Continuous Rate Infusion of Ketamine Hydrochloride and Dexmedetomidine for Maintenance of Anesthesia during Laryngotracheal Surgery in New Zealand White Rabbits (Oryctolagus cuniculus)

Abstract

New Zealand white rabbits (Oryctolagus cuniculus) are an established in vivo model for the study of structural and functional consequences of vocal-fold vibration. Research design requires invasive laryngotracheal procedures, and the presence of laryngospasms or pain responses (or both) hinder phonation-related data collection. Published anesthesia regimens report respiratory depression and muscle tone changes and have been unsuccessful in mitigating autonomic laryngeal responses in our protocol. Infusion of ketamine hydrochloride and dexmedetomidine hydrochloride in pediatric medicine provides effective analgesia and sedation for laryngotracheal procedures including intubation and bronchoscopy; however, data evaluating the use of ketamine–dexmedetomidine infusion in rabbits are unavailable. This study reports a new infusion regimen, which was used in 58 male New Zealand white rabbits that underwent a nonsurvival laryngotracheal procedure to induce phonotraumatic vocal-fold injury. Animals were sedated by using ketamine hydrochloride (20 mg/kg IM) and dexmedetomidine (0.125 mg/kg IM). Maintenance anesthesia was provided by using continuous rate intravenous infusion of ketamine hydrochloride (343 µg/kg/min) and dexmedetomidine (1.60 µg/kg/min). A stable plane of anesthesia with no autonomic laryngeal response (laryngospasm) was achieved in 32 of the 58 rabbits (55%). Laryngospasms occurred in 25 of 58 animals (43%) and were controlled in 20 cases (80%) by providing 0.33 mL 2% topical lidocaine, incremental increase in infusion rate, or both. Continuous rate infusion of ketamine hydrochloride–dexmedetomidine with prophylactic topical lidocaine provides a predictable and adjustable surgical plane of anesthesia, with minimal confounding respiratory and autonomic laryngeal responses, during extended-duration laryngotracheal surgery in rabbits. This regimen should be considered as an alternative to injection maintenance for prolonged, invasive procedures.

New Zealand white rabbits (Oryctolagus cuniculus) have become a popular model for the investigation of structural, molecular, and functional consequences of hoarseness (dysphonia). Phonotrauma is the leading cause of hoarseness, responsible for 43% of short-term disability claims for voice disorders.¹² Dysphonia affects an estimated 20 to 23 million adults in the United States each year, with an economic burden of approximately $13 billion dollars,¹² thus representing a significant public health concern. Phonotrauma involves excessive or abusive vocal behaviors that lead to damage to the vocal folds. As such, the molecular, structural, and functional changes resulting from phonotrauma have been the focus of ongoing investigation.

Our lab has developed and validated experimental procedures to surgically induce vocal fold phonation that can be visualized endoscopically.²¹ This methodology allows for the concurrent investigation of both the structural, molecular, and functional consequences of phonotraumatic behavior. Rabbits have become an important model for this line of investigation because of their histologic similarity in vocal fold structure to humans⁵⁶ and their lower cost in comparison to larger animal models. Rabbits typically do not vocalize unless they are threatened or in pain (that is, nonhabitual vocalizers); therefore, this model has the notable benefit of allowing investigators to surgically control for both dose magnitude and duration of vocalization experimentally.

Phonotrauma-induced changes and vocal-fold wound healing have been investigated in the rabbit model,^39,40,58 and experimental findings align with those described in human phonotrauma literature.^8,70,71 Studies using the rabbit phonation model have characterized mRNA changes in response to phonotraumatic behaviors,^29,39 described changes elicited in epithelial barrier function,³⁹ and elucidated timelines for the recovery of the vocal-fold epithelium after phonotrauma.⁴⁸ Increased understanding of the molecular and cellular response to phonotrauma will contribute to evidence-based medicine and therapy practices for this condition.

Currently, our group uses a rabbit surgical model to characterize the safety and efficacy of glucocorticoid steroid use for the treatment of phonotraumatic vocal-fold damage. In these experiments, rabbits are anesthetized and surgically phonated by using cricothyroid electrical stimulation and forced, humidified airflow through the glottis for 120 min.²¹ Autonomic laryngeal response (for example, laryngospasm) to upper airway stimulation during general anesthesia is widely reported,^20,42,59 particularly in the presence of upper airway manipulation.^42,59 These autonomic laryngeal responses posed a significant concern during the development of our surgical protocols, due to the considerable negative effect they could have on functional vocal-fold vibration and related downstream biologic processes. Therefore, we needed an anesthetic regimen that would provide depth of anesthesia sufficient both to prevent surgical pain responses (for example, movement, increased heart rate in response to surgical manipulation) and to mitigate break-through, local, autonomic laryngeal responses during sustained phonation.

