Public support for neonatal screening for Pompe disease, a broad-phenotype condition

New treatments as well as high-throughput and multiplex screening methods are stimulating policymakers and legislators to consider adding new diseases to blood-based neonatal screening panels. An internationally accepted goal of neonatal screening is benefit for the neonate [1]. Lately, neonatal screening policy discussion is also addressing the value of genetic knowledge for family members besides the infant [2] and the value of a predictive diagnosis without the (immediate) possibility or need for intervention [3]. Several lysosomal storage disorders are being considered in the context of expanded neonatal screening (reviewed in [4]), including Pompe disease.

Pompe disease (MIM ID #232300) is an autosomal recessive enzyme deficiency due to mutations in the gene coding for acid alpha-glucosidase (GAA; MIM ID *606800). Insufficient alpha-glucosidase activity leads to accumulation of glycogen in the cells and eventually to progressive muscle weakness. Pompe disease has a broad geno- and phenotypic spectrum. The most severe, classic infantile form of disease has a birth prevalence of about 1:138,000 in the Netherlands [5]. It presents at a median age of 1.6 months but is usually diagnosed between 4.5 and 5.3 months [6]. The natural course of classic infantile disease includes cardiac hypertrophy and rapidly progressive muscle weakness; without treatment, infants rarely survive beyond 1 year of age [7]. Pre- and post-marketing studies taken together show that enzyme replacement therapy has a positive effect on the heart, prevention of muscle weakness and life expectancy, especially when treatment starts early. However, some children with classic infantile Pompe disease do not get long-term benefit from enzyme replacement therapy, even when it is given promptly. One predictive factor is CRIM status (cross-reactive immunological material), which reflects whether patients produce any endogenous acid alpha-glucosidase [8, 9].

Besides the classic infantile form of Pompe disease, there is a milder, more slowly progressive form often referred to as late-onset which affects about 1:57,000 people [5]. It is often referred to as late-onset Pompe disease and this term will be used in this paper, though it must be stressed that this slowly progressive form can manifest at almost any age, from infancy through the fifth decade, often with a considerable delay between first complaints and diagnosis [10]. The heart is not affected but there is progressive, proximal muscle weakness and often respiratory problems. Many patients become dependent on assisted ventilation and need a wheelchair. An 18-month randomized, controlled trial showed a modest but significant positive effect of enzyme replacement therapy on the ability to walk and on the stabilization of pulmonary function [11]. This and earlier studies suggest that the better the condition of the patient at start of therapy, the greater the benefit. Late-onset patients currently receiving therapy may have been diagnosed decades before therapy was available. Therefore the full potential of therapy will not be known until greater numbers of patients have started treatment soon after diagnosis. A separate question is how best to shorten the delay between first complaints and diagnosis.

The first large-scale experience of neonatal screening for Pompe disease, including clinical follow-up, comes from Taiwan. Over 300,000 newborns have been screened using a fluorimetric assay and evolving diagnostic algorithms which include the relative amount of acarbose-resistant acid alpha glucosidase [12–14]. Full performance statistics have not been published since the pilot phase of the program when false positive rates were rather high [12]. Clinical benefit of neonatal screening has been reported for classic infantile cases (all CRIM positive) [13]. The screening and diagnostic algorithms in Taiwan have so far led to identification of 13 infants without cardiac involvement, who have been classified with 'later-onset' Pompe disease. They were put under surveillance, and some have started treatment with enzyme replacement therapy [14]. Pilot studies without clinical follow-up have been done using various techniques, in Japan [15], Austria [16] the United States [17] and northern Germany [18]. At the time of writing, four U.S. states have mandated screening for Pompe disease to start by 2012 [4].

