SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. What’s known
  4. Introduction
  5. Material and methods
  6. Results
  7. Statistical analysis
  8. Discussion
  9. Funding
  10. Author contributions
  11. References

Aim:  Fabry’s disease is an X-linked inherited lysosomal storage disorder caused by the deficient activity of alpha-galactosidase A. The interrelationships between clinical symptoms in Fabry patients have not yet been fully established. Using cluster and multivariate analysis, the aim of the study was to determine the relationships among clinical symptoms and organ involvement, and predictive clinical symptoms for disease severity.

Methods:  Clinical data obtained from 108 French Fabry patients were retrospectively collected and analysed using multiple correspondence analysis and hierachical ascendant classification. Multivariate analysis was also performed to determine among clinical symptoms predictors for cardiac disease (HRT), renal involvement (KDN) and brain complication (STR).

Results:  The cohort comprised 41 male patients (aged 28.9 ± 11.6 years) and 67 female patients (aged 40.4 ± 15.5 years). Three main clusters of clinical symptoms could be delineated, characterising disease progression: the first cluster grouped digestive disorders (found in 30% of the patients) and exercise intolerance (32%), the second, cluster dyshidrosis (47%), acroparesthesia (67%), angiokeratoma (44%) and cornea verticillata (54%), the third, cluster grouped KDN (30%), HRT (39%) and STR (25%) and hearing loss (44%). In univariate analysis, the patient age predicted HRT and KDN, dyshidrosis predicted HRT and STR, angiokeratoma predicted KDN and cornea verticilla and hearing loss predicted KDN, HRT and STR. In multivariate analysis, hearing loss and age were independent predictors of organ complication.

Conclusion:  Among the various interrelated clinical symptoms occurring in Fabry disease, patients with dyshidrosis and particularly hearing disorders appear to be at higher risk of organ complications.


What’s known

  1. Top of page
  2. Summary
  3. What’s known
  4. Introduction
  5. Material and methods
  6. Results
  7. Statistical analysis
  8. Discussion
  9. Funding
  10. Author contributions
  11. References
  • • 
    Fabry’s disease is a X-linked lysosomal storage disorder, affecting both female patient and male patient, and characterised by a multisystemic disease. Cardiac disease, renal involvement and cerebrovascular disease are major complications, which can be in part prevented by early enzyme replacement therapy.
  • • 
    The interrelationships between clinical symptoms and predictors for major complications have not yet been fully established in Fabry patients.

What’s new

  • • 
    Advanced age of the patient and presence of dyshidrosis, angiokeratoma, cornea verticilla and hearing loss are all statistically linked to one or more major complications in univariate analysis. Hearing loss was an independent predictive factor for cardiac disease, renal involvement and cerebrovascular disease.
  • • 
    In Fabry patients, hearing disorder should be systematically screened and should be considered as a clinical marker of disease severity.

Introduction

  1. Top of page
  2. Summary
  3. What’s known
  4. Introduction
  5. Material and methods
  6. Results
  7. Statistical analysis
  8. Discussion
  9. Funding
  10. Author contributions
  11. References

Fabry’s disease (OMIM 301500) is an X-linked lysosomal storage disorder caused by deficiency of the enzyme alpha-galactosidase A, leading to the accumulation of its principal substrate, the neutral glycosphingolipid globotriaosylceramide (Gb3) in the walls of small blood vessels, renal glomerular and tubular epithelial cells, cardiomyocytes, as well as in nerves and dorsal root ganglia (1). This results in a complex, multisystemic and clinically heterogeneous disease in which involvement of the neurovegetative system explains acroparesthesia, hypo- or hyperhidrosis, gastrointestinal disturbances, exercise intolerance and fever attacks. Other characteristics include angiokeratoma (ANG) caused by Gb3 deposits in cutaneous vessels and corneal opacities (cornea verticillata). However, the main complications of Fabry’s disease are ischaemic stroke, cardiomyopathy, as well as progressive renal failure usually associated with significant proteinuria (1). Both hemizygous males and heterozygous females can be affected, although male patients are usually considered to develop a more severe form of the disease with a lower age at onset than female patients.

Enzyme replacement therapy (ERT) with intravenous infusions of recombinant human alpha-galactosidase A consistently clears lysosomal inclusions from vascular endothelial cells (2). However, some complications of the disease are not responsive to ERT, suggesting that treatment must be initiated early in the course of the disease to be optimally effective (1). Usual recommendations are to treat all male patients and all female patients with substantial disease manifestations as early as possible (3), although these manifestations are not homogeneously defined worldwide. In fact, the relationships between clinical symptoms in individual Fabry patients have not been well documented, and a better identification of intercorrelations between clinical symptoms as well as clarifying those symptoms at risk could significantly help clinicians in their decision to initiate ERT, especially in women.

