Article Text

Natural course of Fabry disease: changing pattern of causes of death in FOS – Fabry Outcome Survey
  1. A Mehta1,
  2. J T R Clarke2,
  3. R Giugliani3,
  4. P Elliott1,
  5. A Linhart4,
  6. M Beck5,
  7. G Sunder-Plassmann6
  1. 1
    University College London, London, UK
  2. 2
    Hospital for Sick Children, Toronto, Canada and Centre Hospitalier Universitaire, Sherbrooke, Canada
  3. 3
    Medical Genetics Service/HCPA and Department of Genetics/UFRGS, Porto Alegre, RS, Brazil
  4. 4
    Charles University, Prague, Czech Republic
  5. 5
    University of Mainz, Mainz, Germany
  6. 6
    Division of Nephrology and Dialysis, Department of Medicine III, Medical University Vienna, Vienna, Austria
  1. Dr A Mehta, Department of Haematology, Royal Free Hospital, Pond Street, London, NW3 2QG, UK; atul.mehta{at}


Background: Fabry disease is a rare X-linked lysosomal storage disorder characterised by severe multisystemic involvement that leads to major organ failure and premature death in affected men and women. Over the past 7 years, the Fabry Outcome Survey (FOS) has collected data on the natural history of Fabry disease, and the long-term efficacy and safety of enzyme-replacement therapy. This paper provides an update on the first analysis of FOS data.

Design: Baseline data on clinical manifestations and causes of death in a cohort of 1453 patients (699 male, 754 female) from 19 countries worldwide were analysed. Causes of death of affected relatives were analysed separately.

Results: The most frequently reported signs and symptoms of Fabry disease were neurological. Cardiac, ocular, gastrointestinal, dermatological, auditory and renal manifestations were also common. The principal causes of death among 181 affected relatives of patients in FOS (most of whom had died before 2001) were renal failure in males (42%) and cerebrovascular disease in females (25%). In contrast, of the 42 patients enrolled in FOS whose deaths were reported between 2001 and 2007, cardiac disease was the main cause of death in both male (34%) and female (57%) patients.

Conclusion: These data suggest that the importance of renal disease as a cause of death in patients with Fabry disease is decreasing while the importance of cardiac disease is increasing. This pattern probably reflects improvements in the management of renal disease in patients with Fabry disease.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Fabry disease is a rare X-linked lysosomal storage disorder caused by deficient activity of the enzyme α-galactosidase A. It is characterised by severe multisystemic involvement that ultimately leads to major organ failure and premature death in affected men and women.1 The condition has conventionally been considered to be rare, affecting only 1 in 40 000 to 1 in 117 000 live male births.1 2 However, a recent screening study suggested a much higher incidence of 1 in 3100–4600 individuals.3

α-Galactosidase A deficiency is caused by a number of mutations affecting the GLA gene, many of which are private (ie, occurring in only single or small numbers of families).4 The α-galactosidase A deficiency results in progressive accumulation of undegraded glycosphingolipids, predominantly globotriaosylceramide (Gb3), within cell lysosomes throughout the body. Common clinical manifestations of Fabry disease include acroparaesthesia and pain crises, gastrointestinal symptoms, angiokeratomas and corneal dystrophy.

Despite the fact that Fabry disease follows an X-linked pattern of inheritance, it is now widely accepted that heterozygous females can be severely affected, although progression of the disease to organ failure generally occurs later in life and symptom severity tends to be milder and more variable than in males.57 Renal failure has been found to be a frequent cause of death in several studies, particularly in males.5 8 9 In females, data on small patient groups suggest that cardiac disease and cerebrovascular disease are the most frequent causes of death.5 911 Death occurs on average 15 years earlier in female patients and 20 years earlier in male patients compared with the general population.6 10 12

Enzyme-replacement therapy (ERT) has become an established treatment for Fabry disease, following demonstrations of both efficacy and safety in adults and children.1316 Treatment guidelines concerning ERT have now been published.17 Early diagnosis of Fabry disease is therefore important, and both general practitioners and clinicians in a range of specialties should be aware of the signs, symptoms and natural history of the disease.

