Early prediction of preeclampsia in pregnancy with cell-free RNA

Abstract

Liquid biopsies that measure circulating cell-free RNA (cfRNA) offer an opportunity to study the development of pregnancy-related complications in a non-invasive manner and to bridge gaps in clinical care^1-4. Here we used 404 blood samples from 199 pregnant mothers to identify and validate cfRNA transcriptomic changes that are associated with preeclampsia, a multi-organ syndrome that is the second largest cause of maternal death globally⁵. We find that changes in cfRNA gene expression between normotensive and preeclamptic mothers are marked and stable early in gestation, well before the onset of symptoms. These changes are enriched for genes specific to neuromuscular, endothelial and immune cell types and tissues that reflect key aspects of preeclampsia physiology^6-9, suggest new hypotheses for disease progression and correlate with maternal organ health. This enabled the identification and independent validation of a panel of 18 genes that when measured between 5 and 16 weeks of gestation can form the basis of a liquid biopsy test that would identify mothers at risk of preeclampsia long before clinical symptoms manifest themselves. Tests based on these observations could help predict and manage who is at risk for preeclampsia-an important objective for obstetric care^10,11.

Fig. 1. Comparing sample, maternal and pregnancy characteristics for normotensive and preeclampsia groups across cohorts.

a, Matched sample collection time both across gestation (left) and after delivery (right). b, Maternal characteristics (P = 0.02 comparing BMI in the discovery cohort). c, Matched gestational age at preeclampsia onset regardless of preeclampsia symptom severity. d, Gestational age at delivery (P = 6 × 10⁻⁷, 0.04, 8 × 10⁻⁹) for the discovery (n = 49 normotensive [37, 36, 39, 30]; n = 24 with preeclampsia [13, 17, 20, 17]), validation 1 (n = 32 normotensive [19, 27, 19, 19]; n = 7 with preeclampsia [3, 8, 6, 5]) and validation 2 (n = 61 normotensive [61]; 26 with preeclampsia [28]) cohorts. Square brackets indicate the sample number per collection group. For a, statistics were calculated by sample group. For b–d, statistics were calculated by cohort group (NS = not significant, *P < 0.05, **P ≤ 10⁻⁷; two-sided (a–c) and one-sided (d) Mann–Whitney rank test).

Fig. 2. Before 20 weeks of gestation, cfRNA measurements segregate preeclampsia and normotensive samples and are enriched for neuromuscular, endothelial and immune cell types and tissues.

a, Distribution of log₂(FC) for DEGs (n = 544) with dashed lines at log₂(FC) = ±1. b, Before 20 weeks of gestation, a subset of DEGs can separate preeclampsia (PE) and normotensive samples despite differences in symptom severity, preeclampsia onset subtype and gestational age (GA) at delivery. HIST2H2BE is also known as H2BC21. See Supplementary Table 3 for more information on genes included in heatmaps. c, Comparison of log₂(FC) for DEGs between the discovery and the validation 2 cohorts reveals strong agreement. d, DEGs for preeclampsia as compared to normotensive samples can be described as either increased (orange) or decreased (dark blue) in preeclampsia over gestation. Points indicate median per trend and shaded region indicates 95% CI. e, Approximately 13% of DEGs are tissue- or cell-type-specific when compared with the Human Protein Atlas (HPA) and an augmented Tabula Sapiens (TSP+) atlas.

Fig. 3. A subset of cfRNA changes can predict risk of preeclampsia early in gestation.

a, Classifier performance as quantified by receiver operator characteristic curve (ROC) for samples collected in early gestation between 5 and 16 weeks, with AUROC and corresponding 90% CI noted per cohort. b, Prediction of preeclampsia incorporates cfRNA levels for 18 genes for which normalized centred log₂(FC) trends hold across the discovery (n = 61 normotensive, 24 preeclampsia), validation 1 (n = 35 normotensive, 8 preeclampsia), validation 2 (n = 61 normotensive, 28 preeclampsia) and Del Vecchio (n = 17 normotensive or other complication, 5 preeclampsia) cohorts as confirmed using univariate analysis (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.005; one-sided Mann–Whitney rank test with Benjamini–Hochberg correction). See Supplementary Table 5 for exact P values. For box plots, centre line, box limits, whiskers and outliers represent the median, upper and lower quartiles, 1.5× interquartile range and any outliers outside that distribution, respectively. Plot limits are −8 to 4 to better visualize the main distribution. log₂(CPM), log₂-transformed counts per million reads.

Fig. 4. cfRNA measurements reflect the multifactorial nature and pathogenesis of preeclampsia over pregnancy before diagnosis.

a, DEGs for preeclampsia with as compared to without severe features (n = 503) can be described by four longitudinal trends. b, Comparison of organ and cell-type changes over gestation for eight organ systems reflects the multifactorial nature of preeclampsia and provides a possible means to comprehensively monitor maternal organ health. Points indicate median per sample group and shaded region indicates 95% CI in a and 75% CI in b. LSECs, liver sinusoidal endothelial cells; NK cells, natural killer cells.

