FormalPara Key Points

Since migalastat is a competitive inhibitor of α-galactosidase A (AGAL), at least 24 h after the last migalastat intake should pass before blood sampling for AGAL measurements.

A strong correlation of higher plasma migalastat concentrations with decreasing eGFR values 24 h after intake may point toward an accumulation of migalastat in patients with end-stage renal disease.

1 Introduction

Fabry disease (FD) is an X-linked inherited disease due to various mutations within the α-galactosidase A (GLA/AGAL) gene, resulting in deficient lysosomal enzymatic AGAL activity. As a consequence, glycolipids such as globotriaosylceramide (Gb3) accumulate in various tissues and cell types, leading to FD-specific manifestations and symptoms. Affected patients are at high risk of early onset of stroke, life-threatening cardiac arrhythmia, myocardial infarction, or cardiac and renal failure, significantly reducing their life expectancy. Currently approved FD-specific treatment includes enzyme replacement therapy (ERT) with biweekly infusion of agalsidase-alfa (0.2 mg/kg body weight (BW), Takeda), agalsidase-beta (1.0 mg/kg BW, Sanofi Genzyme) or pegunigalsidase-alfa (1.0 mg/kg BW, Chiesi) [1,2,3]. Owing to the mode of action, all FD patients can in principal be treated by ERT. Another approved approach is oral treatment with migalastat (123 mg orally every other day, Amicus Therapeutics), which is an oral chaperone [4]. Treatment with migalastat results in an increase of endogenous AGAL activity by reversible binding to the catalytic center of the mutant protein, leading to a correction of the misfolding [5, 6]. Owing to this mechanistic action, use of migalastat is limited to patients with missense mutations, which are “amenable” to migalastat. Treatment of amenable patients with migalastat was demonstrated to be safe and was able to prevent disease progression [4, 7, 8]. For clinical monitoring of therapy efficacy, it is recommended to measure the AGAL activity of migalastat-treated patients, which should increase significantly if the individual mutation is clinically amenable and the patient shows appropriate compliance [9, 10]. However, this relatively simple method raised some questions, as the results of AGAL activity measurements in migalastat-treated patients are inconsistent and therefore cannot be utilized in routine clinical practice. In detail, some studies demonstrated an AGAL activity increase over time in peripheral blood mononuclear cells (PBMCs) [4, 11], while other studies based on measurements from dried blood spots (DBS) did not [7]. This observation might be explained by the mechanistic action of migalastat, which is a competitive inhibitor of AGAL [12, 13]. Depending on the method or source of AGAL enzyme activity measurements, such as intracellular AGAL determination from PBMCs isolated from whole blood or complete AGAL activity measurements from DBS, this might lead to different results (Fig. 1). Furthermore, the timepoint of blood sampling between two migalastat intakes might be important owing to its elimination from plasma.

Fig. 1
figure 1

Methodologic differences between AGAL activity measurements from PBMCs and DBS in patients under migalastat. Top: AGAL enzyme measurement from PBMCs or leukocytes. PBMCs or leukocytes can be isolated by either dense gradient centrifugation or erythrocyte lysis. Cell pellets are washed (which eliminates/dilutes residual migalastat in the sample) and lysed in appropriate buffers. Cytosolic components are subsequently used for AGAL enzyme activity measurements using 4-methylumbelliferyl-galactopyranoside (4-MUG) as artificial substrate for AGAL. Bottom: AGAL enzyme measurement from DBS. Blood is spotted onto DBS cards and dried. The entire contents are extracted from the punches, which contain all blood components, including cells and serum/plasma and thus also migalastat. A subsequent cell lysis reveals intracellular proteins. During this step, migalastat from serum may bind to AGAL, leading to competitive inhibition in subsequent AGAL activity measurements, resulting in decreased or even undetectable AGAL activities, since 4-MUG cannot be catalyzed. PBMC peripheral blood mononuclear cell, DBS dried blood spot

Full size image

The aim of the current study is to analyze why measurement of AGAL activity to monitor treatment efficacy in migalastat-treated patients is not successful and to develop a strategy to avoid false-negative results. To this end, we determined the time-dependent migalastat concentration in sera from 43 patients with 58 visits and compared AGAL activities measured from PBMCs and DBS.

