- Email: [email protected]
A Droplet Digital PCR Method for Severe Combined Immunodeficiency Newborn Screening
D J M ram 17 rog 20 E P CM
The Journal of Molecular Diagnostics, Vol. 19, No. 5, September 2017
jmd.amjpathol.org
A Droplet Digital PCR Method for Severe Combined Immunodeficiency Newborn Screening Noemi Vidal-Folch,* Dragana Milosevic,* Ramanath Majumdar,* Dimitar Gavrilov,*y Dietrich Matern,*yz Kimiyo Raymond,*y Piero Rinaldo,*yz Silvia Tortorelli,*y Roshini S. Abraham,*x and Devin Oglesbee*y From the Departments of Laboratory Medicine and Pathology,* Clinical Genomics,y Pediatric and Adolescent Medicine,z and Allergy and Immunology,x Mayo Clinic College of Medicine, Rochester, Minnesota CME Accreditation Statement: This activity (“The JMD 2017 CME Program in Molecular Diagnostics”) has been planned and implemented in accordance with the Essential Areas and policies of the Accreditation Council for Continuing Medical Education (ACCME) through the joint providership of the American Society for Clinical Pathology (ASCP) and the American Society for Investigative Pathology (ASIP). ASCP is accredited by the ACCME to provide continuing medical education for physicians. The ASCP designates this journal-based CME activity (“The JMD 2017 CME Program in Molecular Diagnostics”) for a maximum of 36 AMA PRA Category 1 Credit(s). Physicians should only claim credit commensurate with the extent of their participation in the activity. CME Disclosures: The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose.
Accepted for publication May 30, 2017. Address correspondence to Devin Oglesbee, Ph.D., Clinical Biochemical Genetics Laboratory, Department of Laboratory Medicine and Pathology, Mayo Clinic, 200 First St SW, Rochester, MN 55905. E-mail: [email protected].
Severe combined immunodeficiency (SCID) benefits from early intervention via hematopoietic cell transplantation to reverse T-cell lymphopenia (TCL). Newborn screening (NBS) programs use T-cell receptor excision circle (TREC) levels to detect SCID. Real-time quantitative PCR is often performed to quantify TRECs in dried blood spots (DBSs) for NBS. Yet, real-time quantitative PCR has inefficiencies necessitating normalization, repeat analyses, or standard curves. To address these issues, we developed a multiplex, droplet digital PCR (ddPCR) method for measuring absolute TREC amounts in one DBS punch. TREC and RPP30 levels were simultaneously measured with a Bio-Rad AutoDG and QX200 ddPCR system. DBSs from 610 presumed-normal, 29 lymphocyte-profiled, and 10 clinically diagnosed infants (1 X-linked SCID, 1 RAG1 Omenn syndrome, and other conditions) were tested. Control infants showed 14 to 474 TREC copies/mL blood. SCID infants, and other TCL conditions, had 15 TREC copies/mL. The ddPCR lower limit of quantitation was 14 TREC copies/mL, and the limit of detection was 4 TREC copies/mL. Intra-assay and interassay imprecision was <20% CV for DBSs at 54 to 60 TREC copies/mL. Testing 29 infants with known lymphocyte profiles resulted in a sensitivity of 88.9% and a specificity of 100% at TRECs <20 copies/mL. We developed a multiplex ddPCR method for the absolute quantitation of DBS TRECs that can detect SCID and other TCL conditions associated with absent or low TRECs and validated this method for NBS. (J Mol Diagn 2017, 19: 755e765; http://dx.doi.org/10.1016/j.jmoldx.2017.05.011)
Severe combined immunodeficiency (SCID) disease is a group of inherited immunodeficiencies characterized by T- and B-cell lymphopenia. The overall incidence of SCID is reported to be approximately 1:58,000.1 Early detection of SCID is paramount; otherwise, severe and recurrent infections may be fatal within the first few years of life.2 Intervention by hematopoietic stem cell transplantation has changed the clinical outcome for early diagnosed SCID infants, as recent studies have shown.3 Survival rates of >80% for SCID infants treated by hematopoietic stem cell transplantation within the first few months of life with, or without, resolved infections are in stark contrast to reported
survival rates of only 50% for older infants with an active infection during transplant. Hence, early diagnosis of SCID is vital for achieving improved treatment success. Given the severity of SCID-related symptoms and improved outcomes after an early diagnosis, newborn screening (NBS) fulfills classic Wilson-Jungner screening criteria, and SCID was deemed a suitable candidate for NBS by the US Department of Health and Human Services, per a Supported by Mayo Clinic Department of Laboratory Medicine and Pathology internal funding. Disclosures: None declared.
Copyright ª 2017 American Society for Investigative Pathology and the Association for Molecular Pathology. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jmoldx.2017.05.011
Vidal-Folch et al letter issued on February 25, 2010, to the Secretary of Health and Human Services. Early detection and diagnosis is currently available by newborn screening programs in most US states.1 Newborn screening for SCID has also been evaluated by studies at several other locations in North America, Europe, and Asia.4e9 Current SCID screening methods exploit T-cell receptor excision circles (TRECs) that are a DNA by-product of T-cell receptor recombination and reflect T-cell maturity.10e12 Low TREC amounts mirror concurrent T-cell lymphopenia (TCL), and neonatal dried blood spot (DBS) TREC levels can be used to detect impaired T-cell development, and SCID by newborn screening.13,14 After a positive NBS result, diagnostic follow-up is required to determine whether the infant has typical SCID, or another form of TCL, and to establish a specific diagnosis.15 SCID can be detected in DBSs using real-time quantitative PCR (qPCR) for TRECs as a single-plex,16e19 duplex,5,20 triplex,21 or combined with targets for other conditions, such as an SMN1 exon 7 deletion for spinal muscular atrophy.22 Presently, there is no standardized SCID newborn screening method and the ideal newborn screening assay for SCID must be able to distinguish specimens with absent, or low, TREC content from those with insufficient DNA yield. The variability in newborn DBS specimen collection adds to the challenge of distinguishing samples with abnormal TREC levels. Because most SCID assays are based on PCR amplification, low results could be because of artifacts, such as an inadequate DBS sample, failure to elute a sufficient DNA amount, or the presence of a DNA polymerase inhibitor. To ensure that a low TREC result is not because of an artifact (causing a false-positive screening result), most TREC assays include simultaneous, or sequential, amplification of a conserved genomic sequence, like the ribonuclease P gene, RPP30, or the b-actin gene, ACTB.13,20 Detection of a reference gene within the expected range indicates that the tested specimen is amplifiable and TREC quantification is reliable. Another sensitive method to detect low concentrations of a DNA target within a single analysis is droplet digital PCR (ddPCR).23 With ddPCR, a single DNA sample is partitioned into thousands of uniform-size droplets and subsequently amplified by standard PCR, and a signal from a fluorescently labeled probe within each droplet is recorded as either positive or negative, depending on the presence or absence of a target’s PCR product. This results in specific and sensitive fluorescent signals derived from thousands of individual droplets containing target product. Negative droplets do not provide a fluorescent signal as they are void of target DNA amplicon. The application of a Poisson distribution calculation across all droplets permits an absolute concentration determination for a DNA target and eliminates the need for standard curves of reference standards or endogenous controls. Thousands of replicate measurements of each sample occur during each run. Last, the application of differentiated labeled fluorescent probes allows for
multiplexing PCR targets, such as TRECs and RPP30, within a single DNA tube and from the same DBS punch. Given the efficiencies promised by ddPCR, we designed a ddPCR assay to simultaneously measure the absolute level of TRECs and RPP30. Validation of this method is described herein.
756
jmd.amjpathol.org
Materials and Methods Infant Specimens Infant and cord blood specimens were collected with Mayo Clinic Institutional Review Board approval. A population-based TREC reference range was determined from 610 (541 full-term and 69 preterm) residual, anonymized DBSs collected onto Whatman 903 Protein Saver Cards (GE Health Care, Pittsburgh, PA) whose demographics are shown in Table 1. Infants with a gestational age <37 weeks were considered preterm. DBSs were also prepared from 10 infants diagnosed with X-linked SCID, Omenn syndrome, idiopathic TCL, chromosome 22q11.2 deletion-negative TCL, DiGeorge syndrome, CHARGE (coloboma of the eye, heart abnormalities, coanal atresia, retardation of growth and development, genitourinary anomalies, ear anomalies and deafness) syndrome, ataxia telangiectasia, and cartilage hair hypoplasia, whose demographic data are listed in Table 1 and TREC results in Table 2. Moreover, 29 blood samples submitted to the Mayo Clinic for follow-up testing, after NBS results, were anonymized and analyzed with data shown in Table 3.
Control Specimens Three levels of controls were prepared as DBSs to evaluate the performance characteristics of the TREC assay: i) an SCID-like sample, with TREC content lower than the expected range for newborns, but normal RPP30 levels; ii) a borderline negative control with TREC content adjacent to the assay’s threshold between normal and abnormal (decreased but not absent TREC), and normal RPP30 levels; iii) a negative control with TREC content within the expected newborn range, and normal RPP30 levels. In addition, DBSs were prepared for stability, accuracy, linearity, imprecision, clinical sensitivity, and specificity studies by spotting 75 mL of EDTA whole blood onto Whatman 903 cards and drying overnight at room temperature. DBSs were stored at 20 C with desiccant in sealed plastic bags. Positive-control DBS cards were made from anonymized adult EDTA whole blood, and negative-control DBS cards were made from donor umbilical cord blood, both collected with institutional review board approval. Last, 30 DBSs were obtained from the CDC’s Newborn Screening Quality Assurance Program for TREC Analysis in DBS to assess performance against other laboratories using qPCR or an EnLite Neonatal TREC kit (Perkin Elmer, Waltham, MA). Every quarter, the CDC distributes a panel of five unknown
-
The Journal of Molecular Diagnostics
Droplet Digital PCR for SCID NBS Table 1
Demographics of Presumed Normals and Low TREC Subjects
Cohort/specimen
Sex
Age at collection, days
Gestational age, weeks
Birth weight, g
Full-term infants (n Z 541) Preterm infants (n Z 69) DBS001 DBS002 DBS003 DBS004 DBS005 DBS006 DBS007 DBS008 DBS009 DBS010
272 females, 264 males, 5 not provided 30 Females, 39 males Male Female Male Male Male Female Female Male Male Male
1e133 (3)
37e41 (38)
1660e4680 (3020)
1e74 (7) 15 41 8 8 6 8 8 6 22 12
27e36 (36) NA NA 40 NA 37 NA NA NA NA NA
1350e3885 (2290) 4069 NA 4454 3852 3300 NA NA NA NA NA
Median values are shown in parenthesis. NA, not available; TREC, T-cell receptor excision circle.
DBS specimens to domestic, international, and manufacturer laboratories to analyze TREC content in peripheral blood.
DNA Extraction An overview of the ddPCR assay workflow is shown in Figure 1. A DNA extraction protocol was established for this study. Briefly, a single 3.2-mm disk was punched from DBSs into a semiskirted PCR 96-well plate (Eppendorf, Hamburg, Germany). DBS disks were washed twice by adding 120 mL Generation DNA Purification solution (Qiagen, Valencia, CA) and heating to 50 C for 30 minutes. After washing, 120 mL Generation DNA Elution buffer (Qiagen) was added, and the plate was incubated at room temperature for 15 minutes. After removing the Generation Table 2
DNA solution, 15 mL of water was distributed per well, and the plate was heated to 99 C for 25 minutes to release DNA. Last, the plate was stored overnight at 4 C, followed by a high-speed centrifugation at 3500 g for 10 minutes to remove debris.
ddPCR The ddPCR TREC primers and probe sequences were the same as those described by Douek et al.11 RPP30 primers and probes were newly designed for this assay (Table 4). Both TREC and RPP30 ddPCR assay mixes were custom synthesized by IDT (Coralville, IA). Amplicon sizes for the expected PCR products for TREC and RPP30 were 109 and 65 bp, respectively. The ddPCR mixture was assembled by an automated liquid handler system. Reaction mixtures of 22 mL
SCID Screen-Positive Cohort DBS results (reference range)*
Whole blood results (reference range)* CD16þ56þ NK cells, CD45RA CD45RO % of CD4 % of CD4 cells/mL T cells (43e256) T cells
CD4þ31þ 45RAþ % of CD4þ 45RAþ T cells Clinical diagnosis
RPP30 copies/mL TREC blood copies/mL blood (<20) (1500)
842
1480
80
15
52
10
14,550
229
202
158
84
11
57
5
23,550
429
379
417
549
73
20
46
15
17,775
986 248 116 914 3 307 408
314 94 42 495 18 69 15
1226 252 795 113 325 1006 2
446 159 1681 1418 42 317 1420
54 68 50 59 0 74 <1
30 12 41 39 94 26 95
82 57 76 63 0 47 0
6 4 <4 11 <4 <4 <4
13,413 10,500 11,925 7513 3013 5438 8150
TREC copies/106 CD3þ T cells, CD4þ T cells, cells/mL CD3 T cells cells/mL (1484e5327) (733e3181) Specimen (6794)
CD8þ T cells, cells/mL (370e2555)
DBS001
1220
1044
729
319
DBS002
998
912
686
DBS003
2201
814
DBS004 DBS005 DBS006 DBS007 DBS008 DBS009 DBS010
2192 1241 1090 5698 <5 1836 <5
1293 344 174 1436 24 382 424
CD19þ B cells, cells/mL (370e2306)
Idiopathic T-cell lymphopenia 22q11Deletion-negative TCL Idiopathic T-cell lymphopenia DiGeorge syndrome DiGeorge syndrome Cartilage hair hypoplasia Ataxia telangiectasia X-linked SCID CHARGE syndrome RAG1 SCID/Omenn syndrome
*Reference ranges listed are age specific. CHARGE, coloboma of the eye, heart abnormalities, coanal atresia, retardation of growth and development, genitourinary anomalies, ear anomalies and deafness; DBS, dried blood spot; NK, natural killer; SCID, severe combined immunodeficiency; TCL, T-cell lymphopenia; TREC, T-cell receptor excision circle.
