5/22/2018 Rciu 2014 Clinic

1/13

E-Mail karger@karger.com

Review

Fetal Diagn Ther

DOI: 10.1159/000357592

Update on the Diagnosis and Classificationof Fetal Growth Restriction and Proposalof a Stage-Based Management Protocol

Francesc Figueras Eduard Gratacs

Barcelona Center of Maternal-Fetal Medicine and Neonatology (Hospital Clinic and Hospital Sant Joan de Deu), IDIBAPS,

University of Barcelona, and Centre for Biomedical Research on Rare Diseases (CIBER-ER), Barcelona, Spain

agement of FGR and the decision to deliver aims at an opti-

mal balance between minimizing fetal injury or death versus

the risks of iatrogenic preterm delivery. We propose a proto-

col that integrates current evidence to classify stages of fetal

deterioration and establishes follow-up intervals and opti-

mal delivery timings, which may facilitate decisions and re-

duce practice variability in this complex clinical condition.

2014 S. Karger AG, Basel

Introduction

Fetal growth restriction (FGR) is defined as a failure toachieve the endorsed growth potential. The diagnosis offetal smallness is currently performed on the basis of anestimated fetal weight (EFW) below a given threshold,most commonly the 10th centile. It is likely that this def-

inition lacks sensitivity, so that it misses cases of growthrestriction that do not fall below the 10th centile, but itidentifies a subset of pregnancies at high risk of poorerperinatal outcome. Thus, detection of small fetuses isclinically relevant because as a whole this group of fetuses

Key Words

Fetal growth restriction update Stage-based management

protocol Constitutional small-for-gestational age

Early-severe versus late-mild fetal growth restriction

Abstract

Small fetuses are defined as those with an ultrasound esti-

mated weight below a threshold, most commonly the 10th

centile. The first clinically relevant step is the distinction of

true fetal growth restriction (FGR), associated with signs of

abnormal fetoplacental function and poorer perinatal out-

come, from constitutional small-for-gestational age, with a

near-normal perinatal outcome. Nowadays such a distinc-

tion should not be based solely on umbilical artery Doppler,

since this index detects only early-onset severe forms. FGR

should be diagnosed in the presence of any of the factors

associated with a poorer perinatal outcome, includingDoppler cerebroplacental ratio, uterine artery Doppler, a

growth centile below the 3rd centile, and, possibly in the

near future, maternal angiogenic factors. Once the diagnosis

is established, differentiating into early- and late-onset FGR

is useful mainly for research purposes, because it distin-

guishes two clear phenotypes with differences in severity,

association with preeclampsia, and the natural history of fe-

tal deterioration. As a second clinically relevant step, man-

Received: November 19, 2013

Accepted: November 19, 2013

Published online: January 23, 2014

Francesc Figueras and Eduard Gratacs

Maternal-Fetal Medicine DepartmentHospital Clinic, University of BarcelonaSabino de Arana 1, ES08028 Barcelona (Spain)E-Mail ffiguera @ clinic.ub.es and egratacos@fetalmedicinebarcelona.org

2014 S. Karger AG, Basel

10153837/14/00000000$39.50/0

www.karger.com/fdt

E. Gratacs and F. Figueras contributed equally to this work.

5/22/2018 Rciu 2014 Clinic

2/13

Figueras/Gratacs

Fetal Diagn Ther

DOI: 10.1159/000357592

2

is associated with poorer perinatal outcome, and this rep-resents opportunities for preventing cases of intrauterinefetal death, perinatal brain injury and severe intrapartumfetal distress. In addition, evidence accumulating over thelast 20 years has consistently demonstrated how beingborn small has important implications for the quality of

health during adulthood.Population-based series show that prenatal identifica-

tion of small-for-gestational age (SGA) babies results in areduction of adverse perinatal outcomes and stillbirth [1,2]. However, most SGA babies remain unnoticed untilbirth, even when routine third-trimester ultrasound isperformed [3, 4]. Moreover, according to pregnancy au-dits, most instances of avoidable stillbirth are related witha failure to antenatally detect SGA [5]. In order to definestrategies to improve detection, our understanding of thedifferent clinical forms and its risks factors should also beimproved. Furthermore, there is a wide variability in the

management of FGR in terms of monitoring and recom-mended gestational age at delivery, mainly in early-onsetforms. Clinical practice variability, a marker of poor qual-ity of care, is partially accounted for by the lack of com-parability among studies and the lack of clear recommen-dations in the medical literature, combined with the dif-ficulty of integrating the meaning of the myriad ofparameters described for FGR, with the risks of deliveryat different gestational age periods.

In this review we analyze current evidence and suggesta systematic approach to FGR, which entails a proper

identification of FGR versus SGA, and a stage-basedmanagement protocol which may help in reducing clini-cal variability.

Distinction between Fetal Growth Restriction and

(Constitutional) Small-for-Gestational Age

While overall fetal smallness is associated with pooreroutcome, clinical evidence suggests that there are, at least,two groups of small fetuses. By arbitrary convention,these two groups are normally referred to as FGR versus

constitutional SGA, the latter category henceforth re-ferred as SGA.

