Whole genome sequencing of newborn babies is a complex intervention with interdependent factors (such as gene selection, variant pathogenicity, penetrance, expressivity and benefits of earlier treatment) contributing to the balance of benefits and harm. There is a paucity of evidence around these factors for the screening context, and we have very low levels of knowledge about the balance of benefit and harm. We have highlighted many research gaps that span the complete evaluation process of WGS for newborn screening and propose possible research approaches to address these. We do not propose an order of priority for these recommendations, as all the research is needed to inform future policy decisions about WGS. It is imperative that this research is undertaken as part of large joined-up and possibly international collaborations to produce the evidence that is needed to thoroughly assess the benefits and harms of WGS for newborn screening for rare and ultra-rare conditions.
The research recommendations are as follows with justifications outlined below:
- a systematic review of gene and variant selection approaches reported in newborn genomic studies
- large studies to establish the penetrance of pathogenic variants in the newborn screening population [either cohort studies, healthy adult biobank studies, i.e. looking for absence of variants believed to be highly pathogenic in childhood or studies of newborn dried blood spots (DBSs)]
- clinical effectiveness studies comparing newborn genetic screening with no screening using rare disease registries
- comparative test accuracy studies of gene panel tests, WES, WGS and expanded NBS testing
- UK micro-costing studies to help underpin future cost-effectiveness evaluations of WGS
- broader research about public perception and understanding of WGS, views on risks and uncertainty and acceptability and accuracy of WGS in different populations and ethnic groups.
A systematic review of gene and variant selection approaches
A systematic review of studies reporting gene and variant selection approaches will help in the understanding of prioritisation and reporting decision of genomic projects. This may aid in the formulation of a consensus, best practice gene curation approach for newborn genetic screening and inform the evaluation of GEL’s gene list and classification of pathogenicity of included variants as part of the future evaluation of WGS for newborn screening.
Large studies to establish the penetrance of pathogenic variants in the newborn screening population
We recommend addressing the key question of penetrance in a screening population and produce the evidence needed for an evaluation. The evaluation of the subsequent interdependent factors can then focus on conditions only where there is good evidence of high penetrance for at least some pathogenic variants. Large research studies implementing newborn screening using WGS, and reporting what they detect (i.e. diagnostic yield) and offering treatment, do not provide the penetrance and expressivity data required for decisions about whether to implement. This is because if variants are reported to parents as a positive diagnosis in their newborns and they are offered treatment, and they go on to have a good health outcome, there is no way of knowing whether treatment was curative or whether they would never have had any symptoms or effects and did not require treatment. So, it is unknown if they benefited or were harmed. Current newborn genomic projects report results for variants to parents, whose babies will receive clinician-guided management. This treatment provision precludes the estimation of penetrance from the study. If studies were to choose to report fewer pathogenic variants, only those where there is existing good evidence that penetrance is high, that would provide the evidence that policy-makers need on the penetrance of other pathogenic variants (which can be accurately measured through follow-up to symptomatic disease in the absence of diagnosis and treatment). Another key challenge in this area is that even with a very large sample of 100,000 newborns, the power to analyse results by the condition or variant is extremely low due to the rarity of most relevant conditions.
To understand penetrance, the following combinations of research studies could be employed:
- The ideal studies would be large cohort studies where either screening with WGS is given to newborn babies, without reporting results to parents or very few conditions of well-evidenced penetrance and expressivity are reported to ensure more good than harm. Other potential conditions/genes should not be reported to participants unless they present symptomatically to measure the penetrance and clinical significance of pathogenic variants and therefore the benefits and overdiagnosis harms of revealing the test results. This would enable the accumulation of evidence about penetrance for conditions/variants, which could be assessed on an ongoing basis for addition to the study/programme, if there is sufficient evidence of more good than harm. This may generate the evidence needed to allow a gradual or stepped implementation in the future based on levels of evidence on pathogenicity and penetrance/expressivity, burden and cost of available treatment, availability of confirmatory tests and disease onset. However, these studies would raise considerable ethical questions and would only be possible with parents’ complete understanding, agreement and consent. There would be concern that once parents were asked whether they want to participate in a trial of WGS without telling them what you find, none of them will be in equipoise.
