Intraluminal tracheal stent fracture in a Yorkshire terrier

Abstract

An 8-year-old Yorkshire terrier was presented with tracheal collapse. Two intraluminal nitinol stents were implanted. The implanted stents were found to be fractured 4 weeks after implantation. The fractured stents were removed. To restore the collapsed trachea, ring prostheses were applied. However, the dog was euthanized because of a bad outcome following surgery.

Tracheal collapse (TC) is characterized by flattened C-shaped cartilaginous tracheal rings and the development of a loose redundant dorsal tracheal membrane, with subsequent tracheal narrowing and obstruction. Several treatments for TC have been well documented elsewhere (1–3). Described surgical corrections include plication of the dorsal tracheal membrane, tracheal ring chondrotomy, and intra- and extraluminal stabilization with polypropylene rings (1–3). Generally, only extraluminal polypropylene stabilization gives a favorable outcome. Despite the known success of extraluminal stabilization, the method requires open surgery that may lead to additional complications, such as persistent coughing, iatrogenic laryngeal paralysis, and dyspnea.

Intraluminal prosthetic restoration by using a self-expanding nitinol (nickel-titanium alloy) stent has been used recently to stabilize the collapsed trachea (4). This prosthesis was originally designed to either maintain the tracheal lumen from tumors that compressed the trachea, or expand collapsed arteries in humans. Because nitinol is flexible and elastic and has physical properties similar to tracheal cartilage, most self-expanding stents are made of nitinol. After the deployment of the stent inside the tracheal lumen, the stent gradually adapts to size of the lumen (5). The advantages of this procedure compared with surgical stabilization are that it is noninvasive, does not require intensive care after implantation, and takes only 5 to 10 min, depending on the skill of the practitioner. However, this stent has been associated with several complications (transient coughing, laryngeal spasm, perforation of the tracheal mucosa, stent fracture) due to progressive shortening of the stent (4).

Several other types of intraluminal stent have been developed for management of airway collapse in humans. The Palmaz-Schatz stent is a rigid type of stent that is not associated with shortening of the stent (5). However, this type of stent is associated with other complications (pneumothorax, infection, tracheal obstruction by mucous plugs, stent migration, compression of the upper stent ends) and is not commonly recommended for use in canine tracheal collapse (5). Thermal shape memory nickel titanium (NiTi) alloys have also been applied in the treatment of tracheal collapse. Although it has good compatibility, it is not commonly used in the veterinary field, because the implantation procedure is technically complex. Recently, bioabsorbable airway stents (poly-L-lactic acid stents) have been developed and used experimentally in dogs.

Case description

An 8-year-old, castrated male, Yorkshire terrier, weighing 4.07 kg, was referred from a local practitioner because of persistent coughing and dyspnea. At presentation, the dog was cyanotic with a persistent goose-honking cough, especially after excitement. The dog had mild polycythemia (8.2 × 10¹²/L; reference range, 5.5 to 8.2 × 10¹²/L), mild neutropenia (2.93 × 10⁹/L; reference range, 3.0 to 11.8 × 10⁹/L), lymphocytosis (5.82 × 10⁹/L; reference range, 1.0 to 4.8 × 10⁹/L), and eosinophillia (3.09 × 10⁹/L; reference range, 0.1 to 1.3 × 10⁹/L). Blood gas analysis showed the presence of mild hypoxia (P_aO₂: 72 mmHg; reference range, 80 to 104 mmHg) and mildly elevated hepatic enzymes, as determined from the results of blood chemical analysis (alanine transaminase 139 IU/L; reference range, 3 to 100 IU/L; alkaline phosphatase 514 IU/L, reference range, 20 to 300 IU/L). The radiographic and endoscopic examinations revealed that the trachea had collapsed from the levels of the caudal aspect of the 5th cervical to the 2nd thoracic vertebra (Figure 1A).

Figure 1.

Lateral radiograph A: The trachea was severely collapsed from the caudal aspect of the 5th cervical to the 2nd thoracic vertebra. B: Radiograph taken at the 4-week reevaluation. Radiograph revealed the fracture of stent at the junction where 2 stents faced each other.

