Expansion of variant panels in newborn screening algorithms to identify cystic fibrosis: A retrospective analysis of 25 years of genotypes and implications on diagnosis.

Abstract

We seek to understand the incremental value of applying expanded variant panels or sequencing in population-based screening algorithms for a well-understood condition, such as cystic fibrosis (CF). We compared newborn screening methods to determine at what point do attempts to increase sensitivity of second-tier testing meet diminishing returns. Using the genotypes of all Massachusetts CF-affected patients who were born between February 1, 1999 and January 31, 2024, we retrospectively applied screening algorithms that used (1) 39 CF gene (CFTR) variants, 139 CFTR variants, or CFTR sequencing and (2) different algorithms for referral to CF Center. Sensitivity, specificity, and timing of diagnosis were evaluated. Our current 39 CFTR variant panel and referral algorithm yielded a clinical sensitivity of 98.7%. In Massachusetts, expanding the variant panel might result in limited sensitivity improvement, but if the referral algorithm requires detection of 2 CFTR variants, it might decrease the sensitivity. Expanding the CFTR variants genotyped does not necessarily guarantee an increase in screening sensitivity. Using a conservative cutoff for DNA testing might accomplish as much. Screening turnaround time, costs, and geographic location of CF Centers should be factored into decisions about the benefit of NGS technology within newborn screening.