This paper provides an update of the demographic and baseline clinical characteristics of a large cohort of patients with Fabry disease enrolled in the Fabry Outcome Survey (FOS) and examines the causes of death among patients and their affected relatives.5


FOS has been approved by the Ethics Institution Review Boards of participating centres (appendix A) and all patients gave written informed consent.

Patients and data collection

To minimise bias, data from all consenting patients are entered in FOS. All patients enrolled in FOS are receiving or are candidates for ERT with agalsidase alfa. On enrolment, each patient’s medical history is documented by a physician or nurse specialist, including the year of diagnosis of Fabry disease, signs and symptoms of the disease, treatment, demographic details and family history. All measurements performed routinely in clinical practice are entered into the database. Assessments of cardiac and renal function, ophthalmological, gastrointestinal and audiological examinations are optional. Anonymised data are submitted electronically by participating physicians to the central FOS database. Disease severity is systematically categorised in all patients using an adapted version of the Mainz Severity Scoring Index (FOS-MSSI).18

Data analysis

Student’s two-sample and one-sample t-tests and the Fisher exact test or χ2 test were used for statistical analyses, using SAS V.9.1. Values are presented as mean (SD) unless otherwise stated and p<0.05 was considered significant.


Patient demographics

Table 1 shows the baseline characteristics of the 1453 patients (699 male, 754 female) enrolled in FOS at the time of analysis (December 2007). Most were Caucasian (95.7%).

Table 1 Mean baseline characteristics of the 1453 patients with Fabry disease in the Fabry Outcome Survey

Disease severity

The severity of Fabry disease was greater in males than in females in FOS, as indicated by FOS-MSSI total score (mean 16.9 (SD 10.8) vs 11.0 (SD 9.0), respectively; p<0.001). This was true for all four subcategories of FOS-MSSI score (general, neurological, cardiovascular and renal).

Signs and symptoms in male and female patients

The prevalence and age at onset of the main signs and symptoms of Fabry disease in males and females at entry into FOS are shown in table 2. As in the previous analysis of FOS data, the most frequently reported signs and symptoms of Fabry disease were neurological, affecting 75% of males and 61% of females (table 2). Cardiac, ocular, gastrointestinal, dermatological, auditory and renal manifestations were also common.

Table 2 Prevalence and age at onset of organ system involvement in patients with Fabry disease enrolled in the Fabry Outcome Survey

Renal signs and symptoms of Fabry disease were reported in 693 patients in FOS. Proteinuria was the commonest renal manifestation, being seen in 332 males and 246 females. Renal failure was present in 134 males and 21 females. In total, 89 males and five females had end-stage renal disease (ESRD) (mean age of onset 32.3 (SD 19.7) and 23.3 (SD 21.3) years, respectively).

Cardiac manifestations were recorded in 422 male and 376 female patients. Of these patients, 284 males and 180 females had left ventricular hypertrophy (LVH), which was reported to have begun at a mean age of 28.7 (SD 19.3) and 34.1 (SD 24.9) years, respectively.

Cerebrovascular events occurred with approximately the same frequency in males and females (25% (n = 172) and 21% (n = 159), respectively). Stroke was reported in 39 females and 63 males (mean age of onset 51.4 (SD 14.1) and 39.2 (SD 11.5) years, respectively). Transient ischaemic attack (TIA) was recorded in 46 females and 60 males (mean age of onset 46.6 (SD 14.8) and 38.4 (SD 13.6) years, respectively).


Deaths in patients in the Fabry Outcome Survey

Data on age and cause of death in patients who have received ERT at any point in time have been collected in FOS since 2001. More recently (since August 2006) data have been collected on deaths in patients who had never received ERT. Information on any death can be entered retrospectively. At the time of this analysis, age and cause of death had been reported for 42 patients (35 male, 7 female) in FOS (table 3). Mean age at death was 51.8 (SD 9.3) years and 64.4 (SD 10.0) years for males and females, respectively. The primary cause of death in both male and female patients was cardiac disease, reported in 34% (n = 12) and 57% (n = 4) of cases, respectively. Of the 42 patients who died, 33 (28 males) had received ERT at some point during their life, and among these cardiac disease remained the commonest cause of death (n = 12; 36%), followed by infection (n = 5; 15%) and multisystemic disease (n = 4; 12%). Renal failure was responsible for three deaths (n = 3; 9%). Cardiac disease was also the commonest cause of death (n = 4 of 9; 44%) among patients who died but had not received ERT at any point. However, data on deaths in patients not receiving ERT have only been collected since August 2006 and therefore the untreated group is not entirely representative of all untreated patients.

