Circulating neutrophil gelatinase-associated lipocalin and preeclampsia: a meta-analysis

Abstract

Background

Previous studies evaluating the association between circulating neutrophil gelatinase-associated lipocalin (NGAL) and the risk of preeclampsia (PE) showed inconsistent results. A systematic review and meta-analysis was performed to summarize the relationship between circulating NGAL and PE.

Methods

Studies comparing the circulating NGAL between pregnant women with PE and controls with no PE were found by searching Medline, Web of Science, Cochrane’s Library, and Embase. Pooling results was performed using a random-effects model incorporating heterogeneity.

Results

In the study, 1293 women with PE and 1773 healthy pregnant women were enrolled in 18 case-control studies, and the gestational age was matched between cases and controls. Pooled results showed that compared to controls, women with PE had a significantly higher blood level of NGAL (standardized mean difference [SMD]: 0.95, 95% confidence interval [CI]: 0.63–1.28, p < .001; I² = 92%). Subgroup analyses showed consistent results in studies of NGAL measured at the first (SMD: 0.47, 95% CI: 0.15–0.80, p = .004), the second (SMD: 0.87, 95% CI: 0.55–1.19, p < .001), and the third trimester (SMD: 1.06, 95% CI: 0.63–1.24, p < .001) of pregnancy. In addition, women with mild (SMD: 0.78, 95% CI: 0.13–1.44, p = .02) and severe PE (SMD: 1.19, 95% CI: 0.40–1.97, p = .003) both had higher circulating NGAL as compared to controls.

Conclusions

High circulating NGAL is associated with PE, which may be independent of the trimesters for blood sampling and the severity of PE.

Keywords:

• Neutrophil gelatinase-associated lipocalin
• circulating
• preeclampsia
• meta-analysis

Introduction

Preeclampsia (PE) is a severe pregnancy complication which is characterized by elevated blood pressure and proteinuria after 20 weeks of pregnancy [1,2]. According to previous studies, PE incidence in pregnant women varies between 3% and 8% [3]. Accumulating evidence has confirmed a series of adverse outcomes of women with PE, which involve kidney injury, neurological compromise, and eclamptic seizures for women during pregnancy, as well preterm birth and low birth weight in the offspring [4,5]. In addition, the chances of suffering an adverse cardiovascular event later in life are also higher for women who suffer from PE [6,7]. Moreover, as the most effective treatment for PE is delivery, PE could also lead to iatrogenic preterm births [8,9]. Accordingly, efforts are constantly being made to understand the key mechanisms underlying PE’s pathogenesis in order to develop potentially effective preventative and therapeutic measures [10].

Neutrophil gelatinase-associated lipocalin (NGAL) is a 25-kDa secreted adipokine which is also known as lipocalin-2 [11,12]. As a member of the lipocalin superfamily, NGAL was firstly identified in human neutrophils, and subsequently found to be expressed in many tissues such as the kidney, liver, colon, breast, and adipose tissue [13]. Functions as a regulator of energy metabolism, NGAL has been shown to be closely associated with endocrine and metabolic disorders such as diabetes, obesity, endocrine hypertension, dyslipidemia, and insulin resistance [12,14,15]. Besides, as a key mediator of systemic inflammation, NGAL has also been related to cardiovascular and renal injuries [16,17]. Interestingly, some previous studies have suggested that a high NGAL level in the blood may be a risk factor of PE [18–29], while other studies did not show similar results [30–35]. In view of the inconsistent results of previous studies evaluating the association between circulating NGAL and PE, an evidence-based quantitative meta-analysis was performed in this study to investigate the correlation between circulating NGAL levels and PE.

Methods

This study adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guideline [36,37] and the Cochrane Handbook for Systematic Reviews and Meta-analyses (Handbook for Systematic Reviews and Meta-analysis) [38].

Literature search

We obtained studies by searching Medline, Web of Science, and Cochrane’s Library, and Embase with a combined key word strategy: (1) “neutrophil gelatinase-associated lipocalin” OR “NGAL” OR “siderocalin” OR “lipocalin”; and (2) “pre-eclampsia” OR “preeclampsia” OR “eclampsia” OR “PIH” OR “toxemia” OR “pregnancy-induced hypertension” OR “EPH” OR “edema-proteinuria-hypertension gestos”. The included studies were only published in English or Chinese and involved human subjects. To complement this process, we hand-screened the citations of the related original and review articles. Literature searches were last conducted on October 21, 2022.