FGR is normally used to refer to small fetuses withhigher risk for fetal in utero deterioration, stillbirth andoverall poorer perinatal outcome as compared with nor-mally grown fetuses. These fetuses are thought to havetrue growth restriction. In general, FGR is associatedwith Doppler signs suggesting hemodynamic redistribu-tion as a reflection of fetal adaptation to undernutrition/

hypoxia, histological and biochemical signs of placentaldisease and a higher risk of preeclampsia. The term SGAhas been used to differentiate a subgroup of small fetusesthat do not present the changes described above, so thatthere appears to be no fetal adaptation to an abnormalenvironment and with perinatal outcomes similar to

those of normally grown fetuses.Irrespective of whether these diagnostic labels are a

proper reflection of the underlying pathophysiology,from a clinical point of view the distinction between FGRversus SGA is relevant because of the correlation withperinatal outcome. There is wide consensus that it is rea-sonable to deliver electively FGR when lung maturationcan be presumed, or earlier if signs of fetal deteriorationare observed. On the contrary, SGA fetuses are associatedwith virtually normal perinatal outcome and it is gener-ally considered that active management or elective deliv-ery before full term offers no benefit. While the concepts

of FGR and SGA may be clear, the distinction of trueFGR in clinical practice can be challenging. The differen-tiation of these two forms has long been based on Dopplersigns reflecting fetal adaptation to increased placental re-sistance and/or hypoxia.

Umbilical Artery as a Standalone Standard Is NotValid AnymoreFor almost 20 years, the umbilical artery (UA) has

been widely accepted as the standard to identify FGR.During the 1980s and 1990s, a substantial number of

studies demonstrated that abnormal UA Doppler indi-cated a poorer outcome among small fetuses. In addition,meta-analyses have shown that UA Doppler could im-prove mortality and perinatal outcome in FGR [6]. Thisled to identify UA as a surrogate of placental disease andconsequently to consider small fetuses with normal UADoppler as SGA with no placental disease [7]. However,this assumption was based on false premises, because itextended observations that are valid in the most severesubset of FGR fetuses to the whole group of FGR. WhileUA identifies severe placental disease, it fails to pick upinstances of mild placental disease which constitute a

proportion of early-onset cases and virtually all instancesof late-onset FGR [8].

Evidence during the last two decades has demonstrat-ed that SGA, as defined by a normal UA pulsatility index(PI), contains a large proportion of fetuses with worseperinatal outcomes than normally grown fetuses [3,911]. Thus, UA Doppler cannot be used as a standalonecriterion to differentiate FGR from SGA.

5/22/2018 Rciu 2014 Clinic

3/13

Update on FGR Fetal Diagn Ther

DOI: 10.1159/000357592

3

Improving the Definition of Fetal Growth Restriction:Parameters Identifying the Small Fetus with PoorOutcomeResearch over the last 10 years has investigated predic-

tors of poor outcome among mild-late forms of FGR.Current evidence suggests that there is not a single pa-

rameter to best differentiate FGR from SGA. The best in-dividual candidate is the Doppler cerebroplacental ratio(CPR). CPR is calculated by dividing the middle cerebralartery (MCA) PI by the UA Doppler PI. This index re-flects in a combined fashion mild increases in placentalresistance with mild reductions in fetal brain vascular re-sistance. In animal [12] and clinical [13] models this ratiohas been demonstrated to be more sensitive to hypoxiathan its individual components and correlates better withadverse outcome [14]. Since the CPR includes the UA, itcan replace its use for the detection of FGR at any gesta-tional age.

Aside from CPR, the uterine artery Doppler PI (UtAPI) can be abnormal in the presence of a normal UADoppler in small fetuses and predicts a poorer outcomein small fetuses. If used in combination with the cerebralor umbilical Doppler, its independent predictive value isreduced, but evidence still suggests that it can marginallyimprove the identification of poor outcome [11, 15, 16].Another predictor of poor outcome is a very small EFW.Among fetuses below the 10th centile, those with an EFW

5/22/2018 Rciu 2014 Clinic

4/13

Figueras/Gratacs

Fetal Diagn Ther

DOI: 10.1159/000357592

4

Table 1.Summary of the main differences between early- and late-onset forms of FGR

Early-onset FGR (12%) Late-onset FGR (35%)

Problem: management Problem: diagnosisPlacental disease: severe (UA Doppler abnormal, high associationwith preeclampsia)

Placental disease: mild (UA Doppler normal, low associationwith preeclampsia)

Hypoxia ++: systemic cardiovascular adaptation Hypoxia +/: central cardiovascular adaptation

Immature fetus = higher tolerance to hypoxia = natural history Mature fetus = lower tolerance to hypoxia = no (or very short)natural history

High mortality and morbidity; lower prevalence Lower mortality (but common cause of late stillbirth); poorlong-term outcome; affects large fraction of pregnancies

0

20a 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

20

40

60

80

100

120

140

UA normal160

UA abnormal

Fig. 1.Distribution of a large population ofsmall fetuses (n = 656) in FGR or SGA.aFGR is defined by an UA PI >95th centileonly. bFGR is defined using a combination

of CPR 95th centileand an EFW

5/22/2018 Rciu 2014 Clinic

5/13

Update on FGR Fetal Diagn Ther

DOI: 10.1159/000357592

5

hypoxia and acidosis, which is reflected by escalating ab-normalities in the UA and increased PIs in the precordialveins, mainly the ductus venosus (DV). The latency ofsevere fetal deterioration can vary in individual cases, butit normally lasts weeks [27] and it often follows a cascadeof changes which are reflected in a pattern of Doppler

changes that allows to monitor the progression of fetaldeterioration and tailor elective delivery (fig. 1).