Therefore, this study was aimed at determining (i) the relationships among clinical symptoms and organ involvement in a cohort of Fabry patients using multidimensional cluster and multivariate analyses and (ii) predictive clinical symptoms for disease severity.

Material and methods

  1. Top of page
  2. Summary
  3. What’s known
  4. Introduction
  5. Material and methods
  6. Results
  7. Statistical analysis
  8. Discussion
  9. Funding
  10. Author contributions
  11. References

Patients

Routine investigations required for considering a patient for inclusion were: careful medical history and physical examination, audiogram, ophthalmological examination, electrocardiogram and echocardiography, determination of serum creatinine and of creatinine clearance using MDRD estimation, assessment of microalbuminuria and proteinuria. The study was approved by local ethics committees and the Commission Nationale de l’Informatique et des Libertés (approval number 1310844), and informed consent was obtained from all patients.

The following data were recorded and classified into two binary categories (presence or absence of symptom) defining the descriptive variables in the study: acroparesthesia (ACR), exercise intolerance (INT), angiokeratoma (ANG), hypo- or hyperhidrosis (HID). Digestive disorders (DIG) included abdominal pain or diarrhoea. All the above data were labelled as ‘clinical symptoms’ in the study. Routine investigations also comprised:

  •  Audiometry permitting the diagnosis of sensorineural hearing loss (SHL) defined as a hearing threshold above the 95th percentile for age- and gender-matched normal controls.
  •  Ophthalmological examination performed for the presence of cornea verticilatta (CVR).
  •  Electrocardiogram and echocardiography allowing the diagnosis of hypertrophic cardiomyopathy, defined according to the usual recommendations (4). Cardiac disease (HRT) was defined as the presence of a significant hypertrophic cardiomyopathy or rhythm disturbances.
  •  Standard brain magnetic resonance imaging (MRI) evaluation for presence of stroke and white matter lesions. Brain complications (STR) were defined as antecedent of transient ischaemic attacks or stroke or evidence of asymptomatic white matter lesions.
  •  Determination of serum creatinine levels and proteinuria. Glomerular filtration rate (GFR) was evaluated using MDRD estimation. Renal involvement (KDN) was defined as a GFR lower than 60 ml/min or proteinuria higher than 0.3 g/day.

Renal involvement, brain complication and cardiac disease were labelled as ‘organ complications’ in the study. In patients treated with ERT, age at inclusion and clinical data were collected just before starting the treatment. In other patients, the more recently available data were considered for analysis. Only patients for whom complete information was available were included. Patients exhibiting mutations considered to be polymorphisms in the α-galactosidase A gene were excluded.

Multiple correspondence analysis (MCA)

Multiple correspondence analysis is a descriptive analysis of multidimensional qualitative data (5,6). It allows the analysis of a matrix of N individuals (each patient in this study) depicted by P qualitative variables (all above-defined binary variables in this study). Essentially, projections of these individuals in a P-dimensional space are used to calculate factorial axes, the first axis retaining the maximum variance, and the subsequent axes retaining the residual variance and being perpendicular to each other. Briefly, MCA provides a type of graphic representation that allows studying the relationship between the categories of variables in a multiway contingency table. The MCA permits continuous quantitative coordinates to be attributed to variables, as well as the most significant factorial axes to be selected, to reduce the number of dimensions of the initial space (5,6). The importance of the categories of variables for constructing each axis is measured by their absolute contribution and aids in the interpretation of the axis. Categories with high absolute contribution show greater importance in the formation of the factorial axis. Coordinates can also be calculated for additional (supplementary) variables, which do not, however, contribute to the determination of factorial axes. Graphic analysis of association of variable categories is performed by considering their geometric proximity and the separation of categories by quadrants; hence, the closer the variables, the more interrelated they are, whereas those separated by quadrants display groups with opposite profiles.

Hierarchical ascendant classification (HAC)

Hierarchical ascendant classification was used in this study as a complement of MCA to determine suspected variable clustering by the proximity of variable projections observed on factorial axes (5,7). Indeed, HAC allows individuals to be grouped according to their coordinates, by calculating pair-wise distances between cases and aggregating the closest ones. Intra-cluster inertia (variance) is measured by the sum of the squares of distances between cluster cases and the cluster centroid. The closer the cases are grouped around the cluster centroid, the lower the intra-cluster inertia. The number of clusters to consider was determined classically by inspecting the global intra-cluster inertia as a function of cluster numbers.

