A NEW BIOMARKER FOR PREECLAMPSIA

Abstract

This invention relates to the use of biomarker LRG1 as a biomarker for preeclampsia for use from the first trimester. Elevated levels of leucine-rich alpha 2 glycoprotein 1 (LRG1) can predict risk for the future development of preeclampsia or other hypertensive disorders of pregnancy. The predictive test will comprise the measurement of LRG1 protein, peptide fragment, DNA or RNA, from either blood, plasma, serum, urine, saliva or amniotic fluid. The invention provides a method and a test kit to assess the risk of pre-eclampsia in pregnant woman. The method or test can utilise antibodies to measure levels of LGR1 in a sample.

Claims

1. An in vitro method for assessing the risk of preeclampsia in a subject, comprising detecting a level of LRG1 in a biological sample from said subject, wherein an elevated LRG1 level is indicative of risk of hypertensive disorders of pregnancy.

2. A method as claimed in claim 1 wherein the disorder is pre-eclampsia.

3. A method as claimed in claim 1 wherein the method comprises the steps of obtaining a biological sample from the subject; and detecting a level of LRG1 in the biological sample; wherein an elevation in the detected level of LRG1 in the biological sample relative to a control value indicates that the subject is at increased risk of preeclampsia.

4. The method of claim 1, wherein the biological sample is whole blood, serum, plasma, urine, saliva or amniotic fluid.

5. The method as claimed in claim 1, wherein the biological sample is a sample obtained from the subject during a first trimester.

6. The method as claimed in claim 1, wherein the biological sample is a sample obtained from the subject during a second trimester.

7. The method as claimed in claim 1, wherein the biological sample is a sample obtained from the subject during a third trimester.

8. The method as claimed in claim 3 wherein the control value is a reference value representative of a level of LRG1 in a sample from a subject who will not develop preeclampsia.

9. The method as claimed in claim 3 wherein the control value is a reference value representative of a level of LRG1 in a sample obtained from the same subject prior to pregnancy.

10. The method of claim 3 , wherein the elevation in the detected level of LRG1 in the biological sample relative to the control value is at least a 15% elevation.

11. The method of claim 3 , wherein the elevation in the detected level of LRG1 in the biological sample relative to the control value is at least a 30% elevation.

12. The method of claim 1 wherein the level of LRG1 is measured using an immunoreagent to detect the level of LRG1.

13. The method of claim 1, further comprising detecting said level of LRG1 in said biological sample with a test device comprising: a housing a test strip contained within the housing, the test strip comprising one or more immunoreagents, wherein one of the one or more immunoreagents detects the level of LRG1.

14. The method of claim 12, wherein the immunoreagent that measures the level of LRG1 is an anti-LRG1 antibody.

15. The method of claim 14, the test device further comprising means for quantifying binding of the anti-LRG1 antibody to LRG1 in the biological sample.

16. (canceled)

17. An antibody to LRG1 for use in an in vitro method for assessment of the risk of pre-eclampsia or another hypertensive disorder in a pregnant woman.

18. (canceled)

19. The method of claim 14, wherein the antibody binds to epitope regions of LRG1 protein selected from GLKALGHLSGNRLRKL (SEQ ID NO: 1) and AGPEAVKGQTLLAVAKSQ (SEQ ID NO:2).

20. The antibody of claim 17, wherein the antibody binds to epitope regions of LRG1 protein selected from GLKALGHLSGNRLRKL (SEQ ID NO:1) and AGPEAVKGQTLLAVAKSQ (SEQ ID NO:2).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0047] The invention is exemplified in the following non limiting studies and figures.

[0048] In the figures:

[0049] FIG. 1 shows levels of anti-angiogenic (sFlt1) and angiogenic (PIGF) markers prior to preeclampsia diagnosis. These biomarkers and their ratio are effective in predicting preeclampsia late in pregnancy

[0050] FIG. 2 shows levels of LRG1 prior to preeclampsia diagnosis

[0051] FIG. 3 illustrates ROC curves for predictive models of Preeclampsia at each study visit

[0052] FIG. 4 illustrates ROC curves for predictive models of Preeclampsia, after adjustment of covariates, at each study visit

[0053] FIG. 5 shows levels of BNP, a cardiac biomarker, prior to preeclampsia diagnosis

[0054] FIG. 6 shows box plots of the levels of cardiac biomarkers BNP and troponin per study visit, prior to preeclampsia diagnosis

BACKGROUND STUDIES

[0055] The present inventors conducted a prospective study, choosing to study women with and without Type 1 diabetes (‘Markers and Mechanisms for Pre-Eclampsia in Diabetes’ or MAMPED; PI: Lyons). Diabetic women are of interest because of their high PE case yield: a major advantage in a prospective study Women were studied at 12, 22, and 32 weeks’ gestation. Clinical as well as biochemical data were collected. We assessed some known candidate risk factors. As expected from prior studies of non-diabetic women, anti-angiogenic factors including sFlt and Endoglin, and angiogenic PIGF were predictive of PE, but only after 28 weeks gestation, i.e. soon before clinical onset of PE. For the reasons outlined above, we were particularly interested to identify first- or second-trimester characteristics associated with the later onset of PE. Several factors have emerged that are related to maternal dyslipidaemia and insulin resistance. We assessed markers of heart failure/cardiac dysfunction (BNP and Troponin) but found these to be unrelated to subsequent PE.

