Window Anterior Commissure Relaxation Laryngoplasty in the Management of High-Pitched Voice Disorders

Objective  To present the success rate of a less invasive modification of Isshiki type III anterior commissure relaxation laryngoplasty technique in patients with high-pitched voice disorders.

Design  Prospective case series.

Setting  KBB Major Private Clinic of Istanbul Surgery Hospital and the University of Yeditepe Hospital.

Patients  Twenty-one adult patients who believed that their high-pitched voices conflicted with their body image and/or gender identity.

Intervention  Type III thyroplasty for pitch alteration.

Main Outcome Measures  Comparison of preoperative and postoperative (>6 months) fundamental frequency levels, diplophonia, perception of body image and pitch, and subjective ratings of comfort during vocalization.

Results  The patients were mostly male (mean age, 30.5 years). The most frequent cause of high-pitched voice was sulcus vocalis (n = 14), followed by constitutional causes (n = 5), mutational falsetto (n = 1), and severe glottic scarring secondary to childhood diphtheria (n = 1). After surgery, the fundamental frequency dropped significantly from a mean of 213.81 Hz to 149.86 Hz (P < .001), equaling a mean postoperative semitone drop of 6.23. Misperception leading to an abnormal body image was reduced by 86%. Fourteen patients who originally had feelings of tension and fatigue during phonation and vocalization gained comfort postoperatively. Diplophonia with subharmonic signals observed in 11 cases preoperatively was reduced or disappeared in 6 cases. No complications or failures were observed during the follow-up period.

Conclusion  Window anterior commissure relaxation laryngoplasty is an efficient, easy, less invasive, and safe procedure in the surgical management of organic and functional high-pitched voice disorders.

The habitual vocal pitch is a major factor characterizing the vocal quality, and it has a role in defining one's body image. It is correlated with the patient's sex, age, airway size, and laryngeal morphometry.^1-5 Atrophic vocal folds, functional problems, and even normal-looking vocal folds may produce inappropriately high-pitched levels that may lead to a feminine or childlike voice and perception of body image when the patient talks on the telephone or communicates in other nonvisual manners.⁶

Surgical correction or reduction of pitch was first defined by Isshiki et al⁷ as the type III thyroplasty technique. This technique aims to shift the anterior portion of the thyroid cartilage with its anterior commissure attachment to obtain relaxed vocal folds.⁷ The technique is often used to manage atrophic vocal fold abnormalities and mutational falsetto.^8-10 The results are consistent, and good functional outcomes in terms of vibration quality have been reported.¹¹ However, instability is still a potential problem of this technique and can interfere with the fine adjustments made during surgery and result in postoperative complications related to displacement.¹² This report presents a less invasive modification of the standard Isshiki technique that provides better stability and ease of adjustability.

Between February 10, 1999, and November 16, 2006, 21 patients underwent window anterior commissure relaxation laryngoplasty at the KBB Major Private Clinic of Istanbul Surgery Hospital and the University of Yeditepe Hospital. All patients gave their informed consent. The study protocol was approved by the Ethics Committee of Yeditepe University. In the preoperative and postoperative periods, all patients underwent a vocal function evaluation that included videolaryngostroboscopy, acoustic analysis, and voice recording.

The study population included 19 men and 2 women with a mean (SD) age of 30.5 (4.6) years (range, 23-42 years). The primary concern of these patients was that their voices were consistently being perceived as having a feminine quality for men or as childlike for women in nonvisual audio or telephone communications. Fourteen of the patients (67%) also complained of vocal tension and fatigue.

The cause was sulcus vocalis in 14 cases, as diagnosed by laryngeal stroboscopy. Twelve of these patients had a bilateral asymmetric “vergeture” type II sulcus and 2 had bilateral “pocket” type III sulcus vocalis. Three patients underwent sulcus vocalis resection with microsuturing, 1 underwent Reinke space augmentation with hyaluronic acid, 1 had temporal muscle fascia implantations and suturing, and 1 had microweb excision. We did not perform surgery until at least 1 year after the previous procedure.

Cases with normal anatomy of the phonation organs and vocal folds with normal modal vibratory behavior, except for the measured high fundamental frequency, were defined as constitutional (n = 5). These patients did not show any improvement after extensive voice therapy.

