Hillel AT, Giraldez L, Samad I, Gross J, Klein AM, Johns MM. Voice Outcomes Following Posterior Cordotomy With Medial Arytenoidectomy in Patients With Bilateral Vocal Fold Immobility. JAMA Otolaryngol Head Neck Surg. 2015;141(8):728-732. doi:10.1001/jamaoto.2015.1136
Voice quality-of-life (VQOL) and perceptual voice outcomes are presumed to worsen following posterior cordotomy with medial arytenoidectomy for bilateral vocal fold immobility (BVFI); however, subjective and objective voice outcomes are not well studied in this postsurgical patient population.
To evaluate VQOL and perceptual voice outcomes following posterior cordotomy with medial arytenoidectomy for BVFI.
Design, Setting, and Participants
Retrospective medical record review of 15 patients with BVFI who underwent posterior cordotomy with medial arytenoidectomy at a tertiary care academic hospital from 2009 to 2012.
Suspension microlaryngoscopy was performed to expose the posterior glottis. A posterior cordotomy and medial arytenoidectomy was performed anterior to the vocal process of the vocal fold in a medial to lateral fashion.
Main Outcomes and Measures
Data included age, sex, tracheostomy status, number of cordotomies, and voice outcomes. Voice-Related Quality of Life (VRQOL) and Consensus Auditory Perceptual Evaluation of Voice (CAPE-V) data were collected preoperatively and postoperatively surrounding a single procedure. Comparisons within a single group were performed with a paired t test. Statistical significance was determined at P ≤ .05.
Eight patients (53%) were male, and 7 (47%) were female. Six patients (40%) required a tracheotomy at some point during treatment, 4 were successfully decannulated. For all 15 patients, the mean VRQOL scores improved 12 points from 47.33 to 59.33 after posterior cordotomy (P = .12). Mean CAPE-V overall severity scores in 13 patients increased 26 points after posterior cordotomy with medial arytenoidectomy from 38.12 to 62.77 (P = .01), indicating further deviance from normal.
Conclusions and Relevance
To our knowledge, this is the first study to compare VQOL with perceptual voice outcomes following posterior cordotomy with medial arytenoidectomy in a series of patients with BVFI. Patients who underwent posterior cordotomy in this study had significantly reduced perceptual voice outcomes with unchanged VQOL. While postcordotomy patients have a dysphonia that is noticeable to voice professionals, most patients in this study subjectively felt as though their voice improved after surgery. Surgeons should be aware of these factors when counseling patients considering cordotomy for BVFI.
Bilateral vocal fold immobility (BVFI) is a condition that can affect breathing, voice, and swallowing with potentially a severe impact on quality of life (QOL). Etiologies of BVFI are numerous and can broadly be categorized into either neurogenic or mechanical causes. Neurogenic causes are a result of surgical trauma from damage to bilateral recurrent laryngeal nerves, such as from previous thyroid, parathyroid, or mediastinal surgery.1 Mechanical causes are a result of physical damage to components of the larynx, such as the cricoarytenoid joint, which may result from a neoplasm or prolonged intubation causing posterior glottic stenosis.2 Both ultimately result in a narrowed glottal airway, and patients may present with a variety of symptoms including shortness of breath, dyspnea on exertion, stridor, inability to control vocal pitch, breathy voice quality, or coughing or choking while eating.
Treatment options for BVFI are principally surgical and include posterior cricoid split with graft, posterior cordotomy, medial arytenoidectomy, total arytenoidectomy, tracheostomy, or a combination of these procedures.2 Most of these are destructive, irreversible procedures aimed at increasing the glottic aperture, providing symptomatic relief, with the goal of preserving swallow and at least a functional voice. Posterior cordotomy is a particularly effective endoscopic procedure used to treat BVFI. First described by Dennis and Kashima3 in 1989, the endoscopic procedure involves using a laser to make a transverse cut in the posterior aspect of the more immobile vocal fold. This allows the incised vocal fold to shorten and displace anteriorly, increasing the size of the posterior glottic airway to improve air flow through the glottis while reducing the area of contact between vocal folds on phonation. Posterior cordotomy is commonly combined with medial arytenoidectomy, which destroys the vocal process and medial arytenoid cartilage to create more space posteriorly. The combined procedures are performed with the goal of avoiding a tracheotomy or to assist with decannulation, and according to one study,4 results in decannulation rates as high as 92%. Complications with posterior cordotomy are rare; however, because it is a destructive procedure, patients may complain about voice changes, and some may develop dysphagia.2,4,5
While many otolaryngologists assess voice through the use of validated QOL instruments and perceptual evaluation, few studies have compared preoperative and postoperative voice outcomes surrounding posterior cordotomy.6- 8 The objective of our study is to assess voice outcomes in patients who undergo posterior cordotomy for BVFI, using the Voice-Related Quality of Life (VRQOL) and Consensus Auditory Perceptual Evaluation of Voice (CAPE-V) instruments. We hypothesize that VRQOL and CAPE-V scores will correlate, and voice quality will worsen (VRQOL to decrease, CAPE-V to increase) postoperatively.
