Independent risk factors for twin pregnancy adverse fetal outcomes before 28 gestational week by first trimester ultrasound screening

Abstract

BACKGROUND

The incidence of multiple pregnancies has increased worldwide recently and women with a twin pregnancy are at higher risk of adverse outcomes compared with women with a singleton pregnancy. It is important to understand the risk factors for adverse fetal outcomes in twin pregnancy in order to guide clinical management.

AIM

To identify the independent risk factors, including maternal personal and family medical histories and first trimester ultrasound screening findings, for adverse fetal outcomes of twin pregnancy before 28 weeks of gestation.

METHODS

The data of 126 twin pregnancies in our hospital, including pregnancy outcomes, first trimester ultrasound screening findings and maternal medical history, were retrospectively collected. Twenty-nine women with adverse outcomes were included in the abnormal group and the remaining 97 women were included in the control group.

RESULTS

Patients in the abnormal group were more likely to be monochorionic diamniotic (13/29 vs 20/97, P= 0.009), with a higher mean pulsatility index (PI, 1.57 ± 0.55 vs 1.28 ± 0.42, P = 0.003; cutoff value: 1.393) or a higher mean resistance index (0.71 ± 0.11 vs 0.65 ± 0.11, P = 0.008; cutoff value: 0.683) or early diastolic notch of bilateral uterine arteries (UtAs, 10/29 vs 15/97, P = 0.024) or with abnormal ultrasound findings (13/29 vs 2/97, P < 0.001), compared with the control group. Monochorionic diamnioticity, higher mean PI of bilateral UtAs and abnormal ultrasound findings during first trimester screening were independent risk factors for adverse fetal outcomes (P < 0.05).

CONCLUSION

First trimester ultrasound screening for twin pregnancy identifies independent risk factors and is useful for the prediction of fetal outcomes.

Key Words: Twin pregnancy; First trimester ultrasound screening; Uterine artery; Pulsatility index; Monochorionic diamniotic twin

Core Tip: Recently, the incidence of twin pregnancy has increased worldwide. Understanding the risk factors for adverse fetal outcomes in twin pregnancy would be helpful for clinical management. We retrospectively analyzed the data of 126 twin pregnancies in the first trimester, including first trimester ultrasound screening findings. Our results showed that monochorionic diamnioticity, higher mean pulsatility index of bilateral uterine arteries and abnormal ultrasound findings during first trimester screening were independent risk factors for adverse fetal outcomes. First trimester ultrasound screening in twin pregnancy is useful for predicting fetal outcomes.

INTRODUCTION

The incidence of multiple pregnancy has recently increased worldwide with a reported twin birth rate of 3.21% in the United States and 3.69% in China in 2019[1,2]. Women with a twin pregnancy are at higher risk of adverse outcomes compared to women with a singleton pregnancy, including miscarriage, congenital abnormalities, intrauterine fetal death, hypertension disorders, preterm delivery, etc.[3-5].

It is important to understand the risk factors for adverse fetal outcomes in twin pregnancies in order to guide clinical management. However, the risk factors for adverse outcomes in twin pregnancies are still unclear. Clinical guidance for twin pregnancy is usually based on studies related to singleton pregnancy. The probable risk factors reported include chorionicity, maternal age, body mass index (BMI), smoking, use of assisted conception, gestational diabetes, and hypertensive disorders of pregnancy[6-8]. Also, crown-rump length (CRL) discordance at 11-13 weeks gestation is considered to be associated with adverse outcomes, but its value has not been confirmed[2,9,10]. Higher pulsatility index (PI) and resistance index (RI) of the uterine artery (UtA) in the first trimester are considered to be related to a higher risk of pregnancy losses in singleton pregnancy[11]. However, the role of UtA parameters for adverse outcomes in twin pregnancy is still unknown.

