Spontaneous resolution of an umbilical cord arteriovenous fistula

Introduction

Variations in the number, configuration, size, or patency of umbilical vessels may be associated with adverse pregnancy outcomes [1, 2]. In this report, we describe the prenatal sonographic characteristics, diagnosis and potential complications of an uncommon umbilical vascular anomaly, namely, umbilical cord arteriovenous (AV) fistula.

Case report

A 36-year-old woman, gravida 1, with dichorionic-diamniotic twins, was referred at 24 weeks of gestation for ultrasound (US) after the detection of a cord abnormality for twin A. There was a normal low risk combined first trimester screening result and a normal fetal morphology scan reported for both twins. At the time of referral, on US, twin A was found to have an abnormally thick appearance of the umbilical cord near its insertion into the abdominal wall, measuring 18×18×17 mm (Figure 1). Color flow and spectral Doppler examinations demonstrated the lesion to be a dilated umbilical vein (UV) leading to a narrower segment with a high velocity pulsatile flow (up to 100 cm/s), a jet-like form that connects to one of the umbilical arteries (Figure 2). Turbulent blood flow was noted within the area of the jet while normal arterial flow and non-pulsatile venous flow were noted on either side of the jet (Figure 3). These US features were interpreted as a small restrictive AV fistula between the umbilical artery (UA) and UV. Fetal echocardiography demonstrated mild cardiomegaly and markedly distended inferior vena cava (IVC). Detailed fetal anatomical survey revealed no other abnormalities. Chromosomal testing revealed a normal karyotype.

Figure 1

Gray-scale ultrasound shows an enlarged abnormal appearance of the umbilical cord at the level of cord insertion into the abdominal wall (white arrow).

Figure 2

Color flow ultrasound shows a dilated umbilical vein with a high velocity, jet-like flow, originating from one of the umbilical arteries (white arrow).

Figure 3

Color flow and spectral Doppler ultrasound shows a normal arterial flow (A), a shunt velocity pattern (arterialized venous flow) within the area of the jet (B), and a non-pulsatile venous flow (C).

A follow-up US assessment at 25 weeks showed oligohydramnios (by measuring the maximal vertical pocket), an increase in the size of the umbilical cord AV fistula measurement (24×19×23 mm), and deep A waves in the ductus venosus (DV). The couple asked that delivery would not be considered until at least 28 weeks of gestation due to the risks of prematurity. Surprisingly, during the period of outpatient follow up, the oligohydramnios resolved, the AV fistula measurement appeared unchanged, the DV showed no A waves, and no growth restriction or fetal compromise was observed in either twin. At 31 weeks, the AV fistula, and the dilatation of the IVC, could no longer be visualized. At 36 weeks, the affected twin developed a persistent deep A wave in the DV combined with abnormal UA PI (0.44). Due to the possibility of hemodynamic effect from the umbilical cord AV fistula, despite the technical inability to visualize it, the patient was admitted for an induction of labor followed by a vaginal delivery. Twin A weighed 2620 g and Twin B weighed 2660 g, both with Apgar scores of 9 at 1 and 5 min. An immediate examination of twin A after delivery showed a thick and wide umbilicus, presumably the fistula area (Figure 4). Both twins’ neonatal course was uneventful. Pathologic examination of twin A’s cord revealed area of congestion and dilatation at its distal segment, measuring 16 mm in diameter. Both placentas and cords showed no other abnormalities.

Figure 4

Photograph of twin A after delivery shows a thick and wide umbilicus, presumably the fistula area.

Discussion

AV fistula of the umbilical cord is an exceedingly rare condition. An OvidSP Medline database search was performed with subject headings “umbilical cord”, and “arteriovenous fistula”. Only one case report of umbilical cord AV fistula was found, which was associated with an umbilical cord aneurysm in a fetus with trisomy 18 [3]. In the only report found, as well as in our case, the umbilical cord lesion measured approximately 20 mm, and Doppler studies showed a jet originating from one of the umbilical arteries entering the dilated UV. The differential diagnosis of an umbilical cord vascular lesion included a cord malformation, umbilical vein varix, cord knot, fistula, or an aneurysm of the cord. Aneurysm, but not a fistula of the cord, was reported to be associated with trisomy 18, a single UA and a hyperechogenic rim [2, 4], none of which were present in our case. Variceal dilatation of the UV is also rare and usually located within the fetal abdomen [5]. In our case, the variceal dilation of the UV was purely extra abdominal and presumably occurred as a consequence of the umbilical cord AV fistula. Nevertheless, due to the possible association of abnormalities of the cord, the patient was counseled regarding the increased risk of aneuploidy.

The mechanism leading to the formation of AV fistula of the umbilical cord has yet to be fully understood. In contrast to umbilical aneurysms, which are usually located near the placental cord insertion [2], the AV fistula in our case is located at the fetal cord insertion. Due to the risk for fetal complications (e.g., fetal distress and intrauterine fetal death) arising from vascular anomalies of the umbilical cord, most authors recommend delivery as early as possible [4]. Early delivery may be essential due to the sudden change in fetal hemodynamics, which can be caused by these lesions, and the absence of reliable tools for the monitoring of fetal well-being [3, 6]. With an AV fistula of the cord, there is a large volume shunt from the UA to the UV, with secondary high venous return and cardiac output status. This can lead to dilatation of the IVC and fetal cardiomegaly, followed by heart failure, hydrops, and even fetal death. The possibility of a locally occurring thrombosis, disseminated intravascular coagulation or even a rupture and hemorrhage of the vascular lesion exists, but this seems to be low.

As implicated with dural AV fistulas [7], it is possible that thrombosis or hemorrhage served as a mechanism in the presumably spontaneous resolution of the umbilical AV fistula in our case. It is difficult to understand when the fetus with an umbilical cord AV fistula is at greatest risk for fetal death. Although our case was associated with uneventful outcomes, we cannot advise delay of delivery, to term, with the diagnosis of an umbilical cord AV fistula. Due to the limitation in options for intrauterine management, without spontaneous resolution of this lesion and in any case of fetal status deterioration, an early delivery may be a better option.

In conclusion, we have presented a rare case of an umbilical AV fistula detected in utero which resolved spontaneously. However as fetal therapeutic interventions for these lesions are not possible, close observation for fetal deterioration and early delivery may be warranted.