[0056] The prevalence of PE is increased 3- to 6-fold by maternal Type 1 diabetes mellitus (T1DM), even in women with no pre-gestational microalbuminuria or hypertension (Yu et al., 2009; Powers et al., PLoS One, 2010; Holmes et al., 2013). This high prevalence of PE in T1DM women brings an opportunity in that it facilitates human prospective studies with a high case-yield, and thus a manageable time-frame. We designed and conducted such a study (‘Markers and Mechanisms for Pre-Eclampsia in Diabetes’ or MAMPED; PI: Lyons) to seek new knowledge of biomarkers and mechanisms for PE in general, by capitalizing on the high case yield from women with diabetes. This rationale is supported by the congruence of findings between diabetic and non-diabetic women for antiangiogenic predictive factors, as described above. For reference values, MAMPED included a group of healthy, non-diabetic women who had normal, uncomplicated pregnancies.

Markers and Mechanisms of Pre-Eclampsia in T1DM (MAMPED; Pl Lyons)

[0057] MAMPED enrolled 151 pregnant women with documented uncomplicated pre-gestational T1DM (free of hypertension or microalbuminuria) and 24 non-diabetic pregnant women, recruited from six medical centers in three countries (Norway, Australia, USA). Of the T1DM subjects, 20% developed PE, vs ~4% (i.e. one) of the non-diabetic subjects, consistent with the high risk conferred by diabetes. Complete clinical data and specimens (plasma, serum and urine) were collected at three study visits (~11-13, 20-22, and 30-32 wks of gestation) and at term (~38 wks). The non-diabetic control group was included to provide reference values, enabling comparisons of data between T1DM patients and the low-risk non-diabetic group. MAMPED has resulted in many high-impact publications (Yu et al., 2009; Azar et al., 2011; Basu et al., 2012; Yu et al., 2012; Du et al. 2013), including the first to define predictive roles of angiogenic/anti-angiogenic factors in early third-trimester diabetic pregnancy (Yu et al., 2009). The present inventors also found some first and second trimester parameters that were significantly associated with subsequent PE, but not to an extent that would be useful clinically. For example, we observed lower plasma levels of soluble receptor for advanced glycation end-products (sRAGE), and its ratio to advanced glycation endproducts (AGEs) at 12 wks in T1DM women who subsequently developed PE, possibly reflecting a higher burden of AGEs in diabetes and resultant compromise of scavenger capacity (Yu et al., 2012). In a detailed analysis of plasma lipoproteins, measuring conventional lipids, apolipoproteins, and subclasses defined by size (NMR) and apolipoprotein content (Basu et al. 2012), we found that in women with vs. without subsequent PE, low-density lipoprotein (LDL-C), particle concentrations of total LDL and large LDL, Apolipoprotein B (ApoB), and ApoB:Apolipoprotein Al (ApoAl) ratio were all increased in the 1.sup.st and 2.sup.nd trimesters, while peripheral lipoprotein lipolysis was decreased. This suggested that increased levels of certain cholesterol-rich particles and impaired peripheral lipolysis early in pregnancy are associated with subsequent PE. MAMPED samples were similarly used to test the utility of a novel protein, leucine-rich alpha 2 glycoprotein 1 (LRG1). This protein has never been studied during pregnancy as a predictor of PE.

Leucine-rich Alpha 2 Glycoprotein 1 (LRG1)

Background

[0058] Leucine-rich alpha 2 glycoprotein 1 (LRG1) was identified in 1977 as a trace component of human serum, and resolution of the primary structure in 1985 indicated that it exists as a single polypeptide chain of MW approximately 45 kDa (Takahashi et al., 1985, Haupt and Baudner, 1977). The presence of a leucine at every seventh position in segments of this protein suggests the possibility of forming a leucine-zipper structure, which has been implicated in protein-DNA and protein-protein interactions (Takahashi et al., 1985). Although the precise function of LRG1 has yet to be fully elucidated, evidence to date suggests that it is associated with inflammatory responses and neutrophilic differentiation, implicated in cell migration, and linked to cellular responses to the pro-fibrotic cytokine transforming growth factor beta (TGFβ) (Zhong et al., 2015, Takemoto et al., 2015, Ha et al., 2014, Lynch et al., 2013, Serada et al., 2012, O’Donnell et al., 2002, Codina et al., 2010) . In studies examining LRG1 in the context of established disease, the directional change of LRG1 levels varies in comparison to associated experimental controls. For example, it has been shown that in patients with hepatitis C, LRG1 expression in patients decrease with increasing severity of fibrotic change in diseased livers, and in a separate study involving patients with allergic airway disease, circulating LRG1 levels and TGFβR2 decrease compared to controls (Hao et al., 2016, Zhang et al., 2015b). However, in other pathological settings, circulating levels increase, including in sepsis, neurodegenerative disease, heart failure, and cancer (Hashida et al., 2016, Cavalcante Mde et al., 2016, Zhang et al., 2015a, Furukawa et al., 2015, Miyajima et al., 2013, Watson et al., 2011). Thus it is difficult to predict how expression levels of LRG1 protein might change in particular disease states, and whether or not it plays a causal role in various disease settings. In addition, all published studies have investigated LRG1 in established disease, and thus little is known about the predictive power of LRG1 as a biomarker for future disease in otherwise disease-free patients.