In this study, we retrospectively analyzed maternal personal and family medical history, first trimester ultrasound screening results and pregnancy outcomes in twin pregnancies before 28 weeks of gestation. Our aim was to explore the independent risk factors for adverse fetal outcomes in twin pregnancies before 28 weeks of gestation.

MATERIALS AND METHODS

Setting and ethics

This was a retrospective study and was approved by the Ethics Committee of our hospital (No. CNFBLLKTY-IEC-2023-023). Written informed consent was obtained from each patient before ultrasound examination.

Patient enrollment

We retrospectively collected the data of women with twin pregnancies at outpatient department of our hospital from January 2019 to December 2022. Inclusion criteria were: (1) Confirmed twin pregnancy before 10 weeks gestation with two heart beats using transvaginal ultrasound; (2) Prenatal ultrasound screening performed at 11-13 weeks gestation including fetal biometry, assessment of fetal anatomy and detection of double uterine arteries; and (3) Complete data on pregnancy outcome, personal history and family history. Exclusion criteria were: (1) Induced abortion for non-medical reasons; and (2) Intrauterine fetal death of one or both twins after 28 weeks of gestation.

The participants with adverse pregnancy outcomes before 28 weeks of gestation including one or both intrauterine fetal death, spontaneous abortion, induced abortion for fetal anomaly and selective fetal reduction for fetal anomaly were assigned to the abnormal group. The participants who had two live births after 28 weeks of gestation were assigned to the control group.

Data collection

Height and weight of the participants before pregnancy were recorded in order to calculate BMI (kg/m²) = weight/height². Systolic and diastolic blood pressure values were measured twice at 11-13 weeks gestation and the smaller values were recorded for calculation of mean arterial pressure = (systolic blood pressure + 2 × diastolic blood pressure)/3. Diastolic blood pressure ≥ 85 mmHg or systolic blood pressure ≥ 130 mmHg were considered high blood pressure. Personal history included age, method of conception (natural or in vitro fertilization), parity (nulliparous or parous if a previous pregnancy resulted in birth at ≥ 24 weeks gestation), smoking status, history of hypertension, diabetes, systemic lupus erythematosus, anti-phospholipid syndrome or pre-eclampsia, history of hysteroscopy, and family history of hypertension.

Ultrasound examination

Prenatal ultrasound screening at 11-13 weeks of gestation followed the practice guideline of the International Society of Ultrasound in Obstetrics and Gynecology (ISUOG)[12-14]. All the examinations were performed by radiologists with a certificate of competence for the 11-13 week scan offered by the Fetal Medicine Foundation using a Voluson^TM E8 diagnostic ultrasound system (GE HealthCare) and a transabdominal C4-8 probe. Twin chorionicity and amnionicity were determined using ultrasound following the ISUOG practice guideline. The parameters of bilateral UtAs including PI, RI and notch were measured and recorded. The mean PI and mean RI were calculated. CRL discordance was defined as (CRL_large - CRL_small)/CRL_large ≥ 10%[15]. Nuchal translucency (NT) ≥ 3.0 mm was considered abnormal. Any fetal structural anomalies were observed and recorded.

Statistical analysis

SPSS version 24.0 software (IBM Corporation, Chicago, IL, United States) was used for statistical analysis. P < 0.05 was considered statistically significant. Numerical variables are presented as mean ± standard error and the independent sample t-test was used for the comparisons if they showed normal distribution (using the Kolmogorov-Smirnov test). The cut-off points for PI and RI were calculated by a receiver operating characteristic (ROC) curve. Enumeration data are presented as a number (percentile) and compared in the two groups using the χ² test. Bivariate logistic regression analysis was performed to determine independent risk factors for abnormal pregnancy outcome.

RESULTS

Basic patient information

One hundred and twenty-eight twin pregnancies were included in this study. One was excluded as one fetus died in utero at 30 weeks of gestation; another was excluded because of induced abortion for non-medical reasons at 15 weeks of gestation.

In total, 126 twin pregnancies were included in the study; 29 in the abnormal group and 97 in the control group. Pregnancy outcomes in the abnormal group are shown in Table 1.

