Prenatal Diagnosis on Ultrasound, Management and Prognosis of Umbilical Artery Thrombosis: A Retrospective Study of 11 Cases

Umbilical artery thrombosis (UAT) is a rare occurrence during pregnancy and is responsible for most of the fetal demises. It is usually dicult to diagnose and easy to be missed or misdiagnosed. The purpose of this study is to investigate the prenatal ultrasonographic and clinical features of UAT. in 3 to the Doppler of the terminated by 2 birth 1


Background
A normal umbilical cord (UC), as the only connection between the placenta and the fetus, contains two umbilical arteries (UAs) and one umbilical vein (UV) surrounded by Wharton jelly. Umbilical artery thrombosis (UAT), a rare occurrence during pregnancy, which stops the blood ow from the fetus towards the placenta, is associated with adverse perinatal outcomes, such as fetal growth restriction (FGR), abnormal UA ow, neonate thrombotic diseases and even perinatal death [1,2]. The etiology of UAT is not yet fully understood. Early prenatal diagnosis, enhancement of fetal monitoring and termination of pregnancy at proper time are the key points to reduce perinatal mortality. Ultrasound is the rst choice for the prenatal detection of UC abnormalities. However, UAT is easy to be missed or misdiagnosed in the routine ultrasound examination. So far, the studies focusing on UAT are rare, and most of them are case reports. The prenatal ultrasonic diagnosis and clinical management of UAT remain challenging.
In this article, we summarized the ultrasonic image characteristics and the perinatal data in 11 UAT cases, so as to provide some useful information for clinical management. Association. Written informed consent for participation and publication was obtained from each pregnant woman included in this research.
All of them were examined by serial routine ultrasound in our department during the second and third trimester and the diagnosis was con rmed postnatally. All the scans were performed transabdominally using 3-5MHz transducers (GE Voluson E10/E8). The diagnosis of UAT was con rmed independently by two senior sonologists. Fetuses in the cases diagnosed as UAT were assessed by follow-up ultrasound in our department weekly or every other week. Fetal biometry included biparietal diameter, head and abdominal circumference and femur length, and all of them were measured on the basis of guidelines of the International Society of Ultrasound in Obstetrics and Gynecology (ISUOG). Estimated fetal weight was calculated. Fetal gestational age was estimated based on the last menstrual period of pregnant women or ultrasound scan performed during early pregnancy. Examinations of UA were successfully achieved in all cases. Doppler measurements were obtained from both umbilical and middle cerebral arteries (MCA). FGR and fetal distress were evaluated. Torsion of UC, numbers of vessels in UC, thrombi in UA were observed and recorded.
The maternal clinical data were collected, including maternal complications of pregnancy, D-dimer level, mode of delivery, gestational age at delivery. After birth, the length of UC was measured, and the number of complete coils was counted. Placentas with abnormal UC insertions were recorded. Placentas and cords underwent standard histopathological techniques. Stained slides of the placenta and cord microscopic features had been reviewed by two experienced pathologists. The neonatal information was also recorded, including birth weight, newborn sex, Apgar scores and neonatal intensive care unit (NICU) admission.

Results
Maternal clinical data, pathological ndings and neonatal outcomes were available in Table 1. The ultrasonic examination reports were illustrated in Table 2.  In all cases, only one UA was incidentally found at the level of fetal bladder by routine ultrasound examination in the third trimester (range 29 +1 to 39 +4 weeks) ( Figure 1B, 4A, 6B, 6E), however with a history of two UAs previously detected in the second trimester ( Figure 1A, 2A, 6A, 6D). 8 cases (72.72%) (Case 1-8) were successfully and correctly diagnosed as UAT before delivery. Among them, three vessels, one UV and two UAs, could be seen in the cross and longitudinal section of UC by prenatal ultrasound. More importantly, thrombus in one UA and blood ow obstructed by the thrombi were also observed ( Figure 1C, 2B-2E, 3A-3B, 4B-4E, 5A-5B, 5D-5E). Case 7 was misdiagnosed as single umbilical artery (SUA) at 34+ weeks, since only one UA at the level of fetal bladder was found ( Figure 6B), but the thrombus in the other UA was missed ( Figure 6C). The diagnosis was corrected to UAT at 37+ weeks by follow-up ultrasound, when thrombus, hyper-coiling, and FGR were detected at the same time. No hyper-coiling was found by ultrasound in the other cases.
3 cases (27.27%) (Case 9-11) were misdiagnosed as SUA in the third trimester, of which the history of two UAs in previous ultrasound scans were ignored by examiners ( Figure 6D) and thrombi were out of detection ( Figure 6F).
U-fetal neck pressure traces could be seen before or at the same time with the detection of UAT in 9 cases (Case 1, 3-10) (81.8%).
