The Laryngoscope VC 2009 The American Laryngological, Rhinological and Otological Society, Inc. Observations of Recurrent Laryngeal Nerve Injury and Recovery Using a Rat Model Belachew Tessema, MD; Rick M. Roark, PhD; Michael J. Pitman, MD; Philip Weissbrod, MD; Sansar Sharma, PhD; Steven D. Schaefer, MD Objectives/Hypothesis: To evaluate standar- motor system. This model is useful to evaluate the ef- dized recurrent laryngeal nerve (RLN) injuries using ficacy of systemic and local neurotropic agents in the a rat model via minimally invasive transoral electro- treatment of RLN injury. myography (ToL EMG) and histologic studies. Key Words: RLN injury, rat model, histology, Methods: Forty-two female Sprague Dawley electromyography. rats weighing 200 g to 250 g underwent crush injury Laryngoscope, 119:1644–1651, 2009 to the right RLN using a calibrated pressure clip (0.61 N or 1.19 N) for 60 seconds. Following injury, serial ToL EMGs were performed on abductor and INTRODUCTION adductor laryngeal muscles during respiratory cycles The recurrent laryngeal nerve (RLN) is sometimes and spontaneous vocal fold abduction on day 4 and injured during surgeries of the head and neck region. then weekly for 6 weeks. Vocal fold motion associated Prognosis of functional recovery is challenging and often with spontaneous respiration was graded from 0 to 4. innacurate.1 Recovery from RLN injury is dependent on Rats were sacrificed at different time points for histo- the mode and extent of injury. As our knowledge of the logic evaluation of injured nerves. biology of peripheral nerve regeneration increases, there Results: EMG signals showed fibrillation poten- tials on day 4 in all experimental conditions. Crushed has been increasing interest in developing models of RLN, regardless of force, exhibited polyphasic poten- nerve regeneration that are pertinent to specific clinical tials at 2 weeks postinjury. Normal motor unit poten- questions. Succesful return of normal function is vari- tials and recruitment patterns were observed in EMG able. Neurotrophic factors are promising but, so far, signals at 4 weeks for all 0.61 N clip animals. Six have inconsistencies in the results. A highly focused weeks following crush injury, motor unit potentials study with predictable results is needed to effectively having normal appearance were observed in most ani- evaluate trophic factors. Various experimental animal mals. Synkinetic EMG signals were observed at 5 models have been used to study peripheral nerve injury weeks and 6 weeks in the 1.19 N clip animals. Endo- and functional recovery.2–6 Recently, several studies scopic evaluation of vocal fold mobility was consis- have used the rat larynx to evaluate potential agents tently normal at 6 weeks only following 0.61 N crush 7–11 injury. that promote recovery of the RLN following injury. Conclusions: This model is useful to simulate Aneurysm clips provide a consistent and reliable intraoperative RLN injuries and to better understand crush injury to peripheral nerves with a calibrated force 12,13 the electrophysiologic events during nerve recovery. that is applied to an area over a specific time. A The severity of injury to the RLN dictates histologic, standard method of compression will serve to ameliorate neurologic and functional recovery of the laryngeal discrepancies among research teams when studying crush injury of the RLN. The aim of the present study was to evaluate the regeneration and functional recovery From the Department of Otolaryngology, the New York Eye and Ear of RLN following controlled crush injury. This model is Infirmary, New York and New York Medical College, Valhalla, New York, U.S.A. (B.T., R.M.R., M.J.P., P.W., S.D.S.) and the Department of Cell Biology, intended to simulate one type of intraoperative RLN New York Medical College, Valhalla, New York, U.S.A. (S.S.). injury and to facilitate understanding of the histological Editor’s Note: This Manuscript was accepted for publication on and myoelectrical events that are observed during nerve February 24, 2009. recovery. Presented at the Triological Society Combined Sections Meeting, Orlando, Florida, U.S.A., May 1–4, 2008. This research was performed at New York Medical College, Val- MATERIALS AND METHODS halla, NY. Send correspondence to Steven D. Schaefer, MD, FACS, 310 E. Animals 14th St., 6th Fl, New York Eye and Ear Infirmary, New York, NY 10003. Forty-two Sprague Dawley rats weighing 200 g to 250 g E-mail: [email protected] were used in the present study. The animals were divided into DOI: 10.1002/lary.20293 two groups. The first group underwent crush injury using a Laryngoscope 119: August 2009 Tessema et al.: RLN Injury and Recovery in Rats 1644 commercially available Sugita calibrated 0.61 N aneurysm clip ports that are spaced 200 lm apart in a square pattern.16 Bipo- (Mizuho Ikakogyo, Tokyo, Japan), the second group underwent lar EMG signals were recorded from two of the four electrode crush injury using a 1.19 N aneurysm clip (Mizuho