Is pathological fetal Doppler cerebroplacental ratio (CPR) associated with adverse delivery outcomes in pregnancies complicated by fetal growth restriction undergoing labor induction? A retrospective cohort study

Abstract

Objective

The optimal delivery approach for fetal growth restriction (FGR) with pathological cerebroplacental ratio (CPR) remains uncertain. This study evaluated the association between isolated pathological CPR (<5th percentile) and delivery outcomes, specifically the rate of cesarean delivery (CD) due to non‐reassuring fetal heart rate (NRFHR), in FGR cases undergoing induction of labor (IOL).

Methods

In this retrospective cohort study (2014–2023), pregnancies with FGR (estimated fetal weight < 10th percentile) undergoing IOL were stratified by CPR values: pathological (<5th percentile) and normal CPR. Exclusion criteria included multiple gestations, abnormal umbilical artery or ductus venosus Doppler, abnormal non‐stress test or biophysical profile, major fetal anomalies, and genetic disorders.

Results

Among 432 FGR, 86 had pathological CPR. This group had higher rates of second‐trimester human chorionic gonadotropin (HCG) >2.5 multiples of the median (11.7% vs. 3.8%, P = 0.01) and delivered earlier (37.1 vs. 38.0 weeks, P < 0.01). Overall CD rates and CD due to NRFHR were comparable between groups (19.8% vs. 18.8%, P = 0.84; 14.0% vs. 15.9%, P = 0.66, respectively). Neonates with pathological CPR had lower birthweights (2174 ± 279 g vs. 2279 ± 278 g, P < 0.01) and higher neonatal intensive care unit (NICU) admission rates (17.4% vs. 9.2%, P = 0.03). Logistic regression analysis adjusted for potential confounders showed no independent association between pathological CPR and increased risk of CD due to NRFHR (P = 0.74).

Conclusion

In FGR pregnancies undergoing IOL, isolated pathological CPR is not associated with an increased risk of cesarean delivery for NRFHR. However, these fetuses remain at higher risk for NICU admissions, emphasizing the need for individualized management and close monitoring.

1. INTRODUCTION

Fetal growth restriction (FGR) is a significant complication in obstetric care contributing to increased rates of perinatal morbidity and mortality. ¹ These fetuses are particularly vulnerable to severe early neonatal complications, such as metabolic disturbances, respiratory distress, and organ dysfunction, as well as long‐term consequences like neurodevelopmental impairments and metabolic syndrome. ² , ³ Early identification and tailored management are essential for optimizing outcomes in FGR pregnancies, yet strategies for labor and delivery in the context of pathological Doppler findings, particularly cerebroplacental ratio (CPR), remain an area of clinical uncertainty.

Doppler velocimetry is pivotal in evaluating placental function and fetal well‐being in FGR cases, with the CPR serving as a critical marker of fetal adaptation to placental insufficiency. A pathological CPR, defined as <5th percentile, reflects a brain‐sparing response, indicating that the fetus is redistributing blood flow toward the brain in response to compromised placental function. ⁴ , ⁵ , ⁶ This adaptation has been linked to increased risks of perinatal morbidity, including stillbirth, and adverse neonatal outcomes. ⁶ More recently, attention has focused on how a pathologic CPR might affect intrapartum outcomes, especially with respect to the fetus's ability to tolerate labor.

In cases of placental‐mediated FGR, fetal intolerance to labor is a significant concern, often resulting in non‐reassuring fetal heart rate (NRFHR) patterns and necessitating emergent cesarean delivery (CD). ⁷ , ⁸ Notably, in severe FGR cases with late‐stage umbilical artery Doppler abnormalities, such as absent or reversed end‐diastolic velocity, CD rates exceed 80%, making primary cesarean the favored approach of delivery in these situations. ⁹ , ¹⁰ However, when FGR is complicated by “isolated” pathologic CPR without additional Doppler abnormalities, management becomes less clear. These fetuses may not show overt signs of placental insufficiency, yet their ability to tolerate labor remains a concern.

The consensus supports vaginal delivery in cases of FGR with normal Doppler studies and with estimated fetal weight (EFW) between the 3rd and 10th percentiles. ⁷ , ¹¹ Yet, the clinical course of FGR fetuses with “isolated” pathologic CPR is less well understood. Although pathologic CPR has been associated with an increased risk of adverse neonatal outcomes, its predictive value for CD due to intrapartum complications and distress, specifically NRFHR, has not been fully elucidated. This presents a significant challenge in determining whether induction of labor (IOL) is a safe option for these pregnancies or whether CD should be considered to preempt potential fetal compromise during labor.