Selection of studies

The following criteria were used to include studies in the analysis: (1) full-length articles; (2) women with PE and healthy pregnant women without PE were included as cases and controls, who were matched for the gestational age (GA); (3) measured the circulating level (plasma or serum) of NGAL; and (4) blood NGAL in cases and controls were reported as means with standard deviations (SDs), or these values could be calculated. A PE diagnosis was made according to the criteria in the studies that were included. Preclinical studies, review articles, studies without PE cases, studies with no normal pregnant controls, or studies that did not report the circulating NGAL levels between cases and controls were excluded. Because circulating NGAL changes according to the GA of the pregnant women [39], studies of women with and without PE who were not matched for GA of blood sampling were excluded from the current meta-analysis. In addition, to minimize the potential influences of comorbidities on the circulating NGAL, studies with women of superimposed PE were also excluded, such as those including women with diabetes, essential hypertension, and kidney diseases. Studies with the largest sample sizes were included in the meta-analysis of studies with overlapped patients.

Data collection and study quality assessment

Two independent authors searched, collected, and assessed the data. Consensus between the two authors was indicated to resolve disagreements. The data collected were: (1) author, year, and location of the study; (2) participant characteristics, such as numbers of the cases and controls, as well as their mean ages; (3) diagnostic criteria for PE in each study; (4) source and characteristics of controls; (5) timing of blood sampling, and methods for measuring circulating NGAL; and (6) variables matched or controlled between cases and controls. An assessment of study quality was conducted according to the Newcastle-Ottawa Scale (NOS) [40]. Based on this scale, each study is rated on three broad criteria: selection of cases and controls, comparability between groups, and exposure measurement. The total score was 1–9, and a higher score indicates a better study quality.

Statistical methods

The difference of circulating NGAL between pregnant women with and without PE were presented as a standardized mean difference (SMD) and the corresponding 95% confidence interval (CI) [41]. An evaluation of the extent of heterogeneity between studies was conducted by performing Cochrane’s Q test and estimating the I² statistic as explained previously [41,42]. An I² >50% indicates heterogeneity. The results were pooled using a random-effects model that took into account possible between-study heterogeneity [38]. An influencing analysis was conducted by excluding one study at a time in order to assess the stability of the results. Moreover, subgroup analyses were conducted in order to examine whether the results were consistent according to the timing of blood sampling (the first, second, or third trimester of pregnancy) and the severity of PE. Generally, if a pregnant woman with

PE had at least one or more those risk factors such as headache, visual or cerebral disturbance, elevated liver enzymes, thrombocytopenia, dyspnea due to pulmonary edema, progressive renal failure, systolic blood pressure ≥160 mmHg and/or diastolic blood pressure ≥110 mmHg, those patients were classified as severe PE; otherwise, mild PE was considered [43]. In order to determine whether publication bias exists, funnel plots were constructed and visually examined for symmetry [44]. In addition, for the purpose of testing for publication bias, Egger’s regression analysis was conducted [44]. RevMan (Version 5.1; Cochrane Collaboration, Oxford, UK) and Stata were used for the statistical analysis. p values above .05 indicate statistical significance.

Results

Study retrieval

As shown in Figure 1, following a search of electronic databases, 383 articles were retrieved, while 337 remained after duplications were removed. Among the 337 titles and abstracts screened for the meta-analysis, 298 were excluded due to their noncompliance with the meta-analysis’s criteria. Twenty-one of the remaining 39 studies were subsequently excluded due to the reasons listed in Figure 1 after full texts were read by two authors independently. As a result, 18 observational studies were enrolled in the meta-analysis [18–35].

Figure 1. Diagram illustrating the process of searching databases and identifying studies.

Study characteristics

Table 1 provides a summary of the characteristics of the studies included in the meta-analysis. Overall, 18 case-control studies [18–35] which consisted of 1293 women with PE and 1773 healthy pregnant women were available for the subsequent meta-analysis. These studies were conducted in Italy, Germany, Turkey, China, Korea, India, Greece, Denmark, and Iraq between 2008 and 2022. The mean ages of the included women were 26–37 years. The diagnosis of PE was in accordance with the American College of Obstetricians and Gynecologists (ACOG) criteria in 16 studies [19–29,31–35], and with the International Society for the Study of Hypertension in Pregnancy (ISSHP) criteria in the other two studies [18,30]. The circulating NGAL levels of the included women were measured in the first trimester in five studies [19,22,24,25,30], in the second trimester in two studies [18,22], and in the third trimester of pregnancy in 13 studies [20–24,26–29,31–34]. In 14 of the included studies [19–29,31,33,34], the circulating NGAL was obtained by enzyme-linked immunosorbent assay; while other methods such as sandwich enzyme immunoassay [18,30,32] and Multiplex immunoassay were applied in the remaining studies [35]. For all of the included studies, GA for blood sampling was matched between women with and without PE; while other factors, such as maternal age, body mass index, and parity were also matched in some of the included studies. In general, the included studies received NOS of eight or nine stars, which indicate a good quality study (Table 2).

