Lung lobe torsion in association with a chronic diaphragmatic hernia and haemorrhagic pleural effusion in a cat

Abstract

Clinical summary This report describes torsion of the right cranial lung lobe in a cat with haemorrhagic pleural effusion and a chronic diaphragmatic hernia. Surgical treatment comprising lung lobectomy without de-rotation, and repair of the diaphragmatic defect, led to an uneventful recovery.

Practical relevance Lung lobe torsion is a rare condition in cats. While spontaneous lung lobe torsions may occur, a frequent association with underlying thoracic disease has been recognised in cats. However, neither haemorrhagic pleural effusion nor diaphragmatic hernia have been previously described in cats with lung lobe torsions, although they have been documented in dogs and humans. In a cat with suspected lung lobe torsion, a thorough search for an underlying disease should be undertaken.

Lung lobe torsions (LLTs) are rare in cats, with just 13 cases reported over the past 40 years.^1–9 Affected breeds have included domestic mixed breeds (nine cases), Persians (two), Siamese (one) and a Himalayan.^1–9 Males (10/13) appear to be overrepresented. The median age of affected cats is 9 years (range 10 months to 19 years).

LLT involves axial rotation of the lung lobe, resulting in occlusion of the bronchus and pulmonary vein. While LLT may arise spontaneously, a strong association with pre-existing thoracic disease has been recognised in cats. Any condition that results in atelectasis, pleural effusion or disruption of the normal spatial relationship between adjacent lung lobes can theoretically predispose to LLT. ¹⁰ Ten of the 13 (77%) reports documenting feline LLTs cited underlying thoracic disease including sterile pleuritis (three cases), idiopathic chylothorax with prior thoracic duct ligation (two), peritoneopericardial hernia (one), chronic bronchitis with bronchiectasis and bronchial occlusion (one), pericardial haemorrhage (one), pyothorax (one) and mediastinal lymphosarcoma (one).^1–6,9

Once torsed, the affected lobe becomes congested and necrotic. Subsequent clinical signs include tachypnoea, increased respiratory effort, coughing, lethargy and anorexia.^1–6,9 Haemoptysis and epistaxis may occur if the bronchus is only partially occluded.^8,11,12 Concurrent pleural effusion has been noted in 92% (12/13) of feline and 100% of canine LLT cases.^1–20 The type of effusion may reflect underlying disease. In cats with LLT, reported effusions have included chylous effusions (five cases, three of which were pre-existing), modified transudates (three) and exudates (three cases, one non-septic, one septic and one neoplastic).^1–7,9

This case report describes LLT in conjunction with chronic diaphragmatic herniation and haemorrhagic effusion. Neither of these findings have previously been associated with LLT in cats, although they have been reported in dogs and humans.^7,11–13,20

Clinical report

A 10-year-old, female neutered Chinchilla presented to the referring veterinarian with acute onset of respiratory distress. The cat resided indoors and there was no history of recent trauma. Thoracic radiographs revealed a bilateral pleural effusion. Thoracocentesis yielded 200 ml of grossly haemorrhagic effusion. Bilateral thoracostomy tubes (8F feeding tube, Tyco Healthcare) were placed under sedation. Oxygen supplementation was provided over the following 8 h.

Cytological analysis of the pleural fluid revealed a predominance of erythrocytes with moderate numbers of non-degenerate neutrophils and macrophages. Erythrophagocytosis was evident. No infectious organisms were identified. Aerobic and anaerobic bacterial culture was negative. Post-thoracocentesis thoracic radiographs revealed consolidation of the right cranial lung lobe (Figure 1). Intravenous fluid therapy was commenced along with clavulanic acid/amoxicillin, 13.8 mg/kg SC q12h (Clavubactin S; Bomac Animal Health) and metronidazole, 15 mg/kg IV q12h (Metronidazole BP IV infusion; Baxter Healthcare). Although the respiratory signs improved, the cat remained depressed and inappetent. Repeat aspiration of the thoracostomy tubes was unproductive.

