Management of hepatic haemangioma in pregnancy
- Anthony Spartan Casabianca 1,
- Ana Ivette Hernandez Caballero 2,
- Loralei L Thornburg 3 and
- Darren Carpizo 4
- 1 General Surgery, University of Rochester, Rochester, New York, USA
- 2 Pathology and Laboratory Medicine, University of Rochester, Rochester, New York, USA
- 3 Obstetrics and Gynecology, University of Rochester, Rochester, New York, USA
- 4 Surgical Oncology, University of Rochester Wilmot Cancer Institute, Rochester, New York, USA
- Correspondence to Dr Darren Carpizo; Darren_Carpizo@urmc.rochester.edu
Abstract
Hepatic cavernous haemangioma is a benign tumour of vascular origin found within the liver. Often incidentally diagnosed, the management of these vascular masses is frequently determined by the size of the mass and symptoms associated with its compression of adjacent structures. Tumours >10 cm are known as giant haemangiomas and are associated with increased risks of compression symptoms, coagulopathies and haemorrhage. Known to express hormone receptors for oestrogen, intervention for these masses remains controversial in the setting of pregnancy where concerns for tumour growth and life-threatening complications are increased. Here we present the case of a woman in her 30s recently diagnosed with a giant haemangioma who is found to be pregnant, their management and a review of the literature.
Background
Hepatic cavernous haemangioma is a benign tumour of vascular origin found within the liver. These tumours are most commonly asymptomatic and incidentally diagnosed. They originate from alterations in vasculogenesis during development and are common (accounting for 73% of all benign liver tumours).1 It has been reported that at autopsy liver haemangioma is seen in 0.4%–20% of patients, making it the most common tumour of liver origin, and second most common liver tumour overall, after metastases.2 Histologically, these tumours consist of wide endothelium lined blood-filled spaces with multiple septations that can be fed from multiple arteries.3
Often diagnosed incidentally, these lesions are frequently seen on ultrasound or CT imaging obtained for non-specific abdominal symptoms or in the evaluation of unrelated pathologies.1 These tumours range in size and are most commonly <4 cm.4 At this size these tumours are commonly managed with observation and repeat imaging.5 Some authors categorise tumours >5 cm as giant haemangiomas with reported increased risks of secondary effects which can include pain, early satiety, nausea, vomiting and risk of spontaneous or traumatic rupture with this distinction indicating the need for treatment.2 However, one review of the risk of spontaneous rupture failed to find an associated increase in risk for tumours <11 cm.6 This has led some to use a definition of ≥10 cm for giant haemangiomas that require treatment as this is when it most likely that tumours could cause secondary symptoms or be associated with spontaneous life-threatening haemorrhage.4 Other surgical indications include worsening secondary symptoms related to the tumours size and compression of adjacent organs, rapid growth, the development of consumptive coagulopathies such as Kasabach-Merritt syndrome and life events which may increase the risk of spontaneous or traumatic rupture.5
The management of haemangiomas that are discovered during pregnancy is unclear. This is due to a relatively poor understanding of the effects of pregnancy on the growth of these tumours and their risk of rupture. While the majority of haemangiomas remain stable during pregnancy, case reports have described enlargement during pregnancy, between pregnancies and during oestrogen and progesterone therapy suggesting a role for hormones such as oestrogen on their growth.7–9 It is common for these tumours to express the ER receptor, which may explain the greater preponderance in female patients with reported frequencies as high as 5:1 compared with male patients.6 Further complicating clinical decision making is that the risk of rupture must be weighed against the risk of surgery on the fetus as well as the optimal timing and route of delivery of the child.
Here we present a case of a woman diagnosed with a giant cavernous haemangioma who became pregnant very soon after this diagnosis. We discuss the management of this patient and outcomes as well as the review of the literature surrounding liver haemangioma in pregnancy.
Case presentation
A woman in her 30s who first presented to her primary care provider with complaints of right upper quadrant pain associated with a visible swelling in the right side of her abdomen noticed while laying down. At rest, she felt a fullness in the abdomen and the pain was described as a dull ache.