Early severe FGR is associated with severe injury and/or fetal death before term in many cases [28]. Manage-ment is challenging and aims at achieving the best balancebetween the risks of leaving the fetus in utero versus thecomplications of prematurity.

Late-Onset Fetal Growth RestrictionLate-onset FGR represents 7080% of FGR [25]. A first

distinction with early-onset forms is that the associationwith late PE is low, roughly 10% [25]. The degree of pla-

cental disease is mild, thus UA Doppler is normal in vir-tually all cases [8]. Despite normal UA PI Doppler, thereis a high association with abnormal CPR values [8]. Inaddition, advanced brain vasodilation suggesting chronichypoxia, as reflected by an MCA PI 700 cases deter-mined that 32 weeks at diagnosis and 37 weeks at deliverybest classified two groups where the differences in termsof adverse perinatal outcome are maximized [36].

Clinical Management of Fetal Growth Restriction

and Small-for-Gestational Age Fetuses

Over the following sections we first briefly review ex-isting evidence on the correlation between methods forassessing fetal well-being in FGR and the risk of fetal in-jury or death. Subsequently, we propose a unique manag-ing guideline for FGR as a whole, which integrates exist-ing evidence to define stages of fetal deterioration, andproposes follow-up intervals, and ideal gestational age atand mode of delivery for each stage.

5/22/2018 Rciu 2014 Clinic

6/13

Figueras/Gratacs

Fetal Diagn Ther

DOI: 10.1159/000357592

6

Brief Description of Methods and Indices for FetalAssessment and Their Correlation with PerinatalOutcomesFetal well-being tests and indices could be roughly

classified as chronic or acute. Whilst the former becomeprogressively abnormal due to increasing hypoxemia

and/or hypoxia, the latter correlate with acute changesoccurring in advanced stages of fetal compromise, char-acterized by severe hypoxia and metabolic acidosis, andusually precedes fetal death in a few days. Some of themeasures and indices discussed below are essentially usedfor the diagnosis/identification of FGR from SGA, andconsequently they are relevant for the decision as towhether delivery is indicated when pregnancy term isreached. Another set of indices have a prognostic value,since they are useful to determine that there is a high riskof deterioration, and consequently they are used to indi-cate delivery before term is reached.

Umbilical Artery DopplerUA Doppler is the only measure that provides both

diagnostic and prognostic information for the manage-ment of FGR. On the one hand, increased UA Doppler PIhas a great clinical value for the identification of FGR,alone or combined in the CPR ratio. On the other hand,the progression of UA Doppler patterns to absent or re-verse end-diastolic flow correlates with the risks of injuryor death.

There is compelling evidence that using UA Doppler in

high-risk pregnancies (most of them SGA fetuses) im-proves perinatal outcomes, with a 29% reduction (248%)in perinatal deaths [6]. Absent or reversed end-diastolic

velocities, the end of the spectrum of the abnormalities ofthe UA Doppler, have been reported to be present on aver-age 1 week before the acute deterioration [37]. Up to 40%of fetuses with acidosis show this umbilical flow pattern[37]. There is an association between reversed end-diastol-ic flow in the UA and adverse perinatal outcome (with asensitivity and specificity of about 60%), which seems to beindependent of prematurity [38]. After 30 weeks the risk ofstillbirth of a fetus with isolated reversed end-diastolic ve-

locities in the UA Doppler overcomes the risks of prema-turity [3941], and therefore delivery seems justified.

Middle Cerebral Artery DopplerMCA informs about the existence of brain vasodila-

tion, a surrogate marker of hypoxia. MCA is considereda rather late manifestation, with acceptable specificity butlow sensitivity, which is improved by the use of CPR, asdiscussed below. There is an association between abnor-

mal MCA PI and adverse perinatal and neurological out-come, but it is unclear whether delivering before termcould add any benefit. MCA is particularly valuable forthe identification [8] and prediction [42, 43] of adverseoutcome among late-onset FGR, independently of theUA Doppler, which is often normal in these fetuses. Fe-

tuses with abnormal MCA PI had a sixfold risk of emer-gency cesarean section for fetal distress when comparedwith SGA fetuses with normal MCA PI [44], which is par-ticularly relevant because labor induction at term is thecurrent standard of care of late-onset FGR [45, 46]. LateFGRs with abnormal MCA PI have poorer neurobehav-ioral competence at birth and at 2 years of age [42, 47].