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
757
Vidal-Folch et al Table 3
SCID Newborn Screen Follow-Up Cohort
Sample Sex DBS017 DBS019 DBS020 DBS021 DBS022 DBS023 DBS024 DBS025 DBS026 DBS027 DBS029 DBS030 DBS031 DBS032 DBS037 DBS054 DBS055 DBS058 DBS059 DBS060 DBS061 DBS062 DBS063 DBS064 DBS065 DBS066 DBS067 DBS068 DBS069
Original NBS Age, days result
Male 5 Male 10 Female 18 Male 18 Male 9 Male 9 Male 17 Female 64 Female 48 Male 33 Female 7 Male 9 Male 115 Female 9 Female 6 Female 21 Male 62 Male 15 Female 17 Male 32 Male 65 Female 11 Female 27 Male 53 Male 5 Male 14 Female 37 Female 46 Female 10
Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal NA NA NA NA NA Abnormal NA Abnormal Abnormal NA Abnormal Abnormal
Whole blood results (reference range)
DBS results (reference range)
LymphocyteTREC copies/106 CD3þ T cells CD4þ T cells CD8þ T cells based cells (6794) (1484e5327) (733e3181) (370e2555) interpretation
RPP30 copies/mL TREC blood copies/mL blood (20) (1500)
4560 5628 7037 3227 5233 5809 5900 9830 7369 3155 5144 4911 4670 7957 7888 4359 839 22,188 NA NA 2998 16,406 <5 7487 17,552 7469 6247 2948 22,566
190 98 201 106 116 148 43 34 69 85 29 50 101 94 79 24 <4 145 80 25 14 33 <4 11 4 15 4 <4 155
4096 3463 4074 3189 3097 3418 3669 1091 1794 3230 1209 1090 2511 2039 1884 1831 150 4612 2022 717 2156 1266 963 1135 51 956 116 321 2751
2493 2718 2960 2646 2048 2649 2228 785 1534 2396 898 900 1775 1444 1658 1297 81 3424 1611 567 1108 843 898 687 39 672 70 277 2288
1444 778 1124 518 1036 760 1337 291 238 795 270 199 595 583 251 516 43 1115 412 152 926 418 24 362 12 286 37 42 451
Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal TCL Normal Normal TCL* TCL Normal SCID iTCL Secondary TCLy TCL TCL Leaky SCID Normal
22,188 19,863 21,700 17,988 14,763 19,350 9000 24,450 7813 17,463 11,150 35,375 17,175 25,488 49,500 11,375 6238 17,050 14,625 19,850 12,300 13,150 5025 54,875 7388 10,488 19,275 7438 19,538
*Final TCL diagnosis was unestablished because of absent TREC copies/106 cell levels. y Final secondary TCL diagnosis was unconfirmed as CD4 recent thymic emigrant % and clinical history were unavailable. iTCL, idiopathic TCL; NA, not available; NBS, newborn screening; SCID, severe combined immunodeficiency; TCL, T-cell lymphopenia; TREC, T-cell receptor excision circle.
were prepared in semiskirted PCR 96-well plates with 11 mL of 2 ddPCR Supermix for probes (no dUTP) (Bio-Rad, Pleasanton, CA), 1.1 mL of TREC assay mix, 1.1 mL of RPP30 assay mix, and 8.8 mL of DBS DNA. The final concentrations of primers and probes in the reaction were 900 and 250 nmol/L, respectively. Multiwell plates were sealed with foil at 180 C for 4 seconds using a Plate Sealer (Bio-Rad), vortexed briefly, centrifuged, and placed on the automated droplet generator, AutoDG (Bio-Rad), where the generation of droplets occurred by loading 20 mL of Reaction Mix and 70 mL of droplet oil into a DG-32 cartridge (Bio-Rad). After droplet generation, plates were resealed with the Plate Sealer. PCR amplification was performed on a Veriti Thermal Cycler (Applied Biosystems, Foster City, CA) using end point PCR conditions: 95 C for 10 minutes, followed by 40 cycles of denaturation at 94 C for 30 seconds, annealing and extension
at 58 C for 1 minute, and a final extension step at 98 C for 10 minutes. Plates were stored at 4 C until read on a QX200 Droplet Reader (Bio-Rad).
758
jmd.amjpathol.org
Data Analysis QuantaSoft software version 1.7.4 (Bio-Rad) was used for droplet cluster classification, and multiwell threshold was applied on a plate-by-plate basis. For each fluorophore, the fraction of positive droplets was fitted into a Poisson distribution equation, thereby providing absolute quantification of TREC and RPP30 PCR products per well without a standard curve. Rejection criteria for excluding a result from the data summary included a clog detected by the QX200 reader or <1500 acceptable droplets measured per well. An external calculation was performed to convert target copies
-
The Journal of Molecular Diagnostics
Droplet Digital PCR for SCID NBS female and male populations. TREC limits were determined from decision-threshold curves using Analyze-it Method Validation Edition Software version 4.81 (Analyze-it, Leeds, UK) for Microsoft Excel (Microsoft, Redmond, WA), which was also used for data management and statistical analysis.
Results Multiplex ddPCR Four droplet clusters were observed for every normal control specimen, as illustrated in Figure 2A. These clusters correspond to negative, TREC-positive, RPP30-positive, and TREC-positive/RPP30-positive (double-positive) droplets. Specimens from SCID patients resulted in the observation of only two clusters (Figure 2B). These clusters consisted of RPP30-positive and RPP30-negative droplets, as TREC-positive droplets were absent. If template DNA was withheld, negative droplets were only obtained (Figure 2C). A small amount of RPP30 signal, and no TREC-specific signal, was observed while investigating potential carryover from the entire analytical process (Figure 2D). Droplet saturation ranged from 30% to 96%, with an average of 70%. The number of droplets generated per reaction ranged from 3425 to 18,299, with an average of 11,670 droplets. Nine presumed-normal specimens were excluded from analysis because of an inadequate number of droplets generated with the first attempt and an insufficient amount of residual DBS specimen to repeat the analysis. Figure 1 Workflow of TREC ddPCR analysis. Isolated genomic DNA from DBSs is added to a PCR and partitioned into droplets by the AutoDG. After end-point amplification, droplets are singularized by the QX200 reader, and fluorescence is measured as FAM and HEX signal amplitudes. Pink lines indicate FAM and HEX signal amplitude thresholds applied to the analysis for designating positive or negative droplets.
per 20 mL well into copies/mL blood considering that one 3.2-mm disk contains 3 mL of blood, DNA is eluted in 15 mL of water, and 8 mL of DNA is used in the final PCR: (copies per 20 mL well/8 mL input DNA) (15 mL elution/3 mL blood per punch) Z copies/mL blood. Specimens were reported as SCID-screen negative (TREC content > the assay threshold) or SCID-screen positive (TREC content < the threshold). The detection capability for this assay was established following guidelines by the Clinical and Laboratory Standards Institute.24 The imprecision of the assay was evaluated following guidelines in the Clinical and Laboratory Standards Institute document EP05-A3.25
Statistical Analysis Pearson’s correlation test was used to analyze correlations between TREC and RPP30 levels. A Wilcoxon-Mann-Whitney test was used to compare TREC level mean values between
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
Performance Characteristics The limit of blank was established at 0 as TREC-positive droplets were not observed with 20 replicate measurements of blank Whatman 903 card punches. The limit of detection was established at 4 TRECs/mL using the Clinical and Laboratory Standards Institute formula calculation, where Limit of Detection Z Limit of Blank þ 1.645 SD,where the SD is the standard deviation from 20 replicate measurements of an adult DBS sample with only 5 TREC copies/mL blood.
Table 4
Primers and Probes for ddPCR
Reagent TREC Forward primer Reverse primer Probe RPP30 Forward primer Reverse primer Probe
Sequence 50 -CACATCCCTTTCAACCATGCT-30 50 -GCCAGCTGCAGGGTTTAGG-30 50 -FAM-ACACCTCTGGTTTTTGTAAAGGTGCCCACT-ZEN/IBFQ-30 50 -AGATTTGGACCTGCGAGCG-30 50 -GAGCGGCTGTCTCCACAAGT-30 50 -HEX-TTCTGACCTGAAGGCTCTGCGCG-ZEN/IBFQ-30
TREC, T-cell receptor excision circle.
759
Vidal-Folch et al
Figure 2
The clustering of droplets in the ddPCR analysis shown as a heat map. Larger numbers of droplets are represented in red. Droplet cloud thresholds are shown as pink lines. A: Negative droplets (NEG; absent TREC or RPP30 PCR products) are shown; blue represents positive droplets containing TREC PCR product, RPP30 PCR product, or double-positive droplets with both TREC and RPP30 PCR products. B: Lack of TREC-positive droplets is seen for an SCID-positive infant. C: Lack of TREC- or RPP30-positive droplets is seen for a sample without template DNA. D: The level of ddPCR carryover observed for a blank paper specimen run after a DBS with our NBS process. Pink lines indicate FAM and HEX signal amplitude thresholds applied to the analysis for designating positive or negative droplets.
Reportable Range
Imprecision
The theoretical dynamic range in ddPCR is largely determined by the number of partitions analyzed. Pinheiro et al26 established a ddPCR theoretical dynamic range at 105 target copies. However, we performed linearity studies by diluting extracted DNA from healthy infant DBSs into moleculargrade water. Twofold serial dilution experiments on multiple samples resulted in TREC levels fitted to a trend line, with R2 correlation coefficients >0.946 and a range of 400 to <4 TREC copies/mL blood.
Three DBS controls were prepared to span an expected analytical measuring range (an SCID-positive, an SCIDnegative, and a borderline SCID-negative specimen). Twenty replicate measurements were collected for each experiment. Intra-assay measurements resulted in a 17.2% CV at 34 TRECs/mL and a 16.9% CV at 389 TRECs/mL, whereas interassay measurements resulted in a 19.1% CV at 59 TRECs/mL and a 15.5% CV at 441 TRECs/mL. The SCID-positive control specimen remained at <4 TREC copies/mL blood in the intra-assay studies and ranged from <4 to 9 TREC copies/mL blood during interassay studies.
Stability Analyte stability on DBSs was evaluated as DNA fragmentation was reported to increase over time within stored dried blood spot cards.27 Seven DBS samples were tested after 1 to 120 days of storage (one SCID positive, four SCID negative, and two borderline SCID negative), and these specimens showed identical TREC recoveries when stored at room temperature, 4 C, 20 C, or up to four freeze-thaw cycles (data not shown). Yet, storage of DBSs at 37 C and 50 C affected assay performance by reducing TREC recovery after 7 days, indicating that long-term specimen storage is not recommended for temperatures higher than ambient.
