(FD) is a rare, X-linked lysosomal storage disease caused by a deficiency of alpha galactosidase A enzyme activity due to mutations in the GLA gene. Accumulation of the glycosphingolipid globotriaosylceramide (GL-3) and its derivative globotriaosylsphingosine (lyso-Gb3) causes multisystemic disease, life-threatening complications, and a reduced life expectancy in both males and females.1 GL-3 is predominantly intracellularly and lyso-Gb3 is predominantly in plasma. Target organ manifestations include progressive renal failure, cardiomyopathy and recurrent strokes/transient ischaemic attacks (TIAs), resulting in an increased risk for early death if untreated.2,3 Enzyme replacement therapy (ERT) with agalsidase (human recombinant alpha galactosidase A) reduces GL-3 inclusions and lyso-Gb3 levels.4,5 Two agalsidase preparations were approved in Europe in 2001, both administered intravenously every other week (EOW): agalsidase-alfa (Replagal®) administered at the licensed dose of 0.2mg/kg; and agalsidase-beta (Fabrazyme®) administered at the licensed dose of 1mg/kg body weight.6,7 Although ERT has been in clinical use since 2001, many questions remain regarding timing if treatment initiation, optimal dose, and treatment goals.8,9 Thus, although ERT was shown to slow disease progression,6 recent long-term analyses of agalsidase-alfa as well as agalsidase-beta treatment outcomes suggested substantially diverse frequencies of clinically relevant events while on therapy.3,10–12 A target organ affected is usually the kidney. In children, podocyte glycolipid deposits are associated with podocyte injury (foot process effacement), which precedes pathological albuminuria, which is usually the first evidence of Fabry nephropathy.13,14 This is followed by a progressive decrease in glomerular filtration rate (GFR), leading to the need for renal replacement therapy (RRT) at a mean age of 40 years in classical FD, that is, 5–15 years before the first cardiac event.15,16 The magnitude of proteinuria is a key determinant of nephropathy progression in treated and non-treated patients.12,17

In the present study we analysed the incidence of clinically relevant renal, cardialogical and neurological FD-related events in 69 patients with FD treated at Spanish centres and explored the factors predicting outcomes.

In this retrospective multicentre longitudinal study, a total of 69 (27 females, mean age at diagnosis 42±15 years) genetically confirmed patients with a clinical classical or later-onset phenotype of FD from centres in Spain were analysed. We assessed the incidence of first renal, cardiac, or neurological event or death during long-term ERT (median time on ERT 60 (24–120) months). Table 1 provides a comprehensive overview of the baseline characteristics of the study group divided by gender. Males were younger (p=0.006) and had lower eGFR (p=0.004) at the time of ERT initiation but did not display higher IVSD and males more frequently carried later-onset phenotype (p=0.018) (Table 1).

An overview of GLA mutations and their clinical phenotypes is provided in Table 2. No differences were found among baseline characteristics of FD patients divided according to ERT type (data are not showed).

Outcomes

During ERT, 21 patients (30%) suffered a first incident clinical event: 13 males (31%) and 8 females (30%). The distribution of clinical events was: six renal, 13 cardiac (3 cardiovascular deaths included) and 2 neurological events. New cardiac events included four cases of heart failure, four cases of arrhythmia/atrial fibrillation, two implantable cardioverter device/pacemaker implantations and three myocardial infarctions. Neurological events included two strokes and renal events included patients starting renal replacement therapy (5 dialysis and 1 renal transplant). Patients suffering from an event while on ERT presented a significantly lower eGFR at baseline compared with those without events (p=0.014) and higher ACR values (p=0.043) (Table 3). Most patients who suffered a clinical event after ERT carried mutations related with likely classic phenotype (19 out of 21 patients) (p=0.011). The age at the start of ERT, gender or previous clinical events were not different between patients who suffered or not incident clinical events (Table 3). Patients with CKD before the start of ERT had more frequent events (log Rank 12.423, p=0.001), even after excluding incident renal events (log Rank 4.086, p=0.043) and these differences were more relevant in females (log Rank 18.514, p<0.001 for all events and females and logRank 3.442, p=0.064 for males) (Fig. 2a–d). Baseline CKD predicted the incidence of clinical events in different Cox models, adjusted for gender, treatment type, age at start of treatment, phenotype, and RAASi treatment (Table 4).

