Materials and Methods

Criteria for selection of cases—Medical records of dogs and cats evaluated because of NPS between 2005 and 2007 were evaluated. Animals were included in the present study if CT and rhinoscopy had been performed, a diagnosis of benign NPS had been made, and a BEMS was placed.

Medical records review—Signalment, history, physical examination findings, previous treatments, diagnostic imaging features on CT scan and rhinoscopy, balloon and stent type and size, and procedure time were extracted from the medical records of cases included in the study. Cases included 3 dogs (animals 1, 2, and 3) and 3 cats (animals 4, 5, and 6).

Imaging—Computed tomography was performed with a third-generation helical CT scanner^a prior to rhinoscopy. A scan was performed in 1-mm slices at 2-mm intervals from the nares to the proximal portion of the trachea before and after IV administration of iohexol^b (350 mg of iodine/mL; 2.2 mL/kg [1 mL/lb]). On axial images the stenotic lesion was identified, and orthogonal measurements were obtained of the diameter of the nasopharynx directly rostral and caudal to the stenosis. Sagittal images were used to determine the length of the stenosis (Figure 1). An appropriate stent size was chosen via minimally oversizing the measured height of the nasopharynx by 1 to 2 mm, with stent length determined by measurement beyond the extent of the stenosis via axial and sagittal CT images.

View Full Size

Computed tomographic images of the head of a cat with NPS. White boxes contain values of various measurements as indicated. Large arrows indicate locations of transverse views. A—Transverse view just caudal to the NPS. B—Transverse view at the NPS. C—Transverse image just rostral to the NPS. D—Sagittal view (rostral is to the right) of the NPS located just caudal to the junction of the hard (HP) and soft palate (SP). NP = Nasopharynx.

Citation: Journal of the American Veterinary Medical Association 233, 9; 10.2460/javma.233.9.1432

• Download Figure
• Download figure as PowerPoint slide

Stenting—Each animal was positioned in lateral recumbency. A 6-mm flexible endoscope^b was placed in a retroflexed manner transorally into the caudal aspect of the nasopharynx, and the NPS was identified (Figure 2). Under fluoroscopic guidance,^c a 0.035-inch hydrophilic angle-tipped guidewire^d was advanced through the naris into the ventral nasal meatus and directed caudally to the stenosis (Figure 3). For animals with an existing stenotic orifice (animals 1, 4, 5, and 6), the guidewire was directed through the small opening and into the esophagus. For animals in which no orifice was visible and a complete membrane was present (animals 2 and 3), a 4-F access sheath^e was advanced over a 0.035-inch angle-tipped guidewire in an antegrade fashion through the ventral nasal meatus via fluoroscopic guidance (Figure 4). The NPS was viewed via retroflexed rhinoscopy, and a guidewire was advanced through the endoscope to determine the caudal aspect of the stenotic lesion (Figure 5). An 18-gauge, 15-cm access needle^f was advanced through the access sheath and pierced the membrane seen rhinoscopically and fluoroscopically. The needle was directed in a dorsomedial direction. The stylette was removed, and the 0.035-inch hydrophilic guidewire was advanced through the needle and down the esophagus. The spinal needle and sheath were removed, and a 5-F marker catheter^g was advanced over the wire for further dilatation. Subsequently, an 8-F vascular dilator^h was used to open the orifice large enough to accept the balloon delivery system. The distal end of the stenosis was appreciated rhinoscopically and its location confirmed fluoroscopically. In all cases, a PTA balloonⁱ was advanced over the guidewire. The diameter of the balloon was approximately 50% to 60% of the nasopharyngeal diameter at the stenotic lesion.

View Full Size

Endoscopic view of BEMS placement in a cat with an NPS. A—Retroflexed rhinoscopic image of the NPS. B—A guidewire is placed in a normograde fashion through the NPS. C—A marker catheter is advanced over the guidewire and down the esophagus. D—A PTA balloon is inflated at the NPS for minimal dilatation and (E) ease of BEMS placement. F—Retroflexed rhinoscopic view of the stenotic area after BEMS placement.