Statistics

Results are expressed as mean ± standard deviation. Categorical variables were compared using the chi-squared test. Comparisons of means were performed using the nonparametric Mann–Whitney U-test. Multivariate analysis was carried out using multiple logistic regression analysis to study binary variables. Iterative receiver-operating characteristic (ROC) analyses were applied to calibrate cut-off values for variables of interest, which were identified as those that optimised the sum of sensitivity and specificity. Sensitivity, specificity and relative risk were calculated with 5% confidence interval (5% CI). All statistical tests were two-sided; p-values less than 0.05 were considered significant. Statistical analyses were performed using the XLSTAT2009 statistical software package (Addinsoft).

Results

  1. Top of page
  2. Summary
  3. What’s known
  4. Introduction
  5. Material and methods
  6. Results
  7. Statistical analysis
  8. Discussion
  9. Funding
  10. Author contributions
  11. References

One hundred and eight patients from 51 families were included in the study, comprising 41 men (aged 28.9 ± 11.6 years at inclusion) and 67 women (aged 40.4 ± 15.5 years). Age at diagnosis was 25.9 ± 14.2 years (range 3–60 years) for male patients and 37.5 ± 16.3 years (range 4–77 years) for female patients. Fabry’s disease was diagnosed by familial screening in 41 female patients (61.2%) and 10 male patients (24.4%). Ten patients were asymptomatic when considering clinical symptoms analysed in our study: seven female patients (age range 18–47 years) and three male patients (age range 28–37 years), whereas 57 patients (52.3%) presented at least one major symptom.

Prevalences of clinical symptoms and organ involvement are given in Table 1. Acroparaesthesia, dyshidrosis and particularly ANG were more frequent in male patients than in female patients, whereas prevalence of organ complications was similar in both genders.

Table 1. Prevalence of clinical symptoms and organ complications in 108 patients with Fabry’s disease
Clinical symptomsMales n = 41 n (%)Females n = 67 n (%)All n = 108 n (%)p-value
Acroparaesthesia33 (80.5)39 (58.2)72 (66.7)0.017
Angiokeratoma28 (68.3)20 (29.9)48 (44.4)< 0.0001
Dyshidrosis25 (61.0)26 (38.8)51 (47.2)0.025
Exercise intolerance17 (41.5)17 (25.4)34 (31.5)0.081
Gastrointestinal disorders13 (31.7)19 (28.4)32 (29.6)0.711
Hearing loss20 (48.8)27 (40.3)47 (43.5)0.388
Cornea verticillata21 (51.2)37 (55.2)58 (53.7)0.685
Cardiac disease16 (39.2)26 (38.8)42 (38.9)0.982
Renal involvement14 (34.1)18 (26.9)32 (29.6)0.421
Stroke10 (24.4)17 (25.4)27 (25.0)0.909

Descriptive analysis: correspondence analysis and hierarchic ascendant classification

The array of 108 individuals × 10 variables was treated using MCA. The variables ‘gender’ and ‘age’, which was categorised in four intervals, were projected on factorial axes as additional variables. The first two factorial axes accounted for 87% of total variance, whereas the other axes were not significantly contributive. The first factorial axis (76.4% of total variance) divided all negative (absence) categories of variables with all positive (presence) categories of variables (Figure 1). All variables contributed significantly to this axis. The second axis (10.6% of total variance) divided the variables DIG and INT with all organ complications (HRT, KDN and STR). The variable SHL was projected in proximity to that of organ complications in the upper right quadrant of the first factorial plane. Other variables representing clinical symptoms (ACR, ANG, CVR, HID) projected on an intermediate position between DIG/INT and the variables representing organ complications, near the 1st factorial axis, hence indicating a weak relationship among these two groups of variables. The age category corresponding to the younger patients projected near the same quadrant as that of the ‘absence’ of all clinical symptoms category. In contrast, the presence of organ complications and hearing loss projected in the same area as the diagnosis age > 50 years category, indicating that organ complications tended to be observed in patients with late diagnosis, whereas patients with early diagnosis tended to present less severe symptoms. A similar reasoning could be made to a certain extent regarding genders: the ‘male’ category projected on the same side as the ‘presence’ of clinical signs category on the first factorial axis, in contrast to the ‘female’ category, which was characterised by absence of clinical symptoms. This indicated a tendency towards a more severe form of the disease in male patient than in female patients.

