2018 Section 6 - Laryngology, Voice Disorders, and Bronchoesophalogy

may be beneficial for the patients to understand their postoperative condition. However, there are relatively few longitudinal studies of vocal outcome in patients after TLM for early glottic carcinoma. 16,17 In this regard, we hypothesized that longitudinal analysis of vocal out- comes after TLM according to the surgical extent may provide some rationale to decide the optimal treatment for patients with T1 glottic carcinoma from the perspec- tive of vocal outcomes. We conducted a retrospective review of the medical records of 211 patients who underwent TLM for glottic carcinoma at our center. Among them, 134 patients underwent TLM as an initial treatment for T1 glottic carcinoma, and 57 patients who underwent voice analysis before the surgery, within 3 months after surgery (early postoperative period), and more than 6 months after surgery (late postoperative period) were enrolled in this study. Among the 134 patients who underwent TLM for T1 glottic carcinoma, 21 among 35 patients (60.0%) who under- went lesser-extent cordectomy (type I, II) were included, whereas 36 among 99 patients (36.4%) who underwent larger- extent cordectomy (type III, IV, V) were enrolled in this study. The patients included 50 T1a cases (87.7%) and 7 T1b cases (12.3%) and all were treated with TLM as a single modality treatment without any additional laser resection or radiation therapy. Patients were recommended to have voice rest for 1 week after surgery, and antireflux medicines were prescribed. On the day of the visit for the first postoperative voice exam, an experienced speech therapist educated the patients regarding vocal hygiene and exercise for vocal rehabilitation. The patients were followed monthly for the first year, bimonthly for the sec- ond year, quarterly for the third year, and biannually for the fourth and fifth years. Videostroboscopic examination was per- formed at every postoperative visit. This study was approved by the institutional review board of our center. Transoral Laser Microsurgery TLM was conducted as it has been described in our previous study. 1 The cordectomies were classified into six types according to the recent Proposal for Revision of the European Laryngologi- cal Society Classification of Endoscopic Cordectomies 18 : sub- epithelial (type I), subligamental (type II), transmuscular (type III), total (type IV), extended (type V), and anterior bilateral cordectomy and commissurectomy (type VI). Type I and type II cordectomy were applied in patients who had midcordal microin- vasive carcinoma presented as a very early lesion with superficial or exophytic tumors. Voice Analysis Voice analysis was conducted by a speech pathologist using Computerized Speech Lab (CSL) (Model 4500; KayPEN- TAX, Montvale, NJ). The patient was comfortably seated on a chair and told to say /a/ for 3 seconds with a comfortable level of effort that allowed the voice to be recorded with a micro- phone 10 cm away from the lips. Fundamental frequency, Jitter, Shimmer, noise-to-harmonic ratio (NHR), and maximal and minimal energy (dB, voice intensity) were analyzed by the Mul- tidimensional Voice Program (MDVP), which is a software pro- gram for acoustic analysis. Maximal phonation time (MPT) was measured during phonation of /a/ at a constant pitch and inten- MATERIALS AND METHODS Patients

sity after full inspiration. The examination was conducted three times consecutively, and the maximum value was used for anal- ysis. Two experienced speech pathologists who were blinded to the type of cordectomy conducted perceptual analysis using the Grade, Roughness, Breathiness, Asthenia, Strain (GRBAS) scale 19 and rated conversational speech and sustained vowels. The outcomes of each patient were classified from 0 to 3 (0 5 normal, 1 5 mild, 2 5 moderate, 3 5 severe). The Korean Voice Handicap Index (VHI) 20 was used to evaluate self-perception analysis by each patient, which comprised 30 questions of three categories (i.e., Functional, Physical, and Emotional). 21 A score from 0 to 4 (from least to most disability) was assigned to each question. Statistical Analysis All data were expressed as mean 6 standard deviation, and Student t test or Wilcoxon signed rank test was used for analysis as appropriate. Correlations between Grade scale and other parameters related to voice quality were determined by Pearson correlation coefficient ( r ). Statistical analysis was per- formed using the PASW 18 software program (SPSS Inc., Chi- cago, IL), with P < .05 considered significant. RESULTS The patients included 54 male patients and three female patients with a median age of 59 years (range, 34– 70 years). Type of cordectomy included 10 (17.5%) type I, 11 type II (19.3%), 21 type III (36.8%), one type IV (1.8%), and 14 type V (24.6%). The median follow-up period was 30 months (range, 9–159 months) and the median time of voice evaluation during the early postoperative period was 10 weeks after surgery (range, 8–13 weeks), whereas the median time of the late postoperative period was 12 months (range, 6–41 months) after surgery. Wound heal- ing was completed between 4 to 8 weeks after TLM with- out any significant complications such as hemorrhage or infection. Among 19 patients who underwent resection of the anterior commissure or bilateral vocal cord, eight (42.1%) showed a mild to moderate degree of stenosis of the anterior glottis. One of them underwent laser resec- tion of the stenosis with keel insertion. The MPT, VHI-Total, VHI-Functional, VHI-Physical, Asthenic, and Strained scale showed significant deterio- ration at the early postoperative period. However, these parameters showed no significant difference more than 6 months after the surgery, whereas the Grade and Rough- ness scales showed significant improvement (Table I, Fig. 1). Because Grade of the GRBAS scale showed signif- icant improvement, we conducted a correlation analysis to verify whether it might represent the quality of voice in patients after TLM. Grade of the GRBAS scale at the late postoperative period ( > 6 months) showed significant correlation with most of the parameters related to postop- erative voice quality (Table II). Late postoperative voice quality correlated with the extent of resection and showed worse outcomes in extended cordectomies (type III, IV, V) with respect to Jitter (%), Shimmer (%), NHR, VHI-Physical, and all parameters of GRBAS scale (data not presented). Although patients who underwent type I or type II cordectomy showed significant improvement in VHI-Physical and Grade, those with type III, IV, and V

Laryngoscope 126: September 2016

Lee et al.: Voice After Transoral Laser Microsurgery

16