• rheumatoid arthritis
• gene polymorphism
• epidemiology

Fabry disease (FD) is a rare lysosomal storage disorder caused by mutations of the alpha-galactosidase A gene (GLA) which causes premature morbidity due to organ dysfunction following deposition of globotriaosylceramide (Gb3).1 Presenting symptoms include musculoskeletal pain such as acroparesthesias2 which often result in rheumatological consultations.3 Concerns have been raised that patients with FD are often falsely attributed with classical rheumatological diagnoses, precluding early effective treatment.3

We therefore performed GLA sequencing (Centogene, Rostock, Germany) in a multicentre national cohort of 798 patients with early undifferentiated arthritis (500 female, mean follow-up 19.5±7.4 months) (Course And Prognosis of Early Arthritis study, CAPEA,4 ethics approval 3368). Patients aged above 18 years with clinical arthritis for ≤26 weeks in ≥2 joints, or 1 joint with morning stiffness >30 min were eligible for inclusion.

As detailed in table 1, eight heterozygous variants in female patients were found, none of which was considered to be pathogenic for FD (resulting prevalence in women [age 55.3±14.7 years]: 1.6%). Globotriaosylsphingosine values were tested in these patients by tandem mass spectrometry (Centogene) to further exclude active disease.5 Seven of the patients were diagnosed with rheumatoid arthritis (RA) and one patient with systemic sclerosis. In order to assess whether the presence of GLA variants (exposition) were associated to a distinct phenotype (outcome), patients with RA with GLA variants were compared with patients with RA without GLA variants. Generalised linear mixed modelling with a random intercept for matched cases was performed after 1:3 propensity-score matching including age, sex and disease duration (matching variables) within the patients diagnosed with RA (n=660 including 7 GLA-variant cases described in table 1). The presence of a GLA variants was associated to a reduced swollen joint count (OR 0.085, p=0.047), but not to a higher frequency of symptoms typically associated to FD, that is, congestive heart failure, renal insufficiency, pulmonary disease, gastrointestinal complaints (all p=1.0), cerebral ischaemia (p=0.76) or a combination thereof (p=0.98). In a longitudinal analysis (follow-up 23.1±2.3 [GLA variants] and 20.3±6.7 months [wild type]), cases with GLA variants favourably responded to standard-of-care treatment for RA with reduction in the DAS28 (OR 0.92, p<0.001), Simplified Disease Activity Index (OR 0.49, p<0.001) and Clinical Disease Activity Index (OR 0.54, p<0.001). These trends over time were also observed in patients with RA without GLA variants. Interestingly, patients carrying GLA variants also showed improvements in several items including joint function (‘Health Assessment Questionnaire’, OR 0.98, p=0.02), tender joint count (OR 0.78, p<0.001) and duration of morning stiffness (OR 0.08, p=0.01). The latter items were not significantly improved in patients with RA without GLA variants.

Our data thus suggest that, at least under the present inclusion criteria (≥1 clinically swollen joint), FD is unlikely to be systematically overlooked in clinical rheumatology practice in a large national cohort. It has to be stressed, however, that rheumatologists should nevertheless be alert to the rare differential diagnosis of FD as they are often among the first specialists consulted due to acral pain.2 3 FD should especially be considered in individuals complaining about acral pain without any confirmed joint swelling. Helpful clinical signs include angiokeratomas, multiorgan dysfunction, especially chronic heart failure, renal insufficiency, stroke and a family history thereof. Of note, X-chromosomal heritage results in a preponderance for more severe symptoms in men,6 which may explain the lack of male cases in our study. Interestingly, patients with GLA variants improved in more items compared with wild-type RA. We are not aware of any systematic differences of treatment strategies in either group. So far, GLA variants have not been identified as modifiers of RA disease severity; verification in independent cohorts is hence warranted.