Minithyrotomy With Radiofrequency-Induced Thermotherapy for the Treatment of Adductor Spasmodic Dysphonia

Transcription

1 The Laryngoscope VC 2016 The American Laryngological, Rhinological and Otological Society, Inc. Minithyrotomy With Radiofrequency-Induced Thermotherapy for the Treatment of Adductor Spasmodic Dysphonia Shaun C. Desai, MD; Andrea M. Park, MD; Rebecca D. Chernock, MD; Randal C. Paniello, MD, PhD Objectives/Hypothesis: A simple, safe and effective surgical alternative for treating adductor spasmodic dysphonia (ADSD) would appeal to many patients. This study evaluates a new option, using radiofrequency-induced thermotherapy (RFITT) of the thyroarytenoid muscle (TA) via the minithyrotomy approach to reduce the force of adduction. Methods: Fifteen dogs were used. In part 1, the optimal RFITT power settings, exposure time, probe location, and number of passes were determined. Part 2 compared laryngeal adductor pressures (LAPs) at baseline; immediately postintervention; and at 1, 3, or 6 months postintervention. Interventions included RFITT via the transcervical minithyrotomy approach (n 5 15), transoral RFITT (n 5 3), botulinum toxin (Botox) injection (n 5 3), or no-intervention controls (n 5 3). Postintervention induced phonation and histologic analyses were performed as well. Results: In the minithyrotomy RFITT group, the mean LAP was 30.3% of baseline immediately posttreatment. At 1, 3, and 6 months postoperatively, the mean LAPs were 24.9%, 44.8%, and 43.5%, respectively. Transoral RFITT reduced LAP to 56.6% of baseline immediately posttreatment, but returned to normal in the 1 and 3 month animals. The Botox injections dropped the LAP to 57% of baseline at 1 month, but returned to normal at 3 months. Mucosal waves, based on induced phonation stroboscopy, were present at the terminal date in all animals. Thirteen of 15 transcervical RFITT preparations (87%) showed no injury to the lamina propria, whereas 80% showed evidence of TA muscle atrophy and fibrosis. Conclusion: Minithyrotomy RFITT is a feasible technique that shows encouraging long-term results for the potential treatment of patients with ADSD. Key Words: Spasmodic dysphonia, radiofrequency, surgery. Level of Evidence: N/A. Laryngoscope, 126: , 2016 INTRODUCTION Spasmodic dysphonia (SD) is a primary taskspecific focal dystonia involving the laryngeal muscles during speech, affecting over 50,000 people in North America alone. 1 Over 90% of cases are of the adductor type, characterized by voice breaks in vowels during speech and a tremulous shaking of the voice. 2 Depending on the severity, patients voices can sound strained, tight, strangled, breathy, or whispery, causing significant impairment in quality of life. The remaining rarer cases are of a mixed or abductor type, characterized by voiceless breaks due to prolonged voiceless consonants before initiation of the following vowels. 3 From the Department of Otolaryngology Head and Neck Surgery (S.C.D., A.M.P., R.C.P.); and the Department of Pathology (R.C.), Washington University School of Medicine, St. Louis, Missouri. Currently at the Department of Otolaryngology Head and Neck Surgery, Johns Hopkins University (S.C.D.), Baltimore, Maryland, U.S.A. Editor s Note: This Manuscript was accepted for publication February 26, Presented at the annual meeting of the American Laryngological Association, Orlando, Florida, U.S.A., April 10 11, Supported by the National Institutes of Health grant R01 DC The authors have no other funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Randal C. Paniello, MD, PhD, Department of Otolaryngology Head & Neck Surgery, Washington University School of Medicine, 660 South Euclid Avenue, Box 8115, St. Louis MO paniellor@ent.wustl.edu DOI: /lary Although the exact cause of adductor spasmodic dysphonia (ADSD) remains uncertain, it is likely to originate from the basal ganglia. Unregulated neural stimulation of one of the main laryngeal adductors, the thyroarytenoid (TA) muscle, results in the persistent strained and strangled voice. To date, treatment options, both surgical and nonsurgical, have aimed at peripheral muscle denervation. 