55 (0%) 7 (12.7%) 2998.93 35.7 (17.3–88.8)

0 (0%) 3 (8.1%) 5 (13.5%) 1/30(3.3%) 9 (24.3%) 1699.93 15.7 (8.3–34.4)

1 (1.1%) 8 (8.7%) 7 (7.6%) 1/85(1.2%) 16 (17.4%) 4698.87 24.5 (15.1–42.8)

0.015 0.0026 0.79 0.33 0.0039

0 (0%) 3 (13%) 1 (4.4%) 0/21 (0%)

2 (3.2%) 11 (17.5%) 4 (6.4%) 0/51 (0%)

2 (2.3%) 14 (16.3%) 5 (5.8%) 0/72 (0%)

0 (0%) 2 (3.6%) 2 (3.6%) 0/55 (0%)

0 (0%) 2 (5.4%) 4 (10.8%) 0/33 (0%)

0 (0%) 4 (4.4%) 6 (6.5%) 0/85 (0%)

0.23 0.012 1 1

−0.43 (0.86) 2.05 (2.09)

−2.45 (1.23) 1.72 (1.82)

−1.87 (0.88) 1.81 (1.4)

0.21 (0.65) 1.43 (2.05)

−1.97 (2.5) −1 (2.03)

−0.25 (0.67) 0.77 (1.58)

0.18 0.63

−1.11 (3.68) 2.06 (8.99)

−4.49 (0.57) 0.22 (3.12)

−3.73 (1.04) 0.44 (3.03)

−0.1 (0.6) −0.49 (2.91)

−3.74 (1.07) 3.02 (3.16)

−1.93 (0.68) 1.12 (2.05)

0.19 0.83

96.5 (32.4) 6 (27.3%) 16 (72.7%) 121.95 (17.14) 72.73 (8.08) 13 (59.1%) 16 (72.7%) 19 (86.4%)

80.7 (29.7) 31 (51.7%) 32 (51.6%) 117.45 (14.59) 72.45 (10.81) 42 (67.7%) 34 (54.8%) 46 (74.2%)

84.9 (31.3) 37 (45.1%) 48 (57.1%) 118.63 (15.32) 72.52 (10.12) [ 55 (65.5%) 50 (59.5%) 65 (77.4%)

91.5 (32.4) 16 (30.2%) 32 (58.2%) 121.05 (14.88) 73.6 (9.22) 35 (63.6%) 37 (67.3%) 47 (85.5%)

88.4 (28.1) 12 (35.3%) 16 (45.7%) 125.26 (13.89) 77.97 (8.58) 16 (47.1%) 16 (45.7%) 24 (68.6%)

90.3 (30.5) 28 (32.2%) 48 (53.3%) 122.66 (14.58) 75.27 (9.18) 51 (57.3%) 53 (58.9%) 71 (78.9%)

0.26 0.11 0.65 0.08 0.06 0.28 1 0.86

Excludes patients with end stage renal disease on study entry. Mean (standard error). ACEI angiotensin converting enzyme inhibitor ARB angiotensin receptor blocker.

p-Value (1a vs. 1b)

0.08

S.M. Sirrs et al. / Molecular Genetics and Metabolism 111 (2014) 499–506

N Clinical endpoints At any time Death Cardiac event Acute neurological event Need for renal replacement therapya Composite (any of the 4 above) Person time at risk (years) Patient years per composite clinical event (95% confidence interval) By 3 years Death Cardiac event Acute neurological event Need for renal replacement therapya Annual change from baseline: proteinuria at entry ≤0.5 g/dayb Change in eGFR (ml/min/1.73 m2/year) Change in left ventricular mass index (g/1.73 m2/year) Annual from baseline: proteinuria at entry N0.5 g/dayb Change in eGFR (ml/min/1.73 m2/year) Change in left ventricular mass index (g/1.73 m2/year) Risk factor modification (as of January 31, 2013) LDL-cholesterol (mean+/−SD; mg/dL) % of cohort with LDL-cholesterol b77.2 mg/dL % of cohort on statin Systolic BP (mm Hg; mean (SD)) Diastolic BP (mm Hg; mean (SD)) % of cohort with BP b130/80 % of cohort on ACEI/ARBc % of cohort on ASA