While multiple studies have reported continuous rate infusion (CRI) of ketamine hydrochloride plus dexmedetomidine hydrochloride for procedural sedation and analgesia in pediatric populations,^{11,24,46,50,69} there is an absence of data regarding this route of administration in rabbits. In this report, we describe the development of a ketamine–dexmedetomidine (KD) CRI anesthetic regimen with topical lidocaine for use in New Zealand white rabbits undergoing extended-duration procedures that require extensive manipulation of the trachea and larynx. Within-procedure anesthetic doses and animal vital signs during this regimen are reported in comparison to normative values for this species. Parameters reported include intraoperative respiratory rate, heart rate, mean arterial blood pressure (MAP), and SpO₂; we also describe the incidence and management of clinically salient changes in vital signs and autonomic laryngeal responses (laryngospasms).

Materials and Methods

Research design.

Data presented were obtained from animal records of surgical procedures that occurred during the development of the KD anesthetic regimen in our laboratory. These data reflect the development and refinement of our anesthetic regimen for our induced phonation procedures.

Animals.

Animal records from 58 male adolescent New Zealand white rabbits (Oryctolagus cuniculus; age, 4 to 6 mo; weight, 3.21 ± 0.20 kg) were reviewed for this study. All procedures for original studies were approved by the Vanderbilt University Medical Center Animal Care and Use Program, a Public Health Service–assured and AAALAC–accredited facility in compliance with the Guide for the Care and Use of Laboratory Animals³⁴ and the Animal Welfare Act.² Rabbits were purchased from Charles River Laboratory (Senneville, Quebec, Canada), where they were bred in an environment free from Encephalitozoon cuniculi, rabbit hemorrhagic disease virus, Pasteurella multiocida, Salmonella spp., Treponema spp., and Clostridium piliforme. Rabbits were ear-tattooed for identification and maintained in a closed colony. The rabbits were single-housed in caging (Allentown, Allentown, NJ) in temperatures ranging from 62 to 70° F (16.7 to 21.1 °C) and in a humidity-controlled (35% to 75%) environment on a 12:12-h on:off light cycle. The daily diet consisted of chow (Laboratory Rabbit Diet HF 5326, Purina, St Louis, MO) and a variety of produce and other food enrichments, with municipal water provided through a drinking valve. The rabbits were acclimated prior to experimental procedures.

Selection of route of administration.

Standard anesthetic approaches for rabbits call for the administration of nebulized inhalational anesthesia agents (for example, isoflurane) through either oral mask or laryngeal mask airway with spontaneous respiration^6,33 or placement of an endotracheal tube with mechanical ventilation.^15,51 However, the complexity of our surgical methods and the future need for animal recovery after the procedure prevented the use of these standard anesthetic regimens and necessitated careful consideration of alternative routes of administration. To surgically induce phonation, controlled airflow must be forced through the glottis as the vocal folds are adducted through electrical stimulation. To introduce airflow through the glottis, the trachea is transected, and the cranial segment is redirected to receive positive airflow from an external air supply (Figure 1). This diverted airway precludes oral administration of inhaled anesthesia, because the resulting airway (caudal aspect of the trachea) bypasses the larynx, nasopharynx, and oral cavity. The nebulized anesthesia agent delivered by tracheostomy with mechanical ventilation is a standard alternative to oral administration, but this approach requires a cuffed endotracheal tube to seal the caudal airway. Lumen pressure created by an inflated cuff can induce tissue degradation over the course of these long procedures, thus potentially introducing complications during tracheal anastomosis in survival procedures.³⁸ Because a cuffed endotracheal tube is required in the cranial tracheal segment to create the sealed subglottis necessary for vocal-fold vibration, we prioritized finding an alternative to isoflurane with mechanical ventilation.

Figure 1.

Procedural setup for an induced-phonation procedure. (A) A close view of the incision site, demonstrating endotracheal tube and electrode placement. (B) Anesthetized rabbits are placed in a supine position on a circulating-water heat mat, and a surgical incision is made to expose the larynx and trachea. Oxygen is supplied directly to the airway through an uncuffed endotracheal tube in the caudal tracheal segment, while humidified air is directed through the cranial tracheal segment via a cuffed endotracheal tube. Custom stainless-steel electrodes are placed in the cricothyroid membrane and muscle to deliver trains of electrical stimulation. A pediatric laryngoscope is used to elevate the glottis such that vocal folds can be visualized, and a cushion is used to support the rabbit's neck and head.

