Carrier Screening, Incidence of Cystic Fibrosis, and Difficult Decisions

Unplanned, uncontrolled, and unblinded experiments pose tremendous analytical challenges when treated as retrospective studies,^{1 - 2} but on those occasions when biases are successfully controlled,^{3 - 4} results can be informative and thought provoking. In this issue of JAMA, Castellani et al⁵ present their analysis of one such natural experiment, comparing the incidence of cystic fibrosis (CF) from adjacent regions in northern Italy that used different approaches to carrier detection. Screening of newborns for CF was performed throughout the entire region during the study period (1993-2007).⁶ In the western region near Verona, Alto-Adige, selective testing for CF carrier status was the practice, with detection efforts concentrating on relatives of patients confirmed to have CF and couples planning in vitro fertilization procedures. In contrast, in the eastern region near the University of Padua, Veneto, a more general program of screening for CF carrier status was offered.

Due to the difference in selectivity, the increase in the number screened in the Veneto region far outpaced those screened in the Alto-Adige region, reaching a ratio of 30:1 by 2007. The number of couples consisting of 2 individuals who were CF carriers detected in the Veneto region outnumbered those detected in the Alto-Adige region by 9:1. Most importantly, birth rates for infants with CF were significantly lower in Veneto compared with Alto-Adige, and heterogeneity in the respective populations, effects of neonatal screening for CF, and differences in general birth rates could not explain the result.⁵ By making a series of straightforward estimates, the authors suggest that approximately 113 fewer infants were born in Veneto to carrier couples. Most of these appear to be due to prevention of pregnancy; only 8 interrupted pregnancies during the study period were directly attributable to prenatal diagnosis of CF.

Before the advent of carrier screening for any disease, the choice facing couples with access to methods of pregnancy prevention was whether to attempt to have children. With the possibility of screening for carrier status, choices become more difficult. Prospective parents must consider the risk of conceiving a child with disease and, for those electing to undergo prenatal diagnosis, the acceptability of selective abortion. Providers of genetic testing face the daunting task of providing truly informed consent that accounts for cultural norms and personal beliefs.

For Tay-Sachs disease, similar to CF in its genetic nature and amenity to preconception screening, the number of children being born with this disease has decreased by an estimated 90%.⁷ In contrast with CF, treatments for Tay-Sachs disease are poor or nonexistent. Children born with Tay-Sachs inexorably quickly fail and die. Disease is coincident with birth, and prevention of disease is only synonymous with prevention of birth. In a starkly different situation from Tay-Sachs disease, the development of numerous therapies for CF has made extended survival probable and has made that survival much more normal in quality. Prevention of disease by prevention of birth is not the only option.

During the study period, CF therapies and outcomes dramatically improved. Randomized, placebo-controlled trials demonstrated that recombinant human DNase,⁸ inhaled tobramycin,⁹ chronic oral azithromycin,¹⁰ and inhaled hypertonic saline¹¹ all improved outcomes in CF. Projected survival by life-table and cohort survival analyses show enormous gains in survival; projected median survival from birth is now more than 37 years and more than 45% of those living with CF are adults.¹²

In some ways, the shifting timing of morbidity and mortality toward young adulthood in CF creates issues similar to those associated with familial cancers due to BRCA1 and BRCA2 mutations.^{13 - 14} Women with either or both of these mutations have greatly increased odds of breast and ovarian cancer early in adulthood, and men are at increased risk for breast or prostate cancers. Great morbidity can occur against a background of previously normal health and well-being. For women, disease prevention efforts have focused on aggressive cancer screening and elective risk-reducing surgeries, but recent studies have examined attitudes toward prevention of BRCA1/BRCA2-related disease by prevention of births for the sake of cancers that may—or may not—appear in the fourth to fifth decades of life. As might be expected, the opinions on the propriety of such a prevention program vary widely.^{13 - 14}

For individuals with CF, no cure exists. As therapies have been added and survival has improved, the burden to patients and their families in time, effort, and money has increased.¹⁵ To use all possible data-supported therapies, patients must devote up to 4 waking hours per day, or sometimes more, sustain great determination toward adherence, and spend many tens of thousands of dollars per year, either personally or through insurance or various other support programs. Outcomes over time are greatly improved but the personal and financial tolls continue to make CF a fearsome disease. Informed of the 1 in 4 chance of having children with CF, the majority of carrier couples in northeastern Italy contemplating parenthood chose either to avoid or, in some cases, to terminate pregnancies.⁵

For parents, raising children to adulthood promises numerous intangible rewards but exacts substantial personal and financial sacrifices. The years of effort are punctuated by moments worthy of celebration and others that can only be described as harrowing. Raising and caring for a child with CF poses additional complications in all aspects of personal life for all members of a family, and the desire to avoid those extra costs is easy to understand. On the other hand, the improvements in care have allowed many patients with CF to successfully reach adulthood, complete education, and start families and careers. Thus, a suggestion to avoid all births of children with CF seems distinctly discordant with the increasingly common successes within individual lives affected by CF due to continuing progress in treatment.

More and more, prospective parents and clinicians must face difficult decisions regarding the appropriate response to increased risk of disability, familial cancer, or a disease like CF in the context of whatever the present knowledge suggests may be true about these entities in the future. To be clear, Castellani et al⁵ have taken no position with regard to the appropriateness of prevention of CF by prevention of birth, nor should such a judgment be made. Instead, the clear response of couples in this study toward the risk of CF communicates the message that despite great and many successes, dramatically decreasing the morbidity and treatment burdens, and finding the elusive cure itself, remain as urgent and compelling as they were 71 years ago when CF was first described.¹⁶