Video-Assisted Thoracoscopic Surgery for Pneumothorax Induced by Migration of a K-Wire to the Chest
A 2 yr old female English setter dog was admitted for acute dyspnea. The dog underwent treatment of a T9T10 thoracic vertebral fracture subluxation at the authors' institution 15 mo earlier. Upon admission, a chest X-ray revealed a pneumothorax and a metallic foreign body in the left hemithorax. An emergency video-assisted thoracoscopic surgery was successfully performed to remove a 4.6-mm long Kirschner wire that migrated from the thoracic vertebral column to the thoracic cavity. The operating time was 27 min. The dog made an uneventful recovery and was discharged on the third day after surgery. Pneumothorax should be considered in patients that develop acute dyspnea and have a history of wire fixation in the thoracic vertebral column. Video-assisted thoracoscopic surgery is a safe and effective treatment of this condition.
Introduction
Kirschner wires (K-wires) are commonly used in human and veterinary surgeries to aid in the reduction and stabilization of long bone fractures and vertebral fractures or luxations. The wires can migrate in the postoperative period and can cause significant problems. In human medicine, pericardial tamponade, arrythmias, pericarditis, pneumothorax, hemoptysis, subclavian steal syndrome, and dysphagia have been reported after pin migration within the chest.1 Recently, a case report described the migration of a K-wire to the heart in a Yorkshire terrier.2 To the authors' knowledge, there have been no reported cases of pneumothorax because of implant migration in the veterinary literature. This report describes the clinical and radiographic features of a dog with a pneumothorax because of migration of a K-wire into the thoracic cavity and its removal by video-assisted thoracic surgery (VATS).
Case Report
A 2-yr-old female English setter dog weighing 25 kg was referred from a private veterinarian to the Surgical Referral Center Clinique Vétérinaire Alliance following acute dyspnea that occurred the day before. No traumatic event was reported. The dog's medical history included a T9-T10 thoracic vertebral fracture subluxation with dorsal displacement of T10 after being hit by a car 15 mo earlier (Figure 1). The neurologic examination revealed nonambulatory paraparesis and upper motor neuron signs to the pelvic limbs. The vertebral fracture subluxation was treated using seven K-wires and polymethylmethacrylate (PMMA) at the authors' institution (Figures 2 and 3). The dog was found to be neurologically normal 2 mo after surgery. The dog was admitted 3 mo after surgical intervention as one K-wire emerged from the skin associated with a wound seroma. Vertebral thoracic radiographs at this time showed that three pins had migrated, one of which partially protruded into the left thoracic cavity. Removal of the PMMA and four of the K-wires was performed, but the three migrating pins could not be found or extracted (Figure 4). The pin protruding into the thoracic cavity could not be extracted during removal of the PMMA and needed a specific intercostal thoracotomy approach that was declined by the owners for financial reasons. The owners were aware of the potential risks of developing serious complications, such as pneumothorax or life-threatening hemorrhaging because of laceration of the great vessels within the chest.
Citation: Journal of the American Animal Hospital Association 47, 4; 10.5326/JAAHA-MS-5648
Citation: Journal of the American Animal Hospital Association 47, 4; 10.5326/JAAHA-MS-5648
Citation: Journal of the American Animal Hospital Association 47, 4; 10.5326/JAAHA-MS-5648
Citation: Journal of the American Animal Hospital Association 47, 4; 10.5326/JAAHA-MS-5648
Upon admission at the authors' institution, the dog showed mild dyspnea with a respiratory rate of 34 breaths/min; the heart rate was 130 beats/min. The mucous membranes were slightly congested with a capillary refill time of less than 2 sec. Chest auscultation revealed muffling of lung sounds dorsally in the left hemithorax. Pulse oximetry on the lip indicated an indirect hemoglobin saturation measurement (SpO2) of 93%. A chest radiograph revealed a pneumothorax and showed that a K-wire was located in the dorsal part of the left eighth intercostal space (Figure 5). A retrospective examination of chest radiographs obtained 12 mo before presentation showed no migration of the two other K-wires that were not extracted. The suspected diagnosis was that the K-wire migrated from the thoracic vertebral column to the left chest and was complicated by a pneumothorax. Removal of the migrating K-wire by VATS was planned with the owners with the potential need to convert to open thoracotomy.
