Transnasal Laryngoscopy for the Diagnosis of Laryngeal Paralysis in Dogs

Introduction

Idiopathic laryngeal paralysis usually occurs in older, large-breed dogs and is a slowly progressive cause of upper airway obstruction.^1–3 Clinical signs of abnormal phonation and inspiratory stridor may progress to heat or exercise intolerance, dyspnea, and collapse.¹² The initial diagnosis is based on clinical signs, neurological examination, thoracic radiography, and laboratory tests to rule out polyneuropathies, central nervous system disorders, cervical or thoracic neoplasia, trauma, or inflammatory processes.^1–4 A definitive diagnosis is made with laryngoscopy per os (PO) under heavy sedation or general anesthesia. Controversy exists as to the best method of sedation and anesthesia for laryngoscopy. While visualization of normal function is not disputed, lack of function or diminished function may result from the sedative or anesthetic agents used, or the depth of anesthesia achieved.⁵ Specific, published studies of the effects of premedications and anesthetic agents on laryngeal function in dogs are lacking. One study demonstrated greater ease of evaluation of the larynx using thiopental or propofol compared to ketamine-diazepam in normal dogs.⁶ Two abstracts investigated the effects of different anesthetic agents on laryngeal function in normal dogs.⁷⁸ Those studies reported that standard oral laryngoscopy under thiopental induction and intravenous doxapram allowed for the largest change in normalized glottic gap area.⁷⁸

Transnasal endoscopy is used to assess laryngeal structure and function in awake or sedated horses and cattle.^9–11 The objective of this study was to assess the feasibility of transnasal laryngoscopy to evaluate laryngeal structure and function in sedated dogs.

Materials and Methods

The dogs in this report were presented to the Kansas State University Veterinary Medical Teaching Hospital for evaluation of laryngeal function, mass removal (case no. 1), or elective ovariohysterectomy. A physical examination, neurological examination, and complete blood count (CBC) were done on each dog. Thoracic radiography and a serum biochemical profile were performed on dogs with signs of upper airway obstruction. An opioid analgesic and acepromazine were given to each dog (with or without an anticholinergic agent). All premedications were based on the anesthesiologist’s preference for laryngeal examination and were given intramuscularly [see Table].

Lidocaine (2% solution) was applied to the left nasal passage through a 3.5-French (Fr) polypropylene catheter inserted into the ventral nasal meatus approximately 30 minutes after premedication. The dose of lidocaine varied from 1 to 2 mg/kg with each dog receiving approximately 2 mL of lidocaine. A flexible video endoscope with a 2.5-mm outer diameter (OD) with two-way deflection capability^a was passed through the left nasal passage until the larynx was visualized. The position of the endoscope was adjusted so that the larynx occupied most of the viewing field and was in a dorsoventral position on the video monitor. An assistant restraining the dog announced when each inspiratory phase of ventilation occurred, and the larynx was observed until the endoscopist was comfortable with the laryngeal assessment.

All but one of the dogs was induced with either thiopental or propofol after transnasal laryngoscopy, and laryngoscopy was performed PO. Endotracheal intubation followed, and three of the four dogs with laryngeal paralysis underwent left arytenoid lateralization. Ovariohysterectomy or a mass removal was later performed on the normal dogs. All dogs were monitored for complications after laryngeal evaluation, including epistaxis and stertor, and were fed the morning after surgery. Dogs with laryngeal paralysis were fed small balls of canned food, and dogs that underwent ovariohysterectomy were fed a combination of canned and dry food.

Results

Four dogs with suspected laryngeal paralysis and three normal dogs were evaluated [see Table]. Case no. 1 had a 6.8 × 4.0 × 4.3-cm mass in the subcutaneous tissue of the ventral neck. Prednisone (0.3 mg/kg PO) had been administered every other day for the week prior to examination to treat the cervical swelling. The CBC and biochemical profile in this dog revealed changes consistent with corticosteroid administration, including elevated serum alkaline phosphatase (1721 U/L; reference range, 12 to 122 U/L) and alanine amino-transferase (175 U/L; reference range, 13 to 79 U/L), and lymphopenia (690/μL; reference range, 1.5 to 5.0 × 10³/μL). No abnormalities were found on any CBC, biochemical profile, or thoracic radiographs of the remaining dogs.

