- Email: [email protected]
Newborn screening for severe combined immunodeficiency using a novel and simplified method to measure T-cell excision circles (TREC)
Clinical Immunology 175 (2017) 51–55
Contents lists available at ScienceDirect
Clinical Immunology journal homepage: www.elsevier.com/locate/yclim
Newborn screening for severe combined immunodeficiency using a novel and simplified method to measure T-cell excision circles (TREC) Laura Tagliaferri a,b, Joachim B. Kunz a, Margit Happich a, Susanna Esposito b, Thomas Bruckner d, Daniel Hübschmann a, Jürgen G. Okun c, Georg F. Hoffmann c, Ansgar Schulz e, Judit Kappe e, Carsten Speckmann f, Martina U. Muckenthaler a, Andreas E. Kulozik a,⁎ a
Department of Pediatric Oncology, Hematology and Immunology, Children's Hospital, University of Heidelberg, Heidelberg, Germany Pediatric Highly Intensive Care Unit, Department of Pathophysiology and Transplantation, Università degli Studi di Milano, Fondazione IRCCS Ca'randa, Ospedale Maggiore Policlinico, Milan, Italy c Division of Metabolic Diseases and Newborn Screening Center, Department of Paediatrics I, Children's Hospital, University of Heidelberg, Heidelberg, Germany d Institute of Medical Biometry and Informatics, University of Heidelberg, Heidelberg, Germany e Department of Pediatrics, University Medical Center Ulm, Germany f Center for Pediatrics and Adolescent Medicine and Center of Chronic Immunodeficiency, University Medical Center Freiburg, Germany b
a r t i c l e
i n f o
Article history: Received 3 April 2016 Received in revised form 25 October 2016 Accepted with revision 28 November 2016 Available online 2 December 2016 Keywords: Newborn screening Severe combined immunodeficiency T-cell receptor excision circles (TRECs)
a b s t r a c t The prognosis of children with severe combined immunodeficiency (SCID) depends on a presymptomatic diagnosis and early treatment before complications occur. We established and tested a simplified, practical and economic newborn screening method based on the quantification of T-cell receptor excision circles (TRECs) on dried blood spots (DBSs) through qPCR. Our method was validated by the analysis of 11 positive controls, which all showed an absence of TRECs, thus yielding a sensitivity of 100%. Further, we analyzed 6034 anonymized newborns of whom 6031 (99,95%) showed a normal TREC qPCR with a median of 600 estimated TREC copies/1.6 mm punch. The test showed a recall-rate of 0.05%. We present a highly sensitive, specific and time- and cost-effective method of TREC quantification, which is suitable for SCID newborn screening. In comparison to established methods, our test requires only 25% of the input material, doesn't require DNA purification and significantly reduces time and cost requirement. © 2016 Elsevier Inc. All rights reserved.
1. Introduction SCID is a group of N 20 disorders caused by different genetic defects [1]. The incidence has been calculated at approx. 1:58,000 live births [2]. All types of SCID have in common the lack of functional T-cells, leading to a combined cellular and humoral immunodeficiency. Without allogeneic hematopoietic stem cell transplantation (HSCT), children die in the first years of life because of severe infections. The only way to prevent death is early diagnosis followed by HSCT, before infections occur [3]. Indeed, studies have shown that the most important prognostic value for the primary outcome and the long-term survival rate is the clinical status of the patient (patients with active infection at the time Abbreviations: SCID, Severe Combined Immunodeficiency; DBSs, dried blood spots; NBS, newborn screening; TRECs, T-cell receptor excision circles; ADA-SCID, Adenosine deaminase deficiency; JAK3, Janus kinase 3, Tyrosine-protein kinase; ATG, antithymocyte globulin. ⁎ Corresponding author at: Department of Pediatric Oncology, Hematology, Immunology and Pulmonology, University Medical Center for Children and Adolescents, Angelika Lautenschläger Children's Hospital, Im Neuenheimer Feld 430, D-69120 Heidelberg, Germany. E-mail address: [email protected] (A.E. Kulozik).
http://dx.doi.org/10.1016/j.clim.2016.11.016 1521-6616/© 2016 Elsevier Inc. All rights reserved.
of transplantation have a survival rate of 50%, whereas those with no infection 82–90%) [4,5]. Because of the value of early pre-symptomatic diagnosis, SCID fulfills all the criteria for a disease to be targeted by newborn screening and was recommended being added to the panel of NBS illnesses in 2011 [6]. While a method to screen neonates for SCID in a high throughput format in dried blood spots has not been available until recently, the development of a practicable test by Chan and Puck in 2005 has been a breakthrough [7]. Recently, a number of studies in the US and in Europe proved the importance and the validity of a SCID newborn screening [2,3]. Efforts have been made in several research laboratories in order to establish an optimal screening test combining good sensitivity and specificity at costs affordable in a massive-scale format [8]. Nevertheless, no agreement on the method and the panel of diseases to be screened has been achieved so far. The most promising method is the quantification of T-cell Receptor Excision Circles (TRECs) on dried blood spots (DBS) by real-time quantitative PCR (qPCR) [9]. TRECs are small episomal pieces of DNA which are generated in the thymus during the VDJ-T-cell receptor gene rearrangement and are therefore good markers for naïve T-cells [10].
52
L. Tagliaferri et al. / Clinical Immunology 175 (2017) 51–55
Several efforts have been made to implement methods for the early diagnosis of SCID [5,11–17]. We established and tested a robust method to detect SCID in newborns through the quantification of TRECs by qPCR. Our method showed a very high sensitivity and specificity and compared favorably to existing methods in terms of time and costs, which plays an important role in the perspective of a population screening on large numbers.
amplification. Calibration curves for TREC quantification were obtained by using a 10-fold serially diluted TREC plasmid. β-actin was used as a semi quantitative internal control to assess successful DNA extraction in patients whose TREC copy number did not achieve the fixed cutoff (see Results section).
2.5. Statistical analysis 2. Material and methods ROC curves were generated in order to fix an appropriate cutoff for TREC-absolute value and TREC-Cq value.
