Expert Reviews ajog.

org

Diagnosis and surveillance of late-onset fetal

growth restriction
Francesc Figueras, PhD; Javier Caradeux, MD; Fatima Crispi, MD; Elisenda Eixarch, MD;
Anna Peguero, MD; Eduard Gratacos, PhD

without obvious cardiovascular changes

By consensus, late fetal growth restriction is that diagnosed >32 weeks. This condition is beyond these findings.5,6 Contrary to
mildly associated with a higher risk of perinatal hypoxic events and suboptimal neuro- early-onset FGR, in late-onset FGR the
development. Histologically, it is characterized by the presence of uteroplacental vascular association with preeclampsia is weak.7
lesions (especially infarcts), although the incidence of such lesions is lower than in The Table shows the main differential
preterm fetal growth restriction. Screening procedures for fetal growth restriction need to features between both clinical subtypes.
identify small babies and then differentiate between those who are healthy and those who Another major source of terminolog-
are pathologically small. First- or second-trimester screening strategies provide detec- ical confusion is the distinction between
tion rates for late smallness for gestational age <50% for 10% of false positives. pathologically and constitutionally small
Compared to clinically indicated ultrasonography in the third trimester, universal fetuses. By convention, both clinical
screening triples the detection rate of late smallness for gestational age. As opposed to forms have been termed as “fetal growth
early third-trimester ultrasound, scanning late in pregnancy (around 37 weeks) increases restriction” and constitutional “small-
the detection rate for birthweight <3rd centile. Contrary to early fetal growth restriction, ness for gestational age” (SGA), respec-
umbilical artery Doppler velocimetry alone does not provide good differentiation between tively. Whereas FGR represents a
late smallness for gestational age and fetal growth restriction. A combination of biometric pathological condition (mainly associ-
parameters (with severe smallness usually defined as estimated fetal weight or ated with placental insufficiency8) asso-
abdominal circumference <3rd centile) with Doppler criteria of placental insufficiency ciated with adverse perinatal outcome,
(either in the maternal [uterine Doppler] or fetal [cerebroplacental ratio] compartments) constitutional smallness is associated
offers a classification tool that correlates with the risk for adverse perinatal outcome. with near-normal perinatal outcomes as
There is no evidence that induction of late fetal growth restriction at term improves it represents the lowest end of the size
perinatal outcomes nor is it a cost-effective strategy, and it may increase neonatal spectrum of normal fetuses.
admission when performed <38 weeks.
Short- and long-term consequences
Key words: fetal growth restriction, infant, late-onset disorders, newborn, small-for- of late FGR
gestational age, term birth Neonatal and infant consequences
Approximately one third of the medically
indicated late preterm births are compli-
Definition of “late-onset” fetal that a cut-off of 32 weeks at diagnosis or 34 cated with FGR.9 Late FGR is associated
growth restriction weeks at delivery maximized the clinical with cesarean delivery for fetal distress,
Late fetal growth restriction (FGR) is differences between early- and late-onset neonatal acidosis, and admission to the
usually defined as that diagnosed >32 FGR, in terms of perinatal mortality neonatal unit.10 The association with
weeks of pregnancy. One study1 showed (7.1% vs 0%; P < .001), adverse perinatal harder hypoxic events emerges when
outcome (13.4% and 4.6%; P <.001), and large cohorts are analyzed. Mendez-
association with preeclampsia (35.1% vs Figueroa et al,11 in a cohort of 5416
From the Barcelona Center for Maternal-Fetal 12.1%; P < .001). More recently, a survey term, uncomplicated pregnancies with
and Neonatal Medicine (Hospital Clínic and
Hospital Sant Joan de Deu), Institut
was conducted on 45 experts aiming SGA (birthweight [BW] <10th centile)
d’Investigacions Biomèdiques August Pi i at reaching consensus on the definition found a higher incidence of neonatal
Sunyer, University of Barcelona; and Center for of late vs early FGR.2 There was good death (1.1 vs 0.4/1000 births; adjusted
Biomedical Research on Rare Diseases, Madrid, agreement (89%) in defining late FGR odds ratio [OR], 2.56; 95% confidence
Spain. as that diagnosed >32 weeks. interval [CI], 1.83e3.57). In another
Received Oct. 3, 2017; revised Nov. 16, 2017; While in early-onset FGR the typical recent study, Chauhan et al12 evaluated in
accepted Dec. 1, 2017.
pattern of deterioration progresses from a cohort of 115,502 uncomplicated
The authors report no conflict of interest. escalating abnormalities in Doppler pregnancies of nonanomalous singletons
Corresponding author: Francesc Figueras, PhD. parameters to abnormal biophysical born at term the association between SGA
ffiguera@clinic.cat
parameters,3,4 in late-onset FGR there is a (<10th centile of BW: n ¼ 4983) and
0002-9378/$36.00 hypoxic composite neonatal morbidity
common pattern of normal or minimally
ª 2017 Elsevier Inc. All rights reserved.
https://doi.org/10.1016/j.ajog.2017.12.003 elevated umbilical Doppler indices with including 5-minute Apgar score <5
mildly abnormal cerebral Doppler, but (prevalence among SGA 0.4%), hypoxic

S790 American Journal of Obstetrics & Gynecology FEBRUARY 2018

ajog.org Expert Reviews

ischemic encephalopathy (prevalence

0.4%), seizures (prevalence 0.1%), and TABLE
neonatal death (prevalence 0.1%). After Main differential features between both clinical phenotypes of fetal
adjusting for potential confounders, growth restriction
hypoxic composite neonatal morbidity Early FGR Late FGR
was significantly higher in SGA (1.1%) Prevalence 7
0.5e1% 5e10%
compared with normally grown babies 10
Challenge Management (gestational Detection and diagnosis
(0.7%; adjusted relative risk [RR], 1.44; age at delivery)
95% CI, 1.07e1.93). A large case-control
Evidence of placental High Low
study13 including 493 babies with cerebral disease 1,7,a 70% Abnormal umbilical <10% Abnormal umbilical
palsy born 35 weeks found severe Doppler Doppler
smallness (BW <2 SD) to be associated 60% Association with 15% Association with
with an OR of 4.81 (95% CI, 2.7e8.5). preeclampsia preeclampsia
It has been shown by spectroscopy that Severe angiogenic Mild angiogenic disbalance
disbalance
late SGA fetuses14 and, to a greater extent,
late FGR infants15 (defined by BW <3rd Pathophysiology Hypoxia þ/þ Hypoxia þ/e
centile or Doppler abnormalities) have and oxygen delivered to Systemic cardiovascular Central cardiovascular
brain6 adaptation adaptation
brain metabolite differences vs normally
grown babies that are correlated with later Clinical impact10 High mortality and morbidity Low mortality/morbidity þ high
prevalence ¼ large etiological
neurodevelopment. A meta-analysis16 on fraction of adverse outcomes
neurodevelopment in term SGA babies
FGR, fetal growth restriction.
including 28 studies (7861 SGA babies) a
Crispi F, Dominguez C, Llurba E, Martin-Gallan P, Cabero L, Gratacos E. Placental angiogenic growth factors and uterine artery
found that SGA-born infants had 0.32 SD Doppler findings for characterization of different subsets in preeclampsia and in isolated intrauterine growth restriction. Am J
poorer (95% CI, 0.25e0.38) standardized Obstet Gynecol 2006;195:201-7.
neurodevelopmental scores. Figueras. Late-onset fetal growth restriction. Am J Obstet Gynecol 2018.

Long-term consequences
At long term, the effects of SGA are >30 years) diabetes mellitus (OR, 2.42; percentile (compared to 9.9% of 142
more difficult to disentangle from other 95% CI, 1.44e4.07, adjusted for body placentas from normally grown babies; P
environmental factors. However, a mass index and parental history of < .001). Only 21.8% (31/142) of SGA
recent cohort17 (n ¼ 1,100,980) study diabetes). placentas were free of histological
that adjusted for maternal and paternal abnormalities, while it was 74.6% (106/
educational level found that term SGA Placental histopathological findings 142) in the normally grown group
was significantly associated with an in late FGR (P < .001). In the abnormal SGA pla-
increased risk of poor school perfor- Placentas from FGR fetuses delivered at centas (111/142) there were a total of 161
mance at the time of graduation from term have significantly increased fre- lesions (classified according standard-
compulsory school (grades <10th quencies of uteroplacental vascular ized criteria25) attributable to maternal
percentile), with adjusted OR and 95% lesions (especially infarcts) compared to underperfusion in 64% (103/161), fetal
CI ranging from 1.85 (1.65e2.07) for normal controls, although the incidence underperfusion in 15.5% (25/161), and
severe SGA (<3 SD of BW) to 1.5 of such lesions is much lower than in inflammation in 20.5% (33/161). Inter-
(1.43e1.58) for moderate SGA (BW e2 preterm FGR.20-22 Furthermore, it has estingly, those pregnancies with signs of
to e3 SD). In a subanalysis, all BW been reported that compared to normal underperfusion25 had a significantly
groups were associated with an term pregnancies, placentas from FGR at higher incidence of emergency cesarean
increased risk of poor school perfor- term may have an increased incidence of delivery for nonreassuring fetal status
mance among boys with short stature other villous lesions including fibrosis, (44.1% vs 21.4%, respectively; P ¼ .013)
(10.1% of those individuals born with a hypovascularity, and avascularity, sug- and neonatal metabolic acidosis at birth
BW <2 SD) compared to those with gestive of fetal thrombotic events.23 (33.3% vs 14.3%, respectively; P ¼ .023)
nonshort stature. Finally, it has been Hence, differences in placental histo- than did those without signs of under-
suggested that fetal programing also pathological findings between late and perfusion.26 Furthermore, neonatal
operates in term SGA babies,18 predis- early FGR are more quantitative (in morbidity (as assessed by the Morbidity
posing them to a higher incidence of severity and extension) rather than Assessment Index For Newborns
metabolic syndrome. Another recent qualitative.24 A series of 142 placentas score27) differed significantly between
large cohort study19 on 49,927 female from singleton SGA pregnancies born those with and without placental signs of
nurses found that term SGA (<10th >34 weeks with normal umbilical artery underperfusion (89 vs 0, respectively;
centile of BW) was associated with an (UA) Doppler velocimetry found that P ¼ .025). Finally, 83 infants of the same
increased risk of adult-onset (diagnosed 54.2% had placental weights <3rd cohort were followed up for 2-year

