Uterine arteriovenous malformation in pregnancy with uterine artery embolization and viable birth: A case report

Abstract

Uterine arteriovenous malformation (UAVM) is a life-threatening condition in which there is an abnormal vascular connection to the uterus. Few cases of UAVM during an active pregnancy have been described in the literature. This report presents a case in which a 37-year-old woman (gravida 5, para 1031) presented for recurrent vaginal bleeding. Digital subtraction angiography established the diagnosis of UAVM. The patient underwent embolization of the cervical branches of the uterine arteries and non-occlusive embolization of both main uterine arteries. She had a vaginal delivery at 28 weeks of gestation and required embolization of the internal iliac arteries postpartum. This case provides a unique illustration of the presentation and sequelae of UAVM in pregnancy, and demonstrates how uterine artery embolization remains the first-line treatment.

Graphical abstract

Highlights

• •Uterine arteriovenous malformation can lead to vaginal bleeding in pregnancy.
• •Digital subtraction angiography is the gold standard for diagnosis.
• •Uterine artery embolization should be considered first-line management.

1. Introduction

Uterine arteriovenous malformation (UAVM) is a condition involving an abnormal connection between arteries and veins. Of the several hundred cases reported in the literature, few have involved pregnancy [1]. They may be either congenital or acquired following surgical procedures. The most common symptom is painless vaginal bleeding without another etiology [2]. Ultrasonography is the modality most often used to diagnose UAVM [3]. For patients who are pregnant or of reproductive age, uterine artery embolization (UAE) is generally considered first-line management as it is both highly effective and fertility-sparing [1]. In the case reported here, a pregnant patient presented for vaginal bleeding in the second trimester and was diagnosed with a UAVM before undergoing UAE and eventually delivering at 28 weeks of gestation.

2. Case Presentation

A 37-year-old woman (gravida 5, para 1031) presented for vaginal bleeding. She had a history of one previous term delivery, followed by three losses in the late first trimester/early second trimester suspicious for cervical insufficiency. Each pregnancy loss required surgical management via dilation and curettage. The case was complicated by homozygosity for factor XIII and plasminogen activator inhibitor PAI-1 deficiency. During the current pregnancy, the patient was briefly admitted at 11 weeks of gestation following an episode of vaginal bleeding of unknown etiology. In light of the pregnancy losses suspicious for cervical insufficiency, at 12 weeks and 1 day, she underwent placement of a McDonald cerclage. She returned at 14 weeks and 6 days of gestation for another episode of vaginal bleeding, for which she was monitored overnight and discharged.

At 15 weeks and 4 days, the patient had another admission for vaginal bleeding. Her hemoglobin had fallen to 8.5 (normal range 12.0–16.0) from 10.0 four days prior, and she received intravenous tranexamic acid. Trans-vaginal ultrasound revealed an area of increased vascularity over the anterior uterine wall concerning for a possible UAVM, though this could not be established definitively. The next day the patient had more heavy vaginal bleeding and was transfused 4 units packed red blood cells. She was taken to the operating room for exam under anesthesia and possible uterine evacuation and hysterectomy if needed for hemorrhage control. At the time of exam under anesthesia, there was no active bleeding. Intra-operative hemoglobin was 5.2 and no obvious etiology could be identified. As this was a desired pregnancy, more aggressive management could not be justified. In the recovery area, the patient underwent a repeat trans-vaginal ultrasound, which demonstrated a 1 × 1 cm vascular lesion in the upper endocervical canal again concerning for a UAVM. Computed tomography (CT) angiography was obtained but was inadequate for definitive diagnosis.

Due to strong suspicion for UAVM, diagnostic angiography was undertaken. Digital subtraction angiography (DSA) was able to visualize branches from the bilateral uterine arteries supplying the endocervical area where the suspected UAVM was located, confirming the diagnosis. Coil embolization of these proximal cervical branches was performed. Additional gelfoam (Ethicon Surgifoam™) pledget embolization (non occlusive) of both uterine arteries was performed to reduce arterial pressure to the uterus to promote thrombosis of the UAVM. Gelfoam is absorbed within 4–6 weeks [4]. The patient was discharged home the next day.

The patient was felt to be at high risk for preterm delivery and received a course of antenatal steroids at 24 weeks of gestation. At 26 weeks and four days she came to the labor and delivery department (L&D) for recurrent vaginal bleeding. Her hemoglobin was 13.2. She was given a rescue course of antenatal corticosteroids and neuro-protective magnesium prophylaxis. Initial trans-vaginal ultrasound was unable to detect a UAVM. Consideration was given to repeat embolization, given that DSA had previously demonstrated a UAVM. Ultimately the patient had no subsequent bleeding and was discharged home in stable condition. At 28 weeks and two days of gestation, she presented to L&D with preterm, pre-labor rupture of membranes (PPROM) without vaginal bleeding. The following day she went into labor and had a vaginal delivery. The child was delivered with APGAR scores of 7 and 9. She was intubated at 14 min of life and transported to the neonatal intensive care unit (NICU), where she was extubated after one day. Initial birth weight was 1140 g (61%). She did well in the NICU and was discharged home in stable condition after 54 days.

