Bridging the notch: quantification of the end diastolic notch to better predict fetal growth restriction

Abstract

Purpose We aimed to evaluate several quantitative methods to describe the diastolic notch (DN) and compare their performance in the prediction of fetal growth restriction.

Materials and Methods Patients who underwent a placental scan at 16–26 weeks of gestation and delivered between Jan 2016 and Dec 2020 were included. The uterine artery pulsatility index was measured for all of the patients. In patients with a DN, it was quantified using the notch index and notch depth index. Odds ratios for small for gestational age neonates (defined as birth weight <10th and <5th percentile) were calculated. Predictive values of uterine artery pulsatility, notch, and notch depth index for fetal growth restriction were calculated.

Results Overall, 514 patients were included, with 69 (13.4%) of them delivering a small for gestational age neonate (birth weight<10th percentile). Of these, 20 (20.9%) had a mean uterine artery pulsatility index >95th percentile, 13 (18.8%) had a unilateral notch, and 11 (15.9%) had a bilateral notch. 16 patients (23.2%) had both a high uterine artery pulsatility index (>95th percentile) and a diastolic notch. Comparison of the performance between uterine artery pulsatility, notch, and notch depth index using receiver operating characteristic curves to predict fetal growth restriction <10th percentile found area under the curve values of 0.659, 0.679, and 0.704, respectively, with overlapping confidence intervals.

Conclusion Quantifying the diastolic notch at 16–26 weeks of gestation did not provide any added benefit in terms of prediction of neonatal birth weight below the 10th or 5th percentile for gestational age, compared with uterine artery pulsatility index.

Zusammenfassung

Ziel Unser Ziel war es, verschiedene quantitative Methoden zur Beschreibung des diastolischen Notchings (DN) zu evaluieren und deren Leistung bei der Vorhersage einer fetalen Wachstumsrestriktion zu vergleichen.

Material und Methoden Eingeschlossen wurden Patientinnen, bei denen zwischen der 16. und 26. Schwangerschaftswoche eine Plazenta-Untersuchung durchgeführt wurde und die zwischen Januar 2016 und Dezember 2020 entbunden hatten. Bei allen Frauen wurde der Pulsatilitätsindex der A. uterina gemessen. Bei Patientinnen mit DN wurde dieses mittels Notch-Index und des Notch-Tiefenindex quantifiziert. Es wurden Odds-Ratios für SGA-Neugeborene (small for gestational age: Geburtsgewicht <10. und <5. Perzentile) berechnet. Die prädiktiven Werte der Pulsatilität der A. uterina, des Notch-Index und des Notch-Tiefenindex für fetale Wachstumsrestriktion wurden berechnet.

Ergebnisse Insgesamt wurden 514 Patientinnen eingeschlossen, von denen 69 (13,4%) ein SGA-Neugeborenes (Geburtsgewicht <10. Perzentile) zur Welt brachten. Von diesen hatten 20 (20,9%) einen durchschnittlichen Pulsatilitätsindex der A. uterina, der über der 95. Perzentile lag, 13 (18,8%) hatten ein unilaterales Notching und 11 (15,9%) ein bilaterales Notching. Sechzehn Frauen (23,2%) hatten sowohl einen hohen Pulsatilitätsindex der A. uterina (>95. Perzentile) sowie ein diastolisches Notching. Die Leistung des Pulsatilitätsindex der A. uterina sowie des Notch- und des Notch-Tiefenindex bezüglich der Vorhersage einer fetalen Wachstumsrestriktion (<10. Perzentile) wurde mittels der ROC-Kurve (receiver operating characteristic curve) verglichen. Die Werte für die AUC (area under the curve) betrugen 0,659 (PI der A. uterina), 0,679 (Notch) bzw. 0,704 (Notch-Tiefenindex), mit jeweils überlappenden Konfidenzintervallen.

Schlussfolgerung Die Quantifizierung des diastolischen Notchs in der 16.–26. Schwangerschaftswoche brachte im Vergleich zum Pulsatilitätsindex der A. uterina keinen zusätzlichen Nutzen in Bezug auf die Vorhersage eines Geburtsgewichts unterhalb der 10. oder 5. Perzentile.