Since the tracheal collapse was extremely severe, intraluminal nitinol stents were applied to reconstruct the collapsed trachea. The lateral and dorsoventral views of the thoracic radiographs were used to determine the length and diameter of the trachea, using a radiographic ruler based on the method described previously (4). A 10-mm (diameter) × 100 mm (length) self-expanding intraluminal nitinol stent (Zilver 635 biliary stents; Cook, Bloomington, Indiana, USA) was implanted inside the collapsed trachea with the dog under heavy sedation following the administration of atropine (Daewoo pharmaceuticals, Pusan, Korea), 0.02 mg/kg body weight (BW), diazepam (Daewon pharmaceuticals), 0.2 mg/kg BW, and propofol (Dongkwang Pharmaceuticals, Seoul, Korea), 5 mg/kg BW, IV. After recovery from the sedation, the dog showed marked respiratory improvement. However, radiography revealed that the implanted nitinol stent was not covering the proximal end of the collapsed trachea (Figure 2A). Therefore, a 2nd nitinol stent was implanted in the trachea to extend from the level of the 3rd cervical vertebra to the proximal end of the 1st stent (Figure 2B). No clinical signs associated with tracheal collapse were observed after the 2 nitinol stent implantations, except for occasional coughing. The dog was discharged with instructions to the owner to administer the cough suppressants methylephedrine (Codewon; Daewon Pharmaceuticals), 1 mg/kg BW, PO, q12h, prednisolone (Solondo; Yuhan Pharmaceuticals, Seoul, Korea), 1 mg/kg BW, PO, q12h, and cefazoline (Cephradine; Myungmoon Pharmaceuticals, Seoul, Korea), 10 mg/kg BW, PO, q12h for 2 wk.

Figure 2.

Lateral radiograph. A: Radiograph taken after the 1st stent was deployed showing that the implanted stent did not cover the proximal end of collapsed trachea. B: Radiograph taken after the 2nd stent was deployed showing that the 2 implanted stents covered the entire length of the trachea, including the collapsed area of trachea.

At the 2-week reevaluation, the owner reported noticing marked improvement in the dog’s breathing and that the dog coughed infrequently during periods of excitement. Results from a CBC and blood biochemical panel showed that values had returned to normal. Cough suppressants and prednisolone, reduced to 0.5 mg/kg BW, q12h, were prescribed for a further 2 wk.

At the 4-week reevaluation, the owner reported persistent coughing but no respiratory difficulty. Radiography revealed a stent fracture at the junction between the intraluminal stents (Figure 1B). A hematological examination revealed mild leukocytosis (21.82 × 10⁹/L; reference range, 6.0 to 17.0 × 10⁹/L), otherwise the general condition of the dog was good. However, the dog was admitted for further investigation. During the 1st night of hospitalization, the dog’s condition worsened, with severe respiratory difficulty.

Since the dog’s condition continued to deteriorate, the fractured stents were surgically removed and replaced with prosthetic rings. First, an intratracheal tube (4 Fr) was inserted via the mouth for inhalation anesthesia. After resecting the affected tracheal ring, where both ends of the fractured stents were located, a 2nd intratracheal tube (6 Fr) was inserted via the incised tracheal ring. The fractured stents were easily removed and replaced with 5 prosthetic rings, placed individually around the tracheal rings from the 5th cervical to the 2nd thoracic vertebra. After surgery, the dog was medicated with cough suppressant methylephedrine, 1 mg/kg BW, PO, q12h, butorphanol (Butophan; Myungmoon, Myungmoon pharmaceuticals), 0.3 mg/kg BW, SC, q12h, prednisolone, 0.5 mg/kg BW, IM, q12h, and gentamicin (Samyang pharmaceuticals, Seoul, Korea), 5 mg/kg BW, IM, q12h, with saline nebulization. After the dog recovered from anesthesia, there were no signs associated with airway obstruction and the frequency of coughing was markedly reduced. However, 2 d after surgery, the dog suddenly suffered inspiratory difficulty. Thoracic radiographs revealed progressive collapse of the intrathoracic trachea and pneumothorax due to the rupture of the sutures where the trachea had been incised. With the owner’s consent, the dog was euthanized.

Discussion

The etiology of TC is unknown. Most TC cases are the acquired type, which usually occurs in middle-aged to aged dogs, although congenital cases have been reported in young dogs (6). A recent study found that deficiency of glycoprotein and glycosaminoglycan leads to progressive chondromalacia in trachea and bronchi and is associated with TC (6).