Cohort 1b

Females

S.M. Sirrs et al. / Molecular Genetics and Metabolism 111 (2014) 499–506

subjects, crossovers of drug or cohort, and subjects lost to follow up. The allocation of subjects within Cohort 1b to the two different ERT formulations is unequal due to shortage of agalsidase beta. Table 1 shows the characteristics of subjects in the ERT treated cohorts as of January 31, 2013. Information on genotype is available for 92% of subjects and the majority of these mutations are known to be associated with a classical Fabry phenotype [13]. Information on mutations other than the Nova Scotia mutation has been previously reported [13] and is not included here. This mutation (c.427G N C) may interfere with substrate binding to the active site [14] and is associated with a classical Fabry phenotype [15–17]. Subjects in Cohort 1a had received ERT prior to the start of the CFDI for the mean of 34+/− 24(males) and 35 +/−25 (females) months. Subjects in Cohort 1a were more often male, had more advanced renal disease, and more comorbidities like previous transplantation, or pacemaker insertion than those subjects in Cohort 1b. These are expected results as ERT in Canada prior to the start of the CFDI was available only as part of clinical trials in males or through a compassionate use program for subjects with advanced disease. There are differences in the indications to start ERT in Cohorts 1a and 1b reflecting entry criteria for the clinical trials involving Cohort 1a subjects prior to the CFDI. For example, one of the trials of ERT prior to the CFDI had pain as the primary outcome so all subjects enrolled in that trial had pain as an indication for ERT but pain as the sole indication to start ERT was uncommon after the start of the CFDI as it is amenable to other therapies [18].

a

503

Clinical endpoints for the ERT treated cohorts are shown in Table 2 combining data from both arms of Cohort 1b for this comparison. The composite clinical outcome developed in 48 ERT-treated subjects (Fig. 2a–c) with the individual events contributing to the composite outcomes shown in Table 2. In Cohort 1a (median follow-up 64 months), 8 subjects died (9.3%) and 26.7% of the cohort had cardiac events, 9.3% neurologic events, and 4.2% required renal replacement therapy. In Cohort 1b (median follow-up 59 months), 1 subject died (1.1%), 8.7% of the cohort had cardiac events, 7.8% neurologic events, and 1.2% required renal replacement therapy. Subjects in Cohort 1b were less likely to achieve the composite clinical outcome (HR 1a versus 1b 1.93 p = 0.032) than subjects in Cohort 1a, but this difference was no longer significant when controlled for age, gender, and baseline proteinuria (adjusted HR 1a versus 1b 1.84 p = 0.09) suggesting that the CFDI monitoring protocol helped to target intervention with ERT to subjects at an earlier stage of disease. As expected, the rate of clinical events varies with gender, but was low overall even in males with 1 male patient meeting the composite endpoint per 13.9 (Cohort 1a) to 15.7 (Cohort 1b) patient years of follow-up. Cohorts 1a and 1b have a different gender distribution as expected by the study design. To ensure that the differences in outcomes between Cohort 1a and Cohort 1b do not simply reflect this gender difference, Tables 1 and 2 also include comparative data on demographic characteristics clinical outcomes in Cohorts 1a and 1b broken down by gender

b 100 80 70 60 50 Log rank P-value = 0.03

40 30 20

Cohort 1a Cohort 1b

90

Event-Free Rate (%)

90

Event-Free Rate (%)

100

Cohort 1a Cohort 1b

80 70 60 50 Log rank P-value = 0.56

40 30 20 10

10

0

0 0

10

20

30

40

50

60

70

80

0

10

20

Time (Months) 86 92

Cohort 1b

78 85

73 73

70 69

62 58

56 47

45 36

8 6

Cohort 1b

d 100

Cohort 1a Cohort 1b

60

70

80

56 31

52 24

49 23

45 18

40 15

32 12

6 2

100

Agalsidase beta Agalsidase alfa

80

80 70 60 50 Log rank P-value = 0.08

40 30 20

70 60 50 40

Log rank P-value = 0.52

30 20 10

10

0

0 0

10

20

30

40

50

60

70

80

0

10

20

Time (Months)

30

40

50

60

70

14 22

2 4

80

Time (Months)

Number at risk

Cohort 1b

50

90

Event-Free Rate (%)