image

Figure 1.  Multiple analysis of correspondence, showing the projections of clinical data on the first factorial plane, which represents 87% of the total variance, and results of the hierarchical ascendant classification. On the left, categories of variables corresponding to the absence of clinical symptoms aggregate into two clusters, dividing those of clinical symptoms and those of organ complications, including hearing loss. On the right, categories of variables corresponding to the presence of clinical symptoms merge in three distinct clusters corresponding to organ complications, and to clinical symptoms separating digestive disorders and exercise intolerance on the one hand, and the others on the other hand. Abbreviations: ACR: acroparesthesia, ANG: angiokeratoma, CVR: cornea verticillata, DIG: digestive disorders, HID: dyshidrosis, HRT: heart disease, INT: exercise intolerance, KDN: kidney involvement, SHL: hearing loss, STR: cerebrovascular complications. The labels ‘0’ and ‘1’ indicate, respectively, absence and presence of the clinical symptoms

Download figure to PowerPoint

Hierarchic ascendant classification was performed using the Euclidian distance and the Ward criterion for aggregation with the initial array of 108 rows × 10 columns and with the coordinates of the 108 individuals calculated using MCA on the first two factorial axes. The results were similar, showing a distribution into five main clusters (Figure 1): the first two clusters pooled all ‘absence’ of variables categories. A third cluster merged the categories ‘presence’ of the variables representing clinical symptoms (ACR, ANG, CVR, HID). A fourth cluster grouped INT and DIG. Finally, the fifth cluster merged the categories ‘presence’ of HRT, KDN and STR, which also included SHL.

Statistical analysis

  1. Top of page
  2. Summary
  3. What’s known
  4. Introduction
  5. Material and methods
  6. Results
  7. Statistical analysis
  8. Discussion
  9. Funding
  10. Author contributions
  11. References

Iterative ROC analyses indicated that the best cut-off values of age at inclusion to predict organ involvement in our cohort were age greater than 25 years in men and age greater than 35 years in women, with a relative risk (RR) of cardiac disease of 10.0 (5% CI 3.0–33.7) and of 2.7 (5% CI 1.3–10.3) of kidney involvement in the older patients compared with others. However, age did not significantly predict brain complications. Gender was not a significant predictor of organ involvement. Sensibility, specificity and predictive values to predict organ complications of all aforementioned variables representing clinical symptoms, which merged into both clusters (ACR, ANG, CVR, HID) and (DIG, INT) are summarised in Table 2a–c. Using univariate analysis, dyshidrosis was a significant predictor for cardiac disease and brain complications (RR: 1.8; 5% CI 1.1–3.0 and 2.2; 5% CI 1.1–4.4 respectively), cardiac disease and kidney involvement were significantly, but weakly predicted by cornea verticillata (RR 1.7; 5% CI 1.0–2.9 and 1.9; 5% CI 1.0–3.6 respectively) and ANG significantly predicted for kidney disease (RR 2.1; 5% CI 1.2–3.8). Moreover, patients with hearing loss have an increased RR of 3.2 (5% CI 1.9–5.6) to present cardiac disease, of 3.9 (5% CI 2.0–7.7) to present kidney involvement and of 3.1 (5% CI 1.5–6.3) to present brain complications. Using multiple logistic regression, including the age of the patients, gender and the clinical symptoms (ACR, ANG, CVR, HID, DIG, INT), age and/or dyshidrosis appeared to be independent factors to explain the different organ involvements (Table 2a–c). After addition of the variable hearing loss in the analysis, cardiac disease was independently explained by age, dyshidrosis and hearing loss, whereas hearing loss was only the independent factor explaining kidney involvement and brain complication. Moreover, age and dyshidrosis were independent factors explaining the presence of at least one of the three organ complications, and after addition of hearing loss in analysis, age and hearing loss remained the sole independent factors with odds ratio of 7.1 (5% CI 2.6–19.2) and 5.8 (5% CI 2.2–15.2) respectively.

