Case 1

A 10 yr old male neutered golden retriever presented to the referring veterinarian with a 12 hr history of dyspnea and exercise intolerance. There was no history of trauma. Pleural effusion was diagnosed by thoracic ultrasound and 400 mL of bloody fluid was removed by needle thoracocentesis.

At presentation to Queen Mother Hospital for Animals, Royal Veterinary College, London, the dog was tachypneic with a respiratory rate of 40 breaths per min with tachycardic, hypodynamic pulses with a rate of 160 beats per min (bpm). Respiratory sounds were bilaterally dull ventrally. Venous blood gas and biochemical analysis revealed marked hyperlactatemia (40 mg/dL reference interval [RI] 5–22 mg/dL) and mild azotemia with urea 3.6 mg/dL (RI 1.05–3.5 mg/dL) and creatinine 2.7 mg/dL (RI 0.9–2.5 mg/dL). Manual packed cell volume (PCV) was 25% with total solids (TS) of 4.6 g/dL. Thoracic ultrasonography, performed by a European College of Veterinary Radiology certified radiologist, demonstrated pleural effusion but no pericardial effusion or cardiac masses were visible. Thoracocentesis was repeated and a further 210 mL of fluid was removed. The PCV of the thoracic effusion was 35%. Blood typing was performed and the dog was found to be dog erythrocyte antigen (DEA) 1.1 positive. There was no evidence of a coagulopathy. Coagulation times were within the control reference intervals with a prothrombin time (PT) of 13.1 sec (control RI 11–17 sec) and activated partial thromboblastin time (aPTT) of 96 sec (RI 72–102 sec).

The dog received a transfusion of 2 units of packed red blood cell (PRBC) and two intravenous crystalloid fluid boluses of lactated Ringer's solution^a at a rate of 20 mL/kg over 20 min. When cardiovascularly stable as determined by a sustained reduction in heart rate to 100 bpm, the dog was anesthetised for a computed tomographic (CT)^b scan of the thorax and abdomen. Methadone (0.1 mg/kg)^c was administered and anesthesia was induced by alfaxalone^d 0.005 mg/kg and midazolam^e 0.4 mg/kg coinduction. A second unit of PRBC were administered after induction. On CT^b, there was a large volume of pleural effusion with collapsed lung lobes. No pulmonary lesions were identified. A 9 cm heterogeneous soft tissue attenuating mass was visible in the dorsal mediastinum, ventral to thoracic vertebrae 8 to 10, displacing the aorta and esophagus ventrally and compressing the vena cava (Figure 1A). Iohexol^ff, an iodine based contrast medium, at a dose of 2 mL/kg was administered IV. The mass had heterogenous uptake of IV contrast. The origin of the mass could not be determined. Abdominal CT was unremarkable.

View Figure

• Download as Powerpoint
• Download Figure

The primary differential diagnosis was a malignant vascular neoplasm, such as HSA. Exploratory thoracotomy was performed via median sternotomy. The mass was adherent to the aorta; therefore, dissection of the mass was not possible and the dog was euthanized intraoperatively. Postmortem was declined; however, histopathology of biopsy samples was consistent with HSA and revealed populations of proliferative partially vasoformative spindle cells forming channels containing erythrocytes. There were several large areas of extravasated erythrocytes consistent with recent hemorrhage.

Case 2

A 6 yr old male neutered Rottweiler presented to the Emergency service at Queen Mother Hospital for Animals with a 2 day history of progressive dyspnea, lethargy, polydipsia, and anorexia. The respiratory rate was 60 breaths per min and lung sounds were dull bilaterally over the ventral lung fields. The pulse rate was 200 bpm with poor peripheral pulse quality. The PCV on presentation was 25% with TS of 5.2 g/dL. After IV crystalloid resuscitation, with compound sodium lactate^a at a rate of 10 mL/kg as two boluses, the PCV fell to 16% and the TS to 4.0 g/dL. There was marked hyperlactatemia (102 mg/dL RI 5–22 mg/dL) on serum biochemical analysis. Ultrasound analysis was used to confirm pleural effusion and 1760mL of hemorrhagic fluid was drained; PCV of the fluid was 25% and TS was 3.6 g/dL. The dog had a blood type performed and was DEA 1.1 positive. The dog was stabilized with lactated Ringer's solution^a fluid resuscitation (10 mL/kg bolus) and one unit of PRBCs. The dog was not coagulopathic with a PT of 14 sec (RI 11–17 sec) and aPTT of 87 sec (RI 72–102 sec).

