Interest in newborn screening for Mucopolysaccharidosis type I (MPS I) has increased following the development of dried blood spot screening assays [1
], availability of effective treatments [2
], and increasing evidence that early intervention improves patient outcomes [4
]. In early 2016, following systematic evidence-based review and subsequent approval from the Secretary of the Department of Health and Human Services, MPS I was added to the Recommended Universal Screening Panel. Numerous pilot and full-population screening programs using measurement of α-l
-iduronidase (IDUA) activity have been underway in several countries and US states—many of which have experienced a high false-positive rate due to the overlap of IDUA activity between unaffected and affected patients and a high incidence of pseudodeficiency alleles in IDUA
, resulting in overall poor specificity for MPS I [5
]. While several programs have implemented second-tier molecular testing in an attempt to address the poor specificity of enzymatic screening alone, this approach has proved to be problematic due to the discovery of private mutations, variants of unknown significance, identification of carrier status and limitations of utilized methodology [8
]. In addition, the high incidence of genotypic variants leading to pseudodeficiency of IDUA results in excessive performance of potentially unnecessary molecular genetic testing [10
]. Pseudodeficiency for IDUA was first reported in 1985 [11
], and was thought to be a rare occurrence prior to newborn screening.
Measurement of glycosaminoglycans (GAGs), in particular dermatan sulfate (DS) and heparan sulfate (HS), in dried blood spots by several methodologies has been used for the evaluation of at risk patients as well a primary newborn screen for MPS I [12
]. With both prenatal evidence of GAG storage [15
] and confirmation that DS and HS levels can accurately discriminate newborns with severe and attenuated forms of MPS I from unaffected newborns [12
], implementation of biomarker testing in the newborn period is a logical next step to address the shortcomings experienced with enzyme screening and second-tier molecular testing. Our laboratory has developed a stepwise approach to newborn screening for MPS I by incorporating second-tier biomarker testing of the GAGs, DS and HS, following identification of decreased IDUA activity. Second-tier testing utilizes the original newborn screening sample, thereby eliminating recall of the infant for additional testing [16
]. This model reduces the number of false-positive results and improves the specificity of newborn screening not only for MPS I [10
] but also for other conditions [16
]. Testing of dried blood spot GAGs can also be performed in the clinical setting for evaluation of screen-positive infants referred by the respective newborn screening program for possible MPS I. Dried blood spots are an ideal sample type to obtain from newborns, as urine collection can be difficult. Here, we report our experience with the incorporation of DS and HS measurement in the newborn screening process for MPS I.
Newborn screening programs that include lysosomal storage disorders utilizing a one-tiered, enzymatic approach require that all cases with decreased enzyme activity be referred for confirmatory testing and evaluation. Reports from several newborn screening programs utilizing this model suggest that the positive predictive value (PPV) for MPS I newborn screening using IDUA enzyme activity alone is approximately 3% [5
]. Hopkins et al. [6
] reports Missouri’s experience over 4 years of newborn screening for lysosomal storage disorders. During this time period, 133 infants out of 308,000 screened were referred for confirmatory testing and evaluation due to decreased IDUA activity. Out of those referred, 71 cases were attributed to pseudodeficiency, (1/4338) while only two cases were confirmed as MPS I (1/154,000). Similar results were reported by Burton et al. [5
], with 151 infants out of 219,793 screened referred for confirmatory MPS I testing and an incidence of 1/7326 for IDUA pseudodeficiency and 1/219,793 for MPS I. These findings are reproduced in several additional reports [24
]. With the overall incidence of pseudodeficiency being approximately 16 times higher than true disease, newborn screening programs and clinicians involved in the evaluation of referred infants will continue to be overwhelmed with false-positive results if the enzymatic screen alone is utilized.