Citation: Journal of the American Animal Hospital Association 47, 4; 10.5326/JAAHA-MS-5648
The dog was initially stabilized with intravenous fluids and thoracocentesis in the dorsal third of the seventh left intercostal space. Three hundred cc of air were removed. Cephalexina (22 mg/kg) and buprenorphineb (0.01 mg/kg) were administered, and the dog received oxygen (100 mL/kg) through a nasal tube pending surgical treatment. The respiratory rate decreased to 24 breaths/minute and the SpO2 normalized to 99%. Surgery was performed 2 hours later. The dog had a normal presurgical complete blood count and serum biochemistry profile. Thoracocentesis was repeated before the induction of general anesthesia and 150 cc of air was removed. Anesthesia was induced intravenously with 0.2 mg/kg diazepamc and 10 mg/kg thiopentald. Anesthesia was maintained in 100% oxygen with two-lung ventilation and positive end expiratory pressure (PEEP) in a semi-closed circle system. PEEP was maintained at 4 cm H20 during the VATS procedure and acceptable SpO2 measurements (>98%) were obtained. Pain relief was addressed subcutaneously with 0.2 mg/kg morphine chlorhydratee at the induction of anesthesia. Anesthetic monitoring included heart and respiratory rates, oxygen saturation, capnography, electrocardiography, and systolic and diastolic blood pressures.
The entire left thoracic wall from the prescapular area to the second lumbar vertebra was clipped and an initial aseptic skin preparation was performed. The dog was placed in a right lateral recumbent position and the left hemithorax was aseptically prepared and draped. An intercostal nerve block was performed with 0.5 mg/kg lidocainef from the seventh to the ninth intercostal spaces. Two thoracoscopic portals were established for the VATS procedure using triangulation to observe and provide access to the migrating K-wire. A 1-cm stab incision was made in the ventral aspect of the eighth intercostal space near the costochondral junction. Tissues were bluntly dissected with a Kelly forceps, and after a pneumothorax, a 5-mm trocar/threaded cannula assembly was inserted into the thoracic cavity. A 2.7-mm outer diameter 30° rigid endoscope with a video camerag was passed through the cannula to view and inspect the left thoracic cavity. The pin was located at the eighth intercostal space. One end of the pin had pierced the intercostal musculature and was inserted approximately at a 30° angle relative to the thoracic wall (Figure 6). The authors searched carefully for any lung parenchyma injury, but did not see any lesions that were thought to be foreign body lacerations.
Citation: Journal of the American Animal Hospital Association 47, 4; 10.5326/JAAHA-MS-5648
The telescope was used to facilitate the placement of the second portal. A 1-cm incision was made in the left ninth intercostal space at middle distance between the costovertebral junction and the sternum, and an instrument portal was established using a 5-mm trocar/assembly. Straight grasping forceps were placed through this caudal portal access and were used to grasp the wire (Figure 7). After several tries, the pin was dislodged from the intercostal muscles and dropped over the caudal lung lobe. The pin was grasped with caution by its nonpointed end with the axis of the grasping forceps parallel to the axis of the pin. Finally, the pin was brought out of the port site without removing the instrument cannula (Figure 8). Intermittent cessation of ventilation throughout the procedure was required to improve observation. Dislodging the wire induced a slight hemorrhage due to damage to the intercostal vein that was resolved spontaneously. A 10-French chest tube with a trocarh was inserted in a routine manner under thoracoscopic guidance. A small skin incision in the dorsal third of the left thoracic wall was made at the level of the 11th intercostal space. A tunnel under the latissismus dorsi muscle was created in a cranioventral direction over three intercostal spaces, avoiding cannula sites, and the tube was introduced into the seventh intercostal space in the direction of the opposite shoulder. The tube was sutured directly to the skin in a Chinese finger-cuff suture pattern.