[0059] The present inventors originally set out to study LRG1 in the context of diabetes and its complications. As part of this, we explored LRG1 levels in the MAMPED pregnancy cohort of type 1 diabetic women: these women had no evidence of prior renal or hypertensive problems, as described above. This was the first-ever study to investigate LRG1 during pregnancy. Within this well-characterised cohort of pregnant women, we found little change in LRG1 level as pregnancy progressed, and we found no association of LRG1 with diabetes per se. However to our surprise, from the earliest time point (12 weeks’gestation), LRG1 levels were associated with future hypertensive complications of pregnancy. Thus, LRG1 predicted the (much) later development of preeclampsia, well before the onset of high blood pressure and clinical disease, and over a time-frame distinctly different from established angiogenic biomarkers. In women with apparently normal pregnancies who developed preeclampsia with onset after 33 weeks gestation, LRG1 levels were significantly elevated as early as 12 weeks gestation, and remained elevated at 22 and 32 weeks (FIG. 2). LRG1 thus acts as a biomarker for the future development of PE much earlier than the gold-standard PE diagnostic biomarkers, sFlt-1 and PIGF (FIG. 1). The discovery of LRG1 as an early biomarker for the future development of PE was only possible because of the longitudinal nature of the MAMPED cohort, facilitating analyses throughout the duration of pregnancy.

[0060] Various statistical models highlighting the prognostic strengths of LRG1 and how it out performs the gold standard PE biomarkers sFlt-1 and PIGF are shown in FIGS. 3 & 4 and in Tables 1-4 below. Collectively, these unique findings provide strong evidence that LRG1 could have clinical utility in the prediction, detection, and monitoring of hypertensive complications of pregnancy. It would also have a novel application in longitudinal screening in disease-free pregnant individuals as a risk predictor for preeclampsia and other hypertensive complications of pregnancy.

LRG1 as a Biomarker for PE

[0061] The following results from the MAMPED study show that LRG1 has the capacity to serve as a useful biomarker for preeclampsia from as early as 12 weeks gestation.

[0062] FIG. 1 shows levels of anti-angiogenic (sFlt1) and angiogenic (PIGF) markers prior to preeclampsia diagnosis. These biomarkers and their ratio are effective in predicting preeclampsia late in pregnancy (MAMPED study, Yu et al. 2009)

[0063] More specifically FIG. 1 shows longitudinal changes of a sFlt, and b PIGF during gestation. Values (means±SEM) were plotted against the average gestational age at each visit. All data are prior to the onset of PE. For illustrative purposes, three groups of participants are shown. Blue symbols, non-diabetic normotensive healthy controls; green symbols, Type 1 diabetic who remained normotensive; red symbols, Type 1 diabetic who subsequently developed pre-eclampsia. The primary analysis compared diabetic pre-eclampsia and diabetic normotensive groups (*p<0.05). Secondary analyses compared diabetic normotensive vs non-diabetic normotensive groups (ns).

[0064] For sFlt1, in non-diabetic women, serum sFlt1 (FIG. 1a) was stable from visit 1 to visit 3. Among type 1 diabetic women, there was a pronounced increase (doubling) in sFlt1 between the second and third study visits in those who later developed preeclampsia, whereas in those remained normotensive throughout pregnancy, there was little change.

[0065] For PIGF, serum PIGF (FIG. 1b) increased progressively in non-diabetic normotensive women from visit 1 to visit 3. In type 1 diabetic women the increase between the second and third study visits was blunted, most markedly in women who subsequently developed pre-eclampsia.

[0066] FIG. 2 shows unpublished levels of LRG1 prior to preeclampsia diagnosis (MAMPED study, unpublished).

[0067] Longitudinal changes of LRG1 during gestation are plotted. Values (means ± SEM) were plotted against the average gestational age. All study visits took place prior to the onset of PE. Values significantly different between participant groups (p<0.05) are indicated: the primary analysis compared diabetic pre-eclampsia and diabetic normotensive groups (*p<0.05, **p<0.01, ***p<0.001). There were no significant differences in secondary analyses comparing between diabetic normotensive vs non-diabetic normotensive groups at any study visit.

[0068] Women with type 1 diabetes who later developed preeclampsia had significantly higher LRG1 than diabetic normotensive women from the first study visit, and this persisted at the second and third study visits. No significant difference was noted between the diabetic women who remained normotensive and the non-diabetic controls at any study visit during pregnancy. Longitudinal analyses revealed significant increase in LRG1 between visit 1 and visit 2 (p<0.05), and a significant decrease between visit 2 and visit 3 (<0.05) for the diabetic women with preeclampsia. There were no longitudinal changes for the women with diabetes who remained normotensive. LRG1 significantly decreased between the first and second study visits in non-diabetic women (<0.05), but levels remained similar between the second and third visits.