FEBRUARY 2018 American Journal of Obstetrics & Gynecology S791

Expert Reviews ajog.org

neurological assessment28 when early-onset growth restriction in up to is recommended as a simple and inex-
adjusted neurodevelopmental outcomes 90%.7 However, late-onset growth re- pensive screening tool.38,39 Only 1 ran-
(Bayley scale) were significantly poorer striction is still largely unpredicted. Most domized controlled trial40 addressed the
in births involving placental under- studies addressing first-trimester incremental yield of fundal height mea-
perfusion (relative to SGA infants screening for late FGR, used SGA as a surement over abdominal palpation,
without these signs) for all 3 domains of proxy, reporting detection rates (DR) for a showing a nonsignificant improvement
the Bayley scale: cognitive (105.5 vs 96.3, 10% of false positives ranging from 25% of 32% (95% CI, e8% to 90%) in
adjusted P ¼.03), language (98.6 vs 87.8, (with only uterine artery [UtA] Doppler detecting neonatal SGA. Furthermore,
adjusted P < .001), and motor (102.7 vs velocimetry)32 to 51% (combining the meta-analysis of 34 studies (most of
94.5, adjusted P ¼ .007). The adjusted maternal characteristics, blood pressure, them hospital-based, which may bias the
ORs of abnormal cognitive, language, uterine Doppler velocimetry, pregnancy- results toward overoptimistic perfor-
and motor competencies in instances of associated plasma protein-A, and mance) showed a DR of SFH for SGA of
underperfusion were 9.3-, 17.5-, and placental growth factor).33 At first w60% for a false-positive rate of w15%,
1.44-fold higher, respectively, differing trimester, prediction of late FGR (when concluding that the method is unsuitable
significantly for the former 2 domains. defined as severe smallness [<3rd centile] for primary screening.41 Although SFH
or the presence of Doppler signs suggestive determination is of limited value in
Screening for late SGA of placental insufficiency) was better than routine obstetrical care, it continues to
Since failure to achieve growth potential prediction of late SGA (for a 10% of false- be the only physical examination
is a concept difficult to gauge, fetal size is positive rate 65.8% [95% CI, 64.8e66.8] screening test available. Furthermore,
used as a proxy, with all its limitations. vs 23% [95% CI, 18.8e27.2]).7,34 Predic- the studies do not report separate in-
Therefore, traditionally, a SGA fetus has tion for late FGR with preeclampsia was formation on early and late clinical
been regarded as equivalent of FGR. only slightly better (nonsignificantly) than subtypes, nor differentiate between SGA
However, a distinction between FGR for normotensive FGR (for a 10% of false and FGR.
(with an increased risk of perinatal positives, 70.2% [95% CI, 56e81.4] vs
complications) and low-risk SGA would 63.5% [95% CI, 58.4e68.3]. At second Ultrasound fetal biometry
be desirable. trimester, a combination of fetal biometry, A meta-analysis of randomized trials
estimated fetal weight (EFW) <10th cen- failed to demonstrate benefit from
The importance of diagnosing late tile, and abnormal uterine Doppler routine third-trimester scan.42 It may be
small fetuses velocimetry or cerebroplacental ratio argued that the older pooled data hold
An audit on 1543 cases of perinatal death (CPR) (which combines UA and middle limited contemporary validity. Technol-
(>28 weeks of gestational)29 revealed that cerebral artery [MCA] pulsatility indices) ogy and expertise in the 1970s and 1980s
a failure to detect FGR accounted for 10% only detected 20.1% of late SGA for 25% of do not translate legitimately into current
of the avoidable cases (those in which false positives.35 Similarly, in a large se- practice. The most recent study43 was
panel agreement was met that suboptimal ries36 of 8024 pregnancies (nulliparous dated 2003 and it claimed a 30%
care contributed to the fatal outcome), women) uterine Doppler velocimetry reduction in SGA. Some of studies also
although no stratification was provided (elevated pulsatility index) was evaluated relied on outdated surrogates of fetal
for late vs early FGR. A large study30 at 16-22þ6 weeks showing limited per- growth or formulas to EFW.44,45
reported a significantly increased risk of formance in predicting BW <5th centile: Furthermore, many of the studies
fetal death in SGA delivered >37 weeks (DR of 45.5% for 25% of false positives). involved no change in management if a
compared to those delivered in the 37th Incorporating maternal age, early preg- diagnosis of FGR was made, which does
week (47/10,000; 95% CI, 34.6e62.5 vs nancy body mass index, race/ethnicity, not reflect current practice.
21/10,000; 95% CI, 13.0e32.1; RR, 2.2; smoking status prior to pregnancy, A total of 13 series37,46-57 on routine
95% CI, 1.3e3.7). Another study chronic hypertension, and pregestational ultrasound screening that performed the
including 92,218 singletons found fetal diabetes in the prediction model resulted scan at a mean gestational age >32 weeks
death rates of 9.7 vs 18.9/1000 with in only modest improvements. Finally, have been published since 2012,
antenatally detected vs nondetected combining maternal characteristic, first- including a total of 22,927 pregnancies
FGR.31 Gestational age in both groups trimester blood pressure, and second- with 1776 SGA babies. SGA was variably
differed by only 10 days (270 vs 280 days), trimester biometrics and uterine Doppler defined as BW <10th centile or <5th
which underscores the relevance of velocimetry, 43.3% of DR for 10% of false centile. The summary receiver operating
detection and timely delivery. positives has been reported.37 Therefore, characteristic curve showed an area un-
early screening for late FGR is of limited der the curve of 88.2% (95% CI,
Early screening value. 85.4e91%) (Figure 1 and Appendix).
First- or second-trimester screening For a false-positive rate of 10%, the
with uterine Doppler velocimetry, Fundal height resulting DR was 70% (95% CI,
biochemical markers (angiogenic factors), In low-risk pregnancies, serial measure- 62e78%). Only in 1 study37 FGR (severe
and maternal characteristics may detect ments of symphysis-fundal height (SFH) smallness or abnormal Doppler

S792 American Journal of Obstetrics & Gynecology FEBRUARY 2018

ajog.org Expert Reviews

velocimetry) was addressed (including

FIGURE 1
1303 pregnancies and 83 FGR), showing
a DR of 74.4% (CI, 63.6e83.4%) for
Performance of routine ultrasound screening for the prediction of late SGA
10% of false positives. Sensitivity SROC Curve
1
The main questions for the use of
ultrasound in the detection of small Symmetric SROC
0,9 AUC = 0,8818
fetuses are: (1) whether it must be SE(AUC) = 0,0143
systematic or only focused on a selected Q* = 0,8123
SE(Q*) = 0,0145
0,8
population by high-risk factors; (2)
which is the optimal parameter, ie, EFW,
0,7
abdominal circumference (AC), or both;
and (3) what the optimal gestational age
0,6
would be.
0,5
Systematic vs selective ultrasound
Serial scanning is recommended in high-
0,4
risk pregnancies,38,39 based on expert
opinion. Although up to 25% of the 0,3
maternity population fall into the cate-
gory that would require serial scanning 0,2
based on risk factors,58 the DR for SGA
of such a strategy is <50%.59 A large 0,1
prospective study57 has been published
involving 3977 nulliparous women, in 0
0 0,2 0,4 0,6 0,8 1
which serial scanning was performed at 1-specificity
28, 32, and 36 weeks and the results were Summary receiver operating characteristic curve (SROC)* of studies on routine ultrasound for
concealed to participants and treating smallness for gestational age screening performed at mean gestational age >32 weeks. *Random
clinicians. This series reported that uni- (DerSimonian method) effects model: Walter SD. Properties of summary receiver operating char-
versal screening triples the DR of SGA as acteristic (SROC) curve for diagnostic test data. The methodology used to construct the data can be
compared with screening based on found in Stat Med 2002;21:1237-56.
clinical risk factors (from 69 [20%] of AUC, area under curve; SGA, small for gestational age.
352, to 199 [57%] of 352). Several factors Figueras. Late-onset fetal growth restriction. Am J Obstet Gynecol 2018.
could account for the variability of the
reported performances of third-
trimester ultrasound screening, such as 5% of false-positive rate the detection could be questionable. A recent ran-
baseline risk of the population, the for severe SGA (<3rd centile) requiring domized study56 compared in low-risk
gestational age at scan (being the per- delivery within 2 weeks of assessment pregnancies routine screening at 32
formance better late in the third (that could be considered a proxy for (n ¼ 1272) vs 36 (n ¼ 1314) weeks. For
trimester47,56), and the parameter used FGR) was 83.3% (95% CI, 67.2e93.6) FGR at birth (<10th customized centile
for growth assessment AC alone vs EFW. for a combination of fetal biometrics at birth), the false-positive rates for both
(AC, head circumference, and femur strategies were similar (6.4% vs 8.2%),
EFW vs AC length) plus maternal characteristics and but the DR were superior at 36 vs 32
While the American Congress of Ob- 91.7% (95% CI, 77.5e98.2) for EFW weeks’ gestation (38.8% vs 22.5%;
stetricians and Gynecologists supports plus maternal characteristics. Individual P ¼ .006). Likewise, for severe smallness
only the use of EFW <10th percentile, performance of each biometric measure (<3rd customized centile at birth), for a
the Royal College of Obstetricians and was not provided. similar rate of false positives (8.5% vs
Gynecologists supports the use of AC 8.7%), DR was superior at 36 vs 32
<10th percentile as an additional Optimal gestational age for screening weeks’ gestation (61.4% vs 32.5%;
criterion.38,39 Indeed, a recent system- A meta-analysis of randomized trials P ¼ .008). Two prospective studies46,62
atic review and meta-analysis on this failed to demonstrate benefit from carried out on the same population
issue60 found that AC is comparable to routine third-trimester scan.42 However, showed that while universal screening
EFW in predicting SGA. The largest of the included studies only 3 of (maternal characteristics plus EFW) at
prospective study46 (including 5515 them43,45,61 (contributing only 12% of 30-34 weeks had a DR of 65% (10% of
singleton pregnancies routinely scan- subjects overall) performed the scan >34 false positives) of 1727 babies born with a
ned at 35-37 weeks) favors EFW over weeks. Therefore, the relevance of this BW <5th centile, at 34-37 weeks the DR
individual biometric measures: for a meta-analysis to current clinical practice was 80% (10% of false positives) of 278.

