The correlation between serum levels and placental tissue expression of PLGF and sFLT-1 and the FIGO grading of the placenta accreta spectrum disorders

Abstract

Objective

This study aims to further explore the role of angiogenic vs anti-angiogenic factors in placenta accreta spectrum (PAS).

Methods

This cohort study included all patients with placenta previa and placenta accreta spectrum (PAS) disorders undergoing surgery at Dr. Soetomo Hospital (Academic Hospital of Universitas Airlangga, Surabaya, Indonesia) from May to September 2021. Venous blood samples for PLGF and sFlt-1 were drawn immediately prior to surgery. Placental tissue samples were taken during surgery. The FIGO grading was diagnosed intraoperatively by an experienced surgeon and confirmed by the pathologist and followed by immunohistochemistry (IHC) staining. The sFlt-1 and PLGF serum were performed by an independent laboratory technician.

Results

Sixty women were included in this study (20 women with placenta previa; 10 women with FIGO PAS grade 1; 8 women with FIGO PAS grade 2; 22 women with FIGO PAS grade 3). The median with 95% Confidence interval of PLGF serum values in placenta previa, FIGO grade I, grade II, and grade III were 233.68 (0.00–2434.00), 124.39 (10.42–663.68), 236.89 (18.83–418.99) and 237.31 (2.26–3101.00) (p = .736); the median values with 95% CI of serum sFlt-1 levels in placenta previa, FIGO grade I, grade II, and grade III were 2816.50 (418.00–12925.00), 2506.00 (227.50-16104.00), 2494.50 (888.52–20812.00), and 1601.00 (662.16–9574.00) (p = .037). Placental PLGF expression in placenta previa, FIGO grade 1, grade II, and grade III showed median values (with 95% CI) of 4.00 (1.00-9.00), 4.00 (2.00–9.00), 4.00 (4.00–9.00), and 6.00 (2.00–9.00) (p = .001); sFlt-1 expression median values (with 95% CI) were 6.00 (2.00–9.00), 6.00 (2.00–9.00), 4.00 (1.00–9.00), and 4.00 (1.00–9.00) (p = .004). Serum PLGF and sFlt-1 levels did not correlate with placental tissue expression (p = .228; p = .586).

Conclusion

There are differences in PAS's angiogenic processes ​according to the severity of trophoblast cell invasion. But there is no overall correlation between serum levels and PLGF and sFlt-1 expression in the placenta, suggesting the imbalance between angiogenic and anti-angiogenic are local mechanisms in the placental and the uterine wall.

Keywords:

• PLGF
• sFlt-1
• placenta previa
• placenta accreta spectrum
• FIGO grading
• maternal health

Introduction

Placenta accreta spectrum (PAS) disorders are defined as partial or complete disruption of the trophoblast attachment to the myometrial uterine wall [1]. PAS disorders nowadays represent a growing obstetrical challenge since the patient mostly requires specialized multi-disciplinary teams to prevent catastrophic hemorrhage and maternal death. The main problem in PAS is massive bleeding during surgery which causes morbidity and mortality [2]. Bleeding problems during surgery are due to a massive increase in neo-vascularization, which could be explained by an imbalance of angiogenic and anti-angiogenic in PAS [3].

PLGF (Placental Growth Factor) and sFlt-1 are pro-angiogenic and anti-angiogenic factors, which might play a role in placental vascular growth and maturation during PAS neovascularization [4]. This angiogenic and anti-angiogenic imbalance in PAS may be a local event but could also occur in systemic circulation. However, the mechanism leading to neovascularization in PAS is still unclear, and whether it is systemic or local at the placental-myometrial interface is still controversial[5–7].

The aim of this study is to understand the relationship between serum PLGF and sFlt-1 levels, and placental tissue expression of PLGF and sFlt-1 in PAS in order to better understand the local and/or systemic role of PLGF and sFlt-1 in the different FIGO PAS grades.

Methods

This is a cohort study using standardized placental sampling. Pregnant women diagnosed with placenta previa and PAS disorders undergoing surgery at Dr. Soetomo Hospital from May to September 2021 (Five months) were recruited. The primary outcomes of this study were the maternal serum levels and placental expression of sFlt-1 and PLGF in the different stages of PAS (based on FIGO grading). Pre-operative diagnosis using ultrasonography, which has more than two markers, indicates a high suspicion of PAS [8,9]. The diagnosis of placenta previa was made on the basis of the placenta being implanted into the lower uterine segment and close to the internal cervical ostium as demonstrated on abdominal and/or transvaginal using ultrasound. Hypertension, preeclampsia, IUGR, multiple gestations, metabolic disorder like obesity and gestational diabetes were excluded.

