Fabry disease (FD) is a X-linked systemic disease characterized by the accumulation of globotriaosylceramide (Gb-3) in the extracellular space secondary to decreased activity of the lysosomal enzyme alpha-galactosidase-A (α-gal-A).1 Inflammation and subsequent fibrosis play an essential role in developing and progressing nephropathy in FD. The main culprits of the fibrotic process in Fabry nephropathy are still unclear. Previous studies showed that the deposition of Gb-3 and globotriaosylsphingosine (lyso-Gb3) may be responsible for these undesirable results.2 While glycosphingolipid accumulation is initially reversible, as the accumulation increases over time, it causes irreversible tissue injury and organ dysfunction. In Fabry nephropathy, glycosphingolipid accumulate in tubular and vascular structures, mainly glomerular structures, and progress to the fibrosis process which leads to chronic kidney injury over time.3 Early enzyme replacement therapy (ERT) slows the deterioration of kidney function, but ERT cannot equally affect the progression of glomerulosclerosis or proteinuria in all patients and often cannot prevent progressive kidney injury, especially in patients with proteinuria greater than 1g/day.4 To prevent kidney injury in FD, the search for new treatments continues.

Periostin (osteoblast-specific factor 2) is an extracellular matrix protein expressed in various tissues during embryonic development and mostly synthesized from collagen-rich connective tissues in adults.5 While it is expressed extensively in renal tissues during nephrogenesis, its expression is very low in healthy renal tissues in adults and increases significantly after kidney injury.6 Previously, it has been shown that there is periostin expression from cyst epithelium in autosomal dominant polycystic kidney patients, periostin expression in interstitial areas and tubule epithelium, especially in fibrotic areas in glomerulopathies, and increased periostin expression in chronic kidney disease patients.7 The proposed mechanism in the relationship between periostin and kidney injury is that periostin mediates and amplifies kidney inflammation and fibrosis in the kidney epithelium. As a result of all studies, it was concluded that increased expression of periostin is associated with kidney injury, and deficiency or inhibition of periostin is associated with better preservation of kidney structure and function.

In FD, primary factors that initiate kidney injury and secondary factors that continue and advance the process to fibrosis are different. Therefore, replacing only the deficient enzyme may not be able to stop kidney injury in all patients. We planned this study based on our hypothesis that periostin is one of the perpetuating factors causing kidney injury. If we know the secondary factors that cause kidney injury, we can reach new treatment options and be more successful in treating Fabry nephropathy.

This is a cross-sectional study. Ethics committee approval was obtained from our institution (2018/1266). Written informed consent was obtained from FD patients and control patients.

The study included eighteen FD patients (10 males, 8 females) diagnosed between 2014 and 2020. Twenty-two healthy adults (12 males, 10 females) with demographic characteristics similar to the FD patient group were included in the study as the control group. At the time of diagnosis, 5 cc venous blood samples were taken from all FD patients, centrifuged, and serum samples were stored at −80°C. All FD patients included in the study had an indication for ERT. When the target number of patients was reached (June 2022), serum periostin levels were studied in serum samples stored at −80°C. Plasma α-gal-A and lyso-Gb3 levels at the time of diagnosis were recorded from the hospitals’ system. In addition, a review of medical records (including information on age, sex, weight, height, disease duration, medications, history of other diseases, proteinuria and kidney function tests at the time of diagnosis) was undertaken from our hospitals’ system.

Inclusion criteria were (1) 18–70 years of age (2) decreased (<2.5nmol/ml/h) α-gal-A activity in dried blood spots (DBS) in male patients, (3) presence of GLA gene mutation associated with FD in female patients, (4) patients with an indication for initiation of ERT due to FD. Exclusion criteria were (1) Active infection, (2) presence of diabetes mellitus, (3) presence of glomerulonephritis, lupus nephritis, systemic vasculitis proven by kidney biopsy. Inclusion criteria as a control group were 18–70 years of age, non-smoker, and no history of chronic disease. Having a FD family history, smoking, a history of chronic disease, and having signs of acute infection at the time of blood tests were considered exclusion criteria for control groups.