2.1. Screened samples DBS specimens were obtained from Guthrie cards collected by the Newborn Screening Laboratory of Heidelberg University in the period between October and December 2012. After completing the other routine screening tests, DBSs were de-identified and punched for the TREC assay. After having obtained a written consent from the parents, anonymized DBS specimens of newborns and children with different types of T-cell deficiency were used as positive controls (Supplemental Table 1). 2.2. Ethical approval
3. Results We established and tested a new high performance assay for SCID screening based on a qPCR assay for the quantification of TRECs. A very fast DNA extraction phase requiring only few minutes was followed by the qPCR analysis. Calibration curves for TREC quantification were obtained by serially diluting a plasmid coding for the TREC sequence (Fig. 1a–b). β-actin served as semi quantitative internal control. The method was highly reproducible and quantitative over a range from 10 to 100,000 copies per punch and showed some advantages in comparison to other existing methods [8].
This pilot study was approved by the ethics committee of the University of Heidelberg. 3.1. Establishing TREC cutoff value in a newborn screening population 2.3. DNA extraction A new protocol for DNA extraction was established for the purpose of this study. 1.6 mm punches were incubated in sealed 96-well plates with 13 μl lysis reagent (DNA Extract All Reagents Kit, Cat. Number 4402599, Applied Biosystems) for 3 min at 95 °C. Lysis was stopped by adding 13 μl of stabilizing reagent (DNA Extract All Reagents Kit, Cat. Number 4402599, Applied Biosystems). 2.4. qPCR Real-time quantitative PCR for TREC quantification was performed in a total final volume of 20 μl containing 10 μl Mastermix (SensiFAST Probe Hi-ROX Kit, Bioline), 2 μl TREC forward-primer 10 μM, 2 μl TREC reverse-primer 10 μM, 0.3 μl FAM-TAMRA-labeled TaqMan TREC probes 1:10 (Life Technologies), 0.8 μl BSA 1% and 2 μl DNA extract. Real-time quantitative PCR for β-actin quantification was also performed in a total final volume of 20 μl containing 10 μl Mastermix (SensiFAST Probe Hi-ROX Kit, Bioline), 0.7 μl 1:3 VIC-labeled Beta Actin TaqMan Gene Expression Assay (Nr Hs03023880, Order Nr 4,448,484, Life Technologies) containing β-actin primers and probes and 2 μl DNA extract. The DNA sequences of primers and probes are listed in Table 1. The qPCR reactions were carried out on a StepOnePlus real-time PCR System (Applied Biosystems) in 96-well plates and underwent 1 cycle of 2 min at 50 °C, 1 cycle of 10 min at 95 °C and 45 cycles of 30 s at 95 °C and 30s at 60 °C. A fixed quantification cycle (Cq) was set for data collection and amplification analysis during the exponential phase of the PCR
Table 1 Sequences of TREC primers and probes [11]. Sequences of β-actin primers and probe according to the Gene expression Assay Nr Hs03023880, Order Nr 4448484, Life Technologies. Name
Sequence
TREC forward primer TREC reverse primer TREC probe
5′-CACATCCCTTTCAACCATGCT 5′-GCCAGCTGCAGGGTTTAGG 6-FAM-ACACCTCTGGTTTTTGTAAAGGTGCCCACT-3′-TAMRA
We collected Guthrie cards of 6046 consecutive unselected samples after the standard NBS program had been completed. Because of the low frequency of SCID (b 1:60,000) these samples were expected not to include SCID patients. Twelve samples had to be excluded as they contained too little material. Fig. 1 shows an example of the results obtained from the qPCR TREC assay. We thus analyzed 6034 samples, in which we found a distribution of TREC copy number from 0.1–5109 copies/punch with an average of 814 copies/punch and a median of 600 copies/punch; the 1st percentile laid by 243 copies/punch and the 99th percentile by 3067 copies/punch. Through statistical analysis based on ROC-curves (Fig. 2) we analyzed the distribution of the results and calculated an appropriate cutoff for the interpretation of the results. Within the positive controls we observed a single TREC-value of 94 copies/ punch (possibly due to sample contamination), so that the cutoff was fixed at 95 TREC copies per 1.6-mm DBS punch. Based on the cutoff-value, we established an algorithm for the sample analysis (Fig. 3). The quantification of simply TREC copy number was used as a first tier. For samples that did not pass the first tier, a second tier included actin as an internal control. Only samples that failed the second tier would be recalled in a nonanonymized setting. Fig. 4 illustrates the distribution of TREC values in our study population in comparison to the positive controls. 5964 (98.8%) of 6034 tested infants had an initial TREC value above the cutoff. 70 infants had an initial TREC value below the cutoff, which led us to perform a second TREC analysis including an internal control with beta actin to distinguish between real TREC-negative results and “DNA amplification failure” (retest-rate 1.2%). The majority of these samples (59 of 70; 85%) passed the second tier with a TREC copy number above the cutoff and clearly detectable actin (Cq Value b 30). For 11 newborns (15%) the analysis of a second punch from the same blood sample was needed, because of “DNA amplification failure”, which identified TRECs above the cutoff in 8. Therefore, in 3 of 6034 samples no TRECs above the cutoff could be detected. These newborns must therefore be suspected to have T-cell deficiency and would have been recalled for further investigations in a non-anonymized setting (recall-rate 0.05%; Fig. 3).
L. Tagliaferri et al. / Clinical Immunology 175 (2017) 51–55
53
Fig. 1. Example of the results obtained from the qPCR analysis of Guthrie card samples in a 96-well plate including SCID patients, NBS samples and a TREC plasmid. The amplification plot (1a) shows the exponential amplification of DNA during the PCR reaction in the different groups and the standard curve for TREC copy number (1b) an example of the distribution of the patient populations in relation to the TREC plasmid. ΔRn = delta normalised reporter; Cq = quantification cycle.