Citation: Journal of the American Veterinary Medical Association 233, 9; 10.2460/javma.233.9.1432

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Fluoroscopic images of BEMS placement in a cat (left lateral recumbency; rostral is to the left) with NPS. A—A guidewire (*) is passed through the NPS with a PTA balloon (the 2 black marks are the rostral and caudal aspects of the balloon) over the wire centered at the stenotic area; the arrow points to the endoscope. B—After partial balloon dilatation, the BEMS is mounted over a PTA balloon, advanced over the guidewire, and centered at the NPS (arrow). C—By filling of the balloon with 50% contrast medium, the balloon is inflated under fluoroscopic guidance (arrow points to the waist of the stricture). D—The waist of the stricture is defaced, and the balloon fully expands the stent. E—The balloon is deflated, and the stent remains in place at the stenosis location (within indicated oval).

Citation: Journal of the American Veterinary Medical Association 233, 9; 10.2460/javma.233.9.1432

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Fluoroscopic images of a dog with a closed nonpatent NPS during stent placement (rostral is to the right). A—A 6-mm endoscope is retroflexed over the caudal end of the soft palate, and a guidewire is advanced through the working channel of the endoscope and positioned at the stenotic membrane. A second guidewire is advanced through the naris in an antegrade fashion to the level of the NPS, documenting the length of the stenosis as the distance between the 2 guidewires. B—A 15-cm renal access needle (asterisk) is used to puncture the stenotic membrane via fluoroscopic and endoscopic guidance. C—When the needle can be confirmed as being through the membrane via endoscopy, the stylette is removed, and a guidewire is passed through the needle down the esophagus. D—A BEMS (indicated by an oval) is advanced over the wire through the created orifice in the membrane covering the length of the NPS. E—The balloon is inflated with 50% contrast medium to break the stenotic lesion. F—The balloon is deflated and removed over the wire with the expanded stent left in place (oval). A bony narrowing at the junction of the hard and soft palate is evident (arrowhead), which does not allow the stent to fully expand in this area. The remainder of the stenosis is fully defaced.

Citation: Journal of the American Veterinary Medical Association 233, 9; 10.2460/javma.233.9.1432

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Retroflexed endoscopic images of a dog with a membrane that completely covers an NPS. A—An access needle pierces the closed membrane with the guidewire (black star) that was placed through the endoscope to document the caudal aspect of the stenosis fluoroscopically. B—A marker catheter is advanced over the guidewire for dilatation. C—A PTA balloon is used to dilate the stenosis prior to stent placement. D—The stenosis is opened by the stent; notice the guidewire passing through the lumen of the stent.

Citation: Journal of the American Veterinary Medical Association 233, 9; 10.2460/javma.233.9.1432

• Download Figure
• Download figure as PowerPoint slide

The balloon was inflated with 50% iohexol^j and 50% saline (0.9% NaCl) solution via fluoroscopic guidance. The waist of the stenosis was viewed (Figure 3), and the balloon was deflated and removed over the wire. The premounted BEMS^k was advanced over the wire, centered across the stenosis, and inflated via fluoroscopic and rhinoscopic guidance until the stenosis was effaced by use of the associated balloon. Care was taken not to exceed the rated burst pressure of the balloon. The balloon was deflated and removed over the wire, leaving the expanded stent in place across the now expanded lesion. A catheter was placed over the wire, the wire was removed, and a nasal flush was performed with sterile saline solution before a local anesthetic (bupivicaine, 1 mg/kg [0.45 mg/lb], for dogs; 0.2 mg/kg [0.9 mg/lb], for cats) was injected at the level of the stenosis.

Postoperative care—For the first 12 hours, buprenorphine (0.008 to 0.01 mg/kg [0.004 to 0.005 mg/lb], IV) was given as needed. The animals were discharged with an antimicrobial (amoxicillin-clavulanic acid, 13 to 18 mg/kg [5.9 to 8.2 mg/lb], PO, q 12 h, for 7 to 14 days) and a 4- to 6-week tapering dose of corticosteroid (prednisone, 0.5 mg/kg [0.23 mg/lb], PO, q 12 h, initially).