We aimed to investigate the association between “isolated” pathologic CPR (<5th percentile) and delivery outcomes, specifically focusing on CD rates due to NRFHR in FGR cases undergoing IOL. By exploring the relationship between pathologic CPR and intrapartum outcomes, we seek to provide clearer guidance on the management of this common clinical scenario in modern obstetric practice.

2. METHODS

This retrospective cohort study included all singleton pregnancies complicated by FGR ≥34 weeks, that underwent IOL, at the high‐risk unit, between January 2014 and December 2023 at Meir Medical Center. FGR was defined as an EFW below the 10th percentile. ¹² All prenatal ultrasound EFW measurements and Doppler flow analyses were performed at the departmental prenatal ultrasound unit. Doppler measurements were routinely performed in all fetuses once an EFW below the 10th percentile was identified. The Doppler results used were obtained from the last sonographic examination performed up to 7 days before labor. Color pulsed Doppler studies were performed with 2–6 MHz curved array probes. In all participants, gestational age was confirmed by first‐trimester ultrasonography. CPR was calculated by dividing the Doppler PI of the middle cerebral artery and the UA. EFW percentiles were determined using the Hadlock formula. ¹² Follow‐up Doppler measurements were performed weekly.

Notably, only neonates born small for gestational age (SGA), defined as birthweight below the 10th percentile for gestational age, were included in the study evaluation. ¹²

The study excluded multiple pregnancies; cases with pathological UA flow, defined as absent or reversed end‐diastolic velocity (UA A/REDV) or elevated PI; ductus venosus Doppler abnormalities; non‐reassuring fetal status, as indicated by abnormal non‐stress test or oligohydramnios; major fetal anatomical or genetic anomalies; and cases with confirmed intrauterine fetal infection.

The cohort was divided into two groups based on the CPR measurements obtained within 1 week prior to delivery: pathological CPR group, defined as CPR <5th percentile for gestational age and normal CPR group, defined as CPR ≥5th percentile. ¹³

Indications for IOL included maternal hypertensive disorders (gestational hypertension, chronic hypertension, and preeclampsia), term gestational age, preterm premature rupture of membranes or premature rupture of membranes, and decreased fetal movements. In accordance with departmental protocol, labor induction was indicated at 37.0 weeks of gestation in pregnancies with EFW below the 10th percentile and pathological CPR or in cases of EFW below the 3rd percentile. Otherwise, expectant management was pursued, patients were monitored with weekly Doppler measurements, and labor induction was postponed until 39.0 weeks of gestation.

Method of induction was selected by the attending physician based on established clinical guidelines and individual patient characteristics. Induction techniques included pharmacological cervical ripening with prostaglandins (either dinoprostone vaginal insert or misoprostol), mechanical dilation using an extra‐amniotic double‐balloon catheter (Atad catheter) inflated with 80 mL of saline, and intravenous oxytocin administered following cervical ripening, titrated to achieve effective uterine contractions.

The primary outcome was CD rate due to NRFHR. NRFHR was diagnosed in the presence of recurrent late decelerations, repetitive prolonged decelerations, recurrent variable decelerations with loss of variability or fetal tachycardia, or sustained fetal bradycardia.

Secondary outcomes included a composite measure of adverse perinatal outcomes, defined as the presence of at least one of the following: neonatal intensive care unit (NICU) admission, neonatal hypoglycemia, need for phototherapy, or requirement for neonatal respiratory support.

2.1. Data

Maternal demographic data, obstetric history, and fetal ultrasonographic measurements were extracted from medical records. Maternal demographic data included age, body mass index (BMI, kg/m²), medical history, gestational diabetes mellitus (GDM) A1 and A2 as well as pre GDM, hypertensive disorders (including preeclampsia, chronic and gestational hypertension), and obstetric history. Obstetrics characteristics and outcomes included gestational age at delivery, mode of induction, indication for induction, mode of delivery, and CD indication. Neonatal characteristics and outcomes included birthweight, neonatal gender, betamethasone treatment, Apgar scores, NICU admission, hypoglycemia (defined as blood glucose<40 gr/dl), need for phototherapy, and/or any need for respiratory support.

2.2. Sample size calculation

To detect a significant difference between the normal CPR group and the pathological CPR group in the primary outcome of CD due to NRFHR, with the primary outcome rate expected to be 15% in the normal CPR group and a potential doubling to 30% in the research group, ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ the required sample size for each group is approximately 60 participants. This calculation assumes a two‐tailed test with a significance level of 0.05 and 80% power. The resulting sample size ensures that the study will have a sufficient number of participants to identify reliably any meaningful difference in the primary outcome between the two groups, thereby providing robust statistical evidence for the effectiveness of the intervention.