- The existing large genomic screening cohort studies can produce evidence on the penetrance for variants which are not included on their panels for reporting and treatment. This may provide some useful data because of the large variation in gene lists between studies. However, a note of caution should be applied because some of the rationale for different inclusion of variants may be differences between populations, which would mean we would be primarily measuring the penetrance for variants that are of lesser importance in that population and penetrance measures may not be generalisable to the UK population. A further complication is that good quality follow-up to confirmed disease is necessary, such as using robust disease registries, which is not always available.
- Genetic information from healthy adult cohorts, such as the UK Biobank and worldwide datasets like GNOMAD, could be used to identify low penetrance variants which are not suitable for inclusion on a screening panel. If large numbers of healthy adults have a variant believed to be associated with childhood-onset disease, then it will not be a highly penetrant variant suitable for use in newborn screening. However, these studies are less useful for identifying which variants to include because they will exclude people with the disease who have died before reaching adulthood.
- Whole genome or exome sequencing of stored DBS samples could provide excellent data on penetrance if the samples are of sufficient quality for sequencing and if there is a robust system of follow-up to ascertain symptomatic disease status. An example of this approach has been successfully applied in California, where Adhikari et al.182 performed WES on DBS samples and achieved exomes comparable to exomes from fresh blood in 1090/1416 samples. However, only samples from IEM-affected and MS/MS false-positive samples were included and WES results were compared to follow-up testing precluding the estimates of penetrance for the screening context. The challenges of doing these studies at scale are the cost and linking to good phenotype data.
Clinical effectiveness studies using rare disease registries
For evidence on the clinical effectiveness of genetic screening in newborns, future research could explore the role of rare disease registries. A recent study from Germany evaluated the effectiveness of genetic newborn screening for SMA by comparing the outcomes in screen-detected and symptomatically detected patients (asymptomatic vs. symptomatic treatment start) within the same healthcare system.183 This was feasible because two pilot projects for genetic SMA newborn screening were performed in Germany in two federal states before its nationwide implementation in 2021. This represents an excellent example of how a study can be designed within a rare disease registry. However, registry data are not collected to address specific RQs and limitations will be common. Registry studies rely on the high quality of data collection and reporting. The registry studies identified in our reviews were of low quality because none was comparative, no information on the type of test was provided and the information on the definitions of disease was insufficient.
Comparative test accuracy studies
Comparative (test accuracy) studies of gene panel tests, WES, WGS and expanded NBS testing are required to provide the evidence on which type of test is most promising for newborn screening. A technology-centric approach (WGS for all conditions) has been identified as inappropriate for population screening172 and the best-suited test needs to be identified for each condition under consideration. For instance, WES has been shown to have insufficient sensitivity and specificity to replace MS/MS for IEMs in general, but effectiveness varied among individual IEMs.182
United Kingdom micro-costing studies
The UK micro-costing studies are needed to help underpin future cost-effectiveness evaluations of WGS, particularly when conducted in a routine clinical practice setting, rather than a research setting and across a range of clinical contexts, to understand more fully when and why costs may differ.
Studies on public perception, views and acceptability
Finally, there is a need for broader research about public perception and understanding of WGS and whether parents would still be in equipoise once they are fully informed. Further research should aim to map diverse views on complex topics in genomics and screening and associated views on risks and uncertainty to social characteristics, backgrounds and particular lived experiences to facilitate a nuanced understanding of the distributions of harms and benefits of genomic screening to inform policy recommendations and implementation. Future research also needs to investigate the acceptability and accuracy of WGS in different populations (including island populations who are genetically diverse from mainland UK) and ethnic groups, including those where consanguinity is common, and any resultant ethical and equity issues. Population genomic testing experience is skewed towards Western Europeans and has resulted, for instance, in poorer prediction of polygenic risk scores in non-European populations.172 Concerns over deepening health disparities with WGS are warranted.