4 Such treatments have included unilateral section of the recurrent laryngeal nerve (RLN), 5,6 unilateral RLN avulsion, 7,8 TA myectomy, 9,10 and TA denervation with reinnervation. 11,12 These treatments have had mixed success, with some improvement in voice breaks but somewhat unpredictable side effects such as vocal harshness. 4 The only treatment option shown to be clinically effective in a double-blind controlled study was the use of botulinum toxin (Botox). 13 In that study, patients were injected with Botox directly into the TA muscle, with over 90% of patients experiencing some beneficial effect. However, limitations include the need for repeated injections every several months, technical difficulty in placing the needle, the need for possible electromyography, and significant cost (often not covered by health insurance and prohibitive to the uninsured). In addition, longitudinal follow-up studies have showed that patients have reduced quality of life for a large proportion of the cycle from one injection to the next. 14,15 In 2005, Remacle et al. introduced a new type of neuromyectomy using bipolar radiofrequency-induced 2325

2 thermotherapy (RFITT) in three patients. 16 Using a transoral approach, they used RFITT on the TA in one site just lateral to the tip of the vocal process. This was expected to severely traumatize the TA branches of the recurrent laryngeal nerve, which would weaken the force of laryngeal closure. They reported significant improvement in the Vocal Handicap Index (VHI) in all three patients. The bipolar radiofrequency (RF) probe prevents the diffusion of the current and protein coagulation with lower temperatures, eliminating unnecessary damage to adjacent tissues, which helps preserve the mucosal wave. In 2008, Kim et al. reported a modified technique of transoral RFITT by treating the TA in two sites rather than one, first in dogs and then in 20 patients. 17 The mean VHI improved significantly; however, the glottic gap that was present in 80% at 2 months persisted in only 35% at 6 months and 25% at 12 months. Half of the patients returned to Botox injections at 1 year postoperatively. In this study, we proposed that a longer benefit could be obtained by treating more than two sites; instead, we sought to treat the entire TA muscle by use of Gray s minithyrotomy approach. 18 This allows the RF probe to follow the long axis of the TA muscle along its full length, from the thyroid to the arytenoid cartilages. In part 1 of this canine study, we systematically determined the optimal settings for RFITT by comparing various combinations of power, exposure time, and probe locations within the TA. In part 2, we took the optimal settings obtained from part 1 and compared the minithyrotomy approach to the transoral approach as well as to the current gold standard, Botox injection. MATERIALS AND METHODS Fifteen purpose-bred, conditioned, female mongrel dogs weighing 20 to 25 kg were obtained and maintained in a facility approved by the American Association for the Accreditation of Laboratory Animal Care. Guidelines of the Public Health Service Policy on Humane Care and Use of Laboratory Animals, the National Institutes of Health Guide for the Care and Use of Laboratory Animals, and the Animal Welfare Act (7 U.S.C. et seq.). The Institutional Animal Care and Use Committee of Washington University (St. Louis, MO) approved the protocol. Procedures were performed under general anesthesia, induced via intravenous thiopental sodium, and maintained with 2% halothane inhalant. Part 1: Determining the Optimal Settings and Approach In three dogs, a tracheostomy was performed through rings 10 to 13, as previously described. 