Rostral direction virtual laryngoscopy in canine glottis under sedation with butorphanol, medetomidine, and alfaxalone

Abstract

The glottis of 26 dogs with respiratory signs was examined by virtual laryngoscopy for peering into the throat from within the trachea (RDVL: rostral direction virtual laryngoscopy) with and without sedation of butorphanol, medetomidine, and alfaxalone (BMA). These dogs are divided into two groups of with or without sedation. In 9 unsedated dogs, there were many Chihuahuas, and the primary symptom was coughing. In 17 sedated dogs, Toy Poodles and Pomeranians were the most common breeds, and the primary symptoms were coughing and worsening respiratory symptoms at excitement. RDVL provided a reliable view of glottic motion, and the airway space was interfered with for flaccid adductors. RVDL without sedation is recommended. However, BMA sedation is suggested when necessary.

The larynx opens and closes airway through the balances of contraction and relaxation of the internal and external muscles. The internal muscles are composed of several muscles as follows; cricothyroid, posterior cricoarytenoid, lateral cricoarytenoid, transverse arytenoid, oblique arytenoid, aryepiglottic, vocalis and thyroarytenoid muscles. The posterior cricoarytenoid muscle (abductor of arytenoid) opens the glottis during breathing, and the lateral cricoarytenoid, transverse arytenoid, oblique arytenoid and aryepiglottic muscles (adductors of arytenoid) close the glottis during swallowing [9]. It is not yet known whether these flaccid adductors coordinate with respiratory movements during breathing, nor to what extent they interfere with the airway. In humans, it is suspected that the morphological and functional abnormalities of adductors interfere with the patency of the laryngeal airway [9].

Echolaryngography is an unsedated method of functional assessment of the larynx [16, 20]. Asymmetry or absence of motion of the cuneiform processes is mainly confirmed as a sign of laryngeal paralysis in dogs. Moreover, echolaryngography can diagnose paralysis and laryngeal mass, cyst, and everted saccules in dogs and cats [19, 21, 22]. However, the swallowing reflex, which would interfere with the motion of cuneiform processes, has been reported to occur if the elongated soft palate located on the epiglottis is inserted into the airway between the left and right arytenoid cartilages during inspiration [2, 5, 18]. Therefore, it is necessary to assess the relationship between the structures constituting the throat when abnormal laryngeal motion is suspected. Echolaryngography is an essential method for screening; however, laryngoscopy remains the most standardized examination for the pharyngeal and laryngeal areas in dogs. In laryngoscopy per os, immobilization is necessary to maintain the position of the head and open the mouth. For immobilization, propofol or alfaxalone is usually administered to achieve a light plane of anesthesia [1, 2, 12,13,14,15,16, 25]. They have the potential to inhibit laryngeal motion in some dogs in the light plane of anesthesia. Therefore, the administration of doxapram, which is a respiratory stimulant, and can increase laryngeal motion, is required to confirm laryngeal paralysis [2, 15, 17, 29]. It has also been reported that isoflurane anesthesia after induction with propofol and alfaxalone maintains arytenoid abduction activity [6]. This switch to inhalation anesthesia requires a waiting time of 20−40 min to improve laryngeal function. Recently, the examination of laryngeal function with a protocol containing dexmedetomidine, which is an α-2 adrenergic receptor agonist, and butorphanol has been reported to produce sufficient immobilization without affecting arytenoid abduction [3, 26]. Medetomidine, which provides analgesic effects like dexmedetomidine, is also commonly used for sedation in Japan. In addition, alfaxalone has a depressant effect similar to that of propofol, whereas alfaxalone has a chance to affect laryngeal function compared with propofol at induction doses [1, 14]. Therefore, at our hospital, when performing screening examinations using computed tomography (CT) scans from the head to the chest in dogs with respiratory diseases, sedation using a combination of low-dose medetomidine and low-dose alfaxalone followed by butorphanol (BMA) is administered as necessary. Furthermore, CT examination of the glottis has already been reported to be useful for predicting airway obstruction [23, 28]. In this retrospective study, we reconstructed three-dimensional images from two dimensional-CT images and devised a virtual laryngoscope for peering into the glottis from within the trachea (RDVL: rostral direction virtual laryngoscopy) to investigate whether flaccid adductors interfere with the laryngeal airway under the sedation with butorphanol, medetomidine, and alfaxalone in the canine epiglottis.