Patients were grouped into four categories: placenta previa as a control group, FIGO grade I PAS, grade II PAS, and grade III PAS [10]. The clinical classification of FIGO grade 1: abnormally adherent placenta (placenta accreta) with the uterus showing no obvious distension over the placental bed (placental “bulge”), no placental tissue invading through the surface of the uterus, and no or minimal neovascularity. FIGO Grade 2: Abnormally invasive placenta (increta) with findings at laparotomy: Abnormal macroscopic findings over the placental bed: bluish/purple coloring, distension (placental “bulge”), significant amounts of hypervascularity (dense tangled bed of vessels or multiple vessels running parallel cranio-caudally in the uterine serosa), no placental tissue seen to be invading through the uterine serosa, gentle cord traction results in the uterus being pulled inwards without separation of the placenta (the so-called ‘dimple sign’). FIGO grade 3: abnormally invasive placenta (percreta) divided into three categories: grade 3a, 3b, and 3c. FIGO grade 3a: placental invasion limited to the uterine serosa, FIGO grade 3b: placenta invades the urinary bladder, and FIGO grade 3c: placenta invades other pelvic organ/tissues [10].

The management of PAS depends on the placental topography; patient with placental invasion in the upper trigonal bladder can be successfully managed using uterine conservative-resective surgery, as recently published in large series of this Indonesian center [11]. Placental invasion in the lower bladder, parametrial invasion were managed using cesarean hysterectomy [12].

Venous blood samples were taken immediately prior to surgery, centrifuged serum was then stored in a refrigerator until all research samples were collected, followed by ELISA assays. Serum sFlt-1 and PlGF were analyzed using specific reagen-kits: sFlt-1 (Human sVEGFR 1/sFlt1 ELISA Kit Picokine) and PlGF (Human PLGF ELISA Kit PicoKine^TM). These Kits use a quantitative sandwich enzyme immunoassay technique. The minimum detectable doses in the assays for sFlt-1and PlGF were <30 and <1 pg/ml, respectively, with Intra- Assay precision of the sFlt-1 and PLGF reagen-kits were (4.1–7.8 and 5.6–7.9%, respectively), while the Inter- Assay precision was (4.7–9.7 and 5.6–9.5%, respectively). This assay method is the same as used in the original Levine et al. study [13]. The laboratory processes and analyses were all performed by independent laboratory technician in the Clinical Pathology Laboratory, Dr. Soetomo Academic Hospital, Surabaya.

All tissue sampling was taken intraoperative carefully and analyzed immediately by the pathologist to distinguish between invasive and dehiscence areas of PAS including the PAS managed by uterine conservative – resective surgery [14]. The results of immunohistochemical were calculated using scores from previous studies by Tseng et al. Cell percentage: Score 1: if stained 0-10% trophoblast cells, Score 2: 10-50%, Score 3: ≥50%; Cell intensity: score 1: weak intensity, score 2: moderate, score 3: strong. Final score; percentage score multiplied by intensity score, which is 1 to 9 ⁴.

Written consent was obtained from all patients. SPSS was used for the statistical analysis. Spearman test was used to determine the correlation between PLGF in serum and placental tissue and sFlt-1 in serum and placental tissue. Kruskal Wallis test was performed to determine the difference between PLGF and sFlt-1 serum levels in the placenta previa group, the FIGO grade I, grade II, and grade III groups.

This study was approved by the Health Research Ethics Committee of Dr. Soetomo General Academic Hospital, Surabaya, Indonesia (Approval number: 0197/KEPK/V/2021).

Results

During the five-month study periods 60 patients were recruited, 20 patients with placenta previa, 8 patients FIGO grade I, 10 patients FIGO grade II, and 22 patients FIGO grade III. Table 1 shows the clinical characteristics of four groups.

Table 1. Clinical characteristics of subjects with placenta previa totalis and with FIGO 1-3 PAS disorders.

