We read with interest the article by Maasen et al.1 on the complications and failure of uterine artery embolisation for intractable postpartum haemorrhage. We would like to share our experience of a case with a similar complication, thromboembolism of the leg, following balloon occlusion of the uterine arteries.

A 45-year-old primigravida had an elective caesarean section at 37 weeks. She has had subfertility for 6 years and previously had uterine artery embolisation (UAE) for treatment of leiomyomata. Her pregnancy was complicated by a large cervical leiomyoma (7 × 4 × 6 cm), as well as several smaller ones in the lower uterine segment. Ultrasound scan suggested the presence of an anterior placenta accreta. Because of religious beliefs, the patient was refusing transfusion of all blood products.

Prior to the procedure bilateral prophylactic balloon (8.5 mm) placement into the internal iliac arteries across the aortic bifurcation was undertaken under fluoroscopic guidance. After delivery of the fetus abnormal placentation was confirmed and it was decided to leave the placenta undisturbed in situ. The right femoral sheath was left in situ perfused with heparin infusion.

Six hours later the patient developed significant postpartum haemorrhage and it was decided to proceed to a caesarean hysterectomy. Occlusion balloons were further deployed to minimise intraoperative haemorrhage. The origins of both internal iliacs were occluded at the level of the iliac bifurcation. On the right the existing sheath was used. On the left a new sheath was introduced through a different puncture.

After the procedure was completed a check angiography demonstrated a pericatheter thrombus around the sheath in the right external iliac artery necessitating a formal surgical embolectomy.

Postoperatively, the patient developed neuropathic pain in her right leg which was treated with gabapentin. When discharged her haemoglobin was 10 g/l while pre-delivery it was 10.8 g/l. Six weeks postpartum although there was improvement she was still symptomatic with pain. Histology confirmed the presence of placenta accreta at the side of the previously embolised leiomyoma.

A relatively small number of pregnancies following previous UAE have been reported.2 It appears that arterial embolisation can be used successfully in controlling or preventing obstetric haemorrhage in these cases.

In our case, the complication developed on the side where the femoral sheath was left in situ. The prolonged presence of the sheath in a hypercoagulable pregnant patient may have provoked thrombus initiation and propagation. The complication may have been prevented if the sheath and external iliac segments were perfused intraoperatively with pressurised heparinised saline solution. It is logical to suggest prompt removal of the sheaths, unless significant ongoing risk of haemorrhage is apparent.

As embolisation of the uterine arteries is increasingly being used both prophylactically and therapeutically there are concerns raised regarding safety and efficacy.3 Existing evidence is limited thus teams should be encouraged to publish their experiences. Caution should be exercised while interventional radiology techniques develop but in selected cases they can prove indispensable and life saving.


  1. Top of page
  2. References