Newborn screening for spinal muscular atrophy: The views of affected families and adults

Spinal muscular atrophy (SMA) is one of the leading genetic causes of infant death worldwide. However, due to a lack of treatments, SMA has historically fallen short of Wilson‐Jungner criteria. While studies have explored the acceptability of expanded newborn screening to the general public, the views of affected families have been largely overlooked. This is in spite of the potential for direct impacts on them and their unique positioning to consider the value of early diagnosis. We have previously reported data on attitudes toward pre‐conception and prenatal genetic screening for SMA among affected families (adults with SMA [n = 82] and family members [n = 255]). Here, using qualitative interview [n = 36] and survey data [n = 337], we report the views of this same cohort toward newborn screening. The majority (70%) of participants were in favor, however, all subgroups (except adults with type II) preferred pre‐conception and/or prenatal screening to newborn screening. Key reasons for newborn screening support were: (1) the potential for improved support; (2) the possibility of enrolling pre‐symptomatic children on clinical trials. Key reasons for non‐support were: (1) concerns about impact on the early experiences of the family; (2) inability to treat. Importantly, participants did not view the potential for inaccurate typing as a significant obstacle to the launch of a population‐wide screening program. This study underscores the need to include families affected by genetic diseases within consultations on screening. This is particularly important for conditions such as SMA which challenge traditional screening criteria, and for which new therapeutics are emerging.

increasingly acknowledged that traditional Wilson-Jungner criteria (now over 40 years old) do not adequately accommodate the very specific challenges posed by genetic disorders (Andermann, Blancquaert, Beauchamp, & Dery, 2008). In response to this, various attempts have been made to develop focused genetic screening criteria, however, uptake has been inconsistent and there appears to be no universally accepted standards to appraising potential genetic screening programs (Cornel et al., 2012;Walters, 1992).
As the criteria used to guide genetic screening policies come under scrutiny, the views and perspectives of stakeholder groups set to be affected by them have gained significance. Various studies have been undertaken exploring attitudes to expanded newborn screening, however, these have tended to focus on the views of clinicians (Hiraki, Ormond, Kim, & Ross, 2006) and/or (expectant) parents (e.g., Hasegawa, Fergus, Ojeda, & Au, 2011), with far less attention paid to the views of families living with the conditions that are potential screening candidates (with a few notable exceptions: Fragile X (Skinner, Sparkman, & Bailey, 2003), Mucopolysaccharidoses (Hayes, Collins, Sahhar, Wraith, & Delatycki, 2007), and Duchenne/Becker Muscular Dystrophy (Wood et al., 2014). This lack of consultation with affected families is surprising given that they are set to be directly impacted by the introduction of newborn screening, both through the change in public profile of the disease, but also through potential advances in research as affected children come to be enrolled earlier (and potentially presymptomatically) onto clinical trials. Aside from these impacts, families living with potentially screened-for conditions are also in a privileged position to consider the impact that an early diagnosis would have had for their lives, and consequently have much to offer studies considering the effects and desirability of expanded newborn screening (Wood et al., 2014). This paper addresses this identified gap in literature by presenting attitudes toward newborn genetic screening (NGS) among families and individuals living with a condition for which NGS could feasibly soon be offered-Spinal Muscular Atrophy (SMA) (Phan, Taylor, Hannon, & Howell, 2015;Swoboda, 2010). Indeed, in light of emerging therapies for SMA, NGS for the condition is receiving renewed interest, evidenced by the formation of the "newborn screening working group" and the submission of SMA for consideration by the federal Recommended Uniform Screening Panel (RUSP) in early 2017. SMA is a neuromuscular disorder and one for which NGS has been described as particularly critical, not only because of the serious impact SMA has on families (Klug et al., 2016) and the acknowledged difficulties with obtaining a timely diagnosis (Lin, Kalb, & Yeh, 2015), but also because developing treatments for the condition requires children to be entered into clinical trials prior to the onset of symptoms, which is typically early in life (Prior & Nagan, 2016;Swoboda, 2010). While a limited number of studies have been conducted to explore public attitudes toward NGS for SMA (Rothwell, Anderson, Swoboda, Stark, & Botkin, 2013), there is very little evidence on the views of affected families, bar one study which included the views of five parents of SMA-affected children (Wood et al., 2014).
Spinal Muscular Atrophy (SMA) is an autosomal recessive neuromuscular disorder and is a leading genetic cause of infant death (Munsat & Davies, 1992). Although presenting symptoms are due to the loss of the alpha motor neurones of the spinal cord (Munsat & Davies, 1992), recent reports have shown more systemic pathology (Somers et al., 2016;Thomson et al., 2016). It is sub-classified into four main types, based on age of onset, severity, and inability to reach defined motor milestones (Munsat & Davies, 1992;Prior & Nagan, 2016;Prior & Russman, 1993;Prior, Nagan, Sugarman, Batish, & Braastad, 2011). Type I SMA is the severe form, with onset within the first few months of life and death usually occurring before 18 months through respiratory failure (Munsat & Davies, 1992). Type II SMA (intermediate) is the most divergent form, with onset usually within the first 2 years of life (Munsat & Davies, 1992). The impact on lifespan for individuals living with type II is dictated by the degree of respiratory involvement, with affected individuals facing end of life events anywhere from adolescence to late adulthood. Although mildly progressive, type II disease pathways tend to involve long "static" periods where symptoms do not change significantly (Glanzman et al., 2011;Munsat & Davies, 1992). Type III SMA is usually diagnosed after the age of 4 years, with the majority of able to sit and stand unaided (Dunaway et al., 2012;Glanzman et al., 2011;Munsat & Davies, 1992;Oh, Kim, Shim, & Sunwoo, 2011). Type IV SMA is diagnosed in adulthood, with patients developing generalized muscle weakness (Clermont et al., 1995). In both type III and IV there is a gradual deterioration in abilities over time, although life span is usually unaffected Clermont et al., 1995;Munsat & Davies, 1992).
We have previously reported data from the SMA Screening Survey (UK), which tested the views of 337 adults associated with SMA on three separate screening programs for the condition: (1) PCGS; (2) PNGS; and (3) NGS .
Our initial study reported the data on PCGS and PNGS; here we report the cohort's views on NGS. To the best of our knowledge, this is the largest study to date to systematically describe the views of SMA-affected families and adults toward NGS. We also explore their perceptions of the key social and ethical concerns which currently surround NGS more broadly.