FEBRUARY 2018 American Journal of Obstetrics & Gynecology S793

Expert Reviews ajog.org

SGA vs FGR smallness by itself could be seen as a parameters for predicting adverse
Once the diagnosis of a small fetus has stand-alone criterion for FGR. Fetuses outcome was not explored in this series.
been established, the next step is the with severe smallness are over- In a series75 of 308 nulliparous women
differential diagnosis between FGR and represented among the prenatally sus- who subsequently gave birth to
constitutional smallness. Indeed, SGA pected late SGA fetuses (w30%) because adequate-for-gestational age infants,
represent a heterogeneous population they are more likely to be detected reduced growth velocity between 28-36
that comprises fetuses that fail to achieve antenatally.7,57,68 weeks’ (EFW decline >30 centiles) was
their growth potential mainly due to found associated with abnormal CPR
placental insufficiency along with a Velocity of biometric parameters (RR, 2.80; 95% CI, 1.25e6.25) and
fraction of the small babies who are only Some reports69-72 in high-risk pop- nonsignificantly with UA pH <7.15
are constitutionally small (ie, they have a ulations have shown that defective lon- (RR, 2.34; 95% CI, 0.89e6.14). How-
low growth potential). Further gitudinal fetal growth is associated with ever, the predictive performance of
complexity is added by the fact that adverse perinatal outcome. In these se- reduced fetal growth was poor (for
histological63 and biochemical64 signs of ries, “high-risk” was variably defined by abnormal CPR, DR of 23% [95% CI,
placental insufficiency can occur in maternal characteristics (age >35 years, 9e43.7] for 8.2% of false positives; and
apparently uncomplicated pregnancies smoking), obstetric history (previous for pH <7.15, DR of 21% [95% CI,
with normal fetal/neonatal weight. FGR or hypertensive diseases), or EFW/ 6.1e45.6] for 9% of false positives).
The distinction between pathological AC <10th centile. Also recently, we reported on a cohort
(FGR) and constitutional smallness is Several recent studies used serial of 472 SGA-suspected fetuses that
clinically relevant because of the corre- ultrasound evaluations of fetal growth as longitudinal growth assessment from
lation with perinatal outcome. Whereas a strategy to improve diagnosis of FGR. diagnosis to delivery does not add to
FGR represents a pathological condition The first, conducted by Sovio et al57 in a other Doppler parameters in predicting
associated with adverse perinatal large cohort of 3977 unselected nullipa- adverse perinatal outcome.76 Indeed,
outcome, SGA babies are associated with rous women, found a significant associ- recent evidence on retrospective
near-normal perinatal outcomes. While ation of second- to third-trimester series77,78 shows that there is a correla-
conceptually FGR and SGA are different growth velocity to the occurrence of tion between growth velocity and the
conditions, their clinical differentiation adverse outcome. Upon stratification, umbilical/cerebral Doppler. However,
is difficult. the association remained only significant when the association to adverse out-
for those cases (n ¼ 560) with EFW comes has been addressed, only the
The diagnosis of severe SGA based on <10th centile (RR, 1.96; 95% CI, Doppler parameter remained indepen-
biometric parameters 1.21e3.19). The effect of SGA on dently associated.77
Pilliod et al65 analyzed a retrospective adverse outcome was significantly There is not a clear definition of slow
cohort of all births in the United States modified by the presence of low growth growth. A recent consensus on the defi-
during 2005 (n ¼ 3,349,816 non- velocity, to a greater extent than by the nition of late FGR2 proposed defining
anomalous singletons), as recorded in a presence of abnormal umbilical or slow growth when AC/EFW crosses 50
national database. An increase in the uterine Doppler velocimetry. centiles between 2 measurements (for
number of fetal deaths >37 weeks was On the other hand, Karlsen et al,73 in instance, from the 75th centile in a first
reported, mainly for those babies <3rd a prospective cohort of 211 pregnancies measurement to the 25th centile in sec-
centile: while the risk of fetal death (per at risk of SGA (24.6% of them with an ond measurement). However, it could be
10,000 at-risk pregnancies) at 36 and 37 EFW <5th centile), aimed at evaluating argued that this definition is arbitrary
weeks were 21.4 and 18.7, respectively, it whether the use of conditional growth and lacking interval time frame: the
was 23.2, 32.3, 32.4, and 58 at 38, 39, 40, centiles could improve the prediction of same amount of decline over a short
and 41 weeks, respectively. A prospective adverse outcome. This method uses a time interval would be seen as more
cohort of 132 term SGA fetuses with previous measurement to condition concerning. Despite fetal growth, veloc-
normal umbilical, uterine, and cerebral individualized ranges for the subse- ity assessment is an intuitive notion that
Doppler velocimetry found that very- quent measurement.74 Using a 5th is in keeping with how postnatal growth
low EFW centile (<3rd centile) pre- centile threshold, for adverse outcome, assessment is performed; the evidence
dicted a higher risk of adverse perinatal the specificity of 78% (95% CI, supporting its use is unclear.
outcome.66 In addition, the follow-up of 70e84%) using size centile as a pre-
a cohort of 292 late SGA babies found dictor was improved to 94% (95% CI, Doppler parameters
that severe smallness (<3rd centile) was 89e97%) when conditional growth There is a sizable body of evidence
the most predictive factor for neurolog- centile was added to the model, whereas showing that the UA velocimetry does
ical problems at 2 years of life as defined the sensitivity was not significantly not reliably reflect placental insufficiency
by an abnormal score in the Ages and changed (60% [95% CI, 49e69%] vs and does not reliably predict adverse
Stages Questionnaire (ASQ) (OR, 3.6; 39% [95% CI, 30e50%]). The combi- outcome in late-onset FGR.5,69,79 It is
95% CI, 1.5e8.8).67 Thus, severe nation of growth velocity with Doppler intriguing that while most cases of