To address the complex anesthesia needs of this protocol, we chose to investigate intravenous CRI for multiple reasons: (1) CRI would not interfere with the surgical site, (2) once achieved, steady-state anesthesia is readily maintained through CRI, and (3) CRI results in predictable effects, because it avoids the absorption-related variability associated with intramuscular maintenance. Furthermore, if negative anesthesia-associated consequences occur, CRI dosing can be quickly and easily adjusted, although the effects of these changes may lag somewhat, depending on the pharmacokinetic properties of the drugs used.

Selection of anesthetic agents.

Drug selection for this protocol was informed through the clinical expertise of our research veterinarian and relevant literature review. Pentobarbital is widely used for intravenous anesthesia in rabbits, but we excluded this drug due to its known respiratory and cardiovascular depression^32,74 and described immunosuppression,¹⁹ which was likely to affect experimental outcomes during the evaluation of vocal-fold injury and subsequent glucocorticoid steroid treatment. Similarly, propofol was not considered for this experiment in light of reported bradycardia and bradypnea associated with higher doses and subtherapeutic anesthesia at lower doses.⁶² Thiopental was excluded due to its tendency to increase heart rate in rabbits,⁴⁹ as well as its minimal analgesic action, which indicated that the drug was unlikely to prevent autonomic laryngeal responses and surgical pain for the duration of the procedure.

Ketamine hydrochloride is a N-methyl-D-aspartate receptor antagonist with combined sedative and dose-dependent analgesic properties and provides fast-acting and long-lasting anesthesia. Ketamine was selected for CRI because (1) its distribution half-life is only 10 min, (2) it does not significantly decrease functional, residual, respiratory capacity, and (3) it does not increase pain sensitivity after dissipation of its analgesic effect.^28,36 However, high doses of ketamine can have unwanted effects, including tachycardia, ventricular arrhythmia,⁴⁴ respiratory depression,^35,63 increased muscle tone,^35,54 suppression of proinflammatory cytokines,⁵⁴ and antagonistic action at opioid receptors.^4,63,72 In addition, multiple reports describe laryngospasm during ketamine anesthesia;^5,9,13 however, the causal relationship between ketamine and laryngospasm is not well defined.⁷² The unwanted side effects associated with high doses of ketamine have precipitated a need to identify drug combinations that result in potent anesthesia at lower doses.^25,41

Dexmedetomidine hydrochloride is a selective α2 adrenergic receptor agonist and was chosen as a promising complement to ketamine in this study. Dexmedetomidine has a rapid onset of action (less than 5 min), with a peak effect within 15 min.³⁷ Several reports demonstrate that the KD drug combination provided safe and effective analgesia, with reduced incidence of the tachycardia, hypertension, and salivation associated with the administration of ketamine only and absence of the bradycardia and hypotension observed with high doses of dexmedetomidine alone.^22,26,37,67 Advantageous to our study addressing the effectiveness of glucocorticoid administration on recovery from phonotraumatic injury, single-dose intramuscular injection of KD in a rabbit resulted in stable serum glucocorticoid levels over a 24-h period, whereas the use of ketamine–buprenorphine, with or without dexmedetomidine, resulted in a decrease in cortisol.²³

Because an autonomic response to laryngeal stimulation (for example, laryngospasm) is a commonly observed occurrence in procedures involving manipulation of the upper airway,^20,42,59 management of these autonomic responses was a driving factor in modifying our anesthetic protocol. Initial procedures using KD alone revealed that additional local anesthetic agents were necessary to mitigate these responses. Lidocaine hydrochloride is the most common topical anesthetic used in the airway.⁵⁵ Lidocaine's initial anesthetic effect onsets at 1 min, with a therapeutic effect lasting as long as 45 min.⁵⁷ When administered to the upper airway, lidocaine has been found to prevent airway protective reflexes (for example, laryngospasm).⁴⁷ Therefore, topical administration of 2% lidocaine solution was added to our protocol, initially to be administered as a therapeutic agent in response to laryngospasm and ultimately as prophylactic agent to prevent laryngospasm.

Sedation.

On the day of the surgery, rabbits were weighed and transferred from the housing room to the operating room by using a covered pet carrier. Rabbits were immobilized in a sternal recumbent position, supplied with blow-by-nose oxygen, and sedated by injecting ketamine (20 mg/kg IM; Vedco, St Joseph, MO) and dexmedetomidine (0.125 mg/kg IM; Zoetis, Parsippany, NJ) into either the longissimus or illicostalis muscle. Sedation was determined by the following indicators: absence of pedal withdrawal reflex in response to toe pinch, stable reduced frequency of respiration (reduced from more than 150 breaths per minute to 30 to 60 evenly paced breaths per minute), reduction of smooth muscle tone, and loss of righting reflex when placed in a lateral recumbent position.

Maintenance anesthesia.