Results

Signalment and history—Three dogs and 3 cats with NPS met the criteria for inclusion in the study. Animals 1, 2, 4, and 5 were evaluated at the Matthew J. Ryan Veterinary Hospital of the University of Pennsylvania, animal 3 at the Advanced Critical Care and Internal Medicine Referral Hospital in Tustin, Calif, and animal 6 at the Purdue University Veterinary Teaching Hospital. Animals 1, 2, and 3 were dogs (a Toy Poodle, a mixed-breed dog, and a Beagle, respectively) ranging in age from 7 months to 16.5 years and in weight from 6.5 to 18 kg (14.3 to 39.6 lb). Animals 4, 5, and 6 were cats (a Domestic longhair, a Tonkinese, and a Domestic shorthair, respectively) ranging in age from 6 to 11 years and in weight from 3.5 to 7.7 kg (7.7 to 16.9 lb).

History and clinical signs—Inspiratory stertor and severe nasal discharge (mucoid to mucopurulent) were present in all animals for a median of 9 months (ranging from 6 weeks to 10 years) prior to evaluation. Animal 1 had a ventral rhinotomy performed for removal of a nasal adenoma. The resulting palate defect required multiple corrective surgeries for complete repair. Animal 2 had a tentative diagnosis of canine distemper as a pup and since that time had been unable to pass air through the nose and had intermittent severe mucopurulent nasal discharge that was moderately responsive to antimicrobial treatment and nasal flushing. Rhinoscopy performed by the referring university hospital revealed a complete membrane without an orifice. A hole had been created through the thick membranous tissue by use of rhinoscopic guidance and a holmium:yttrium-aluminum-garnet laser, which alleviated the clinical signs for a few days before they recurred and the dog was referred. Animal 3 had an abrupt onset of inspiratory stertor attributed to suspected aspiration rhinitis after ovariohysterectomy. At the time of anesthetic recovery, gastric contents were found coming out of the nares, and regurgitation during anesthesia was suspected as the cause, which has been reported.⁸ All 3 cats had a previous diagnosis of chronic lymphoplasmacytic rhinitis made on the basis of nasal mucosal biopsy specimens. All cats had chronic copious mucoid to mucopurulent nasal discharge. Retroflex rhinoscopy was performed prior to evaluation and revealed NPS. Animal 5 had a previous diagnosis of feline asthma and had been treated intermittently via inhalation of a corticosteroid. Animal 6 had a previous diagnosis of bartonellosis. Two cats (animals 4 and 6) had 2 treatments each of fluoroscopicguided balloon dilatation of their stenotic lesions. Both cats had recurrence of clinical signs within 1 to 3 months after each balloon procedure. All 6 animals had severe continuous stertorous breathing on physical examination with an exaggerated inspiratory effort. Animal 5 was in respiratory distress and was placed in an oxygen cage at evaluation.

Diagnostic tests—Histologic examination of the stenotic tissue was not diagnostic in any case because the tissue collected was too small for meaningful evaluation. The histopathologic diagnosis obtained from biopsied nasal turbinates was chronic lymphoplasmacytic rhinitis in all animals examined (animals 2, 4, 5, and 6).

A CT scan was performed on all 6 animals prior to stent placement. In animal 1, there was an oblique band of soft tissue along the left side of the rostral portion of the nasopharynx, immediately caudal to the level of the sphenopalatine sinus and hard palate. The osseous dimensions also appeared axially narrowed. The size of the nasopharynx was 13 × 5 mm rostrally and 8.5 × 4 mm caudally. The stenotic area measured 6 mm in length. The CT scan for animal 2 revealed a stenotic lesion that extended caudally for 16 mm from just dorsal to the caudal aspect of the hard palate. The airway proximal to the stenosis measured 4 × 11 mm and caudal to the stenosis measured 10 × 10 mm. Some osseous narrowing of the rostral portion of the nasopharynx was evident just caudal to the choanae. There was voluminous fluid accumulation in the frontal sinuses. The CT scan of animal 3 revealed a stenotic lesion with no orifice, located 3 to 4 mm caudal to the junction of the hard and soft palate within the nasopharynx and extending 18 mm caudally. There was a large accumulation of fluid in both nasal passages. In animal 4, there was severe bilateral turbinate destruction in the rostral and middle nasal cavity with multifocal areas of fluid accumulation and mucosal enhancement. An NPS composed of contrast-enhancing tissue was detected just dorsal to the rostral aspect of the soft palate, causing a marked narrowing of the nasopharynx over a distance of 6.4 mm. Animal 5 had an NPS 11 mm rostral to the caudal edge of the soft palate. The diameter of the nasopharynx rostral and caudal to the NPS was 8 × 8 mm, and its length was 5.5 mm. The CT scan of animal 6 revealed moderate turbinate destruction bilaterally, moderate contrast enhancement of the nasal mucosa, and an NPS located 3.5 mm caudal to the junction of the hard and soft palate and extending 4.5 mm caudally (Figure 1). There was a moderate amount of fluid accumulation in both tympanic bullae and the nasal passages.