S794 American Journal of Obstetrics & Gynecology FEBRUARY 2018

ajog.org Expert Reviews

late-onset SGA present histological signs accomplished in the majority of cases. growth within normal ranges. There is
of placental underperfusion (mainly Such an algorithm allows profiling the evidence showing that normally growing
vascular occlusion and villous hypopla- general population of late SGA fetuses in babies with abnormal CPR have a higher
sia),80 this is not reflected in the UA 2 risk-differing groups. While small frequency of placental insufficiency as
Doppler. One could speculate that the fetuses with moderate growth restriction determined by abnormal UtA Doppler.99
degree of the extension is what accounts (>3rd centile) and normal placental There is evidence coming from large
for this finding. Indeed, animal81 and function on both the fetal (normal CPR) retrospective series on pregnancies
mathematical82 experimental models of and maternal (normal uterine Doppler) attended in a tertiary center that
placental vessel obliteration have sug- sides are classified as low-risk SGA and abnormal CPR is associated with peri-
gested that UA Doppler becomes managed as constitutionally small babies, natal morbidity77,100 and mortality,101
abnormal only if an extensive part of the those with either severe smallness or independently of BW. Furthermore,
placenta is involved. Since the UA Doppler evidence of placental dysfunc- some smaller prospective series on low-
velocimetry is not a sensitive parameter tion are considered high-risk FGR. risk populations also suggest this
to detect late forms of FGR, other Among late FGR babies, further evidence notion.102 However, the clinical appli-
Doppler measures have been explored to is needed to determine what the contri- cation of these findings remains to be
reflect the fetal adaptation to placental bution is of each component of the high- determined. The largest prospective se-
insufficiency. risk FGR definition into the overall ries on unselected pregnancies failed to
The use of MCA Doppler in this performance. find predictive value of CPR adverse
setting is supported by recent studies Interestingly, a study comparing outcome when performed at 30-34103 or
that have demonstrated that 15-20% of routine induction at 37 weeks for all SGA 35-37104 weeks. Therefore, even if CPR is
term SGA fetuses with normal UA detected cases with selective induction a marker of placental insufficiency in-
Doppler have reduced impedance in only for those classified as FGR accord- dependent of size, the effectiveness of a
MCA blood flow, and that this sign is ing to the previously mentioned criteria strategy based on CPR assessment in the
associated with poorer perinatal (EFW <3rd centile, CPR and UtA) overall population is still to be proven.105
outcome83,84 and neurobehavior, both at showed not only less intervention (ce-
birth85 and at 2 years of age.86 Further- sarean delivery rate of 25% of 143 vs Antepartum surveillance
more, the CPR, which combines the 40% 138; P < .06) but also improved Contrary to early-onset FGR, late-onset
pulsatility index of the MCA and UA, has neonatal outcomes (composite neonatal FGR is not associated with a progression
been demonstrated to be more sensitive morbidity of 9% vs 22%; P < .01).94 In of hemodynamic changes, and fetuses
to hypoxia (defined as a reduced partial this series, the qualifying criteria for FGR only exceptionally display Doppler
pressure of oxygen [pO2] in the arterial were EFW <3rd centile in 35.9% (101/ changes in the UA or ductus venosus.
system) than its individual compo- 281), an abnormal uterine Doppler However, progression to severe fetal
nents87 and it correlates better with velocimetry in 31% (29/93), an deterioration and even fetal death can
adverse outcome.88-90 In addition to abnormal CPR in 19.6% (55/281), and occur rapidly. Evidence from animal
these brain Doppler parameters, first-trimester low pregnancy-associated models106 shows that prolongation of
abnormal UtA Doppler has been asso- plasma protein-A in the remaining 5% pregnancy renders to significant growth
ciated with an increased risk of intra- (14/281). restriction (14% fetal weight reduction)
partum fetal distress, emergency It is likely that in future years maternal and hypoxia (evaluated by magnetic
cesarean delivery, and admission to blood biomarkers are incorporated as a resonance imaging and immunohisto-
intensive care unit.84,91,92 diagnostic criterion of FGR in composite chemistry) because of differential loss of
algorithms, as a marker of placental placental mass rather than any
Combination of fetal biometry and involvement.95 Indeed, it has been compromise in fetoplacental blood
Doppler velocimetry shown in late-onset FGR that the flow. This finding validates the inability
A model combining severe smallness angiogenic factors correlate with of UA Doppler to safely monitor such
(ie, EFW <3rd centile) with Doppler placental findings secondary to fetuses.
parameters (CPR and UtA) remarkably underperfusion.96,97 This might be explained by reduced
improves the risk profiling (for neonatal tolerance to hypoxia of the term in
acidosis, 11.7% of 307 vs 5% of 202 in Can FGR be present in normally grown comparison with the preterm fetus and
SGA without criteria, P ¼ .009; for babies? the more common presence of uterine
cesarean delivery for nonreassuring fetal It is biologically implausible that all cases contractions. Therefore, the strategy in
status 29.3% vs 7.9%, P <.001)93 without of placental insufficiency occur in babies the management of late-onset FGR is
excessive technical sophistication; with BW <10th percentile. In fact, essentially based on risk assessment.
the estimation of the fetal weight is an perinatal mortality remains higher in Concerning surveillance, a relevant
integrated part of the third-trimester babies between the 10th-50th percentile notion is that the status low-risk vs high-
echography and Doppler interrogation of BW,98 suggesting that a proportion of risk, and consequently the recom-
of the UA, MCA, and UtA is easily cases of placental insufficiency exhibit mended management, can change after

FEBRUARY 2018 American Journal of Obstetrics & Gynecology S795

Expert Reviews ajog.org

an initial diagnosis, and for these reason phenomena of circulatory redistribu- review90 found that in fetuses with SGA
serial measurements of the biometrics tion mainly characterized by the born >32 weeks of gestational age, CPR
and Doppler are recommended. centralization of blood flow. The better (9 studies) adds value to assessment over
There is evidence from 1 randomized oxygenated blood goes toward the most MCA (8 studies) in predicting adverse
trial107 that when compared with vital organs (brain, heart, adrenals), outcomes. Therefore, the CPR could be
monitoring every 2 weeks, twice-a-week while vasoconstriction limits the blood’s seen as the primary surveillance tool in
monitoring results in more inductions arrival at the organs considered less late SGA. Several references have been
(82% of 70 vs 66% of 54; P ¼ .02), indispensable. This constellation of published on CPR and different cut-offs
without any improvement in the peri- changes is known as hemodynamic have been used in the literature.115 A
natal outcomes. Thus, the standard of redistribution. Among them, Doppler CPR calculator that included most used
care for those low-risk SGA would be MCA assessment has been the most references is available at: http://www.
this latter regimen. However, in late- widely studied parameter. Figure 2 ajog.org/pb/assets/raw/Health%20Advance/
onset FGR such definition of “low-risk shows the progression of changes in the journals/ymob/CPR/index.htm.
SGA” could not be reliably trusted on MCA.
the umbilical Doppler, because it does Longitudinal studies in late-onset Uterine Doppler
not reflect any progression from diag- SGA cases have shown that MCA Uterine Doppler is a noninvasive surro-
nosis to delivery.5 Therefore, some becomes abnormal in about 15% of gate of the placental function of the
other monitoring markers are needed. the cases.5 It has been shown83,113 that in maternal compartment. In late SGA it
near-term SGA fetuses, MCA could be has been found to be associated with a
Serial growth assessment useful to predict adverse outcome, higher frequency of placental sign of
Ultrasound growth assessment should not independently of the UA Doppler. maternal underperfusion.116 Figure 3
be performed more frequently than every Remarkably, a study where clinicians shows the spectrum of abnormalities in
2 weeks because the inherent error were blinded to the Doppler indices the UtA waveform.
associated with ultrasonographic mea- compared SGA babies with normal and Interestingly, about a third of preg-
surements can preclude an accurate abnormal MCA, reporting 6 times as nancies with abnormal third trimester
assessment of growth.108 The advantage of many instances of emergent cesarean uterine Doppler had normal values at the
longitudinal over cross-sectional assess- delivery for fetal distress (29% vs 4.8%; beginning of the pregnancy, and this
ment has not been clearly demonstrated. P < .001) and a 3-fold increased risk of group still has an exceedingly high inci-
neonatal metabolic acidosis (7.6% vs dence (w30%) of placenta-related dis-
Amniotic fluid volume 2.4%; P ¼ .03) at labor induction.114 eases.117 This suggests that the uterine
In a large randomized controlled trial This is a relevant issue because accord- Doppler has the potential advantage of
on late SGA,109 one third had oligohy- ing current guidelines,38,39 labor induc- capturing placental insufficiency from
dramnios, as defined by an amniotic tion at term is the current standard of differing pathways: that resulting from
fluid index <5. However, compared care of late-onset SGA. Therefore, while defective trophoblastic invasion early in
with the single deepest vertical pocket, an association exists between abnormal pregnancy, but also that emerging late in
the amniotic fluid index results in an MCA and adverse outcome that would pregnancy and probably related to other
overdiagnosis of oligohydramnios.110 In justify its use as surveillance tool, this pathological mechanisms.
a prospective series of pregnancies with sign is rather a late manifestation, with At diagnosis, late SGA with abnormal
oligohydramnios and FGR,111 the fetal acceptable specificity but low sensitivity uterine Doppler has a 2-fold increased
growth centile remained stable over the for clinical applicability.90 risk (62.7% vs 34.6%; P < .01) of
8 weeks following the diagnosis, sug- Due to increased impedance in the developing abnormal brain Doppler
gesting that the oligohydramnios is not placental vasculature in combination indices before induction of labor.118
reflecting any risk of progression. with a decrease in cerebral resistance This may be of value in planning the
Furthermore, a meta-analysis112 of 18 secondary to vasodilation, the ratio be- timing of fetal surveillance. In addition,
trials showed an association with tween the MCA and UA (ie, the CPR) is serial uterine Doppler assessment as a
abnormal 5-minute Apgar, but not with decreased even with UA and MCA values surveillance tool has doubtful value
acidosis or perinatal death in SGA very close to normal.89 Consistently, because longitudinal studies fail to show
(RR, 1.6; 95% CI, 0.9e2.6). Because of animal models demonstrated that this any progression from diagnosis to
the limited evidence, the inclusion of ratio is better correlated with hypoxia delivery.5
oligohydramnios assessment in man- than its individual components.87 In late
agement protocols of SGA/FGR is not SGA fetuses, CPR becomes abnormal in Timing and mode of delivery
recommended. w20% of cases.5,94 The CPR improves Several guidelines recommend delivery at
the sensitivity of UA and MCA alone, 37-38 weeks.38,39 This recommendation
Brain Doppler because it is already decreased when its is based on the findings of the DIGITAT
Hypoxemia secondary to placental individual components are still within study,109 in which 650 women with
insufficiency sets into motion normal ranges.87,88 A recent systematic SGA >36 weeks were randomized to

