Newborn blood spot screening for cystic fibrosis with a four-step screening strategy in the Netherlands

Abstract

Background: Newborn screening for cystic fibrosis (NBSCF) was introduced in the Dutch NBS program in 2011 with a novel strategy.

Methods: Dutch NBSCF consisted of four steps: immuno-reactive trypsin (IRT), Pancreatitis-associated Protein (PAP), DNA analysis by Inno-LiPa (35 mutations), extended gene analysis (EGA) as fourth step and as safety net. Only samples with two CFTR-variants were considered screen-positive, but samples with one disease-causing variant were considered also screen-positive from April 2013. The first 5 years of NBSCF were evaluated during a follow-up ranging from 2 to 6.8 years for sensitivity, specificity, positive predictive value (PPV), ratio of CF/Cystic Fibrosis Screen Positive infants with an Inconclusive Diagnosis (CFSPID) and median age at diagnosis, and were compared to other novel strategies for NBSCF and European Cystic Fibrosis Society (ECFS) Best Practice Standards of Care.

Results: NBSCF achieved a sensitivity of 90% (95% CI 82%-94%), specificity of 99.991% (95% CI 99.989%-99.993%), PPV of 63% (95% CI 55%-69%), CF/CFSPID ratio of 4/1, and median age at diagnosis of 22 days, if samples with two variants as well as samples with one disease-causing variant were considered screen-positive.

Conclusion: The program achieved the goal to minimize the number of false positives and showed a favourable performance but sensitivity and CF/CFSPID ratio did not meet criteria of EFCS Best Standards of Care. Changed cut-off values for PAP and IRT and classification of R117H-7T/9T to non-pathogenic may improve sensitivity to ≥95% and CF/CFSPID ratio to 10/1. PPV is estimated to be around 60%.

Keywords: DNA-analysis; Extended CFTR-gene analysis; Immuno-reactive trypsin; Newborn screening; Pancreatitis-associated protein; Validity.