|
Abstract INTRODUCTION:
Newborn screening (NBS) identifies rare treatable conditions in infants pre symptomatically to enable early intervention. While the Recommended Uniform Screening Panel (RUSP) includes over 30 core conditions that are universally screened in the U.S., many rare inherited disorders with effective therapies or emerging therapies in late-stage development remain unscreened due to limited evidence and lack of robust analytical methods. To address this gap, we developed a highly multiplexed dried blood spot (DBS) assay that currently measures 12 analytes to support biomarker evaluation and future NBS pilot studies across over 20 neurometabolic disorders.
METHODS:
An isotope dilution LC MS/MS assay was developed to quantify 12 chemically diverse analytes from DBS: 2 pyrrolidinone, 25 tetrol glucuronide (GlcA tetrol), 3-O-methyldopa, 4-hydroxybutyric acid (GHB), N-acetylputrescine, N6-succinyladenosine, orotic acid, pipecolic acid, ribitol, S-sulfocysteine, thymidine, and 2 OPP (R/S). Two 3.2 mm DBS punches were extracted for 1 hour at 40°C in a single solvent mixture containing equal volumes of water, methanol, and acetonitrile with stable isotope labeled standards for each target analyte, then split, dried under nitrogen, and reconstituted for two chromatographic workflows (HILIC and reverse phase). The HILIC method quantified nine analytes using a 10 min gradient on a BEH Amide column with polarity switching on a Sciex 6500 triple quadrupole. The reverse phase method quantified three analytes using an 8 min gradient on an HSS T3 C18 column in positive ESI on a Sciex API 4000. Calibration used 1/x weighted linear regression for all analytes except sulfocysteine and thymidine, where 1/x weighted quadratic regression was used. Preliminary evaluation of the assay began with an analytical validation that included assessment of linearity, accuracy, precision, detection capability, analytical specificity, carryover, matrix effects, and stability according to CLSI guidelines. Reference intervals were established using 307 newborn DBS samples. Biomarker concentrations were compared between the normal newborn cohort and disease state, notably aromatic L-amino acid decarboxylase (AADC) deficiency, adenylosuccinate lyase (ADSL) deficiency, and succinic semialdehyde dehydrogenase (SSADH) deficiency.
RESULTS:
Preliminary assessment of the panel revealed three clinically validated biomarkers demonstrating clear discrimination from the newborn reference range. 3-O-methyldopa was significantly elevated in AADC deficiency cases (N=5) where a selected cutoff of 440 ng/mL achieved 100% sensitivity and specificity. N6-succinyladenosine was significantly elevated in ADSL deficiency cases (N=5) where a selected cutoff of 625 ng/mL achieved 100% sensitivity and specificity. 4-hydroxybutyric acid achieved 98.7% accuracy in correctly classifying SSADH deficiency samples from normal samples. Using a combination biomarker approach with 4-hydroxybutyric acid and 2-pyrrolidinone, 100 % sensitivity and specificity were achieved. Across forty-five SSADH samples spanning the newborn period (< 28 d; N=10) and childhood (ages 1–18 y; N=25), and adulthood (> 18 y; N=10), GHB concentrations were above the 99th percentile of the newborn reference interval (disease mean: 16,823 ng/mL; normal mean: 54 ng/mL). Notably, two disease state samples, both 1 day of age, showed substantially lower GHB (mean 527 ng/mL) than the remaining SSADH cases (2,665–107,011 ng/mL), illustrating a potential age-dependent presentation of GHB concentrations. In contrast, 2 pyrrolidinone was elevated above the 99th percentile only in newborn SSADH samples, with a general tendency to have highest concentrations in samples collected at less than 24 hours (mean: 5,599 ng/mL) and decreases with age towards the second screen window (mean: 937 ng/mL) and samples collected during childhood measuring within the normal reference interval (< 509 ng/mL).
DISCUSSION:
Our work demonstrates the feasibility and utility of a multiplex targeted metabolomics approach to expand newborn screening for rare metabolic disorders. With use of a single extraction and two short chromatographic workflows, this assay offers a scalable platform for systematically assessing candidate markers across multiple neurometabolic disorders. Establishing newborn reference intervals alongside preliminary analytical validation enables quantitative comparisons between the newborn population and disorders being evaluated.
Across conditions evaluated to date, the panel identified biomarkers with strong discrimination from the newborn reference range, supporting its potential as a fit-for-purpose screening tool. The observed separation for clinically validated markers in AADC deficiency (3-O-methyldopa) and ADSL deficiency (N6-succinyladenosine), together with the ability to correctly classify SSADH deficiency using a dual biomarker marker strategy, illustrates the utility of our assay. The SSADH findings are a direct outcome of this approach, where inclusion of 2-pyrrolidinone alongside GHB strengthened diagnostic performance and highlights how multi-biomarker profiling can refine screening strategies. In our cohort, GHB alone classified SSADH with high accuracy, but a dual marker strategy achieved 100% sensitivity and specificity and addresses the observed early life presentation in very young infants who may not yet show the classic high-GHB profile while 2 pyrrolidinone is markedly elevated.
As additional candidate disorders and biomarkers are prioritized, the panel can be expanded to incorporate new targets without fundamentally changing the sample preparation or analysis method, enabling data generation for large pilot studies. Collectively, these results connect biomarker discovery and clinical implementation, providing the quantitative evidence needed to evaluate clinical utility and support future NBS expansion efforts.
|
= Discovery stage. (57.21%, 2026)
= Translation stage. (23.38%, 2026)
= Clinically available. (19.40%, 2026)
