Reid Emma S, Papandreou Apostolos, Drury Suzanne, Boustred Christopher, Yue Wyatt W, Wedatilake Yehani, Beesley Clare, Jacques Thomas S, Anderson Glenn, Abulhoul Lara, Broomfield Alex, Cleary Maureen, Grunewald Stephanie, Varadkar Sophia M, Lench Nick, Rahman Shamima, Gissen Paul, Clayton Peter T, Mills Philippa B
Genetics and Genomics Medicine Programme, UCL Institute of Child Health, London, UK.
Neurology Department, Great Ormond Street Hospital for Children NHS Foundation Trust, London, UK.
Brain. 2016 Nov 1;139(11):2844-2854. doi: 10.1093/brain/aww221.
Neurometabolic disorders are markedly heterogeneous, both clinically and genetically, and are characterized by variable neurological dysfunction accompanied by suggestive neuroimaging or biochemical abnormalities. Despite early specialist input, delays in diagnosis and appropriate treatment initiation are common. Next-generation sequencing approaches still have limitations but are already enabling earlier and more efficient diagnoses in these patients. We designed a gene panel targeting 614 genes causing inborn errors of metabolism and tested its diagnostic efficacy in a paediatric cohort of 30 undiagnosed patients presenting with variable neurometabolic phenotypes. Genetic defects that could, at least partially, explain observed phenotypes were identified in 53% of cases. Where biochemical abnormalities pointing towards a particular gene defect were present, our panel identified diagnoses in 89% of patients. Phenotypes attributable to defects in more than one gene were seen in 13% of cases. The ability of in silico tools, including structure-guided prediction programmes to characterize novel missense variants were also interrogated. Our study expands the genetic, clinical and biochemical phenotypes of well-characterized (POMGNT1, TPP1) and recently identified disorders (PGAP2, ACSF3, SERAC1, AFG3L2, DPYS). Overall, our panel was accurate and efficient, demonstrating good potential for applying similar approaches to clinically and biochemically diverse neurometabolic disease cohorts.
神经代谢紊乱在临床和遗传方面都具有显著的异质性，其特征是伴有提示性神经影像学或生化异常的可变神经功能障碍。尽管早期有专科医生介入，但诊断延迟和开始适当治疗的情况仍很常见。新一代测序方法仍有局限性，但已能使这些患者更早、更有效地得到诊断。我们设计了一个针对614个导致先天性代谢缺陷的基因的基因检测板，并在一个由30名表现出可变神经代谢表型的未确诊儿科患者组成的队列中测试了其诊断效力。在53%的病例中发现了至少能部分解释观察到的表型的基因缺陷。在存在指向特定基因缺陷的生化异常的情况下，我们的检测板在89%的患者中明确了诊断。13%的病例出现了可归因于多个基因缺陷的表型。我们还研究了包括结构导向预测程序在内的计算机工具对新型错义变体进行特征描述的能力。我们的研究扩展了已充分表征的疾病（POMGNT1、TPP1）和最近发现的疾病（PGAP2、ACSF3、SERAC1、AFG3L2、DPYS）的遗传、临床和生化表型。总体而言，我们的检测板准确且高效，显示出将类似方法应用于临床和生化多样的神经代谢疾病队列的良好潜力。
