Minimally Invasive Murine Laryngoscopy for Close-Up Imaging of Laryngeal Motion during Breathing and Swallowing

Abstract

The larynx is an essential organ in mammals with three primary functions - breathing, swallowing, and vocalizing. A wide range of disorders are known to impair laryngeal function, which results in difficulty breathing (dyspnea), swallowing impairment (dysphagia), and/or voice impairment (dysphonia). Dysphagia, in particular, can lead to aspiration pneumonia and associated morbidity, recurrent hospitalization, and early mortality. Despite these serious consequences, existing treatments for laryngeal dysfunction are largely aimed at surgical and behavioral interventions that unfortunately do not typically restore normal laryngeal function, thus highlighting the urgent need for innovative solutions. To bridge this gap, we have been developing an experimental endoscopic approach to investigate laryngeal dysfunction in murine (i.e., mouse and rat) models. However, endoscopy in rodents is quite challenging due to their small size relative to current endoscope technology, anatomical differences in the upper airway, and the necessity for anesthesia to optimally access the larynx. Here, we describe a novel transoral laryngoscopy approach that permits close-up, unobstructed video imaging of laryngeal motion in mice and rats. Critical steps in the protocol include precise anesthesia management (to prevent overdosing that abolishes swallowing and/or risks respiratory distress-related mortality) and micromanipulator control of the endoscope (for stable video recording of laryngeal motion by a single researcher for subsequent quantification). Importantly, the protocol can be performed over time in the same animals to study the impact of various pathological conditions specifically on laryngeal function. A novel advantage of this protocol is the ability to visualize airway protection during swallowing, which is not possible in humans due to epiglottic inversion over the laryngeal inlet that obstructs the glottis from view. Rodents therefore provide a unique opportunity to specifically investigate the mechanisms of normal versus pathological laryngeal airway protection for the ultimate purpose of discovering treatments to effectively restore normal laryngeal function.

Figure 1:. Murine endoscopy platform.

(A) Side and (B) top views of the custom murine endoscopy platform are shown, with essential components labeled. Note the tabletop beneath the heating pad is size-adjustable. Shown here is the tabletop and heating pad sizes used with rats, which are easily removed to expose a mouse-sized tabletop that accommodates a smaller heating pad (not shown). A custom adapter secures an endoscope to a micromanipulator that is attached to the platform base. This strategic design allows the entire platform to be moved as a unit during the endoscopy procedure, without risking injury to the animal from inadvertent/uncontrolled endoscope motion. The micromanipulator permits gross and micro adjustments of the endoscope tip in multiple directions, including x (left/right), y (forward/back), z (up/down), as well as rotation around y (pitch) and z (yaw).

Figure 2:. Otoscope and custom sheath for murine laryngoscopy.

(A) Disassembled components of a commercial otoscope and custom stainless-steel sheath with adapter for murine laryngoscopy. (B) When assembled, the otoscope tip extends 1 mm beyond the metal sheath but is adjustable up to 5 mm as needed. This strategic design facilitates advancement of the narrow otoscope tip into the rodent’s laryngeal inlet while the slightly larger diameter (2.4 mm) metal sheath sufficiently holds the velum and epiglottis open for optimal visualization of the entire larynx during breathing and swallowing.

Figure 3:. Minimally invasive electrophysiological recording during endoscopy.

A respiratory sensor is taped to the rodent’s abdomen; an EMG electrode is inserted through the skin into the genioglossus muscle of the tongue; and a ground electrode is inserted subcutaneously at the hip. This approach permits investigation of swallowing, breathing, and swallow-breathing coordination in synchrony with endoscopy. Note the skin is shaved and cleaned/disinfected at the electrode insertion sites. Yellow star = aluminum foil wrapped around the electrode lead connection sites to improve signal-to-noise ratio in the electrophysiological recordings.

Figure 4:. Transoral endoscopy to visualize the larynx from a distance.

(A) After gently retracting the tongue with a light finger grip, the endoscope is inserted between the tongue and central incisors at the red star location (i.e., same side as retracted tongue to maintain anatomical alignment with the endoscope shaft). (B) As the endoscope is advanced past the hard palate, (C) the epiglottis and velum come into view. (D) To visualize the glottis, the velum and epiglottis must be “decoupled” by applying pressure against the surface of the velum (at the location of the black outlined star in image C).

Figure 5:. Close-up endoscopic visualization of the larynx.

(A) The endoscope tip is gently guided via micromanipulator control between the decoupled velum and epiglottis (at the location of the black outlined star). As the endoscope advances, (B) the larynx comes into view and the glottal space (yellow star) is centered in the camera field of view via micromanipulator adjustments. (C) Continued micromanipulator advancement of the endoscope results in visualization of the entire ventral-dorsal and lateral dimensions of the larynx. Abbreviations: VC = ventral commissure of the larynx (i.e., ventral junction point between the vocal folds); DC = dorsal commissure of the larynx (i.e., dorsal junction point between the arytenoids); VFs = vocal folds; A = arytenoid.

Figure 6:. Visualization of the murine larynx during breathing and swallowing.