S796 American Journal of Obstetrics & Gynecology FEBRUARY 2018

ajog.org Expert Reviews

induction or expectant management. In-

FIGURE 2
duction group infants were delivered
10 days earlier (266 vs 277 days; mean
Middle cerebral artery Doppler changes
difference e9.9 days; 95% CI, e11.3 to
e8.6) and weighed 130 g less (mean dif-
ference e130 g; 95% CI, e188 to e71 g)
than babies in the expectant monitoring
group. There was no difference in the
perinatal outcomes except that women in
the expectant group had a 2-fold increase
in risk of developing preeclampsia (3.7%
of 321 vs 7.9% of 329; mean difference
e4.2; 95% CI, e7.7 to e0.6). A total of
17/321 (5.3%) infants in the induction
group experienced the composite adverse
neonatal outcome, compared with 20/
329 (6.1%) in the expectant monitoring
group (difference e0.8%, 95% CI, e4.3
to 3.2%). This was defined as 5-minute
Apgar <7, UA pH <7.05, or admission
to neonatal intensive care. The prevalence
of these outcomes was 9/650 (1.4%),
14/567 (2.5%), and 22/650 (3.4%),
respectively.
In a secondary analysis,119 a higher
percentage of neonatal admissions was
found after induction <38 weeks’
gestational age: 125 (61%) admissions
vs 92 (44%) after expectant manage-
ment, difference 16% (95% CI,
6.7e26%; P ¼ .001). This suggests that
if induction is considered, it is reason-
able to delay until 38 weeks. Cesarean
deliveries were performed on 45
(14.0%) mothers in the induction group
and 45 (13.7%) in the expectant
monitoring group (difference 0.3%,
95% CI, e5.0 to 5.6%).
After a response rate of w50%, 2-year
evaluation was performed of neuro-
development (ASQ) and neurobehavior
(Child Behavior Checklist [CBCL]). A
total of 27% of 274 infants had an
abnormal score on the ASQ and 13% of
265 on the CBCL. Results of the ASQ and
the CBCL did no differ between expec-
tant and induction arms.67 The most
Color Doppler assessment of middle cerebral artery (MCA) A, at level of circle of Willis and flow
predictive factors for abnormal ASQ was
velocity waveforms B, in control fetus with normal waveform and C, in growth-restricted fetus with
a BW <2.3 centile (OR, 3.6; 95% CI,
high diastolic velocities and decreased pulsatility index.
1.5e8.8).
ACA, anterior cerebral artery; PCA, posterior cerebral artery
A health economics analysis120 Figueras. Late-onset fetal growth restriction. Am J Obstet Gynecol 2018.
demonstrated that both strategies (in-
duction vs expectant management)
generated comparable costs: on average expectant management group (N ¼ 329) In the DIGITAT study all SGA fetuses
V7106 per patient for the induction with a cost difference of V111 (95% CI, were managed under a common proto-
group (N ¼ 321) and V6995 for the V1296e1641). col without any attempt to differentiate

FEBRUARY 2018 American Journal of Obstetrics & Gynecology S797

Expert Reviews ajog.org

delivery of high-risk FGR at 38 weeks is

FIGURE 3
justified, a more expectant management
Uterine artery Doppler changes could be offered to low-risk SGA.
An observational study94 compared a
strategy of systematic induction of late
SGA at 37 weeks (n ¼ 138) with risk
stratification and induction indicated by
severe smallness, abnormal CPR, or
abnormal uterine Doppler velocimetry
(n ¼ 143). The incidence of neonatal
composite adverse outcomes was lower
after selective induction (9% vs 22%;
P < .01). This was defined as the
presence of at least 1 of the following
conditions: Apgar <7 at 5 minutes, cord
arterial pH <7.10, hypoglycemia (blood
glucose <2.5 mmol/L), and ventilation.
The individual prevalence of these
outcomes was 3/281 (1.1%), 5/98 (5%),
25/240 (10.4%), and 28/281 (10%),
respectively. Furthermore, neonatal
admission was also more frequent after
systematic induction (13% v 42%;
P < .01). Furthermore, cesarean delivery
rates (25% vs 40%; P ¼ .06) were lower
after selective induction. This study
suggests that protocol-based manage-
ment of SGA babies may improve out-
comes and that identification of
moderate SGA should not alone prompt
delivery.

Conclusions
First- or second-trimester screening
strategies provide limited DR for late
SGA. At third trimester, universal
screening triples the DR of late SGA
compared to clinically indicated scan-
ning. As opposed to early third-trimester
ultrasound, scanning late in pregnancy
(around 37 weeks) increases the DR for
BW <3rd and 10th centile.
UA Doppler velocimetry alone does
not provide good differentiation
between low-risk SGA and high-risk
FGR. A combination of biometric pa-
rameters (with severe smallness usually
defined as EFW or AC <3rd centile)
with Doppler criteria of placental
Site of insonation of uterine artery with color Doppler A, at crossover of iliac artery. B, Normal and C,
insufficiency (either in the maternal
abnormal (increased impedance to flow with early diastolic notching) waveforms.
[uterine Doppler] or fetal [CPR] com-
Figueras. Late-onset fetal growth restriction. Am J Obstet Gynecol 2018.
partments) offers a classification tool
that correlates with the risk for adverse
between low-risk SGA and high-risk parameters allows profiling a subgroup perinatal outcome. For surveillance
FGR. As discussed above, a combina- of fetuses that concentrates most in- purposes, CPR is sensitive to reflect
tion of biometric and Doppler stances of adverse outcomes.10 While progression from diagnosis until term.