One patient had a deep bilateral, asymmetric scar secondary to a laryngeal diphtheria infection in childhood. This patient had undergone suspension laryngoscopy with hyaluronic acid injection and temporal muscle fascia implantation 1 year before the modified Isshiki type III thyroplasty. One patient had a persistent mutational falsetto disorder resistant to long-term extensive voice therapy and other conservative forms of management (Table 1).

The period of follow-up was 6 to 80 months postoperatively, with a mean (SD) of 32.6 (22.4) months.

This technique is a modification of Isshiki type III thyroplasty. Instead of separating the thyroid cartilage into 3 pieces, only the anterior commissure–thyroid cartilage complex (ACTC) is separated. This spares the thyroid cartilage to increase stability and allow additional procedures, such as medialization and arytenoid adduction. The shape of the cartilage is rhomboid to facilitate surgical manipulation through a small anterior midline incision. The ACTC alone can be moved in coronal and sagittal directions as a composite tissue block. Coronal displacement provides symmetric tension or relaxation, while sagittal displacement provides asymmetric tension or relaxation.

The surgery was performed using local sedation anesthesia. Patients were sedated with intravenous fentanyl citrate and midazolam hydrochloride under the control of an experienced anesthesiologist. The tissue overlying the anterior thyroid region was infiltrated with 2% lidocaine containing 1:200 000 epinephrine. Oxygen was supplied via a nasal cannula throughout the procedure.

During the operation, computerized voice analysis systems were available and the patients were monitored during the tuning process and positional changes. The fundamental frequency levels in phonation and speech tasks (sustained vowel and counting from 1 to 12) were analyzed. When a medialization procedure was added, the patients were monitored by fiberoptic nasolaryngoscopy to determine the exact location and to prevent undesired augmentation volume and sites, such as the supraglottis and anterior glottis.

The patients were placed in the supine position with the head extended to expose the thyroid cartilage. A 2-cm horizontal incision was made anterior to the midthyroid level inside a skin crease. The sternohyoid muscles were separated and retracted laterally to expose the anterior thyroid cartilage. The external perichondrium of the thyroid cartilage was incised vertically with the superior and inferior margins of the thyroid cartilage in the shape of a vertically rotated H. The anterior commissure was localized by either measuring the midpoint or finding the red point of the anterior glottic vessels.

A rhomboid-shaped cartilage window was prepared, while keeping the anterior commissure centered in the operative field. The superior and inferior corners were not less than 3 mm from the inferior and superior margins of the thyroid cartilage to maintain the stability of the entire thyroid cartilage framework. The lateral corners were extended laterally not less than 10 mm in the direction of a glottic line parallel to the inferior border of the inferior thyroid cartilage border (Figure 1). The cartilage window was incised with a No. 15 scalpel or by drilling with 2- to 3-mm diamond burrs (Figure 2). After the internal perichondrium was reached, the window was partially mobilized by circumferential elevation of the thyroid lamina away from the internal perichondrium, while keeping the ACTC window and its internal perichondrium attachments intact (Figure 3). The lateral cartilage borders of the flanges were removed further by drilling or cutting. Care was taken to preserve the internal perichondrium. The lateral corners were reduced while following the glottic line for symmetry. Each reduction resulted in passive posterior displacement of the window ACTC, resulting in a drop in the fundamental frequency (Figure 4).

After displacement of 4 to 5 mm was obtained, the lateral contact between the cartilage window and thyroid cartilage frame was lost, but further relaxation was allowed until a certain stable and balanced tension was reached and no further fundamental frequency change was measured. The cartilage window was grasped gently and moved in sagittal and coronal directions to obtain better voice quality. Coronal movements resulted in symmetric tension, while sagittal movements resulted in asymmetric tension. If the passive relaxation was insufficient, a force was applied to displace the ACTC, which was then stabilized. After the desired or optimal pitch levels were reached, the quality of the voice was reevaluated. In the presence of breathiness and diplophonia, sagittal positioning and medialization procedures were planned. In the case of bowed, atrophic, or incompetent vocal fold or folds, a medialization pocket was undermined through the inferior border of the frame. The frame corner was the upper limit of the augmentation pocket. The pocket was extended through the midglottic level, which usually did not need to be exceeded for atrophic abnormalities.