Cerebroplacental RatioThe CPR is essentially a diagnostic index. The CPR

improves remarkably the sensitivity of UA and MCAalone, because increased placental impedance (UA) is of-

ten combined with reduced cerebral resistance (MCA).Thus, the CPR is already decreased when its individualcomponents suffer mild changes but are still within nor-mal ranges [13, 48]. In late SGA fetuses, abnormal CPR ispresent before delivery in 2025% of the cases [49], andit is associated with a higher risk of adverse outcome atinduction, although to a lesser degree than MCA [44].There are no long-term studies evaluating the neurobe-havioral or neurodevelopmental consequences of lateSGA with abnormal CPR. However, it is remarkable thateven in the general population, an abnormal CPR pre-

dicts neurobehavioral problems at 18 months of age [50].Interestingly, the anterior cerebral artery-CPR ratherthan the MCA-CPR showed the stronger association,demonstrating a differential impact of regional altera-tions in cerebral blood flow impedance on development,which is consistent with findings in early FGR [51, 52].

Ductus Venosus DopplerDV is the strongest single Doppler parameter to pre-

dict the short-term risk of fetal death in early-onset FGR.Longitudinal studies have demonstrated that DV flowwaveforms become abnormal only in advanced stages of

fetal compromise [27, 37, 38, 53]. Consistently, there is agood correlation of abnormal DV waveform with late-stage acidemia at cordocentesis [54]. Absent or reversedvelocities during atrial contraction are associated withperinatal mortality independently of the gestational ageat delivery [55], with a risk ranging from 40 to 100% inearly-onset FGR [41, 56]. Thus, this sign is normally con-sidered sufficient to recommend delivery at any gesta-tional age, after completion of steroids. A DV above the

5/22/2018 Rciu 2014 Clinic

7/13

Update on FGR Fetal Diagn Ther

DOI: 10.1159/000357592

7

95% centile is associated with higher risks but not as con-sistently as when atrial flow is reverse. Overall, the sensi-

tivity for perinatal death is still 4070% [28, 55, 57]. Asystematic review of 18 observational studies (including2,267 fetuses) found that DV Doppler has predictive ca-pacity for perinatal mortality [58]. In about 50% of cases,abnormal DV precedes the loss of short-term variability(STV) in computerized cardiotocography (cCTG) [27],and in about 90% of cases it is abnormal 4872 h beforethe biophysical profile (BPP) [53]. Hence, it is consideredto provide a better window of opportunity for deliveringfetuses in critical conditions at very early gestational ages.

Aortic Isthmus Doppler

The aortic isthmus (AoI) Doppler is associated withincreased fetal mortality and neurological morbidity inearly-onset FGR [59]. This vessel reflects the balance be-tween the impedance of the brain and systemic vascularsystems [60, 61]. Reverse AoI flow is a sign of advanceddeterioration and a further step in the sequence startingwith the UA and MCA Dopplers (fig. 2, 3). Remarkably,the AoI can be found abnormal also in a small proportionof late-onset FGRs [29].

AoI has a strong association with both adverse perina-tal [62] and neurological outcome [59]. However, longi-

tudinal studies show that the AoI precedes DV abnor-malities by 1 week [29, 63], and consequently it is not asgood to predict the short-term risk of stillbirth [41]. Incontrast, AoI seems to improve the prediction of neuro-logical morbidity [59]. Among early-onset FGR with pos-itive DV atrial velocities, a reverse AoI indicated a veryhigh risk of late neonatal neurological injury (57 vs. 9.7%)[41]. In the opinion of the authors, reverse AoI could al-ready be incorporated in clinical protocols as a sign ofsevere placental insufficiency and could justify consider-ing elective delivery beyond 34 weeks of gestation. If fu-ture studies confirm the strong association with neuro-

logical morbidity, reverse AoI flow could be used to indi-cate delivery even earlier, but more data are required.

Fetal Heart Rate Analysis by Conventional andComputerized CardiotocographyEarly studies on high-risk pregnancies showed that,

though highly sensitive, CTG has a 50% rate of false pos-itives for the prediction of adverse outcome [64]. In addi-tion, a meta-analysis [65] on high-risk pregnancies failed

Growth restriction

DV PI >p95

UA PI >p95

AoI PI >p95MCA PI

5/22/2018 Rciu 2014 Clinic

8/13

Figueras/Gratacs

Fetal Diagn Ther

DOI: 10.1159/000357592

8

to demonstrate any beneficial effect in reducing perinatalmortality. Hence, there is no evidence to support the useof traditional fetal heart rate (FHR) monitoring or non-stress tests in FGR fetuses. We must acknowledge thatthese studies were conducted in the early 1980s, and thecontrol group had no fetal well-being assessment or out-dated techniques such as biochemical tests. However, ingeneral, a silent FHR pattern or the presence of spontane-ous decelerations represent a very late event precedingfetal demise, and consequently measures that allow ear-lier identification and delivery must be used. A main lim-

itation of conventional CTG is the subjective interpreta-tion of the FHR, which is extremely challenging in verypreterm fetuses with a physiologically reduced variability.

cCTG has represented a step forward and has providednew insights into the pathophysiology and managementof FGR. cCTG evaluates STV of the FHR, an aspect thatsubjective evaluation cannot assess. Current evidencesuggests that cCTG is sensitive to detect advanced fetaldeterioration, and it provides a value similar to DV re-

verse atrial flow for the short-term prediction of fetaldeath. STV closely correlates with acidosis and severe hy-poxia as demonstrated by cord blood sampling at the timeof a cesarean section [6668]. Recent longitudinal serieshave pointed to a potential role as an acute marker [27].STV becomes abnormal, coinciding with the DV, where-

as in about half of cases, abnormal DV precedes the lossof short-term FHR variability, the latter is the first to be-come abnormal in the other cases [27].