Table 1 Pregnancy outcomes in the abnormal group (n = 29).

Pregnancy outcomes	n
One of the twins died in utero	9
Selective fetal reduction, for one fetus with omphalocele (Figure 1)	1
Both of the twins died in utero	1
Spontaneous abortion	14
Induced abortion for fetal anomaly	4

Abnormal ultrasound findings in the first trimester screening are shown in Table 2 and Figure 1.

Figure 1 Twin pregnancy ultrasound scan at the 13 + 2 gestational week. A: A dichorionic diamniotic twin pregnancy with two placentas and two distinct amniotic cavities; B: Fetus A with normal ultrasound findings; C: Fetus B with omphalocele shown on crown-rump length section as a mass protrusion of the intestine and liver through a navel hernia in the abdominal wall; D: Color Doppler showed that the blood supply of the mass was from Fetus B. CRL: Crown-rump length; NT: Nuchal translucency.

Table 2 Abnormal ultrasound findings during first trimester screening in the two groups.

Abnormal group	Control group
Seven cases of one intrauterine fetal death	One case with one fetus showing NT ≥ 3.0 mm
One case of twin reversed arterial perfusion sequence	One case with both fetuses showing NT ≥ 3.0 mm
One case of one fetus with omphalocele
One case of one fetus with gastroschisis
One case of one fetus with hydrops fetalis
One case of one fetus with NT ≥ 3.0 mm
One case of one intrauterine fetal death and the other with NT ≥ 3.0 mm

NT: Nuchal translucency.

Univariate analysis

None of the women in both groups were smokers or had a history of pre-eclampsia, diabetes or systemic lupus erythematosus. Only 1 patient in the control group had a history of hypertension.

Personal history and first trimester screening findings between the abnormal group and the control group are shown in Table 3. The mean PI of bilateral UtAs and mean RI of bilateral UtAs in the abnormal group were significantly higher than those in the control group. The patients in the abnormal group were more likely to be monochorionic diamniotic, with early diastolic notch in bilateral UtAs or abnormal ultrasound findings in the first trimester screening, compared with patients in the control group.

Table 3 Comparison of first trimester screening results and personal history between the abnormal group and the control group.

Characteristics	Abnormal group (n = 29)	Control group (n = 97)	P value
Maternal age (years)	32.39 ± 3.65	31.62 ± 4.07	0.333
Maternal age (≥ 35 years)	6 (20.69)	28 (28.87)	0.384
Assisted conception	11 (37.93)	46 (47.42)	0.368
Systolic blood pressure	113.17 ± 9.74	111.24 ± 9.77	0.351
Diastolic blood pressure	73.00 ± 7.40	71.82 ± 8.01	0.482
Mean arterial pressure	86.39 ± 7.59	84.96 ± 7.74	0.383
High blood pressure	3 (10.34)	8 (8.25)	0.714
BMI (kg/m2)	22.42 ± 3.40	22.21 ± 3.13	0.753
BMI ≥ 24 kg/m2	6 (20.69)	30 (30.93)	0.284
BMI ≥ 30 kg/m2	1 (3.44)	2 (2.06)	0.547
History of hysteroscopy	8 (27.59)	29 (29.90)	0.811
History of anti-phospholipid syndrome	2 (6.90)	2 (2.06)	0.333
Family history of high blood pressure	6 (20.69)	30 (30.93)	0.284
Nulliparous	16 (55.17)	47 (48.45)	0.525
Chorionicity			0.009
Dichorionic diamniotic	16 (55.17)	77 (79.38)
Monochorionic diamniotic	13 (44.83)	20 (20.62)
Mean PI of bilateral uterine arteries	1.57 ± 0.55	1.28 ± 0.42	0.003
Mean RI of bilateral uterine arteries	0.71 ± 0.11	0.65 ± 0.11	0.008
Early diastolic notch in bilateral uterine arteries	10 (34.48)	15 (15.46)	0.024
Early diastolic notch in at least one uterine artery	15 (51.72)	39 (40.21)	0.271
Abnormal ultrasound findings during first trimester screening	13 (44.83)	2 (2.06)	< 0.001
Crown-rump length discordance ≥ 10%	3 (3/20, 15)	6 (6.19)	0.181

Data are given as mean ± standard error or n (%). BMI: Body mass index; PI: Pulsatility index; RI: Resistance index.

