Newborn screening and carrier screening for cystic fibrosis: alternative or complementary?

Effects of carrier screening on newborn screening

Carrier screening detects couples who for each pregnancy have a 25% probability of having a child with CF. These couples can choose from various reproductive alternatives, which include the full range of options if screening occurs before conception, and more limited choices if a pregnancy has already started. Several of these options, i.e. change of partner, pre-implantation or prenatal diagnosis, affected foetus termination and adoption will lower the live birth prevalence of CF, as has been reported in the UK [10], USA [11], and Italy [4].

The decrease in the numbers of neonates with CF would not be associated with an equivalent decrease of false positives, with the final effect of lowering the positive predictive value (ratio between true positives and the total of true positives and false positives) of the newborn screening system.

Furthermore, carrier screening would not reduce the number of infants in which CF can neither be confirmed nor excluded after newborn screening [12]. These equivocal situations are considered by most an unwanted side-effect of newborn screening. Over time a few of these infants will develop CF but many others will remain healthy. It is unlikely that carrier screening will reduce the incidence of this phenomenon, because gene sequence variants associated with it are usually not included in standard screening mutation panels. Moreover, it has been anticipated that the potentially milder or absent clinical picture should have no consequence on parents' reproductive plans. Thus, the ratio between CF and these inconclusive diagnoses, which in parts of the USA has been reported to be close to 2:1 [13], may be expected to decrease and the number of equivocal conditions detected by newborn screening to approximate that of new CF cases.

Effects of newborn screening on carrier screening

Families of affected infants diagnosed through newborn screening are offered genetic counselling. Mutation analysis in relatives has a higher yield than carrier screening in the general population, both because of the higher a priori carrier risk and of the greater awareness of the significance of the test in families with a personal experience of CF. Even so, most babies with CF are born to families with no history of CF, and family testing following newborn screening is expected to have modest effects on CF incidence [14].

Depending on the configuration of the newborn screening protocol, a variable number of false positives are detected, who usually turn out to be carriers. Although heterozygotes tend to have higher IRT levels than non-heterozygotes [15], the number of carriers detected by newborn screening is small compared with the entire neonatal heterozygous population, as only the blood spots from the relatively few babies with an elevated IRT proceed to mutation analysis. Families where a carrier is found are also offered genetic counselling, but even after expanding the test to relatives, the vast majority of carriers are not detected by newborn screening [16].

Costs effectiveness

One of the advantages of the introduction of CF newborn screening is that it has been added to existing newborn screening programmes, which significantly defrays the cost. The IRT assay is relatively cheap and the more expensive CFTR mutation analysis is performed only in the small subset of IRT-positive newborns. The cost per infant diagnosed can be calculated, and it can even be estimated that savings are possible by getting infants into care early and delaying the onset of expensive medical requirements for advanced CF disease [17]. Conversely, the introduction of carrier screening for CF involves the establishment of a service as there is very little in the way of inherited disease screening for couples planning a pregnancy or for pregnant females in most countries. In contrast to newborn screening, established carrier screening programmes will result in a lower live birth prevalence of CF and the potential for an alternative source of cost savings. The cost effectiveness data, with regard to carrier screening, suffer from methodological variations in terms of expenditure for care and difficulties in estimating what the uptake would be for this service. However, two recent studies in Australia [18, 19], using different estimates of the cost of care, indicate that, after the initial investment to establish a carrier screening programme, cost savings are possible.

There are no studies to calculate the direct relationship between CF carrier and newborn screening. It is plausible that the decrease in incidence following carrier screening implementation would have no impact on the total cost of newborn screening programmes, but the cost per infant detected would increase.

Could carrier screening replace newborn screening?

Depending on their attitudes and on the preconception or prenatal timing of the test, some carrier couples will elect to have no children with CF and others will pursue their original reproductive plans. The latter scenario would result in prompt CF diagnoses, either at birth, if prenatal diagnosis had been performed, or shortly afterwards by testing the newborn for the parental mutations and/or by carrying out an early sweat test. Thus, it is theoretically possible for carrier screening to replace newborn screening. However, a number of practical issues make it quite unlikely to happen.

First, newborn screening coverage rates include nearly all the neonatal population, while even the best planned community-wide carrier screening programmes do not reach the entire target group. An exception to this occurs in premarital programmes in some motivated communities, such as Ashkenazi Jews, a paradigm difficult to adopt in the general cultural framework [20].

Secondly, the sensitivity of CF newborn screening programmes is usually >95%, whereas in many carrier screening programmes is close to 85%. Since it is necessary to identify both partners as carriers to offer a meaningful risk assessment of having an affected pregnancy, an 85% detection rate would identify only approximately three quarters of the carrier couples tested.

Finally, the guidelines recommend only testing for CF-causing mutations found in at least 0.5–1% of the local CF population [2]. This can further decrease the carrier test sensitivity in regions where the distribution of mutations is more heterogeneous, thus making the number of undetected heterozygotes and the residual carrier risk too high to justify the implementation of carrier screening.

While carrier screening is unlikely to replace newborn screening, there are some potential advantages to the infant diagnosed in utero. There is growing evidence that CF lung damage starts very early in life, and may even be present at the time of diagnosis by newborn screening at 4–6 weeks of age [21]. For 5% of the CF population (those with the p.Gly551Asp mutation) there is a newly developed medication (ivacaftor) that improves chloride transport through CFTR [22]. Further compounds aimed at other mutations are under development. Although ivacaftor is currently licensed to be used from 6 years of age, it is reasonable to suppose that the earliest possible initiation of therapy would provide the greatest benefits. A phase III clinical trial on the same compound in children aged 2–5 years is currently recruiting patients [23]. Furthermore, 20–30% of infants with CF have residual exocrine pancreatic function at birth that is then lost over 6–12 months. Early therapy with CFTR function restoring medications might prevent such loss and obviate the need for replacement enzymes. Taking this argument one step further, treatment of pregnant females with an affected foetus may possibly prevent some of the in utero development of CF complications, such as meconium ileus, absent vas deferens and pancreatic exocrine insufficiency.

Conclusion

CF newborn and carrier screening have complementary roles and neither can replace the other. In contrast to newborn screening, carrier screening allows informed reproductive choices before the birth of a child with CF. Carrier screening may get infants into care earlier than newborn screening, but would miss more affected babies.