Biophysical ProfileBPP is calculated by combining ultrasound assessment

of fetal tone, respiratory and body movements, with am-niotic fluid index and a conventional CTG. It was de-signed to improve the performance of FHR. Observation-al studies show an association between abnormal BPPand perinatal mortality and cerebral palsy [69]. Studies inwhich a cordocentesis was performed demonstrated a

good correlation with acidosis [70], with fetal tone andgross motor movements the best correlated components.However, as with FHR, a high false-positive rate (50%)limits the clinical usefulness of the BPP [71]. Early obser-vational studies reported a very low risk of false positivesfor acidosis and perinatal death, but more recent studieson early-onset very preterm FGR fetuses raise concernsover the false-positive rate, with up to 23% of instances ofintrauterine fetal death in fetuses with BPP >6 and 11%in those with BPP >8 [72]. A meta-analysis [73] showedno significant benefit of BPP in high-risk pregnancies.

Consequently, whenever Doppler expertise and/or cCTGare available, the incorporation of BPP in managementprotocols of FGR is questionable.

Amniotic Fluid IndexAmniotic fluid index (AFI) is used essentially as part

of BPP. Amniotic fluid volume is believed to be a chronicparameter. In fact, among the components of BPP, it isthe only one that is not considered acute. A meta-analysis[74] of 18 randomized studies demonstrated that a re-duced AFI is associated with an abnormal 5-min Apgarscore, but there was no association with acidosis or peri-

natal death in SGA (RR 1.6, 95% CI 0.92.6). Longitudi-nal studies in early-onset FGR fetuses have shown that theAFI fluid index progressively decreases [27, 38]. Oneweek before acute deterioration, 2030% of cases haveoligohydramnios [38, 53]. There is limited evidence onthe role of oligohydramnios to predict perinatal compli-cations in FGR fetuses managed with Doppler so that itsinclusion in management protocols is questionable.

Growth restriction

MCA PI p95

CPR p95

CTG decelerations

AcidosisInjury/death

HypoxiaCentralization

Placental diseaseReduced placental reserve

Acute deteriorationHours

Diagnostic markersWeeks

AoI reverse

Fig. 3.Fetal deterioration and monitoring in late-mild FGR. Pla-cental disease is mild and UA Doppler values are not elevated

above the 95th centile. The effects of fetal adaptation are best de-tected by the CPR, which can pick up mild changes in the AU andMCA Doppler. An important fraction of cases do not progress tobaseline hypoxia so that they remain only with abnormal CPR.Once baseline hypoxia is established, placental reserve is minimaland progression to fetal deterioration may occur quickly, as sug-gested by the high risk of severe deterioration or intrauterine fetaldeath after 37 weeks in these cases, possibly due to a combinationof a higher susceptibility to hypoxia of the term-mature fetus andthe more common presence of contractions at term (see text forabbreviations).

5/22/2018 Rciu 2014 Clinic

9/13

Update on FGR Fetal Diagn Ther

DOI: 10.1159/000357592

9

Evidence about Timing of Delivery in Fetal GrowthRestrictionSince no treatment has been demonstrated to be of ben-

efit in growth restriction [7579], assessment of fetal well-being and timely delivery remain as the main managementstrategy. The aim behind a clinical protocol for managing

FGR is to combine existing evidence on various methodsfor monitoring fetal well-being in order to establish therisks of fetal injury or death, and to balance them againstthe risks of prematurity if the fetus is delivered. Added intothe equation is the awareness that leaving pregnancies withFGR to deliver at term may also lead to perinatal morbid-ity and delayed effects such as cerebral palsy [80].

The Growth Restriction Intervention Trial (GRIT)study [40] was a multicenter randomized controlled trialaimed to compare the effect of delivering early with delay-ing birth for as long as possible. A total of 588 fetuses be-tween 24 and 36 weeks were randomized to immediate

delivery or delayed delivery until the obstetrician was nolonger certain. The study observed that when obstetri-cians were uncertain about timing of delivery based onUA Doppler, they were prepared to vary the timing byabout 4 days, and although such delay caused some still-births, earlier delivery resulted in an almost equal numberof additional deaths. Moreover, at 2 years of age, there wasa trend towards more disability in the immediate deliverygroup [39]. Concerns regarding the external validity ofthis study exist [81], since only 5% of the eligible popula-tion was recruited, which raises doubts about the repre-

sentativeness of the sample.One randomized equivalence trial exists comparing theeffect of induction of labor or expectant monitoring inwomen after 37 weeks of gestation with suspected SGA.They found negligible differences in peri- and neonataloutcomes between induction of labor and expectant mon-itoring [45, 46]. At 2 years of age, about half of the cohortwere evaluated for neurodevelopmental and neurobehav-ioral assessment, with no differences between both strate-gies [82]. Based on these results, it was concluded that itseems reasonable to offer delivery after 37 weeks in SGAinfants. As the authors pointed out in the discussion of

this study, further studies differentiating true FGR fromother causes of SGA not associated with poor perinataloutcome are required to clarify this question.