Citation: Journal of the American Animal Hospital Association 47, 4; 10.5326/JAAHA-MS-5648
Citation: Journal of the American Animal Hospital Association 47, 4; 10.5326/JAAHA-MS-5648
An inspection of the left thoracic cavity at the end of the procedure did not reveal any abnormalities. The cannulas were removed, and the thoracoscopic sites were closed in two layers: the subcutis layer with 3-0 polyglyconate in a cruciate pattern, and the skin with 4-0 nylon in a simple interrupted pattern. The chest was evacuated for any residual air at the time of closure via a thoracostomy tube. The chest tube was secured in place relative to the thoracic wall by incorporating it entirely into a bandage. Its position was confirmed with a thoracic radiograph. An extension set was added and sealed with a three-way stopcock. The three-way stopcock was connected to a closed collection system with a bottle functioning as a collection reservoiri. This system allowed for continuous pleural drainage, but needed to be checked regularly to determine the amount of negative suction applied into the bottle via a suction control indicator. The total surgical time was 27 min. Ten minutes after the end of anesthesia, the dog began to breathe spontaneously. After 30 min, the endotracheal tube was removed.
The dog made an uneventful recovery. Morphine was administered postoperatively at 0.1 mg/kg q every 4 hr for 24 hr. The dog received 22 mg/kg cephalexin q every 8 hr for 24 hr, and 0.1 mg/kg meloxicamj subcutaneously q every 24 hr for 7 days after a loading dose of 0.2 mg/kg the first day. The chest tube was checked every hour the first 4 hr and every 4 hr thereafter. The volume of air removed from the chest tube decreased gradually from 150 cc the first hour to 100 cc at 8 hr, and 50 cc at 16 hr after surgery. After 24 hr, the dog had normal clinical parameters, a thoracic auscultation revealed normal lung sounds, and no air was aspirated from the chest tube. The thoracostomy tube was monitored every 4 hr for a further 24 hr. Two days after surgery, no air was aspirated from the thoracostomy tube and chest radiographs confirmed resolution of the pneumothorax. The thoracostomy tube was subsequently removed. The incisions were inspected daily, and any postoperative complications were recorded. The dog was closely observed for a further 24 hr after chest tube removal for signs of respiratory distress, and was then discharged from the hospital on the third day after the VATS procedure. At follow-up examinations at 2, 4, and 8 weeks, and 9 mo, there were no clinical signs of respiratory disease and there was no radiographic evidence of pneumothorax. No migration of the two remaining K-wires was observed. The owners were informed that these pins could also potentially migrate in the future.
Discussion
Migration of K-wires within the chest is uncommon, but is a known complication in human surgeries, particularly when fixation wires are used around the shoulder.1,3,4 The migration of a titanium rod used to stabilize a comminuted fracture of the first lumbar vertebra was also described recently in a male human patient.5 In the dog, only one case report describing the migration of a K-wire to the chest after femoral osteosynthesis has been reported. In that case, the K-wire migrated to the heart, causing acute pulmonary edema and death two days after its surgical removal by thoracotomy.2 Therefore, the migration of orthopedic fixation devices into the thoracic cavity requires specific consideration.