Clinical Validation
760
jmd.amjpathol.org
DBSs from 10 infants diagnosed with SCID, or other TCL conditions, were confirmed to contain 15 TREC copies/ mL blood (Table 2 and Figure 3A). Conditions analyzed included X-linked SCID, RAG1 Omenn syndrome, idiopathic TCL, chromosome 22q11.2 deletion-negative TCL, DiGeorge syndrome, CHARGE syndrome, ataxia telangiectasia, and cartilage hair hypoplasia. The TREC content of SCID patients was lower than that of infants with other TCL conditions (Table 2). The clinical TREC threshold
-
The Journal of Molecular Diagnostics
Droplet Digital PCR for SCID NBS
Figure 3
Distribution of population-based reference ranges and affected cases. A and B: Dot plots show the range of TREC (A) and RPP30 (B) levels in the normal reference range study and for low TREC patients. Data plotted as log10. Dashed lines represent the TREC and RPP30 assay thresholds. C and D: Histograms illustrating the distribution of TREC and RPP30 copies in full-term infants observed during the population-based study.
was established at <20 TREC copies/mL blood. This threshold was chosen as a conservative value to capture TCL cases while limiting false-positive results, retesting, and unnecessary referrals (Figure 3). This threshold is consistent with other NBS programs using qPCR.28 RPP30 content ranged from 3013 to 23,550 copies/mL blood (Figure 3B). A minimum content of 1500 RPP30 copies/mL blood was set as a limit to ensure that sufficient DNA was obtained for TREC quantification.
NBS CDC Proficiency Testing TREC clinical assessment of 30 DBS specimens from the CDC’s Newborn Screening Quality Assurance Program (NSQAP) was 100% concordant with CDC results using our ddPCR method. Table 5 summarizes the reported frequency of clinical assessment by all participants in the CDC NSQAP over several challenges. Six specimens of 30 were expected to be SCID positive (TREC assessment classified as follow-up required, and RPP30 assessment classified as within reference range). However, two false negatives (specimens 216R2 and 416R1) and 16 false positives were reported by other NSQAP participants using either qPCR or the EnLite Neonatal TREC kit.
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
Reference Range Wilcoxon-Mann-Whitney analysis of 610 infant TREC levels showed no significant TREC mean differences between males or females. Thus, TREC and RPP30 thresholds were established regardless of sex. TREC content in infants ranged from 14 to 474 copies/mL blood, with a mean of 144 copies/mL blood and a median of 128 copies/mL blood, similar to previously published literature on TREC DBS content.28 RPP30 content ranged from 4000 to 62,500 copies/mL blood, with a mean of 17,690 copies/mL blood and a median of 16,269 copies/mL blood (Figure 3, C and D). The TREC 2.5% and 97.5% percentiles were calculated at 48 and 331 copies/mL blood, respectively. The RPP30 2.5% and 97.5% percentiles were calculated at 7134 and 35,857 copies/mL blood, respectively. The 610 infants’ ages ranged from a minimum of 1 to 133 days, and there was a moderate correlation between TREC and RPP30 levels, with a Pearson’s correlation coefficient of 0.352, as expected.
NBS Referral Specimens A total of 29 infants were referred for lymphocyte characterization at Mayo Clinic, ostensibly for abnormal NBS
761
Vidal-Folch et al Table 5
CDC SCID NBS Proficiency Results Mayo results (reference range)
CDC results RPP30 TREC assessment assessment
TREC copies/mL Specimen no. blood (20)
RPP30 copies/ mL blood (1500)
1, no follow-up required (screen negative); 2, follow-up required
415R1 415R2 415R3 415R4 415R5 116R1 116R2 116R3 116R4 116R5 216R1 216R2 216R3 216R4 216R5 316R1 316R2 316R3 316R4 316R5 416R1 416R2 416R3 416R4 416R5 117R1 117R2 117R3 117R4 117R5
18,988 26,750 329 6700 12,275 7800 12,213 34,625 186 18,200 17,950 12,413 16,375 12,675 391 19,550 4975 11,163 38 1638 8600 26,375 8125 29 12,900 16,150 23 52,500 9238 9138
1 1 2 2 1 2 1 1 2 1 1 2 1 1 2 1 2 1 2 2 2 1 1 2 1 1 2 1 1 2
66 169 <4 14 146 14 131 210 <4 108 179 <4 153 140 <4 165 15 63 <4 9 <4 188 29 <4 63 120 <4 224 70 <4
RPP30 TREC assessment assessment
1, no follow-up required (screen 1, within reference range; negative); 2, follow-up 2, outside reference range required
Participants with correct 1, Within NSQAP clinical reference range; assessment, % 2, outside (n incorrect/ reference range total)
1 1 2 1 1 1 1 1 2 1 1 1 1 1 2 1 1 1 2 1 1 1 1 2 1 1 2 1 1 1
1 1 2 1 1 1 1 1 2 1 1 1 1 1 2 1 1 1 2 1 1 1 1 2 1 1 2 1 1 1
1 1 2 2 1 2 1 1 2 1 1 2 1 1 2 1 2 1 2 2 2 1 1 2 1 1 2 1 1 2
100 98 (1/46) 100 100 100 100 100 100 100 100 100 98 (1/47) 100 100 100 100 100 100 100 100 98 (1/53) 96 (2/53) 98 (1/53) 100 98 (1/53) 94 (3/53) 100 98 (1/53) 87 (7/53) 100
NBS, newborn screening; NSQAP, Newborn Screening Quality Assurance Program; SCID, severe combined immunodeficiency; TREC, T-cell receptor excision circle.
results (Table 3). The reason for referral was unavailable in seven cases, whereas 22 were confirmed to have an abnormal NBS result for SCID. Our TREC assay confirmed five positive infants, whereas 17 infants were normal, with TREC levels ranging from 24 to 201 TREC copies/mL blood. In this cohort, 77% of NBS referrals sent to our department were false positives, as determined by follow-up lymphocyte characterization that was performed by experienced clinical immunologists using TREC copies/106 cells and CD3þ, CD4þ, and CD8þ T-cell criteria. Based on these NBS referral specimens, we calculated an 88.9% clinical sensitivity and a 100% clinical specificity. Unfortunately, the infant (DBS060) who presented with low T-cell counts was unable to be assigned a final diagnosis because of absent TREC values, which were likely obtained at another
institution. With this information, our calculated clinical sensitivity would likely be even higher. In turn, increasing our positive TREC threshold to 25 TREC copies/mL blood would reduce the specificity of our assay to 99.4% but have a sensitivity of 100%.
762
jmd.amjpathol.org
Discussion In this study, we demonstrated that ddPCR can be used to sensitively and specifically to measure TREC levels in a single DBS punch. The method’s limit of detection (4 TREC copies/mL blood) and imprecision (<20% CV) are sufficient to distinguish all typical SCID and leaky SCID as well as most other T-cell lymphopenia disorders, such as
-
The Journal of Molecular Diagnostics
Droplet Digital PCR for SCID NBS Newborn TREC Assay
TREC ≥ 20
TREC < 20
RPP30 <1500
RPP30 ≥ 1500
Suspected Positive Screen
Possible Unsatisfactory Specimen Repeat Assay
TREC ≥ 20
TREC < 20
RPP30 < 1500
RPP30 ≥ 1500
New Specimen Required
Positive
If preterm, follow NICU protocol
Negative
If full-term, order diagnostic testing
TBBS CD4RT TREC,B
Figure 4
The proposed SCID newborn screening algorithm with ddPCR. CD4RT, CD4 T-cell recent thymic emigrant; NICU, neonatal intensive care unit; TBBS, T- and B-cell quantification by flow cytometry; TREC, B, T-cell receptor excision circles analysis, blood.
idiopathic TCL, DiGeorge syndrome, CHARGE syndrome, and ataxia telangiectasia, tested to date. The addition of a simultaneously measured second analyte, RPP30, enhances confidence in a low TREC level (screen-positive result) without the necessity to punch multiple DBSs for repeat testing because of inefficient DNA extraction or PCR inhibitors. An additional benefit of this method is the ability to measure absolute TREC levels and remove reliance on the performance of a standard curve, which is integral to other published TREC qPCR methods. High-throughput newborn screening can be achieved with the proposed follow-up algorithm, illustrated in Figure 4. Unlike other published methods, our assay used an automated DNA extraction process that included an overnight DNA elution step. The addition of an extended elution was observed to increase our TREC yields, and to lower the limit of detection, than when using a shorter elution time. In contrast, TREC limit of detection by qPCR, and another DNA isolation method, has been reported to be as low as 19 TREC copies/mL blood.20 The elution step enabled our
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
technologists to complete specimen accessioning, DBS punching, and DNA extraction up to the point of elution, during 1 working day. The ddPCR section of the method can be completed within 3 hours the following day, and SCID-positive results were reportable in a manner that was consistent with the urgency of diagnostic confirmation and therapeutic intervention, while also fitting into our laboratory workflow. During the course of our study, we observed, as previously reported, that TREC levels diminish with increasing age.12 However, the TREC levels found in our reference range study were within TREC concentration ranges described for a handful of SCID newborn study experiences,7,29 even though our reference range population of full-term infants had a median age of 3 days at sample collection, and a range of 132 days, which is higher than a recommended testing age of 24 to 48 hours. In addition, as this was a retrospective study, our DBS samples were stored up to 1 month at ambient temperature. Nevertheless, our population study and related TREC threshold were sufficient to obtain 100% clinical specificity. There were two infant specimens tested during our population study with TREC levels at, or lower than, our screen-positive threshold of 20 TREC copies/mL blood (Figure 3A). Unfortunately, our population study was accomplished with deidentified, and anonymized, specimens, and we were unable to confirm whether these infants experienced concurrent T-cell lymphopenia at the time of specimen collection. We also could not collate these specific results with clinical specimens sent for characterization at our Cellular and Molecular Immunology Laboratory. Among prospectively tested specimens, we found only one TCL infant whose DBS TREC level of 25 copies/mL blood was higher than our initial screen-positive threshold. Blood T-cell characterization of this infant showed reduced levels of CD3þ, CD4þ, and CD8þ T cells within the same specimen. However, a cell-specific TREC level was not obtained, and we were unable to resolve whether this infant’s TREC copies/106 cells would be near normal range, as suggested by the DBS analysis. Thus, a final diagnosis for this infant remains unconfirmed. We included this infant’s test result in our calculation of clinical sensitivity at 88.9%, even though we were unable to confirm a final diagnosis of a condition that would be detectable by our testing method. Updating our threshold value to 25 TREC copies/mL blood would lead to the identification of this infant, and 100% sensitivity for the potential TCL conditions tested to date. However, there are advantages and disadvantages to increasing our screening threshold to include this unresolved case. A TREC concentration threshold at 25 TREC copies/mL blood would affect the specificity of the screening assay, and lead to more false-positive test results, with a calculated specificity of 99.4% after a single test. We aimed to keep our screening threshold at a level that would identify all confirmed cases of SCID or leaky SCID, while excluding any false-positive results (100%
763
Vidal-Folch et al specificity). Because of this balance, there is the potential that infants with immune defects might be missed, or that normal infants are preventably referred for additional laboratory testing. Past qPCR studies have reported specificities of >99%5,16,17,21 that were dependant on repetitive, multispecimen testing, and sensitivities that also rely on a balance, ranging from 87.5% to 100% contingent on the selected TREC threshold.5,21 In contrast, the CDC NSQAP showed that qPCR’s clinical sensitivities can range from 98% to 100% and clinical specificities from 87% to 100%, depending on a sample’s TREC level and the laboratory performing testing (Table 5). From a limited CDC NSQAP experience, ddPCR exceeded qPCR in screening performance, with 100% clinical sensitivity and 100% clinical specificity when assessing the same specimens. TREC levels alone are regrettably unable to identify all infants with primary immune defects.1,18,30 The inclusion of other analytical targets, such as B-cellespecific, k-deleting recombination excision circles or DBS adenosine, may aid to increase the sensitivity of NBS for extra primary immunodeficiencies, as reported elsewhere.6,21,31 Although the addition of molecular analyses to NBS may help to potentially detect a greater number of primary immunodeficiencies, current genomic methods are fraught with an increasing number of variants of unknown significance that will need functional characterization and disease correlation before widespread use of genomic methods for NBS.32,33 One shortfall of our study is that we did not sample a sufficient number of preterm infant DBS specimens to identify whether a specific preterm-infant threshold for TREC levels would be desirable for this method. Based on previous experience, we predict that using covariates and additional newborn screening marker results with an integrated report will aid SCID detection, as we described for other NBS conditions.34,35 Thus, a supplementary study with a larger prospective sample collection is ongoing and aims to determine whether the addition of covariates, such as sex, birth weight, gestational age, or other NBS condition test results, can add further improvements to SCID NBS detection. Another concern about our described method is the fact that the technology, ddPCR, is uncommon for Public Health Laboratories, and that associated assay costs might be inhibitory for population screening. In our limited proof-ofprinciple study, which did not have clinical volumes to appreciate any cost savings gained by an economy of scale, the material costs for our ddPCR method were calculated to be $0.46 for DNA extraction, and $3.80 for ddPCR, for a total of $4.30 per test. This figure is within per-test costs reportedly ranging from $4.22 to $6.00 for other SCID screening methods.13,36 The current hurdles for laboratories seeking to adopt ddPCR will be an initial investment into the clinical ddPCR testing platform and the scarceness of Federal Drug Administrationeapproved clinical testing kits for NBS. Regardless of the outcome of forthcoming studies, the ddPCR method we described herein is able to measure low
TREC levels in DBSs, is able to detect infants with SCID or leaky SCID and other TCLs, requires material expenditures that are comparable to current qPCR methods, and can be implemented for high-throughput screening.