After obtaining Emory University institutional review board approval, a retrospective medical record review was performed. Patients diagnosed as having BVFI who had undergone at least 1 posterior cordotomy with medial arytenoidectomy from July 2009 to July 2012 at the Emory Voice Center were included in the study. Patients were excluded from the study if they had undergone any additional procedures besides a posterior cordotomy with medial arytenoidectomy, other than tracheostomy, for the treatment of the BVFI. Data collected included age and sex of the patient, date of surgical procedure(s), preoperative and postoperative VRQOL/CAPE-V scores, preoperative and postoperative tracheostomy status, and etiology of BVFI.
Posterior laser cordotomy and medial arytenoidectomy was performed in all patients using either a Lumenis UltraPulse carbon dioxide SurgiTouch laser (Lumenis) or a potassium titanyl phosphate (KTP) laser (American Medical Systems). Suspension microlaryngoscopy was performed to expose the posterior glottis. A posterior cordotomy was performed anterior to the vocal process of the vocal fold in a medial to lateral fashion. The vocal fold was completely incised toward the thyroid cartilage laterally, displacing it anteriorly, and leaving a wedge-shaped posterior glottic defect. Next, medial arytenoidectomy was performed beginning with destruction of the vocal process and extending posterolaterally until an adequate airway is created. The Figure demonstrates a preoperative appearance of limited glottic airway (Figure A), an immediate postoperative photograph (Figure B), and 5-week postoperative photograph (Figure C) following right posterior cordotomy with medial arytenoidectomy. The preferred method of ventilation for the procedure was pressurized supraglottic jet ventilation with a canula or via a laser safe tube through the tracheostomy.
Two validated scoring instruments were used to measure voice outcomes: VRQOL and CAPE-V.6,7 Data included both preoperative and postoperative results surrounding a single procedure.8 If multiple VRQOL scores were recorded, the closest one prior to surgery was taken as the “preoperative” score, while the first score obtained at least 4 weeks postoperatively, was taken to be the “postoperative” score. This postoperative “grace period” was given to allow for immediate postoperative voice changes from swelling and inflammation to settle down, prior to patient self-evaluation.
Overall CAPE-V scores were obtained via high-quality preoperative and postoperative audio voice recordings. Two experienced speech-language pathologists with expertise on voice disorders were blinded to speaker identity, the nature of the voice problem, and the reason for rating the voice recordings. All audio recordings were randomized and evaluation by the raters was performed independent of each other. Scores for each patient were then averaged to produce a single preoperative and postoperative CAPE-V overall severity score.
Statistical analysis was performed using Graph Pad (GraphPad Software Inc). The Pearson Product Moment Coefficient Correlation (PPMC) was used to determine interrater reliability between different rater’s scores for preoperative CAPE-V assessment and postoperative CAPE-V assessment independently. Comparisons within a single group over time were performed with a paired t test. Statistical significance was determined at P ≤ .05.
Fifteen patients who were diagnosed as having BVFI subsequently underwent at least 1 cordotomy procedure. Eight patients (53%) were male, and 7 (47%) were female. Patients ranged in age from 36 to 80 years (mean age, 56.9 years). Eight patients (53%) had a neurologic etiology, while 7 (47%) had a mechanical etiology of their paralysis. Of the 8 patients with BVFI of neurologic origin, 3 (20%) were a complication of prior thyroidectomy, 3 (20%) were idiopathic, 1 (6%) was associated with a Chiari malformation, and 1 (6%) was a complication of mediastinal surgery. Of the 7 patients who had a mechanical etiology for their paralysis, 6 (40%) were a result of posterior glottic stenosis from prolonged or traumatic intubation, and 1 (6%) had postradiotherapy fixation of the cricoarytenoid joints.