Variable	Placenta previa (n = 20)	FIGO grade I (n = 10)	FIGO grade 2 (n = 8)	FIGO grade 3 (n = 22)	p
Characteristic
Maternal age (year)*	32 (27–41)	33,00 (27–38)	32,00 (30–38)	35,00 (22–44)	.0506
BMI (Kg/m^2)*	27,40 (20–44)	25,79 (23–34)	25,24 (19–32)	28,25 (22–50)	.255
Gravida*	3 (1–6)	4 (2–5)	3 (2–6)	3 (2–6)	.255
Delivery pregnancy week (week)*	37 (30–38)	36 (32–38)	36 (32–37)	36 (27–38)	.014
History of uterine operation* : no cs	5 (25%)	0	0	0	.032
Previous CS 1×	9 (45%)	3 (30%)	4 (50%)	8 (36,4%)
Previous CS 2×	6 (30%)	7 (70%)	4 (50%)	12 (54,5%)
Previous CS 3×	0	0	0	2 (9,1%)
Operation Procedure (n%)** : Cesarean section	20 (100%)	1 (10%)	0 (0)	0 (0)	.000
Uterine resective-reconstructive surgery	0 (0)	9 (90%)	7 (87,5%)	6 (27,3%)
Hysterectomy	0 (0)	0 (0)	1 (12,5%)	16 (72,2%)
Total blood loss (ml)*	450 (200–1500)	1200 (300–2500)	2100 (1000–4500)	2050 (1000–8000)	.000
Transfusion during surgery* PRBC (250cc/bag)	0 (0–2)	1,00 (0–4)	1,50 (0–4)	2,50 (1–8)	.000

PRBC: Packed Red Blood Cell; *Data presented as median (Min-Max); **Data presented as a number of sample; ***Statistical analysis using Kruskal–Wallis.

The maternal age, BMI, and gravidity were similar between the placenta previa and placenta accreta spectrum groups. However gestational ages at delivery were significantly different between the placenta previa, grade I, grade II, and grade III PAS groups (Table 1), with the PAS group delivering earlier (placenta previa 37 (30–38) vs 36 grade I (32–38) vs grade II 36 (32-37) vs grade III 36 (27-38). In this series, 17 patients required a hysterectomy because the placental invasion was in the lower part of the uterus involving the bladder making conservative surgery impossible [11].

Levels of serum PLGF were similar among the four groups. However, PLGF expression showed a trend to increase with increasing FIGO grading. Serum sFlt-1 was significantly decreased in the patients with the grade III invasion, while placental sFlt-1 expression was significantly lower in the higher FIGO PAS grades (Table 2).

Table 2. Results of serum levels and expression of PLGF and sFlt-1 in placenta previa totalis, grade I, grade II and grade III.

Variable	Placenta previa (n = 20)	Grade I (n = 10)	Grade II (n = 8)	Grade III (n = 22)	p-Value
	Median (min–max)	Median (min–max)	Median (min–max)	Median (min–max)
Serum PLGF (pg/ml)*	233.69 (0.00–2343)	124.39 (10.42–663.68)	236.89 (18.83–418.99)	237.31 (2.26–3101)	0.736
Serum sFlt-1 (pg/ml)*	2816.50 (418–12925)	2506.00 (227.5–16104)	2494.50 (888.52–20812)	1601.00 (662.16–9574)	0,037
PLGF Expression*	4 (1–9)	4 (2–9)	4 (4–9)	6 (2–9)	0.001
sFLT-1 Expression*	6 (2–9)	6 (2–9)	4 (1–9)	4 (1–9)	0.004
Adjusted >34 weeks of GA	n = 20	n = 9	n = 6	n = 19
Serum PLGF (pg/ml)*	228.66 (0.01–2434)	124.39 (10.42–663.68)	236.89 (55.01–419)	238.24 (2.26–3101)	0.693
Serum sFlt-1 (pg/ml)*	2941 (418–12925)	2506 (227.50–16104)	2494.5 (1398–4642)	1695 (679–9574)	0.083
PLGF Expression*	4 (1–9)	4 (2–9)	5 (4–9)	6 (2–9)	0.003
sFLT-1 Expression*	6 (2–9)	4 (1–9)	4 (1–9)	4 (1–9)	0.014

Grade I, II dan III: grading FIGO Placenta Accreta Spectrum; PLGF: Placental growth factor; sFlt-1 : Soluble Fms-like tyrosin kinase 1. Data presented in median (min-max) & mean ± SD; *Statistical analysis using Kruskal–Wallis.