Once sedation was confirmed, rabbits were transferred to a supine position on a circulating water heat pad coupled to a heat therapy pump (model HTP-1500, Adroit Medical Systems, Loudon, TN) to maintain body temperature, and the forelimbs were gently lateralized and restrained. Monitoring devices were attached after sedation but prior to the initiation of maintenance CRI anesthesia; indirect blood pressure was measured by using a Cardell multiparameter monitor (Midmark, Dayton, OH), with a size 2 cuff secured around the right hindleg; a probe was placed on a digit of a hindleg for monitoring SpO₂, heart rate was assessed by using a digital system (V5 Vet View, Scil Veterinary Excellence, Gurnee, IL), temperature was measured by using a rectal thermometer, and respiratory rate was determined visually. Heart rate and SpO₂ were monitored continuously and reported every 15 min. Temperature, blood pressure, and respiratory rate were monitored and recorded at 15-min intervals for the duration of the procedure. Saline-flushed 24-gauge catheters (Surflash IV, Terumo, Tokyo, Japan) were inserted into the marginal vein of each ear. One catheter was used to provide warmed 0.9% sodium chloride through a 70-in. macrobore primary intravenous line (Zoetis) at a rate of 30 drops per minute (30 mL/h). The second catheter was used to supply maintenance intravenous KD anesthesia by CRI (ketamine, 200 to 1000 µg/kg/min; dexmedetomidine, 0.65 to 3.0 µg/kg/min). Separate syringe infusion pumps (model AS50m Baxter, Chicago, IL) were used for each drug for the fidelity of dose adjustment. A surgical plane of anesthesia was determined by absence of a corneal or laryngeal response under maintenance of independent diaphragmatic breathing.

Surgical procedures.

The nonsurvival phonation surgical procedure has previously been described.²¹ After confirmation of a surgical plane of anesthesia, the chest and neck were clipped to expose the skin from the sternal notch to the mandible. Lidocaine hydrochloride 2% (Hospira, Lake Forest, IL) was injected subcutaneously along the midline, to provide local anesthesia. A surgical incision was made from sternum to submentum, and the fascia and muscle were dissected away to expose the larynx and trachea. To maintain a constant supply of oxygen for spontaneous respiration, the trachea was transected approximately 4 cm below the larynx, and oxygen was supplied directly to the caudal segment of the trachea by using an uncuffed endotracheal tube. A cuffed endotracheal tube was placed in the upper tracheal segment, approximately 1cm below the larynx, and the cuff was inflated to seal the upper airway (Figure 1 A).

To induce vocal-fold vibration, the vocal folds must be partially or completely adducted in the presence of transglottal airflow. A ConchaTherm Neptune system (Hudson RCI, Temecula, CA) was used to deliver humidified air to the upper airway, forcing air through the glottis at a controlled rate. The vocal folds were medialized into the airstream by using electrical stimulation of intrinsic laryngeal muscles. Custom, stainless-steel, hooked electrodes were inserted into the cricothyroid membrane and the belly of the cricothyroid muscle bilaterally to produce adduction of the vocal folds (Figure 1 A). Electrical stimulation of the larynx was achieved by using a Grass S88 stimulator (SA Instrumentation, Encinitas, CA) and constant current isolation unit (model PSIU6, Grass Telefactor, West Warwick, RI).

For visual evaluation of the vibratory consequences of prolonged phonation, a pediatric laryngoscope (Karl Storz Endoscopy, El Segundo, CA) was used to suspend the larynx for vocal-fold visualization, and the neck was supported by using a Vac-Lock cushion (CIVCO, Coralville, IA). Images of vocal-fold vibration were captured by using a 0°, 4.0-mm rigid endoscope (KayPENTAX, Montvale, NJ) coupled to a monochrome high-speed (8000 frames per second) camera (FASTCAM MC 2.1, KayPENTAX) and 300-W continuous xenon light source (Karl Storz Endoscopy; Figure 1 B). After 120 min of surgically induced phonation, rabbits were euthanized by intravenous injection of 780 mg sodium pentobarbital–phenytoin (Virbac, Carros, France), and the larynges were harvested for assessment of structural and molecular changes resulting from prolonged phonation.

