Cystic Fibrosis Diagnosis in Newborns, Children, and Adults

Carlo Castellani, MD; Barry Linnane, MB, BCh, BAO, DCH, MRCPI, MRCPCH, MD; Iwona Pranke, PhD; Federico Cresta, MD; Isabelle Sermet-Gaudelus, MD, PhD; Daniel Peckham, MD


Semin Respir Crit Care Med. 2019;40(6):701-714. 

In This Article

Improving Specificity With Second-tier Testing

Recognition that raised IRT on the initial bloodspot sample was a sensitive screening test for CF but lacked specificity has resulted in the inclusion of (at least) a second tier of testing.[54–57,82] An early strategy involves a second IRT sample (IRT/IRT), measuring IRT at two ages, and using two different IRT cut-off values.[64,67,83,84] Advantages of this approach are that it uses a single biochemical test, and avoids the identification of carriers of a single CF mutation, or mutations of uncertain clinical significance.[64,67,83,84] Disadvantages are the requirement for a second sample from the infant, challenges in setting the second IRT value, and a concern that the process performs suboptimally when compared with IRT/DNA model.[73,76,85]

Identification of the defective gene causing CF in 1989 opened a second strategy, the IRT/DNA model. This involves referring a certain percentage of raised IRT samples, typically the top 0.5 to 5%, for DNA analysis as the second tier of testing. This approach is now employed in approximately 90% of the US states, most of Europe, and all of Australia and New Zealand.[9,43,45–47,71,85–91] Initial models involved identification of F508del (the most common CF mutation worldwide) with further mutations added over time.[9,45–47,85,86,89,91,92] Some NBS programs create customized panels to reflect the mutation prevalence of their target population.[71] Advantages of the IRT/DNA model are that the second tier of testing, the DNA analysis, can be performed on the original dried blood spot, eliminating the need for a second blood sample. The results are available sooner, the number of infants identified as being at increased risk of CF is diminished resulting in fewer false positives. The identification of two disease-causing mutations allows a presumptive diagnosis; however, a sweat test is still required to confirm the diagnosis.[8,93,94] Disadvantages are the cost, the complexity, and the inadvertent identification of CF carriers, and of infants in whom the diagnosis is uncertain.[9,46,95]

There are variations to the IRT/DNA model. Some programs recommend sweat testing when no mutations are identified but the IRT is very high (>99.8th centile).[96,97] However, Massie et al did not demonstrate an advantage in this approach. Some programs use extended mutation panels, or employ expanded gene analysis, but there is a constant trade-off between sensitivity, specificity, and positive predictive value. No program can reach 100% sensitivity and attempts to achieve it would result in unacceptable numbers of unaffected infants being identified, with associated psychological, economic, and resource costs.[16,70,71,98,99] Therefore, all programs are faced with a judgment decision regarding setting the sensitivity of their program within the constraints of their organizational resources.[71]

The measurement of the pancreatitis-associated protein (PAP) in the dried bloodspot is another option for the second tier of testing.[100–106] The combination of IRT/PAP has shown some promise, with sensitivity and specificity reported as similar to that achieved with IRT/DNA but without the unwanted effect of identifying unaffected carriers.[100–106] The positive predictive value can be improved by employing a multitier approach involving IRT, PAP, and selected, plus expanded, DNA analysis as demonstrated by the CHOPIN study.[107] The approach is yet to gain widespread acceptance but work in this area is ongoing.