Stage-Based Protocol for Managing Fetal GrowthRestrictionFGR is probably among the obstetrical entities with

the greatest variation in clinical practice. This resultsfrom a combination of the lack of strong supportive evi-

dence, the complexity of the variables and indices thatneed to be integrated for assessing fetal deterioration, andthe variable risks associated with prematurity at differentgestational ages.

Although when considered as groups there are cleardifferences between early- and late-onset forms, on an

individual basis there is important overlapping of clinicalfeatures at borderline gestational ages. In addition, caseswith the same gestational age at onset are often detectedat different time points during gestation. Figure 1 illus-trates how FGR represents a continuum with increasingnumber of instances as gestational age progresses. Con-sequently, a management scheme that establishes follow-up intervals and timing of delivery on the basis of fetalrisks can include both early- and late-onset forms in anintegrated fashion.

The main aims behind clinical management of FGRshould be firstly to distinguish FGR from SGA and sec-

ondly to ascertain whether there is risk of in utero fetalinjury or death. Thus, a first step is to identify within thesmall fetuses the subset of FGR, because they will have anincreased risk of adverse outcome and stillbirth andshould be managed actively once term is reached. A sec-ond step is to identify the presence of any sign suggestingthat there is risk of fetal injury or death that may recom-mend delivery before term.

While strong evidence is lacking to support firm rec-ommendations on the timing of delivery, a protocol thatintegrates the best available evidence can help reducing

clinical practice variation. One approach is to group instages those indices or signs that are associated with sim-ilar fetal risks, since they should indicate similar follow-up intervals and timing of delivery. Thus, based on theexisting evidence extensively discussed above and, whereno evidence is available on experts opinion, we suggestto profile several stages, or prognostic groups, which de-fine different management strategies (table 2; fig. 4).

In a first step, once a small fetus (i.e. EFW

5/22/2018 Rciu 2014 Clinic

10/13

Figueras/Gratacs

Fetal Diagn Ther

DOI: 10.1159/000357592

10

Stage IV: DV absent/reversed EDV (persisting 12 h apart) orpathological CTG (reduced STV or deceleration pattern)

26 weeksYes

No

No

Yes

Repeat in 1224 h

Repeat in 2448 h

Stage III: DV pulsatility index >95th centile orUA reversed EDV (both persisting 12 h apart)

30 weeksYes

No

Yes

YesRepeatin2w

eeks

No 40 weeks

No

Laborinduction

NoRepeat in 23 days

Stage II: UA absent EDVorAoI reversed diastolic velocities (both persisting 12 h apart)

34 weeksYes

No

Yes

NoRepeat in 1 week

Stage I: EFW

5/22/2018 Rciu 2014 Clinic

11/13

Update on FGR Fetal Diagn Ther

DOI: 10.1159/000357592

11

Stage I Fetal Growth Restriction (Severe Smallness orMild Placental Insufficiency).Either UtA, UA or MCADoppler, or the CPR are abnormal. In the absence of oth-er abnormalities, evidence suggests a low risk of fetal de-terioration before term. Labor induction beyond 37 weeksis acceptable, but the risk of intrapartum fetal distress is

increased [44]. Cervical induction with Foley catheter isalso recommended. Weekly monitoring seems reason-able.

Stage II Fetal Growth Restriction (Severe Placental In-sufficiency).This stage is defined by UA absent-end dia-stolic velocity (AEDV) or reverse AoI. Although evidencefor UA AEDV is stronger than that for AoI, observation-al evidence suggests an association between the latter toabnormal neurodevelopment, so that both criteria be-come a single category. Delivery should be recommendedafter 34 weeks. The risk of emergent cesarean section atlabor induction exceeds 50%, and, therefore, elective ce-

sarean section is a reasonable option. Monitoring twice aweek is recommended.

Stage III Fetal Growth Restriction (Advanced Fetal De-terioration, Low-Suspicion Signs of Fetal Acidosis). Thestage is defined by reverse absent-end diastolic velocity(REDV) or DV PI >95th centile. There is an associationwith a higher risk of stillbirth and poorer neurological

outcome. However, since signs suggesting a very high riskof stillbirth within days are not present yet, it seems rea-sonable to delay elective delivery to reduce as possible theeffects of severe prematurity. We suggest delivery shouldbe recommended by cesarean section after 30 weeks.Monitoring every 2448 h is recommended.

Stage IV Fetal Growth Restriction (High Suspicion ofFetal Acidosis and High Risk of Fetal Death).There arespontaneous FHR decelerations, reduced STV (

5/22/2018 Rciu 2014 Clinic

12/13

Figueras/Gratacs

Fetal Diagn Ther

DOI: 10.1159/000357592

12

21 Larroque B, et al: School difficulties in20-year-olds who were born small for gesta-tional age at term in a regional cohort study.Pediatrics 2001; 108: 111115.

22 Crispi F, et al: Fetal growth restriction resultsin remodeled and less efficient hearts in chil-dren. Circulation 2010; 121: 24272436.