[0069] FIG. 3 shows ROC curves for predictive models of preeclampsia at each study visit. The dash line: reference line indicating AUC = 0.5. PIGF and sFlt-⅟PIGF are logarithmically transformed to the base 10

[0070] FIG. 4 shows ROC curves for predictive models of preeclampsia, after adjustment of covariates, at each study visit. The dash line: reference line indicating AUC = 0.5. PIGF and sFlt-⅟PIGF are logarithmically transformed to the base 10. Adjusted for established risk factors (clinical covariates) BMI and Mean arterial pressure (MAP)

[0071] FIG. 5 shows levels of BNP, a cardiac biomarker, prior to preeclampsia diagnosis.

[0072] FIG. 5 plots longitudinal changes of BNP over time. Values (geometric mean ± 95% Cl of geometric mean) were plotted against the average gestational age at each visit prior to the onset of PE. See FIG. 1 for keys for both primary and secondary analyses

[0073] BNP was not normally distributed. For analysis, data was logarithmically transformed. There were no significant differences in BNP levels between any of the groups at any time point, and therefore no predictive effect.

[0074] FIG. 6 shows box plots of the levels of cardiac biomarkers BNP and troponin per study visit, prior to preeclampsia diagnosis. The lower limit of detection of Troponin is 3 ng/L. Any sample not detected was denoted as 0.5 ng/L for observational purposes. There were no significant differences in BNP or troponin levels between any of the groups at any time point

[0075] Table 1 below shows the Unadjusted/Adjusted odds ratio, areas under the ROC curve, IDI indices, and NRI indices for predictive models of preeclampsia in women with type 1 diabetes at different gestational ages (MAMPED study, unpublished)

TABLE-US-00001 LRG1 (mcg/ml) Unadjusted odds ratio (95% Cl) Adjusted odds ratio (95% Cl)* Area under the ROC curve without covariates [AUC (†p value)] Area under the ROC curve with covariates* [AUC (p‡)] IDI (¶) NRI (§) ~12 weeks gestation 1.08 (1.01-1.16) 1.08 (1.00-1.16) 0.707 (0.019) 0.769 (0.201) 0.097 (0.035) 0.455 (0.118) ~22 weeks gestation 1.13 (1.04-1.23) 1.20 (1.05-1.38) 0.824 (<0.001) 0.919 (0.08) 0.0252 (<0.001) 1.01 (<0.001) ~32 weeks gestation 1.06 (1.00-1.13) 1.10 (1.01-1.20) 0.638 (0.108) 0.900 (0.180) 0.122 (0.014) 0.848 (0.001) ROC: receiver operating characteristic; IDI: integrated discrimination improvement; NRI: net reclassification improvement. *Adjusted for established risk factors (‘covariates’): Body Mass Index (BMI) and Mean arterial pressure (MAP). to AUC of 0.5, for a logistic model containing only LRG1 at current visit. ‡ Relative to AUC of 0.676, 0.781, and 0.843 for a logistic model containing covariates only per visit Addition of LRG1 resulted in statistically significant improvement in predicted risk of preeclampsia at all three study visits. § Addition of LRG1 resulted in statistically significant improvement in the reclassification of preeclampsia risk at the second and third study visits.

[0076] The unadjusted odds ratios (i.e. assessing LRG1 by itself) showed that for every unit increase in LRG1, the risk of developing PE increased by 8%, 13%, and 6% at study visits 1, 2, and 3 respectively. In order to derive the best model, the LRG1 data were corrected for two simple maternal factors, BMI and MAP. The adjusted odds ratios indicate that, keeping all other covariates equal, for every unit increase in LRG1, the risk of developing PE increased by 8%, 20%, and 10% at study visits 1, 2, and 3 respectively. LRG1 remains significantly associated with preeclampsia after controlling for established risk factors. The addition of LRG1 to a model of clinical maternal risk factors improved the predictive value at all three study visits. The IDI and NRI statistic show that the

TABLE-US-00002 Variable(s) in model Visit 1 Visit 2 Visit 3 Area under the ROC curve P value* Area under the ROC curve p value* Area under the ROC curve p value* sFlt only 0.556 0.528 0.541 0.644 0.702 0.020 PIGF only 0.561 0.496 0.496 0.961 0.692 0.027 sFlt/PIGF only 0.511 0.903 0.530 0.734 0.728 0.009 LRG1 only 0.703 0.022 0.820 <0.001 0.629 0.139 LRG1 + sFLt 0.703 0.022 0.827 <0.001 0.821 <0.001 LRG1 + PIGF 0.721 0.013 0.818 <0.001 0.734 0.007 LRG1 + sFlt/PIGF 0.703 0.022 0.825 <0.001 0.784 0.001

addition of LRG1 significantly improved the discrimination between women who developed PE and those who remained normotensive (at all three visits), and the correct reclassification of women who developed PE and those who remained normotensive.

[0077] Table 2 (see also FIG. 3) - shows the areas under the ROC curve models including angiogenic/anti-angiogenic factors and/or LRG1 (MAMPED study, unpublished)

[0078] *p value relative to area under the ROC curve of 0.5 PIGF and sFlt-⅟PIGF are logarithmically transformed to the base 10

[0079] In general, the ROC curves show that LRG1 alone is a better predictor of PE than the angiogenic/anti-angiogenic factors, or their ratio, at the first two study visits. The area under the ROC curve of a model including the angiogenic/anti-angiogenic factors improves after the addition of LRG1 at all study visits.