S798 American Journal of Obstetrics & Gynecology FEBRUARY 2018

ajog.org Expert Reviews

There is no evidence that induction of uncomplicated pregnancies. Obstet Gynecol newborn health. Am J Obstet Gynecol
late FGR at term improves the perinatal 2017;130:511-9. 1999;181:701-8.
13. Blair EM, Nelson KB. Fetal growth restriction 28. Parra-Saavedra M, Crovetto F, Triunfo S,
outcomes nor is it a cost-effective strat- and risk of cerebral palsy in singletons born after et al. Neurodevelopmental outcomes of near-
egy, and it may increase neonatal admis- at least 35 weeks’ gestation. Am J Obstet term small-for-gestational-age infants with and
sion when performed <38 weeks. - Gynecol 2015;212:520.e1-7. without signs of placental underperfusion.
14. Sanz-Cortes M, Egana-Ugrinovic G, Placenta 2014;35:269-74.
Simoes RV, Vazquez L, Bargallo N, Gratacos E. 29. Richardus JH, Graafmans WC, Verloove-
Association of brain metabolism with sulcation Vanhorick SP, Mackenbach JP; EuroNatal
REFERENCES and corpus callosum development assessed by International Audit Panel; EuroNatal Working
1. Savchev S, Figueras F, Sanz-Cortes M, et al. MRI in late-onset small fetuses. Am J Obstet Group. Differences in perinatal mortality and
Evaluation of an optimal gestational age cut-off Gynecol 2015;212:804.e1-8. suboptimal care between 10 European regions:
for the definition of early- and late-onset fetal 15. Simoes RV, Cruz-Lemini M, Bargallo N, results of an international audit. BJOG
growth restriction. Fetal Diagn Ther 2014;36: Gratacos E, Sanz-Cortes M. Brain metabolite 2003;110:97-105.
99-105. differences in one-year-old infants born small at 30. Trudell AS, Cahill AG, Tuuli MG,
2. Gordijn SJ, Beune IM, Thilaganathan B, et al. term and association with neurodevelopmental Macones GA, Odibo AO. Risk of stillbirth after 37
Consensus definition for placental fetal growth outcome. Am J Obstet Gynecol 2015;213:210. weeks in pregnancies complicated by small-for-
restriction: a Delphi procedure. Ultrasound e1-11. gestational-age fetuses. Am J Obstet Gynecol
Obstet Gynecol 2016;48:333-9. 16. Arcangeli T, Thilaganathan B, Hooper R, 2013;208:376.e1-7.
3. Cosmi E, Ambrosini G, D’Antona D, Saccardi C, Khan KS, Bhide A. Neurodevelopmental 31. Gardosi J, Madurasinghe V, Williams M,
Mari G. Doppler, cardiotocography, and biophys- delay in small babies at term: a systematic Malik A, Francis A. Maternal and fetal risk factors
ical profile changes in growth-restricted fetuses. review. Ultrasound Obstet Gynecol 2012;40: for stillbirth: population based study. BMJ
Obstet Gynecol 2005;106:1240-5. 267-75. 2013;346:f108.
4. Crimmins S, Desai A, Block-Abraham D, 17. Lindstrom L, Wikstrom AK, Bergman E, 32. Melchiorre K, Leslie K, Prefumo F, Bhide A,
Berg C, Gembruch U, Baschat AA. Lundgren M. Born small for gestational age and Thilaganathan B. First-trimester uterine artery
A comparison of Doppler and biophysical find- poor school performanceehow small is too Doppler indices in the prediction of small-for-
ings between liveborn and stillborn growth- small? Horm Res Paediatr 2017;88:215-23. gestational age pregnancy and intrauterine
restricted fetuses. Am J Obstet Gynecol 18. Newsome CA, Shiell AW, Fall CH, Phillips DI, growth restriction. Ultrasound Obstet Gynecol
2014;211:669.e1-10. Shier R, Law CM. Is birth weight related to later 2009;33:524-9.
5. Oros D, Figueras F, Cruz-Martinez R, Meler E, glucose and insulin metabolism? A systematic 33. Poon LC, Syngelaki A, Akolekar R, Lai J,
Munmany M, Gratacos E. Longitudinal changes review. Diabet Med 2003;20:339-48. Nicolaides KH. combined screening for
in uterine, umbilical and fetal cerebral Doppler 19. Katanoda K, Noda M, Goto A, Mizunuma H, preeclampsia and small for gestational age
indices in late-onset small-for-gestational age Lee JS, Hayashi K. Impact of birth weight on at 11-13 weeks. Fetal Diagn Ther 2013;33:
fetuses. Ultrasound Obstet Gynecol 2011;37: adult-onset diabetes mellitus in relation to cur- 16-27.
191-5. rent body mass index: the Japan nurses’ health 34. Crovetto F, Crispi F, Scazzocchio E, et al.
6. Zhu MY, Milligan N, Keating S, et al. The he- study. J Epidemiol 2017;27:428-34. First-trimester screening for early and late small-
modynamics of late-onset intrauterine growth 20. Apel-Sarid L, Levy A, Holcberg G, Sheiner E. for-gestational-age neonates using maternal
restriction by MRI. Am J Obstet Gynecol Term and preterm (<34 and <37 weeks serum biochemistry, blood pressure and uterine
2016;214:367.e1-17. gestation) placental pathologies associated with artery Doppler. Ultrasound Obstet Gynecol
7. Crovetto F, Triunfo S, Crispi F, et al. First- fetal growth restriction. Arch Gynecol Obstet 2016;43:34-40.
trimester screening with specific algorithms for 2009;282:487-92. 35. Hernandez-Andrade E, Maymon E, Erez O,
early- and late-onset fetal growth restriction. 21. Salafia CM, Vintzileos AM, Silberman L, et al. A low cerebroplacental ratio at 20-24
Ultrasound Obstet Gynecol 2016;48:340-8. Bantham KF, Vogel CA. Placental pathology of weeks of gestation can predict reduced fetal
8. Lackman F, Capewell V, Gagnon R, idiopathic intrauterine growth retardation at size later in pregnancy or at birth. Fetal Diagn
Richardson B. Fetal umbilical cord oxygen term. Am J Perinatol 1992;9:179-84. Ther 2017 Sep 20. https://doi.org/10.1159/
values and birth to placental weight ratio in 22. Stanek J. Comparison of placental pathol- 000479684. [Epub ahead of print].
relation to size at birth. Am J Obstet Gynecol ogy in preterm, late-preterm, near-term, and 36. Parry S, Sciscione A, Haas DM, et al. Role of
2001;185:674-82. term births. Am J Obstet Gynecol 2013;210: early second-trimester uterine artery Doppler
9. Carreno CA, Costantine MM, Holland MG, 234.e1-6. screening to predict small-for-gestational-age
Ramin SM, Saade GR, Blackwell SC. Approxi- 23. Kovo M, Schreiber L, Ben-Haroush A, et al. babies in nulliparous women. Am J Obstet
mately one-third of medically indicated late The placental factor in early- and late-onset Gynecol 2017;217:594.e1-10.
preterm births are complicated by fetal growth normotensive fetal growth restriction. Placenta 37. Triunfo S, Crovetto F, Scazzocchio E, Parra-
restriction. Am J Obstet Gynecol 2011;204:263. 2013;34:320-4. Saavedra M, Gratacos E, Figueras F. Contingent
e1-4. 24. Mifsud W, Sebire NJ. Placental pathology in versus routine third-trimester screening for late
10. Figueras F, Gardosi J. Intrauterine growth early-onset and late-onset fetal growth restric- fetal growth restriction. Ultrasound Obstet
restriction: new concepts in antenatal surveil- tion. Fetal Diagn Ther 2014;36:117-28. Gynecol 2016;47:81-8.
lance, diagnosis, and management. Am J 25. Redline RW. Classification of placental le- 38. American College of Obstetricians and Gy-
Obstet Gynecol 2011;204:288-300. sions. Am J Obstet Gynecol 2015;213:S21-8. necologists. Fetal growth restriction. ACOG
11. Mendez-Figueroa H, Truong VT, Pedroza C, 26. Parra-Saavedra M, Simeone S, Practice bulletin no. 134. Obstet Gynecol
Khan AM, Chauhan SP. Small-for-gestational- Triunfo S, et al. Correlation between histo- 2013;121:1122-33.
age infants among uncomplicated pregnancies logical signs of placental underperfusion and 39. Royal College of Obstetricians and
at term: a secondary analysis of 9 Maternal-Fetal perinatal morbidity in late-onset small-for- Gynecologists. The investigation and manage-
Medicine Units Network studies. Am J Obstet gestational-age fetuses. Ultrasound Obstet ment of the small-for-gestational age fetus.
Gynecol 2016;215:628.e1-7. Gynecol 2014;45:149-55. Greenetop Guideline No 31. 2013. Available at:
12. Chauhan SP, Rice MM, Grobman WA, et al. 27. Verma A, Okun NB, Maguire TO, https://www.rcog.org.uk/en/guidelines-research-
Neonatal morbidity of small- and large-for- Mitchell BF. Morbidity assessment index for services/guidelines/gtg31/. Accessed December
gestational-age neonates born at term in newborns: a composite tool for measuring 21, 2017.