Currently published first-tier screening methods using blood spots cannot distinguish between classic infantile and late-onset Pompe disease. Confirmatory testing, which includes clinical and laboratory procedures, will identify classic infantile cases but will alert even more parents to the possibility of late-onset disease in their child. Although a severity-rating scale for GAA mutations has been developed [19], genotyping offers little certainty in predicting the age of symptom onset and the rate of disease progression for late-onset patients [20, 21]. Thus some infants flagged by screening will become "patients-in-waiting" [22, 23]. The ethical complexity of neonatal screening for Pompe disease has indeed been recognized in the literature [24–28]. Neonatal screening for Pompe disease will detect more than twice as many late-onset cases as classic infantile cases; in fact it is likely that the proportion of late-onset cases is currently underestimated, as neonatal screening for other lysosomal storage disorders has identified larger than expected numbers of (probable) late-onset cases [29–31]. This prospect raises the question of whether the expected benefits of neonatal screening for Pompe disease do indeed outweigh its possible drawbacks. This issue of 'proportionality' is especially important if the screen falls under a directive paradigm of protection through public health. It might be less problematic under a paradigm of individual choice by consumers. Although previous studies have addressed parental acceptability of neonatal screening for a treatable late-onset disease [32, 33] or a hypothetical, untreatable late-onset disease, [34] there is a lack of studies which address acceptability of neonatal screening for a treatable disease of broad phenotype such as Pompe disease.

The aims of this study were to measure support for neonatal screening for Pompe disease in the general public and to test whether (parents of) patients differ from the general public in their support for neonatal screening for their condition. A questionnaire specific to neonatal screening for Pompe disease was developed. The study aimed at a broad age range of informants, in contrast to previous studies on hypothetical acceptability of expanded neonatal screening in the general population which targeted prospective parents or parents with children under 18 [34, 35].

A broad age base was sought because neonatal screening for Pompe disease touches on long-term predictive testing, an emerging field which may affect people in many phases of life. The questionnaire was grounded in descriptive ethics [36]. Screening can be studied as a public health issue (should the state implement it for the benefit of the population?) or as a matter of individual choice (does the informed consumer want to use the test?). In this study, informants were asked to think along both perspectives. This approach is relevant for jurisdictions like the Netherlands where although newborn screening is a routine procedure with high uptake, it is not mandatory. The main outcome measures were (1) informed judgement on whether the government should offer neonatal screening for Pompe disease and (2) informed, hypothetical choice to use the screening. The size of the respondent groups permitted exploration of a few extra demographic variables which might explain the main outcome measures. Educational level was chosen because it has been shown previously that lower educational level is associated with stronger willingness to extend neonatal screening criteria [35]. Gender was also included because male respondents are underrepresented in many surveys on issues in neonatal screening e.g. [35, 37]. The survey included two other major features. It explored the moral reasoning for the main outcome measures, focussing on ethical principles which often appear in discourse on neonatal screening frameworks. Finally, the survey offset benefits and harms of neonatal screening for Pompe disease in case of a positive first-tier blood spot test.

The impetus for this study was that several jurisdictions are considering the expansion of neonatal screening to include treatable lysosomal storage disorders. Sometimes patient advocates have promoted the expansion of neonatal screening for particular conditions, while evidence-based reviews by professional experts have been more hesitant [47]. Pompe disease is of special interest because it exemplifies disorders where screening will flag a population of diverse phenotypes. The opinion of citizens is generally unknown. To our knowledge this is the first quantitative study which measured the general public's opinion about neonatal screening for a specific disorder and directly compared it to the opinion of (parents of) patients with that particular disorder.

This study found evidence for rather high support for neonatal screening for Pompe disease in the general public, as measured among consumer panel members who had been effectively informed prior to completing a questionnaire. Their support was consistent, whether questions were framed from the public health perspective (should Pompe disease be added to the neonatal screening program?) or the personal perspective (would you probably use the screening?), and it was confirmed after standardization for demographics of the Dutch population. The expected annual numbers of false positives and (probable) late-onset cases were acceptable to most respondents in the consumer panel. Furthermore, balancing benefits and harms within these two unsought screening outcomes came out on the side of benefit for the child, both in the short- and long term. Moderate harm was expected for parents.

Public opinion as measured in this study can be compared to policies formulated by experts. By far the most common decisive reasons to use screening in this study were the chance for a better quality of life of the child and the chance for health gain. These preferences fit fairly well with the criteria for neonatal screening as formulated by expert advisory boards. For example, the Health Council of the Netherlands formulated the main goal of neonatal screening as prevention of health damage, while it also recognized other advantages though according them lesser weight, e.g. faster diagnosis and better care [48]. In the U.S.A. the Secretary's Advisory Committee on Heritable Disorders in Newborns and Children uses a very similar prioritization [49].