Representative endoscopic images depicting laryngeal motion during breathing and swallowing in an adult Sprague Dawley rat (A–C) before and (D–F) after surgical transection of the right RLN. Note that resting posture of the larynx appears unchanged (D) following RLN injury compared to (A) baseline. (B,E) During maximum inspiration, laryngeal asymmetry becomes obvious following RLN injury. Instead of both arytenoids abducting to enlarge the glottal space (yellow star), (B) as shown at baseline, (E) the ipsilateral (right) arytenoid (black asterisk) and vocal fold appear immobilized throughout the respiratory cycle following RLN injury. Right-sided asymmetry is also evident during swallowing. (C) At baseline, the arytenoids approximate at midline during swallowing, leaving a small ventral glottal gap between the vocal folds. (F) Following RLN injury, the ipsilateral arytenoid and VF move paradoxically (i.e., in the same direction as the unaffected side, red arrow) during swallowing, leaving a large glottal gap (yellow star) extending from the ventral to posterior laryngeal commissures. (F) This image provides direct evidence of impaired laryngeal airway protection in a rat model of iatrogenic RLN injury. (C,F) Note the larynx moves closer to the endoscope during swallowing, as indicated by the epiglottis and velum no longer being visible in the camera field of view. Black arrows indicate the direction of normal laryngeal motion whereas the red arrow indicates paradoxical motion; yellow star = glottal space. Abbreviations: VFs = vocal folds; A = arytenoid; RLN = recurrent laryngeal nerve.

Figure 7:. Using serial laryngoscopy to investigate laryngeal dysfunction during breathing and swallowing in a rat model of iatrogenic RLN injury.

A Likert scale ranging from −2 to +2 was used to estimate laryngeal motion distance and direction in eight adult Sprague-Dawley rats over a 4 month period. After baseline laryngoscopy, the rats underwent a surgical procedure to transect the right RLN, followed by serial laryngoscopy at 1 week post-surgery, then again at 1 month intervals from 1 to 4 months post-surgery. All eight rats survived the procedure, thus demonstrating the effectiveness of our anesthesia regimen for serial laryngoscopy. (A) Videos were analyzed in real time and frame-by-frame/slow motion to quantify laryngeal motion during breathing, where 0 = no motion, 1 = some motion, and 2 = normal motion distance of the affected (right) side compared to the intact (left) side. (B) For swallowing, the glottal gap size was estimated as follows: 0 = no reduction in the glottal gap size (i.e., no laryngeal airway protection), 1 = some glottal gap reduction (i.e., incomplete airway protection), and 2 = complete adduction of the arytenoids, with only a small ventral glottal gap between the vocal folds (i.e., complete airway protection). Negative values for breathing and swallowing indicate laryngeal motion in the opposite direction than expected (i.e., paradoxical). Note that following RLN injury, both breathing and swallowing were negatively affected. Interestingly, laryngeal airway protection was complete (albeit paradoxical) at the 1 WPS timepoint but worsened thereafter, ranging from no protection to incomplete protection. Abbreviations: WPS = week post-surgery; MPS = months post-surgery; RLN = recurrent laryngeal nerve.

Figure 8:. Swallowing inhibited by ISO in rodents.

(A) Image of a rodent undergoing laryngoscopy under ISO anesthesia, with labeled components of the custom ISO delivery system designed for this purpose. A major caveat of this innovative approach is the risk of personnel exposure to ISO. (B) Another downside to this approach is ISO suppression of swallowing. This side-by-side boxplot and scatterplot summarizes unpublished data comparing the effect of ISO versus KX anesthesia in mice (9 per group) undergoing direct electrical stimulation of the right superior laryngeal nerve to evoke swallowing. Shown here is the number of swallows evoked during a 5 min trial consisting of 20 s trains of 20 Hz stimulation followed by 10 s of rest. Compared to KX, mice anesthetized with ISO (as low as 2%) had significantly fewer swallows (p < 0.001, independent samples t-test), and swallowing was even abolished in 4/9 mice. Similar findings emerged from non-surgical experiments with both mice and rats (data not shown). Abbreviations: ISO = isoflurane; KX = ketamine-xylazine.

Figure 9:. Objective quantification of murine laryngeal motion using tracking software.

The same images from Figure 6 showing breathing versus swallowing in a rat at baseline versus post RLN injury are shown here, with laryngeal motion tracking lines added by our custom software. Tracking lines were manually added to the first video frame along the medial border of the arytenoids for automated tracking of left (blue line) versus right (red line) laryngeal motion in the remaining video frames. Corresponding laryngeal motion graphs generated by our custom software from 2.5 s video clips show individual left/right motion versus derived global laryngeal motion, with labels corresponding to (A,D) laryngeal resting posture, (B,E) maximum glottal gap during inspiration, and (C,F) glottic closure during swallowing. Note the paradoxical motion of the right side (red arrows) post RLN injury, as well as the large glottal gap shown in the corresponding derived global motion graph. Representative outcome measures are included in Table 1. Abbreviation: RLN = recurrent laryngeal nerve.

Figure 10:. Electrophysiology-based outcome measures for correlation with laryngoscopy data.

(A) Electrophysiology recordings during breathing and swallowing are shown for a healthy rat. The top window shows a respiratory trace (from a respiratory sensor taped to the rodent’s abdomen), the middle window shows EMG activity in the genioglossus muscle, and the bottom window shows the filtered EMG activity. Note the rhythmic respiratory and EMG pattern during breathing, which is interrupted during swallowing events. Swallow events are readily detected via jagged motion in the respiratory trace (black arrows) that is immediately followed by brief apnea (red asterisk). (B) An expanded window of the dashed rectangular box in A shows how several outcome measures are quantified from the electrophysiological recordings. (A) Note that during inspiration (yellow panels), the respiratory trace (top window) is delayed ^~150 ms (blue double arrow) compared to EMG bursting activity, which highlights temporal differences between the two electrophysiological methods. Representative electrophysiology-based outcome measures include: 1) inspiratory phase duration (i); 2) inter-respiratory-interval (ii, calculated via the respiratory and filtered EMG channels); swallow area under the curve (iii); and swallow apnea (iv; calculated via the respiratory and filtered EMG channels). Abbreviation: EMG = electromyography.