Introduction:
Gaucher disease (GD) presents on a continuum of severity ranging from non-neuronopathic (GDI) to severe neuronopathic disease (GDII, GDIII) with approved treatments existing for GDI and GDIII but not yet for GDII. Several newborn screening (NBS) programs in the US and elsewhere have begun screening for GD by enzyme analysis. While this approach is highly sensitive, it is not definitive in determining disease or severity. Moreover, GBA genotype/phenotype correlations are limited. While the N409S variant either in homozygous state or compound heterozygous with another pathogenic allele predicts GDI, and the L483P variant either in homozygous state or in combination with a null allele is typically associated with neuronopathic disease (GDII or GDIII), only 29% of variants listed currently by ClinVar are of known significance. With over 500 GBA variants associated with GD, determining phenotype in presymptomatic newborns, which is necessary to guide treatment decisions, is therefore difficult.

Methods:
Since 2014, our laboratory has performed glucosylsphingosine (LGB1) testing as an aid to the diagnosis and monitoring of patients with GD. A subset includes 142 LGB1 specimens from unique infants (0-5 months old) identified by NBS. These cases were reviewed in more detail for this report.

Results:
108 of 142 specimens were collected within 3 weeks of birth and the remaining were collected before 6 months old. Confirmation for affected and carrier cases was by molecular genetic analysis or physician report, and for unaffected infants by molecular genetic and/or enzyme analysis. Of the 44 affected patients, 22 had known non-neuronopathic variants, while 7 had neuronopathic variants. In 15 affected cases, phenotype was unknown at the time of follow up including 6 cases where genotype was known. Of the remaining patients, 17 were carriers, 55 unaffected non-carriers, and 26 with undetermined outcome (lost to follow up). LGB1 levels were compared for unaffected patients and those with GDI vs GDII/GDIII. LGB1 levels largely correlated with the diagnosis. Patients with GD had LGB1 elevations which correlated with phenotype severity, although patients with GDI exhibited some variability. Those with GDII/GDIII had significantly higher LGB1 levels (median: 0.88, range: 0.29-1.79 nmol/mL) than cases with GDI (median 0.06; range: 0.02-0.25 nmol/mL). Of the 15 affected cases with unknown phenotype, 1 had a significantly elevated LGB1, similar to those with GDII/GDIII. All unaffected, non-carrier individuals had normal LGB1 (≤0.04 nmol/mL) as did the remaining 14 with unknown phenotype. Of note, 2 carriers had mildly elevated LGB1, one of which normalized at later testing. The other, with a normal enzyme result and single variant, was discharged from care.

Conclusion:
LGB1 analysis achieves separation of neuronopathic from non-neuronopathic GBA genotypes. With over 500 GBA variants reported in GD, obtaining LGB1 levels either as a second-tier NBS test or following an abnormal NBS for GD should be useful to reduce false positive NBS results and timely elucidation of abnormal NBS results.