Comparisons using the Mann–Whitney test showed that sFlt-1 tissue expression in the placenta previa group was not significantly different compared with PAS grade I and II groups. However, placenta previa had a significantly higher sFlt-1 expression when compared with grade III (6 (2–9) vs 4 (1–9); p = .008); grade I and grade II (6 (2-–) vs 4 (1–9); p = .046), and grade I (Figure 1(A)) versus grade III (6 (2-9) vs 4 (1-9); p = .001) respectively (Figure 1(B)).

Figure 1. (A) sFlt-1 expression at 400x magnification in Placenta Previa. Percentage of cells 0-10% with weak intensity. Blue arrow indicates IHC stained in the villous trophoblast. Red arrow indicates the mesenchymal cell. Yellow arrow indicates the blood vessel in the placental villous. (B) Strong intensity of sFlt-1 expression of villous trophoblast (blue arrow) at 400x magnification in FIGO grade 3. Red arrow indicates the mesenchymal cell.

PLGF expression followed a similar pattern. Placenta previa had a significantly lower PLGF expression when compared with grade II (4 (1–9) vs 4 (4–9); p = .014) and grade III. (4 (1–9) vs 6 (2–9); p < 0.001) In contrast, there was no significant difference in PLGF expression between placenta previa with grade I (Figure 2(A)) and grade II with grade III (figure 2(B)). The Spearman test results showed a P-value of p = .228 and p = .586) indicating no correlation between serum PLGF and sFlt-1 levels and PLGF and sFlt-1 expression in placental tissue.

Figure 2. (A) PLGF expression at 200× magnification in Placenta Previa show weak intensity with percentage of cells 0-10%. Blue arrow indicates IHC stained in the villous trophoblast. Red arrow indicates the mesenchymal cell. Yellow arrow indicates the blood vessel in the placental villous. (B) Strong intensity of PLGF expression at 200x magnification in FIGO grade 3. Percentage of cells ≥ 50% with strong intensity of villous trophoblast (blue arrow). Red arrow indicates the mesenchymal cell. Yellow arrow indicates the blood vessel in the placental villous.

In the Sub-analysis to exclude extreme gestational age, we analyzed cases of gestational age > 34 weeks, and the result showed similar patterns

Levels of serum PLGF were similar among the four groups; however, serum sFlt-1 showed a trend to decrease with increasing FIGO grading. While placental PLGF expression showed a trend to increase with increasing FIGO grading and sFlt-1 expression showed the same trend as serum, where expression decreased with increasing FIGO grading (Table 2).

These results show that the sFlt-1 serum and sFlt-1 expression may have an important role in PAS grading, but there is no correlation between systemic and intra-placental sFlt-1 and PlGF.

Discussion

Dr. Soetomo Academic General Hospital, Surabaya, Indonesia, has been confronted with a massive increase in the number of PAS cases, reaching 4% of all deliveries in Dr. Soetomo Hospital during 2013–2016 [15]. Over the period 2013-2020, 396 confirmed cases of PAS were managed in our center [11]. This rapid increase in the incidence of PAS patients is accompanied by increased morbidity due to massive bleeding related to the massive neo-vascularizationin PAS.

The results of this study in 40 patients with PAS disorders demonstrate that PLGF expression increases with increasing FIGO grading and serum sFlt-1 significantly decreases in patients with the grade III invasion, with placental sFlt-1 expression also significantly lower in the higher PAS grades. PAS grade III had the highest level of serum PLGF and the lowest level of serum sFlt-1. The pro-angiogenic profile is possibly involved with the increase in the process of angiogenesis in cases of PAS increasing the formation of massive new blood vessels in the placenta. The placenta is the largest organ that plays a role in producing the angiogenic factor PLGF. PLGF is expressed at the human implantation site by maternal uterine NK (uNK) and fetal trophoblast cells [16,17]. PLGF promotes trophoblast growth and maturation of the placental vascular system [13,18].

Unfortunately, the differences in placental expression of PLGF and sFlt-1 do not correlate with the serum levels of these biomarkers. When comparing the median serum levels of PLGF and sFlt-1 in cases of placenta previa and placenta accreta in this study with the result in normal pregnancies in the Levine et al. study [13] (same assay used) we found that (for a mean gestational age of 36–37 weeks) the median serum levels of PLGF in PAS grade III cases are higher than normal pregnancies at similar gestational ages (about 400 pg/ml) at >37 weeks), while the median serum level of sFlt-1 was lower (2000 pg/ml) [13]. When stratified by >34 weeks, the results showed a similar pattern to the Levine study but had higher serum PlGF levels and lowered sFlt-1 levels at the same gestational age. It indicates that this study has a consistent result on different categories of gestational age.

