AbstractPhonomicrosurgical treatment of premalignant vocal fold epithelium and microinvasive cancer combines principles of surgical oncology with advanced laryngoscopic microsurgical‐techniques. This treatment is guided by mucosal‐wave theory of voice production and strives not only to cure the disease but also to achieve optimal vocal function.Surgical techniques developed during the past two centuries have improved methods for vocal fold visualization, tissue retrieval, and tissue evaluation. Examination of the evolution of these surgical techniques reveals the incomplete convergence of laryngoscopic surgical theory with both the concept of premalignancy and the anatomical‐physiological principles of voice production. This historical review, which helps to explain the lack of consensus about current treatment options, led to a series of four investigations. They were conducted with the aim of developing a laryngoscopic (phonomicrosurgical) management approach for improving the treatment of premalignant and microinvasive vocal fold epithelium.In the first of four investigations, 42 patients (each of whom had a significant smoking history) underwent microlaryngoscopic biopsy of 52 vocal fold lesions. These lesions, which were suspicious for atypia or malignancy and were confined to the musculomembranous vocal fold, were mapped according to surface involvement and depth of penetration. Review of the maps revealed that 27 of the 52 lesions involved only the superior/ventricular surface. For these patients, the entire layered vocal fold structure could potentially be preserved on the medial/vocalizing surface. Twenty‐five of the 52 lesions involved both the superior/ventricular surface and the medial/vocalizing surface. No lesion involved only the medial surface.These data suggest that (in smokers) geographic localization of keratotic and erythroplastic lesions on the superior/ventricular surface of the musculomembranous vocal fold are likely to contain atypia. This characteristic facilitates the appropriate selection of patients for biopsy and may spare individuals, who have lesions resulting from hyperfunctional dysphonia and/or gastroesophageal reflux, from unnecessary biopsy. These two disorders typically result in pathology on the medial and/or posterior glottal surfaces.In order to determine whether a directed biopsy or an excisional biopsy approach is preferable for obtaining an accurate diagnosis, all specimens underwent whole‐mount sectioning for three‐dimensional histopathological analysis. Keratosis was noted: without atypia in 14; with atypia in 27; and with carcinoma in 11. The severity of the atypia usually varied throughout each specimen. The surface appearance of the lesion was not a reliable prognosticator of the severity of dysplasia either between patients or in different areas of the same lesion; therefore, excisional biopsy and whole‐mount, multiple‐section histopathological analysis were necessary for obtaining an accurate diagnosis.In the second study, submucosal vocal fold dye infusions were performed to identify where the glottal mucosa was supported by the superficial lamina propria (SLP). The histological sections of these vocal folds demonstrated that the caudal limitation of the SLP was in close proximity to the epithelial surface change from ciliated to nonciliated epithelium (inferior arcuate line). It is generally thought that the inferior arcuate line denotes the proximal subcordal vibratory epithelium. The infusions also demonstrated that dye‐induced hydrostatic distension of the SLP would provide more room for surgical dissection in the SLP and would reduce instrumental trauma to the layered vocal‐fold microstructure. These features assisted in the design of a procedure that, by simulating Reinke's edema, would help to preserve the vocal fold's layered anatomy.The third investigation employed the anatomical SLP‐infusion studies in a clinical trial. A submucosal infusion of saline and epinephrine in conjunction with microflap excisional biopsy was performed on 43 vocal folds. Complete excisional biopsy was accomplished in all. The infusion produced hydrostatic distension of the SLP and the overlying epithelium. This improved the visualization of the lesion's perimeter and resulted in precise placement of incisions around the lesion. The increased depth of the SLP and the hydrodissection of the epithelium from the vocal ligament facilitated both the development of a precise microflap and the maximal preservation of the vocal fold's layered microstructure. The infusion effectively demonstrated invasion through the SLP into the vocal ligament by causing distension of the epithelium and of the SLP not involved with cancer, while leaving a relative depression where the SLP was replaced by cancer.In the fourth study, a pilot group of 17 patients was assessed to determine which postoperative objective vocal‐measurement criteria would be valuable, and to begin to establish expected postsurgical outcome data. The results generally suggested that as the epithelial resection extended from the superior surface to include the medial surface as well, the measures used began to reflect deteriorating vocal function. Specifically, the acoustic data revealed that the most likely aberrations postoperatively were increased jitter (cycle‐to‐cycle irregularity of pitch) and reduced maximal ranges for both loudness (intensity) and pitch (frequency). These changes appeared to be related to higher levels of stiffness and/or tension in the vocal mechanism. With these changes, there was often an increase in subglottic aerodynamic support (pressure), which could be a compensatory adjustment or the manifestation of persistent (presurgical) hyperfunctional behaviors. While patients did not describe discomfort in conjunction with voice production, the stroboscopic evaluations visually confirmed the presence of increased stiffness/restriction in the glottal valve. The most common abnormal sequela seen with stroboscopy was a decrease in the magnitude of the mucosal wave on the superior/ventricular vocal‐fold surface. If a medial/vocalizing surface resection was done, the amplitude of vocal fold vibration (excursion) was also likely to be diminished. The perceptual vocal assessment parameters were typically normal or revealed mild dysphonia.Phonomicrosurgical treatment of premalignant and microinvasive vocal fold epithelium integrates principles of surgical oncology with those of voice production. The use of submucosal infusion with microflap excisional biopsya.provides a tissue specimen suitable for multiple‐section histopathological analysis;b.establishes an unequivocal diagnosis;c.treats the lesion;d.reserves all cancer treatment options;e.is simple, repeatable, and cost‐effective;f.preserves the layered vocal‐fold anatomy, andg.avoids the medial aspect of the vocal fold (vocalizing surface) as much as is oncologically possible for optimal postoperative voice. The pilot voice evaluations that were done suggest that while patients treated by this approach did not typically achieve “normal” vocal function (according to measures of underlying physiology), most did manage to produce voices that were, acoustically and perceptually, within normal li