Keywords

pulsatility index - uterine artery - fetal growth restriction - end diastolic notch

Introduction

Fetal growth restriction (FGR) complicates up to 5–15% of pregnancies and may lead to poor pregnancy outcomes and high rates of maternal and neonatal morbidity and mortality [1] [2]. One of the most clinically relevant etiologies underlying FGR is potential placental disorder. Placental disorders are thought to arise from impedance in uteroplacental blood flow secondary to impaired trophoblastic invasion of the spiral arteries [3]. Changes in perfusion of the placenta and fetus can be reflected as increased vascular resistance of more proximal vessels. The uterine artery (UtA) is of particular interest as fetal activity does not affect its waveform, unlike those of the umbilical artery, umbilical vein, and middle cerebral artery [4] [5] [6] [7]. 

Given the morbidity and mortality associated with placental disorders, uterine artery Doppler (UtA-Dop) screening has been playing a growing role in predicting FGR [4] [5] [6]. The most studied parameters are the pulsatility index (PI), resistance index (RI), and the presence of an early diastolic notch (DN) [4] [5] [6]. While PI and RI are quantitative measures, attempts to quantify the DN for the prediction of FGR are limited. 

The DN is defined as a reduction in forward flow at the start of diastole. It is present in 46–64% of normal gestations in the first trimester and is expected to disappear by 13 weeks of gestation upon establishing low-resistance flow through the placenta [8] [9]. Thus, persistence of the DN beyond 20 weeks is thought to represent abnormal uteroplacental flow and has been associated with adverse outcomes including FGR, preeclampsia, increased risk of preterm delivery, and fetal distress in labor [10] [11] [12]. As early as 1993, Bower and colleagues introduced the systolic/early diastolic ratio as a method to quantify the notch [13]. Others have attempted to quantitatively define the notch [7] [13] [14] [15], but these efforts date back more than two decades and used limited sample sizes with different definitions of FGR. 

Thus, we aim to quantify the diastolic notch in an attempt to better predict the development of FGR by comparing the UtA-Dop characteristics of patients with and without placental disorders. We hypothesize that by numerically analyzing characteristics across the UtA-Dop waveform, we will be able to provide a quantitative assessment of the diastolic notch and correlate it with the development of FGR.

Methods

We conducted a retrospective cohort study of all patients who underwent a placental scan between 16–26 weeks of gestation and who delivered at two tertiary academic centers during a 4-year period. 

Patients with gestational age < 22^0/7 weeks at the time of delivery, multiple pregnancies, or pregnancies complicated by known fetal abnormalities were excluded from the study. Additionally, patients who were treated and transferred to or from another center or missing institutional standard antenatal and postpartum data were excluded as well.

Maternal demographic information, past and current pregnancy history, pre-existing and pregnancy-associated maternal medical conditions (e.g., hypertensive disorders of pregnancy (HDP), gestational diabetes mellitus (GDM)), and pregnancy exposures (e.g., alcohol, smoking, etc.) were extracted. Diagnostic imaging findings from routine ultrasound, anatomy scan, and placental scan were reviewed along with results from first trimester screening for aneuploidy (FTS) and GDM. The UtA-Dop PI was measured for all patients. When a DN was present, it was quantified using the notch index (NI) and notch depth index (NDI) as defined by the relationship between the Doppler wave obtained from the uterine artery ([Fig. 1]).

Fig. 1 Illustration of notch depth index and notch index.

Information surrounding labor, including Bishop score, induction, type of delivery, medications, postpartum complications, and maternal and neonatal outcomes (e.g., congenital abnormalities, neonatal intensive care unit admission) were recorded.

Our primary outcome was the need for induction of labor due to estimated fetal weight < 10th percentile. NI and NDI were compared for patients who developed placental disorders and those who did not.

There are several definitions for fetal growth restriction. For example, the guidelines of the International Society for Ultrasound in Obstetrics and Gynecology from 2020 define early FGR as either EFW/AC below the 3^rd percentile, or below the 10^th percentile with an abnormal UtA PI or UA PI. In comparison, the American College of Obstetricians and Gynecologists defines both SGA and FGR as below the 10^th percentile, but SGA is reserved for birth weight, and FGR for fetuses. As such, we felt the use of FGR in the context of EFW below the 10^th percentile is appropriate in order to apply the findings of the study to a wide variety of guidelines.

All data was analyzed using the SPSS package version 29.0 [IBM, Armonk, NY, USA]. Continuous variables were tested for normal distribution. Variables with normal distribution were compared using the Student T-test, and the rest were compared using the Mann-Whitney U test. Categorical variables were compared using the Chi-square test. The level of significance was pre-determined to be p-values of less than 0.05. Logistic regression and ROC curves were calculated as appropriate. 