Stent placement—Retroflex rhinoscopy and simultaneous fluoroscopy were used in all animals at the time of balloon dilatation and stent placement (Figures 2 and 3). In animals 2 and 3, a slightly different approach was used to gain access across the stenotic lesion because of the presence of a complete intact membrane within the NPS. Predilatation with a PTA balloon was performed in all cases (dogs, 4 × 20 mm or 5 × 40 mm; cats, 4 × 20 mm or 6 × 40 mm). Stent sizes were then chosen on the basis of CT measurements and stent availability. Stenosis location was identified via fluoroscopy and rhinoscopy, and the stent was centered over the lesion (Figures 2–5). Stent sizes in the dogs were 10 × 20 mm, 10 × 27 mm, or a combination of both and in cats were 8 × 16 mm, 9 × 17 mm, or 9 × 19 mm, respectively. All stents were inflated to the balloons' individually recommended pressures (8 to 12 atmospheres of pressure). The procedure times ranged from 22 to 70 minutes with a median of 38 minutes. Animal 3 needed 2 stents at the initial procedure to span the entire stenosis. The initial stent (10 × 27 mm) was minimally displaced as the balloon was being retracted after dilatation. Because there was a small area at the caudal aspect of the NPS not covered by the stent, another 10 × 20-mm stent was placed to span the most caudal aspect of the stenosis, overlapping the first stent by approximately 50%.

Complications—There were no immediate post-procedural complications. Five of 6 animals did not have any evidence of pain or discomfort associated with the stent placement from the time of placement to the end of the study (8 to 22 months). Animal 5 had some exaggerated swallowing when eating, which the owner noticed, particularly after the cat vomited a hairball. For the first month, the owner did not consider this to be severe or concerning because the signs were intermittent. The cat was maintained on a strictly moist food diet. Radiographs obtained at 4 weeks following BEMS placement revealed no stent migration, and the cat was breathing normally. At 6 weeks, the owner noticed abrupt worsening of inspiratory noise, particularly when the cat was eating. Oral examination revealed a palpable bulge dorsal to the soft palate. Repeat rhinoscopy revealed that the distal aspect of the stent had partially collapsed, was not incorporated into the nasopharyngeal epithelium, and was filled with entrapped hair. Forceps were used to remove some of the hair and assess stent mobility. The stent could not be moved because of incorporation of metal into the naso pharyngeal epithelium rostrally. A balloon was inserted over a guidewire and inflated in the stent lumen (9 × 20 mm) to redilate the stent. The cat recovered without complications and was eating well at home at the 2- and 4-week follow-up telephone calls. Two months later, the cat acutely developed signs of severe intermittent inspiratory noise and exaggeration when swallowing. Radiography revealed no stent migration or narrowing. Retroflex rhinoscopy revealed further entrapment of hair within the stent lumen. Hair was removed as described, and the caudal third of the stent was trimmed with a wire cutter through a midline incision into the caudal soft palate. This portion of the stent had not been covered by nasopharyngeal epithelium and was suspected to be contributing to the dysphagia, hair entrapment, and discomfort during swallowing. The soft palate was sutured closed, and the cat recovered without complications and was able to eat normally the following morning. One year later, the owner reported that the cat was breathing and eating normally and was comfortable at home.

No animal with an open stenotic membrane had evidence of tissue in-growth through the stent. The 2 animals in which an orifice needed to be created (animals 2 and 3) had evidence of restenosis on reevaluation endoscopically. Animal 2 was clinically normal for 6 weeks and then abruptly developed upper respiratory stertor.