19 The RLN on each side was dissected, an electrode was applied and connected to a custom constant-current laryngeal nerve stimulator (WR Medical Electronics, Stillwater, MN), and baseline laryngeal adductor pressure (LAP) was measured using a previously described method. 20 Briefly, the RLNs are stimulated while the balloon of an endotracheal tube is passed between the vocal folds. The lowest current that gives the maximum LAP at 80 Hz is determined and used subsequently. Each stimulation lasts 3 to 5 seconds while LAP is recorded at 10-Hz intervals; plateau values from 70 Hz to 100 Hz are averaged. A recovery time of 30 seconds is allowed between stimulations Fig. 1. Minithyrotomy radiofrequency-induced thermotherapy. Cutaway of thyroid cartilage reveals lateral TA muscle with embedded radiofrequency probe. False vocal fold is also removed to show true vocal folds. Probe is passed through minithyrotomy into TA muscle and advanced posteriorly to reach body of arytenoid (shadowed), and then a series of lesions is created as probe is withdrawn anteriorly. TA 5 thyroarytenoid muscle. Gray s minithyrotomy was performed as described, 18,19 except that the soft thyroid cartilage was opened with a 3-mm punch rather than a drill. The radiofrequency probe used was one designed for turbinate reduction (Reflex Ultra 45, Arthro- Care, Austin, TX), with a 45-degree bend midshaft; the tip creates lesions about 1 cm in length. The probe was passed through the minithyrotomy and into the entire length of the TA muscle (Fig. 1), with careful attention paid not to disrupt the epithelial surface. A second operator viewed the vocal folds by direct laryngoscopy to monitor this process. The probe was activated according to the parameters given in Table I; it was then advanced anteriorly approximately 1 cm at a time, creating additional lesions at each position using the same parameters until the probe was completely withdrawn. The number of lesions in each pass depended on the size of the larynx and the amount of overlap between them. Immediately after the RFITT treatment was complete, the LAP was measured as described above. The wound was closed, and the animal recovered. After 2 or 4 weeks, the dog was anesthetized, the RLN dissected, and LAP measurements repeated. Induced phonation was performed (below section). After euthanasia, the larynx was harvested for gross and histologic exam. Induced Phonation. Induced phonation was performed in order to determine whether mucosal waves could still form after RFITT. A second endotracheal tube was directed rostrally through the tracheostomy and positioned with the balloon close to the stoma to maximize the distance between the tip and the vocal folds. Humidified, warmed air was passed through the larynx, with the flow controlled by a rotameter. In some cases, electrical RLN stimulation did not elicit adequate adduction to induce phonation, so the vocal folds were approximated by use of arytenoid adduction sutures. Videostroboscopy (Kay-Pentax, Montvale, NJ) via direct laryngoscopy enabled photodocumentation of mucosal waves. Histological Analysis. Immediately after the dog was sacrificed, the larynx was harvested. Vertical cross-sectional slices immediately anterior to the vocal process and midvocal fold

3 TABLE I. Part 1 Experiments and LAP Results. Optimal Parameters for Use in Part 2 Were Found in Experiment 2R (in bold). Setting Exposure Lesions Passes LAP Interval Dog (watts) (sec) No. No., Location Initial Terminal Weeks 1 L , med R , med L , med R , med and lat 3 L , lat R , lat L 5 left; lat 5 lateral aspect of TA; med 5 medial aspect of TA; R 5 right; TA 5 thyroarytenoid muscle. were obtained, making sure to include the epithelial surface and the TA muscle. These specimens were placed in separate containers, frozen in liquid nitrogen, and sectioned at 50 lm. The sections were stained with hematoxylin and eosin. Light microscopy was performed to evaluate for any epithelial or muscle change. A dedicated