Most informants in this study were not concerned about lack of autonomy of a child being screened for late onset disease. This may be due, at least in part, to the fact that so-called late-onset Pompe disease may manifest in childhood. Among the portion of the general public who would not use screening, the most common decisive reason was concern that a test outcome of 'possibly late-onset Pompe disease' is too burdensome for a growing child, which relates to the ethical principle of avoiding harm. It is difficult to relate this finding to current policy, because screening children for broad phenotype conditions is not addressed in guidelines such as the ESHG guideline for genetic testing of children [50] or guidelines for genetic screening for chronic conditions [51]. Notably, testing of asymptomatic children is discouraged unless it has consequences for preventive, medical actions during childhood.

This study had complex findings on whether (parents of) patients with Pompe disease differ from the general public in their support for neonatal screening for their disease. On the one hand (parents of) patients did not exceed the position of the general public when considering neonatal screening from a public health- or population perspective (approval of a government offer of screening, acceptability of rates of false positives and late-onset detection). On the other hand, (parents of) Pompe patients were more supportive of early detection of late-onset disease than the general public; they expected more benefit for such children on a lifetime scale and less harm for their parents. Moral reasoning around screening was similar for the general public and (parents of) patients, with one exception. (Parents of) patients were less concerned than the general public that screening might not improve children's quality of life sufficiently to justify screening. Taken together, these results show that (parents of) patients are as sensitive as the general public to most concerns around neonatal screening for Pompe disease, but (parents of) patients expect more benefit from early detection of late-onset disease than the general public.

Univariate regression analysis showed that educational level explained approval of screening. A post-hoc multivariate analysis, in which some categories were combined, showed that independent of Pompe status, people with high educational level were more likely to be reserved about screening than people with middle or lower education. Further studies should try to confirm this finding, which may be related to the phenomenon that lower educational level is associated with approval for expanding neonatal screening to include untreatable diseases [35]. Regression analyses also suggested that women might be more likely to be reluctant to use screening than men, but the broad confidence interval does not support a strong effect of gender.

This study has several limitations. First, standardization was only done for age and gender or educational level, while other demographic variables may also explain attitude towards neonatal screening for Pompe disease. For the group familiar with Pompe disease, this study did not differentiate between parents and patients, who might well have distinctive perspectives. Next, selection bias may have occurred against the least educated stratum of the targeted population, i.e. people with prevocational secondary education who were classified within lower education. The questionnaire was not pre-tested on readers from this stratum. In the consumer panel, lower education was a risk factor for non-response, for failing the comprehension quiz and for skipping the question on decisive reason to use screening. It is not clear whether lower education also played a role in non-response for the Pompe group. Since the comprehension cut-off used for inclusion excluded a fair proportion (26%) of consumer panel members from analysis, it is noted that over 90% of the excluded group supported an offer of screening and would probably use screening (5 to 8% missing). This suggests that more lenient inclusion criteria would not have changed the study's main conclusions.

Although comprehension was an inclusion criterion, there was some evidence of information bias. First, the fact that some respondents from the general public would decline screening because Pompe does not occur in their family shows that the introductory information did not always succeed in conveying the risk of recessive inheritance. Misconceptions in this area are common, and the written information provided in this study was not followed up by face-to-face discussion or counselling. Second, readers' comments in the pre-test phase suggested confusion with the question "What do you think is the net effect of the screening on this child in its first year of life?" on the late-onset scenario. Responders were unsure which specific benefits and harms they should be weighing, or they found it unacceptable to explicitly offset harms and benefits. Despite possibly weak validity, the question was retained because it created a temporal contrast with the following question, which addressed the life-time perspective.

The rather high public support measured in the Netherlands for neonatal screening for Pompe disease may be shared by citizens of other countries with a similar level of health care. Yet in other countries too some people, including a portion of the patient community, may have reservations about adding this condition to a state-run public health program. Results of this study suggest that it would be challenging to effectively inform less-well educated parents about the opportunities and risks of neonatal screening for a broad-phenotype condition like Pompe disease. This is especially important for jurisdictions like the Netherlands, which aim for informed consent from parents prior to neonatal screening. Yet regardless of whether screening is mandatory, citizens' views are important for policy-makers and legislators who are judging the benefit-to-risk balance of screening for a new condition. Therefore additional studies should be undertaken to confirm which determinants explain the range of public opinion on neonatal screening for a broad-phenotype condition like Pompe disease. Finally, the hesitation expressed by some respondents for screening for a broad phenotype at birth may be construed as encouragement for exploring alternative ways to shorten the diagnostic delay of Pompe disease.