Improving the Sensitivity and Positive Predictive Value in a Cystic Fibrosis Newborn Screening Program Using a Repeat Immunoreactive Trypsinogen and Genetic Analysis

Abstract

To evaluate the performance of a new cystic fibrosis (CF) newborn screening algorithm, comprised of immunoreactive trypsinogen (IRT) in first (24-48hours of life) and second (7-14days of life) dried blood spot plus DNA on second dried blood spot, over existing algorithms. A retrospective review of the IRT/IRT/DNA algorithm implemented in Colorado, Wyoming, and Texas. A total of 1 520 079 newborns were screened, 32 557 (2.1%) had abnormal first IRT; 8794 (0.54%) on second. Furthermore, 14 653 mutation analyses were performed; 1391 newborns were referred for diagnostic testing; 274 newborns were diagnosed; and 201/274 (73%) of newborns had 2 mutations on the newborn screening CFTR panel. Sensitivity was 96.2%, compared with sensitivity of 76.1% observed with IRT/IRT (105ng/mL cut-offs, P<.0001). The ratio of newborns with CF to heterozygote carriers was 1:2.5, and newborns with CF to newborns with CFTR-related metabolic syndrome was 10.8:1. The overall positive predictive value was 20%. The median age of diagnosis was 28, 30, and 39.5days in the 3 states. IRT/IRT/DNA is more sensitive than IRT/IRT because of lower cut-offs (∼97 percentile or 60ng/mL); higher cut-offs in IRT/IRT programs (>99 percentile, 105ng/mL) would not achieve sufficient sensitivity. Carrier identification and identification of newborns with CFTR-related metabolic syndrome is less common in IRT/IRT/DNA compared with IRT/DNA. The time to diagnosis is nominally longer, but diagnosis can be achieved in the neonatal period and opportunities to further improve timeliness have been enacted. IRT/IRT/DNA algorithm should be considered by programs with 2 routine screens.