Placental Growth Factor (PlGF) associated with compromised Fetal
Growth and Perinatal Outcomes in a High-Risk Pregnancy Population: A
Retrospective Cohort
Introduction:
Fetal Growth Restriction (FGR) is a failure of the fetus to reach its
genetically determined growth potential(1), mostly associated with
placental dysfunction(2). Growth-restricted fetuses have an increased
risk of stillbirth, neonatal morbidity, and mortality(1, 3, 4).
Identification of circulating Placental Growth Factor (PlGF) in maternal
serum is important for the accurate prediction and timely diagnosis of
FGR(1, 2, 5-7).
While FGR refers to a fetus that does not reach its growth potential
(regardless of growth centile), Small for Gestational Age (SGA) includes
constitutionally small but healthy fetuses at lower risk of adverse
perinatal outcomes(5). SGA is typically defined by the deviation of
fetal size from reference percentiles. The overlap between these
definitions can cause some confusion, leading to the development of the
Delphi Criteria(8) to better differentiate FGR from SGA. These criteria
use a combination of abdominal circumference, Estimated Fetal Weight
(EFW), and end-diastolic flow in the umbilical artery, rather than EFW
alone, to define FGR(8). FGR is also further classified as either
early-onset (< 32 weeks) or late-onset (> 32
weeks). The Canadian guidelines have adopted the Delphi criteria for FGR
and SGA, defining SGA as an EFW or birth weight below the
10th percentile(5). The Delphi ultrasound-based
criteria are used for diagnosis of FGR, requiring one of three criteria
for early-onset FGR and two of three criteria for late-onset FGR(5).
Recent work has connected antenatal serum PlGF with the development of
placental-mediated FGR(2, 9, 10), performing better than other markers
of placental dysfunction for antenatal diagnosis(7). This suggests that
measuring PlGF levels can not only help to identify higher-risk
pregnancies in SGA fetuses but can also allow for early detection and
monitoring(2, 6). The Canadian guidelines for screening, diagnosis, and
management of FGR reflect the literature on PlGF and recommend
measurement of PlGF to help identify growth-restricted fetuses affected
by underlying placental disorders(8).
The objective of this study was to evaluate the association between
Fetal Growth Restriction defined by Delphi Criteria(8), Placental Growth
Factor and perinatal outcomes in a high-risk pregnancy population.
Methods:
2.1) Study design: Retrospective cohort study of 292 patients seen at
the Fetal Assessment Unit, Regina General Hospital, from December 2021
until September 2024. The study received REB approval #Bio3702.
2.2) Inclusion criteria: High-risk pregnant patients who had an
indication to test PlGF during the period of the study and followed the
Clinical Protocol at the Fetal Assessment Unit, described below.
2.3) Exclusion criteria: low-risk pregnancies, non-singleton
pregnancies, pregnancies with fetal malformations or fetal aneuploidies,
patients who do not consent to participate in the study and minors (age
< 18yo).
2.4) Clinical Protocol:
PlGF was ordered from 12w-36w at the initial visit at the Fetal
Assessment Unit, following the indications below:
1) Previous obstetrical history of term or preterm severe
preeclampsia/HELLP(1);
- Fetal Growth Restriction currently or previously (defined by Delphi
Criteria)(8);
- Low PAPP-A (<0.4 MoM), High AFP (> 2.3 MoM),
during 1st/2nd trimester screening(9);
- Previous history of pregnancy complications due to suspected placental
insufficiency: previous FGR < 5th,
previous abruption, previous stillborn, significant placental
pathology (maternal vascular malperfusion, fetal vascular
malperfusion)(11);
- Chronic Hypertension (PlGF can help to separate those patients that
might develop superimposed preeclampsia)(11);
- To differentiate other clinical conditions from preeclampsia(12).