Maintenance anesthesia doses were initiated toward the low end of the range (mean initiation dose: ketamine, 331.3 ± 21.5 µg/kg/min; dexmeditomidine, 1.54 ± 0.38 µg/kg/min; Table 1) and were modified as required to provide a stable plane of anesthesia with minimal autonomic reflexes. Anticipated adverse events in this procedure included surgical pain response, autonomic laryngeal response (laryngospasm), reduced respiratory rate (fewer than 30 breaths per minute),³ and bradycardia (below 130 bpm).³ When autonomic laryngeal response occurred, the larynx first was treated with 0.33 mL of 2% topical lidocaine. If symptoms did not abate within 2 min, CRI maintenance anesthesia doses were increased in a step-wise fashion (increases of 50 µg/kg/min for ketamine and 0.5 µg/kg/min for dexmeditomidine) at 5-min intervals, with observation for symptom abatement. When monitored parameters increased, CRI maintenance anesthesia doses were increased again as described, followed by the administration of lidocaine as needed. Conversely, when anesthesia side effects such as bradycardia or significantly reduced respiratory rate with a concomitant drop in SpO₂ occurred, the CRI maintenance anesthesia dose was decreased in a step-wise fashion until symptom abatement.

Table 1.

Vital signs before and after and median change during transition from sedation to intravenous anesthesia maintenance

	Before IV anesthesia		After IV anesthesia		Sedation to anesthesia change			Effect size
	Median	IQ	Median	IQ	Median	Range	P	rs	P
Body temperature (°C)	40.0	0.5	39.90	0.52	0	−0.60 to +0.40	0.0182	0.9282	<0.0001
HR (bpm)	157.0	30.3	158.5	27.0	4	−19.00 to +52.00	0.0015	0.8439	<0.0001
RR (breaths/min)	48	22	42	16	−4	−40.00 to +48.00	0.0199	0.6026	<0.0001
SpO2 (%)	98.5	1.0	99.0	1.0	0	−7.00 to +7.00	0.1998	0.2701	0.0202
MAP (mm Hg)	82.33	16.00	82.17	21.25	3.67	−50.33 to +49.00	0.0138	0.6141	<0.0001

HR, heart rate; IQ, interquartile range; MAP, mean arterial pressure; RR, respiratory rate

Significant P values (that is, P < 0.05) are bolded.

Statistical analysis

Procedural durations are reported as mean ± 1 SD. All vital signs data are reported as median ± interquartile range and compared with published normative physiologic values for New Zealand white rabbits.^{3,10,18,27,45} Statistics were calculated by using Prism version 8.2.0 (GraphPad Software, San Diego, CA). Baseline and final vital signs were compared by using the Wilcoxon matched-pair signed-rank test, with Spearman correlation to calculate effect size. Initial and final doses of CRI were compared by using the Wilcoxon matched-pair signed-rank test. The effect of transitioning from sedation to maintenance anesthesia was assessed; vital signs before and after the introduction of intravenous anesthesia were compared by using the Wilcoxon matched-pair signed-rank test, with Spearman correlation to calculate effect size. The χ² test of independence was used to evaluate the relationship between prophylactic administration of lidocaine and the presence of laryngospasms. Significance was defined as a P value less than 0.05.

Results

Sedation was of sufficient depth, with no autonomic laryngeal responses or surgical pain responses observed at the time of catheterization or surgical incision. Time from sedation to the start of intravenous maintenance anesthesia start was 34 ± 10 min (mean ± 1 SD). The transition between sedation and anesthesia was smooth, with no observed pain responses or autonomic laryngeal responses. Minimal changes in the following vital signs were observed between before and after the introduction of intravenous anesthesia (Table 1): body temperature decreased from 40 ± 0.5 to 39.9 ± 0.5 °C (P = 0.0182), heart rate increased from 157 ± 30 to 158.5 ± 27 bpm (P = 0.0015), respiratory rate decreased from 48 ± 22 to 42 ± 16 breaths per minute (P = 0.0199), and MAP decreased from 82 ± 16 to 82 ± 21 mm Hg (P = 0.0138). A large effect size was identified for all 4 of these variables (r_s = 0.9282, 0.8439, 0.6026, and 0.6141, respectively; P < 0.0001 for all 4 variables), indicating a strong positive linear relationship between vital signs before and after transition to intravenous anesthesia. No significant difference between before and after transition to intravenous anesthesia introduction was found for SpO₂ (P = 0.2701).

The average procedure duration from sedation to termination was 221 ± 80 min. Mean vital signs at 15-min intervals for the duration of the procedure are shown in Figure 2, and summary vital signs are reported in Table 2. Heart rate increased from an average of 146 bpm at baseline to 163 bpm at surgery termination (P < 0.0001), with a procedural average of 164 bpm and moderate pairwise correlation (r_s = 0.3198). SpO₂ increased from 97% to 99% over the duration of sedation and anesthesia (P < 0.0001), with a procedural average of 98% and a small pairwise correlation (r_s = 0.2478). The respiratory rate decreased throughout the procedure, from 48 breaths per minute at baseline to 34 breaths per minute at surgery termination (P < 0.0001), with a procedural average of 37 breaths per minute and moderate effect size (r_s = 0.309). Body temperature was 39.6 °C at baseline and decreased throughout the procedure, to 38.2 °C at surgery termination (P < 0.0001), with a procedural average of 38.1 °C and moderate effect size (r_s = 0.3149). MAP fell from 80 mm Hg at baseline to 67 mm Hg at procedure termination, with a procedural average of 70 mm Hg and a moderate effect size (r_s = 0.4916). Baseline, final, and procedural averages for heart rate, respiratory rate, and temperature were within published normal limits for New Zealand white rabbits.^{3,10,18,27,45}

Figure 2.