Investigations
An ultrasound at that time revealed two right liver masses measuring 12 × 9 × 9 cm and 7 × 5 × 5 cm. This finding prompted a CT scan which identified the lesion as a single mass in the right lobe of the liver measuring 16 × 14 × 23 cm which displaced her right kidney inferiorly. An MRI at that time described a T2 hyperintense lesion with multiple septations and peripheral progressive nodular enhancement consistent with a 23 × 13.5 × 16 cm haemangioma (figure 1). After this finding she was scheduled for surgical consultation. However, just prior to this visit the patient reported a positive pregnancy test and serum human chorionic gonadotropin and fetal ultrasound confirmed her to be approximately 7 weeks.
(A) Preoperative MRI demonstrating right sided giant hepatic haemangioma measuring 23 × 13.5 × 16 cm enlarging the right hepatic lobe and compressing adjacent structures. (B) Transcatheter arterial embolisation identifying two feeding arteries branching from the right hepatic artery (arrows). Embolisation of feeding arteries (C).
Treatment
The patient was referred to the maternal-fetal medicine service for consultation, and the patient was presented at a multidisciplinary management group meeting where the consensus was to recommend surgical resection. Timing was discussed with the team, and the patient elected to proceed in the first trimester due to symptoms, with a plan to proceed prior to the uterus becoming enlarged and out of the pelvis, as well as the need to coordinate care. Given the super large size, early surgery was done to avoid having enlarging uterus interfere with the surgery.
On the day before surgery the patient was admitted to the hospital and underwent transarterial embolisation of the haemangioma which identified two feeding arteries from the anterior and posterior right hepatic artery (figure 1B). One day following embolisation she underwent an open right hepatectomy. The right hepatic inflow and outflow vascular structures were dissected and divided extrahepatically. The parenchymal transection was performed with a clamp-crush technique using a vessel sealing device for smaller intrahepatic vessels and a vascular stapler to divide the main right pedicle. Total operative time and blood loss was 6 hours and 600 mL. A 1518 g right hepatectomy segment was resected containing the haemangioma measuring 24 × 18 × 7 cm (figure 2). In the recovery room the patient underwent a fetal ultrasound which detected a normal fetal heart rate. Her postoperative course was uneventful, and she was discharged in good condition on postoperative day 6. On postoperative visit 1 week later, the patient reported excellent pain control and the fetus was found to be in good health.
(A) Gross specimen image depicting a 1518 g haemangioma measuring 24 × 18 × 7 cm. Note the necrosis due to preoperative transarterial embolisation. (B) H&E section at 200× magnification depicting characteristic dilated vascular spaces lined with bland, flat endothelial cells with intervening paucicellular fibrous bands.
Outcome and follow-up
She subsequently underwent routine pregnancy care and went into spontaneous labour at 41 2/7 weeks gestation, delivering a female infant by spontaneous vaginal delivery without anaesthesia, weighing 3390 g, APGARS 9, 9. There was a retained placenta and the patient was unable to tolerate manual removal, necessitating curettage. Placental pathology was unremarkable. The dyad was discharged home on postoperative/postpartum day 2 in excellent condition. She received an intrauterine device for postpartum contraception. Follow-up MRI 11 months after surgery demonstrated interval removal of the large haemangioma with the presence of two small, T2 hyperintense lesions at the surgical margin consistent with small benign residual haemangioma. The patient reports being in good health at follow-up. The infant continues to develop normally without complication.
Discussion
The optimal management of giant cavernous hepatic haemangiomas continues to be debated. These masses present significant risk of surgical complications owing to their size, association with neighbouring structures, as well as from significant blood loss. Non-operative interventions such as radiofrequency ablation and interventional radiology embolisation have been reported, but complications such as liver injury or abscess formation have precluded their adaptation as first-line therapy.1 Surgery remains the primary definitive therapy for symptomatic patients, or those where bleeding is suspected. Symptoms are primarily related to mass effect of the tumour on neighbouring organs. These can include abdominal discomfort, distension and early satiety.2 More serious complications warranting operative intervention include spontaneous rupture and Kasabach-Merritt syndrome.10–12 The management of asymptomatic giant haemangiomas is controversial with some recommending resection in all patients with an adequate performance status to a selective approach based on symptoms, size and growth rate.