The thoracic vertebral column is usually considered an immobile section of the spine because of inherent stability provided by the ribs, intercostal muscle attachments, ligamentous support, and the well-developed epaxial muscle mass.6 It is, therefore, unlikely that the K-wire migration observed in the present study occurred due to repetitive movement of the thoracic vertebral column. In addition, there were no features on the postoperative radiographs that suggested that this pin might be susceptible to loosening. In the authors' opinion, the surgical technique used and the inherent limitations of the smooth pins used were factors that contributed to the migration of the pins. In the thoracic vertebrae, it is recommended that the K-wires be placed in the pedicles and driven into the vertebral bodies for maximal bone purchase.6 In the present case, a modified technique was used where each pin was placed in a ventrolateral direction from the base of the spinous processes, crossing the articular facets and terminating in the tubercles of the ribs. This technique is commonly used in the authors' hospital for thoracic vertebral fracture subluxations and presents two major advantages: each pin engages six cortices, and the more horizontal pin position allows the pin to avoid entering the pleural cavity or injuring the aorta. Nevertheless, the authors hypothesized that the surface area of the implant that was in contact with bone with this technique was reduced compared with the conventional technique (data not shown). This modified technique might decrease the strength and rigidity of the fixation device, subsequently causing the wires to migrate. Furthermore, in this case, the authors did not use positive threaded pins, and the exposed part of the wires were not bent. Bending the wire and the use of positive threaded pins have been proposed to prevent migration.7,8 It is commonly accepted that the bone purchase of smooth pins (and the pull out strength) is not as effective as positive threaded pins.9 However, these methods do not completely eliminate the risk of migration, which has also been reported in cases with threaded pins and even after fixation of the pin with bone cement.10,11
It is of academic interest to speculate on the course taken by the K-wire as it passed from the site of insertion to the site of removal. The wire usually follows a retrograde path, protruding near their entry point. Occasionally, however, the wire can migrate in another direction. The most likely path of wire migration from the spinal column to the intrapleural space or to the lung parenchyma is per continuitatem. The authors presumed that the wire reached the chest cavity through the epaxial musculature and terminated in the intercostal muscles. Respiratory movements, negative intrathoracic pressure associated with respiration, gravitational forces, muscular activity, and regional resorption of bone were reported as factors contributing to pin migration in humans treated for shoulder fractures.1,4,7 The authors suspect that these factors also played a role in the migration of the wire to the chest in the present case.
Pneumothorax has many causes. Chronic pneumonia, pulmonary abscess, pulmonary adenocarcinoma, infestation by Dirofilaria immitis, and rupture of pulmonary blebs or bullas have been reported as potential etiologies.12–14 In the present case, there was no clinical or radiographic evidence of pulmonary disease. Furthermore, even if the authors did not examine both hemithoraces during VATS, there was no identification of any blebs or bullas on the left lung parenchyma. Because the pneumothorax resolved completely after the removal of the wire and the placement of the thoracostomy tube, the authors strongly suspect that the pneumothorax was related to the presence of the wire. The pointed end of the wire was found to point in the direction of the lung parenchyma during the VATS. Consequently, the authors hypothesized that lung movement during respiration allowed for laceration of the lung parenchyma and visceral pleura, subsequently causing a pneumothorax. With the removal of the pin and the continuous suction applied to the pleural cavity for 48 hr after surgery, the pulmonary laceration has probably healed. Continuous suction has the advantage of keeping pleural surfaces in contact, which can aid sealing of pleuro-pulmonary fistulas. One other advantage of the device used in the present case was that it permitted measurements of air volume removed from the pleural cavity.
VATS has been used to remove migrating wires from human patients.15,16 Thoracoscopic removal of a wire can be accomplished safely if the wire can be withdrawn through a portal site, if it does not traverse the mediastinum, and if the patient can tolerate one-lung ventilation.15 In this dog, the pin was small (2 mm in diameter), one end of the pin was inserted in the intercostal musculature, and no adhesions of the pulmonary parenchyma to the thoracic wall were noted, so removal was successfully achieved without recourse to open thoracotomy. In addition, having an instrumental portal that was located caudally and was in close proximity to the K-wire probably facilitated the surgical procedure by allowing easy direct manipulation. No complications were noted as a direct result of the VATS procedure. The dog recovered rapidly and required only short-term postoperative analgesic drugs. However, the VATS procedure might be difficult if there was inadequate visualization of the pin, extensive bleeding or laceration of the lung parenchyma, if the pin dropped into the thoracic cavity and the surgeon was not able to locate it, or if there were significant adhesions of the pulmonary parenchyma to the thoracic wall. These features must be considered preoperatively, and the owners must be informed of the potential need to convert to open thoracotomy.
The authors believed that a VATS approach might allow for a more rapid recovery with earlier ambulation in comparison with an intercostal thoracotomy. The advantages of VATS over thoracotomy have been well documented in human patients and include improvements in detailed observations, decreased postoperative complications, and a more rapid return to normal function.17–19 There have been limited objective comparisons of “open” versus VATS procedures in veterinary patients, but similar advantages are likely in dogs and cats.20,21 A little cut-down over the pin location with assistance of fluoroscopy would have been an interesting alternative minimally invasive approach in the present case. However, a fluoroscopy unit was not available in the authors' hospital. In addition, no inspection of the thoracic cavity and lung parenchyma could be accomplished with this technique compared with the VATS procedure.