Doppler measurements of the unobstructed UA such as systolic to diastolic ratio (S/D) and pulsatility index (PI) values were completely achieved. When UAT was diagnosed, the S/D values of UA (UA-S/D) were less than the 25th percentile in 4 cases, less than the 10th percentile in 2 cases and less than the 5th percentile in 5 cases (Case 1, 6, 7, 9, 11) [3]. The PI values of UA (UA-PI) were less than the 5th percentile in all cases ( Figure 7A) [4]. Doppler measurements of MCA were measured in only 3 cases (Case 1, 4, 8) [4]. The PI values of MCA were less than the 5th percentile in 2 cases (Case 1, 4) ( Figure 7B), nevertheless, the cerebroplacental ratios (CPR) were higher than the 25th percentile in all of those 3 cases (Case 1, 4, 8) [4].
7 women (Case 1-7) were admitted into hospital due to the nding of UAT by ultrasound (6 women at term, Case 1-5, 7; 1 woman at 36+ weeks of gestational age, Case 6), and 6 of them were given emergency cesarean section to terminate the pregnancy (Case 1-4, 6-7). The woman in Case 5 was ever admitted to hospital at 31+ weeks of gestational age due to the reduction of fetal movements but the electronic fetal heart rate monitoring (EFM) manifested normal during hospitalization. Later at 37+ weeks, she delivered the baby via vagina rapidly after admission, although an emergency cesarean section was suggested by the obstetrician due to the ultrasound report of UAT.
EFM was performed in all cases. 4 of them (Case 1-4) had unsatisfactory EFM (category II or III according to a 3-tiered system for categorization, ACOG guideline, 2009 [5]) before the termination of pregnancy. Case 8 was diagnosed as UAT at 29+ weeks with a normal estimated fetal weight, blood ow by ultrasound. Considering the poor prognosis of premature infant, the obstetrician advised the woman to follow-up with ultrasound scan and EFM more frequently and pay more attention to fetal movement until delivery. Finally, the fetus was delivered at term (38+ weeks) by elective cesarean section with scarred uterus as the indication.
In those 3 cases (Case 9-11) misdiagnosed as SUA by prenatal ultrasound, the pregnant women had no pregnant complications, and their follow-up ultrasound scans and EFM were normal. Finally, they gave birth at full term (Case 9: spontaneous vaginal delivery; Case 10: vaginal forceps delivery; Case 11: elective cesarean section with scarred uterus as the indication).
Meconium-stained amniotic uid was found in 3 cases (Case 2, 9, 10). Gross examination of placenta and UC showed that all the cords were in normal length, and hyper-coiling cords were found in 4 cases (Case 3, 6, 7, 11). The pathological ndings showed the structure of cord vessels (two UAs and one UV) was normal in all cases, and only 1 pathological report (Case 8) described the existence of thrombus in one of the UAs and the infarction of arterial vascular wall ( Figure 5C, 5F). In accordance with the intrapartum ndings, one UA was dark red in 7 cords (Case 1-7) ( Figure 1D, 3C) and earthy-yellow in 1 cord (Case 8) ( Figure 4F), with thrombus found in them. Unfortunately, the intrapartum description about the UA appearance could not be obtained in Case 9-11. Velamentous placenta was reported in Case 8.
All newborns were born alive with normal birth weights (2922±399g), except Case 7 with a low-birth-weight infant (2180g). In Case 4, Apgar score at 1 and 5 minutes was 6 and 10, respectively, and those in the other cases were >7 at 1 and 5 minutes. Discussion adverse perinatal outcomes. Since it has been rarely reported, its exact incidence remains uncertain. . All neonates had been through an abnormal process and hypoperfusion was shown in all placentas [8]. Even though the blood ow of the unobstructed UA, amniotic uid volume and fetal growth parameters in the third trimester are in normal range, UAT is still a life threaten event and the fetus could die suddenly without being noticed [9]. Thus, it is necessary to identify UAT more effectively by prenatal ultrasound, so proper clinical management can follow up as soon as possible.