Six dogs (case nos. 1–4, 6, 7) were mildly or moderately sedated by the premedications, and one dog (case no. 5) was severely sedated. The dogs were calm and required minimal restraint, but they resisted the nasal manipulations. Administration of intranasal lidocaine caused sneezing in five of the seven dogs, but passage of the endoscope was achieved in each case without complication. Laryngeal structure and function could be observed in every dog, and manipulation of the epiglottis was not required in any dog [see Figure]. All dogs were comfortable and did not resist examination once the tip of the endoscope was stable in the caudal pharynx. Bilateral laryngeal paralysis was identified in all four dogs (case nos. 1–4) with signs of upper airway obstruction. The gross appearance of the mucosa was evaluated by moving the endoscope closer to the larynx. The vocal folds remained in a paramedian position, and no motion of the arytenoid cartilages was observed with spontaneous ventilation in three dogs (case nos. 1, 2, 3); paradoxical arytenoid movement occurred in one dog (case no. 4). One dog (case no. 1) had no obvious movement of the epiglottis with swallowing attempts. The epiglottis could not be observed during swallowing in the other six dogs as a result of pharyngeal motion. A closed biopsy forceps was passed down the biopsy channel of the endoscope in one dog with laryngeal paralysis (case no. 4); contact with the arytenoid mucosa caused swallowing but did not induce cough or arytenoid motion. At the request of the owners, no further surgery was performed in this dog.

One dog (case no. 5) became extremely sedate from the premedications and could not actively lift its head or resist initial manipulations. Arytenoid motion was absent upon initial evaluation of the larynx. The dog allowed passage of an 8-Fr polypropylene catheter through the oral cavity without resistance. Immediate, vigorous exhalation and inspiration occurred with extreme adduction and abduction of the arytenoid cartilages in phase with ventilation when the catheter contacted the left arytenoid mucosa. Normal ventilation and arytenoid movement then occurred. Normal arytenoid motion was absent in a second clinically normal dog (case no. 7) that was only mildly sedated during passage of the endoscope. A closed biopsy forceps was then passed down the biopsy channel of the endoscope, and contact with the arytenoid mucosa resulted in immediate arytenoid adduction and abduction, which continued with normal ventilation.

Left arytenoid lateralization was performed on three of the four dogs with laryngeal paralysis. The cervical mass in case no. 1 was excised, and histopathological examination of the tissue was consistent with an infected sebaceous adenoma. The trachea, recurrent laryngeal nerves, and carotid sheaths were not in contact with or displaced by the cervical mass. No epistaxis or stertor occurred in any dog during or following recovery from anesthesia. Dysphagia was not noted upon feeding any of the dogs the day after surgery.

Discussion

Transnasal laryngoscopy was successfully performed on all seven dogs in this study. The method of examination was standardized to the left nostril based on previous equine studies that demonstrated the left nostril was the best route for repeatable examinations.¹²

Transnasal laryngoscopy was performed in this study using published doses of medications commonly used prior to induction of general anesthesia. Each animal was sedated but responded to passage of the endoscope, and all but one maintained normal, deep respiration. The clinically normal dogs that initially failed to adequately abduct the arytenoid cartilages responded vigorously to topical stimulation of the arytenoid mucosa. This maneuver did not result in arytenoid motion in one dog with laryngeal paralysis. The use of a catheter was not possible in the other animals, because they were too alert and responsive to allow passage of a catheter PO. Topical stimulation was performed in two of these dogs using a closed biopsy forceps passed down the operating channel of the endoscope, as described in horses and cattle.¹¹¹³ Contact with the arytenoid mucosa by the forceps failed to induce arytenoid motion in one dog with laryngeal paralysis, but it resulted in normal arytenoid motion in a clinically normal dog. Based on these findings, it is recommended that topical stimulation of the arytenoids be attempted in any dog lacking arytenoid motion on initial evaluation, prior to concluding that laryngeal paralysis is present.

The size of the endoscope used in this study may limit its use in smaller dogs. A 2.2-mm OD, 30-cm long rhinoscope^b is available and has been used by the authors in dogs in which the 2.5-mm endoscope cannot be passed. The rhinoscope passes easily but does not contain a biopsy channel, thereby eliminating the ability to topically stimulate the larynx through the endoscope.

