Zhong Qing, Wagner Ulrich, Kurt Henriette, Molinari Francesca, Cathomas Gieri, Komminoth Paul, Barman-Aksözen Jasmin, Schneider-Yin Xiaoye, Rey Jean-Philippe, Vassella Erik, Rogel Uwe, Diebold Joachim, McKee Thomas, Jochum Wolfram, Kashofer Karl, Hofman Paul, Zischka Melanie, Moch Holger, Rechsteiner Markus, Wild Peter J
Department of Pathology and Molecular Pathology, University Hospital Zurich, 8091, Zurich, Switzerland; Children's Medical Research Institute, University of Sydney, 2145, Westmead, Australia.
Department of Pathology and Molecular Pathology, University Hospital Zurich, 8091, Zurich, Switzerland.
Pathol Res Pract. 2018 Jul;214(7):957-963. doi: 10.1016/j.prp.2018.05.020. Epub 2018 May 22.
Next-generation sequencing (NGS) enables parallel analysis of multiple genomic targets. The increasing demand for NGS-based multiplexed molecular diagnostics requires standardized protocols and recommendations to ensure reproducibility and accuracy of test results for routine clinical decision making. However, the lack of clinical NGS data from multi-laboratory tests and the absence of inter-laboratory comparisons have hampered the establishment of instructive clinical NGS standards. To fill the gap, we set up Proficiency Testing (PT) for inter-laboratory comparison, in which formalin-fixed paraffin-embedded specimens from eight lung and eight colon cancers were analyzed by 15 European molecular diagnostic laboratories on three different platforms using multiple target enrichment systems. We first performed platform, test, and informatics pipeline validation and conducted sensitivity and specificity analysis by random in silico down-sampling. We then implemented a multi-level filtering strategy based on performance tests of base substitution, replicate runs, and Sanger sequencing verified variants. We finally applied the filter criteria to the NGS data from the respective PT participants and obtained high inter-laboratory agreement. We demonstrated accuracy, scalability, and robustness of NGS by means of PT, serving as a benchmark for detecting clinically actionable molecular alterations in research and diagnostic laboratories. In conclusion, this study strongly highlights the importance of establishing standards for NGS-based testing, particularly when the test results impact on clinical decisions, and systematically provides data sets from multiple different labs to infer such standards.
下一代测序（NGS）能够对多个基因组靶点进行平行分析。基于NGS的多重分子诊断需求不断增加，这就需要标准化方案和建议，以确保常规临床决策中检测结果的可重复性和准确性。然而，缺乏来自多实验室检测的临床NGS数据以及实验室间比较的缺失，阻碍了具有指导意义的临床NGS标准的建立。为了填补这一空白，我们设立了实验室间比较的能力验证（PT），其中15个欧洲分子诊断实验室使用多种靶向富集系统，在三个不同平台上对来自8例肺癌和8例结肠癌的福尔马林固定石蜡包埋标本进行了分析。我们首先进行了平台、检测和信息学流程验证，并通过随机虚拟下采样进行了敏感性和特异性分析。然后，我们基于碱基替换、重复运行和桑格测序验证变异的性能测试，实施了多级过滤策略。我们最终将过滤标准应用于各自PT参与者的NGS数据，并获得了较高的实验室间一致性。我们通过PT证明了NGS的准确性、可扩展性和稳健性，为研究和诊断实验室检测临床可操作分子改变提供了一个基准。总之，本研究强烈强调了建立基于NGS检测标准的重要性，特别是当检测结果影响临床决策时，并系统地提供来自多个不同实验室的数据集以推断此类标准。