Ikakogyo). detectors. Assuming that diameters of rat laryngeal muscle Both aneurysm clips were 1.4 mm in width. Animal procedures fibers are approximately 25 lm,7 detectors of the quadrifilar were approved by the institutional review board of New York electrode, therefore, spanned multiple muscle fibers while pro- Medical College and performed in facilities approved by the viding an increased selectivity of individual motor unit action National Academy of Sciences and the National Society for potential morphologies over concentric needle electrodes. We Medical Research. Humane care was provided for these animals also tested clinically-available bipolar concentrilc needle and and all institutional and national guidelines were observed. monopolar needle electrodes for obtaining EMG recordings from the rat larynx; however, the quadrifilar electrode provided the most consistent and reliable recordings for the rat model. An Surgery ear clip electrode served as reference for the EMG recordings. Animals were sedated with isoflurane and then anesthe- Myoelectric signals were routed to a preamplifier module tized with an intramuscular injection of a mixture of Ketaset located at the needle head and then to high-gain differential and Rompun (1:1). A final dosage of 45 mg ketamine/100 g body amplifiers located nearby (in-house construction). Myoelectric weight and 0.9 mg/100 g body weight of xylazine plus 1 mg/kg signals were bandpass filtered from 15 Hz to 1,000 Hz (Krohn- acepromazine maleate produced sufficient anesthesia. Hite Model 3360, Brockton, MA) and then digitized at a depth With animals placed in a supine position in a stereotactic of 12 bits and a rate of 20 kilosamples per second (Data Trans- apparatus, a vertical midline neck incision was performed, and lation Model DT2821G, Marlboro, MA). Digitized data were the strap muscles were separated. The right recurrent laryngeal stored to the internal disc drive of a portable computer in real nerve was crushed in all animal groups and the left side was time (Dolch Computer Systems Model LPAC-PT, Freemont, used as control. The RLN was carefully identified in the tra- CA). EMG signals were displayed, analyzed, and compared cheoesophageal groove under a dissecting microscope. In the using Matab software (The Mathworks, Release 2007b, Natick, first group of animals, crush injury was performed at the level MA). of the seventh tracheal ring with a 0.61 N closing force for Myoelectric activity was also monitored visually using a exactly 60 seconds. The second group underwent an identical digital real-time oscilloscope (Tektronix TDS210, Tektronix, procedure using a Sugita 1.19 N closing force. Strap muscles Beaverton, OR) and audibly by means of an amplifier/speaker and overlying fascia were closed with 4-0 chromic suture and unit connected in parallel to the EMG data recoding system. the skin reapproximated with 5-0 nylon sutures. Kinematic activity (e.g., vocal fold movement and position) was observed visually by one investigator at a given time by means of ToL endoscopy. In some experiments, kinematic activity was Transoral Evaluation of Vocal Fold observed by all investigators by means of a video monitor that Movement EMG received its signal from a solid-state color camera (Richard Wolf Functional behavior was studied using transoral laryngeal S5366.40, Richard Wolf GmbH, Knittlingen, Germany) mounted electromyography (ToL EMG) obtained during spontaneous re- on the eyepiece of the endoscope. Simultaneous recordings of ki- spiratory cycles and vocal fold adduction under anesthesia. The nematic and myoelectric activities were achieved for some methodology of ToL EMG has been published in our earlier pa- experiments by feeding the audio/video signals to a VHS tape per,14 with salient details provided here. The animals were recorder (Panasonic AG-6300MD, Panasonic, Knoxville, TN). placed supine on a modified stereotactic operating table with a The audio/video signals were digitized to AVI files by a com- 15 incline in a Faraday cage recording room. The tongue was puter workstation following the experimental session. AVI files retracted using a 3-0 silk suture placed midline in the anterior facilitated postexperimental review and discussion. This tech- one third of the tongue and suspended. A 0 Storz operating nique helped to confirm judgments that were made during the endoscope equipped with an epiglottis elevator was transorally experimental session, including incidences of synkinesis. inserted to visualize the endolarynx and provide excellent visu- alization of the glottal region. Vocal fold movements observed via transoral endoscopy were independently graded by two Histology investigators using a continuous scale from 0 to 4 prior to per- Following ToL EMG, animals were sacrificed with an in- forming each ToL EMG. Scoring of vocal fold movement and traperitoneal injection of lethal dose of phenobarbital at position was guided by the following criteria: 0 no vocal fold postoperative day 4, 7, 14, 21, 28, 33, and 42.