A silicone or polytetrafluoroethylene (PTFE) (GORE-TEX; W. L. Gore & Associates, Flagstaff, Arizona) stripe was introduced into the medialization pocket and used to stabilize the ACTC. After any bleeding was controlled, the strap muscles and skin flaps were approximated with 3-0 polyglactin and the skin was closed with 4-0 nylon sutures. A Penrose drain was left in the neck for 1 or 2 days. The sutures were removed on the sixth postoperative day.

In the postoperative period, the patients were instructed not to make any vocal effort or participate in sports activities. Antireflux medication was added for 3 weeks. Vocal fold edema and voice quality were checked between 4 and 10 days postoperatively, in the sixth postoperative week, and more than 6 months postoperatively.

Voice analysis was performed with the Multi-Dimensional Voice Program developed by Kay Elemetrics (Lincoln Park, New Jersey). Each voice sample was recorded in a quiet room (noise level, <40 dB) with the use of a unidirectional (cardioid) microphone (SM58; Shure, Niles, Illinois) placed 10 cm from the lips. The subjects were instructed to sound the vowel a at a comfortable level for 3 seconds. The first and last 200 milliseconds of the sustained phonation were discarded and the fundamental frequency was measured. Each trial was captured digitally at 16-bit resolution with a sampling rate of 44 100 Hz. A narrow-band spectrogram (Real-Time Spectrogram, Model 5129, Version 2.4; Kay Elemetrics) was used to evaluate the subharmonics and diplophonia in the voice. Semitones were calculated from the corresponding notes that represent the frequency value. A narrow-band spectrogram was used to measure the degree of subharmonic type 2 signals.

Routine questionnaires were used to evaluate 3 main points in the preoperative and postoperative periods (more than 6 months): (1) self-perceived pitch as high, normal, or low; (2) caller response on routine telephone communications as “Mr,” “Mrs,” “Miss,” or “Get your Mommy or Daddy”; and (3) the comfort level in speech as bad or good.

All computations and statistical analyses were performed with SPSS 8.0.0 (SPSS Inc, Chicago, Illinois). Paired t tests were used to compare differences between the preoperative and postoperative values of the voice analysis. Statistical significance was set at P < .05.

The overall preoperative mean (SD) fundamental frequency was 213.81 (32.48) Hz, and it decreased postoperatively to 149.86 (27.36) Hz, which was highly significant (P < .001). The mean decrease was 63.95 (20.35) Hz, corresponding to a mean of 6.23 (1.98) semitones. The greatest decrease (12.22 semitones) was seen in the patients with mutational falsetto, followed by the patients with the scarred vocal fold with 6.93 semitones, the sulcus vocalis cases with 6.00 semitones, and the constitutional cases with 5.56 semitones (Table 2).

Without implants, a considerable decrease in the fundamental frequency was obtained. For further relaxation, PTFE was used in 6 cases and silicone blocks in 7 cases for ACTC compression and stabilization. Although the best relaxation results were achieved with silicone, there was no significant difference between the implant and nonimplant groups (Figure 5).

A subjective pitch drop was noted by all patients in the constitutional (n = 5), mutational falsetto (n = 1), and glottic scarring (n = 1) groups. In the sulcus group (n = 14), 3 patients reported no improvement in their voice quality postoperatively despite changes in the fundamental frequency (Table 3). The overall change in high-pitched voice perception was 86% (in 18 of the 21 patients), which was statistically significant (P < .05).

On the questionnaires, a feeling of tension or fatigue was noted by 12 patients in the sulcus group, the patient with a mutational falsetto disorder, and the patient with glottic scarring secondary to a laryngeal diphtheria infection in childhood. It was not a problem for the patients in the constitutional group. All of the patients who experienced the feeling of tension or fatigue gained a certain level of comfort postoperatively. The feeling disappeared in 6 of the 14 patients (43%) and was reduced in the remaining 8 (57%).