Two days after initial presentation, thoracic and abdominal CT under general anesthesia was performed to determine the etiology of the hemothorax. Methadone^b was administered (0.1 mg/kg) and anesthesia was induced by propofol^g 1.2 mg/kg and midazolam^d 0.4 mg/kg coinduction. The CT showed bilateral moderate pleural effusion with collapse of the left cranial and left caudal lung lobes. There was an 8 cm multilobulated peripherally hyperattenuating mass lesion ventral to the right middle lung lobe. The mass lesion was adjacent to the right atrium but was separated from the cardiac silhouette by a thin layer of pericardial fat (Figure 1B). The origin of the mass lesion could not be determined based on the CT. There was a 5 mm pulmonary nodule in the right caudal lung lobe. The sternal lymph node was enlarged as it measured 8 mm in diameter. A single focal hypoattenuating 7 mm nodule was present in the caudal aspect of the spleen. It was not aspirated.

Surgical exploration via a right sixth intercostal thoracotomy was performed under general anesthesia. On entry to the thorax, 1.5 L of hemorrhagic effusion were removed with suction. The pulmonary nodule identified on CT was not visible or palpable. There was an 8 cm spherical, multiple lobulated, firm, dark purple mass in the caudoventral mediastinum attached to the ventral sternum. Dorsally the mass surrounded the vena cava but was not adherent to it. The mass was circumferentially dissected from its pleural attachments using a bipolar tissue sealing device^h and blunt dissection. The entire mass was removed; however, there were multiple small 1 to 2 mm dark coloured nodules throughout the pleura, adjacent and distant to the mass.

A right-sided thoracic drainⁱ was placed exiting in the eighth intercostal space. The dog recovered well and the thoracic drain was removed 2 days postoperatively. Radiographs taken 5 days postoperatively revealed minimal pleural effusion and the dog was discharged. Histopathology was consistent with HSA with neoplastic spindyloid cells forming blood-filled channels. Neoplastic cells were present at the surgical margins and as were hemosiderin-containing macrophages indicating historical hemorrhage.

While an appointment 2 wk after surgical discharge was recommended with the oncology service, the owners returned 4 wk postoperatively for re-evaluation. The incision had healed well and no further dyspneic episodes had been noted. A chemotherapy protocol was initiated; doxorubicin^j (30mg/m²) was administered IV. Cyclophosphamide^k (200mg/m²) was dispensed and was to be given in a divided dose per os for 2 days, 10 days after each doxorubicin^j administration.¹¹

The dog represented 2 days after doxorubicin^j administration. Physical examination findings were consistent with hypovolemic shock. The dog was moderately dyspneic with bilateral pleural effusion confirmed with thoracic ultrasound. Thoracocentesis revealed hemorrhagic fluid with a PCV of 24%. The owners elected for euthanasia. No further diagnostic tests were performed. Postmortem examination was declined.

