Representative photographs depict CO2 laser evaporation of a vocal polyp and the postoperative findings.
The CO2 laser was used to make a mucosal incision into the superior lateral surface of the vocal fold to vaporize redundant mucosa and accumulated fluid, resulting in meaningful improvement 6 weeks after surgery.
Blue area indicates field of view. A, Openings for the laser fiber and camera lens are located on the left and right sides of the patient, respectively. B, Surgery to treat the right vocal fold lesion was hindered by a limited field of view. C, Pulling the tip of the laryngoscope away from the lesion increased the field of view but reduced the precision of the technique.
Hu H, Lin S, Hung Y, Chang S. Feasibility and Associated Limitations of Office-Based Laryngeal Surgery Using Carbon Dioxide Lasers. JAMA Otolaryngol Head Neck Surg. Published online February 16, 2017. doi:10.1001/jamaoto.2016.4129
Can laryngeal surgery performed using a carbon dioxide laser in an office setting be considered a feasible treatment option for laryngeal lesions?
In this study of 49 consecutive laryngeal surgical procedures performed using a carbon dioxide laser in 40 patients in an office-based setting in Taiwan, patient tolerance was good in 47 procedures, and one patient experienced a complication in 1 procedure.
With meticulous patient selection, office-based laryngeal surgery performed using a carbon dioxide laser appears to be a feasible treatment option for various types of vocal lesions.
There are few reports evaluating awake, office-based carbon dioxide (CO2) laser surgery for laryngeal lesions. To date, this study was the largest reported case series of office-based laryngeal surgery by fiber delivery CO2 laser. Office-based laryngeal surgical procedures have become increasingly popular. Technical problems and treatment outcomes associated with the use of a CO2 laser for office-based laryngeal surgery have yet to be fully addressed.
To discuss a single institution’s clinical experience with office-based CO2 laser laryngeal surgery and the feasibility and limitations associated with this procedure.
Design, Setting, and Participants
This retrospective study evaluated 49 laryngeal surgical procedures performed using a CO2 laser in 40 consecutive adult patients at a single institution in Taiwan from July 1, 2014, through September 30, 2015. Laryngeal lesions treated included vocal fold leukoplakia (n = 13), benign vocal fold lesions (n = 10), Reinke edema (n = 4), recurrent respiratory papillomatosis (n = 6), and lesions outside the vocal folds (n = 7).
Office-based laryngeal surgery performed using a CO2 laser under topical anesthesia.
Main Outcomes and Measures
Videolaryngoscopy was performed on all patients at each follow-up point. Among patients with benign vocal lesions and Reinke edema, videolaryngostroboscopy, voice laboratory measurements, perceptual measurements of vocal quality, and subjective evaluations were conducted before and after surgery.
Among the 40 patients included in this study (28 men [70%] and 12 women [30%]; median [range] age, 56 [29-83] years), median follow-up time was 6.5 months (range, 1-21 months). Among the 49 procedures, 2 (4%) could not be tolerated by patients owing to severe gag reflex and laryngeal hypersensitivity, 6 (12%) could not completely evaporate lesions owing to an inadequate surgical field or laryngeal instability, and 1 (2%) led to a complication (ie, mild vocal fold wound stiffness). In addition, 2 patients with premalignant vocal fold leukoplakia showed lesion recurrence in the subglottic area. Among patients with benign vocal lesions and Reinke edema, postoperative phonatory function showed large improvements in jitter (effect size, 0.61; median difference, −0.98%; 95% CI, −1.57% to −0.11%), noise to harmonic ratio (effect size, 0.63; median difference, −0.02; 95% CI, −0.07 to −0.01), maximal phonation time (effect size, 0.61; median difference, 3.6 seconds; 95% CI, 1.9 to 8.8 seconds), and Voice Handicap Index–10 score (effect size, 0.60; median difference, −7; 95% CI, −12 to −2).
