Predicting the local outcome of glottic squamous cell carcinoma after definitive radiation therapy: value of computed tomography-determined tumour parameters.

Abstract

The T-classification has shortcomings in the prediction of local outcome of glottic squamous cell carcinoma (SCC) treated by definitive radiation therapy. In this regard, the value of several CT-derived tumour parameters as predictors of local outcome was investigated. The pretreatment CT studies of 119 patients with glottic SCC (T1, n=61; T2, n=40; T3, n=14; T4, n=4) treated with curative intent by radiation therapy were reviewed for tumoral involvement of specific laryngeal anatomic subsites (including laryngeal cartilages). Tumour volume was calculated with the summation-of-areas technique. Actuarial (life-table) statistical analysis was done for each of the covariates; multivariate analysis was performed using the Cox proportional hazards model. In the actuarial analysis tumour volume was significantly correlated with local recurrence rate (P=0.0062). Involvement of the cricoid cartilage (P=0.0052), anterior commissure (P=0.0203), subglottis (P=0.0481) and preepiglottic space (P=0.0134) and degree of involvement of the true vocal cord (P=0.0441) and paraglottic space at the level of the true vocal cord (P=0.0002) were also significantly correlated with local recurrence rate. In the multivariate analysis, only degree of involvement of the paraglottic space (at the level of the true vocal cord) (P=0.0001) and preepiglottic space (P=0.02) were found to be independent predictors of local recurrence. The T-category was significantly correlated with local outcome in the actuarial analysis (P=0.0001), but not in the multivariate analysis (P=0.5915). Several CT-derived parameters are powerful predictors of local outcome in glottic cancer treated with radiation therapy; some of these parameters are stronger linked to the local control rate than the T-classification.