STEVEN M. ZEITELS, MD; ROBERT E. HILLMAN, PHD; RAMON A. FRANCO, MD; GLENN W. BUNTING, MS
BOSTON, MASSACHUSETTS

Phonosurgical management of early glottic cancer has evolved considerably, and objective vocal outcome data are sparse. A prospective clinical trial was done on 32 patients with unilateral cancer (T1a in 28 and T2a in 4) who underwent ultranarrow-margin resection; 15 had resection superficial to the vocal ligament, and 17 deep to it. The subepithelial infusion technique facilitated selection of these patients for the appropriate procedure. All are cancer-free without radiotherapy or open surgery. Involvement of the anterior commissure (22/32) or the vocal process (15/32) of the arytenoid cartilage did not influence local control. Nine of 17 patients had resection of paraglottic musculature, and all underwent medialization reconstruction by lipoinjection and/or Gore-Tex laryngoplasty. Eight of the 17 had resections deep to the vocal ligament, but without vocalis muscle, and 1 of the 8 underwent medialization. Posttreatment vocal function measures were obtained for all patients. A clear majority of the patients displayed normal values for average fundamental frequency (72%) during connected speech, and normal noise-to-harmonics ratio (75%) and average glottal airflow (91 °I°) measures during sustained vowels. Smaller majorities of patients displayed normal values for average sound pressure level (SPL; 59%) during connected speech, and maximum ranges for fundamental frequency (56%) and SPL (59%). Fewer than half of the patients displayed normal values for sustained vowel measures of jitter (45%), shimmer (22%), and maximum phonation time (34°l°). Almost all patients had elevated subglottal pressures and reduced values for the ratio of SPL to subglottal pressure (vocal efficiency). There were significant improvements in a majority of patients for most vocal function measures after medialization reconstruction. By means of a spectrum of resection and reconstructive options, normal or near-normal conversation-level voices were achieved in most cases regardless of the disease depth. These favorable results are based on establishing aerodynamic glottal competency and preserving the layered microstructure of noncancerous glottal tissue.