Conclusions and Relevance
Office-based laryngeal surgery performed using a CO2 laser was shown to be a feasible treatment option for various types of vocal lesions. However, patients should not undergo this procedure if they have multiple bulky lesions or lesions involving the subglottic area, the laryngeal ventricle, or (in cases of inadequate laryngeal stability) the free edge of a vocal fold.
Office-based laryngeal surgery, including intralesional corticosteroid injection, injection laryngoplasty, and laryngeal laser surgery, have become increasingly popular in recent years.1- 3 These types of surgical procedures are less precise than those involving general anesthesia; however, they appear to be a worthwhile alternative treatment option based on their low cost and low morbidity rates.3,4 Photoangiolytic lasers, such as the 585-nm pulsed-dye laser and the 532-nm potassium titanyl phosphate laser, have been widely used in office-based laryngeal surgical procedures. Indeed, previous studies3,5- 9 have reported successful treatment of multiple types of vocal lesions, including papilloma, vocal process granuloma, ectasia, varix, polyp, Reinke edema, and glottal dysplasia, using these methods. The angiolytic properties of the aforementioned lasers shrink lesions through photothermolysis9; however, outcomes are hard to anticipate immediately after the procedure owing to the fact that disease regression can take several weeks.10
Conversely, carbon dioxide (CO2) lasers have long been used as a fundamental tool for the treatment of laryngeal disease. The wavelength (10 600 nm) of the CO2 laser is absorbed by water, thereby making it possible to vaporize lesions with a high percentage of intracellular water.3 Physicians began to use CO2 lasers for office-based laryngeal surgery when flexible, mechanically robust, biocompatible, low–optical-loss CO2 laser fibers became available. These devices feature an omnidirectional mirror structure around their hollow core, which allows the CO2 laser to fit through the working channel of a flexible laryngoscope.10 However, previous reports pertaining to the use of CO2 lasers in office-based laryngeal surgical procedures3,10,11 are anecdotal. Indeed, technical problems and treatment outcomes associated with office-based CO2 laser laryngeal surgery have yet to be fully addressed. In this report, we sought to help fill this research gap by relaying our clinical experience with this surgical procedure and its feasibility and limitations.
We conducted a retrospective analysis of 40 consecutive patients who received office-based CO2 laser laryngeal surgery at the Voice Clinic of Cheng Hsin General Hospital, Taipei, Taiwan, from July 1, 2014, through September 30, 2015. All procedures were performed by one of us (S.-Y.C.). This study was approved by the institutional ethics and research committee of the Cheng Hsin General Hospital. Because the study was a retrospective medical record review, informed consent was not required by the decision of the institutional ethics and research committee.
All patients received an intramuscular injection of 1 mg of atropine sulfate 30 minutes before surgery to reduce the production of saliva. The procedure began with the spraying of lidocaine hydrochloride solution, 4%, over the nasal cavity, nasopharynx, oropharynx, hypopharynx, and larynx. Bilateral superior laryngeal nerve blocks were then applied (using 1 mL of lidocaine hydrochloride solution, 2%, delivered bilaterally). After completion of the nerve block, transnasal insertion of a flexible laryngoscope enabled the delivery of 5 to 10 mL of lidocaine hydrochloride solution, 4%, to the larynx through the working channel.
After the area was sufficiently anesthetized, the CO2 laser waveguide (AcuPulse 40WG with FiberLase CO2 fiber; Lumenis) was used to perform laryngeal surgery. The laser waveguide was first passed through the protective sheath, whereupon the protective sheath was inserted into the working channel of the flexible laryngoscope (ENF-VT2; Olympus). Power settings were initially set at 5 W in super-pulse mode with timing set at 0.05 second on and 0.01 second off. The laser settings were adjusted according to the location of the lesion being evaporated. Specifically, for lesions involving the free edge of the vocal fold, power was decreased to 3 W. For lesions outside the vocal folds, power was increased until optimal ablative effects were observed (range, 5-10 W). During surgery, patients had control over a saliva ejector (a hollow, perforated suction tube used to evacuate saliva and liquid debris from the oral cavity), which allowed them to remove saliva and smoke.