We found some important results in the study in which serum periostin levels were evaluated in Fabry patients. First, serum periostin was negatively correlated with age of the first symptom and GFR; and positive correlated with proteinuria and lyso-Gb3. The other significant result, the only independent predictor of proteinuria, one of the most critical kidney failure progression indicators, was serum periostin, and we found that patients with high proteinuria had high serum periostin levels. To the best of our knowledge, our study is the first to examine the relationship between serum periostin and FD, and for the first time in our research, it was shown that serum periostin is higher in FD patients than in healthy individuals.

We found that age of the first symptom and GFR were negatively correlated with serum periostin. Renal expression of periostin increases in correlation with increasing age and decreased kidney function.8 FD patients are divided into classical and atypical FD patients according to their clinical findings, and symptom onset and organ dysfunctions are earlier in patients with classical FD.9 The mean age of symptom onset in our FD patient group is 27, and we can consider that all of our patients are late-onset. Periostin is a marker that increases in inflammatory processes; we may think that inflammation is more in FD patients with early onset of symptoms, and therefore periostin increases.10 Periostin levels increase in chronic kidney disease.11 GFR and periostin are expected to be negatively correlated in FD patients, like in our study. However, only 5 (27.7%) of the patients in our study had a GFR below 60ml/min. Most of our patients consisted of patients with normal/near-normal GFR and whose kidney functions were not severely reduced shows us that a process independent of chronic kidney disease is observed in the relationship between serum periostin and FD.

The expression of periostin in healthy and damaged kidneys and its role on kindey function are not yet fully understood. While its expression in healthy kidney tissues is extremely low in adults, periostin, which is re-expressed in renal tissues in case of inflammatory injury, has an important role in the process of renal inflammation and fibrosis.12,13 Many studies have shown that renal periostin expression increases in different patient groups, and renal inflammation and fibrosis decrease with measures to reduce the level of periostin.12–16 Our study showed for the first time that periostin levels are higher in FD patients than in the healthy population.

There are similarities between the pathogenesis of diabetic nephropathy and Fabry nephropathy. Annual GFR losses are similar, kidney injury begins as proteinuria in both diseases, and the presence of secondary factors that continue the injury after the primary metabolic components that initiate renal impairment are similar in the two diseases. Even after the metabolic components that initiated the injury are eliminated, the injury mechanism may continue to progress in both disorders. Following early podocyte injury in both diseases, the fibrotic process by extracellular matrix elements begins.17 In FD, levels of lyso-Gb3, an active lipid metabolite, are elevated, and lyso-Gb3 induces the proliferation of vascular smooth muscle cells.2 Previously, the secondary pathway inducing renal fibrosis in FD was shown to be TGF-β activation following lyso-Gb3 deposition.17,18 ERTs reduce endothelial lyso-Gb3 levels, but sphingolipid deposits and their clinical repercussions are not completely normalized.2 In addition, starting ERT after the onset of tissue fibrosis does not regress the fibrosis and does not completely prevent organ dysfunction.4 While all these data confirm the role of sphingolipid deposits as the primary trigger causing FD nephropathy, the secondary pathways that still sustain the injury are not clearly known. The fact that both sphingolipid depositions in Fabry nephropathy and periostin expression during kidney injury begin first in the same kidney areas (glomerular areas) raises doubts about the possible role of periostin in Fabry nephropathy.13 The idea that periostin may be a secondary mechanism that induces TGF-β and perpetuates kidney injury caused by sphingolipid accumulation in FD patients seems to be a plausible explanation for the mechanism of worsening kidney function in these patients. We think that periostin may also have an essential role in answering the question of why ERTs do not prevent progression to end-stage kidney disease in some patients.

