A miRNA signature for early non-invasive diagnosis of pre-eclampsia. A miRNA signature for early non-invasive diagnosis of pre-eclampsia

Background: Preeclampsia (PE) is a multi-systemic maternal syndrome with substantial maternal and fetal morbidity and mortality. Currently, there is no clinically viable non-invasive biomarker assay for early detection, thus limiting the effective prevention and therapeutic strategies for PE. Methods: We conducted a discovery-training-validation three-phase retrospective and prospective study with cross-platform and multicenter cohorts. The initial biomarkers were discovered and verified in tissue specimens by small RNA sequencing and qRT-PCR. A miRNA signature (miR2PE-score) was developed using the Firth’s bias-reduced logistic regression analysis, and subsequently validated in two independent multinational retrospective cohorts and two prospective plasma cohorts. Results: We initially identified five PE-associated differentially expressed miRNAs from miRNA sequencing data and subsequently validated two miRNAs (miR-196b-5p and miR-584-5p) as robust biomarkers by association analysis with clinical characteristics and qRT-PCR in tissue specimens in the discovery phase. Using the Firth’s bias-reduced logistic regression analysis, we developed the miR2PE-score for the early detection of PE. The miR2PE-score showed a high diagnostic performance with an area under the receiver operating characteristic curve (AUROC) of 0.920, 0.848, 0.864 and 0.812 in training, internal and two external validation cross-platform and multicenter cohorts, respectively. Finally, we demonstrated the non-invasive diagnostic performance of the miR2PE-score in two prospective plasma cohorts with AUROC of 0.933 and 0.787. Furthermore, the miR2PE-score revealed superior performance in non-invasive diagnosis compared with previously published miRNA biomarkers. Conclusions: We developed and validated a novel and robust blood-based miRNA signature, which may serve as a promising clinically applicable non-invasive tool for the early detection of PE. Overall design: Isolated total RNA from placental tissue samples of 10 women (5 normotensive women, denoted with a NC, and 5 preeclamptic women, denoted with a PE) and used it for miRNA-sequencing on the Nova6000 platform (Illumina). Sequences were aligned to the sequences of human mature microRNAs downloaded from the miRbase database. Aligned sequences were then used to quantify the miRNAs and perform differential expression analysis.

背景：先兆子痫（preeclampsia, PE）是一种多系统性孕产妇综合征，可引发严重的孕产妇与胎儿发病率及死亡率。目前尚无临床可行的非侵入性生物标志物检测手段用于早期诊断，这限制了子痫前期的有效预防与治疗策略制定。
方法：本研究开展了一项发现-训练-验证三阶段回顾性与前瞻性研究，涵盖跨平台多中心队列。初始生物标志物通过小RNA测序（small RNA sequencing）与定量实时聚合酶链反应（quantitative real-time polymerase chain reaction, qRT-PCR）在组织标本中筛选并验证。采用弗思偏差缩减logistic回归分析构建了miRNA特征评分模型（miR2PE-score），随后在两个独立的跨国回顾性队列以及两个前瞻性血浆队列中完成验证。
结果：本研究首先从miRNA测序数据中鉴定出5个与PE相关的差异表达miRNA，随后在发现阶段通过临床特征关联分析与qRT-PCR在组织标本中验证了两个稳定性优异的生物标志物：miR-196b-5p与miR-584-5p。借助弗思偏差缩减logistic回归分析，我们构建了用于PE早期检测的miR2PE-score模型。该模型在训练队列、内部验证队列以及两个跨平台多中心外部验证队列中，受试者工作特征曲线下面积（Area Under the Receiver Operating Characteristic curve, AUROC）分别为0.920、0.848、0.864与0.812，展现出优异的诊断性能。最后，我们在两个前瞻性血浆队列中证实了miR2PE-score的非侵入性诊断性能，其AUROC分别为0.933与0.787。此外，相较于此前已发表的miRNA生物标志物，miR2PE-score在非侵入性诊断中表现更优。
结论：本研究开发并验证了一种新型且稳定的血液源性miRNA特征模型，有望成为极具临床应用前景的非侵入性工具，用于PE的早期检测。
整体实验设计：从10名女性的胎盘组织标本中提取总RNA，其中5名为血压正常女性（设为NC组），5名为子痫前期女性（设为PE组），随后在Nova6000测序平台（Illumina）上进行miRNA测序。将测序序列比对至从miRbase数据库下载的人类成熟microRNA序列，再利用比对完成的序列进行miRNA定量与差异表达分析。