Article Text

Download PDFPDF

Thyroid haemangiosarcoma in a seven-year-old female Shih Tzu
  1. Luis Pedro Rocha Moreira1,
  2. Emma Scurrell2,
  3. Paul Mahoney3 and
  4. Stephen Baines4
  1. 1Interns, Willows Veterinary Centre and Referral Centre, Solihull, UK
  2. 2Cytopath Ltd, Ledbury, Herefordshire, UK
  3. 3Idexx Teleradiology Services, IDEXX Laboratories Ltd, Wetherby, West Yorkshire, UK
  4. 4Soft Tissue Surgery, Willows Referral Service, West Midlands, UK
  1. Correspondence to Dr Luis Pedro Rocha Moreira; lprmoreira{at}


Canine thyroid tumours are uncommon and the majority of tumours are carcinomas or adenomas, with only very few mixed tumours or metastases from distant sites described to date. A primary thyroid haemangiosarcoma has never been reported in veterinary medicine. In this case report, we describe a dog with a history of a large, non-painful, mobile ventral neck mass in the right paralaryngeal region. CT and ultrasound-guided fine needle aspirates were used for clinical staging. The mass was surgically excised and histopathological examination indicated a haemangiosarcoma. Abdominal ultrasound revealed the presence of splenic nodules and splenectomy indicated the presence of haemangiosarcoma. Chemotherapy with doxorubicin was started, but the dog was euthanased after three rounds of therapy, 97 days after the mass was discovered.

  • canine
  • thyroid gland
  • haemangiosarcoma

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, an indication of whether changes were made, and the use is non-commercial. See:

View Full Text

Statistics from


Canine thyroid tumours are uncommon in veterinary medicine, accounting for 1.1–3.8 per cent of all canine tumours.1 2 No sex predilection has been reported and these tumours tend to affect middle age to older animals, with boxers, beagles, golden retrievers and Siberian huskies being over-represented.3 4 Affected animals are normally asymptomatic, until a large, mobile, generally non-painful mass in the ventral aspect of the neck is noticed.5–7 However, these patients can present with a wide array of clinical signs depending on the size of the tumour, the presence of unilateral or bilateral disease, the presence of local tissue invasion and distant metastasis and whether the serum T4 is elevated.3 Thyroid tumours tend to metastasise to the lungs and regional lymph nodes,7 with the mandibular, retropharyngeal and superficial cervical lymph nodes being the most affected, but other organs can also be involved.3 5

Adenocarcinomas and adenomas are by far the most common tumour types, with an incidence of 99.3 per cent1 and 100 per cent8–10 of all tumours in several studies. In one study of 545 thyroid tumours, only one was identified as haemangiosarcoma (HSA), which was a metastasis from a non-specified distant site.1 Furthermore, eight reports of thyroid carcinosarcomas have been described, but none of them had an HSA component.11 12

In human medicine, a diagnosis of angiosarcoma, an umbrella term used to describe HSA or haemangioendotheliomas, or in other words, a vascular endothelial malignant neoplasm, affecting the thyroid gland is rare, with only around 50 cases described to date.13 Most human cases appear to be distributed to the Alpine regions of Europe and are thought to be associated, with iodine deficiency and long-standing goitre.13 Thyroid angiosarcomas in human beings are usually highly malignant tumours with a poor prognosis and the most common presenting symptom is a fast-growing neck mass.14

Case presentation

An 8.2 kg, 7.5-year-old neutered female Shih Tzu presented with a unilateral, large, firm, rounded mobile mass in the right jugular groove extending caudally from the paralaryngeal region (figure 1). The mass had been noticed one week before referral, with no other clinical signs. This animal was treated in the eight months before referral for suspected spinal pain, due to intermittent, sudden, unexplained episodes of neck pain, initially with prednisolone (Prednidale, Dechra Veterinary Products Limited), followed by meloxicam (Metacam, Boehringer Ingelheim) and underwent dental prophylaxis, with five dental extractions, two months before the referral with no abnormalities noted on physical examination or on a pre-anaesthetic blood biochemical panel.

Figure 1

Presurgical view of the ventral neck mass.


A CT scan using a 16 slice multidetector unit (Brightspeed, General Electric Medical Systems; Milwaukee) of the neck and chest was performed under sedation. After evaluation of the precontrast series, iohexol (Omnipaque 350 mg/ml; GE Healthcare) was administered at a dose of 600 mgI/kg delivered through the cephalic vein using an automatic angiographic injection system at a fixed injection rate of 2 ml/s (MEDRAD Stellant CT Injection System; Bayer). The CT revealed a 6.5 cm (craniocaudal) × 3.6 cm (diameter) oval-shaped soft tissue mass, with a central focal area of mineralisation, lying adjacent to and continuous with the lateral aspect of the right thyroid gland (figures 2–4). The mentioned thyroid gland was reduced in size, with its cranial half appearing to be replaced by mass tissue. The right retropharyngeal lymph node was not visible and the right prescapular lymph node was not enlarged. There was no evidence of pulmonary metastasis.

