Identifying newborns at risk for cystic fibrosis (CF) by newborn screening (NBS) using dried blood spot (DBS) specimens provides an opportunity for presymptomatic detection. All NBS strategies for CF begin with measuring immunoreactive trypsinogen (IRT). Pancreatitis-associated protein (PAP) has been suggested as second-tier testing. The main objective of this study was to evaluate the analytical performance of an IRT/PAP/IRT strategy versus the current IRT/IRT strategy over a two-year pilot study including 68,502 newborns. The design of the study, carried out in a prospective and parallel manner, allowed us to compare four different CF-NBS protocols after performing a post hoc analysis. The best PAP cutoff point and the potential sources of PAP false positive results in our non-CF newborn population were also studied. 14 CF newborns were detected, resulting in an overall CF prevalence of 1/4, 893 newborns. The IRT/IRT algorithm detected all CF cases, but the IRT/PAP/IRT algorithm failed to detect one case of CF. The IRT/PAP/IRT with an IRT-dependent safety net protocol was a good alternative to improve sensitivity to 100%. The IRT × PAP/IRT strategy clearly performed better, with a sensitivity of 100% and a positive predictive value (PPV) of 39%. Our calculated optimal cutoffs were 2.31 µg/L for PAP and 167.4 µg2/L2 for IRT × PAP. PAP levels were higher in females and newborns with low birth weight. PAP false positive results were found mainly in newborns with conditions such as prematurity, sepsis, and hypoxic-ischemic encephalopathy.
Keywords: cystic fibrosis; dried blood spot (DBS); immunoreactive trypsinogen (IRT); newborn screening (NBS); pancreatitis-associated protein (PAP).
© 2019 by the authors.