2 Materials and Methods

2.1 Patients and Study Design

All investigations were performed after approval by the Medical Association of Westphalian-Lippe and the Ethics Committee of the Medical Faculty of the University of Muenster (project no. 2011-347-f, date of report 7 July 7 2011) and in accordance with the Declaration of Helsinki. Written informed consent was obtained from all included patients for analysis and publication. Between March 2023 and May 2024, a total of 42 migalastat-treated patients were consecutively recruited at the Interdisciplinary Fabry Center (IFAZ) at the University Hospital Muenster in Germany.

The inclusion criteria for this prospective study were (a) age ≥ 18 years at enrolment, (b) a genetically confirmed GLA mutation, amenable to migalastat, (c) continuous treatment with migalastat for ≥ 1 months prior to blood sampling, and (d) no participation in another clinical trial. Fluorometric-based AGAL activity measures from DBS were performed by Centogene (Rostock, Germany). A lower limit of detection (LOD) was reported with < 2.8 µmol/L/h, and a reference value of 15.3 µmol/L/h was reported as the lower limit of normal.

2.2 Biochemical Analyses

Serum-mediated inhibition assays were performed as previously reported [14,15,16]. In detail, 5 μL of patients’ sera was pre-incubated with 1 ng agalsidase-beta (Sanofi) for 10 min at room temperature. Subsequently, 4-methylumbelliferyl-α-d-galactopyranoside (4-MUG) (Biosynth, Staad, Switzerland) was added to measure α-galactosidase A (AGAL) activity via fluorescence measurement [17]. N-Acetylgalactosamine (Biosynth) was used to specifically block endogenous α-galactosidase B activity [18]. The reaction was set in sodium acetate buffer with a final concentration of 100 mM at pH 4.6. After 1 h incubation at 37 °C, the reaction was stopped by adding one equal volume of 0.5 M Na2CO3 buffer (pH 10.8), and fluorescence activity was measured at 460 nm. Five microliters of FCS, instead of human serum, was used as a control. Detected AGAL activity was expressed as a percentage compared with activity measured in the control. Measurements were performed in triplets.

To determine the AGAL inhibition by migalastat, inhibition assays with increasing concentrations (end concentration 0.005–5 µM) were performed. In short, the different concentrations of migalastat were diluted in 5 µl FCS and pre-incubated with 1 ng agalsidase-beta for 10 min at room temperature. AGAL activity was measured using 4-methylumbelliferone-α-d-galactopyranoside. To express enzyme inhibition, AGAL activities were normalized against negative control (FCS only). Enzyme inhibition was than plotted against the amount of migalastat to prepare a standard curve.

AGAL activity measures from PBMCs were performed as follows: PBMCs were extracted from 7.5 mL EDTA-stabilized blood using erythrocyte lysis buffer (150 mM NH4Cl, 10 mM KHCO3, 1 mM EDTA; pH 7.4) and incubated for 30 min on ice. Remaining cell pellets were washed twice with PBS (pH 7.4). Cells were subsequently lysed with 1× passive lysis buffer (Promega; Walldorf, Germany) according to manufacturer’s instruction. AGAL activities were measured as described above, and values were compared with a standard curve of serial dilution of agalsidase-beta activity. For AGAL quantification from PBMCs by western blots, 20 µg total protein of whole protein extract was separated on a 12% SDS gel. Subsequently, samples were blotted onto PVDF membranes and blocked with 5% semi-skimmed milk powder (Roth, Karlsruhe, Germany) in Tris-buffered saline supplemented with 0.1% Tween 20 (AppliChem, Darmstadt, Germany) for at least 2 h. The primary antibody was a monoclonal rabbit anti-galactosidase alpha antibody (Abcam, Cambridge, UK, ab168341, 1:10,000).

The secondary antibody was a goat anti-rabbit antibody conjugated with horseradish peroxidase (Merck Millipore, Burlington, Massachusetts, USA, 12-348, 1:20,000). For detection, Clarity Western ECL Substrate (Bio Rad, California, USA) was used. ImageJ (Rasband, W.S., ImageJ, U.S. National Institutes of Health, Bethesda, Maryland, USA, https://imagej.net/ij/, 1997-2018) was used to analyze the individual signal intensities from western blots, which were expressed and compared as the individual areas under the curve (AUC).