Clinical history, demographics, diagnostic imaging, and lab data were obtained retrospectively from the patient’s electronic charts following institutional research ethics board approval. There was no direct patient involvement during the study. All patient data was anonymized and coded. The study protocol was approved by the local institutional review board (REB #4984, approved June 7, 2021). Given the nature of the study, individual informed consent was waived.

Results

Overall, 514 pregnant persons met the inclusion criteria and were included in the analysis, with 69 (13.4%) of them having neonates with birth weight <10^th percentile (FGR). [Table 1] depicts the maternal and pregnancy characteristics. Individuals in the study group had a lower median gravidity and parity. They were also more likely to receive antenatal corticosteroids, or to undergo induction of labor (for any reason, including FGR). FGR neonates were also more likely to be delivered at <28 and <32 weeks of gestation compared to non-FGR neonates (5.8% vs. 0.2%, p<0.001 and 5.8% vs. 1.6%, p=0.02, respectively). No other significant demographic differences were found between the groups ([Table 1]).

Table 1 Maternal and pregnancy characteristics.

Variable	Birth weight < 10th percentile (n=69)	Birth weight ≥ 10th percentile (n=445)	P-value
Data is presented as either median [min.-max., intraquartile range] or n (%) PTB: preterm birth; CD: cesarean delivery; BMI: body-mass index; HTN: hypertension; DM: diabetes mellitus; ART: artificial reproductive technologies; HDP: hypertensive disorder of pregnancy; PE: preeclampsia; w/o: without; w/: with; VD: vaginal delivery Birth weight percentiles are based on Kramer MS, Platt RW, Wen SW, Joseph KS, Allen A, Abrahamowicz M, Blondel B, Bréart G; Fetal/Infant Health Study Group of the Canadian Perinatal Surveillance System. A new and improved population-based Canadian reference for birth weight for gestational age. Pediatrics. 2001 Aug;108(2):E35. DOI: 10.1542/peds.108.2.e35.
Maternal age, years	32 [22–47, 6]	34 [22–51, 6]	0.06
Gravidity, n	2 [1–8, 1]	2 [1–14, 2]	0.01
Parity, n	0 [0–2,1]	1 [0–9,1]	<0.001
Previous PTB, n (%)	4 (5.8)	61 (13.7)	0.07
Previous stillbirth, n (%)	0 (0)	18 (4.0)	0.09
Previous CD, n (%)	7 (10.1)	90 (20.2)	0.05
Pre-pregnancy BMI, kg/m2	23.7 [17.9–56.0, 5.8]	24.1 [14.7–56.4, 7.3]	0.08
Chronic HTN, n (%)	4 (5.8)	10 (2.2)	0.09
Pre-gestational DM, n (%)	2 (2.9)	14 (3.1)	0.91
ART use, n (%)	4 (5.8)	32 (7.2)	0.67
Daily use of aspirin, n (%)	44 (63.8)	259 (58.2)	0.38
Smoking in pregnancy, n (%)	2 (2.9)	8 (1.8)	0.54
Gestational DM, n (%)	8 (11.6)	62 (13.9)	0.60
HDP, n (%)	10 (14.5)	60 (13.5)	0.82
Gestational HTN, n (%)	9 (13.0)	49 (11.0)	0.62
PE w/o severe features, n (%)	1 (1.4)	9 (2.0)	0.75
PE w/ severe features, n (%)	0 (0)	1 (0.2)	0.69
Eclampsia, n (%)	0 (0)	1 (0.2)	0.69
Antenatal administration of corticosteroids, n (%)	8 (11.6)	19 (4.3)	0.01
Induction of labor, n (%)	44 (63.8)	176 (39.6)	<0.001
Due to suspected fetal growth restriction, n (%)	22 (31.9)	4 (0.9)	<0.001
Due to hypertensive disorder of pregnancy, n (%)	3 (4.3)	25 (5.6)	0.67
Gestational age at the time of delivery, weeks	38.4 [25.6–41.3, 1.8]	38.9 [27.7–41.6, 1.6]	0.02
Birth weight, grams	2518 [521–3160, 465]	3201 [839–4692, 525]	<0.001
Spontaneous VD, n (%)	35 (50.7)	228 (51.2)	0.94
Assisted VD, n (%)	7 (10.1)	33 (7.4)	0.43
CD, n (%)	27 (39.1)	184 (41.5)	0.73
PTB <37 weeks of gestation, n (%)	10 (14.5)	49 (11.0)	0.40
PTB <32 weeks of gestation, n (%)	4 (5.8)	7 (1.6)	0.02
PTB <28 weeks of gestation, n (%)	4 (5.8)	1 (0.2)	<0.001
Neonatal gender at birth
Female, n (%)	39 (56.5)	218 (49.0)
Male, n (%)	30 (43.5)	227 (51.0)	0.24
5-minute Apgar score < 5, n (%)	0 (0)	1 (0.2)	0.69

