The prevalence of Fabry disease in a statewide chronic kidney disease cohort – Outcomes of the aCQuiRE (Ckd.Qld fabRy Epidemiology) study

Fabry Disease (FD) is a rare genetic condition caused by the absence or deficiency of the enzyme alpha-galactosidase A (α-Gal A), which leads to build up of glycosphingolipids, in particular globotriaosylceramide (GB3), throughout the body in a variety of cell types and tissues [1]. FD affects multiple systems in the body and symptoms and signs vary in type and severity amongst patients. Major causes of morbidity and mortality amongst patients affected by FD include renal, cardiac and cerebrovascular manifestations [2‐4]. Other affected systems include skin, ocular, nervous, gastrointestinal, respiratory, lymphatic and skeletal (1). FD has also been found to be associated with poorer quality of life (QOL) [5‐7]. Although an X-linked disorder affecting the GLA gene [8, 9], females have also been shown to be affected [10‐12], with 69.4 to 80.0% of affected females reporting some degree of symptoms [13, 14] with greater variability of severity and age at onset compared to males with FD [15].

Overall population prevalence of FD is estimated at 0.04% (range = 0.0 to 0.4%) [16]. Large scale screening studies of neonates have estimated prevalence at between 0.013% [17] and 0.05% [18] for males and 0.0% [18] and 0.002% [19] for females, though some populations experience increased prevalence of particular disease-causing alleles sich as Taiwan. Overall prevalence of FD in haemodialysis populations world-wide is estimated to be 0.12% [20, 21] to 0.36% [22]. In a systematic review of FD screening in dialysis populations, Linthorst, et al. [23], estimated prevalence at 0.33% for males and 0.1% for females. Doheney, et al. [24], conducted a reanalysis of 63 studies that estimated prevalence of FD in haemodialysis, kidney transplant, stroke & cardiac populations with GLA mutations (total n = 36,820 patients; 23,954 males, 12,866 females). Revised prevalence estimate for haemodialysis patients was 0.21% in males and 0.15% in females. Revised prevalence rates for kidney transplant patients based upon α-Gal A enzymatic screening was 0.24% in males and 0.00% in females.

The prevalence of FD within the broader CKD population is less well studied. Two Turkish studies of pre-terminal CKD patients (i.e. patients with CKD stages 1 through 5 [25] and not on kidney replacement therapy (KRT) identified FD in 1.8% [26] and 0.4% [27] male patients but not in female patients [26, 27]. Favalli, et al. [28], examined the prevalence of FD in patients selected from multiple settings and also identified FD in male patients (2.7%) but not female patients. Sample sizes for these studies ranged from 72 [28] to 1453 [27]. Lack of identified cases in female patients may be due to true lack of undiagnosed female FD cases, less sensitive screening methods, or insufficient sample size.

FD can be identified through screening, including Dried Blood Spot (DBS) testing to analyse blood levels of the α-galactosidase A enzyme, lyso-GB3 testing and genetic testing [29]. Whilst DBS-based non-genetic screening is imperfect amongst women, the feasibility of a direct genetic sequencing screening approach is not feasible in current local operational settings. As an X-linked disorder with variable phenotype amongst women, we anticipate that there might be two genetically affected women in the broader community for every male identified. Several treatments are available for FD, including Enzyme Replacement Therapy (ERT). Two preparations of ERT shown to stabilise or improve FD symptomatology [30, 31] are currently funded under the Life Saving Drugs Program in Australia for FD patients.

The Ckd. Qld fabRy Epidemiology (aCQuiRE) study [32] targeted a significantly larger screening sample of 3000 CKD patients. We aimed to identify FD in patients with CKD in public, specialty nephrology practices.

The detailed aCQuiRE study protocol has been previously published [32].

A total of 3000 CKD patients were screened across seven sites. Eight patients withdrew consent after screening. Of the remaining 2992 screened patients, six (0.20%) patients were diagnosed with FD (Supplementary Table 1), 2902 (96.99%) patients did not have FD, and 84 (2.81%) patients received inconclusive results (See Fig. 1).