Following delivery the patient had significant bleeding post-partum (850 cc) requiring misoprostol, methylergonovine and tranexamic acid with balloon tamponade placement. The uterus was firm and bleeding was suspected to be from the UAVM. Interventional radiology was contacted urgently, and the patient underwent empiric occlusive gelfoam embolization of the bilateral internal iliac arteries. She had no significant bleeding afterwards and was discharged on postpartum day 2. She was doing well at the time of her postpartum visit 6 weeks later. She had no further bleeding and did not require any further interventions, and did not have any further postpartum visits.

At the time of writing, the patient has continued to be seen for her annual visits and has not had any long-standing complications following her delivery. She has not reported any abnormal uterine bleeding and her menses are within normal limits. She has not had any subsequent pregnancies and has not had any further visits for vaginal bleeding.

3. Discussion

UAVM was first described in 1926 [5]. A search of PubMed showed approximately 300 articles discussing UAVM. Only two described cases where the patient was diagnosed with an UAVM during or prior to pregnancy. Rebarber et al. describe a case of UAVM in pregnancy in which the patient underwent uterine artery embolization at 20 weeks of gestation and went on to deliver at 35 weeks [6]. Nasu et al. describe a case of UAVM in pregnancy complicated by postpartum hemorrhage following delivery at term, requiring UAE 3 months postpartum [7].

Though UAVMs may be congenital, most described cases are associated with previous uterine trauma or surgery [2]. In the present case, the patient had three previous dilation and curettage procedures which predisposed her to develop a UAVM. UAVM is often noted as an area of abnormal vascularity on ultrasound. CT or CT angiography may also be considered, but may not be diagnostic. Digital subtraction angiography (DSA), a radiologic technique wherein surrounding radiopaque structures can be eliminated from the image, is considered optimal for diagnosis [3].

This case is also noteworthy in that the patient required a second embolization of her internal iliac arteries to manage her postpartum hemorrhage. Interestingly, no active bleed in the pelvis or the internal iliac artery was visualized at time of angiography, though embolization was carried out empirically. Several studies describe postpartum hemorrhage attributable to UAVM [8], [9], [10]. Ligation of the internal iliac arteries has been described as an uncommon technique to manage hemorrhage related to UAVM [11]. It is possible that bleeding in labor was mitigated due to tamponade of the UAVM from the fetal head.

The described patient was homozygous for factor XIII and plasminogen activator inhibitor (PAI-1) deficiency. Both factor XIII and PAI-1 affect the stability of fibrin clots. This condition is associated with recurrent pregnancy loss and postoperative bleeding [12]. One review recommends having a factor XIII activity level as high as 10% in early pregnancy and 30% in labor [13]. The patient in this report did not have a level checked while in labor. One study describes two pregnant patients with PAI-1 deficiency who had antenatal bleeding and preterm labor, while another report describes a patient with PAI-1 deficiency who had two preterm placental abruptions. [14], [15] This condition may have exacerbated the bleeding from the UAVM and possibly contributed to her miscarriages and postpartum hemorrhage.

UAE is the first-line minimally invasive intervention, though its sequalae on subsequent pregnancies have not been well studied. UAE carries a theoretical risk of fetal demise or fetal growth restriction owing to decreased placental perfusion, and at least one case series describes placental abnormalities in subsequent pregnancies [16]. However, a recent metanalysis by Matsuzaki et al. reported that patients who underwent UAE had increased incidence of postpartum hemorrhage and placenta accreta spectrum disorders, but complications such as fetal growth restriction and preterm birth did not differ significantly from the general population [17]. A review of over 300 cases of UAVM treated with UAE reported a success rate of 88.4% with a complication rate of 1.8% [18]. 20.7% of patients in that same study went on to have another pregnancy, and among those 77 patients there were 18 complications reported; no recurrences of UAVM were noted. A recent metanalysis by Yan et al. reported that UAE decreased subsequent pregnancy rates in the total population when compared with a control group, though this effect did not hold in a sub-analysis comparing UAE with other treatment options [19].

In the present case, the fetus does not appear to have been adversely affected by the embolization; however, as delivery was preterm it is unclear what might have happened if the pregnancy progressed to term. PPROM is an established risk factor for preterm delivery. The Yan et al. metanalysis suggested that UAE may decrease the risk of preterm delivery in subsequent pregnancies, but pooled results were not enough to achieve statistical significance [19]. It is uncertain how much PPROM, UAVM, and UAE each independently contributed to this patient's preterm delivery.

This case suggests that UAE may be considered for treatment and pregnancy prolongation in cases complicated by UAVM.

Contributors

Osamah Khan contributed to conception of the case report, acquiring and interpreting the data, drafting the manuscript, undertaking the literature review and revising the article critically for important intellectual content.

Lauren Hutka contributed to patient care, conception of the case report, drafting the manuscript, undertaking the literature review and revising the article critically for important intellectual content.

Prabhcharan Gill contributed to patient care and revising the article critically for important intellectual content.

Mitryan Kar contributed to patient care, conception of the case report, drafting the manuscript, undertaking the literature review and revising the article critically for important intellectual content.

All authors approved the final submitted manuscript.

Patient consent

Written informed consent was obtained from the patient for publication of the case report.

Provenance and peer review

This article was not commissioned and was peer reviewed.

Funding

This work did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declaration of competing interest

The authors declare that they have no competing interest regarding the publication of this case report.