Comparison of placental scan characteristics between FGR and non-FGR groups are shown in [Table 2]. In the FGR group, the mean UtA PI and proportion of abnormal (>95^th percentile) bilateral and unilateral UtA PI measurements were higher compared to the non-FGR group. The proportion of bilateral DN of the uterine arteries, but not unilateral DN, was also higher in the FGR group, as was the proportion of the combination of UtA PI > 95^th percentile and a DN.

Table 2 Placental scan characteristics.

Variable	Birth weight < 10th percentile (n=69)	Birth weight ≥ 10th percentile (n=445)	p-value
Data is presented as either median [min.-max., intraquartile range] or n (%). GA: gestational age at the time of the placental scan; EFW: estimated fetal weight; UA: umbilical artery; PI: pulsatility index; UtA: uterine artery; DN: diastolic notch. *Notch depth index and notch index were measured only when a notch was present (24 cases in the birthweight < 10th percentile group, and 76 cases in the control group).
GA, weeks	21.0 [16.7–24.3, 3.0]	20.6 [16.0–26.0. 2.7]	0.27
EFW percentile, %	33.0 [2.0–87.0, 27.0]	48.0 [2.0–98.0, 34.0]	<0.001
UA PI	1.18 [0.87–2.41, 0.3]	1.15 [0.59–1.90, 0.23]	0.62
Mean UtA PI	1.04 [0.49–3.25, 0.53]	0.96 [0.48–2.21, 0.35]	0.002
Mean UtA PI >95th percentile, n (%)	13 (18.8)	22 (4.9)	<0.001
UtA PI >95th percentile in one or both arteries, n (%)	20 (29.0)	51 (11.5)	<0.001
Unilateral UtA PI >95th percentile, n (%)	13 (18.8)	43 (9.7)	0.02
Bilateral UtA PI >95th percentile, n (%)	7 (10.1)	8 (1.8)	<0.001
DN in one or both of the uterine arteries, n (%)	24 (34.8)	76 (17.1)	0.001
Unilateral DN, n (%)	13 (18.8)	54 (12.1)	0.12
Bilateral DN, n (%)	11 (15.9)	22 (4.9)	0.001
Any UtA PI >95th percentile AND any DN, n (%)	16 (23.2)	29 (6.5)	<0.001
Notch depth index*	0.193 [0.068–0.631, 0.182]	0.135 [0.029–0.811, 0.118]	0.007
Notch index*	0.111 [0.044–0.307, 0.069]	0.082 [0.021–0.519, 0.055]	0.018
Notch depth index > 90th percentile, n (%)*	4 (5.8)	3 (0.7)	0.008
Notch index > 90th percentile, n (%)*	2 (2.9)	1 (0.2)	0.049

NDI and NI were measured only when a notch was present. In total, there were 100 patients with a diastolic notch with 24 cases (13 unilateral, 11 bilateral) in the FGR group and 76 cases (54 unilateral, 22 bilateral) in the non-FGR group. Both NDI and NI were higher in the FGR group, as well as the proportion of NDI >90^th percentile and NI >90^th percentile ([Table 2]).

The independent contribution of each variable as a predictor of FGR was analyzed using a multivariable logistic regression model and is presented in [Table 3]. The presence of a unilateral or bilateral DN (OR=2.06; 95% CI 1.06–4.02, p=0.03), unilateral or bilateral UtA PI >95^th percentile (OR=2.65, 95% CI 1.31–5.37, p=0.007), NDI >90^th percentile (OR=9.21, 95% CI 1.50–56.74, p=0.02), NI >90^th percentile (OR=5.38, 95% CI 1.06–27.19, p=0.04), and chronic hypertension (OR=5.39, 95% CI 1.37–21.24, p=0.02) were significantly associated with FGR. Additionally, the presence of mutually existing UtA PI > 95^th percentile and DN was also found to be highly associated with FGR (OR 5.66, 95% CI 2.71–11.78, p<0.001). Other variables included in the model were not found to be significantly associated with FGR.

