Genetic risk for Huntington Disease and reproductive decision‐making: A systematic review

Huntington Disease (HD) is an incurable autosomal dominant single gene neurodegenerative disorder. Typical onset is between 30 and 40 years and characterised by motor difficulties, cognitive impairment, and behavioural and personality changes. The availability of reproductive testing means that affected and at‐risk individuals can make reproductive decisions with genetic risk in mind. We aimed to summarise the literature on reproductive decision‐making in the context of HD risk in terms of outcomes and the subjective experiences of at‐risk individuals. Five databases were searched. Findings were synthesised using Framework analysis to identify common factors across results of quantitative and qualitative studies. Twenty five studies met inclusion criteria. Framework analysis identified the following key areas: ‘The relationship between reproductive intentions and HD genetic risk’, ‘Views on assistive options’, ‘Complexity and challenges in reproductive decision‐making’, ‘Actual reproductive outcomes’, and ‘Other factors influencing reproductive decision‐making’. Quality of included studies was mixed. Reproductive decision making in the context of HD risk was found to be a complex and emotionally challenging process. Further research is required into reproductive decision‐making and outcomes among those not utilising assistive options, and in developing a model of reproductive decision‐making in HD.

demanding, and navigating emotional and relationship changes emotionally complex. 9,10 HD is a single-gene disease with an autosomal dominant pattern of inheritance with the disease-causing variant showing full penetrance. The huntingtin gene which codes for the huntingtin protein contains a short, repeated section of the three nucleotides, cytosine-adenine-guanine (CAG) which contains fewer than 27 repeats in the normal gene. Individuals with 40 or more CAG repeats in this section will develop the disease and is classified as a positive result. Children of a parent with HD have a 50% chance of inheriting the disease-causing variant and are thus considered genetically 'at risk' of HD. 5 The HD gene was mapped in 1983 and presymptomatic genetic tests (PT) to confirm a diagnosis of HD have been available in the form of family linkage tests since 1986. However, this required having intact families and living affected family members. This type of testing was generally only available through select research centres. 11 Direct testing of the HD gene has been available since since isolation of the gene in 1993 and in 2009 the correlation between the CAG repeat size and age of symptom onset was recognised. Given this long history of genetic testing for HD and it being the first heritable disease for which PT were widely available, robust guidelines for PT in HD have been developed, including pre-test genetic counselling and posttest support following a positive result. 12 Given HD's high risk of almost 100% penetrance heredity, age of onset, progressive and incurable status, HD genetic risk knowledge has profound implications for affected and at-risk individuals. At-risk individuals are often aware of their own potential disease trajectory, and the risk of inheritance to their children. 3 At-risk individuals have several available reproductive options 13 : (1) natural conception without attempts to mitigate genetic risk; (2) use of prenatal diagnosis (PND)-natural conception, followed by in utero genetic testing, with the option to terminate or continue affected pregnancy; (3) use of pre-implantation genetic diagnosis (PGT)-genetic testing of in in vitro fertilised embryos, followed by implantation of gene-negative results; (4) non-biological routes to parenthood (e.g., sperm or egg donation); and (5) abstinence from parenthood. Clinical reviews of HD identify reproductive decision-making (RDM) as extremely challenging, involving multiple complex decisions regarding reproductive intent, use of assistive technologies and knowledge regarding own genetic status. 3 The PT protocol of HD 12 specifies that support in RDM should focus on providing clear and useful information on options and potential outcomes.
The development of PT and late reproductive assistive options has lead to a small but developing body of research exploring the reproductive intentions and outcomes in the context of HD, as well as attitudes and uptake related to PND and PGT. These range from straightforward reports on reproduction in this cohort over time to exploration of the impact of PT on outcomes, as well as a small number of qualitative studies focused on the experience of reproductivedecision making in the context of HD risk. However, this research has not yet been synthesised to allow for identification of themes and patterns regarding the impact of HD genetic risk on RDM. Therefore, this review aimed to (1) integrate research regarding reproductive intentions, decision-making and outcomes among those at genetic risk for HD; (2) report relevant findings on attitudes towards and uptake of developing assistive technologies available to aid reproduction in HD; and (3) summarise research on the subjective experience of RDM in the context of HD genetic risk. Care were searched from 1983 to 4th November 2021. The search strategy involved combining terms from previous reviews related to RDM 14 and HD 7 respectively. Relevant study reference lists were also hand searched and relevant 'grey literature' was included if the results were available, and all inclusion criteria were met. Grey literature was included to counteract publication bias, 15 which can be problematic when published research is relatively sparse. 16