Between April 2024 and March 2025, we retrospectively investigated cases in which RDVL was performed in dogs that were referred and presented to the Service of Respiratory Medicine, Animal Medical Center, Nihon University, with complaints of respiratory signs, and in which four dimensional (4D)-CT was performed for morphological and functional examination of the upper airway or tracheal bifurcation. All dogs were positioned in sternal recumbency. Images during spontaneous respiration (at least one breath) were acquired for 6 sec using a dynamic 4D-CT system equipped with 320 detector rows, each 0.5 mm in width with a gantry rotation time of 0.275 sec, covering a 16 cm range (24 cm field of view) per rotation (Aquilion ONE, Toshiba Medical Systems, Tokyo, Japan) [4]. For the RDVL, CT images were reconstructed using the virtual endoscopy mode of INTAGE Realia (image processing software, KGT Inc., Tokyo, Japan). The camera position of the virtual laryngoscope was set at the level of the cricoid cartilage in the trachea (Fig. 1).

Fig. 1.

Camera positions for rostral-directed virtual laryngoscopy. The right and left images show the coronal and sagittal views, respectively, of the head computed tomography images. The cross points of the purple lines and blue wedges indicate the location and direction of the camera, respectively.

This study was approved by the Animal Hospital Committee for Clinical Study (ANMEC-1-027). All examinations were performed after written informed consent of the dogs’ owners. Samples were collected for routine clinical purposes by well-educated and trained veterinarians in Animal Medical Center, Nihon University. All dogs were also diagnosed and treated by these veterinarians at the hospital and provided with customary high standards of medical care and welfare. In addition, the study was performed in accordance with the Guide for Animal Experimentation published by Nihon University for exemption from the requirement for ethics approval.

During the study period, 29 dogs were examined using 4D-CT, of which 26 were unsedated (N=9) or sedated with BMA (N=17) and included in this study. One dog that was administered both butorphanol and medetomidine and two dogs that were administered both butorphanol and alfaxalone were excluded from the study. The BMA protocol is as follows. At first, dogs were administered intravenously medetomidine (0.5 µg /kg, Dorbene Vet, Kyoritsu Seiyaku, Co., Tokyo, Japan) and alfaxalone (0.1 mg/kg, Alfaxan^®, Meiji Seika Pharma. Co., Ltd., Tokyo, Japan) with prior administration of butorphanol (0.2 mg/kg, Vetorphale, Meiji Seika Pharma Co., Ltd.). After that, medetomidine (0.5 µg/kg) and alfaxalone (0.1 mg/kg) were administered in additional doses according to the level of consciousness until the CT scan was performed. The total median dose of medetomidine and alfaxalone was 2.5 µg/kg (range: 1.0–9.0 µg/kg) and 0.5 mg/kg (range: 0.2–2.2 mg/kg). In the sedated state, dogs remained motionless but were unable to open their mouths, and oxygen (1–2 L/min) was supplied to the tip of their nose via a flow-by system. The breeds of unsedated dogs were Chihuahuas (N=5), Toy Poodles (N=1), Pomeranians (N=1), Miniature Dachshunds (N=1) and Pugs (N=1). The breeds of sedated dogs were Toy Poodles (N=5), Pomeranians (N=3), Shiba (N=2), mongrels (N=2), Chihuahuas (N=1), Labrador Retrievers (N=1), Miniature Pinschers (N=1), Papillons (N=1) and Yorkshire Terriers (N=1). In unsedated dogs, the genders were 5 males (castrated: 3) and 4 females (spayed: 4), with a median age of 11 years (range: 6–14) and median body weight of 3.6 kg (range: 2.9–7.3). In sedated dogs, the genders were 7 males (castrated: 5) and 10 females (spayed: 8), with a median age of 11 years (range: 7–13) and median body weight of 5.1 kg (range: 2.9–19.2). The main clinical sign in unsedated dogs was cough (N=5), while the others were stertor (N=1), stridor (N=1), inspiratory effort (N=1), and tachypnea (N=2). The main clinical sign in sedated dogs was cough (N=7) also, while the others were reverse sneezing (N=1), wheezing at excitement (N=2), dyspnea at excitement (N=2), voicelessness (N=1), and tachypnea (N=3).