[0080] Table 3 (see also FIG. 4) - Areas under the ROC curve models including angiogenic/anti-angiogenic factors and/or LRG1 after adjustment of covariates (MAMPED study, unpublished).

[0081] A) Comparing to a model of covariates (BMI, mean arterial pressure) only, per visit

TABLE-US-00003 Variables in model Visit 1 Visit 2 Visit 3 Area under the ROC curve p value Area under the ROC curve p value Area under the ROC curve p value Covariates only 0.680 0.773 0.873 sFlt-1 + covariates 0.682 0.777 0.790 0.472 0.899 0.361 PIGF + covariates 0.714 0.345 0.777 0.480 0.877 0.853 sFlt-⅟PIGF + covariates 0.699 0.524 0.777 0.762 0.877 0.879 LRG1 + covariates 0.768 0.201 0.916 0.078 0.905 0.429 LRG1 + sFlt-1 + covariates 0.771 0.218 0.916 0.046 0.946 0.105 LRG1 + PIGF + covariates 0.768 0.224 0.909 0.082 0.905 0.411 LRG1 + sFlt-⅟PIGF + covariates 0.775 0.202 0.907 0.070 0.911 0.371 *p value relative to area under the ROC curve of model of covariates (BMI and MAP) only, per visit PIGF and sFlt-⅟PIGF are logarithmically transformed to the base 10

[0082] B) Comparing to a model of equal chance (i.e. area under ROC curve = 0.5)

TABLE-US-00004 Variables in model Visit 1 Visit 2 Visit 3 Area under the ROC curve p value Area under the ROC curve p value Area under the ROC curve p value Covariates only 0.680 0.044 0.773 0.002 0.873 <0.001 sFlt-1 + covariates 0.682 0.041 0.790 0.001 0.899 <0.001 PIGF + covariates 0.714 0.016 0.777 0.002 0.877 <0.001 sFlt-⅟PIGF + covariates 0.699 0.025 0.777 0.002 0.877 <0.001 LRG1 + covariates 0.768 0.003 0.916 <0.001 0.905 <0.001 LRG1 + sFlt-1 + covariates 0.771 0.002 0.916 <0.001 0.946 <0.001 LRG1 + PIGF + covariates 0.768 0.003 0.909 <0.001 0.905 <0.001 LRG1 + sFlt-⅟PIGF + covariates 0.775 0.002 0.907 <0.001 0.911 <0.001

[0083] In general, the ROC curves show that addition of LRG1 to a model containing established risk factors, is better at predicting PE than the addition of the angiogenic/anti-angiogenic factors, or the ratio, at all study visits. After the adjustment of covariates, the area under the ROC curves of models including the angiogenic/anti-angiogenic factors improve after the addition of LRG1, at all study visits. All models are better than a model of equal chance.

TABLE-US-00005 Variable in model Visit 1 Area under the ROC curve Equal chance (classification cutoff 0.5) Setting specificity -90% Setting specificity~95% Sensitivity Specificity Sensitivity Specificity Sensitivity Specificity Covariates only 0.680 63.64 77.27 50.00 90.91 18.13 96.46 sFlt-1 + covariates 0.682 66.67 77.27 52.38 90.91 10.08 96.46 PIGF + covariates 0.714 61.90 68.18 52.38 90.91 36.10 96.46 sFlt-1 /PIGF + covariates 0.699 61.90 68.18 52.38 90.91 28.57 96.46 LRG1 + covariates 0.768 68.18 72.73 45.45 90.91 27.27 96.46 LRG1 + sFlt-1 + covariates 0.771 66.67 77.27 38.10 90.91 28.67 96.46 LRG1 + PIGF + covariates 0.768 66.67 68.18 42.36 90.91 28.67 95.45 LRG1.sup.-+ sFlt-1/PIGF + covariates 0.775 61.90 68.18 42.36 90.91 23.81 95.45

TABLE-US-00006 Variable in model Visit 2 Area under the ROC curve Equal chance (classification cutoff0.5) Setting specificity -90% Setting specificity~95% Sensitivity Specificity Sensitivity Specificity Sensitivity Specificity Covariates only 0.773 73.91 80.95 43.48 90.48 13.04 96.24 sFlt-1 + covariates 0.790 77.27 66.67 60.00 90.48 22.73 95.24 PIGF + covariates 0.777 72.73 80.95 40.91 90.48 9.09 95.24 sFlt-1 /PIGF + covariates 0.777 72.73 71.43 40.91 90.48 13.64 95.24 LRG1+ covariates 0.919 86.96 76.19 88.96 90.48 88.98 95.24 LRG1 +sFlt-1 + covariates 0.916 81.82 85.71 81.32 90.48 72.73 95.24 LRG1 + PIGF + covariates 0.909 86.36 76.19 77.27 90.48 77.27 95.24 LRG1 +sFlt-1 /PIGF + covariates 0.907 77.27 90.48 77.27 90.48 88.08 95.24