Table 1. Characteristics of the included studies.

Study	Country	Study design	Maternal age (years)	Diagnosis of PE	Source of control	No. of PE	No. of Control	Timing of sampling	Methods for plasma NGAL measuring	Variables matched or adjusted
D’anna et al. [18]	Italy	CC	28.7	ISSHP	Non-hypertensive pregnant women	48	96	At diagnosis (2nd trimester)	Sandwich enzyme immunoassay	Maternal age, nulliparous, BMI, and GA
D’anna et al. [30]	Italy	CC	28.3	ISSHP	Women with single pregnancy without hypertension or renal disease	37	74	Before diagnosis (1st trimester)	Sandwich enzyme immunoassay	Maternal age, nulliparous, BMI, and GA
Stepan et al. [31]	Germany	CC	30.7	ACOG	Healthy pregnant women	22	22	At diagnosis (3rd trimester)	ELISA	Maternal age, BMI, and GA
Arikan et al. [32]	Turkey	CC	28.7	ACOG	Healthy pregnant women	44	22	At diagnosis (3rd trimester)	Sandwich enzyme immunoassay	Maternal age, BMI, and GA
Youssef et al. [19]	Italy	CC	30.8	ACOG	Healthy pregnant women	13	515	Before diagnosis (1st trimester)	ELISA	Maternal age, BMI, and GA
Cao et al. [20]	China	CC	31.2	ACOG	Healthy pregnant women	64	25	At diagnosis (3rd trimester)	ELISA	Maternal age and GA
Xiao et al. [21]	China	CC	27.2	ACOG	Healthy pregnant women	25	25	At diagnosis (3rd trimester)	ELISA	GA
Kim et al. [33]	Korea	CC	33	ACOG	Healthy pregnant women	37	37	At diagnosis (3rd trimester)	ELISA	Maternal age, BMI, and GA
Sachan et al. [23]	India	CC	26.3	ACOG	Healthy pregnant women	97	31	At diagnosis (3rd trimester)	ELISA	Maternal age, gravida, BMI, and GA
Karampas et al. [22]	Greece	CC	33.5	ACOG	Healthy pregnant women	12	33	Before and at diagnosis (1st, 2nd, and 3rd trimester)	ELISA	Maternal age and GA
Simonazzi et al. [34]	Italy	CC	37	ACOG	Healthy pregnant women	22	18	At diagnosis (3rd trimester)	ELISA	Maternal age, nulliparous, BMI, and GA
Cui et al. [35]	China	CC	39.7	ACOG	Healthy pregnant women	110	62	At diagnosis (3rd trimester)	Multiplex immunoassay	Maternal age, nulliparous, BMI, and GA
Zhang et al. [24]	China	CC	27.1	ACOG	Healthy pregnant women	128	183	Before diagnosis (1st trimester)	ELISA	Age, BMI, and GA
Pihl et al. [25]	Denmark	CC	28.8	ACOG	Healthy pregnant women	268	449	Before diagnosis (1st trimester)	ELISA	Maternal age, BMI, and GA
Li et al. [27]	China	CC	30	ACOG	Healthy pregnant women	186	72	At diagnosis (3rd trimester)	ELISA	Maternal age and GA
Yalman et al. [28]	Turkey	CC	27.9	ACOG	Healthy pregnant women	52	25	At diagnosis (3rd trimester)	ELISA	Maternal age, parity, BMI, and GA
Ali et al. [26]	Iraq	CC	27.4	ACOG	Healthy pregnant women	88	44	At diagnosis (3rd trimester)	ELISA	Maternal age, parity, BMI, and GA
Parasher et al. [29]	India	CC	22.8	ACOG	Healthy pregnant women	40	40	At diagnosis (3rd trimester)	ELISA	Maternal age, parity, and GA

CC, case-control; PE, preeclampsia; NGAL, neutrophil gelatinase- associated lipocalin; ISSHP, the International Society for the Study of Hypertension in Pregnancy; ACOG, the American College of Obstetricians and Gynecologists; ELISA, enzyme-linked immunosorbent assay; BMI, body mass index; GA, gestational age;.