PlGF classification:
Using the table from Mclaughlin et al. (2021)(13):
Low PlGF (≤ 5% for gestational age)
Normal PlGF (≥ 10% for gestational age)
2.5) Groups, Exposure and Outcomes:
The patients were divided into three groups depending on Fetal Growth.
Patients presenting with compromised fetal growth were divided according
to Delphi Criteria(8) (FGR < 3% and SGA 4-10%). Patients
with compromised fetal growth were subdivided into < 32w
(early onset) and > 32w (late onset), depending on the
gestational age they were at their first visit to our Fetal Assessment
Unit (time of diagnosis). Normal Fetal Growth was defined by Estimated
Fetal Weight (EFW) assessed by ultrasound from 11%-90%ile (Normal
Fetal Growth)(14).
A total of 292 participants were enrolled. The normal-grown fetuses’
group (EFW11-90%) had 213/292 (73%) participants. The FGR <
3% and SGA 4-10% groups had in total 79/292 participants (27%). The
FGR < 3% had 46 participants (36 participants <
32w; 10 participants > 32w); the SGA 4-10% group had 33
participants (21 participants < 32w; 12 participants
> 32w).
The primary exposure was maternal PlGF levels, which were divided into
Normal (≥ 10%) and Low (≤ 5%), based on a previous study that
classified PlGF values in percentiles for each gestational age from 12w
to 36w(13).
The primary outcome assessed was fetal growth. The other outcomes
assessed were abnormal placental morphology in the initial ultrasound
(US) at the Fetal Assessment Unit (thin placenta, presence of echogenic
cystic lesions, “bulky” placentas, small placentas or calcified
placentas), the development of preeclampsia (PET) at any gestational
age, fetal demise (IUFD), and preterm birth (PTB) < 34w.
Secondary perinatal outcomes included birthweight < 3%(15),
admission to the NICU, and findings in placental pathology consistent
with placental insufficiency (fetal and maternal vascular
malperfusion)(16).
2.6) Statistical analysis:
Assuming that 10% of the subjects in the reference population have
FGR(5), the study would require a total sample size of at least 150
participants at a 1:2 rate (test: reference group) to achieve a power of
80% for detecting a difference in proportions of 0.22 between the two
groups(17) at a two-sided p-value of 0.05(18). Exposure and outcomes
were analyzed in contingency tables with a chi-square test
(X2 ) or Fisher’s Exact test to assess
associations. Odds Ratio (OR) with a 95% Confidence Interval (CI 95%)
was also calculated for the studied outcomes.
Parametric data (average of gestational age at first visit, gestational
age at delivery and interval between gestational age at first visit and
delivery) were analyzed with one-way ANOVA to compare variances and the
differences between the means between the studied groups.
The software PRISM 10 for macOS version 10.4.1 was used for all
analyses. Associations were significant if p < 0.05.
Results:
Table 1 summarizes the associations between exposure and outcomes. The
association between maternal PlGF levels and compromised fetal growth
was highly statistically significant with p < 0.0001
(OR 4.9 CI 95% 2.9-8.5). Patients with compromised fetal growth have
almost 5-fold chances of having low maternal PlGF levels. Prenatally,
abnormal sonographic placental morphology was also highly associated
with fetal growth restriction (p< 0.0001; OR 5.3 CI 95%
3.1-9.1) (Figure 1).
Although patients with compromised fetal growth had an increased
frequency of preeclampsia compared to patients with normal fetal growth
(34% vs 26%), this association was not significant (p=0.18). On the
other hand, there were 12.5-fold chances of fetal demise (IUFD) in
fetuses with compromised growth compared to normal-grown fetuses (OR
12.5 CI 95% 2.08-75.42; p=0.001). The likelihood of low birthweight
< 3% was also increased (OR 10.3 CI 95% 3.28-29.46) in
growth-restricted fetuses, compared to normal-grown cases. The chances
of preterm birth < 34w (OR 3.99 CI95% 2.12-7.22) and neonatal
admission to NICU (OR 3.81 CI95% 2.15-6.52) were also augmented among
patients with FGR/SGA fetuses.