| MATERIALS AND METHODS
An exploratory sequential mixed methods research design was adopted to address the complex and multi-faceted question of screening for SMA. This design involved the use of qualitative interviews (n = 36) which were used to inform the development of a survey which was subsequently administered to a larger sample of families and adults with SMA (n = 337), as set out below. the newsletter of the main support and advocacy group for families living with SMA in the UK, SMA Support UK. The interviews were designed to explore experiences with SMA, views around and previous/anticipated use of reproductive genetic technologies, as well as perceptions of NGS for SMA.

| Qualitative interviews
Interviews were either completed over the telephone (n = 31) or face-to-face (n = 5), depending on participant preference and geographical location. The interview recordings were transcribed verbatim (with names and identifiers removed or changed) and the data analyzed using qualitative data analysis software, Nvivo10. A constructivist approach to grounded theory data analysis was used in order that the participants' own meanings and interpretations guided the analysis, rather than those of the researcher. Initially, "open coding" of the data was carried out which was largely descriptive, before hierarchical coding was undertaken. A process of coding, refinement of concepts (through data interpretation), followed by re-coding was carried out over a period of 5 months until "theoretical saturation" had occurred (Glaser, 1967). The qualitative analysis was completed by an experienced qualitative researcher, under the supervision of two senior academic mentors with expertise in qualitative methodology.

| SMA Screening Survey (UK)
The SMA Screening Survey (UK) was developed directly from the qualitative data in order to ensure that the priorities of SMA families were reflected in the survey questions. The survey assessed views on PCGS, PNGS, and NGS. The survey was developed through single sentence "attitude/belief" statements derived from the qualitative interviews, which were in turn developed into quantitative survey questions through the use of a Likert scale. As such, the seven key themes from the qualitative analysis were directly used to delineate the key domains of the survey. In this way, the qualitative analysis directly informed the content of the survey (see Table 1 for a list of statements). Questions designed to capture demographic information from respondents (such as educational attainment, religious faith, and ethnicity) were either directly replicated from, or appear as modified versions of, questions used in the 2011 UK Census survey.
As well as the underpinning qualitative work, the survey was also passed through three expert panels, made up of professionals working with families affected by SMA (SMA Support UK/SMA Patient Registry) as well as people living with SMA themselves. Ethical approval for the survey was granted (separately to that for the qualitative interviews) by the Biomedical and Scientific Research Ethics Committee in July 2014.
Quantitative data collection was carried out over a period of 10 months, from September 1st, 2014 to June 30th, 2015. Two versions of the survey were made available, an online version (hosted on a secure website) and a paper copy. The survey was made available online via UK SMA Support and the Imaging Future research website.
Potential participants were invited to complete the survey if they were aged 18 or over and either had SMA themselves, or at least one diagnosis of SMA in the family. People affected by one the variant forms of SMA (Spinal Muscular Atrophy and Respiratory Distress, Spinal Bulbar Muscular Atrophy) were also invited to take part. No restrictions were placed on the type of family members invited to take part: step-, adopted and fostered family members were included. The recruitment strategy for family members was kept broad (and included non-biological relatives) as the social relationship to the person with SMA was considered as important as the biological relatedness of the person.
While the SMA Screening Survey (UK) also included questions on PCGS and PNGS, due to the very specific social and ethical issues pertaining to these types of screening (i.e., those of selective reproduction), data on attitudes to NGS are the focus of this paper. Data on the other screening programs are discussed elsewhere .