3.2. Positive controls The second analysis was performed on 11 positive controls to calculate the sensitivity of the test in recognizing patients with low or absent T-cells. These were newborns and children affected by different forms of primary immunodeficiency (8 SCID, 1 Di George Syndrome and 1 Nijmegen-Breakage-Syndrome, Supplemental Table 1) and 1 T-cell depleted patient after serotherapy with ATG (antithymocyte globulin, Supplemental Table 1). In order to evaluate the reproducibility of the method we repeatedly retested the samples in different runs, for a total of 122 analyses. In all of these analyses no or only traces of TRECs were detectable. The results distribution of the positive controls in comparison to the NBS population is showed in the Box-Whisker-Plot in Fig. 4. All of the positive controls showed TREC values below the cutoff. The TREC median value was below the detection limit for all the samples and only few samples were found to be near to the cutoff. The 95th
Fig. 2. ROC curve obtained after statistical analysis of the results used to fix an appropriate cutoff of TREC copy number for the data analysis.
percentile for the positive controls laid by 6 TREC copies/punch, quite far from the cutoff of 95 copies/punch. Finally, we observed a normal TREC copy number in the Guthrie card of a newborn affected by lateonset ADA-SCID, as expected because these patients do not show any immunodeficiency at birth (Supplemental Table 1). Amplification failure because of lack of DNA could be excluded by normal β-actin amplification in all of the samples. 4. Discussion 4.1. Test quality All of the positive controls were found to have TREC numbers below the cutoff (estimated sensitivity 1.0). The specificity was estimated based on the proportion of unselected newborns with a TREC number above the cutoff to be 0.990 ± 0.003, CI 99%. The re-test rate of 1.2% and the recall rate of 0.05% lie within the range of other existing tests for SCID and below the range generally accepted in newborn screening [11,12]. As our study was retrospective and anonymous, we cannot comment on the clinical status of three newborns with no detectable TRECs. These newborns may have included patients with SCID, but based on the low frequency of SCID (b1:60,000), there may have been other reasons for low T-cell counts such as prematurity, sepsis, hearth surgery, chylothorax, maternal drug use in pregnancy and various clinical conditions including DiGeorge, CHARGE and Jacobsen syndromes [2,12,18,19]. Alternatively, there may have been unrecognized technical reasons for a failure of TREC-amplification. As the number of false positive results is not known, the positive predictive value can only be estimated. Assuming that none of the newborns in the NBS population was affected by a T-cell defect, we estimate a false positive rate of 0.05%. The positive predictive value largely
54
L. Tagliaferri et al. / Clinical Immunology 175 (2017) 51–55
Fig. 3. Flow chart for the process of sample analysis in the SCID screening.
depends on the frequency of SCID in the population screened. Based on an estimated frequency of 1:60,000, we calculate a positive predictive value of 0.03. Interestingly, the method reported here also identified a patient with Nijmegen-Breakage Syndrome, who was included in the panel of
positive controls. This result supports recent data [20], showing that the quantification of TRECs can also be used for the early diagnosis of relevant non-SCID lymphopenic disorders. 4.2. Time and costs The method proved to be technically simple and highly time- and cost-effective. Thanks to the simplified DNA extraction phase requiring only few minutes, the test showed excellent time-efficiency and resulted to be faster than any other published method [6,11,12]. Considering that the analysis of 96 samples in a multi-well plate requires approx. 1.5 h, we estimated that an experienced technician can process up to 425 samples with one real-time PCR machine per one working day, adding up to N2000 samples per week and N 100,000 samples per year. Based on current list prices for equipment, reagents, the material costs for DNA extraction and qPCR for TREC quantification amount to approx. € 1.14 per sample. For a qPCR for β-actin quantification € 1.83 are required. Taking into account that approx. 1.2% of samples require both the quantification of TREC and of β-actin, the average material costs for each sample amounts to € 1.18. The cost for personnel was calculated to be around € 0.80 per sample and included a full time technician, who can process 100,000 samples annually and 2 h per week of a physician, who is responsible for medical validation and tracking. This time requirement is extrapolated from the experience with other rare diseases that are screened for by our large metabolic screening center. The total cost is thus estimated at about € 2.0 per sample, which is considerably lower than the cost of currently used assays that are calculated at approx. $ 4 per sample [8,11,21,22]. Finally, using 1.6 mm punches instead of the commonly used 3.2 mm punches [11,12] reduces the required material, which is an important aspect in newborn screening particularly when considering the continuously increasing number of screened diseases. 5. Conclusions
Fig. 4. Box-Whisker-Plot showing the distribution of TREC values in the NBS population and in the positive controls (TREC copy nr./punch). The boxes represent the values included between the 1st and the 3rd quartile and the whiskers the minimum and maximum of all the data. The band inside the box shows the median value. The dashed line marks the cutoff-value, which was fixed by 95 TREC copies/punch. Finally, the spots indicate the single values below the 1st percentile for the NBS Population and above the 3rd percentile for the positive controls.
In sum, this anonymized pilot phase presents and evaluates a new reliable, rapid and cost-effective method of newborn screening for SCID. We currently plan a prospective non-anonymous screening phase of all newborns that is meant to prepare for a general newborn screening for SCID in Germany.