TABLE-US-00007 Variable in model Visit 3 Area under the ROC curve Equal chance (classification cutoff 0.5) Setting specificity -90% Setting specificity~95% Sensitivity Specificity Sensitivity Specificity Sensitivity Specificity Covariates only 0.873 71.43 87.50 42.86 91.67 33.33 96.83 sFlt-1 + covariates 0.899 71.43 91.67 71.43 91.87 47.62 96.83 PIGF + covariates 0.877 71.43 87.50 57.14 91.87 33.33 96.83 sFlt-1 /PIGF + covariates 0.877 71.43 87.50 71.43 91.87 33.33 96.83 LRG1 + covariates 0.905 80.95 87.50 71.43 91.87 52.23 96.83 LRG1 + sFlt-1 + covariates 0.946 85.71 91.67 86.71 91.87 76.19 96.83 LRG1 + PIGF + covariates 0.905 80.95 87.50 57.14 91.67 57.14 96.83 LRG1 +sFlt-⅟PIGF + covariates 0.911 85.71 87.50 66.67 91.67 52.38 96.83 *covariates are BMI and MAP, per visit! PIGF and sFlt-⅟PIGF are logarithmically transformed to the base 10

[0084] Table 4 shows Model fits were compared for different predictive algorithms, using the established factors for predicting PE (anti-angiogenic factors:sFlt, PIGF and sFlt/PIGF), before and after the addition of LRG1 by plotting receiver operating characteristic (ROC) diagrams and assessing the area under the curve (ROC-AUC). Initially the sensitivity and specificity of the models were calculated at a cut-off probability of 50%. Next we fixed the specificity to 90% or 95% (i.e. 1 in 10 or 1 in 20 tests is a false positive). At visit 1, although the ROC-AUC increased after the addition of LRG1 to all models, there was no significant improvement in the sensitivities and specificities. At visit 2, the model for predicting PE using LRG1 and covariates (BMI and MAP) had the highest AUC of 0.919, with a sensitivity of 82% and specificity of 86%, at a cut-off probability of 50%. When we fixed the specificity to either 90% or 95%, the specificity was 87%. This model performed better than other models using the anti-angiogenic markers with/without the addition of LRG1. At visit 3, the ROC-AUC tended to increase and the sensitivities improve, after the addition of LRG1 to any of the models using anti-angiogenic markers. Overall, LRG1 is a better predictor of PE compared to established anti-angiogenic factors at visit 2 and at visit 3, the addition of LRG1 to a model with anti-angiogenic factors, may improve the prediction of PE

[0085] Although certain embodiments have been illustrated and described herein, it will be appreciated by those of ordinary skill in the art that a wide variety of alternate and/or equivalent embodiments or implementations calculated to achieve the same purposes may be substituted for the embodiments shown and described without departing from the scope. Those with skill in the art will readily appreciate that embodiments may be implemented in a very wide variety of ways. This application is intended to cover any adaptations or variations of the embodiments discussed herein. Therefore, it is manifestly intended that embodiments be limited only by the claims and the equivalents thereof.

[0086] The invention is not limited to the embodiments described herein but can be amended or modified without departing from the scope of the present invention.

REFERENCES

[0087] Askie LM, Duley L, Henderson-Smart DJ, Stewart LA: Antiplatelet agents for prevention of pre-eclampsia: a meta-analysis of individual patient data. Lancet 2007;369:1791-1798

[0088] American College of Obstetricians & Gynecologists, Task Force on Hypertension in Preeclampsia: Hypertension in pregnancy. Report of the American College of Obstetricians and Gynecologists’ Task Force on Hypertension in Pregnancy. Obstet Gynecol 2013;122:1122-1131

[0089] Brown MA, Lindheimer MD, de Swiet M, Van Assche A, Moutquin JM: The classification and diagnosis of the hypertensive disorders of pregnancy: statement from the International Society for the Study of Hypertension in Pregnancy (ISSHP). Hypertens Pregnancy 2001;20:lX-XlV

[0090] Duley L, Henderson-Smart DJ, Meher S, King JF: Antiplatelet agents for preventing pre-eclampsia and its complications. Cochrane Database Syst Rev 2007:CD004659

[0091] Hakim J, Senterman MK, Hakim AM: Preeclampsia is a biomarker for vascular disease in both mother and child: the need for a medical alert system. International J. Pediatrics 2013;2013:953150

[0092] Henderson JT, Whitlock EP, O’Connor E, Senger CA, Thompson JH, Rowland MG: Low-dose aspirin for prevention of morbidity and mortality from preeclampsia: a systematic evidence review for the u.s. Preventive services task force. Ann Intern Med 2014;160:695-703

[0093] Roberge S, Giguere Y, Villa P, Nicolaides K, Vainio M, Forest JC, von Dadelszen P, Vaiman D, Tapp S, Bujold E: Early administration of low-dose aspirin for the prevention of severe and mild preeclampsia: a systematic review and meta-analysis. Am J Perinatol 2012;29:551-556

[0094] Akolekar R, Syngelaki A, Beta J, Kocylowski R, Nicolaides K. Maternal serum placental protein 13 at 11-13 weeks of gestation in preeclampsia Prenat Diagn 2009; 29: 1103-1108.

