1. Introduction

Newborn screening (NBS) for lysosomal storage disorders (LSDs) has been a topic of considerable interest for over twenty years, especially in the United States (US). The first published method to show proof of principle for an LSD test applicable to dried blood spots (DBS) was an enzymatic assay for α-l-iduronidase (IDUA) developed by Chamoles, et al. and applied to Hurler syndrome (MPS I) ^[1]. The method used reagents that released a fluorophore (4-methylumbelliferone; 4-MU) when exposed to the residual enzymes in an aqueous extract from a DBS. This method was basically an adaptation of existing methods used by biochemical genetics laboratories to measure enzyme activity in extracts from whole blood. Scaling down this method to yield a robust assay requiring just a few µL of blood from a 3 mm diameter punch from a DBS was a considerable achievement that was subsequently replicated for other LSDs, including Pompe disease ^[2]. Methods for LSD enzymatic assays based on tandem mass spectrometry (MS/MS) with synthetic reagents were also being developed by Gelb, et al. at about this time. The first published MS/MS enzymatic method targeted Krabbe disease ^[3], which was soon expanded to include several other LSDs ^[4], including MPS I ^[5]. The emergence of these methods and promising new therapeutic strategies such as enzyme replacement therapy (ERT) and hematopoietic stem cell transplantation (HSCT) provided the impetus to develop high-throughput methodologies to screen newborns for LSDs using dried blood spots ^[6][7].

Digital microfluidics (DMF), based on the principle of electrowetting, is a method that manipulates aqueous droplets on printed circuit boards using electrical impulses ^[8]. Programmable droplet dispensing and mixing from reagent and sample reservoirs on a disposable cartridge facilitated a variety of assays applicable to clinical diagnostics ^[9]. The first exploratory application of DMF to NBS was a collaborative project between Advanced Liquid Logic, Inc., the North Carolina Public Health Laboratory, and scientists at Duke University Medical Center, published over ten years ago ^[10]. Shortly thereafter, a critical step forward was taken when a multiplex assay for five LSD enzymes was developed using a prototype cartridge that could process 12 NBS samples ^[11]. In principle, the DMF method miniaturizes and automates the aforementioned 96-well plate format fluorometric assay using reagents specific for each targeted enzyme. The first NBS pilot study using this DMF prototype was performed in the Illinois Newborn Screening Laboratory on over 8000 specimens ^[12]. Shortly thereafter, an improved cartridge that could process 48 specimens was developed and validated for a multiplex assay of five LSDs ^[13].

Newborn screening programs in the US generally follow a protocol for the addition of new conditions for NBS referred to as the recommended uniform screening panel (RUSP) that was established by a consensus committee of experts from the American College of Medical Genetics in 2006 ^[14]. At that time, there were 29 core conditions listed as primary targets for NBS. The ACMG Committee was later encompassed by the U.S. Department of Health and Human Services Secretary’s Advisory Committee on Heritable Disorders in Newborns and Children (ACHDNC) ^[15], which subsequently added to the RUSP severe combined immune deficiency (SCID) in 2009, critical congenital cyanotic heart disease (CCHD) in 2010, and more recently Pompe disease (Glycogen Storage Disease-II) in 2013, Hurler Syndrome (MPS I) in 2016, X-linked adrenoleukodystrophy (X-ALD) in 2016, and spinal muscular atrophy (SMA) in 2018, making a total of 35 primary (core) conditions as of the time of writing. Despite the late addition of the two LSDs (Pompe disease and MPS I) and the fact that the first application to add Pompe disease to the RUSP in 2008 was unsuccessful, special interest groups lobbied to add LSDs to state NBS menus and were successful in several states, including New York, Illinois, Missouri, and others. Consequently, several state programs in the US began screening for LSDs prior to their appearance on the RUSP. Mandates were issued by state legislators without due consideration for the added costs and infrastructure needed to implement NBS for LSDs. It was in this setting that both the MS/MS and DMF based multiplex LSD enzyme assays began to emerge in NBS. Currently, 22 programs in the US are performing full-population NBS for MPS I plus at least one additional LSD; about one-third of them are using DMF and the rest employ MS/MS.

2. MPS I Newborn Screening Using Digital Microfluidics

2.1. Platform Overview

The DMF platform originally developed by Advanced Liquid Logic consists of a disposable cartridge that is loaded with the reagents and samples for analysis and is inserted into a benchtop instrument containing electronics that automate the assay protocols. All operations within the cartridge are controlled by software loaded onto an attached computer, which is capable of controlling up to four instruments simultaneously. The cartridge consists of two plates: the lower plate is a printed circuit board (PCB) with discrete electrodes coated with a dielectric and hydrophobic material, and the top plate is a clear plastic sheet with a conductive hydrophobic coating on the inner surface. These two plates form a sandwich that is filled with an oil to prevent evaporation of the aqueous droplets. Samples and reagents are loaded through the top plate into reservoirs within the cartridge. Droplets are drawn from the reservoirs onto the PCB, attracted by the principle of electrowetting, and manipulated along paths of electrodes using electrical impulses controlled by a computer program ^[9]. (A brief video illustrating the movement of droplets in a DMF LSD newborn screening assay is available online at: https://bit.ly/3iinddB).

The volume of each sample or reagent droplet is determined simply by the area of the electrode adjacent to each reservoir and in this particular system is less than 200 nL, dispensed with an imprecision of <2% error.