Figure 2

Dorsal plane reconstruction showing right thyroid mass and both thyroid glands (precontrast sequence). Soft tissue algorithm. There is a large, mixed attenuating mass (yellow arrow) lateral and craniolateral to the right thyroid gland (blue arrow). The left thyroid gland (red arrow) is seen as an elongated hyperattenuating structure adjacent to the air-filled trachea. The right thyroid gland is more caudally located, shows similar attenuation to the left gland, is continuous with the caudomedial border of the mass and is reduced in size.

Figure 3

Transverse plane image through the caudal region of the mass. Soft tissue algorithm. This image demonstrates the intimate location of the mass in contact with the right thyroid gland. Both thyroid glands are visible on this image, seen as the paired hyperattenuating structures adjacent to the trachea. The cross-sectional area of the right thyroid gland is less than that of the left.

Figure 4

Dorsal plane reconstruction showing the right thyroid mass after administration of contrast. Its craniocaudal dimension is measured at 65 mm in length. There is marked, patchy, rim enhancement over the cranial and medial margins of the mass, moderate rim enhancement over the caudal and lateral margins of the mass, and mild heterogenous post contrast enhancement throughout the remainder of its structure (pre 25–30 Hu, post 35–50 Hu).

A further ultrasound (US) examination, while under sedation, confirmed the gross appearance of the mass as being consistent with a thyroid tumour (figure 5) and several US-guided fine needle aspirates (FNAs) were taken. An in-house cytological examination of the aspirates was deemed suspicious of an epithelial neoplasia, as they revealed few bare nuclei cells on a blood rich background, but considered to be non-diagnostic.

Figure 5

Longitudinal plane US image through the central region of the mass. US, ultrasound.

Differential diagnosis

The clinical findings suggested a diagnosis of a thyroid carcinoma and surgical excision of the mass was planned.


A ventral midline cervical approach was made, confirming a large right thyroid mass with some invasion to the sternothyroid and sternohyoid muscles (figure 6). Through blunt and sharp dissection using a vessel sealing device (LigaSure Small Jaw Open Sealer/Divider, Medtronic) surgical device, the mass was excised while preserving the neurovascular structures in the neck and sent for histological analysis (figure 7). No ipsilateral thyroid gland was present after the removal of this mass and the left thyroid gland was examined and deemed unremarkable. Due to an absent CT and surgically detectable lymphadenopathy, no lymph node was excised at the time of surgery.

Figure 6

Mid-line ventral neck incision with minimal dissection. A portion of the mass can be seen right to the mid-line (the head of the animal is towards the left in the picture) with a focal invasion of the adjacent musculature.

Figure 7

Excised thyroid haemangiosarcoma.

Outcome and follow-up

On gross examination, the mass was sectioned in its entirety and no normal tissue was identified. Histopathological examination of the mass indicated a HSA characterised by malignant endothelial cells lining irregularly branching and blood-filled vascular channels (figure 8). The tumour was well-circumscribed and no normal tissue, including that of thyroid gland, was identified in the multiple sections examined. Given the multicentric nature of HSA, further clinical staging was advised and chemotherapy was discussed.

Figure 8

Histopathology image of the mass showing irregular blood-filled vascular channels lined by multiple layers of neoplastic spindeloid cells indicating a haemangiosarcoma. H&E. Bar=100 µm.

Abdominal US was performed nine days after surgery, at the referring veterinary clinic, which revealed a 1.2 cm diameter nodule in the head of the spleen. Two weeks later, a further scan showed that the original splenic mass had increased in size to 1.5 cm in diameter and a second 1.2 cm diameter splenic mass was present in the tail of the spleen. A splenectomy was performed and no additional lesions were identified in the abdominal cavity. Histolopathological examination of the splenic masses showed areas of neoplastic cellular proliferation, consisting of blood-filled clefts and channels with spindle to polygonal cells with dark eosinophilic cytoplasm and oval nuclei with finely stippled chromatin and large central prominent nucleoli, compatible with HSA, with an average mitotic rate of one per high power field, as well as areas of nodular hyperplasia.

Chemotherapy was discussed and a single agent chemotherapy protocol with doxorubicin (1 mg/kg intravenous every three weeks) was initiated two weeks after the splenectomy. Abdominal and cardiac US scans were performed two weeks after the first round of chemotherapy, and no abnormalities were detected. Haematological and biochemical panels taken before the second and third round of chemotherapy were unremarkable. Unfortunately, due to the deterioration of this patient’s quality of life, she only underwent three rounds of chemotherapy, before being euthanased 97 days after diagnosis, due to a general loss of quality of life with lethargy, weakness, anorexia, adipsia and haematochezia being reported by the owner. Further investigation or a postmortem examination were not performed.