[0095] Audibert F, Boucoiran l, An N, Aleksandrov N, Delvin E, Bujold E, Rey E: Screening for preeclampsia using first-trimester serum markers and uterine artery Doppler in nulliparous women. Am J Obstet Gynecol 2010;203:383 e381-388

[0096] Bredaki FE, Mataliotakis M, Wright A, Wright D, Nicolaides KH. Maternal serum alpha-fetoprotein at 12, 22 and 32 weeks’ gestation in screening for pre-eclampsia. Ultrasound Obstet Gynecol 2016; 47: 466-471

[0097] O’Gorman N; Wright D; Syngelaki A; Akolekar R; Wright D; Poon LC; Nicolaides KH. Competing risks model in screening for preeclampsia by maternal factors and biomarkers at 11-13 weeks gestation. Am J Obstet Gynecol 2016; 214:103.e1-12.

[0098] Verlohren S, Herraiz l, Lapaire O, Schlembach D, Zeisler H, Calda P, Sabria J, Markfeld-Erol F, Galindo A, Schoofs K, Denk B, Stepan H: New gestational phase-specific cutoff values for the use of the soluble fins-like tyrosine kinase-⅟placental growth factor ratio as a diagnostic test for preeclampsia. Hypertension 2014;63:346-352

[0099] Wright D, Syngelaki A, Akolekar R, Poon LC, Nicolaides KH. Competing risks model in screening for preeclampsia by maternal characteristics and medical history. Am J Obstet Gynecol 2015;213:62.e1-10.

[0100] Azar M, Basu A, Jenkins AJ, Nankervis AJ, Hanssen KF, Scholz H, Henriksen T, Garg SK, Hammad SM, Scardo JA, Aston CE, Lyons TJ: Serum carotenoids and fat-soluble vitamins in women with type 1 diabetes and preeclampsia: a longitudinal study. Diabetes Care 2011;34:1258-1264

[0101] Basu A, Alaupovic P, Wu M, Jenkins AJ, Yu Y, Nankervis AJ, Hanssen KF, Scholz H, Henriksen T, Lorentzen B, Clausen T, Garg SK, Menard MK, Hammad SM, Scardo JA, Stanley JR, Dashti A, Aston CE, Lyons TJ: Plasma lipoproteins and preeclampsia in women with type 1 diabetes: a prospective study. J Clin Endocrinol Metab 2012;97:1752-1762

[0102] Basu A, Yu JY, Jenkins AJ, Nankervis AJ, Hanssen KF, Henriksen T, Lorentzen B, Garg SK, Menard MK, Hammad SM, Scardo JA, Aston CE, Lyons TJ: Trace elements as predictors of preeclampsia in type 1 diabetic pregnancy. Nutrition Research 2015;35:421-430

[0103] Du M, Basu A, Fu D, Wu M, Centola M, Jenkins AJ, Hanssen KF, Garg SK, Hammad SM, Scardo JA, Aston CE, Lyons TJ: Serum inflammatory markers and preeclampsia in type 1 diabetes: a prospective study. Diabetes Care 2013, 36:2054-2061

[0104] Holmes VA, Young IS, Patterson CC, Maresh MJ, Pearson DW, Walker JD, McCance DR: The role of angiogenic and antiangiogenic factors in the second trimester in the prediction of preeclampsia in pregnant women with type 1 diabetes. Diabetes Care 2010;36:3671-3677

[0105] Powers RW, Jeyabalan A, Clifton RG, Van Dorsten P, Hauth JC, Klebanoff MA, Lindheimer MD, Sibai B, Landon M, Miodovnik M: Soluble fms-Like tyrosine kinase 1 (sFlt1), endoglin and placental growth factor (PIGF) in preeclampsia among high risk pregnancies. PLoS One 2010;5:e13263

[0106] Yu Y, Hanssen KF, Kalyanaraman V, Chirindel A, Jenkins AJ, Nankervis AJ, Torjesen PA, Scholz H, Henriksen T, Lorentzen B, Garg SK, Menard MK, Hammad SM, Scardo JA, Stanley JR, Wu M, Basu A, Aston CE, Lyons TJ: Reduced soluble receptor for advanced glycation end-products (sRAGE) scavenger capacity precedes pre-eclampsia in Type 1 diabetes. Br J Obs Gyn 2012;119:1512-1520

[0107] Yu Y, Jenkins AJ, Nankervis AJ, Hanssen KF, Scholz H, Henriksen T, Lorentzen B, Clausen T, Garg SK, Menard MK, Hammad SM, Scardo JC, Stanley JR, Dashti A, May K, Lu K, Aston CE, Wang JJ, Zhang SX, Ma JX, Lyons TJ: Anti-angiogenic factors and pre-eclampsia in type 1 diabetic women. Diabetologia 2009;52: 160-168.

[0108] CAVALCANTE MDE, S., TORRES-ROMERO, J. C., LOBO, M. D., MORENO, F. B., BEZERRA, L. P., LIMA, D. S., MATOS, J. C., MOREIRA RDE, A. & MONTEIRO-MOREIRA, A. C. 2016. A panel of glycoproteins as candidate biomarkers for early diagnosis and treatment evaluation of B-cell acute lymphoblastic leukemia. Biomark Res, 4, 1.

[0109] CODINA, R., VANASSE, A., KELEKAR, A., VEZYS, V. & JEMMERSON, R. 2010. Cytochrome c-induced lymphocyte death from the outside in: inhibition by serum leucine-rich alpha-2-glycoprotein-1. Apoptosis, 15, 139-52.