Of the 15 patients, 12 (80%) underwent 1 posterior cordotomy, 2 (13%) underwent 2 cordotomies, and 1 (7%) underwent 4 cordotomies. Patients who underwent multiple procedures had a cordotomy of their contralateral vocal fold performed the additional times. The patient who underwent 4 posterior cordotomies was a 61-year-old woman with bilateral vocal fold paralysis from a Chiari malformation. She required multiple procedures on both vocal folds owing to a pattern of exuberant granulation tissue that rapidly replaced each surgically created defect. Ultimately an adequate glottic airway was maintained, and the patient never required a tracheostomy. The patient’s VRQOL was 50 preoperatively and gradually decreased to a 15 after completing her series of operations. Six patients (40%) required a tracheotomy at some point during their surgical care. Four (27%) were successfully decannulated on follow-up visits, while 2 (13%) remained with a tracheotomy at final review. Of these 2 patients, 1 continued to require mechanical ventilation while asleep, while the other was followed for 1 month before being subsequently lost to follow-up. Follow-up times ranged from 1 to 6 months postoperatively, with a mean time to follow-up of 2.6 months.
All 15 patients had complete VRQOL data sets collected (Table). Preoperative VRQOL scores ranged from 0 to 77.5, with a mean (SD) score of 47.3 (23.8). Postoperative VRQOL scores ranged from 25.0 to 100, with a mean score of 59.3 (23.5) (P = .12). Mean VRQOL scores improved 12.0 points from 47.3 to 59.3 after posterior cordotomy (P = .12). The greatest improvement in VRQOL score was seen in a patient with a gain of 55.0 points (preoperative score of 0 and postoperative score of 55). Eight patients (53%) reported an increase in VRQOL scores of 10.0 points or greater, and scores in 4 (27%) remained unchanged with a variation of ±10 points, whereas 3 patients (20%) reported a drop in VRQOL score postoperatively of more than 10 points. Of these 3 patients, only 1 went on to have an additional cordotomy. Two patients (13%) reported a preoperative VRQOL score of 0, with postoperative scores improving an average of 43.8 points.
Matched preoperative and postoperative voice recordings were available for 13 of 15 patients for CAPE-V assessment (Table). Preoperative CAPE-V scores ranged from 12.5 to 96.0, with an overall mean score of 39.5. Postoperative CAPE-V scores ranged from 28.5 to 89.0 with an overall mean score of 65.5. Mean CAPE-V scores increased 26.0 points after posterior cordotomy from 38.11 (28.6) to 62.3 (18.4) (P = .01). The greatest improvement in CAPE-V score was seen in a patient with a decrease of 13.5 points (preoperative: 51.0, postoperative: 37.5). This was the only patient (8%) with a decrease, whereas 8 (62%) demonstrated an increase in postoperative scores, and 4 (31%) remained unchanged within a range of ±10 points on postoperative assessment of overall severity of voice. Interrater analysis was performed for the patients who had preoperative and postoperative CAPE-V scores by both raters. Analysis was performed on the cohort of 13 patients and demonstrated a very strong positive interrater correlation at both the preoperative (r = 0.86; P < .001) and postoperative time points (r = 0.82; P = .001).