head and neck pathologist (R.C.), blinded to the intervention, graded the amount of mucosal, submucosal, and TA muscle damage. Special attention was paid to judge any disturbance of the lamina propria. Part 2: Comparing Transcervical RFITT, Transoral RFITT, and Botox Twelve mongrel dogs (24 vocal cords) were used. Initial (pretreatment) data was obtained as described in part 1. Then, vocal cords were treated as outlined in Table II: Minithyrotomy (MT) RFITT: Was performed, as described in part 1, using the optimal parameters determined in part 1 (entries for dog 2R in bold in Table I) (n 5 15). Transoral (TO) RFITT: Using direct laryngoscopy, the RF probe was passed into the vocal cord so that the full 1-cm active tip was intramuscular in the TA, with two passes as per Kim et al. 17 Parameters used were an exposure time of 10 seconds with a power setting of 6, as used for the MT-RFITT group (n 5 3). Botox injection: A 27-gauge needle attached to a syringe containing Botox 25U/cc was passed through the cricothyroid membrane and into the vocal fold, confirmed by observation by direct laryngoscopy. Five units of Botox were injected into the TA muscle. (n 5 3). Normal controls: No treatment. (n 5 3). At 1, 3, or 6 months, the dogs underwent RLN dissection; determination of LAP and induced phonation, as described for part 1; and then euthanasia, followed by larynx harvest for histology. TABLE II. Experimental Group Assignments and Endpoint Times. Dog No. Endpoint (months) Left TA Right TA 1 2 MT RFITT No Rx (control) 3 1 MT RFITT Botox 4 MT RFITT TO RFITT 5 6 MT RFITT No Rx (control) 7 3 MT RFITT Botox 8 MT RFITT TO RFITT 9 10 MT RFITT No Rx (control) 11 6 MT RFITT Botox 12 MT RFITT TO RFITT Botox 5 botulinum toxin; MT, minithyrotomy; Rx 5 treatment; TA 5 thyroarytenoid muscle; RFITT 5 radiofrequency-induced thermotherapy; TO 5 transoral. In Part 2, the 15 minithyrotomy-rfitt hemilarynges all showed significant reduction in LAP immediately after treatment, to a mean of 30.3% % of baseline. This reduction was sustained, to 24.9% % at 1 month, 44.8% % at 3 months, and 43.5% % at 6 months (Fig. 3). These data include all 15 experiments at time zero (immediately posttreatment), and five each at subsequent time points. By comparison, the three transoral-rfitt trials only dropped the LAP to 56.6% % of baseline at time zero, and the 1- and 3-month experiments returned to baseline (Fig. 3). Interestingly, the 6-month trial had an LAP of 67.1%. The Botox-treated hemilarynges had the expected LAP results: no reduction immediately postinjection and a drop to 57% at 1 month, followed by a return to normal by 3 months. Controls showed no change in LAP. Note that these small comparison trials had only one experiment at each time point, so statistical comparisons with the MT-RFITT group are not meaningful. In the MT-RFITT group, induced phonation with normal-appearing mucosal waves was achieved in 14 of RESULTS There were no major surgical complications, and there were no unintentional postoperative deaths. In part 1, the three trials that utilized a pass into the lateral aspect of the TA showed greater initial LAP reduction than those using medial passes (Table I), but only the hemilarynx that was treated both medially and laterally (2R) sustained the drop in LAP to less than 25% of baseline at the endpoint (Fig. 2). Thus, this approach was chosen for use in part 2. Fig. 2. LAP results of dog 2R from part 1. The mean LAP at 70 Hz to 100 Hz immediately following radiofrequency-induced thermotherapy was 24% of baseline, and remained 22% 2 weeks later. LAP 5 laryngeal adductor pressure; Rx 5 treatment. 2327