Six cases of FD were identified among 2992 adult, male and female CKD patients. This is an overall estimated FD prevalence of 0.20%, which is consistent with previously published estimates for dialysis populations and a working hypothesis of 0.2%. Twenty-eight at-risk family members were identified (4–7/FD case identified) who would benefit from family screening.

Of the six CKD patients diagnosed with FD, 5 were male (0.29%) and 1 was female (0.8%). The identification of a case in a female patient highlights the importance of including women in a FD screening program to enable a true estimate of prevalence and for targeted treatment and management. Three of the six cases of FD had been previously identified with all three receiving enzyme replacement therapy. Exclusion of these previously identified cases would have potentially underestimated prevalence.

An upper age limit was not stipulated and all FD cases in the adult CKD population had the possibility of being included. The ages of patients diagnosed with FD ranged from 23 to 70 years, indicating FD as a potential causative factor for CKD patients of all ages. The CKD patients diagnosed with FD were significantly younger than the CKD patients without FD, consistent with persons with genetically driven kidney disease and those with glomerulonephritis in the broader adult CKD.QLD population having a younger age at diagnosis than those with CKD due to diabetes and renovascular disease (42).

FD cases were identified in patients with early through to late-stage kidney disease, though all newly identified cases had already progressed to KF and were receiving KRT. While four patients with FD were receiving KRT, two patients were in earlier stages (Stage 1, Stage 3B), when ERT may be of most benefit [38]. Therefore, this study supports a potential benefit to screening patients in all stages of CKD. No particular pattern of symptoms was identified. This may be due to the small number of cases identified, however is consistent with research that indicates symptom presentation varies among patients [1]. Widespread screening of CKD patients for FD may be beneficial.

Genetic testing is more sensitive for diagnosis of FD but much more costly and time-consuming compared to DBS testing and was not feasible for a large-scale pragmatic screening study such as aCQuiRE. DBS testing of α-Gal A was chosen given cost, time and efficiency. Some issues with the quality of DBS testing were experienced for a proportion of patients. Given that there are known challenges with sole use of α-Gal A testing amongst potentially heterozygous women, this is likely to have impacted underestimation of prevalence amongst females. Other limitations of this study include that 2.81% of participants had an unresolved inconclusive result, that in-depth biochemical data collection from clinial pathology results was not performed, and that individual family pedigrees were not recorded for participants.

Prospective learnings from other FD screening studies were integrated, including sample collection pre-heparinisation for participants receiving haemodialysis [33]. Where still required, cascade testing, including a repeat DBS sample and/or plasma lyso-GB3, resolved sampling issues. There are additional emerging and novel screening strategies for FD as well, including urinary studies [39]. The anticipated number of genetically affected women likely to be present in the community for every identified affected male is 2:1. As such, in this study there may statistically be 9 such genetically affected women present but not identified by the study screening methodology. This might hypothetically be due to the variable FD phenotype experienced by genetically affected women, resulting in a CKD phenotype that may not have yet or ever become present. Further, all three newly identified FD cases had previously undergone kidney biopsy without their clinical diagnosis being revealed. This highlights the need for close and fortified consideration of electron microscopy or toluene blue staining of semi-thin sections as part of kidney biopsy evaluation.

While it is not possible to make statistical conclusions, given the small number of FD cases identified, our results can inform estimates of cost-effectiveness of screening of patients of all ages in adult CKD populations, patients from all stages of CKD and the inclusion of females in screening programs. Sample quality should be consistently monitored. The Queensland Statewide FD Treatment Service identified 28 at-risk family members. These family members were eligible for screening and genetic counselling, with anecdotal evidence that many accessed this. Benefits of wide-spread screening of CKD populations for FD is likely to extend benefits to wider populations through family screening. Initially however, our findings suggest that for every currently identified patient with FD complicated by CKD, there may be a further FD case affected by CKD that is yet to be diagnosed.