2.3. Statistical analysis

Descriptive statistical analysis was performed to assess differences in maternal and neonatal characteristics between pregnancies with normal and pathological CPR. Categorical variables were presented as numbers and percentages, while continuous variables were summarized using means and standard deviations. Group differences for categorical variables were evaluated using Pearson's χ ²‐test or Fisher's exact test, as appropriate. An independent sample t‐test was used for continuous variables to compare normally distributed data, while the Mann–Whitney U‐test was applied to non‐normally distributed data.

Multivariable logistic regression model was performed to examine the association between pathological CPR and the primary outcome, adjusting for confounders. Results were reported as adjusted odds ratios (aORs) with 95% confidence intervals (CIs). Statistical significance was set at a two‐sided P‐value <0.05. All analyses were performed using R statistical software version 4.3.2.

The study was approved by the institutional review board of Meir Medical Center, and the need for informed consent was waived due to the retrospective nature of the study.

3. RESULTS

During the study period, 432 pregnancies complicated with FGR cases met the inclusion criteria; of them, 86 were included in the pathological CPR group. Table 1 presents maternal characteristics and obstetrics outcomes of the study groups. There were no between‐group differences in maternal age, gravidity, parity or maternal gestational diabetes and hypertensive disorders. As compared to the normal CPR group, the pathological CPR group was characterized by a higher rate of second trimester human chorionic gonadotropin >2.5 multiples of the median, 11.7% versus 3.8%, respectively, P = 0.01 and lower gestational age at delivery, 37.1 weeks versus 38.0 weeks, respectively, P < 0.01. Overall, CD and specifically CD due to NRFHR rates did not differ between the normal and the pathological CPR groups,19.8% versus 18.8%, respectively, P = 0.84, and 14.0% versus 15.9%, respectively, P = 0.66. Notably, the mode of IOL did not differ as well between the normal CPR versus the pathological CPR group: prostaglandins (PG) 39.5% versus 35.8%, respectively, P = 0.52; Atad catheter 25.6% versus 28.9%, respectively, P = 0.54; both methods (PG and subsequently Atad catheter) 17.4% versus 17.1%, respectively, P = 0.93; oxytocin 17.5% versus 18.2%, respectively, P = 0.93.

TABLE 1.

Maternal characteristics and primary outcome of the study groups.

Variables	Normal CPR (N = 346)	Pathologic CPR (N = 86)	P‐value
Maternal age (years)	31.1 ± 5.5	31.8 ± 5.3	0.29
Gestational age (weeks)	38.0 ± 1.3	37.1 ± 1.5	<0.01
Gravidity	2.2 ± 1.4	2.2 ± 1.6	0.38
Parity	0.8 ± 1.0	0.6 ± 0.9	0.17
BMI (k/m2)	27.4 ± 5.2	27.6 ± 4.8	0.74
BMI >30 (kg/m2)	81 (23.4%)	19 (22.1%)	0.80
GDM, n (%)	18 (5.7%)	8 (10.3%)	0.15
Maternal hypertensive disorder, n (%)	16 (4.6%)	6 (7.0%)	0.41
Second trimester HCG >2.5 MoM	12 (3.8%)	9 (11.7%)	0.01
Maternal smoking, n (%)	43 (12.5%)	8 (9.3%)	0.42
Total CD, n (%)	65 (18.8%)	17 (19.8%)	0.84
CD due to NRFHR, n (%)	55 (15.9%)	12 (14.0%)	0.66
Instrumental delivery, n (%)	28 (8.1%)	12 (14.0%)	0.09

Note: Data are presented as n (%) or mean ± standard deviation.

Abbreviations: BMI, body mass index (kg/m²); CD, cesarean delivery; CPR, cerebroplacental ratio; GDM, gestational diabetes mellitus; HCG, human chorionic gonadotropin; hypertensive disorders—include preeclampsia, gestational and chronic hypertension; MoM, multiples of the median; NRFHR, non‐reassuring fetal heart rate.

Neonatal characteristics and outcomes are presented in Table 2. The pathological CPR group had lower birthweights (2174 ± 279 g vs. 2279 ± 278 g, respectively, P < 0.01), with no significant difference in birthweight below the 3rd percentile (77.9% vs. 79.7%, respectively, P = 0.92) and higher NICU admission rate (17.4% vs. 9.2%, respectively, P = 0.03) as compared to the normal CPR group. Neonates from the pathological CPR group had a tendency toward an increased rate of composite adverse neonatal outcomes as compared to the normal CPR group (52.6% vs. 40.3%, P = 0.05). Notably, the pathological CPR group had a higher rate of betamethasone treatment (24.4% vs. 13.0%, respectively, P = 0.01).

