Thoracoscopic Lung Lobectomy for Treatment of Lung Tumors in Dogs
Abstract
Objective— To report use of thoracoscopic lung lobectomy (TLL) for treatment of lung tumors (LT) in dogs.
Study Design— Retrospective study.
Animals— Nine dogs.
Methods— Dogs that had TLL for tumor removal were included. Using general anesthesia and 1-lung ventilation, TLL was performed using a 30–60 mm endoscopic gastrointestinal anastomosis stapler. If the visual field was obscured, lobe resection was completed via thoracotomy.
Results— Metastatic and primary LT were resected by thoracoscopic lobectomy in 9 dogs (6 male, 3 female; mean (±SD) weight, 29±7 kg; mean age, 10.7±1.9 years). Six dogs had a solitary mass and 3 dogs had 2 masses within a single lobe. The left caudal lobe was removed in 3 dogs. In 5 dogs, TLL was used alone whereas conversion to thoracotomy was required in 4 dogs because of poor visibility. There were 7 metastatic LT and 2 primary LT. Mean duration of thoracoscopic surgery was 108.8±30.3 minutes compared with 150.75±55.4 minutes in dogs requiring conversion to thoracotomy. Mean hospitalization was 3.1±1.3 days.
Conclusion— Provided the visual field is not obscured, TLL can be performed effectively in dogs.
Clinical Relevance— Dogs with metastatic or primary LTs should be considered for TLL, particularly for small masses positioned away from the hilus in the left caudal lung lobe.
INTRODUCTION
LUNG TUMORS (LTs) occur in a wide variety of dog breeds, affecting mainly older dogs (median age, 11 years).1,2 Primary LTs occur rarely (incidence approximately 1% of all canine tumors) and are solitary in 54% of cases.1–3 Malignant carcinomas represent 97% of cases.2 Surgical resection is the only potentially curative option for primary tumors. Conversely, pulmonary metastatic masses are commonly seen with many tumor types.4 Metastatic lesions can be seen at initial presentation in dogs diagnosed with thyroid cell carcinoma, hemangiosarcoma, melanoma, and osteosarcoma.4 Although prognosis associated with metastatic masses depends on the nature of the primary tumor and the degree of lung involvement, in selected cases, surgical resection of metastatic lesions can prolong survival time.2,5
Thoracoscopy is routinely performed as a diagnostic and therapeutic procedure in humans.6–8 When compared with thoracotomy, similar procedures performed using thoracoscopy in humans result in reduced chest wall trauma and deformity, reduced postoperative pain, decreased patient morbidity, and shorter hospitalization stay.8 Thoracoscopy has been used in dogs for biopsy of lung, mediastinum and pleura,7 treatment of persistent right aortic arch,9,10 identification and ligation of the thoracic duct,11 pericardiectomy,12,13 and occlusion of patent ductus arteriosus.14 In human medicine, excisional biopsy plus evaluation of hilar and mediastinal lymph nodes with thoracoscopy is the modality of choice for diagnosis and treatment of lung masses.8 Lymph node staging and surgical resection margins attained by thoracoscopy are identical to those obtained by thoracotomy.8 Lung lobectomy with thoracoscopy has been reported in healthy dogs15; however, to our knowledge, experience with multiple clinical cases has not been reported.
Our objective was to report our experience with use of thoracoscopic lung lobectomy (TLL) for treatment of LTs, and outcome, in 9 dogs.
MATERIALS AND METHODS
Criteria for Inclusion
Medical records (1999–2003) of all dogs that had TLL for tumor removal were reviewed. Retrieved data were: signalment, body weight, past medical history, preoperative diagnosis, number and location of lobes removed, results of histologic examination of resected lobe(s), and hospitalization. Surgical data included: surgical technique(s), duration of surgery, intra- and postoperative complications. Duration of surgery was defined as the time from initial incision to closure, determined from the anesthetic report.
Preoperative evaluation included complete blood count, serum biochemical profile, urinalysis, and three projection thoracic radiographs.