Table 3 Cause of death in affected male (M) and female (F) patients in the Fabry Outcome Survey

The 42 patients in FOS who died were severely affected by Fabry disease. From the available absolute data shown in table 4, both men and women who died had higher MSSI scores than patients who were alive. Data were also available on whether patients had specific manifestations either before or at/after the start of ERT (information recorded only as yes/no). Of the 42 deceased patients, LVH was reported in 29 (25 male, 4 female) cases. LVH was present before the start of treatment in all 22 of the patients who went on to receive ERT and in 7 of the 9 patients who never received ERT. Absolute data on left ventricular mass (LVM) were available for 19 of the 42 patients: mean LVM at entry into FOS was higher in patients who died than in patients who were still alive (table 4). Conduction abnormalities were evident in 19 patients (15 male, 4 female), 15 of whom received ERT; these abnormalities were present before the commencement of ERT in 11 (9 male, 2 female) of these 15 patients. Valvular disease was recorded in 16 patients (13 male, 3 female): in three males, this was recorded after the start of treatment only.

Table 4 Mean (SD) data at entry into the Fabry Outcome Survey according to whether patients were alive or dead at the time of the analysis

Proteinuria was recorded as present at some point in time (yes/no) in 27 of the 42 patients (21 male, 6 female). Of the 22 patients in this group who received ERT, proteinuria was present before the commencement of ERT in 20 (16 male, 4 female). Absolute data were available for 23 patients, which showed higher proteinuria levels in patients who died than in those who were alive (table 4). Mean proteinuria levels based on the values available were also higher in patients who received ERT (females (n = 4), 1521.9 mg/24 hours (range 52.5–2495.0 mg/24 hours); males (n = 17), 900.4 mg/24 hours (range 21.0–3955.0 mg/24 hours)) than in those who did not (females (n = 1), 250.0 mg/24 hours; males (n = 1), 970.0 mg/24 hours). Hypertension was reported in 20 (15 male, 5 female) of the patients who died. Sixteen of these patients are known to have received ERT and in 15 cases (12 male, 3 female) hypertension was present before the commencement of ERT. In patients who received ERT at some point, haemodialysis was reported before the commencement in four, peritoneal dialysis in two and renal transplantation in seven (all male). After the start of ERT, one further male patient started haemodialysis and five underwent renal transplantation. Of the deceased patients who never received ERT, two males were on haemodialysis.

Stroke was reported in 19 of the 42 patients who died (16 male, 3 female) and TIAs in 13 patients (11 male, 2 female). Of the 14 stroke patients who had received ERT, 10 (nine male, one female) experienced strokes both before and after commencement of ERT. Two males had no further strokes after the commencement of ERT, and two males had their first stroke at/after the start of ERT. The remaining five patients did not receive ERT (three male, two female).

Chronic pain was reported in 25 (21 male, 4 female) and asthma in 14 (10 male, 4 female) of the 42 patients who died.

Deaths in affected relatives of patients in the Fabry Outcome Survey

Information was available on the cause of death of 181 affected relatives (118 male, 63 female) of patients in FOS (table 5). Precise data on the year of death were not available; however, most of these deaths occurred before ERT became available in 2001. The commonest causes of death in affected male family members were renal failure (n = 50; 58% aged 21–50 years) and cardiac disease (n = 31; 71% aged 21–50 years). The cause of death was unknown in 16 male relatives, but Fabry disease-related causes were suspected. Affected female relatives tended to live longer than affected male relatives. The commonest causes of death in female relatives were cerebrovascular disease (n = 16; 38% aged 21–50 years, 50% aged >50 years), followed by cardiac disease (n = 12; 92% aged >50 years) and malignancy (n = 10). Seven female relatives died from renal disease. In 16 cases where the cause of death was unknown, the family member was known or suspected to have Fabry disease.

