目的:评估中国实验室扩大的无创产前筛查(NIPS)拷贝数变异(CNV)的现状,国家临床实验室中心进行了外部质量评估(EQA)计划。
方法:EQA小组由12个与不同综合征相关的人工样本组成,将其与从孕妇收集的母体血浆和来自具有5%至15%的不同胎儿分数(FF)的细胞系的酶消化的无细胞DNA(cfDNA)混合。该小组通过下一代测序进行了验证,并分发给实验室,以及问卷和案例场景。
结果:69个实验室参与了EQA计划,91.30%(63/69)的实验室正确识别了所有样本。共有7.25%(5/69)的实验室报告假阴性结果,2.90%(2/69)的实验室报告了意外的CNVs。22q11.2缺失综合征的正确率,Cri-du-chat综合征,1p36缺失综合征和Angelman/Prader-Willi综合征样本占97.46%,98.55%,100%,100%,分别。随着FF的增加,删除大小,和阅读深度,检出率提高。对于结果报告,只有五个实验室报告了FF值,一个实验室报告了CNV分类类型,没有人报告敏感性,特异性,阳性预测值,和阴性预测值。
结论:NIPS对CNVs的检测能力仍需改进和标准化,FF,删除大小,和读取深度是影响检测率的因素。
OBJECTIVE: To evaluate the current situation of expanded noninvasive prenatal screening (NIPS) for copy number variations (CNVs) in laboratories in China, the National Center of Clinical Laboratories conducted an externalqualityassessment (EQA) program.
METHODS: The EQA panel consisted of 12 artificial samples associated with different syndromes, which were mixed with maternal plasma collected from pregnant women and enzyme-digested cell-free DNA (cfDNA) from cell lines with different fetal fractions (FFs) ranging from 5% to 15%. The panel was validated by next-generation sequencing and distributed to laboratories, along with questionnaires and case scenarios.
RESULTS: Sixty-nine laboratories participated in the EQA program, and 91.30% (63/69) of laboratories correctly identified all samples. A total of 7.25% (5/69) of the laboratories reported false-negative results, and 2.90% (2/69) of the laboratories reported unexpected CNVs. The correct rates of the 22q11.2 deletion syndrome, Cri-du-chat syndrome, 1p36 deletion syndrome and Angelman/Prader-Willi syndrome samples were 97.46%, 98.55%, 100%, and 100%, respectively. With the increase in the FF, deletion size, and read depth, the detection rate increased. For results reports, only five laboratories reported FF values, one laboratory reported the CNV classification type, and none reported sensitivity, specificity, positive predictive values, and negative predictive values.
CONCLUSIONS: The detection capabilities of NIPS for CNVs still need to be improved and standardized, and FF, deletion size, and read depth are factors that affect the detection rate.