Anesthesia
Dogs were administered a combination of morphine (1 mg/kg, subcutaneously [SC]), and glycopyrrolate (0.01 mg/kg SC) or atropine (0.04 mg/kg SC), with or without acepromazine (0.02–0.03 mg/kg SC). Oxymorphone (0.05 mg/kg SC) instead of morphine was used in 1 dog. Anesthesia was induced with propofol (1.3–3.5 mg/kg, intravenously [IV]) and diazepam (0.1 mg/kg IV) in 7 dogs or midazolam (0.2 mg/kg IV) in 1 dog, and in the other dog with thiopental (4 mg/kg IV) and diazepam (0.4 mg/kg IV). Intraoperative IV fentanyl was used either as bolus injection (2–3 μg/kg) in 7 dogs or constant rate infusion in 3 dogs (10–20 μg/kg/h). Anesthesia was maintained using isoflurane (0.5–2.5% inspired) or sevoflurane (0.75–2.5% inspired) in oxygen with 1-lung ventilation and positive end expiratory pressure (PEEP) in a semi-closed circle system.
One-lung ventilation was achieved by inserting an endobronchial blocker (Arndt endobronchial blocker, Cook Critical Care, Bloomington, IN) into the right or left main-stem bronchus to deflate the lungs on the side with the affected lobe. The endobronchial blocker was passed through the lumen of the endotracheal tube and advanced to the level of the main-stem bronchus under direct observation with a 5.3 mm flexible video bronchoscope (Olympus BF type 240 series, Olympus America Inc., Melville, NY). The silicone balloon of the endobronchial blocker was inflated with air until complete occlusion of the bronchus to the affected lobe was achieved. Correct placement of the endobronchial blocker was verified throughout the surgery by direct observation of lung deflation with the thoracoscope. PEEP was maintained at 5–10 cm H2O during thoracoscopy.
Surgery
Dogs were positioned in lateral recumbency with the unaffected lung side located ventrally. Three or four 12 mm cannulae (Ethicon Disposable rigid trocars, Ethicon Endosurgery Inc., Cincinnati, OH) were inserted between the 4th and 10th intercostal spaces using triangulation to observe and provide access to the diseased lobe. For resection of caudal lung lobes, the cannula for the camera was inserted in the ventral aspect of the 8th intercostal space. One instrument cannula was inserted dorsally in the 10th intercostal space to gain access to the dorsal ligament of the caudal lung lobe and a 2nd instrument cannula was inserted in the 7th intercostal space. Stapling equipment was introduced through a cannula in the middle of the 6th intercostal space. For resection of cranial lung lobes, the telescope with video camera (5 mm, 30° telescope, Veterinary Video Camera II, Storz Veterinary Endoscopy, Goleta, CA) was inserted in the ventral part of the 7th intercostal space. The cannula for the stapling device was located in the dorsal half of the 8th intercostal space and an instrument cannula was inserted in the middle of the 5th intercostal space.
Dissection was performed using electrocautery and endoscopic Metzenbaum scissors (Endo shear, Ethicon Endosurgery Inc., Cincinnati, OH) as needed and retraction was performed using atraumatic Babcock forceps (Endo Babcock, Ethicon Endosurgery Inc.). Lobectomy was achieved using a 30–60 mm endoscopic gastrointestinal anastomosis stapler with 3.5 mm cartridges (EndoGIA, United States Surgical Corporation, Tyco Healthcare Group Norwalk, CT). A 30-mm thoracic abdominal stapler (TA 30, United States Surgical Corporation) was used if conversion to thoracotomy occurred. Resected lobes were placed in an endoscopic retrieval bag (Endobag, United States Surgical Corporation) in 2 dogs. Lobes were removed through an extended incision of 1 instrument portal without retraction of ribs in dogs that did not require thoracotomy. Three dogs were administered local intercostal bupivicaine (1.5 mg/kg total dose divided between cannula sites) at skin closure. Fentanyl (1–4 μg/kg/h) was administered by IV constant rate infusion postoperatively until dogs were subjectively evaluated to be comfortable by the attending clinician.
A chest tube was placed in a routine manner from dorsal to ventral, caudal to cranial traveling over 3 intercostals spaces, avoiding cannula sites in all dogs. Short-term survival was defined as discharge from the hospital. Long-term follow-up was performed using documented client communications in medical records and by telephone conversation with owners where possible. Data were reported as mean±SD.
RESULTS
There were 6 male and 3 female dogs (mean weight, 29 kg; range, 15–37 kg). Dog breeds were: 3 Labrador Retrievers, 1 Australian shepherd, 1 standard poodle, 1 Alaskan malamute, 1 golden Retriever, and 1 flat-coated Retriever. Mean age was 10.7 years (range, 8.2–13.25 years).
