Adequate utero-placental vascular development in the first trimester of pregnancy is essential to ensure optimal placental development and function, and consequently to achieve a healthy pregnancy outcome. In this study, increased maternal serum PlGF at 11 weeks is associated with increased first-trimester PV development, reflecting trophoblast tissue development and increased uPVV and uPVS development, reflecting volumetric and morphologic (branching) development of the utero-placental vasculature, respectively. Moreover, increased maternal serum sEng/PlGF ratio is associated with decreased first-trimester PV and decreased uPVV and uPVS development. These associations are not affected by conception method, fetal sex or the occurrence of placenta-related complications.

The positive association between maternal serum PlGF at 11 weeks and the novel uPVV and uPVS imaging markers substantiates that uPVV and uPVS measurements reflect development of the utero-placental vasculature in the first trimester of pregnancy.

Interestingly, our study revealed that the participants with an uneventful pregnancy had higher serum concentrations of sFlt-1 than participants that later experienced a placenta-related complication.

PlGF, a member of the vascular endothelial growth factor (VEGF)-family, stimulates angiogenesis and vasodilatation not only in the placenta, but also in other organs outside of gestation [32]. Animal and in vitro studies show PlGF is involved with branching angiogenesis (blood vessel formation) as well as non-branching angiogenesis (blood vessel elongation and enlargement) and plays an important role in the ability to employ compensatory mechanisms in pathologic conditions, such as ischemic tissue damage and tumor vascularization [32,33,34,35,36]. PlGF knockout mice exhibit aberrant utero-placental vascular development at gestational day 8.5, which corresponds to the late first-trimester in human pregnancy, including reduced branching [34]. Our study is the first to show the associations between circulating PlGF levels on first-trimester utero-placental vascular volumetric and morphologic development in vivo in humans.

Notably, we observed the strongest associations between PlGF and PV development. The PV measurements include but are not limited to the utero-placental vasculature. In fact, the PV measurements represent the whole placenta, which consists mainly of trophoblast cells. It is well known that trophoblast cells produce PlGF, which might explain why the association between PlGF and PV is stronger than associations with the uPVV or uPVS [37].

We did not observe any associations between sFlt-1, the soluble antiangiogenic receptor of PlGF, and first-trimester placental volume, utero-placental vascular volume or morphologic development. Although a few animal studies show increased early and mid-gestational levels of sFlt-1 are associated with impaired spiral artery remodeling [38, 39], these associations have not been studied in humans. One study found positive associations between sFlt-1 levels and the uterine artery pulsatility index, as an indirect measure of utero-placental vascular development, in the third trimester [40], but these associations were not observed in the first and second trimester [41, 42].

Studies comparing first-trimester sFlt-1 levels between healthy pregnancies and pregnancies with preeclampsia or fetal growth restriction report ambiguous results [43,44,45,46]. Importantly, clinically relevant rises of sFlt-1 are observed only from 5 weeks before the onset of preeclampsia [47]. As a consequence, elevated sFlt-1 levels are almost exclusively observed in the second and third trimester. Our study’s findings seem to confirm that circulating sFlt-1 does not have a substantial influence on utero-placental vascular development in the first trimester.

Remarkably, in our study we observed lower sFlt-1 serum concentrations in pregnancies that later developed a placenta-related complication compared to the group with an uneventful pregnancy, which is in line with previous findings by Schiffer et al. [31]. However, other studies found first-trimester sFlt-1 to be higher in pregnancies that later developed preeclampsia [29, 48]. These seemingly paradoxical results pose interesting conjecture for future research.

We observed no associations between sEng, a small receptor with anti-angiogenic properties by limiting the bioavailability of proangiogenic TGFβ, and first-trimester placental volume, utero-placental vascular volume or morphologic development. Although several studies have investigated the predictive value of sEng in the first trimester for the risk of preeclampsia or fetal growth restriction, no studies have researched its associations with utero-placental vascular development. We found one study that found a negative correlation between sEng levels and the uterine artery pulsatility index in the second trimester, but not in the first trimester, which is in accordance with our results [42].