For patients with benign vocal lesions and Reinke edema, videolaryngostroboscopy, voice laboratory measurements, perceptual measurements of vocal quality, and subjective evaluations were conducted before and after surgery. For patients with other lesions (ie, laryngeal lesions outside the vocal fold, vocal fold leukoplakia, and recurrent respiratory papillomatosis), videolaryngoscopy was performed at each follow-up point. Acoustic recordings and phonatory function were evaluated by 2 of us who are speech pathologists (S.-Y.L. and Y.-T.H.). A computerized speech laboratory (model 4500; KayPENTAX) was used to analyze 3-second samples of voice data in which jitter, shimmer, and noise to harmonic ratio values were tabulated. The maximum phonation time was measured using a waveform display of the acoustic signals. We obtained perceptual ratings of voice quality by grade, roughness, breathiness, asthenia, and strain (GRBAS rating; scores range from 0-3, with higher scores indicating a severer condition),12 which were also recorded by the 2 speech pathologists. The Voice Handicap Index–10 questionnaire13 was used to collect subjective ratings from patients (range, 0-40, with higher scores indicating severer voice disorder).
All statistical analyses were performed using SPSS (version 18.0.0; SPSS Inc). Nonparametric statistics were used because the assumptions of parametric tests were not met. Results that involve continuous data are presented as the median (range). We calculated the effect size, median difference, and 95% CI around the difference. We used Wilcoxon signed rank tests to compare paired nonparametric data.
Forty-nine consecutive office-based laryngeal surgical procedures were performed using a CO2 laser. These procedures were performed to treat a variety of vocal lesions in 40 patients, including vocal fold leukoplakia (n = 13), benign vocal fold lesions (n = 10), recurrent respiratory papillomatosis (RRP) (n = 6), laryngeal lesions outside the vocal fold (n = 7), and Reinke edema (n = 4). The sample included 28 men (70%) and 12 women (30%), with a median age of 56 years (range, 29-83 years). Follow-up data were available for all patients, and median follow-up time was 6.5 months (range, 1-21) months. Among the 49 procedures, 2 (4%) could not be completed because patients exhibited a severe gag reflex or laryngeal hypersensitivity. In 6 of the 49 procedures (12%), we were unable to eliminate the lesions completely. Specifically, these instances involved the treatment of RRP (n = 2), Reinke edema (n = 2), vocal fold leukoplakia (n = 1), and prolapsed ventricle (n = 1). Only 1 of the 49 procedures (2%) was complicated by mild vocal fold wound stiffness (Table 1). The reasons underlying the incomplete surgical procedures and complications are discussed in later sections.
A total of 16 procedures were performed on 13 patients with vocal fold leukoplakia (1-2 procedures per patient). Previous pathologic reports (obtained through transoral or transnasal flexible fiberoptic biopsies) indicated that among these 13 patients, 1 had carcinoma in situ; 4, severe dysplasia; 2, moderate dysplasia; 2, mild dysplasia; and 1, keratosis with focal atypia. Pathologic reports were not available for the remaining 3 patients.
In every case, a CO2 laser was used to evaporate all visible lesions. Median follow-up time was 10 months (range, 1-20 months) for patients in this group. One patient was unable to tolerate the procedure, which necessitated surgery performed in an operating theater under general anesthesia. One patient with keratosis and focal atypia could not undergo the entire procedure because the surgical field was partially hidden by the right false vocal fold. This patient later had the lesion removed using a cold instrument in an in-office procedure. Among the remaining 11 patients, 9 (82%) had complete remission after office-based CO2 laser surgery (1 procedure in 6 patients and 2 procedures in 3 patients). Carcinoma in situ recurred in the subglottic region of the larynx in 1 patient at 9 months, and moderate dysplasia recurred in the subglottic region of the larynx in 1 patient at 7 months. Pathologic reports indicated that the recurrent lesions observed in both patients were squamous cell carcinoma. Both patients underwent subsequent surgery in the operating theater. No complications were observed in this group.
