Prognostic significance of conventional and volumetric PET parameters with and without partial volume correction in the assessment of head and neck squamous cell carcinoma.

Abstract

The optimal quantification of PET in assessment of head and neck squamous cell carcinoma (HNSCC) is still under development. The effect of partial volume correction (PVC) on the evaluation of survival in the HNSCC patients has not been investigated yet. Pretreatment 18F-FDG-PET/CT scans of a selected group of 57 patients with advanced stage HNSCC were collected. Conventional (SUVmean and SUVmax) and volumetric [total lesion glycolysis (TLG) and metabolic tumor volume (MTV)] PET metrics were calculated. The ROVER software (ABX GmbH, Radeberg, Germany) automatically applied PVC to the PET metrics. Cox proportional hazards regression model calculated hazard ratio (HR) for assessment of predictive parameters of progression-free survival (PFS). In multivariate Cox regression analysis, including age, gender, race, human papillomavirus status, and stage, the only significant predictors of PFS were the volumetric PET parameters (TLG: HR, 1.003; 95% CI, 1.001-1.005; P = 0.02), pvcTLG (HR, 1.002; 95% CI, 1.001-1.004; P = 0.01) and MTV (HR, 1.050; 95% CI, 1.024-1.077; P < 0.01). The partial volume-corrected values were significantly higher than the noncorrected values (Wilcoxon sign test; P < 0.05). However, there was not a statistically significant difference between the nonpartial volume corrected and partial volume-corrected PET metrics for assessment of PFS. Volumetric PET metrics were predictors of PFS in Cox regression analysis. Applying PVC could not significantly improve the accuracy of PET metrics for assessment of PFS.