Management of venous thromboembolism secondary to ovarian hyperstimulation syndrome: A case report documenting the first use of a superior vena caval filter for upper limb venous thromboembolism in pregnancy, and the difficulties and complications relating to anticoagulation in antenatal and peri-partum periods

Abstract

The management of venous thromboembolism and subsequent pulmonary embolism in pregnancy remains hugely challenging. In this case, we report the first use of a superior vena caval filter in pregnancy as an adjunct to pharmacological anticoagulation. This is the first reported use of a superior vena caval filter in pregnancy. We discuss the complexities of managing thromboembolism in pregnancy and the peri-partum period.

Introduction

Ovarian hyperstimulation syndrome (OHSS) is a complication of assisted reproduction characterised by gross cystic enlargement of the ovaries and fluid shift into the third space.¹ The reported incidence of OHSS varies between 0.6% and 14%.² The most frequently observed complication of OHSS is thromboembolism, the majority of which are venous. Venous thromboembolism (VTE) remains a leading cause of maternal death in the UK.³ The incidence of VTE in women conceiving via IVF is 0.2%; this increases to 1.7% in association with OHSS representing a 100-fold increase compared to background risk.⁴ These thrombotic events tend to occur in uncommon sites such as the upper limb, cerebral and cardiac vessels.²

The use of anti-coagulation in pregnancy represents a huge challenge, particularly peri-partum. Recent RCOG guidance suggests therapeutic low-molecular-weight heparin (LMWH) dosage should be based on pre-pregnancy weight rather than measurement of anti-Xa levels.⁵ However, it has been found that weight-based regimens of LMWH are associated with a higher incidence of thromboembolism in high-risk women with mechanical heart valves.⁶ Consequently, many clinicians advocate use of anti-Xa monitoring with dose adjustment given the dynamic physiological changes in pregnancy.⁷

Despite adequate anti-coagulation, pulmonary embolus may still occur. In this circumstance or where LMWH is contraindicated, inferior vena caval filters have been used successfully in pregnancy as an adjunct or alternative to anticoagulation.^8,9

Case report

A previously fit and well 33-year-old nulliparous lady underwent in vitro fertilisation treatment for primary subfertility of unknown aetiology. Pre-treatment investigations including pelvic ultrasound scan and hormone profile were normal. Antagonist protocol ovarian stimulation was undertaken and subsequently 21 eggs were collected. Prior to egg collection the oestradiol level was approximately 18,000 pmol/L. Given the high risk of subsequent OHSS, the patient was given cabergoline 500 µg daily for five days. Progesterone luteal phase support was commenced and single embryo (blastocyst) transfer was performed. One week following embryo transfer the patient presented with abdominal pain and bloating. A diagnosis of OHSS was made following initial investigation including pelvic ultrasound. The largest ovarian diameter was 11 cm. Admission was not initially required. Prophylactic enoxaparin was commenced at 40 mg daily. There were no pre-existing risk factors for thrombosis and body mass index was within normal range. The cycle was successful and an intrauterine pregnancy was seen on routine ultrasound scan at six weeks of gestation.

At seven weeks of gestation, the patient presented with severe neck pain and swelling. A Doppler ultrasound identified an extensive, occlusive thrombus in the right internal jugular vein. Treatment was changed from prophylactic to therapeutic enoxaparin based on a standard weight-based dose of 60 mg bd. Ultrasound scan demonstrated an ongoing pregnancy.

The patient returned five days later with shortness of breath with associated pleuritic chest pain. On initial assessment, she had an increased respiratory rate of 18 breaths per minute and oxygen saturations of 96% on room air. These parameters returned to normal with 2 L of supplementary oxygen. Electrocardiogram demonstrated normal sinus rhythm. An urgent computerised tomography pulmonary angiogram identified bilateral pulmonary emboli. Given embolism had occurred despite presumed therapeutic dose LMWH, vascular surgical input was requested. Subsequently, a superior vena caval (SVC) filter was inserted endovascularly via the femoral vein using fluoroscopic guidance. This was placed under local anaesthesia without immediate complication. At the time of the filter insertion, the patient was anti-coagulated with adjusted dose intravenous (IV) unfractionated heparin infusion which was discontinued 24 h post-procedure. The dose of enoxaparin was then titrated according to peak and trough anti-factor Xa levels, the stable dose during the remainder of the pregnancy was 55 mg tds. Screening for hereditary and acquired thrombophilia conditions was negative.

The patient remained stable with satisfactory fetal growth in the second and third trimesters. Labour was induced at 38 weeks of gestation. The last dose of LMWH was given approximately 10 h prior to delivery, at the time of vaginal prostaglandin gel insertion. A remifentanil PCA was given for pain relief during the first stage of labour. Following a second stage of less than an hour, a female infant weighing 3.6 kg was delivered vaginally. A second degree tear was identified and sutured. Total blood loss at delivery was estimated to be 500 mL. Therapeutic enoxaparin treatment was recommenced at 6-h post-partum at 60 mg bd.