Table 5 Number of deaths in relatives of patients enrolled in the Fabry Outcome Survey reported up to December 2007


This report describes the baseline clinical manifestations of Fabry disease in a large cohort of patients from 19 countries worldwide who were enrolled in FOS. The number of patients enrolled in FOS has increased steadily over time since the first report in 2004.5 The present analysis confirmed that the major clinical manifestations of this multisystemic disease in early adulthood are pain (acroparaesthesia), angiokeratomas, and renal, cardiac and cerebrovascular disease. Gastrointestinal manifestations and sensory organ abnormalities were also common. Although some signs and symptoms, including pain and gastrointestinal disturbances, were evident from an early age, end-stage organ damage was not seen in children in FOS.

Analysis of data on causes of death among patients currently in FOS and among their relatives suggests that the clinical course of Fabry disease may be changing. Consistent with previously published historical data,12 renal failure was the commonest cause of death in affected relatives of patients in FOS (31%), most of whom had died before 2001. When analysed according to sex, renal failure was the major cause of death in affected male relatives (42%) and cerebrovascular disease was the most frequently reported cause of death in affected female relatives (25%), as has been reported previously. Cardiac disease was the second most important cause of death in both affected male and affected female relatives.

Data obtained for patients enrolled in FOS whose deaths were reported between 2001 and 2007 are somewhat different. In these 42 patients, cardiac disease was the most frequent cause of death in both males (34%) and females (57%), and only 3 patients died from renal disease (7%). These data suggest that deaths from renal failure are becoming less common over time and most patients are now dying from cardiac disease. This was true irrespective of whether or not patients had received ERT, although formal comparison of deaths in ERT-treated and non-ERT-treated patients was not possible because of the limited data available from the latter group.

A recent retrospective review of 447 patients (279 male, 168 female) with Fabry disease19 confirms that Fabry nephropathy has a higher prevalence, occurs earlier and progresses more rapidly in the absence of ERT in males and females. This study also found a high prevalence of cardiac events (mainly arrhythmias) in 49% of males and 35% of females before commencement of ERT; strokes/TIAs occurred in 11% and 6% of males, and 8% and 4% of females, respectively.

Evidence from our own study suggests that cardiac involvement was well established in many of the patients who died. LVH was present in almost 70% of patients enrolled in FOS who died and in seven of the nine patients who died without receiving ERT. Increased LVM, valvular heart disease, conduction disturbances and hypertension were also present before commencement of ERT in most of the deceased patients. It seems likely therefore that most deaths attributed to cardiac causes resulted from the progression of pre-existing cardiac involvement associated with Fabry disease. However, a proportion of these cardiac deaths may be unrelated to Fabry disease, as cardiovascular disease is a major cause of death in normal adults, accounting for nearly half of all deaths in Europe (48%).20 In addition, most of the patients who died from cardiac-related causes also had severe manifestations affecting other organ systems (primarily renal and central nervous systems).

Considering that there is a high prevalence of renal disease, including ESRD, among patients currently in FOS, it is striking that there were few deaths associated with renal failure. Improvements in supportive care, greater access to dialysis facilities and improved management of hypertension may all have contributed to the lower number of deaths due to renal disease. Although ERT has not been shown to be effective in reversing ESRD, several studies confirm that both agalsidase alfa and agalsidase beta can slow the decline, or perhaps even stabilise, renal function in patients with lesser degrees of renal impairment.2124 Current guidelines for the treatment of Fabry disease, emphasise the importance of angiotensin-converting enzyme inhibitors/angiotensin receptor blockade, careful management of hypertension and early initiation of ERT.25 As cardiac disease is now the most commonly reported cause of death, its treatment is clearly important. Studies have shown that both agalsidase alfa and agalsidase beta improve cardiac structure and function in males and females with Fabry disease.2629 If treatment of cardiac and renal dysfunction are optimised and patients receive optimal supportive care, it is also conceivable that early stroke may be prevented, although recent reports suggest that premature stroke may still occur in patients receiving ERT.30 Unfortunately, owing to the lack of an appropriate control group, it was not possible to evaluate specifically whether ERT changed the progression of disease manifestations or increased the life expectancy of patients with Fabry disease in the present study.