| Statistical analysis
The attitudes of families and adults with SMA toward NGS were compared to determine if there were any statistical differences. The following subgroup analyses were performed: All participants were analyzed collectively to identify any overriding trends (all participants).
Responses from families (all) and adults with SMA (all) were compared to determine if living with the disease directly altered views. Subanalyses on participants associated with the three most prevalent childhood forms of SMA (types I, II, and III) were then performed.
Responses from families associated with type I were compared with responses from families with milder forms (type II/III SMA (combined), type II alone, and type III alone)-to determine if severity altered families' views. Responses were compared between families and adults with SMA, to determine if the relationship to SMA affects views (when severity is standardized). This analysis was split into three: (1) type II-associated participants; (2) type III-associated participants; and (3) type II/III combined (the combined analysis was performed to facilitate logistic regression analysis based on the relatively low number of adults with SMA in the two subgroups. Finally, responses from adults with type II were compared to adults with type III, and responses form type II families were compared to type III families. This assessed whether the severity and age of diagnosis impacts views, and whether any differences were seen in both families and adults living with the disease. For the subgroup analysis, families members associated with more than one form of the disease were classified according the most severe form within their family (e.g., a family with a type I and type II child would be classified as a type I family).
In each of the subgroup analyses, the individual questions were assessed and then responses correlated against support for screening.
For each question the number of "agree" vs. "other" responses were reported and statistical differences between the subgroups were assessed using a chi-squared analysis (Graphpad Prism software, v6).
Associations between positive "agree" responses to each question were assessed using binary logistic regression (performed against survey Q20l (I would support a newborn genetic screen for SMA).