At three days’ post-natal, the patent reported severe perineal pain and on examination was found to have a tense perineal haematoma. Immediate surgical exploration was performed and identified a large ischiorectal haematoma. Evacuation, haemostatic suturing and packing were undertaken. Estimated total blood loss was 3500 mL. Eight units of red blood cells, two units of fresh frozen plasma and one pool of platelets were given to correct coagulopathy. Haemostasis was maintained and LMWH was re-commenced 12 h post-operatively. The patient received post-operative IV antibiotics for 48 h. She was clinically well and fit for discharge home on day 7 post-natal. Enoxaparin was continued at a dose of 60 mg bd at the time of initial discharge.

Readmission was required the following day (day 8 post-natal) due to a complaint of abdominal pain which was associated raised inflammatory markers. Ultrasound and magnetic resonance imaging scans identified a large right-sided pelvic side wall haematoma measuring 10 cm × 5 cm × 15 cm. Enoxaparin was discontinued for 24 h but restarted when a repeat USS confirmed no extension of the haematoma. Conservative management was employed and clinical symptoms resolved within 72 h. Broad spectrum IV antibiotics were given for five days.

Subsequent follow-up imaging has shown complete resolution of the haematoma. The SVC filter was removed without complication at eight weeks post-partum at which time enoxaparin treatment was discontinued.

Discussion

This case is interesting for a number of reasons, the most important of which is the successful insertion of an SVC filter during pregnancy. This is the first reported use in pregnancy. It sets a precedent for the management of pregnant women with jugular and/or upper limb thrombosis, complicated by pulmonary embolism. The risk of embolism is lower in cases of upper extremity thrombosis (5.6%) compared to lower extremity (25.1%),¹⁰ so this will not be a common occurrence. The risk of embolism from either site is likely to be reduced further if the dose of LMWH is titrated to peak and trough anti-factor Xa targets of 1–1.2 (peak) and >0.6 (trough). This approach contrasts to the current, recent RCOG advice, which suggests that a weight-based dose should be used. However, the evidence for the RCOG position is limited and in our experience, the weight-based dosage usually results in sub-therapeutic anticoagulation. Indeed, others have reported a similar experience.¹¹

Achieving and maintaining adequate anticoagulation in the context of the dynamic physiological changes of pregnancy is complex and a weight-based approach to the dose of LMWH may be inadequate, particularly in context of an ongoing thrombotic inducing condition such as OHSS. The pharmacokinetics of LMWH are significantly altered by pregnancy with increased renal clearance.¹² In this particular case, the weight calculated dose of LMWH was much lower than the final dose achieved after titration to peak and trough anti-Xa levels. To achieve these targets, LMWH needed to be administered three times per day. Although the RCOG has not advocated this type of monitoring, other clinical groups have suggested that it provides more effective anticoagulation. The American College of Chest Physicians recommends target trough anti-Xa levels of 0.6–0.7 IU/mL and peak levels of around 1.0–1.2 IU/mL or less than 1.5 IU/mL.⁷

Anticoagulation in the peri-partum period is challenging and requires tight control: balancing the need for adequate anticoagulation versus the risk of bleeding at the time of delivery. Rather than an increase in peri-partum haemorrhage, obstetric patients on anti-coagulation have a higher risk of wound haematoma relating to caesarean section or perineal injury, as we found in this case,¹³ which may have life-threatening consequences. Despite this, one must not forget that the risk of thromboembolism increases post-partum and hence the need to restart therapeutic LMWH.¹⁴ The use of unfractionated heparin infusion around delivery may provide more flexibility,⁷ as this can be easily reversed with protamine sulphate in a titrated manner to avoid difficulty re-introducing anticoagulation. In our case, we opted to stop the LMWH and to use non-regional approach to pain relief with a remifentanyl PCA, as we had an SVC filter in place that would prevent further embolic events and our patient had been stable through last six months of pregnancy. However, the major haemorrhagic complications which occurred prompt consideration of the time frame for stopping and re-introducing therapeutic dose LMWH. Given the in situ filter, one may consider delaying the reintroduction of therapeutic dose LMWH for 24–48 h and instead using a prophylactic dose. Similarly, prophylactic dose could be given prior to anticipated delivery which would enable the use of regional neuroaxial anaesthesia.

Overall, the peri-partum management of a woman on anticoagulation is complex and needs to be individualised and meticulously planned to reduce the risk of thrombotic and haemorrhagic complications.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethical approval

Written consent from the patient was obtained for publication.

Guarantor

NM-J

Contributionship

NMJ, MM and MJ all contributed to writing the paper.