FEBRUARY 2018 American Journal of Obstetrics & Gynecology S799

Expert Reviews ajog.org

40. Lindhard A, Nielsen PV, Mouritsen LA, 53. Simcox LE, Myers JE, Cole TJ, underperfusion: a longitudinal case-cohort study.
Zachariassen A, Sorensen HU, Roseno H. Johnstone ED. Fractional fetal thigh volume Am J Obstet Gynecol 2015;214:629.e1-17.
The implications of introducing the in the prediction of normal and abnormal 65. Pilliod RA, Cheng YW, Snowden JM,
symphyseal-fundal height-measurement. A fetal growth during the third trimester of Doss AE, Caughey AB. The risk of intrauterine
prospective randomized controlled trial. Br J pregnancy. Am J Obstet Gynecol 2017;217: fetal death in the small-for-gestational-age
Obstet Gynaecol 1990;97:675-80. 453.e1-12. fetus. Am J Obstet Gynecol 2012;207:318.
41. Goto E. Prediction of low birthweight and 54. Skrastad RB, Eik-Nes SH, Sviggum O, et al. e1-6.
small for gestational age from symphysis-fundal A randomized controlled trial of third-trimester 66. Savchev S, Figueras F, Cruz-Martinez R,
height mainly in developing countries: a meta- routine ultrasound in a non-selected popula- Illa M, Botet F, Gratacos E. Estimated weight
analysis. J Epidemiol Community Health tion. Acta Obstet Gynecol Scand 2013;92: centile as a predictor of perinatal outcome in
2013;67:999-1005. 1353-60. small-for-gestational-age pregnancies with
42. Bricker L, Medley N, Pratt JJ. Routine 55. Sokol Karadjole V, Agarwal U, normal fetal and maternal Doppler indices.
ultrasound in late pregnancy (after 24 weeks’ Berberovic E, Poljak B, Alfirevic Z. Does serial Ultrasound Obstet Gynecol 2011;39:
gestation). Cochrane Database Syst Rev 3rd trimester ultrasound improve detection of 299-303.
2015;6:CD001451. small for gestational age babies: comparison of 67. van Wyk L, Boers KE, van der Post JA, et al.
43. McKenna D, Tharmaratnam S, Mahsud S, screening policies in 2 European maternity Effects on (neuro)developmental and behavioral
Bailie C, Harper A, Dornan J. A randomized trial units. Eur J Obstet Gynecol Reprod Biol outcome at 2 years of age of induced labor
using ultrasound to identify the high-risk fetus in 2017;215:45-9. compared with expectant management in in-
a low-risk population. Obstet Gynecol 56. Roma E, Arnau A, Berdala R, Bergos C, trauterine growth-restricted infants: long-term
2003;101:626-32. Montesinos J, Figueras F. Ultrasound screening outcomes of the DIGITAT trial. Am J Obstet
44. Bakketeig LS, Eik-Nes SH, Jacobsen G, for fetal growth restriction at 36 vs 32 weeks’ Gynecol 2012;206:406.e1-7.
et al. Randomized controlled trial of ultrasono- gestation: a randomized trial (ROUTE). Ultra- 68. Fratelli N, Valcamonico A, Prefumo F,
graphic screening in pregnancy. Lancet 1984;2: sound Obstet Gynecol 2015;46:391-7. Pagani G, Guarneri T, Frusca T. Effects of
207-11. 57. Sovio U, White IR, Dacey A, antenatal recognition and follow-up on perinatal
45. Neilson JP, Munjanja SP, Whitfield CR. Pasupathy D, Smith GC. Screening for fetal outcomes in small-for-gestational age infants
Screening for small for dates fetuses: a growth restriction with universal third delivered after 36 weeks. Acta Obstet Gynecol
controlled trial. Br Med J (Clin Res Ed) 1984;289: trimester ultrasonography in nulliparous Scand 2012;92:223-9.
1179-82. women in the Pregnancy Outcome Predic- 69. Chang TC, Robson SC, Spencer JA,
46. Fadigas C, Saiid Y, Gonzalez R, Poon LC, tion (Pop) study: a prospective cohort study. Gallivan S. Prediction of perinatal morbidity at
Nicolaides KH. Prediction of small-for- Lancet 2015;386:2089-97. term in small fetuses: comparison of fetal growth
gestational-age neonates: screening by fetal 58. Gardosi J, Williams M. Serial ultrasound and Doppler ultrasound. Br J Obstet Gynaecol
biometry at 35-37 weeks. Ultrasound Obstet scanning in pregnancies at risk of intrauterine 1994;101:422-7.
Gynecol 2015;45:559-65. growth restriction. Birmingham: Perinatal Insti- 70. De Jong CL, Francis A, Van Geijn HP,
47. Souka AP, Papastefanou I, Pilalis A, tute; 2014. Gardosi J. Customized fetal weight limits for
Michalitsi V, Panagopoulos P, Kassanos D. 59. McCowan LM, Thompson JM, Taylor RS, antenatal detection of fetal growth restriction.
Performance of the ultrasound examination in et al. Prediction of small for gestational age in- Ultrasound Obstet Gynecol 2000;15:36-40.
the early and late third trimester for the prediction fants in healthy nulliparous women using clinical 71. Smith-Bindman R, Chu PW, Ecker JL,
of birth weight deviations. Prenat Diagn and ultrasound risk factors combined with early Feldstein VA, Filly RA, Bacchetti P. US evaluation
2013;33:915-20. pregnancy biomarkers. PLoS One 2017;12: of fetal growth: prediction of neonatal outcomes.
48. Tarca AL, Hernandez-Andrade E, Ahn H, e0169311. Radiology 2002;223:153-61.
et al. Single and serial fetal biometry to detect 60. Blue NR, Yordan JMP, Holbrook BD, 72. Barker ED, McAuliffe FM, Alderdice F, et al.
preterm and term small- and large-for- Nirgudkar PA, Mozurkewich EL. Abdominal The role of growth trajectories in classifying fetal
gestational-age neonates: a longitudinal cohort circumference alone versus estimated fetal growth restriction. Obstet Gynecol 2013;122:
study. PLoS One 2016;11:e0164161. weight after 24 weeks to predict small or large for 248-54.
49. Hammad IA, Chauhan SP, Mlynarczyk M, gestational age at birth: a meta-analysis. Am J 73. Karlsen HO, Johnsen SL, Rasmussen S,
et al. Uncomplicated pregnancies and ultra- Perinatol 2017;34:1115-24. Kiserud T. Prediction of adverse perinatal
sounds for fetal growth restriction: a pilot 61. Duff GB. A randomized controlled trial in a outcome of small-for-gestational-age preg-
randomized clinical trial. AJP Rep 2016;6: hospital population of ultrasound measurement nancy using size centiles and conditional growth
e83-90. screening for the small for dates baby. Aust N Z J centiles. Ultrasound Obstet Gynecol 2016;48:
50. Papastefanou I, Souka AP, Eleftheriades M, Obstet Gynaecol 1993;33:374-8. 217-23.
Pilalis A, Chrelias C, Kassanos D. Predicting fetal 62. Bakalis S, Silva M, Akolekar R, Poon LC, 74. Owen P, Ogston S. Conditional centiles for
growth deviation in parous women: combining Nicolaides KH. Prediction of small-for- the quantification of fetal growth. Ultrasound
the birth weight of the previous pregnancy and gestational-age neonates: screening by fetal Obstet Gynecol 1998;11:110-7.
third trimester ultrasound scan. J Perinat Med biometry at 30-34 weeks. Ultrasound Obstet 75. MacDonald TM, Hui L, Tong S, et al.
2015;43:485-92. Gynecol 2015;45:551-8. Reduced growth velocity across the third
51. Rad S, Beauchamp S, Morales C, 63. Wright E, Audette MC, Ye XY, et al. Maternal trimester is associated with placental insuffi-
Mirocha J, Esakoff TF. Defining fetal growth re- vascular malperfusion and adverse perinatal ciency in fetuses born at a normal birthweight: a
striction: abdominal circumference as an alter- outcomes in low-risk nulliparous women. Obstet prospective cohort study. BMC Med 2017;15:
native criterion. J Matern Fetal Neonatal Med Gynecol 2017;130:1112-20. 164.
2017 Aug 17:1-6 [Epub ahead of print]. 64. Korzeniewski SJ, Romero R, 76. Caradeux J, Eixarch E, Mazarico E,
52. Reboul Q, Delabaere A, Luo ZC, et al. Pre- Chaiworapongsa T, et al. Maternal plasma Basuki TR, Gratacos E, Figueras F. Longitudinal
diction of small-for-gestational-age neonate by angiogenic index-1 (placental growth factor/sol- growth assessment for the prediction of adverse
third-trimester fetal biometry and impact of uble vascular endothelial growth factor receptor- perinatal outcome in SGA-suspected fetuses.
ultrasound-delivery interval. Ultrasound Obstet 1) is a biomarker for the burden of placental le- Ultrasound Obstet Gynecol 2017 Aug 7 [Epub
Gynecol 2017;49:372-8. sions consistent with uteroplacental ahead of print].