| RESULTS
The cohort characteristics have been previously reported . Briefly, of the 337 participants, 255 were family members of people with SMA (75.7%) and 82 had SMA themselves (24.3%).
Overall, 70% of survey participants were in favor of NGS, with no statistical differences between any of the analyzed sub-groups (Tables   1 and 2). However, the overall levels of support were lower than the previously reported levels of support in the same participants for both PCGS (77%) and PNGS (76%) . Response breakdowns are shown for family subgroups (all, type I, type II/III combined, type II and type III) and patient subgroups (all, type II/III combined, type II and type III). Responses for each question were stratified as "agree" versus "other" (other = disagree and neither disagree nor agree). This stratification was used to enable binominal logistic regression of each subgroup.  (Tables   3 and 4).
We compared the levels of support for NGS against support for two alternative programs (PCGS and PNGS). As reported here and elsewhere , in general there is more support for PCGS and PNGS than NGS in all analyzed subgroups (Table 5). The kappa analysis suggests there is a minimal-weak agreement within each subgroup; this is important, because it highlights that participants are not simply infavor of all tests, instead there are subtle differences between the different groups that reflect their views and experiences.
As highlighted in the analysis, adults with type II SMA are the only subgroup that preferentially support NGS over the other groups (Table   5). This is in keeping with our previous report, which demonstrates these participants have a comparatively positive view on their condition, believing they have fulfilling lives and can have a valuable impact on society . Therefore, their support for NGS is understandable, because it is the one test that would not result in fewer children with type II SMA being born (this was highlighted in our previous study as one of the main reasons adults with type II SMA were opposed to PCGS and PNGS programs) . screening. This support of trials was fairly even across all types of SMA as well as between family members and adults with SMA (Table   1). It is unclear whether participants perceived a direct benefit to trial enrollment for SMA children or whether they accepted the indirect benefits. However, the importance of supporting such trials, as well as the earlier introduction of support and healthcare, the importance of an earlier diagnosis, and the benefits in terms of future reproductive decisions all featured as positive drivers for NGS support ( Table 1).
The importance of an early SMA diagnosis and trial enrollment has received increased attention recently following the preliminary reports from a phase 2, open-label, dose-escalation study of Nusinersen (an antisense oligonucleotide that modifies SMN2 RNA splicing) Finkel et al., 2016;Hache et al., 2016).
The trial involved 20 participants, with 2-3 copies of SMN2 and age of onset ranging from 21 to 154 days (Finkel et al., 2016). Data from this trial demonstrated that pre-symptomatic infants at high genetic risk of type I SMA responded well to Nusinersen, achieving motor milestones in timelines more consistent with normal development (Finkel et al., 2016). These findings suggest that improved outcomes (motor function, achieved motor milestones, and increased time to ventilation) could be achieved if pre-symptomatic patients (identified through NGS) could be enrolled and treated with Nusinersen (or similar ASOs).
This therapeutic has been approved by the U.S. FDA and may be prescribed for newborns with high genetic risk for type I SMA.  Harper, 1990;Grob, 2008). Given the gravity of an SMA diagnosis, the lack of available treatments and difficulties associated with accurate prognostic information, this was also an issue that emerged as significant for SMA families.
Concerns about the impact of NGS on the early experiences of the family were also evident in attitudes toward the impact of a pre-  NGS did so not out of a rejection of screening per se, but rather because they wanted screening in a different form. In contrast, the data highlight that for 22% of adults with type II SMA rejected all forms of screening for SMA. It is noteworthy, however, that this view was not evidenced among adults with type III SMA, and seems to be related to the perceptions of the condition among adults with type II. Shakespeare postulates that people with fixed impairments from birth or early childhood are often better adjusted to their disabilities than those whose impairments are later onset, fluctuate, or involve periods of decline or deterioration (Shakespeare, 2006). Our analysis grouped participants as "families" or "adults with SMA." This means we have not reported whether there are differences Odds ratios are presented for response breakdowns are shown for family sub-groups (all, type I, type II/III combined, type II and type III). Responses for each question were stratified as "agree" versus "other" (other = disagree and neither disagree nor agree). Odds ratios (OR) show the likelihood that responders who agreed with each individual question (variable) were also in favor of newborn genetic screening. Positive drivers are indicated by a odds ratio >1 and a p-value <0.05); negative drivers are indicated by an odds ration <1 and a p-value <0.05). Significant drivers are highlighted in bold. Several ORs are presented as NR (not returned); this is because of the low number of responders and included "other" responses. (other = disagree and neither disagree nor agree). Odds ratios (OR) show the likelihood that responders who agreed with each individual question (variable) were also in favor of newborn genetic screening. Positive drivers are indicated by a odds ratio >1 and a p-value <0.05); negative drivers are indicated by an odds ration <1 and a p-value <0.05). Significant drivers are highlighted in bold. Several ORs are presented as NR (not returned); this is because of the low number of responders and included "other" responses. between close (parents, siblings) and distant (cousins, uncles etc) family members. This was because the low numbers involved for some of the family members reduced the significance of the analysis.
In conclusion, this study highlights that for families living with SMA, NGS is viewed favorably by the majority of participants, irrespective of the availability of treatments and irrespective of the screen's ability to accurately determine the type of SMA affecting the infant. This finding is in contrast to policy reviews and criteria where the absence of accurate typing and treatment for SMA were seen as fatal flaws to screening implementation (Cartwright, 2012). It is also in contrast to attitudes toward other forms of screening for SMA (PCGS and PNGS), where inability to determine type was controversial, particularly among adults with type II SMA .
Unlike PCGS and PNGS, which potentially involve the prevention, or termination, of lives affected by SMA , NGS, through its focus on early detection, is the least emotive, and consequently the least divisive, form of screening for SMA. It has, furthermore, been identified by the SMA research community as the form of screening most likely to yield the most progress in terms of treatment development, through its concomitant increase in infants participating in clinical trials (Phan et al., 2015).

This research was funded by the Economic and Social Research
Council (Grant Number: ES/K002090/1). The authors would like to acknowledge with gratitude the guidance and support with recruitment provided by SMA Support UK (formerly the Jennifer Trust for SMA) and the UK SMA Patient Registry. Special thanks go to the families and adults living with SMA who both advised on, and participated in this study.

CONFLICTS OF INTEREST
The authors, Felicity Boardman, Philip Young, and Frances Griffiths each declare that they have no competing or conflicting interests.