On endoscopic reexamination by the referring hospital, there was evidence of in-growth of tissue into the lumen of the stent, and the edges of the stent were deformed. Placement of a covered balloon-expandable stent to prevent aggressive tissue in-growth through the interstices of the stent was recommended but declined by the owner. Similarly, in animal 3, tissue was seen rhinoscopically to be growing through the interstices 2 weeks after BEMS placement, but a patent orifice was still present. This animal also had a completely covered NPS membrane, which required piercing at the time of BEMS placement, at which time a covered stent was unavailable but recommended. It was recommended for this dog to be reevaluated at 2 weeks to determine whether aggressive tissue in-growth was present as in animal 2. A 10 × 20-mm PTA balloon was used to redilate the stenosis inside the uncovered BEMS while a covered stent was ordered. At the time of this balloon dilatation, MMC^l was applied to the stenotic region topically (5 mL of 0.1% solution). The MMC solution was injected through a catheter at the newly dilated stricture site and allowed to dwell for 5 minutes before being flushed with saline solution. Approximately 1 month later, a 10 × 30-mm covered BEMS^m was placed through the stenotic area and through the previously placed BEMS. Three and 12 months following covered stent placement, the dog lacked clinical signs of airway obstruction and was able to breathe with a closed mouth. On occasion, the dog had bouts of mild intermittent inspiratory noise that resolved after sneezing.

Reexamination and clinical outcome—All animals were reexamined at 6 to 12 weeks with repeat rhinoscopy (n = 5), radiography (6), and CT (1). All stents remained in place with no evidence of migration. On repeat rhinoscopy, the stents were incorporated into the epithelium of the nasopharyngeal mucosa. Animal 5 had the rostral two thirds of the stent incorporated, but not the caudal third, as described. Animals 2 and 3 had more advanced tissue in-growth. Animal 3 had evidence of tissue in-growth at the time of the 2-week endoscopic examination at which the NPS was treated via balloon dilatation and MMC was topically applied as described. Approximately 1 month later, a 10 × 30-mm covered balloon-expandable stent was placed to prevent further in-growth. Continuous stertorous noise did not recur in this patient. Animal 2 had evidence of complete tissue in-growth at 6 weeks, and a covered stent was recommended but declined by the owner. All animals had immediate resolution of inspiratory stertor. Five of 6 animals lacked signs of stertor and were breathing normally at the end of the study period (12 to 28 months after stent placement). Animals 4 and 5 required treatment with prednisone every other day for persistent mild nasal discharge, which resolved with treatment. This discharge was present prior to the development of NPS, was much worse while the stenosis was present, and improved substantially after BEMS placement.

Discussion

Benign NPS is a rare acquired condition that has been reported in humans secondary to uvulopalatopharyngoplasty, palatal surgery, trauma, adenotonsillectomy, and severe syphilis or tuberculosis infection.^15–17 It is usually associated with a thin membrane that partially or completely obstructs airflow through the nasophayrnx, resulting in inspiratory stertor that is relieved with open-mouth breathing. Nasopharyngeal stenosis has been described in a small number of dogs and cats, with dyspnea more common in cats because they are more reluctant to open-mouth breathe. Nasopharyngeal stenosis has never been described with a complete occlusion of the nasopharynx, as seen in 2 of the 6 animals of this report. Typically, a small lumen is detected via rhinoscopy.

Clinical signs in the 6 client-owned animals reported here were similar to those described previously for dogs⁸ and cats with NPS.^1–6 A diagnosis of NPS was made in all animals with a combination of CT, fluoroscopy, and retroflex rhinoscopy. A particular benefit of CT is its accurate measurements of the nasopharyngeal dimensions just proximal and caudal to the area of stenosis, as well as its length, for determination of appropriate stent size. Computed tomography may also provide important information about the extent, tissue type, and exact location of the stenosis and other concurrent abnormalities such as bony proliferation, excessive numbers of nasal turbinates, and choanal atresia. Retroflex rhinoscopy does not reveal the most rostral aspect of the stenosis, but it can usually be visualized via CT and contrast fluoroscopy. It is important to know that mucus accumulation rostral to the NPS can make recognition of the rostral aspect difficult. This makes the combination of CT, rhinoscopy, and contrast fluoroscopy useful if the rostral aspect is not clearly visible via CT.³ Rhinoscopy and CT have been the diagnostic tools of choice for humans with NPS or choanal atresia for decades.¹⁸ Measurements can also be obtained with the use of magnetic resonance imaging or by use of a marker catheter inserted under fluoroscopic guidance into the nasopharynx.³ The marks are 1 cm apart, and with a contrast study rostral and caudal to the stenosis, magnification can be adjusted for, and the length of the stenosis can be accurately discerned. It is important that the patient is intubated and the endotracheal tube cuff is fully inflated to prevent aspiration of contrast medium.