As is generally the case for most non-interventional observational databases, there are several limitations to this study. First, as all clinical decisions and data entry are at the discretion of the attending physician, datasets are often incomplete and there is a lack of standardisation in the assessments between the patients. There is also variability between patients in terms of the treatment protocols used and the supportive care provided. Thirdly, patient-reported data on the causes of death in relatives are subject to selective recall and these data have not been verified with reference to source documents. In addition, only simple data are available for some variables, such as cause of death. Finally, the nature of the database makes it likely that untreated and mildly affected patients are under reported.

In conclusion, these data suggest that the natural clinical course of Fabry disease may be changing. The importance of renal failure as a cause of death is decreasing, whereas the importance of cardiac disease is increasing. This is partly a result of significant improvements in supportive care, particularly for renal failure and hypertension, which have occurred in recent years. Such improvements, in addition to the provision of ERT, may be contributing to changes in morbidity and mortality associated with Fabry disease. It is hoped that in the future these natural history data will enable the accurate assessment of the effect of ERT in patients with Fabry disease. There remains, however, a very real need for new treatments, particularly for oral therapies and for the development of treatment approaches that are better able to access the central nervous system and organs such as the heart and kidney.


All the FOS Investigators who have submitted data from their patients to the FOS database (see for complete list of contributors). The FOS database is under the independent control of the FOS International Board. Editorial assistance to the authors was provided by Dr Harriet Crofts (Oxford PharmaGenesis Ltd, Oxford, UK).

Appendix A

Ethics committee approval

The FOS database has been approved by the institutional ethics review boards of the >100 participating centres and all patients gave written informed consent. As so many investigators and participating centres are involved in the FOS, we have provided a list of the investigators who submitted data to the FOS (correct as of December 2007), together with their location.

Argentina: P Rozenfeld (La Plata).

Australia: K Nicholls (Melbourne).

Austria: G Sunder-Plassmann, J Kleinert, A-C Hauser, O Bodamer, W Kristoferitsch, W Schreiber (Vienna); B Plecko,T Kroepfl, I P Kotanko, C Binder, M Brunner-Krainz (Graz) Belgium: C Verellen-Dumoulin, F Dehout, D Roland, F Henry, J Jaumotte (Charleroi); F Eyskens, A Neirinck (Antwerp); M C Nassogne, Y Pirson, A Ballaut, P Goyens, M Libert (Brussels); N Mazoin (Verviers).

Brazil: R Giugliani, L Kalakun, C Netto, T Vieira, L Jardim (Porto Alegre).

Canada: B Lemieux (Sherbrooke); D Bichet, M F Arthus, C Fortier (Montreal); J Clarke, C Azcona (Toronto); M West, K LeMoine (Halifax); A Chan, E Kowal-Zuk (Edmonton); L Clarke (Vancouver).

Czech Republic: A Linhart, L Golánv, J Bultas, J-C Lubanda, S Magage, D Karetová, G Dostalova.

France: G Choukroun (Amiens); J Berthelot (Angers); S Carey Reomonnay (Besançon); D Lacombe (Bordeaux); S Benziane (Cambrai); A Khau van Kien (Montpellier); J M Mittelberger (Freyming Merlebach); C Thevenot (Laon); E Hachulla, D Dobbelaere (Lille); B Dussol (Marseille); R Reade (Maubeuge); P Kaminsky (Nancy); C Guyot, M Lino (Nantes); T Ghafari (Nice); E Monlun (Pau); D P Germain, N Ouali, G Touati, O Lidove, B Knebelmann (Paris); R Perrichot (Vannes); R Jaussaud (Reims); B Richalet (Saint Lô); V Klotz (Selestat); E Andres (Strasbourg); D Caraman (Thionville); J Bazex (Toulouse); P Hardy (Arras); P Bataille (Boulogne/mer).