TABLE 2.

Neonatal characteristics and outcomes of the study groups.

Variables	FGR with normal CPR (n = 46)	FGR with pathologic CPR (n = 86)	P‐value
Birth weight (grams)	2279 (278)	2174 (279)	<0.01
Birth weight <3rd percentile	276 (79.7%)	67 (77.9%)	0.92
Gender male	176 (51.0%)	40 (46.5%)	0.45
Apgar at 5 min <7	4 (1.2%)	1 (1.2%)	0.99
NICU admission	32 (9.2%)	15 (17.4%)	0.03
Hypoglycemia	41 (12.0%)	13 (15.3%)	0.41
Phototherapy	44 (12.9%)	13 (15.1%)	0.58
Neonatal respiratory support	35 (10.3%)	7 (8.2%)	0.57
Composite adverse neonatal outcome	127 (40.3%)	41 (52.6%)	0.05

Note: Data are presented as n (%) or mean ± standard deviation. Hypoglycemia, defined as neonatal blood glucose levels <40 mg/dL. Composite adverse neonatal outcome included the presences of at least one of the following outcomes: NICU hospitalization, neonatal hypoglycemia, phototherapy, neonatal respiratory support.

Abbreviation: NICU, neonatal intensive care unit.

Multivariable logistic regression models were conducted to examine the associations with CD due to NRFHR, adjusting for maternal age, parity, BMI (kg/m²), gestational age, GDM, birth weight percentile, and pathological CPR. Only maternal age, BMI (kg/m²), and parity were found to be independently associated with CD due to NRFHR (Table 3).

TABLE 3.

Association between maternal and neonatal variables and CD due to NRFHR.

Variable	Adjusted OR (95% CI)	P‐value
Pathologic CPR	0.87 (0.36, 1.95)	0.74
Maternal age	1.08 (1.02, 1.15)	0.02
Gestational age	1.09 (0.87, 1.39)	0.44
Parity	0.54 (0.34, 0.80)	<0.001
BMI	1.12 (1.05, 1.19)	<0.001
GDM	1.81 (0.58, 5.10)	0.28
Birth weight <3rd percentile	1.92 (0.78, 5.53)	0.18

Note: Adjustment was performed for maternal age, parity, body mass index (BMI, kg/m²), gestational age, gestational diabetes mellitus (GDM), birth weight percentile below 3rd, and pathological cerebroplacental ratio (CPR).

Abbreviations: CD, cesarean delivery; CI, confidence interval; NRFHR, none reassuring fetal heart rate; OR, odds ratio.

4. DISCUSSION

The current study demonstrates that IOL in fetuses with fetal growth restriction (FGR) and “isolated” CPR Doppler values below the 5th percentile was not associated with an increased risk of CD due to NRFHR, compared to FGR fetuses with normal CPR values.

The optimal mode of delivery for FGR cases with isolated pathological CPR remains a subject of ongoing debate. Doppler velocimetry, particularly CPR, plays a central role in the assessment and management of pregnancies complicated by FGR. ⁴ A CPR below the 5th percentile reflects placental insufficiency, which leads to fetal blood flow redistribution. This hemodynamic adaptation might indicate an increased risk of labor intolerance, often manifesting as NRFHR patterns necessitating emergent CD. ⁷ CPR might provide additional information when UA Doppler is normal, reflecting subtle redistribution of blood flow. ¹⁸ However, while some studies suggest that pathological CPR is associated with labor intolerance, others argue that its predictive value for CD due to NRFHR is limited. ¹⁹

A meta‐analysis by Conde‐Agudelo et al., ²⁰ which aimed to evaluate the predictive performance of CPR for adverse perinatal outcomes in FGR fetuses, found CPR to be a strong predictor of perinatal death and acidosis ⁶ but a poor predictor of CD due to NRFHR, with a sensitivity of only 59%.

Consistent with our findings, Jo et al. ²¹ studied a cohort of 184 SGA fetuses (<5th percentile), categorized into normal CPR, pathological CPR (<1.08), and pathological CPR combined with elevated uterine artery pulsatility index (UA PI). They found that pathological CPR alone was not independently associated with an increased rate of CD due to NRFHR, although the combination with elevated UA PI was significantly associated.