Nasopharyngeal stenosis has traditionally been treated surgically via the technique described by Mitten⁴; however, recurrence of the stenosis is considered common.^1,4,6,7,11 Choanal atresia in humans is different from NPS because this usually involves bony and membranous tissue proliferation at the atretic site of the choanae, requiring aggressive tissue removal rather than use of a simple balloon that can break the fibrous lesion as seen in NPS. Stenting for choanal atresia in humans has been met with controversy, with endoscopic laser repair holding the best promise for cure.^16,17,19 Choanal atresia in animals has been reported occasionally, also as a separate disease condition from NPS, though it is most often associated with a similar type of membranous scar tissue, rather than a bony membrane, making it more difficult to distinguish between the 2 conditions.^1,2 Balloon dilatation with or without stent placement may also be a consideration for this condition. Balloon dilatation of NPS has been reported in numerous cats,^3,5,6 a dog,⁸ and children.^20,21 Multiple interventions are often required to achieve long-term resolution of clinical signs.^3,5,6,8 This was evident in 2 of 3 cats and 1 of 3 dogs of the present report prior to stent placement. Serial procedures require multiple anesthetic episodes and the associated cost. There is 1 description of a temporary catheter that was used as a stent in a balloon-dilated unilateral choanal atresia in a cat.² Recently, in humans, successful endoscopic repair by use of laser technology, obturators, or plastic tubing for temporary stent placement in choanal atresia and NPS has been reported.^16,17,19 The present report is the first description of the use of BEMS for permanent treatment of NPS in dogs and cats, to our knowledge. This is also the first report of a completely closed membrane in veterinary patients with NPS and the use of a covered BEMS for such a condition.

The minimally invasive treatment for NPS reported here was both a successful and effective therapeutic modality in 5 of the 6 cases. The authors suspect that the use of a covered stent in the animal that had tissue ingrowth through the stent would probably have resulted in a positive outcome as well. Repeated procedures after BEMS placement will hopefully be limited in the future if covered stents are placed for closed membranes initially and a caudal stenosis is serially ballooned prior to stent placement and if, in the event a stent is placed, 1 cm of unstented soft palate is left caudal to the stent. The procedure was short (22 to 70 minutes) and minimally invasive. All stenotic lesions were predilated with the PTA balloon, which facilitated placement of the premounted BEMS across the stenotic lesion, thus reducing the risk of the stent sliding off the balloon when being advanced through the stenosis. A long 18-gauge access needle, access sheath, and covered stent are recommended for treatment of completely occluded lesions without a preexisting orifice.

Additional procedures were needed in 2 dogs with a complete membrane and a cat with a caudal stenosis. For NPS with complete membranes that separate rostral and caudal portions of the nasopharynx and do not have a preexisting orifice, reexamination with endoscopy in 1 to 2 weeks after stent placement may be performed if an open BEMS is used to evaluate whether partial in-growth is occurring prior to complete occlusion. If tissue in-growth is found, a covered stent is recommended. It makes sense that those animals with a closed membrane would have a higher risk of restenosis because the lumen that is created is not surrounded by nasopharyngeal epithelium and is made of fibrous tissue. The tissue, once opened with the BEMS, is a new lumen that is not coated with epithelial mucosa, and the fibrous tissue is likely to continue to grow. When a covered BEMS is used, the fibrous tissue cannot reocclude the lumen of the nasopharynx. Without having a large number of animals with covered stents in place, it is difficult to assume that their use is without complications, but in the 1 dog of this study treated in that manner, the stent was well tolerated. The authors recommend placing a covered BEMS initially for animals with closed membranes, rather than a regular BEMS, although covered stents are nearly double the price of a noncovered BEMS and need to be specially ordered at least 2 to 4 weeks in advance.