From the RDVL shown in Fig. 2, the image with the largest opening of the glottis at inspiration was selected from each dog. The shape of the glottis was classified into four types; diamond (the gap extends from the dorsal inner margin to the ventral inner margin of the glottis, and its horizontal length at the central portion exceeds at least one-fourth of the vertical length of the gap), flat (the central portion has a horizontal length less than one-fourth of the vertical length of the gap), hole (no gaps on the dorsal and ventral inner margins of the glottis), and closed, as shown in Fig. 3. In addition, the areas of airway space at the glottis were measured using the open-source software ImageJ (1.54 g) during the expiration and inspiration (Fig. 3). Table 1 summarizes a summary of findings. In the unsedated dogs, the shapes of diamonds and flats were eight (89%) and one, respectively. There were no cases of hole or closed glottis. In the sedated dogs with BMA, the shapes of diamond, flat, hole, and closed were 5 (29%), 9 (53%), 2, and 1, respectively. The three sedated dogs showed paradoxical laryngeal movements from expiration to inspiration. In addition, one of the three dogs with paradoxical laryngeal movements had a closed glottis at inspiration, and was diagnosed with laryngeal paralysis (stage 2) by standard laryngoscopy per os. The area of airway space during the inspiration in unsedated dogs and sedated dogs without exacerbation at excitation is significantly larger than the expiration. In addition, the area of inspiration in unsedated dogs is significantly larger than the area of inspiration in sedated dogs and no significant difference in sedated dogs without exacerbation at excitation. The area of inspiration in sedated dogs with exacerbation at excitation tended to be small (Fig. 4).

Fig. 2.

Rostral direction virtual laryngoscopy of the glottis during respiration (the lower right of each image shows the elapsed times). In the sedated dog of the upper row, the hole shape of the glottis shows the diamond shape at inspiration (1.5 sec), and the attachment of the dorsal surface of epiglottis (Ep) and the ventral surface of pharynx (Ph) closes the airway. In the unsedated dog of the lower row, the caudal edge of the soft palate (Sp) is located at the front of the epiglottis (Ep). The diamond-shaped glottis widened further at 1st inspiration (1.0 sec), and the glottis showed a flat at expiration (1.5 sec) after that. Then, the pharynx (Ph) drops just before the 2nd inspiration (2.5 sec). L: left, R: right, D: dorsal and V: ventral.

Fig. 3.

The shapes of glottis classification (shape A: diamond, B: flat, C: hole, and D: closed) from the image showing during inspiration. In the shape of a diamond (A), the soft palate located under the epiglottis (Ep) and the nasopharyngeal airway can peek between the pharynx (Ph) and the epiglottis. Relaxation of the cricothyroid muscle (Cr) causes it to bulge inward, interfering with the glottis. L: left, R: right, D: dorsal, and V: ventral. The area inside the yellow dotted line indicates the airway space at the glottis.

Table 1. Shape of the largest opening of the posterior glottis at inspiration with and without sedation.

Shape type	Findings of virtual laryngoscopic rostral direction imaging
	Diamond	Flat	Hole	Closed
Unsedated (9)	8	1	0	0
Sedated (17)	5	9	2	1*

The numbers in the table indicate the number of dogs. This contingency table is analyzed with the Fisher exact test and the proportion (Diamond vs. Flat, Hole, and Closed) among unsedated and sedated groups shows a significantly different (P=0.011). *: This case was diagnosed with a laryngeal paralysis (Stage 2) by the standard laryngoscopy.

Fig. 4.

The area of airway space at the glottis during expiration and inspiration in unsedated (US) and sedated (SD) dogs with and without exacerbation (EX) at excitation. The groups were compared using Mann-Whitney rank sum test, and a P-value of <0.05 was considered statistically significant. a: expiration vs. inspiration; b: vs. inspiration of US. Solid horizontal bars represent median values and boxes represent the 25th to 75th percentiles. Whiskers represent 5th to 95th percentiles and outliers are indicated by black dots.

In this study, BMA sedation appeared to influence laryngeal function and interfered with opening the laryngeal airway for flaccid adductors. In humans, virtual laryngeal endoscopy has already been used to evaluate organic disorders such as laryngeal tumors and adhesions, but the direction of the view is the same as that of a conventional endoscope, and there appears to be no report of its use in evaluating laryngeal movement function [31]. However, a report on virtual endoscopes in the veterinary field diagnosed laryngeal paralysis and collapse with preserved inspiration, but the direction of view was still the same as that of conventional endoscopes [26]. To the best of our knowledge, this is probably the first time that RDVL has been used to analyze glottal motion in the rostral direction. The RDVL in this study may be suitable for investigating the throat in a state where the dog is keeping its mouth closed and aperture is not possible due to jaw tension in intolerant patients to laryngoscopy and high-risk anesthesia cases.