Eleven of 21 patients (52%) had a diplophonic voice quality verified by the subharmonic signal output in narrow-band spectrograms. The majority of these cases were in the sulcus group (n = 8). In the late postoperative evaluation, a diplophonic voice quality and subharmonic measures were still observed in 6 cases, including 1 patient who did not have this condition preoperatively. The technique improved the diplophonic voice quality by 45%, but the change was not statistically significant.

No complications, displacement, or implant extrusion were observed during the follow-up, and this was confirmed by computed tomography (Figure 6). In the first 2 to 3 weeks after surgery, all patients had minor glottic edema at the anterior third without compromising the airway, and this showed complete spontaneous resolution after 5 weeks.

The fundamental frequency is one of the major voice measures and is a determinant of vocal perception. A shift in one's frequency to the level of men, women, or children is possible because of anatomic variation or various abnormalities.¹³ This may lead to confusion in terms of the sex or age of the speaker in the listener's mind. The most common disorders leading to a high fundamental frequency are functional mutational falsetto, atrophic glottis abnormalities, and sulcus vocalis. Regardless of the cause, relaxation techniques are effective at reducing the fundamental frequency.⁶ However, the desired values are still controversial. Putting the fundamental frequency within the normal range of a desired population or group often results in success in terms of the correct perception and identification of sex, while levels outside these normal ranges are often misperceived.

Window anterior commissure relaxation laryngoplasty is less invasive than the standard technique for pitch-lowering surgery. However, the change and relaxation have limits. After a certain level is reached or exceeded, the voice quality deteriorates, which may require additional augmentation.

The comfort levels have not been well studied in vocal disorders or in terms of the surgical outcome. In our patients, it was apparent that, despite the perceptual quality projected by society, most patients, especially those with sulcus vocalis, were content to gain comfort during phonation after surgery. The feeling of tension or fatigue is an expected symptom in disorders with less efficient vibratory glottic closure and stiff vocal folds that are characteristics of patients with sulcus vocalis. The reduction in glottic efficiency necessitates extra effort by extralaryngeal muscles and breath support that results in voicing and speech discomfort. Relaxation laryngoplasty techniques provide some advantage while reducing the glottic tension and length; they are also expected to increase the mucosal pliability at the midglottis, where efficient vibratory dynamics occur. Another problem observed in the sulcus group was the presence of a diplophonic voice quality. Although no phonosurgical technique has been designed specifically to reduce diplophonia, the relaxation technique may provide some advantages to overcome this problem.

For most laryngoplasty techniques, the extent of the surgical area over the laryngeal cartilage and deep tissues determines the extent of the intralaryngeal edema.¹⁴ This may lead to minor vocal changes or severe airway problems postoperatively. When the surgical procedure is prolonged, intraoperative edema may result in difficulties with intraoperative voice monitoring and adjustment. The anterior commissure region of the larynx is resistant to edema formation because the tissue there is firmer. Therefore, window anterior commissure relaxation laryngoplasty as a less invasive modification of the standard technique seems advantageous in allowing additional fine adjustments.

Our study showed that the standard relaxation laryngoplasty technique devised by Isshiki can be refined to localize the surgery to the anterior commissure region, and that satisfactory results can be obtained with minimal complications.

Correspondence: Ismail Kocak, MD, MSc, Department of Otorhinolaryngology, Head and Neck Surgery, Yeditepe Universitesi Hastanesi, Bagdat Cad Poliklinigi, Bagdat Cad No:238 Goztepe, 34728 Istanbul, Turkey (ikocak@yeditepe.edu.tr).

Submitted for Publication: December 15, 2007; final revision received February 17, 2008; accepted March 2, 2008.

Author Contributions: Dr Kocak had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Kocak and Bengisu. Acquisition of data: Kocak, Dogan, Alkan Cakir, and Akpinar. Analysis and interpretation of data: Tadihan. Drafting of the manuscript: Kocak, Dogan, Alkan Cakir, Bengisu, and Akpinar. Critical revision of the manuscript for important intellectual content: Kocak and Tadihan. Statistical analysis: Tadihan. Administrative, technical, and material support: Kocak, Dogan, Alkan Cakir, Bengisu, and Akpinar. Study supervision: Kocak.

Financial Disclosure: None reported.