Table 3 Logistics regression of independent contributors as predictors of fetal growth restriction.

Variable	Odds ratio	95% CI	p-value
UtA: uterine artery; PI: pulsatility index; NDI: notch depth index; NI: notch index Other variables in the model (non-significant): maternal age, use of aspirin, pre-gestational diabetes mellitus, preeclampsia, gestational hypertension, smoking and body-mass index
Unilateral/bilateral diastolic notch	2.06	1.06–4.02	0.03
Unilateral/bilateral UtA PI > 95th percentile	2.65	1.31–5.37	0.007
NDI > 90th percentile	9.21	1.50–56.74	0.02
NI > 90th percentile	5.38	1.06–27.19	0.04
Any UtA PI>95th percentile AND any DN, n (%)	5.66	2.72–11.78	<0.001
Chronic hypertension	5.39	1.37–21.24	0.02

The performance characteristics of UtA PI, notching, NDI, and NI are presented in [Table 4]. All of the parameters had poor sensitivity and good specificity. The PPV of NDI and NI > 90^th percentile was 57% and 67%, respectively, higher than the other parameters, with a comparable NPV. Subgroup analysis of patients presenting with a unilateral or bilateral diastolic notch is presented in [Table 5]. NDI and NI > 90^th percentile had a moderate PPV (67%) and good NPV (77%).

Table 4 Test characteristics of different Doppler parameters for the prediction of fetal growth restriction in the entire study population.

Test	Sensitivity (%)	Specificity (%)	PPV (%)	NPV (%)	LR+	LR-
LR+: positive likelihood ratio; LR-: negative likelihood ratio; UtA: uterine artery; PI: pulsatility index; NDI: notch depth index; NI: notch index
Unilateral diastolic notch	18.8	87.9	19.4	87.6	1.56	0.92
Bilateral diastolic notch	15.9	97.1	45.8	88.2	5.49	0.87
Any notch	34.8	82.8	24.0	89.0	2.02	0.79
Unilateral UtA PI > 95th percentile	18.8	90.4	23.2	87.9	1.96	0.90
Bilateral UtA PI > 95th percentile	10.1	98.2	46.7	87.6	5.68	0.91
Any UtA PI > 95th percentile	29.0	88.6	28.2	89.0	2.55	0.80
NDI > 90th percentile	5.8	99.3	57.1	87.3	8.66	0.95
NI > 90th percentile	2.9	99.8	66.7	87.0	12.99	0.97
Any UtA PI>95th percentile AND any DN, n (%)	23.2	93.5	35.6	88.7	3.56	.082

Table 5 Test characteristics of different Doppler parameters for the prediction of fetal growth restriction only in patients with a diastolic notch.

Test	Sensitivity (%)	Specificity (%)	PPV (%)	NPV (%)	LR+	LR-
LR+: positive likelihood ratio; LR-: negative likelihood ratio; UtA: uterine artery; PI: pulsatility index; NDI: notch depth index; NI: notch index
Unilateral UtA PI > 95th percentile	37.5	68.4	27.3	77.6	1.19	0.91
Bilateral UtA PI > 95th percentile	29.2	93.4	58.3	80.7	4.43	0.76
Any UtA PI > 95th percentile	66.7	61.8	35.6	85.5	1.75	0.54
NDI > 90th percentile	8.3	98.7	66.7	77.3	6.33	0.93
NI > 90th percentile	8.3	98.7	66.7	77.3	6.33	0.93
Any UtA PI >95th percentile AND any DN, n (%)	66.7	61.8	35.6	85.5	1.75	0.54

The area under the ROC curve was calculated to compare the mean UtA PI, mean NDI, and mean NI for predicting FGR <10th, 5th, and 3rd percentiles ([Fig. 2]). For all three thresholds, the predictive ability of the mean UtA PI, mean NDI, and mean NI was moderate to poor. The AUC for predicting FGR <10th percentile was 0.659 for UtA PI, 0.704 for mean NDI, and 0.679 for mean NI ([Fig. 2]a). The AUC for FGR <5th percentile and FGR <3rd percentile was similar ([Fig. 2]b,c).