Although one-lung ventilation was not induced during the VATS procedure, the authors succeeded in safely and quickly extracting the wire using an intermittent cessation of ventilation. Nevertheless, selective ventilation of the right lung (“one-lung ventilation”) would have provided improved visualization during the VATS procedure. In several clinical studies, deflation of the surrounding lobes reduced the surgical time because the visualization of the operative field was improved and the risk of iatrogenic trauma to the lung lobes was substantially reduced.22,23 Application of PEEP during the VATS procedure increased airway pressure during exhalation to exceed atmospheric pressure and prevented exhalation of the entire tidal volume, helping to prevent atelectasis and keeping more alveoli open for effective gas exchange.24,25 In this dog, PEEP of 4 cm H20 was used. This level of PEEP allowed the authors to maintain acceptable SpO2 measurements (>98%) throughout the procedure without obstructing the surgical manipulations. Higher levels of PEEP were not considered because PEEP has negative effects on hemodynamic variables, including reduced cardiac output and increased peak inspiratory pressure.26,27
Conclusion
Although this dog underwent a successful removal of wire and recovered uneventfully, this case report suggested that the migration of K-wires could cause potentially fatal complications. Clinicians caring for patients with orthopedic K-wires inserted previously for thoracic vertebral stabilization should be aware of the potential for wire migration to the chest. Regular radiographic follow-ups are mandatory, and the wire must be removed immediately when wire migration is observed to prevent dangerous complications. If a patient developed acute dyspnea and has a history of wire fixation in the thoracic vertebral column, the potential for pneumothorax should be considered. VATS is safe, feasible, and effective, and requires minimal surgical trauma for the extraction of wires that have migrated to the chest.
Laterolateral view of the thoracic vertebral column before surgery. A fracture subluxation of T9-T10 with dorsal displacement of T10 was initially present on a plain radiograph.
Intraoperative images (the head is above on the picture). A: Placement of the pins (n = 7) into the thoracic vertebral column after reduction of the T9-T10 vertebral fracture subluxation. Each pin was placed in a ventrolateral direction from the base of the spinous processes, crossing the articular facets and terminating in the tubercles of the ribs. B: After placement of the polymethylmethacrylate (PMMA).
Laterolateral (A) and dorsoventral (B) views of the thoracic vertebral column after vertebral stabilization. A stabilization of T9-T10 by means of Kirschner wires (K-wire) and PMMA was completed with good realignment of the affected vertebrae.
Laterolateral (A) and dorsoventral (B) views of the thoracic vertebral column 3 mo following vertebral stabilization and after removal of the PMMA and four of the K-wires. Three pins could not be found or extracted. One pin protruded partially into the left thoracic cavity.
Chest radiographs shown in laterolateral (A) and dorsoventral (B) views upon admission 12 mo after removal of the PMMA and four of the K-wires. A poor positioning quality on the dorsoventral view was obtained, and no further radiography was performed to minimize stress of the patient. A pneumothorax was diagnosed and a wire was observed in the dorsal part of the left eighth intercostal space. The presence of free pleural air allowed the heart to shift in a recumbent lateral view, giving the heart the appearance of being elevated from the sternum.
Video-assisted thoracic surgery (VATS) showed the K-wire on the intercostal musculature at the eighth intercostal space. One end of the pin had pierced the intercostal musculature and was inserted approximately at a 30° angle relative to the thoracic wall.
Video-assisted thoracic procedure for the extraction of the K-wire via a 5-mm port.
The K-wire removed from the interscostal space was 4.6 mm long and 2 mm in diameter.
Contributor Notes
J-G. Grand's present affiliation is Surgical Referral Center Clinique Vétérinaire de la Plage, Marseille, France.
S. Bureau's present affiliation is Surgical Referral Center Clinique Vétérinaire Alliance, Bordeaux, France.