Event-Free Rate (%)

90

63 37

Cohort 1a

c

Cohort 1a

40

# at risk

Number at risk

Cohort 1a

30

Time (Months)

Number at risk

23 55

22 54

21 49

21 46

17 40

16 32

13 24

2 4

Agalsidasebeta Agalsidasealfa

30 62

26 59

23 50

22 47

20 38

17 30

Fig. 2. Kaplan–Meier plot demonstrating the likelihood of obtaining the composite clinical endpoint of ERT-treated subjects in the CFDI. a. Kaplan–Meier plot showing the composite clinical endpoints of subjects in Cohorts 1a and 1b. b. Kaplan–Meier plot showing the composite clinical endpoints of male subjects in Cohorts 1a and 1b. c. Kaplan–Meier plot showing the composite clinical endpoints of female subjects in Cohorts 1a and 1b. d. Kaplan–Meier plot showing the composite clinical endpoints of subjects on agalsidase alfa and agalsidase beta in Cohort 1b.

504

S.M. Sirrs et al. / Molecular Genetics and Metabolism 111 (2014) 499–506

although this analysis is limited in statistical power. The comparison of cohorts 1a and 1b broken down by gender did not show significant differences in age, renal function, or left ventricular mass at baseline. There were slight differences in LDL-cholesterol (higher in 1b females than 1a females) and diastolic blood pressure (higher in 1b males than 1a males) which likely reflected the fact that all of the Cohort 1a patients were under specialty care for Fabry disease at the time of entry whereas some of the Cohort 1b patients were newly diagnosed at the time of entry and may not yet have received intensive risk factor modifications. The rates of all clinical events stratified by gender, except stroke, were numerically lower in Cohort 1b than those in Cohort 1a although the only one of these that reached statistical significance was the rate of cardiovascular events (1a females 30.4%;1b females 9.1% p = 0.03; 1a males 25.4%;1b males 8.1% p = 0.04).

3.2. Comparison of agalsidase alfa and agalsidase beta We observed no statistical difference in endpoints between the agalsidase alfa and agalsidase beta arms of Cohort 1b (HR alfa versus beta 1.29; p = 0.67) but power is limited. The separation between the two arms is shown in Fig. 2d. Within Cohort 1b, 19.4% of the subjects on agalsidase alfa and 13.3% of the subjects on agalsidase beta met the composite clinical endpoint (p = 0.57). Data on clinical and antibody responses in those patients who were switched from agalsidase beta to agalsidase alfa during the drug shortages will be reported separately.

3.3. Clinical outcomes in subjects who did not initially meet criteria for ERT (Cohort 1c) At enrollment, 209 subjects (median age 35.5 years, 40 males and 169 females, median time in cohort 52.2 months) did not initially meet criteria or declined ERT (6 females and 1 male declined ERT and a second male was randomized to cohort 1b but died before ERT could be started and is analyzed as part of Cohort 1c; Table 3). Cohort 1c is predominantly female as expected by study design. With follow-up, 25 subjects developed indications for ERT and were moved to Cohort 1b (demographics included in Table 1). Ten subjects who did not receive ERT met the composite clinical endpoint (Table 4). Of these 10 subjects, 3 were eligible for but declined ERT and 1 subject was enrolled in Cohort 1b but died before ERT was started so is analyzed as part of the non-ERT cohort. Of the remaining 6 subjects, 4 (3 with stroke and 1 with myocardial infarction at the age of 73 years) developed complications which are not thought to respond to ERT [19–24]. The final 2 subjects developed arrhythmias within 2 months of enrollment so it is unlikely that the use of ERT for this short time frame would have prevented these events. Thus we did not identify adverse outcomes in Cohort 1c which the earlier use of ERT might have prevented.

