PROJECT SUMMARY
Newborn screening (NBS) has been of incalculable benefit to infants since its inception in the 1960s. The
laudable goal of NBS programs has been to identify infants who will develop lethal or debilitating childhood
disorders at a time when they are pre-symptomatic and when treatment is maximally effective. However,
problems with the accuracy of the diagnostic paradigms for screened disorders represent a potential source of
harm to infants and parents alike. Excessive false positive rates of some newborn screens, which cause
insufficiently high positive predictive values (PPVs), contribute to diagnostic uncertainty. This uncertainty can
lead to 1) a delay in diagnosis and worsening of treatment outcomes, and 2) morbidity and mortality resulting
from unnecessary treatment. Furthermore, significant parental anxiety needlessly accompanies false positive
NBS results. The broad goal of the current application is to improve the accuracy of NBS for Krabbe disease
(KD), Pompe disease (PD), and Mucopolysaccharidosis type I (MPSI). During our recently completed R21
grant, it was established that for KD, an approach to NBS consisting of the development of bivariate normal
limits (BVNL) for the amounts of two biomarkers, psychosine (PSY) and the enzyme galactocerebrosidase
(GalC), in newborn dried blood spots (DBS) can predict symptoms before they occur. Retrospective testing of
this tool resulted in a very high PPV of 98.5%, essentially eliminating the existing false positive problem for KD.
(KD screening in New York State resulted in 1.4 % PPV.) Preliminary studies indicate that BVNL tools, with
PPVs approaching 100%, can also be developed for MPSI and PD.
This proposal will test novel BVNL tools for KD, MPSI, and PD using specific NBS biomarkers, and will assess
these tools for effective pre-symptomatic identification of these disorders. Aim 1a will prospectively collect
results of PSY and GalC from DBS of infants who screen positively for KD in New York, Ohio, Missouri, and
metropolitan Chicago, Illinois. These areas, as well as Gifu, Osaka, and Shimane, Japan (where only data for
MPSI will be collected), comprise the proposed Bivariate Analysis for Newborn Screening (BANS) Network of
collaborating genetics referral sites. This acronym reflects the anticipated use of BVNL tools in an approach
that eliminates the potential harm of excessive false positives after NBS of KD, PD and MPSI. Aim 1b will
utilize the BANS Network for prospective monitoring of infants who have screened positive for KD to determine
whether application of the BVNL tool to their pre-symptomatically collected blood spots predicts subsequent
symptom emergence. Aim 2 will utilize the BANS network and newborn blood spots obtained from the Virtual
Repository of Dried Blood Spots to further develop BVNL tools for MPSI and PD. Aim 3 will again employ the
BANS Network to prospectively determine the predictive capacity of the BVNL tools for MPSI and PD. If the
improved prediction of KD, MPSI, and PD after NBS is achieved, treatment will be enhanced for these
devastating illnesses, and life-threatening treatments for infants will be prevented.