Extended Data Fig. 1. Samples with outlier values for at least one QC metric cluster separately from most non-outlier samples.

a–c, For discovery (a), validation 1 (b), and validation 2 (c), hierarchical clustering (left) and PCA (right) reveals that most outlier samples cluster with negative control (NC) samples and separately from non-outlier samples. d, e, Visualization of other QC metrics like the amount of cfRNA extracted (d) and the percent of reads that align uniquely to the human genome (e) (n = 209, 106, 89 samples for discovery, validation 1, and validation 2, respectively). For PCA in a–c, sample outliers and poorly detected genes drive PCA and serve as leverage points. The top two principal components are visualized when performed using all samples and all genes (leftmost PCA) or only samples that pass QC metrics (middle PCA) reveals that certain samples can act as leverage points. Once sample outliers and lowly detected genes are removed from the cfRNA gene matrix (rightmost PCA), the top two principal components reflect natural variance in the data and are no longer driven by a few leverage points. For box plots, centre line, box limits, whiskers and outliers represent the median, upper and lower quartiles, 1.5× interquartile range and any outliers outside that distribution, respectively.

Extended Data Fig. 2. Across gestation before diagnosis, changes in the cfRNA transcriptome segregate preeclampsia and normotensive samples and reflect known preeclampsia biology.

a, Distribution of CVs with dashed line at CV = 1 for all DEGs (n = 544) between preeclampsia as compared to normotensive samples across gestation. b, At ≥23 weeks of gestation and post-partum, a subset of DEGs can separate preeclampsia (n = 20, 17) and normotensive (n = 37, 29) samples despite differences in symptom severity, preeclampsia onset subtype, and GA at delivery. c, Comparison of log₂(FC) for DEGs for preeclampsia as compared to normotensive between discovery and validation 1 reveals good agreement across gestation but not post-partum. d, The genes in each longitudinal trend group reflect known preeclampsia aetiology as highlighted across four databases (GO biological processes, KEGG, the reactome, and GO cellular compartment). Some preeclampsia associated terms are emphasized in bold, coloured text that corresponds to group colour from Fig. 2d (Dark blue and orange indicate decreased and increased in preeclampsia versus normotensive, respectively) (p ≤ 0.05; one-sided hypergeometric test with multiple hypothesis correction, see Methods). e, Comparison of log₂(FC) for DEGs for preeclampsia without severe features versus normotensive and preeclampsia with severe features versus normotensive in the discovery cohort reveals good agreement along the y=x axis with a slope of 0.93, 1.03, 0.77, and 0.86 at ≤12 weeks, 13–20, ≥23 weeks, and post-partum, respectively.

Extended Data Fig. 3. Across gestation and before diagnosis, changes in the cfRNA transcriptome identified at one time point can moderately segregate preeclampsia and normotensive samples at other time points.

DEGs with |log₂(FC)| ≥ 1 and CV < 0.5 or 0.4 for the 13–20 week time point were identified at each time point across gestation. Each row visualizes how well a specific DEG subset from a given sample collection period can separate preeclampsia (n = 13, 16, 20, 17) and normotensive (n = 36, 33, 37, 29) samples in all other collection periods (≤12, 13–20, ≥23 weeks of gestation and post-partum respectively). The number of genes identified per sample collection period is highlighted along the main diagonal.

Extended Data Fig. 4. k-means clustering reveals meaningful longitudinal patterns.

a, c, The chosen k clusters (dashed line) comparing a performance metric, the sum of squared distances, and values of k for clustering of DEGs for preeclampsia versus normotensive related to Fig. 2d (a) and DEGs for preeclampsia with versus without severe features related to Fig. 4a (c). b, d, Following permutation of the data matrix prior to k-means clustering, longitudinal changes over gestation are replaced by 2 flat lines for clustering of log₂(FC) for preeclampsia versus normotensive (b) and 4 uninformative lines for clustering of log₂(FC) for preeclampsia with versus without severe features (d). For b, d, points indicate median per trend and shaded region indicates 95% CI.

Extended Data Fig. 5. Examining the logistic regression model used to predict risk of preeclampsia early in gestation.

a, Comparison of gestational age at sample collection (weeks) for incorrectly predicted (yellow) or correctly predicted (green) samples across normotensive and preeclampsia groups in the discovery, validation 1, validation 2 and Del Vecchio cohorts shows that incorrectly predicted preeclampsia samples (false negatives) are collected at later gestational ages. b, Estimated probability of preeclampsia as outputted by logistic regression for both preeclampsia and normotensive samples shows that the model is well-calibrated across most predictions. Dashed line at 0.35 indicates classifier cut-off where P(PE) ≥ 0.35 constitutes a sample predicted as preeclampsia. c, Logistic regression models trained on subsets of 1–18 genes of the initial 18 genes can moderately predict future preeclampsia onset in the validation 2 cohort with improving performance as subset size increases and as characterized by PPV, NPV, sensitivity, specificity and AUROC (left to right).