3.4. Cardiovascular risk factor modification The use of ancillary medications was high in ERT treated subjects (Table 2) with 78% on ASA, 59% on ACEI, and 53–57% on statins. Mean

Table 3 Characteristics of subjects not treated initially with ERT (Cohort 1c). Parametera

N Age at baseline (years; mean (SD)) Time in CFDI (months) — median (IQR) Gender Females Males Patients with Nova Scotia mutationb Sitting systolic blood pressure (mm Hg; mean (SD)) Sitting diastolic blood pressure (mm Hg; mean (SD)) Serum creatinine (mg/dL; mean (SD))c MDRD GFR (ml/min/1.73 m2) Proteinuria (g/day) Left ventricular mass index (g/m2) LDL-cholesterol (mg/dL) Alpha galactosidase levels (leukocytes; nmol/h/mg protein) Chronic kidney disease stage 1 2 3 4 5 Health conditions present at baseline High blood pressure Pacemaker Dialysis Transplant TIA Stroke Concomitant medication use at baseline ACEI/ARBd at baseline Statin at baseline ASA at baseline a

Cohort 1c Females

Males

1C combined

169 37.6 (16.8) 53.3 (31.5–63.2)

40 26.7 (16.6) 48.6 (30.8–62.4)

209 35.5 (17.3) 52.2 (31.5–63.1)

169 0 52/169 (32.7%) 118.5 (15.8) 74.3 (10.4) 0.74 (0.13) 95.5 (20.5) 0.2 (0.5) 84.1 (24.6) 100.5 (27)

0 40 8/40 (20.0%) 124.6 (11.5) 76.5 (10.1) 0.98 (0.25) 100.1 (27.3) 0.3 (0.4) 95.8 (20.2) 96.7 (38.7)

169 (80.9%) 40 (19.1%) 60/199 (30.2%) 119.5 (15.3) 74.6 (10.4) 0.78 (0.17) 96.1 (21.6) 0.2 (0.5) 85.9 (24.2) 100.3 (30.9) 19.6 (25)

80 (55.2%) 62 (42.8%) 3 (2.1%) 0 (0%) 0 (0%)

16 (64%) 6 (24%) 3 (12%) 0 (0%) 0 (0%)

96 (56.5%) 68 (40%) 6 (3.5%) 0 (0%) 0 (0%)

34 (20.4%) 1 (0.6%) 1 (0.6%) 0 (0%) 7 (4.2%) 6 (3.6%)

7 (17.5%) 1 (2.5%) 1 (2.5%) 1 (2.5%) 2 (5%) 1 (2.5%)

41 (19.8%) 2 (1%) 2 (1%) 1 (0.5%) 9 (4.4%) 7 (3.4%)

28 (18.3%) 25 (16.3%) 40 (26.1%)

8 (28.6%) 3 (10.7%) 10 (35.7%)

36 (19.9%) 28 (15.5%) 50 (27.6%)

Percentage values where presented are based on total number of subjects for whom information on a given datapoint is available. The prevalence of the Nova Scotia mutation in Cohort 1c is not significantly different from the prevalence in the ERT treated cohorts (P values for comparison: 1c vs 1a p = 0.66; 1c vs 1b p = 0.39; 1c vs 1a + 1b p = 0.35). c Excludes subjects with end stage renal disease on study entry. d ACEI angiotensin converting enzyme inhibitor ARB angiotensin receptor blocker. b

S.M. Sirrs et al. / Molecular Genetics and Metabolism 111 (2014) 499–506

505

Table 4 Subjects who were not treated initially with ERT (Cohort 1c) who progressed to a composite clinical endpoint. Age at enrollment

Age at time of event

Gender

Event

47

47

Male

Death

31 70 68 51 71 35 69 34 76

31 70 69 53 73 35 71 34 78

Female Female Female Female Female Female Male Male Female

Pacemaker or ICDa Pacemaker or ICD Pacemaker or ICD Cardioversion Myocardial infarction Stroke Stroke Stroke Stroke

a

Time between enrollment and event (months) 1

2 1 10 26 25 11 31 5 21

Comment

Subject was enrolled in cohort 1b but died before started on ERT so was analyzed as part of non-ERT treated cohort Event caused randomization to Cohort 1b Event caused randomization to Cohort 1b Declined ERT Declined ERT Event caused randomization to Cohort 1b Event caused randomization to Cohort 1b Declined ERT Event caused randomization to Cohort 1b Event caused randomization to Cohort 1b

ICD implantable cardiac defibrillator.