| Study selection
Study selection was an iterative process following PRISMA guidance. 17 Duplicate study records were identified and removed. Next, studies were screened by title, abstract and then full-text. A second reviewer (CR) screened 10% of studies at each stage to ensure reliability of selection criteria, with an overall agreement rate of 95%.
Studies were included if they met the following criteria: involved individuals at genetic risk of developing HD; at least one aim related to RDM; published on or after 1983, the year in which genetic testing for HD became possible 18 ; peer-reviewed studies reporting novel quantitative and qualitative research, in addition to relevant 'grey literature' consisting of publicly accessible Masters and Doctoral theses meeting all other inclusion criteria; full text available in English. Studies were excluded if they did not report findings on HD or reported aggregated results with other conditions.

| Data extraction and quality appraisal
The key characteristics of all included studies were extracted. Study quality was assessed using the QualSyst tool, 19 as it allows for the appraisal of both qualitative and quantitative study designs, including cross-sectional and observational designs, and has clear, replicable guidance for quality assessment. Resultant quality ratings are expressed between 0.0 and 1.0, based on several quality assessment criteria. For the purpose of this review, the following quality ranges were used: 'High' quality, 0.80 or higher, 'Good' quality between 0.79 and 0.70, 'Medium' quality between 0.69 and 0.60, and 'Low' quality, 0.60 and lower. Low quality studies were excluded from data synthesis.

| Data synthesis
Extracted study characteristics and results were reviewed, and given the resultant heterogeneity of study designs, outcome measurement, along with the need to account for both quantitative and qualitative research, a narrative synthesis using a 'Framework Analysis' approach 20 was identified as an appropriate method of data synthesis.
This provides a flexible approach in accounting for both qualitative and quantitative data, and a robust iterative process of stepped analysis to ensure relevant data is accounted for both within and between studies. Framework analysis consists of five stages: (1) familiarisation with full text of included studies and relevant extracted data; (2) development of thematic framework based on previous research and patterns identified, reflexively adapted to proceeding analysis; (3) indexing of extracted data to identified framework, using textual codes to connect specific data to different themes; (4) charting of data across all studies to headings from developed thematic framework; and (5) mapping of patterns and associations between the data across studies, and interpretation of the dataset as a whole.
This thematic framework was developed in response to broad areas of reported data, covering both attitudes towards RDM and assistive technologies, actual reproductive outcomes and uptake of assistive technologies. The subjective challenges and complexities of RDM, both practical and emotional were also synthesised. Where possible, themes attempted to synthesise qualitative and quantitative results with equal weight in the development of overarching themes, though certain sub-themes emerge as containing only qualitative or quantitative data. On completion of the initial analysis process, the thematic structure and content was compared to the dataset to ensure representativeness. Role of others: often decision between people, but conflict between lots of sources (e.g., spouse, family, HCWs, societal pressures)-power in relationships affecting this decision; role of HCWs, especially in optimism re potential cure, or source of judgement.
Other options: adoption-mixed, helping someone but also burdening them; PGT-generally positive, though not resolving question of what happens when HD appears; CVS/Abortiontension of responsibility to look after potential child but not engender suffering, also ambivalence to idea of abortion because of HD, need for things to 'sit right'.
Guilt and shame: pouring over current and past reproductive decisions, concerns about future views of, for example, not testing or having children from self and others, responsibility to children complicated. Abstience: sometimes considered, from children and relationships, as way out-also leaving it 'up to God' as relief from complex and heavy decisions. Abstract screening left 64 papers screened at full text level. 27 met inclusion criteria and underwent quality analysis (see Figure 1 for summary). Study quality findings are in Table 1, 33% were rated as high quality, 37% as good and 22% as medium. 8% (n = 2) of studies 30,31 were excluded based on low quality ratings, leaving 25 papers included in analysis.  Table 1 for full study summary).

Child desire and future reproductive intentions
Six studies reported on future reproductive intentions. Three were of 'High' quality-one quantitative, 42 one qualitative 26 and one mixedmethods. 22 Three were 'Good' quality quantitative studies. 34,37,38 Earlier studies reported high levels of future reproductive intention-82% of participants at genetic risk for HD intended to have at least one child, and 78% two or more, 37 and 80% intended to have a further child at time of survey. 38 In comparison, rates of future reproductive intentions in more recent studies ranged between 51.59% 22 and 38%. 42 McCormack et al. 34 identified that at-risk males were more likely to express future reproductive intention than comparable controls, while atrisk females were less likely than control to express the same. Gong