Fig. 2 Receiver operating characteristic curves for the prediction of birth weight below the 10th percentile a, 5th percentile b, and 3rd percentile for gestational age. UtA – uterine artery; NI – notch index; NDI – notch depth index.

Discussion

The aim of the current study was to quantify the early diastolic notch using NI and NDI to assess its predictive ability for FGR. Our results demonstrate that although NDI and NI were predictors of FGR, neither provided additional benefit when compared to UtA PI for predicting FGR. However, in the presence of a DN, UtA PI, NI, or NDI >90th percentile can aid in the diagnosis of FGR.

An abnormal UtA-Dop, defined as a PI >95th percentile for gestational age, or the presence of a DN, has been associated with adverse pregnancy outcomes including FGR. Previous studies found the persistence of a DN past 20 weeks to be strongly associated with uteroplacental insufficiency and FGR [16] [17] [18]. Similar to those studies, we also found the mean UtA PI, the proportion of UtA PI >95th percentile, and the prevalence of DN to be higher among the FGR group compared to the non-FGR group. The mean NDI, NI, and proportion of NDI and NI >90th percentile were similarly higher in the FGR group.

An ideal screening program includes a high sensitivity to identify at-risk patients and a high specificity to avoid implementation of expensive, excessive, or possibly harmful interventions. Our study demonstrated that among patients at risk for FGR based on second-trimester uterine Doppler, UtA PI and notching had poor sensitivity, a moderate positive predictive value, and a positive likelihood ratio for FGR. All parameters had a high specificity and strong negative predictive value. As such, it may not be necessary to calculate all uterine artery Doppler parameters in patients with abnormal first trimester screening if another parameter with comparable sensitivity and specificity is known.

Previous studies have similarly shown poor clinical utility of the DN and UtA PI in predicting FGR. A systematic review by Chien et al. including 12,994 patients across 27 studies found limited diagnostic accuracy in predicting FGR using UtA-Dop flow velocity [19]. Stratification of FGR by low-risk pregnancies (14 studies) and high-risk pregnancies (9 studies) to examine the predictability of flow waveform ratio and diastolic notch found a pooled likelihood ratio of 3.6 (95% CI 3.2–4.0) and 2.7 (95% CI 2.1–3.4) for a positive result, and a pooled likelihood ratio of 0.8 (95% CI 0.8–0.9) and 0.7 (95% CI 0.6–0.9) for a negative result, respectively. Examining the DN in isolation showed improved applicability. However, the patient sample size was significantly smaller (2 studies) [19]. More recent cohort studies continue to support the limited use of UtA PI and DN studies for predicting FGR babies [17] [20] [21].

We demonstrated that quantifying the DN with NI and NDI indices as part of initial screening for FGR has limited clinical utility due to its poor sensitivity, moderate positive predictive value and positive likelihood ratio. However, for patients with a diastolic notch, quantifying the notch offers a more accurate approach for ruling out or ruling in FGR. While the notch, in and of itself, provides limited predictive value ([Table 4]a), when the notch is quantified, by using PI>95^th percentile, NDI >90th percentile, or NI>90th percentile, the test characteristics improve. This can be of clinical importance and help reassure patients with a DN when the quantitative elements are favorable or trigger enhanced surveillance if they are less favorable.

An early study by Ohkuchi et al. investigated 288 patients between 16–23 weeks of gestation with healthy pregnancies in Japan and reported an optimal NDI cut-off of 0.14 for predicting an FGR neonate [7]. NDI >0.14 yielded a higher PPV (22%) than that of other parameters, namely RI (9%) and A/C ratio (12%). However, UtA PI and the presence of DN, which are more commonly used predictors of FGA, were not investigated. A similar study by Aardema et al. also investigated a new method of diastolic notch quantification, comparing PI and NI as predictors of HDP [15]. While the results suggest that both PI and NI are poor predictors of HDP, they do not indicate any association with FGR. Our study used a larger and more multi-ethnic sample to build upon these previous studies, demonstrating superior PPV of NI >90th percentile (66.7%) when compared with NDI >90th percentile (57.1%), UtA PI >95th percentile (28.2%), and the presence of DN (24.0%).