Vital signs variables recorded from sedation (n = 58, time, 0 min) to termination of the surgical procedure. (A) Heart rate, SpO₂, and respiratory rate. (B) Mean arterial blood pressure. (C) Rectal body temperature. All values are reported as median ± interquartile range.

Table 2.

Vital signs at procedure start and termination, procedural change, and normative reference values

	Procedure (median [IQ])			Median change over procedure duration		Effect size
	Mean	Start	End		P	rs	Pairing effect P	Normative value
Body temperature (°C)	38.07 (1.33)	39.65 (0.50)	38.25 (1.00)	−1.4	<0.0001	0.3149	0.008	38-403
HR (bpm)	163.5 (13.6)	146.0 (19.5)	162.5 (34.5)	15.5	<0.0001	0.3120	0.0072	130-3253
RR (breaths/min)	37.04 (5.06)	48 (20)	34 (10)	−14	<0.0001	0.309	0.0091	30-603
SpO2 (%)	98.45 (0.66)	97.00 (4.25)	99 (1)	1	<0.0001	0.2478	0.0304	>9718
MAP (mm Hg)	69.69 (12.02)	80.33 (19.50)	66.83 (24.00)	−12.83	<0.0001	0.4916	<0.0001	60-9710,27,45

HR, heart rate; IQ, interquartile range; MAP, mean arterial pressure; RR, respiratory rate

Significant P values (that is, P < 0.05) are bolded.

The duration of CRI anesthesia was 187 ± 81 min (Table 3). Ketamine was initiated at an average starting dose of 331 µg/kg/min and increased to a final dose of 372 µg/kg/min (P < 0.0001), with a procedural average of 343 µg/kg/min. Dexmedetomidine was initiated at an average starting dose of 1.5 µg/kg/min and increased to a final dose of 1.8 µg/kg/min (P < 0.0001), with an average procedural dose of 1.6 µg/kg/min. Eight rabbits demonstrated acute changes in heart rate within a 15-min interval (mean increase, 52 bpm). One animal demonstrated an acute increase in respiratory rate, from 60 to 80 breaths per minute. These animals were all managed by increases in CRI, with heart rates stabilizing and respiratory rate returning to a physiologic level within 15 min.

Table 3.

CRI dosage at procedure start and termination and change in dosage

		Initial CRI dosage (µg/kg/min)			Final CRI dosage (µg/kg/min)			Change over procedure duration (µg/kg/min)		Effect size
	Induction dose (mg/kg)	Mean	95% CI	Range	Mean	95% CI	Range	Mean	P	rs	P
K	20	331.25	321.00–342.80	200–350	372.41	346.3–398.5	250–1000	41.16	<0.0001	0.2374	0.0364
D	0.125	1.54	1.435–1.639	0.50–2.15	1.76	1.641–1.890	0.65–3.00	0.22	<0.0001	0.5345	<0.0001

D, dexmeditomidine; K, ketamine

Significant P values (that is, P < 0.05) are bolded.

Laryngospasm events and their management are summarized in Figure 3. To minimize the occurrence of autonomic laryngeal responses, lidocaine was administered prophylactically in 31 of the 58 subjects (53%); 22 of these 31 animals (70%) experienced no laryngospasm, and the use of prophylactic lidocaine significantly (P = 0.009) reduced the likelihood of adverse events. Of the 9 animals that experienced adverse events despite prophylactic lidocaine, 8 had autonomic laryngeal responses, and one underwent cardiac arrest, which later was determined to be due to a congenital heart defect. Autonomic laryngeal response was successfully treated with lidocaine alone in 1 of 8 cases (12.5%) and with lidocaine plus CRI dosage increase in 5 of the 8 cases (62.5%). Among the 27 rabbits that did not receive prophylactic lidocaine, 17 (63%) experienced laryngospasm; these laryngospasms were managed successfully with lidocaine alone in 6 cases (35%), with increased CRI dosage alone in 1 case (6%), and with lidocaine plus CRI dosage increase in 7 cases (41%). In 5 subjects, autonomic laryngeal responses could not be resolved by using any of the management techniques described. Two of these 5 animals had received prophylactic lidocaine. The starting dosages of ketamine and dexmedetomidine did not have statistically significant or clinically salient effects on the manifestation of autonomic laryngeal responses (P = 0.199 and P = 0.527, respectively).