Case 3

A 9 yr old male neutered Labrador retriever presented for management of hemothorax diagnosed by the referring veterinarian and a 24 hr history of dyspnea and lethargy. On presentation, respiratory rate was 72 breaths per min with muffled respiratory and cardiac sounds ventrally. The heart rate was 180 bpm with hyperdynamic pulses. The PCV was 22% and TS 4.8 g/dL. Emergency blood gas and serum biochemical analysis was unremarkable. Hemothorax was diagnosed on the basis of hemorrhagic pleural fluid on thoracocentesis with PCV of 27% and TS of 5.4 g/dL. The dog was stabilized with intravenous crystalloid boluses of lactated Ringer's solution^a at 10 mL/kg for 2 boluses. A blood type was performed and was DEA 1.1 positive. One unit of PRBC (DEA 1.1 positive) was administered. The dog was not coagulopathic with a PT of 15 sec (RI 11–17 sec) and aPTT of 85 sec (RI 72–102 sec).

Methadone^c was administered (0.1 mg/kg) and anesthesia was induced by propofol^g titrated to effect and midazolam^e 0.4 mg/kg coinduction. A CT scan revealed a discrete, smoothly marginated, multilobular, 6 cm diameter mass lesion in the caudo-ventral thorax, which was heterogeneously enhancing following intravenous contrast administration^f (Figure 2). The mass was adjacent to the xiphoid process at the level of the sixth and seventh ribs. It was not possible to determine the origin of the mass from the CT scan.

View Figure

• Download as Powerpoint
• Download Figure

As a result of ongoing hemorrhage, exploratory surgery was performed. Due to the location of the mass on the dorsal aspect of the sternum, a cranial midline laparotomy and transdiaphragmatic approach was performed. A discrete mass was present in the caudoventral thorax arising from a broad base on the right dorsal aspect of the xiphisternum. The mass was purple in colour with a highly vascular surface. No other adhesions or areas of invasion to the pulmonary parenchyma or pleura were present making surgical excision of the mass feasible, although a midline sternotomy from the xiphoid to the level of the fifth sternebrae was necessary to improve exposure. The mass was not cut into using this approach, but an ostectomy of the xiphisternum and sixth sternebrae was necessary as they were firmly attached to the broad base of the mass. An oscillating saw^l was used for removal of the sternebrae. The intercostal vessels and any aberrant vasculature to the mass were ligated. A right-sided silicone thoracic drainⁱ was placed in the eighth intercostal space. Tension-free reconstruction of the thoracic wall deficit was achieved by diaphragmatic advancement and apposing the rectus abdominus muscle.

The dog received a further unit of PRBC, but made an uneventful recovery from general anesthetic and the surgery. The thoracic drain was removed 12 hr postoperatively and the dog was discharged 4 days postoperatively. Histopathology of the mass was consistent with an intramuscular HSA with spindyloid neoplastic cells forming sheets and irregular vascular channels containing erythrocytes. The neoplastic cells extended to the surgical margins and there were large areas of necrosis present.

The dog presented to the Oncology service for a scheduled re-examination 2 wk postsurgery. Physical examination and in-house hematology were unremarkable. The surgical incision had healed well. The same chemotherapy protocol as case 2 was employed.¹¹

The dog presented 3 wk after the initial doxorubicin^j dose. The owners reported mild polyuria and polydipsia. The surgical site had healed well and no abnormalities were detected on clinical examination. Hematology revealed a leukopenia with a white blood count of 2.82 × 10⁹/l (RI 5.5–16.9 × 10⁹/l) and neutrophil count of 1.23 × 10⁹/l (RI 2–12 × 10⁹/l). The doxorubicin^j was postponed for two days and prophylactic oral amoxicillin-clavulanate^m 16 mg/kg was initiated twice daily.

Five mo postoperatively, at the time of writing, the dog had no reoccurrence of dyspnea or lethargic episodes. No further imaging studies were performed. The dog received four cycles of doxorubicin^j and cyclophosphamide^k protocol. Following the second cycle of Doxorubicin/ Adriamycin and cyclophosphamide chemotherapy the dog experienced grade III neutropenia (Veterinary cooperative oncology group – common terminology criteria for adverse events v1.1), so the doxorubicin dosage was reduced to 27 mg/m2 and the cyclophosphamide dosage was reduced to 150 mg/m2. This was successful and no further postchemotherapy episodes of neutropenia were noted.