The area under the ROC curve of mean PI of bilateral UtAs was 0.650 with a Youden index of 0.309 and the cutoff value was 1.393. Twenty-one cases (21/29, 72.4%) in the abnormal group had a PI higher than the cutoff value and 60 cases (60/97, 61.9%) in the control group had a PI lower than the cutoff value. The area under the ROC curve of mean RI of bilateral UtAs was 0.661 with a Youden index of 0.319 and the cutoff value was 0.683 (Figure 2). Twenty cases (20/29, 96.0%) in the abnormal group had a RI higher than the cutoff value and 61 cases (61/97, 62.9%) in the control group had a RI lower than the cutoff value.

Figure 2 Receiver operating characteristic curve for mean pulsatility index of bilateral uterine arteries and mean resistance index of bilateral uterine arteries. The area under the receiver operating characteristic (ROC) curve of mean pulsatility index (PI) of bilateral uterine arteries (UtAs; blue line) was 0.650 with a Youden index of 0.309 and the cutoff value as 1.393. The area under the ROC curve of mean resistance index (RI) of bilateral UtAs (pink line) was 0.661 with a Youden index of 0.319 and the cutoff value as 0.683. Mean PI of bilateral UtAs larger than 0.650, or mean RI larger than 1.393 suggested a higher risk of adverse fetal outcomes. ROC: Receiver operating characteristic; PI: Pulsatility index; RI: Resistance index.

Bivariate logistic regression analysis

The results of bivariate logistic regression analysis are shown in Table 4. Monochorionic diamniotic, higher mean PI of bilateral UtAs and abnormal ultrasound findings during first trimester screening were independent risk factors for adverse pregnancy outcomes (P < 0.05).

Table 4 Results of bivariate logistic regression analysis suggestive of adverse pregnancy outcome.

Characteristics	Odd ratios	95% confidence interval	P value
Mean PI of bilateral uterine arteries	3.524	1.015-12.236	0.047
Abnormal ultrasound findings during first trimester screening	56.146	10.086-312.547	< 0.001
Chorionicity	0.276	0.088-0.862	0.027

PI: Pulsatility index.

DISCUSSION

Our study results showed that the first trimester ultrasound screening findings, including monochorionic diamniotic, higher mean PI of bilateral UtAs and abnormal ultrasound findings were independent risk factors in twin pregnancy for adverse pregnancy outcomes before 28 weeks gestation. Personal and family medical history were not independent risk factors for adverse pregnancy outcomes.

Monochorionic twin pregnancies are associated with a high risk of adverse pregnancy outcomes, including spontaneous intrauterine death, congenital anomalies, preterm labor and unique complications due to shared single placenta, unequal placental sharing, velamentous cord insertion and vascular anastomoses in the placenta[16,17]. In our study, 39.4% (13/33) monochorionic twin pregnancies had adverse outcomes before 28 weeks gestation. Our results were similar to those in the study by Glinianaia et al[16], which reported a fetal loss rate of 31.8% in monochorionic twin pregnancies before 28 weeks gestation.