L. Tagliaferri et al. / Clinical Immunology 175 (2017) 51–55
Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.clim.2016.11.016. Acknowledgements We would like to thank Alessandra Sottini and Luisa Imberti for providing us a TREC-plasmid, as well as all the colleagues of the Pediatric Centers of Ulm and Freiburg who contributed with sending positive control-samples. We are finally grateful to the Dietmar Hopp Stiftung, St. Leon-Rot, Germany, for the generous funding of this project. References [1] C. Picard, W. Al-Herz, A. Bousfiha, J.L. Casanova, T. Chatila, M.E. Conley, C. Cunningham-Rundles, A. Etzioni, S.M. Holland, C. Klein, S. Nonoyama, H.D. Ochs, E. Oksenhendler, J.M. Puck, K.E. Sullivan, M.L.K. Tang, J.L. Franco, H.B. Gaspar, Primary immunodeficiency diseases: an update on the classification from the International Union of Immunological Societies Expert Committee for primary immunodeficiency 2015, J. Clin. Immunol. 35 (2015) 696–726. [2] A. Kwan, R.S. Abraham, R. Currier, A. Brower, K. Andruszewski, J.K. Abbott, M. Baker, M. Ballow, L.E. Bartoshesky, V.R. Bonagura, F.A. Bonilla, C. Brokopp, E. Brooks, et al., Newborn screening for severe combined immunodeficiency in 11 screening programs in the United States, JAMA 312 (2014) 729–738. [3] L. Brown, J. Xu-Bayford, Z. Allwood, M. Slatter, A. Cant, E.G. Davies, P. Veys, A.R. Gennery, H.B. Gaspar, Neonatal diagnosis of severe combined immunodeficiency leads to significantly improved survival outcome: the case for newborn screening, Blood 117 (2011) 3243–3246. [4] R.H. Buckley, Molecular defects in human severe combined immunodeficiency and approaches to immune reconstitution, Annu. Rev. Immunol. 22 (2004) 625–655. [5] S.Y. Pai, B.R. Logan, L.M. Griffith, H.R. Buckley, R.E. Parrott, C.C. Dvorak, N. Kapoor, I.C. Hanson, A.H. Filipovich, S. Jyonouchi, K.E. Sullivan, T.N. Small, L. Burroughs, S. SkodaSmith, et al., Transplantation outcomes for severe combined immunodeficiency 2000–2009, N. Engl. J. Med. 371 (2014) 434–446. [6] R.R. Howell, Report on newborn screening for severe combined immunodeficiency, secretary's advisory committee on heritable disorders in newborns and children, http://www.hrsa.gov/advisorycommittees/mchbadvisory/heritabledisorders/ recommendations/correspondence/severeimmunodeficiency.pdf2011. [7] K. Chan, J.M. Puck, Development of population-based newborn screening for severe combined immunodeficiency, J. Allergy Clin. Immunol. 115 (2005) 391–398. [8] J. van der Spek, R. Groenwold, M. van der Burg, J.M. van Montfrans, TREC based newborn screening for severe combined immunodeficiency disease: a systematic review, J. Clin. Immunol. 35 (2015) 416–430. [9] J.M. Puck, Laboratory technology for population-based screening for severe combined immunodeficiency in neonates: the winner is T-cell receptor excision circles, J. Allergy Clin. Immunol. 129 (2012) 607–616.
55
[10] M.C. van Zelm, M. van der Burg, A.W. Langerak, J.J.M. van Dongen, PID comes full circle: applications of V(D)J recombination excision circles in research, diagnostics and newborn screening of primary immunodeficiency disorders, Front. Immunol. 2 (2011) 12. [11] M.W. Baker, W.J. Grossman, R.H. Laessig, G.L. Hoffman, C.D. Brokopp, D.F. Kurtycz, M.F. Cogley, T.J. Litsheim, M.L. Katcher, J.M. Routes, Development of a routine newborn screening protocol for severe combined immunodeficiency, J. Allergy Clin. Immunol. 124 (2009) 522–527. [12] A. Kwan, J.A. Church, M.J. Cowan, R. Agarwal, N. Kapoor, D.B. Kohn, D.B. Lewis, S.A. McGhee, T.B. Moore, E.R. Stiehm, M. Porteus, C.P. Aznar, R. Currier, F. Lorey, J.M. Puck, Newborn screening for severe combined immunodeficiency and T-cell lymphopenia in California: results of the first 2 years, J. Allergy Clin. Immunol. 132 (2013) 140–150. [13] J.W. Verbsky, M.W. Baker, W.J. Grossman, M. Hintermeyer, S. Datu, B. Bonacci, S. Reddy, D. Margolis, J. Casper, M. Gries, K. DeSantes, G.L. Hoffman, C.D. Brokopp, C.M. Seroogy, J.M. Routes, Newborn screening for severe combined immunodeficiency; the Wisconsin experience (2008–2011), J. Clin. Immunol. 32 (2012) 82–88. [14] M.W. Baker, R.H. Laessig, M.L. Katcher, J.M. Routes, W.J. Grossman, J.W. Verbsky, D.F. Kurtycz, C.D. Brokopp, Implementing routine testing for severe combined immunodeficiency within Wisconsin's newborn screening program, Public Health Rep. S2 (125) (2012) 88–95. [15] N.M. Chase, J.W. Verbsky, J.M. Routes, Newborn screening for SCID: 3 years of experience, Ann. N. Y. Acad. Sci. 1238 (2011) 99–105. [16] N. Nakagawa, K. Imai, H. Kanegane, H. Sato, M. Yamada, K. Kondoh, S. Okada, M. Kobayashi, K. Agematsu, H. Takada, N. Mitsuiki, K. Oshima, O. Ohara, Quantification of κ-deleting recombination excision circles in Guthrie cards for the identification of early B-cell maturation defects, J. Allergy Clin. Immunol. 128 (2011) 223–225. [17] S. Borte, Ü. von Döbeln, A. Fasth, N. Wang, M. Janzi, J. Winiarski, U. Sack, Q. PanHammarström, M. Borte, L. Hammarström, Neonatal screening for severe primary immunodeficiency diseases using high-throughput triplex real-time PCR, Blood 119 (2012) 2552–2555. [18] R. Correa-Rocha, A. Pérez, R. Lorente, S. Ferrando-Martínez, M. Leal, D. Gurbindo, M.Á. Muños-Fernández, Preterm neonates show marked leukopenia and lymphopenia that are associated with increased regulatory T-cell values and diminished IL-7, Pediatr. Res. 71 (2012) 590–597. [19] S. Chabra, C. Cottrill, M.K. Rayens, R. Cross, D. Lipke, M. Bruce, Lymphocyte subsets in cord blood of preterm infants: effect of antenatal steroids, Biol. Neonate 74 (1998) 200–207. [20] J.P. Patel, J.M. Puck, R. Srinivasan, C. Brown, U. Sunderam, K. Kundu, S.E. Brenner, R.A. Gatti, J.A. Church, Nijmegen breakage syndrome detected by newborn screening for T cell receptor excision circles (TRECs), J. Clin. Immunol. 35 (2015) 227–233. [21] K. Chan, J. Davis, S.Y. Pai, F.A. Bonilla, J.M. Puck, M. Apkon, A Markov model to analyze cost-effectiveness of screening for severe combined immunodeficiency (SCID), Mol. Genet. Metab. 104 (2011) 383–389. [22] C.A. Saavedra-Matiz, J.T. Isabelle, C.K. Biski, S.J. Duva, M.L. Sweeney, A.L. Parker, A.J. Young, L.L. DiAntonio, L.M. Krein, M.J. Nichols, M. Caggana, Cost-effective and scalable DNA extraction method from dried blood spots, Clin. Chem. 59 (2013) 1045–1051.