23 Verkauskiene R, et al: Birth weight and long-term metabolic outcomes: does the definitionof smallness matter? Horm Res 2008; 70: 309315.

24 Lackman F, et al: Fetal umbilical cord oxygenvalues and birth to placental weight ratio inrelation to size at birth. Am J Obstet Gynecol2001; 185: 674682.

25 Crovetto F, et al: Performance of first-trimes-ter integrated screening for early and late smallfor gestational age newborns. Ultrasound Ob-stet Gynecol 2013 (E-pub ahead of print).

26 Turan OM, et al: Progression of Doppler ab-normalities in intrauterine growth restric-tion. Ultrasound Obstet Gynecol 2008; 32:160167.

27 Hecher K, et al: Monitoring of fetuses with in-

trauterine growth restriction: a longitudinalstudy. Ultrasound Obstet Gynecol 2001; 18:564570.

28 Baschat AA, et al: Predictors of neonatal out-come in early-onset placental dysfunction.Obstet Gynecol 2007; 109: 253261.

29 Cruz-Martinez R, et al: Changes in myocar-dial performance index and aortic isthmusand ductus venosus Doppler in term, small-for-gestational age fetuses with normal um-bilical artery pulsatility index. UltrasoundObstet Gynecol 2011; 38: 400405.

30 Kady S, Gardosi J: Perinatal mortality and fe-tal growth restriction. Best Pract Res Clin Ob-stet Gynaecol 2004; 18: 397410.

31 Figueras F, et al: Small-for-gestational-age fe-

tuses with normal umbilical artery Dopplerhave suboptimal perinatal and neurodevelop-mental outcome. Eur J Obstet Gynecol Re-prod Biol 2008; 136: 3438.

32 Van Vliet EO, et al: Placental pathology andlong-term neurodevelopment of very preterminfants. Am J Obstet Gynecol 2012; 206: 489.e17.

33 Chan PY, et al: The long-term effects of pre-maturity and intrauterine growth restrictionon cardiovascular, renal, and metabolic func-tion. Int J Pediatr 2010; 2010: 280402.

34 Spinillo A, et al: Placental histopathologicalcorrelates of umbilical artery Doppler velo-cimetry in pregnancies complicated by fetal

growth restriction. Prenat Diagn 2012; 32:12631272.35 Llurba E, et al: Emergence of late-onset pla-

cental dysfunction: relationship to the changein uterine artery blood flow resistance be-tween the first and third trimesters. Am JPerinatol 2013; 30: 505512.

36 Savchev S, et al: Evaluation of an optimal ges-tational age cut-off for the definition of early-and late-onset fetal growth restriction. FetalDiagn Ther 2013, Epub ahead of print.

37 Ferrazzi E, et al: Temporal sequence of abnor-mal Doppler changes in the peripheral andcentral circulatory systems of the severelygrowth-restricted fetus. Ultrasound ObstetGynecol 2002; 19: 140146.

38 Cosmi E, et al: Doppler, cardiotocography,and biophysical profile changes in growth-restricted fetuses. Obstet Gynecol 2005; 106:12401245.

39 Thornton JG, et al: Infant well-being at 2 yearsof age in the Growth Restriction InterventionTrial (GRIT): multicentred randomised con-trolled trial. Lancet 2004; 364: 513520.

40 GRIT Study Group: A randomised trial oftimed delivery for the compromised pretermfetus: short-term outcomes and bayesian in-terpretation. BJOG 2003; 110: 2732.

41 Cruz-Lemini M, et al: Risk of perinatal deathin early-onset intrauterine growth restrictionaccording to gestational age and cardiovascu-lar Doppler indices: a multicenter study. FetalDiagn Ther 2012; 32: 116122.

42 Eixarch E, et al: Neurodevelopmental out-come in 2-year-old infants who were small-

for-gestational age term fetuses with cerebralblood flow redistribution. Ultrasound ObstetGynecol 2008; 32: 894899.

43 Hershkovitz R, et al: Fetal cerebral blood flowredistribution in late gestation: identificationof compromise in small fetuses with normalumbilical artery Doppler. Ultrasound ObstetGynecol 2000; 15: 209212.

44 Cruz-Martinez R, et al: Fetal brain Doppler topredict cesarean delivery for non-reassuringfetal status in term small-for-gestational-agefetuses. Obstet Gynecol 2011; 117: 618626.

45 Boers KE, et al: Neonatal morbidity after in-duction vs. expectant monitoring in intra-uterine growth restriction at term: a subanaly-sis of the DIGITAT RCT. Am J Obstet Gyne-

col 2012; 206: 344.e17.46 Boers KE, et al: Induction versus expectant

monitoring for intrauterine growth restric-tion at term: randomised equivalence trial(DIGITAT). BMJ 2010; 341:c7087.

47 Oros D, et al: Middle versus anterior cerebralartery Doppler for the prediction of perinataloutcome and neonatal neurobehavior in termsmall-for-gestational-age fetuses with normalumbilical artery Doppler. Ultrasound ObstetGynecol 2010; 35: 456461.