blood pressures met the high risk target range of b130/80. Lipid control did not meet targets in that only 45% of subjects in Cohort 1a and 32% of subjects in Cohort 1b had LDL-cholesterol values in the target range (b2 mmol/L). Data are available for key parameters like eGFR (90% of subjects), proteinuria (80% of subjects), echocardiogram (81%), and brain imaging (74%) reflecting a high rate of compliance with the monitoring protocol. 4. Discussion We present data from a large prospective treatment cohort which documents clinical outcomes in subjects followed using a 4 point protocol which includes use of ERT in subjects who meet evidence-based criteria, cardiovascular risk factor modification, centralization of care to experienced centers, and close follow-up of patients not on ERT to allow the early identification of disease progression which might be amenable to the use of ERT. The indications for ERT used in the CFDI require that early disease manifestations be present, similar to protocols used in the United Kingdom [25] and Australia [26]. Despite this restriction on the use of ERT, the outcomes in our ERT-treated subjects compare favorably with published data and there are considerable cost savings from this approach. It is difficult to compare data on clinical outcomes from the CFDI with that from the large international registries [6,7] as the voluntary nature of those registries could lead to the inclusion of more severely affected subjects [13]. Also, some retrospective cohort studies [5,27] encompass decades of follow-up so outcomes may be affected by advances in supportive care. Two recent large cohort studies [28,29] with demographic characteristics including age and duration of follow-up similar to Cohort 1b described major clinical events (stroke, cardiac, end stage renal disease, sudden death) in 37–45% of ERT treated male subjects with a total of 8 deaths amongst the 58 ERT-treated males described in those two studies. By contrast, none of CFDI Cohort 1b males have died and 24.3% have had a major event. While the extent of impact of ERT on Fabry related outcomes has not previously been well defined leading some authors to question the utility of this costly therapy [12,30], our data suggest that the protocol used by the CFDI which includes targeted ERT and cardiovascular risk factor modification is beneficial in reducing Fabry-related events. The results of ongoing primary prevention trials may further inform clinicians about the optimal timing for initiation of ERT in the future. Stroke continues to be common in Fabry patients despite ERT. We noticed no difference in acute neurological event rates in our two ERT treated cohorts, suggesting that intervention earlier in the course of the disease with ERT (Cohort 1b) does not reduce stroke risk. Our findings are consistent with those of other studies [15,18,29] that existing therapies including ERT do not prevent stroke in Fabry disease.

The CFDI is an analysis of the effects of a 4 point protocol which is a “strategy” of care. We cannot separate out the impact of risk factor modification and close follow-up from that of ERT on patient outcomes in Cohort 1a and 1b subjects in the CFDI. Although the literature is conflicting on attributable risks related to traditional cardiovascular risk factors [31], patients with Fabry disease have abnormalities of vascular function [32] which can be modified with drugs like ACEI and statins [33,34]. Thus, the use of these agents may provide benefit which is not quantified by measurement of blood pressure or cholesterol levels. However, although the contributions of each part of the CFDI protocol cannot be separated, it is the cumulative impact of this approach, similar to the multifactorial approach recommended for subjects with other complex health conditions like diabetes mellitus [35], which has resulted in the outcomes reported here. One strength of the CFDI is that ascertainment bias is minimal with 91% of known Canadian subjects with Fabry disease enrolled. A second strength is the high rate of collection of major data points which exceeds that found in the larger international registries [36]. The CFDI was initiated as a means of providing treatment to Canadian patients and decisions have been made (such as during the drug shortages where patients randomized to agalsidase beta were given agalsidase alfa) which may not be desirable from a research standpoint but which were necessary to ensure ongoing care of Canadians with Fabry disease. As other clinicians around the world have been grappling with the same issues, we do not view this as a limitation of our data but would emphasize that our data show what happens to patients treated in a “realworld” environment. The CFDI has limitations. The CFDI does not have a placebo arm as all subjects who met treatment criteria are given ERT. Cohort 1b is the longest running head-to-head trial of commercially available ERT products but actions required during the drug shortages and lower than expected event rates have restricted the power of this cohort. The comparison of these two drugs within Cohort 1b is ongoing. Cohort 1b was randomized for gender only and not for other variables known to affect prognosis [4,36] as most known Canadian subjects who met treatment criteria were already on therapy (Cohort 1a) at the start of the CFDI and not available for randomization. Finally, while many different mutations are found within the Canadian cohort, there is a founder effect of the Nova Scotia mutation. However, as the Nova Scotia mutation is associated with a classical disease phenotype [15–17] and the prevalence of the Nova Scotia mutations was equal across the cohorts, we do not feel that the Nova Scotia founder effect has a major impact on the generalizability of our findings. 5. Conclusions A protocol which includes cardiovascular risk factor modification, centralization of care, targeted use of ERT, and close follow-up of