At present, the etiology and pathogenesis of UAT have not been completely clari ed. Hypercoagulability, blood ow stasis and endothelial injury are supposed to be the main reasons for thrombosis formation. As for UAT, it may be the same. It is well known that pregnant women are in a physiologic hypercoagulable state that can result in thromboembolism more easily. Hypercoagulability may be related to maternal genetic or acquired factors, for example, the unstable level of blood glucose of diabetes disrupting the coagulation function by affecting the expression of endothelial factors [10]. Not only maternal gestational diabetes mellitus, but also smoking and other complications during pregnancy such as hypertensive disease and thrombophilia can induce the formation of clots in UA [1,2,11]. Abnormal UC anatomy, mechanical injury of the UC, and pathological state of fetus or pregnant woman may be the high-risk factors for thrombosis of UC [6]. Cord abnormalities such as anatomical dysplasia, abnormal length (too long or too short), reduction of diameter, unusual Wharton jelly morphology, anomalous placental insertion, twisting, tieing, true knot, excessive helix or compression may induce ow stasis and lead to thrombi of UA [2,7,[11][12][13]. According to studies in recent years, UAT might be closely related to accidental injury of the artery via unpremeditated UV puncture when inserting the needle into the cord [14]. Coincidentally, in our series, 2 out of 11 cases had undergone UV puncture. Consequently, we speculate that UV puncture is also a thrombophilia-related inducer. And there were 9 cases with U-fetal neck pressure traces before or when UAT was diagnosed by ultrasound, which implied cord-around-neck might cause blood ow stasis relating to thrombus formation. Avagliano et al. [11] considered the intrauterine infection could damage the vascular endothelium and caused in ammation and thrombus of umbilical vessels. Additionally, it was believed that protein S de ciency of the fetus which in uenced the state of anticoagulant material in the endothelium of blood vessels might play an important part in the thrombosis of UA [1]. In our study, 1 newborn (Case 7) suffered from NEC, which might be related to the formation of intestinal microthrombi. 2 fetuses (Case 2, 3) were complicated with cardiac malformation (VSD and cTGA, respectively). However, whether intestinal microthrombi and fetal heart diseases are associated with UAT or not needs to be analyzed in further studies with larger sample.
Before UAT occurs, UC usually develops normally. Two-dimensional ultrasound imaging shows that the transverse section of UC includes three circular structures, of which the larger one is UV while the two smaller ones are UAs (Figure 2A), and the longitudinal section of the cord is arranged in a spiral form, which shows all of the three vessels are unobstructed. Color doppler imaging shows two UAs on both sides of the bladder and the vessels in the cord are in "one red two blue" or "one blue two red" structure in which different color means the opposite directions of blood ow. As shown in the ultrasound images of the cases in our study, when thrombosis in one UA occurs, the structure of three vessels can still be seen in the transverse and longitudinal section of UC. One UA is normal and the other one becomes thinner with thrombus lling inside in a hypoechoic or hyperechoic way by two-dimensional ultrasound ( Figure 2B, 2D, 3A, 4B-4C, 4E, 5A, 5D, 6F). Color Doppler imaging shows only one UA is displayed on either side of the bladder, while the blood ow of the other UA is obstructed by thrombus in the bladder level ( Figure 1B, 4A, 6B, 6E) and in the free segment of UC ( Figure 1C, 2C, 2E, 3B, 4D, 5B, 5E, 6C). Some cases were accompanied by excessive helix of UC ( Figure 3A-B, 4B, 4D, 6C). Thrombus observed in the lumen of one UA can be used as a direct sign to diagnose UAT. It is worth noting that ultrasonographic appearance of thrombus may be different depending on the occurrence time and the range of thrombosis.
It is easy to misdiagnose one single UAT as SUA when only one UA is found on either side of the bladder by prenatal ultrasound in the third trimester. Persutte et al. [15] inferred that the main reasons for the occurrence of SUA might be congenital agenesis or dysplasia and gradual disappearance of one UA in the later development, even both probabilities might coexist. The incidence of SUA reported in literature was around 0.31% [16], which was signi cantly higher than that of UAT. Due to the fact that SUA is associated with the congenital dysplasia of UC, it is detectable as early as 12 weeks of gestation age by modern ultrasonography equipment [17], while UAT usually occurs during the third trimester. Thus, it becomes possible for making the prenatal diagnosis of UAT via comparing the patient's initial ultrasound scans, in which two arteries could be clearly seen in the cord, with later ultrasound images in the third trimester, which shows the pathological manifestation of UAT. However, when it comes to the cases without any routine ultrasound examination data before, the differentiation can only rely on the sonographic manifestation in late pregnancy. Compared to one single UAT, two-dimensional ultrasound imaging of SUA shows only two vessels, one artery and one vein, in the cord through the whole pregnancy. Color Doppler ow imaging shows only one UA on either side of the fetal bladder and the umbilical vessels are in "one red and one Blue", which represents different directions of blood ow.
Besides thrombi observed in the artery, which was recognized as the direct sign of UAT, a distinctive ultrasonic sign of UAT found by Klaritsch et al. [18] was that one clogged UA was in parallel with the other unobstructed one and surrounded by UV, like 'an orange grabbed by a hand' in the transverse section of the umbilical cord. Moreover, Tanaka et al. [19] regarded 'Orange grabbed sign' might stand for hyper-coiling of UC, meanwhile, it was suggested as a repeatable and innovative way of UAT diagnosis. In our series, 'Orange grabbed sign' could be observed in Case 7 ( Figure 4E) and Case 9 ( Figure 6F). Since this sign is totally different from the ultrasound manifestation of SUA in the transverse section, it can be used for the differentiation between UAT and SUA.