48 Arbeille P, et al: Assessment of the fetal PO2changes by cerebral and umbilical Doppler onlamb fetuses during acute hypoxia. Ultra-sound Med Biol 1995; 21: 861870.

49 Cruz-Martinez R, et al: Longitudinal brainperfusion changes in near-term small-for-gestational-age fetuses as measured by spec-tral Doppler indices or by fractional movingblood volume. Am J Obstet Gynecol 2010;203: 42.e16.

50 Roza SJ, et al: What is spared by fetal brain-sparing? Fetal circulatory redistribution andbehavioral problems in the general popula-tion. Am J Epidemiol 2008; 168: 11451152.

51 Figueroa-Diesel H, et al: Doppler changes inthe main fetal brain arteries at different stagesof hemodynamic adaptation in severe intra-uterine growth restriction. Ultrasound ObstetGynecol 2007; 30: 297302.

52 Dubiel M, Gunnarsson GO, Gudmundsson S:Blood redistribution in the fetal brain duringchronic hypoxia. Ultrasound Obstet Gynecol2002; 20: 117121.

53 Baschat AA, Gembruch U, Harman CR: Thesequence of changes in Doppler and biophys-ical parameters as severe fetal growth restric-tion worsens. Ultrasound Obstet Gynecol2001; 18: 571577.

54 Hecher K, et al: Fetal venous, intracardiac,and arterial blood flow measurements in in-trauterine growth retardation: relationshipwith fetal blood gases. Am J Obstet Gynecol1995; 173: 1015.

55 Schwarze A, et al: Qualitative venous Dopplerflow waveform analysis in preterm intrauter-ine growth-restricted fetuses with ARED flowin the umbilical artery correlation withshort-term outcome. Ultrasound Obstet Gy-

necol 2005; 25: 573579.56 Baschat AA, et al: Qualitative venous Doppler

waveform analysis improves prediction ofcritical perinatal outcomes in prematuregrowth-restricted fetuses. Ultrasound ObstetGynecol 2003; 22: 240245.

57 Bilardo CM, et al: Relationship betweenmonitoring parameters and perinatal out-come in severe, early intrauterine growth re-striction. Ultrasound Obstet Gynecol 2004;23: 119125.

58 Morris RK, et al: Systematic review and meta-analysis of the test accuracy of ductus venosusDoppler to predict compromise of fetal/neo-natal wellbeing in high-risk pregnancies withplacental insufficiency. Eur J Obstet Gynecol

Reprod Biol 2010; 152: 312.59 Fouron JC, et al: The relationship between

an aortic isthmus blood flow velocity indexand the postnatal neurodevelopmental sta-tus of fetuses with placental circulatory in-sufficiency. Am J Obstet Gynecol 2005; 192:497503.

60 Fouron JC, et al: Relationship between flowthrough the fetal aortic isthmus and cerebraloxygenation during acute placental circula-tory insufficiency in ovine fetuses. Am J Ob-stet Gynecol 1999; 181: 11021107.

61 Makikallio K, Jouppila P, Rasanen J: Retro-grade aortic isthmus net blood flow and hu-man fetal cardiac function in placental insuf-

ficiency. Ultrasound Obstet Gynecol 2003; 22:351357.62 Del Rio M, et al: Doppler assessment of the

aortic isthmus and perinatal outcome in pre-term fetuses with severe intrauterine growthrestriction. Ultrasound Obstet Gynecol 2008;31: 4147.

63 Figueras F, et al: Monitoring of fetuses withintrauterine growth restriction: longitudinalchanges in ductus venosus and aortic isthmusflow. Ultrasound Obstet Gynecol 2009; 33: 3943.

5/22/2018 Rciu 2014 Clinic

13/13

Update on FGR Fetal Diagn Ther

DOI: 10.1159/000357592

13

64 Evertson LR, et al: Antepartum fetal heart ratetesting. I. Evolution of the nonstress test. AmJ Obstet Gynecol 1979; 133: 2933.

65 Pattison N, McCowan L: Cardiotocographyfor antepartum fetal assessment. CochraneDatabase Syst Rev 2000:CD001068.

66 Dawes GS, Redman CW: Automated analysisof the FHR: evaluation? Am J Obstet Gynecol1992; 167: 19121914.

67 Grivell RM, et al: Antenatal cardiotocographyfor fetal assessment. Cochrane Database SystRev 2010:CD007863.

68 Bracero LA, Morgan S, Byrne DW: Compari-son of visual and computerized interpretationof nonstress test results in a randomized con-trolled trial. Am J Obstet Gynecol 1999; 181:12541258.

69 Manning FA, et al: Fetal assessment based onfetal biophysical profile scoring. VIII. The in-cidence of cerebral palsy in tested and untest-ed perinates. Am J Obstet Gynecol 1998; 178:696706.

70 Manning FA, et al: Fetal biophysical profilescore. VI. Correlation with antepartum um-

bilical venous fetal pH. Am J Obstet Gynecol1993; 169: 755763.

71 Miller DA, Rabello YA, Paul RH: The modi-fied biophysical profile: antepartum testing inthe 1990s. Am J Obstet Gynecol 1996; 174:812817.

72 Kaur S, et al: Biophysical profile in the treat-ment of intrauterine growth-restricted fetuseswho weigh