506

S.M. Sirrs et al. / Molecular Genetics and Metabolism 111 (2014) 499–506

subjects not on ERT is associated with low rates of cardiovascular and renal events and death in male and female subjects with Fabry disease. Conflict of interest The Canadian Fabry Disease Initiative is funded jointly by the provincial governments of Canada, Genzyme, a Sanofi Company, and Shire Human Genetics Therapies. The funding bodies did not have any role in data analysis, interpretation, or manuscript preparation. SMS, DGB, RC, JTRC, KL and MLW have served on advisory boards, received fees for speaking or travel support, and participated in other clinical trials and registries sponsored by Genzyme and Shire. Acknowledgments CFDI Investigators: Principal Investigators: Dr. M.L. West, Dr. D. Bichet, Dr. J.T.R. Clarke, Dr. R. Casey, Dr. S.M. Sirrs. Site investigators: Dr. A. Chan, Dr. S. Dyack, Dr. C. Greenberg, Dr. J. MacKenzie, Dr. B. Maranda, Dr. A. Mhanni, Dr. C. Morel, Dr. S. Murphy, Dr. C. Prasad, Dr. J. Raiman, and Dr. L Turner. The investigators would like to thank the Canadian Fabry Association, subjects, and families for their support of this project. The investigators would like to thank Dr. D. Moore, Dr. C. AurayBlais, Dr. D. Sinasac, Dr. B. Chodirker, and Dr. A. Willan for their contributions to this project. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ymgme.2014.01.014. References [1] R.J. Desnick, R. Brady, J. Barranger, A.J. Collins, D.P. Germain, M. Goldman, G. Grabowski, S. Packman, W.R. Wilcox, Fabry disease, an under-recognized multisystemic disorder: expert recommendations for diagnosis, management, and enzyme replacement therapy, Ann. Intern. Med. 138 (2003) 338–346. [2] R. Schiffmann, J.B. Kopp, H.A. Austin, S. Sabnis, D.F. Moore, T. Weibel, J.E. Balow, R.O. Brady, Enzyme replacement therapy in Fabry disease: a randomized, controlled trial, JAMA 285 (2001) 2743–2749. [3] C.M. Eng, N. Guffon, W.R. Wilcox, D.P. Germain, P. Lee, S. Waldek, L. Caplan, G.E. Linthorst, R.J. Desnick, Safety and efficacy of recombinant human alfa-galactosidase: a replacement therapy in Fabry's disease, N. Engl. J. Med. 345 (2001) 9–16. [4] M. Banikazemi, J. Bultas, S. Waldek, W.R. Wilcox, C.B. Whitley, M. McDonald, R. Finkel, S. Packman, D.G. Bichet, D.G. Warnock, R.J. Desnick, Agalsidase-beta therapy for advanced Fabry disease, Ann. Intern. Med. 146 (2007) 77–86. [5] A.C. Vedder, G.E. Linthorst, G. Houge, J.E.M. Groener, E.E. Ormel, B.J. Bouma, J.M.F.G. Aerts, A. Hirth, C.E.M. Hollak, Treatment of Fabry disease: outcome of a comparative trial with agalsidase alfa or beta at a dose of 0.2 mg/kg, PLoS ONE 7 (2007) e598. [6] A. Mehta, R. Ricca, U. Widmer, F. Dehout, A.G. de Lorenzo, C. Kampmann, A. Linhart, G. Sunder-Plassmann, M. Ries, M. Beck, Fabry disease defined: baseline clinical manifestations of 366 patients in the Fabry Outcome Survey, Eur. J. Clin. Invest. 34 (2004) 236–242. [7] C.M. Eng, J. Fletcher, W.R. Wilcox, S. Waldek, C.R. Scott, D.O. Sillence, F. Breunig, J. Charrow, D.P. Germain, K. Nicholls, M. Banikazemi, Fabry disease: baseline medical characteristics of a cohort of 1765 males and females in the Fabry registry, J. Inherit. Metab. Dis. 30 (2007) 184–192. [8] Canadian Expert Drug Advisory Committee, CEDAC final recommendation on reconsideration and reasons for recommendation agalsidase alfa, http://www. cadth.ca/media/cdr/complete/cdr_complete_Replagal_2004Nov24.pdf (Accessed August 26 2013). [9] Canadian Expert Drug Advisory Committee, CEDAC final recommendation and reasons for recommendation agalsidase beta resubmission, http://www.cadth.ca/ media/cdr/complete/cdr_complete_Fabrazyme_Resubmission_may2005.pdf (Accessed August 26 2013). [10] R.P. El Dib, P. Nascimento, G.M. Pastores, Enzyme replacement therapy for Anderson-Fabry disease, Cochrane Libr. 2 (2013) 1–64. [11] H. Tahir, L.L. Jackson, D.G. Warnock, Antiproteinuric therapy and Fabry nephropathy: sustained reduction of proteinuria in patients receiving enzyme replacement therapy with agalsidase beta, J. Am. Soc. Nephrol. 18 (2007) 2609–2617.