Table 2. Sensibility, specificity and predictive values of clinical symptoms for cardiac disease, kidney involvement and brain complications
Cardiac diseaseUnivariate analysisMultivariate analysis (hearing loss excluded)
VariablesSensibilitySpecificityOdds ratio (5% CI)VariablesOdds ratio (5% CI)Probability
Age: M > 25 years; F: > 35 years92.251.5 21.3 (5.4–83.1)Age29.1 (6.1–140)p < 0.0001
Male38.162.11.0 (0.5–2.2)Dyshidrosis4.3 (1.6–11.2)p = 0.003
Acroparesthesia69.034.81.2 (0.5–2.7)   
Angiokeratoma52.460.61.7 (0.8_3.7)   
Dyshidrosis61.962.1 2.7 (1.2–5.9)   
Exercise intolerance26.265.21.5 (0.6–3.5) Multivariate analysis (hearing loss included)
Gastrointestinal disorders28.669.71.1 (0.5–2.5)Age21.9 (4.4–109)p = 0.0002
Cornea verticillata66.754.5 2.4 (1.1–5.3)Hearing loss4.3 (1.5–11.5)p = 0.005
Hearing loss71.474.2 7.2 (3.1–16.9)Dyshidrosis3.0 (1.1–8.3)p = 0.04
Kidney involvementUnivariate analysisMultivariate analysis (hearing loss excluded)
VariablesSensibilitySpecificityOdds ratio (5% CI)VariablesOdds ratio (5% CI)Probability
Age: M > 25 years; F: > 35 years84.440.8 3.7 (1.3–10.3)Age4.0 (1.4–11.7)p = 0.01
Male43.864.51.4 (0.6–3.2)Dyshidrosis2.2 (0.9–5.3)p = 0.08
Acroparesthesia78.138.22.2 (0.9–5.6)   
Angiokeratoma62.563.2 2.9 (1.2–6.6)   
Dyshidrosis59.457.92.0 (0.9–4.6)   
Exercise intolerance34.469.71.2 (0.5–2.9)   
Gastrointestinal disorders28.669.71.1 (0.5–2.5) Multivariate analysis (hearing loss included)
Cornea verticillata66.754.5 2.4 (1.1–5.3)Hearing loss6.9 (2.7–17.7)p < 0.0001
Hearing loss71.474.2 7.2 (3.1–16.9)Dyshidrosis3.0 (1.1–8.3)p = 0.04
Brain complicationsUnivariate analysisMultivariate analysis (hearing loss excluded)
VariablesSensibilitySpecificityOdds ratio (5% CI)VariablesOdds ratio (5% CI)Probability
  1. Values in bold characters are statistically significant.

Age: M > 25 years; F: > 35 years56.126.91.6 (0.6–4.1)Dyshidrosis3.7 (1.4–9.8)p = 0.008
Male37.061.71.1 (0.4–2.5)   
Acroparesthesia63.032.11.2 (0.5–3.0)   
Angiokeratoma55.659.31.8 (0.8–4.3)   
Dyshidrosis66.759.3 2.9 (1.2–7.1)   
Exercise intolerance29.667.90.9 (0.4–2.3)   
Gastrointestinal disorders22.267.91.7 (0.6–4.5)   
Cornea verticillata66.754.5 2.4 (1.1–5.3) Multivariate analysis (hearing loss included)
Hearing loss71.474.2 7.2 (3.1–16.9)Hearing loss4.5 (1.7–11.6)p = 0.002
  • image(2)

[  Relationship between pair-wise clinical data. The intersection between a line and a column yields the result of the chi-squared function calculated for the two variables. Greyscale indicates the p-value of the chi-squared test: the darker the colour, the more interrelated the variables (chi-squared > 15.1: p < 0.0001; chi-squared > 10.8: p < 0.001; chi-squared > 6.6: p < 0.01; chi-squared > 3.8: p < 0.05). Abbreviations: ACR: acroparesthesia, ANG: angiokeratoma, CVR: cornea verticillata, DIG: digestive disorders, HID: dyshidrosis, HRT: heart disease, INT: exercise intolerance, KDN: kidney involvement, SHL: hearing loss, STR: cerebrovascular complications ]

Discussion

  1. Top of page
  2. Summary
  3. What’s known
  4. Introduction
  5. Material and methods
  6. Results
  7. Statistical analysis
  8. Discussion
  9. Funding
  10. Author contributions
  11. References

In this study, we report the prevalence of clinical signs and complications in a cohort of Fabry patients. We found interrelations among clinical symptoms as well as among major complications. Three clusters of clinical signs were identified, one notably grouping organ complications and hearing loss. Of particular interest is the findings that patients with hearing impairment - and less conclusively with dyshidrosis had higher risk of major complications.

