Spargo Thomas P, Iacoangeli Alfredo, Ryten Mina, Forzano Francesca, Pearce Neil, Al-Chalabi Ammar
Department of Basic and Clinical Neuroscience, Maurice Wohl Clinical Neuroscience Institute, King's College London, London WC2R 2LS, UK.
Department of Biostatistics and Health Informatics, King's College London, London WC2R 2LS, UK.
Biomedicines. 2025 Apr 23;13(5):1018. doi: 10.3390/biomedicines13051018.
Genomic sequencing enables the rapid identification of a breadth of genetic variants. For clinical purposes, sequencing for small genetic variations is considered a solved problem, while challenges remain for structural variants, given the lower sensitivity and specificity. Interest has recently risen among governing bodies in developing protocols for population-wide genetic screening. However, usefulness is constrained when the probability of being affected by a rare disease remains low, despite a positive genetic test. This is a common scenario in neurodegenerative disorders. The problem is recognised among statisticians and statistical geneticists but is less well-understood by clinicians and researchers who will act on these results, and by the general public who might access screening services directly without the appropriate support for interpretation. We explore the probability of subsequent disease following genetic screening of several variants, both single nucleotide variants (SNVs) and larger repeat expansions, for two neurological conditions, Huntington's disease (HD) and amyotrophic lateral sclerosis (ALS), comparing these results with screening for phenylketonuria, which is well-established. The risk following a positive screening test was 0.5% for in ALS and 0.4% for in HD when testing repeat expansions, for which the test had sub-optimal performance (sensitivity = 99% and specificity = 90%), and 12.7% for phenylketonuria and 10.9% for ALS when testing pathogenic SNVs (sensitivity = 99.96% and specificity = 99.95%). Subsequent screening confirmation via PCR for led to a 2% risk of developing ALS as a result of the reduced penetrance (44%). We show that risk following a positive screening test result can be strikingly low for rare neurological diseases, even for fully penetrant variants such as , if the test has sub-optimal performance. Accordingly, to maximise the utility of screening, it is vital to prioritise protocols with very high sensitivity and specificity, and a careful selection of markers for screening, giving regard to clinical interpretability, actionability, high penetrance, and secondary testing to confirm positive findings.
基因组测序能够快速识别多种遗传变异。就临床目的而言，小的遗传变异测序被认为是一个已解决的问题，而结构变异仍存在挑战，因为其敏感性和特异性较低。最近，管理机构对制定全人群基因筛查方案的兴趣有所增加。然而，当尽管基因检测呈阳性但患罕见病的概率仍然很低时，其有用性就受到了限制。这在神经退行性疾病中是一种常见情况。统计学家和统计遗传学家认识到了这个问题，但临床医生、将依据这些结果采取行动的研究人员以及可能直接获得筛查服务而没有适当解读支持的普通公众对这个问题的理解较少。我们探讨了对两种神经系统疾病——亨廷顿病（HD）和肌萎缩侧索硬化症（ALS）——进行几种变异（包括单核苷酸变异（SNV）和较大的重复扩增）基因筛查后发生后续疾病的概率，并将这些结果与成熟的苯丙酮尿症筛查结果进行比较。当检测重复扩增时，ALS的筛查阳性后患病风险为0.5%，HD为0.4%，该检测的性能欠佳（敏感性 = 99%，特异性 = 90%）；当检测致病性SNV时，苯丙酮尿症的患病风险为12.7%，ALS为10.9%（敏感性 = 99.96%，特异性 = 99.95%）。通过PCR对进行后续筛查确认，由于外显率降低（44%），导致患ALS的风险为2%。我们表明，对于罕见的神经系统疾病，即使对于像这样完全外显的变异，如果检测性能欠佳，筛查阳性后的患病风险也可能极低。因此，为了使筛查的效用最大化，至关重要的是优先选择具有非常高的敏感性和特异性的方案，并仔细选择筛查标记，同时考虑临床可解释性、可操作性、高外显率以及用于确认阳性结果的二次检测。