Table 3.

Differences between patients who suffered or not incident.

	Clinical events (n=21)	No clinical events (n=48)	p
Sex	13 males, 8 females	29 males, 19 females	1.000
eGFR (ml/min/1.73m2)	55.1±31.9	77.2±33.8	0.014
ACR(mg/g)	440(47–1839)	115(34–629)	0.043
IVSD(mm)	15.6±4.7	13.7±5.0	0.167
Age at start of ERT	48.1±11.2	43.0±14.6	0.162
Phenotype (classic/later-onset)	19/2	28/20	0.011
Previous clinical events (%)	38%	31%	0.591

Fig. 2.

Survival free clinical events according chronic kidney disease. 2a: all population and clinical events, logRank: 12.423, p=0.001, 2b: non-renal events, logRank: 4.086, p=0.043, 2c: female patients and all clinical events, logRank: 18.514, p=0.001, 2d: male patients and all clinical events, logRank: 3.442, p=0.064.

(0.2MB).

Table 4.

Factors influencing incidence of clinical events after ERT. Cox regression analysis.

Variables	HR (95% IC)	p
Baseline age (years)	1.041 (1.003–1.080)	0.034
Gender (female)	1.060 (0.432–2.603)	0.898
Phenotype(later-onset)	0.294 (0.068–1.270)	0.101
Treatment type (agalsidase beta)	1.672 (0.690–4.056)	0.255
ACR (mg/g)	1.001 (1.000–1.001)	0.026
CKD G3–G5 (yes)	4.568 (1.794–11.733)	0.001
RAASi (yes)	2.012 (0.672–6.023)	0.212
Pre-ERT clinical events (yes)	1.215 (0.483–3.057)	0.679
CKD vs age	4.205 (1.641–10.775)	0.003
CKD vs gender (male)	7.214 ((2.393–21.752)	0.001
CKD vs ERT (agalsidase-beta)	4.549 (1.772–11.678)	0.002
CKD vs RAASi	4.323 (1.668–11.203)	0.003
CKD vs phenotype(later-onset)	3.979 (1.545–10.248)	0.004
CKD vs ACR	2.816 (0.992–7.994	0.052

ACR: albumin/creatine ratio, CKD G3-G5: chronic kidney disease with eGFR<60ml/min/1.73m2. RAASi: renin–angiotensin–aldosterone system inhibitors.

In this study, we analysed the incidence of clinical events in patients with FD on long-term ERT and identified differences between patients suffering or not suffering events. Our main finding was the negative impact of baseline renal function at the start of ERT on incident clinical events in a population with lower baseline eGFR and higher use of RAS blockers than prior reports. CKD at the start of ERT increased 3- to 7-fold the risk of clinical events while on therapy. Another key and novel finding is that the impact of CKD on outcomes was also present, and even to a higher extent, in females as compared to males.

Thirty per cent of our patients on ERT during a median of 60 months developed a clinically relevant event. This percentage is slightly higher than the incidence of clinical events reported for patients on agalsidase-beta (17%–19%),19,20 but lower than published for agalsidase-alfa (50%).21,22 Our patient characteristics differ from these prior reports mainly in the lower eGFR at the start of ERT, both in men and in women (61±36 and 84±27ml/min/1.73m2, respectively), probably as patients were treated mostly in nephrology units. In this regard, and again different from some prior reports, a high percentage of patients were on RAS blockers. In prior reports, age, gender and classic phenotype were strong predictors of the risk to develop clinical events while on ERT.10,11,20,23–26 An insufficient sample size or a higher prevalence of lower eGFR may have precluded to reproduce some of these findings. Thus, only age, baseline renal function and albuminuria were predictors of incident clinical events, and not only of renal events, but also of cardiological and neurological events.