Six dogs had previous neoplastic disease: 1 dog each with salivary gland adenocarcinoma, mammary gland adenocarcinoma, osteosarcoma, and synovial cell sarcoma and 2 dogs had complex neoplastic disease including melanoma plus hepatic carcinoma, and fibrosarcoma plus malignant histiocytoma. Three dogs did not have a history of neoplasia: 1 had acute respiratory distress and no previous medical history, 1 had been treated for immune-mediated hemolytic anemia and was in remission at admission, 1 had bilateral laryngeal paralysis.
In addition to thoracic radiography, thoracic fluoroscopy was used in 2 dogs, and nuclear scintigraphy, fine needle aspirate, and bronchoalveolar lavage cytology in 1 dog each to establish a diagnosis. Preoperative diagnosis was metastatic lung disease in 6 dogs and primary LT in 3 dogs; 7 diagnoses were presumptive based on history. Six dogs had a solitary mass and 3 had at least 2 masses within the same lung lobe. Tumor locations were right cranial lobe (1 dog), right middle lobe (1), right caudal lobe (2), left cranial lobe (2), and left caudal lobe (3).
Conversion to thoracotomy was required in 4 dogs because of poor intraoperative visibility caused by: bleeding from an intercostal artery (1 dog), failure of 1-lung ventilation to maintain lung deflation (2), and poor access (1), which occurred in a dog with a mass in the right middle lung lobe (2 cm × 2 cm mass located centrally).
Histopathologic examination identified 7 metastatic tumors including: malignant histiocytoma, osteosarcoma, undifferentiated carcinoma (2 dogs), undifferentiated sarcoma (2), and adenocarcinoma (1). The dog admitted for acute respiratory distress had an undifferentiated carcinoma. Two primary LTs, bronchogenic carcinoma and bronchioalveolar carcinoma, were identified. Microscopically complete resection margins were achieved in 4 thoracoscopic procedures and 3 thoracotomy procedures. One complete lobe was removed in all dogs: 2 right cranial, 1 right middle, 2 right caudal, 1 left cranial, and 3 left caudal lobes. All 3 left caudal lung lobe masses were removed by thoracoscopy alone. One right cranial lobe mass was preoperatively determined to be on the left side. Lobectomy was completed by thoracoscopy, without complication, by dissection through the mediastinum to approach the left side.
Mean duration of surgery was 109 minutes (range, 75–150 minutes) in dogs that had thoracoscopy only and 151 minutes (range, 85–210 minutes) in dogs that had thoracotomy after thoracoscopy. A chest tube was inserted in all dogs before recovery from anesthesia.
Mean duration of fentanyl administration for postoperative analgesia was 29.5 hours (range, 17–58 hours). Seven dogs were also administered lidocaine (1–1.5 mg/kg) and bupivicaine (1–1.5 mg/kg) combined through the chest tube. Another dog was also administered piroxicam (0.3 mg/kg per OS every 24 hours) and 1 dog also had a fentanyl patch (0.7 μg/kg).
Postoperative complications were hemorrhage from an intercostal artery requiring anesthesia to place a suture to ligate the artery (1), thrombocytopenia and hyperthermia (1), corneal ulceration (1), subcutaneous emphysema associated with a chest tube (1), aspiration pneumonia with poor oxygenation requiring nasal oxygen supplementation (1), and poor oxygenation requiring nasal oxygen supplementation (2). All dogs had air and fluid removed through their chest tubes using a syringe every hour for the first 4 hours postoperatively and then every 4–6 hours or as needed. Duration of chest tube placement ranged from 10 to 43 hours (mean, 23.6 hours).
Mean duration of hospitalization was 3.1 days (range, 2–6 days). The dog that was hospitalized for 6 days had a unilateral laryngeal lateralization at the time of thoracoscopy. Although this dog had a lung lobe removed by thoracoscopy without conversion to thoracotomy, postoperative aspiration pneumonia occurred, requiring nasal oxygen therapy and prolonged hospitalization. Exclusion of this dog yielded a mean hospitalization of 2.8 days (range, 2–4 days). All were discharged from hospital. One dog was lost to long-term follow-up the day of discharge. Documented client communication in medical records provided follow-up for 3 dogs; all had returned for repeat radiographs and were reported alive at 2, 5, and 13 months postoperatively. There were no indications of complications from TLL and all dogs were alive at last communication. Owners of 5 dogs were available for telephone interview (mean, 3.8 years; range, 2–5 years). Three dogs had thoracoscopy only. Mean survival time was 1.3 years (range, 6 months–2 years). Four dogs had been euthanatized and 1 dog died after prolonged status epilepticus. All owners reported excellent short-term recovery without complications and further stated that based on their experience they would elect surgery using a minimally invasive approach again.