[0110] FURUKAWA, K., KAWAMOTO, K., EGUCHI, H., TANEMURA, M., TANIDA, T., TOMIMARU, Y., AKITA, H., HAMA, N., WADA, H., KOBAYASHI, S., NONAKA, Y., TAKAMATSU, S., SHINZAKI, S., KUMADA, T., SATOMURA, S., ITO, T., SERADA, S., NAKA, T., MORI, M., DOKI, Y., MIYOSHI, E. & NAGANO, H. 2015. Clinicopathological Significance of Leucine-Rich alpha2-Glycoprotein-1 in Sera of Patients With Pancreatic Cancer. Pancreas, 44, 93-8.

[0111] HA, Y. J., KANG, E. J., LEE, S. W., LEE, S. K, PARK, Y. B., SONG, J. S. & CHOI, S. T. 2014. Usefulness of serum leucine-rich alpha-2 glycoprotein as a disease activity biomarker in patients with rheumatoid arthritis. J Korean Med Sci, 29, 1199-204.

[0112] HAO, L., XIE, H., ZHANG, B., CHEN, D., WANG, S., ZHANG, H. & HE, S. 2016. LRG1 downregulation in allergic airway disorders and its expression in peripheral blood and tissue cells. J Transl Med, 14, 202.

[0113] HASHIDA, T., NAKADA, T. A., SATOH, M., TOMITA, K, KAWAGUCHI, R., NOMURA, F. & ODA, S. 2016. Proteome analysis of hemofilter adsorbates to identify novel substances of sepsis: a pilot study. J Artif Organs.

[0114] HAUPT, H. & BAUDNER, S. 1977. [lsolation and characterization of an unknown, leucine-rich 3.1-S-alpha2-glycoprotein from human serum (author’s transl)j. Hoppe Seylers Z Physiol Chem, 358, 639-46.

[0115] LYNCH, J., MEEHAN, M. H., CREAN, J., COPELAND, J., STALLINGS, R. L. & BRAY, l. M. 2013. Metastasis suppressor microRNA-335 targets the formin family of actin nucleators. PLoS One, 8, e78428.

[0116] MIYAJIMA, M., NAKAJIMA, M., MOTOI, Y., MORIYA, M., SUGANO, H., OGINO, l., NAKAMURA, E., TADA, N., KUNICHIKA, M. & ARAI, H. 2013. Leucine-rich alpha2-glycoprotein is a novel biomarker of neurodegenerative disease in human cerebrospinal fluid and causes neurodegeneration in mouse cerebral cortex. PLoS One, 8, e74453.

[0117] O’DONNELL, L. C., DRUHAN, L. J. & AVALOS, B. R. 2002. Molecular characterization and expression analysis of leucine-rich alpha2-glycoprotein, a novel marker of granulocytic differentiation. J Leukoc Biol, 72, 478-85.

[0118] SERADA, S., FUJIMOTO, M., TERABE, F., IIJIMA, H., SHINZAKI, S., MATSUZAKI, S., OHKAWARA, T., NEZU, R., NAKAJIMA, S., KOBAYASHI, T., PLEVY, S. E., TAKEHARA, T. & NAKA, T. 2012. Serum leucine-rich alpha-2 glycoprotein is a disease activity biomarker in ulcerative colitis. Inflamm Bowel Dis, 18, 2169-79.

[0119] TAKAHASHI, N., TAKAHASHI, Y. & PUTNAM, F. W. 1985. Periodicity of leucine and tandem repetition of a 24-amino acid segment in the primary structure of leucine-rich alpha 2-glycoprotein of human serum. Proc Natl Acad Sci U S A, 82, 1906-10.

[0120] TAKEMOTO, N., SERADA, S., FUJIMOTO, M., HONDA, H., OHKAWARA, T., TAKAHASHI, T., NOMURA, S., INOHARA, H. & NAKA, T. 2015. Leucine-rich alpha-2-glycoprotein promotes TGFbeta1-mediated growth suppression in the Lewis lung carcinoma cell lines. Oncotarget, 6, 11009-22.

[0121] WATSON, C. J., LEDWIDGE, M. T., PHELAN, D., COLLIER, P., BYRNE, J. C., DUNN, M. J., MCDONALD, K. M. & BAUGH, J. A. 2011. Proteomic analysis of coronary sinus serum reveals leucine-rich alpha2-glycoprotein as a novel biomarker of ventricular dysfunction and heart failure. Circ Heart Fail, 4, 188-97.

[0122] ZHANG, H., CAO, J., Ll, L., LIU, Y., ZHAO, H., Ll, N., Ll, B., ZHANG, A., HUANG, H., CHEN, S., DONG, M., YU, L., ZHANG, J. & CHEN, L. 2015a. Identification of urine protein biomarkers with the potential for early detection of lung cancer. Sci Rep, 5, 11805.

[0123] ZHANG, Y., LUO, Q., WANG, N., HU, F., JIN, H., GE, T., WANG, C. & QIN, W. 2015b. LRG1 suppresses the migration and invasion of hepatocellular carcinoma cells. Med Oncol, 32, 146.

[0124] ZHONG, D., HE, G., ZHAO, S., Ll, J., LANG, Y., YE, W., Ll, Y., JIANG, C. & Ll, X. 2015. LRG1 modulates invasion and migration of glioma cell lines through TGF-beta signaling pathway. Acta Histochem, 117, 551-8.