Contents lists available at ScienceDirect
Clinical Immunology journal homepage: www.elsevier.com/locate/yclim
Newborn screening for severe combined immunodeficiency using a novel and simplified method to measure T-cell excision circles (TREC) Laura Tagliaferri a,b, Joachim B. Kunz a, Margit Happich a, Susanna Esposito b, Thomas Bruckner d, Daniel Hübschmann a, Jürgen G. Okun c, Georg F. Hoffmann c, Ansgar Schulz e, Judit Kappe e, Carsten Speckmann f, Martina U. Muckenthaler a, Andreas E. Kulozik a,⁎ a
Department of Pediatric Oncology, Hematology and Immunology, Children's Hospital, University of Heidelberg, Heidelberg, Germany Pediatric Highly Intensive Care Unit, Department of Pathophysiology and Transplantation, Università degli Studi di Milano, Fondazione IRCCS Ca'randa, Ospedale Maggiore Policlinico, Milan, Italy c Division of Metabolic Diseases and Newborn Screening Center, Department of Paediatrics I, Children's Hospital, University of Heidelberg, Heidelberg, Germany d Institute of Medical Biometry and Informatics, University of Heidelberg, Heidelberg, Germany e Department of Pediatrics, University Medical Center Ulm, Germany f Center for Pediatrics and Adolescent Medicine and Center of Chronic Immunodeficiency, University Medical Center Freiburg, Germany b
a r t i c l e
i n f o
Article history: Received 3 April 2016 Received in revised form 25 October 2016 Accepted with revision 28 November 2016 Available online 2 December 2016 Keywords: Newborn screening Severe combined immunodeficiency T-cell receptor excision circles (TRECs)
a b s t r a c t The prognosis of children with severe combined immunodeficiency (SCID) depends on a presymptomatic diagnosis and early treatment before complications occur. We established and tested a simplified, practical and economic newborn screening method based on the quantification of T-cell receptor excision circles (TRECs) on dried blood spots (DBSs) through qPCR. Our method was validated by the analysis of 11 positive controls, which all showed an absence of TRECs, thus yielding a sensitivity of 100%. Further, we analyzed 6034 anonymized newborns of whom 6031 (99,95%) showed a normal TREC qPCR with a median of 600 estimated TREC copies/1.6 mm punch. The test showed a recall-rate of 0.05%. We present a highly sensitive, specific and time- and cost-effective method of TREC quantification, which is suitable for SCID newborn screening. In comparison to established methods, our test requires only 25% of the input material, doesn't require DNA purification and significantly reduces time and cost requirement. © 2016 Elsevier Inc. All rights reserved.
1. Introduction SCID is a group of N 20 disorders caused by different genetic defects [1]. The incidence has been calculated at approx. 1:58,000 live births [2]. All types of SCID have in common the lack of functional T-cells, leading to a combined cellular and humoral immunodeficiency. Without allogeneic hematopoietic stem cell transplantation (HSCT), children die in the first years of life because of severe infections. The only way to prevent death is early diagnosis followed by HSCT, before infections occur [3]. Indeed, studies have shown that the most important prognostic value for the primary outcome and the long-term survival rate is the clinical status of the patient (patients with active infection at the time Abbreviations: SCID, Severe Combined Immunodeficiency; DBSs, dried blood spots; NBS, newborn screening; TRECs, T-cell receptor excision circles; ADA-SCID, Adenosine deaminase deficiency; JAK3, Janus kinase 3, Tyrosine-protein kinase; ATG, antithymocyte globulin. ⁎ Corresponding author at: Department of Pediatric Oncology, Hematology, Immunology and Pulmonology, University Medical Center for Children and Adolescents, Angelika Lautenschläger Children's Hospital, Im Neuenheimer Feld 430, D-69120 Heidelberg, Germany. E-mail address: [email protected] (A.E. Kulozik).
http://dx.doi.org/10.1016/j.clim.2016.11.016 1521-6616/© 2016 Elsevier Inc. All rights reserved.
of transplantation have a survival rate of 50%, whereas those with no infection 82–90%) [4,5]. Because of the value of early pre-symptomatic diagnosis, SCID fulfills all the criteria for a disease to be targeted by newborn screening and was recommended being added to the panel of NBS illnesses in 2011 [6]. While a method to screen neonates for SCID in a high throughput format in dried blood spots has not been available until recently, the development of a practicable test by Chan and Puck in 2005 has been a breakthrough [7]. Recently, a number of studies in the US and in Europe proved the importance and the validity of a SCID newborn screening [2,3]. Efforts have been made in several research laboratories in order to establish an optimal screening test combining good sensitivity and specificity at costs affordable in a massive-scale format [8]. Nevertheless, no agreement on the method and the panel of diseases to be screened has been achieved so far. The most promising method is the quantification of T-cell Receptor Excision Circles (TRECs) on dried blood spots (DBS) by real-time quantitative PCR (qPCR) [9]. TRECs are small episomal pieces of DNA which are generated in the thymus during the VDJ-T-cell receptor gene rearrangement and are therefore good markers for naïve T-cells [10].
52
L. Tagliaferri et al. / Clinical Immunology 175 (2017) 51–55
Several efforts have been made to implement methods for the early diagnosis of SCID [5,11–17]. We established and tested a robust method to detect SCID in newborns through the quantification of TRECs by qPCR. Our method showed a very high sensitivity and specificity and compared favorably to existing methods in terms of time and costs, which plays an important role in the perspective of a population screening on large numbers.
amplification. Calibration curves for TREC quantification were obtained by using a 10-fold serially diluted TREC plasmid. β-actin was used as a semi quantitative internal control to assess successful DNA extraction in patients whose TREC copy number did not achieve the fixed cutoff (see Results section).
2.5. Statistical analysis 2. Material and methods ROC curves were generated in order to fix an appropriate cutoff for TREC-absolute value and TREC-Cq value.