Uterine spiral artery remodeling during normal pregnancy can transform the narrow myometrial arteries into uterine placental vessels with low resistance and a large caliber and impairment in maternal uterine spiral artery remodeling could lead to abnormal placental function[18]. UtA parameters by Doppler ultrasound including PI, RI and early diastolic notch may reflect the changes in uteroplacental circulation during the remodeling process and are usually used as predictors of fetal growth restriction or pre-eclampsia[19,20]. UtA parameters may be independent risk factors for adverse pregnancy outcomes[21,22]. The study by Bao et al[21] showed that the increase in uterine radial artery RI at 8 weeks gestation was associated with spontaneous abortion in women with recurrent pregnancy losses and thrombophilia; the study by Awor et al[22] showed that bilateral end-diastolic notch in UtAs at 16-24 weeks gestation was a predictor of stillbirth at 24 + weeks. Although UtA parameters in twin pregnancies were lower than in singletons, studies showed that increased UtA PI in the first trimester was still useful for the identification of early-onset pre-eclampsia and small-for-gestational age of both twins, but not useful for the identification of late-onset pre-eclampsia and small-for-gestational age of one twin[23,24]. However, the value of UtA parameters for the prediction of adverse pregnancy outcomes before 28 weeks gestation in twin pregnancies is still uncertain. Our study showed that twin pregnancies with adverse pregnancy outcomes before 28 weeks gestation were more likely to have higher UtA PI, higher UtA RI and bilateral UtA early diastolic notch than those with normal pregnancy outcomes. Our results suggest that higher UtA PI was an independent risk factor for adverse pregnancy outcomes in twin pregnancy.

First trimester ultrasound between 11 and 13 weeks gestation for twin pregnancy is very important and does not only establish gestational age and determine the chorionicity and amnionicity accurately, but also evaluates fetal anatomy, risk of aneuploidy and unique complications in twin pregnancy[12-14]. In our study, we detected eight cases of one intrauterine fetal death, one case of twin reversed arterial perfusion sequence, three cases of one fetal structural abnormality as omphalocele, gastroschisis and hydrops fetalis and two cases of one fetal NT thickening. The early detection of fetal abnormalities by first trimester ultrasound was an independent risk factor for adverse pregnancy outcomes in twins and was helpful for further management and prognosis.

CRL discordance at 11-13 weeks gestation is usually defined as (larger CRL - smaller CRL)/larger CRL ≥ 10%[15]. The study by Janssen et al[9] showed that CRL discordance at 11-14 weeks gestation was associated with a high risk of stillbirth or miscarriage. However, our study showed that CRL discordance did not predict adverse pregnancy outcomes. The value of CRL discordance still requires further large sample studies to confirm this association[2].

Some maternal personal factors were considered risk factors for adverse outcomes in twin pregnancy[6-8]. In our study, none of the patients in both groups had a history of pre-eclampsia, diabetes or systemic lupus erythematosus. Only 1 patient in the control group had a history of hypertension. None of the patients were smokers. These factors may have been associated with the better socio-economic and educational status of the patients. In addition, these factors were not evaluated in this study. Our results showed higher maternal age, higher BMI, higher blood pressure, assisted conception and family history of hypertension were not associated with adverse outcomes in twin pregnancy.

Our study showed that monochorionic diamnioticity, higher mean PI of bilateral UtAs and abnormal ultrasound findings during first trimester screening were independent risk factors for adverse pregnancy outcomes. These results highlighted the importance of first trimester ultrasound screening and when above-mentioned results are detected, adverse pregnancy outcomes should be considered and measures should be taken to monitor the fetuses closely for early diagnosis of adverse pregnancy outcomes or preventing the adverse pregnancy outcomes.

Our study had some limitations. First, patient numbers were insufficient, especially in the abnormal group. Second, the participants in our study may deviate from the general population as the proportion of severe maternal abnormalities, such as BMI ≥ 30 kg/m² or high blood pressure, was very low. Further prospective larger studies are needed to verify our results.

CONCLUSION

Monochorionic diamnioticity, higher mean PI of bilateral UtAs and abnormal ultrasound findings during first trimester screening were independent risk factors for adverse pregnancy outcomes. First trimester ultrasound screening is important and the findings were useful for predicting pregnancy outcomes and for guiding clinical management.