Quality considerations and major pitfalls for high throughput DNA-based newborn screening for severe combined immunodeficiency and spinal muscular atrophy

Abstract

Background: Many newborn screening programs worldwide have introduced screening for diseases using DNA extracted from dried blood spots (DBS). In Germany, DNA-based assays are currently used to screen for severe combined immunodeficiency (SCID), spinal muscular atrophy (SMA), and sickle cell disease (SCD).

Methods: This study analysed the impact of pre-analytic DNA carry-over in sample preparation on the outcome of DNA-based newborn screening for SCID and SMA and compared the efficacy of rapid extraction versus automated protocols. Additionally, the distribution of T cell receptor excision circles (TREC) on DBS cards, commonly used for routine newborn screening, was determined.

Results: Contaminations from the punching procedure were detected in the SCID and SMA assays in all experimental setups tested. However, a careful evaluation of a cut-off allowed for a clear separation of true positive polymerase chain reaction (PCR) amplifications. Our rapid in-house extraction protocol produced similar amounts compared to automated commercial systems. Therefore, it can be used for reliable DNA-based screening. Additionally, the amount of extracted DNA significantly differs depending on the location of punching within a DBS.

Conclusions: Newborn screening for SMA and SCID can be performed reliably. It is crucial to ensure that affected newborns are not overlooked. Therefore a carefully consideration of potential contaminating factors and the definition of appropriate cut-offs to minimise the risk of false results are of special concern. It is also important to note that the location of punching plays a pivotal role, and therefore an exact quantification of TREC numbers per μl may not be reliable and should therefore be avoided.

Fig 1. Effect of freezing in the CXCE-buffer extraction.

Comparison of Cq values without freezing (not frozen, yellow) and with freezing (frozen, blue) in the TREC (A) and the SMN1 (B) assays.

Fig 2. MagNA pure 96 vs. CXCE-buffer extraction.

Comparison of TREC amplification signals from MagNA Pure 96 extraction (black amplification curves) vs. CXCE-buffer extraction (red amplification curves).

Fig 3

Evaluation of pre-analytic DNA carry-over in TREC (A) and SMN1 (D) amplification curves. In total 364 samples, including 40 white/empty samples, 80 TREC negative samples (dark blue lines: white/empty punches; light blue dashed lines: TREC negative punches; orange lines: positive samples (for TREC≈300 000 copies/ml)). Boxplots and the defined cut-off values based on the Cq values of the TREC (B) and SMN1 (C) assays with TREC and SMN1 positive samples in orange, negative samples in blue. Further 110 samples showed no amplification for TREC and 8 no amplification for SMN1 and therefore cannot be included in the Boxplot. (D) represents 364 samples, including 40 white/empty (blue lines) and SMN1 positive samples (orange lines).

Fig 4

Comparison of amplification curves (A) of the middle/inner DBS area (M) vs. the periphery/edge DBS area (P) with the CXCE-buffer extraction. The performance of the amplification curves for the nucleic acids of TRECs shows the impact of the chosen punching area within a DBS of the same sample (blue lines/boxes: middle DBS Area (M), orange lines/boxes: periphery DBS area (P)). Direct comparison of the respective Cq values for areas M and P for TREC (B) and SMN1 (C) DNA.