In contrast, Figueras et al. ²² observed in a cohort of 509 pregnancies with FGR fetuses (<10th percentile) a significantly higher rate of fetal distress and CD during labor induction in those with pathological CPR (<10th percentile). The discrepancy between these results and ours might be explained by differences in study populations and methodologies; for instance, in their study, fetuses with pathological CPR had significantly lower birthweight percentiles compared to those in the normal CPR group. Similarly, García‐Simón et al., ¹⁷ in a smaller cohort of 164 SGA fetuses (EFW <10th percentile), reported an increased risk of emergency CD for fetal distress in those with pathological CPR (<5th percentile) compared to fetuses with normal CPR during labor induction. In this study, CPR measurements were performed within 24 h before labor induction.

In our study, the overall rates of CD and CD due to NRFHR were similar between FGR fetuses with and without pathological CPR. Notably, although the pathological CPR group had significantly lower birthweights, both groups had similar rates of birthweight below the 3rd percentile (approximately 80% in each group). To adjust for residual confounders, we conducted a multivariable analysis including maternal age, gestational age at delivery, parity, BMI, gestational diabetes, and birthweight percentile. Pathological CPR was not found to be independently associated with an increased risk of CD for NRFHR.

As expected, we observed a higher rate of NICU admission among neonates in the pathological CPR group. This finding is likely attributable to the lower gestational age at delivery in this group, despite a higher rate of antenatal corticosteroid (betamethasone) administration. ²³ Our results align with previous studies reporting a strong association between pathological CPR and adverse neonatal outcomes, particularly NICU admission. ⁵ , ²² , ²⁴ These findings underscore the vulnerability of growth‐restricted fetuses with Doppler abnormalities and the importance of individualized delivery planning and close monitoring.

The clinical relevance of our findings lies in the reassurance they provide regarding the safety of labor induction in pregnancies complicated by FGR with isolated pathological CPR Doppler findings. These fetuses appear to tolerate labor induction well. Nonetheless, clinicians should remain vigilant, as even SGA fetuses with normal Doppler studies demonstrate a reduced ability to withstand intrapartum stress compared to appropriately grown fetuses. ⁷ , ⁸

A key strength of our study lies in its focus on a clearly defined population of FGR pregnancies with isolated pathological CPR, without additional Doppler abnormalities (as described in previous studies ¹⁷ , ²¹ , ²² ), contributing to the standardized studied population. Second, data was retrieved from a single medical center with unified treatment protocols, making the studied population homogenous. Third, there were similar rates of methods of induction, with prostaglandins and mechanical methods, emphasizing the similar effectiveness of the methods. Nevertheless, there are some limitations that should be acknowledged. First, the retrospective nature of our study might introduce selection bias and limit the ability to establish causality. Another limitation of our study is the lack of systematic cord acid–base data, which would have allowed a more direct assessment of intrapartum hypoxia. These measurements were not consistently available across the 10‐year study period and could not be analyzed. However, neonatal condition was indirectly assessed by Apgar scores at 1 and 5 min, the need for NICU admission, hypoglycemia, respiratory support, and phototherapy, which were reliably documented. In addition, intrapartum monitoring was performed with continuous cardiotocography, and non‐reassuring fetal heart rate was clearly defined in our methods. While these measures cannot replace cord gases, they provide meaningful clinical proxies of neonatal compromise. Future prospective studies should incorporate systematic cord blood gas analysis to validate these findings. Second, while we adjusted for multiple confounders, there might still have been an influence on our findings. Moreover, the fact that CPR was measured within 1 week prior to delivery might limit the accuracy of our findings. This limitation is derived from our departmental protocols of performing weekly Doppler measurement, consistent with internationally accepted guidelines. ²⁵ In addition, Doppler measurements were not repeated once a pathological CPR was identified, which might have increased the likelihood of false‐positive cases due to interobserver variability. ¹⁸ , ²⁶

In conclusion, induction of labor in FGR fetuses with isolated pathological CPR appears to be safe and is not associated with an increased risk of CD due to non‐reassuring fetal heart rate. However, these fetuses remain at higher risk for NICU admission and might be more vulnerable to other adverse neonatal outcomes, highlighting the importance of individualized management and close intrapartum monitoring in this population.

AUTHOR CONTRIBUTIONS

Bar Noy‐Traub Nofar: Conceptualization, project administration, methodology, investigation, manuscript writing and editing. Farladansky‐Gershnabel Sivan: Conceptualization, project administration, methodology, writing—review and editing. Friedman Eden, Grinblatt Eynit: Data curation, writing—reviewing and editing. Yarza Shaked: Formal analysis, writing—reviewing and editing. Biron‐Shental Tal, Kovo Michal, Dorit Ravid, Arnon Shmuel: Interpretation of data, writing—review and editing. All authors agree with the final version of the manuscript and its submission to the journal.

FUNDING INFORMATION

No external funding was received for this research.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.