2.1. Screened samples DBS specimens were obtained from Guthrie cards collected by the Newborn Screening Laboratory of Heidelberg University in the period between October and December 2012. After completing the other routine screening tests, DBSs were de-identified and punched for the TREC assay. After having obtained a written consent from the parents, anonymized DBS specimens of newborns and children with different types of T-cell deficiency were used as positive controls (Supplemental Table 1). 2.2. Ethical approval
3. Results We established and tested a new high performance assay for SCID screening based on a qPCR assay for the quantification of TRECs. A very fast DNA extraction phase requiring only few minutes was followed by the qPCR analysis. Calibration curves for TREC quantification were obtained by serially diluting a plasmid coding for the TREC sequence (Fig. 1a–b). β-actin served as semi quantitative internal control. The method was highly reproducible and quantitative over a range from 10 to 100,000 copies per punch and showed some advantages in comparison to other existing methods [8].
This pilot study was approved by the ethics committee of the University of Heidelberg. 3.1. Establishing TREC cutoff value in a newborn screening population 2.3. DNA extraction A new protocol for DNA extraction was established for the purpose of this study. 1.6 mm punches were incubated in sealed 96-well plates with 13 μl lysis reagent (DNA Extract All Reagents Kit, Cat. Number 4402599, Applied Biosystems) for 3 min at 95 °C. Lysis was stopped by adding 13 μl of stabilizing reagent (DNA Extract All Reagents Kit, Cat. Number 4402599, Applied Biosystems). 2.4. qPCR Real-time quantitative PCR for TREC quantification was performed in a total final volume of 20 μl containing 10 μl Mastermix (SensiFAST Probe Hi-ROX Kit, Bioline), 2 μl TREC forward-primer 10 μM, 2 μl TREC reverse-primer 10 μM, 0.3 μl FAM-TAMRA-labeled TaqMan TREC probes 1:10 (Life Technologies), 0.8 μl BSA 1% and 2 μl DNA extract. Real-time quantitative PCR for β-actin quantification was also performed in a total final volume of 20 μl containing 10 μl Mastermix (SensiFAST Probe Hi-ROX Kit, Bioline), 0.7 μl 1:3 VIC-labeled Beta Actin TaqMan Gene Expression Assay (Nr Hs03023880, Order Nr 4,448,484, Life Technologies) containing β-actin primers and probes and 2 μl DNA extract. The DNA sequences of primers and probes are listed in Table 1. The qPCR reactions were carried out on a StepOnePlus real-time PCR System (Applied Biosystems) in 96-well plates and underwent 1 cycle of 2 min at 50 °C, 1 cycle of 10 min at 95 °C and 45 cycles of 30 s at 95 °C and 30s at 60 °C. A fixed quantification cycle (Cq) was set for data collection and amplification analysis during the exponential phase of the PCR
Table 1 Sequences of TREC primers and probes [11]. Sequences of β-actin primers and probe according to the Gene expression Assay Nr Hs03023880, Order Nr 4448484, Life Technologies. Name
Sequence
TREC forward primer TREC reverse primer TREC probe
5′-CACATCCCTTTCAACCATGCT 5′-GCCAGCTGCAGGGTTTAGG 6-FAM-ACACCTCTGGTTTTTGTAAAGGTGCCCACT-3′-TAMRA
We collected Guthrie cards of 6046 consecutive unselected samples after the standard NBS program had been completed. Because of the low frequency of SCID (b 1:60,000) these samples were expected not to include SCID patients. Twelve samples had to be excluded as they contained too little material. Fig. 1 shows an example of the results obtained from the qPCR TREC assay. We thus analyzed 6034 samples, in which we found a distribution of TREC copy number from 0.1–5109 copies/punch with an average of 814 copies/punch and a median of 600 copies/punch; the 1st percentile laid by 243 copies/punch and the 99th percentile by 3067 copies/punch. Through statistical analysis based on ROC-curves (Fig. 2) we analyzed the distribution of the results and calculated an appropriate cutoff for the interpretation of the results. Within the positive controls we observed a single TREC-value of 94 copies/ punch (possibly due to sample contamination), so that the cutoff was fixed at 95 TREC copies per 1.6-mm DBS punch. Based on the cutoff-value, we established an algorithm for the sample analysis (Fig. 3). The quantification of simply TREC copy number was used as a first tier. For samples that did not pass the first tier, a second tier included actin as an internal control. Only samples that failed the second tier would be recalled in a nonanonymized setting. Fig. 4 illustrates the distribution of TREC values in our study population in comparison to the positive controls. 5964 (98.8%) of 6034 tested infants had an initial TREC value above the cutoff. 70 infants had an initial TREC value below the cutoff, which led us to perform a second TREC analysis including an internal control with beta actin to distinguish between real TREC-negative results and “DNA amplification failure” (retest-rate 1.2%). The majority of these samples (59 of 70; 85%) passed the second tier with a TREC copy number above the cutoff and clearly detectable actin (Cq Value b 30). For 11 newborns (15%) the analysis of a second punch from the same blood sample was needed, because of “DNA amplification failure”, which identified TRECs above the cutoff in 8. Therefore, in 3 of 6034 samples no TRECs above the cutoff could be detected. These newborns must therefore be suspected to have T-cell deficiency and would have been recalled for further investigations in a non-anonymized setting (recall-rate 0.05%; Fig. 3).
L. Tagliaferri et al. / Clinical Immunology 175 (2017) 51–55
53
Fig. 1. Example of the results obtained from the qPCR analysis of Guthrie card samples in a 96-well plate including SCID patients, NBS samples and a TREC plasmid. The amplification plot (1a) shows the exponential amplification of DNA during the PCR reaction in the different groups and the standard curve for TREC copy number (1b) an example of the distribution of the patient populations in relation to the TREC plasmid. ΔRn = delta normalised reporter; Cq = quantification cycle.