Evaluation of the larynx in this study was carried out at a lighter plane of sedation than previously described. As a result, the technique may decrease the number of false-positive results or questionable diagnoses of laryngeal paralysis compared to standard oral laryngoscopy. Assessment of laryngeal function must always be considered in conjunction with the plane of sedation or anesthesia utilized. The plane of anesthesia is not an adverse issue if arytenoid motion is obvious. However, anesthetic drugs may cause temporary loss of laryngeal function and may falsely allow a diagnosis of laryngeal paralysis if the arytenoid cartilages fail to abduct. With the passage of time, partial recovery from these anesthetic effects and return of arytenoid function may occur.⁸ With respect to changes in the size of the glottic aperture, there are no reported differences between induction and recovery from anesthesia in normal dogs.⁸

Different sedatives have also been shown to diminish laryngeal function in horses and cattle.¹¹¹² Xylazine caused diminished or abolished laryngeal function in horses, as well as increased upper airway resistance and increased work of breathing.^12–15 Xylazine also diminished laryngeal function and arytenoid position in cattle.¹¹ The use of xylazine was avoided in this study because of its potential effect on laryngeal function and its increased risk of side effects (i.e., hypertension, bradycardia) in older dogs. Acepromazine was determined to be an acceptable sedative for transnasal laryngoscopy in large animals and was therefore included in this study’s premedication regimens.¹¹¹² However, the dog (case no. 5) in this report that received the highest dose of acepromazine had diminished laryngeal motion on initial examination. Future study is needed to evaluate the ideal dosage of acepromazine to be used for laryngeal examination. Dosages causing abolition of laryngeal reflexes have been determined for different rapid-acting barbiturates.¹⁶ Thiopental and isoflurane via mask were deemed acceptable methods of anesthetic induction for laryngeal evaluation in one study.⁸ These drugs may represent supplemental methods of restraint for transnasal laryngoscopy; however, mask administration of isoflurane would have to be temporarily discontinued for passage of the endoscope. Research into the effects of thiopental, ketamine, topical lidocaine, and inhaled halothane on laryngeal reflexes in cats has mainly studied their effects on laryngospasm and not their effects on arytenoid abduction.^17–19

The effects of sedatives, anesthetic induction agents, and inhalant agents on laryngeal function in people have also been studied. Laryngeal competence was best maintained with ketamine alone compared to ketamine after premedication with parasympatholyics, opioids, and benzodiazepine, or compared to induction with thiopental or methohexital.²⁰ Propofol and thiopental induction have been examined in humans, and propofol had a more detrimental effect on laryngeal activity than thiopental.²¹ People in that study were premedicated with a benzodiazepine and parasympatholytic.²¹ A study of propofol alone showed no change in reflex activity of the upper airway except after fentanyl infusion, which decreased laryngeal reflexes.²²²³ It appears that both premedication and induction agents alone and in combination can alter laryngeal activity in people and in dogs. The study reported here suggests that both the drugs administered and the dosages used may affect laryngeal function.

Topical stimulation of the laryngeal mucosa or stimulating the animal to breathe more deeply may be attempted if the arytenoid cartilages fail to abduct upon initial examination. In this study, topical stimulation was performed PO or through the endoscope. Other methods of stimulating respiration include intercostal muscle stimulation or application of pressure to a digit with forceps. An increase in the size of the glottic aperture was seen in normal dogs that were anesthetized with isoflurane and given doxapram in one study; however, doxapram’s use has not been evaluated in dogs with laryngeal paralysis.⁷

Topical lidocaine was administered after sedation to facilitate passage of the endoscope through the nasal cavity and avoid the need for induction of general anesthesia. A significant amount of lidocaine was expelled from the nares by sneezing; however, its use is still recommended. The effects of systemic absorption of lidocaine and the effects on laryngeal reflexes associated with nasal application of lidocaine should be considered. Significant plasma drug concentrations of lidocaine to induce toxic side effects were not expected at the dosages used, which were based on studies performed in children and infants.²⁴²⁵ The effectiveness of topical stimulation may also be lost if large amounts of lidocaine reach the larynx, decreasing the ability to induce laryngeal motion with topical stimulation in dogs obtunded by sedation. A recent study in horses showed that complete anesthesia of the larynx did not ablate normal arytenoid motion.²⁶

Ultrasonographic examination of the larynx has been developed in dogs as an alternative to direct visualization.²⁷²⁸ Based on the results of the study reported here, direct visualization of larynx via transnasal laryngoscopy eliminated the difficulty of performing ultrasonography and any problems with interpreting ultrasonographic findings. Transnasal laryngoscopy also eliminated the adverse effects of deep sedation and general anesthesia, which are often required for oral laryngoscopy.

Conclusion

This study demonstrated that transnasal laryngoscopy allowed direct visualization of the larynx in awake, sedated dogs and verified both normal and abnormal laryngeal function. The technique was successfully performed on all seven dogs, and the endoscope used was an appropriate size for the large-breed dogs that typically are presented with idiopathic laryngeal paralysis. The authors concluded that transnasal laryngoscopy is useful as a minimally invasive tool for evaluating large-breed dogs with signs consistent with laryngeal paralysis.