[12] W. Terryn, P. Cochat, R. Froissart R, A. Ortiz, Y. Pirson, B. Poppe, A. Serra, W. Van Biesen, R. Vanholder, C. Wanner, Fabry nephropathy: indications for screening and guidance for diagnosis and treatment by the European Renal Best Practice, Nephrol. Dial. Transplant. 28 (2013) 505–517. [13] S. Sirrs, J.T.R. Clarke, D.G. Bichet, R. Casey, K. Lemoine, G. Flowerdew, D.S. Sinasac, M.L. West, Baseline characteristics of subjects enrolled in the Canadian Fabry disease initiative, Mol. Genet. Metab. 99 (2010) 367–373. [14] S. Garman, D.N. Garboczi, The molecular defect leading to Fabry disease: structure of human α-galactosidase, J. Mol. Biol. 337 (2004) 319–335. [15] M.L. West, K. LeMoine, C. Simms, S. Dyack, Cross-sectional analysis of renal disease in the Nova Scotia Fabry disease kindred, J. Am. Soc. Nephrol. 13 (2002) 517A. [16] M. West, K. LeMoine, R. McKillup, J. Harnett, Renal transplantation in the Nova Scotia Fabry disease kindred, Acta Paediatr. 19 (2002) 133. [17] M.L. West, S. Dyack, C. Riddell, K. LeMoine, C. Camfield, P. Camfield, A Nova Scotia kindred with Fabry disease, Amer. J. Hum. Gen. 71 (2002) 418. [18] M.L. West, J.T.R. Clarke Casey, R.M. Iwanochko, D.F. Moore, S.M. Sirrs, Canadian Fabry disease Treatment Guidelines 2012, http://www.garrod.ca/wp-content/uploads/ Canadian-Fabry Disease-Treatment-Guidelines-2012.pdf (Accessed August 26 2012). [19] A. Mehta, J.T.R. Clarke, R. Giugliani, P. Elliott, A. Linhart, M. Beck, G. Sunder-Plassmann, on behalf of the FOS investigators, Natural history of Fabry disease: changing patterns and causes of death in FOS — Fabry Outcomes Survey, J. Med Genet. 46 (2009) 548–552. [20] D.A. Hughes, M.A. Romero, C.E.M. Hollak, R. Giugliani, P.B. Deegan, Response of women with Fabry disease to enzyme replacement therapy: comparison with men, using data from FOS—the Fabry outcome survey, Mol. Genet. Metab. 103 (2011) 207–214. [21] A. Mehta, M. Beck, P. Elliott, R. Giugliani, A. Linhart, G. Sunder-Plassmann, R. Schiffmann, F. Barbey, M. Ries, J.T.R. Clarke, on behalf of the Fabry Outcome Survey Investigators, Enzyme replacement therapy with agalsidase alfa in subjects with Fabry's disease: an analysis of registry data, Lancet 274 (2009) 1986–1996. [22] K. Sims, J. Politei, M. Banikazemi, P. Lee, Stroke in Fabry disease frequently occurs before diagnosis and in the absence of other clinical events: natural history data from the Fabry registry, Stroke 40 (2009) 788–794. [23] A. Salviati, A.P. Burlina, W. Borsini, Nervous system and Fabry disease, from symptoms to diagnosis: damage evaluation and follow-up in adult subjects, enzymereplacement, and support therapy, Neurol. Sci. 31 (2010) 299–306. [24] S.M. Rombach, T.B. Twickler, J.M.F.G. Aerts, G.E. Linthorst, F.A. Wijburg, C.E.M. Hollak, Vasculopathy in subjects with Fabry disease: current controversies and research directions, Mol. Genet. Metab. 