In our study, thrombi could not be found in 3 cases (27.27%) (Case 9-11) by prenatal ultrasound, and they were misdiagnosed as SUA. 1 case (Case 7) had been misdiagnosed as SUA and the diagnosis was corrected to UAT in later examination. The possible explanation for this might be as follows: (1)  decreased, CPR (MCA-PI/UA-PI) tends to be normal. These changes might be related to fetal brain-sparing effect when one UA was obstructed. When intrauterine hypoxia occurs, fetal brain-sparing effect will be activated, with the cerebral arteries dilating and vascular resistance in ventricular diastole reducing, so that the cerebral perfusion can be maintained [21]. Hershkovitz et al. [22] found the resistance index of MCA in fetuses with SUA were similar to that in fetuses with a normal UC including two arteries. Currently, the vascular resistance indices of UA and MCA are being used to judge if there is intrauterine hypoxia. Whether they can also be used to differentiate UAT from SUA remains to be further con rmed.
SUA is associated with the high incidence of fetal anomaly. It was reported by Murphy-Kaulbeck et al. [23] that the fetuses and neonates with SUA had greater risks of congenital malformations and chromosomal abnormalities, 6.77 and 15.35 times respectively. Trisomy 18 and 21 were the most common chromosomal abnormalities [17]. Therefore, the prenatal identi cation of SUA by ultrasound may be helpful in the antenatal detection of congenital anomalies and aneuploidy. Xu et al. [24] implied that isolated SUA dramatically raised the risk for fetus of small for gestational age. Thus, when SUA fetuses are found by prenatal ultrasound, it needs not only to nd out structural abnormalities of the fetuses, but also to assess their growth and development. Prenatal diagnosis for chromosomal anomalies and genetic counseling should be suggested. For the cases with the coexistence of only one UA displayed in the third trimester and fetal structural abnormalities, SUA should be suspected, which may be helpful in the differentiation between UAT and SUA.
According to the existing data, the standard of ultrasound scan and EFM frequency after the prenatal UAT diagnosis and the timing of pregnancy termination are absent. In Zhu's opinion [25], when the signs of fetal intrauterine distress (reduction of fetal movement, abnormal EFM) appeared, the occurrence of UC thrombosis should be vigilant of, and ultrasound scan should be conducted seriously as well. For term pregnancy, emergency cesarean section should be performed without hesitation in order to avoid serious fetomaternal conditions [25]. For preterm cases, the duration of expectant management remained uncertain, and the timing to terminate pregnancy depended on the supportive capacity of NICU for premature neonates [25]. Dussaux et al. [26] reported a case of UV thrombosis diagnosed at 32 weeks of gestation with a slightly increased resistance index of UA found by Doppler ultrasound, in which close follow-up (daily EFM and weekly ultrasound scans) was suggested instead of emergency cesarean section after the diagnosis. Although the fetus died in utero 3 days later, the authors still insisted on expectant management and intensive fetal monitoring for the UC thrombosis cases before 37 weeks of gestational age on the basis of risk-bene t balance [26]. Luckily, in our study, the incidence of perinatal complications was not as high as reported in the literature, with only one fetus (9.09%) suffering from FGR. All fetuses were born alive, and their prognosis was good. The main reason might be most of the thrombosis were found by ultrasound prenatally, so that fetuses could bene t from the immediate clinical managements by obstetricians. Even if 3 cases (Case 9-11) were misdiagnosed as SUA, whose pregnancies were lucky to last without close surveillance until spontaneous delivery or elective cesarean section at term, and 1 case (Case 8) was given expectant management until full-term, we still recommended that emergency cesarean section should be performed once thrombus was found by ultrasound scan during term pregnancy. Fetal demise may happen all of a sudden without any warning. For preterm cases, even though EFM and ultrasound indices are normal during expectant management, the pregnancy should also be terminated by cesarean section once the gestational age exceeding 32 weeks, after the promotion of fetal lung maturation by corticosteroid therapy.

Conclusion
In summary, UAT is a rare occurrence and cannot be easily diagnosed during antenatal examination. For cases with only one UA was displayed for the rst time in the third trimester by prenatal ultrasound, especially for cases underwent UV puncture, UAT should be highly suspected. Double-checking the previous ultrasound examinations for two vascular structures of UA and the existence of thrombi in one UA was the most useful way to differentiate one single UAT from SUA. For suspected or diagnosed UAT cases, the frequency of follow-up EFM and ultrasound examination should be increased, and the termination of pregnancy should be performed at the proper time.