764
jmd.amjpathol.org
Acknowledgments We thank Myra Wick, M.D., Ph.D., and the Mayo Clinic Umbilical Cord and Placental Biobank, for study sample collection and the CDC for the Newborn Screening Quality Assurance Program (NSQAP) experience.
References 1. Kwan A, Abraham RS, Currier R, Brower A, Andruszewski K, Abbott JK, et al: Newborn screening for severe combined immunodeficiency in 11 screening programs in the United States. JAMA 2014, 312:729e738 2. Dvorak CC, Cowan MJ, Logan BR, Notarangelo LD, Griffith LM, Puck JM, Shearer WT, O’Reilly RJ, Fleisher TA, Pai SY, Hanson IC, Pulsipher MA, Fuleihan R, Filipovich A, Goldman F, Kapoor N, Small T, Smith A, Chan KW, Cuvelier G, Heimall J, Knutsen A, Loechelt B, Moore T, Buckley RH: The natural history of children with severe combined immunodeficiency: baseline features of the first fifty patients of the primary immune deficiency treatment consortium prospective study 6901. J Clin Immunol 2013, 33:1156e1164 3. Pai SY, Logan BR, Griffith LM, Buckley RH, Parrott RE, Dvorak CC, et al: Transplantation outcomes for severe combined immunodeficiency, 2000-2009. N Engl J Med 2014, 371:434e446 4. Jilkina O, Thompson JR, Kwan L, Van Caeseele P, RockmanGreenberg C, Schroeder ML: Retrospective TREC testing of newborns with severe combined immunodeficiency and other primary immunodeficiency diseases. Mol Genet Metab Rep 2014, 1:324e333 5. Audrain M, Thomas C, Mirallie S, Bourgeois N, Sebille V, Rabetrano H, Durand-Zaleski I, Boisson R, Persyn M, Pierres C, Mahlaoui N, Fischer A: Evaluation of the T-cell receptor excision circle assay performances for severe combined immunodeficiency neonatal screening on Guthrie cards in a French single centre study. Clin Immunol 2014, 150:137e139 6. Barbaro M, Ohlsson A, Borte S, Jonsson S, Zetterström RH, King J, Winiarski J, von Döblen U, Hammarström L: Newborn screening for severe primary immunodeficiency diseases in Sweden: a 2-year pilot TREC and KREC screening study. J Clin Immunol 2017, 37:51e60 7. Adams SP, Rashid S, Premachandra T, Harvey K, Ifederu A, Wilson MC, Gaspar HB: Screening of neonatal UK dried blood spots using a duplex TREC screening assay. J Clin Immunol 2014, 34:323e330 8. Somech R, Lev A, Simon AJ, Korn D, Garty BZ, Amariglio N, Rechavi G, Almanshanu S, Zlotogora J, Etzioni A: Newborn screening for severe T and B cell immunodeficiency in Israel: a pilot study. Isr Med Assoc J 2013, 15:404e409 9. Morinishi Y, Imai K, Nakagawa N, Sato H, Horiuchi K, Ohtsuka Y, Kaneda Y, Taga T, Hisakawa H, Miyaji R, Endo M, Oh-Ishi T, Kamachi Y, Akahane KKobayashi C, Tsuchida M, Morio T, Sasahara Y, Kumaki S, Ishigaki K, Yoshida M, Urabe T, Kobayashi N, Okimoto Y, Reichenbach J, Hashii Y, Tsuji Y, Kogawa K, Yamaguchi S, Kanegane H, Miyawaki T, Yamada M, Ariga T, Nonoyama S: Identification of severe combined immunodeficiency by T-cell receptor excision circles quantification using neonatal guthrie cards. J Pediatr 2009, 155:829e833 10. Kong F, Chen CH, Cooper MD: Thymic function can be accurately monitored by the level of recent T cell emigrants in the circulation. Immunity 1998, 8:97e104 11. Douek DC, Vescio RA, Betts MR, Brenchley JM, Hill BJ, Zhang L, Berenson JR, Collins RH, Koup RA: Assessment of thymic output in
-
The Journal of Molecular Diagnostics
Droplet Digital PCR for SCID NBS
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
adults after haematopoietic stem-cell transplantation and prediction of T-cell reconstitution. Lancet 2000, 355:1875e1881 Douek DC, McFarland RD, Keiser PH, Gage EA, Massey JM, Haynes BF, Polis MA, Haase AT, Feinberg MB, Sullivan JL, Jamieson BD, Zack JA, Picker LJ, Koup RA: Changes in thymic function with age and during the treatment of HIV infection. Nature 1998, 396:690e695 Baker MW, Grossman WJ, Laessig RH, Hoffman GL, Brokopp CD, Kurtycz DF, Cogley MF, Litsheim TJ, Katcher ML, Routes JM: Development of a routine newborn screening protocol for severe combined immunodeficiency. J Allergy Clin Immunol 2009, 124:522e527 Routes JM, Grossman WJ, Verbsky J, Laessig RH, Hoffman GL, Brokopp CD, Baker MW: Statewide newborn screening for severe Tcell lymphopenia. JAMA 2009, 302:2465e2470 Picard C, Al-Herz W, Bousfiha A, Casanova JL, Chatila T, Conley ME, Cunningham-Rundels C, Etzioni A, Holland A, Klein C, Nonoyama S, Ochs HD, Oksenhendler E, Puck JM, Sullivan KE, Tang ML, Franco JL, Gaspar HB: Primary immunodeficiency diseases: an update on the classification from the International Union of Immunological Societies Expert Committee for Primary Immunodeficiency 2015. J Clin Immunol 2015, 35:696e726 Chien YH, Chiang SC, Chang KL, Yu HH, Lee WI, Tsai LP, Hsu MH, Hwu WL: Incidence of severe combined immunodeficiency through newborn screening in a Chinese population. J Formos Med Assoc 2015, 114:12e16 Verbsky J, Thakar M, Routes J: The Wisconsin approach to newborn screening for severe combined immunodeficiency. J Allergy Clin Immunol 2012, 129:622e627 Kwan A, Church JA, Cowan MJ, Agarwal R, Kapoor N, Kohn DB, Lewis DB, McGhee SA, Moore TB, Stiehm ER, Porteus M, Aznar CP, Currier R, Lorey F, Puck JM: Newborn screening for severe combined immunodeficiency and T-cell lymphopenia in California: results of the first 2 years. J Allergy Clin Immunol 2013, 132:140e150 Vogel BH, Bonagura V, Weinberg GA, Ballow M, Isabelle J, DiAntonio L, Parkar A, Young A, Cunningham-Rundels C, Fong CT, Celestin J, Lehman H, Rubinstein A, Siegel S, Weiner L, SaavedraMatiz C, Kay DM, Caggana M: Newborn screening for SCID in New York State: experience from the first two years. J Clin Immunol 2014, 34:289e303 Gerstel-Thompson JL, Wilkey JF, Baptiste JC, Navas JS, Pai SY, Pass KA, Eaton RB, Comeau AM: High-throughput multiplexed T-cell-receptor excision circle quantitative PCR assay with internal controls for detection of severe combined immunodeficiency in population-based newborn screening. Clin Chem 2010, 56:1466e1474 Borte S, von Döbeln U, Fasth A, Wang N, Janzi M, Winiarski J, Sack U, Pan-Hammarström Q, Borte M, Hammarström L: Neonatal screening for severe primary immunodeficiency diseases using highthroughput triplex real-time PCR. Blood 2012, 119:2552e2555 Taylor JL, Lee FK, Yazdanpanah GK, Staropoli JF, Liu M, Carulli JP, Sun C, Dobrowolski SF, Hannon WH, Vogt RF: Newborn blood spot screening test using multiplexed real-time PCR to simultaneously screen for spinal muscular atrophy and severe combined immunodeficiency. Clin Chem 2015, 61:412e419 Hudecova I: Digital PCR analysis of circulating nucleic acids. Clin Biochem 2015, 48:948e956
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
24. Clinical and Laboratory Standards Institute. Evaluation of detection capability for clinical laboratory measurement procedures. EP17-A2. Wayne, PA: CLSI, 2012. Available at https://clsi.org/standards/ products/method-evaluation/documents/ep17 25. Clinical and Laboratory Standards Institute. Evaluation of precision of quantitative measurement procedures: approved guideline, ed 3. EP05A3. Wayne, PA: CLSI, 2014. Available at https://clsi.org/ standards/products/method-evaluation/documents/ep05 26. Pinheiro LB, Coleman VA, Hindson CM, Herrmann J, Hindson BJ, Bhat S, Emslie KR: Evaluation of a droplet digital polymerase chain reaction format for DNA copy number quantification. Anal Chem 2012, 84:1003e1011 27. Makowski GS, Davis EL, Hopfer SM: The effect of storage on Guthrie cards: implications for deoxyribonucleic acid amplification. Ann Clin Lab Sci 1996, 26:458e469 28. van der Spek J, Groenwold RH, van der Burg M, van Montfrans JM: TREC based newborn screening for severe combined immunodeficiency disease: a systematic review. J Clin Immunol 2015, 35: 416e430 29. Baker MW, Laessig RH, Katcher ML, Routes JM, Grossman WJ, Verbsky J, Kurtycz DF, Brokopp CD: Implementing routine testing for severe combined immunodeficiency within Wisconsin’s newborn screening program. Public Health Rep 2010, 125 Suppl 2:88e95 30. Kuo CY, Chase J, Garcia Lloret M, Stiehm ER, Moore T, Aguilera MJ, Lopez Sriles J, Church JA: Newborn screening for severe combined immunodeficiency does not identify bare lymphocyte syndrome. J Allergy Clin Immunol 2013, 131:1693e1695 31. la Marca G, Canessa C, Giocaliere E, Romano F, Duse M, Malvagia S, Lippi F, Finghini S, Bianchi L, Della Bona ML, Valleriani C, Ombrone D, Moriondo M, Villanelli F, Speckmann C, Adams S, Gasper BH, Hershfield M, Santisteban I, Fairbanks L, Ragusa G, Resti M, de Martino M, Guerrini R, Azzari C: Tandem mass spectrometry, but not T-cell receptor excision circle analysis, identifies newborns with late-onset adenosine deaminase deficiency. J Allergy Clin Immunol 2013, 131:1604e1610 32. Narravula A, Garber KB, Askree SH, Hegde M, Hall PL: Variants of uncertain significance in newborn screening disorders: implications for large-scale genomic sequencing. Genet Med 2017, 19:77e82 33. Friedman JM, Cornel MC, Goldenberg AJ, Lister KJ, Sénécal K, Vears DF; Global Alliance for Genomics and Health Regulatory and Ethics Working Group Paediatric Task Team: Genomic newborn screening: public health policy considerations and recommendations. BMC Med Genomics 2017, 10:9 34. Marquardt G, Currier R, McHugh DM, Gavrilov D, Magera MJ, Matern D, et al: Enhanced interpretation of newborn screening results without analyte cutoff values. Genet Med 2012, 14:648e655 35. Hall PL, Marquardt G, McHugh DM, Currier RJ, Tang H, Stoway SD, Rinaldo P: Postanalytical tools improve performance of newborn screening by tandem mass spectrometry. Genet Med 2014, 16: 889e895 36. Chan K, Davis J, Pai SY, Bonilla FA, Puck JM, Apkon M: A Markov model to analyze cost-effectiveness of screening for severe combined immunodeficiency (SCID). Mol Genet Metab 2011, 104: 383e389
765
The Journal of Molecular Diagnostics, Vol. 19, No. 5, September 2017
jmd.amjpathol.org
A Droplet Digital PCR Method for Severe Combined Immunodeficiency Newborn Screening Noemi Vidal-Folch,* Dragana Milosevic,* Ramanath Majumdar,* Dimitar Gavrilov,*y Dietrich Matern,*yz Kimiyo Raymond,*y Piero Rinaldo,*yz Silvia Tortorelli,*y Roshini S. Abraham,*x and Devin Oglesbee*y From the Departments of Laboratory Medicine and Pathology,* Clinical Genomics,y Pediatric and Adolescent Medicine,z and Allergy and Immunology,x Mayo Clinic College of Medicine, Rochester, Minnesota CME Accreditation Statement: This activity (“The JMD 2017 CME Program in Molecular Diagnostics”) has been planned and implemented in accordance with the Essential Areas and policies of the Accreditation Council for Continuing Medical Education (ACCME) through the joint providership of the American Society for Clinical Pathology (ASCP) and the American Society for Investigative Pathology (ASIP). ASCP is accredited by the ACCME to provide continuing medical education for physicians. The ASCP designates this journal-based CME activity (“The JMD 2017 CME Program in Molecular Diagnostics”) for a maximum of 36 AMA PRA Category 1 Credit(s). Physicians should only claim credit commensurate with the extent of their participation in the activity. CME Disclosures: The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose.