Common and unwanted side-effect of anesthetics, sedatives and analgesics is respiratory depression. These drugs act on the respiratory center to suppress respiratory movements and have the effect of relaxing the respiratory muscles. BMA sedation, therefore, has a possibility to reduce the pharyngeal tone and alter the inspiratory effort. These may affect RDVL findings. In addition, the abductor and adductors of arytenoid cartilages are all innervated by the same recurrent nerves [8]. However, currently available research on laryngeal function is limited to the following studies, which focus on abduction during sedation in veterinary medicine. Thiopental was often used for laryngoscopy in the past, but since it is no longer available, propofol and alfaxalone are now used as substitutes [1, 2, 12,13,14,15,16, 25]. These drugs are used in veterinary medicine as induction agents at doses higher than those used in this study and are excellent agents for intubation. Sedatives such as dexmedetomidine and medetomidine are commonly used to immobilize animals. Compared to the aforementioned induction agents, they may be more suitable for laryngoscopy, which involves the examination of laryngeal movement [3, 26]. Unfortunately, there are no reports examining the effects on laryngeal movement using a laryngoscope with a combination of BMA, as in this study. The protocol for alfaxalone with acepromazine and butorphanol was less likely to affect laryngeal motion, and alfaxalone alone was not recommended [13]. Several reports have examined the effects of these three agents on the cardio-respiratory system, although the doses and routes of administration differed from those of BMA in this study [7, 8, 10, 11, 27]. It appears that some dogs experience respiratory effects, potentially leading to hypoxemia, hypercapnia, and acidosis [7, 8, 27]. These findings indicate hypoventilation due to BMA, suggesting a decrease in both respiratory rate and tidal volume. RDVL also detected to interfere with the glottis for flaccid adductors as an inward bulge by sedation, forming flat, hole-shaped and closed glottis, which may be a contributing factor to airflow limitation on hypoventilation, whereas at protocol including dexmedetomidine was found to have little or no effect on laryngeal motion testing [3, 17]. In this study, we encountered one case of laryngeal paralysis. In cases where the glottis appears as a hole or closed shape on RDVL under BMA sedation, examination with doxapram administration is likely necessary, as in conventional laryngoscopy.

This study had limitations. First, there were a small number of cases, heterogeneity of clinical status and no controlled RDVL in healthy dogs, and the reconstructed virtual endoscope image may have been under- or overestimated because it was not compared to the general endoscope image. Laryngoscopy through the tracheostomy is required to confirm this [4, 30]. In addition, it is unknown whether adductors relax and interfere with airway at the glottis during inspiration in healthy dogs and may be able to demonstrate whether the four shapes classification of glottis are related to the severity of the disease by selecting a more homogeneous group of dogs. Second, laryngeal function was not assessed in same dogs before and after sedation. Finally, the unsedated dogs in the control group were docile and tolerated crate confinement without any issues. However, it is unknown whether psychological factors cause changes in throat movement in dogs. In humans, it has been speculated that the sensation of breathlessness may lead to a depressed state [24]. Therefore, nonsedated dogs may have been breathing with greater effort and attempting to open their glottis more widely. Future work should include more cases, healthy controls, and a direct comparison of RDVL with standard endoscopy. Evaluating the same dogs before and after sedation will help clarify sedative effects on glottic movement. Studies should also examine whether stress or psychological factors influence breathing patterns in nonsedated dogs.

In summary, RDVL provided a reliable view of glottic motion in both sedated and unsedated dogs, indicating that the technique can complement traditional laryngoscopy for evaluating laryngeal function. Since it allows for a more accurate representation of the dog’s physiological state, RVDL without sedation is recommended. However, if the dog has difficulty opening its mouth or if imaging without sedation is difficult, the BMA sedation is preferable. Further studies, including healthy controls and quantitative comparison with standard laryngoscopy, are recommended to validate diagnostic thresholds

CONFLICT OF INTEREST

The authors have no conflicts of interest to declare.