Major role of HD knowledge in reproductive decision making
Five studies explore the major impact of risk knowledge on RDM.
Three 'high' quality studies-one quantitative, 42 one qualitative 35 and one mixed-methods 22 -as well as one 'good' quality quantitative study 28 and one 'medium' quality qualitative study. 23 In Quaid et al. 35

Impact of genetic testing on reproductive intentions
Seven quantitative studies of various quality reported on anticipated and actual impact of positive PT result on reproductive intentions-two 'high' quality, 24,41 three 'good' quality, 33,37,38 and two 'medium' quality. 21,27 In terms of anticipated response to PND, between 33.5% 33

Views on PT, PND and PGT
Seven studies reported on views regarding PT, PND and PGT. Two were high quality, one quantitative 42 and one qualitative, 26 three were good quality quantitative studies, 28,33,38 and two were good quality qualitative studies. 25,29 In studies conducted as PT was first becoming available, between 63.2%, 33 73% 38 and 78% 28 reported willingness to pursue PT when possible, rising to 86.5% in the event of improved treatment options. 33 However, actual uptake rates of PT were lower than suggested by these studies. Prospective PND willingness was between 48% 33 and 65%, 28 with qualitative results highlighting the emotional challenges of potential termination of an affected pregnancy. 29 Similarly, Fowler 25 highlighted the interpersonal difficulties raised by pregnancy termination in PND as the main reason for non-utilisation in one case study. Views of PGT were generally positive, qualitatively characterised as the preferred option for having unaffected children, 26 and the 'least bad' option, specifically contrasted with the emotional challenges of PND. 29 Tsang reported 88% awareness of PGT, and 58% future intention to utilise. Commonly identified negative aspects of PGT were prohibitive cost, 26,42 and previous negative experiences of the process. 42 Views and use of other assistive options (adoption, donation) Three 'good' quality qualitative studies 25

Response to changing technological options
Five studies-one 'high' quality mixed methods, 22 two 'good' quality quantitative, 39,43 one 'good' quality qualitative 44  Maat-Kievit et al. 32 found that 84% of participants continued to use the option they were familiar with from previous experience. A second area of concern was utilisation of direct PND/PGT (involving parents knowing their own genetic status) versus exclusion PND/PGT (involving parents avoiding knowledge of own risk status)-between 32% 43 and 35% 39 of those seeking PND or PGT opted for exclusion testing. Qualitative results highlighted the choice of exclusion testing as an attempt to balance the strong desire to avoid inheritance to children with the desire to avoid own risk knowledge, so as to avoid stigma, hopelessness or 'life being overshadowed' by HD.

Balancing desire for a child with responsibility
Five studies report experiences of difficulty balancing child desire with concerns regarding impact of HD inheritance-one 'high' quality mixed-methods, 22 three 'good' quality qualitative 25,29,44 and one 'medium' quality qualitative. 23 Five year post-PT, participants struggled to reconcile their desire for biological children with the responsibility to the child to avoid risk inheritance and future caring burden. 22 Klitzman et al. 29 outlined a similar 'push-pull' dynamic between these two concerns, characterised by rumination and uncertainty. 22,29 Some participants expressed a wish to return to a state of pre-risk knowledge 'ignorance', to avoid contending with challenges of managing responsibility. 23,29 This desire to achieve a balance of 'fairness' is highlighted in Downing 23 where each case study involved challenges in establishing the 'responsible' reproductive choice, desire to maintain consistency of approach across pregnancies, and attempts to establish responsibility by demonstrating aptitude for parenthood.

Risk acceptance and optimism
Five qualitative studies explore acceptance of genetic risk in pursuit of strong reproductive intentions-one was 'high' quality, 35 three were 'good' 25,29,44 and one 'medium'. 23 Across studies, participants viewed risk of HD inheritance as one risk among many, often referenced against potential, unavoidable disasters (e.g., 'could be hit by a bus tomorrow'). Overall risk was characterised as fundamentally unavoidable, and the value of pursuing strongly held reproductive intentions emphasised. 29,35,44 A process of 'positive denial' was sometimes utilised, where participants 'decide' no inheritance has occurred, or treatments will be developed, as way of managing concerns. 25,35 Other emphasised parenting suitability characterising 'good parenting' holistically. 23

Guilt, regret and rumination on past decisions
Four qualitative studies-one 'high' quality, 35 two 'good' quality 25,29 and one 'medium' quality 23 -explore rumination, guilt and regret regarding reproductive decisions. These included worries regarding the acceptability of decision to others, to future selves, and to children as they became aware of HD genetic risk later in life. 29 This was particularly challenging for those who became parents prior to risk knowledge and described looking for potential 'signs' of illness they might have missed 35 or a desire to 'start over, make different decisions'. 25 Some participants, previously 'at peace' with their reproductive decisions, begin to stressfully re-evaluate these in the light of the emerging HD symptomology of affected relatives, sometimes leading to regret. 23