The present results agree with some prior reports but also add novel findings. Thus, Lenders et al.21 reported that in a cohort of 54 FD patients (28 males) on ERT for 81±21 months, a decreased baseline eGFR (<75ml/min/1.73m2) was associated with an increased risk for cardiovascular [HR 3.59 (95% CI 1.15–11.18); p=0.0273] or combined renal, cardiac and neurological end points [HR 4.77 (95% CI 1.93–11.81); p=0.0007] while on ERT. However, no results were reported independently for males or females and their patients had less severe kidney disease (higher baseline eGFR at 87ml/min/1.73m2 and lower baseline ACR) and lower use of RAS blockers. In a recently published retrospective study of 293 FD patients on ERT by Arends et al.,27 while age, gender and phenotype were predictors of event rates, eGFR was the most important predictor. Thus, the hazard ratios increased from 2 at eGFR<90ml/min/1.73m2 to 4 at eGFR<30, compared to patients with an eGFR>90ml/min/1.73m2. These results are like ours, however in Arends’ study, the influence of baseline renal function was only significant in classical male patients. Thus, for the first time, we show that initiation of ERT in women before renal function deteriorates has a similar or even larger impact as in Fabry males to prevent clinical events.

No significant changes were observed in renal function and cardiac parameters (eGFR or IVSD) during the observation period, while trends to improvement were observed for ACR. In addition, we did not observe an influence of the type of treatment or gender on changes in these parameters. In the literature, some studies show no change in kidney or cardiac parameters over time while on ERT.22 Since Fabry disease is a lifelong progressive debilitating disease associated with premature mortality, no change in these variables is an indication of disease stabilisation.

An issue not addressed in the present report is whether the increased risk of clinical events is direct consequence of more advanced Fabry disease in the different target organs, being the severity of kidney disease a surrogate marker of overall FD severity, versus an alternative explanation in which the increased risk of clinical events would be related to CKD itself. This is not a mere theoretical exercise: if it represents advanced FD, patients with FD-related CKD might benefit from aggressive initial therapy for FD, perhaps even involving the simultaneous initial use of two different FD treatments (e.g. ERT and chaperone or ERT and substrate reduction therapy when this becomes available). However, CKD itself is associated with an increased risk of CKD progression, cardiac events and stroke.28 If this were the main driver of the increased incidence of clinical events, then aggressive therapy of CKD-related risk factors by nephrologists may improve outcomes. In favour of this later hypothesis, a low GFR was an independent predictor of events when adjusted for age, which may be considered another surrogate for overall severity of end-organ damage.

Our study has several limitations. First, it is a retrospective and multicentric study which can influence the data collection. Second, the median follow-up was relatively short (5 years) for a disease that have a natural history of 40 years up to the first severe clinical event. In this regard, the Canadian Fabry Disease Initiative Trial required a follow-up of 10 years to demonstrate statistically significant difference for the renal events.3 Third, the sample size is relatively small, although larger than other observational studies independent from industry and lyso-Gb3 levels were not available. By last, recent studies demonstrated the clinical impact of neutralizing anti-drug antibodies in male patients with Fabry disease,29 but in our study, blood samples were not available to determine antibody titers. This fact could be a bias about the effect of agalsidase-alfa or beta in the outcomes.

In conclusion, renal function at the start of ERT is the main predictor of the development of long-term clinical events even when differences related to type of treatment or type of mutation were not observed. As a novel finding, this is also the case in female patients, suggesting that ERT should be initiated early in the course of the disease, early meaning according to the present report, at least before renal function decreases. Further studies should address whether more aggressive therapy of FD and/or CKD-related risk factors improves outcomes (Fig. 3).

Fig. 3.

Relationship among Fabry disease, chronic kidney disease and outcomes. CKD: chronic kidney disease, CNS: central nervous system.

(0.12MB).

1. Baseline renal function is the main predictor of the long-term clinical events.

2. Enzyme replacement therapy should initiate early before renal function decreases.

3. Renal function predicts outcomes in males and females.

MG, JT, AO, RT are supported by ISCIII RETIC REDINREN, RD016/009 and FEDER funds.

Informed consent was obtained from all individual participants included in the study.

MG has received research support from Sanofi Genzyme and has received speaker honoraria and travel support from Sanofi Genzyme and Shire. MLM, AO, RT are consultants for Amicus Therapeutics, Sanofi Genzyme and Shire; have received research support from Sanofi Genzyme and Shire and have received speaker honoraria and travel support from Amicus Therapeutics, Sanofi Genzyme, and Shire. RMA has received speaker honoraria from Shire and travel support from Sanofi Genzyme. FD has received travel grants and speaker fees from Takeda and Amicus. IM received speaker honoraria and travel support from Sanofi Genzyme and Shire. IA has received travel and courses supported from Shire.