4 Fig. 3. LAP results from part 2 experiments. Minithyrotomy RFITT (solid line) reduced LAP to 30% of baseline (n 5 15) and remained significantly reduced at 1, 3, and 6 months (n 5 5 each time point). Transoral RFITT dropped to 57% of baseline initially (n 5 3), but this was not retained in two of three later time points (n 5 1 each). Botox did not cause an initial decrease in LAP (n 5 3), fell to 57% of baseline at 1 month, but returned to normal at 3 months (n 5 1 each). Controls showed no change in LAP at any time point. Botox 5 botulinum toxin; LAP 5 laryngeal adductor pressure; MT 5 minithyrotomy; RFITT 5 radiofrequency-induced thermotherapy; Rx 5 treatment; TO 5 transoral. 15 experiments at fundamental frequencies of 90 Hz to 180 Hz. One dog developed a granuloma, which restricted phonation. Arytenoid adduction sutures were needed on six dogs to close the posterior glottic gap to achieve vocal fold vibration; six successfully achieved induced phonation without them. Histologic exams of the MT-RFITT groups showed that 13 of 15 vocal cords (87%) had no evidence of injury to the lamina propria (Fig. 4). The other two vocal cords had evidence of fibrosis in this layer. The TA muscles in 12 of 15 (80%) showed evidence of focal muscle atrophy and fibrosis. The TO-RFITT group showed no evidence of lamina propria injury in any of the three vocal cords tested, and all three showed TA muscle atrophy, although to a lesser degree. As expected, the Botox and control groups showed no evidence of significant mucosal/submucosal damage or TA muscle atrophy and fibrosis. DISCUSSION Many patients with ADSD grow tired of the Botox roller-coaster and seek a surgical alternative. Radiofrequency-induced thermotherapy showed promise in the clinical reports of Remacle et al. 16 and Kim et al., 17 although for some patients symptomatic relief did not last longer than the typical Botox cycle. We hypothesized that the introduction of the RF probe via Fig. 4. Sample histologic findings. Top left: normal TA muscle from a control. Top right: atrophic muscle trapped in scar 1 month after MT- RFITT. Bottom left: normal vocal fold lamina propria from a control. Bottom right: 6 months after MT-RFITT, lamina propria remains undamaged. Hematoxylin and eosin stain, MT5 minithyrotomy; RFITT 5 radiofrequency-induced thermotherapy; TA 5 thyroarytenoid muscle. [Color figure can be viewed in the online issue, which is available at

5 the minithyrotomy approach would enable treatment of the entire TA muscle, and thereby lead to greater reduction in adductor strength and longer-lasting benefit. The data from this limited animal study support this hypothesis. Another key finding is that the MT-RFITT procedure did not significantly damage the lamina propria (LP) and did not cause a loss of mucosal waves. Although two of 15 specimens did show some histologic evidence of fibrosis of the LP, mucosal waves were still produced in these animals. This suggests that a clinical trial of this new procedure would entail reasonable risk, but it will be critical to make certain that the probe remains intramuscular throughout the treatment and does not contact the LP when activated. This would best be achieved by endoscopy, viewing the vocal folds while performing the RFITT. With the minithyrotomy approach, the vocal fold mucosa is intentionally not punctured by the probe tip (as is necessary in the transoral approach), and thus less injury of the LP would be expected. The minithyrotomy approach does involve a small (< 2 cm) skin incision, but this is well-tolerated by patients 18,21,22 and is less invasive than other surgical treatments used for ADSD. Compared with the small transoral group, we found that the minithyrotomy afforded better access to the TA muscle and better control of the exact position of the RF probe. The histologic findings showed that the minithyrotomy approach did result in a more complete treatment of the TA muscle, as well as greater atrophy. Both approaches likely traumatize the distal RLN, as well as the TA, given the nerve s location in the center of the muscle. This local denervation likely contributes to both the reduced adductor strength and the atrophy. The minithyrotomy approach also offers the potential to