In our study, there were no differences in sEng levels between women with and without placenta-related complications. Some studies show higher first-trimester levels of sEng in women ultimately developing preeclampsia [29, 49]. Notably, in our study, placenta-related complications comprised a heterogeneous group of clinical conditions defined as the presence of PIH, PE and/or FGR, SGA or PTB, which could explain why we observed no associations. Further, other studies show elevated sEng levels in pregnancies which later developed preeclampsia from 17 weeks onwards [50], which may also explain the lack of associations in our study.

We found no associations between the sFlt-1/PlGF ratio and utero-placental vascular development. Our results imply that, at least in the first trimester, free PlGF levels are relevant for physiological angiogenesis (utero-placental vascular development) irrespective of sFlt-1 levels. These findings are in line with previous research, which suggests free PlGF concentrations are diminished in patients with preeclampsia independent from sFlt-1 concentrations [51]. Our findings are further substantiated by a recent study that shows the imbalance of sFlt-1/PlGF ratio in preeclampsia is mainly caused by diminished PlGF production as opposed to increased sFlt-1 levels [52]. These findings give rise to the possibly that, in the first trimester, an increase of sFlt-1 in is accompanied by a proportional rise in PlGF, maintaining the concentration of free PlGF constant. In later pregnancy, during the late second and third trimester, excessive rises in sFlt-1 are no longer matched by PlGF, which leads to a sFlt-1/PlGF ratio imbalance and gives rise to placenta-related complications.

Although previous studies have found associations between the sEng/PlGF ratio and pregnancy outcomes [26], no studies have investigated its associations with angiogenesis. The sEng/PlGF ratio can be interpreted as a composite ratio reflecting both the VEGF- and TGFβ-pathways involved with angiogenesis. In our study, associations with PlGF are stronger than associations with the sEng/PlGF ratio. In addition, we found no associations with sEng. Therefore, the associations between the sEng/PlGF ratio and imaging markers of utero-placental vascular development are likely contributable to the effect of PlGF and we consider the possibility the TGFβ-pathways is involved with first-trimester utero-placental vascular development unconvincing.

One of the major strengths in our study is the use of validated test kits for the analyses of PlGF, sFlt-1 and sEng. Further, we made use of longitudinal first-trimester 3D power Doppler ultrasounds to perform validated 3D virtual reality based segmentations for the measurements of uPVV and uPVS [12, 13].

Unfortunately, all serum biomarker concentrations were determined only once during these pregnancies. Some studies show longitudinal changes in PlGF, sFlt-1 and sEng are potentially more clinically relevant than absolute levels [30, 53]. Longitudinal analysis may provide additional insights in the interplay between serum biomarker concentrations and first-trimester imaging markers of utero-placental vascular development.

The number of imaging markers used in this study might introduce some questions regarding the issue of multiple testing. However, our research has repeatedly demonstrated positive strong correlations between the uPVV and uPVS and an inverse association with density of vascular branching. Therefore, we argue the consistency of the presence or absence of these relationships can be viewed as internal validation of our findings.

This study’s findings suggest higher free PlGF concentrations likely contribute to increased first-trimester utero-placental vascular development as reflected by uPVV and uPVS, thereby providing new insights into potential mechanisms underlying placenta-related complications. Moreover, these results substantiate a role for PlGF in early prediction models for placenta-related complications. In contrast, sFlt-1 and sEng appear not to have a substantial influence on first-trimester utero-placental vascular development and therefore the added value of first-trimester sFlt-1 and sEng for early detection of placenta-related complications will be limited.

Importantly, the uPVV and uPVS can be used as 3D power Doppler imaging markers to monitor the volumetric and morphologic development of the utero-placental vasculature throughout the first trimester of pregnancy. Future application of the uPVV and uPVS measurements could benefit research on early vascular development to improve our understanding of the pathophysiology of placenta-related complications and the mechanisms underlying both therapeutic and preventative regimens, such as prophylactic administration of aspirin.

Finally, as maternal serum PlGF is currently used in multiple prediction models for preeclampsia, we recommend investigating the added value of the uPVV and uPVS as non-invasive and uncostly predictors for preeclampsia.