3.2. Positive controls The second analysis was performed on 11 positive controls to calculate the sensitivity of the test in recognizing patients with low or absent T-cells. These were newborns and children affected by different forms of primary immunodeficiency (8 SCID, 1 Di George Syndrome and 1 Nijmegen-Breakage-Syndrome, Supplemental Table 1) and 1 T-cell depleted patient after serotherapy with ATG (antithymocyte globulin, Supplemental Table 1). In order to evaluate the reproducibility of the method we repeatedly retested the samples in different runs, for a total of 122 analyses. In all of these analyses no or only traces of TRECs were detectable. The results distribution of the positive controls in comparison to the NBS population is showed in the Box-Whisker-Plot in Fig. 4. All of the positive controls showed TREC values below the cutoff. The TREC median value was below the detection limit for all the samples and only few samples were found to be near to the cutoff. The 95th
Fig. 2. ROC curve obtained after statistical analysis of the results used to fix an appropriate cutoff of TREC copy number for the data analysis.
percentile for the positive controls laid by 6 TREC copies/punch, quite far from the cutoff of 95 copies/punch. Finally, we observed a normal TREC copy number in the Guthrie card of a newborn affected by lateonset ADA-SCID, as expected because these patients do not show any immunodeficiency at birth (Supplemental Table 1). Amplification failure because of lack of DNA could be excluded by normal β-actin amplification in all of the samples. 4. Discussion 4.1. Test quality All of the positive controls were found to have TREC numbers below the cutoff (estimated sensitivity 1.0). The specificity was estimated based on the proportion of unselected newborns with a TREC number above the cutoff to be 0.990 ± 0.003, CI 99%. The re-test rate of 1.2% and the recall rate of 0.05% lie within the range of other existing tests for SCID and below the range generally accepted in newborn screening [11,12]. As our study was retrospective and anonymous, we cannot comment on the clinical status of three newborns with no detectable TRECs. These newborns may have included patients with SCID, but based on the low frequency of SCID (b1:60,000), there may have been other reasons for low T-cell counts such as prematurity, sepsis, hearth surgery, chylothorax, maternal drug use in pregnancy and various clinical conditions including DiGeorge, CHARGE and Jacobsen syndromes [2,12,18,19]. Alternatively, there may have been unrecognized technical reasons for a failure of TREC-amplification. As the number of false positive results is not known, the positive predictive value can only be estimated. Assuming that none of the newborns in the NBS population was affected by a T-cell defect, we estimate a false positive rate of 0.05%. The positive predictive value largely
54
L. Tagliaferri et al. / Clinical Immunology 175 (2017) 51–55
Fig. 3. Flow chart for the process of sample analysis in the SCID screening.
depends on the frequency of SCID in the population screened. Based on an estimated frequency of 1:60,000, we calculate a positive predictive value of 0.03. Interestingly, the method reported here also identified a patient with Nijmegen-Breakage Syndrome, who was included in the panel of
positive controls. This result supports recent data [20], showing that the quantification of TRECs can also be used for the early diagnosis of relevant non-SCID lymphopenic disorders. 4.2. Time and costs The method proved to be technically simple and highly time- and cost-effective. Thanks to the simplified DNA extraction phase requiring only few minutes, the test showed excellent time-efficiency and resulted to be faster than any other published method [6,11,12]. Considering that the analysis of 96 samples in a multi-well plate requires approx. 1.5 h, we estimated that an experienced technician can process up to 425 samples with one real-time PCR machine per one working day, adding up to N2000 samples per week and N 100,000 samples per year. Based on current list prices for equipment, reagents, the material costs for DNA extraction and qPCR for TREC quantification amount to approx. € 1.14 per sample. For a qPCR for β-actin quantification € 1.83 are required. Taking into account that approx. 1.2% of samples require both the quantification of TREC and of β-actin, the average material costs for each sample amounts to € 1.18. The cost for personnel was calculated to be around € 0.80 per sample and included a full time technician, who can process 100,000 samples annually and 2 h per week of a physician, who is responsible for medical validation and tracking. This time requirement is extrapolated from the experience with other rare diseases that are screened for by our large metabolic screening center. The total cost is thus estimated at about € 2.0 per sample, which is considerably lower than the cost of currently used assays that are calculated at approx. $ 4 per sample [8,11,21,22]. Finally, using 1.6 mm punches instead of the commonly used 3.2 mm punches [11,12] reduces the required material, which is an important aspect in newborn screening particularly when considering the continuously increasing number of screened diseases. 5. Conclusions
Fig. 4. Box-Whisker-Plot showing the distribution of TREC values in the NBS population and in the positive controls (TREC copy nr./punch). The boxes represent the values included between the 1st and the 3rd quartile and the whiskers the minimum and maximum of all the data. The band inside the box shows the median value. The dashed line marks the cutoff-value, which was fixed by 95 TREC copies/punch. Finally, the spots indicate the single values below the 1st percentile for the NBS Population and above the 3rd percentile for the positive controls.
In sum, this anonymized pilot phase presents and evaluates a new reliable, rapid and cost-effective method of newborn screening for SCID. We currently plan a prospective non-anonymous screening phase of all newborns that is meant to prepare for a general newborn screening for SCID in Germany.