Primary thyroid tumours are uncommon in dogs, and the vast majority are epithelial tumours, mainly adenocarcinomas. To the author’s knowledge, this is the first report on an HSA in the canine thyroid gland. HSA of the thyroid gland is described infrequently in human medicine.13

Since HSA is a tumour of vascular endothelium, they may arise from any tissue with a blood supply, but certain organs are more commonly affected than others.15 Primary nodal HSA affecting the cervical lymph nodes, although infrequent, have been described previously in dogs.16 Considering the absence of normal thyroid tissue in any histopathology sections, it can be considered as a potential differential diagnosis. However, a disparity in tumour location, clinical signs and the replacement of the cranial half of the thyroid gland by the mass tissue in the CT scan would be more supportive of the diagnosis of thyroid HSA. Histopathological absence of thyroid tissue could be due to an effacement of the gland by the tumour or to the absence of any minimal thyroidal tissue remaining in the particular planes of sections examined.

When animals present with HSA at multiple sites, it can be difficult to determine which was the primary site and which sites represent metastasis. It is not clear whether the thyroid gland or the spleen was the site of the initial lesion, although the thyroid lesion was larger and detected earlier. In dogs, splenic HSA tends to spread primarily to the lungs, liver, omentum, right atrium and mesentery, whereas non-splenic HSAs typically spread initially to the lungs and spleen, followed by liver, right atrium, kidneys, brain and then other less common sites.17 The thyroid gland has never been listed as a site of metastasis in dogs. Hence, metastasis of a thyroid HSA to the spleen would appear to be more likely than metastasis of a splenic HSA to the thyroid gland. In addition, the thyroid mass was considerably larger (32 cm3; 6cm×3.4 cm×3 cm), with associated areas of haemorrhage and a large area of central necrosis, than the splenic nodules. As such, we are inclined to believe the thyroid tumour to be the primary lesion. However, due to the multicentric nature of HSA,17 the absence of abdominal imaging during the initial staging and the main site of metastasis in human thyroid angiosarcomas being the lungs,14 it is impossible to be certain.

The CT scan and US examination supported the diagnosis of thyroid carcinoma given the location, CT18 and US19 appearance of a thyroid carcinoma. However, neither a CT nor an US scan are 100 per cent specific for this or any histological type of tumour.20 While HSA can have a few precontrast and postcontrast CT morphological characteristics similar to a thyroid carcinoma, such as attenuation or texture, cavitation and intratumoural vascularisation, thyroid carcinomas can also have a varying appearance on CT examination.18 21 Dystrophic calcification, as detected in this study, is commonly seen in epithelial tumours such as prostatic adenocarcinomas, thyroid carcinomas, primary lung tumours and hepatocellular carcinomas but is not a common finding with HSA.21

In both human and veterinary medicine, while FNAs of suspected thyroid tumours, even those performed with US guidance, may be supportive of thyroid carcinoma, they can be non-diagnostic in more than half of the cases,8 14 namely due to haemodilution of the sample, and may not allow differentiation between adenoma and carcinoma.3 6 A tissue biopsy is normally the only way to obtain such differentiation.3 However, due to a potential risk of tumour seeding during the biopsy, haemorrhage due to high vascularity of thyroid carcinomas and local coagulopathies, if the thyroid mass is largely mobile, not invading the surrounding tissues and in the absence of visible metastasis, surgical excision is preferred to biopsy.3 6 22

A mean survival time cannot be extrapolated from a singular case only, but the witnessed thyroid haemangiosarcoma in this patient, carried a poor prognosis in spite of surgery and the use of other adjuvant treatments, with a survival time of 3 months which is similar to the poor prognosis seen in some human reports.23

Learning points

  • Thyroid malignant neoplasias are rare in dogs and, normally, consist of either carcinomas or adenocarcinomas. Nonetheless, other neoplasias such as the one described in this paper must always be considered as a differential diagnosis.

  • While a discussion regarding the importance of abdominal imaging should always take place with clients that present a patient for thyroid tumour staging, the multicentric and metastatic nature of haemangiosarcomas (HSAs),17 would deem it necessary, if the thyroid neoplasia is determined to be an HSA.

  • Finally, it seems that the outcome for a thyroid HSA in dogs is similarly poor, compared with that reported for thyroid carcinosarcomas, with short survival times.11 12


View Abstract


  • Contributors LM: main author responsible for writing the article. SB: main clinician involved with the case, provided the clinical data and helped in writing the case report. ES: pathology specialist responsible for the histopathology report of the excised mass and also reviewed the case report. PM: imaging specialist responsible for the CT and US report and also reviewed the case report.

  • 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.

  • Patient consent for publication Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data availability statement All data relevant to the study are included in the article.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.