The results of this study support that posterior cordotomy with medial arytenoidectomy may offer patients improved or unchanged voice quality of life, (VQOL) despite the decrease in overall voice severity perceived by professionals. The mean change VRQOL (47 to 59) demonstrated a trend toward improvement. Individually, 8 patients demonstrated an improvement, 4 had no change, with only 3 subjectively reporting a worse VQOL. This is in contrast to the perceptual evaluation of overall voice severity, which showed a significant decrease following posterior cordotomy. One possible explanation for this distinction is the impact of airway stenosis on patient self-assessment of voice. Following cordotomy, many patients experience relief from symptoms related to expiratory air trapping under the adducted vocal folds, resulting in a sense of breathlessness when phonating. Item 2 in the VRQOL is “I run out of air and need to take frequent breaths when talking,” and improvement in this question may reflect improvement in glottic airflow.6 The results in this study show a similar trend to the improved VQOL seen in patients with laryngotracheal stenosis following dilation procedures.9
These results are in agreement with other literature, which report unchanged or improved VQOL following cordotomy and/or medial arytenoidectomy. Gorphe et al10 demonstrated that the mean emotional parameter of voice handicap index (VHI) of 120 statistically improved in 8 (of 20) patients following laser medial arytenoidectomy for bilateral vocal fold paralysis, while the global, physical, and functional parameters did not. Similarly, Bosley et al11 showed no significant difference in VHI of 10 for 4 of 11 patients who underwent cordotomy and 5 of 6 patients who underwent medial arytenoidectomy. This study included a perceptual assessment of voice quality; however, it did not report results on change in global voice severity. Finally, Bernstein et al12 included 3 patients with preoperative and postoperative QOL and perceptual voice evaluation data, which were unchanged for both outcomes following cordotomy. In contrast, the larger sample size included in our study demonstrated clinically significant worsening in CAPE-V scores; an expected outcome from the surgeon’s perspective following transection of the vocal fold.
The findings in this study and others may help better counsel patients with BVFI considering posterior cordotomy and/or medial arytenoidectomy. Current preoperative counseling includes discussion of the balance between improved breathing through a larger glottic airway and the subsequent reduced voice quality and volume, as well as the possibility of requiring multiple procedures to achieve an adequate end result.5 While the patient’s voice may lessen in quality and volume to others, there is a good possibility that he or she may feel as though the voice has improved or not changed after surgery. Surgeons should be aware of this when counseling patients with BVFI on the benefits and risks of posterior cordotomy, especially in those patients who are particularly bothered by running out of air or taking frequent breaths when talking.
This study does have limitations. While it is a larger study assessing voice quality and QOL following cordotomy, it is a small retrospective study. Our study had relatively short follow-up times. Nevertheless, it was sufficient time to decannulate 4 of 6 patients with tracheostomy. Furthermore, voices would not be expected to deteriorate with further time removed from cordotomy; usually a small amount of scar fills the surgical defect, and patients learn to generate compensatory phonatory techniques that have a tendency to improve voice. Another drawback is that individual questions in this QOL instrument were not available for retrospective review, so the effect of item 2 in the VRQOL could not be assessed. However, it should be noted that the improvement in postcordotomy VRQOL change is not be explained by a single question alone.6 Finally, 2 different lasers were used to perform the cordotomy and medial arytenoidectomy. Laser choice of KTP or carbon dioxide was determined by surgeon preference. While each laser is a surgical instrument that cuts and coagulates, it is possible that the laser results in distinct scarring patterns that may contribute to different voice outcomes. What is more likely is that each laser cuts through the vocal fold leaving a foreshortened vocal fold combined with a posterolateral aperture that leads to the reduced quality of voice.
Patients who underwent posterior cordotomy for BVFI in this study had significantly reduced perceptual voice outcomes with unchanged VQOL. While postcordotomy patients have a dysphonia that is noticeable to voice professionals, most patients in this study subjectively felt as though their voice improved after surgery. Surgeons should be aware of these factors when counseling patients considering cordotomy for BVFI.
Corresponding Author: Michael M. Johns III, MD, Emory Voice Center, Department of Otolaryngology, Emory University School of Medicine, MOT–Ste 9-4400, 550 Peachtree St, Atlanta, GA 30308 (email@example.com).
Submitted for Publication: March 5, 2015; final revision received April 13, 2015; accepted May 7, 2015.
Published Online: June 25, 2015. doi:10.1001/jamaoto.2015.1136.
Author Contributions: Drs Hillel and Giraldez had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs Hillel and Giraldez shared equally in this research and are to be considered co–first authors.
Study concept and design: Giraldez, Johns.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Hillel, Giraldez, Samad, Gross, Johns.
Critical revision of the manuscript for important intellectual content: Giraldez, Klein, Johns.
Statistical analysis: Hillel, Giraldez, Gross.
Administrative, technical, or material support: Giraldez, Johns.
Study supervision: Klein, Johns.
Conflict of Interest Disclosures: None reported.