A total of 10 procedures were performed in 6 patients with RRP (1-2 procedures per patient) who had undergone previous surgical procedures for the disease (>10 times per patient). In the present study, a CO2 laser was used to evaporate all visible lesions observed in each of the patients with RRP. Median follow-up time was 5 months (range 2-21 months) for patients in this group. One patient was unable to tolerate the procedure and received conventional surgery in an operating theater. Surgery in 2 patients could not be completed owing to the presence of multiple bulky lesions. The stability of patients with bulky lesions gradually decreased during longer procedures (ie, even if the patients could bear the uncomfortable sensation of the procedure, they may still develop frequent involuntary vocal movement, swallowing, or cough, which could reduce the effectiveness of the surgery), thereby necessitating conventional surgery in an operating theater. No complications were observed in this group.
A total of 8 procedures were performed on 7 patients (1-2 procedures per patient) who presented with false vocal fold granuloma (n = 2), arytenoepiglottic fold cyst (n = 2), prolapsed ventricle (n = 2), or arytenoid granuloma (n = 1). All lesions were evaporated using a CO2 laser. Median follow-up time was 10 months (range, 1-19 months) for patients in this group. The procedure could not be completed when performed on 1 patient with a prolapsed ventricle (resulting from a previous autologous fat injection laryngoplasty) owing to sheltering of the surgical field by the right false vocal fold. That patient subsequently underwent conventional surgery in an operating theater. Recurrence and complications were not observed in this group.
A total of 10 procedures were performed on 10 patients who presented with vocal polyp (n = 5), vocal cyst (n = 3), and vocal nodule (n = 2). Eight of the 10 patients had previously received office-based intralesional corticosteroid injections but did not achieve complete remission. In the present study, a CO2 laser was used to evaporate all lesions in patients belonging to this group (Figure 1). Median follow-up time was 3.5 months (range, 1-18 months) for patients in this group. Recurrence was not observed in any patient.
One patient who had a nodule on the right vocal fold showed obvious involuntary vocal fold movement during surgery. That patient experienced mild right vocal fold wound stiffness (diagnosed using postoperative videolaryngostroboscopic findings); however, postoperative voice outcomes still improved. The patient was treated by injecting corticosteroids into the stiff area (in an office-based procedure), with subsequent signs of improvement.
A total of 5 procedures were performed on 4 patients with Reinke edema (range, 1-2 procedures per patient). During each procedure, a mucosal incision was created by applying the CO2 laser to the superior lateral surface of the vocal fold to vaporize redundant mucosa and accumulated fluid (Figure 2). Median follow-up time was 7.5 months (range, 1-16 months) for patients in this group. Notable edema reductions were observed in vocal folds during follow-up. However, mucosal incisions in 2 patients were incomplete owing to sheltering of the surgical field by the right false vocal fold. Those patients were treated by injecting corticosteroids into the remnant edematous area in an office-based procedure. No complications were observed in this group.
Table 2 lists the results of voice laboratory measurements, perceptual measurements of vocal quality, and subjective evaluations for patients with benign vocal lesions and Reinke edema. Postoperative values for maximum phonation time (median difference, 3.6 seconds; 95% CI, 1.9-8.8 seconds), jitter (median difference, −0.98%; 95% CI, −1.57% to −0.11%), noise to harmonic ratio (median difference, −0.02; 95% CI, −0.07 to −0.01), and the Voice Handicap Index–10 score (median difference, −7; 95% CI, −12 to −2) showed large improvements compared with preoperative values.