Early symptoms of Fabry’s disease in children include acroparesthesia, hypohidrosis, heat and exercise intolerance and gastrointestinal disorders (1). After age 20 years, these symptoms tend to progress, while renal involvement, heart disease and cerebrovascular manifestations lead to substantial morbidity (1). Clinical symptoms are usually considered to be more prevalent and more severe in male Fabry patients than in female Fabry patients (1). As a result, prevalences of clinical symptoms or complications vary widely according to the age of the studied patients as well as the ratio of female patients and children included in the study. In the present cohort, prevalences of clinical symptoms or of complications were overall consistent with those reported in other series of similar sizes despite some minor differences. Pain has been reported in 62–82% of male patients and in 23–70% of female patients (3,8–13), dyshidrosis in more than 50% of male patients and 25% of female patients (1) and ANG in 31–93% of male patients and 12–63% of female patients (9,11–16). Gastrointestinal disorders have been evaluated using different means according to the various reports, with a prevalence ranging from 19% in male patients and 12% in female patients to more than 50% in both genders (9,13,16–18). On the other hand, we found only about 50% of corneal opacities in the present cohort, which was lower in comparison with the majority of previous reports (13,14,16,19,20), but similar to those noted in another French ophthalmological study (21) and in a recent report stemming from the large FOS database (18). Although acroparesthesia and especially ANG were significantly more frequent in male patient than in female patients in this study, we did not find any significant differences between genders for the other clinical symptoms, even though we did observe a trend towards more severe disease in male patients. Likewise, prevalences of renal and cardiac involvements as well as of cerebrovascular complications were similar between genders in our cohort. The higher mean age of female patients by comparison with that of male patients probably explains these results.

Three clusters of symptoms were underlined. Organ complications were found merged with hearing disorders in a first group of intercorrelated variables. Relationships between left ventricular hypertrophy and cerebral complications or proteinuria have been reported in one-third of male Fabry patients (13), with similar findings also observed in female Fabry patients (14). One noteworthy observation in the present study is that projections corresponding to organ complications in MCA were in proximity to the category ‘late diagnosis’. This indicates that, as expected, patients with late diagnosis often present organ complications. Gastrointestinal disorders and exercise intolerance appeared in a singular group and were not correlated with other variables. Finally, other clinical symptoms (ACR, ANG, CVR, HID) amalgamated in a third cluster. These symptoms, as a whole, were interrelated and weakly correlated with organ complications. This confirms the conclusions of a previous study, which found only limited relationship between various disease manifestations (13), and could be interpreted in terms of disease evolution. Indeed, clinical symptoms are first to appear in the time course of the disease, but do not presuppose organ complications, which occur much later. Nevertheless, some clinical symptoms significantly predict organ complications in the present study: as instance Fabry patients with dyshidrosis had 1.8 times higher risk of presenting cardiac disease and 2.2 times higher risk of demonstrating brain complications. This indicates that dyshidrosis probably reflects more severe disease. In the same way, the presence of ANGs may be correlated with the severity of systemic manifestations (15), and our results corroborate these findings, as in this study, patients with ANGs had 2.1 times higher risk of presenting kidney involvement. From a practical standpoint, this means that a physician who establishes a diagnosis of ANGs or dyshidrosis in a given Fabry patient should systematically and totally screens the latter for cardiac, brain and kidney complications.

Also, the present study highlights hearing disorders as predictors of organ complications and consequently of disease severity. Otological symptoms are common in patients with Fabry’s disease, although they are not considered as life threatening (22). Hearing disturbances were found in almost half of male patients and in 43.5% of female Fabry patients in the present study, a result similar with those reported in earlier studies (23–25). We found that Fabry patients with hearing disorders had three to four times higher risk of presenting organ complications. These results emphasise those found in a small cohort of 22 male patients, in which hearing disorders were found to be correlated with kidney failure and cerebrovascular lesions (24). Furthermore, Reas and collaborators reported that hearing loss in male patients was associated with higher microvascular cerebral white matter lesion load, lower kidney function and lower residual alpha-galactosidase A activity (25).

Indeed, Fabry’s disease is not only a storage disorder as demonstrated by Gb3 deposits in endothelial cells, but pathophysiology of Fabry’s disease also includes ischaemic features in central nervous system, heart and kidney, as well as fibrosis, such as myocardial fibrosis or glomerulosclerosis (26,27). The vasculopathy associated with Fabry’s disease may be considered as a ‘peudovascultis’ of the small vessels. Thus, foamy vacuolised cytoplasm of endothelial cells lining the internal auditory artery has been documented (28).

Inherent limitations of this study lie in the design based on a transversal analysis of patients recruited only in departments of internal medicine. This bias may have ruled out some patients with very severe disease, such as those with end-stage renal failure, who are generally followed in subspeciality departments. Another limitation of the study resides in the fact that the clinical picture of Fabry’s disease was not fully examined. However, the main clinical symptoms involved in disease prognosis were considered. Moreover, binary variables were used in this study, and this could be debated. Extent of ANG, severity of pain, or deepness of hearing loss are often put forward to evaluate disease severity. In the same way, the definition of cardiac disease in this study included cardiac hypertrophy, which could be a relatively stable and benign condition as well as severe cardiomyopathy complicated by heart failure. Also, the definition kidney involvement used in this study merged moderate proteinuria with normal glomerular filtration rate and severe renal failure. Scoring systems for patients with Fabry’s disease have been developed to take into account severity of symptoms (29). However, even though such scores are useful in clinical research, they appeared to be too complex in clinical practice. Taking charge of a Fabry patient, and making a decision to treat or not is usually based on a binary reasoning, such as the presence or not of cardiac hypertrophy, or of documented renal disorder.