L. Tagliaferri et al. / Clinical Immunology 175 (2017) 51–55
Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.clim.2016.11.016. Acknowledgements We would like to thank Alessandra Sottini and Luisa Imberti for providing us a TREC-plasmid, as well as all the colleagues of the Pediatric Centers of Ulm and Freiburg who contributed with sending positive control-samples. We are finally grateful to the Dietmar Hopp Stiftung, St. Leon-Rot, Germany, for the generous funding of this project. References [1] C. Picard, W. Al-Herz, A. Bousfiha, J.L. Casanova, T. Chatila, M.E. Conley, C. Cunningham-Rundles, A. Etzioni, S.M. Holland, C. Klein, S. Nonoyama, H.D. Ochs, E. Oksenhendler, J.M. Puck, K.E. Sullivan, M.L.K. Tang, J.L. Franco, H.B. Gaspar, Primary immunodeficiency diseases: an update on the classification from the International Union of Immunological Societies Expert Committee for primary immunodeficiency 2015, J. Clin. Immunol. 35 (2015) 696–726. [2] A. Kwan, R.S. Abraham, R. Currier, A. Brower, K. Andruszewski, J.K. Abbott, M. Baker, M. Ballow, L.E. Bartoshesky, V.R. Bonagura, F.A. Bonilla, C. Brokopp, E. Brooks, et al., Newborn screening for severe combined immunodeficiency in 11 screening programs in the United States, JAMA 312 (2014) 729–738. [3] L. Brown, J. Xu-Bayford, Z. Allwood, M. Slatter, A. Cant, E.G. Davies, P. Veys, A.R. Gennery, H.B. Gaspar, Neonatal diagnosis of severe combined immunodeficiency leads to significantly improved survival outcome: the case for newborn screening, Blood 117 (2011) 3243–3246. [4] R.H. Buckley, Molecular defects in human severe combined immunodeficiency and approaches to immune reconstitution, Annu. Rev. Immunol. 22 (2004) 625–655. [5] S.Y. Pai, B.R. Logan, L.M. Griffith, H.R. Buckley, R.E. Parrott, C.C. Dvorak, N. Kapoor, I.C. Hanson, A.H. Filipovich, S. Jyonouchi, K.E. Sullivan, T.N. Small, L. Burroughs, S. SkodaSmith, et al., Transplantation outcomes for severe combined immunodeficiency 2000–2009, N. Engl. J. Med. 371 (2014) 434–446. [6] R.R. Howell, Report on newborn screening for severe combined immunodeficiency, secretary's advisory committee on heritable disorders in newborns and children, http://www.hrsa.gov/advisorycommittees/mchbadvisory/heritabledisorders/ recommendations/correspondence/severeimmunodeficiency.pdf2011. [7] K. Chan, J.M. Puck, Development of population-based newborn screening for severe combined immunodeficiency, J. Allergy Clin. Immunol. 115 (2005) 391–398. [8] J. van der Spek, R. Groenwold, M. van der Burg, J.M. van Montfrans, TREC based newborn screening for severe combined immunodeficiency disease: a systematic review, J. Clin. Immunol. 35 (2015) 416–430. [9] J.M. Puck, Laboratory technology for population-based screening for severe combined immunodeficiency in neonates: the winner is T-cell receptor excision circles, J. Allergy Clin. Immunol. 129 (2012) 607–616.
55
[10] M.C. van Zelm, M. van der Burg, A.W. Langerak, J.J.M. van Dongen, PID comes full circle: applications of V(D)J recombination excision circles in research, diagnostics and newborn screening of primary immunodeficiency disorders, Front. Immunol. 2 (2011) 12. [11] M.W. Baker, W.J. Grossman, R.H. Laessig, G.L. Hoffman, C.D. Brokopp, D.F. Kurtycz, M.F. Cogley, T.J. Litsheim, M.L. Katcher, J.M. Routes, Development of a routine newborn screening protocol for severe combined immunodeficiency, J. Allergy Clin. Immunol. 124 (2009) 522–527. [12] A. Kwan, J.A. Church, M.J. Cowan, R. Agarwal, N. Kapoor, D.B. Kohn, D.B. Lewis, S.A. McGhee, T.B. Moore, E.R. Stiehm, M. Porteus, C.P. Aznar, R. Currier, F. Lorey, J.M. Puck, Newborn screening for severe combined immunodeficiency and T-cell lymphopenia in California: results of the first 2 years, J. Allergy Clin. Immunol. 132 (2013) 140–150. [13] J.W. Verbsky, M.W. Baker, W.J. Grossman, M. Hintermeyer, S. Datu, B. Bonacci, S. Reddy, D. Margolis, J. Casper, M. Gries, K. DeSantes, G.L. Hoffman, C.D. Brokopp, C.M. Seroogy, J.M. Routes, Newborn screening for severe combined immunodeficiency; the Wisconsin experience (2008–2011), J. Clin. Immunol. 32 (2012) 82–88. [14] M.W. Baker, R.H. Laessig, M.L. Katcher, J.M. Routes, W.J. Grossman, J.W. Verbsky, D.F. Kurtycz, C.D. Brokopp, Implementing routine testing for severe combined immunodeficiency within Wisconsin's newborn screening program, Public Health Rep. S2 (125) (2012) 88–95. [15] N.M. Chase, J.W. Verbsky, J.M. Routes, Newborn screening for SCID: 3 years of experience, Ann. N. Y. Acad. Sci. 1238 (2011) 99–105. [16] N. Nakagawa, K. Imai, H. Kanegane, H. Sato, M. Yamada, K. Kondoh, S. Okada, M. Kobayashi, K. Agematsu, H. Takada, N. Mitsuiki, K. Oshima, O. Ohara, Quantification of κ-deleting recombination excision circles in Guthrie cards for the identification of early B-cell maturation defects, J. Allergy Clin. Immunol. 128 (2011) 223–225. [17] S. Borte, Ü. von Döbeln, A. Fasth, N. Wang, M. Janzi, J. Winiarski, U. Sack, Q. PanHammarström, M. Borte, L. Hammarström, Neonatal screening for severe primary immunodeficiency diseases using high-throughput triplex real-time PCR, Blood 119 (2012) 2552–2555. [18] R. Correa-Rocha, A. Pérez, R. Lorente, S. Ferrando-Martínez, M. Leal, D. Gurbindo, M.Á. Muños-Fernández, Preterm neonates show marked leukopenia and lymphopenia that are associated with increased regulatory T-cell values and diminished IL-7, Pediatr. Res. 71 (2012) 590–597. [19] S. Chabra, C. Cottrill, M.K. Rayens, R. Cross, D. Lipke, M. Bruce, Lymphocyte subsets in cord blood of preterm infants: effect of antenatal steroids, Biol. Neonate 74 (1998) 200–207. [20] J.P. Patel, J.M. Puck, R. Srinivasan, C. Brown, U. Sunderam, K. Kundu, S.E. Brenner, R.A. Gatti, J.A. Church, Nijmegen breakage syndrome detected by newborn screening for T cell receptor excision circles (TRECs), J. Clin. Immunol. 35 (2015) 227–233. [21] K. Chan, J. Davis, S.Y. Pai, F.A. Bonilla, J.M. Puck, M. Apkon, A Markov model to analyze cost-effectiveness of screening for severe combined immunodeficiency (SCID), Mol. Genet. Metab. 104 (2011) 383–389. [22] C.A. Saavedra-Matiz, J.T. Isabelle, C.K. Biski, S.J. Duva, M.L. Sweeney, A.L. Parker, A.J. Young, L.L. DiAntonio, L.M. Krein, M.J. Nichols, M. Caggana, Cost-effective and scalable DNA extraction method from dried blood spots, Clin. Chem. 59 (2013) 1045–1051.