In the cat that required additional procedures, there was mild intermittent exaggeration of swallowing. Continued entrapment of hair at the caudal aspect of the stent, which was not incorporated into nasopharyngeal mucosa, led to gagging and inspiratory noise several months after stent placement. The lack of epithelialization could have been caused by the movement of soft palate and nasopharyngeal walls when swallowing, which interrupted contact between the metal and mucosa. The exaggerated swallowing was likely associated with the presence of the stent caudally, which prevented the caudal aspect of the soft palate from closing while swallowing or vomiting, although those signs were intermittent. The cat had exaggeration of clinical signs after vomiting a hairball on 2 occasions. This could have been caused by aspiration-induced rhinitis in the nasopharyx because the caudal portion of the soft palate was no longer protective. Once the stent was minimally trimmed, enough palate was left unstented to allow those signs to resolve. The remainder of the stent being in place did not cause this animal a problem. The proximity of the relatively rigid, metallic stent over the soft palate could affect palate pliability and interfere with the process of swallowing. Therefore, if the NPS is located in the caudal third of the nasopharynx, balloon dilatation alone (with or without the use of MMC) or, when available, use of a biodegradable or retrievable stent may be advisable. This would allow the stenosis to open up without associated complications of a metallic stent that could interfere with swallowing or vomiting. The authors feel that if there is at least 1 cm of soft palate remaining caudal to the stent closure of the nasopharynx, swallowing can occur more easily. Fortunately, the most commonly reported location for an NPS is immediately caudal to the junction of the hard and soft palate, as seen in all other animals of this report. The stent location was well tolerated by all 5 of these animals. If serial balloon dilatation is declined or contraindicated, stent placement at this more typical location can be recommended. It was immediately effective and safe for the cases reported here.

Stent placement was selected for treatment of NPS in these animals because of multiple failed balloon dilatation (n = 2) or laser procedures (1) or serial procedures declined by the owner (3). This modality was also less invasive and less expensive than surgical intervention or serial balloon dilatation. All animals were discharged from the hospital the same or following day with normal breathing patterns and no appreciable postoperative discomfort. Long-term resolution of clinical signs was accomplished in 5 of 6 animals (12 to 28 months following stent placement). The authors do not recommend stenting the most caudal aspect of the nasopharynx (caudal third of the nasopharynx) because of possible interference with the swallowing function of the soft palate. Serial balloon dilatation or biodegradable or retrievable stent placement should be attempted prior to considering a BEMS in this distal location. At least 1 cm of unstented nasopharynx dorsal to the soft palate distal to the stent is suggested to allow for nasopharyngeal closure during swallowing.

The local use of triamcinolone⁸ was not attempted in any of the reported animals because of the small working channel of the endoscope and the difficulty of submucosal injection between the interstices of the stent, but MMC was topically applied in 1 animal after balloon dilatation while awaiting the delivery of the covered stent. There has been some anecdotal topical use of MMC after balloon dilatation of esophageal, rectal, or urethral strictures in a small number of small animals, none of which are reported in the literature. This has been described in humans for esophageal, tracheal, and urethral strictures^22,23 and was used in animal 3 when excessive in-growth of tissue was seen after initial stent placement. The dose used here was extrapolated from the human literature and anecdotal use in veterinary patients. In hindsight, this could have been performed in all animals after BEMS placement (particularly animal 2), but discovery of the idea was not made until animal 3 was evaluated.

Mitomycin C, an anthracyclin antimicrobial, is an antineoplastic agent that inhibits fibroblast proliferation, reduces collagen cross-linking, and is effective in reducing scar formation.²⁴ Its successful use for choanal atresia in humans²⁵ has been recently reported. Animal 3 had MMC applied after balloon dilatation while awaiting the delivery of the covered stent. Its benefit in this patient was difficult to assess because uncovered stent placement was performed and the stricture remained static during the interim. The authors recommend intralesional injection of triamcinolone in animals in which the endoscope is large enough to accept an appropriate needle⁸ or topical application of MMC, as described.

Balloon-expandable metallic stent placement is an effective and reliable treatment modality for fixation of NPS at the junction of the hard and soft palate in dogs and cats. Immediate placement of a covered BEMS should be considered in animals with a closed membrane, although the extra cost associated with this stent should be taken into account. Another option is to repeat balloon dilatation at 1 to 2 weeks following BEMS placement for animals with complete occlusion of the nasopharynx, along with the topical application of MMC. Tissue in-growth seems uncommon in animals with an open NPS at the time of diagnosis, and a covered BEMS is not considered necessary for the typical case of NPS.