It is still uncertain if the overall pro-angiogenic status in PAS disorders is just a marker of the pathological process or a causative factor. There are conflicting data regarding the role of PLGF in trophoblast invasion [19,20]. Cytotrophoblast cells typically invade deeper when sensing an increasing oxygen gradient. Conversely, PLGF expression also increases with increased placental oxygenation, but it is not clear whether these two events are directly related or not. Decidual NK cells, critically involved in regulating early cytotrophoblast invasion into the decidua also produce PLGF [21].

The role of sFlt-1 in PAS disorders is still not completely understood. McMahon et al. in a smaller study, suspected that sFLT-1 plays a role in the regulation of placental invasion. They suggested that sFlt-1 has a role in controlling placental invasion, creating a favorable environment for placental separation and reducing the risk of postpartum hemorrhage. They conducted a retrospective case-control study over 10 years with a sample of just 10 patients with placenta accreta spectrum and 10 patients with placenta previa as controls. This study showed a possible relationship between low tissue immunohistochemistry sFlt-1 expression and the degree of placental invasion [22]. In contrast, Biberoglu et al. in a study on 68 placenta previa patients with and without concomitant placenta acreta spectrum showed no significant differences in maternal serum sFlt-1, PLGF, sFlt-1/PLGF, and VEGF ratios between placenta previa and placenta accreta spectrum [23].

SFlt-1, as a soluble receptor, binds to PLGF and VEGF, creating a critical deficiency in both pro-angiogenic factors, which leads to endothelial cell disease. sFlt has received particular attention and recognition as a syncytiotrophoblast (STB) stress marker, which increased levels in states of inflammation and ischemia-reperfusion [24–26]. There are contradictory differences between the pathophysiology of preeclampsia and placenta accreta. There may be differences in the cause of the decrease in sFlt-1 in these placental invasion disorder cases. It might be that the reduced sFlt-1 levels in high-grade accreta may reflect that the deeply embedded trophoblast is not exposed to too much oxidative stress. Alternatively, the remaining surface area of STB exposure is low; most of the intervillous space being absent in patients with high-grade PAS and major invasion areas. Ali and Chandraharan 2021 posed that the increasing levels of angiogenic factors (VEGF and PLGF) compared to normal placentation would drive the degree of placenta invasion of the uterine wall [27]. Although Wang et al. suggested that PLGF could potentially identify pregnancies at high risk for placenta accreta [28], our data demonstrate that PLGF levels are not different in PAS disorders.

Our study also revealed no differences in the angiogenic factor (PLGF) and anti-angiogenic factor (sFlt-1) between placenta previa and FIGO grade I and II, while there are significant differences with FIGO grade III. In contrast, Munoz et al. found that PLGF levels did not differ between placenta previa and placenta accreta spectrum [29]. The findings in FIGO grade III probably reflects the highly increased neovascularization process in PAS FIGO grade III penetrating beyond the uterine tissue. Whereas in placenta previa and PAS grades I, II, the attachment of the placenta is still limited in the uterus similar to placenta previa [10].

Even though appears to be a tendency for an increase in PLGF and a decrease in sFlt-1 levels proportional to the rise of PAS grading in both serum and tissue expression, there is, unfortunately, no clear correlation between serum levels and placental tissue expression of PLGF and sFlt-1.

This study indicates that the angiogenic process of placenta accreta occurs in the vicinity of the placenta and the uterine wall, suggesting that the imbalance of angiogenic and anti-angiogenic are local mechanisms. However clearly, more research is needed to fully understand the involvement of angiogenic processes in the maternal blood circulation of the PAS, which could involve a variety of mechanisms.

Conclusion

This study demonstrated differences in PAS's angiogenic processes according to the severity of trophoblast cell invasion. FIGO grade 3 patients have significantly lower sFlt-1 levels. However, there is no overall correlation between serum level and PLGF and sFlt-1 expression in the placenta, suggesting the imbalance between angiogenic and anti-angiogenic are local mechanisms in the placental and the uterine wall.

Author contribution

The study was design by RAA, ERN, LA. It was planned by RAA, ERN, GAD, EGD. It was conducted by LAA, RAA, GAR, BAG. Data were analyzed by LAA, RAA, ERN, GAD, AKB. All authors contributed to writing the manuscript.

Disclosure statement

The author reports no declarations of interest.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.