Newborn screening programs incorporating second-tier molecular genetic testing also have poor specificity for MPS I and refer at risk babies based on decreased α-l
-iduronidase levels alone, because molecular results may take days or weeks to return. Because of high referral numbers requiring second-tier molecular analysis during their testing pilot, the North Carolina newborn screening program incorporated postanalytical interpretation using Collaborative Laboratory Integrated Reports (CLIRs) into their testing algorithm, which provides on-demand web-based analysis using evidence-based segregation of laboratory results from affected patients vs. false-positive cases. Although this was shown to decrease the number of screen positive cases by 64%, only one confirmed case of MPS I was identified out of 19 cases referred for second-tier IDUA
molecular testing [7
Since February 2016, we tested 1213 specimens from cases identified as at risk for MPS I by newborn screening. Based on available follow-up data, we believe to have demonstrated that blood spot GAG analysis in either the initial newborn screening card (performed as a second-tier test) or clinically submitted specimens from infants with decreased IDUA activity identified by newborn screening accurately discriminates between patients with confirmed MPS I and false-positive cases due to IDUA
pseudodeficiency or heterozygosity. All samples tested from infants confirmed by IDUA
molecular analysis to have MPS I had significantly elevated levels of DS and HS compared to those with confirmed IDUA
pseudodeficiency and/or heterozygosity. If newborn screening referrals were limited to patients identified with elevated DS and HS by our laboratory, only 23 patients would have actually required genotyping, which translates to a PPV of 74% (17/23). Accordingly, we have shown in a prospective cohort that blood spot GAG analysis and postanalytical interpretation can significantly increase the specificity of newborn screening for MPS I and should be incorporated into the screening process in order to improve overall program performance. Recently, Burlina et al. [29
] reported that retrospective incorporation of second-tier GAG analysis in dried blood spots resulted in a PPV of 100% for MPS I, with elevated DS and HS in 2/26 cases with positive first-tier results.
Analysis of our testing population and correlation with molecular results identified few false-positive results and uncovered no evidence of false-negative cases at the time of manuscript submission. While several reports [12
] demonstrate elevation of GAGs in newborns with MPS, there is little data on attenuated cases with respect to the timeline of GAGs accumulation and reliability of newborn screening for diagnosis. We acknowledge the possibility that newborns with attenuated MPS I may escape diagnosis using this screening algorithm, and therefore must rely on retrospective and prospective data collection as well as the exploration of alternative and possibly more sensitive testing methodologies [30
] to address this limitation in the newborn screening setting.
Our findings suggest that second-tier GAGs testing and postanalytical interpretation has several important benefits compared to molecular testing, such as avoidance of unnecessary patient recall, decreased assay cost and turnaround time, the prevention of identification of pseudodeficiency, carriers, and cases with inconclusive molecular results which may ultimately receive unnecessary clinical follow up, as well the provision of an actual biochemical phenotype of GAGs metabolism. As molecular analysis alone can be an unreliable tool for prediction of disease severity, GAG analysis combined with the use of postanalytical tools can help to determine pathogenicity of those dreaded variants of uncertain clinical significance. A recent publication by Clarke et al. [31
] discusses the complexity of using genotypic information to predict phenotype in MPS I, with unique genotypes in 12.4% and 40% of patients with severe and attenuated MPS I, phenotypic variability associated with certain missense variants, and inability of current biochemical assessments to predict phenotype. It is also important to note that while the data originated from a voluntary patient registry, 18 patients had only one IDUA
variant reported, highlighting the limitations of molecular testing without accompanying biochemical phenotyping as a component of newborn screening.
As there is evidence that GAGs accumulation in mucopolysaccharidoses occurs prenatally [15
], measurement of DS and HS in newborn screening blood spots can be successfully applied in the newborn screening setting for conditions other than MPS I. Several states have recently added MPS II, also known as Hunter syndrome, to their testing panels and newborn screening has already been underway in Japan and Taiwan for several years. Both published and recent/stated experience [32
] has shown a high false-positive rate for MPS II newborn screening, attributable to the high frequency of pseudo and pseudo-like alleles in the IDS