Figure 1

Left lateral thoracic radiograph taken by the referring veterinarian on the day of initial presentation showing an air bronchogram (arrowheads) within the consolidated right cranial lung lobe. Pleural effusion has obscured the cardiac silhouette and resulted in leafing of the caudal lung lobes. The diaphragmatic crura are not clearly defined. The thoracostomy tubes are poorly positioned

One week after initial presentation, the cat was referred to the Melbourne Veterinary Specialist Centre. Tachypnoea with intermittent open-mouth breathing was noted. A haemogram and biochemistry panel were unremarkable apart from mild monocytosis (0.9 x 10⁹/l; reference interval [RI] 0–0.7 x 10⁹/l) and hyperkalaemia (5.8 mmol/l; RI 3.7–5.4 mmol/l). Prothrombin and activated partial thromboplastin times were normal.

Thoracic ultrasonography (LOGIQ P5; GE Healthcare) revealed right cranial lung lobe consolidation (Figure 2a). No blood flow could be demonstrated within the lobe using pulsed wave Doppler. The spleen and intestinal loops were identified within the left caudal thorax (Figure 2b). Echocardiography was unremarkable. Thoracic radiographs (Fujifilm FCR Capsula XL; Fujifilm Corporation) were repeated (Figure 3).

Figure 2

Ultrasound images obtained from the right fourth intercostal space (a) and left sixth intercostal space (b). The arrowheads in (a) delineate the margins of the consolidated, heterogeneous right cranial lung lobe. No blood flow could be demonstrated within this lung lobe using Doppler (not shown in this image). The white star marks a blood vessel adjacent to the medial aspect of the lung lobe. The arrow in (b) points to an intestinal loop within the thorax

Figure 3

Preoperative dorsoventral thoracic radiograph revealing persistent consolidation of the right cranial lung lobe (white arrow). The diaphragmatic crura were not clearly defined. The soft tissue density within the left hemithorax (black arrow) is likely to represent the herniated spleen. Loops of gas-filled intestine are evident within the right thorax (arrowheads). The left thoracostomy tube is poorly positioned (angled across the caudal thorax). The right thoracostomy tube is located subcutaneously

Intravenous fluid therapy (0.9% sodium chloride, 10 ml/kg/h) was recommenced following premedication with buprenorphine, 0.005 mg/kg SC (Temgesic; Reckitt Benckiser). Oxygen supplementation was provided for 5 mins prior to anaesthetic induction with intravenous propofol (Fresofol 1%; Pharmatel Fresenius Kabi). General anaesthesia was maintained with isoflurane (Isorrane; Baxter Healthcare) and oxygen. Intermittent positive pressure manual ventilation was provided using an in-line pressure manometer (peak pressure <18 cm H₂O, Surgivet; Smiths Medical). Both thoracostomy tubes were removed.

A ventral midline abdominal approach revealed a radial diaphragmatic tear extending from the central tendon through the pars costalis. The edges of the defect were irregular, which was consistent with a traumatic diaphragmatic rupture. The diaphragmatic lesion was enlarged to facilitate reduction of the herniated contents, which included the colon, spleen, omentum and small intestine. Biopsies of several hyperaemic nodules on the serosal surface of the colon were obtained via blunt dissection.

A 4 mm rigid 30^o oblique Storz thoracoscope was introduced into the right hemithorax through the diaphragmatic defect. The right cranial lung lobe was congested and a torsion was identified at the hilus (Figure 4). The midline abdominal incision was advanced cranially and a caudal median sternotomy was performed. Lung lobectomy was carried out using a TA premium 30 V3 stapler (Tyco Healthcare/United States Surgical) placed around the hilus of the lobe. A new thoracostomy tube (12F Trocar catheter; Tyco Healthcare/Kendall Argyle) was placed. The thorax was lavaged with sterile saline. The median sternotomy was then closed with interrupted alternating figure-of-eight transsternal 1 polypropylene sutures (Prolene; Johnson & Johnson).

Figure 4

Intraoperative thoracoscopy reveals a consolidated and congested right cranial lung lobe. The hilus was torsed (not visible in this image)

The edges of the diaphragmatic hernia were debrided and the defect was closed with a combination of simple interrupted and continuous 3/0 polydioxanone sutures (PDS II; Johnson & Johnson). The abdomen was lavaged with warm sterile saline and closed routinely. Postoperative analgesia was provided via a fentanyl continuous rate infusion, 0.003 mg/kg/h (DBL Fentanyl Injection; Hospira) and a buprenorphine patch (Norspan 5 Transdermal Patch; Mundipharma). Subsequent gentle intermittent aspiration of the thoracostomy tube was unproductive. The thoracostomy tube and oxygen therapy were withdrawn 12 h postoperatively. The cat was discharged 36 h after surgery. No further clinical signs were observed over the following 10 months.