We found that, in an office-based setting, the CO2 laser is effective in treating various types of laryngeal lesions. For example, in cases of vocal fold leukoplakia, 11 of 13 patients (85%) received their entire treatment in the office, and 9 of these (82%) showed complete resolution. Koufman et al3 previously reported that 80% of 25 patients did not require subsequent treatment for vocal fold leukoplakia after receiving office-based pulsed-dye laser surgery; and Zeitels and Burns14 reported that, among 29 patients, the complete remission rate of dysplasia achieved through office-based potassium titanyl phosphate laser surgery was 62%. Despite this study only including a limited number of cases, we found that CO2 laser surgery can achieve treatment outcomes comparable to those reported in the aforementioned studies on photoangiolytic laser surgery. Moreover, in most cases, use of a CO2 laser to treat granulation and cystic lesions in the laryngeal region outside the vocal fold also resulted in a good prognosis on par with that reported for pulsed-dye laser surgery.3
Office-based CO2 laser surgery immediately evaporated lesions in patients with RRP and showed treatment effectiveness similar to that of laryngoscopic microsurgery (which is performed under general anesthesia), although laryngoscopic microsurgery is a more precise technique. Office-based CO2 laser surgery provides enormous benefits for patients with RRP because multiple procedures have usually been required to control the disease. Office-based procedures can decrease the time and cost of treatment and reduce the need for general anesthesia.
Most of the patients treated in our clinic for benign vocal lesions had previously received office-based injections of intralesional corticosteroids, which led to satisfactory outcomes, as reported in previous articles.1,15 Specifically, the initial corticosteroid injections were used to shrink lesions to tiny superficial nodules, which could then be evaporated using the CO2 laser. To achieve this, we set the power to 3 W in super-pulse mode to minimize thermal effects as much as possible. In contrast, treatment of patients with Reinke edema involved using the CO2 laser to make mucosal incisions in the superior lateral surface of the vocal fold to vaporize redundant mucosa and accumulated fluid. Although office-based laryngeal surgery performed using a CO2 laser is less precise than procedures that are performed in an operating theater,16 follow-up results were fair. Furthermore, in patients with Reinke edema or benign vocal lesions, maximum phonation time, jitter, and noise to harmonic ratio showed major improvements that were comparable to results reported in a previous study on CO2 laser–assisted phonosurgery (performed under general anesthesia).17 For these patients, subjective ratings (the Voice Handicap Index–10 score) also showed significant improvements after surgery. Finally, results of perceptual analysis (ie, roughness and asthenia measured by the GRBAS rating) indicated improved outcomes; however, these findings lacked statistical significance, perhaps owing to the small number of patients included in our study.
Despite the overall success of CO2 laser surgery, the procedure nonetheless has some limitations. Office-based surgical procedures are less precise than those involving general anesthesia, which may lead to local recurrence or extrathermal injury. Furthermore, lesions could be excised and removed by CO2 laser in the operating theater, whereas vaporization of lesions is the rule in the office setting. These drawbacks should be weighed against potential benefits before the procedure is performed.
Technically, the laser was only able to access lesions located on the perpendicular axis of the delivery fiber.9 In our review, 1 patient with carcinoma in situ and 1 patient with moderate dysplasia experienced recurrence in the subglottic area. These results highlight the importance of identifying the correct location of premalignant and malignant lesions before surgery. Indeed, we suggest that office-based laryngeal surgery using CO2 lasers should not be performed on lesions located below the free margin of vocal folds or inside laryngeal ventricles.
On the other hand, a number of lesions were visible through a flexible laryngoscope but remained beyond the easy access of the laser beam because the camera lens and working channel used different openings at the tip of scope. In the flexible laryngoscope used in the present study, the working channel opening (used by the laser) and the camera lens were located on the left and right sides of the patient, respectively, during surgery. Thus, lesions on the right side of the larynx were sometimes outside the field of view, hidden by the right arytenoepiglottic fold or the right false vocal fold. The problem could be partially resolved by pulling the tip of the scope away from the lesion and extending the laser fiber, thereby providing the surgeon with a better surgical field. Unfortunately, this resolution would detract from the accuracy of the laser by increasing the distance between the camera lens and lesion (Figure 3). Flipping the scope 180° may theoretically resolve the problem, but we need more clinical experience to verify it. In the present study, 1 patient with vocal leukoplakia, 1 patient with a prolapsed ventricle, and 2 patients with Reinke edema outside the field of view were partially treated in this manner.