S800 American Journal of Obstetrics & Gynecology FEBRUARY 2018

ajog.org Expert Reviews

77. Khalil A, Morales-Rosello J, Khan N, et al. Is 90. Schreurs CA, de Boer MA, Heymans MW, population. Fetal Diagn Ther 2017 Jul 27 [Epub
cerebroplacental ratio a marker of impaired fetal et al. Prognostic accuracy of cerebroplacental ahead of print].
growth velocity and adverse pregnancy ratio and middle cerebral artery Doppler for 103. Bakalis S, Akolekar R, Gallo DM,
outcome? Am J Obstet Gynecol 2017;216:606. adverse perinatal outcomes: a systematic Poon LC, Nicolaides KH. Umbilical and fetal
e1-10. review and meta-analysis. Ultrasound Obstet middle cerebral artery Doppler at 30-34
78. Regan J, Masters H, Warshak CR. Estima- Gynecol 2017 Jul 14 [Epub ahead of print]. weeks’ gestation in the prediction of adverse
tion of the growth rate in fetuses with an 91. Ghosh GS, Gudmundsson S. Uterine and perinatal outcome. Ultrasound Obstet Gynecol
abnormal cerebroplacental ratio compared to umbilical artery Doppler are comparable in pre- 2015;45:409-20.
those with suspected growth restriction without dicting perinatal outcome of growth-restricted 104. Akolekar R, Syngelaki A, Gallo DM,
evidence of centralization of blood flow. fetuses. BJOG 2009;116:424-30. Poon LC, Nicolaides KH. Umbilical and fetal
J Ultrasound Med 2015;34:837-42. 92. Vergani P, Roncaglia N, Andreotti C, et al. middle cerebral artery Doppler at 35-37 weeks’
79. Vergani P, Andreotti C, Roncaglia N, et al. Prognostic value of uterine artery Doppler gestation in the prediction of adverse perinatal
Doppler predictors of adverse neonatal velocimetry in growth-restricted fetuses deliv- outcome. Ultrasound Obstet Gynecol 2015;46:
outcome in the growth restricted fetus at 34 ered near term. Am J Obstet Gynecol 2002;187: 82-92.
weeks’ gestation or beyond. Am J Obstet 932-6. 105. Figueras F, Gratacos E, Rial M, et al.
Gynecol 2003;189:1007-11. 93. Figueras F, Savchev S, Triunfo S, Revealed versus concealed criteria for placental
80. Parra-Saavedra M, Crovetto F, Triunfo S, Crovetto F, Gratacos E. An integrated model insufficiency in an unselected obstetric popula-
et al. Placental findings in late-onset SGA births with classification criteria to predict small-for- tion in late pregnancy (RATIO37): randomized
without Doppler signs of placental insufficiency. gestational fetuses at risk of adverse perinatal controlled trial study protocol. BMJ Open
Placenta 2013;34:1136-41. outcome. Ultrasound Obstet Gynecol 2015;45: 2017;7:e014835.
81. Morrow RJ, Adamson SL, Bull SB, 279-85. 106. Rahman A, Cahill LS, Zhou YQ, et al.
Ritchie JW. Effect of placental embolization on 94. Veglia M, Cavallaro A, Papageorghiou A, A mouse model of antepartum stillbirth. Am J
the umbilical arterial velocity waveform in fetal Black R, Impey L. Small for gestational Obstet Gynecol 2017;217:443.e1-11.
sheep. Am J Obstet Gynecol 1989;161: age babies after 37 weeks: an impact study 107. McCowan LM, Harding JE, Roberts AB,
1055-60. of a risk stratification protocol. Ultrasound Barker SE, Ford C, Stewart AW. A pilot ran-
82. Thompson RS, Stevens RJ. Mathematical Obstet Gynecol 2017 Jun 10 [Epub ahead of domized controlled trial of two regimens of fetal
model for interpretation of Doppler velocity print]. surveillance for small-for-gestational-age fe-
waveform indices. Med Biol Eng Comput 95. Aviram R, T BS, Kidron D. Placental etiol- tuses with normal results of umbilical artery
1989;27:269-76. ogies of fetal growth restriction: clinical and Doppler velocimetry. Am J Obstet Gynecol
83. Hershkovitz R, Kingdom JC, Geary M, pathological differences. Early Hum Dev 2000;182:81-6.
Rodeck CH. Fetal cerebral blood flow redistri- 2010;86:59-63. 108. Mongelli M, Ek S, Tambyrajia R. Screening
bution in late gestation: identification of 96. Benton SJ, Hu Y, Xie F, et al. Angiogenic for fetal growth restriction: a mathematical
compromise in small fetuses with normal um- factors as diagnostic tests for preeclampsia: a model of the effect of time interval and
bilical artery Doppler. Ultrasound Obstet Gyne- performance comparison between two com- ultrasound error. Obstet Gynecol 1998;92:
col 2000;15:209-12. mercial immunoassays. Am J Obstet Gynecol 908-12.
84. Severi FM, Bocchi C, Visentin A, et al. 2011;205:469.e1-8. 109. Boers KE, Vijgen SM, Bijlenga D, et al. In-
Uterine and fetal cerebral Doppler predict the 97. Triunfo S, Lobmaier S, Parra-Saavedra M, duction versus expectant monitoring for intra-
outcome of third-trimester small-for-gesta- et al. Angiogenic factors at diagnosis of late- uterine growth restriction at term: randomized
tional age fetuses with normal umbilical artery onset small-for-gestational age and histological equivalence trial (DIGITAT). BMJ 2010;341:
Doppler. Ultrasound Obstet Gynecol 2002;19: placental underperfusion. Placenta 2014;35: c7087.
225-8. 398-403. 110. Nabhan AF, Abdelmoula YA. Amniotic fluid
85. Cruz-Martinez R, Figueras F, Oros D, et al. 98. Vasak B, Koenen SV, Koster MP, et al. index versus single deepest vertical pocket as a
Cerebral blood perfusion and neurobehavioral Human fetal growth is constrained below screening test for preventing adverse pregnancy
performance in full-term small-for-gestational- optimal for perinatal survival. Ultrasound Obstet outcome. Cochrane Database Syst Rev 2008;3:
age fetuses. Am J Obstet Gynecol 2009;201: Gynecol 2015;45:162-7. CD006593.
474.e1-7. 99. Khalil A, Morales-Rosellõ J, Townsend R, 111. Zhang J, Troendle J, Meikle S,
86. Eixarch E, Batalle D, Illa M, et al. Neonatal et al. Value of third-trimester cerebroplacental Klebanoff MA, Rayburn WF. Isolated oligohy-
neurobehavior and diffusion MRI changes in ratio and uterine artery Doppler indices as pre- dramnios is not associated with adverse peri-
brain reorganization due to intrauterine growth dictors of stillbirth and perinatal loss. Ultrasound natal outcomes. BJOG 2004;111:220-5.
restriction in a rabbit model. PLoS One 2012;7: Obstet Gynecol 2016;47:74-80. 112. Chauhan SP, Sanderson M, Hendrix NW,
e31497. 100. Khalil AA, Morales-Rosello J, Morlando M, Magann EF, Devoe LD. Perinatal outcome and
87. Arbeille P, Maulik D, Fignon A, et al. et al. Is fetal cerebroplacental ratio an indepen- amniotic fluid index in the antepartum and
Assessment of the fetal Po2 changes by cere- dent predictor of intrapartum fetal compromise intrapartum periods: a meta-analysis. Am J
bral and umbilical Doppler on lamb fetuses and neonatal unit admission? Am J Obstet Obstet Gynecol 1999;181:1473-8.
during acute hypoxia. Ultrasound Med Biol Gynecol 2015;213:54.e1-10. 113. Eixarch E, Meler E, Iraola A, et al. Neuro-
1995;21:861-70. 101. Khalil A, Morales-Rosello J, Townsend R, developmental outcome in 2-year-old infants
88. Bahado-Singh RO, Kovanci E, Jeffres A, et al. Value of third-trimester cerebroplacental who were small-for-gestational age term fetuses
et al. The Doppler cerebroplacental ratio and ratio and uterine artery Doppler indices as pre- with cerebral blood flow redistribution. Ultra-
perinatal outcome in intrauterine growth re- dictors of stillbirth and perinatal loss. Ultrasound sound Obstet Gynecol 2008;32:894-9.
striction. Am J Obstet Gynecol 1999;180: Obstet Gynecol 2015;47:74-80. 114. Cruz-Martinez R, Figueras F, Hernandez-
750-6. 102. Bligh LN, Al Solai A, Greer RM, Kumar S. Andrade E, Oros D, Gratacos E. Fetal brain
89. Gramellini D, Folli MC, Raboni S, Vadora E, Diagnostic performance of cerebroplacental Doppler to predict cesarean delivery for non-
Merialdi A. Cerebral-umbilical Doppler ratio as a ratio thresholds at term for prediction of low reassuring fetal status in term small-for-
predictor of adverse perinatal outcome. Obstet birthweight and adverse intrapartum and gestational-age fetuses. Obstet Gynecol
Gynecol 1992;79:416-20. neonatal outcomes in a term, low-risk 2011;117:618-26.

FEBRUARY 2018 American Journal of Obstetrics & Gynecology S801

Expert Reviews ajog.org

115. DeVore GR. The importance of the cere- onset placental dysfunction: relationship to the 119. Boers KE, van Wyk L, van der Post JA,
broplacental ratio in the evaluation of fetal well- change in uterine artery blood flow resistance et al. Neonatal morbidity after induction vs
being in SGA and AGA fetuses. Am J Obstet between the first and third trimesters. Am J expectant monitoring in intrauterine growth re-
Gynecol 2015;213:5-15. Perinatol 2012;30:505-12. striction at term: a subanalysis of the DIGITAT
116. Parra-Saavedra M, Crovetto F, Triunfo S, 118. Cruz-Martinez R, Savchev S, Cruz- RCT. Am J Obstet Gynecol 2012;206:344.
et al. Association of Doppler parameters with Lemini M, Mendez A, Gratacos E, Figueras F. e1-7.
placental signs of underperfusion in late-onset Clinical utility of third-trimester uterine artery 120. Vijgen SM, Boers KE, Opmeer BC, et al.
small-for-gestational-age pregnancies. Ultra- Doppler in the prediction of brain hemodynamic Economic analysis comparing induction of labor
sound Obstet Gynecol 2014;44:330-7. deterioration and adverse perinatal outcome in and expectant management for intrauterine
117. Llurba E, Turan O, Kasdaglis T, small-for-gestational-age fetuses. Ultrasound growth restriction at term (DIGITAT trial). Eur J
Harman CR, Baschat AA. Emergence of late- Obstet Gynecol 2014;45:273-8. Obstet Gynecol Reprod Biol 2013;170:358-63.

S802 American Journal of Obstetrics & Gynecology FEBRUARY 2018

ajog.org Expert Reviews

Appendix imaging"[All Fields] OR "ultra- ("pregnancy"[All Fields] AND "trimes-

Systematic review of studies on sonography"[All Fields] OR "ultra- ter"[All Fields] AND "third"[All Fields])
routine ultrasound screening perform- sonography"[MeSH Terms])) AND (SGA OR "third pregnancy trimester"[All
ing the scan at a mean gestational [All Fields] OR ("Proc Int Conf Autom Fields] OR ("third"[All Fields] AND
age >32 weeks Face Gesture Recognit"[Journal] OR "trimester"[All Fields]) OR "third tri-
Query syntax: "fgr"[All Fields] OR "Fungal Genet mester"[All Fields]) OR 3rd[All Fields])
(("diagnostic imaging"[Subheading] Rep"[Journal] OR "fgr"[All Fields]) OR Inclusion criteria:
OR ("diagnostic"[All Fields] AND (("Small"[Journal] OR "small"[All Population: low-risk or nonselected
"imaging"[All Fields]) OR "diagnostic Fields]) AND gestational[All Fields] AND pregnancies
imaging"[All Fields] OR "ultrasound"[All ("Age"[Journal] OR "age"[All Fields] OR Predictor: estimated fetal weight or
Fields] OR "ultrasonography"[MeSH "Age (Omaha)"[Journal] OR "age"[All abdominal circumference
Terms] OR "ultrasonography"[All Fields] Fields] OR "Age (Dordr)"[Journal] OR Outcome: smallness for gestational
OR "ultrasound"[All Fields] OR "ultra- "age"[All Fields] OR "Adv Genet Eng"[- age or fetal growth restriction
sonics"[MeSH Terms] OR "ultra- Journal] OR "age"[All Fields]))) AND Design: observational (prospective or
sonics"[All Fields]) OR US[All Fields] OR ((detection[All Fields] AND ("J Rehabil retrospective cohorts) or interventional
("radionuclide imaging"[MeSH Terms] Assist Technol Eng"[Journal] OR Mean gestational age at ultrasound
OR ("radionuclide"[All Fields] AND "rate"[All Fields])) OR ("sensitivity >32 weeks
"imaging"[All Fields]) OR "radionuclide and specificity"[MeSH Terms] OR Publication date 2012
imaging"[All Fields] OR "scan"[All ("sensitivity"[All Fields] AND "specific- Sources: PubMed
Fields]) OR ultrasonographic[All Fields] ity"[All Fields]) OR "sensitivity and Query results: 132 Studies, of which
OR ("diagnostic imaging"[Subheading] specificity"[All Fields] OR "sensitivi- 13 met the inclusion criteria after review
OR ("diagnostic"[All Fields] AND ty"[All Fields])) AND (("pregnancy of full-text document by 2 independent
"imaging"[All Fields]) OR "diagnostic trimester, third"[MeSH Terms] OR reviewers (J.C. and F.F.).

FEBRUARY 2018 American Journal of Obstetrics & Gynecology S802.e1