Histopathology of the right cranial lung lobe revealed diffuse haemorrhage, necrosis and inflammation. The bronchi contained haemorrhagic casts and fibrosis of the pleural serosa was evident. These findings were consistent with pulmonary congestion, haemorrhage and infarction secondary to LLT. The nodules on the surface of the colon were ectopic splenules.

Discussion

While LLT is commonly associated with underlying thoracic disease in cats, there have been no prior reports of LLT in association with diaphragmatic herniation. In this case, it would appear most likely that the diaphragmatic hernia was acquired as a consequence of blunt trauma, given the irregular nature of the defect and the presence of splenules (which suggest prior splenic rupture). ²¹ The concurrent pleural fibrosis suggests a chronic duration.^20,22

The presence of abdominal viscera within the thoracic cavity can lead to atelectasis, bronchial displacement and, in 10–29% of cases, a pleural effusion.^20,22–27 Subsequent disruption of the spatial relationship between adjacent lung lobes and the thoracic wall could predispose to LLT. The cranial and middle lung lobes are prone to torsion as they lack supportive pulmonary ligaments which secure the caudal lung lobes to the caudal mediastinum. ²⁸ Torsion of the left cranial and right middle lobes has been reported in conjunction with diaphragmatic herniation in dogs. In this case, torsion of the right cranial lobe was observed.

A review of the available literature on LLT (due to any cause) in dogs and cats revealed that the right middle lobe is affected most commonly (43.2% dogs, 46% cats), followed by the left cranial lobe (35.1% dogs, 31% cats) and then the right cranial lobe (10.8% dogs, 23% cats).^{1–20,29–31} Involvement of caudal, accessory and multiple lobes has also been reported in dogs.^{7,10,11,15–18,30,31}

In cases of chronic diaphragmatic herniation without LLT, the pleural effusion is usually associated with entrapment of abdominal viscera within the hernia and subsequent congestion or necrosis.^20,26 In this case, the haemorrhagic pleural effusion was likely to be the result of congestion and necrosis of the lung lobe itself, given that there was no evidence of congestion or necrosis of the abdominal viscera or diaphragm. Iatrogenic haemorrhage during thoracocentesis was also unlikely as the fluid failed to clot on exposure to air and microscopy revealed erythrophagocytosis and an absence of platelets. ³² While haemorrhagic effusions have been documented in dogs and humans with LLTs, this is the first report of a haemorrhagic effusion in association with LLT in a cat.

Cats are more likely to have underlying thoracic disease in association with LLT (77%) compared with dogs (18–43%).^1–10,13 When LLT is suspected, a thorough search for a possible underlying disease should be undertaken. Preoperative identification of the concurrent diaphgramatic hernia significantly altered the surgical approach in this case. Thoracoscopy was also used to confirm the presence of LLT before the median sternotomy was performed.

Lung lobectomy is the treatment of choice for LLT. The torsed lung lobe was removed without de-rotation because prior studies have suggested an increased incidence of perioperative mortality if the lung lobe is de-rotated prior to resection. ¹³ De-rotation may release cytokines, toxins and oxygen radicals, resulting in reperfusion injury. ¹⁰ In dogs, peri-operative mortality rates following lung lobectomy have ranged from 10–47%.^10,13 Outcome was reported in 10 of the published feline LLT cases. Four cats were euthanased before a definitive diagnosis was made.^2,3,6 All six cats that underwent lung lobectomy survived.^1,3–5,9 Only two major complications were noted; one cat had an isolated grand mal seizure within 24 h of surgery (cause not determined) and one cat developed a recurrent chylothorax 7 days postoperatively which required thoracic duct ligation.^4,5

No postoperative complications were observed in this cat, and it would appear that the prognosis for cases of feline LLT is good. While recurrent LLT has been reported in the canine literature (approximately 11%), recurrence has not been reported in cats.^10,15,16