Role in relationships with others
Four qualitative studies-two 'high' quality 26,35 and two 'good' quality 25,29 -comment on the role of interpersonal relationships in RDM.
The risk of developing HD and resultant RDM is influenced by the information received through healthcare professionals. 29,35 The healthcare professionals are relied upon to share this important information and when this did not occur, it was viewed as extremely harmful. 35 As well as being an important source of clinical information, these relationships could also serve as a source of implicit judgement of reproductive choices. 29 RDM is seen as inherently interpersonal, requiring negotiation with partners, 25,26,29 with RDM sometimes viewed as entirely dependent on the relationship, and not considered outside of this context. 26 Conflicting opinions, with a partner 25 or through multiple strong, differing opinions with a family 29,35 are challenging to navigate, and cause stress in the RDM process. Navigating complex relational dynamics sometimes became the prime factor in RDM, limiting options utilised and overshadowing issues of inheritance. 25

| Actual reproductive outcomes
Ten studies report on actual reproductive outcomes among this population. Three were 'high' quality-one mixed-methods 22 and two quantitative. 41,46 Four were 'good' quality quantitative studies. 36,39,43,45 Three were 'medium' quality quantitative studies. 32,40,47 Tibben et al. 41 50 and is important to acknowledge as a major factor in RDM, sometimes overshadowing risk knowledge.
'Actual reproductive outcomes' summarised research on births and utilisation of assistive options among the target population. A general trend towards avoidance of HD inheritance where PND or PGT is pursued emerges, with a minority of HD-inheriting pregnancies continued by choice. It should be noted that these studies only reported outcomes among a self-selecting sub-population, and are therefore not necessarily representative of the broader at-risk population. Uptake of PGT, and to a lesser extent PND, appears to be increasing in the population over time from 1993 when PGT first became available, but remains low as percentage of the eligible population, similar to PT. 51 Findings suggest that awareness of PGT is not total among the eligible population, and cost issues limit utilisation. However, risk knowledge may be associated with reduced perceived responsibility.
'Other factors influencing RDM', outlines demographic and experiential influences on RDM, and important indicators of the inherently contextual and reflexive nature of RDM. Sex has contradictory findings, with both females 26 and males 33 reported to be more likely to express reproductive intentions, though lack of significance reporting and overall lower study quality in Markel et al. 33  higher quality and more recent data collection, they may be more representative of current cultural attitudes. Previous parents are less likely to consider risk inheritance as important, and less likely to utilise assistive options. It may be the case that, having successfully navigated parenthood they have broader ideas of what has been and therefore will be influential on parenting. Those with increased experience of HD within their family are more hesitant regarding risk inheritance, 13 and Catholics were significantly more hesitant to pursue termination of an affected pregnancy. 33 Educational levels are associated with a range of views related to affected pregnancies, mirroring a general trend towards greater acceptance of termination as an option generally associated with greater educational attainment. 53  and re-evaluated across the lifespan, suggests the importance of facilitating access to genetic counselling reflexively to need across time, rather than discretely during testing process and assistive option utilisation only. There is an important role for patient support organisations to play in reaching the whole at-risk population with good quality information regarding RDM and encouraging early referral for genetic counselling. There is also a need to educate primary care physicians that a referral for genetic counselling does not require the individual to have PST but allows the opportunity for discussion with a professional with knowledge of HD. Lastly, the emotional complexity of these processes suggest genetic counselling services require the capacity to provide space for emotional processing, and time given to consider implications of decisions, rather than focussing solely information provision.

| Limitations
Existing literature focusses on the minority of individuals who engage with genetic counselling and therefore suffers from selection bias and may not reflect the views of the total at-risk population. Future research could focus on establishing reasons for lack of uptake of assistive options among the eligible population and how it might be facilitated, as well as exploration of RDM and outcomes among those who do not utilise these options.
In addition, there is a lack of an overarching psychological model of RDM in HD, despite both a rich body of qualitative research findings exploring its complexities, and the development of similar models in heritable physical conditions. 54 Given that HD is the longest tested for and most researched heritable neurogenerative disorder, and that on which the approach to others is often based, the lack of explanatory model of this key issues is an area for future consideration.

ACKNOWLEDGMENTS
I would lie to thank all researchers and staff at the ADAPT Lab who provided their time, support and expertise to this review.