99 (2010) 99–108. [25] Lysosomal Storage Diseases Expert Advisory Group, Adult Fabry disease standard operating procedures, http://www.specialisedservices.nhs.uk/document/ guidelines-anderson-fabry-disease (Accessed August 26 2013). [26] Anonymous, Guidelines for the treatment of Fabry disease through the life saving drugs program, http://www.health.gov.au/internet/main/publishing. nsf/Content/lsdp-info/$File/Fabry-Guidelines-29-July-2013.doc (Accessed August 26 2013). [27] R. Schiffmann, D.G. Warnock, M. Banikazemi, J. Bultas, G.E. Linthorst, S. Packman, S.A. Sorensen, W.R. Wilcox, R.J. Desnick, Fabry disease: progression of nephropathy, and prevalence of cardiac and cerebrovascular events before enzyme replacement therapy, Nephrol. Dial. Transplant. 24 (2009) 2102–2111. [28] S.M. Rombach, B.E. Smid, M.G. Bouwman, G.E. Linthorst, M.G. Dijkgraaf, C.E. Hollak, Enzyme replacement therapy for Fabry disease: effectiveness on kidney, heart, and brain, Orphanet J. Rare Dis. 8 (2013) 47. [29] F. Weidemann, M. Niemann, S. Störk, F. Breunig, M. Beer, C. Sommer, S. Herrmann, G. Ertl, C. Wanner, Long-term outcome of enzyme-replacement therapy in advanced Fabry disease: evidence for disease progression towards serious complications, J. Intern. Med. 274 (2013) 1–13. [30] T. Alegra, F. Vairo, M. de Souza, B.C. Krug, I.V.D. Schwartz, Enzyme replacement therapy for Fabry disease: a systematic review and meta-analysis, Genet. Mol. Biol. 35 (2012) 947–954. [31] K. Nicholls, Increased arterial stiffness is associated with high cardiovascular mortality in male Fabry subjects, J. Inherit. Metab. Dis. 35 (2012) 885–889. [32] S.M. Rombach, B. van den Bogaard, F. de Groot, J.E.M. Groener, B.J. Poorthuis, G.E. Linthorst, B.J.H. van den Born, C.E.M. Hollak, J.M.F.G. Aerts, Vascular aspects of Fabry disease in relation to clinical manifestations and elevations in plasma globotriaosylsphingosine, Hypertension 60 (2012) 998–1005. [33] K. Ramasubbu, J. Estep, D.L. White, A. Deswal, D.L. Mann, Experimental and clinical basis for the use of statins in subjects with ischemic and nonischemic cardiomyopathy, J. Am. Coll. Cardiol. 51 (2008) 415–426. [34] J.L. Probstfield, K.D. O'Brien, Progression of cardiovascular damage: the role of renin-angiotensin system blockade, Am. J. Cardiol. 105 (2010) 10A–20A. [35] P. Gaede, P. Vedel, N. Larsen, G.V.H. Jensen, H.H. Parving, O. Pedersen, Multifactorial intervention and cardiovascular disease in subjects with type 2 diabetes, N. Engl. J. Med. 348 (2003) 383–393. [36] D.G. Warnock, A. Ortiz, M. Mauer, G.E. Linthorst, J.P. Oliveira, A.L. Serra, L. Marodi, R. Mignani, B. Vujkovac, D. Beitner-Johnson, R. Lemay, J.A. Cole, E. Svarstad, S. Waldek, D.P. Germain, C. Wanner, on behalf of the Fabry Registry, Renal outcomes of agalsidase beta treatment for Fabry disease: role of proteinuria and timing of treatment initiation, Nephrol Dial Transplant. 27 (2012) 1042–1049.