For an enzymatic assay, a sample droplet is mixed with a reagent droplet, the mixed droplet is incubated for 1 h, the reaction is stopped by adding a droplet of quenching reagent, and the resulting droplet is moved to a fixed location where a detector reads the fluorescence from the reaction product. Finally, the droplet is moved to a waste reservoir. On the current production cartridge, the maximum number of samples is 44 (four fixed reservoirs are required for calibration) and the maximum number of enzyme reagent wells is five ^[11][13]. Each enzymatic reaction is independently carried out within a discrete droplet; thus, there would be up to 220 individual reactions taking place on the cartridge. It is noteworthy that sample droplets may traverse the same electrodes with minimal cross-contamination or carry-over because the moving droplets are not in direct physical contact with the electrodes ^[9]. The cartridge cannot be reused; it is designed to be disposable.

2.2. Development and Commercialization of the DMF Platform

After the pilot study in Illinois ^[12], which was performed as mentioned earlier with a prototype cartridge limited to only 12 sample inputs, modifications and improvements were made that led to a much more practical model for high-throughput NBS ^[11][13]. This product, developed by Advanced Liquid Logic in 2012, was offered to the Missouri Public Health Laboratory (MSPHL) as a practical solution to their mandate to screen for multiple LSDs and included new reagents prepared under good laboratory practice (GLP) conditions. Consisting of two workstations, each with four DMF instruments controlled by a single computer, it had the capacity to analyze 640 DBS samples per 8 h shift. It was installed in the MSPHL in less than one day, whence the program launched a full-population pilot study in January 2013 for four LSDs—Fabry, Gaucher, Pompe, and MPS I. This model subsequently was named SEEKER^® by the company. At the time of its deployment in Missouri, it was not an FDA approved medical device.

Advanced Liquid Logic was acquired by Illumina, Inc. in July 2013, leading to a hiatus of approximately two years in further clinical DMF assay development until a new company, Baebies Inc., was founded in late 2014 and was able to resume research and development of DMF for NBS in mid-2015. The MSPHL NBS program was fully supported during this hiatus as a condition of the acquisition. In February 2017, SEEKER became the first United States Food and Drug Administration (FDA) authorized platform to screen newborns for LSDs ^[16]. It was also the first DMF device to be cleared by the FDA for clinical applications. It is noteworthy that the MSPHL program has been running for over 7.5 years with minimal modifications to the originally installed platform and is the longest-running prospective screening lab for MPS I in the US. Their publications are testimony to the effectiveness and robustness of DMF for LSD NBS ^[17][18].

2.3. Procedure for Specimen Analysis on the DMF Platform

The DMF NBS workflow has been previously described ^[13]. Briefly, in practice, newborn screening platforms are designed to accept 3 mm diameter samples from DBS punched in 96-well plates; thus, each section of a NBS laboratory receives their samples from a central punching station in this format. When a loaded 96-well plate containing DBS punches arrives in the LSD screening section, the first step is to add extraction solvent (100 µL) to 88 wells using a multi-channel pipet (the remaining eight wells are reserved for on-cartridge calibration solutions). The plate is then placed on an orbital shaker for 30 min at 600 rpm. During the extraction, DMF cartridges are manually loaded with filler fluid and prepared reagents for each assay. Four calibrators are then loaded onto each cartridge, followed by 40 specimens and 4 controls from the 96-well plate (using a multichannel pipettor). Each 96-well plate is loaded across two DMF cartridges. The cartridges are then ready to be loaded into the SEEKER instruments to begin the analytical program. The entire process of reagent, sample and filler loading takes approximately 5 min per cartridge and the assay is completed within 3 h.

Following the completion of the DMF assay, the provided software indicates the status of the quality control specimens relative to the control ranges. It also flags the newborn screening specimens that have activity values below the cut-off specified by the laboratory. Typically, samples with abnormal values are re-punched and re-analyzed in duplicate prior to reporting as an initial screen-positive (if the mean value is below the cut-off).

2.4. Practical Considerations of Digital Microfluidics as a Platform for NBS

2.4.1. Advantages

The features of DMF that render it a valuable addition to NBS platforms have been reviewed elsewhere ^[19]. Its key attributes, compared with other NBS platforms, include minimal capital outlay for the equipment and for the infrastructure to support it, virtually maintenance free operation with no service contracts, and very low power consumption. A detailed cost analysis of DMF NBS is beyond the scope of this review, however, its cost effectiveness and the simplicity of the workflow relative to MS/MS has been previously discussed ^[20]. In addition, the platform requires very low sample and reagent volumes; literally hundreds of assays could be performed on the extract of a single 3 mm punch from a DBS, with very little waste to be disposed of. Variable hemoglobin concentration has no effect on fluorescence values determined by DMF, which is ascribed to the very short path length in a droplet (<0.3 mm) compared with that of a microtiter plate well ^[19].

A simple workflow comprising of a few straightforward steps is accessible to entry-level technicians. There is little room for human error because most of the assay steps are fully automated. Due to “spatial multiplexing” on the DMF cartridge, each enzymatic assay is performed at its optimum pH and buffer conditions—no compromises are required to meet ideal performance criteria for each enzyme assay. As results are generated within 3 h of sample preparation, it is possible to review initial results, repeat any screen-positive samples, and report results within a single 8 h shift.

2.4.2. Limitations

Limitations of the current SEEKER DMF platform include the requirement for two DMF cartridges per 96-well plate, each performing up to four different enzymatic assays on 40 NBS specimens and four controls. Each workstation can accommodate 160 specimens per run. In high-throughput programs, several workstations may be required to accommodate the maximum daily sample volume; alternatively, the sample load can be distributed over one or more additional periods within the working day using fewer platforms, given the relatively short run time of 3 h per cartridge. The addition of more assays to the current platform is practicable with modifications, however, each modification does require FDA submission and clearance prior to deployment. The current platform is not adaptable to any modifications by the end-user, except to change the cut-offs for each analyte.