Although no correlation can be established between serum or urine Gb3 levels and kidney dysfunction, there is a relationship between heart and brain lesions and serum free sphingolipid levels.18 These circulating molecules, primarily lyso-Gb3, are the main responsible factors for initiating the process of kidney dysfunction that causes vasculopathy and vascular remodeling in FD, thus resulting in renal fibrosis.17 In our study, we found that lyso-Gb3 and serum periostin was correlated. In patients with high Lyso-Gb3 levels, an already elevated inflammatory state is expected. Lyso-Gb3 increases the expression of TGF-β and macrophage inhibitory factor receptor CD74, which causes fibrosis in podocytes.17 Periostin-induced TGF-β-mediated fibrosis process following the inflammatory pathway initiated by lyso-Gb3 may be a new hypothesis for the development of Fabry nephropathy.

In Fabry nephropathy, glomerular sclerosis and fibrosis process has already begun in the kidney before significant proteinuria or kidney dysfunction is reflected in the clinic.19 Considering the lack of correlation of circulating Gb3 accumulations with kidney dysfunction, there is a need for valuable markers to predict renal survival. It is essential to identify predictors of poor clinical outcomes in Fabry disease. Proteinuria greater than 1g/day, GFR less than 45ml/min, or the presence of glomerulosclerosis demonstrated by kidney biopsy is important indicators of inadequate response to ERT.4 The presence of multiple factors influencing proteinuria or GFR makes it unreliable to estimate renal survival from these parameters. Therefore, there is a need for reliable, non-invasive, and easy-to-predict biomarkers of renal survival and fibrosis in Fabry patients. In our study, we found that serum periostin was higher in patients whose has high proteinuria levels, and serum periostin was correlated with lyso-Gb3 and proteinuria. Previously, it has been shown that proteinuria levels and urinary periostin levels are correlated in diseases such as diabetic and hypertensive nephropathy, and lupus nephritis.12,20,21 However, this study revealed this relationship for the first time in Fabry nephropathy. Considering that serum periostin levels are an important indicator of the progression to kidney fibrosis, this correlation between serum periostin levels and proteinuria in Fabry patients is important both for the indicator of renal survival and a possible future treatment target.

While beneficial effects of ERT alone on GFR have been demonstrated, it has no proteinuria-reducing effect alone.22,23 Proteinuria significantly regresses in patients who receive RAS blockade in addition to ERT. This ensures long-term preservation of kidney functions.24 One of the mechanisms by which periostin plays a role in kidney injury is the RAS activation pathway. While periostin can activate the RAS, RAS activation causes increased periostin expression.11 Reducing periostin also reduces kidney injury via integrin-linked kinase, integrin-B3, and oxidative stress pathways.11 Given the reducing effect of RAS inhibition on periostin, RAS blocker therapy may also inhibit the inflammatory and fibrotic process induced by periostin in FD patients, resulting in a more significant reduction in proteinuria in addition to ERT.15

Our study has some limitations. First, the study sample is relatively small. Second, the results of our study consisted of families that included only Turkish Fabry patients. Therefore, applying these results to different races would not be correct. Although many factors that could affect serum periostin levels were excluded during the inclusion process, we think that much more valuable information can be obtained in the development of nephropathy in FD about periostin by looking at the expression of renal periostin. And finally, since our study was a cross-sectional study, it is impossible to draw clear conclusions about the role of periostin in the pathogenesis of Fabry nephropathy.

In the development of Fabry nephropathy, some secondary mechanisms may be in addition to the primary inflammatory process that starts with the accumulation of lyso-Gb3, and continue the kidney injury processes after ERT. Periostin seems to be one of the molecules that may have an essential role in the management of the fibrotic process in Fabry nephropathy. We think that the role of periostin among these mechanisms is worth investigating. In addition to standard ERTs, periostin-reducing therapies may contribute to better kidney survival in Fabry disease. Progressive fibrosis processes caused by periostin in patients with Fabry disease are still a hidden issue waiting to be clarified.

Ethics committee approval was obtained from the institution for the study and written consent was obtained from all patients.

All authors declare that there is no conflict of interest in this study.