Moreover, multidimensional analyses, such as MCA, are usually applied to the study of a large number of individuals or variables. However, such statistical analyses were historically developed and mathematically validated in relatively small array of data (5), as limited by the power of the available computer. In this study, more than 85% of total variance was explained by only two factorial axes, indicating that our data could consequently be reduced to a two-dimensional model with good approximation, thus implying that the relatively small size of our cohort was sufficient for such analyses.

In summary, among the various clinical symptoms occurring in patients with Fabry’s disease, ANG, dyshidrosis and especially hearing disorders appear as singular signs, which should alert clinicians to possible organ involvement and disease severity. This study also suggests that hearing loss could be interpreted as a major clinical symptom in Fabry’s disease. Otological examination must be systematically performed in Fabry patients.

Author contributions

  1. Top of page
  2. Summary
  3. What’s known
  4. Introduction
  5. Material and methods
  6. Results
  7. Statistical analysis
  8. Discussion
  9. Funding
  10. Author contributions
  11. References

Study concept and design: Kaminsky, Lidove. Acquisition of data: Kaminsky, Noel, Jaussaud, Leguy-Seguin, Hachulla, Zenone, Lavigne, Marie, Maillot, Masseau, Serratrice, Lidove. Analysis and interpretation of data: Kaminsky, Noel, Jaussaud, Leguy-Seguin, Hachulla, Zenone, Lavigne, Marie, Maillot, Masseau, Serratrice, Lidove. Drafting of the manuscript: Kaminsky, Lidove. Critical revision of the manuscript for important intellectual content: Kaminsky, Noel, Jaussaud, Leguy-Seguin, Hachulla, Zenone, Lavigne, Marie, Maillot, Masseau, Serratrice, Lidove. Approval of submitted version Kaminsky, Noel, Jaussaud, Leguy-Seguin, Hachulla, Zenone, Lavigne, Marie, Maillot, Masseau, Serratrice, Lidove. Administrative, technical, and material support: Kaminsky, Lidove. Study supervision: Kaminsky, Lidove.