Figure 3.

Frequency of occurrence and applied interventions for periprocedural autonomic laryngeal responses.

Discussion

This study demonstrates that intramuscular injection of ketamine–dexmedetomidine for sedation followed by CRI infusion of ketamine–dexmedetomidine for maintenance, combined with prophylactic topical lidocaine, provides sufficient anesthesia to manage autonomic laryngeal responses and surgical pain responses during invasive laryngotracheal surgery in rabbits. All 58 animals in this study were sedated within 10 min of administration of the intramuscular injection; we anecdotally observed this rate to be slower than in previous experiments using sedation with ketamine–acepromazine–xylazine. Autonomic laryngeal responses were successfully managed in all but 5 cases, supporting the use of this anesthetic regimen for procedures like this one.

KD sedation was successful and well tolerated in all animals. Although ketamine is known to elicit tachycardia and bradypnea, among other side effects, coadministration with the α2-adrenergic agonist dexmedetomidine has been demonstrated to minimize respiratory and cardiovascular effects, providing a stable plane of anesthesia without affecting endogenous steroid levels.²³ The effects of dexmedetomidine, including increased sedation, dose-dependent bradycardia, bradypnea, and decreased SpO₂,⁷⁵ result from α₂-adrenoreceptor agonism⁶⁶ and appear to negate the negative symptoms associated with ketamine; we noted these effects in the current regimen, given that all vital signs remained within normative limits for the duration of anesthesia. Moreover, an antagonist (atipamezole) is available for dexmedetomidine, and dexmeditomidine has a dose-sparing effect for many anesthesia drugs,^7,64,65 reducing the total volume of anesthetic agent administered during procedures and likely improving recovery outcomes—important considerations for future survival procedures.

Significant changes in vital signs were observed from sedation to procedure termination for all parameters measured in the study. Consistent with reports that dexmedetomidine causes decreased respiratory rate in a nonlinear dose-dependent manner,⁵² we noted a modest reduction in the respiratory rate between sedation and study termination; however, respiratory rate remained within normal physiologic limits³ and, given that SpO₂ remained within expected levels for the study duration, the decreased respiratory rate was not indicative of pathologic change. Similarly, increases in heart rate and decreases in rectal temperature were observed over the course of the procedure but remained within physiologically normal limits. Acute increases in heart rate (greater than 50 bpm in15 min) and respiratory rate (more than 20 breaths per minute in 15 min) were ameliorated through stepwise increases in CRI. In addition, MAP decreased over study duration but remained within physiologically normal limits.^10,27,45 These vital signs suggest that a KD anesthetic regimen provides similar respiratory rate, SpO₂, heart rate, and rectal temperature but a more stable MAP compared with a 4-h intravenous ketamine–xylazine anesthetic regimen.⁷³ Furthermore, MAP averages were consistent with a combined protocol of inhaled isoflurane with intramuscular injection of either ketamine–xylazine–buprenorphine or ketamine–medetomidine–buprenorphine.¹⁶ These findings indicate that a CRI KD anesthetic regimen with topical lidocaine produces vital sign stability comparable with other intravenous, inhaled, or combination anesthetic regimens.

Consistent with previous studies evaluating ketamine and dexmedetomidine,⁴⁶ coadministration of these agents appears to minimize the cardiorespiratory effects associated with the independent use of each agent, and body temperature was regulated within normative values over an extended-duration surgical procedure.^{3,10,18,27,45} Of note, within the current anesthesia regimen, weak to moderate correlation was observed between baseline and final vital signs, whereas the pairing effect was significant across all measured variables, suggesting that although individual animals had different physiologic baselines, animal-specific vital signs remained stable and within physiologic norms for the duration of the procedure.

The findings of this study supporting the use of intravenous CRI of KD with topical lidocaine are promising; however, limitations within the methodology of this study warrant consideration. We chose to use only male rabbits in these experiments to control for known differences in laryngeal architecture and functional vocal-fold vibration between sexes; however, this methodology excluded analysis of sex as a biologic variable in this anesthesia regimen. Previous studies have identified sex-associated differences in ketamine anesthesia: female rabbits of a mixed breed backgrounds that received injection anesthesia ketamine and medetomidine required more frequent supplementation with isoflurane than male rabbits,⁵³ and studies have identified that females are more susceptible to the antidepressant effects of low-dose ketamine.⁶¹ The nonrandomized approach to treatment with prophylactic topical lidocaine is a shortcoming of this study. Although the introduction of prophylactic lidocaine partway through the study reduced incidence of autonomic laryngeal response, consistent with findings in previous studies and meta-analyses,¹⁴ the lack of blinding prevented analysis of the risk factors for developing autonomic laryngeal responses. Furthermore, dose selection in our study was driven by the need to maintain sufficient anesthesia depth to minimize laryngeal responses. Consequently, we did not seek to evaluate excessive anesthesia unless negative consequences (apnea, decreased heart rate) occurred. As such, the implementation of this approach mainly resulted in unidirectional anesthesia adjustment. Therefore, this study does not identify an effective upper dosing limit or a mean effective dose. Finally, vital signs were recorded at relatively long intervals (15 min), which may result in masking of negative anesthesia effects in subjects that is experienced an unstable plane of anesthesia.