Table 2. Study quality evaluation via the Newcastle-Ottawa Scale.

Study	Adequate definition of the cases	Representativeness of the cases	Selection of Controls	Definition of Controls	Controlled for GA	Controlled for other factors	Ascertainment of the exposure	Same method of ascertainment of exposure for cases and controls	Non-response rate	Overall
D’anna et al. [18]	1	1	1	1	1	1	1	1	1	9
D’anna et al. [30]	1	1	1	1	1	1	1	1	1	9
Stepan et al. [31]	1	1	1	1	1	1	1	1	1	9
Arikan et al. [32]	1	1	1	1	1	1	1	1	1	9
Youssef et al. [19]	1	0	1	1	1	1	1	1	1	8
Cao et al. [20]	1	1	1	1	1	1	1	1	1	9
Xiao et al. [21]	1	1	1	1	1	0	1	1	1	8
Kim et al. [33]	1	1	1	1	1	1	1	1	1	9
Sachan et al. [23]	1	1	1	1	1	1	1	1	1	9
Karampas et al. [22]	1	1	1	1	1	1	1	1	1	9
Simonazzi et al. [34]	1	1	1	1	1	1	1	1	1	9
Cui et al. [35]	1	1	1	1	1	1	1	1	1	9
Zhang et al. [24]	1	1	1	1	1	1	1	1	1	9
Pihl [25]	1	1	1	1	1	1	1	1	1	9
Li [27]	1	1	1	1	1	1	1	1	1	9
Yalman [28]	1	1	1	1	1	1	1	1	1	9
Ali [26]	1	1	1	1	1	1	1	1	1	9
Parasher et al. [29]	1	1	1	1	1	1	1	1	1	9

Difference of circulating NGAL between women with and without PE

Since 9 studies included two comparisons of circulating NGAL between women with mild or severe PE versus control [20,23,26–28,32–35], these comparisons were included separately as independent datasets. Overall, 27 datasets from 18 studies were available for the meta-analysis [18–35]. Pooled results showed that compared to controls, women with PE had a significantly higher blood level of NGAL (SMD: 0.95, 95% CI: 0.63–1.28, p < .001; I² = 92%; Figure 2). A consistent result was obtained by excluding one study at a time in the influencing analysis (data not shown). Subgroup analyses showed consistent results in studies of NGAL measured at the first (SMD: 0.47, 95% CI: 0.15–0.80, p = .004), the second (SMD: 0.87, 95% CI: 0.55–1.19, p < .001), and the third trimester (SMD: 1.06, 95% CI: 0.63–1.24, p < .001; p for subgroup difference = 0.08; Figure 3) of pregnancy. In addition, subgroup analyses also suggested that compared to controls of women without PE, the circulating levels of NGAL were significantly higher in women with mild (SMD: 0.78, 95% CI: 0.13–1.44, p = .02) and severe PE (SMD: 1.19, 95% CI: 0.40–1.97, p = .003; p for subgroup difference = 0.44; Figure 4).

Figure 2. Forest plots comparing the circulating level of NGAL between women with and without PE.

Figure 3. Forest plots for the subgroup analyses according to the timing of blood sampling.

Figure 4. Forest plots for the subgroup analyses for comparisons of mild or severe PE versus control.

Publication bias

The funnel plots for the meta-analyses comparing circulating NGAL between women with and without PE are shown in Figure 5. There is a low risk of publication bias based on the symmetry of this plot. Moreover, Egger’s regression test did not reveal any significant publication bias (p = .29, respectively).

Figure 5. An analysis of the publication bias of the meta-analyses based on funnel plots.

Discussion

In this systematic review and meta-analysis, we pooled the results of 18 case-control studies, and the results showed that compared to controls of healthy women with pregnancy, women with PE had a significantly higher circulating level of NGAL. In addition, subsequent influencing analysis by excluding one study at a time showed the consistent result, reflecting the stability of the finding. Moreover, further subgroup analyses according to the timing of blood sampling showed that PE was associated with a higher circulating NGAL not only in the third trimester of the pregnancy but also in the first and second trimester of the pregnancy, which was before the diagnosis of PE. Finally, subgroup analysis according to the severity of PE showed that women with mild and severe PE both had higher circulating NGAL as compared to controls. Taking together, these findings indicate that high circulating NGAL is associated with PE, which may be independent of the trimesters for blood sampling and the severity of PE.