References

  1. Top of page
  2. Summary
  3. What’s known
  4. Introduction
  5. Material and methods
  6. Results
  7. Statistical analysis
  8. Discussion
  9. Funding
  10. Author contributions
  11. References
  • 1
    Zarate YA, Hopkin RJ. Fabry’s disease. Lancet2008; 372: 142735.
  • 2
    Brady RO, Murray GJ, Moore DF, Schiffmann R. Enzyme replacement therapy in Fabry disease. J Inherit Metab Dis2001; 24 (Suppl. 2): 1824.
  • 3
    Desnick RJ, Brady R, Barranger J et al.Fabry disease, an under-recognized multisystemic disorder: expert recommendations for diagnosis, management, and enzyme replacement therapy. Ann Intern Med2003; 138: 33846.
  • 4
    Maron BJ, McKenna WJ, Danielson GK et al.CC/ESC clinical expert consensus document on hypertrophic cardiomyopathy: a report of the American College of Cardiology Task Force on Clinical Expert Consensus Documents and the European Society of Cardiology Committee for Practice Guidelines (Committee to Develop an Expert Consensus Document on Hypertrophic Cardiomyopathy). Eur Heart J2003; 24: 196591.
  • 5
    Benzécri JP. L’Analyse des données, Tome 1: La taxinomie, Tome 2: l’analyse des correspondances, 2nd edn. Paris: Dunod, 1976.
  • 6
    Greenacre MJ. Correspondence analysis in medical research. Stat Methods Med Res1992; 1: 97117.
  • 7
    Lebart L, Morineau A, Piron M. Statistique exploratoire multidimensionnelle, 3rd edn. Paris: Dunod, 2000.
  • 8
    Deegan PB, Baehner AF, Barba Romero MA et al.Natural history of Fabry disease in females in the Fabry Outcome Survey. J Med Genet2006; 43: 34752.
  • 9
    Eng CM, Fletcher J, Wilcox WR et al.Fabry disease: baseline medical characteristics of a cohort of 1765 males and females in the Fabry Registry. J Inherit Metab Dis2007; 30: 18492.
  • 10
    Hoffmann B, Beck M, Sunder-Plassmann G et al.Nature and prevalence of pain in Fabry disease and its response to enzyme replacement therapy – a retrospective analysis from the Fabry Outcome Survey. Clin J Pain2007; 23: 53542.
  • 11
    MacDermot KD, Holmes A, Miners AH. Anderson-Fabry disease: clinical manifestations and impact of disease in a cohort of 60 obligate carrier females. J Med Genet2001; 38: 76975.
  • 12
    MacDermot KD, Holmes A, Miners AH. Anderson-Fabry disease: clinical manifestations and impact of disease in a cohort of 98 hemizygous males. J Med Genet2001; 38: 75060.
  • 13
    Vedder AC, Linthorst GE, van Breemen MJ et al.The Dutch Fabry cohort: diversity of clinical manifestations and Gb3 levels. J Inherit Metab Dis2007; 30: 6878.
  • 14
    Gupta S, Ries M, Kotsopoulos S, Schiffmann R. The relationship of vascular glycolipid storage to clinical manifestations of Fabry disease: a cross-sectional study of a large cohort of clinically affected heterozygous women. Medicine (Baltimore)2005; 84: 2618.
  • 15
    Orteu CH, Jansen T, Lidove O et al.Fabry disease and the skin: data from FOS, the Fabry outcome survey. Br J Dermatol2007; 157: 3317.
  • 16
    Galanos J, Nicholls K, Grigg L et al.Clinical features of Fabry’s disease in Australian patients. Intern Med J2002; 32: 57584.
  • 17
    Hoffmann B, Keshav S. Gastrointestinal symptoms in Fabry disease: everything is possible, including treatment. Acta Paediatr Suppl2007; 96: 846.
  • 18
    Mehta A, Clarke JT, Giugliani R et al.Natural course of Fabry disease: changing pattern of causes of death in FOS – Fabry Outcome Survey. J Med Genet2009; 46: 54852.
  • 19
    Nguyen TT, Gin T, Nicholls K et al.Ophthalmological manifestations of Fabry disease: a survey of patients at the Royal Melbourne Fabry Disease Treatment Centre. Clin Experiment Ophthalmol2005; 33: 1648.
  • 20
    Sodi A, Ioannidis AS, Mehta A et al.Ocular manifestations of Fabry’s disease: data from the Fabry Outcome Survey. Br J Ophthalmol2007; 91: 2104.
  • 21
    Orssaud C, Dufier J, Germain D. Ocular manifestations in Fabry disease: a survey of 32 hemizygous male patients. Ophthalmic Genet2003; 24: 12939.
  • 22
    Keilmann A, Hegemann S, Conti G, Hajioff D. Fabry disease and the ear. In: Mehta A, Beck M, Sunder-Plassman G, eds. Fabry Disease: Perspectives from 5 years of FOS. Oxford: Oxford Pharmagenesis, 2006. Available from: http://www.ncbi.nlm.nih.gov/books/NBK11606/.
  • 23
    Conti G, Sergi B. Auditory and vestibular findings in Fabry disease: a study of hemizygous males and heterozygous females. Acta Paediatr Suppl2003; 92: 337.
  • 24
    Germain DP, Avan P, Chassaing A, Bonfils P. Patients affected with Fabry disease have an increased incidence of progressive hearing loss and sudden deafness: an investigation of twenty-two hemizygous male patients. BMC Med Genet2002; 3: 10.
  • 25
    Ries M, Kim HJ, Zalewski CK et al.Neuropathic and cerebrovascular correlates of hearing loss in Fabry disease. Brain2007; 130: 14350.
  • 26
    Beer M, Weidemann F, Breunig F et al.Impact of enzyme replacement therapy on cardiac morphology and function and late enhancement in Fabry’s cardiomyopathy. Am J Cardiol2006; 97: 15158.
  • 27
    Weidemann F, Niemann M, Breunig F et al.Long-term effects of enzyme replacement therapy on fabry cardiomyopathy: evidence for a better outcome with early treatment. Circulation2009; 119: 5249.
  • 28
    Schachern PA, Shea DA, Paparella MM, Yoon TH. Otologic histopathology of Fabry’s disease. Ann Otol Rhinol Laryngol1989; 98: 35963.
  • 29
    Beck M. The Mainz Severity Score Index (MSSI): development and validation of a system for scoring the signs and symptoms of Fabry disease. Acta Paediatr Suppl2006; 95: 436.