2.3 Statistical Analyses

Unless otherwise indicated, measures were performed in triplets. Categorical data are expressed as numbers, and relative frequencies as percentages. Two-tailed Student’s t tests were used for statistical analysis, and statistical significance was considered at two-sided p < 0.05. Cartoons were created with BioRender.com. GraphPad PRISM V8.4 software (GraphPad Software Inc, La Jolla, California) was used for appropriate statistical analyses and visualization.

3 Results

Current expert opinion and guidelines for patient management recommend measuring AGAL enzyme activity to monitor biochemical amenability in migalastat-treated patients as part of routine clinical practice [9, 10]. However, the results are often inconclusive. Table 1 demonstrates the DBS-based measurements of AGAL enzyme activity from 43 patients (18 females; 58 visits) with amenable AGAL mutations, treated with migalastat. Importantly, AGAL activity was below the lower limit of detection (LOD; < 2.8 µmol/L/h) in 21 (42.0%) measurements, although distinct enzymatic activity should be measureable in these patients (due to the residual activities of the individual mutations) even without migalastat stimulation. Furthermore, data from the GLP-HEK assays used to detect amenable AGAL mutations [5] clearly demonstrate the baseline activity and the theoretical increase in activity after incubation with migalastat of the respective AGAL mutants in HEK-293T cells (Table 1). This led to the assumption that migalastat (which is a competitive inhibitor of AGAL) might interfere within the AGAL activity assays from DBS (Fig. 1).

Table 1 α-Galactosidase A activity measurements from dried blood spot in patients (n = 43, 58 visits) with amenable AGAL mutations under migalastat
Full size table

3.1 Serum-Mediated Inhibition Assays to Detect Migalastat-Mediated Inhibition

To analyze whether AGAL activity from DBS is dependent on the timing of last migalastat intake and subsequent blood sampling, we assessed these intervals based on patients’ intake patterns and the timing of blood sampling at our center (Fig. 2A).

Fig. 2
figure 2

AGAL activities from DBS and serum-mediated inhibition assays as a tool to measure migalastat concentrations indirectly in serum samples from migalastat-treated patients. A Overview of the intake behavior of patients receiving migalastat. Black circles: all; blue triangles: males; red squares: females. B Correlation between individual AGAL activities measurable by DBS and the time after migalastat intake. C Mean serum-mediated AGAL inhibition in patients receiving migalastat. D Schematic overview of the experimental setup. E Serum-mediated AGAL inhibition as a function of time interval of activity determination and migalastat intake. Cmax at 3–4 h and plasma half-life of 4 h according to literature is indicated on the X axis. DBS dried blood spot

Full size image

Most patients took their capsules late in the evening/night or early in the morning, which seems to be most convenient due to the strict recommendation for migalastat intake (abstinence from food for at least 2 h before and after migalastat intake).

Of note, we did not see any difference in capsule intake behavior between females and males. A linear regression analysis between time after migalastat intake and measurable DBS-based AGAL activities revealed a slight correlation, showing increasing activities with increasing time (r2 = 0.2089; p < 0.0165; Fig. 2B). Recently, we reported that serum samples from migalastat-treated patients can show markable AGAL inhibition in serum-mediated inhibition assays without the presence of neutralizing anti-drug antibodies [16]. Thus, we used the well-established serum-mediated inhibition assays, which are generally used to detect neutralizing anti-drug antibodies against recombinant AGAL to measure a potential inhibition in sera of migalastat-treated patients (Fig. 2C, D). Following this approach, we detected significant AGAL inhibition of the measured serum samples, confirming our previous observations [16] (Fig. 2C). A correlation between time after migalastat intake until blood sampling and the measurable AGAL inhibition revealed a strong correlation, showing a decreasing inhibition with increasing time interval (r2 = 0.8140; p < 0.0001; Fig. 2E). After 24 h, only slight inhibition was still measurable, which reflects the Cmax of 3–4 h after intake and a plasma half-life of ~ 4 h [6]. Thus, serum-mediated AGAL inhibition was significantly higher in blood samples drawn before 24 h after migalastat intake (Supplementary Fig. 1A; p < 0.0001), while measurable AGAL activities from DBS were significantly lower in samples drawn before 24 h (Supplementary Fig. 1B, p = 0.0319). To further substantiate our findings that migalastat might interfere with DBS-based AGAL activity measurements, we retrospectively analyzed DBS-based AGAL activities from 17 patients before migalastat treatment was initiated (i.e., migalastat-naïve). The DBS-based AGAL activities from migalastat-naïve patients (n = 10) and their AGAL activities under migalastat treatment (when AGAL activities were measured < 24 h after migalastat intake) demonstrate that measured AGAL activities were significantly reduced compared with the treatment-naïve baseline (p = 0.0148; Supplementary Fig. 2A). Vice versa, if DBS-based AGAL activities were measured ≥ 24 h after migalastat intake, AGAL activities were significantly increased (p = 0.0151; Supplementary Fig. 2B).