Abnormal placental pathology findings, Maternal Vascular Malperfusion
(MVM) and Fetal Vascular Malperfusion (FVM) were also increased 7-fold
in patients with growth-restricted fetuses, compared to normal-grown
cases (OR 7.06 CI95% 3.92-12.62).
Table 2 summarizes the average gestational age at the first visit, at
delivery, and the difference between the gestational age at the initial
visit and delivery. For normal-grown fetuses, the average gestational
age at the first visit was 26.9±6.16 weeks; for FGR, 3% was 27.8±4.80
weeks, and for SGA, 4-10% was 30.2±4.16 weeks. The average gestational
age at delivery for normal-grown fetuses, FGR 3% and SGA 4-10%, were
respectively 36.5±3.0 weeks, 32.8±5.1 weeks, 35.9±2.9 weeks. The
interval in days from initial visit until delivery for the studied
groups was 66.5 days (9.5 weeks) for normal-grown fetuses, 35.5 days
(5.07 weeks) for FGR 3% and 39.5 days (5.6 weeks) for SGA 4-10%. The
differences between the averages of gestational ages were statistically
significant (p < 0.05) and are summarized and broken down
(< 32 weeks and > 32 weeks) in Table 2.
Discussion:
This retrospective cohort studied perinatal outcomes associated with
normal fetal growth and compromised fetal growth in patients from a
high-risk pregnancy clinic.
In our studied population, the timelines of FGR/SGA pregnancies
significantly differed from those of normal-grown fetuses, which had
longer pregnancies. Pregnancies with normal-grown fetuses were on
average initially seen at 27w and delivered at 36-37w (9.5w interval).
While early onset (< 32 w) FGR/SGA had an average of initial
diagnosis at 27-28w and delivery at 33-35w (5-7.5w interval), late-onset
FGR/SGA was initially seen around 33-34w and delivered at 36-37w (2.5w
interval). The significant differences between the initial visit and
diagnoses of FGR and SGA and the delivery timing should be carefully
analyzed, as timely delivery in infants with FGR/SGA should balance
between preventing stillbirth and reducing risks associated with
prematurity(19).
We demonstrated a statistically significant association between maternal
low PlGF (< 5%) and compromised fetal growth, consistent with
current literature(1, 2, 6, 7, 20). In a study evaluating 47 biomarkers
and ultrasound parameters, PlGF had a sensitivity of 93% and a negative
predictive value (NPV) of 90% for predicting delivery of an SGA/FGR
infant in women presenting with suspected preeclampsia(21). Similarly,
Kingdom et al. also showed an association between PlGF and
placental FGR with a sensitivity of 70% and NPV of 60%(7). These
results show a promising role of PlGF testing as a screening method for
FGR in settings where access to ultrasound for diagnosis is limited(22).
Of note, our Maternal Fetal Medicine Unit is the second centre in Canada
utilizing real-time PlGF to manage high-risk pregnancies.
Our findings showed that abnormal sonographic placental morphology was
seen in 62% of cases with FGR/SGA, which was a significant association.
Thin or globular placentas and the presence of echogenic cystic lesions
were among the most frequently documented placental abnormalities
(Figure 1). Placental thickness is known to increase with gestational
age, and a thin placenta is an early echographic sign of FGR(20).
Additionally, our results show a significant association between growth
restriction and confirmed insufficiency in placental pathology. PlGF has
been shown to mark placental insufficiency(3). Comparably, recent work
by Shinar et al. has also demonstrated a significant association
between low PlGF and MVM in SGA fetuses(7).