DISCUSSION
Thoracoscopic lung resection without conversion to thoracotomy was accomplished in 5 of 9 dogs. Left caudal lung lobe masses were removed successfully by thoracoscopy without intraoperative complications. Thoracoscopic caudal lung lobectomy was easiest although cranial lung lobectomy was also successfully accomplished. The most common reason to change from thoracoscopy to thoracotomy was poor operative vision with failure of 1-lung ventilation being the most common reason.
One-lung ventilation is a technique for providing optimal hilus visualization during thoracoscopic procedures and is frequently used in human thoracoscopy.16 Deflation of the surrounding lobes reduces surgical time because visualization of the operative field is improved and the risk of iatrogenic trauma to unaffected lobes is substantially reduced. We induced 1-lung ventilation by inserting, under bronchoscopic guidance, an endobronchial blocker into the bronchi we wanted to occlude. We did not achieve adequate 1-lung ventilation in 2 dogs because the blocker moved during patient positioning for surgery. Initially, our protocol involved placing the endobronchial blockers while the dog was in the anesthesia induction area before transport to the operating room. Subsequently, we induced 1-lung ventilation in the operating room after final surgical positioning without any further failures. Therefore, we recommend placing the blocker with endoscopic guidance after the patient has been positioned for surgery but before sterile draping is complete.
One-lung ventilation causes massive right to left shunt in the lungs resulting in a low V/Q mismatch.17 Thus 1-lung ventilation has the potential to induce a substantial reduction in oxygen saturation. To prevent desaturation a positive end expiratory pressure (PEEP) of 5 cm H2O was used. PEEP maintains the alveoli of the ventilated-dependent lung open, which prevents further deterioration of gas exchange. One-lung ventilation for open thoracic surgery does not reduce the oxygen delivery index in dogs,17 and use of PEEP does not seem to affect oxygen delivery and cardiac output in dogs during thoracoscopy.18 Detrimental side effects of the anesthesia technique we used were not observed.
Hemorrhage from an intercostal artery, most likely from trauma during cannula insertion, resulted in poor visualization in 1 dog. Whereas bleeding was minimal, it was directly over the camera at the 8th intercostal space and it was not possible to ligate the artery under thoracoscopic direction, so we converted to thoracotomy where the vessel was ligated and lobectomy performed. Hemorrhage from a cannula occurred in 1 dog in the postoperative period and required a second surgery to ligate the intercostal artery. We recommend monitoring the cannula sites at the time of cannula removal and the placement of sutures to control any bleeding. The risk of fatal hemorrhage during thoracoscopy in dogs is unknown. In 1239 human thoracoscopic lung lobectomies only 10 were converted to thoracotomy because of hemorrhage; no deaths occurred from hemorrhage and blood transfusion was not required in every patient.19
Poor access occurred only for the right middle lobe; thoracoscopic removal of this lobe has not been reported in dogs. We found that placement of the cannulae did not provide enough working space to manipulate the instruments and the middle lobe for safe application of staples at the hilus. Most likely cannulae would have to be inserted in the 10th, 11th, or 12th intercostal spaces to be able to visualize the hilus and apply staples. With more experience, resection of the right middle lobe should be possible with thoracoscopy. It is our impression that the accessory lung lobe located between the right and left caudal lobes would also be difficult to resect using thoracoscopy. Hilar lymph nodes can be biopsied by thoracoscopic technique if they are enlarged. None of our dogs had enlarged lymph nodes most likely because the tumors were small and had not yet metastasized to regional lymph nodes.
One of the benefits of thoracoscopy is that a magnified view is provided by the endoscope and this improved visualization. In the dog where the contralateral side was mistakenly entered, the tumor in the cranial lung lobe could clearly be seen through the transparent mediastinum with the endoscope. Our decision to dissect through the mediastinum was made because we could confidently view and avoid mediastinal structures. Because the mass was small and the lobe could readily be accessed it was removed from the contralateral side without modifying portal placement. Thoracoscopic-assisted lung lobectomy could be an alternative when there is poor visualization during thoracoscopy.20 The thoracoscope would be inserted into the thoracic cavity to localize the hilus, after which a limited thoracotomy would be performed at the level of the hilus to exteriorize and then resect the lung with staples placed at the base of the lobe under thoracoscopic observation.