Accepted for publication May 30, 2017. Address correspondence to Devin Oglesbee, Ph.D., Clinical Biochemical Genetics Laboratory, Department of Laboratory Medicine and Pathology, Mayo Clinic, 200 First St SW, Rochester, MN 55905. E-mail: [email protected].
Severe combined immunodeficiency (SCID) benefits from early intervention via hematopoietic cell transplantation to reverse T-cell lymphopenia (TCL). Newborn screening (NBS) programs use T-cell receptor excision circle (TREC) levels to detect SCID. Real-time quantitative PCR is often performed to quantify TRECs in dried blood spots (DBSs) for NBS. Yet, real-time quantitative PCR has inefficiencies necessitating normalization, repeat analyses, or standard curves. To address these issues, we developed a multiplex, droplet digital PCR (ddPCR) method for measuring absolute TREC amounts in one DBS punch. TREC and RPP30 levels were simultaneously measured with a Bio-Rad AutoDG and QX200 ddPCR system. DBSs from 610 presumed-normal, 29 lymphocyte-profiled, and 10 clinically diagnosed infants (1 X-linked SCID, 1 RAG1 Omenn syndrome, and other conditions) were tested. Control infants showed 14 to 474 TREC copies/mL blood. SCID infants, and other TCL conditions, had 15 TREC copies/mL. The ddPCR lower limit of quantitation was 14 TREC copies/mL, and the limit of detection was 4 TREC copies/mL. Intra-assay and interassay imprecision was <20% CV for DBSs at 54 to 60 TREC copies/mL. Testing 29 infants with known lymphocyte profiles resulted in a sensitivity of 88.9% and a specificity of 100% at TRECs <20 copies/mL. We developed a multiplex ddPCR method for the absolute quantitation of DBS TRECs that can detect SCID and other TCL conditions associated with absent or low TRECs and validated this method for NBS. (J Mol Diagn 2017, 19: 755e765; http://dx.doi.org/10.1016/j.jmoldx.2017.05.011)
Severe combined immunodeficiency (SCID) disease is a group of inherited immunodeficiencies characterized by T- and B-cell lymphopenia. The overall incidence of SCID is reported to be approximately 1:58,000.1 Early detection of SCID is paramount; otherwise, severe and recurrent infections may be fatal within the first few years of life.2 Intervention by hematopoietic stem cell transplantation has changed the clinical outcome for early diagnosed SCID infants, as recent studies have shown.3 Survival rates of >80% for SCID infants treated by hematopoietic stem cell transplantation within the first few months of life with, or without, resolved infections are in stark contrast to reported
survival rates of only 50% for older infants with an active infection during transplant. Hence, early diagnosis of SCID is vital for achieving improved treatment success. Given the severity of SCID-related symptoms and improved outcomes after an early diagnosis, newborn screening (NBS) fulfills classic Wilson-Jungner screening criteria, and SCID was deemed a suitable candidate for NBS by the US Department of Health and Human Services, per a Supported by Mayo Clinic Department of Laboratory Medicine and Pathology internal funding. Disclosures: None declared.
Copyright ª 2017 American Society for Investigative Pathology and the Association for Molecular Pathology. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jmoldx.2017.05.011
Vidal-Folch et al letter issued on February 25, 2010, to the Secretary of Health and Human Services. Early detection and diagnosis is currently available by newborn screening programs in most US states.1 Newborn screening for SCID has also been evaluated by studies at several other locations in North America, Europe, and Asia.4e9 Current SCID screening methods exploit T-cell receptor excision circles (TRECs) that are a DNA by-product of T-cell receptor recombination and reflect T-cell maturity.10e12 Low TREC amounts mirror concurrent T-cell lymphopenia (TCL), and neonatal dried blood spot (DBS) TREC levels can be used to detect impaired T-cell development, and SCID by newborn screening.13,14 After a positive NBS result, diagnostic follow-up is required to determine whether the infant has typical SCID, or another form of TCL, and to establish a specific diagnosis.15 SCID can be detected in DBSs using real-time quantitative PCR (qPCR) for TRECs as a single-plex,16e19 duplex,5,20 triplex,21 or combined with targets for other conditions, such as an SMN1 exon 7 deletion for spinal muscular atrophy.22 Presently, there is no standardized SCID newborn screening method and the ideal newborn screening assay for SCID must be able to distinguish specimens with absent, or low, TREC content from those with insufficient DNA yield. The variability in newborn DBS specimen collection adds to the challenge of distinguishing samples with abnormal TREC levels. Because most SCID assays are based on PCR amplification, low results could be because of artifacts, such as an inadequate DBS sample, failure to elute a sufficient DNA amount, or the presence of a DNA polymerase inhibitor. To ensure that a low TREC result is not because of an artifact (causing a false-positive screening result), most TREC assays include simultaneous, or sequential, amplification of a conserved genomic sequence, like the ribonuclease P gene, RPP30, or the b-actin gene, ACTB.13,20 Detection of a reference gene within the expected range indicates that the tested specimen is amplifiable and TREC quantification is reliable. Another sensitive method to detect low concentrations of a DNA target within a single analysis is droplet digital PCR (ddPCR).23 With ddPCR, a single DNA sample is partitioned into thousands of uniform-size droplets and subsequently amplified by standard PCR, and a signal from a fluorescently labeled probe within each droplet is recorded as either positive or negative, depending on the presence or absence of a target’s PCR product. This results in specific and sensitive fluorescent signals derived from thousands of individual droplets containing target product. Negative droplets do not provide a fluorescent signal as they are void of target DNA amplicon. The application of a Poisson distribution calculation across all droplets permits an absolute concentration determination for a DNA target and eliminates the need for standard curves of reference standards or endogenous controls. Thousands of replicate measurements of each sample occur during each run. Last, the application of differentiated labeled fluorescent probes allows for
multiplexing PCR targets, such as TRECs and RPP30, within a single DNA tube and from the same DBS punch. Given the efficiencies promised by ddPCR, we designed a ddPCR assay to simultaneously measure the absolute level of TRECs and RPP30. Validation of this method is described herein.
756
jmd.amjpathol.org
Materials and Methods Infant Specimens Infant and cord blood specimens were collected with Mayo Clinic Institutional Review Board approval. A population-based TREC reference range was determined from 610 (541 full-term and 69 preterm) residual, anonymized DBSs collected onto Whatman 903 Protein Saver Cards (GE Health Care, Pittsburgh, PA) whose demographics are shown in Table 1. Infants with a gestational age <37 weeks were considered preterm. DBSs were also prepared from 10 infants diagnosed with X-linked SCID, Omenn syndrome, idiopathic TCL, chromosome 22q11.2 deletion-negative TCL, DiGeorge syndrome, CHARGE (coloboma of the eye, heart abnormalities, coanal atresia, retardation of growth and development, genitourinary anomalies, ear anomalies and deafness) syndrome, ataxia telangiectasia, and cartilage hair hypoplasia, whose demographic data are listed in Table 1 and TREC results in Table 2. Moreover, 29 blood samples submitted to the Mayo Clinic for follow-up testing, after NBS results, were anonymized and analyzed with data shown in Table 3.
Control Specimens Three levels of controls were prepared as DBSs to evaluate the performance characteristics of the TREC assay: i) an SCID-like sample, with TREC content lower than the expected range for newborns, but normal RPP30 levels; ii) a borderline negative control with TREC content adjacent to the assay’s threshold between normal and abnormal (decreased but not absent TREC), and normal RPP30 levels; iii) a negative control with TREC content within the expected newborn range, and normal RPP30 levels. In addition, DBSs were prepared for stability, accuracy, linearity, imprecision, clinical sensitivity, and specificity studies by spotting 75 mL of EDTA whole blood onto Whatman 903 cards and drying overnight at room temperature. DBSs were stored at 20 C with desiccant in sealed plastic bags. Positive-control DBS cards were made from anonymized adult EDTA whole blood, and negative-control DBS cards were made from donor umbilical cord blood, both collected with institutional review board approval. Last, 30 DBSs were obtained from the CDC’s Newborn Screening Quality Assurance Program for TREC Analysis in DBS to assess performance against other laboratories using qPCR or an EnLite Neonatal TREC kit (Perkin Elmer, Waltham, MA). Every quarter, the CDC distributes a panel of five unknown
-
The Journal of Molecular Diagnostics
Droplet Digital PCR for SCID NBS Table 1
Demographics of Presumed Normals and Low TREC Subjects
Cohort/specimen
Sex
Age at collection, days
Gestational age, weeks
Birth weight, g
Full-term infants (n Z 541) Preterm infants (n Z 69) DBS001 DBS002 DBS003 DBS004 DBS005 DBS006 DBS007 DBS008 DBS009 DBS010
272 females, 264 males, 5 not provided 30 Females, 39 males Male Female Male Male Male Female Female Male Male Male
1e133 (3)
37e41 (38)
1660e4680 (3020)
1e74 (7) 15 41 8 8 6 8 8 6 22 12
27e36 (36) NA NA 40 NA 37 NA NA NA NA NA
1350e3885 (2290) 4069 NA 4454 3852 3300 NA NA NA NA NA
Median values are shown in parenthesis. NA, not available; TREC, T-cell receptor excision circle.
DBS specimens to domestic, international, and manufacturer laboratories to analyze TREC content in peripheral blood.
DNA Extraction An overview of the ddPCR assay workflow is shown in Figure 1. A DNA extraction protocol was established for this study. Briefly, a single 3.2-mm disk was punched from DBSs into a semiskirted PCR 96-well plate (Eppendorf, Hamburg, Germany). DBS disks were washed twice by adding 120 mL Generation DNA Purification solution (Qiagen, Valencia, CA) and heating to 50 C for 30 minutes. After washing, 120 mL Generation DNA Elution buffer (Qiagen) was added, and the plate was incubated at room temperature for 15 minutes. After removing the Generation Table 2
DNA solution, 15 mL of water was distributed per well, and the plate was heated to 99 C for 25 minutes to release DNA. Last, the plate was stored overnight at 4 C, followed by a high-speed centrifugation at 3500 g for 10 minutes to remove debris.
ddPCR The ddPCR TREC primers and probe sequences were the same as those described by Douek et al.11 RPP30 primers and probes were newly designed for this assay (Table 4). Both TREC and RPP30 ddPCR assay mixes were custom synthesized by IDT (Coralville, IA). Amplicon sizes for the expected PCR products for TREC and RPP30 were 109 and 65 bp, respectively. The ddPCR mixture was assembled by an automated liquid handler system. Reaction mixtures of 22 mL
SCID Screen-Positive Cohort DBS results (reference range)*
Whole blood results (reference range)* CD16þ56þ NK cells, CD45RA CD45RO % of CD4 % of CD4 cells/mL T cells (43e256) T cells
CD4þ31þ 45RAþ % of CD4þ 45RAþ T cells Clinical diagnosis
RPP30 copies/mL TREC blood copies/mL blood (<20) (1500)
842
1480
80
15
52
10
14,550
229
202
158
84
11
57
5
23,550
429
379
417
549
73
20
46
15
17,775
986 248 116 914 3 307 408
314 94 42 495 18 69 15
1226 252 795 113 325 1006 2
446 159 1681 1418 42 317 1420
54 68 50 59 0 74 <1
30 12 41 39 94 26 95
82 57 76 63 0 47 0
6 4 <4 11 <4 <4 <4
13,413 10,500 11,925 7513 3013 5438 8150
TREC copies/106 CD3þ T cells, CD4þ T cells, cells/mL CD3 T cells cells/mL (1484e5327) (733e3181) Specimen (6794)
CD8þ T cells, cells/mL (370e2555)
DBS001
1220
1044
729
319
DBS002
998
912
686
DBS003
2201
814
DBS004 DBS005 DBS006 DBS007 DBS008 DBS009 DBS010
2192 1241 1090 5698 <5 1836 <5
1293 344 174 1436 24 382 424
CD19þ B cells, cells/mL (370e2306)
Idiopathic T-cell lymphopenia 22q11Deletion-negative TCL Idiopathic T-cell lymphopenia DiGeorge syndrome DiGeorge syndrome Cartilage hair hypoplasia Ataxia telangiectasia X-linked SCID CHARGE syndrome RAG1 SCID/Omenn syndrome
*Reference ranges listed are age specific. CHARGE, coloboma of the eye, heart abnormalities, coanal atresia, retardation of growth and development, genitourinary anomalies, ear anomalies and deafness; DBS, dried blood spot; NK, natural killer; SCID, severe combined immunodeficiency; TCL, T-cell lymphopenia; TREC, T-cell receptor excision circle.