When surgical management of hepatic haemangiomas is undertaken, it can include tumour enucleation in which the haemangioma is dissected off the surrounding hepatic stroma or partial hepatectomy.4 One retrospective study of patients undergoing surgery for giant liver haemangioma found no significant differences existed between blood loss or overall complications between these two surgical modalities.4 Transarterial embolisation (TAE) is a less invasive modality used in the treatment of giant haemangiomas. Short-term outcomes have been reported demonstrating good results in shrinking tumours; however, long-term follow-up revealed this therapy was incapable of preventing future tumour growth in lesions >10 cm.1 Additionally, TAE can result in the formation of large intrahepatic abscesses following degeneration of the embolised mass.1 We typically only consider TAE as the definitive treatment for giant haemangiomas in patients who are not surgical candidates. TAE can also be used as an adjunct to resection when used in the preoperative setting. It is our practice to use preoperative TAE in patients with haemangiomas >20 cm which we have termed ‘super haemangiomas’. The purpose of the TAE is to shrink the tumour substantially which will facilitate the resection by improving mobilisation of the liver, decreasing blood loss and operative time.
It is common for haemangiomas to express the ER receptor, which may explain the greater preponderance in female patients with reported frequencies as high as 5:1 compared with male patients.6 This finding has also led to great concern over the risk of growth/rupture in response to hormonal stimulation. In fact a multitude of conflicting evidence exists, mainly in the form of case reports regarding this.3 8 9 In one study of 94 female patients with hepatic haemangiomas, an increase in size was detected in 22.7% of patients receiving hormonal contraceptives compared with 9.7% in their control population.8 No mechanism is described to account for these differences though it is believed to be related to increased ER receptor expression within the tumour epithelium.13 In a separate case review of pregnant patients, tumour growth was observed in six of eight pregnant patients suggesting a hormonal influence.7 Conversely, in a retrospective review of 149 women, no association was found between the use of oral contraceptives and prevalence of liver haemangioma.3 This study found that age alone was a predictor of haemangioma, and that other factors associated with increased oestrogen exposure such as early menarche, number of pregnancies and late menopause were not indicators of haemangioma prevalence.3 Reports of complications associated with hepatic haemangiomas during pregnancy remain limited to case reports. Few published examples of spontaneous rupture during pregnancy exist which commonly resulted in emergency delivery and surgical resection.7 More frequently resection is offered during pregnancy due to worsening symptoms related to the size of the mass or the development of consumptive coagulopathy.7
The timing of surgical intervention during pregnancy presents specific risks to both the patient and the developing fetus. Typical surgical practice favours surgery during the second trimester to avoid injury to the fetus. However, the effects of increased oestrogen on haemangioma growth and rupture are unclear and present a risk to the patient for spontaneous life-threatening haemorrhage. For patients with haemangiomas <10 cm, we would typically choose to observe these during pregnancy and recommend following them with ultrasound every 3 months for growth. For giant haemangiomas (10 cm) we recommend these be followed more frequently by ultrasound and if there is growth that surgical resection be strongly considered, ideally in the early second trimester. For super haemangiomas >20 cm, first-trimester intervention with appropriate patient counselling may also be considered, especially in symptomatic patients.
Learning points
-
Hepatic cavernous haemangioma is a benign, often incidentally discovered liver tumour of vascular origin which may require resection in cases where compressive symptoms or coagulopathies are present.
-
Hepatic haemangiomas <10 cm may be safely monitored with regular ultrasounds provided patients are asymptomatic.
-
Giant hepatic cavernous haemangiomas should be considered for surgical resection, ideally in the early second trimester; however, masses >20 cm warrant first-trimester intervention with appropriate patient counselling.
Ethics statements
Patient consent for publication
Footnotes
-
Contributors ASC: conducted the background research and initial drafting of the manuscript with significant input and editorial guidance from DC and LLT. LLT and DC: consulted on the medical and surgical management of the patient, conducted the surgical procedures mentioned in the study and conceptualised the report. AIH provided pathology analysis and figure images. All four authors had full access to the data and reports at all times and have all seen and agreed upon the final draft for submission.
-
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
-
Competing interests None declared.
-
Provenance and peer review Not commissioned; externally peer reviewed.
- © BMJ Publishing Group Limited 2022. No commercial re-use. See rights and permissions. Published by BMJ.
References
Use of this content is subject to our disclaimer