3.2 Migalastat Concentrations in Patients’ Samples

To determine remaining migalastat concentration in patients’ sera, inhibition assays with serial dilution of migalastat were performed (Fig. 3A). The inhibition curve shows a sigmoidal course with a half-maximal inhibitory concentration (IC50) of 0.039 µM migalastat in the control serum (bovine fetal calf serum) (Fig. 3A). The calculation of the respective migalastat concentration in patients’ samples revealed a heterogeneous distribution (Fig. 3B). As expected, migalastat concentrations were lowest ≥ 24 h after the last intake (Fig. 3B). Since the excretion of migalastat is mainly mediated via the kidneys [6], we next analyzed whether renal function (eGFR) could have an impact on individual migalastat elimination in our patients. During the first 24 h after intake, renal function (eGFR) seemed to have no significant effect on the migalastat-mediated inhibition and migalastat concentrations in our patients’ sera (p = 0.1859 and p = 0.3086, respectively; Fig. 3C). However, 24 h after intake, renal function (eGFR) was significantly associated with migalastat concentrations in that low eGFR values were associated with the highest remaining migalastat-mediated inhibition and migalastat concentration in individual patients’ samples (p = 0.0064 and p = 0.0102, respectively; Fig. 3D).

Fig. 3
figure 3

Determination of migalastat concentrations and impact of renal function on plasma clearance. A The concentration-dependent inhibition of AGAL can be used as a tool to determine migalastat concentrations in blood samples. B Migalastat concentrations in 54 serum samples from patients under migalastat. C Overall correlations of individual eGFR values with serum-mediated AGAL inhibition and respective migalastat concentrations from serum samples drawn < 24 h after migalastat intake. D Overall correlations of individual eGFR values with serum-mediated AGAL inhibition and respective migalastat concentrations from serum samples drawn ≥ 24 h after migalastat intake

Full size image

3.3 AGAL Measurement in Patients’ PBMCs

To assess whether the measurement of AGAL activity in PBMCs from migalastat-treated patients might be more appropriate than that from DBS, we measured the enzymatic activities in PBMCs from 29 patients (Fig. 4). AGAL activities in PBMCs were heterogeneous, but overall higher when ≥ 24 h elapsed between migalastat intake and blood sampling (p = 0.0317; Fig. 4A). Western blot analyses to control presence of AGAL in the patients’ samples demonstrated a strong correlation between the individual enzymatic AGAL (ng/mg protein) activities and the amount of AGAL determined by western blot (AUC) in a representative subset of samples (Fig. 4B–D).

Fig. 4
figure 4

α-Galactosidase A measurement and quantification in patients’ PBMCs. A AGAL activity in PBMCs (n = 29) seems to be related to the time interval between migalastat intake and blood sampling. The dotted green line at 29.2 ng/mg protein marks the reference AGAL activity (minimum) of normal values. B Western blot dependent AGAL quantification in selected samples; 20 µg total protein from PBMCs lysates were loaded in individual lanes. C AGAL quantity is also related to the time interval between migalastat intake and blood sampling. D AGAL quantity corresponds to the enzymatic activity in PBMCs of migalastat-treated patients. PBMC peripheral blood mononuclear cell

Full size image

4 Discussion

Treatment with migalastat is safe, and most patients present with a stable disease course under treatment [4, 8]. To assess the individual biochemical response to therapy, the current guidelines recommend (repeated) measurements of enzymatic AGAL activity in migalastat-treated patients [9, 10]. However, it is currently not clear when exactly AGAL activity should be measured after taking migalastat and from which source to obtain clinically relevant, valid values. In this study, we tried to address a suitable timepoint and method to measure AGAL activity in treated patients. In this respect, our main findings are: (1) in DBS-derived AGAL measurements, migalastat seems to interfere with AGAL measurements, potentially leading to false-negative results, (2) since migalastat is a competitive AGAL inhibitor, serum samples from patients can be used to indirectly measure the migalastat concentration in the blood, (3) more valid results in enzymatic AGAL measurements can be obtained when using samples collected at least 24 h after the last migalastat intake, (4) AGAL activity measurements from PBMCs seem to be more valid than those from DBS, (5) even after 24 h, reduced renal function leads to significantly reduced migalastat clearance and thus to potentially reduced therapeutic AGAL activity.