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
757
Vidal-Folch et al Table 3
SCID Newborn Screen Follow-Up Cohort
Sample Sex DBS017 DBS019 DBS020 DBS021 DBS022 DBS023 DBS024 DBS025 DBS026 DBS027 DBS029 DBS030 DBS031 DBS032 DBS037 DBS054 DBS055 DBS058 DBS059 DBS060 DBS061 DBS062 DBS063 DBS064 DBS065 DBS066 DBS067 DBS068 DBS069
Original NBS Age, days result
Male 5 Male 10 Female 18 Male 18 Male 9 Male 9 Male 17 Female 64 Female 48 Male 33 Female 7 Male 9 Male 115 Female 9 Female 6 Female 21 Male 62 Male 15 Female 17 Male 32 Male 65 Female 11 Female 27 Male 53 Male 5 Male 14 Female 37 Female 46 Female 10
Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal Abnormal NA NA NA NA NA Abnormal NA Abnormal Abnormal NA Abnormal Abnormal
Whole blood results (reference range)
DBS results (reference range)
LymphocyteTREC copies/106 CD3þ T cells CD4þ T cells CD8þ T cells based cells (6794) (1484e5327) (733e3181) (370e2555) interpretation
RPP30 copies/mL TREC blood copies/mL blood (20) (1500)
4560 5628 7037 3227 5233 5809 5900 9830 7369 3155 5144 4911 4670 7957 7888 4359 839 22,188 NA NA 2998 16,406 <5 7487 17,552 7469 6247 2948 22,566
190 98 201 106 116 148 43 34 69 85 29 50 101 94 79 24 <4 145 80 25 14 33 <4 11 4 15 4 <4 155
4096 3463 4074 3189 3097 3418 3669 1091 1794 3230 1209 1090 2511 2039 1884 1831 150 4612 2022 717 2156 1266 963 1135 51 956 116 321 2751
2493 2718 2960 2646 2048 2649 2228 785 1534 2396 898 900 1775 1444 1658 1297 81 3424 1611 567 1108 843 898 687 39 672 70 277 2288
1444 778 1124 518 1036 760 1337 291 238 795 270 199 595 583 251 516 43 1115 412 152 926 418 24 362 12 286 37 42 451
Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal TCL Normal Normal TCL* TCL Normal SCID iTCL Secondary TCLy TCL TCL Leaky SCID Normal
22,188 19,863 21,700 17,988 14,763 19,350 9000 24,450 7813 17,463 11,150 35,375 17,175 25,488 49,500 11,375 6238 17,050 14,625 19,850 12,300 13,150 5025 54,875 7388 10,488 19,275 7438 19,538
*Final TCL diagnosis was unestablished because of absent TREC copies/106 cell levels. y Final secondary TCL diagnosis was unconfirmed as CD4 recent thymic emigrant % and clinical history were unavailable. iTCL, idiopathic TCL; NA, not available; NBS, newborn screening; SCID, severe combined immunodeficiency; TCL, T-cell lymphopenia; TREC, T-cell receptor excision circle.
were prepared in semiskirted PCR 96-well plates with 11 mL of 2 ddPCR Supermix for probes (no dUTP) (Bio-Rad, Pleasanton, CA), 1.1 mL of TREC assay mix, 1.1 mL of RPP30 assay mix, and 8.8 mL of DBS DNA. The final concentrations of primers and probes in the reaction were 900 and 250 nmol/L, respectively. Multiwell plates were sealed with foil at 180 C for 4 seconds using a Plate Sealer (Bio-Rad), vortexed briefly, centrifuged, and placed on the automated droplet generator, AutoDG (Bio-Rad), where the generation of droplets occurred by loading 20 mL of Reaction Mix and 70 mL of droplet oil into a DG-32 cartridge (Bio-Rad). After droplet generation, plates were resealed with the Plate Sealer. PCR amplification was performed on a Veriti Thermal Cycler (Applied Biosystems, Foster City, CA) using end point PCR conditions: 95 C for 10 minutes, followed by 40 cycles of denaturation at 94 C for 30 seconds, annealing and extension
at 58 C for 1 minute, and a final extension step at 98 C for 10 minutes. Plates were stored at 4 C until read on a QX200 Droplet Reader (Bio-Rad).
758
jmd.amjpathol.org
Data Analysis QuantaSoft software version 1.7.4 (Bio-Rad) was used for droplet cluster classification, and multiwell threshold was applied on a plate-by-plate basis. For each fluorophore, the fraction of positive droplets was fitted into a Poisson distribution equation, thereby providing absolute quantification of TREC and RPP30 PCR products per well without a standard curve. Rejection criteria for excluding a result from the data summary included a clog detected by the QX200 reader or <1500 acceptable droplets measured per well. An external calculation was performed to convert target copies
-
The Journal of Molecular Diagnostics
Droplet Digital PCR for SCID NBS female and male populations. TREC limits were determined from decision-threshold curves using Analyze-it Method Validation Edition Software version 4.81 (Analyze-it, Leeds, UK) for Microsoft Excel (Microsoft, Redmond, WA), which was also used for data management and statistical analysis.
Results Multiplex ddPCR Four droplet clusters were observed for every normal control specimen, as illustrated in Figure 2A. These clusters correspond to negative, TREC-positive, RPP30-positive, and TREC-positive/RPP30-positive (double-positive) droplets. Specimens from SCID patients resulted in the observation of only two clusters (Figure 2B). These clusters consisted of RPP30-positive and RPP30-negative droplets, as TREC-positive droplets were absent. If template DNA was withheld, negative droplets were only obtained (Figure 2C). A small amount of RPP30 signal, and no TREC-specific signal, was observed while investigating potential carryover from the entire analytical process (Figure 2D). Droplet saturation ranged from 30% to 96%, with an average of 70%. The number of droplets generated per reaction ranged from 3425 to 18,299, with an average of 11,670 droplets. Nine presumed-normal specimens were excluded from analysis because of an inadequate number of droplets generated with the first attempt and an insufficient amount of residual DBS specimen to repeat the analysis. Figure 1 Workflow of TREC ddPCR analysis. Isolated genomic DNA from DBSs is added to a PCR and partitioned into droplets by the AutoDG. After end-point amplification, droplets are singularized by the QX200 reader, and fluorescence is measured as FAM and HEX signal amplitudes. Pink lines indicate FAM and HEX signal amplitude thresholds applied to the analysis for designating positive or negative droplets.
per 20 mL well into copies/mL blood considering that one 3.2-mm disk contains 3 mL of blood, DNA is eluted in 15 mL of water, and 8 mL of DNA is used in the final PCR: (copies per 20 mL well/8 mL input DNA) (15 mL elution/3 mL blood per punch) Z copies/mL blood. Specimens were reported as SCID-screen negative (TREC content > the assay threshold) or SCID-screen positive (TREC content < the threshold). The detection capability for this assay was established following guidelines by the Clinical and Laboratory Standards Institute.24 The imprecision of the assay was evaluated following guidelines in the Clinical and Laboratory Standards Institute document EP05-A3.25
Statistical Analysis Pearson’s correlation test was used to analyze correlations between TREC and RPP30 levels. A Wilcoxon-Mann-Whitney test was used to compare TREC level mean values between
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
Performance Characteristics The limit of blank was established at 0 as TREC-positive droplets were not observed with 20 replicate measurements of blank Whatman 903 card punches. The limit of detection was established at 4 TRECs/mL using the Clinical and Laboratory Standards Institute formula calculation, where Limit of Detection Z Limit of Blank þ 1.645 SD,where the SD is the standard deviation from 20 replicate measurements of an adult DBS sample with only 5 TREC copies/mL blood.
Table 4
Primers and Probes for ddPCR
Reagent TREC Forward primer Reverse primer Probe RPP30 Forward primer Reverse primer Probe
Sequence 50 -CACATCCCTTTCAACCATGCT-30 50 -GCCAGCTGCAGGGTTTAGG-30 50 -FAM-ACACCTCTGGTTTTTGTAAAGGTGCCCACT-ZEN/IBFQ-30 50 -AGATTTGGACCTGCGAGCG-30 50 -GAGCGGCTGTCTCCACAAGT-30 50 -HEX-TTCTGACCTGAAGGCTCTGCGCG-ZEN/IBFQ-30
TREC, T-cell receptor excision circle.
759
Vidal-Folch et al
Figure 2
The clustering of droplets in the ddPCR analysis shown as a heat map. Larger numbers of droplets are represented in red. Droplet cloud thresholds are shown as pink lines. A: Negative droplets (NEG; absent TREC or RPP30 PCR products) are shown; blue represents positive droplets containing TREC PCR product, RPP30 PCR product, or double-positive droplets with both TREC and RPP30 PCR products. B: Lack of TREC-positive droplets is seen for an SCID-positive infant. C: Lack of TREC- or RPP30-positive droplets is seen for a sample without template DNA. D: The level of ddPCR carryover observed for a blank paper specimen run after a DBS with our NBS process. Pink lines indicate FAM and HEX signal amplitude thresholds applied to the analysis for designating positive or negative droplets.
Reportable Range
Imprecision
The theoretical dynamic range in ddPCR is largely determined by the number of partitions analyzed. Pinheiro et al26 established a ddPCR theoretical dynamic range at 105 target copies. However, we performed linearity studies by diluting extracted DNA from healthy infant DBSs into moleculargrade water. Twofold serial dilution experiments on multiple samples resulted in TREC levels fitted to a trend line, with R2 correlation coefficients >0.946 and a range of 400 to <4 TREC copies/mL blood.
Three DBS controls were prepared to span an expected analytical measuring range (an SCID-positive, an SCIDnegative, and a borderline SCID-negative specimen). Twenty replicate measurements were collected for each experiment. Intra-assay measurements resulted in a 17.2% CV at 34 TRECs/mL and a 16.9% CV at 389 TRECs/mL, whereas interassay measurements resulted in a 19.1% CV at 59 TRECs/mL and a 15.5% CV at 441 TRECs/mL. The SCID-positive control specimen remained at <4 TREC copies/mL blood in the intra-assay studies and ranged from <4 to 9 TREC copies/mL blood during interassay studies.
Stability Analyte stability on DBSs was evaluated as DNA fragmentation was reported to increase over time within stored dried blood spot cards.27 Seven DBS samples were tested after 1 to 120 days of storage (one SCID positive, four SCID negative, and two borderline SCID negative), and these specimens showed identical TREC recoveries when stored at room temperature, 4 C, 20 C, or up to four freeze-thaw cycles (data not shown). Yet, storage of DBSs at 37 C and 50 C affected assay performance by reducing TREC recovery after 7 days, indicating that long-term specimen storage is not recommended for temperatures higher than ambient.