The findings of our study are in keeping with those of Bower et al., who found the sensitivity and positive predictive value for NDI to be 94% and 41%, respectively [13]. They also found an NDI cut-off value of 0.15 for differentiating between mildly abnormal and moderate-severe abnormal waveforms. However, the study was a small retrospective study with 50 highly selected participants, thus reducing the generalizability of the results. A large retrospective study completed by Becker et al. found poor sensitivity and positive predictive value for FGR using NI even when stratified by degree of DN depth [22]. The use of NDI measurements in the third trimester is also limited [23]. Given these results, there is limited evidence to suggest the implementation of UtA PI, notching, NDI, or NI variables as screening markers for FGR. Currently, its use for routine clinical management of early- or late-onset FGR is not recommended [24].

Lastly, area under the curve calculations comparing UtA PI, mean NDI, and mean NI for predicting FGR <10th, 5th and 3rd percentiles were poor to moderate, with there being no difference between the three variables. Our findings are supported by other studies evaluating a combination of these indices, but no study evaluating all three exists [13] [22].

Future research should continue to evaluate the predictive capacity of NI and NDI with comparison to UtA PI and the presence of DN in a prospective manner and in the context of multiple pregnancies. In addition, the study of biomarkers such as Pregnancy Associated Placental Protein A (PAPP-A), Alpha-FetoProtein (AFP), and Placental Growth Factor (PlGF) together with uterine artery Doppler waveforms would be a pertinent area of further study to aid the development of more sensitive and specific predictive models for FGR.

The main strengths of the current study lie in the large sample size taken from two large tertiary academic hospitals catering to a multi-ethnic patient population. Potential confounding was minimized by accounting for multiple pregnancies, known fetal abnormalities, and patients with gestational age less than 22 weeks. This is also the most recent study to have quantified the diastolic notch, with previous efforts dating back more than two decades. Thus, the current study follows the most updated antenatal care model and employs current ultrasonography technologies. Our study is also the first to have compared different models of end diastolic notch quantification (NDI vs. NI) at the same time as comparisons to current predictors of FGR, including UtA PI >95th percentile and the presence of DN. Finally, all Doppler waveform calculations were performed with research staff blinded to the pregnancy outcomes.

There are limitations to our study, including its retrospective nature and relatively small sample size. Additionally, the geographic proximity of the two tertiary hospitals and the fact that some FGR infants were not referred introduce potential selection bias towards patients of certain SES or those with greater risk factors or more severe presentations for FGR who were more likely to be identified antenatally.

Our data demonstrate that both NI and NDI >90th percentile measured between 16–26 weeks of gestation are predictors of FGR <10th, <5th, and 3rd percentiles but were moderate to poor in their predictive ability and less sensitive than UtA PI >95th percentile as a predictor for FGR. Future research should continue to evaluate the predictive capacity of NI and NDI in a prospective manner and in the context of multiple pregnancies. The results of our study suggest that routine screening of FGR does not offer additional benefit to UtA PI >95th percentile and the presence of DN. However, given the presence of a DN, an NI or NDI >90th percentile can help to rule in FGR.

Conflict of Interest

The authors declare that they have no conflict of interest.