Given the continuous strain placed on the tissue during this procedure, local autonomic responses such as laryngospasm are common. With the regimen evaluated in this study, 55% of animals did not exhibit autonomic laryngeal responses, and the majority of these (21, 66%) had received prophylactic lidocaine to the larynx at 40-min intervals, consistent with the duration of lidocaine's therapeutic effect.⁵⁷ Although CRI starting dose did not influence whether animals experienced involuntary responses, the administration of topical prophylactic lidocaine did reduce the likelihood of autonomic reflexes. However, prophylactic lidocaine did not eliminate autonomic laryngeal responses, suggesting that interindividual differences may contribute to the presentation of these responses during this stressful procedure.

Autonomic laryngeal response was the leading complication in this study, affecting 25 of 58 animals. The presence of these reflexes is not unexpected, given the extensive manipulation of the larynx and proximal areas, and has occurred in previous studies from our group. Nevertheless, the current anesthesia regimen was successful in managing these reflexes in 80% (20 of 25) of these animals. For other procedures that do not require such extensive upper airway manipulation, this anesthetic regimen likely would adequately maintain a surgical plane of anesthesia in the vast majority of rabbits.

Insufficient depth of anesthesia and superior laryngeal nerve activation are the most likely causes of intraoperative laryngospasm.³¹ We were unable to elucidate why the regimen described was unsuccessful in controlling autonomic laryngeal responses in 5 cases, but it is possible that these animals were particularly sensitive to stimulation from laryngoscope placement and endoscopy or that these animals were more resistant to the effects of the anesthesia regimen. Deepening of anesthesia by using either propofol or midazolam has been demonstrated as effective in overcoming intraoperative laryngospasm;⁶⁰ however, these methods were not pursued due to concerns for bradycardia with propofol⁶² and airway obstruction with midazolam.¹⁷ The application of positive pressure was not pursued, given that supplemental oxygen bypassed the larynx and was administered through an uncuffed tube to the aboral portion of the trachea. We did not consider the use of the Larson maneuver,⁴³ applying pressure to the ‘laryngospasm notch’ while performing a jaw thrust, because to our knowledge this technique has not been performed in rabbits and is contraindicated for our protocol because it would require removal of the laryngoscope. Other potential therapeutic avenues, such as the administration of the muscle relaxant succinylcholine,^1,59 were not considered because moderate tone of the thyroarytenoid and cricothyroid muscles is required for normal phonation.^30,68

Overall, analgesia was most successful with the administration of prophylactic topical lidocaine with adjuvant CRI dosage increase plus additional topical lidocaine as needed. In the single animal that experienced cardiac arrest, postmortem necropsy revealed a congenital ventricular septal defect, which likely contributed to this outcome; however, other factors cannot be excluded.

In conclusion, we describe a regimen for CRI of ketamine–dexmedetomidine, with topical lidocaine, that provides an adequate surgical plane of anesthesia for invasive laryngotracheal procedures in New Zealand white rabbits. Furthermore, this regimen can easily be adjusted to manage involuntary reflexes and mitigate complications of underdosing. Optimal surgical outcomes were observed with induction through 20 mg/kg IM ketamine hydrochloride and 0.125 mg/kg IM dexmedetomidine, followed by CRI intravenous maintenance anesthesia of 300 to 350 µg/kg/min ketamine hydrochloride and 1.5 to 1.75 µg/kg/min dexmedetomidine, with prophylactic topical 2% lidocaine administered every 40 min. Adverse events were best managed with a combination of CRI dosage increase paired with additional application of 2% topical lidocaine.

Circumventing known difficulties in the provision of rabbit anesthesia, such as challenges intubating for ventilated anesthesia and variable anesthesia plane with anesthesia maintenance through intramuscular injection, this method represents an easily implementable and reliable alternative to existing anesthesia techniques and facilitates superior animal welfare and experimental rigor. In addition, this regimen has previously been demonstrated to have no effect on endogenous steroid levels,²³ positioning this anesthesia protocol as a strong option for studies investigating injury, repair, and inflammation pathways. Further experiments are in progress to evaluate postprocedural recovery with this anesthesia regimen.