titrate the treatment to the desired effect in order to avoid overreduction of adductor ability. The procedure can easily be repeated, if needed, using the same skin incision and same cartilage fenestration. Limitations of this study include the small sample sizes of the comparison groups. Furthermore, because this is a simulated canine model, one cannot predict the actual human voice quality based on experimental LAP. The study was limited to just 6 months, and future experiments should look at periods of 12 months or longer to further assess longterm efficacy. CONCLUSION Minithyrotomy-RFITT effectively reduced adductor force in a canine model, maintaining intact mucosal waves, and appears to offer advantages over the transoral approach. Further studies are warranted to evaluate its potential use as a treatment for adductor spasmodic dysphonia. BIBLIOGRAPHY 1. National Spasmodic Dysphonia Association. Spasmodic Dysphonia. Available at: Accessed Nov 1, Parnes SM, Lavorato AS, Myers EN. Study of spastic dysphonia using videofiberoptic laryngoscopy. Ann Otol 1978;87: Edgar JD, Sapienza CM, Bidus K, et al. Acoustic measures of symptoms in abductor spasmodic dysphonia. J Voice 2001;15: Ludlow, CL. Treatment for spasmodic dysphonia: limitations of current approaches. Curr Opin Otolaryngol Head Neck Surg 2009;17: Dedo HH. Recurrent laryngeal nerve section for spastic dysphonia. Ann Otol Rhinol Laryngol 1976;85: Aronson AE, De Santo LW. Adductor spastic dysphonia: three years after recurrent laryngeal nerve resection. Laryngoscope 1983;93: Netterville JL, Stone RE, Rainey C, et al. Recurrent laryngeal nerve avulsion for treatment of spastic dysphonia. Ann Otol Rhinol Laryngol 1991; 100: Weed DT, Jewett BS, Rainey C, et al. Long-term follow-up of recurrent laryngeal nerve avulsion for the treatment of spasmodic dysphonia. Ann Otol Rhinol Laryngol 1996;105: Genack SH, Woo P, Colton RH, et al. Partial thyroarytenoid myectomy: an animal study investigating a proposed new treatment for adductor spasmodic dysphonia. Otolaryngol Head Neck Surg 1993;108: Woo P. Carbon dioxide laser-assisted thyroarytenoid myomectomy. Lasers Surg Med 1990;10: Berke GS, Blackwell KE, Gerratt RR, et al. Selective laryngeal adductor denervation-reinnervation: a new surgical treatment for adductor spasmodic dysphonia. Ann Otol Rhinol Laryngol 1999;108: Allegretto M, Morrison M, Rammage L, et al. Selective denervation: reinnervation for the control of adductor spasmodic dysphonia. J Otolaryngol 2003;32: Troung DD, Rontal M, Rolnick M, et al. Double-blind controlled study of botulinum toxin in adductor spasmodic dysphonia. Laryngoscope 1991; 101: Paniello RC, Barlow J, Serna JS. Longitudinal follow-up of adductor spasmodic dysphonia patients after botulinum toxin injection: quality of life results. Laryngoscope 2008;118: Novakovic D, Waters HH, D Elia JB, Blitzer A. Botulinum toxin treatment of adductor spasmodic dysphonia: longitudinal functional outcomes. Laryngoscope 2011;121: Remacle, Plouin-Gaudon I, Lawson G, et al. Bipolar radiofrequency-induced thermotherapy (RFITT) for the treatment of spasmodic dysphonia. A report of three cases. Eur Arch Otorhinolaryngol 2005;262: Kim HS, Choi HS, Lim JY, et al. Radiofrequency thyroarytenoid myothermy for treatment of adductor spasmodic dysphonia: how we do it. Clin Otolaryngol 2008;33: Paniello RC, Sulica L, Khosla SM, Smith ME. Clinical experience with Gray s minithyrotomy procedure. Ann Otol Laryngol 2008;117: Dahm JD, Paniello RC. Tracheostomy for long-term laryngeal examination. Otolaryngol Head Neck Surg 1998;118: Paniello RC, West SE. Laryngeal adductory pressure as a measure of postreinnervation synkinesis. Ann Otol Rhinol Laryngol 2000;109: Gunderson M, Bauer B, Glab RC, Dailey SH. Technical refinements to the minithyrotomy procedure. J Voice 2014;28: Mallur PS, Gartner-Schmidt J, Rosen CA. Voice outcomes following the Gray minithyrotomy. Ann Otol Rhinol Laryngol 2012;121: