Oesophageal angioleiomyosarcoma in a cat

Abstract

A female spayed domestic longhair cat aged 3 years and 9 months was referred for investigation of regurgitation and weight loss of 2 months’ duration. Thoracic radiographs revealed a soft tissue mass within the cranial mediastinum causing focal oesophageal dilation. Computed tomography confirmed a contrast-enhancing mass located cranial to the heart base, possibly originating from the oesophagus. Exploratory thoracotomy revealed an intramural soft tissue mass within the ventral oesophageal wall. Surgical excision of the mass and reconstruction of the oesophagus around an oesophageal tube was performed successfully. Histopathological examination of the mass was compatible with a spindle cell tumour with a prominent vasoformative component. Immunohistochemistry was positive for α-smooth muscle actin and von Willebrand factor protein, and negative for CD117/c-kit protein. Both histopathological and immunohistochemical findings confirmed the diagnosis of an angioleiomyosarcoma. The cat was clinically well 6 months postoperatively. To our knowledge, this is the first report of an oesophageal angioleiomyosarcoma in a cat.

Case Report

A female spayed domestic longhair cat aged 3 years and 9 months was referred to the Queen Mother Hospital for Animals for investigation into regurgitation and weight loss of 2 months’ duration. During this time, the primary complaint was regurgitation of food, of varying severity, a few minutes after or during meals. Within the week prior to referral the cat was also immediately regurgitating water. The cat had a normal-to-increased appetite. The referring veterinarian had noted a 0.9 kg weight loss over the course of these 2 months. Preceding referral,a barium swallow study was performed, which revealed a focal oesophageal dilation cranial to the cardiac silhouette.

On presentation, the cat was tachycardic (heart rate 210 beats per minute), tachypnoeic (48 breaths per minute) and pyrexic (rectal temperature 40°C). A grade II/VI sternal systolic cardiac murmur and harsh lung sounds were appreciated on thoracic auscultation. The cat weighed 3.2 kg, with a body condition score of 3/9.

Haematology revealed a moderate leukopenia (3.89 × 10⁹/l; reference interval [RI] 5.5–19.5 × 10⁹/l) with mild lymphopenia (0.97 × 10⁹/l; RI: 1.5–7.0 × 10⁹/l). The band count was 0.23 × 10⁹/l (RI: 0–0.3 ×10⁹/l) and the polymorphonuclear leukocyte count was 2878.6 cells/μl. The neutrophils had mild toxic change, and there was evidence of a left shift. Serum biochemical analysis revealed a mild hypoalbuminaemia (25.4 g/l; RI: 28–42g/l), mild hyponatraemia (144 mmol/l; RI: 148–160 mmol/l), mild hypokalaemia (3.6 mmol/l; RI: 3.8–5.5mmol/l) and mild hypercholesterolaemia (2.1 mmol/l; RI: 2.2–6.7 mmol/l). Urea concentration was markedly decreased (2.5 mmol/l; RI: 6.1–12.0 mmol/l). Taking into account the history and presentation, the haematological changes were consistent with infection, which increased our suspicion of aspiration pneumonia. The cat was maintained on intravenous fluid therapy using compound sodium lactate supplemented with potassium chloride prior to diagnostics.

Initial diagnostic imaging included radiographs, imaging of the thorax by computed tomography (CT) and oesophagoscopy in order to further assess the origin and extent of the lesion. The cat was sedated with medetomidine (Seda-start, 0.005 mg/kg, IV; Animalcare) and butorphanol (Alvegiesic, 0.2 mg/kg, IV; Dechra). Orthogonal thoracic radiographs confirmed the presence of a soft tissue mass within the cranial mediastinum and associated focal oesophageal dilation (Figures 1a,b). There was also evidence of an alveolar pattern (consolidation) in the right cranial lung lobe (Figure 1b). In order to perform the CT and oesophagoscopy, general anaesthesia was induced using intravenous propofol (Vetofol; Norbrook Laboratories) titrated to effect and maintained with isoflurane (Isoba, 1.5%; MSD Animal Health) in 100% oxygen. CT confirmed the presence of an alveolar infiltrate affecting the right cranial and middle lung lobes consistent with aspiration pneumonia. There was no evidence of pulmonary metastasis. The oesophageal lumen contained residual barium contrast material, and was severely dilated and stretched over a large soft tissue mass cranial to the cardiac silhouette (Figure 2a). Post-contrast (Iohexol, Omnipaque 300; GE Healthcare) images showed moderate patchy contrast enhancement within the mass (Figure 2b). The presence of stretching of the oesophageal lumen was suggestive of a close relationship between the mass and the oesophageal wall, and an oesophageal mural mass was considered likely. Oesophagoscopy was then attempted in order to visualise the mass and obtain biopsies; however, this was unsuccessful owing to a large amount of food material at the level of the mass, hindering visualisation.

Figure 1

(a) Right lateral view of the thorax showing a cranial mediastinal mass causing dorsal displacement of the thoracic trachea. Oesophageal dilation, cranial to the heart, is highlighted by the presence of residual oesophageal barium contrast administered by the referring veterinary surgeon approximately 36 h prior to this radiograph being taken. Colonic barium can also be seen. (b) Dorsoventral view of the thorax shows a rounded, soft tissue, cranial mediastinal mass. There is residual barium contrast within the oesophageal lumen, which is dilated and positioned on the left side of the mass. There is also an alveolar pattern within the right cranial lung lobe. At this stage, it was unclear whether the cause of these radiographic findings was an external mass deviating the oesophagus (such as a thymoma or enlarged mediastinal lymph node) or a mass directly involving the oesophagus

Figure 2

(a) Pre-contrast thoracic computed tomography (CT). The oesophageal lumen contains residual contrast material and is dilated and stretched over a large soft tissue-attenuating mediastinal mass cranial to the cardiac silhouette. The mass occupies most of the cranial thorax. (b) Post-contrast CT. The image shows moderate patchy contrast enhancement within the mass. The presence of stretching of the oesophageal lumen is suggestive of a close relationship between the mass and the oesophageal wall; oesophageal mural mass was considered more likely

To further characterise the lesion and potentially resect it, we opted to perform an exploratory thoracotomy and surgical exploration of the oesophagus. However, as aspiration pneumonia was suspected, surgical exploration was delayed. To facilitate enteral nutrition, a 16 French mushroom-tipped, silicone percutaneous gastrostomy feeding tube (Surgivet; Smiths Medical PM) was inserted surgically via a ventral mid-line coeliotomy after the oesophagoscopy was performed. Postoperative analgesia was maintained with buprenorphine (Buprecare, 0.02 mg/kg IV q8h for the first 24 h and then reduced to 0.01 mg/kg IV q8h; Animalcare). Antimicrobial therapy was initiated using intravenous amoxicillin/clavulanic acid (Augmentin, 20 mg/kg q8h; GlaxoSmithKline). This was then changed to clindamycin (Villertion, 10 mg/kg IV q12h; Invest) and marbofloxacin (Marbocyl 1% SA injection, 2 mg/kg IV q24h; Vétoquinol UK) as the rectal temperature had increased to 40.7°C the next day. Enteral nutrition was maintained through the gastrostomy tube with blended Hill's Prescription Diet i/d Feline. During this time, no food was ingested by mouth. Repeat serum electrolytes performed 24 h later were noted to be within normal limits.

Two days after admission, the cat was noted to be normothermic with a normal resting heart rate and respiratory rate. Four days later, prior to surgery, ultrasonography of the cranial thorax was performed. The cat was anaesthetised by the same regimen as described earlier. The mass was found to be heterogenous in echogenicity, with hypoechoic cavitary lesions. Transthoracic ultrasound-guided fine needle aspirates of the mass were obtained. These were examined immediately and were found to reflect spindle cell proliferation, but we could not confirm or exclude neoplasia; however, the presence of lymphoma was ruled out (Figure 3).

Figure 3

Examination of the smears of the mass lesion revealed a low number of dense clusters of cells with spindle-shaped nuclei and finely stippled chromatin. There was minimal anisokaryosis, and the cytoplasm was noted to be basophilic

Owing to the size of the mass, a median sternotomy approach was performed to allow thorough thoracic exploration. A firm, multi-nodular, soft tissue mass measuring 5 cm × 3.5 cm × 4 cm located cranial to the heart involving the ventral oesophageal wall was identified. An oesophageal stethoscope was used to confirm the position of the lumen of the oesophagus. Stay sutures were placed in the mass using 3/0 polydioxanone (PDS; Ethicon). An oesophagostomy was performed in normal oesophageal wall approximately 1.5 cm distant to the mass lesion; the mass was confirmed to be intramural. An oesophageal stethoscope was then advanced through the oesophageal lumen into the distal oesophagus and via the cardia into the stomach, and therefore confirming patency of the lumen at this level. The mass was excised sharply approximately 1.5 cm from the gross margins and placed in 10% neutral buffered formalin. The oesophagus was then reconstructed over the oesophageal stethoscope using a simple interrupted suture pattern in a single layer using 3/0 polydioxanone (PDS; Ethicon). The large size of the mass had distended the oesophageal wall so that adequate tissue was present for reconstruction, despite the removal of the large mass. The thoracic cavity was flushed copiously with saline, and an 18 F fenestrated 14 Ga × 20 cm chest tube (MILA International) was placed prior to routine closure of the thoracic cavity.

Haematoxylin and eosin-stained sections of the mass revealed a neoplasm of variable cellular density extending from the muscularis mucosa, elevating the surface mucosa and infiltrating the underlying tissues. Neoplastic cells were elongated and fusiform with abundant eosinophilic cytoplasm and elongated nuclei. Cell-dense areas were composed of cells arranged in interlocking streams, bundles and herringbone patterns (Figure 4).

Figure 4

Cell-dense areas composed of neoplastic smooth muscle cells elevate the mucosa and expand within the submucosa. Haematoxylin and eosin stain, × 40

Less cell-dense areas were populated by cells arranged in loose vascular channels and streams containing clusters of erythrocytes (Figure 5). Moderate-to-marked anisocytosis and anisokaryosis was noted throughout, and mitotic figures were 1 per 10 high-powered fields. In addition, areas of focal necrosis and haemorrhage were observed within the neoplastic tissue. Neoplastic cells extended within 5 mm of the lateral and deep margins of the sections examined. These findings were consistent with a low-grade sarcoma, with features highly suggestiveof an angioleiomyosarcoma. Immunohistochemically,neoplastic smooth muscle cells within cell-dense areas were immunopositive for α-smooth muscle actin (SMA) antibody (Figure 6). In less cell-dense areas, neoplastic vasoformative endothelial cells were immunopositive for von Willebrand factor antibody (Figure 7). CD117 (also known as c-kit) antibody immunoreactivity was not observed, confirming that the mass was an angioleiomyosarcoma.

Figure 5

Less cell-dense areas populated by cells arranged in loose vascular channels containing clusters of erythrocytes with abundant surrounding myxomatous material. Haematoxylin and eosin stain, × 10

Figure 6

α-smooth muscle actin (SMA) immunohistochemical stain revealed positive smooth muscle bundles (arrow) and smooth muscle surrounding vessels (asterix), × 200

Figure 7

Von Willebrand factor immunohistochemical stain revealed positive staining of vasoformative endothelial cells, × 200

Postoperative therapy included methadone (Physeptone 0.2 mg/kg IV q4–6h for 48 h and then reduced to 0.1mg/kg IV for 24 h; GlaxoSmithKline) and antimicrobial therapy was continued, as described previously, for 3 days and then switched to oral formulations (Kesium [Alstoe Animal Health] and Antirobe [Pfizer]) for a further 7 days. Intrapleural ropivacaine (Naropin 1 mg/kg; Astrazeneca UK) was instilled through the chest drain every 6 h; however, this was stopped 6 h postoperatively as the cat had pulled out the chest drain. The chest was drained every hour for the first 2 h and then every 2 h until the chest drain was removed 6 h postoperatively. Enteral nutrition was continued via the gastrostomy tube, as described previously. All oral medication was administered via the gastrostomy feeding tube. Water consumption was allowed 3 days postoperatively, and food was introduced the following day, at which time the cat voluntarily consumed 100% of its resting energy requirement. The gastrostomy tube was removed, and the cat was discharged 8 days postoperatively. At this stage, the cat was no longer on medication, and could eat soft food normally. As this was a low-grade sarcoma with clean (but narrow) surgical margins, a follow-up CT scan to monitor for local recurrence and evidence of haematogenous spread was recommended. However, the client declined follow-up monitoring owing to financial concerns. The cat was reportedly clinically normal at 6 month e-mail follow-up with the owner.

Angioleiomyosarcomas are very rare soft tissue tumours in animals and humans. In cats, these tumours have been most commonly reported to affect the dermis and subcutis. ¹ In this tumour, neoplastic smooth muscle cells blended imperceptibly with muscularis mucosae smooth muscle, possibly suggesting their derivation from the oesophageal muscularis mucosa. The main features of malignancy in this case included invasiveness and the presence of areas of necrosis. No information is currently available about the biological behaviour of these tumours at this site. In this cat, at the time of writing, the patient had not exhibited signs of metastatic disease. To our knowledge, oesophageal angioleiomyosarcomas have not been reported in domestic animals, and this is the first case study to be documented in a cat.

The categorisation of different smooth muscletumours is achieved through histological observations. Angioleiomyomas and angioleiomyosarcomas are often described as perivascular wall tumours made of closely packed interlacing bundles of neoplastic smooth muscle fibres originating from the smooth muscle tunica media of blood vessels.^1,2 Angioleomyosarcomas are identified by a higher mitotic rate, atypia and higher cellularity compared with angioleiomyomas. ¹ Another defining area is the presence of areas of necrosis. Mesenchymal tumours, such as leiomyoma, leiomyosarcoma and gastrointestinal stromal tumours (GISTs), have a similar cytological and histological appearance.^3–6 Aspirates of these tumours are often highly cellular and consist of long, thin mesenchymal cells with ‘cigar-shaped’ nuclei arranged in aggregates and linear bundles.^3–8 The most common locations for smooth muscle tumours in cats include gastrointestinal, cutaneous and the female reproductive tract.^1,3,4,8

Leiomyosarcomas and leiomyomas are known to readily express SMA (up to 84% ¹ ), desmin, to a lesser extent, and are usually negative for CD117 protein.^9,10 In humans and dogs, GISTs are known to consistently express CD117, as well as vimentin and CD34.^3–5,9 Up to 30% of GISTS in humans express SMA, while desmin and S-100 are not usually expressed.^5,9 A single report of a GIST in a cat has also been found to show a similar immunohistochemical pattern. ⁶ GISTs carry a much better prognosis than most malignant smooth muscle sarcomas owing to their ready response to tyrosine kinase inhibitors.^11,12 This prompts the need to accurately define these tumours and separate them from other mesenchymal neoplasms, which can be achieved through immunohistochemistry.^4,5

Oesophageal neoplasia is rare in cats and dogs. Although there are no recent epidemiological studies assessing the prevalence of disease, it is still believed to account for <0.5% of all documented cancers in dogs.^11,13,14 In cats, there have been no epidemiological studies to date, and the current literature consists of only a few individual case reports or case series.^10–15 The most common histological types in dogs include squamous cell carcinoma, leiomyosarcoma, fibrosarcoma and osteosarcoma.^11,14 Of the latter two, these neoplastic lesions have been reported to arise secondarily to spirocercosis in endemic regions. In cats, squamouscell carcinomas have been reported, and there has alsobeen one report of a neuro-endocrine carcinoma.^{10,11,12,15,16} Similar to humans, most primary oesophageal cancers in dogs and cats are usually found to be malignant, with moderate-to-high metastatic potential. ¹¹

Oesophageal surgery is typically associated with a high incidence of complications, including dehiscence, local necrosis and stricture formation. ¹⁷ In addition, intrathoracic resections are further complicated by poor exposure, lengthy resections and tension on anastomosis.^11,18 The unavoidable movement and friction caused by swallowing and breathing postoperatively also affects healing. ¹⁷ In this case, the stretching of the oesophageal wall caused by the mass allowed easy closure of the defect after resection of the tumour, with little tension. Two-layer, single-layer suture closure and surgical stapling of the oesophagus have been described with equal success in veterinary patients.^17,19 A single-layer closure was chosen in this case for speed and simplicity, and because the space at the surgical site would not have permitted the use of stapling equipment. Withholding food and water after oesophageal surgery is routine and may last for 24 h to 7 days, depending on different reports, and this is usually done in combination with facilitated feeding via gastrostomy tubes. In this case, we withheld oral water and food for 3 days, with a slow introduction to soft food thereafter. In this case, oral water and food were withheld for 3 days with slow introduction to soft food thereafter. The placement of the gastrostomy tube prior to thoracic surgery in this case was to aid in the stabilisation of the patient in order to reduce risks of anaesthesia and surgery.

Reports of chemotherapy in veterinary patients with oesophageal cancer are scarce. The use of cyclophosphamide and doxorubicin for dogs with gastrointestinal leiomyosarcomas has been described, but these reports are few in number. ¹⁴ Radiotherapy for the cervical oesophagus can be attempted, but is of limited value for the intrathoracic oesophagus because of the poor tolerance of surrounding normal tissues such as the lung and heart. ¹¹ In humans, surgical excision appears to be the best option for cure in patients with early-stage disease, and local control in patients with advanced disease. ¹⁸ However, like in humans, the majority of canine and feline patients have advanced disease at the time of diagnosis, making surgical resection difficult or impossible.^{10,11,14–17}

To our knowledge, reports of treatments of angioleiomyosarcomas in animals are few, and have only described surgical excision.^1,20 Local recurrence has been reported in 2/3 dogs diagnosed with cutaneous angeioleiomyosarcomas with follow-up information. ¹ Reports of adjuvant chemotherapeutic treatments following surgical excision in humans with angioleiomyosarcomas are scarce, but have described the use of ifosfamide, vepeside, doxorubicin and vincristine. ²¹

In this case, as the tumour was a low-grade sarcoma and clean margins were obtained, local recurrence was considered less likely. Because of this, adjuvant medical therapy or further local therapy was not warranted as the risks of adverse effects of these treatments were considered to be higher than the potential benefits. Further monitoring is indicated, and as haematogenous spread would be most likely, a repeat CT scan 6 months postoperatively would have been ideal to monitor for evidence of local recurrence or metastasis. However, as future follow-up monitoring was declined, we are unable to determine the progression of the disease.

Conclusions

Primary oesophageal cancer is reported rarely in cats, and little is known of its prognosis and treatment options. This report describes the successful surgical treatment of a large oesophageal angioleiomyosarcoma in a cat.