Depth of Invasion Predicts Vocal Fold Motility and Muscle Invasion in Glottic Carcinoma.

Abstract

Vocal fold motility guides staging of glottic squamous cell carcinoma (SCC) but is subjective and operator dependent. Because impaired mobility reflects tumor depth of invasion (DOI) into the vocal muscle and paraglottic space, objective DOI measurement may better capture aggressiveness and functional impairment. This two-stage study assessed radiologic-pathologic agreement of DOI, derived pathologic DOI (pDOI) thresholds for altered motility and deep muscle infiltration, and externally validated these thresholds, aiming to support a more reproducible staging surrogate. Surgically treated glottic SCC from Ospedale Policlinico San Martino (Genoa, Italy) formed the derivation cohort; external validation was performed at the European Institute of Oncology (Milan, Italy). Outcomes included correlation/agreement between radiologic DOI (rDOI) and pDOI, and between pathologic depth of vocal muscle invasion (pDMI). Spearman's ρ, Bland-Altman analysis, and Lin's concordance correlation coefficient (CCC) assessed agreement. ROC analysis (Youden index), likelihood ratios, and Fagan nomograms evaluated diagnostic performance and clinically meaningful thresholds. In 262 derivation cases, rDOI strongly correlated with pDOI (ρ = 0.75; CCC = 0.824). pDOI predicted altered vocal fold motility (AUC 0.725; cutoff 2.1 mm) and accurately predicted muscle invasion (AUC 0.94; cutoff 2.2 mm), whereas pDMI showed moderate prediction of motility (AUC 0.69). In 38 validation cases, correlation remained significant (ρ = 0.56; CCC = 0.56), and Genoa-derived DOI cutoffs were reproducible (sensitivity 85%-94%, specificity 76%-83%). DOI provides a robust, objective surrogate of functional impairment and muscle infiltration in glottic SCC, with externally validated thresholds that may complement or replace subjective mobility assessment in future staging algorithms.