• References

• 1 Hutcheon JA, Lisonkova S, Joseph KS. Epidemiology of pre-eclampsia and the other hypertensive disorders of pregnancy. Best Pract Res Clin Obstet Gynaecol 2011; 25: 391-403
• 2 Suhag A, Berghella V. Intrauterine Growth Restriction (IUGR): Etiology and Diagnosis. Curr Obstet Gynecol Rep 2013; 2: 102-111
• 3 Prefumo F, Sebire NJ, Thilaganathan B. Decreased endovascular trophoblast invasion in first trimester pregnancies with high-resistance uterine artery Doppler indices. Hum Reprod 2004; 19: 206-209
• 4 Kennedy AM, Woodward PJ. A Radiologist’s Guide to the Performance and Interpretation of Obstetric Doppler US. Radiographics 2019; 39: 893-910
• 5 Tercanli S, Kagan KO, Pertl B. Integrating Doppler Ultrasound into Obstetrics Management. Ultraschall in Med 2023; 44 (01) 10-13
• 6 Faber R, Heling KS, Steiner H. et al. Doppler ultrasound in pregnancy – quality requirements of DEGUM and clinical application (part 2). Ultraschall in Med 2021; 42 (05) 541-550
• 7 Ohkuchi A, Minakami H, Sato I. et al. Predicting the risk of pre-eclampsia and a small-for-gestational-age infant by quantitative assessment of the diastolic notch in uterine artery flow velocity waveforms in unselected women. Ultrasound Obstet Gynecol 2000; 16: 171-178
• 8 Coppens M, Loquet P, Kollen M. et al. Longitudinal evaluation of uteroplacental and umbilical blood flow changes in normal early pregnancy. Ultrasound Obstet Gynecol 1996; 7: 114-121
• 9 Harman CR, Baschat AA. Comprehensive assessment of fetal wellbeing: which Doppler tests should be performed?. Curr Opin Obstet Gynecol 2003; 15: 147-157
• 10 Figueras F, Caradeux J, Crispi F. et al. Diagnosis and surveillance of late-onset fetal growth restriction. Am J Obstet Gynecol 2018; 218: S790-S802.e1
• 11 Kingdom JC, Audette MC, Hobson SR. et al. A placenta clinic approach to the diagnosis and management of fetal growth restriction. Am J Obstet Gynecol 2018; 218: S803-S817
• 12 Khalil A, Thilaganathan B. Role of uteroplacental and fetal Doppler in identifying fetal growth restriction at term. Best Pract Res Clin Obstet Gynaecol 2017; 38: 38-47
• 13 Bower S, Kingdom J, Campbell S. Objective and subjective assessment of abnormal uterine artery Doppler flow velocity waveforms. Ultrasound Obstet Gynecol 1998; 12: 260-264
• 14 Aquilina J, Barnett A, Thompson O. et al. Comprehensive analysis of uterine artery flow velocity waveforms for the prediction of pre-eclampsia. Ultrasound Obstet Gynecol 2000; 16: 163-170
• 15 Aardema MW, DE Wolf BT, Saro MC. et al. Quantification of the diastolic notch in Doppler ultrasound screening of uterine arteries. Ultrasound Obstet Gynecol 2000; 16: 630-634
• 16 Dave A, Joshi R, Sooruthiya S. et al. Role of uterine artery Doppler in prediction of FGR in high risk pregnancies in 20–24 weeks. Int J Reprod Contraception, Obstet Gynecol 2017; 6: 1388
• 17 Phupong V, Dejthevaporn T, Tanawattanacharoen S. et al. Predicting the risk of preeclampsia and small for gestational age infants by uterine artery Doppler in low-risk women. Arch Gynecol Obstet 2003; 268: 158-161
• 18 Papageorghiou AT, Yu CK, Bindra R. et al. Multicenter screening for pre-eclampsia and fetal growth restriction by transvaginal uterine artery Doppler at 23 weeks of gestation. Ultrasound Obstet Gynecol 2001; 18: 441-449
• 19 Chien PF, Arnott N, Gordon A. et al. How useful is uterine artery Doppler flow velocimetry in the prediction of pre-eclampsia, intrauterine growth retardation and perinatal death? An overview. BJOG 2000; 107: 196-208
• 20 Cruz-Martinez R, Savchev S, Cruz-Lemini M. et al. Clinical utility of third-trimester uterine artery Doppler in the prediction of brain hemodynamic deterioration and adverse perinatal outcome in small-for-gestational-age fetuses. Ultrasound Obstet Gynecol 2015; 45: 273-278
• 21 Parry S, Sciscione A, Haas DM. et al. Role of early second-trimester uterine artery Doppler screening to predict small-for-gestational-age babies in nulliparous women. Am J Obstet Gynecol 2017; 217: 594.e1-594.e10
• 22 Becker R, Vonk R. Doppler sonography of uterine arteries at 20–23 weeks: depth of notch gives information on probability of adverse pregnancy outcome and degree of fetal growth restriction in a low-risk population. Fetal Diagn Ther 2010; 27: 78-86
• 23 Park YW, Cho JS, Choi HM. et al. Clinical significance of early diastolic notch depth: uterine artery Doppler velocimetry in the third trimester. Am J Obstet Gynecol 2000; 182: 1204-1209
• 24 Martins JG, Biggio JR. Society for Maternal-Fetal Medicine (SMFM). et al. Society for Maternal-Fetal Medicine Consult Series #52: Diagnosis and management of fetal growth restriction: (Replaces Clinical Guideline Number 3, April 2012). Am J Obstet Gynecol 2020; 223: B2-B17

Publication History

Received: 11 August 2023

Accepted after revision: 31 January 2024

Accepted Manuscript online:
31 January 2024

Article published online:
29 April 2024