Page 63 - Read Online

Aguiar. Rare Dis Orphan Drugs J 2024;3:13 https://dx.doi.org/10.20517/rdodj.2023.56 Page 7 of 29

the analysis of plasma lyso-Gb3 over a lifetime is biased by the age effect on disease severity.

Therefore, the correlations between plasma lyso-Gb3 and clinical variables are limited, with contradictory
results among studies and most of the studies not reporting a sub-analysis by gender and phenotype, biasing
the results and precluding comparisons between them. Moreover, no proper longitudinal studies with
repeated measurements of plasma lyso-Gb3 are available within the literature in order to evaluate its
prognostic value, and even patients with plasma lyso-Gb3 within the normal range may present clinical
events .
[51]

Role in the assessment of treatment response
Lyso-Gb3 decreases significantly with any of the three available ERTs (with a more pronounced decrease in
classically affected males), but does not reach normal values, even in female patients, independently of ERT
preparation [30,40,42,52,61-63] . Moreover, in classically affected males developing anti-ERT antibodies, reduction of
plasma lyso-Gb3 is relatively poor [40,42,52,61,64] , mainly in patients treated with agalsidase α (with a significant
[60]
difference between patients with and without anti-drug antibodies) . One study reported that classically
affected males starting treatment before 25 years of age have a greater decrease in plasma lyso-Gb3
compared with patients starting ERT later in life, with this difference remaining significant even after
adjustment for plasma lyso-Gb3 at baseline, ERT preparation, and ERT dose. Furthermore, the patients
starting ERT at an earlier age presented a trend for less formation of anti-ERT antibodies, and there were
fewer clinical events in the group of patients treated earlier, but the clinical meaning of the former finding is
[65]
unclear due to an age bias .

In patients treated with the chaperone migalastat, there is also a significant decrease in plasma lyso-Gb3 in
the first six months of therapy in treatment-naïve patients [66,67] . Moreover, in patients previously treated with
ERT, plasma lyso-Gb3 remained stable after switching from ERT to migalastat [67-70] . In patients treated with
migalastat but non-amenable in vivo, there is no decrease or increase in plasma lyso-Gb3 in treatment-naïve
patients or patients switching from ERT, respectively [66,68,71,72] . However, in a single study, no significant
correlation was found between plasma lyso-Gb3 change during treatment and the increase in leucocyte
activity after migalastat initiation .
[71]
The reduction in plasma lyso-Gb3 during ERT was found to correlate with the correction of LV mass in
female patients and with a lower hazard ratio for developing new cerebral white matter lesions in male and
female patients . However, in another longitudinal study, with a subanalysis by gender and phenotype and
[61]
comprising mainly late-onset phenotype patients, there was no correlation between the change of plasma
lyso-Gb3 and the change of LV mass in both classical and late-onset FD patients . This absence of
[73]
correlation between plasma lyso-Gb3 and treatment outcomes was further confirmed in a large cohort of
FD patients, showing that plasma lyso-Gb3 before treatment initiation, plasma lyso-Gb3 absolute decrease,
or plasma lyso-Gb3 relative decrease did not predict the risk of event during ERT . The same results, with
[74]
no correlation between changes in lyso-Gb3 and LV mass, GFR, or clinical events, were found for patients
treated with migalastat . These results are in conflict with a single study showing a correlation between
[75]
baseline lyso-Gb3 and the risk of an event during a median follow-up of 68 months, but all but one event
occurred in classical phenotype patients, and when baseline lyso-Gb3 was added to a multivariable logistic
regression model containing age, sex, phenotype and ERT as other covariates to identify the risk of an event,
no significant improvement was found . Thus, the value of plasma lyso-Gb3 to monitor response to DST
[76]
seems at least questionable, and its correlation with clinical outcomes clearly needs further investigation in
larger cohorts of FD patients.