To the best of our knowledge, this is the first meta-analysis which evaluated the association between circulating NGAL and PE. We compressively searched four electronic databases and retrieved up-to-date literatures regarding the association between NGAL and PE. We only included studies with matched GA of blood sampling for cases and controls. This is important to minimize the confounding effect of GA on the outcomes of the meta-analysis because circulating NGAL has been shown to change dynamically during pregnancy [39]. Moreover, we exclude studies of women with comorbidities or superimposed PE, because comorbidities such as diabetes [45], primary hypertension [46], cardiovascular disease [47], or renal disease [16] may also affect the NGAL level and confound the results of the meta-analysis. In addition to the main meta-analysis which showed significant higher circulating NGAL in women with PE compared to healthy women with pregnancy, we performed several sensitivity and subgroup analyses, which also showed consistent results. We found the high circulating NGAL in PE not only in blood samples obtained in the third trimester at the diagnosis of PE, but also blood samples obtained in the first and second third trimester before the diagnosis of PE. These findings are important, which may suggest that a high NGAL in the first trimester may predict the incidence of PE. However, only five and two studies reported the circulating NGAL levels in women with and without PE in the first and second trimester respectively. The limited number of available studies in these subgroups may require the result to be validated in large-scale studies in the future. Besides, results of the meta-analysis also showed that circulating NGAL increased in both the women with mild and severe PE as compared to control, which may suggest that NGAL is a sensitive biomarker of PE, even for women with mild PE.

The potential mechanisms underlying the association between increased circulating NGAL and PE remain to be determined. Since the placenta plays a central role in the pathogenesis of PE, it could be hypothesized that NGAL may interact with the placenta, therefore, participate in the pathogenesis of PE [48]. An early study showed that increased apoptosis of trophoblasts at the maternal-fetal interface might result from an activated NGAL in the nitrofen model of congenital diaphragmatic hernia [49]. Interestingly, increased apoptosis of trophoblasts has been recognized as a key process involved in the pathogenesis and progression of PE [50]. In addition, via increasing the activity of matrix metalloproteinase-9, NGAL has been shown to enhance the invasion of extravillous trophoblasts in the placenta, thereby increasing the vulnerability to PE [51]. Studies are warranted in the future to determine the potential molecular pathways underlying the association between NGAL and PE.

There are also limitations to our study. In the first place, the included studies were all case-control studies. It is necessary to conduct large-scale prospective studies to determine if increased circulating NGAL is an independent risk factor of PE. Furthermore, as a meta-analysis of observational studies, this study was not able to confirm a causal relationship between PE incidence and changes of NGAL. Additionally, although some potential confounding factors were matched or controlled between women with and without PE, there may still be residual factors that affect the association between NGAL and PE, such as sodium intake [52] and concurrent medications [53]. Moreover, besides diabetes, essential hypertension, and kidney diseases, many other maternal diseases or conditions may affect NGAL, such as infection during pregnancy [54], maternal chorioamnionitis [55], or women with pregnancies affected by fetal growth disorders such as intrauterine growth restriction or small for gestational age [56], etc. We did not apply restrictions to these conditions during study selection, which may affect the findings of the meta-analysis. However, the retrospective reveal of the included studies found that none of them reported the inclusion women of the above conditions which may affect NGAL. Finally, to the best of our knowledge, no study has been performed to evaluate the changes of circulating NGAL in pregnant women with eclampsia. Studies are warranted in the future for further investigations.

Conclusion

In conclusion, results of the meta-analysis indicate that women with PE are associated with significantly higher circulating NGAL as compared to those without PE, and the increment of circulating NGAL in PE may occur as early as in the first trimester of pregnancy. Although large-scale studies are needed to validate the findings, results of the meta-analysis support that circulating NGAL may be a biomarker of PE.

Author contributions

Yaxian Wei designed the study. Yaxian Wei and Li Li performed literature search, quality evaluation and data collection. Fuchuan Wang, Lihua Fu, Zhenhua Li, and Yuhong Hu performed statistical analyses and interpreted the results. Yaxian Wei drafted the manuscript. Li Li, Fuchuan Wang, Lihua Fu, Zhenhua Li, and Yuhong Hu revised the manuscript. All authors approved the submission of the manuscript.

Ethical approval

This article does not contain any studies with human participants or animals performed by any of the authors.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data used to support the findings of this study are included within the article.

Additional information

Funding

The article has no sources of funding.