Tracheal collapse (TC) has been managed by medical therapeutics, surgical correction, and intra- and extra-luminal prostheses, depending on the degree of the collapse (1–3). Medical management in association with a weight loss program has generally been used to control moderate cases of tracheal collapse. One retrospective study reported that 71% (71/100) of TC cases responded to medical management and loss of weight (7). Although an extraluminal prosthesis can be applied to cervical TC, it can not be applied to the intrathoracic TC, which occurs more commonly in dogs. Recent studies found that intraluminal stents have advantages over extraluminal prosthesis. However, long-term medical treatment is still required, because bronchial collapse can not be fixed by intraluminal stenting and metaplasia of the tracheal mucosa with mucociliary clearance may not be reversed by intraluminal stenting. Because all forms of the stenting are palliative, affected dogs should be managed medically first. The intraluminal stenting may be applied, if medical treatment alone does not manage signs of TC.

Most unsuccessful stent implantations result from underestimating the diameter of the stent required. In the literature, the stent size is usually estimated from radiographs obtained at inspiration and expiration (8). Since these measurements can be inaccurate (8), it is generally recommended that the following calculation be used for determining the size of the stent: ([Mean tracheal diameter in lateral view + mean tracheal diameter in dorsoventral view] × 2)/3.14 (4). Recently, investigators have shown that the selection of a stent having a diameter 3 to 5 mm larger than the estimated tracheal diameter might be beneficial. This approach results in stretching the dorsal tracheal membrane, thus helping to hold the stent firmly in its position (8). In our case, the estimated diameter of the collapsed trachea was 8.1 mm on the radiograph. Therefore a 10 mm diameter stent was selected as being appropriate to open the collapsed trachea. However, in the replacement surgery, we found that the actual diameter of the collapsed trachea was 2–3 mm larger than that of unaffected trachea. Since the estimation from the radiograph was based on the unaffected part of the trachea, we did not realize that we had underestimated the diameter of the trachea. Unless a better method is devised to measure the actual affected area of the trachea, underestimation will continue to be a major problem. In humans, computed topography (CT) and bronchoscopy are used to determine the diameter and length of the tracheal lesion (8). However, CT is not generally recommended initially, because of serious concerns related to the general anesthesia required for CT and bronchoscopy (8). Radiographic measurement under positive pressure ventilation has been successfully applied and the method is available (9).

Intraluminal stabilization of the trachea should be extended about 1–2 cm beyond either end of the affected area (personal communication, Dr. Flander). However, as a result of recent studies, it is recommended that the intraluminal stabilization of the trachea be performed along its entire length (7), because the problem in stenting only a small section is that, often, a secondary collapse in the region of the proximal and distal ends of the stent occurs (8). Since shortening of the intraluminal prosthesis, in addition to the increased diameter of the stent, is commonly associated with a self-expanding intraluminal prosthesis (4), it is more appropriate to cover the entire trachea.

In our case, we planned to implant a stent that would extend from 1 cm cranial to the proximal end of the affected trachea to the carina to prevent a 2nd collapse of the trachea at the carina by using a 10-cm stent, which was the longest stent available in Korea. However, this stent did not cover the proximal end of the affected trachea, as required. Despite the circumstances, the dog initially showed marked clinical improvement. The 2nd stent (4-cm long with the same diameter) was implanted to support the uncovered area of the trachea. For the first 4 wk postimplantation, the dog maintained normal respiration with an occasional cough (but no goose-honking cough). However, both stents were later found to have fractured at the junction, located at the thoracic inlet, a highly mobile area, but there was no evidence of stent migration from the radiographic analysis. Therefore, the high mobility of the area where the 2 stents joined ultimately resulted in cyclical fatigue and failure of the metal stent.

An interesting finding arising from this case is that the original conformation of the stent can be progressively distorted, if any part of the stent is broken. The stent implanted in our study was originally designed to restore patency of a large artery or bile duct. Both ends of the stent have gold plated anchors that prevent the migration of the stent and help to maintain its conformation after it expands. Therefore, if the anchors are severely damaged, the whole stent can be progressively distorted and eventually lose stability. As a result, this type of stent is inappropriate for reconstructing a collapsed trachea where the anchor is located at the thoracic inlet (a highly mobile area). Additionally, the anchor should not be located in the severely redundant area of the trachea, since this may lead to stent instability.

In conclusion, from our experience pertaining to this case, the success of intraluminal stenting is heavily dependent on the correct selection (especially diameter and length) of the stent. Although recent studies have proposed improved methods of stent selection, more studies are still required to optimize the methodology.