Clinical Validation
760
jmd.amjpathol.org
DBSs from 10 infants diagnosed with SCID, or other TCL conditions, were confirmed to contain 15 TREC copies/ mL blood (Table 2 and Figure 3A). Conditions analyzed included X-linked SCID, RAG1 Omenn syndrome, idiopathic TCL, chromosome 22q11.2 deletion-negative TCL, DiGeorge syndrome, CHARGE syndrome, ataxia telangiectasia, and cartilage hair hypoplasia. The TREC content of SCID patients was lower than that of infants with other TCL conditions (Table 2). The clinical TREC threshold
-
The Journal of Molecular Diagnostics
Droplet Digital PCR for SCID NBS
Figure 3
Distribution of population-based reference ranges and affected cases. A and B: Dot plots show the range of TREC (A) and RPP30 (B) levels in the normal reference range study and for low TREC patients. Data plotted as log10. Dashed lines represent the TREC and RPP30 assay thresholds. C and D: Histograms illustrating the distribution of TREC and RPP30 copies in full-term infants observed during the population-based study.
was established at <20 TREC copies/mL blood. This threshold was chosen as a conservative value to capture TCL cases while limiting false-positive results, retesting, and unnecessary referrals (Figure 3). This threshold is consistent with other NBS programs using qPCR.28 RPP30 content ranged from 3013 to 23,550 copies/mL blood (Figure 3B). A minimum content of 1500 RPP30 copies/mL blood was set as a limit to ensure that sufficient DNA was obtained for TREC quantification.
NBS CDC Proficiency Testing TREC clinical assessment of 30 DBS specimens from the CDC’s Newborn Screening Quality Assurance Program (NSQAP) was 100% concordant with CDC results using our ddPCR method. Table 5 summarizes the reported frequency of clinical assessment by all participants in the CDC NSQAP over several challenges. Six specimens of 30 were expected to be SCID positive (TREC assessment classified as follow-up required, and RPP30 assessment classified as within reference range). However, two false negatives (specimens 216R2 and 416R1) and 16 false positives were reported by other NSQAP participants using either qPCR or the EnLite Neonatal TREC kit.
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
Reference Range Wilcoxon-Mann-Whitney analysis of 610 infant TREC levels showed no significant TREC mean differences between males or females. Thus, TREC and RPP30 thresholds were established regardless of sex. TREC content in infants ranged from 14 to 474 copies/mL blood, with a mean of 144 copies/mL blood and a median of 128 copies/mL blood, similar to previously published literature on TREC DBS content.28 RPP30 content ranged from 4000 to 62,500 copies/mL blood, with a mean of 17,690 copies/mL blood and a median of 16,269 copies/mL blood (Figure 3, C and D). The TREC 2.5% and 97.5% percentiles were calculated at 48 and 331 copies/mL blood, respectively. The RPP30 2.5% and 97.5% percentiles were calculated at 7134 and 35,857 copies/mL blood, respectively. The 610 infants’ ages ranged from a minimum of 1 to 133 days, and there was a moderate correlation between TREC and RPP30 levels, with a Pearson’s correlation coefficient of 0.352, as expected.
NBS Referral Specimens A total of 29 infants were referred for lymphocyte characterization at Mayo Clinic, ostensibly for abnormal NBS
761
Vidal-Folch et al Table 5
CDC SCID NBS Proficiency Results Mayo results (reference range)
CDC results RPP30 TREC assessment assessment
TREC copies/mL Specimen no. blood (20)
RPP30 copies/ mL blood (1500)
1, no follow-up required (screen negative); 2, follow-up required
415R1 415R2 415R3 415R4 415R5 116R1 116R2 116R3 116R4 116R5 216R1 216R2 216R3 216R4 216R5 316R1 316R2 316R3 316R4 316R5 416R1 416R2 416R3 416R4 416R5 117R1 117R2 117R3 117R4 117R5
18,988 26,750 329 6700 12,275 7800 12,213 34,625 186 18,200 17,950 12,413 16,375 12,675 391 19,550 4975 11,163 38 1638 8600 26,375 8125 29 12,900 16,150 23 52,500 9238 9138
1 1 2 2 1 2 1 1 2 1 1 2 1 1 2 1 2 1 2 2 2 1 1 2 1 1 2 1 1 2
66 169 <4 14 146 14 131 210 <4 108 179 <4 153 140 <4 165 15 63 <4 9 <4 188 29 <4 63 120 <4 224 70 <4
RPP30 TREC assessment assessment
1, no follow-up required (screen 1, within reference range; negative); 2, follow-up 2, outside reference range required
Participants with correct 1, Within NSQAP clinical reference range; assessment, % 2, outside (n incorrect/ reference range total)
1 1 2 1 1 1 1 1 2 1 1 1 1 1 2 1 1 1 2 1 1 1 1 2 1 1 2 1 1 1
1 1 2 1 1 1 1 1 2 1 1 1 1 1 2 1 1 1 2 1 1 1 1 2 1 1 2 1 1 1
1 1 2 2 1 2 1 1 2 1 1 2 1 1 2 1 2 1 2 2 2 1 1 2 1 1 2 1 1 2
100 98 (1/46) 100 100 100 100 100 100 100 100 100 98 (1/47) 100 100 100 100 100 100 100 100 98 (1/53) 96 (2/53) 98 (1/53) 100 98 (1/53) 94 (3/53) 100 98 (1/53) 87 (7/53) 100
NBS, newborn screening; NSQAP, Newborn Screening Quality Assurance Program; SCID, severe combined immunodeficiency; TREC, T-cell receptor excision circle.
results (Table 3). The reason for referral was unavailable in seven cases, whereas 22 were confirmed to have an abnormal NBS result for SCID. Our TREC assay confirmed five positive infants, whereas 17 infants were normal, with TREC levels ranging from 24 to 201 TREC copies/mL blood. In this cohort, 77% of NBS referrals sent to our department were false positives, as determined by follow-up lymphocyte characterization that was performed by experienced clinical immunologists using TREC copies/106 cells and CD3þ, CD4þ, and CD8þ T-cell criteria. Based on these NBS referral specimens, we calculated an 88.9% clinical sensitivity and a 100% clinical specificity. Unfortunately, the infant (DBS060) who presented with low T-cell counts was unable to be assigned a final diagnosis because of absent TREC values, which were likely obtained at another
institution. With this information, our calculated clinical sensitivity would likely be even higher. In turn, increasing our positive TREC threshold to 25 TREC copies/mL blood would reduce the specificity of our assay to 99.4% but have a sensitivity of 100%.
762
jmd.amjpathol.org
Discussion In this study, we demonstrated that ddPCR can be used to sensitively and specifically to measure TREC levels in a single DBS punch. The method’s limit of detection (4 TREC copies/mL blood) and imprecision (<20% CV) are sufficient to distinguish all typical SCID and leaky SCID as well as most other T-cell lymphopenia disorders, such as
-
The Journal of Molecular Diagnostics
Droplet Digital PCR for SCID NBS Newborn TREC Assay
TREC ≥ 20
TREC < 20
RPP30 <1500
RPP30 ≥ 1500
Suspected Positive Screen
Possible Unsatisfactory Specimen Repeat Assay
TREC ≥ 20
TREC < 20
RPP30 < 1500
RPP30 ≥ 1500
New Specimen Required
Positive
If preterm, follow NICU protocol
Negative
If full-term, order diagnostic testing
TBBS CD4RT TREC,B
Figure 4
The proposed SCID newborn screening algorithm with ddPCR. CD4RT, CD4 T-cell recent thymic emigrant; NICU, neonatal intensive care unit; TBBS, T- and B-cell quantification by flow cytometry; TREC, B, T-cell receptor excision circles analysis, blood.
idiopathic TCL, DiGeorge syndrome, CHARGE syndrome, and ataxia telangiectasia, tested to date. The addition of a simultaneously measured second analyte, RPP30, enhances confidence in a low TREC level (screen-positive result) without the necessity to punch multiple DBSs for repeat testing because of inefficient DNA extraction or PCR inhibitors. An additional benefit of this method is the ability to measure absolute TREC levels and remove reliance on the performance of a standard curve, which is integral to other published TREC qPCR methods. High-throughput newborn screening can be achieved with the proposed follow-up algorithm, illustrated in Figure 4. Unlike other published methods, our assay used an automated DNA extraction process that included an overnight DNA elution step. The addition of an extended elution was observed to increase our TREC yields, and to lower the limit of detection, than when using a shorter elution time. In contrast, TREC limit of detection by qPCR, and another DNA isolation method, has been reported to be as low as 19 TREC copies/mL blood.20 The elution step enabled our
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
technologists to complete specimen accessioning, DBS punching, and DNA extraction up to the point of elution, during 1 working day. The ddPCR section of the method can be completed within 3 hours the following day, and SCID-positive results were reportable in a manner that was consistent with the urgency of diagnostic confirmation and therapeutic intervention, while also fitting into our laboratory workflow. During the course of our study, we observed, as previously reported, that TREC levels diminish with increasing age.12 However, the TREC levels found in our reference range study were within TREC concentration ranges described for a handful of SCID newborn study experiences,7,29 even though our reference range population of full-term infants had a median age of 3 days at sample collection, and a range of 132 days, which is higher than a recommended testing age of 24 to 48 hours. In addition, as this was a retrospective study, our DBS samples were stored up to 1 month at ambient temperature. Nevertheless, our population study and related TREC threshold were sufficient to obtain 100% clinical specificity. There were two infant specimens tested during our population study with TREC levels at, or lower than, our screen-positive threshold of 20 TREC copies/mL blood (Figure 3A). Unfortunately, our population study was accomplished with deidentified, and anonymized, specimens, and we were unable to confirm whether these infants experienced concurrent T-cell lymphopenia at the time of specimen collection. We also could not collate these specific results with clinical specimens sent for characterization at our Cellular and Molecular Immunology Laboratory. Among prospectively tested specimens, we found only one TCL infant whose DBS TREC level of 25 copies/mL blood was higher than our initial screen-positive threshold. Blood T-cell characterization of this infant showed reduced levels of CD3þ, CD4þ, and CD8þ T cells within the same specimen. However, a cell-specific TREC level was not obtained, and we were unable to resolve whether this infant’s TREC copies/106 cells would be near normal range, as suggested by the DBS analysis. Thus, a final diagnosis for this infant remains unconfirmed. We included this infant’s test result in our calculation of clinical sensitivity at 88.9%, even though we were unable to confirm a final diagnosis of a condition that would be detectable by our testing method. Updating our threshold value to 25 TREC copies/mL blood would lead to the identification of this infant, and 100% sensitivity for the potential TCL conditions tested to date. However, there are advantages and disadvantages to increasing our screening threshold to include this unresolved case. A TREC concentration threshold at 25 TREC copies/mL blood would affect the specificity of the screening assay, and lead to more false-positive test results, with a calculated specificity of 99.4% after a single test. We aimed to keep our screening threshold at a level that would identify all confirmed cases of SCID or leaky SCID, while excluding any false-positive results (100%
763
Vidal-Folch et al specificity). Because of this balance, there is the potential that infants with immune defects might be missed, or that normal infants are preventably referred for additional laboratory testing. Past qPCR studies have reported specificities of >99%5,16,17,21 that were dependant on repetitive, multispecimen testing, and sensitivities that also rely on a balance, ranging from 87.5% to 100% contingent on the selected TREC threshold.5,21 In contrast, the CDC NSQAP showed that qPCR’s clinical sensitivities can range from 98% to 100% and clinical specificities from 87% to 100%, depending on a sample’s TREC level and the laboratory performing testing (Table 5). From a limited CDC NSQAP experience, ddPCR exceeded qPCR in screening performance, with 100% clinical sensitivity and 100% clinical specificity when assessing the same specimens. TREC levels alone are regrettably unable to identify all infants with primary immune defects.1,18,30 The inclusion of other analytical targets, such as B-cellespecific, k-deleting recombination excision circles or DBS adenosine, may aid to increase the sensitivity of NBS for extra primary immunodeficiencies, as reported elsewhere.6,21,31 Although the addition of molecular analyses to NBS may help to potentially detect a greater number of primary immunodeficiencies, current genomic methods are fraught with an increasing number of variants of unknown significance that will need functional characterization and disease correlation before widespread use of genomic methods for NBS.32,33 One shortfall of our study is that we did not sample a sufficient number of preterm infant DBS specimens to identify whether a specific preterm-infant threshold for TREC levels would be desirable for this method. Based on previous experience, we predict that using covariates and additional newborn screening marker results with an integrated report will aid SCID detection, as we described for other NBS conditions.34,35 Thus, a supplementary study with a larger prospective sample collection is ongoing and aims to determine whether the addition of covariates, such as sex, birth weight, gestational age, or other NBS condition test results, can add further improvements to SCID NBS detection. Another concern about our described method is the fact that the technology, ddPCR, is uncommon for Public Health Laboratories, and that associated assay costs might be inhibitory for population screening. In our limited proof-ofprinciple study, which did not have clinical volumes to appreciate any cost savings gained by an economy of scale, the material costs for our ddPCR method were calculated to be $0.46 for DNA extraction, and $3.80 for ddPCR, for a total of $4.30 per test. This figure is within per-test costs reportedly ranging from $4.22 to $6.00 for other SCID screening methods.13,36 The current hurdles for laboratories seeking to adopt ddPCR will be an initial investment into the clinical ddPCR testing platform and the scarceness of Federal Drug Administrationeapproved clinical testing kits for NBS. Regardless of the outcome of forthcoming studies, the ddPCR method we described herein is able to measure low
TREC levels in DBSs, is able to detect infants with SCID or leaky SCID and other TCLs, requires material expenditures that are comparable to current qPCR methods, and can be implemented for high-throughput screening.