4.1 Migalastat Concentrations in Patients’ Sera

Migalastat is a competitive inhibitor of AGAL and is mainly excreted via the kidneys, but data concerning Cmax and the plasma half-life of migalastat in FD patients are scarce. The use of a modified serum-mediated inhibition assay enabled us to determine Cmax and the corresponding plasma half-life of migalastat in patients relatively easily. Our data confirmed a Cmax after 3–4 h in plasma and a plasma half-life of ~ 4 h [6]. Interestingly, we observed a strong correlation of higher plasma migalastat concentrations with decreasing eGFR values 24 h after intake. This observation is important, because proper migalastat clearance from plasma is important for enhancing AGAL activities. In case of severely impaired renal function, this can no longer be achieved and migalastat could accumulate in the patient, leading to AGAL inhibition. Thus, migalastat treatment is currently not approved in patients with eGFR < 30 mL/min per 1.73 m2. At a first glance, this value appears to be relatively arbitrary and relates more to chronic kidney disease (CKD) stages than to a real functional test in treated patients or healthy volunteers. In future studies, our approach might be suitable to determine a real functional threshold for eGFR under migalastat, potentially leading to a more personalized treatment strategy for patients with FD.

4.2 AGAL Activity Measurement in PBMCs

Leucocytes or PBMCs were used as a reliable source for intracellular enzymatic AGAL activities after it was demonstrated that activity measures from plasma or serum were not reliable for some mutations. An important example is p.D313Y, where low enzymatic activities can be measured in the plasma and serum of male carriers, but intracellular activities are within the normal range [19]. However, over the last years, DBS have replaced measurements from leucocytes or whole PBMCs in many areas of clinical routine in FD. The use of DBS eliminates the need for a cold chain when shipping whole blood samples, and the cards can be stored for longer without any effort. Furthermore, AGAL activity, lyso-Gb3 concentration, and the corresponding genetics for the detection and confirmation of FD can be carried out from one single sample. However, despite these benefits, the use of DBS for AGAL activity measurements in migalastat-treated patients seems to be limited. This is particularly important when AGAL activities are used to assess the biochemical amenability in migalastat-treated patients, as is recommended [10]. First studies analyzing the effect of migalastat in treated patients demonstrated a measurable increase of intracellular AGAL activities over time [4], an effect that which could not always be demonstrated by other studies [7,8,9]. The difference between the studies could be methodological, since the more recent studies used DBS and the residual migalastat in the extract from the DBS preparation may have interfered with the subsequent measurement of AGAL activity. By measuring migalastat concentrations in serum samples, we were able to demonstrate that 24 h after migalastat intake, the remaining migalastat concentrations have less effect on measured AGAL activities from DBS. However, the use of PBMCs as a source for intracellular AGAL activity showed more valid results.

Our study has some limitations. The aim of this study was to elucidate the optimal timepoint for AGAL activity measures under migalastat treatment and not to draw conclusions about the efficiency of the therapy itself. Since 4-MUG is an artificial substrate for AGAL, it cannot be excluded that 4-MUG-based assays will not fully reflect the real endogenous AGAL activity. In addition to disease-causing missense mutations, also patients with genetic variants of unknown significance and probably benign mutations such as p.D313Y were included in this study. In this respect, these additional recruited patients served to increase the statistical power for our biochemical analyses. Nonetheless, we would like to emphasize once again that we discontinued migalastat treatment of the appropriate patients with p.A143T and p.D313Y (recent reclassification of these variants) after the last visit.

5 Conclusions

The optimal time for enzymatic AGAL activity measurement in migalastat-treated patients appears to be 24 h after the last migalastat intake at the earliest. Treating physicians should take this into account when assessing biochemical amenability in their migalastat-treated patients. In addition, AGAL activity measurements should be performed from PBMCs rather than DBS to reduce migalastat-mediated interference.