Ensuring that patients are able to tolerate the procedure in a stable manner is key to ensuring success in surgery. Halum and Moberly10 reported that patients were equally tolerant of office-based CO2 laser laryngeal surgery and pulsed-dye laser surgery. In that study, the tolerance rate for both types of surgery was 100%; however, only a small number of patients were considered. In our study, 38 of 40 patients (95%) tolerated the procedure, which is comparable to the tolerance rate reported in other studies. However, we found it difficult to anticipate which patients would be unable to tolerate the procedure or to complete the procedure owing to laryngeal instability.
Multiple factors may affect the stability of patients. First, the vaporization caused by the laser generates smoke, to which some patients are sensitive. These patients exhibited a cough reflex after exposure to smoke, even when adequate topical anesthesia was administered. Nonetheless, the patients were given control of a saliva ejector to ameliorate this problem. Second, secretions can lead to choking and blurred scope vision during surgery. Thus, instruments with thinner fibers would be better able to remove secretions through residual space in the working channel.18 In the present study, the diameter of the CO2 laser fiber was 1.04 mm, which is larger than that of photoangiolytic laser fibers. The CO2 laser fiber and its protection sheath occupy nearly all the space in the working channel, which prevented the working channel from being used to remove secretions. As an alternative, we used atropine injections and a saliva ejector to ameliorate this problem; however, these solutions proved to be less than ideal. The problem of secretion removal during CO2 laser surgery should be addressed by future research.
Third, in cases where lesions were located on the right side of the larynx, the scope tip risked coming into contact with the right arytenoepiglottic fold or the right false vocal fold and causing a gag reflex (owing to the aforementioned aspect of the scope tip design). Fourth, some of the patients presented paroxysmal involuntary vocal movement during surgery, which reduced the precision of ablation in cases where the lesion was limited to the free edge of the vocal fold. Vocal movement can also cause scar formation by exacerbating thermal injuries. One patient in this review experienced vocal fold wound stiffness due to this problem. Thus, we suggest that CO2 laser surgery not be performed on unstable patients whose lesions involve the free edge of the vocal fold.
Furthermore, some patients exhibited involuntary glottic or supraglottic hyperadduciton during the procedure, which restricted access to lesions located at the anterior commissure. Unconscious emotional stress may have caused these phenomena; however, further study will be needed to identify strategies that can address this problem. Finally, we found that it was better to treat multiple bulky lesions in an operating theater under general anesthesia because the stability of patients gradually decreases during long procedures under topical anesthesia. However, determining the threshold size at which lesions can be definitively classified as bulky is difficult owing to different levels of stability among patients.
This study was limited by the small number of patients undergoing evaluation and the lack of control groups. Additional studies will be required to better validate this approach as a therapeutic strategy.
This article describes our experience performing office-based laryngeal surgery using a CO2 fiber laser. We determined that this technique is a feasible treatment option for vocal fold leukoplakia, RRP, benign vocal fold lesions, Reinke edema, and laryngeal lesions outside of the vocal fold. However, additional studies are required to further validate the efficacy of this procedure. Conversely, office-based CO2 laser laryngeal surgery proved to be unsuitable for lesions involving the subglottic area or the laryngeal ventricle, multiple bulky lesions, and lesions located at the free edge of the vocal fold in unstable patients. Indeed, the location of lesions and the stability of patients have critical influence over the success of surgery.
Corresponding Author: Shyue-Yih Chang, MD, Voice Clinic, Department of Otolaryngology, Cheng Hsin General Hospital, No. 45, Cheng Hsin Street, Pai-Tou, Taipei, Taiwan 112 (email@example.com).
Accepted for Publication: October 22, 2016.
Published Online: February 16, 2017. doi:10.1001/jamaoto.2016.4129
Author Contributions: Drs Lin and Hung contributed equally to this work. Dr Chang 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: Hu, Chang.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Hu.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Hu.
Administrative, technical, or material support: All authors.
Study supervision: All authors.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
Additional Contributions: Tao-Hsin Tung, PhD, Department of Medical Research and Education, Cheng Hsin General Hospital, and Faculty of Public Health, School of Medicine, Fu-Jen Catholic University, Taipei, Taiwan, assisted in revising the manuscript. He was not compensated for this work.