Fabry Nephropathy: Indications for Screening and Guidance for Diagnosis and Treatment by the European Renal Best Practice

Wim Terryn; Pierre Cochat; Roseline Froissart; Alberto Ortiz; Yves Pirson; Bruce Poppe; Andreas Serra; Wim Van Biesen; Raymond Vanholder; Christoph Wanner


Nephrol Dial Transplant. 2013;28(3):505-517. 

In This Article

Epidemiology and the Need for Screening

1.1 We do not recommend screening in the general population. (Ungraded statement)

1.2 We recommend obtaining informed consent from the patient before screening, using an information form drafted in collaboration with a clinical geneticist. (Ungraded statement)

1.3 We recommend screening for FD in male chronic kidney disease (CKD) patients below 50 years of age in whom a reliable renal diagnosis is absent. (Ungraded statement)

1.4 We suggest screening for FD in females with unexplained CKD, irrespective of age, with other unexplained symptoms potentially associated with FD. (Ungraded statement)

1.5 We recommend discussing with the patient the implications of diagnosing a genetic disease and the possible implications for the at-risk relatives. (Level 1C)


Classical FD is a progressive multisystem disease predominantly presenting in males, characterized by angiokeratoma, hypohidrosis and acroparesthesia (neuropathic pain) in childhood, followed by renal failure, left ventricular hypertrophy (LVH), stroke and premature death in the fourth or fifth decade of life.[8] In male patients, levels of α-Gal A activity are classically very low or undetectable. However, as a result of screening studies during the past decade, clinical variants of FD in male patients with varying degrees of residual activity of α-Gal A have been described. The first described was the 'cardiac variant' with isolated LVH and/or cardiomyopathy presenting in the sixth or seventh decade, lacking the classical disease symptoms and time course.[9,10] Patients suffering from this variant may have proteinuria, but their renal function is typically normal for their age. Later a 'renal variant' phenotype was described in a screening study in a dialysis population, where patients again were lacking the classical manifestations. This phenotype was described as 'intermediate' between the cardiac variant and the classic phenotype.[11] These patients with cardiac and renal variants are called 'atypical' or 'attenuated' FD patients. The genetic basis of this variable penetrance and expression is unclear. It is believed that the atypical cases are the result of missense mutations that encode mutant enzyme protein or intronic lesions that reduce transcript levels, both resulting in a reduced but significant residual enzyme function (1–12% of normal),[12] although this has been debated, and others found no genotype–phenotype correlation.[13] Heterozygous women, in spite of having a mutation compatible with typical disease, can also present this attenuated phenotype as it was hypothesized that skewed X-inactivation can result in significant residual enzyme function. However, it must be stressed that most females have the classical phenotype, but with a delayed and/or milder presentation of symptoms.[14]

As a consequence, reported prevalence varies with the population studied and the test used for screening, and genetic screening might find female index cases that are not found by enzyme-based methods.[15] The prevalence of classical FD has been estimated at 1 in 117 000 births[14] and 1 in 40 000 males.[8] In several screening studies in high-risk populations, the frequency was up to 1% or even higher, especially in populations with unexplained LVH.[16] In newborns,[17–19] the incidence of α-Gal A deficiency was 1 in 3100 with an 11 to 1 ratio of patients with the later-onset versus the classic phenotype. In the haemodialysis population, a prevalence of 0.33% in male and 0.10% in female patients has been found in a cross-sectional screening study.[16] Only two studies screened kidney transplant patients. In cryptogenic stroke, a prevalence of 0.8%[20] up to 2.4%[21] and 3.9%[22] was found; however, in the second study,[21] half of the patients had the p.D313Y mutation, which is now generally regarded as a pseudo-deficiency, and in the last study,[22] the specific mutations were not mentioned and could also have been polymorphisms. Many screening studies are not conclusive for the female population, as they most frequently used α-Gal A activity screening, which is in women, as described above, not a sensitive screening tool.

Although there are no studies in the CKD population not on dialysis, we recommend screening for FD in patients with CKD without a clear diagnosis. In classical FD, most males reach CKD Stage 5 or die before the age of 50.[12,23] As a consequence, we recommend screening in males only below the age of 50 years. We recommend screening even in the case of a negative family history as de novo mutations can occur, and the family history is not always suggestive for FD, given the broad phenotypic spectrum of the disease. Arterial hypertension should not be an exclusion criterion as more than 50% of FD patients have mild to moderate hypertension, especially when estimated glomerular filtration rate (eGFR) is <60 mL/min/1.73 m2.[23–25] In women, disease onset can be later, so when there is unexplained kidney disease associated with manifestations suggestive of FD, we suggest screening for FD regardless of age.

The real prevalence should be derived from screening in the healthy population at a young age; this has been done in four studies in newborns.[17–19,26] However, this approach remains problematic for several reasons. The American College of Medical Genetics (ACMG) has proposed newborn screening for 29 disorders, but screening for FD was not included in this list (available online at: https://mchb.hrsa.gov/screening/). Although measurement of α-Gal A has a good sensitivity and specificity in males, it has a low positive predictive value in the healthy population. This will result in unnecessary expensive tests. In addition, the majority of the detected cases in the newborn studies are 'atypical' mutations, giving an attenuated phenotype or a cardiac variant. The finding of a genetic predisposition for a possible late-onset disease where the treatment effectiveness is unclear has ethical and legal implications that constrain a systematic screening of newborns. In these cases, it would be difficult to decide on ERT, as the natural history of patients carrying atypical mutations is poorly characterized, effects of ERT in mild cases have not been studied, and a lifelong treatment is a psychological burden for the patient and a financial one for both the individual and society with, on top of that, uncertain results. As a consequence, we do not recommend screening for FD in the general population.

As FD is an X-linked disease with variable but significant morbidity both in males and females, its diagnosis might have profound consequences for the proband and his relatives. As a consequence, we recommend obtaining informed consent from the proband before screening, when possible in cooperation with an expert in genetic counselling. (Example in Supplementary appendix.)

Once the diagnosis is made, it is important to make up a pedigree in order to identify all relatives at risk. FD is an X-linked disease where all carriers can be symptomatic. It should be kept in mind that 'skipping' of a generation is possible because of variable expression.

The patient should receive further guidance in communication with his family. He must be able to provide sufficient information (e.g. by using flyers written by the treating team), and one must anticipate a number of possible problems in the communication with his family. Some people do not want a work-up to the diagnosis of FD, and it should be explained to the patient that they do have the right not to know their genetic status.