We used stapling equipment designed specifically for minimally invasive surgery to perform the lung lobectomy. This stapling instrument simultaneously applies 6 rows of staples and cuts, leaving 3 rows of staples on the hilus and 3 rows on the lobe to be removed. The instrument has an articulation that facilitates correct placement of the staple line, allowing the bronchus, pulmonary artery, and pulmonary vein to be occluded with 1 cartridge of staples. Successful lung lobectomy using staples has been reported in dogs using a stapling device designed for open surgery that applies 4 rows of staples.21 The device we used comes in 3 different lengths: 30, 45, and 60 mm, with staples sized 2.0, 2.5, 3.5, or 4.8 mm long and can accommodate all cartridge sizes. We recommend using the 60 mm cartridges with 3.5 mm long staples to perform lung lobectomy in dogs. We found that two 30 mm cartridges were required to obtain complete resection of most masses, thus using one 60 mm cartridge is more economical. In small dogs thoracoscopic lobectomy may be limited by the size of the cartridges.
Seeding of the portal sites after thoracoscopic removal of LTs is a theoretical concern in our patients. The risk of cancer recurrence at the cannula site in humans is considered low (3 of 934 operations)19; however, because seeding can be fatal, use of an endoscopic retrieval bag is the standard of care in human surgery.8 We used a retrieval bag in 2 dogs without difficulty. The bag facilitated retrieval of the lung lobe through a very small incision in 1 dog and through a cannula site without extending the incision in 1 dog. Recurrence of tumor at portal sites when a retrieval bag was used occurred in 1 of 308 humans with after a median follow-up of 25 months.22
LT removal is most commonly performed through a left or right lateral thoracotomy in dogs.3 The procedure is associated with significant postoperative morbidity and pain unless substantial postoperative analgesia is administered.23,24 Pain after lateral thoracotomy has been associated with rib spreading, transection of muscles, and neurovascular compression after closure.23,24 As thoracoscopy does not involve these events, it should be less painful. Thoracoscopic pericardiectomy in dogs has decreased morbidity and postoperative pain compared to pericardiectomy by lateral thoracotomy.12 Analgesia was not blinded nor was a pain scale used to evaluate our dogs but it was our clinical impression that dogs that had thoracoscopy only were more comfortable immediately postoperatively and had a faster recovery than dogs that had thoracotomy. Although 3 of our thoracoscopy only cases, where owners were contacted all reported excellent short-term recovery with minimal postoperative morbidity, a prospective study would be required to prove a difference in the level of comfort and recovery of dogs after thoracoscopy versus thoracotomy.
Eight dogs recovered well from the surgery, however because of the nature of a retrospective study it is possible that minor complications such as delayed healing or seroma formation were not recalled by owners or recorded in medical records. Because our purpose was to describe a surgical technique, the value of the follow-up is difficult to determine, particularly in patients with varying complex disease. The potentially faster recovery after thoracoscopic lobectomy may be an important consideration, particularly in patients with a poor long-term prognosis for survival.
Contraindications for thoracoscopic lobectomy in humans are markedly enlarged mediastinal lymphadenopathy, endobronchial tumor, chest wall or mediastinal invasion, or use of neo-adjuvant treatment.7 These contraindications would also apply to dogs. Additional contraindications would include dogs that would not tolerate 1-lung ventilation, e.g., dogs with concurrent pulmonary disease.
Metastatectomy remains controversial in human surgery, despite some indication of success for certain diseases.25–28 More research is required before definitive conclusions can be made. In light of the fact that the consequences of metastatectomy in dogs is unknown, it is our view that stable patients with single masses, hypertrophic osteopathy, or masses isolated to 1 lobe are potential candidates for metastatectomy.
Thoracoscopy may indirectly improve postoperative immune function by reducing systemic inflammatory response.29,30 Thoracoscopy has been reported to increase survival rate after resection of stage I non-small cell lung cancers in humans compared with resection via thoracotomy.31 It is thought that thoracotomy results in elaboration of more humoral factors that may induce metastatic cell growth and increase inflammatory mediators that impair anti-tumor activity.32
The goals of surgical treatment or palliation of neoplastic lung masses in dogs are similar to those in human surgery: an effective procedure, decreased postoperative morbidity, and rapid return to normal activity. Over the past decade, thoracoscopic techniques have advanced to become the gold standard for many human surgical procedures.6 The use of thoracoscopic surgical procedures are increasing in veterinary medicine with the advantage of reduced postoperative morbidity and rapid recovery.10,33–35 In our experience, TLL was a useful technique for treatment of small LTs in dogs and it is our impression that small masses positioned away from the hilus of the lung are good candidates for thoracoscopic removal.