Germany: M Beck, K Baron, S Delgado-Sanchez, R Hartung, G Kalkum, C Whybra, C Kampmann, C Wiethoff, A Schwarting, A Keilmann, S Pitz, K Bruns, F Bähner, K Lackner (Mainz); B Koletzko, B Hoffmann, B Pontz (Munich); H P H Neumann, M Cybulla, G Schluh, A Winter (Freiburg); A Das, S Illsinger, T Lücke (Hannover); A von Arnim-Baas, J Hennermann, N Weinhold, J-M Berger (Berlin); S Stolz (Cottbus); S Grabbe, T Jansen, E Chrobot (Essen); A Gal, E Schäfer, K Ullrich, N Muschol, I Berkau (Hamburg); A Rolfs, T Böttcher, S Miethe, F Rimmele (Rostock); C Wanner, I Davydenko, R Wössner (Würzburg); S Al Sawaf, B Hoffmann (Düsseldorf); R Schueuermann (Mannheim); C Haase (Jena).

Hungary: L Maródi (Debrecen).

Italy: O Gabrielli (Ancona); D Concolino (Catanzano); W Borsini (Firenze); R Parini, S Sala, M Rigoldi (Monza); R Di Vito (Ortona); A P Burlina, A Burlina (Padova); C Feliciani, R Ricci, D Antuzzi, M Castorina, A Zampetti (Rome); S Feriozzi (Viterbo).

The Netherlands: C Hollak, A Vedder, E Ormel, A van Duinen, F Wijburg (Amsterdam).

Norway: G Houge (Bergen); Ø Kaarbøe (Ålesund).

Slovenia: B Vujkovac (Slovenj Gradec).

Spain: V Fernández (Clínico Santiago); J González (Puerta del Mar-Cádiz); A Rivera (Xeral Cies-Vigo); J Andreu (Virgen de Rocío-Sevilla); J Paniagua, F Rodriguez (El Bierzo-Ponferrada); I Martin (JR Jimenez Huelva); J Torras (Bellvitge-Barcelona); V Torregrosa (Clinic-Barcelona); F J Barbado, J Garcia-Consuegra, A Garcia de Lorenzo, M López (La Paz-Madrid); S Hernández (San Agustin-Linares); M A Barba (Albacete); A Perez García, I Febrer (Gral Univ Valencia); E Gómez Huertas (Central-Asturias); J Herrera (General-Asturias); J Ara, J Bonal, G Pintos (Trias i Pujol-Badalona); R Torra, J Ballarin (Puigvert-Barcelona); R Botella, A Franco (Alicante); V Valverde, J Martin (Elda); F Cabadés (Vinaroz); J M de Toro (Ourense).

Switzerland: U Huyen-Do (Bern); F Barbey, J Theytaz, D Hayoz (Lausanne); G Schulthess, U Widmer, K Walter, B Steinmann, M Schärer, F Ruggieri (Zurich).

UK: T M Cox, P Deegan, U Ramaswami, N Wright, P Desveaux (Cambridge); A Mehta, D Hughes, P Jeevaratnam, L Richfield, S Goodwin, S Pringle, A Milligan, R Bruce, P Elliott, A Muthumala, C Orteu, L Ginsberg, R Baker, D Hajioff, S Keshav, A Ioannides, B Cadge (London).

USA: E Boyd, K Crandall (Asheville); G Pastores, N Barnett (New York); R Schiffmann, M Ashmus (NIH); H Lien, M Van Skiver (Arizona); R Martin (Washington).


Supplementary materials


  • ▸ An appendix is available online only at

  • Funding: Data collection and analysis in FOS are supported by Shire Human Genetic Therapies (HGT), Danderyd, Sweden. The sponsor had no role in the interpretation of data or writing of the report.

  • Competing interests: AM received speaker fees, research funds and funding for symposium attendance from Shire HGT; JTRC received funding for symposium attendance from Shire HGT; RG received speaker fees and funding for symposium attendance from Shire HGT; PE received speaker fees, funding for symposium attendance and unrestricted educational grants from Shire HGT; AL and GSP received speaker fees, travel grants and research support from both Shire HGT and Genzyme; MB received speaker fees, funding for symposium attendance and unrestricted scientific grants from Shire HGT, Genzyme and Biomarin. Honoraria were not paid in relation to FOS data entry or for the writing of this report.