764
jmd.amjpathol.org
Acknowledgments We thank Myra Wick, M.D., Ph.D., and the Mayo Clinic Umbilical Cord and Placental Biobank, for study sample collection and the CDC for the Newborn Screening Quality Assurance Program (NSQAP) experience.
References 1. Kwan A, Abraham RS, Currier R, Brower A, Andruszewski K, Abbott JK, et al: Newborn screening for severe combined immunodeficiency in 11 screening programs in the United States. JAMA 2014, 312:729e738 2. Dvorak CC, Cowan MJ, Logan BR, Notarangelo LD, Griffith LM, Puck JM, Shearer WT, O’Reilly RJ, Fleisher TA, Pai SY, Hanson IC, Pulsipher MA, Fuleihan R, Filipovich A, Goldman F, Kapoor N, Small T, Smith A, Chan KW, Cuvelier G, Heimall J, Knutsen A, Loechelt B, Moore T, Buckley RH: The natural history of children with severe combined immunodeficiency: baseline features of the first fifty patients of the primary immune deficiency treatment consortium prospective study 6901. J Clin Immunol 2013, 33:1156e1164 3. Pai SY, Logan BR, Griffith LM, Buckley RH, Parrott RE, Dvorak CC, et al: Transplantation outcomes for severe combined immunodeficiency, 2000-2009. N Engl J Med 2014, 371:434e446 4. Jilkina O, Thompson JR, Kwan L, Van Caeseele P, RockmanGreenberg C, Schroeder ML: Retrospective TREC testing of newborns with severe combined immunodeficiency and other primary immunodeficiency diseases. Mol Genet Metab Rep 2014, 1:324e333 5. Audrain M, Thomas C, Mirallie S, Bourgeois N, Sebille V, Rabetrano H, Durand-Zaleski I, Boisson R, Persyn M, Pierres C, Mahlaoui N, Fischer A: Evaluation of the T-cell receptor excision circle assay performances for severe combined immunodeficiency neonatal screening on Guthrie cards in a French single centre study. Clin Immunol 2014, 150:137e139 6. Barbaro M, Ohlsson A, Borte S, Jonsson S, Zetterström RH, King J, Winiarski J, von Döblen U, Hammarström L: Newborn screening for severe primary immunodeficiency diseases in Sweden: a 2-year pilot TREC and KREC screening study. J Clin Immunol 2017, 37:51e60 7. Adams SP, Rashid S, Premachandra T, Harvey K, Ifederu A, Wilson MC, Gaspar HB: Screening of neonatal UK dried blood spots using a duplex TREC screening assay. J Clin Immunol 2014, 34:323e330 8. Somech R, Lev A, Simon AJ, Korn D, Garty BZ, Amariglio N, Rechavi G, Almanshanu S, Zlotogora J, Etzioni A: Newborn screening for severe T and B cell immunodeficiency in Israel: a pilot study. Isr Med Assoc J 2013, 15:404e409 9. Morinishi Y, Imai K, Nakagawa N, Sato H, Horiuchi K, Ohtsuka Y, Kaneda Y, Taga T, Hisakawa H, Miyaji R, Endo M, Oh-Ishi T, Kamachi Y, Akahane KKobayashi C, Tsuchida M, Morio T, Sasahara Y, Kumaki S, Ishigaki K, Yoshida M, Urabe T, Kobayashi N, Okimoto Y, Reichenbach J, Hashii Y, Tsuji Y, Kogawa K, Yamaguchi S, Kanegane H, Miyawaki T, Yamada M, Ariga T, Nonoyama S: Identification of severe combined immunodeficiency by T-cell receptor excision circles quantification using neonatal guthrie cards. J Pediatr 2009, 155:829e833 10. Kong F, Chen CH, Cooper MD: Thymic function can be accurately monitored by the level of recent T cell emigrants in the circulation. Immunity 1998, 8:97e104 11. Douek DC, Vescio RA, Betts MR, Brenchley JM, Hill BJ, Zhang L, Berenson JR, Collins RH, Koup RA: Assessment of thymic output in
-
The Journal of Molecular Diagnostics
Droplet Digital PCR for SCID NBS
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
adults after haematopoietic stem-cell transplantation and prediction of T-cell reconstitution. Lancet 2000, 355:1875e1881 Douek DC, McFarland RD, Keiser PH, Gage EA, Massey JM, Haynes BF, Polis MA, Haase AT, Feinberg MB, Sullivan JL, Jamieson BD, Zack JA, Picker LJ, Koup RA: Changes in thymic function with age and during the treatment of HIV infection. Nature 1998, 396:690e695 Baker MW, Grossman WJ, Laessig RH, Hoffman GL, Brokopp CD, Kurtycz DF, Cogley MF, Litsheim TJ, Katcher ML, Routes JM: Development of a routine newborn screening protocol for severe combined immunodeficiency. J Allergy Clin Immunol 2009, 124:522e527 Routes JM, Grossman WJ, Verbsky J, Laessig RH, Hoffman GL, Brokopp CD, Baker MW: Statewide newborn screening for severe Tcell lymphopenia. JAMA 2009, 302:2465e2470 Picard C, Al-Herz W, Bousfiha A, Casanova JL, Chatila T, Conley ME, Cunningham-Rundels C, Etzioni A, Holland A, Klein C, Nonoyama S, Ochs HD, Oksenhendler E, Puck JM, Sullivan KE, Tang ML, Franco JL, Gaspar HB: Primary immunodeficiency diseases: an update on the classification from the International Union of Immunological Societies Expert Committee for Primary Immunodeficiency 2015. J Clin Immunol 2015, 35:696e726 Chien YH, Chiang SC, Chang KL, Yu HH, Lee WI, Tsai LP, Hsu MH, Hwu WL: Incidence of severe combined immunodeficiency through newborn screening in a Chinese population. J Formos Med Assoc 2015, 114:12e16 Verbsky J, Thakar M, Routes J: The Wisconsin approach to newborn screening for severe combined immunodeficiency. J Allergy Clin Immunol 2012, 129:622e627 Kwan A, Church JA, Cowan MJ, Agarwal R, Kapoor N, Kohn DB, Lewis DB, McGhee SA, Moore TB, Stiehm ER, Porteus M, Aznar CP, Currier R, Lorey F, Puck JM: Newborn screening for severe combined immunodeficiency and T-cell lymphopenia in California: results of the first 2 years. J Allergy Clin Immunol 2013, 132:140e150 Vogel BH, Bonagura V, Weinberg GA, Ballow M, Isabelle J, DiAntonio L, Parkar A, Young A, Cunningham-Rundels C, Fong CT, Celestin J, Lehman H, Rubinstein A, Siegel S, Weiner L, SaavedraMatiz C, Kay DM, Caggana M: Newborn screening for SCID in New York State: experience from the first two years. J Clin Immunol 2014, 34:289e303 Gerstel-Thompson JL, Wilkey JF, Baptiste JC, Navas JS, Pai SY, Pass KA, Eaton RB, Comeau AM: High-throughput multiplexed T-cell-receptor excision circle quantitative PCR assay with internal controls for detection of severe combined immunodeficiency in population-based newborn screening. Clin Chem 2010, 56:1466e1474 Borte S, von Döbeln U, Fasth A, Wang N, Janzi M, Winiarski J, Sack U, Pan-Hammarström Q, Borte M, Hammarström L: Neonatal screening for severe primary immunodeficiency diseases using highthroughput triplex real-time PCR. Blood 2012, 119:2552e2555 Taylor JL, Lee FK, Yazdanpanah GK, Staropoli JF, Liu M, Carulli JP, Sun C, Dobrowolski SF, Hannon WH, Vogt RF: Newborn blood spot screening test using multiplexed real-time PCR to simultaneously screen for spinal muscular atrophy and severe combined immunodeficiency. Clin Chem 2015, 61:412e419 Hudecova I: Digital PCR analysis of circulating nucleic acids. Clin Biochem 2015, 48:948e956
The Journal of Molecular Diagnostics
-
jmd.amjpathol.org
24. Clinical and Laboratory Standards Institute. Evaluation of detection capability for clinical laboratory measurement procedures. EP17-A2. Wayne, PA: CLSI, 2012. Available at https://clsi.org/standards/ products/method-evaluation/documents/ep17 25. Clinical and Laboratory Standards Institute. Evaluation of precision of quantitative measurement procedures: approved guideline, ed 3. EP05A3. Wayne, PA: CLSI, 2014. Available at https://clsi.org/ standards/products/method-evaluation/documents/ep05 26. Pinheiro LB, Coleman VA, Hindson CM, Herrmann J, Hindson BJ, Bhat S, Emslie KR: Evaluation of a droplet digital polymerase chain reaction format for DNA copy number quantification. Anal Chem 2012, 84:1003e1011 27. Makowski GS, Davis EL, Hopfer SM: The effect of storage on Guthrie cards: implications for deoxyribonucleic acid amplification. Ann Clin Lab Sci 1996, 26:458e469 28. van der Spek J, Groenwold RH, van der Burg M, van Montfrans JM: TREC based newborn screening for severe combined immunodeficiency disease: a systematic review. J Clin Immunol 2015, 35: 416e430 29. Baker MW, Laessig RH, Katcher ML, Routes JM, Grossman WJ, Verbsky J, Kurtycz DF, Brokopp CD: Implementing routine testing for severe combined immunodeficiency within Wisconsin’s newborn screening program. Public Health Rep 2010, 125 Suppl 2:88e95 30. Kuo CY, Chase J, Garcia Lloret M, Stiehm ER, Moore T, Aguilera MJ, Lopez Sriles J, Church JA: Newborn screening for severe combined immunodeficiency does not identify bare lymphocyte syndrome. J Allergy Clin Immunol 2013, 131:1693e1695 31. la Marca G, Canessa C, Giocaliere E, Romano F, Duse M, Malvagia S, Lippi F, Finghini S, Bianchi L, Della Bona ML, Valleriani C, Ombrone D, Moriondo M, Villanelli F, Speckmann C, Adams S, Gasper BH, Hershfield M, Santisteban I, Fairbanks L, Ragusa G, Resti M, de Martino M, Guerrini R, Azzari C: Tandem mass spectrometry, but not T-cell receptor excision circle analysis, identifies newborns with late-onset adenosine deaminase deficiency. J Allergy Clin Immunol 2013, 131:1604e1610 32. Narravula A, Garber KB, Askree SH, Hegde M, Hall PL: Variants of uncertain significance in newborn screening disorders: implications for large-scale genomic sequencing. Genet Med 2017, 19:77e82 33. Friedman JM, Cornel MC, Goldenberg AJ, Lister KJ, Sénécal K, Vears DF; Global Alliance for Genomics and Health Regulatory and Ethics Working Group Paediatric Task Team: Genomic newborn screening: public health policy considerations and recommendations. BMC Med Genomics 2017, 10:9 34. Marquardt G, Currier R, McHugh DM, Gavrilov D, Magera MJ, Matern D, et al: Enhanced interpretation of newborn screening results without analyte cutoff values. Genet Med 2012, 14:648e655 35. Hall PL, Marquardt G, McHugh DM, Currier RJ, Tang H, Stoway SD, Rinaldo P: Postanalytical tools improve performance of newborn screening by tandem mass spectrometry. Genet Med 2014, 16: 889e895 36. Chan K, Davis J, Pai SY, Bonilla FA, Puck JM, Apkon M: A Markov model to analyze cost-effectiveness of screening for severe combined immunodeficiency (SCID). Mol Genet Metab 2011, 104: 383e389
765