One interesting finding of our study was the non-statistically
significant association between compromised fetal growth and
preeclampsia, even though low PlGF levels were associated with
compromised fetal growth in the analyzed sample. A recent literature
review introduced the novel concept of normotensive Fetal Growth
Restriction (n-FGR) versus Fetal Growth Restriction associated with
hypertensive disorders (HDP-FGR)(21). Preeclampsia is the most frequent
clinical presentation of placental dysfunction(1, 13, 16), but not the
exclusive phenotype for placental insufficiency(23, 24). The results of
this study confirm that compromised fetal growth is also strongly
representative of placental dysfunction(24), associated or not with
preeclampsia, reinforcing the concept of normotensive FGR(25).
Fetal demise (IUFD) remains a significant concern for pregnancies
complicated by SGA and FGR(7). It is well documented that perinatal
morbidity and mortality is increased for SGA infants born at any
gestational age compared to healthy infants(26). In this study, we found
the chance of IUFD was 12.5 times higher in fetuses with compromised
growth. However, it should be noted that our study population was a
high-risk one. With an overall rate of IUFD of 2.4%, our results are
similar to other work examining high-risk pregnancies(27).
Preterm birth (PTB) before 34 weeks, low birth weight, and the need for
NICU admission are critical complications frequently observed in
pregnancies characterized by placental insufficiency(11). Reduced levels
of PlGF have been strongly associated with these adverse outcomes, as
low PlGF levels indicate impaired angiogenesis and suboptimal placental
development(28). This disruption in placental function contributes to
FGR, which often necessitates preterm delivery to prevent further
complications, resulting in low birth weight and increased NICU
admissions due to prematurity-related morbidities (e.g. respiratory
distress syndrome)(5).
Our results show rates of PTB under 34 weeks were 4 times higher in
SGA/FGR infants, with early onset FGR being higher risk at 56%. We also
observed increased NICU admissions (OR 3.81) for FGR/SGA infants. A
randomized controlled trial looking at perinatal outcomes with the
addition of enoxaparin to treat high-risk pregnancies with low PlGF
showed a NICU admission rate of approximately 20%, which is similar to
our overall rate of admission(27). However, it is worth noting that our
data was limited, as ethics approval restricted our ability to follow
infants after birth. It is likely that there are additional NICU
admissions that are not accounted for in our data.
This study is a retrospective chart review, which inherently carries
limitations such as missing or incomplete data and potential biases
related to data collection and patient selection(29). While the cohort
size is relatively small, our findings provide valuable insights into a
high-risk pregnancy population, underscoring the potential relevance of
these results for clinical practice. It is important to note that
healthy, low-risk pregnancies were not assessed, limiting the
generalizability of our findings to the broader population. As a cohort
study, our work is also subject to confounding factors and cannot
establish causality(29).
Nevertheless, these results lay the groundwork for future research,
which aims to build on these findings. We are currently undertaking
additional prospective work. A particularly intriguing avenue for
further investigation is the potential role of enoxaparin in augmenting
PlGF levels(13, 30, 31). In a small study by McLaughlin et al. ,
the daily administration of enoxaparin to patients with high-risk
pregnancies and low PlGF showed promise for restoring deficient PlGF
levels(13). However, a trial in New Zealand showed that the addition of
enoxaparin did not reduce the risk of recurrence of preeclampsia or
having an SGA fetus in patients with a known history(27). Hence, further
studies are necessary to investigate the role of enoxaparin in
augmenting PlGF levels and potentially preventing detrimental perinatal
outcomes.
Conclusion:
The results show that real-time PlGF for the management of high-risk
cases with compromised fetal growth is useful. The associated risks of
imminent preterm birth, early-onset preeclampsia, and IUFD may warrant
referral of high-risk women with low PlGF levels to maternal fetal
medicine centers(11). The proposed clinical protocol for the use of PlGF
in high-risk centres is effective in screening patients who need
increased maternal-fetal surveillance. In this context, PlGF testing has
the potential to overcome some of the real challenges healthcare systems
in Canada and elsewhere face in providing effective obstetrical care to
women(32), especially in remote or low-income settings, which is a
reality in Saskatchewan(33).